review and evaluation of fisheries stock management models

[From the U.S. Government Printing Office, www.gpo.gov]

A R eview. and Evaluation of
Vis'heries Stock Management Models

Y F N. t

aP) expression of in
ri6 (t-i) growth rate due
COASTAL ZONE ENIT[R q1 fishing mortalit
RMATION C
1-M
eb[l-M]t PO e

0 0.2 0.. O.'s -W Z2

SH
328
R48
1980
ny

CZIC COLLECTION

A REVIEW AND EVALUATION OF

FISHERIES STOCK MANAGEMENT MODELS

Part I -Text

W.A. Richkus;
J.K. Summers
T.T. Polgar
A.F. Holland
Martin Marietta Corporation
Environmental Center
1450 South Rolling Road
Baltimore, Maryland 21227

Prepared for

Coastal Resources Division
Tidewater Administration
Maryland Department of Natural Resources
Tawes State Office Building
Annapolis, Maryland 21401

COASTAL ZONE
INFORMATION CENTER

TABLE OF CONTENTS

Page

List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . v

List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . vi

I. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . .

II. REVIEW OF FISHERIES MODEL LITERATURE . . . . . . . . . . . . .

A. Objective . . . . . . . . . . . . . . . . . . . . . . .

B. Selection Criteria and Search Procedures . . . . . . . .

III. CHARACTERIZATION OF STOCK @IkNAGDENT MODELS . . . . . . . . .

A. Model Reviews . . . . . . . . . . . . . . . . . . . . .

B. Major Categorization of Stock Management Models . . . .

C. Categorization of Biological Stock Managemnt Models. . 111-4

1. Statistical Stock Management Models . . . . . . . . 111-4

2. Surplus Production Models . . . . . . . . . . . . . 111-5

3. Yield-Per-Recruit Models . . . . . . . . . . . . . . III-11

4. Simulation Stock Management Models . . . . . . . . . 111-14

D. Data Dependence of Stock Management @-Tbdel Types . . . . 111-16

IV, INPUT SUN-IODELS AND PARAMETER ESTIMATION . . . . . . . . . . . IV-1

A. Selection Criteria and Search Procedures . . . . . . . . IV-1

B. Categorization of Input Development . . . ... . . . . . IV-1
1. Parameter Estimation of Specific Yield Models . . . IV-2

2. Functional Submodels . . . . . . . . . . . . . . . . IV-2

3. Data Acquisition Methods . . . . . . . . . . . . . . IV-3

V. LIFE HISTORY CHARACrERIZATION OF EXPLOITED MARYLAND SPECIES . V-1

A. Objective . . . . . . . . . . . . . . . . . . . . . . . V-1

B. Selection of Species . . . . . . . . . . . . . . . . . . V-1

C. Categorization of Selected Species with Respect Page
to Model Applicability . . . . . . . . . . . . . . . V-19

VI. EVALUATION OF MODEL APPLICABILITY . . . . . . . . . . . . VI-1

A. Introduction . . . . . . . . . . . . . . . . . . . . VI-q1

B. Model Selection Scheme . . . . . . . . . . . . . . . VI-1

C. Species-Mbdel Evaluations . . . . . . . . . . . . . VI-5

1. Hard Clam (Meqrcenaria mercenaria) . . . . . . . VI-5

2. Soft Clam (Mya arenaria) . . . . . . . . . . . . VI-7

3. Oyster (Crassostrea virginica) . . . . . . . . . VI-8

4. White Perch (Morone americana) . . . . . . . . . . . . . VI-10

5. Yellow Perch (Perca flavescens) . . . . . . . . VI-12

6. Carp (Cyprinus carpia) . . . . . . . . . . . . . VI-12

7, White Catfish (Ictalurus catus) . . . . . . . . VI-13

8. Alewife (Alosa pseudoharengus) and Blueback
(Alosa aestivalis) . . . . . . . . . . . . . . . VI-14

9. American Shad (Alosa sapidissima) . . . . . . . VI-16

10. Striped Bass (Morone saxatilis) . . . . . . . . VI-17

11. Winter Flounder (Pseudopleuronectes americanus) VI-19

12. Bluefish (Pomatomus saltatrix) . . . . . . . . . VI-20

13. Atlantic Menhaden (Brevoortia tyrannus . . . . VI-21

14. Blue Crab (Callinectes dapidus) . . . . . . . . VI-22

15. American Eel (Anguilla rostrata) . . . . . . . . VI-23

16. Spot (Leiostomus xanthurus)
(Micropogonias undulatus), and Weakfish
(Cynoscion regalis) . . . . . . . . . . . . . . VI-24

17. Sumner Flounder (Paralichthys dentatus) . . . . VI-26

VII. COMMUNICATIONS WITH MANAGEMENT AGENCIES . . . . . . . . . VII-1

A. Objective . . . . . . . . . . . . . . . . . . . . . VII-1

iii

Page

B. Interview Procedure . . . . . . . . . . . . . . . . VII-1

C. Results . . . . . . . . . . .. . . . . . . . . . . VII-1

D. General Overview . . . . . . . . . . . . . . . . . VII-5

VIII. REFERENCES . . . . . . . . . . . . . . . . . . . . . . . VIII-1

LIST OF TABLES

Table Page

II-1 Results of library retrieval searches employed in
the review of fisheries model literature . . . . . . . 11-2

11-2 Plartin Marietta-oivned journals and books used in
the review of fisheries model literature . . . . . . . 11-3

III-I Data requirements and examples of biological
stock management models . . . . 4 . I . # . # . . . . 111-6

111-2 Applicable stock management model types as
determined by data availability . . . . . . . . . . . 111-17

IV-1 Relative data requirements of functional submodels
and their potential utility as input to stock
management models . . . . . . . . . . . . . . . . . . IV-4

V-1 Finfish and shellfish species reported in commercial
fisheries statistics for Maryland, 1971-197S . . . . . V-2

V-2 Life history data on selected Maryland species . . . . V-3

VII-1 Individuals and agencies providing information on
current use of fisheries models in management
programs (only those contacts from which informa-
tion was received are listed) . . . . . . . . . . . . VII-2

LIST OF FIGURES

Figure Page

I-1 A Ricker-type stock recruitment curve, illustrating
the concept of surplus production . . . . . . . . . . .

1-2 Change in size and total weight of a yearclass of
fish as a function of age, showing the concept of
maximum biomass yield from such a yearclass. .. . . . . 1-4

III-1 Hierarchical organization and basic data require-
ments of stock management model types . . . . . . . . . 111-3

V-1 Procedure.followed in grouping selected Maryland
exploited finfish and shellfish species according
to major life history characteristics . . . . . . . . . V-20

VI-1 Conceptual overview of model selection process . . . . VI-2

VI-2 Detailed conceptual overview of model evaluation
scheme ... . . . . . . . . . . . . . . . . . . . . . . VI-3

I. INIRODUCTION

The major objective of this study is to assess the suitability of
existing fisheries production and population models for use in the management
of Maryland's tidewater fisheries. Vo lume I of this report contains Phase I
of the project, consisting primarily of 1) a literature review of quantitative
fisheries yield and management models and 2) life history data of selected
Maryland species Volume II contains bibliographies of literature used
in the Phase I activities. This interim report presents the results of work
performed by the Martin Marietta Environmental Center (EC) for the Coastal
Resources Division (CRD) of the Maryland Tidewater Administration, under
contract C12-80-430.
Quantitative models serve several important functions in the management
of living resources. By serving as a means of organizing, causally connecting,
and relating various kinds of population, exploitation and economic data, they
provide the basis for conceptualizing problems in management policy. Because
data-input and model structures are goal oriented, models also reveal
information deficiencies and inadequacies.
The application of numerical schemes for catch pTediction in a fishery
was initiated by Baranovi it was he who first proposed the concept of
surplus production, stating that."... a fishery, by thinning out a fish
population, itself creates the production by which it is maintained"
(Baranov, 1927).*
The concept of surplus production and its use can be illustrated
by a diagram, such as shown in Figure I-1 (from Ricker, 1978). This curve
relates magnitude of the parent stock (as measured alternatively in numbers,
weight, egg production, etc.) to numbers of individuals recruited into the
adult or exploited population. A population can be viewed as being atan
equilibrium level that is consistent with the carrying capacity of its
environment. If the population size deviates from this level, the population
acts to restore itself to its equilibrium by increased or decreased survival
and growth. If harvesting of an unexploited stock causes the population
to decline to level C in Figure I-1, the population responds with an increase
of magnitude AB.

from Ricker (1978)

1.0-

0.8-

Z
UJ06
U9
0

0.4

0 0.2 0.4 C 0.8 ID 1.2

SPAWNERS

Figure I-1. A Ricker-type stock recruitment curve, illustrating the
concept of surplus production; the distance A-B represents
the amount of progeny which can be harvested at population
ID
'B

size C without decreasing the population si ze further
(from Ricker, 1978).

1-2

As long as a harvest generated by recruitment of magnitude AB is taken
each year, the population remains around a constant, equilibrium level C.
Of course, growth of the recruits in a single age group or cohort will
influence the magnitude of biomass harvest from that cohort, depending on
the age at harvest. Figure 1-2 shows the influence of natural mortality
and growth upon cohort biomass. The relationship depicted in Figures
I-1 and 1-2 form the basis of many mathematical fisheries models.
The population level at which surplus biomass production is maximiZed
will generate the Maximum Sustainable Yield MY), a term frequently used
in fisheries literature. Many fisheries management programs have
focused on attaining MSY. More recently, the alternative concept of
Optimum Sustainable Yield (OSY) has been promoted in fisheries management
(Larkin, 1977). In this approach, the major management goal is not only the
maximization of biomass yield but, alternatively, the maximization of
economic return,,or "benefit" to the ecosystem, or optimum recreational
activity, for example. Much of the current literature, particularly in
bioeconomics,- deals with optimization of fisheries yields from different
perspectives.
The concepts presented here, described in a somewhat simplistic
manner, form the basis for most mathematical formulations of exploited
populations presented in the fisheries literature. Generally, these
formulations relate the reproductive potential of a population and growth
of members of that population to mortalities caused by man as well as by
natural phenomena. They can thus be used to examine consequences of
various levels of exploitation and of other environmental perturbations
caused by man to resource stocks. The major limitation.in model applications ,
however, is that certain simplifying assumptions about population dynamics
must generally be made, and the validity of model output will be heavily
dependent upon whether various assumptions employed are reasonable in a
variable real world. Thus, a comparison of model characteristics to life
history characteristics of the species considered provides a first step
in assessing model applicability.
Also, data requirements of models vary widely, depending on their
level of complexity. The most simplified models, which tend to be those
based on the greatest number of assumptions, usually require the least data.
Models which attempt to account for most factors influencing fisheries popula-
tions require the greatest amounts of data. Thus, assessment of model suitability

1-3

Size of Yearclass Declines
from Mortality

Maximum Biomass

Total Biomass of Yearclass

Weight of Individuals Increases
with Growth

------------------

Age of Yearclass (years)

Figure 1-2. Change in size and total weight of a year class of fish as a function of age, showing the
concept of maximum biomass yield from such a year class (modified from Rounsefell, 1975).

M. WX M AM am

involves a review of data availability, data quality, and means of
acquiring and estimating properties of the necessary data.
The organization of this project is based on the relationship noted
above. The Phase I work reported here consists of: 1) an extensive
review of literature on current and classical fisheries models and their
application; 2) documentation of life history characteristics of
selected, exploited Maryland species; 3) an evaluation of suitability of
existing models that use assumptions and mathematical structures
consistent with the life histories of particular Maryland species; and
4) a review of literature dealing with input* models to production
models or means of deriving parameters for such models. Phase II activities
will center on 1) a review of fisheries data currently available in Maryland;
2) an assessment of data needs for the application of management models;
3) the development of recommendations; concerning data acquisition and
processing procedures; and 4) the detailed assessment of the applicability
of potentially suitable production or population models to selected Maryland
species.

Since the study is ongoing, the results of Phase I presented here may
be modified or appended during the course of Phase II investigations. Any
such modifications will appear in the final project report.

An input model is generally considered a mathematical formulation which
represents processes (e.g., growth or recruitment) other than changes in
numbers of individuals in the population.

1-5

II. RE-VIEW OF FISHERIES MODEL LITERATURE

A. Objective

To evaluate the suitability of yield and management models for
use with Maryland species, we first had to identify potentially useful
models and document their mathematical structure. The first step in this
task was to locate literature on mathematical models which have been used
in the management of fisheries or which, because of their tractable and useful
mathematical structure, might have management applicability.

B. Selection Criteria and Search Procedures

Initial selection of potentially relevant articles was basea. on the
titles of papers and reports. Key words used in both computerized library
retrieval searches and in the direct review of readily available journals
and books were:

a fish management (1)

a fish yield (2)

9 fish population (3)

a model of stock recruitment (4)

e model of fish management (5)

a model of fish population (7)

0 mathematical model of (4) to (7)

The library retrieval services employed in the literature review al-e
listed in Table II-1. A total of 249 abstracts was obtained as a result
of this search. From examination of the titles, we culled many articles
that appeared irrelevant to the study and did not appear in bibliographies
of other selected articles.
Issues of all applicable in-house journals and books were directly
revie.ived, based on the key words noted above. Journals and books
covered in this review are listed in Table 11-2.

II-1

M man, wM M MM M M M

Table II-1. Results of library retrieval searches employed in the review of fisheries model literature

Retrieval Service Sources Covered Inclusive Dates Number of Abstracts
Obtained

DIALOG Biosis Biological Abstracts 1969 - present 103
Previews and Bioresearch Index:
2,265,000 citations

DIALOG Fhviroline Periodicals@ government 1971 - present 52
publications, industry
reports, proceedings
of meetings:
7S,,000 citations

DIALOG Oceanic Journals) books, tech- 1964 present 35
Abstracts nical reports, confer-
ence proceedings,
governmental and trade
publications;
110.,SOO citations

NTIS Government-sponsored 1964 present S9
research..development,
and engineering reports:
72S,,000 citations

Table 11-2. Martin Marietta-owned journals and books used in the review
Of fisheries model literature

Journals
(P = Present)

Bulletin of Mathematical Biology (1973-P; Volumes 35-41)
Chesapeake Science (Complete; Volumes 1-18)
Coastal Zone Management Journal (1975-P; Volumes 2-5)
Deep Sea Research (1973-P; Volumes 20-26)
Ecology (1972-P; Volumes 53-60)
Ecological Modelling (Complete; Volumes 1-7)
Estuaries (Complete; Volumes 1-2)
Estuarine and Coastal Marine Science (1974-P; Volumes 2-8)
Fish Bulletin (1974-P; Volumes 72-77)
Journal of Experimental and Marine Biology (197S-P; Volumes 32-39)
Journal of Fish Biology (1974-P; Volumes 6-14)
Journal of Fisheries Research Board, Canada (1973-P; Volumes 30-36)
Journal of Marine Biological Association, U.K. (1973-P; Volumes (53-59)
Journal of Marine Research (1974-P; Volumes 32-36)
Marine Biology (1974-P; Volumes 24-52)
Oecologia (197S-P; Volumes 18-34)
Proceedings of National Shellfisheries Association (1973-1977;
Volumes 63-67)
Transaction of American Fisheries Society (1971-P; Volumes 100-108)

Books

Cushing, D. H. 1973. Marine Ecology and Fisheries, London: Cambridge
University Press. 278 pp.
Gulland, J.A. (ed.). 1977. Fish Population Dynamics. New York: John
Wiley and Sons. 372 pp.
Harden-Jones, F.R. (ed.). 1974. Sea Fisheries Research. New York:
John Wiley and Sons. 510 pp.
Knauss, J. (ed.). 1979. Climate and Fisheries. Center for Ocean
Management Studies. Kingston, Rhode island. 135 pp.
Ricker, W.E. (ed.). 1971. Methods for Assessment of Fish Production
in Fresh Waters. Oxford: Blackwell Scientific Publications. 348 pp.
Ricker, W.E. 1978. Ccmputation and Interpretation of Biological Statistics
of Fish Populations. Bull. Fish. Res. 3d. Canada 191. 382 pp.
Smith, R.F. (ed.). 1966. A Symposium on Estuarine Fisheries. Am.
Fish. Soc. Spec. Publ. 3. 154 pp.
Van Winkle, W. (ed.). Assessing the Effects of Power-Plant-Induced
Mortality on Fish Populations. Sponsored by Oak Ridge National
Laboratory, Energy Research and Development Administration,
and Electric Power Research Institute.

II-3

Considerable overlap with the titles acquirIed in the library retrieval
process was noted. Bibliographies of all articles obtained in-house
were also searched.
An attempt was made to acquire copies of all the possibly relevant
articles. Acquisition was restricted to those articles published
since 1960., with the exception of some which are considered "classics"
based on their frequency of citation (e.g., Schaefer, 1954, 1957; Beverton
and Holt, 1957; and Graham, 1935). Two hundred and twenty-three were
ordered through interlibrary loan. Since considerable time lags occurred
between distribution, 28 of the articles ordered had not been received by
the end of Phase I. These articles will be reviewed and added to the
annotated bibliography (Appendix A) upon receipt.

1.1-4

III. CHARACTERIZATION OF STOCK MAMGEMENT MODELS

A. Model Reviews

Of the 397 relevant manuscripts, journal articles, and reports selected
by the criteria discussed in the previous section, all but 14 of the 123
articles were reviewed. The review of incoming articles was discontinued
on October 5 to maintain the project and report schedule. The remaining
articles will be reviewed and presented in a supplement to the appended
bibliography in the final report at the end of the second phase of this
project.
The 383 available articles concerning the determination, management,
or optimization of fishery yields were evaluated, and "potentially useful"
models were retained for further examination. Results of this review were
documented in Appendix A, which contains comments generally describing
each model's applicability to the management of Maryland fishery yield, and
abstracts of the individual articles. Of the 383 articles reviewed... 83
contained information pertaining to models "potentially useful" for the
management of Mary land fish and shellfish populations.
Criteria used to determine the potential usefulness ofindividual
models were somewhat subjective, but the primary ones included:

a the direct applicability of the model structure to a Maryland
fishery
a the general applicability of the model structure
o the complexity of the data base necessary for the construction
of the model
o the rationality of the biological and economic assumptions
.on which the-model structure was based
a the potential applicability of the model structure to estuarine
and/or coastal environments.

B. Major Categorization of Stock Management Models

All stock management models for the management of a resource can be
partitioned into three hierarchical groups (Figure III-1):

q optimization models as applied to functioning stock management
models (that is, models which integrate population biology and
economics of a fishery such that management activities could be
developed to optimize a certain aspect of that fishery, such as
economic return)

III-1

Figure 111-1. Hierarchical organiZation and basic data requirements
of stock nunagement model types

Fi a Goa s Optimization Model
(e.g., maximiza- pplied to Bioeconomic
tion of yield or Population Biology
and minimiza-
tionof effort) Stock Management
Models

Economic
Data Base on Bioeconomic Stock
Fishery and > -Management Model

Market

Biological
Data Base on Population Biology
Stock and Stock Management
Fishery Model

111-2

e bioeconomic models for stock management (i.e., those models
which deal with the economic costs and returns of fisheries)

0. population biology models for stock management (i.e., those
models dealing only with the biological aspects of fish
populations).

Because both optimization and bioeconomic stock management models
depend on the existence of a population level biological model, biological
models were chosen as the most basic form of fishery yield formulations (Fig-
ure III-1), and further effort was concentrated on the analysis of this type
of stock management model. Biological stock management models take many
forms (e. g. , stock population, population dynamics) , the complexity of which
depends on the objectives of the modeler and the availability of data; thus,
population models can be relatively simplistic (e.g., involving only statistical
relationships between yield and environmental variables (KE@lccmme, 1976; Adams
and Olver, 1977) or very complex, encompassing numerous aspects of a
population (Parrish, 1975; Sissenwine, 1977) or several interacting
populations (Andersen and Ursin,, 1977).
Although economic functions are the primary output of bioeconomic models,
these models, as shown in Figure III-IP require a biological population model
or submodel of the species being managed as a base on which to determine
economic relationships (Plourde, 1970; Southey, 1972; Clark, 1973; Anderson,
1975; Clark et al.) 1973; Dow et al., 1975; Silvert, 1977; Grant and Griffin,,
1979). As a result, this class of models is considered to be secondary in the
development of stock management plans for Maryland fisheries: their.construction
should be approached in most cases only after a model of the population
dynamics of the various fisheries has been formulated or determined irrelevant
to the succes's of the fishery. An example of the latter situation would
be a shellfishery in which the cost of harvesting restricts fishing effort
when stocks are at low density, and reproductive capacity at that density
is sufficiently high for stock maintenance.
Optimization procedures represent the highest order of complexity in
this hierarchy (Figure III-1) because their use, which depends on the goals
and objectives' of the manager,requires a functional stock management model
on either the bioeconomic or biological population level. These models determine
a set of parameters that produce a maximal or minimal result for the control
functions (e.g., yield, effort, catch/effort) (Clark et al., 1973; Palm., 1975;

111-3

Saila and [less, 1975; Hilborn, 1976: Huang et al., 1976; Peterman, 1977;
Agnew, 1979), =d represent a tertiary step in the management scheme, i.e.,
they cannot be considered for Maryland stocks until functional biological
stock management models are constructed.
Because bioeconomic models and optimization procedures are considered
secondary goals for developing management plans of Maryland stocks, management
models were further evaluated in Phase I. This model type represents a basic,
first-level formulation of the stock and its dynamics and is necessary to any
management plan. Bioeconomic and optimization procedures may be invest-
gated at a later date after population formulations are completed.

C. Categorization of Biological Stock Management Models

Biological stock management models of resource populations can be
minimally classified into five different groups, depending primarily on
existing data describing the fishery and its dynamics and the assumptions
of the various model structures. The five biological model types in order
of increasing complexity are:

* statistical models

0 surplus production models

0 yield-per-recruit models

0 population simulation models

0 interacting multispecies simulation models.

1. Statistical Stock Management Models

Statistical stock management models include all model formulations in
which yield or production is statistically related to some variable. There
are basically two types of statistical yield models which could be useful in
the modeling of Maryland fisheries:

0 autoregressive stock management models

0 single and multiple regressive stock management models.

111-4

The objective of the autoregressive yield model is to predict the next
year's yield (not necessarily the maximum sustainable yield [IMSY] or
optimal sustainable yield [OSYI)from a set of previous yield data. This
method, adapted from linear and non-linear forecasting Nntgomery and
Johnson, 1976), has been utilized with particular success in areas where
environmental variability is low (Dyer and Gillooly, 1979). The model's
predictive power is based on the primary assumption that the correlative
statistical relationship between yield and previous yield(s) remains un-
changed in time. The nature of the data required, estimates of the size of
the data set, and examples of autoregressive yield models are presented
in Table III-L
Regressive stock management models are generally expressed as a
statistical relationship (generally a linear or non-linear single or multi-
ple regression) between yield or stock size and an environmental variable(s)
(e.g., water temperature, water transport). The-main objective of
these models is to estimate the potential annual yield expected from a fishery
(i.e., not MSY OR OSY), and the only major assumption is a constant correlative
relationship between yield and the environmental parameter(s) measured.
The parameter may be represented as the value of the variable coincident with
the yield measurement (Sutcliffe et al... 1977) or lagged by some time quantity.,
usually the length of time between spawning and recruitment (Flowers and Saila,
1972).
A major use of these density-independent stock management models has been
to characterize potential yield of lakes (e.g. , Matuszek, 1978) using the
concentrations of total dissolved solids and lake depth. Ryder et al. (1974)
developed a morpho-edaphic index as a predictor of fisheries yield of lakes
based on these relationships. The adaptation of the lake morpho-edaphic
index to rivers by Welcomme (1976) may be usable in some altered form for
estuarine situations. The nature of data required, estimates of the size
of the data set necessary to implement this model type, and examples of its
use are given in Table III-1.

2. Surplus Production INIodels

Surplus production models represent a class of fishery yield models
which are not heavily data dependent. The major objective of this model

W@m mom @ M @ M M M " M M M an

Table III-I. Data requirements and examples of biological stock management models

Model Type Data Requirements Time Span of Data Examples

Statistical Stock
Management t4odels
a) Autoregressive Yield 5 to 10 years Orach-Meza and Saila
(1978)
Dyer and Gillooly
(1979)
b) Single and multiple Yield and environmental S to 10 years Talbot (1954)
regressive data (e.g., water tem- Flowers and Saila
perature or freshwater (1972)
discharge) Clady (197S)
Welcomme (1976)
Adams and Olver (1977)
Patriarche (1977)
Sutcliffe et al.(1977)
Matuszek (1978)

Surplus Production Models
a) Basic Biodel without Yield and effort Span must include Graham (1935)
recruitment lag complete range of Schaefer (1954, 1957,
effort data and 1968)
yield data for the Pella and Tomlinson
same period; the (1969)
longer the time span, Fox (1970)
the more likely Jensen (1972)
assumptions are to Schnute (1977)
be invalidated Fletcher (1978)
May et al. (1979)
b) Basic model inclu@ing Yield, effortY estimate Span must include com- Walter (1973, 1976,
recruitment delay of recruitment rate, and plete range of effort 1978)
time lag between spawning data and yield data Marchesseault et al.
and recruitment for the same period; (1976)
the longer the time
span, the more likely
assumptions are to be
invalidated

Table III-1. Continued.

Model Type Data Requirements Time Span of Data @@an
ALles

Yield-Per-Recruit Models
a) Beverton-Holt Age of recruitment into Time span should be Beverton (1953)
exploitable phase (t r) L life span of fish Beverton and Holt
Number of individuals species considered (19S7)
at time of initiali- Walters (1969)
zation (to) Sissenwine and Tibbetts
Maximum age of fish (tX) (1977)
Estink-ite of instantaneous
fishing mortality rate (F)
Estimate of natural
mortality rate (NI)
Maximal length or weight
of fish (L or W
Time fish spend in exploited
phase (t), - tr)
Growth constant from von
Bertalanffy analysis (K)

b) Ricker Number of recruits to each Time span should be
cohort > fishable life span Ricker (19S8, 1973)
Estimate of instantaneous of fish species con- Paulik and Bayliff
fishing mortality rate (F) sidered (1967)
of each cohort Schaaf and Huntsman
(1972)
Growth and natural mortal- Balsinger (1974)
ity rates of each cohort Lett et al. (197S)
i-oucks, andSutcliffe
Biomass of each cohort (1978)

mm-womew''M on wow M mom mamm"N@Nam

W S" no @ M M M .M

Table 111-1. Continued.

Model Type Data Requirements TimeSpan of Data Examples
Simulation Models
a) Dynamics of a single I to 10 years for all Leslie (1945,1959)
population Data requirements depend parameters Lefkovitch (1965)
on model structure, Pope (1972)
assmptions, and objectives; Balsinger (1974)
many different types of Be'land (1974)
data required. Hackney and Minns (1974)
Kitchell et al. (19114)
Caddy (197S)
Rudd (1975)
DeAngelis (1976)
Englert et al. (1976)
Lewis (1976)
Ulltang (1976)
Winters (1976, 1978)
DeAngelis et al. (1977)
Eberhardt and Siniff
(1977)
Horst (1977)
Sissenwine (1977)
Johnson (1978)
Orth (1979)

b) Multi-population Data requirements depend 1 to 10 years for all Swartzman and Van
dynamics on model structure, parameters Dyne (1972)
assunptions, and objectives; Regier and Henderson
many different types of (1973)
data required. Parrish (1975)
Andersen and Ursin
(1977)

type is the determination of maximum sustainable yield. This type of
yield model, developed initially by Graham (1935) and expanded into a
more applicable form by Schaefer (1954, 19S7), requires only annual
catch and effort data (although estimates of catchability must be
calculated or measured independently). Because its data require-
ments are limited (Table III-I), it is generally the most frequently
encountered form of standardized yield model found in the scientific
literature.
The most generalized form of this model may be expressed as follows:
the change in catchable fish biomass per unit biomass (e.g., so many
harvestable pounds per 100 pounds of stock) with respect to time is a
function of population growth reduced by natural and fishing mortality.
This can be mathematically expressed as:

1 dP = b - aP@'_ 1 -qf (1)
P dt

where

P biomass of catchable fish in the fishery
aand b E regression coefficients related to population
growth and mortalityrespectively, derived
using effort and yield data
0
M = a parameter based on catch data and related
to the carrying capacity of the environment
f = fishing effort
q = catchability.

The solution of equation (1), which represents stock size at a given time,
t' is:
P (t) (1-e b[l-M]t)+ Pol-M e b[l-M]t 1 1M
b (2)

These mathematical formulations represent the generalized stock production
model of Pella and Tomlinson (1969). The other major surplus production
models, namely the Schaefer and Fox models, are simply special cases of these
formulations with M = 2 and 14- 1, respectively-

111-9

Surplus production models generally ignore age, size, and sex-
population parameters and essentially treat a population as a single unit.
This approach requires several simplifying assumptions so that
MSY levels can be determined from only catch and effort data. These assump-
tions are:

9 The growth pattern of the stock is asymptotic.

9 There is no time lag between spawning and recruitment; thus, the
rate of population increase reacts instantaneously to.changes in
population size. .

0 The fishery is characterized by a stable age structure.

# Absence of fishing mortality would result in a fishery at a steady-
state carrying capacity.

9 The fishery is a definable unit stock.

o There is a functional relationship between yield and stock size
Cial as in Pella and
(i.e. , linear as in Schaefer [1954], exponew
Tomlinson [1969], or logarithmic as in Fox 11970].

o Natural mortality is independent of age and time.

e Population recruitment is constant.
9 The stock is at equilibrium level (Graham, 1935, only)

o The annual variability of economic, enviromental, and interspecific
factors are reflected in the annual catch data.

o The distribution of stock and exploitation effort is homogeneous.

9 The rate of natural increase at a given population biomass is inde-
pendent of the population's age composition.

Several modifications of the basic Schaefer surplus production model have
altered the basic equation to include the time lag that generally exists be-
tween the time of spawning and the time of recruitment (Walter, 1973, 1976
and 1978). This new form allows the use of a surplus production model for
fisheries in a state of disequilibrium. Mathematically, the equation includes
the same terms as eq. (1) above, except that population growth attributable to
recruitment is explicity accounted for. This model is represented as:

00
d in P = b - aP + ria (t-i) - qf (3)
dt 00

III-10

where

(b aP) expression of individual growth and natural mortality rate
Eri 6 (t-i) growth rate due to recruitment
qf fishing mortality.

Submodels are generally unnecessary for the application of surplus
production models. However, submodels for the normalization of fishing
effort are often required for fisheries where technological advances have
significantly altered relative catch per unit effort (i.e., relative catch-
abili ty) over time, such as in the case of the menhaden fishery (Schaaf and
Huntsman, 1972).
Table III-I presents the basic data requirements of all the types of
surplus production models, along with estimates of the size of the data set
and examples. In most equilibrium applications, surplus production models
can be utilized interchangably (i.e., Graham, Schaefer, Pella-Tomlinson, and Fox
models), depending on which yield-stock size relationship provides the "best
fit" with the curve for stock or catch versus effort.

3. Yield-Per-Recruit Models
There are two well-known yield-per-recruit models in use in fisheries
research and management. The first and less complex is the Beverton and
Holt (1957) model, which incorporates a specific growth function, par-
titions mortality losses into natural and fishing mortalities, and includes
an explicit accounting of the time lag between spawning and recruitment.
Mathematically, the Beverton-Holt approach defines
Y = F t=ty Rw e -Z(t.-tr) dt
t

where

Y weight of yield Cusually annual)

F instantaneous rate of fishing

R yearly number of recruits which enter fishery at age tr
Z E instantaneous total mortality rate (= F + instantanc=s natural
mortality rate)

t E age in years

t age of recruitment to fishery
r

t7 maximum age attainable
wt W.(l-e-K[.t-to])3 (the von Bertalanffy growth equation)
where

W. _E: asymptotic weight of a fish

K Brady growth coefficient (Brody, 1927, 1945)

to hypothetical age when fish would have zero length
according to Brody - von Bertalanffy growth relationship.

The primary difficulty associated with the application of the
Beverton-Holt yield-per-recruit model centers around its assumption of
the applicability of the von Bertalanffy growth submodel and the constancy
of natural and fishing mortality after recruitment. The additional
assumptions of a stable age distribution and unit stock status of the fishery
and of constant growth, mortality, and/or recruitment parameters as implicit
reflections of economic and environmental variability also limit the
applicability of the model.
The Beverton-Holt model is generally not used in management situations
due to the constraints of its model structure and the relatively large
amount of data necessary to implement its use (Table III-1). Examples of
its practical use are also given in the table.
The second approach to yield-per-recruit modeling is generally attributed
to Ricker (1958, 1978). This model allows for age-specific population func-
tions and also allows population parameters to vary throughout the year,
usually as a function of fishing season. For example, in a case where
a species is fished only during one season of the year, the fishing
mortality rate is zero during the remainder of the year. The Ricker model
is generally applied under steady-state or equilibrium conditions. Although
this ideal condition never really occurs'in nature, equilibriun conditions
may be approximated on a long-term average.

111-12

Under equilibrium conditions, the annual yield from all yearclasses
will be the same as the yield of a single cohort over its entire life
span and, according to the Ricker formulation, can be calculated as
Y (W) t'Y F(t) N(t) w(t) dt (5)
f
tr

where

Y(w) annual yield in weight

F(t) instantaneous fishing mortality coefficient at time t

N(t) number of individuals at time t

w(t) mean weight of an individual at time t
tr age of recruitment
t'Y maximum age.

Since the functions F(t), N(t), and w(t) are not generally smooth, well-
defined functions, Ricker has approximated the above equation by

F At (6)
Y(W) = E i I
i=tr

where, for each time interval t
i.1 the parameters are constant, and"Ni and vi
are the mean number and mean weight for the time interval ti.. This modeling
procedure generally uses,submodel@ of growth (e.g. von Bertalanffy growth
equation),,natural and fishing mortality, and recruitment submodels that are
age or size specific.
The approximation by Ricker effectively removes the major assumptions
of the Beverton-Fblt model, and holds the parameters of growth, natural and
fishing mortality, and recruitment constant only over short, user-defined
intervals. That is, the year can be divided into arbitrary time intervals
within which certain parameters can be considered constant. This approach

111-13

still assumes a unit stock with a stable age distribution and does not
account for partial recruitment.
The Ricker yield-per-recruit model is generally not used in manage-
ment situations because the data required for its implementation (Table
III-1) are not available. With a data collection program for these types
of data, this could prove to be a useful management tool for populations which
display stable age distributions and unit stocks. Examples of its use in the
management of fisheries (e.g. , Balsinger, 1974; Lett et al. , 1975 are
given in Table III-1.

4. Simulation Stock Management Nbdels_

Computer simulation models have been increasingly used in fisheries
research afid management programs over the past decade. In the context presented
here, the simulation model attempts to represent fishery yield or population
dynamics as a set of functionally related equations, such as

dp f (P,G,R,F,11@,MF,T,IPYPYlP TJ (7)
ft

where

dP/dt = change through time of the population level in either weight
or numbers

P = population level in biomass or numbers
G = individual or population growth

R = recrdtment

F = fecundity

natural mortality
MN
MF fishing mortality

T migratory behavior

Y annual catch

Y1 state history of the fishery (the historical data record of
the fishery, such as yield or effort)

111-14

T2 state history of the stock (the historical data record of
stock dynamics, such as historical population levels or
reproductive levels).

The simulation stock management model has a very flexible structure
in that its assumptions and mathematical descriptions are controlled by the
modeler. Various modifying factors on population growth, natural
mortality, and recruitment, such as habitat destructionenvirormental
pollution, and unusual meteorological events, can be incorporated into the
model structure. The drawback to this type of yield modeling is the heavy
data dependency of these models. In most cases,, large data collection efforts
would have to be initiated to formulate simulation models of Maryland
fisheries.
Single population models represent the least complex of the simulation
models used in predicting yield and/or population dynamics. Nbdel
structures have been formulated that summarize population dynamics in
terms of mass flow (Sissenwine, 1977; Orth, 1979), spatial dynamics
(Caddy, 1975), network theory (Lewis, 1976), bioenergetics (Kitchell et al.,
B61and) 1974). These simulation models often rely heavily on submodels for
the determination of growth, mortality, and recruitment rates. Examples
of single population simulation model structures which may be potentially
applicable to Maryland fisheries are given in Table III-1. In addition.,
new model structures could be formulated based on the collection of pertinent
data on Maryland fisheries.
The most complex simulation models for yield determinations are multi-
species or ecosystem level simulation models. Few examples
(Table III-1) of this type of model structure are available (e.g., Andersen
and Ursin., 1977), primarily due to the very large data sets necessary to
implement the models. For example, the Andersen and Ursin (1977) ecosystem
model of the North Sea requires over 800 sets of inputs and parameters to
estimate the yield of 12 fisheries. While large-scale simulation models
are not generally used in management, single population simulation models
represent the most used and often most useful model structure for the
determination of fishery dynamics and yields.

III-is

D. Data Dgendence of Stock Management Nbdei Types

Although not specifically stated in the characterization of stock
management model categories, data requirements and availability are an
important constraint on the choice of model structures for Maryland
fisheries. While data availability does not necessarily restrict model
usage as long as the data are,realistically obtainable at some future
time, time and economic constraints may complicate the collection of
additional data.

Table 111-2 shows the applicability of stock management model types
based on data availability, regardless of model assumptions and species
biology. This @ramework shows the pyramid-like compounding of usable
model st-ructures with increases in data. A large data base (i.e.,
#5) allows the use of all model structures, constrained only by
individual model assumptions and the population characteristics of the
individual fisheries.

111-16

Table 111-2. Applicable stock management model types as determined by data availability

WDEL TYPES
Statistical
Data Availability Dens ity Surplus Yield-Per- Simulation
Autoregressive Independent Production Recruit

1) Yield data only

2) Yield and environ-
mental data

3) Yield, effort, and
environmental data*

4) Specific popula-
tion parameters,
yield, effort, and PV
environmental data*

S) All major popula-
tion parameters,
yield, effort, and
environmental data

Environmental data not necessary for construction of some of the indicated models.

IV. INPUT SUBMODELS AND PARAMETER ESTDIATION

A. Selection Criteria and Search Procedures

The initial selection of articles potentially relevant to input sub-
models for yield models was based primarily on the title of the paper or
report. The key words used in this initial screening process of current
journals, reports, and books were:
0 parameter estimation
0 submodel
0 growth
0 mortality
0 fishing effort
0 recruitment
0 migration
0 fecundity
0 age structure
0 tagging studies
0 stock assessment.
The in-house journals and books reviewed were presented in Table 11-2.
in addition, we conducted bibliographic searches of the literature cited in
these papers and those selected as possibly relevant to yield modeling,
using the same key words.
TWo hundred and three articles concerning input submodels, data acquisi-
tion, and life history parameters were obtained through this search (including
those obtained through interlibrary loan). Fifty-seven articles were deter-
mined, through inspection of abstracts, to directly concern various input
submodels or the estimation of parameters for specific yield models. The
remaining 167 articles and reports concerned life history information on
INfaryland fish and shellfish species and methods for the collection of life
history and population information.

B. Categorization of Input Development

All procedures for developing inputs to yield models can be placed in
three major categories:

IV_l_

e parameter determination for specific yield models. e.g.,
methods for determining parameters a and b in eq. (1)

o functional submodels

9 data acquisition methods

Functional SUbModels can be further partitioned according to the
mechanism modeled into:

o age structure submodels (determination of age from length)

e allometric submodels (determination of biomass from length or width)

o catchability submodels (normalization of biomass from length or width)

9 effort submodels (normalization of effort through time)

@ fecundity submodels (egg production from size distribution data)

e growth submodels (change of growth rate with age)

o mortality submodels (change of mortality rate with age)

* recruitment submodels (change of recruitment rate with age structure
and time)

0 yearclass strength submodels (yearclass size from environmental
variables or stock size)

1. Parameter Determination for Specific Yield Nbdels

Parameter estimation procedures were reviewed using biological realism
and ease of application as criteria for potential utility to Maryland management
plans for fisheries. All of the parameter estimation articles reviewed concerned
the application of surplus production models and the estimation of parameters
a., bVand M in eq. (1), as dis'cussed earlier. Comments concerning the
usefulness of the procedures and abstracts of the individual articles
appear in Appendix B.

2. Functional Submodels

Fifty-seven articles were reviewed concerning submodels which could be
used as inputs for yield models. The most important submodels for input to

IV-2

standard yield models (surplus production and yield-per-recruit models)
are those for growth, mortality, recruitment, and effort. In addition, submodels
for age structure, fecundity, and yearclass strength could provide important
inputs for the construction of simulation yield models. Comments concerning
the utility of functional submodels and abstracts of the individual articles
evaluated appear in Appendix B. Relative data requirements of the
functional submodel types are given in Table IV-1.

3. Data Acquisition ','4ethods_

The collection of data acquisition procedures is docunented in'Appendix
C, subdivided according'to type of information. This bibliography is
directly applicable to Phase II of this project, where the present data
collection procedures will be reviewed and improvements and/or the
addition of new data acquisition techniques will be suggested. The
categories of data acquisition presently encompassed by the review include:

e stock assessment

* tagging studies

* deteimination of growth rates

9 determination of mortality rates

0 migration studies

* recruitment analyses

a population age structure deteiminations.

IV-3

Table IV-1. Relative data requirements of functional submodels and their potential utility as
input to stock management models

Submodel Type Applicable Yield Effo-rt Required Applicability Examples
Models for Data to Maryland
Acquisition Fisheries

Age Structure Simulation, yield - Limited Generally not useful Kimura, 1977
per-recruit (age-length keys are Kumar and Adams,
biased); information 1978
can be collected McNew and Sumer-
directly felt, 1978
Westrheim and
Ricker) 1978

,Allometric Simulation Limited Information easily Dame,, 1972
collected; Pienaar and Thom-
useful, particularly son, 1969
with shellfish
Catchability Surplus production Limited;2 Potentially usefu@ Paloheimo, 1961
intensive particularly for Rafail)@ 1977
(depending surplus production
on specific model applications
procedure)

Effort
Cominercial Surplus production Limited Useful for Maryland Nicholson, 1971
yield-per-recruit, fisheries, particu- Schaaf and Hunts-
simulation larly where techno- man) 1972
logical advances have
been prevalent
Recreational SurPlus production Intensive Useful for future Malvestuto et al.,
yield-per-recruit, recreational surveys 1978A 1979
simulation in Maryland Robson, 1961

1,2 See last page of table. M @ M M 0

Table IV-1. Continued.

Submodel Type Applicable Yield Effort Required Applicability Examples
Models For Data to Maryland
Acquisition. Fisheries

Fecundity Simulation Intensive May prove useful Brousseau) 1978a,,b
for the construc- Tsai and GibsonV
tion of simulation 1971
yield models

Growth Yield-per-recruit Limited to Necessary for the Allen, 1966
simulation intensive construction of Bayley, 1977
these yield models Brousseau, 1978, 1979
Chadwick et al., 1978
Cloern and Nichols,
1978
Dame, 1975
DeAngelis and
0
Contant, 1979
Gallucci and Quinnp
1979
Kitchell et al., 1977
Knight, 1969
Pratt and Campbell,
1956
Richards, 19S9
Taylor, 1962
Ursin, 1967
Yamaguchi, 1975

Mortality
Fishing Yield-per-recruit., Intensive Necessary to con- Brown et al., 1979
simulation struct these yield Francis, 1974
models Van Winkle et al.
1978
1,2 See last page of table. Young, 197S

Table IV-1. Continued.

Submodel Type Applicable Yield Effort Required Applicability Examples
Models for Data to Maryland
Acquisition Fisheries

Mortality (Cont.)
Natural Yield-per-recruit, Intensive Necessary to con- Brousseau, 1978
simulation struct these yield Butler and
models McDonald, 1979
Marten, 1978
Paloheimo., 1961
Polgar, 1978
Robson and Chapman,
1961
Ursin) 1967
Van Sickle., 1977
Ware, 1975

Recruitment Yield-per-recruit Intensive Necessary to con- Allen, 1968
simulation struct simulation Beverton and Holt,
models 19S7
Christensen et al.,1
1978
Ricker, 1954

Yearclasses Strength Simulation Intensive Procedure is potent- Stevens, 1977
ially applicable, Walter and Hoagman,
but existing sub- 1975
models are not appli-
cable

I Data set required for implementation is small., well defined,,
and easily obtainable.
2 Data set required for implementation is large, poorly defined,
and/or difficult to acquire.
3 In a combined structure which includes yield-per-recruit and
stock recruitment models.
M M M M M M M M @ M M M M M M M M M =

V. LIFE HISTORY CHARACTERIZATION OF EXPLOITED MARYLAND SPECIES

A. Objective

This portion of the study was designed to obtain information on certain
life history characteristics of selected exploited finfish and shellfish
species. Life history characteristics of interest were those frequently
incorporated into various mathematical fisheries models.

B. Selection of Species

Table V-1 lists all finfish and shellfish species which were reported
taken in commercial landings in Maryland tidewaters between 1971 and 1975.
Many of these species are not sought-7after.and are taken only.incidentally during
harvests of more desirable species. Others are taken in very small quantities
and have little importance in either commercial or recreational fisheries
in Maryland.
Before proceeding further, we decided that only a limited number of species
from Table V-1 should be considered for the project. Accordingly, species
for consideration were selected in consultation with personnel of
Maryland's Tidewater Fisheries Division and Coastal Resources Division and are
listed in Table V-2. Some of the criteria considered in the selection
process were: commercial and sport importance, residence characteristics,
longevity, and current abundance.
Pertinent life history characteristics chosen for inclusion in Table V-2
are presented as coluun headings. Highest priority was placed on obtaining
life history information for Maryland stocks of the selected species. If
such data could not be located., data from other Atlantic coast stocks
were included. Sources of life history information are listed in Appendix D.
To augment the in-house search for life history data, a draft copy of Table V-2
was distributed to the staff of the Tidewater Fisheries Division and other
fisheries research workers in Maryland and Virginia. Nunerous additions to
the table were obtained from these sources. Additional information is being
collected and this table will be regularly updated as the study progresses.

V-1

Table V-1. Finfish and shellfish species reported in commercial fisheries
statistics for Maryland, 1971-1975 (NOAA, 1971-1975)

Species Atlantic Ocean Chesareake Bav
and Tiibutarie's

Finfish

Alewife x x
Bluefish x x
Butterfish x x
Carp x
Catfish x x
Cod x
Crappie X
Croaker x x
Drum x x
Eels x
Blackback flounder x x
Fluke x x
Other flounders x x
Gizzard shad x
Red hake x
Sea herring x x
Hickory shad x x
Fba.choker x
King whiting x
Mackerel x
Menhaden x X
Mullet x x
Scup x
Sea bass x x
Sea trout x x
Shad x x
Gray shark x
Other sharks x
Spanish vackere! x x
Spot x x
Striped bass x x
Sturgeon x x
Suckers x
Sunfish x
Tauto@g x
*,.edte perch x x
Yellow Perch x x

Shellfish

Blue crab x x
Lobster x
Hard clam x x
Soft-shell clam x
Surf clam x x
Conch x x
Oyster x x
Squid x x
Terrapin x
Snapper x
1-brseshoe crab
I x

Total Finfish Soecies 31 30
Total Shellfish Species 8

V-2

Table V-2. Life history data on selected Maryland species; reference
numbers correspond to sources listed in Appendix D; migra-
tion refers to movements of exploitable age groups; blank
columns indicate that data have not yet been obtained or
do not exist.

V-3

Stock- Fecundity Effective Age of Type of Location of Catch-Size Catch-Age Mortality Growth
Species Recruitment Eggs Per Exploited Sexual Fishery in Migration Spawning Distribution Distribution Rates Rates
Relationship Female Age Group Maturity Maryland (yr)
(in Md.) ( ) (yr) (yr)

Alewife No strong Mean of 6 yr 3 - 4 Commercial: Anadromous; In Maryland, In Long Pond, Mostly age In Long Pond, Length by
Alosa relation- 100,000 [94] [27,52] pound nets, juveniles in selected Maine: group IV and Maine, annual age data:
pseudoharengus ship [57] gill nets, move seaward tributaries Age Length V in Long expected mor- juveniles in
between 50,000 - haul seines, from spawning of 0-3 ppt I 5.3 in. Pond, Maine V to VI = R.I.
adult stock 100,000 fyke nets, grounds in salinity II 8.7 in. [33] 78.68, age VI [27,48],
size and [27] hoop nets summer: adults [58,69] III 10.8 in. In North to VII = 74.78; ages I-IV in
production 48,000 - [49] migrate to IV 11.9 in. Carolina, age post-spawning Maine freshwater
niles in mean of spawining areas V 12.4 in. range: males mortality be- [33]
Virginia 229,000 in spring [33] = 3 to 7, tween 1954 - ages 0-IV in
rivers [20] [27,48] Date available females 1959 averaged Ches. Bay
[82] from Choptank = 3 to 8 41%, age groups [59]
and Susque- [71] combined juveniles in
hanna river [33] North Carolina
systems Freshwater [71]
[75] adult annual Monthly growth
mortality in rates, 1971:
R.I. = 38% James River =
after spawning 8.5 mm,
[27] Pamunkey River =
In Bride Lake, 9.7 mm,
Conn., adult Mattaponi
annual River = 10.7
mortalities: mm, Rappa-
1966 - 57.4% hannock River -
1967 - 48.6% 7.7 mm,
Tagged fish Potomac River
overall = = 12 mm
57.4% mor- [38]
tality
[20]
In North Carol-
lina 1978,
adult annual
mortality rate
= 44%
[71]

Stock- Vec fffective Age of Type of
SFQL'i V.1 Rec 1-111 t 11ten t it, undity lixploited Sexual Fishery in Migration Location of Catch-Size Catch-Age Wrtality (;I'O%VLli
telationshil- 'gs elNge Grouli Kiturity Kiryland Spawning Distribution DistrIbution Rates RaLes
Female (in W.) (V) (yr) (y r) (yr)

American Eel 413.000- 5 - 18 S - IS Commercial; Catadromous; Atlantic OLeart Length by
Ango i I la 561,000 (S31 JS31 eel pots, leave coastal (in Sargasso age data,
1531 pound nets rivers and Sea) juveniles
149,S81 estuaries in 158,S91 and adults in
fal I for Ches. Ray
spawning IS!))
grounds;
juveniles
enter
estuaries and
streams in.
spring
(531

American Shad Density- 2SO,000 2 4 ':ommercial: Anadromous; In Maryland, Data availabic In North In North Caro- @tontfily
Alosa dependent [651 JS21 611 nets, move occanward selected from Choptank Carolina, age lina, 1978, growth
.@K) R! ss (791 262,000 nound nets in fall, enter tributaries of and Susque- range: total annual rates, 1971;
JS2) Recreatiolill c')aStal w;1tel's 0 - I ppt hanna river male - 3-.7, mortality Jalives River
58,SOO (38,49) and the Ches. salinity systems female = 4-7 36% 12. 3 dwi,
659,000 Bay in spi ing [S8,59) 17S) 1711 1711 Nudunkey
(S21 [23,44,S81 River =
9. 7 wid,
Mattaponi
River = 9.0
did, Rappa-
hannock
River = t4l)
ldin, Potonlac
River - 123
tied
1381

lVeCtivc Age of Typl@ of
Stock- @eclfnd i ly I,
Spec i VS RecrOtirtent [-,Fg xploite,1 Sexual ri,;Ilery in Migration Location of Outch-Size Catch-Age Mor ta I i uy GrOwill
s Per Nge Grour Maturity l,laryland Spawning Distribution Distribution Ra tes Ra ( C.-
1(clation%bil runale (it' W.) (Y) tyr) (yr) (yr)
111tieback No strong Mean of 4 ConTiercial: Anadromotis; In Maryland, In North Caro- III Connecticut In North Caro- Monthly
llerrIng relation- 100,000 152 poulid nets, juveniles mi- in selected lina, 1978: River, both lina, 1978, growth rates,
Alosa ship JS71 gill nets, grates to tributaries 3U% were sexes present 411 adult 1971'. Jaiijes
ae--va I is Ulu River = 7.5
between 45,800 - hatil occan in fall, of 0 - 2 ppt males from in age classes ai al mor-
adidt stock 349,700 fail, PaImInkey
seines, adiilts teturn salinity 140-293 mm, III - Vil tality rate
size and 1701 fyke nets, to spawn in IS81 30% females [70) 1711 River = 6.3
production hoop nets spring from 140-310 Same range in film, @kltl_:ipolli
of jtive- [381; IS21 fall, 40% North Carolina River = 5.7
niles in rec rea - juveniles (III-VII) fall, Rappa-
Viyginia tional: from 58-110 171) haiuiock River
rivers dip nets fall = 7.0 fail,
[821 1711 171) Pototpiac River
Data available 9.0 fail
from Choptank 1381
and Sttsqije- Length 1)),
hajuia river age data,
systellLs juveniles ill
175) North
Carol ina
(711

Bhtefkh Circulation 112,000- 2 Conviercial: Migrate along Ti-x) spawning 24 - 30 i n. TL I'wo popula-
Pomatomus of conti- 195,OOC. [S91 pound nets; coast (south periods [18] tions defilled
;Z11T_F1_L r_1_x nental [S41 recre@l_ in fall, coastal water! > 12 i n. TI, by the size
she I f tional north in in Spring in -in recrea- of fish When
waters de- [541 spring) , move the Gulf tional fish- the first
termines inshore into Stream, sulffnel cry aitnual ring
magnil(Ide (lies. Bay in over the 195] rolins, in ftly:
of year May through Continental fish spatiied
class size Atigtist, and Shelf ill SI)rjllg
141 agaill in 14,28,581 south of Cape
September Ind.. liatteias that
October reach about
128,54,SBI 260 fail by end
of first
winter, fish
spawned in
summer ill the
Middle
Atlantic

Stock- 11M liffective Age of Type of
i(!S ",dily Hxploited Sexisal. Fishery in Migration 1AXatioll of Citch-Size Catch-Age Wrtatily G1 owl 11
Rec nki I ment Eggs Per ge Grour @Jiturity
@c I a t tonsil i f1tryland Spawning
Oll K1.) M (yr) (yr) Vilti-ibution Rates Ra I es
Billefish (yr)
Polliatolmls
saltatrix 14 i gill
(Cont. reach about
120 Ault by
end of
first
willLer

call) 36,000
z - 5 Coumercial , Primarily Near surface
208,ooo 15z] rec rea - inshore - in shal low.
fSzj t ional offshore weedy areas
(59) 1%21 of lakes,
ponds, and
s t roams
(521

-R-i _ck-e-
At I ant ic r 38,000- 1 3 3 Convnercial:_
Krdladen OrvLrwintering Polylial ille; 10-30 un 1 6 (in Flom Long Length by
spawner- 631,000 (24,S81 (SzJ pound nets, movements out Atlantic Mean length: Clies. Bay Island Somd age data:
Brevoortia rec ru i t , [S21 gill nets, Of @t'IrYland; Occan over Age I - 173 1-2) (Cape to Florida ages [-IV,
Mwin 113,000 purse no adults tile Continen- to 204m Cod, 3 and froin 1966- along tile
transport (121 scilles return ; tat Shelf and Age 2 - 221 older) 1968: histall- Atlantic
of larvae 1491 Migrate near mouth of to 275 nin (241 tanems Coast
ill1portant
Recruitment northward in Ches. Bay (24] fishing mor- (241
low in Atlantic in 1521
periods of spring, south- tality = t)s% Growth rate
ward in fall natural mor- slows
envi ronmen- tallty = 52i after age
tat 122,30,45) 1451 IV
fillctua-
tions. 1471
1121 Length by
age data
in Ches.
Pay
1591

Stock- Age of Type of
Fectindit.) ;x
I ploited SCxIIa I I:iqljcrY tocation of Catch -Size Ca tc li -Age Mortal ity Crowth
Species Recruitment 1: in Migration
.gg' Per e Grouli Katurity thryland Spaim ing Distrilmition Di st ri butiol I Rates PaLCS
te I a t. i onshi 1 F& a I c (in Ikki. (yr) (y r) (yr)
Spot 70,000 2 Come rc i at: Migrate off- Polylialine; In lower Ches. III Potomac
1,PiOSLOIOIIS !)0,000 ISS) linul shore in fall Atlantic Bay: in i ve I-: age
@ I _Ul t F, I -i _rI i--. f62) seines, for spawning, OLean spring from 5 (4) to
otter retnill to 1581 to 8 in., fil 13 on, adulis
trawls, (lies. Bay in fall from 8 up to 33-3S
fyke and spring to 14 in. C111
hoop nets; [14,29,S8] [Sq] IS81
rvcl,c;i - In Clies. Bay,
t i ona I rapid
[491 growth
during the
first
staiinier

C@ "'1011ner Flounder 967,000- 3 3 Comnercial Unave Clies. Ba) Polyflatine; In Ches. Bay:
Paralichtlyj 1,700,00( (Comier- (591 pound nets, in fall and Atlantic age I = 12
jentatus JSIJ cial) otter overwinter Occan - 18 c1n, age
> 2 trawls, along edge of Spawning 1 -3/4 = 24)
(Recrea- hatil. seines Continental occ(irs during - 26 un, age
tional) r(ICI'Va - Shelf (return offshore If = 27 - Z8
tional in spring) migration (311
[S6.Sgj f56,58,S91

Mite Catfish 1,000 1 - 2 Cotimcrcial: No apparent Still or nin- In Patumit
Ictaltims 3,500 1521 potind and seasonal mi- ning water in River:
catus (521 fyke nets gratory tell- coastal animal growti
Recrea- dencies streams, or = 25-45 imi,
t ional [SZJ tidal estu- length by
1591 aries age data,
[I'S2,591 ages II-XII

MMM Mom= M MON M M M w M M

I:ffcctivc Age of Type of
Stock- Fectuldity,"
!XPlO1tCk1 SCXual 11iSliery in Location of C.1tch -Size Ca tcl t.- Age Wrta I i ty GI owl It
Species Recru i ument liggs Per Nge Grotili Kiturity tUrylaild Migrition 'Spa'mling Distribution Distribution Rates Rat eg
1;cmale (in KI.) (9) Oq-) (y r) r)
White Perch Dominant S'210 - 2 - 3 Commercial: Anadromous; lit Maryland, In James Annual total Length by
Morone yearclasses 321,0100 IS41 potind nets, iiow shore- in selected River, range mortaliLN age data,
@!@tjcana occur Mostly fyke nets, ward and ill)- tributaries is 70-2S4 min; Jaines River: all ages,
(Virginia) between haul seines stream in of 0-11 1-,pt in York 69% for males James and
(31 SO'000 gill nets spring salinity River; 75-270 after Age IV York RiverF,-
and I'(1crea - 1291 (tidal fresh- min females after first year
IS0,000 tional oligolialille) 13) Age V]; York growth
154) [581 [58,Sgj River. mates greater titan
Age III = 591, in any later
frimiles Age V year, in
57% hoLh SeXe@;
[31 awrige
annual
gr(Aqth
decreases
rapidly
after Age

131
Length by
age data,
ages I-X'
Potomac
River
1581
Growth late
sex
dependent at
least to age
5 yr
J88,3)
Growth rate
influenced
by density.
dependent
factors,
primarily
intra-
specific
cootpetition
fo r fo(YJ
1881

SLoc k- Fect ind i t .Ffertive Age of 7@1)e 0 f
S LxPlotted Sextial 1:1.glicry in location of Ca tch- Size Catch-Age Wrtal ity Growl 11
i 0.% Recrid tmcnt Eggs Per %ge Groul, Maturity K-trylauki Migration
Ru ]a t i onsh i Fmalc Spavining Dist.rikution Distribution Rates PaLes
Ull KI-) M (yr) (yr) (yr)
Yellow Perch Dom i nant 3,03S - 3 - 12 2 - 3 Commrciat: Semi -anadro- In Maryland, > 8 in.T1. > 3 yrs In Sevein Lengd) by
Porca yea r 109,000 111 54 j (11,S41 bo L tcon 111011S. lTk1VC ill sviecred 1111 River (M.), age JaLa,
n-ave-scens classes Mean of trawls, from lower tribtutarles fishing mor- ages It-
occm- (in 23,, 0 gill nets tribntaries of 0-5 ppt tality Vill,
One i da IS41 Feck-01 - to slla%,qlillg salinity greater than Patment
Lake, N.Y.) S,000 - tion.11 grounds in 1581 131 (all River
1131 75,000 [26,581 early spring sizes) (21
Year - 121 154,S91 ill) In Severn
classes S'900 - Mean annual Rivet,
variable 109,000 mortality length 1)),
from 19SS- [11] rates: a age dat.a,
1959 in 10,000- I-VI =70% fur ai!cs I-XII:
.Severn 157,000 males and 51% fouales
River, Nkl. 141] for females; reach legal
highest from size (8 ill.
(87%); during age
lowest front 1-11, males
IV-V (15%) (in during It-
CD Missouri ill
River, 196S- Gj Owth
1974) rates dc-
1251 ciease with
a9v I
uspecia I I y
in males
pq
IA-ligtil 11)r
age data,
ages I-IX,
Hissouri
River
reservoi rs.

ike
In Red L,
Milul: total
first-yeal.
growth
calcoilated
for 15
years
(19S2-1907);

Stock- 1:eCtIlKlityliffec-tive Age of T)q)c of
Spec i vs Recritilment Fggs Per Exploited Sexual Fishery in Location of Catch -.Size Catch-Age Mortalily Growth
IZOU1001'shil Conale Nge Groul NI-ittirity p,farylaIKI Migration Spawn i ng Distribution Distribution Rates Rates
(in M.) (yr) (yr) (yr)

Yellow Percl)
Perca
ri-avescells after
second year
(Cont. of life,
females grow
fastev
than males;
average rate
of growth
during mid-
sLainer
0.722 nn/day
pq
W Like
Michigan
females
longer atid
heavier than
males (after
age 11)
Length by
age (II-VII)
data, Lake
Michigan,
Lake Eric,
Green Bay, attJ
Saginow Bay
141)
Length by
age data in
Poto"'ac River.
ages I-XII

liffective Age of 1`ypC of
Stock- 1:ecundity
Spec i es Recruitment E.911S Per lixplolte(I %Cxull Fishery in Migration Location of Gatch-Size Catch-Age Mortality Growth
lelationshil 1-onale Nge Groiij, Katurity I,L-Irylan(l Spaw"ing Distribution Distribution Hates Rat C.;
(in W. (9) (yr) (Y r) (yr)
CLUR Strong re- Number 2 yr in I - 2 Commercial: None In Ches. Bay, In Mass., high In Prince
t!)Y-111 arenaria crultment of Kass. (4S Pail conveyors, salinities of mortality in qi I I imil
in Maryland oocytes (421 shell dredges, 10 ppt in Pelagic, meta- Sotuid, Alaska,
has occurre( produced length, tongs spring and by ' ge
norphic, and length 11
every 10-15 increaseF in Kiss) 149,SO] 15 ppl, in settlei@ent (O-XII) data
years in th(exponent- 1421 fall stages Rates 1341
Potomac ially ps] decrease with Length by age
eswary (2 with in- age, hotvever (O-Vil) data
peaks of creasing 1421 . rom
spawnbig female ;toucester'
per year) body size lass.
JIS,581 Average 1421
Large oocyte Linear shell
annual produc- 'rowth rates,
variations tion in early
in re- 60inii rowth in-
@rld Ownt clamt. :rements, and
in Mass. abou ;ize-specific
and fiwt- 1.20,000 tean seasonal
uations (39] liell growth
largely dw rates; Von
to natural Wrtalaoffy
variat ion equation used
and not to inwsti-
to changes gate rela-
in popula- tionship
tion feciin- between age
dity and linear
1391 size (Glou-
-ester, Mass)
19()l

Stock- 1:ect"Idi liffective Age of of
tT.Xplotted Sexual F, i@sl I
Spec i e r, Rvc n I i 1111011 t I'ggs Per cry in Location of Catch -size Catch-Age Mortal i ty CI ol-11 h
Migration
Ngc Gronli Miturity I'laryland Spawni jig Distribution Distribution Rites Rates
(ill hd (9) (YO (yr) (y r)
llard Clain Little Over 3 Colonercial: None Unbainnents Most mortal i ty Ili Nariagan-
l'iercenaria knoini 1 1/2 in. 1781 dredges; along tile on small (less seLt Bay,
ii@-rccnart5* (setting in 1'(,C [Ta - wcstern than IS nan) average
occurs on a length; tional Atlantic individuals; . glowth (shell
fairly reg- assumed 191 coast frow -ates decrease length) in-
ular annual age Large Gulf of St. with age (in crelliclits
basis in (3-4 yr) @laryland Lawrence to N.C.) inversely
Great South 176,771 recrea- Florida (361 related to
Pay and Ntaximun t iolla 1 i3l)
ifor Natural mor- initial
seshoe age catch tality data length;
Cove, N.,J.) about 8 [951 from South- growth rates
1.111 yrs hampton, Shot-i high
Sporadic (761 lingland,avail- variation in
wi th able
occas i onal different
major sets; 176) parts of tile
spauniing Annual fishing Bay; more
occurs each mortality in than half
year and one location tile year's
r cruitment in Chinco- growth
r0aittlie may teague Bay was occur, be-
be (Ille to 33S fore mid-
factors 1771 July;
operating glowth rates
Oil larval a function
stages of diatom
1761 abinidance
Setting may and sedi-
he influ- Inc-lit type
enced by (lower
substrate growth in
character- hiqlier silt-
istics ; clay content)
d is tr it,&% 191
tion of set In South
j@ay not he Ca rolina
influenced (tray exper-
by distrl- intents)
bution of growth did
Mill ts not signifi-
cantly vary
1671 amon@
densities

Stock- Fecundity Effectivi: Age of I)q)c 0 f
lixploited Sexual r-isli
Spec it: s RecMit"'Ont Eggs Per kge Grouli K Ory tit Migration Location of Catch-Size Catch-Age Wrtality (;I.UWLII
Relationshil Foitale iturity Maryland Spawming Distribution Distribution Itates
On W.) (9) (Yr) (yr) (Yr)

liald ClW11 Average
W-1-cenal-ia 1114milily
utercenaria vates: BtjU
Bay = 0.8
11011, Clark
Sulu Ill - 1. S
mu, Albur -
gottie C'ruvk
1. 8 null
(361
0-awth data
avai lable
fi-till
Sotithailptoll,
lingland
l7ol

'Blue Cral)
Skme expel-- 700,000- 2 4 2 Fertilized Lower portion Maryland StaLt: RAIdlity fl%All
Callinectes iukcntal 2,000,001 [Sal 1581 trot lines, females of Cites. Bay law, 111ilkillualk egg to adult Mq)id gl'L)Wtil
;5 1-)1 a u s evidence 166,741 crab pots. migrate soutil (27 ppt) of 5-111. 0.999999 icaching
that scrapes in Bay to 17,581 carapace I w) I adtilt size
uk!galopae relcFea - more saline width @ in Ilk Chesapeake (S-6 in.)
and juve- tional water (17 ppt); (liesapeake bay, percent I -121 Y r
niles- (49,SOI ju leni les Bay, 1969 wintel after
respons ible Move ulk Bay all ,vointer likodal
for re- mortality of hatching;
and into. class about adults: quai I c rahs
crui tuient tributaries 75 um wide, 1968 = 3.3t shed fre -
to 17,46,661 spring - 1969 = 19.3% (ILICIltly, bilt
estuaries about 140 Ria 1970 = 351 tblic bctlvt@en
vid illkifti- [73,741 1971 = 7.6% molts in-
gration
(71 1972 = 5.1% creasus as
1973 = 11.4% crabs grow
t-/3,741 largei* ;
each nolmal
--hWding,
Width

M mm @m Mao m Im =I M M w

Hiffectivc Age of '1)1)e of
spc@ it! S Recru i twent I' @&ploited Sexual Fisliery III Wcation of Catclk-Size Catclk-Age Mortality Cruwth
.gg Migration
tclationshil@ 1;cau s Pel' Nge GrouF 1,1aturity Maryland Spatming Distribution Distritkiltion Wites Rates
ale (in W.) (yr) (y (y

Blue clab increases
callinectes 1/4 - 1/3
1 11=1 El I I @' tile initial
(colit.) size
loo]
Crabs
hatched fl-CAll
eggs spawned
in late
$L111ulkell
rcquiro Ill -
20 Illuntlis
to attain
full size
174J
GI-cutil peaks
ilk mid-swi-
laul. ilk (lies.
BUY itn't
declines ill
fill
1731
Growth
ceii@cs
'1111 i lie the
winter
11jujiths ( ill
Ches. Bay)
191J

Stock- Vecundity Fiffectiv Age of Tylle of
Spvc. ics, Recriiii.irictit Eggs Per 1-'.xploite( Sexual Fisliery 'tit Migration Location of Catch-Size Catch-Age @lovtatiq Crowdi
Re]-ationshil Fonale @ge Grotil Riturity Kiryland Spal-41ing Distribution Distribution Rates Rates
(in 1,1J.) (9) (yr) (yr) (Yr)
American Oyster Settlement I X 10 5 Over I Commercial: None W. waters, In Delaware Depends on In Clajibank In South
Crassostrea depends on to 7 2 to 3 100,681 hand tongs, Meso- Bay: years of Creek, S.C., Carolina
available I X if" yr patent polyhaline dominant set.; moitality of growtb In
:111tril, [601 [81) tongs, (58) @ge Size [811 larvae up to both length
water temp., dredges Greater than late utibo and width
salinity, 149,50) 5 ppt 1 0. 1-2.4 cm stage equals is faster
and recrea - 168] 11 2. S -6. 9em 611; from egg for yoLuig
dissolved tional III to spat equals oysters,
oxygen fisbery, & 7.0-14.0 0.99997 decl ining
(58,601 limit I older (311 1691 with age
Major sets bushLI/Per- [721 histantan-
in Gies. son/day cous
Bay in 1964 r9sl rrowth
1968, 1974- rates
75, and given
t977, btit show little
nagniti0c difference
of set between
varied interticlat
narkedly by and sub-
Hay region ti'dal
181) oysters
Settlement 1931
variable on
lifferent
hars in any
year and
variable
oil smne bar
in
di fferent
years
1871

M MM M M no M M M W some MMO@ M MM

irrective Age of 1`ypV of
Stock- Fecundity
1'.xploltef-I Sexual I:isllel-y in Location of Catch-Size Gitch-Age tic) rta I i ty Growth
Spec i es Rec I I I i t Iflent Eggs Per Nge Grour Kiturity t1-1 1. yja ld Migration
Relationshil Fella I e I Spawning 1) i s t r I bu t i oil D i s t r i bution Rates Ral es
On tu M (YO (y r) (yr)
Atlantic Croaker Only suc- 180,000 1 - 2 Cctmmercial : Move up bays Atlantic Total allnual Length by
Mict-pp2gonias cessful [Sq) IS,SSJ pound nets, and estuaries Occan, mortality rate age (0-11)
iiil-tlatus yearclass haul during spring; polyhaline 96% in lata in
recnjitmentF seines, occanwayd in IS8,591 Caroliikian :arolinian
in Maryland drift nets fall Province Province
in 1974 recrea- [SS,58,591 (short life
and 197S tional span) In
(941 [581 (51 Louisiana,
Ekinan traits rowth
port may ates of
in f I tience juvelli I es
yea rc las s fron
strength O.S1 to 0.99
1951 imi/day
1171

Striped Bass Dominant 62,000 2 9 4 5 Cominercial: Anadromous In tintryland, Data available Data available Instaittwieous Juveiki le
Morone year - 112,000 [S4,58) [S41 gill nets, with soliie pop- waters of 0 from Choptalik from Choptank fishing and growt1k late!;
@-a-xat I I is classes eggs are Pound nets illation seg- 3.5 ppt and Susque- and SusqLl--- natural mor- average 0.35
occur produced haul ments leaving salinity hanna river hanna river tality rates sim/day ill
(58,401 for ever) seines KI. waters (tidal-fresh systems systems for popula- k1hemarle
pound of recrea - Upriver spawn- to [7S] (751 tions in iouiad, N.C.,
1-dy wgt t ional ing migrations oligobaline) Virginia, Nord iverage 0.46
164] 1491 in April & May IS81 Carolina, and mi/day in
400,000- (29,581 California, it tudson
I'OSS'00( Potomac River. Uver; in
[63) 1961, 40% Potomac
mean of adult moi-- River,aver-
2,462,00( tality in age (1975)
[611 spring fisher); 0.45 aim/day
in Ches. Bay, (1976) =
35% mortality 0.46 inn/day
of 3-yr old [()I]
males in Length by
spring age data
fiShCry ; ill (all age
Iludson River, classes)
mortality (58,611
rates:

Effective Age of Type. of
Stock- Fecundity lixploited Sexual FLshcrj in location of Catch- Size Catch -Age Mor ta I i I y Crowill
Spec i es Recruitment Eggs Per Age Grotil, Maturity Migration
telat ionshil I-en"ll e I'larylaW Spawning Distrilmition Distribution Ra I e s Rat eq
(.ill R1. (yr) (yr) (yr)

Striped Bass Age
Morone 0 = .99.96%
!@,ii a 6 1 is I = .666%
(Cont. 11 = .40%
[oil

Weakfish Conanercial 45,000 - I - 4 1 - 2 Commercial: Move tip bay Atlantic 17S-350,iin I yr - 621
@j,noscion record in 1,726,000 (Va.) 155] pound nets, and return to Ocem, meso- with motlal 2 yr - 28%
rbi:19 I Is Virginia [SSI l8sl otter occan after potylialine length at 3 yr - 8% In (lies.
suggests trawls, hail spaurning [S81 225 11111 4 yr - I% Ray, length
little seines, 129,551 Lower Ches. (pound nets. (Virginia, awrages
dens i ty pin-se nets; Possible Bay inlets, Virginia, 19S4-19S8) 3-1/2 to 6 i n.
dependence rccrea- di fferential colle s 1954-1958) [8sl at 1/2 yr.
(1891)- tional migration ISS] 1851 8 to 10-112 in.
19S911 lIR,SR,S9, by age -it I
IRS1 85) class 1,12 Yr
00

Winter HoLoider -Partial Average > 2 3 - 5 Commercial fit winter, move In Maryland In Cape Cod, Range 2 - 4 [lost - rec ru it
Pscudopleuro-- recrni t- 500,000 (ftrss- 1561 otter toward shallow waters of >5 two peaks: (Cape Cod) ment total
al waters plit salinity, 30 and 40 cm; 1321 mortality 1 .0
5FCtes ment to JS6] [32) trawls, coast, r In (1&-.
an FlEallus Mass. pound nets, and estuaries inesolialine range, 25-SO on Sex ratio 13:1)" new]).
1970 in
fishery in fyke nets to spat-ni sha I I ow 1321 70/30 Cape Or)d hatched
age classes I'MMI - JS6,581 coastal (female to WOS 3 - 3. 5 nun
t- IV tional waters and male in in,. tantancons in length
Age (491 estuaries Mass.) 111or fa I i t age I = 108
y 178
It 401 1581 (891 rates for: Rim to
Ill 80% fishing 24% Him
IV 81% notin-al fit
1321 1321
InstanLaneolis
fishing
mottality
rales = 21%
1891

M M M MIM VWSWWSNWI@@ M Nam

C. Categorization of Selected Species with Respect to Model Applicability

The type of management model applicable to a species depends on many
aspects of the species life history, as noted above. From a regional
perspective, however, two characteristics ass@he major importance: migra-
tion patterns and location of spawning. Both are major determinants of
whether populations can be considered as unit stocks, a basic requirement
of most stock management models.
Many of the select@d species exhibit similar migration and spawning
patterns and can thus be grouped according to these patterns, as, for example,
in Fig. V-1. The first criterion used for separation of species is
motility. Population levels of non-motile species (hard clams, soft clam3'
oysters) in a given location are not influenced by movements or migrations.
Thus, evaluation of exploitation and/or environmental perturbation is not
complicated by abundance changes caused by imnigratibn and emigration.
Motile species can be divided according to the portion of their life
cycle spent in Maryland. Four species (white and yellow perch, carp,
white catfish) complete their entire life cycles in Maryland waters. As a
result, exploitation can be monitored and totally controlled, which is
particularly relevant to recommendations on management options. Consequences
of envirormental degradation can also be clearly defined for these species.
Of species which spend only a part of their life cycles in Maryland, five
are considered to spawn here (alewife, blueback, American shad, striped'bass,
winter flounder). The one non-anadromous species in this group, winter floun-
der, may exhibit only limitecl spawning in Maryland. However, because there
is scme evidence that it spawns in the lower Potcmac Ri,,er (Appendix D, Ref. 58),
we are including it in this category.
Fish stocks are often delineated on the basis of spawning location
because individuals tend to spawn at the location at which they themselves
were spawned. This is dramatically clear in the case of anadromous species, but
also holds for others, including many open-ocean spawners (Harden Jones, 1970).
This fact is particularly important relative to management in Maryland
because spawning success determines the size of a stock available for
exploitation. Thus management options for these species would include
actions impacting on their spawning success.

V-19

Figure V-1. Procedure followed in grouping selected Maryland exploited
finfish and shellfish species according to major life history
characteristics.

All S@ecies Decision Criteria

INTon-motile Mo motility

Hard Clam
Soft clam
Oyster

------------------------- --------------------------------------------------

Entire life Only part Maryland
cycle in of life Residence
Maryland cycle in
Maryland
White perch
Yellow perch
Carp
White catfish

------------------------ ------------ --- --- -------------------------------

spawning Do not spaivm Location of
in in Maryland Spawning
Maryland

Alewife
Blueback
American shad
Striped bass
Winter flounder

--------------------- --------------- ---------- ----------------------------

Distinct No distinct Stock
stock in stock in Definition
Maryland Maryland

Bluefish Blue crab
American eel
Menhaden
Spot
Croaker
Weakfish
ad
,

s
@h
sh
_____ 4
------ ------------ -

not s
Maryl
under/
....... .......

0 js

Summer flounder

V-20

The final species group includes those which spend only part of their
life cycle in Maryland and spawn outside Maryland waters. This group
might be separable, given sufficient data, into two groups: species in
which the same juveniles or adults of a unit stock return each year to
Maryland waters (possibly bluefish), and those in which the individuals
returning each year represent a random segment of a stock with wide
distribution over non-Maryland waters (blue crab, American eel, menhaden,
spot, croaker, summer flounder, weakfish). This latter group presents the
most complex problem in modeling: how to determine impacts of Maryland
exploitation on the source stock.

V-21

VI. EVALUATION OF MODEL APPLICABILITY

A. Introduction

M@del and species categorizations provide a basis for evaluating the
usefulness of various model types as management aids for different
Maryland species or species groups. In assessing suitability, data avail-
ability plays as great a role as do model structure and assum ptions.
A generalized discussion of evaluation procedures is presented below,
followed by specific discussion of individual species in which model
categories are addressed. The resulting evaluations are thus consistent
and objective.

B. Model Selection Scheme

The decision as to which of the stock management model types already
discussed should be applied to any particular s',pecies -must be based on data
availability, management objectives, and the simplifying assumptions concern-
ing the stock life history characteristics of the individual models. Figure
VI-1 presents an example of this process where the management objective is
assumed to be the determination of annual yield (either maximal or optimal).
The selection method is represented as a two-gate process where data avail-
ability singularly determines the potentially applicable model type(s), and
model assumptions and species life history characteristics determine the
applicability of specific models within a major model category.

Figure VI-2 presents the generalized selection scheme expanded
to incorporate the specific model assunptions into the decision criteria.
For example, if growth, fishing and natural mortality, yield and effort data
were available for species j, yield-per-recruit, surplus production, and
statistical stock management models would all be potentially applicable to
the modeling of the dynamics of stock j. In general, assumptions
concerning the motility, growth, and mortality of species j affect the selection
of a specific yield-per Tecruit model. If species j is mobile, displays
age-dependent growth and mortality, shows no major environmental effects on
stock dynamics, and shows no major economic changes in its exploitatior;

VI-1

M M Im W M to @ to W we no,

Figure VI-1. Conceptual overview of model selection process
Data Inputs Model Types Decision Criteria and Result
Selection Process-

Extrinsic Model Assumptions
Data Concerning Species
(e.g., yield, Life History
effort)

Model Categories

A) Statistical
B) Surplus
Production Specific
Data C) Yield-per- Selection Model(s)
Availability recruit Process Selected
D) Simulation

Species Life History
Intrinsic Characteristics
Data Inputs
(e.g., growth,
mortality)

I
I
. I
I
I
i
I
i
Figure VI-2. Detailed conceptual overview of model evaluation scheme
I
I
I
I
i
I
I
I
I
I
VI-3 I

Autoregressive
Yield Unta
Statiqtical

Models

B Correlative

Density-Depen-
A Environmental dent or Inde-
Da ta
pendent Models

Spatial Surplus Basic Surplus

Production Models Production Mode1q

NO YES
Surplus Species YES ZA NO Instantaneous
g
Production .. Fecruitment?
B 6 Ellort Data Mobile? 10K D-pen:;nt
rowth Rate?
Models

A NO
Collect Data Lagged Sur
on Growth and
Recruitment Frc duc t It)"

th and Yleld-Per-
Dat.1 -----*Rccruit models Spatial Yield- Beverton-11olt
Per- Rec ru it Yield-Per-Recruit
Models Model

NO NO

YES VO YES Age
S ecies Bertal@ ffy Dependent
Mobil
rowth? Mortalities?

NO YES

Collect
Data on YES Partial NO Ricker Yield-Pcr-
Population Recruitment? Rer-ruit Model

Parameters
n
an
11 C11

plus

Hodelo.

L_O_@Spe
lo@

NO
itment
L-T@ <@Re@cru
i
E

D,Ita on A] I Simulation
Popu tat ion Para Models
meters, Yield,
and Efrort
@ V1-4

then a Ricker yield -per- recruit model would be the best yield -per- recruit
formulation available. In addition, either autoregressive or correlative
statistical stock management models could be as potentially useful for the
management of stock j as the yield-per-recruit model. According to the
arguments presented in Fig. VI-2, surplus production models would be inap-
propriate because stock j displays age-dependent growth, and a detailed simulation
model could not be constructed without additional sources of data. A simple
simulation model which incorporated only growth and mortality could be used
in this case. Preliminary applications of the procedure described in Fig. VI-2
to each of the selected species are discussed below.

C. Species-Model Evaluations

1. Hard Clam (Nlercenaria mercenaria)

9 Category -- non-motile; entire life history in Maryland
* Ramifications of category
All exploitation of Maryland stocks is under Maryland regulatory
authority.
Quality of environment in Maryland can influence all life stages
occurring in Ma-riland: spawning success, growth, and survival.
The status of stocks might be influenced by the recruitment of
larvae from stocks outside Maryland waters.
The lack of immigration or emigration of harvestable stocks permits
management of harvest on an areal (spatial) basis.
9 Stock characterization
From settlement to harvest, stocks can be defined on an areal
basis.
Because of possible recruitment of larvae from outside Maryland waters,
a "true" unit stock cannot be defined for Maryland alone; but
such a definition might be possible on an interstate basis.
* Stock-recruitment relationships
No stock-recruitment relationship is documented in the literature;
many environjwntal factors appear to influence setting (Appendix
D, Refs. 76, 77),- cy settlement is documented.
,regarious
Low stock densities may limit commercial fishing as a result of
the economics of harvesting. If the fecundity of non-harvestable
densities is sufficient to C, generate large sets, stock-recruitment
0
relationships are irrelevant to management.

VI-5

o Statistical model application
Relationships between success of set and various potentially
important environmental variables could be investigated by use
of correlation analyses.
Statistical models could be used for predictions of years
of good sets, which might be of value as inputs for
bioecon6mic models or as indicators of environmental problems.
o Surplus production model application
The concept of surplus production has not been applied to shell-
fisheries in the literature reviewed. Although a theoretical
upper limit to population size exists simply by virtue of
space limitations, we do not know the growth response of a stock
to changes in density, necessary for the application of this
model type (See Section III.C.2).
The assumption of homogeneous distribution of stock and exploita-
tion is probably not met for this species, particularly in
Maryland (Appendix D; Ref. 77).
Current knowledge of the population biology of this species is
insufficient for assessing the applicability of this model type.
Yield-per-recruit model application
Because natural mortality rates vary with size (Appendix D, Ref. 36),
application of a Beverton-Holt model might be limited to upper size
ranges, or a Ricker model would have to be Used.
The homogeneous distribution of stock and exploitation assumed in
this model structure is certainly violated (Appendix D, Ref. 77), and
the consequences would have to be investigated.

The stock-recruitment relationship is not known; but, if the
stock is considered to be at equilibrium, a single estimate of
number of recruits could be obtained and used with a Ricker
model; some data suggest fairly regular recruitment over periods
of several years (Appendix D, Refs. 67,76).
Yield-per-recruit models would be useful- for eval-uat*l:ng t'h'e con-
sequences of various exploitation strategies (e.g., the size
limits) on yield. They could be used as inputs to bioeconomics
models for maximizing monetary return from sets of variable magnitude.

Simulation model application
The single shellfish management model found in the literature
review was a simulation model of a sea scallop stock (Appendix
A, Caddy, 1977). This model accounts for many distinct charac-
teristics of shellfish stocks and fisheries and might prove to
be a useful prototyDe for hard clams; heterogeneous spatial
distributions can be divided into zones within which density
and exploitation are honogeneous.

VI-6

Simulation models provide a means of accounting for non-homo-
geneous spatial distribution of stock and fishing effort; they
would be most appropriate where precise management decisions
are necessary or where precise inputs to bioeconomics models
are desirable.

9 Summation
The non-motility of clam populations and the possibility that
recruitment might be independent of stock size differentiate clam
fisheries from most of those which have been modeled in the
literature.
The simulation model for scallops described by Caddy (1977),
which incorporates a Ricker yield-per-recruit model and spatial
partitioning of the stock, would appear to be the most fruitful
C,
approach to modeling this species.

2. Soft Clam 04ya arenaria)

e Category - non-motile; entire life history in Maryland
9 Ramifications of category
-- Same as for hard clam (1)
9 Stock characterization
-- Same as for hard clam (1)
* Stock-recruitment relationships
No stock-recruitment relationship documented in the literature.
Large sets have occurred occasionally over the last 20 years, but
are uncommon (Table V-2); this suggests that major sets are
environmentally determined (i.e., stock-recruitment relationships
are very weak); one statement appears in the literature that n@a
is a non-compensatory species (Appendix D; Refs. 39,42); areaarious
settlement is documented.

Low stock densities may limit commercial fishing as a result of
the economics of harvesting; if the reproduction of non-harvest-
able densities is sufficient to generate large sets, stock-recruit
ment relationships are irrelevant to management.
* Statistical model application
-- Same as for hard clam (1)
e Surplus production model application
This 9tock does not fulfill the assumption of homogeneous stock and
exploitation distributions (e.g., Appendix D, Ref. 80); setting is
strongly influenced by substrate type, and fishing.intensity is
influenced by the density of harvestable clams (Appendix D, Ref. 58).
Many other assumptions may not be met, i.e., the possible weakness
of the stock recruitment relationship suggests an equilibrium
stock level may not be attainable.

VI-7

The concept of surplus production has not been applied to shellfish-
eries in the literature reviewed; a theoretical upper limit to
population size exists simply by virtue of space limitations;
since the species may be non-compensatory (Appendix D, Ref. 42),
the concept of surplus production might not apply.

a Yield-per-recruit model application

Because natural mortality rates vary with size (Appendix D, Ref.
36), application of a Beverton-Holt model might be limited to
upper size ranges, or a Ricker model would have to be used.

The homogeneous distribution of stock and exploitation assuned
is certainly violated (Appendix D, Ref. 58), and consequences
would have to be investigated.

Yield-per-recruit models would be useful for evaluating the con-
sequences of various exploitation strategies (e.g., vari-
ous size limits) on yield; they could be used as inputs to bio-
economics models for maximizing monetary return from sets of
variable magnitude.

e Simulation model application

Same as for hard clam (1)

9 Sumnation

The non-motility of clam populations and the possibility that
recruitment might be independent of stock size differentiate
clam fisheries from most of those which have been modeled in
the literature.

The simulation model for scallops described by Caddy (1977)
(Appendix A), which incorporates a Ricker yield-per-recruit
model., would appear to be the most fruitful approach to modeling
this species.

Soft clams in exploited populations have a much shorter life
span than hard clams (1-2 yrs vs 3-8 yrs, Table V-1); thus, approaches
to management would differ for the two species; for
soft clams, where the stock-recruitment relationship is weak,
a bioeconomics model might be useful to maximize financial return
and allocate effort equitably.

3. Oyster (Crassostrea. virginica.)

9 Category - non-motile; entire life history in Maryland
o Ramifications of category
-- Same as for hard clan (1)
9 Stock characterization
Same as for hard clam (1)

VI-8

0 Stock-recruitment relationship
Successful sets.are dependent on the availability of suitable
substrate; however, substrate availabilit)r is not sufficient
to ensure good sets (Appendix D, Ref. 81).
Environmental factors (e.g., temperature, salinity) appear to
strongly influence the magnitude of a set.
The sporadic occurrence of major sets-over the past 30.years-
and the localized character of recent major sets (Appendix D,
Refs. 81, 87), suggest that the stock-recruitment relationship
if any, is very weak.
9 Statistical model application
.-- Same as for hard clam (1)
* Surplus production model application
Many model assumptions-may not be met; for instance, an equi.-
librium level oyster stock may not be definable, and stock
and exploitation distributions- are not homogeneous.
The concept of surplus production has not been applied to shell--
fisheries in the literature reviewed; a theoretical upper limit to
population size exists simply by virtue of space limitations.
However, we do not know the respons-e of a stock to changes in
density, necessary for application of this model type (5ee
Section 111. C.2).
* Yield--oer-recruit model application
Assumptions of homogeneous stock and exploitation distributions
are not met.
Significant by-catch mortality to pre-recruitment oysters may
occur due to fishing of harvestable oysters (Appendix D, Ref. 81);
therefore, age-specific mortality rates not only differ, but
may actually be interdependent; yield-per-recruit functions
cannot directly accommodate such-relationships.
Yield-per-recruit models,do not appear to be a practical modeling
procedure for oyster management.
0 Simulation model application
--.Oysters are unique among exploited stocks in that they occur in
discrete bars; and stock dynamics vary with location and over
time (Appendix D, Ref. 58). Thus, the only feasible modeling
approach is simulation modeling, which can be adapted to accommodate
high levels of complexity.in stock behavior.
@bdeling of stock production might be valuable as a tool in the
management of cultured stocks (i.e., those created by seed
and cultch planting).
A simulation model incorporating economic aspects of the fishery
also provides a basis for objective allocation of fishing effort
among bars.

VI-9

9 Summation
If a weak stock-recruitment relationship is assumed, management of
oyster stocks to ensure successful reproduction appears to have no
biological basis.
Modeling could be directed toward maximizing yield from any set
in a given year, maximizing the financial return from such a set,
and equitably allocating fishing effort; a simulation model would
appear to be the best modeling approach for these purposes.

4. White Perch (Mbrone americana)

9 Category - motile; entire life history in Maryland
e Ramifications of category
All exploitation of Maryland stocks is under Maryland regulatory
authority.
Quality of environment in Maryland can influence all life stages,
spawning success, growth, and mortality.
9 Unit stock characterization
Distinct stocks may occur in different Bay tributaries (Ritchie
et al., 1973).
Degree of intermixing of tributary stocks is unknown, but regional
definition of unit stock appears reasonable (e.g., Potomac River,
upper Bay),

9 Stock-recruitment relationship
No stock-recrtlitment relationships are documented in the literature.
Stunted populations (i.e., high-density populations exhibiting
lower than average growth rates) have been documented in MlarYland
(e.g., Potomac River [Appendix D, Ref. 58]; Susquehanna River,
Foerester' 1976); stunting suggests that biomass of juveniles and
adults may be at the carrying capacity of the environment but that
survival to the juvenile stage is somewhat density independent.
* Statistical model application
Commercial landing data by zone could be partitioned to define
unit stocks (see unit stock characterization above); availability
of effort data is not presently known.
White perch support an extensive recreational fishery (Speir et
al., 1977); only limited recreational landing data
are currently availa4le; however, a current federally sponsored
sportfishing survey (J. Williams, personal communication) may
provide this data.
Envirormental data (e.g., river flow, temperature, productivity)
are commonly used in statistical fisheries models (Appendix A,
Talbot, 1954; Welcomme, 1976). Lengthy records of such data may
be available for many segments of Marylarid's tidewaters (e.g.,

VI-10

Potomac River, Appendix D, Ref. 58). Thus., data bases may
be available for the construction of statistical models.
-- Statistical models, whether autoregressive or correlative,
provide a means of predicting yield for-a certain level
of effort where all other relationshiDs remain constant.
For white perch they could prove to be of value in deteimining
if production of certain areas is below expected levels, which
in turn could be used as an indicator of environmental quality
or exploitable excess production.
Surplus production model application

Availability of yield and effort data has been discussed above;
the absence of extensive data on recreational yield would be
a major problem in the application of surplus production models.
Data over a fairly wide range of effort are necessary for
surplus production modeling; range of effort must be examined.
Assumptions of equilibrium populations and stable age distribu-
tions might be met in regions where the stock is currently
not heavily exploited; however, if dominant yearclasses occur
often,surplus production models would not be applicable.
By applying'surplus production models to white perch populations,
we could estimate sustained yield obtainable at different
stock levels.

9 Yield-per-recruit model application
Assumptions of constant fishing and natural mortality rates over
all ages (see Section III.C.3) may not apply in cases where recreational
fishing effort is high,and mortality may be highest on older fish.
This sua-ests that a Ricker model would be most appropriate.
The stock-recruitment relationship.is not known; but, if a stock
is considered to be at equilibrium, a single estimate of number
of recruits could be obtained and used with a Ricker model.
Yield-per-recruit models would be useful for evaluating the
consequences to yield of various exploitation strategies (e.g.,
size limits).
Simulation model application
-- Simulation models could be developed which would integrate the
management of all stocks defined in Maryland (i.e., a pooled
stock approach). The availability of considerable data on vari-
ous life history aspects of white perch suggests that a simulation
model is feasible.

A simulation model would be most valuable where precise determination
of consequences of an action is necessary and where the effect of
invalid assumption on the precision of the estimates is of
concern.

VI-11

9 Summat 1 on
Relatively large amounts of data are available on white perch
and its habitat; thus, all model approaches appear feasible.
Choice of a modeling approach would depend primarily on the
objective of the management plan followed.

S. Yellow Perch (Perca flavescens)

o Category - motile; entire life history in Maryland
a Ramifications of category
Same as for white perch (4)
Stock characterization
Since this species is semi-anadromous tributary stock are
probably geneti:caily distinct; however, this has not been
documented in the literature.
A regional definition of stocks would appear reasonable, particu-
larly since lack of tolerance to hikher salinities prevents move-
ment between Bay regions.
s Stock-recruitment relationship
Dominant yearclasses are documented in the literature for lake
populations (Appendix D, Ref. 13).
No stock-recruitment relationships have been documented for
estuarine stocks.
o Statistical model application
- - Same as for white perch (4)
e Suiplus production model application
- - Same as for white perch (4)
* Yield-per-recruit model application
-- Same as for white perch (4)
* Simulation model application
Same as for white perch.(4)
9 Summation
Same as for white perch (4)

6. Carp (Cyprinus can2io)

* Category - motile; entire life history in Maryland
a Ramifications of category
Same as for white perch (4)

VI-12

Stock characterization
-- Stocks may be definable by tributary since lack of tolerance
to higher salinities may limit movement between tributaries
throughout the Bay system.
Stock-recruitment relationship
No stock-recruitment relationship is documented in the literature.
No information is available which is applicable to stock-recruit-
ment relationships.

* Statistical model application
-- Same as white perch, except that recreational and commercial
fisheries may be less extensive, and stocks may not be as
definable geographically.
# Surplus production model applicat-ion
-- Most landings of carp are incidental to efforts on other, more
valuable species, such as striped bass; as a result,
relationships between effort and yield may not be constant,
even assuming that stock size remains constant; therefore,
the application of surplus production models would be inappropriate.
The extent of the recreational fishery is unknown; absence of
recreational yield data might invalidate use of these models.
In tributary systemswhere an intensive fishery directed toward
carp exists, this modeling procedure would be useful.
9 Yield-per-recruit model application
-- No data could be located on growth, fishing mortality, or natural
mortality rates in estuarine populations (extensive data may exist
for cultured populations); thus, the compatibility of carp life
history characteristics with yield-per-recruit model structures
cannot be evaluated.
e Simulation model application
A simulation'model could be constructed to integrate models of
regional stocks; however, the lack of data on life history
characteristics makes it difficult to assess the feasibility
of constructing such a model.
9 Summation
-- The lack of life history data prevents a thorough evaluation of
model applicability; surplus production models may be most useful
under suitable circumstances,

7. White Catfish (Ictalurus catus)

9 Category - motile; entire life history in Maryland
9 Ramifications of category
Sam as for white perch (4)

VI-13

* Stock characteriZation
- - Same as for carp (6)
* Stock-recruitment relationship
No stock-recruitment relationshiT) documented in the literature.
No information available which is applicable to stock-
recruitment relationsh@ps.
* Statistical model application
-- Same as for carp (6)

* Surplus production model application
The extent of the recreational fishery is unknown; absence of
recreational yield data might invalidate use of these models,
C@
In tributary systems where an intensi,,,re fishery directed toward
catfish e'xists,this modeling procedure could be useful; one
exa-apl(@ of such'a situation is the Lipper Potomac estuary (.Appen-
dix D, Ref. 58).
e Yield-per-recruit model application
- - Same as for carp (6)
e Simulation model application
-- Same as for carp (6)
0 Suriimation
Same as for carp (6)

8. Alewife (.klosa pseudoharengus) and Blueback (Alosa aestivalis)

9 Category - motile; only part of life cycle spent in Maryland; spawns
in Maryland
e Ramifications of category
Exploitation rates experienced by fish in non-I'Vlaryland waters
cannot be regulated by Maryland.
The greater the fishing and natural mortality rates outside of
C>
Maryland as a proportion of total mortality,the less impact
Maryland regulations can have on stock size.
Environmental conditions on spawning a-rounds in Maryland can
influence reproductive success of the stock and thus affect
harvest in 'Maryland; for this reason, management of the fishery
must include regulation of-water quality.
t4on
a Stock characteriza L

As with most anadromous species, adults return to spawn where
they were spawned (Appendix D, Ref. 59); thus, unit stocks
can be defined in Maryland according to spawning area.

VI-14

-- Because both species tend to spawn in many small tributaries of
the Chesapeake Bay, regional definitions of stock may be somewhat
imprecise; ideally, management should be carried outby tribu-
tary; however, exploitation is not tributary based, which
rules out this possibility.

* Stock recruitment relationship

-- No stock-recruitment relationship is documented in the literature
in several Virginia rivers, no relationship between stock size and
production of juveniles was noted (Appendix D, Ref. 32).

-- For stocks spawning in restricted locations, such as small tribu-
taries, numbers of juveniles produced may be limited by the
carrying capacity of the spawning area.

-- The establishment of a relationship between size of the spawning
stock in Maryland and subsequent recruitment to the fishery in
Maryland is dependent on non-Maryland exploitation rates being
either constant or known.

* Statistical model application

-- Most of the commercial landings of alewives are from pound nets
located some distance from actual spawning locations (e.g., at
the mouth of the Potomac where spawning occurs in the Potomac
tributaries as far upstream as Washington D.C. in Maryland,
Appendix D, Ref. 58) thus, relationships between yield data and
tributary environmental characteristics might not be clearly
definable.

-- If yield data from specific spawning locations could be obtained,
correlations with environmental variables might be established.

-- Autoregressive models, rather than correlative, might be useful
for establishing stock recruitment relationships; such models
assume that exploitation rates and natural mortality are
constant.

-- No information is available on the magnitude of the recreational
fishery.

* Surplus production model application

-- There has been heavy exploitation of Maryland stocks of anadromous
species while they are at sea and/or in non-Maryland waters (Appen-
dix D, Ref. 82).

-- Because surplus production models use vield as a data input, they
would be applicable only if: (1) the contribution of Maryland
stocks to total harvest outside Maryland could be determined, or
(2) if effort outside Maryland was assumed to be constant over
all years.

-- Because of the potential difficulties in determining (1) and (2)
above, application of surplus production models appears inappro-
priate.

VI-15

e Yiold-per-recruit model application
The alewife fishery in 'Maryland ha-s a much higher age of recruit-
ment than the offshore fishery because harvest in Maryland presumably
takes sexually mature fish returning to spawn Cages III to V), while
open ocean fisheries take all ages (Appendix D, Ref. 82); thus, age
of recruitment (tr), a model parameter, is difficult to define.
Because of the existence or' offshore and non-Maryland fisheries,
an instantaneous fishing rate (F) is difficult t@ define for the
Maryland stock simply on the basis of Maryland fishing rates.
If non-Maryland fishing rates could be determined and if fishing
rates varied with acre, then a Ricker model would be more appro-
0
priate than a Beverton-Holt model.
If the non-Maryland fishery were negligible, or non-Maryland fishing
C> CP
effort were constant over many years, it might be possible to
C,
develop a yield-per-recruit model; since both of these situations
are unlikely, yield-per-recruit models would not appear appropriate
for this species,

a Simulation model application

A simulation model could be developed which would treat the
fishery as consisting of several phases, each
C@
with its own dynamics.
For example, submiodels of the oceanic fisheries, inshore non-Maryland
fisheries, and Maryland fisheries could be developed as major com-
partments -for a total stock model.
A re-oroJuction submodel for Maryland stocks could be developed
to simulate population dynamics and account for the effects of
variations in water quality or habitat modification on reproduc-
tive success.

* Summation
Because of the interstate and international nature of the fishery,
simulation modeling appears most appropriate for these species;
however, the availability of data necessary for development of
such a model is questionable.

9. American Shad (Alosa sapidissima)

� Category - motile; only part of life cycle is spent in '41aryland;
spaims in N-laryland
� Rmnifications of cate-go-r@y
Same as for alewife and blueback (8)
e Stock characterization

VI-16

-- Essentially the same for alewife and blueback (3); however
shad spawn in fewer, larger Bay-tributaries (e.g.., the Potomac,
Susquehanna, and Choptank rivers) than do the other two
species; because exploitation is often centered on these water
bodies, relating Nilaryland harvests to distinct spawning stocks
may be possible.

e Stock-recruitment relationship
On the Hudson Raver, 85% of the variation in stock size during
a given year was attributable to escapement (numbers of adults
reaching the spawning ground) 3, 4, and 5 years earlier (Appendix
A; Talbot,,1954); this suggests a deterministic stock-recruilment
relationship; if11.1aryland stocks have similar recruitment medil-ian-
isms, such a relationship might hold here.
In the Potomac, shad spawn in tidal waters (Appendix D, Ref, 58),
while,in other river systems (e.g., the Hudson and Delaware rivers),
spawning occurs in the non-tidal, upper reaches of the rivers
(Appendi-x A, Talbot,1954); such a difference in spauning habitat
may influence the degree of determinism of the stock-recruitment
relationship.
9- Statistical model application
Since landings in INTaryland tend to be spawning ground-specific,
autoregressive or correlative statistical models appear feasible,
as demonstrated by Talbot (19S4) (-Appendix A).

Little information is available on the recreational fishery;
if recreational landings were significant but not knourn, it
would be difficult to develop a statistical model of total yield.
� Surplus production model application
-- Same as for alewife and blueback
� Yield-per-recruit model application
-- Same as for ale%-ife @_:md blueback, except that the extent of the
offshore fishery is less clear, and age of recruitment to major
non-Niaryland fisheries (i.e., Virginia) may be the same as for
the @--Iaryland fishery.
� Simulation model application
-- Same as for alewife and blueback (8)

e Summation

.e
Same as for alew-i4: and bluebac"

10. Striped Bass (@Iorone saxatilis)
9 Category - motile; only part of life cycle in @,Jaryland, spai.,ns in
I
Maryland

VI-11

* Ramifications of category
* Same as for alewife and blUCback (8)
Essentially the same as for American shad, except that
striped bass spawn lower in tributaries than shad
(Appendix D, Ref. 58).
* Stock-recruitment relationship
Dominant yearclasses are observed (Appendix D, Ref'. .5-9j
suggesting that the stock-recruitment relationship is
weak, and that environmental variation accounts for
most of the variation in yearclass success (e.g.,
Ulanoivicz and Polgar, in press).
0
a Statistical model application
Statistical models have been applied to striped bass
stocks as predictors of yearclass strength (e.g., Appendix
C> C,
B, Stevens, 1977); further applications to Maryland
stocks could be developed.
@ Surplus production model
The occurrence of the dominant yearclass phenomenon indicates
that a striped bass stock is not an equilibrium population;
in the absence of a deterministic stock-recruitment
relationship, a surplus production model cannot be
applied.
s Yield-per-recruit model application
Because of the existence of non-Maryland fisheries, an
instantaneous fishing rate (f) is difficult to define
0
for the Maryland stock simply on the basis of Maryland
fishing rates.
If non-Maryland fishing rates could be determined, and if fishing
rates varied with age, then a Ricker model would be more appro-
priate than a Beverton-Holt model.
If the non-Maryland fishery were negligible or the non-Maryland
fishing effort constant over many years, it might be possible to
develop a yield-per-recruit model; since both of these situations
are unlikely, yield-per-recii-iit models would not appear appropri-
ate for this species.
a Simulation model application
Numerous simulation models of striped bass have been
developed for environmental assessment (e.g., Appendix A,
C,
Van Winkle et al., 1974); such models could be expanded and
applied to management problems.
a Summation
Because of the complex migrations of these species and the
interstate nature of the fishery, simulation modeling
would appear most appropriate for management purposes.

VI-18

Winter Flounder (Pseudopleuronectes americanus)

9 Category - motile; only part of life cycle in Maryland; spawns in
@,,Iaryl and.
9 Ramifications of category
-- Same as for alewife and blueback (8)
o Stock characterization
Extent of spawning is not clearly defined; spawning has been noted
in the lower Potomac estuary (Appendix D, Ref. 58);
if specific spawning areas in Maryland were delineated, and if
it could be demonstrated that distinct spawning stocks used these
areas, stocks could be defined by spawning location.
The proportion of Maryland-harvested winter flounder which
were spawned in Maryland is not known; if a significant
portion of the Maryland harvest was spawned in non-Maryland
waters, approaches to stock management would be very complex.

� Statistical model application
-- Because of the poor definition of spawning areas, statistical
models, other than autoregTessive, would appear difficult to
develop.
� Surplus production model application
Movements of flounder spauned in Maryland are unknown; thus,
degree and location of exploitation cannot be determined, and
total yield from stocks spawned in Maryland cannot be estimated.
--'It appears that surplus production models cannot be applied to
this species for management in Maryland.
9 Yield-per-recruit model application
Data are available on growth, natural mortality, and fi hi
mortality rates for this species i(Appendix D, Refs-3 2 1,s9 @8
A yield-per-recruit model could be developed as a basis for
determining optimal harvesting strategies; using empirical infor-
mation on yearclass sizes, such an approach would de-emphasize
the importance of the stock-recruitment relationship.
e Simulation model application
A lack of information on movements of Maryland and non-?,tarylpmd
stocks makes it difficult to construct a simulation model;
partitioning of in-state and out-of-state fishing mortalities
would not be possible.

M
I
igration data would have to be obtained for all spawning stocks
contributing to the Maryland harvest before a simulation model
would be feasible.

e SLmnation
-- A cur-rent lack of stock definition would make application of any
model extremely difficult at the present time.

VI-19

12. Bluefish (Pomatomus saltatrix)

� Category - motile; only part of life cycle (non-spawning) in Maryland;
possible distinct stock in Maryland
� Ramifications of category
If it could be established that the same subunit of a given stock
returned to the same nursery or feeding area in Maryland each
year, a relationship between management actions and responses
of the stock might exist,

The proportion of mortality due to fishing and natural-mortality
in Maryland would determine the importance of Maryland regulatory
activities to stock maintenance.

� Stock characterization
Bluefish spawn along the continental shelf; there is some indi-
cation that the Chesapeake Bay stock may be a distinct subunit
of the east coast bluefish population (Appendix D, Ref. 28).
How the stock partitions itself between Maryland and Virginia
waters is'unknown.
� Stock-recruitment relationship
Yearclass strength may be determined by coastal circulation
pattern (Appendix D, Ref. 4).
Thus, environmental factors may be more important in determining
reproductive success than stock (i.e., the stock-recruitment
relationship is not deterministic).

9 Statistical model application
If sufficient data were available, it would be possible to use
statistical models to investigate environmental variables which
might influence the extent of migration of a stock into Maryland
waters.
Such a relationship could be used to determine possible conse-
quences to the stock of variations in exploitation rates in
Maryland.
9 Surplus production model application
-- Same as for striped bass; also, recreational harvest is large
and undocumented; total yield data do not exist.
9 Yield-per-recruit model application
Because of the existence of non-IMaryland fisheries, an instan-
taneous fishing rate (F) is difficult to define for the Maryland
stock simply on the basis of Maryland fishing rates.
if non-Maryland fishing rates could be determined and if fishing
rates varied with age, then a Ricker model would be more appro-
priate than a Beverton-Holt model.

Vl-20

If the non-NIaryland fishing effort were assumed constant, it
might be possible to develop a yield-per-recruit model which
combines the non-Maryland fishing mortality with natural
mortality; however, no mortality data are available in the
literature.
e Simulation model application
If a distinct Chesapeake Bay stock were assumed, a simulation
model could be developed consisting of Virginia and Maryland
0
compartments and a reproductive submiodel incorporating the
influence of environmental factors, such as coastal currents.
Such a model could be used to examine the consequences of
management actions in both states as well as the interdepen-
C@
dency of these actions; however, data required for development
of such a model appear unavailable.
* Summation
-- Same as for striped bass (10).

-13. Atlantic 14enhaden (Brevoortia t-vrannus)

9 Category - motile; only part of the life cycle (non-spauning) in
I'daryland
9 Ramifications
The proportion.of mortality due to fishin g and natural mortality in
Iv1aryland would determine the importance of I'vlaryland regulatory
activities.

Stock characterization

The Chesapeak e Bay population of menhaden appears to be a
random segment of the east coast stock (Appendix D, Ref. 22).
Currently, Chesapeake Bay landings (i.e., Nlaryland and Virginia) account
for approximately 60% of total landinas from the east coast stock
(Appendix D, Ref. 83); since most of these fish are not sexuallv mature,
it appears that Chesapeake Bay is the major nursery area for th@
total east coast stock; thus, Ni-laryland management decisions could
influence the status of this stoA. g
e Stock-recruitment relationship
Yearclass strength is determined by both stock size and coastal
circulation patterns (Appendix D, Ref. 12).
Because environmental factors (coastal currents) dominate recruitment
to the fishery, the stock- recruitment relationship is Pon-
deterministic.

Statistical model application
same as for bluefish (12)

VI-21

e Surplus production model application
'f there is no deterministic stock-recruitment relationship, the
L
stock cannont attain an equilibrium level; thus, a surplus pro-
duction model cannot be applied.
9 Yield-per-recruit model application
Because of the existence of non-Maryland fisheries, an instan-
taneous fishing rate (F) is difficult to define for the Maryland
stock simply on the basis of Maryland fishing rates.
A menhaden Ricker yield-per-recruit model has been developed at
the NMFS, Beaufort Laboratory (Appendix D, Ref. 83); some modi-
fication of this model might be applicable for management in
Maryland.
* Simulation model application
The NMFS'Beaufort Laboratory, is developing a menhaden stock
simulation model, compartmentalized by geographic region; the
Chesapeake Bay is one of the compartments in the model; a
cooperative effort with MvffS might be possible to further partition
the Chesapeake Bay submodel into Maryland and Virginia portions.
Summation
Same as for striped bass (1)

14. Blue Crab (Callinectes sapidus)
e Category - motile; only part of life cycle (non-spawning) in i'llaryland
* Ramifications of category
The proportion of mortality due to fishing and natural mortality
in Maryland would determine the importance of Maryland regulatory
activities in stock maintenance.

e Stock characterization
The Chesapeake Bay population of blue crab may be a relatively
distinct stock, since spawning occurs in the lower Bay or near
the mouth of the Bay (Appendix D, Ref. 66).
@-Iigrational patterns differ between sexes; males entering Maryland
waters-may remain over their entire life; females migrate between
Maryland and Virginia; thus, the Maryland portion of the stoc.- is
not a clearly defined segment of the Bay stock.
@ Stock-recruitment relationship

No stock-recruitment relationship is documented in the literature.
Large fluctuations in annual landings over the last
100'>Years (Appendix D, Ref. 66) suggest the irregular occur-
rence of dominant yearclasses; the stock-recruitment relationship
may be weak.

VI-22

� Statistical model application
-- Same as for bluefish (12)
� Surplus production model application
Same as for striped bass (10); also recreational harvest is large
and undocumented; total yield data do not exist.
� Yield-per-recruit model application
Since crab growth is through molts, models employing constant
growth functions (e.g. , von Bertalanffy) are inappropriate -
growth rate based on weight increase could possibIy.approiich
a constant.
Since crab growth is influenced by salinity, growth rates
vary by region in the Bay; a single model is inapplicable for
all Maryland waters.
Standard yield-per-recruit models appear inapplicable to blue
crabs, but versions using modified growth functions could be
developed..
� Simulation model application
Same as for bluefish, except that environmental factors influ-
encing reproduction success would include non-tidal up-Bay
transport, salinity, etc., rather than coastal currents.
Also, since growth is salinity dependent, Bay waters would have
to be partitioned according to salinity and different growt,
models developed for each salinity zone.
Sizmation
-- Same as for striped bass (10)

15. American Eel (Anguilla rostrata)

� Category - motile; only part of life cycle (non-spawning) in
Maryland
� Ramifications of category

-- The proportion of mortality clue to fishing and natural
mortality in Maryland would determine the importance'of
Maryland regulatory activities to stock maintenance.

� Stock characterization
Since all eels in North America originate from a single spawning
location in the Sargasso Sea in the Atlantic Ocean (Appendix D,
Ref. 59), all must be considered to be a single stock.
Although not documented, it is highly unlikely that the elvers
(juvenile eels) entering the Bay eac@ spring are progeny solely
of the adult stock which emigrated from the Bay the previous
fall; thus, the "stock" of Chesapeake Bay is not reproductively
distinct.

VI-23

9 Stock-recruitment relationship
A stock-recruitment relationship for the eel could be established
only on a continental basis.
The size of the stock in Maryland at any given time would have
little influence on size of stocks in subsequent years.
e Statistical model application
If numbers of elvers entering different tributaries or portions
of the Bay could be monitored, statistical models could be
developed to investigate environmental factors influencing
recruitment to those locations.
Biomass production rates of eels may differ by location or region;
statistical models could be developed to correlate these differences
with various environmental variables.
e Surplus production model application
Because of the absence of a stock-recruitment relationship, no
equilibrium population could be defined for the eel; thus, a
surplus production model would not be appropriate.
@ Yield-per-recruit model application
If recruitment is defined as the numbers of elvers reaching a
nursery area, a yield-per-recruit model is feasible; however,
little data are available on growth, natural mortality, and
fishing mortality rates, and the eel presents a very difficult
sampling problem.
0 Simulation model application
Fisheries for elvers-have been established in various parts of
the United States; a simulation model would relate fishing
mortality of elvers to loss of potential eel production at the
adult level; a yield-per-recruit model would be likely to be a
major component of such a model.
@ Summation
Because of the unique nature of the life history of this species
and the complex relationship between fisheries for both juveniles
and adults, a simulation model approach would appear most appro-
priate for this species.

16. Spot (Leiostomus xanthprus), Atlantic Croaker (Micropogonias undulatus),
and Weakfish (Cynoscion regalis)

o Category - motile; only part of life history (non-spawning) in
Maryland
o Ramifications of category
The proporrion of mortality due to fishing and natural mortality
in Maryland would determine the importance of Maryland regulatory
activities.

VI-24

0 Stock characterization
These species spawn either offshore or near the mouth of the Bay
(Appendix D, Ref. 59); the exist--ence of distinct Chesapeake Bay
stocks is uncertain; fish present in the Bay and in Maryland
waters may represent random segments of a southeast coast stock
(Appendix D, Ref. 86).
Weakfish, which may spawn in the lower Bay (Appendix D, Ref. 8S),
probably could be defined as having a Bay stock; but the portion
of that stock which enters Maryland waters is probably not dis-
tinct.
* Stock-recruitment relationship
No stock-recruitment relationships have been documented in the
literature for these species.
Several strong yearclasses of croaker and weakfish have appeared
in recent years after long periods of very low stocks (Appendix D,
Refs. S8,85), suggesting a nondeterministic stock-recruitment
relationship.
Abundance of spot juveniles in Maryland waters has remained
consistently high over the last 5 to 7 years (Hixon et al.,
1979), suggesting that the stock may currently be at some
equilibrium level; lower abundances in earlier years indicate that
this might be a transient -phenomenon.
9 Statistical model application
Because'of the nature of spawning location, certain environmental
0
variables might influence the numbers of larvae entering Maryland
waters; statistical models could be developed as predictors of
juvenile abundance.
If sufficient data were available, statistical models could be
used to investigate environmental variables that might
Z@-
influence the extent of migration of a stock.into Maryland
waters.

Such a relationship could determine possible conse-
quences to the stock of variations in exploitation rates in
Maryland.
9 Surplus production model application
For croaker and weakfish, same as striped bass (10); also,
recreational harvest may be large and undocumented; thus,
total yield data do not exist.

For spot, a surplus production model may be applicable if an
equilibrium population currently exists; however, because of
the recreational and interstate nature of the fishery, determination
of yield and effort would be difficult.

VI-2S

* Yield-per-recruit model applica .tion
Same as for bluefish (12)
* Simulation model application
If a distinct Chesapeake Bay stock were assumed, a simulation
model could be developed consisting of Virginia and Maryland
compartments and a reproductive submodel incorporating the
influence of environmental factors,, such as coastal currents.
Such a model could be used to examine the consequences of
management actions in both states as well as the interdepen-
dency of these actions; however, data required for development
of such a model appear unavailable.
Croaker and spot, with offshore spawning, present the same types
of problems for management as do menhaden; given sufficient data,
a simulation model compartmentalized by geographic region could
be developed.
9 Summation
-- Same as for striped bass (10.)

17. Sumer Flounder (Paralichthys dentatus)

o Category - motile; only part of life cycle (non-spawning) in Maryland
9 Ramifications of category
The proportion of mortality due to fishing and natural mortality
in Maryland would determine the importance of Maryland
regulatory activities.
a Stock characterization
This species spawns offshore; the Maryland population in any
year may be a random segment of two distinct east coast spawn-
ing stocks (P. Scarlett, N.J. Division of Fish, Shellfisheries
and Wildlife, personal communication); management of this species.
is currently being reviewed by the State-Federal Marine Fisheries
program because of the wide distribution of these stocks.
e Stock-recruitment relationship
No stock-recruitment relationships documented in the literature,

a Statistical model application
Because no Maryland stock can be defined, application of
statistical models appears inappropriate.

e Surplus production model application
Without knowledge of the population dynamics of the species,
no judgement can be made about the applicability of surplus
production models.

VI-26

� Yield-p6r-recruit model application
Because the individual fish present in Maryland waters at any
given time represent a random segnent of stock distributed
over a much wider range, application of this model type to
fish in Maryland would appear inappropriate.
� Simulation model application
A simulation model could reasonably be applied only to
stocks contributing to Maryland landings, with Maryland waters
representing a geographical compartment of the model; however,
without data on the extensive undocuTiented recreational fishery
for this species, model development would not be possible.
� Suinmation
Same as for striped bass (10)

VI-27

VII. CONHUNICATIONS WITH M-AM- ENT AGENCIES

A.- Objective

The purpose of this task was to determine if mathematical models were
being used in the fisheries management programs of various state, federal, and
international agencies and, if so, the nature of the models in use and the
degree of success of their application.

B. Interview Procedure

Initial contacts were made by telephone with federal laboratories known
to be actively involved in fisheries model development (e.g., Woods Hole
NMFS Northeast Fisheries Laboratory). Individuals contacted were then
asked for names of persons in other organizations who were also involved
in the development and/or application of fisheries models. @bst contacts
were made in this way.
In telephone conversations, information requested included: name and
position of the individual contacted, species being managed, types of
models being employed, and evaluation of success of the application.
Written requests have been made to the United Nations Food and Agri-
culture Organization and to the Fisheries Research Board of Canada
for any available reports on application of fisheries management models.
The requested information has not yet been received.

C. Results

Table VII-1 sumarizes the results of phone and letter communications
with staff members of various fisheries management agencies. Because most
of the information was received verbally, any inaccuracies in representation
occurring are the responsibility of the authors. In many cases, precise
answers to questions were not possible, and some degree of interpretation
of comments was employed by the authors.
Because of the time-consuming nature of these activities and diffi-
culties in contacting many individuals, all potential contacts have not
yet been made. This part of the project will be continued into Phase II.

VII-1

Table VII-1. Individuals and agencies providing information on current use of fisheries models in
management programs (only those contacts from which information was received are listed).

individual Organization Type of Current Comments
Information Use of
Provided Models

Dr. Michael Northeast Marine Verbal plus Stock assessments of Model application here
Sissenidne, Fisheries Center, publications many major species in appears to be as advanced
Deputy Chief, National Marine the northwest Atlantic as at any other agency
Resource Fisheries Service, made using contacted; large data
Assessment Woods Hole, Mass. cohort analysis; bases (both survey and
Division Beverton-Holt yi Peld- landings) are available;
per-recruit models however, management
frequently used to decisions are made by
choose ranges of F's Regional councils; NMFIS
for stock assess- provides technical support
<i ments; model use to the councils.
summarized in Sissen-
wine et al., 1979
(Appendix A).

Walter Nelson., Southeast Marine Verbal Most modeling work Because menhaden is an
William Fisheries Center, deals with menhaden; inshore species, it is
Schaaf National Marine models developed include under regulatory authority
Fisheries Ser- one dealing with recruit- of the States; the Atlantic
vice, Beaufort, ment (Nelson et al., States Marine Fisheries
North Carolina 1977, Appendix A) and Commission is developing
a Ricker yield.-per- a management plan for this
recruit model; develop- species; the NMFS people
ment of a simulation serve as technical advisors
model based on regionial to the Commission.
compartments is under-
way.

mmmm M M M M M MM

TableVII-1. Continued.

Individual Organization Type of Current Comments
Information Use of
Provided Models

Dr. Felix Northwest Marine Verbal An ecosystem model of the Application of this model
Favorite Fisheries Center, (publications eastern Bering Seafish- was not determined.
National Marine not yet received) eries has been developed
Fisheries Service, (documentation of the
Seattle, Wash. model has not yet been
received).

Dr. Guy New England Verbal and Level of modeling varies Much of the work being
Marchesseault, Regional Fisher- publications by species; surplus pro- done is in conjunction
Senior ies Management duction and yield-per- with the Northeast Marine
Biologist Council recruit models have been Fisheries Center; all stock
used in making management assessment data comes from
decisions on the scallops NMFC; the council staff
fishery; extensive bio- mainly evaluates
economic modeling,being biological and economic
done. consequences of various
management alternatives.
Dr. John Mason, Mid-Atlantic Verbal No models are currently Council staff are provided
Ann Williams Regional Fisher- (publications being used by council with stock assessment data
ies Management not yet received) staff. by NMFC; [email protected] also provides
Program technical assistance for
development.of management
plans.

Irwin Alperin Atlantic States Verbal and No modeling being done Chairmen of Scientific and
Marine Fisheries written directly by committees Statistical Committees for
Commission set up by the Commission. menhaden, striped bass, and
summer flounder, were asked
for information on modeling;
being done concerning men-
haden; this work is being
done by 11. Nelson of NMFS
(see above).

Table VII-1. Continued.

Individual Organization Type of Current Comments
Information Use of
Provided Models

Michael Street North Carolina Verbal A predictive statistical The shrimp work is the only
Marine Fisheries, (publications model is currently in modeling work currently
Morehead,, North not yet received) use, with salinity.and going on; work is being done
Carolina temperature as indepen- in cooperation with R. Derisa
dent variables. at the University of North
Carolina.

Lawrence Six Pacific Fishery Verbal Surplus production and Modeling work is being done
Management Council, cohort analyses are being by technical advisors (@'Z,9:S)
Portland, Oregon used in stock assessment to the Council.
of some non-pelagic fish
populations.

@-4

D. General Overview

From the contacts made thus far, it is evident that the level of
modeling effort and the sophistication of models employed vary widely
among agencies and locations. Lack of information on stock characteri-
zation and life history was most often cited as the reason for lack of
greater modeling efforts.

VII-5

VIII. REFERENCES*

Brody S., 1945. Biooriergetics and Grow-th. Peinhold Publishing Corp.,
Neur York. 10 23po.
Brody, S. 1927. Growth rates. University Missouri Ag-ric. Exp., Sta.
Bull. 97.
Foerster, J.W. 1976. Assessment of the effect of hydroelectric water
discharge from the Conowingo Dam on the spawning fish in the
C@
lower Susquehanna River. Final Report. Prepared for Maryland
Department of Natural Resources, Power Plant Siting Program. 320pp.
Hixson, J.H., and T. Capizzi. 1979. Fish bottom trawling. In Non-radio
logical Environmental Monitoring Report , Calvert Cli7t-ts Nuclear
Power Plant, January - December 1978. Baltimore Gas and Electric
Company, Baltimore, Maryland.
Kaplan, W 1962. Operational methods for linear systems. Addison-
@;sley Publishing Company, Reading Mass., S77pp.
C,
Montgomery, D.C., and L.A. Johnson. 1976. Forecasting and time series
analysis. McGraw-Hill, London. 304 pp.
Ritchie, D.E., H.J. King, and A.J. Lippson. 1973. White perch. pp.34-35.
In A.J. Lippson Ced.). Johns Hopkins University Press, Baltimore,
IU
I ryland. The Chesapeake Bay in Maryland - An Atlas of
Natural Resources.

Ryder, R.A., S.R. Kerr, K.H. Loftus, and H.A. Regier. 191-4. The morpho-
Cp
edaphic index, a fish yield estimator -- review and evaluation.
J. Fish. Res. Bd. Canada 31:663-588.
Schaaf,, W.E., and G.R. Huntsman. 1972. Effects of fishing on the Atlantic
menhaden stock: 19SS-1969. Trans. Am. Fish. Soc. 101:290-297.
Spier, H., D.R. Weinrich, and R.S. Early. 1977. 1976 Maryland Chesapeake
Bay sport fishery survey. Final Report to U.S. Department of
Interior., Fish and Wildlife Service., Project NO. * F-3)2-R-
Ulanowicz, R.E., and T.T. Polgar. In press. The influences of anadromous
spawning behavior and optimal environmental conditions upon year
0
class strength. Submitted to J. Fish. Res. Bd. Canada.

This reference list consists of literature not cited in attached appendices.

VIII-1

A REVIE-11 AND EVALUATION OF

FISHERIES STOCK @,,MAGTMENT MODELS

Part II - Appendices

W.A. Richkus
J.K. Summers
T.T. Polgar
A.F. Holland
Martin Marietta Corporation
Environmental Center
1450 South Rolling Road
Baltimore, Maryland 21227

Prepared for
Coastal Resources Division
Tidewater Administration
Maryland Department of Natural Resources
Tawes State Office Building
Annapolis, Maryland 21401

December 1979

TABLE OF CONTENTS

An Annotated Bibliography of Stock Management Iviodels . . . . . . . .A-1
An Annotated Bibliography 6T-'Stock Management Submodels . . . . . . B-1
Age Structure Submodels . . . . . . . . . . . . . . . . . . . . B-2
Allometric Submodels . . . . . . . . . . . . . . . . . . . . . . B-4
Catchability Submodels . . . . . . . . . . . . . . . . . . . . . B-5
Effort Submodels . . . . . . . . . . . . . . . . . . . . . . . . B-6
Fecundity Submodels . . . . . . . . . . . . . . . . . . . . . B-9
Growth Submodels . . . . . . . . . . . . . . . . . . . . . . . . B-11
Mortality Submodels . . . . . . . . . . . . . . . . . . . . . .B-21
Parameter Estimation . . . . . . . . . . . . . . . . . . . . . . B-30
Recruitment Submodels . . . . . . . . . . . . . . . . . . . . . B-32
Yearclass Strength Submodels . . . . . . . . . . . . . . . . . B-33
An Annotated Bibliography of Methods of Data Acquisition . . . . . . C-1
Age Structure . . . . . . . . . . . .. . . . . .. . . . . . . . . C-2
Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C-2
Migration . . . . . . . . . . . . . . . . . . . . . . . . . . . C-6
Stock Assessment. .;. i . . . . . . . . . . . * . . . . . . . . C-11
A Bibliography for Life History Data on Selected Maryland
Species ... ... . . . . . . . . . . . . . . . . . . . . . . . . D-1
A Keyword and Species Index to Appendices A-D . . . . . . . . . . . . E-1

APPENDIX A

An annotated bibliography of stock management models

The categorization indicated in the
upper right corner of each page was
based on a preliminary reading of
each article; those articles deemed
"applicable" were considered to be
potentially useful in real world
situations without further theoreti-
cal development.

A-1

Abrainson, N.J. (ed.). 1971. Computer programs for fish Applicable
stock assessment. F.A.0. Fish. Tech. Paper 101.
F.A.O., Rome, Italy.

Abstract

This book represents a collection of computer appli-
cations for the calculation of a fishery's yield and the
estimation of several parameters such as mortality and
effort.

Comment

This collection will be very helpful in this project,
particularly in an actual application phase.

A-2

Adams, C.F., and C.H. Olver. 1977. Yield properties Applicable
and structure of boreal percid communities in
Ontario. J. Fish. Res. Bd. Canada 34:1613-1625.

Abstract

A synoptic review of yield data for 70 northern
Ontario lakes from 1917 to 1973 showed that percids,
mainly walleye (Stizostedion vitreum vitreum constituted
about one-third, by weight, of the total fish yield.
This relationship, which was independent of fishing
effort, lake size, and lake productivity, is considered
to be an emergent property of this type of fish
community and represents a degree of homeostasis within
the community under exploitation. The relation of
percid yield to theoretical yield (based on the
morphoedaphic index--NEI) reflected organizational
structure and suggested the existence of a community
(percid) component within the NEI, and from this we
recommend upper limits of percid harvest for boreal
percid lakes.

Most (83%) of the 70 lakes had an average total
yield of less than 2.5 kg'ha"1*yr-1, 53% (37 lakes)
yielding less than one-half of the theoretical yield
(average 3.4 kg*ha-1-yr-1). Long-term average yields
exceeded the theoretical maximums in only 11 lakes.
Mesotrophic to slightly eutrophic waters appeared as
optimum for percid yields.

Inferences from the data suggest an unexploited
boreal percid community is characterized by high
community stability and low net community production
with resiliency low because of the low productive
capacity of the waters. A yield index (RYI), which
was assumed to reflect both effort and vulnerability
to exploitation, showed that fishing intensity tended
to be higher on the smaller, less productive lakes
in this study.

Comment

The approach would appear to have limited
applicability to tidal waters, where fish communities
and environmental conditions are very variable. Itmay
be utilizable in the case of recreational fisheries in
relatively enclosed tidal waters.

A-3

Agnew, T.T. 1979. Optimal exploitation of a fishery Applicable
employing a non-linear harvesting function.
Ecol. Modelling 6:47-57.

Abstract

A harvesting function is developed to describe
the rate of removal of fish from a fish population.
The function incorporates the effects of both
the handling or processing time of the catch and the
competition, between boats in the fleet, for the fish.

_'d that the growth rate of the fish
It is assume
population can be modelied with a concave, dome
shaped growth curve. With this assumption, it has
been.shoim that if the rate of harvesting the fish
is linearly related to both effort (which can be
thought of as same measure of the number of boats
in the fleet) and the population size, then the
population will tend towards a single equilibrium
level which is globally stable. This paper shows
that the saturation effects due to the handling
time may generate two equilibrium levels (one stable,
one unstable) rather than a single globally stable
equilibrium. The results of competition between
boats are economically undesirable because of the
decrease in efficiency. However, this competition
may be beneficial to the exploited fish population.

Using the harvesting model derived earlier, the
steady state or long term optimal harvesting policies
as well as the transition paths to these states are
developed. The only constraint is on the maximum
allowable effort which is effectively an upper
limitation on the fleet size or number of man-hours
of fishing.

Comment
This method can possibly be utilized in a fish-
eries management context. It is a theoretical model,
incorporating some novel views of factors influencing
fishing mortality. Its main wealaiess is the
use of only the logistic model for growth response
of the exploited population.

A-4

Ahmed, N.U., and N.D. Georganas. 1973. Optimal Non-applicable
control theory applied to a dynaTrdc aquatic
ecosystem. J. Fish. Res. Bd. Canada
30:576-579.

Abstract

In this report a dynamic model is presented
for an aquatic ecosystem consisting of a single
Fecies living in a polluted environment. Consider-
ing the dynamics of g h of the species and
"rowt
increase in pollution, optimal control theory is
applied to obtain species removal and pollution
reduction rates at minimum cost.

Comment

This paper explores the use of optimal
control theory on a fisheries problem. The
specificity of the problem discussed limits
the general applicability of the material
presented.

A-5

Allen, K.R. 1973. Analysis of the stock recruitment Applicable for
relation in Anarctic fin whales (Balaenoptera parameter
p@ysalus). Con. Per., Inter. LlExploratioiT de la estimation
Mer, Rapports No. 164:132-137.

Abstract

The author discusses a technique for the estimation
of recruitment rates from age compositions of catches
taken from an exploited population and examines the
nature of the stock recruitment relationship in southern
hemisphere stocks of the fin whale. Recruitment is not
shown to increase with a decrease in stock size. However,
inaccuracies in estimates of age distributions may have
caused this unexpected finding.

Comment

The paper presents a means of deriving estimates of
recruitment. It might be useful in developing recruitment
functions for other models.

A-6

Allen, K.R. 1974. The influence of random fluctuations Applicable
in the stock-recruitment relationship on the economic
return fram salmon fisheries. Con. Per., Inter.
L'Exploration de la Mer, Rapports INTo. 164:350-359.

Abstract

The author examines the effects of random variation in
the stock recruitment relationship of a salmon stock on the
optiman strategy for managing a fishery by regulation of
exploitation. Consequences of various management strategies
to future stock sizes and harvests are investigated by runs
of a model based on a 50-year data set. None of the exploi-
tation strategies resulted in extinction of the run. The
runs suggested management strategies which could be adjusted
to achieve various economic objectives.

Comment

The procedures followed in this paper may be useful as
a means of developing optimal strategies. However, the
model and management strategies are specifically oriented
toward salmon, and may not be directly applicable to any
Maryland species.

A-7

Allen, R.L. 1975. Models for fish populations.
A review. N.Z. Oper. Res. Q. 4:1-20.

Abstract

Comment

This paper is unavailable for review but has
been requested through interlibrary loans. It will
be reviewed when received.

A-8

Allen, R.L., and P. Basasibwaki. 1974. Properties Non-applicable
of age structure models for fish populations.
J. Fish. Res. Bd. Canada 31:1119-1125.

Abstract

The behavior of a class of dynamic population
models that can be described as a life table operating
on a population with a stock recruit relation formed
by the product of the egg production and a survival
function was examined. A combination of analytical
and simulation methods were [sic] used to find necessary
conditions for the stability of equilibrium populations
and the properties of fluctuations about the equilibrium.
Regular oscillations that occurred in populations
with an unstable equilibriun were of most interest
and these were considered as possible causes of
the regular fluctuations in populations of fish
such as sockeye salmon, Oncorhynchus nerka, or blue
pike, Stizostedion vitreEF-giaucum.

Comment

This theoretical model requires too many
nonexistent types of data to be applicable to
real situations.

A-9

Andersen, K.P., and E. Ursin. 1977.. A multispecies Applicable
extension to the Beverton and Holt theory of
fishing, with accounts of phosphorus circulation
and primary production. Meddr. Dam. Fisk.
og Havunders 7:319-435.

Abstract

Interaction between species in a marine ecosystem
is described by expressions for food consumption
and grazing mortality which are consistent with each
other and with the Beverton and Holt model of the
population dynamics of individual species. A model
of primary production is.introduced in order to make
possible an account of nutrient circulation (as,
exemplified by phosphorus) within and nutrient flow
through the system. It is demonstrated in an
application to North Sea fisheries that recent
changes in total yield can be described in some
detail under the terms of the model as a function
of fishing mortality alone. The composition of the
North Sea fauna in the virgin state is discussed
and also the conditions under which total yield could
be increased above the 1970 level.

Comment

This paper presents the most advanced multi-
species production model in the literature. It
may be of great value for this program.

A-10

Anderson, L.G. 1973. Optimum economic yield of a Non-applicable
fishery given a variable price of output.
J. Fish. Res. Bd. Canada 30:509-518.

Abstract

The majority of work done in the economics of
fisheries uses the assumption of a fixed price of
output. This paper describes the effects on the
traditional fisheries model of relaxing this
assumption, the most important of which, as far as
regulatory agencies are concerned, are the negation
of marginal cost of effort equaling marginal revenue
of effort as the criterion for a social optimum,
and the introduction of the possibility of multiple
equilibria and multiple industry profit maxima.
Also, sane new insights on fishery management with
variable price and different assumptions about the
number of fisheries and the number of countries
involved are pointed out.

Comment

The paper is a theoretical aiscussion of
the bioeconomics of fisheries. The author
acknowledges that 11 ... empirical investigation
based on the general model will be very difficult....
It will not be useful for our study.

A-11

Anderson, L.G. 1975. Optimun economic yield of Non-applicable
an internationally utilized common property
resource. Fish. Bull. 73:Sl-65.

Abstract

The exploitation of a common property resource,
specifically a fishery, by nationals of two countries
is discussed using a simple general equilibrium analysis.
The interdependence of their production possibility
curves is used to describe the open-access equilibrium
yield, local maximum economic yields, and a true
international maximum economic yield. Finally a
complete description of the conditions necessary for
this international maximum economic yield and why
they are different from those in a national fishery
is presented.

Comment

Principles explored here appear to have little
relevance to Maryland, even if management is viewed
as an interstate problem.

A-12

Anderson, L.G. 1975. Analysis of open-access Applicable
commercial exploitation and maximum. economic
yield in biologically and technologically
interdependent fisheries. J. Fish. Res.
Bd. Canada 32:1825-1842.

Abstract

Fisheries may be interdependent because of
biological relationships that exist between their stocks
or because the gear of one affects mortality in the
stock of the other. The problems of defining a
maximum sustainable yield in these cases are discussed.
A graphical analysis is used to describe the
combinations of effort from both fisheries where
concur-rent exploitation is possible and which of
these combinations will result in a simultaneous
equilibrium. Finally the conditions for a combined
maximum economic yield (NEY) are presented and it is
shown that they will not hold if each fishery is
managed to obtain an individua.L NL,..

Coment

The paper investigates the consequences of species
interactions on a fishery but does not really present
predictive equations. The framework might be used to
integrate other models in some way.

A-13

Austin, H.M., and M.C. Ingham. 1979. Use of Non-applicable
environmental data in the prediction of marine
fisheries abundance. pp. 93-108 In Climate and
Fisheries. U. of R. I. Center for 066-ari-
Management Studies, Kingston, R.I. 135 pp.

Abstract

The authors critically evaluate methods which
have been used in relating fisheries yields to
environmental variables, noting the weaknesses of
methods used and the possible erroneous conclusions
which may result. They then discuss factors which
should be taken into consideration in designing
studies for the purpose of investigating climatic
effects on fisheries populations.

Comment

The paper presents an informative discussion
of the problems encountered in relating environmental
variables to fish abundance. However, no models are
presented.

A-14

Balsinget-, J.1V. 1974. A computcr simulation model Applicable
for the Eastern Bering Sea king crab population.
Ph.D. Thesis. U. Wash., Seattle, Washington.
198 pp.

Abstract

A Ricker yield per recruit model and a simulation
model were constructed for an Alaskan king crab popula-
tion. From the results recommendations were made con-
cerning the fishery policies for the exploitation of
that species. Several submodels (i.e., growth, mortality)
were also developed.

Comment

This paper presents an excellent combination of
classical yield and simulation models and submodels
for the Alaskan king crab fishery. It could prove
helpful in this project.

A-15

Beddington, J.R., and R.M. May. 1977. Harvesting Non-applicable
natural populations in a randomly fluctuating
enviroment. Science 197:463-465.

Abstract

As harvesting effort and yield are increased,
animal populations that are being harvested for
sustained yield will take longer to recover from
enviromentally imposed disturbances. One consequence
is that the coefficient of variation (the relative
variance) of the yield increases as the point of
maximm sustained yield G,1SY) is approached. When
overexploitation has resulted in a population smaller
than that for MSY, high effort produces a low average
yield with high variance. These observations accord
with observed trends in several fish and whaling
industries. We expect these effects to be more
pronounced for a harvesting strategy based on constant
quotas than for one based on constant effort. Although
developed in a MSY cmtext, the conclusions also
apply if the aim is to maximize the present value
of (discounted) net economic revenue.

Coment

This is a theoretical paper looking at the
possible consequences of differing management
strategies, given different population dynamics. Nothing
presented is directly applicable to a real-world
problem.

A-16

B61and, P. 1974. On predicting the yield from brook Applicable
trout populations. Trans. Am. Fish. Soc.
103:353-355.

Abstract

The predicted yield from an unstable population
is compared to that from a stable population. It is
shown that an unstable population will reach a stable
structure, following which the yield will not fluctuate
but go steadily down. Data from the population indicate
that it is on its way towards extinction, even without
additional mortality imposed on any age group.

Comment

This paper represents the application of a yield
model to a set of real data. It will be worthwhile
to evaluate further.

A-17

Beverton, R.J.H. 1953. Some observations on the Applicable
principles of fishery regulation. J. Cons.,
Cons. Int. Enlor. Mer. 19:56-68.

Abstract

An alternative method of calculating equilibrium
yield per recruit is advanced which is age dependent
and related to Brody-von Bertalanffy growth rates..
The relationships of yield and recruitment age,in
conjunction with fishing mortality are discussed graph-
ically. The concepts of eumetric fishing and yield
curves are discussed.

Comment

This paper is one of the classical equilibrium
yield-per-recruit presentations. It will be helpful
in this project for species where Beverton-Holt type
recruitment occurs.

A-18

Beverton, R.J.H.. and S. J. Holt. 1956. A review
of the methcKi for estimating mortality rates in
fish populations, with special reference to
sources of bias in catch sampling. Rapp. P.-V.
Reun., Cons. Int. Explon Mer. 140:67-83.

Abstract

Comment

This paper is unavailable for review but has
been requested through interlibrary loans. It will
be reviewed when received.

A-19

Beverton, R.J.H., and S.J. Holt. 1957. On the
dynamics of exploited fish populations. Min.
Agr. Fish. and Food (U.K.), Fish. Invest. Ser. II,
19. 533 pp.

Abstract

Comment

This paper is unavailable for review but has
been requested through interlibrary loans. It will
be reviewed when received.

A-20

Bledsoe, L.J., J. Buss, N. Ehrhardt, and C. Lee. 1974. Non-applicable
NEPAC -- A system for evaluating the consequences
of regulatory structures in northeastern Pacific
fisheries. Tech. Rept. 55, NR 15, NORFISH, Univ.
of Washington Sea Grant.

Abstract

The report describes specifications for a proposed
model of the northeastern Pacific fisheries. Discussed
are:
1. A series of population dynamic submodels for the
stocks in various locations.
2. A series of assumptions about different strate-
gies employed in fishing by various fishing
vessels.
3. A complex bookkeeping system for accumulating
and summarizing biological and economic yield
data over a period of simulated time.

Comment

The paper is, in essence, a proposal to develop
a detailed simulation model of a fishery. The concepts
presented may be useful, but the detailed information is
not applicable to Maryland fisheries.

A-21

Bledsoe., L.J., and P. Katz. 1976. Management information Non-applicable
for regional fishery systems with special reference
to the northeast Pacific. Tech. Rept. 62, MRFISH
NR 27, University of Washington Sea Grant.

Abstract

The paper discusses factors and processes which must
be incorporated into any modeling effort aimed at a multi-
species, geographically dispersed fishery.

Comment

The paper presents interesting and useful approaches
to development of models for management purposes. However,
no mathematical formulations are presented.

A-22

Bonar, A. 1977. Relations between exploitation, yield, Non-applicable
and community structure in Polish pikeperch
(Stizostedion luciopercq.) lakes, 1966-71. J. Fish.
Res. Bd. Canada 34: lS76-lS80.

Abstract

A review of official fishery records maintained
on 43 Polish pikeperch (Stizostedion lucioperca) lakes
during 1966-71 showed that yield was dependent
on exploitation intensity and community structure.
Community structure was studied on the basis of
three groups of fish: predatory, undesirable,
and valuable nonpredatory. The percent of predatory
fish was used as an index of community structure.
An increase of 1 unit of exploitation intensity
increased yield by 0.42 - 0.50 kg/ha and a 1% increase
in predatory decreased yield by 0.45 - 0.49 kg/ha.
Undesirable fish predominated in lakes with Iligh
yields and predatory fish predominatled in lakes
with low yields. The largest catches of predatory
fish species occurred when their percentage in
the total catch reached 25. A more effective
regulation of predatory and nonpredatory fish
populations may be achieved through controlled
exploitation.

Comment

The relationships presented in this paper.
appear to be site specific with little relevance
to general fisheries yield models.

A-23

Booth, D.E. 1972. A model for optimal salmon-manage- Non-applicable
ment. Fish. Bull. 70:497-506.

Abstract

Considerable attention has been given in the liter-
ature recently to continuous time dynamic maximizing
models for fisheries in general, but the time discrete-
ness and interdependency problems encountered in the
case of most salmon fisheries have been largely ignored.
Hence, a discrete time profit maximizing model for a
salmon fishery is developed in this paper, and it is
shown that a correct salmon management policy takes
the form of an investment decision with respect to
the level of escapement and that a management policy
of maximum sustained yield may be incorrect from an
economic standpoint. It is hoped that continued re-
search including construction of a working model will
provide some indication of the difference between the
magnitude of spawner stocks selected on the basis of
maximum sustained yield and stocks selected on the
basis of economic optimality.

Comment

This paper is too species specific to be of use
in this project. Optimization procedures can be applied
only to a working yield model.

A-24

Brown, B.E., J.A. Brennan, and J.E. Palmer. 1979. Applicable for
Linear programming simulations of the effects parameter 'estimation
of bycatch on the management of mixed
species fisheries off the Northeastern Coast
of the United States. Fish. Bull. 76:51-860.

Abstract

The results of using historic bycatch (inci-
dental catch) ratios in adjusting fishing regula-
tions by linear programung techniques are evaluated.
They used both 1971 and 1973 bycatch ratios
separately to assess the sensitivity of the
results to the reported changes in bycatch ratios
in estimating the total 1975 catch of countries
fishing in the northwest Atlantic. For 4 of the
11 countries for which data were examined, the
difference between the percentage of a country's
species total allowable catches (i.e., those
catches allowed a country by regulation) using
the 1971 and 1973 bycatch ratios, was at least
20%. Only four countries were predicted to catch
at least 80% of their species total allowable
catches. The predicted total catches of all
countries and all species was only 60% of the total
species quotas. The simulated directed fisheries
constituted only 70% of the total catch using
1971 bycatch ratios and only 73% using 1973
bycatch ratios. Examination of the reported 1975
catches indicated that the total allowable
catches for herring were most frequently limiting
a country's catch. Except for U.S.S.R., the
differences between reported and simulated
catches were less than 50 metric tons, with the
difference less than 10 metric tons for 6 of the
11 countries. There was little difference in
reported versus simulated catches between the
schemes using the 1971 and 1973 bycatch.ratios.

Comment

The paper discusses a simple bycatch/yield
ratio, which is important as a management tool
or to incorporate into other models. It falls
into the category of parameter estimation.

A-25

Caddy, J.F. 1975. Spatial model for an exploited Applicable
shellfish population, and its application
to the Georges Bank scallop fishery. J. Fish.
Res. Bd. Canada 32:1305-1328..

Abstract.

Existing fisheries models employ the "unit
.stock" concept that makes no explicit allowance
for spatial distribution of biomass and effort
over a fishing ground. The utility of the unit
stock concept rests largely upon the "dynamic
pool" assumption. However, this is both invalid
for sedentary species, and difficult to apply
when information on spatial distribution by
statistical subunits is available, as for the
Georges Bank (lat. 420N, long. 670W) scallop
population. By reference to the Georges Bank
scallop fishery, more realistic general assumptions
for a spatial model of shellfish populations are:

a) Recruitment occurs in patches of random
size and location with the constraint that local
biomass does not exceed the virgin biomass of
each unit area.

b) The fraction of effort expended within
each statistical area of a fishing ground is either
determined by available local biomass alone
(proportional effort allocation) or in combination
with "traditional fishing practice."

Therefore, a spatial mod el CYRAREA) simulating
nonrandom recruitment and harvesting of sedentary
organisms is postulated and applied to Georges Bank
scallop stocks. Some of the general predictions
of this model differ significantly from those
employing the unit stock concept, as follows:

1) Under proportional effort allocation,
overall yield declines more sharply with increasing
effort subsequent to maximum sustained yield MY) than
under dynamic pool assumptions. 2) Peak mortality
and the apparent point of full recruitment on
the catch curve occur progressively earlier in life
(even at partially recruited ages) with increased
effort or degree of clumping of the population,
and subsequently decline with age, except where
"fully recruited" ages coincide spatially with the
"target" age-group making up the largest biomass
component. This may be of general relevance to

A-26

fisheries under intensive exploitation with
sophisticated methods of navigation and fish finding;
peak mortality may occur earlier in life than
predicted from the gear selection ogive, if year-
classes are independently distributed and there is
no size limit regulation. 3) Variance in
biomass/unit area is predicted to fall with age.

For the Georges Bank scallop population,
the model described herein predicts that a
substantial increase in yield would result from
diversion of effort from the Northern Edge to
less heavily fished areas of the bank.

Coment

This paper describes in great detail a
realistic yield model. It appears to be very
relevant to the management of other shellfish
species.

A-27

Carlander, K.D. 1977. Biomass, production, Non-applicable
and yields of walleye (Stizostedion vitreum
vitreum and yellow perch (Perc flavescens)
in North American lakes. J. Fish. Res.
Bd. Canada 34: 1602-1612.

Abstract

Compilation of available data indicated that
walleye (Stizostedion vitreum.vitreum biomass
in lakes averaged 16 kg/ha, but the data were not
adequate to show relationships with mean depth,
alkalinity, latitude, or morphoedaphic indexes
of the lakes. Yellow perch (Per a flaVescens)
biomas@ also failed to show relationships with
these factors. In small lakes and ponds with
only perch, biomass ranged from 39 to 215 kg/ha,
but in lakes with other species, perch biomass
was under 65 kg/ha. Annual production of walleye
was from 1.2 to 4.1 kg'hg-1-yr-1 and that of
yellow perch was 21.9 kg*ha-1*yr-l. Average
comercial-yields of walleye ranged up to 3.06
kg*ha-l*yr and sport fish yields averaged
3.7 kg*ha-r@r-l. Annual commercial and sport
yields decreased with latitude. Area of lake was
negatively correlated with sport yield, but
positively with commercial yield. The latter
situation is believed to be an artifact of the
sample and not a general trend. Commercial yield
increased with total dissolved solids of the
lakes. Lack of other correlations may be related
to the fact that walleye biomass and yield do
not bear a constant relationship to total biomass
and yield.

Comment

The procedures presented here appear
inapplicable to estuarine and marine environments,
which exhibit wide environmental variability.

A-28

Carlson, E.W. 1969. Bio-economic model of a fishery. Applicable
Working Paper No. 12. Bureau of Commercial
Fisheries, Division of Economic Research.

Abstract

This paper is an attempt,to restructure the eco-
nomic theory of fisheries along the lines of received
microeconomic theory. This approach has many virtues
not the least of which is that the power of traditional
microeconomic theory can be marshalled to help solve
various problems that might arise. Another benefit
is that if used fishery economists will be able to com-
municate among themselves and other economists more
readily.
The main thrust of the model is that there is a
divergence between the social cost and the private costs
of harvesting that is partly because of the probabilistic
nature of fish capture and because of the density depen-
dent growth of the fish population.
Actual implementation of the model in its present
form is perhaps not possible because the growth curve
for a fish population does not exist in such a prede-
termined fashion. Rather for most exploited fisheries
the growth of the population is to a large extent deter-
mined by the birth of the fish which appears to have an
extremely high variance. The proper management of a
fishery will take this into account so that each year
class will be exploited optimally. The method for
fishery management under these conditions is outside
the scope of this paper.

Comment

The paper presents what may be useful approaches
to modeling the economics of a fishery.

A-29

Caswell, H. 1972. A simulation study of a Non-applicable
time lag population model. J. Theor. Biol.
34:419-439.

Abstract

The lack of terms-representing time lags is
a widely recognized weakness of the Volterra-Gause
formulations of population growth and interaction.
A number of analytical attempts to include time
lag terms have been made in the past, but they
have provided relatively little ecological insight.

In this paper the Volterra-Gause equations for
a predator and two competing prey are modified
to include a number of time lags, and studied by
simulation. The modifications include lags in
the response to intra- and inter-specific competition
and food supply, and terms that represent the "hunger"
of the predator population. Predation is made
size-selective as a function of hunger, and
refuges are provided for the prey.

Adding these factors greatly increases the
variety of behavior exhibited by the model.
Refuges for some minimum number of prey are required
to assure persistence of the system. If the
system persists its behavior ranges from damped
to undamped oscillations of varying amplitude.
Time lags in response of the predator to changes
in prey abundance, the physiological food require-
ments of the predator, and the heterogeneity of
the environment (as measured by prey refuges) all
affect the form of the oscillations. The time lag
terms can reverse the outcome of competition in
some cases. In this model selective predation
stabilizes otherwise unstable campetitive relations.

Certain results of laboratory predator-prey
studies are difficult to explain in terms of the
standard Volterra-Gause formalism. It is suggested
that some of these features can be explained in
terms of time lags in various population responses.

Comment

The theoretical material presented does not
appear to be applicable generally to fisheries
yield models.

A-30

Chapman, D.G. 1973. Spawner-recruit models and estimation Applicable
of the level of maximum sustainable catch. Con. Per.,
L'Exploration de la Mer, Rapports No. 164:325-332.

Abstract

The author discusses various mathematical formulations
of spawner-recruit relationships (e.g., Ricker, Beverton-
Holt, etc.) and the relationships of these formulations to
each other. Advantages and.disadvantages of the formulations
are discussed relative to actual application to a fishery.
A non-parametric method of estimating optimum catch is
presented and applied to a fur seal population.

Comment

This paper presents a valuable discussion of stock
recruitment models and develops a method of estimating
optimum catch which may be useful in Maryland.

A-31

('11cli, C.W. , arid G.T. Orlob. 1972. Ecologic simulation Applicable
for aquatic environments. Prepared for Office of
Water Resources Research, U.S. Dept. of the Interior.

Abstract

A mathematical model for computer simulation of
aquatic ecosystems was developed and adapted to lake
and estuarial systems. The model is capable of simula-
ting the annual cycle of ecologic successions involving
algae, bacteria, zooplankton, fish and benthic animals
and the interdependent relationships between biota and
abiotic substances carried in the natural aquatic system.
It is water quality oriented, predicting the temporal
and spatial distributions of temperature, dissolved
oxygen, biochemical oxygen demand, pH, conservative
constituents (e.g., salinity, TDS, etc.), toxicity,
nitrogen (three forms), carbon dioxide, and phosphorus
as well as the biomass of each trophic level in the
system. The basic formulations in the model are based
on kinetic principles and the law of Conservation of
Mass. Algal growth kinetics are governed by a Michaelis-
Menton relationship including light, temperature,
carbon, nitrogen, and phosphorus.
A demonstration of simulation capability of the
Lake Ecologic Model was performed on Lake Washington
using the data of W.T. Edmundson covering the periods
1962-63 and 1967-69 for comparison. The earlier period
corresponded to a condition of incipient eutrophication
and the latter period to recovery of the lake after
diversion of waste water inflows. The model reproduced
with fair reliability the major quality changes that
were observed over the annual cycle in each of the two
periods. A demonstration of the Estuary Ecologic Model
was performed on the San Francisco Bay-Delta System using
data gathered by the University of California. The
model simulated closely the actual response of the system.
It was further tested for several alternative conditions
of water quality control involving waste water treatment,
relocation of outfalls and flow regulation.

Comment

This paper describes a very detailed ecosystem model.
Parts of this model and the approaches used in integrating
its components may be of use in this'study.

A-32

Christensen, S.W., D.L. DeAngelis, and A.G. Clark. Applicable for
1978. Development of a stock-progeny model parameter estimation
for assessing power plant effects on fish
populations. pp. 19S-225 In W. Van Winkle (ed.).
Assessing the Effects of P@o_wer-Plant-Induced
Mortality on Fish Populations. Sponsored by
Oak Ridge Nationa! Laboratory, Energy Research
and Development Administration, and Electric
Power Research Institute.

Abstract

A multi-age-class model, based on simple but
general biological principles, is developed to
assess the impact of power plants on fish populations.
The model is then parameterized in order to produce
a variety of stock-progeny relationships, assuning
that the stock is always at stable age distribution.
The predicted response of the fish stock to power
plant cropping of young-of-the-year fish is
investigated for each of these stock-progeny
relationships. In general, the sensitivity of the
equilibriun stock size to cropping is positively
related to the slope of the stock-progeny curve
at the equilibriun point and, to a lesser extent,
negatively related to the slope of the curve at
the origin. In addition, the timing of power-plant-
induced mortality that can be tolerated by the
stock can be calculated from the sloDe of the
curve at the origin. Application of@ the model
to specific cases will likely need to utilize
time-series simulations in addition to the steady-
state approach investigated here.

Comment

The model described here is not directly
applicable to yield, but has possibilities as a
recruitment submodel.

A-33

Clady, M.D. 1975. The effects of a simulated Applicable
angler harvest on biomass and production
in lightly exploited populations of smallmouth
and largemouth bass. Trans. Amer. Fish.
Soc. 104: 270-276.

Abstract

Harvests of 13 to 21%, mostly of larger, older
0
fish, were imposed for two years on populations
of smallmouth and largemouth bass (Micropterus
dolomieui and M.,salmgde
I ga) that previously had
been virtually unfished. Another population of
smallmouth bass served as a control. Using trap
nets, estimates of the nunber of bass and their
size were made for one year before and two years
after initial harvest. From these data, annual
natural mortality, growth, standing crop and
production were computed. There were no strong
and consistent changes in any parameter which
could be attributed to reductions in numbers of
fish. Changes in growth unrelated to population
density apparently resulted in declines of ratios
of production to biomass following harvest.
Possible causes of the lack of compensatory
responses to harvest include the relatively short
period of study and time lag in changes in growth
and mortality.

Comment

The approach-is possibly useful in situations
where a unit stock is known. The approach is pri-
marily applicable to lake/stream (enclosed waterways)
situations but may.be a-good approach for species
such as @vhite catfish.

A- 34

Clark, CJV. 1972. The dynamics of commercially exploited Applicable
natural animal populations. Math. Bio. 13:149-164.

Abstract

A simple model is analyzed to describe the temporal
behavior of a natural animal population that is subject
to economically optimal exploitation. It is shown that
the population will reach an optimal equilibrium level,
at which the marginal growth rate of the stock equals
the accepted interest rate. (In some cases, an a priori
constraint, prohibiting destruction of the stock, is
required to reach this conclusion.) Since the optimal
equilibrium yield is necessarily smaller than the maximum
sustainable yield, it follows that the exploiting agency
will experience a degree of overcapitalization once the
optimal level is reached. Failure to observe this
phenomenon would lead to overexploitation of the resource.

Comment

The paper describes a means of determining an
optimum (in an economic sense) yield from a resource
population. It may be of value in developing bioeco-
nomics models of Maryland stocks.

A-35

Clark, C.W. 1973. The econanics of over- Applicable
exploitation. Science 181: 630-634.

Abstract

The general economic analysis of a biological
resource presented in this article suggests that
overexploitation in the physical sense of reduced
productivity may result from not one, but two
social conditions: camon-property competitive
exploitation on the one hand,, and private-
property maximization of profits on the other. For
populations that are economically valuable but
possess low reproductive capacities, either condition
may lead even to the extinction of the population.
In view of the likelihood of private firms adopting
high rates of discount, the conservation of renewable
resources would appear to require continual public
surveillance and control of the physical yield
and the ccndition of the stocks.

Cament

This paper is very theoretical in nature adding
no new information beyond that in Clark et al. (1973);
however, itmay prove helpful in the analysis of bio-
economics =dels.

A- 36

Clark, CJV. 1976. A dclayed-r6cruitment model of Non-applicable
population dynamics, with an application to
baleen whale populations. J. Math. Biol..
3:381-391.

Abstract

A simple delay of harvested populations has been
considered. A necessary and sufficient condition
for stability of the natural equilibrium has been
presented. Harvest policies that maximize discounted
economic rent have been described. These policies
are characterized by optimal equilibrium levels
of escapement. It is argued that, although the
theoretically optimal dynamic policy requires an
asymptotic approach to equilibrium, in practice
the 1!most-rapid" (or "bang-bang") approach policy
is at worst only marginally suboptimal. The
results are applied to the Antarctic fin whale
fishery.

Comment

This approach is not useful as it adds little
new information to that presented in earlier works.

A-37

Clark, C., G. Edwards, and M. Friedlaender. 1973.' Applicable
Beverton-Holt model of a commercial fishery:
optimal dynamics. J. Fish. Res. Bd'. Canada
30:1629- 1640.

Abstract

The problem of optimal regulation of a fishery
is discussed. Of special interest is the problem
of regulating an overexploited fishery by reducing
effort to allow the fish population to build up
to a suitable level.

It is argued that the problem requires an
economic analysis based on the concept of maximiza-
tion of present value. From this concept we then
deduce a simple, general rule, the "Fisher Rule,"
which we subsequently use to determine optimal
exploitation. Among the principal results are the
following: (a) an optimal mesh-size is determined,
which, because of the discounting of future revenues,
is smaller than the size corresponding to maximun
sustainable yield; (b) the optimal recovery policy
for an overexploited fishery is deduced; it
consists of a fishing closure permitting the fish
population to reach an optimal age; (c) the
optimal development of an unexploited fishery is
deduced; an initial development stage characterized
by large landings and profits is rapidly transformed
into a situation of optimal sustained yield, in
which both landings and profits may be significantly
reduced; (d) the optimal policy is deduced for a
fishery in which gear limitations are impractical;
the result may be a strongly unstable fishing
industry; (e) the effect of high discount rates,
which might be employed by private fishing interests,
is discussed; such rates may result in overfishing
similar to the case of a ccnnon-property fishery.

Comment

This paper develops the concept of an optimal
sustained economic yield. It may be useful as a
means of adding economics to production models.
Optimal age, length, and weight in an economic
framework can also be determined by this approach.

A-38

Clark, C.W., and G.R. Munro. 1975. The economics Applicable
of fishing and modern capital theory: A
simplified approach. J.: Envaron. - Econ.
Manag. 2:92-106.

Abstract

-While the link between fisheries economics
and capital theory has long been recognized,
fisheries economics has, until the last few years,
developed largely along nondynamic lines. The
purpose of this article is to demonstrate that,
with the aid of optimal control theory, fisheries
economics can without difficulty be cast in a
capital-theoretic framework yielding results that
are both general and readily comprehensible.

We commence by developing a dynamic linear
autonomous model. The static version of the
fisheries economics model is seen to be the
equivalent of a special case of the dynamic
autonomous model. The model is then extended, first
by making it nonautonomous and second, nonlinear.
Problems arising therefrom, such as multiple
equilibria, are considered.

Comment

This model appears useful for the determination
of optimal standing stock level and harvest level in
0
an economic context. It seems to ignore the
concept of steady-state populations.

A-39

Clark, R.D., and R.T. Lackey. 1975. Computer-implemented Non-applicable
simulation as a planning aid for state fisheries
management agencies. Dept. of Fish. and Wildlife.
Virginia Polytechnic Institute and State University.
FIVS-3-75. 179 pp.

Abstract

A basic job of fisheries management agencies is to
forecast the demand and produce the necessary supply of
fishing opportunities. Present day angling consumption
rates often exceed managers' ability to supply fishing
opportunities of the desired quality. Therefore, a
primary means for improving fisheries management may
be to regulate angling consumption. Operations research
techniques are well suited for handling the.complexities
involved with planning multiple action policies for regu-
lating angler consumption.
PISCES is a computer-implemented simulator of the
inland fisheries management system of Tennessee, but is
adaptable for use in any state. The purpose of PISCES
is to aid in planning fisheries management decision
policies at the macro-level. PISCES generates predic-
tions of how fisheries management agency activities will
affect angler use for a fiscal year. Subjective prob-
ability distributions for random variables and Monte
Carlo simulation techniques are employed to produce an
expected value and standard-deviation for each prediction.
Test runs under realistic hypothetical situations and
discussions with personnel of Tennessee Wildlife
Resources Agency suggest that PISCES may help fisheries
management agencies to improve budget allocation decis-
ions, to formulate multiple action policies for regu-
lating angler use, and to enhance fisheries development.
A hypothetical application of PISCES in Tennessee is
given.

Comment

The model presented deals with angler use of areas-
rather than yields. It does not appear useful in this
study.

A-40

Clark, R.D., Jr., and R.T. Lackey. 1976. A Applicable for
technique for improving decision analysis parameter estimation
in fisheries and wildlife management. Va.
J. Sci. 27:199-201.

Abstract

Computer implemented modeling has clearly
benefited managers in industrial and military
management positions. Improvements in the cost
effectiveness of computer use should allow fisheries
and wildlife managers to make increasing use of
computerized decision models. A modeling technique
developed for industrial purposes to utilize the
full state of knowledge of a decision problem is
presented here and proposed for use in fisheries
and wildlife management. The technique uses the
Weibull function for subjective probability
assignment to help solve the problem of making
decisions based upon incomplete or inadequate data.

Comment

This paper presents a good method of estimating
non-quantitative variables and their probability
density functions.

A-41

Cliff, E.M., and T.L. Vincent. 1973. An optimal Non-applicable
policy for a.fish harvest. J. 2ptim._Theory
Appl. 12:485-496.

Abstract

A dynamical model for harvesting a fish
population system is proposed by introducing control
into the known Verhulst-Pearl model. An optimal
control problem including sane parameters is
stated, and the usual necessary conditions are
applied. For specific parameter values, the
candidate control policy is deduced, and optimality
is verified by applying a sufficiency theorem.
The optimal trajectories may contain maximum
and minimum control arcs as well as a singular
subarc. The significance of the singular arc
is interpreted in terms of the system dynamics.

Comment

This paper does not appear very useful. It
uses optimal control theory to minimize cost but
ignores other externalities.

A-42

Crutchfield, J.A. 1973. Economic and political objectives in Non-applicable
fishery management. Trans. Am. Fish. Soc. 102:481-491.

Abstract

The changes in attitudes from MSY to optimal sustained
yields and extensions of bioeconomic fishing theory are
discussed in relation to their inherent social goals. The
role of political and administrative objectives as they
relate to fishery management are also discussed from this
perspective.

Comment

This paper is not useful for this project. It represents
discussion of the social goals of fishery management and is
beyond the present scope of the project..

A-43

Cushing, D.H., and J.G.K. Harris. 1973. Factors Non-applicable
affecting recruitment and mechanisms of recruit-
ment control. @a:4. P.-V. Reun., Cons. Inter..
Explor. Mer. 164:14Z-155.

Abstract

This paper is in three parts: (a) an examination of
the biology of fishes which suggests that the larval drift
in the plankton from spawning ground to nursery ground is the
stage in the life history at which density dependence is most
marked. It will be suggested that density dependent growth and
mortality are linked and that both competition and the natural
regulation of numbers may take place during the larval drift;
(b) Cushing (1968, 1971) related density dependence to
fecundity; fish with low fecundities were expected to have
near-linear curves of recruitment on parent stock, whereas
the dome-shaped curves would be characteristic of fish with
high fecundities. An index of density dependence was derived
from a log-log plot of recruitment on parent stock, which
is a rough way of averaging the multifarious range of data.
The tentative conclusion, that density dependence was a
function of fecundity, should be tested and in this paper the
same data are re-examined and fitted to a stock and recruitment
curve, with an estimate of error; (c) any curve of recruitment
on parent stock expresses the ratio of density dependent to
density independent mortality at different stock levels.
Obviously an estimate of this ratio independent of the data
on recruitment and parent stock is needed. As a first step
towards this, a model of the generation of density dependent
mortality in the North Sea plaice has been developed.

Comment

This paper is a good review of stock-recruitment relationships
but is of little direct use in the application of fish yield models.

A-44

DeAngelis, D.L. 1976. Application of stochastic Applicable
models to a wildlife population. Math. Biosci.
31:227-236.

Abstract

Two stochastic population models, a birth-and-death
model and a stochastic difference equation model, are
compared, using data for a giant-Canada-goose population.
The quasistationary probability distribution predicted by
the stochastic difference equation agrees well with values
of the population mean and variance computed from several
years of data.

Comment

The birth and death model was shown to be inappli-
cable to populations which are very stable.. unlike most
fisheries. The stochastic difference equation model is
very simplistic, but could be derived from empirical
data. It might be useful in some instances.

A-45

DeAngelis, D.L., S.W. Christensen,and A.G. Clark. Applicabli
1977. Responses of a fish population model to young-
of-the-year mortality. J. Fish. Res. BeL. Canada
34:2124-2132.

Abstract

A multiple-age-class model is used to examine the
effects of increases in density-independent young-of-the-
year mortality caused by power plan entrainment of larval
fish. It is demonstrated analytically that in all
realistic cases, an increase in such mortality results in
a smaller equilibrium population density of adult fish.
The stability of the population with respect to perturbations
about its equilibrium point is increased in these cases.
However, situations can occur where a slight increase in
mortality causes a catastrophic population decline.
The model is used to generate autoregression graphs of
population numbers that can be compared with field data.

Comment

The model incorDorates some novel ideas into a Dopulation
model, and it could have some value to this study.

A-46

DeAngelis, D.L., W. Van Winkle, S.W. Christensen, Applicable
S.R. Blum, B.L. Kirk, and C. Ross. 1977.
A generalized fish life-cycle population
model and computer program. ORNL/71,1-6125.
Oak Ridge National Laboratory, Oak Ridge,
Tennessee. 176 pp.

Abstract

A generalized fish life-cycle population model and
computer program have been prepared to evaluate the long-
teTm effect of changes in mortality in age class 0. The
general question concerns what happens to a fishery when
density-independent sources of mortality are introduced
that act on age class 0, particularly entrairment and
impingement at power plants. This paper discusses the
model formulation and computer program, including sample
results.
The population model consists of a system of dif-
ference equations involving age-dependent fecundity and
survival. The fecundity for each age class is assumed
to be a function of both the fraction of females sexually
mature and the weight of females as they enter each age
class. Natural mortality for age classes I and older is
assumed to be independent of populatim size. Fishing
mortality is assumed to vary with the number and weight
of fish available to the fishery. Age class 0 is divided
into six life stages.* The probability of survival for
age class 0 is estimated considering both density-inde-
pendent mortality.(natural and power plant) and density-
dependent mortality for each life stage. Two types of
density-dependent mortality are included. These are
(1) cannibalism of each life stage by older age classes
and (2) intra-life-stage competition.

Comment

This paper represents an excellent example of a
simulation model which utilizes Leslie matrices, incor-
porates fishing mortalities, and explicitly calculates
yield. While the data requirements of this model are
large, its generalized form appears applicable for
several Maryland fisheries such as striped bass and
herring.

A-47

Doubleday, W.G. 1976. Environmental fluctuations and Non-applicable
fisheries management. Int. Commi. Northwest Atl..
Fish. Sel. Pap. 1:141-lTO-

Abstract

The effect of random fluctuations in production on
the success of fisheries management schemes is examined
using a discrete version of Schaefer's (1954) model.
Control of stock biomass, catch, and effort are [sic]
considered. The average yield taken is shown to be
inversely related to yearly fluctuations in yield. Con-
trol of stock biomass maximizes the average yield at
the cost of large fluctuations in catch. Control of
catch requires a large reduction in average yield to
obtain stability. The effects of controlling ' effort
lie between those of controlling biomass and controlling
catch.
The restoring force of an exploited stock to devia-
tions from equilibrium is examined and the presence of
a critical zone of biomass less than one fourth of the
virgin biomass in which further displacement weakens the
restoring force and from which recovery of stock biomass
is slow is noted. It is shown that control of effort at
a level corresponding to an equilibrium biomass of two
thirds the virgin stock instead of one half as is commonly
recommended achieves a reduction in catch variance of from
60% to 75% and an increase.of catch per unit effort of
33-1/3% with a loss in yield of 11%. The biomass buffer
between equilibrium biomass and the critical zone is in-
creased 133%, making the stock more resilient to depletion
by a succession of weakyear classes and reducing the need
for rapid changes of regulations based on preliminary
estimates of incoming year-class strength.

Comment

This paper, while demonstrating the importance of
production and recruitment fluctuations in assessing MSY
levels, includes numerous parameters (such as virgin
equilibrium biomass level) which are unlikely to be
determinable for Maryland fisheries. Thus, while the
paper is instructive, it does not appear directly appli-
cable to this project.

A-48

Dow, R.L., F.W. Bell, and D.M. Harriman. 197S. Applicable
Bioeconomic relationships for the Maine
lobster fishery with consideration of
alternative management schemes. NQAA
Technical Report, NMFS-SSRF-683.

Abstract

A bioeconomic model is formulated using basic
Schaefer relationships coupled with cost-profit analysis.
Plots of steady state functions typifying population
dx
steady-state(F =
t 0) and cost-profit equilibrium
Cdk =0) result in optimal economic popula-
Tt
tion yield level.

Coment

The method presented here may be useful for this
project. The lack of accounting for biological
variables other than population size may be a problem.

A-49

Dyer, T.G.J. and.J.F. Gillooly. 1979. Simulating fish Applicable
production using exponential smoothing. Ecol.
Modelling 6:77-87.

Abstract

The variation over time of thetotal annual production of
pelagic fish for South Africa and the United Kingdom has
been described quantitatively using the exponential
smoothing technique. The exercise was repeated on annual
mackerel landings for the same two countries. It is
suggested that in some cases, the production figure for the
current year can be used to simulate the following year's
value. The greater variations in South Africa's annual
production probably gave rise to the poorer results for these
data.

Comment

The model is very simplistic and unrealistic in many
ways. However, as a technique distinct from standard
fisheries models, it may be worthwhile to evaluate it
further.

A-50

Eberhardt, L.L. 1977. Relationship between two stock- Non Applicable
recruitment curves. J. Fish. Res. Bd. Canada.
34:425-428.

Abstract

The Beverton and Holt and Ricker stock-recruit3nent
curves can be used to generate population growth curves.
The Beverton and Holt curve is then identical to a difference
equation model for the logistic growth curve, and may be
derived in terms of equations for linearly density-
dependent population relations.The same equations lead to
the Ricker curve if the density-regulating effect is assumed
to depend only on population size at the beginning of the
interval between generations. At low rates of population
growth, the Ricker curve approaches that of Beverton and
Holt. The two curves appear to represent certain concepts
known in population biology as I'r and K selection."

Comment

The analysis suggests that Beverton-Holt is appropriate
for populations that can only attain modest increases from
generation to generation, whereas Ricker may be appropriate
for species which can exploit particularly advantageous
conditions. Material presented here provides no methods
or procedures which would appear useful to this project.

A-51

Eberhardt, L.L., and D.B.Siniff. 1977. Population dynamics Applicable
and marine mammal managemert policies. J. Fish. Res.
Bd. Canada 34:193-490.

Abstract

Some criteria for appraising population level relative to
the maximal or carrying capacity point are listed. Simple
population-dynamics models are then used to explore some of
the criteria. Age at first reproduction does not seem as
important as in some more prolific and shorter-lived species,
being at most equivalent to a few percentage points of adult
survival. Age-specific reproductive rates are very much the
same for a number of pinniped species. For most marine
mammal species, survival through immature stages is an
unknown quantity, but appears to be a factor of major
importance'in determining population trend. Data on the
Pribilof fur seals (Callorhinus ursinus) lead to the speculative
conclusion that the maxim=_SUsta`1_neU-yield (NISY) point may
be to the right of the median value frequently assumed, so
that "optimal" population levels may be closer to the asymptotic
or carrying capacity level. Such a view is proposed as a
conservative management policy.

Comment

The model should be evaluated for use here, since it
represents application of a production model. Even though the
population parameters of marine mammals may be very different
fr6Uthose of most exploited fish species, the an alysis could
sti-1-1 prove useful.

A-52

Edwards, R.L., and R.C. Hennemuth. 1975. IMaximun yield: Non-applicable
Assessment and attaiment. Oceanus 18:3-!'9.

Abstract

The authors review fisheries management directed toward
maximum yield and critique the procedures commonly used in
developing management programs.

Coment

The paper presents no actual models, but does provide
some useful insights into problems encountered in developing
management strategies.

A-53

Englert, T.L., J.P. Lawler, F.N. Aydin, and
G. Vachtsevanos. 1976. A model study of striped Applicable
bass population dynamics in the Hudson River.
pp. 137-150 In N1. Wiley (ed.), Estuarine Pro-
cesses, Vol. 7, Academic Press, New York.

Abstract

Present and planned operation of electric power generating
stations along the Hudson River may have an impact on the
Atlantic striped bass population due to the use 6f river water
in once-through cooling systems at the plants. Withdrawal
of river water for cooling purposes can have two principal
effects on the young-of-the-year striped bass spawned in the
Hudson: (1) Eggsj larvae and early juvenile fish may be
entrained in the water which is circulated through the
plant's cooling system and returned to the river; (2) later
juvenile fish may be impinged on the debris screens at the
plant intakes.

Population models of the Hudson River striped bass are
useful in making predictions of the impact of entrainment
and impingement at the power plants. In the model study
presented here, results from a detailed simulation of the
young-of-the-year population are input to a model of the
adult bass population in order to predict short- and long-
range impacts on the population.

The young-of-the-year model traces development of the
early life stages of the bass from eggs through the larvae
and juvenile stages. Egg production rates calculated from
field data are used to initialize the population model. The
temporal, spatial and age distributions of the early life
stages are simulated by equations which include the effects
of hatching period,' natural mortality, plant withdrawal
rates and the convective and dispersive effects of the Hudson's
hydrodynamics. The hydrodynamic simulation is intra-tidal
or real-time. The spatial distribution of the organisms is
calculated at twenty-nine longitudinal grid points in both
the upper and lower layers of the river.

Comparisons of model results and field data provide a
measure of the verification of the model.

Comment

The model requires detailed knowledge of environmental
C,
and population parameters. However, it should be reviewed
in depth to evaluate its potential value.

A-54

Eraslan, A.H., W. Van Winkle, R.D. Sharp, S.W. Christensen,
C.P. Goodyear, R.M. Rush and W. Fulkerson. 1976.
A computer simulation model for the striped bass young-
of-the-year population in the Hudson River.
Oak Ridge Nat. Lab., Oak Ridge, Tem.. ORNLINUREG-8.

Abstract

Cotment

This paper is unavailable for review but has been
requested through interlibrary loans. It will be reviewed
when received.

A-55

Everitt, R.R., N.C. Sonntag, M.L. Putterman, and Non-applic e
P. Whalen. 1978. A mathematical programming model for
the management of a renewable resource system: The
Kemano, II development project. J. Fish. Res. Bd. Canada
35:235-246.

Abstract

This paper considers a mathematical programming model for
the manageinent of a salmon fishery and watershed. Management
objectives are specified as bounds and desirable target
levels for certain important variables.@ The mathematical
program consists of minimizing the deviation from these
targets subject to natural and policy constraints. A
detailed application studying the effect of the proposed
Kemano II development project on the salmon fisheries of
northwestern British Columbia i-5 presented. Parametric
programming is used to compare several development scenarios
on the basis of the sensitivity of the system to reduced runoff
in a specified year.

Comment

The specifit fisheries models used are-not applicable
to this study as production.models. However, the mathematical
programing technique could be applicable in the resolution
of conflicts in resource allocation.

A-56

Fletcher, R. 1978. On the restructuring of the Applicable
Pella-Tomlinson system. Fish. Bull. 76:51S-521.

Abstract

The time-dependent analysis of an earlier work is
extended to the equilibrium case of the Pella-Tomlinson
system, and the relationships between the equilibrium
and nonequilibrium versions of the restructured system
are developed. The dual formulations of the conventional
analysis are avoided and maximum sustainable yield is
separated from the indeterminacy of the system. All
arbitrary coefficients are eliminated and the manage-
ment conponents incorporated directly into the system
equations. The source of the statistical degeneracy
in the model is revealed and explicitly formulated.

Comment

The theoretical discussion presented here suggests
an improved means of applying a Pella-Tomlinson system
to an exploited stock. However, its application is
apparently covered in a follow-up paper (Rivard and
Bledsoe, 1978; -Appendix B) .

A-57

Flowers, J.M., and S.B. Saila. 1972. An analysis of temper- Applicable
ature effects on the inshore lobster fishery. J. Fish.
Res. Bd. Canada 29:1221-1225.

Abstract

In the past, water temperature has been utilized in com-
bination with some measure of fishing effort in the development
of economic estimator or predictor equations for the yield of
the lobster, Homarus; americanus. The hypothesis that the
inshore lobster fisEery in the United States has been overfished
since the end of World War II to the point where increases in
fishing effort since that time have had only minor effects on
the yields was examined. It was shown that suitable yield
prediction equations could be developed using only lagged and
present temperatures as the independent variables. Comparisons
were made of equations developed for the Maine fishery and
sections of the Canadian fishery. Further analyses were done
comparing equations developed using winter vs. summer tempera-
tures and surface vs. bottom temperatures.

Comment

This paper presents a good example of the use of statis-
tical techniques to model fishery yield. The methods may be
of some use in the project.

A-58

Fox, W.W., Jr. 1970. An exponential surplus-yield
model for optimizijig exploited fish populations. Applicable
Trans. Am. Fish. Soc. 99:80-88.

Abstract

A surplus-yield model of fishery dynamics which assumes
the Gompertz growth function is developed, resulting in an
implied exponential relationship between catch per unit
effort and fishing effort, and in an asymmetrical yield curve.
A maximum sustainable yield, predicted by the exponential
model, is obtained froma.population size which is about
37% of the environmentally limited maximum size. Three
methods for estimating the parameters of the exponential
model, adapted from those used for the linear model of
Schaefer (1954, 1957), are presented. The exponential model
is compared with the linear model using examples of the
fisheries for the California sardine, Sardino2s caerulea (Girard.),
and yellowfin tuna, Thunnus albacares (Bonnaterre) F
eastern tropical Pacific-and western Atlantic oceans.
Management implications are discussed.

Comment

This article is excellent, *f1ealing urith derivation of
model parameters as well as the basis-foT using different
model forms.

A-59

Francis, R.C. 1974. TUNPOP, a computer simulation model of Applicable
the yellowfin tuna population and thesurface tuna
fishery of the eastern Pacific Ocean. Inter-Am. Trop.
Tuna Comm. Bull. 16:235-258.

Abstract

Mathematical documentation of TUNPOP, an age-structured
computer simulation model of the yellowfin tuna population
and surface tuna fishery of the eastern Pacific Ocean, is
described. Example runs of the model are presented and
discussed, and the sensitivity of,the model output to chmges
in various parameters is examined.

Comment

This paper presents excellent documentation for TUNPOP,
an age-structured simulation model of yield and population
dynamics which employs variable recruitment levels. This
generalized structure, which subdivides stock into unit areas
of exploitation, may be applicable to Maryland fisheries
which suffer interstate and/or international exploitation.

A-60

Francis,, R.C. 1974. Relationship of fishi.-ng mortality to Applicable
natural mortality at the level of maximum sustaindble for
yield under the logistic stock production model. parameter
J. Fish. Res.-Bd. Canada 31:1539-1542. estimation

Abstract

The often-used approximation that, for a stock of fish
under exploitation, the in-stantaneous fishing mortality rate
equals the instantaneous natural mortality rate at the point
of the maximum sustainable yield is examined with respect to
its mathematical roots and practical utility. Examples from
two diverse fisheries are utilized.

Comment

The paper presents a means by which stock-recruitment
relationships can be evaluated. It also disproves an
assumption commonly made in the application of the Schaefer
model.

A-61

Fransz, H.G. 1979. Estimation of birth rate and juvenile Applicable for
.mortality from observed numbers of juveniles in a parameter
mammal population with normally dispersed reproduction. estimation
Ecol. Modelling 7.-125-133.

Abstract

The increase in the number of juveniles,in a mammal
population with a normally dispersed reproduction is simulated
using a computer. The effect of juvenile mortality and its
age-dependency on the recruitment curve is discussed. The
maximum number of juveniles is reached before all juveniles
are born. The ratio of the period of time between the
beginning of the reproduction period and the maximum of
juveniles and of births (the coeffecient of reduction of the
apparent reproduction period) is related to the juvenile
mortality rate and the ratio of the maximum number of
juveniles to the total number of births. These relationships
can be used to estimate the total number of births and the
juvenile mortality rate from a series of counts of the juveniles.
The simulation model used is programmed in CSMP-III.

Comment

The model presented here would generally not be useful.
The approach is applicable to contained populations with low
seasonal fecundities (1-5 offspring per female) typical of
mammal populations and would thus be of little use in
Maryland fisheries. However, methods for estimating mortality
rates may be of value.

A-62

Gales, L.E.,, and J.A. Buss. 1975. NEPTUNE: A FORTRAN- Non-applicable
based hybrid simulator for fishery systems. Univ.
Wash. Sea Grant, Seattle 1@hsh., Norfish Tech. Rept.
6S. 39 pp.

Abstract

Over the past 15 years a great variety of computer
models have been developed for the study of complex
systems. Despite this variety, most models can be classi-
fied as:

1. CSCT - continuous space, continuous time,
2. DSCT - discrete space, continuous time,
3. DSDT - discrete space, discrete time.

CSCTmodels view the world as a continuous spatial field
which is expressible in terms of partial differential
equations; DSCT models view it as a set of lumped objects
whose behavior is governed by ordinary differential equa-
tions; and DSDT models view i-c a-,:; i sa"Z of discrete
objects tha-_- are activated and de-activated at discrete
points in time.
This paper describes a hybrid DSCT1-DSDT simulation
system which is inplemented in- CDC FORTIUN and CDC UPDATE.
It draws heavily upon the concepts and nomenclature of
SIMULA 67, and provides some of the simulation and data
structuring features of that 'i'maguage, but with the many
practical advantages of FORTRAN.

Comment

This paper presents a discrete space-discrete time,
discrete space-continuous time interactive simulation
model of a fishery. Because the text of the paper dis-
cusses only the progra=ing aspects of the model, an evalu-
ation of applicability to Maryland fisheries based on
model assumptions is not possible.

A-63

Gales, L.E., J.A. Buss, and L.J. Bledsoe. 1977. Non-appliJe
Simulation concepts for fishery systems. J. Fish. Res.
Bd. Canada 34:2374-2380.

Abstr*act

Computer models for analysis of fishery systems may
involve complex applications of mathematics and computer
science. While the mathematical aspects of these models
have been thoroughly studied, less,attention has been paid
to methodologies for their compu@er implementation. As a
result, some models excessively simplify fishery systems to
avoid difficulties in implementation. This study presents
an advanced computer technique (linked lists) which permits
the inclusion of complex time simulations and data structures
in fishery models with little additional effort. Examples
of application to complex information arrays and an age-
structured population model are included.

Comment

This article does not present a model, but could prove
useful in the application of a model or in the computeri-
zation of a data base.

A-64

Gatto, M., and S. Rinaldi. 1976. NL-an value and variability Non-applicable
of fish catches in fluctuating environments. J. Fish.
[Ze-s. B(I. Cant-i(la 33:1.89- M.

Abstract

The mean value of the catch and its variability due to
environmental fluctuationsare analyzed for a very general
stock-recruitment model. Particular attention is devoted to
the comparison of two standard fishing strategies (constant
effort and constant escapement) in terms of mean catch,
variance in catches, and maximum deviation of catch. It is
demonstrated analytically that constant escapement policies
should always give higher mean catch, but should give higher
catch variance and more extreme catches only under certain
conditions of environmental variability.

Comment

The model presented here does not have general utility.
It represents a statistical attenpt to interpret the conse-
quences of constant escapement versus constant effort manage-
ment policies. This evaluation is essentially identical to
earlier literature.

A-65

Gatto, M., S. Rinaldi, and C. Walters. 1976. A predator-
prey model for discrete-time commercial fisheries.
Int. Inst. Appl. Syst. Anal., Laxenburg, Austria,
Res. Rep. Ser. 75-5. 38 pp.

Abstract

Comment

This paper is unavailable for review but has been
requested through interlibrary loans. It will be
reviewed when received.

A-66

Geaghan, J.P. 1978. Development and application of the Non-applicable
Schaefer model in fisheries management. Unpublished
manuscript.

Abstract

The development of the logistic-based surplus produc-
tion model of Schaefer is traced from its inception in
1954 through its adapted form in bioeconomic models.
Comparative evaluations of the equilibrium and non-equili-
brim forms of this model type are made statistical using
the American lobster and tuna as examples.

Comment

While providing an excellent overview of the develop-
ment of the Schaefer surplus production model, this paper
provides little in the way of new information concerning
this model type. It will not be directly applicable to
Maryland fisheries other than as a review of surplus pro-
duction model techniques.-

A-67

Goh, B. S. 1.969. Optimal control of a fish resource.
Malayan Scientist 5:65-70.

Abstract

Coment

This paper is unavailable for review but has been
requested througgh interlibrary loans. It will be reviewed
when received.

A-68

Graham, N1. 1935. Modern theory of exploiting a fishery, Applicable
and application to North Sea trawling. J. Cons., Cons.
Inter. Tx-plor. Mer. 10:2.64-274.

Abstract

This classical.fisheries paper explores that concept
that a "certain proportion of the time and money of fishermen
is ... devoted to reducing their catch or is at least wasted.,,
The theory of the "economy of effort" has essentially tivo
postulates: (1) reduction in mortality (i.e., by reducing
fishing rate) will increase the mean age of the population
and the inverse holds true; and (2) there is a most profitable
age to harvest a fishery. The concept of maximm sustainable
yield is developed.

Comment

This paper represents an early attempt to incorporate the
logistic growth concept into fisheries management and is one of
the initial uses of the concept of MSY. It could be applicable
to Maryland fisheries.

A-69

Grant, W*E. and W.L. Griffin. 1979. A bioeconomic model Applicable
of the Gulf of Mexico shrimp fishery. Ttans. An,. Fish.
Soc. 108:1-13.

Abstract

A bioeconomic model of the brown shrimp (Penaeu aztecus)
fishery in Galveston Bay, Texas, and adjacent offshore waters
accurately predicts the general trends in the seasonality of
shrimp harvet and the distribution of the harvest in relation
to size of shrimp and water depth.

Comment

This paper presents a valuable first attempt to
explicitly integrate population dynamics and economic
price-cost analysis. It could prove useful.

A-70

Gulland, J.A. 1964. Manual of methods for fish populations
analysis. FAO Fish. Tech. Pap. 40. 60 pp.

Abstract

Comment

This paper is unavailable for review but has been
requested through interlibrary loans. It will be reviewed
when received.

A-71

Gulland, J.A. 1969. Manual of methods for fish stock assess-
ment. Part 1. Fish population analysis. FAO Man.
Fish. Sci. 4. 154 pp.

Abstract

Comment

This paper is unavailable for review but has been
requested through interlibrary loans. It will be reviewed
when received.

A-72

Hackney, P.A., and C.K. 16,1inns. 1974. A computer Applicable
model of.biomass dynamics and food competition with
implications for its use in fishery management. Trans.
Am. Fish. Soc. 103:215-225.

Abstract

A model useful for managing interacting fish species,
based on concepts of food competition, is developed and tested.

The model consists of differential equations. One, in
the Pearl-Verhulst idiom, specifies organisms which can serve
only as food resources for consumers of higher trophic
status. A "storage" equation describes ingestion and assimilation
of food by a consumer. Another equation derived with the
storage equation describes growth and attrition (i.e.,
metabolic and mortality costs) in terms of energy.

Further, consumers may prey upon each other in addition
to the resources and the equations are shown to be adequate
for describing food webs. The model is essentially one of
energy flow and. deals with biomass, instead of numerical,
dynamics.

Coded in a computer language (PL/I), the model simulates
competition and clarifies aspects of yield and harvest
strategies in systems of interacting populations.

The model is illustrated with hypothetical populations
which have variable turnover rates and the Lake-Trout-.Burbot
system in Lake Opeongo, Ontario.

Comment

This model may be applicable for competitive species
and predator-prey systems, although it is'useful only
in density -dependent situations. A number of the model
-parameters may be difficult to obtain.

A-73

Hammond, D.E., and T. Lackey. 1976. Analysis Non-applicablo
of catchable trout fisheries management by computer
simulations. Trans. Am. Fish. Soc. 105:48-56.

Abstract

Although strategies to meet most management objectives
are relatively clearcut in single-species, catchable trout
programs, strategies become much more complex when two or
more species are involved. A difficult problem that must be
faced in evaluating catchable trout fisheries management
strategies is defining management objectives. One approach
to testing alternative management strategies in complex
resource systems, such as catchable trout fisheries, is
system simulation. A computer-implemented CAtchable Trout
Fishery Simulator (CATS) was developed to 6v-aluate fi-shery
response7under various management strategies in a multispecies
stocking program. The user of CATS can select alternative
management strategies and functions which generate predictions
of fishing pressure in a particular fishery. To evaluate the
effect of each system component, CATS was exercised over a
wide range of potential, athough entirely hypothetical, system
component alternations. Predominant stocking of brook trout
appreciably increased average catch per angler hour and
percentage return to creel. Altering the stocking ratio to
favor brown trout substantially increased the number of
angler hours. Stocking predominantly rainbow trout produced
results intermediate between those caused by stocking
predominantly brook or brown trout. Estimates of expected
angling pressure.and catchability coefficients of each
species stocked are of primary importance because of their
considerable effect on other system components. A user must
have a sound objective befoKe deciding where, when, which
species, and how many fish to plant. The primary utility of
CATS is to enable the user to evaluate management strategies
prior to implementation.

Coment

The model is not really usable. The methodology presented
is not quantified beyond a functional stage. All functional
parameters are essentially left to the wishes of the user.

A-74

Hashagen, K.A., Jr. 1973. Population structure changes and Non-applicable
yields of fishes during the initial eight years of
impoundment of a warmwater reservoir. Calif. Fish..
and Game 59:221-244.

Abstract

Data frcm creel census and netting programs were used
to study changes in the relative abundance, age, and species
composition of fish during the first 8 years of impoundment
of Merle Collins Reservoir, Yuba County, California. The
effects of the changes on fish yields are discussed.
High survival of the first (1964) year class of large-
mouth bass Oficropterus salmoides) produced a large popula-
tion of slow growing bass that-do-minated the fishery through
1967. These fish limited survival and recruitment of all
centrarchids in 1965 and 1966 and depressed initial yields.
Largemouth bass and green sunfish CLepomis cyanellus)
declined numerically after 1968. Bluegill (L. macrochirus)
and redear sunfish (L. microlophus)increased dramatically.
Catfish remained stable and were probably underexploited.
Salmonids., introduced in late 1966,increased total pressure
and effort and raised annual yields.
Age composition of centrachids changed from adult fish
and fish of the initial year class to a structure in which
several year classes and all sizes were represented by the
end of the 8-year study.
Anticipated chaftges in species composition failed to
develop as numbers of,nongame species remained extremely
low throughout the study.
High initial yields customarily associated with new
inpoundments were not obtained. Yields ranged from 2.3
to 10.7 lb/acre.

Comment

This paper is not applicable to the project as it
simply describes historical alterations in total yield
from the reservoir. No models of production or yield
are presented.

A-75

Hilborn, R. 1976. Optimal exploitation of multiple stocks Applicable
by a common fishery: A new methodology. J. Fish. Res.
Bd. Canada 33:1@5.

Abstract

Optimal harvest rates for mixed stocks of fish are
calculated using stochastic dynamic programming. This
technique is shown to be superior to the best methods
currently described in the literature. The Ricker stock
recruitment curve is assumed for two stocks harvested by the
same fishery. The optimal harvest rates are calculated as a
function of the size of each stock, for a series of possible
parameter values. The dynamic programming solution is similar
to the fixed escapement policy only when the two stocks have
similar Ricker parameters, or when the two stocks are of
equal size. Normally, one should harvest harder than calculated
from fixed escapement analysis.

Comment

The procedure presented here may be usable in situations
where population dynamics are modeled adequately, as a method
of determining multi-stock relationships and for the optimization
of recruitment for several species.

A-76

Horst, T. J. 1977. Use of the Leslie matrix for Applicable
assessing environmental impact with an example for a
fish population. Trans. Am. Fish. Soc. 106:253-257.

Abstract

The Leslie matrix model for discrete population theory
is examined for the assessment of the effects of environmental
alterations on a species population using an eigenvalue
analysis. This analysis provides estimates of population
growth rate and stable age distribution. A sensitivity
analysis is conducted for changes in elements of the population
matrix and the resultant effect on population growth rate
and stable age distribution. An example of this technique
is presented for the cunner (Tautogolabrus adspersus) . This
example considers the effecf-of entrainment of cunner eggs
and larvae at the intakes of power stations.

Comment

The procedure presented here may be applicable if
environmental impact can be considered equivalent to fishing
mortality.

A-77

Hrbacek, J. 1969. Relations between some enviromental
parameters and the fish yield as a basis for a
predictive model. Internationale Vereinigung fur
Theoretische und Angewandte Limnologic Verhandlungen
17:1069-1081.

Abstract

Comment

This paper is unavailable for review but has been
requested through interlibrary loans. It will be reviewed
when received.

A-78

lt=g, C.C., I.B. Vertinsky, and N.J. Wilimovsky. 1976. Applicable
Optimal controls for a single species fishery and the
economic value of research. J. Fish. Res. Bd. Canada
33:793-809.

Abstract

Mathematical proofs and analyses of solution methods are
presented for determining optimal policies for the management
of a single species fishery under equilibrium conditions.
Previous intuitive arguments for solution of optimal policies
controlling mesh size and fishing rate given complete information
are explicitly proven. The analysis is extended to the case
where some of the parameters describing the dynamics of the
population are known only imprecisely to the manager. Using
probability distributions for those unknown parameter values
the problem is cast as a stochastic program where expected
sustained net revenues from the fishery are Tm=ized.
The associated problem of optimal allocation of research
resources under uncertainty conditions is considered by
evaluating the direct value of such information to management
activities.

Examples and algorithms are presented for the class of
problems discussed.

Comment

The model presented here may be useful, but only after a
population dynamics model is completed. It develops a
methodology to determine optimal age of capture and its
associated yield (i.e., by regulation of mesh size).
optimal yield by controlling f ishing rate, or a combination
of the two.

A-79

Jacquette, D.L. 1972. A discrete time population control Non-applicable
model. Math. Biosci. 15:231-252.

Abstract

A discrete time stochastic model is assumed to describe
the growth behavior of a natural animal, pest, or even epi-
demic population. Periodically, control action representing,
for example, harvesting or exterminating can be taken to
modify the future gr(xvth of the population. Dynamic program-
ming can be used to generate optimal control policies for
models where growth and control produce economically measur-
able benefit or cost. This paper uses dynamic programming
to find conditions under which the classically simple "single
critical number policy" is optimal. These conditions are
then expanded to include a wide range of control models and
the optimal policyobtained is characterized as a "single
critical function policy."

Conmnt

.This paper presents an optimal control stochastic
model of population dynamics. It is very theoretical
in nature and its parameters are not biologically rea-
listic. This model approach is beyond the scope of
this study (optimization) and it is somewhat doubtful
that it can be applied to Maryland fisheries at all.

A- 80

Jenkins, R.M. 1976. Prediction of fish production in Applicable
Oklahoma reservoirs on the basis of environmental
variables. Ann. Okla. Acad. Sci. 5:11-20.

Abstract

To maintain productivity of fishery resources in
Oklahoma reservoirs, the managers of these reservoirs
must know the environmental factors that control fish
production and harvest. The present study, using corre-
lation and regression analysis, indicates that of the
ten environmental variables examined, low storage ratios,
i.e., high water-exchahge rates, most favor high standing
crops of fish. Revised estimates of standing crop values,
made in part by correcting for the difference in recover-
ies from coves and from a larger area, indicate much
larger and dominant populations of bottom feeders. The
potential of various reservoirs for production of sport
fish is evaluated on the basis of water quality, and
fishing pressures on the reservoirs are estimated.

Comment

While this paper is not directly related to Maryland
fisheries, it may prove helpful in the adaptation of a
morphoedaphic index (MEI) to estuarine situations. The
relationship used concerns standing crop and discharge
or storage rates rather than total dissolved solids. Since
earlier versions of the MEI assume negligible flushing,
and this approach assumes variable flushing (more realis-
tic when considering estuaries), the latter approach may
be applicable for producing statistical management models
for Maryland fisheries.-

A-81

Jenkins, R.1'4. 1978. Prediction of fish biomass, harvest Non-applicable
and prey-predator relations in reservoirs. pp. 282-
296 In W. Van Winkle (ed.)., Assessiu the Effects
of Poi%@e_-F-Plant-Induced Mortali on Fish Populatio'ns.
Sponsored by Oak Ridge National Laboratory, Energy
and Development Administration and Electric Power
Research Institute.

Abstract

Regression analyses of the effect of total dissolved
solids on fish standing crops in 166 reservoirs produced
formulas with coefficients of determination of 0.63 to
0.81. These formulas provide indexes to average biotic
conditions and help to identify stressed aquatic environments.
Simple predictive formulas are also presented for clupeid
crops in various reservoir types, as clupeids are the
fishes most frequently impinged or entrained at southern
power plants. A method of calculating the adequacy of the
available prey crop in relation to the predator crop is
advanced to further aid in identification of perturbed prey
populations. Assessment of stress as reflected by changes in
sport fishing success can also be approached by comparison of
the predicted harvest potential with actual fish harvest data.
Use of these predictive indexes is recommended until more
elaborate models are developed to identify power plant effects.

Comment

The procedures presented here are not useful in the
sense of a prediction of yield. They would be of little
use in the project.

A-82

Jensen, A.L. 1972. Population biomass, number of individuals, Applicable
average individual weight, and the linear surplus-
production model. J. Fish. Res. Bd. Canada 29:
1651-1655.

Abstract
The' identity Bt = Nt1Rt , where Bt is the population
biomass, Nt is the population size, and @Vt is the
average individual weight for all ages, is applied to
develop simultaneous equations for change in biomass,
number of individuals, and average individual weight for
the linear surplus-production equation. It is shown that
equations for all three variables cannot be simultaneously
logistic. The relation between logeNt and logeWt predicted
by the linear surplus-production model is compared with
observations of bluegill population densities and average
weights estimated from 10 years of cove rotenone sampling in
five large TVA reservoirs. The fit of the model to the data
is fairly good, but it accounts for only a small amount of
the total variation observed.

Comment

This paper presents a good discussion of the inter-
relationships of numbers and biomass in surplus-production
models. However, the material presented is not significantly
different from a Schaefer model.

A-83

Jensen, A.L. 1973- Relation between simple dynamic pool Non-applicab
and surplus production models for yield from a fishery.
J. Fish. Res. Bd. Canada 30:998-1002.

Abstract

Dynamic pool models without self-regenerating properties
are continuous age models, and surplus production models
are continuous time models. Self-regenerating dynamic pool
models are continuous age-discrete generation models and,
also, discrete time-discrete age models. In a steady state,
specification of the regulatory function and direct estimation
of biomass results [sic] in the surplus production model.
Estimation of biomass by specifying the functions with respect
to age for si.ze of a cohort and individual weight and applica-
tion of the coefficient of fishing mortality result in the
dynamic pool model. A third approach, not applied in
fisheries, is to specify the regulatory function and functions
with respect to age of cohort size and individual growth
in weight. In a steady state, all methods for calculating
yield give the same results if the functions specified are
realistic. Specification of the functions requires that
many assumptions be made. The dynamic pool model may be
more accurate than the*surp1us production model because'the
regulatory function may be more difficult to determine than
the functions with respect to age of cohort size and growth
in individual weight.

Comment

This paper presents a good discussion of similarities
of Beverton-Holt and Schaefer-like models., but presents
no material significantly different from those models.

A-84

Applicable for
Jensen, A.L. 1976. Relation between yield and fishing parameter
mortality for Ricker's yield equation. J. Fish. estimation
Res. Bd. Canada 33:275-277.

Abstract

A method is presented that can be applied to examine
the relation between yield and magnitude of fishing mortality
for Ricker's dynamic pool yield equation. The magnitude
of fishing mortality is measured by the norm of the vector
of interval specific instantaneous fishing mortality
coefficients. The method is applied to a bluegill sunfish
(Lepomi macrochirus) population.

Comment

The paper is interesting, but not directly applicable to
yield estimation. It sinply discusses the relationship of total
mortality to yield in a cohort model.

A-85

Johnson, ]_,'. C. 1978. Salmon Fislieries Systems Applicable
Ana.LYSLS. Wasliiiigton Department of Fisheries
Completion Report for Project No. 1-99-D, for
the Period July 1973 - September 1976. 81 pp.

Abstract

The Catch/Regulation Analysis Model (the fishery model)
has been developed by the National Bureau of Standards
for the Washington Department of Fisheries to serve as
a salmon fisheries management tool for analyzing the
economic and biological effects associated with changes
in salmon fishery regulations. The goal of the model
is to provide a common methodology for quantifying the
impact of alternate sets of fishing regulations on
fisheries performance and salmon stock abundance and
stability. It has been used to examine and evaluate over
a total of 100 different sets of fishing regulations for
the 1976, 1977 and 1978 salmon fishing seasons. Personnel
.in the Washington Department of Fisheries have been trained
in data preparation for the model and in the mechanics
of running the model on the Control Data computer
system at the University of Washington, Seattle,
Washington. The purpose of this document is to provide
a non-technical description of the model and to provide
realistic input and output examples. The fishery model
is a large and complex model which contains over 6000
FORTRAN statements.

.Comment

This methodolo gy appears useful when lar data sets
_:@a e
are available. It may not be directly appllMa le to any
I'viaryland species. The actual model is delineated in
an earlier paper and it is impossible to evaluate
assumptions without the earlier paper.

A-86

Jones, R. 1964. Estimating population size froni commercial
statistics when fishing mortality varies with age.
P.%V. Reun_,_, Cons. Int. Explor. Mer. 55:210-214.

Abstract

Comment
Thi@ paper is unavailable for review but has been
requeste@ through interlibrary loans. It will be reviewed
when received.

A-87

Jorgensen, S.E. 1976. A wdel.of fish growth. Ecol. Non-applicable
Modellin 2:303-313.

Abstract

A fish model based upon mass balances is set up.
Several of the parameters have been determined by
experiment. The remaining parameters are based upon
literature values and it was not necessary to find any
parameter by calibration. The model was validated
with a completely acceptable result.

The model includes equations giving the fish growth,
the production of ammonia and organic matter, the
oxygen consumption, and the influence of the temperature.

The model can be used for management of fish farms
and as a submodel for a total aquatic system.

Comment

This model is a simplistic formulation, which is
applicable to very controlled situations. It appears
unusable in a non-stable, natural environment.

A-88

Kitchell, J.F., J.F. Koonce, R.V. O'Neill, H.H. Shugart, Jr.
J.J. Magnuson, and R.S. Booth. 1974. Model of fish
biomass dynamics. Trans. Am. Fish. Soc. 103:786---77F.

.Abstract

A model of fish biomass dynamics is developed based
on principles of physiology, popluation biology and
trophic ecology. The model is designed to incorporate
measurable parameters for simulating seasonal changes in
a natural population. All parameters were implemented for
the bluegill (Lepomis macrochirus) and simulation results
compared with UZ 5-end-en-t-1y--Je-r1ved laboratory and field
data. Applications of the model to thermal enrichment
problems are given as an example of future potential use.

Conment

This model represents a synthesis of many different
factors which influence fish production. Its complexity
may limit its applicability, but it should be considered
for use in this program.

A-89

Kitchell, J.F., D.J. Stewart, and D. Weininger. 1977. Applicable for
Applications of a bioenergetics model to yellow parameter estimation
perch (Perca flavescens) and walleye (Stizostedion
vitreum-NFi-treum). J_ Fish. Res. Bd.- Canada
Tg=-1939.

Abstract

A simple energy budget equation is developed to
C>
yield a bioenergetics model designed to simulate fish
growth. Parameters for the model are estimated from the
literature for application to yellow perch (Perca
flavescens) and walleye (Stizostedion vitreum vitreum
lations are presented that demonstrate model output
as functions of body size, activity level, ration level,
food quality and environmental temperature. Sensitivity
analyses identify the importance of food consumption,
activity and excretion as biological processes represented
in the parameters. On the basis of temperature
conditions in selected lakes and specified feeding levels,
Che importance
simulations are presented to quantify 4
of year-to-year variation.of temperature in determining
growth. In heterothermal systems, temperature selection
by percids can have a significant effect on growth. For
walleye on fixed rations, annual growth can vary from
zero to twofold increments due entirely to differences in
summer temperatures. Variations in food quality have
lesser effects..

"lomment

The models presented require detailed information which
may be unavailable for many species. However, it should be
considered as a possible input model. Sensitivity analysis
is included.

A-90

Knight, W. 1970. A possible optimization experiment Non-applicable
in fishery management. J. Fish. Res. Bd. Canada
27:961-963.

Abstract

A tag-recapture method for deciding whether a
change in effort governing regulations will improve
a fishery or the opposite is proposed. When the
criterion is weight of catch, this reduces to seeing
whether the total weight of fish recaptured in a
certain tag-recapture experiment exceeds the total
weight of fish released. Most of the many practical
objections besetting this experiment are common to
any method depending on tag-recapture methods to
estimate rate of exploitation.

Comment

This paper does not present a model of yield
or production and is thus not suitable for use in
this study.

A-91

Kutty, M.K., and S.Z. Qasim. 1968. The estimation of Applicable
optimum age of exploitation and potential yield
in fish populations. J. Cons. , Cons. Int. LXILI-0-r.
Mer. 32:249-255.

Abstract

Some methods for estimating the optimum age of
exploitation (ty) and the potential yield (Y) in fish
populations havi@ been.described under allometric and
isometric conditions of growth. The relative constancy
of ty and Y under all conditions of exploitation indi-
cates that these parameters, in the absence of a long
series of data, could be used for measuring the fishing
efficiency.

Comment

This paper presents a review of the Beverton-Holt
yield model under the conditions of allometric and iso-
metric growth. I%Ihile offering little new information,
this paper could prove valuable in a management frame-
work for the determination of the optimal acre of exploi-
tation.

A-92

Lackey, R.T. 1975. Fisheries and ecological models in Non-applicable
fisheries resource management. pp. 241-249 In
C.S. Russel (ed.) Ecoloaical Modelina,in a R.-e-source
Management FramewoITY,- R sources tor the Future,
Inc., Washington, D.C.

Abstract

The author reviews the use of fisheries and ecological
models in fisheries management. fie discusses the basic
types of models (i.e., habitat, biological and social) and
how they have been integrated to generate the types of
models in general use. He also discusses problems
encountered in applying models to decision-making
processes.

Comment

This paper presents a good, brief background
discussion of the types of models used in fisheries
management. However, it does not present actual models
and is thus not useful for this program.

A-93

Laevastu, T., and F. Favorite. 1978. Numerical evaluation Applicable
of marine ecosystems. Part I. Deterministic bulk
biomass model (BBM). Northwest and Alaska Fish. Center
Proc. Rept. NIOAA.

Abstract

The wise management of marine fishery resources requires
the consideration of the total marine ecosystem in a given
region, as the components are interacting and the fishery on
one or more species will, in many instances affect the abun-
dance and distribution of other species as well, i.e., the
abundance of one species might be declining, the other
increasing.
A deterministic method to estimate either the minimum
sustainable and/or saturation biomasses of species and/or
groups of ecologically similar species in a given -region is
described. This method can also be applied to unexploited
or underexploited resources. Trophodynamic approaches are
used that permit the quantitative determination of the major
component of the natural mortality -- the predation or graz-
ing mortality. The error limits depend on the accuracy of
some trophodynamic data. Migration is not considered
because detailed migrations are an integral part of the
large ecosystem model -- DYNUMES, for which the present
model serves as one of the initial input subroutines.

Comment

While appealing, the ecosystem model for determining
system carrying capacity (several species or species groups)
cannot be utilized for Maryland fisheries without numerous
studies to evaluate the feeding coefficients necessary for
the model's implementation. The model, assuming the data
are available, provides a good estimate of trophic carrying
capacity and maximum sustained yields.

A-94

Laevastu, T., and F. Favorite. 1978. Numerical evaluation Applicable
of marine ecosystems. Part 2. Dynamical numerical
marine ecosystem model (DYNUMES III) for evaluation of
fishery resources. Northwest and Alaska Fish. Center
Proc. Rept. NOAA.

Abstract

A method of numerical dynamical deterministic reproduc-
tion of a marine ecosystem with emphasis on applications to
fisheries problems is presented. Although such simulations
are location dependent -- i.e., greatly determined by the
nature of the specific ecosystem components present in the
region and by the availability of local research and survey
results as well as by the intensity and nature of the exploi-
tation of marine resources by man in the given region -- the
method gives a general basic framework of one type of marine
ecosystem simulation model. Objectives, principles, and
major computational formulas of the model, which has been
applied by NWAFC to the eastern Bering Sea and the area
around Kodiak Island in the Gulf of Alaska, are also pre-
sented. More detailed discussions on the input of local
knowledge, other specific data and their validity, results
of model applications, and the computer program documentation
are available in NWAFC Processed Reports.

Comments

The model present in this report is very complex
with state variables varying in 4 dimensions (3 spatial
and 1 time). While the approach is ecologically satisfy-
ing (i.e., approaches a realistic biological situation),
the data requirements, particularly for ecosystem processes,
are too large and difficult to obtain to be of much use to
Maryland fisheries. This procedure may be instructive in
the formulation of data acquisition schemes to obtain this
type of information for potential ecosystem management
models at a later date.

A-95

Laevastu, T., F. Favorite, and H.A. Larkins. 1979. Applicable
Resource assessment and evaluation of the dynamics
of the fishery resources in the NE Pacific with
numerical ecosystem models. Northwest and Alaska
Fish. Center Proc.. Rept. 79-17.

Abstract

The development of offshore fisheries in the vast
NE Pacific is relatively recent. Exploratory surveys in
the late 1950's demonstrated the abundance of groundfish
in this region. The resource surveys .in the large area
are expensive and without an inordinantly large field
effort the accuracy of the -results is low. Due both to
the lack of conventional fisheries and biological data
and the inherent shortcomings of single species models,
those models are of questionable value for managing the
multination, multispecies fisheries of the northeastern
Pacific Ocean.
Two biomass-based, holistic, ecosystem models are
being developed and are used at NIVAFC for the evaluation
of the abundance and dynamics of the fishery resources
and for the study of the response of these resources to
exploitation and to environmental changes (anomalies).
The general background of these models is given...
and a simplified version of one of them is given in skele-
ton form in the Appendix..
Equilibrium biomasses, as computed with the PROBUB
model for the eastern Bering Sea and the western Gulf of
Alaska, are given in this paper. These are computed
(validated) with conventional trawling survey results,
which have been converted with catchability and availabil-
ity factors.
The nature of the dynamics of the biomasses in space
and time is briefly described and the effect on the
resource assessment is demonstrated with some results
from the DYNIUMES model,

Comment

This paper presents several ecosystem simulation
models for the Pacific Northwest fisheries. Single
population models are inapplicable in these cases due to
the high degree of species interaction caused by preda-
tion. The large number of parameters necessary to con-
struct this model makes its application difficult to
IMaryland fisheries, but this type of simulation formu-
lation is the most ecologically complete type of manage-
7wnt model described in the literature.

A-96

Larkin, P.A. 1963. Interspecific competition and exploita- Non-applicable
tion. J. Fish. Res. Bd. Canada 20:647-678.

Abstract

The consequences of exploitation of either or both of a
pair of competing species are examined using the Lotka-Vol-
terra equations. The removal of a fixed proportion of a
population on an instantaneous basis shifts the equilibrium
population sizes for both the exploited species and its
competitor. Similar shifts occur when both species are
exploited. The maximum sustained yield of a species can be
estimated under various degrees of exploitation of its com-
petitor. The maximum combined sustained yield can be esti-
mated for various relative values of the two species. From
this analysis it is observed (1) harvesting only one species
may provide a mistaken underestimate of capacity for sus-
tained yield, (2) harvesting two species but relating yield
to the fishing mortality rate of only one of the two may
give a misleading overestimate of further capacity for
sustained yield. Similar conclusions can be drawn if
exploitation rate is proportional to abundance.
A stochastic version of the model is given for study
of the effects of exploitation on small populations of
competitors.
Fixed percentage exploitation and abundance propor-
tional exploitation may be considered in depicting respec-
tively the mode of action of density-independent and density-
dependent factors. Accepting these parallels, the model may
demonstrate some widely discussed properties of mechanisms
of population regulation. Variability in factors both
density dependent and density independent which are extrin-
sic to the bio'logical system can be simulated in the model
by random variates.
A discrete time model is described which was used with
a computer for study of transitions from one steady state
to another and extinction probabilities. The computer
results confirm the theoretical predictions of the model.
In addition it is suggested that there is no apparent dif-
ference in the result when competitors are exposed to the
same or different random sequences of environmental effects
of the same average intensity.
It is concluded that this formulation of interspecific
competition together with variations could be applied to
laboratory or natural situations to test its usefulness as
a basis for prediction.

Comment

This paper represents a novel inclusion of competition
theory in yield exploitation but its data and parameter
requirements are too large to merit its possible use for
Maryland fisheries.

A-97

Larkin, P.A., and A.S. Hourston. 1964. A model for simulation Non-applicable
of the population biology of Pacific salmon. J. Fish. Res.
Bd. Canada 21:1245-1265.

Abstract

A model simulating the population dynamics of the stocks
of salmon spawning in a large river system is constructed,
incorporating (1) a series of theoretical reproduction curves
operating in successive stages during the life history of
each stock producing compensatory or depensatory effects,
(2) a device for simulating environmental variability (extra-
pensatory effects)', (3) the effects of a joint fishery on
mixed stocks, (4) the effects of multiple age of spawning
on fluctuation in abundance of a single stock, (5) the con-
sequences to yield of various degrees of stabilization of
the fishery, and (6) the inheritance of age at maturity in
salmon. Application of the model is illustrated by two
examples. The first simulates the production of cyclic
dominance by the inheritance of age of maturity and depens-
ation in the freshwater stage. In the second example, five
stocks differing in relative abundance, age composition,
environmental variability, and rate of compensation are
subjected to a common fishery which is selective for age
and provides for complete stabilization of the escapement
for the combined run as a whole. Comparisons between
various combinations of the five stocks show the effects
of these factors on the various stocks and on the run as
a whole over a similated period of,120 years.

Coment

This model is probably too data intensive and site-/
species -specific to be useful in this project.

A-98

Laurence, G.C. 1977. A bioenergetic model for the Non-applicable
analysis of feeding and survival potential of
winter flounder, Pseud22leuronectes americanus,
larvae during the period from hatchiRg to meta-
morphosis. Fish. Bull. 75:529-546.

Abstract

A bioenergetic model was developed which simulated
effects of temperature, prey density, and larval size
on ability of winter flounder, Pseudopleuronectes'
americanus, larvae to obtain food energy to provide for
experimentally determined growth and metabolism. Larval
feeding at constant temperature and as a function of
prey concentration was exponential and more sharply
asymptotic in younger fish than in those near metamorphosis.
Specific growth rates were exponentially related to prey
concentrations and ranged from 5.72 to 8.700-,/day at
survival prey concentrations of 2.3 to 21.7 cal/liter.
Daily required feeding time was directly related-to prey
availability. Critical plankton densities below which
larvae did not have enough time during the day to obtain
adequate food for growth and metabolism varied with age
and ranged from 2.1 to 5.7 cal/liter. Simulated
physiological energy utilization and required caloric
food intake were inversely related to prey concentration
and varied with larval stage of development. Food
requirements expressed as numbers of copepod nauplii
consumed per day ranged from 19 for first feeding larvae
to 235 for metamorphosed juveniles. Predicted gross
growth efficiencies were directly related to prey
concentration and increased with age from 5 to 33%. All
indications pointed to a "critical period" of larval
survival during the period of exogenous feeding initiation
and immediately after.

Comment

The model presented here is a standard bioenergetics
model with no real modifications. Its application is to
larval growth, which is beyond the scope of this project.
Because it is similar to other models to be considered
for parameter estimation, it should not be reviewed further.

A-99

Lei-kovitch, L.P. 196S. The study of population growth, Applicable
in organisms grouped by stages. Biometrics 21:1-18.

Abstract

In this extension to the use of matrices in population
mathematics, the division of a population into equal age groups
is replaced by one of unequal stage groups, no assumptions being
made about the variation of the duration of the stage that
different individuals may show. This extension has application
in ecological studies where the age of an individual is
rarely known. The model is briefly applied to three
experimental situations.

Comment

The paper presents a modified Leslie matrix model.
It might be a useful modification for application to
exploited stocks.

A-100

Lesliel P.H. 194S. On the use of matrices in certain popu- Applicable
lation mathematics. Biometrika 33:183-212.

Abstract

The author describes the application of matrices to
analysis of population dynamics, using estimates of age-
specific fecundity and mortality as inputs and changes
in age distribution over time as output.

Comment

This paper is one of the classic papers of Leslie
describing the application of matrices to the study of
animal populations. It will be useful for evaluating
the appropriateness of this approach to fisheries manage-
ment.

A-101

Leslie, P.H. 1959. The properties of a certain lag type Applicable
of population growth and the influence of an external
random factor on a number of such populations.
Physiol; Zool. 32:151-159.

Abstract

The author considers the growth in numbers of a
population of a single species, living in a limited
enviroment,'in which it is assumed that the fertility and
mortality of each age-group depend not only on the existing
state of the population at some given time t but
also on the state of the population at some-previous time
t - x, when the individuals forming a particular age-group.,
iged-between x and x + A-x were born. This is partly
a "lag" form _(@f popu-lati5n growth in which the period
of lag varies from age-group to age-group depending on
the length of time the individuals have lived in the
population. It is shown that this type of population,
living under optimum constant conditions as regards the
environment, would oscillate in numbers, with amplitude
of the oscillations, which initially may be.quite marked,,-
gradually damping out as an approach is made to the
stationary state in numbers.

A further problem which is considered is the
influence of an external seasonal factor on a number of
oscillating populations of this type, which are situated
in the same geographic region. This common factor is
here regarded as being a hierarchy of five possible
classes of "season, ranging from "very good" to
"very bad,"ahich are assumed to follow one another in a
random order. It appears that the effect of an external
random factor of this type, acting independently of
age and numbers, is to bring the oscillations of
such geographically isolated systems gradually into
phase.

Comment

This is one of the original presentations of the Leslie
matrix population model, which is applied to fish populations
in more recent literature. It should be included in this study,
grouped with the application papers.

A-102

Lett, P.F., W.T. Stobo, and W.G. Doubleday. 1975. A Applicable
system simulation of the Atlantic mackerel fishery
in ILNAF subareas 3, 4, and 5 and Statistical
Area 6; with special reference to stock management.
Int. Comm. Northwest Atl. Fish Res. Doc. 75/32.
io pp.

Abstract

The international commercial catch of Atlantic
mackerel (Scomber scombrus) from ICNAF Subareas 3, 41
and 5 and -9-tatistical Area 6 increased from 6,831m.t.
in 1961 to 419,306m.t. in 1973 (Anderson IMS 1975a).
The estimated value of F'in 1973 was 0.60; the
recommended level by the ICNAF ad hoc mackerel working
group (Redbook 1973, Part 1).
A stock recruitment relationship was derived for
Atlantic mackerel which spawn in the Gulf of St. Law-
rence (Lett et al. @B 1975b). Using the determined
relationships, a system simulation was constructed to
investiaate the effects of different levels of fishing
intensity on stock biomass, catch and recruitment.
These effects are examined when temperature is held
constant and treated as a stochastic variable.
It is anticipated that the validity of maximum
sustainable yield being determined at a fishing mor-
tality of 0.6 could also be investigated. Furthermore,
a test is made of Ricker's (1963) hypothesis that small
increases in effort beyond maximum sustainable yield
cause rapid declines in stock biomass, recruitment,
and catch.

comment

This paper presents a good methodology for testing
the consequences of variable fishing pressures, using
a yield-per-recruit Beverton-Holt formulation coupled
with statistical models of egg production, and larval and
prerecruit survival. This model may be helpful in
constructing management models for Maryland fisheries,

A-103

Lett, P.F., and T. Benjaminsen. 1977. A stochastic Non-applicable
.model for the manaaement of the northwestern
Z@'
Atlantic harp seal (PaRophilus uoenlaadicus
population. J. Fish. Res. Bd. Canada 34-1155-1187.

Abstract

Advice from the scientific advisers under the
auspices of IMAF to the international commissioners
for 1977 was that the total allowable catch (TAC) for
harp seals (Pagophilus groenlandicus should not
exceed 170,000. This advice, in part, was based on
the scientific arguments presented in this paper.
A stochastic model is developed that takes into account
the variations in natural mortality and the landsmen's
high arctic and Greenland catches. The Canadian-
Norwegian large vessel hunt is controlled under
quota regulations. The model is nonlinear, a result
of changes in fertility and fecundity rates in response
to shifts in population size. The maximum sustainable
yield (IVISY). I+ population size is determined to be 1.6
million seals, or a breeding stock size of 375,000 seals.
The MSY is approximately 240,000 seals, assuming the
hunt continues its present pattern. The 240,000 can
further be split into 200,000 pups and 40,000 1+
seals. Present stock size is approximately 1.2
million and a TAC of 170,000 seals will allow the
population size to reach to MSY level in 10-15 years.
A number of other management strategies are considered,
in addition to prospects for further research.

Comment

The paper presents a simulation of a harp seal
fishery based on cohort analysis. Techniques presented
here are covered in other papers rejiewed, and other
specific information presented appears irrelevant to this
study.

A-104

Lett, P.F., A.C. Kohler, and D.N. Fitzgerald. 1975. Applicable
Role of stock biomass and temperature in
recruitment of southern Gulf of St. Lawrence
Atlantic cod, Gadus morhua. J. Fish. Res.
Bd. Canada 32:1613-1627.

Abstract

A multivariate approach was used to elucidate the
simultaneous effects of temperature and estimated parent
stock biomass on the recruitment mechanism of Gulf of
St. Lawrence cod., The second order effects of temperature
and estimated stock biomass were key factors in deter-
mining eg abundance levels. In addition, egg abundance
C-9 It,
was closely related to the growth rate of cod.. The
numbers of larvae increased with the interaction of
temperature with egg abundance but decreased with the
interaction of egg abundance and time. The most
important step in the recruitment- mechanism occurs
during the juvenile state, the degree of density
dependence being reliant on total biomass of the adult
cod stock. A system simulation was constructed
amalgamating the equations of early life history of
cod with the effects of exploitation on stock biomass.
Regular 12-yr oscillations were demonstrated at low
levels of catch, while the population became more stable
at higher fishing efforts in the absence of environmental
effects. The optimal fishing mortality for the Gulf of
St. Lawrence cod was found to be.F 0 4with a maximum
sustainable yield. of 42,000 metric * tons.

Comment

This paper presents a very detailed, deterministic
population model which requres an extensive data base
for application. It represents an advanced management
tool.

A-105

Lewis, E.R. 1976.. Applications of discrete and continuous Applicable
network theory to linear population models. Ecolo
57:33-47.

Abstract

The well-established methods of network construction and
analysis are adapted to the problem of modeling single popu-
lations. A major advantag
.,e of the resulting approach is that
it allows explicit incorporation of key processes in the life
cycle of the organism being modeled, with feedback loops pro'-
viding economy of representation where they are allowed. Thus,
network structures provide heuristic vehicles by which popula-
tion models can be developed and modified. When a model is
linear and has parameters that do not vary with time, a char-
acteristic dynamic function can be derived by inspection from
a simple transform of the network representation. The zeros
of the function can be found (analytically or by commonly
available numerical methods) and used directly to deduce the
modeled population's dominant growth pattern and its propensity
to sustain oscillations. In addition, under certain conditions
(i.e., that the network model not contain both time delays
and integrators), a straightforward method (partial fraction
expansion) is available for deduction of the modeled popula-
tion's specific responses to a variety of perturbations.

Comment

This approach offers a secondary method of linear time
scaling through networks where events define the boundaries
of stages rather than age as in the Leslie matrix. This
method offers little new and will probably prove to be of
little use in the project, but it should be reviewed.

A-106

Lewis, E.R. 1977. Linear population models with stochastic Non-applicable
time delays. Ecology 58:738-749.

Abstract

Previously it was shown that reproductive-cycle parameters
such as time to maturity, ovulation interval, gestation period,
duration of regression, duration of nonreproductive lactation
period, and the like, can be incorporated into population models
rather easily through the use of a simple network approach. In
this paper, the network approach is extended to include the same
types of reproductive parameters when their values are not
necessarily fixed, but may vary randomly from one member of a
population to the next and/or for a given member from one time
to the next. It is shown that linear transforms of the para-
meter distribution functions can be incorporated directly into
the network models and that analysis of the resulting dynamics
follows in a straightforward manner, the characteristic dynamical
equation being obtainable by direct analysis in simple cases or
by well-established numerical methods in complicated cases.
1he roots themselves can be interpreted directly in terms of
dominant patterns of population growth and deduced propensity
of the population to sustain oscillations triacrered by exter-
nal stimuli. In the case of a simple natality cycle with
gamma, negative binomial, and binomial distributions of matur-
ation times, it is shown that the dominant growth pattern
approximates rather closely that expected for a nonrandom
maturation time equal to the mean of the distribution, and
that the propensity to.sustain population oscillations de-
creases markedly both with increasing standard deviation
and with increasing (positive) skewness in the distribution.

Comment

This paper offers very little that would be useful in
this project. The concept of stochastic event-initialized
time will probably not be appropriate for yield conceptu-
alizations.

A-107

Lipschultz, F., and G. Krantz. 1978. An analysis of oyster Non-applicable
hatchery production of cultched and cultchless oysters
utilizing linear programming techniques. Proc. Nat..
Shellfish. Assoc. 68:5-10.

Abstract
and Manpower and operational requirements for cuitched
cultchless oyster production schedules in a large-
scale hatchery were compared using a system of linear
equations and a computer optimization program. The
matrix of equations defined the operational sequence
within the oyster hatchery, the resource requirements,
and restrictions., if any. The optimization program
minimized a cost objective function within the constraints
defined by the system matrix.

Data for the calculations of the matrix coefficients
wern tnkpn from records of the University of Marylana
small-scale hatchery and from Dupuy (1973). The hatchery
data provided estimates of manpower requirements for each
activity, oyster mortality rates, and equipment costs.
Dupuy's paper provided space and density requirements.
The temporal sequence of oyster development stages was
based on well-documented literature and observations
at the model hatchery.

Results show that labor was the major cost component
in all types of hatchery schedules. The optimal solution
involved purchase of larcre amounts of equipme@it which
remained idle most of the year, being fully utilized
in two pulses during the year. Constant maximal use
of equipment required less equipment but more labor and
therefore increased production costs. Use bf the cultched
mode of hatchery operation as opposed to the cultchless
resulted in approximately 45% savings in production
costs.

This study represents the first phase of a long-term
project to optimize production scale oyster hatchery
operations. Several problem areas indicated by this
model will be investigated and changes incorporated into
a revised formulation.

Comment

The model presented here deals explicity with hatchery
oyster culture and has little applicability to natural
oyster populations.

A-108

Lord, G.E. 1976. Decision theory applied to the Non-applicable
simulated data acquisition and management of a
salmon fishery. Fish. Bull. 74:837-846.

Abstract

A salmon fishery management model utili zing
statistical decision theory has been constructed.
The model provides for the successive acquisition of
data that can be used to formulate and maintain an
optimum management strategy. The Bayes risk is
defined as the expected economic loss resulting from a
set of fishery management decisions and the criterion
of optimality is taken to be the strategy that
minimizes the Bayes risk. Specific functional forms are
assumed where necessary in order to obtain a closed
fdrm expression for the Bayes risk. Bayes risk in
units of nunbers of fish, can then be computed for any
particular sequence of fishery managment decisions.

Coment

The model presented here deals with management of
a fishery in which there is sequential acquisition of
additive knowledge about stock size during a fishing
season-, and management decisions are made during
acquisition of this information. It also depends on
escapement from the fishery being the sole determinant
of future stock size. The approach appears inappropriate
for Maryland fisheries.

A-109

Loucks, R.H. and W.H. Sutcliffe, Jr. 1978. A simple Applicable
fish-population model including' environmental
influence, for two western Atlantic shelf stocks.
J. Fish. Res. Bd. Canada 35:279-285.

Abstract

For several Scotian Shelf and Gulf ofMaine stocks,
correlations have been observed between ocean temperatures,
subsequent fish catches, and fisHng effort. Ile consider
a simple fish-population model relating these elements.
We suggest that variation in ocean climate triggers
corresponding fluctuations in fish stock-recruitment
and subsequent abundance and catch. Two pathways
between abundance and catch are recognized: (1) catch
is simply proportional to abundance (direct influence);
and (2) in some cases at least, fishermen traditionally
have sensed the abundance of stock and adjusted their
fishing effort accordingly; therefore, in periods of
less favorable climate for a stock, catch would diminish
indirectly because of reduced or "responsive" effort.
In such cases, where the fishermen act to stabilize
the stock, fishing effort and fish catch are correlated
with ocean climate. In any case, ocean climate and
"responsive" effort merit consideration as potentially
significant factors in population models of these stocks.

Comment

The model presented here may possibly be useful for
species which exhibit recruitment as a function of
enviromental variables.

A-110

NlacCall,- A.D. 1978. A note on production modeling Non-applicable
of populations with discontinuous reproduction.
Calif. Fish and Game-64:275-227'.

Abstract

A general exponential population model is adapted
to incorporate situations where discontinuous
reproduction takes place.

Comment

This model might be utilizable in cases of distinct
seasonal reproduction. but overall generates more
assumptions than other models. Its only major
accomplishment isthe removal of the assumption of
continuous reproduction. The method is not applied to
any data in the paper, and it is unknown if it can be
applied to real populations.

A-111

Macketts, D.J. 1973. Manual of methods for fisheries
resource survey and appraisal. FA0 Fish Tech.
Pap. 124. 40 pp.

Abstract

Coment

This paper is unavailabLefor review but has been
requested through interlibrary loans. It will be
reviewed when received.

A-112

Non-applicable
MacKinnon, J.C. 1973. Analysis of energy flow and
production in an unexploited marine flatfish
population. J. Fish. Res. Bd. Canada 30:1717-1728.

Abstract

The seasonal pattern of production processes in an
unexploited resident population of American plaice
(Hippog,lossoides platessoides) in St. Margaret's Bay,
N.S., was analyzed with an energetics model which
represents an extension of the analytical approachused
in fishery theory. During summer, production is about
twice the annual net production of 1.5 hcal/m2 by fish
aged 1 and up. The ecological efficiency is 17% with
larvae and 0+ fish accounting for some 20% of total
population ingestion and 34% of net population production.
Metabolic expenditures constitute the largest fraction
(62%) of population energy intake and about 80% of this
amount is consumed during summer. Plaice ingest about
half the yearly estimated production (2Skcal/m2) of benthos
in the deeper parts of the Bay.

Comment

This model represents a modification of standard
bioenergetic models. Yield is not incorporated as
an explicit variable but is instead incorporated into
the natural mortality term. Also, a large number of
parameters (many of which would be very difficult to
obtain) are necessary for application of the model. It
appears to be unusable.

A-113

Mann, S.H. 1970. A mathematical theory for the harvest Non-applicable
of natural animal populations when birth rates are
dependent on total population size. Math. Biosci.
7:97-110.

Abstract

Natural animal populations are considered in which
members of the population are harvested for their own
value, either esthetic or monetary. Population growth
is assumed to be dependent on total population size.
The parameters of population growth are allowedto be
random variables whose joint distribution fLmction is
described by an associated Markovian distribution pro-
cess. Revenue and cost structures are defined for these
populations and harvesting policies are described that
minimize the expected economic loss to be realized in
the duration of time in which management of the system
is anticipated.

Comment

While this paper presents a population harvest model
incorporating density-dependent, sex-dependent growth,
mortality rates,and economic factors, the model is pre-
sented in a theoretical form which has little utility
for the actual construction of management models.

A-114

Marchesseault, G.D., S.B. Saila 'and W.J. Palm. 1976. Applicable
Delayed recruitment models and their application
to the American lobster (Homarus americ2Lu@
fishery. J. Fish. Res. Bd. Canada 33:1779-1787.

Abstract

A delayed recruitment model intended for use in
developing dynamic strategies for fisheries management
is proposed. The conceptual and analytical properties
of the model are elaborated and compared with those of
the instantaneous model of Schaefer and the delayed
recruitment model recently suggested by Walter. Of the
three models discussed, the delayed recruitment model
proposed herein constitutes the more biologically
meaningful tool for use in management decision making
with fisheries characterized by a multiple year delay
between spawning and recruitment. The proposed delay and
Schaefer models are fitted to catch and effort data from
the Rhode Island inshore pot lobster fishery, and the
generated coefficients are examined with respect tu
their interpretation and relative importance. Values
of optimum equilibrium catch and effort are calculated
for the proposed delay and Schaefer models, and we show
that the delay model's estimates of these management
indices are more conservative than those derived from
Schaefer's model. The proposed delay and Schaefer
models are compared in a dynamic analysis of the fishery,
in which perturbations in the stock level and fluctuations
in the applied effort are simulated to predict the
subsequent behavior of the stock.

Comment

This method may be applicable to a number of Maryland
species, particularly blue crabs. The altered Schaefer
model presented may be useful when little information
other than catch. statistics is available.

A-115

Marchesseault, G., J. Mueller, L. Vidaeus, and W.G. Applicable
Willette. 1979. Bio-economic simulation of the
Atlantic sea scallop fishery. NATO Symp. Appl.
Oper. Res. Fish. August 1979. Trondheim, Norway.

Abstract

A bio-economic simulation fr amework for evaluation
of harvest strategies in the Atlantic sea scallop fishery
is presented. Lacking a time series of stock assessment
data, a Bayesian approach is used to specify a probabi-
listic recruitment function. Stock abundances may be
stochastically simulated over a defined plan period in
response to alternative harvest strategies.
Linked to the biological simulation model is an
economic framework for evaluation of harvesting strate-
gies, consisting of a price forecasting model, a sector
catch model, and a financial (net income) simulator. In
a preliminary application six alternative harvest strate-
gies targeted towards two different terminal stock goals
and describing different annual catch trends are evaluated.
Alternative assumptions regarding fleet size, factor prices,
and the appropriate discount rate are used. To enhance
the usefulness for management decision making, several
aspects of the biological and economic components of the
model are beina refined.

Comment

This paper presents a relatively unique method of
modeling the management of a fishery for maximum
economic return. It incorporates a probability function
of yearclass success (the stock-recruitment relationship
is unknown) stochastically as a biological submodel. The
approach may be applicable for the hard clam, soft clam,
and possibly oyster fisheries in Maryland.

A-116

Marten, G.G. 1978. Calculating mortality rates and Applicable
optimum yields from length samples. J. Fish. Res.
Bd. Canada 35:197-201.

Abstract

An equation is derived for yield per recruit of a
fishery (or other exploited animal population) as a
function of fishing intensity and age of first capture.
The equation has the advantage that it does not require
explicit-estimates of natural mortality or individual
growth rate parameters. Linear length growth is assumed
until maximum size is reache4 and mortality parameters
are expressed relative to growth rate. Mortality
parameters are estimated from average length samples of
separate populations experiencing different fishing efforts
in the same fishery. The equation may be used to compare
existing fishing efforts and age of first capture with
optimal values. Samples of the catfish,Bap-rus docmac,
frcm Lake Victoria (East Africa) are used to illustrate
the method.

Coment

This model may provide a method of estimating potential
yield. The means of estimating mortality and some
* I?
assumptions concerning the similarity of different popu-
lations are subject to some criticism.

A-117

Matuszek, J.E.. 1978. Empirical predictions of fish Applicable
yields of large North American lakes. Trans. Am.-
Fish. Soc. 107:385-394.

Abstract

Regression analyses have been used for over 2
decades to develop useful statistical relationships
between estimates of fish catch and variops members of
� set of abiotic and biotic variables. In this study
� number of conceptual and data refinements have been
attempted with respect to certain lake data. More
complete data are now available than was the case with
earlier authors such as D.S. Rawson and R.A. Ryder.
Rather than use average long-term catch, approximate
estimates of maximum sustainable yield (MSY) were
derived. The MSY of all species combined,
Ct, was estimated as well as the MSY of a preferred
taxa, C , comprising lake trout (Salvelinus namaycush)
plus late whitefish (Coregonus clupeaformis) plus
walleye (Stizostedion vitreum) plus sauger (S.
canadense . The most significant predictive relationships
included only one independent variable, average dry
weight of bottom fauna standing crop, which explained
83% of the variation of Ct per unit water area in a
semilog relationship, and 80% of the variation of Ct
per unit area and Cs per unit area in log-log relation-
ships. Thebest relationships incorporating solely
abiotic variables included mean depth and total dissolved
solids concentration asthe only significant independent
variables, and explained less than 70% of the variation.
Other factors analyzed included the annual cumulative
degree days above 5.6''C, the presence or absence of
thermal stratification, and the average dry weight of
net plankton standing crop.

Comment

This statistical method using regression of morpho-
edaphic factors seems of little use.in estuarine fisheries.
However, the use of benthic standing stock as an indicator
of fish yield may be applicable to the project.

A-118

May, R.M., J.R. Beddington, C.W. Clark, S.J. Holt, Applicable
and R.M. Laws. 1979. Management of multi-
species fisheries. Science 205:267-277.

Abstract

With the overexploitation of many conventional fish
stocks, and growing interest in harvesting new kinds of
food frcm the sea, there is increasing need for managers
of fisheries to take account of interactions among species.
In particular, as Antarctic krill-fishing industries
grow, there is a need to agree upon sound principles
for managing the Southern Ocean ecosystem. Using simple
models, we discuss the way multispecies food webs respond
to the harvesting of species at different trophic
levels. These biological and economic insights are
applied to a discussion of fisheries in the Southern
Ocean and the North Sea and to enunciate some general
principles for harvesting in multispecies systems.

Comment

This is a well-preserted paper demons trat ing, the
importance of the use of multi-species concepts in the
determination of optimal exploitation. The model
presented may be useful for application to the
menhaden-bluefish-striped bass-white perch food web.

A-119

McConnell, W.J., S. Lewis, and E. Olson. 1977. Gross Non-applicable
photosynthesis as an estimator of potential fish
production. Trans. Am. Fish. Soc. 106:417-423.

Abstract

Fish yield and gross photosynthesis were highly
correlated in six ecosystems. Gross photosynthesis of
phytoplankton and attached plants was measured as diel
oxygen changes in unconfined water. Fish yield was
CP
measured by capture of all fish at the end of the
experiments. Fish yield as live weight ranged from
0.54 to 2.48% of gross photosynthesis. The range of
percentages was attributed to differences in foods used
and to differences in stocking densities. Species of
fish were-rainbow trout (Salmo gairdneri , channel
catfish (Ictaluru punctatus), goldfish (Carassius
auratus), and tilapia hybrids (Tilapi mossambica x
Tilapia hornorum

Comment

While the relationship of gross photosynthesis
and yield may prove to be significant, the artificiality
of the ponds used and the differences in stocking
methods and densities make this paper unusable. 'The
paper ignores the effects of reduced circulation on
nutrient availability and thus on photosynthesis.

A-120

Melack, T.M. 1976. Priihary productivity and Non-applicable
fish yields in tropical lakes. Trans. Am. Fish.
Soc. 105:575-580.

Abstract

Measurements of primary productivity can improve
assessment of the fish yields from tropical lakes. In
tropical African and Indian lakes commercial fish yields
increase logarithmically as primary productivity increases
arithmetically. The regression equation describing the
relation between fish yields (FY) and gross photosynthesis
(PG) for eight African lakes is log FY = 0.113 PG + 0.91.
The coefficient of determination is 0.57. The regression
equation based on fifteen tropical Indian lakes, log
FY = 0.122 PG + 0.95, corroborates the relation for
Africa.

Comment

This statistical approach would not be useful in an
estuarine situation due to difficulties introduced by
tidal flow and flushing. The p aper is not directly
applicable.

A-121

Nelson, W.R., M.C. Ingham, and W.E. Schaaf. 1977.' Larval Applicable for
transport and year-class strength of Atlantic menhaden, parameter
Brevoortia tyrannus. Fish. @ull. 75:23-41. estimation

Ahstract

A Ricker spawner-recruit model was developed for Atlantic
menhaden, Brevoortia t)-rannus, from data on the 1955-70.year
classes. The nUTe__r of eggs produced by the spawning stock was
calculated as the independent variable to account for changes
in fecundity due to changes in population size and age structure.
A survival index was developed from deviations around the
Ricker curve and was regressed on several environmental para-
meters to determine their density-independent effects. The
recruit-en-vironment model accounted for over 84% of the vari-
ation in the survival index. Zonal Ekman transport, which
acts as a mechanism to transport larval menhaden from offshore
spawning areas to inshore nursery grounds, was the most sig-
nificant parameter tested. Ricker functions for good and
poor environmental years were developed, indicating the
wide range of recruitment that can be expected at different
stock sizes. Comparisons of spawner-recruit relations for
Pacific sardine and Atlantic menhaden indicated striking
similarities. Surplus yield for the Atlantic menhaden
fishery was calculated from observect and predictect survival,
and compared with the actual performance of the fishery.

Comment

This paper will be useful to estimate successful
recruitment for menhaden. It will have to be adapted
to include yield.

A-122

Oglesby, R.T. 1977. Relationships of fish yield to lake Non-applicable
phytoplankton standing crop, production, and
morphoedaphic factors. J. Fish Res. Bd. Canada
34:2271- 2279.

Abstract

Fish yield is related to annual primary production,
summer phytoplankton standing crop, and the morphoedaphic
index for ,lakes representing a wide variety of typologies
by a series of models in the form of log-log regressions.
Tentative boundary conditions are established by which
lakes inappropriate to the models can be excluded.
Confidence intervals for predicted values about the mean
are given for the fish yield-phytoplankton standing
crop regression. From this relation, potential yields
for the lakes studied are reduced from a range of 10,000
to one of 25-fold. Efficiences with which carbon is
transferred from primary productLon to fish yield vary
by 2 to 3 orders of magnitude and are highest for small,
intensively managed ponds and lowest for large, deep,
cold-water lakes. Models based upon fish yield as a
function of phytoplaikton production or standing crop
are inherently more accurate and subject to fewer
exceptions than are those related to morphoedaphic
factors. The former appear to be capable of substantial
refinement but even in their present state might be employed
to make useful predictions for groups of lakes. A
suggested zupplement to existing approaches in fishery
management involves the following sequence: (1) use of
expectation-variability diagrams to obtain an overview
of the problem; (2) selection of an appropriate model
or models to predict yield; (3) prediction of a range of
yields; and (4) implementation of regulations proved
successful for other lakes in the same yield category.

Comment

The statistical relationship presented in this paper
does not appear to be strong in water bodies of low resi-
dence time (<0.2 yr) as typified by local estuarine situ-
ations. Thus, while the concept may be useful as a secondary
approach, its application may prove fruitless. This same
procedure is described in several other papers.

A-123

O'Heeron, M.K., Jr., and D.B. Ellis. 1975. A Non-applicable
comprehensive time series model for studying the
effects of reservoir management on fish populations.
Trans. Am. Fish. Soc. 104:591-595.

Abstract

We describe a linear modeling technique that derives
functions (parameters) of the predictor variables in a
tabular form using a Fourier transformation technique.
The result is a complex multiple regression model that is
actually a series of models independently describing each
parameter. This technique (1) provides an approximation
to nonlinear modeling that is not restricted to a small
number of variables, (2) does not require precisely
formulated theoretical functions for each variable, and
(3) avoids many of the restricting assumptions normally
required for nonlinear modeling. The estimate of the
total systematic information (variance) in the system
and the amount of information predicted by the model can
be used as a measure of the efficacy of the model. A
production form of the model is being completed for use
on the Missouri River mainstem reservoir system for
predicting the effects of reservoir management practices
on the indigenous fish populations.

Comment

The method is possibly useful, if a large data
base exists, to construct a holistic system model of all
species. The approach seems interesting, but the
authors do not provide sufficient information on
parameter estimation and use of time series analysis to
properly evaluate the technique.

A-124

Orach-Meza, F.L., and S. B. Saila. 1978. Applicable
Application of a polynomial distributed lag
model to the Maine lobster fishery. Trans. Am.
Fish. Soc. 107:402-411.

Abstract

Time series data on catch and effort for the American
lobster (Homaru americanus) from the state of Maine
during the period 1933 to 1974, as well as mean annual
sea surface temperatures for the same period were
examined by means of a polynomial distributed lag model.
The model was described and shown to be sufficiently
flexible to take into account environmental influences
over specific life history stages as well as over the
entire life history of the species. The significance of
lagged temperature effects and the fit of the model to
the observed data were demonstrated.

Comment

The method presented here appears to be a useful
technique, especially for species with a combination of
lagged density-dependent recruitment and environmentally
dominated growth and survival rates during the lag period.
4@ Z@
The technique may prove useful for blue crab, menhaden
(EIcnan transport of larvae), or possib@y shellfish.

A-125

Orth, D.J. 1978. Computer simulation models for predict- Applicable
ing population trends of largemouth bass in large
reservoir's. Proc. Okla. Acad. Sci. 58:35-43.

Abstract

Two conputer simulation models of population dynamics
of largemouth bass (Micropterus salmoides) are described,
Model I is an age-structured, detiermiFistic model with
numbers as the only state vector. Constant age-specific
fecundities and survival rates are required inputs. Model I
simulates population trends based on an equilibrium popula-
tion. Sensitivity analysis of this model indicates that
density of bass is most sensitive to variations in survival
from egg to age I. Multiple regression equations with
water level during spaiming and water level fluctuation
since the end of the previous growing season as predictor
variables resolved 88.2% of the observed variation in year
class strength and 86.7% of the observed variation in
mortality from egg to age I of largemouth bass in Lake
Carl Blackwell. Model II is similar to Model I except
that the effect of reservoir water level and water level
fluctuation on survival of the young-of-the-year is
included. Predictions of number of age I recruits from
Model II agree closely with population estimates for
Lake Carl Blackwell.

Comment

This paper presents a management model for lake
bass which combines a Leslie matrix and statistical
techniques to assess survivorship and age structure.
While the model is not directly applicable to
Maryland species, it may be adaptable to.freshwater
situations for bass,

A-126

Orth, D.J. 1979. Computer simulation model of the popu- Applicable
lation dynamics of largemouth bass in Lake Carl
Blackwell, Oklahoma. Trans. Am. Fish. Soc.
108:229-240.

Abstract

The computer simulation model described predi::ts
year-class strength, production, and yield of large-
mouth bass (Micropterus salmoides) populations. It is
an age-structured, deterministic model with numbers,
average weights and lengths, biomass, yield in weight
and numbers, and gross and net production as state
vectors. Mortality from egg to age I is estimated by a
multiple regression equation; water level during spawning
and water level fluctuation since the end of the
previous growing season are predictor variables.
Predictions of year-class strength, production and
yield compare favorably with estimates made'in previous
studies on Lake Carl Blackwell. Sensitivity analysis
indicates that predictions of production, yield, and
catch (numbers) are most sensitive to variation in rate
of mortality from egg to age I. Growth rates of the
younger age-groups are next in importance in predicting
production, yield and'catch. The model may be useful
for predicting year-class strength, population trends,
and the response of-largemouth bass populations to
minimum length limits, water level fluctuations or
manipulations, and supplemental stockings.

Comment

The paper presents the application of the Leslie
matrix to yield estimation. The method may be appro-
priate for use in this project depending on the avail-
able data base.

A-127

Palm, W.J. 1975. Fishery regulation via optimal Applicable
control theory. Fish. Bull. 73:830-837.

Abstract

This paper attempts to show how control theory
can be used to formulate a regulatory scheme for
fisheries. The regulatory mechanism considered is
a limit imposed on fishing effort. It is shown that
static optimization methods, such as maximum equilibrium
yield analysis, need to be supplemented with dynamic
methods, such as optimal control theory, which take
into account the variable nature of a fishery. The
dynamic analysis is used to show that the size of a
limit on effort should be a feedback function of the
variables in the state of the fishery. The concept
of the Linear-Quadratic Optimal Control Problem is
introduced as a method for devising such a feedback
scheme for fishery regulations.

A single-variable logistic model is used to
introduce the basic concepts. A model with three
variables is then analyzed to show how the techniques
are easily extended to the general multivariable case.
Details of the general method are given in an Appendix.

Comment

This is an excellent paper which explains the
practical uses of optimal control theory for the
determination of optimal regulatory actions in
fisheries management.

A-128

Paloheineo, J.E. 1961. Studies on estimation of
mortalities. 1. Comparison of a method described
by Beverton and Holt and a new linear formula.
J. Fish. Res. Bd. Canada 18:645-662.

Abstract

Comment

This paper is unavailable for review but has
been requested through interlibrary loans. It will
be reviewed when received.

A-129

Parks, W.W. 1975. A Pacific salmon fisheries model for
the study of gear regulation: An application to
the Washington troll fishery. Univ. Wash. Sea
Grant, Seattle, Wash., Norfish Tech. Rep. 58.
165 pp.

Abstract

Comment

This paper is unavailable for review but has been
requested through interlibrary loans. It will be
reviewed when received.

A-130

Parrish, J.D. 1975. @'Iarine trophic inter- Applicable
actions by dynamic simulation of fish species.
Fish. Bull.73:695-715.

Abstract

A mathematical model was developed for performing
dynamic simulations of groups of interacting animal
species. The energy balance of the individual animal
was modeled so that growth and reproduction respond to
food consumption after metabolic expenses are met.
Populations change in response to recruitment (based
on parental spawning) and mortality from natural
causes, predation, starvation and (where applicable)
human exploitation. The forms of the various component
mathematical functions were derived from the available
ecological sources. Functions and parameters are
especially applicable to marine fish species. Trophic
webs of any size or form can be constructed using this
basic species model. Computer solution of the essentially
continuous differential model gives a time history of
trophic and population variables for all species in the
web. Models of trophic webs of 2, 3 and 4 levels were
constructed and exercised. These were used to examine
effects of age class structure, reproductive time lag,
and population regulation by starvation mortality and
fecundity control. Competition between species and
the effects of a top predator on competitors, with and
without human exploitation, were studied.

Comment

The paper presents a good analysis of trophic
relationships in fish food webs. Fishing and harvesting
are only tangentially treated. A very large data set is
necessary, which may make its application in this
project doubtful.

A-131

Patariarche, M.H. 1977. Biological basis for
management of lake whitefish in the Michigan Applicable
waters of northern Michigan. Trans. Am. Fish.
Soc. 106:295-308.

Abstract

Stocks of lake whitefish (Coregonus clupeaformis)
have supported an intensive commercial fishery in
the Michigan waters of Lake Michigan for over a century.
However., certain biological indicators suggested that
recent upsurges in the catch reflect overfishing,
and fish managers should institute measures to assure
a stabilized population and fishery. A biological
basis for establishing quotas is described in this
paper, using information from 1968 - 73 commercial fisheries
in statistical districts MM-1 and NJM-3 and a modification
of Ricker's dynamic pool model. Natural mortality rates
computed for an unfished population of whitefish in the
lower end of nearby Grand Taverse Bay were important
components of the model.

Quota possibilities were based on the premise that
the annual harvest should be confined to weight gained
each year by the harvestable portion of the population.
Six computations of equilibrium yields were made. A
comparison of actual harvests and adjusted yields revealed
an average annual overharvest of 28% during the period
1968-72 in the two statistical districts.

Total biomass for six age groups (I-VI) in three
Michigan statistical districts of northern Lake
Michigan was computed to be 6,600 tonnes in 1972.
Approximately 2,695 tonnes (60%) of the total biomass
in NIVI-1 and n-1-3 were susceptible to exploitation.
Adjusted 1972 yields based on both biomass calculations
and the modified dynamic pool model differed by only
0.3%. A change in the minimum total length limit (from
.432 to 482 mm) to build up a depleted stock also was
discussed. Increased spawning stock, more spawning
opportunities, and greater egg deposition should result
frcm this regulation change.

Comment

The paper presents a possible method for determining
equilibrium yield. However, calculations are based on
numerous assumptions concerning growth and IMSY, which
might complicate its application.

A-132

Patten, B.C. 1969. Ecological systems analysis and Non-applicable
fisheries science. Trans. Am. Fish. Soc, 98:
570-581.

Abstract

Fish population dynamics cannot be separated
operationally from ecosystem dynamics. Modeling
and computer simulation are needed to adapt this
truism for use in fisheries management. To exempli-
fy the relevance of ecosystems to problems in fish
production, a model food web is explored in detail.
Dynamic characteristics are described, and sensitiv-
ity analysis is used to quantify direct and indirect
interactions between components, two of which are
groups of fish populations. Application to optimal
control of a fishery is then considered.

Comment

This paper presents marine ecosystem models
which are generally inapplicable to management
situations. The presentation is primarily a
@heoretical discussion of holistic mechanisms
in systems analysis and the roles of linear and
non-linear systems procedures in ecosystem
analysis.

A-133

Patten, B.C. 1975. A reservoir cove ecosystem model. Non-applicable
Trans. Am. Fish. Soc. 104:596-619.

Abstract

A total ecosystem compartment model of a reservoir cove is
described, with emphasis on the fish submodel. Nominal (unper-
turbed) annual cycles of selected compartments are presented,
and results from three perturbation experiments (thermal pollu-
tion, eutrophication, piscivore invasion) are summarized. The
ability of fishes to influence structure and function of the
entire ecosystem (model) is demonstrated, and the role of top
trophic levels in controlling the design of ecological commu-
nities is discussed. The status and prosDects of total eco-
system modeling as a tool for fisheries science are considered.

Comment

The paper presents a good approach to ecosystem modeling,
assuming the validity of donor-controlled flow. Cursory treat-
ment of harvesting and lack of feedback make this model useless
for this project.

A-134

Paulik, G.J., and W.H. Bayliff. 1967. A generalized computer Applicable
prograrn for the Ricker model of equilibriun yield per
recruitment. J. Fish. Res. Bd. Canada 24:249-259.

Abstract

The Ricker method for predicting the yield per recruit-
nent from a stock of fish.under various conditions is superior
to that of Beverton and Holt (1957) under most conditions be-
cause it permits greater flexibility of the input of growth
and mortality data. Such useful devices as the isopleth
diagram and the eumatic fishing curve, usually associated
with the Beverton and Holt method, can be used also with the
Ricker method. A computer program for the Ricker method is
described, and its use is demonstrated with an example.

Coment

This paper represents an excellent overview of the
Ricker yiel&per-recruit model and will be useful in this
project.

A-135

Pella, J.J., and P.K. Tomlinson. 1969. A generalized Applicable
stock production model. Inter. Am. Trap. Tuna
Comm. Bull. 13:419-496.

Abstract

The problem investigated in this paper is the deter-
mination of the-sustainableyield from fish stocks which
can be anticipated under different effort information. The
yield predictions from the model we discussed should be , I
reasonably accurate provided the harvesting techniques are
the same as those used to generate the data base from which
the parameter estimates are made. A change in size selection
by the fishery or in the time of the year when fishing
occurs (e.g., conpression of fishing seasons due to catch
restrictions) could modify the stock production curve.
Still in these cases the analysis of catch and effort infor-
mation on the basis of the generalized production model should
provide a bench mark from which refinements in yield estimates
can be made on the basis of more detailed studies of the
dynamics of the stock.

Comment

This paper presents a generalization of the Schaefer
Jogistic surplus production model. This form represents
the best application of the surplus production-type model
Cincorporating later alterations in form), but requires the
estimation of five parameters instead of three. It will
Drove useful in this project.

A-136

Peterman, R.M. 197S. New techniques for policy evaluation Non-applicable
in ecological systems: @Lethodology for a case study of
Pacific salmon fisheries. J. Fish Res. Bd. Canada
32:2179-2188.

Abstract

The complexity of exploited ecological systems creates
difficulties for the manag
,er who must decide among alternative
policy options. Some methods for overcoming these difficulties
are presented, using examples from the salmon fishery of the
Skeena River system in British Columbia. The described methods
produced a "desk-top optimizer," a tool that permits decision-
makers to perform fairly sophisticated "optimization" opera-
tions at their desks instead of having to rely on decision
theorists or operations researchers. Also discussed are
various system indices which should become part of the infor-
mation used by managers. These indices include measures of
resilience (ability to absorb the effects of unexpected events),
costs of failures in management policies, and cost of tincer-
tainty of various types.

Comments

The method presented for optimization of regulatory
mechanisms represents a simple approach that can be used
after population models are completed. However, the @s
technique is not as good as dynamic programming. It is
explored further in a follow-up paper.

A-137

Peterman, R.M. 1977. Graphical evaluation of environmental Applicable
management options: Examples from a forest-insect pest
system. Ecol. Modelling 3:133-148.

Abstract

A graphical technique is demonstrated which, when
combined with any resource simulation model, permits the
resource manager to explore the effects of different
management options. Also, this technique (nomogram or
response surface) permits derivation of "optimal solu-
tions" given particular objectives. Examples of the
methodology are given for the spruce budworm -- forest
system in eastern Canada. Effects of several kinds of
uncertainties are discussed, including uncertainties in
model assumptions, management precision, future objectives
and system evolution. The graphical nature of nomograms
helps managers and analysts to grasp more easily the
complicated behavior of ecological system modelsi
Finally, the role of computer models in decision-making
is,discussed.

Comments

This paper presents a good and possibly useful
optimization method which requires little computer
work. It maybe worth evaluating for future use.

A-138

Plourde, C.*G. 1970. A simple model of replenishable Applicable
natural resource exploitation. Amer. Econ. Rev.
60:518-522.

Abstract

A simple economic model is proposed to determine
optimal steady-state population levels and consumption
levels of replenishable natural resources. The model
is biological, based on the'logistic growth equation, and
incorporates a welfare utility function which depicts
the economic dynamics of the exploitation of the re-
source. Optimal control theory is then applied to the
resulting steady state to determine optimal usage levels.

Comments

The methods presented here are possibly applicable
to the project as a simplified method of integrating
biology and economics. SiiTplifying assimptions may
limit its applicability.

A-139

Pope, J.G. 1971. An investigation of the accuracy of
Virtual population analysis. Inter. Comm. North-
west Atl. Fish. Res. Doc. 71/116 Ser. No. 2606:!7-11.

Abstract

Comments

This paper is unavailable for review, but has
been requested through interlibrary loans. It will
be reviewed when received.

A-140

Pope, J.G. 1972. An investigation of the accuracy of virtual Applicable
population analysis using cohort analysis.
Inter. Comm. Northwest Atl. Fish. Bull. 9:65-74.

Abstract

Cohort analysis is a simplified, approximate form of Gulland's
virtual population analysis. As such it may be used to obtain
estimates of the instantaneous rate of fishing morta@@ity and the
population surviving for each age of a year class, given the catch-
at-age data, and an estimate of the instantaneous rate of natural
mortality and an estimate of the fishing mortality at the final
age of exploitation. More importantly, the simplicity of cohort
analysis makes it possible to investigate the errors generated in
such estimates by the arbitrary choice of the rate of fishing
mortality on the last age exploited and by the sampling errors
of the catch-at-age data.

Comwnt

This paper presents an excellent explanation of the use
of virtual population analysis and cohort analysis. It will
be useful in this project.

A-141

Powers, J.E., R.T. Lackey, and J.R. Zuboy. 1975. De- Non-applicable
cision-making in recreational fisheries management.
Trans. Am. Fish. Soc. 104:630-634.

Abstract

.A conceptual model of the decision-making process
in fisheries management is presented in conjunction with
applications of computer and systems analysis to this
process. One of the most difficult problems to solve
is selecting objectives to be used for management evalu-
ation. An objective function based on some or all of the
components of yield, species, size desirability, and
environmental quality is needed. Systems analysis and
computer technology in data processing and simulation
may be used in many situations to evaluate decision
alternatives as an aid in developing management strate-
gies.

Comments

This paper is not useful to this project since it
does not describe an analytical model. A method for
analysis of the decision process may be helpful at a
later date.

A-142

Rafail, S.Z. 1977. A simplification for the study Applicable for
of fish populations by capture data. Fish. Bull. parameter estimation
75:561-569.

Abstract

Expressions given by Rafail for estimating
catchability are modified here to eliminate itera-
tion, for better accuracy, and a large economy in
calculations and time. The evaluation of catch-
ability allows the estimation of other important
parameters with'the useful assumption of their
variabilities according to seasons and recognized
sections of a population.

Comments

The paper presents a method which appears
useful for parameter estimation.

A-143

Regier, H.A., and H.F. Henderson. 1973. Towards a broad Applicable
ecological model of fish communities and fisheries.
Trans. Am. Fish. Soc. 102:56-72.

Abstract

The paper brings together major inferences from:
(1) classical limnology--lake and stream typology, the
role of major abiotic variables; (2) fisheries limol-
ogy--Ryder's morphoedaphic index, Jenkins' reservoir
findings, concepts of habitat niches; (3) studies of
ecological structure of cormunities--succession,
diversity, stability, variability, regulation; (4) re-
cent developments concerning the effects of major
cultural stresses on fish communities. A model is
.proposed to interrelate these and other concepts,
and then relate them all to conventional fisheries
practices and objectives. The model is directed at
events and processes at the community level of organ-
ization and it-is a@gued that much of fisheries
theory and management practices of the future will
perforce need to be directed at the community level.

Comment

This model should be further reviewed.as an
exanple of an ecosystem yield application, although
its usefulness for-this project is doubtful,

A-144

Picker, W.E. 1945. A method of estimating minimum size Applicable
limits for obtaining maximum yield. Copeia 1945:84-94. for
parameter
estimation
Abstract

1. The "critical size" for a year-class of fish is
defined as the size at which the average instan-
taneous rate of natural mortality begins to
exceed the average instantaneous rate of increase
in weight.

2. Assuming that what is true of the year-class as a
whole is also true, on the average, of the fish
individually, the critical size can be used as a
basis for estimating the best minimum size at
which fish should be taken, in order to achieve
maximum production at the existing rate of fishing.

3. The best minimum size varies with the rate of fish-
ing, being zero when the latter is very small, and
approaching the critical size when fishing is 'very
intensive.

4. Possible secondary effects of a change in fishing
C,
intensity, on the rate of growth of the fish, make
it unwise to extrapolate too far from existing
conditions: i.e., the critical size may change.
However such extrapolation is a useful guide to
the minimum size which will probably be desirable,
following a moderate change in rate of fishing.

Comment

This paper may be useful in determining growth and
mortality coefficients.

A-145

ldcker, W.E. 1958. Maximum sustained yields from fluctua- Non-applicable
ting environments and mixed stocks. J. Fish. Res. Bd.
Canada 15:991-1006.

Abstract

Using numerical models, effects of environmental vari-
ability upon yield were tested for six single-age fish stocks
characterized by different kinds and degrees of density-de-
pendent reproduction potential. The two levels of variability
examined had extremes of yield standing in the ratios 7:1 and
18:1, respectively. Close regulation of fishing to the optimum
percentage for each year's stock improves the long-term average
catch taken, the inprovement being the greater, the more vari-
able the environment. With the higher level of variability,
improvement in average catch among five of the stocks ranged
from 26% to 79% increase. However this increase in mean catch
is achieved at the expense of increased variability in catch
from year to year -- in fact, for some kinds of stock there
must be.complete cessation of fishing in some years in order
to get the long-term maximum. The yield of stocks, in which
reproduction per spawner declines at low levels of abundance,
is particularly improved by a close adaptation of fishing
effort to the supply of fish available.
When two or more populations of a species, characterized
by different reproduction potentials, are fished in common,
total potential catch is less than when each can be fished
separately at its optimum level. If a common fishery cannot
be avoided, the achievement of maximum average yield may find
one of two originally-equal stocks as abundant or even more
abundant than before the fishery began, while the other may
persist only at a low level or even be exterminated completely.

Comment

This paper reports no useful new information concerning
the Ricker yield-per-recruit model and will not be useful .
for this project. It does present a good comparison of Ricker
and Beverton-Holt stock-recruitment relationships.

A-146

Ricker, W.E. 1973. Two mechanisms that make it impos- Applicable
sible to maintain peak-period yields from stocks
of Pacific salmon and other fishes. J. Fish. Res.
Bd. Canada 30:1275-1286.

Abstract

'Nechanism 11' has two aspects: catches taken at a
given rate of exploitation are greater when rate of
exploitation has been increasing than when it has been
steady or decreasing; also, the yield taken from the
progeny of a spawning of a given size is greater when
rate of exploitation has been increasing than when it
has been steady or decreasing. "Mechanism 211 is the
fact that mixtures of stocks of unequal productivity,
when harvested together, produce smaller recruitments
than single stocks of the same original size.and having
the same optimum rate of exploitation. In addition,
any fishery for a valuable species is likely to develop
beyond the optimum rate of exploitation because there
is no easily detectable symptom that the optimum is
being passed. When this has happened, maximum sus-
tainable yield (MSY) will not be achieved immediately
if the optimum rate is imposed subsequent to a period
of overexploitation; rather there will be a gradual
approach to @SY that extends over several generations
after the optimum rate is established. Both of the
two mechanisms above, plus the likelihood of unrecog-
nized overfishing, make for a catch maximum while
fishing is still on the increase. For salmon this
maximum is likely to be 30-60% greater than the
sustainable yield. In addition, unavoidable diffi-
culties of management make for even greater differ-
ences between the historical maximum and the mean
equilibrium yield that can be achieved in practice.
Good annual prediction of recruitment can improve
this picture because rate of exploitation can then
be adjusted to the quantity of fish available; how-
ever this procedure too is much.less effective when
mixtures of stocks are fished in common, because in
general the recruitments to different stocks do not
vary in exactly the same way. The phenomena described
may also contribute to an historical early maximum of
catch in fisheries for species such as cod, being in-
dependent of and additional to the maximum caused by
"removal of accumulated stock."

Comments

The paper presents a useful methodology for
comparing single stock and multiple stock applications
of Ricker yield/recruit models. It discusses the reasons
for optimal yield being below maximal yield in the con-
text of population biology. A-147

Riffenburgh, R.H. 1969. A stochastic model of inter- Applicable
population dynamics in marine ecology. J. Fish.
Res. Bd. Canada 26:2843-2880.

Abstract

A nonstationary Markov chain was developed to analyze
and model the passage of energy through an ecological system
composed of the Pacific sardine, the northern anchovy, their
competitors, their predators, and their prey; then to carry
the model forward through time projecting the biomasses of
the relevant species if anchovy or hake fisheries, or both,
have begun. The model seemed to agree adequately with
observed data.
The hypothesis of some earlier work of the author that
the abundances of populations could be completely controlled
by fishing intensities was rejected. Although no measure
was made of the robustness of the ecosystem, it was concluded
to be relatively impervious to artificial pressures, although
there was seen a measurable boundary beyond which the eco-
logical interactions became unstable and the system collapsed.
Specifically, based on 1950-S9 data, the sardine catch
was projected to stabilize in the vicinity of 20 metric
kilotons per year. Overfishing did not seem to have been
a cause of the sardine "disappearance." To maximize sardine
catch, the introduction of anchovy and hake fisheries with
annual catches limited to certain percentages (depending
an the maximizing criterion) of abundances would double or
triple sardine catch and stabilize the fishing industry,
but sardines could not be reinstated as the most abundant
species as a result of selective fishing. To maximize the
combined catch of the three fisheries, annual catches of
30% and 20% (for all criteria) of abundances of anchovy
and hake, respectively, would yield nearly 1800 metric
kilotons of fish, although most of it would be.of less
commercial interest than the sardine.

.Comment

While this paper presents a very complex model of three
interacting species, the application of the model requires
a large numer of parameter values. This modeling approach
(multiple population simulation model) may be applicable
for interacting and/or competitive stocks (e.g., menhaden
and bay anchovies) in Maryland, but the data base required
for its utilization may be prohibitive.

A-148

Robson, D.S., and D.G. Chapman. 1961. Catch
curves and mortality rates. Trans. Amer.
Fish. Soc. 90:181-189.

Abstract

Comments

This paper is unavailable for review, but has
been requested through interlibrary loans. It will
be reviewed when received.

A- 149

Rothschild, B.J., and J.W. Balsiger. 1971. A linear Non-applicable
programming solution to salmon management.
U.S. Fish. Wild. Serv. Fish. Bull. 69:117-139.

Abstract

A linear-programming model was constructed to
allocate the catch of salmon among the days of the
salmon run. The objective of the model was to derive
a management schedule for catching the salmon which
would result in maximizing the value of the landings
given certain constraints. These constraints ensured
that cannery capacity was not exceeded, and that
escapement of both male and female fish was "adequate."
In addition to considering 'the allocation of the catch
in the primal problem, the dual problem considered
the shadow prices or marginal value of the various
sizes of fish, eggs, and cannery capacity, thus enabling
the manager to view his decisions in light of the
marginal values of these entities.- As an example, the
model was applied to a run of sockeye salmon in the
Bristol Bay system. In the particular example, which
was chosen to replicate the 1960 run, the additional
value of the catch owing to optimality amounted to
an ex-vessel value of a few hundred thousand dollars.
In addition it appeared that the required processing
time could be reduced by several days. The optimun
allocation was obtained through conformance to the
linear-programming model. The cost of this confor-
mance was not, howeverP determined.,

Comment

This paper presents a linear programing optimi-
zation procedure for bioeconomic variables in a salmon
management model. This procedure tests various manage-
ment schemes allowing the determination of an optimal
plan to maximize dock value of salmon. The approach
is inappropriate to the project at this stage.

A-150

Rudd, W.G. 1975. Population modeling for pest management Applicable
studies. Math. Biosci. 26:283-302.

Abstract

Modeling for crop insect pest management studies re-
quires detailed attention to age- and stage-specific
effects. Compartmental models with time delays produced
by impulse response functions and proper attention to age
density distribution functions provide for the needed gen-
erality and precision. Data required for such models
include emergence and mortality functions, initial popu-
lations, and initial age density distribution functions.

Comment

This paper represents a good discussion of the use
of discrete and continuous time lag in the population
dynamics descriptions of a pest species. It may be
applicable to harvested species as well.

A-151

Ryder, R.A., and H.F. Henderson. 1975. Estimate of potential Non-applicable
fish yield for the Nasser Reservoir, Arab Republic of
Egypt. J. Fish. Res. Bd. Canada 32:2137-2151.

Abstract

Available biological and hydrographic information on Lake
Nasser was examined to provide guidelines for future develop-
ment of the fisheries. Not enough consistent data were avail-
able to allow precise estimates of fishing-stock interrelations-,
however, an approximation of potential yield was derived from
subjective evaluation of limnological characteristics of the
lake. Morphoedaphic factors were utilized to establish a
numerical expectation of yield in relation to other tropical
lakes and reservoirs. Two other models based on estimated
zooplankton biomass and an assumed net gain in productivity
in Lake Nasser at the expense of a yield loss to the Mediter-
ranean fishery provided order-of-magnitude agreement. Potential
yield estimates for Lake Nasser (including its southern
portion, Lake Nubia, Sudan) were 23,000 metric tons (Mr) of
fish annually, providing the reservoir fills to its expected
maximum (180m), or 12,000 N1T if it remains at its 1973 level
of about 160m. These estimates are considered sufficiently
reliable to guide future development of the fishery.
There is little evidence to support the existence of
stock overexploitation. Tilapia nilotica, however, the major
species entering the fishery, has been subjected to potentially
damaging fishing practices on its breeding grounds. These
practices combined with water level drawdown, inadvertently
timed to perturb Tilapia fry inhabiting the shallow nursery
areas, may unduly stress a prime species largely responsible
for the current success of the fishery.
Management recommendations for the Nasser-Nubia Reservoir
include a regular program of monitoring both stocks and har-
vest, especially for Tilapia. Development emphasis should be
placed on improving tEe 'logistics of the fishery rather than
increasing the number of fishermen, at least until the lake
area expands substantially.

Comment

The material presented here adds little to morpho-
edaphic index applications described in other papers which
have been reviewed. It is also based on poor estimates of
seasonal or annual trends in edaphic variables.

A-152

Saila, S.B., and K.W. [less. 1975. Some applications of optimal Applicable
control theory to fisheries management. Trans. Am. Fish.
Soc. 104:620-629.

Abstract

A review is made of some optimization techniques applied
to fisheries and related problems. Explanations of optimal
control theory as applied to the Schaefer or logistic model
and to the Beverton and Holt model are provided. An optimal
policy for a fishery management model based on the Brody
growth function is developed. The control variable in each
of the above examples is the rate of fishing and the objec-
tive function is the maximum biomass yield. A method for
estimating the parameters of the Brody growth function is
illustrated for the albacore tuna.

Comment

This paper presents a good discussion of optimization
techniques and their application to two major yield models.
The methods may be useful if optimization techniques are
employed after yield models are developed for Maryland stocks.

A-153

Schaaf., W.E., and G.R. Huntsman. 1972. Effects of fishing on Applicable
the Atlantic menhaden stock: 1955-1969. Trans. Am. for
Fish. Soc. 101:290-297. parameter
estimation

Abstract

To determine the effect of purse-seine fishing on the
Atlantic menhaden (Brevoortia taannus) population, we
analyzed data from 1955-69 -on fishing activity and catches.
Changes in fishing efficiency necessitated establishment
of an abstract effort unit, the 1965 vessel-week. Catch
per unit of effective effort in 1965 was one-fifth that of
1955. Our instantaneous natural mortality rate M estimate
was 0.37. At the current recruitment age, 1.5 years,
reducing F, instantaneous fishing mortality, to about 0.8
would slightly decrease the yield per recruit, but would
increase the spawning stock and ultimately allow annual
catches of 400,000-500,000 metric tons, the maximum sus-
tained yield.

Comment

This methodology may be applicable to menhaden stocks,
particularly in conjunction with other menhaden models con-
cerning the effects of environmental variability on yield
and recruitment. This paper also describes a method for the
normalization of total effort when the fishery undergoes
large changes due to technological advances.

A-154

Schaefer, M.B. 1954. Some aspecis of the dynamics of population Applicable
important to the management of the commercial marine fish-
eries. Inter.-Am. Trop. TLma Comm. Bull 1:25-56.

Abstract

A population of oceanic fish under exploitation by a fishery
may be influenced by a great number of elements in the complex
ecological system of which it forms a part. Of these, however,
only one, predation by man, is capable of being controlled or
modified to any significant degree by man's actions. Any
management or control of the fishery, to the extent this may
be possible at all, must, therefore, be effected through control
of the activities of the fishermen. It seems important to elu-
cidate some of the basic principles of the effect of fishing on
a fish population and, conversely, the effect of the fish popu-
lation on the mount of fishing, in order to understand in what
circumstances and in what manner such control of the activities
of the fishermen can influence the fish population and the yield
obtained therefrom.

Comment

This paper presents the logistic surplus production model.
Although severely limited by assumptions concerning stock dyna-
micso it is a very useful model, particularly in early manage-
ment phases, because of its limited data requirements.

A-155

Schaefer, M.B. 1957. A study of the dynamics of the fishery Applicable
for yellowfin tuna in the Eastern Tropical Pacific Ocean.
Inter.-Am. Trop. Tuna Comm. Bull. 2:245-285.

Abstract

The mathematical model employed is essentially the same
as that discussed by Schaefer (1954), with some modifications
in notation. The theory is, however, also extended in appli-
cation to provide estimates of all the essential constants.
from the catch data alone., without recourse to tagging data
for estimating fishing mortality which was required in the
earlier paper.

Comment

This paper provides little additional information con-
cerning the Schaefer surDlus; production formulation. Its
major advance is the determination of a method*of calcu-
lating catchability (q) without utilizing additional data
(e.g., tagging studies). This paper will be useful if
surplus production models are applied in this project.

A-156

Schaefer, M.B. 1968. Methods of estimating effects of fishing Applicable
on fish populations. Trans. Am. Fish. Soc. 97:231-241.

Abstract

The yield from an exploited fish population depends on
the rate of harvesting (fishing mortality rate) and the
magnitude of the standing stock. The latter is determined
by rates of increase from recruitment and growth and rates
of loss from both natural and fishing mortality. Since we
lack information respecting the density-dependence of each
of these rates, various simplifying assumptions are made in
practice in developing mathematical models of exploited fish
populations. Such models are described, with illustrations
of their application to important commercial fisheries.
Models of fisheries involving competing fish species, and
the employment of computer simulation in studies of fishery-
dynamics are also discussed.

Comment

This paper presents excellent reviews and comparisons
of "dynamic pool" and "logistic" models. While it is not
directly applicable to the project in the sense of providing
a new yield model, it describes the relative merits, assump-
tions, and similarities of the two approaches.

A-157

Schnute, J. 1977. Improved estimates from the Schaefer Applicable
production model:Theoretical considerations.
J. Fish. Res. Bd. Canada 34:583-603.

Abstract

The Schaefer production model is converted to a form
directly applicable to a data stream of annual fishing
efforts and catches. The new version is also stochastic;
that-is, it allows for unpredictable influences on the
fishery. A new method for estimating optimum effort and
catch results from this analysis, as well as a way of
measuring uncertainty in these estimates. Equations are
given for predicting the next annual catch and assigning
confidence limits to this prediction. Linear and non-
linear regressions are proposed for this analysis, and
the relationship between them is rigorously demonstrated.
The linear method leads to estimation formulas simple
enough to be applied on a programmable pocket calculator.

comment

The paper presents a modified means of fitting a
Schaefer model to existing effort and yield data, which
provides- better estimatescof the variability associated
with maximum sustainable yield. It may be useful in
this study.

A-158

Sheldon, RX, W.H. Sutcliffe, Jr., and M.A. Paranjape. Non-applicable
1977. Structure of pelagic food chain and rela-
tionship between plankton and fish production.
J. Fish. Res. Bd. Canada 34:2344-2353.

Abstract

Further observations on the standing stocks of
pelagic organism confirm the occurrence of approxi-
mately equal biomass over logarithmically equal size
ranges. A simple theoretical framework is developed
that shows that the structural elements of the pelagic
ecosystem can be described in terms of the sizes of
predator and prey and of the efficiencies of their
interactions. In practice this means that if the
standing stock at any size range is known, the pro-
duction can be estimated. The theory is tested on
three fisheries. For the Gulf of Maine and the
North Sea, phytoplankton production is estimated
from fishery production. For the area off Peru the
fishery production is estimated from the plankton
production.

Comment

The paper applies several theoretical models
to pelagic fish.communities. The applications
do not deal specifically with the development of
management of these fisheries, and do not appear
to be particularly relevant to this study.

A-159

Shute.rP B.J., and J.F. Koonce. 1977. A dynamic model of Non-applicable
the Plestern Lake Erie walleye (Stizostedion vitreum
vitreum) population. J.-Fish. Res. R. Ca-RaTa-
.34:1972 1982.

Abstract

A simple population model of western Lake Erie walleye
was constructed using empirical relationships linking
growth to population density and recruitment to breeding
stock size and the spring water temperature regime. Given
a reasonable set of total survival for the period from
1947 to 1975, the model generated a pattern of behaviour
similar, both qualitatively and quantitatively, to that
exhibited by the real population. Two types of stochastic
models,based on the initial population model, were used
to derive optimal harvest strategies for the population.
Optimal strategies were not sensitive to variations in
catchability and natural mortality. Yields produced were
highly sensitive to variations in both these factors. A
refined and extended version of the model may serve as a
useful tool in developing realistic management policies
for this population.

Comment

The paper applies pre-existing yield models to a
specific fishery based on a large amount of empirical
data. It has little applicability to this study.because
the species modeled is not harvested in significant
numbers in Maryland.

A-160

Silliman, R.P. 1969. Analog computer simulation and Applicable
catch forecasting in commercially fished popula-
tions. Trans. Am. Fish. Soc. 98:S60-S69.

Abstract

An analog computer was programed and used to derive
data on population dynamics of fish and whales. The types
of mathematical models that were applied and the results of
the studies, several of idnich have been published, are
briefly reviewed.
Applications included two types of mathematical models--
models of populations for studying and forecasting yields
and models of competing populations. The yield models
combined fishing and natural mortality rates with a Gompertz
growth curve in a differential equation. Wien integrated
by the computer, this equation produced survival curves for
successive year classes, whose annual values were summed
graphically. Recruitment was determined from a stock-
recruitment curve. Yields for each season were calculatea
from stock weight by using knoi%m rate of exploitation and
were compared with actual yields to test the validity of
the models.
The technique has been demonstrated for Pacific sar-
dine (Sardinq@s sagax , haddock (Melanogrammus aeglefinus),
Atlant cod (Gadus morhua), lake_t_r_o_ut_-_FS_a1vel1_nus namay-
cush) , skipjacF pelamis) -Fir-ge-ye tun-a-_-
(775nnus obesus), blue-vvhale (BaLaenoptera musculus), and
fi-n-Walie (Ba.-aenoptera p@xsalU@7. -Forecasts of s tain-
able catch Fa-ve been made for the two species of tuna and
for the fin whale.
The models of two competing populations use Volterra
equations to generate biomass of one population as a func-
tion of biomass of the other. They have been applied to
data of Pacific sardine -- northern anchovy (Engraulis
mordax: and yellowfin tuna @Thunnus albacares -- skipjack
tuna.

Comment

The paper presents the application of a simple analog
model to several marine species. While giving generally
good results, it is unclear how yield is determined in the
model. Also, its omission of environmeental information
makes a determination of its applicability to Maryland
fisheries difficult. In light of its catch forecasting
success, perhaps this type of simple yield-recruitrient
approach should be investigated further.

A-161

Shuter B.J., and J.F. Koon ce. 1977. A dynamic model of Non-applicable
the Plestern Lake Erie walleye (Stizostedion vitreum
vitreum) population. J. Fish. Res. Bd. Can@ada
.34:19 1982.

Abstract

Comment

A-160

Silvert, W. 1977. The economics of over-fishing. Applicable
Trans. Am. Fish. Soc. 106:121-130.

Abstract

I analyze the problem of determining the optimal
economic strategy for exploitation of a fish popula-
tion when the stock has been driven to a low level and
is threatened with extinction if over-exploitation
continues. Two very simple models are investigated
in detail and only very simple optimization techniques
are used. Whether the optimal strategy, namely that
which maximizes present value, is one that leads to
conservation or extinction depends on economic factors
which are partially determined by public policy, such
as tax structure. In some economic situations the
optimal strategy will always lead to extinction, and
in these cases a policy of conservation requires direct
government intervention through quotas and other re-
strictions; in other cases the choice between conserva-
tion and extinction can be affected by purely economic
policies. I hope that this type of analysis will make
it possible for societies to identify the most efficient
methods of encouraging conservation while minimizing
economic dislocation and direct governmental control.

Comment

41ity
The material here may have applicab-L to stocks
which are over-exploited, if any such populations occur
in Maryland. It also presents a good discussion of the
economics of fisheries.

A-163

Silvert, W. 1978. The price of knowledge: Fisheries Non-applicable
management as a research tool. J. Fish. Res. Bd.
Canada 35:208-212.

Abstract

one of the side effects of fisheries management is
the discovery of new scientific information. Since this
information has economic value, in that it can be used to
improve future management of the fishery, the information
that can be gained through a particular management strategy
should not be ignored in evaluating that strategy. This
paper shows, using a simple model, how the research com-
ponent of fisheries management can be measured and used to
plan an optimal strategy. The management objectives are
taken to include avoidance of risk and maximization of
yield. The results depend critically on the time horizon
for management. Long-term management favors creative risk-
taking and leads to optimal future exploitation, while
management based on short-term considerations may freeze
the fishery in a permanent pattern of suboptimal yields.

Comment

The argments presented in this paper apply only to
situations where the species of concern is already being
managed according to some yield functions. It appears
premature for this study.

A-164

Sinko, J.W., and W. Streifer. 1967. A new model for age- Non-applicable
size structure of a population. Ecology 48:910-918.

Abstract

An equation describing the dynamics of single species
populations is derived. The model allows for variations in
the physiological characteristics of animals of different
ages and sizes. An analytical solution which holds under
certain specific conditions is found. It is shown that
Von Foerster's equation, the logistic equation and other
prior models are special cases of the new model.

Comment

The discussion presented here is theoretical in nature.
The complexity of the required inputs make it inapplicable
for this study.

A-165

Sissenwine, M.P. 1974. Variability in recruitment and Non-applicable
equilibrium catch of the Southern New England
yellowtail flounder fishery. J. Cons. Int. Explor.
Mer. 36:15-26.

Abstract

The significance of biotic and abiotic regulation of
the Southern New England yelloi-Aail flounder fishery was
investigated. Analysis of the published data provided no
e of biotic regulation of the fishery. This analysis
evidenc
included the estimation of the annual equilibrium catch and
recruitment to the fishery during the period 1944-1965.
The multiple correlation coefficients of regressions
fitted between the natural log of the annual equilibrium
catch and recruitment of the fishery with three- and four-
year moving averages of atmospheric temperature at Block
Island, Rhode Island CU.S.A.) ranged from 0-862 to 0-927.
According to these regressions, the decline of the fishery
during the late 1940s resulted from the adverse effect of
a general warming trend in the region.

Comment

This paper adds no new information to the yellowtail
simulation study done by the same author; thus it will not
be included in this study.

A-166

Sisserr,vine, INT.P. 1977. A compartmentalized simulation model Applicable
of the Southern New England yellowtail flounder, Limanda
ferruginea, fishery. fish. Bull. 75:465-482.

Abstract

A compartmentalized simulation of the Southern New England
yellowtail flounder, Limanda ferrugine , fishery was developed.
The population was diVile-dinto 10 age-groups, each of which was
subdivided in 7 size categories. The model simulated discard
mortality as well as natural mortality and fishing mortality.
Fishing and discard mortality rates depended on the level of
fishing and on gear and market selection factors. Both linear
and density independent stock-recruitment functions were con-
sidered. Seasonal variations in growth and exploitation were
incorporated into the model. The influence of fluctuation in
temperature on recruitmnt and growth was also simulated. The
model usina a linear stock-recruitment function accounted for
85.5% of the variability in the yield of the fishery for 1943-
65; with a density-independent stock-recruitment function.
the model explained 83.2% of the variability in yield for the
same period.
The linear stock-recruitment model was used to investigate
the response of the fishery to alternative fishing strategies.
Substantial increases in the past yield of the fishery were
indicated by the model when fishing effort was concentrated
during the second half of the year and when fishing effort
and discard mortality were reduced.

Comment

This paper presents a very detailed and complex simulation
model of an exploited fish stock. It may be of value in this
study as an example of a single population simulation model.

A-167

Sissenwine, M.P., and A.M. Tibbetts. 1977. Simulating the Applicable
effect of fishing on squid (Loligo and Illex popula-
tions of the northeastern United States. Int. Comm.
Northwest Atl. Fish. Sel. Pap. 2:71-84.

Abstract

Models designed to simulate the effect of fishing on
squid (Loligo and Illex) were developed. The instantaneous
growth, fishing and natural mortality rates were varied on
a monthly basis. Spawning was simulated over an extended
period. Recruitment was described by the Beverton and
Holt (1957) stock-recruitment function.
Based on these models, the exploitation rate (over the
lifespan of the species, EMSY) that will result in the max-
imum sustainable yield is 0.75 and 0.63 for Loligo and Illex
respectively, if recruitment is independent of spawning stock
size. If recruitment is moderately dependent on spawning
stock size, the EMSY is probably about 0.40 and 0.37 for
Loligo and Illex, respectively. EMSY is further reduced to
about 0.15 for both species for a population with a stronger
stock-recruitment relationship.

Comment

This paper represents one of the few applications of
Beverton-Holt stock-recruitment relationships. While its
predictive power is in doubt, it should be analyzed further
and may be helpful in this project.

A-168

Sissenwine, M.P., B.E. Brown, and J. Brennan-Hoskins. 1979. Applicable
Brief history and state of the art of fish production
models and some applications to fisheries off the
northeastern United States. pp. 25-48 In Climate and.
Fisheries. Center for Ocean Mianagement-Studies. Univ.
of Mo-de--I s land, Kingston, R.I.

Abstract

Production models applicable to individual fish, cohorts
of fish, and entire populations are reviewed. Hypotheses
(often untested) describing the relationship between produc-
tion and the-biotic and abiotic environment are advanced.
The paper supports the following generalizations:
(1) The effect of the physical environment on fish
production is better understood by considering
fisheries in an ecological context instead of
the context of the traditional fish production
models. On the other hand, fish production
models can sometimes be modified to account for
environmentally induced fluctuations empirically.
(2) @bst of the variability in fish production results.
from presently unexplained variability in repro-
ductive success. Variability in reproductive
success probably reflects a plethora of complex
biotic and abiotic interactions of early life
stages of fish.
(3) Fish production models do not usually consider
the effect of a fluctuating environment explic-
itly, but their application is usually tempered
so that the conclusions based on these models
are generally valid and useful.

Comment

This paper reviews the relationships between
existing yield models and discusses their application
to many Atlantic fisheries. It will be very valuable
to this study.as a source for the comparability of the
many model types.

A-169

Smith, R.H., and R. Mead. 1975. A note on population growth Non-applicable
in a variable environment. Oecologia 20:333-337.

Abstract

Some properties of diffusion approximations to models
of population dynamics in a variable environment are noted
and discussed. A counter-intuitive feature of these models
is that a larger mean population can lead to instability,
given a particular level of variation. The need to consider
different types of stochastic variation is stressed. Migra-
tion is shown to be either stabilising or destabilising,
depending on the balance of effects of different types of
variation.

Comment

The paper explores the conseauences of incorporating
stochastic processes into a logistic growth model. It pre-
sents little information that is directly applicable to
this study as its application-is primarily theoretical in
nature.

A-170

Smith, T.D., and T. Polacheck. 1979. Analysis of a simple Non-applicable
model for estimating historical population sizes.
Fish. Bull. 76:771-779.

Abstract

Estimates of historical abundance of animal populations
are important in mmy management decisions. Historical esti-
mates based on a simple model of population growth have been
made for several populations of dolphin involved with the
yellowfin tuna purse seine fishery. We used the data for the
bridled dolphin, Stenella attenuata, to investigate the behav-
ior of the model by whi these his orical estimates were
calculated. For populations with low net reproductive rates,
the effect of bias in the estimates of the input parameters
on the estimated historical abundances was approximately
linear and additive. When all the input parameters were
independently estimated, the variances of the historical
abundance estimates were dominated by the variance of the
initial abundance estimates and the coefficient of variation
of the historical estimate was less the largest coeffi-
cient of variation of any parameter.

Comment

The procedures described in this paper appear to have
limited applicability for finfish or shellfish fisheries.

A-171

Smith, V.L. 1969. On models of commercial fishing. Non-applicable
J. Polit. Econ. 77:181-198.

Abstract

Models based on a formulation of an economic theory of
fish production are developed. They are founded on three
key economic and technological features: (1) a fishery
resource is replenishable; (2) the resource and the activity
of production from it form a stock-flow relationship; (3) the
recovery or harvesting process is subject to various possible
external effects, all of which represent external diseconomics;
to the firm.. The author presents a generalization, explica-
tion and integration of previous works based on these char-
acterizations to provide one example of a descriptive theory
that transforms any specific pattern of assumptions about
cost conditions, demand externalities, and biomass growth tech-
nology into a pattern of exploitation.

Comment

This paper stresses.the economics of fisheries much
more than the dynamics of the exploited stock. It appears
inappropriate for inclusion in this phase of the study.

A-172

Southey, C. 1972. Policy prescriptions in bionomic models: Applicable
The case of a fishery. J. Polit. Econ. 80:769-775.

Abstract

Analysis of fisheries and their efficient exploitation
has [sic] been subject to successive refinements culminating in
the recent work of Smith'(1969). In the process of refine-
ment a number of conclusions as to the effects of regulation
have been modified. Nevertheless it would appear that at
least in steady-state models there is some agreement to the
effect that to achieve efficiency on a hitherto free-access
fishery, total fishing "effect" must indubitably be reduced,
and fish populations would probably increase and certainly
not be reduced.
The authors demonstrate that contrary to the above,
efficiency may involve a permanent increase in the total
expenditure on factors employed in the fishery. This
result is obtained without modification of the basic
ingredients of the fishery model. We shall also show
how the introduction of a somewhat different biological
mechanism allows for a decrease in the population under
regulation. Throughout our analysis, we abstract from
problems of mesh size and crowding, concentrating exclu-
sively on the effect of effort on population. We also
confine our attention to a steady-state solution.

Comment

This paper presents a rather simplified integration
of population dynamics and economic dynamics. It may be
useful as0a conceptual guide, but some unrealistic assump-
tions, may limit its applicability.

A-173

Southward, G.M. 1968. A simulation of management strategies Applicable
'in the Pacific halibut fishery. Int. Pac. Halibut Comm.
Rpt. 47:S-70.

Abstract

Simulation studies offer a mans of investigating dif-
ferent strategies that might be applied to the management
of the Pacific halibut resource. Since this study was
concentrated on long range aspects, it was necessary to
forego investigation of several interesting short-term
questions that arose. However, simulation does provide
a way of formulating hypotheses about relationships exist-
ing in the population such as density-dependent growth
responses to fishing, and offers a means of determining
the type and amount of field sampling necessary to study
the relationship in the population itself.
In general it appears that each of the three manage-
ment schemes would in the long run stabilize the stock of
halibut at a level where the maximum sustainable yield
would be obtained. However,, Rule 1, the empirical anal-
ysis, seems to be the preferable scheme from the stand-
point of small variance, in most of the situations
studied, the exception being the situation where the
recruitment is highly variable. In this case, Rule 2,
the potential yield analysis, results in a lower standard
deviation of the catch per unit effort.

C=L-nt

This paper presents one of the most extensive single
species man agement models found in the literature. The
model consists of three functioning submodels -- biological,
fishery (primarily economics), and management. The model,
although exemplary in its thorough analysis requires a very
C, 5P
large data base. A single species data base of this type
does not exist for any Maryland fishery; thus, the
direct application of the model is inappropriate. In con-
junction with new data collection programs, this type of
modelin- effort could be applicable to Maryland fisheries.

A-174

Sutcliffe, W.H. , Jr. , K. Drink-water, and B.S. 1"tuir. 1977. Applicable
Correlations of fish catch and environmental factors
in the Gulf of Maine. J. Fish. Res. Bd. Canada
34:19-30.

Abstract

In an investigation of catches of 17 commercial marine
species of fish and shellfish from the Gulf of MaineY 10
showed statistically significant correlations with sea temper-
atures at St. Andrews, N.B., or Boothbay Harbour, Maine. Most
fish records contained at least 40 years of data. Descriptive
equations are produced for four species based first on the
correlation between catch and sea temperatures and second on
the correlation between catch and sea temperature allowing
for fishing effort. Inclusion of fishing effort, not sur-
prisingly, improved the correlations for all of the species
so examined. Ilie equations permitted the "prediction" of
later parts of the records from earlier parts.
Considering the fish species collectively, the-Gulf of
Maine system from 1940 to 1959 appeared to be in equilibrium
with little fluctuation in the total commercial biomass.
We interpret the large fluctuations in individual species
abundance as resulting from a combination of fishing pressure
and to a significant degree oceanic climate as represented
by sea temperatures. The small fluctuations in the total
biomass display the species variation, with their differing
climatic "preferences," as well as possible predator (includ-
ing man) -prey relationships. Environmentally imposed
patterns underlie at least 50% of the fluctuations in catch
of many species and the understanding of these fluctuations
is basic to effective management.

Comment

This statistical model is similar in technique to others
using morphoedaphic variables; however, this example is in a
marine environment. It also includes application to benthic
fisheries such as hard and softshell clams. The paper should
be reviewed for application in this study.

A-175

Swartzman, G.L., and G.M. Van Dyne. 1972. An ecologically Applicable
I)ased simulation-@optimization approac1i to natural
1'(%SOUI'CC P1,11111-hig. Arui. Rev. Ecol. Syst. 3:347-398.

Abstract

The paper presents a simulation model of a perturbed
arid land ecosystem, with the natural ecosystem and sheep
introductions modeled together in a complex fashion.
Optimization methods are coupled to the model to aid in
decision making concerning the maintenance of the natural
ecosystem and the maximization of wool output.

Comwnt

The model presented in this paper represents a good
application of optimization techniques and the combination
of simulation and optimization problems. The data require-
ments for the model may make this approach unfeasible for
Maryland fisheries, but it should be reviewed as a holistic
approach tG the management of renewable resources.

A-176

Talbot, G.B. 1954. Factors associated with fluctuations in abundance
of Hudson Rdver shad. Fish. Bull. 56:373-413.

Abstract

Coments

This paper is unavailable for review, but has been
requested through interlibrary loan. It will be reviewed
when received.

A-177

Taut-7, A., P.A. Larkin, and W.E. Ricker. 1969. Some effects Non-applicable
of simulated long-term environmental fluctuations on
maximum sustained yield. J. Fish. Res. Bd. Canada
26:2715-2726.

Abstract

For a variety of stock recruit system , in which environ-
ment variability is simulated by random norma" -deviates used
as multipliers or divisors, Ricker (J. Fish. Res. Bd. Canada
1S:991-1006, 1958) and Larkin and Ricker (J. Fish. Res. Bd.
Canada 21:1-72 1964) demonstrated the beneTits of 'complete
@@t_abilization of escapement as opposed to removal of a fixed
proportion of the stock each year. The present paper is
primarily concerned with the response of these same systems
to a pattern of stochastic modification that is more regular
in form, a pattern such as might be imagined to result from
long-term trends in environmental conditions. The simulations
indicate that the gains derived from complete stabilization
of escapement are determined solely by the variance of the
modifiers -- thus the pattern of modification, i.e., long-
or short-term., is relevant only in terms of its influence
on the variance. In addition,, a check for maximum equilibrium
catch using catch statistics is described.

Comment

In this model, random an'd.cyclic events are included in
such a way as to be biologically unrealistic (e.g., conforma-
tion to a sine wave). The paper does represent one of the few
attempts to incorporate random environmental events into a pro-
duction model, but it probably will not be useful in this pro-
ject.

A-178

Timin, M.E., and B.D. Collier. 1971. A model incorporating Non-applicable
energy utilization for the dynamics of single species
populations. Theor. Pop. Biol. 2:237-251.

Abstract

A model for the population dynamics of a single species
of animals is developed. The model consists of three ordi-
nary differential equations expressing the rates of change of
the state variables, density of anLiials, mean biomass per
animal and food density, and several algebraic equations.
The model differs from previous models in that mean biomass
is used as an index of the size-age structure of the popula-
tion, energy utilization by the population is considered,
and a dimensionless form is developed. Terms are included
for food supply rates and harvest rates. The results of
computer simulations are presented and the properties of
the model., including its stability and sensitivity to
varying the parameters and initial conditions are discussed.

Comment

The paper presents a good approach to modeling population
dynamics, but the data requirements for its application to a
specific species are so large that use of this model is imprac-
tical for this project.

A-179

Toews, D.R., and J.S. Griffith. 1979. Empirical estimates Non-applicable
of potential fish yield for the Lake Bangweulu System,
Zambia, Central Africa. Trans. Am. Fish. Soc.
108:241-252.

Abstract

Estimates of potential fish yield derived for the Lake
Bangweulu system by three independent methods varied from
10 to 35 kg/hectare. An estimate of 20 kg/hectare predicted
by the regression of yield on morphoedaphic index (MEI) for
31 African lakes is in good agreement with the 1973-1974
estimated yield of 19 kg/hectare. Potential yield was 35 kg/
hectare in the morphoedaphic model for 17 intensively
exploited African lakes. Potential yield based on primary
production covariates was 16 kg/hectare. Standing crop
estimates indicated possible yields of 10 to 17 kg/hectare
based on 100% and 60% sampling recovery, respectively, of
fish from chemical treatment samples and a 50% annual ex-
ploitation rate.

.Commnt

This paper and method offer nothing new to the concept
of morphoedaphic indices (NEI). The concept and methodology
are clearly examined and explained in other papers which
have been reviewed.

A-180

Tomlinson, J.W.C., and P.S. Brown. 1979. Decision analysis Non-applicable
in fish hatchery management. Trans. Am. Fish. Soc.
108:121-129.

Abstract

Most hatchery management decision problems are "wicked,"
that is, they are multilevel and intellectually complex,
involve interactions between different areas of analysis, and
are often ill defined. Techniques of analysis commonly applied
to management problems -- operations research, cost-benefit
analysis, econometrics, Bayesian analysis -- are valuable,, but
limited. Basically, they are not decision-making techniques
but assist participants to assess a situation, before under-
taking the reconciliation process of arriving at decisions.
In complex decision problems, participants often revert to
"dodges and strategies" which attempt to minimize potential
regrets of overall optimality. In order to limit the extent
of ineffectual and varied suboptimizing procedures in hatchery
management it is important to clarify decision procediires
in the context of vertical and horizontal linkages on a system
analytical basis.

Comment

This paper does not deal with fishery yield models, but
rather reviews various techniques (e.g., Bayesian analysis)
used in hatchery management. It will not be useful to
this specific project, but may prove helpful in further
management endeavors.

A-181

TLIVIel', 1977. Intertidal ve,octation and commercial yields Non-applicable
of penaeid shrimp. Trans. Am' Fish. Soc. 106:411-416.

Abstract

A positive relationship is demonstrated for 27 locations
between commercial yields of penaeid shrimp per area intertidal
vegetation and latitude which can be described by the formula
y = 1,58.7e -0*070(x)
where y is kilogram/hectare and x is degrees latitude between
00 and 350. The latitudinal gradient grossly parallels a gra-
dient of heating-degree-days and is twice the slope of the
probable rates of litterfall from estuarine macrophytes. On a
regional basis, the yields inshore are directly related to the
area of estuarine vegetation whereas they are not correlated
with the area, average depth, or volume of estuarine water. A
short example is given to illustrate the utility of this analysis
for the selection of alternative land-use decisions.

Comment

This model is not useful to the project as it is specific
to shrimp and not precise enough to denote small-scale differences
in fishery yield.

.A-182

Ulltang, 1976. Sources of errors in and limitations of Applicable
virtual population analysis (cohort analysis). Int.
Coun. Explor. Sea 04/H:40.

Abstract

The virtual population analysis or cohort analysis is
extensively used in stock assessment both within ICES and
other scientific bodies. It is an extremely useful technique
for estimating past values of fishing mortalities and stock
sizes. These part values may in several ways be utilized
to get indications of the present state of the stock and the
prospects for the coming years. Because of the extensive use
of the method it is, however,*i mportant to know the limitations
of it and the various sources of errors.

Comment

This paper represents a useful review of virtual popu-
lation analysis and its limitations and assumptions. It will
be useful in this project.

A-183

11sliert Nlj@. 1970. Extelisi-Olls to 1110(.1cls, iised in renewable Non-appLicable
resource management, whi-ch incorporate an arbitrary
structure. J. Environ. Manag. 4:123-140.

Abstract

Matrix models of populations can be easily formulated,
simply computed, but have often been neglected in modeling
renewable resources. Two extensions to the basic model are
considered. First, what is the effect of error of estima-
tion of the matrix parameters on the eigenvalue and eigen-
vector? This is investigated by tivo theoretical methods,
one described for the first time, and by simulation of a.
matrix derived for a Blue whale population. Secondly, can
more meaningful criteria be advanced for optimizing yield?
A matrix model for a forest is extended so as not only to
investigate number of trees in different size classes but
also to give expressions for the volume increment and
economic increment of the forest as a whole.

Comment

This paper presents theoretical insights into the
utility of Leslie matrices for management models. It
presents no direct application, but provides theoretical
information concerning its applications. Its direct
applicability to this project is questionable.

A-184

V:m Wiiikle, W. , B.W. lZiLst, C.P. Goodycn', S.1@. Bliafl, and Applicable
11. Thall. 1974. A striped bass population model
and computer programs. Oak Ridge National Laboratory,
Oak Ridge, Tenn. ORNL/DI-4578. 200 pp.

Abstract

The population model consists of a system of difference
equations involving age-dependent fecundity and survival.
The model deals only with females. The fecundity for each
age class is assumed to be a function of both the fraction
of females sexually mature and the length of females as they
enter each age class. Natural mortality for age classes 1
to 15 and over is assumed to be independent of population
size. Fishing mortality is assumed to vary with the weight
of fish available to the fishery according to the logistic
relationship. The probability of survival for the y-o-y is
estimated without and with density-dependent mechanism
In the latter case, the entrainment period is considerea
separately from the remainder of the year. The probability
of survival after the entrainment period from all causes of
mortality other than impingement incorporates (a) canni-
balism of y-o-y striped bass by older striped bass and
(b) dependence of growth rate and, consequently, of mortal-
ity of y-o-y striped bass on availability of food.

Comment

This paper represents an application of the Leslie
matrix to model the population dynamics of a striped bass
population. It does not directly determine yield but it
could possibly be adapted for management -purposes. It
should be included in this project for thIs reason.

A-185

Wallis, I.G. 1975. Modelling the impact of waste on a Non-applicable
stable fish population. Water Res. 9:1025-1036.

Abstract

A simple model is developed to express the effect
of waste discharge on a stable fish population in a
randomly varying environment. The conditions for the
model to have a stable mean and variance are derived
analytically. Simulation studies are used to compare
exponential and logistic terms and three different
frequency distributions of environmental disturbances.
It is concluded that within the range of conditions
likely to represent real populations, the predicted
mean population is virtually the same for all of these
possibilities. The study suggests that simple models,
calibrated by field and laboratory determinations of
the effect of waste, can give order of magnitude esti-
mates of population changes resulting from waste
discharge.

Comment

This paper discusses a model of the effects of organic
waste pollutants on population dynamics. It appears to have
little direct applicability to the project, but may prove
useful in management practices which include habitat and
water quality management.

A-186

Walter, G.G. 1973. Delay -differential equation models for rApplicable
fisheries. J. Fish. Res. Bd. Canada 30:939-945.

Abstract

Wo new "simple" fishery models based on delay-differ-
ential equations are introduced and compared to three cur-
rently used differential equation models. These new models
can account for reproductive lag and allow oscillatory
behavior of population biomass, but require only catch and
effort data for their application. Equilibrium levels are
calculated for both models and examples of various types of
growth curves are given. Levels of fishing effort which
maximize yield are calculated and found in one case to
depend on the previous population and in the other to be
constant.

Comment

This paper presents a method by which the time lag
between spaiming and recruitment can be incorporated into
surplus production and yield models. The paper will be
useful in this project.

A-187

Walter, G.G. 1976. Non-equilibrium regulation of Applicable
fisheries. Int. Comm. Northwest. Atl. Fish.
Sel. P@2. 1:129-140.

Abstract

Many of the world's fisheries have until recently
been in a virgin state or close to it. The exploita-
tion has been marginal and has not severely affected
the stock size. In the last twenty years these fish-
eries have come under increasing exploitation, which
has often reduced the stock size and necessitated some
sort of regulation of the catch.
There are many theories available to the fisheries
biologist for this regulation. They usually give an
estimate for the maximum yield that the fishery can
produce on a continuing basis, i.e. the maximum sus-
tained yield (IMSY). It should be observed that this
value is not the maximum possible catch in a given
year but, particularly in a virgin fishery, is con-
siderably less.
The newly exploited stock is usually difficult
to manage. Few data are available and they are often
unreliable. Moreover, the stock is not in equilibrium
in the presence of fishing. This latter fact is often
overlooked and is one of the reasons for the shrinking
estimates of MSY that are sometimes encountered.
In this work we shall study the yield of a fishery
under nonequilibrium conditions and compare strategies
for bringing the stock size to that required for maxi-
mum sustained yield. We shall consider reduction to
this optimum stock size from above as well as increase
from below.
We introduce a procedure based on the equation of
Schaefer which assumes the growth rate of the total
stock biomass to be a function of the biomass itself
and of fishing effort. Although no delayed effects
are present in the equation, we shall see that there
is considerable delay between the initiation of a
regulation and the attainment of the desired equili-
brium state.

Comment'

This paper presents an excellent exp lanation of
the consequences of the application of equilibrium
surplus production formulations to non-equilibrium
fisheries. A method is developed which is applicable
to these situations and appears useable for Maryland
fisheries.

A-188

Waltcr@ 1978. A surplus yicJ6 model incorporating recruit- Applicable
inent @uid applied to a stock of Atlantic mackerel (Scomber
scombrus). J. Fish. Res. Bd. Canada 35:229-234.

Abstract

A modification of Schaefer's surplus yield model that takes
into account variations in year-class strength is introduced.
Expressions for long-term equilibrium yield under assumptions
of both linear and density-dependent recruitment are derived
and compared. Strategies for exploitation under nonequilibrium
conditions are discussed and equations derived@. The model is
fitted to a stock of mackerel and projections for the stock
biomass in 1980 under various levels of fishing mortality are
made.

Comment

The method presented here is potentially useful for
dominant year-class species such as striped bass or other
species where recruitment relationships are known from
historical data.

A-189

Walter, G., and W. Hogman. 1971. Mathematical models Applicable
for estimating changes in fish populations with
applications to Green Bay. Pages 170-184, In Proc.
14th Conf. Great Lakes Res. Great Lakes R6-sea=rc
Division, University oT-Tdc*-higan.

Abstract

A mathematical model for a multispecies exploited
fishery is developed. The model consists of a system of
first order differential equations similar to the single
equation for one species of Schaefer. The coefficients
of the equation are calculated from abundance and fishing
intensity data by using multiple regression. That level
of fishing intensity for each species which gives the
maximum sustained yield is then calculated. The model is
applied to the eight species fishery of northern Green
Bay using data obtained over a period of 41 years.

Comment

This paper presents an excellent application of
Schaefer-like yield formulations to multispecies fish-
eries in a semi-enclosed water body. It may be
directly applicable to Maryland fisheries.

A-190

Walters, C.J. 1969. A generalized computer simulation model Applicable
C,
for fish population studies. Trans. Am. Fish. Soc.
98:SOS-SI,2.

Abstract

A generalized computer model for fish population simula-
tion and maximum yield determination is described. The model
utilizes age-specific natural mortality rates, growth rates,
relative fecundities, and any desired stock-recruitment rela-
tionship. Best harvest strategies are found by treating long-
term yield as a response surface on the set of age- and year-
specific fishing rates. The model is illustrated using data
on arctic cod, stream brook trout) and on a hypothetical pop-
ulation with strong age-class dominance. Best predicted
management strategies include periodic harvest when age at
entry to the fishery cannot be controlled, but maximum yield
is usually obtained with constant fishing rate.

Comment

This model offers little new information concerning the
Beverton-Holt yield formulation but it does incorporate
age-specific recruitment, growth.,and mortality rates and
may be helpful in this project.

A-191

Walters, C.J., and J.E. Gross. 1972. Development of Non-applicable
big game management plans through simulation
0
modeling. J. Wildl. Manag. 36:119-128.

Abstract

Cur-rent and future demands-on wildlife resources
require greater levels of stewardship from the wildlife
manager. More complex demands and inevitable compro-
mises will require more sophisticated management plans
whose attributes are alternative paths of action and
estimates of the consequences. The core of needed
management plans is visualized as question banks and
data-processing models. Simulation models permit pre-
management experimentation in terms of what if games.
Examples of what if games are discussed to illustrate
critical population conditions, sensitive management
parameters, alternative objectives, consequences of
environmental catastrophes, and procedures for devel-
oping objective measures for management performance.
This paper attempts to show how information generated
from a complex of variables can be channeled into the
decision-making process.

Comment

This paper presents the application of a Schaefer-
like logistic model to white-tail deer in Texas. The
specific application to a non-fish population adds
little to the development of the surplus production
concept of management. The model presented is not
directly applicable to Maryland fisheries.

A-192

Welcomme.t R.B. 1976. Some general and theoretical consider- Applicable
ations on the fish yield of African rivers. J. Fish. Biol.
8:351-364.

Abstract

The factors regulating fish production from river systems
remain poorly studied and understood. Rivers conform to physical
and chemical laws which determine their morphology. From these
laws relationships are calculated ivhich estimate the total length
and number of streams of different order on the African continent.
Edaphic factors vary less than morphological ones and the chemical
and physical conditions in the major river channels tend to
resemble each other closely. The present catch from African rivers,
evaluated from catch statistics, by country and by river system,
resembles a theoretical figure derived from the basin area. How-
ever, these statistics are drawn only from major fisheries and
there remain a very large number of smaller streams whose pro-
duction does not enter into this calculation. A theoretical
approach to this problem is proposed which gives an estimate of
annual yield of 530,000 ton of fish at present levels of catch.
Deviations from the theoretical yield in individual river systems
arise from differences in both edaphic and morphological char-
acteristics.

Comment

This paper may represent the most useful of the morphoedaphic
index (MEI) approaches presented in numerous other papers.
The approach described develops a MEI for river systems as
opposed to lakes. The use of total dissolved solids is less
reliable in river systems than lakes, and, thus, the primary
statistical relationships are constructed between yield and
drainage area,and yield and river length.

A-193

Winters, G.H. 1976. Recruitment mechanisms of southern Gulf Applicable
of St. Laivrence Atlantic-herring (Clupea harengus harengus).
J. Fish. Res. Bd. Canada 33:1751-1-763.

Abstract

Estimates of abundance of the southern Gulf of St. Lawrence
herring complex by cohort analysis indicate that both biomass
and population fecundity were reduced to low levels in the late
1950s following a widespread fungus disease in the mid-1950s.
As a result of tivo strong year-classes in the late 1950s, how-
ever, abundance increased dramatically up to 1964 but has
declined continuously since then due mainly to subsequent poor
recruitment. Mackerel were also at low levels of abundance in
the late 1950s and remained so until the mid-1960s when a series
of strong year-classes produced a rapid increase in abundance
to the extent that mackerel replaced herring as the dominant
pelagic fish in the southern Gulf ecosystem. Changes in her-ring
recruitment, growth, and maturation rates are investigated in.
relation to changes in herring biomass and total pelagic
(herring + mackerel) biomass. Density-dependent changes in all
three parameters have occur-red in herring; mackerel also have
interacted with the growth and recruitment of southern Gulf
herring. This sua ests that the carrying capacity of the
C19
southern Gulf for pelagic fish is limited and that competition
and predation by mackerel intensifies [sic] the logistic response
of herring. Thus, recruitment of southern Gulf herring in the
period under consideration was largely controlled by the total
pelagic biomass, acting mainly through herring up to the mid-
1960s and through mackerel since then. The increase in mackerel
abundance is attributed to a combination of favorable tempera-
ture regime and optimum spawning biomass. That being so, the
probability of large year-classes of her-ring in the near future
will depend heavily on the interaction between favorable sur-
vival conditions for mackerel and the effectiveness of IMAF
regulations to maintain mackerel biomass at maximum recruitment
levels.

Comment

This paper describes an application of cohort analysis
to an existing fishery and will prove useful to the project
for species where adequate data have previously been collected.

A-194

Winters, G.H. 1978. Production, mortality, and sustainable Applicable
yield of Northwest Atlantic harp seals (Pag2Philus
groenlandicus). J. Fish. Res. Bd. Can- 35:1249-1261.

Abstract

From recent and historical data the natural mortality rate
of adult harp seals (Pago@hilus groenlandicus) is estimated to
be 0.10 which is within the range of prievious estimates (0.08 -
0.11). New estimates of bedlamer and O-group natural mortality
rates were not significantly different from those of adult seals.
Pup production estimates from survival indices agreed well with
those from sequential population analyses and indicated a decline
from about 350,000 animals in the early 1950s to about 310,000
animals in the early 1970s. Over the same period the 1+ popu-
lation size declined from 2.5 to 1.1 million animals but has
been increasing at the rate of 3% per year since the introduction
of quotas in 1972. The relative contribution of the "Front"
production of total ("Front" plus Gulf) production during the
past decade has fluctuated from 49 to 87%, the average of 64%
being very similar to the 61% obtained previously. These
fluctuations suggest some interchange between "Front" and
Gulf adults and it is concluded that homing in the breeding
areas is a facultative rather than obligatory aspect of seal
behavior. Thus the heavier exploitation of the "Front" produc-
tion is probably sufficiently diffused into the total population
to avoid serious effects on "Front" production. The maximum
sustainable yield of Northwest Atlantic seals harvested accord-
ing to recent patterns is estimated to be 290,000 animals
(80% pups) from a 1+ population size of 1.8 million animals
producing 460,000 pups annually. The sustainable yield at
present levels of pup production (33S,000 animals) is cal-
culated to be 220.,000 animals which is substantially above
the present TAC of 180,000 animals and coincides with pre-
sent harvesting strategies designed to enable the seal hunt
to increase slowly towards the IMSY level.

Comment

The paper represents the combination of cohort analysis
for the determination of parameters and yield and a Schaefer
model for determinations of surplus yield, while the appli-
cation described (i.e., to seals) is inappropriate to the
project, the utility of the combination of these two methods
should be reviewed.

A-195

APPENDIX B

An annotated bibliography of stock management submodels

and methods of parameter estimation

B-1

Age Structure Submodels

Kimura, D.K. 1977. Statistical assessment of age-length key. J. Fish.
Res. Bd. Canada 34:317-324.

Abstract

Since 1934 when Fridricksson originated the age-leng1th key, it has
been widely used by fisheries biologists to estimate age distributions
of populations. In recent years, there has been a general recognition
that often the key has little value, or even worse, gives biased results.
The analysis presented here indicates why the age-length key is so sus-
ceptible to bias. More importantly, a criterion is presented for deter-
mining whether the age-length key should be used in a particular situation.
If the key is to be used, results from examples indicate that random age
subsamples (i.e.,the number of specimens aged from each length category
proportional to the number in each length category) are superior to fixed
age subsamples (i.e.,a constant nunber of specimens aged from each length
category). Generally, small increases in the age sample will likely
increase the accuracy of an age-distribution determination more effec-
tively than relatively large increases in the length sample.

Comment

This paper discusses the bias introduced to estimated acre
distribution through the use of an age-length key and suggests
ID C,
a test for the determination of the key's applicability to spe-
cific investigations. Due to the bias generally involved with
the use of age-length keys, their use is not suggested for this
project.
Kunar, K.D... and S.M. Admns. 1978. Estimation of age structure of fish
populations from length-frequency data. Pages 256-281 In W. Van
Winkle (ed.), Assessing the Effects of ]Row r-P_lant-IndL0;J lybrtality
on Fish Populations. Sponsored by Oak Ridge National Laboratory,
Energy Research and Development Administration, and Electric
Power Research Institute.
Abstract

A probability model is presented to determine the age structure
of a fish population from length-frequency data. It is shown that
when the age-length key is available, maximum-likelihood estimates of
the age structure can be obtained. When the key is not available,
approximate estimates of the age structure can be obtained. The model
is used for determination of the age structure of populations of chan-
nel catfish and white crappie. Practical applications of the model to
impact assessment are discussed.

Comment

This paper presents a good review of the methods for the deter-
mination of population age structure from length-frequency data. The

B-2

assumptions of the normality of length data and the prior existence
of an age-length key may make this procedure less desirable for use
for @4aryland stocks, but'even without the age-length key, first
approximations are possible. The approach will prove useful for
stocks where a stable age distribution does not exist.

McINIew, R.W., and R.C. Summerfelt. 1978. Evaluation of a maximum-likeli-
hood estimator for analysis of length-frequency distributions. Trans.
Ain. Fish. Soc. 107:730-736.

Abstract

The maximum-likelihood estimation procedure described by Hassel-
blad is a statistical method applicable to estimates of population para-
meters in a mixture of normal distributions of component age-groups.
The method was used to estimate mean lencyth-at-age and percentage
composition of the component age-groups in 10 collections of largemout
bass (IMicropterus salmoides) for which age was determined by the scale
method. Compared to fish aged by the scale method, the error of the
estimates of mean length-at-age averaged 3.2%. About one-third of the
31 frequency distributions, and six of seven distributions with more
than 100 fish, deviated significantly from that of a normal distribu-
tion; many distributions exhibited skewness and kurtosis. However,
the general failure of the samples to fit characteristics of the normal
curve did not greatly influence accuracy in estimating mean length.
The average error of the estimates of percentage composition by age
was 28%; the magnitude of this error was related to the degree of
asymmetry and a large standard deviation of length. The latter was
apparently related to a prolonged and disjunct spawning season which
C,
produced multimodal distributions within each age-group.

Comment

This paper presents an established method of determining age
0
structure from length-frequency data. The assumption of normality
.in the distribution of the length data was tested for reservoir bass
and found to not significantly affect the projected age composition.
This assumption should be tested on any Maryland species on which
this procedure might be utilized in this project.

Ivestrheim S.J. and W.E. Ricker. 1978. Bias in using an age-length key
31 .9 0
to estimate age-frequency distributions. J. Fish. Res. Bd. Canada
35:1847189.

Abstract

Consider two representative samples of fish taken in different
years from the same fish population, this being a population in which

B-3

year-class strength varies. For the "parental" sample the length and
age of the fish are determined and are used to construct an "age-
length key," the fractions of the fish in each (short) length interval
that are of each age. For the "filial" sample only the length is
measured, and the parental age-length key is used to compute the cor-
responding age distribution. Trials show that the age-length key
will reproduce the age-frequency distribution of the filial sample
without systematic bias only if there is no overlap in length between
successive ages. Where there is much overlap, the age-length key will
compute from the filial length-frequency distribution approximately
the parental age distribution. Additional bias arises if the rate of
growth of a year-class is affected by its abundance, or if the survival
rate in the population changes. The length of the fish present in any
given part of a population's range can vary with environmental factors
such as depth of the water; nevertheless, a sample taken in any part
of that range can be used to compute age from the length distribution
of a sample taken at the same time in any other part of the range,
without systematic bias. But this of course is not likely to be true
of samples taken from different populations of the species.

Comment

The paper does not present a submodel but only discusses the
bias introduced through the use of age-length keys. This procedure
appears to be particularly biased in cases where variation in year-
class strenath affects growth rates. Age-length keys do not appear
to be a useful tool in this project.

Allometric Submodels

Dame, R.F. 1972. Comparison of various allometric relationships in inter-
tidal and subtidal American oysters. Fish. Bull. 70:1121-1126.

Abstract

The allometric relationships for the possible combinations of whole
weight, dry body weight, soft body weight, shell weight, height, and
length were computed for intertidal and subtidal South Carolina oysters.
All relationships between intertidal and subtidal oysters involving dry
body weight were significantly different. The percent moisture in the
tissues was 81.1% for subtidal oysters and 83.4% for intertidal oysters
and did not vary with size. Height appears to be the most useful para-
meter for predicting other biomass parameters from field data.

Comment

This paper presents the allometric coefficients for several vari-
ables (e.g., shell weight, height, dry body weight, soft body weight)
for intertidal and subtidal oysters. The procedure is straightforward
and can be used in some cases to estimate biomass for simulation models.

B-4

Pienaar, L.V., and J.A. Thomson. 1969. Allometric weight-length regression
model. J. Fish. Res. Bd. Canada 26:123-131.

Abstract

Wo methods of fitting the allometric weight-length relationship
are described.; one involving the common logarithmic transformation of
variables in a multiplicative model.Pand the other assuming an additive
nonlinear model and general nonlinear estimation procedures. Differences
in the assumptions involved in the two methods are emphasized and the
practical significance of the different methods is demonstrated with the
aid of a sample problem. A number of procedures are suggested to com-
pensate for possibly unjustified assumptions.

Comment

This paper presents several statistical procedures for the esti-
mation of allometric coefficients. The information presented allows
determination of the procedure to be used under certain circumstances.
The allometric equation would probably not prove useful in this pro-
ject unless biomass data were non-existent.

Catchability Submodels

Paloheiro , J.E. 1961. Studies on estimation of mortalities: I. Conparison
of a method described by Beverton and Holt and a new linear formula.
J. Fish. Res. Bd. Canada 18:645-662.

Abstract

A method., described by Beverton (1954) and Beverton and Holt (1956
and 1957), giving estimates of the natural mortality rate, M, and the
catchability coefficient, q, from catch at age and effort data, is
examined. This method requires 4 to 5 iterations to arrive at the esti-
mates. We have derived approximate solutions for q and M in a closed
form. This makes the laborious iterations unnecessary, and gives
virtually the same values as arrived at by iterations.

The effectiveness of the iterative Beverton and Holt method is
evaluated by calculating q and M in 30 hypothetical examples. A new
and simple Clinear formula) method for estimating q and INI is derived.
Application of the new method to these 30 examples resulted in a 48%
reduction of the standard deviation of q and a 45% reduction in that
of M. The new method is in part the same as one suggested by Gulland,
Beverton, and Holt (Beverton et al., NIS, 1958; Holt,'INIS, 1959) to
arrive at initial values in their short-cut (iterative) method of
estimating the mortality rates. We show t:hat these initial values
are actually better estimates than the final values arrived at by the
iteration.

B-5

Neither the Beverton and Holt method nor the linear formula
gives necessarily unbiased estimates; the bias depencLs oil the types
of variability in the data.

To arrive at non-biased, least squares estimates would require
ancillary information not normally available on the distributions of
the three variates: catch at age, effort, and catchability coefficient.

Comment

This paper presents a method of estimating natural mortality and
catchability which simplifies an earlier Beverton-Holt method. The
s* lification requires primarily effort data and could be useful
IMP
for Maryland stocks.

Rafail, S.Z. 1977. A simplification for the study of fish populations by
capture data. Fish. Bull. 75:561-569.

Abstract

F_xpressions given by Rafail for estimating catchability are modi-
fied here to eliminate iteration, for better accuracy, and a large
C,
economy in calculations and time. The evaluation of catchability
allows the estimation of other important parameters with the useful
assumption of their variabilities according to seasons and recognized
sections of a population.

Comment

This paper presents a method for the estimation of catchability giv-
en catch per unit effort and a series of effort data. The procedure
for the calculation of the true value is complicated and requires data
not likely to exist for Maryland fisheries. The estimation procedure
may prove useful for this project.

Effort Submodels (Commercial)

Nicholson, W.R. 1971. Changes in catch and effort in the Atlantic menhaden
purse-seine fishery, 1940-1968. fish. Bull. 69:765-781.

Abstract

The catch, number of vessel weeks, and catch per vessel week in
the Atlantic menhaden fishery increased during the 1950's. During this
period fishing methods improved and the efficiency of vessels increased.
Improvements included use of airplanes for spotting schools, aluminum
purse boats, nylon nets, power blocks, and fish pumps for catching and

B-6

handling fish, and larger and faster carrier vessels that could range
farther from port. The catch and catch per vessel week began declining
north of Chesapeake Bay in the early 1960's. By 1966, fish north of
Chesapeake Bay had become so scarce that plants either closed or oper-
ated far below their capacity. In Chesapeake Bay the number of vessel
weeks increased, and the catch and catch per vessel week decreased
through the early and mid 19601s. Variations in catch, e@fort, and
catch per unit of effort showed no trends in the South Atlantic. The
annual mean number of purse-seine sets per day varied in different areas
-md ranged from about 2.0 to 4.5. The annual mean catch per set
ranged from about 11 to 25 metric tons.
;;F
Comment

The paper presents a method for the calculation of total effort
and a "normaliz ing" fixiction J'or a fishery where technical improvement
has increased effort and catchability at the same time. This approach
should be applied to all long-term simulations using historical
data for 14aryland stocks.

Schaaf, W.E., and G.R. Huntsman. 1972. Effects of fishing on the Atlantic
menhaden stock 1955-1969. Trans. Am. Fish. Soc. 101:290-297.

Abstract

To determine the effect of purse-seine fishing on the Atlantic
menhaden (Brevoortia tXrannus) population, we analyzed data from 1955-
1969 on fishing activity and catches. Chances in fishing efficiency
0
necessitated establishment of an abstract effort unit, the 1965 vessel-
week. Catch per unit of effective effort in 1965 was one-fifth that
of 19SS. Our instantaneous natural mortality rate (M) estimate was
0.37. At the current recruitment age, 1.5 years, reducing F, instan-
taneous fishing mortality, to about 0.8 would slightly decrease the
yield per recruit, but would increase the spawning stock and ultimately
allow annual catches of 400,000-500,000 metric tons, the maximum sus-
tained yield.

Comment

This paper presents a useful and easily applied method for "nor-
malizing" fishing effort which is dependent on relative catchability.
This method could be important for all Nfaryland fisheries in which
effort and stock size were significantly altered over the past decade.

Effort Submodels (Recreational)

plalvestuto, S.P., W.D. Davies, and W.L. Sheldon. 1978. An evaluation of
the roving creel survey with nonuniform probability sampling. Trans.
Am. Fish. Soc. 107.:255-262.

B-7

Abstract

A roving creel survey with nonuniform probability sampling was
conducted on West Point Reservoir, Georgia, for 24 months. The sampling
desig.
,n is described in detail. The assumption that catch per unit
effort (CPE) for incompleted fishing trips is an unbiased estimator
of CPE for completed trips is tested and verified. Coefficients of
variation for monthly estimates of catch and effort are used to measure
the precision of the sampling design. Precision was relatively high
during the summer (April-October), but decreased markedly during the
winter (November-March). This change is largely independent of sample
size within the range of 5-10 sample days per month leading to the
conclusion that sampling effort could be reduced 50% without impairing
the precision of the survey. The method appears capable of detecting
changes in the quality of fishing small enough for management purposes.
The paper is intended to provide guidelines for the implementation,
evaluation, and modification of statistically based creel survey pro-
grams.

Comment

This paper presents a complete sampling design and analysis pro-
cedure for the determination of recreational fishing effort. The
present recreational fishing survey being conducted in Maryland will
make the use of this procedure unnecessary, but further studies of
this type should consider the methodology explained in this paper as
well as the companion paper by the same authors.

Malvestuto, S.P., W.D. Davies, and W.L. Sheldon. 1979. Predicting the
precision of creel survey estimates of fishery effort by use of cli,-
matic variables. Trans. Am. Fish. Soc. 108:43-45.

Abstract

A multiple regression equation was developed which explained 83%
of the variation in the precision (CVE).of monthly creel survey e_@timates
of fishing effort on West Point Reservoir, Georgia-Alabama, over a
24-month period. The equation is CVE = 34.678 - 1.154(MI) + 9.274(SI)T)
+ 84.942 CRAIN) - 38.854(SDR), where M = mean daily air temperature
for a particular month; RAIN = mean daily rainfall for that particular
month; and SDT and SDR = the standard deviations of daily air temperature
and rainfall for that same month, respectively. The constants in the
regression equation may be specific to West Point Reservoir, but the
climatic variables are certainly of general importance. The model re-
flects the environmental basis for the variation associated with creel
survey estimates of fishing effort, and provides a means of optimally
allocating sampling effort prior to implementation of a survey.

Comment

This presentation is completely site specific, but the generality
ofclimactic variables affecting the precision of survey estimates of
recreational effort should be investigated in light of the present
IMaryland survey. B-8

Robson, D.S. 1961. On the statistical theory of a roving creel census
of fishermen. Biometrics 17:415-437.

Abstract

In order to estimate the day's total catch from a fishery an
enumerator roves through the fishing area interviewing fishermen as
he encounters them to determine the number n of fish caught and the
time t expended. The interviewer is assumed to (i) start his trip
at a randomly chosen point along a well defined route which com-
pletely covers the fishery, (ii) choose his initial direction at
random from the two alternatives, and (iii) travel at a constant
rate of c circuits per day. If the catch rate n/t at time of inter-
view is an unbiased estimator of a fisherman's cat ch rate for his
completed trip and if the fishermen's movements relative to the
interviewer's path never exceed the interviewer's rate c, then
rn/ct, sumned over all interviews, is an unbiased estimator of the
day's total catch. The unit of time is one day, r is the number of
times the fisherman was interviewed, and n/t is the catch rate at
the r1th interview.

Unblasedness of n/t implies that the waiting times to first
catch and from first to second catch are identically distributed
chance variables, and that all waiting times between successive
catches have the same expected value. If waiting times are inde-
pendent, then unbiasedness implies that fishing is a Polsson process.

Comment

This paper presents,a complete samplin g design and analysis
scheme for determining recreational fishing effort. The Coastal
Zone Management Unit is presently involved in a complete recrea-
tional effort survey which will make the use of this presentation
unnecessary.

Fecundity Submodels

Brousseau, D.J. 19"78a. Population dynamics of the soft-shell clam, Mya
arenaria. Mar. Biol. 50:63-71.

Abstract

A life table was constructed for NVa arenaria from Gloucester,
Massachusetts, USA, based on schedules of age-specific,fecundity and
mortality determined under natural conditions. Mortality rates de-
crease with size and age in this species, with the period of maximum
mortality occurring during the summer months. Mortality rates during
the fall and winter were considerably lower, perhaps due to the inac-
tivity of natural predators. The survivorship curve for M. arenaria

B-9

approximates the Type 3 curve of Deevey (1947). Ivlean life expectancy
is low in recently-settled clams, peaks when the individual reaches
30.0 to 34.9 mm (1 year of age), and remains fairly high for most of
@he reminder of life. The intrinsic rate of natural increase (rmax
is very high: 4.74. This enormous rate of potential increase is
offset by high rates of larval mortality in the plankton. Unlike the
reproductive values of most animals studied, those in M. arenaria peak
late in lifeP well after the known age of first reprod-5-Etlo-H-7-Mis is
probably the result of increased fecundity with age. The inplications
of this work in the area of resource management are discussed.

Comment

This paper presents the use of life tables to determine the intrin-
sic rate of natural increase (r). Size-specific natality and mortality
were converted to age-specific rates using the von Bertalanffy age-size
relationship. This procedure will be of little use for this project.

Brousseau, D.J. 1978 b. Spm@au* ng cycle, fecundity and recruitment in a
population of soft-shell clam, Mya arenaria, @rom Cape Ann, Massachu-
setts. Fish. Bull. 76:155-166.

Abstract

A population of Mya arenaria in the Annisquam River system,
Gloucester, Mass., was studied fo 3 years to determine spaiming
frequency, fecundity, and recruitment rates under natural conditions.
This population was observed to spawn twice each year, in March-
April and June-July. Temperature appeared to be a more critical
factor in the timing of gonad maturation than in triggering the
release of gametes. Female body sizes and oocyte production were
positively correlated (1973, r = 0.95; 1974, r = 0.90). Regression
lines were compared by analysis of covariance. Slopes of the lines
did not differ significantly between years or between spawning cycles
within years (P > 0.05). Elevations of the lines differed signifi-
cantly from one another (P @@ 0.05) indicating annual and seasonal
variability in fecundity. Sex ratios of M. arenaria 2S-9S mm shell
length did not differ significantly from !71 over th 3-year study
period. In smaller individuals, male and female gonads were indis-
tinguishable. No evidence of hermaphroditism or protandry was
observed. Recruitment -rates of juveniles fluctuated widely between
spawning cycles as well as between years.

Coment

This paper presents a very simple statistical linear relation-
ship between female shell length and oocyte number. This approach
could easily be enployed for Moaryland stocks if fecundity was deter-
mined a necessary component of a simulation model.

Tsai, C., and G.R. Gibson., Jr. 1971. Fecundity of the yellow perch, Perca
flavescens Kitchill., in the Patuxent River, Nlaryi&ici. Ches. Sci.-
12:270-27T.

B-10

Abstract

The fecundity of 114 female yellow perch was studied during the
1969 spawning run (20 through 25 March) in the Patuxent River, Maryland.
The fecundity was proportional to the total weight and gutted weight,
respectively,as expressed by the regression equations Y = 150.5573OX
- 1,424.0878 (r = 0.88) for the former, and Y = 217.65656X -
2,387.9915 (r = 0.87) for the latter (Y, fecundity; X, either total
m
weiffit or gutted weight). It was nearly proportional to the qua-
icate of fork length as expressed by the regression equation, log
Y = 3.7179 log X - 1.50917 (r = 0.90). This quadruplicate phenomenon
was due in part to the heavier proportionate increase in body weight
and the visceral space available for egg development as the fish length
increased.

Comment

This paper presents simple regressions of fecundity, total weight,
gutted weight, and fork length for yellow perch. The approach and/or
actual numbers may prove useful if yellow perch yield is evaluated with
a simulation model.

Groi,rth Submodels

Allen, K.R. 1966. A method of fitting growth curves of the von Bertalanffy
type to observed data. J. Fish. Res. Bd. Canada 23:163-179.

Abstract

A method is described for obtaining the best least-squares esti-
mates of the parameters L.., k, and to when von Bertalanffy curves of
the type
L 1. (1 - e-k[t-to])

are fitted to observed data. This method imposes no restrictions on
the number or size of the samples or on the time intervals between them.
It also provides estimates of the limits of error of the parameters.
The amount of computation is fairly large, but a method of systema-
tizing it is described which makes manual computation practicable for
moderate-Si7ed sets of data. The method has been used to develop a
computer program which seems to have advantages over some existing
methods.

A numerical example is worked out in full to illustrate application
of the method.

B-11

Comment

This paper presents an excellent method for the computation of
von Bertalanffy growth parameters by a least-squares technique regard-
less of sarple size.Equations are presented to determine the error of
the growth parameter estimates. This method can be easily applied
without the use of a computer.

Bayley, P.B. 1 977. A method for finding the limits of application of the
von Bertalanffy growth model and statistical estimates of the parameters.
J. Fish. Res. Bd. Canada 34:1079-1084.

Abstract

The following expression is derived expressing the instantaneous
growth rate (G) in terms of length (L),the power of the weight-length
relationship (b), and the Bertalanffy growth parameters (K and L-):

G = bK(L-/ L - 1) .

Since this is valid for a length (or age) range in which growth conforms
to the Bertalanffy model, a plot of G vs l/ Lshould be linear with the
intercept on the G axis being -bK and on the l/ Laxis being 1/L-. Since
the variables can be measured independently, deviations of points from
the regression can be tested and the limits of validity of the model
ascertained. In addition,, confidence limits of K and L- can be estimated.
Two examples compare results with those using previous methods.

Comment

This paper presents an extension of the von Bertalanffy growth
equation where the parameters which determine K can be measured inde-
pendently. It can be easily applied,and an estimate of the variance
associated with the growth parameter CK) is presented. It could be
applied to Nlaryland species especially where growth records available
@Lre for unequal time periods.

Brousseau, D.J. 1978. Population dynamics of the soft-shell clam, iklya
arenaria. Mar. Biol. 50:63-71.

Abstract

A life table was constructed for IN-Va arenaria from Gloucester,
Massachusetts, USA, based on schedules oT -age-specific fecundity and
mortality determined under natural conditions. Mortality rates de-
crease with size and age in this species, with the period of maximum
mortality occurring during the summer months. Mortality rates during
the fall and winter were considerably lower, perhaps due to the inac-
tivity of natural predators. The survivorship curve for M. arenaria
approximates the Type 3 curve of Deevey (1947). Mean life_765@Fe_ctancy

B-12

i,s low iii ctuiis, 1)caks wfieri the in(lividual reaclies
30.0 to 34.9 mm (I year of age), and remains fairly high for most of
the remainder of life. The intrinsic rate of natural increase (r
is very high: 4.74. This enormous rate of potential increase is Inax
offset by high rates of larval mortality in the plankton. Unlike the
reproductive values of most animals studied, those in M. arenaria peak
late in life, well after the known age of first reproduction. Th s is
probably the result of increased fecundity with age. The implications
of this work in the area of resource management are discussed.

Comment

Brousseau, D.J. 1979. Analysis of growth rate in Mya arenaria using the
von Bertalanffy equation. Mar. Biol. 51:221-2277.

Abstract

Field studies were conducted in Gloucester, Massachusetts, USA, to
determine linear shell growth rates for Mya arenaria. These rates,were
then compared with those reported for the same species from other loca-
tions. Most shell deposition occurred from March through November of
each year. Winter interruptions in growth were not as marked in the
small clam as in the larger ones (> 60.0 mm). Annual variations in
growth were slight during the period 1973-1974. Growth of mature clams
(> 35.0 mm) slowed during the spawning season. No significant sexual
dimorphism in mean annual growth rates was detected. Winter rings were
shown to be a reliable method for determining age in clams from Glou-
cester. Age-size relationships, based on two independent measures of
annual growth, winter rings and tagging experiments, were computed using
the von Bertalanffy growth equation. No well-defined latitudinal pat-
terns in growth could be established for M.'arenaria.

Comment

This paper presents the application of the von Bertalanffy growth
function to several populations of soft-shell clams. The estimates
presented may be helpful in determining if growth rates assessed for
Maryland populations are realistic. No new methods are presented.

M.P. T.R. Porter and P. Downton. 1978. Analysis of growth
Chadwick, E., .9 31 Z>
of Atlantic salmon (Salmo salar) in a small Newfoundland river.
J. Fish. Res. Bd. Cana.TF3ST-07-68-

Abstract

Growth and sea su rvival rates decreased with increasing smolt age,
with survival being 12, 6, and 3% for 3+, 4+, and S+ smolt, respectively.
B-13

All spawiiirii,, fish were gri.1se, which suggests that older smolt became
large Salmon' zuid were thus more vulnerable to the commercial fishery.
A density-dependent relationship was observed for 3+ sinolt in their
Ist year of growth, but not for older smolt; younger smolt probably
spend their juvenile life in a more productive but space-limiting part
of the river. Variation between river-system environments may be
responsible for the opposing results of studies on Atlantic salmon
(Salmo salar) life history.

Comment

This paper explains a method for the determination of instantaneous
growth rates from back-calculated lengths (as determined by a regression
of fork length on scale radius). The presentation provides little new
information, but may prove useful for species which do not spend their
entire life span in Maryland waters.

Cloorn, J.E., and F.H. Nichols. 1978. A von Bertalanffy grow-th model with
0
a seasonally varying coefficient. J. Fish. Res. Bd. Canada 35:1479-1482.

Abstract

The von Bertalanffy model of body growth is inappropriate for
organisms whose groi%th is restricted to a seasonal period because it
assumes that growth rate is invariant with time. Incorporation of a
time-varying coefficient significantly improves the capability of the
von Bertalanffy equation to describe changing body size of both the
bivalve mollusc, Macoma balthica, in San Francisco Bay and the flathead
solel Hippoglosso17_esel`as`-s`o1o_n, in Washington state. This simple
modification of the von Bertalanffy model should offer improved pre-
dictions of body growth for a variety of other aquatic animals.

Comment

This paper presents a simple extension of the von Bertalanffy
growth equation incorporating seasonal variation into the growth co-
efficient, K. It is easily applied to any yield model incorporating
von Bertalanffy growth. It could be useful in determining growth
functions forl4ary-land species where growth data are available over several
years.

Dame, R.F. 197S. Day degree growth models for intertidal oysters. Contr.
Z_
Mar. Sci.- 19:107-112.

Abstract

Linear and polynomial day degree models are shoun to be capable
of predicting oyster growth in terms of height or weight. Polynomial
models predict growth more accurately than linear models, but are less
meaningful biologically. A general linear model for predicting oyster
growtJi in terms of weight is developed for the North Inlet, South
Carolina area. B-14

Comment

This paper presents very simple linear and quadratic growth
CO
equations for oysters based on initial size and day-degrees (mean
temperature multiplied by number of days in period). Growth is easily
determined and may prove useful for non-motile species.
.. I
DeAngelis, D.L., and C.C. Contant. 1979. Growth rates and size distri-
butions of first-year smallmouth bass populations: Some conclusions
from experiments and a model. Trans. Am. Fish. Soc..108:137-141.

Abstract

Growth rates of populations of first-year smallmouth bass (i"Ucrop-
terus dolomieui Lac'epede) were studied during the first few week's-o-f-
life at temperatures ranging from 15.2 to 32.5*C. In all cases, the
average length of fish in each group increased linearly with time, t,
in a range from 10 to 30 mm. The variances about these mean lengths
increased approximately as t2. A partial differential equation model
can be useful in expressing the dynamics of populations in which size
distributions are taken into consideration. Applied to our experiment,
this type of model shows that both of the observed length-versus-time
phenomena are expected if the rates of increase of length of fish are
independent of length and that these rates for individual fish are
normally distributed about some mean rate of growth. The variance
of individual growth rates needed to produce the observed length-
versus-time data can be calculated from the model.

Comment

The paper presents a complex growth function (length) which
accounts for mean individual growth and variance during early larval
stages. While the model appears to be applicable to Maryland species
and accounts for variability, the lack of the inclusion of growth
rates at later stages may complicate its general use.

Gallucci, V.F., and T.J. Quinn. 1979. Reparameterizing, fitting, and
testing a simple growth model. Trans. Am. Fish. Soc. 108:14-25.

Abstract

If the von Bertalanffy growth model is used to statistically com-
pare the properties of growth in two spatial regions by examination of
the estimates of the growth parameter k and the asymptotic length
parameter L,,,,, a possible compound null hypothesis Holp is H - kl = k2
and L., = L-2 for regions 1 and 2. Since the results of tgl:s two-
parameter test may be difficult to interpret, an alternative procedure
is suggested. In addition, the interpretation of the test must be
based upon the nature of the data as well as upon the parameter estimates.

B-15

A regression fit of the model to real but "inappropriate" data may
yield a very "good" statistical fit but unrealistic estimates. The
use of a third parameter (for example, to, the time when length is
zero) is necessary to uniquely specify a solution; its inclusion
always enhances the statistical fit. Because of the interdependence
between parameters k and L,,., we reparameterize the von Bertalanffy
model with a new parameter w = k-L.. The parameter corresponds to
the growth rate near to and is suitable for comparisons because of
its statistical robustness. In general, the standard Ford-Walford
method of estimating the model's parameters is now obsolete and
should be replaced with widely available nonlinear regression pro-
grams. Such programs generally estimate a variety of statistical
criteria that facilitate a quantitative comparison of the growth
parameters in Ho above.

Comment

The paper presents a reparameterization of the von Bertalanffy
parameters incorporating tol, (the time when length is zero) and com-
bining k (growth rate) and ,, (maximal length). The method proposed
is easily implemented and statistically robust for all yield models
including von Bertalanffy growth formulations.

Kitchell, J.F., D.J. Stewart, and D. Weininger. 1977. Application of a
bioenergetics model to yellow perch (Perca flavescens) md walleye
(Stizostedion vitreum vitreum . J. F-i`s`F_.R_e7_._YX_.Canada 34:1922-1935.

Abstract

A simple energy budget equation is developed to yield a bio-
energetics model designed to simulate fish growth. Parameters for
the model are estimated from the literature for application to
yellmv perch (Perca flavescens) and walleye (Stizostedion vitreum
vitreum). SimUl-ations are presented that demonstrate model output
a@@wictioiis oC body size, activity level, ration level, food
quality, and environmental temperature. Sensitivity analyses iden-
tify the importance of food consumption, activity, and excretion as
biological processes represented in the parameters. On the basis of
temperature conditions in selected lakes and specified feeding levels,
simulations are presented to quantify the importance of year-to-year
variation of temperature in determining growth. In heterothermal
systems, temperature selection by periods can have a significant
effect on growth. For walleye on fixed rations, annual groinh can
vary from zero to twofold increments due entirely to differences in
simmer temperatures. Variations in food quality have lesser effects.

Comment

This simulation model represents growth bioenergetically. Its
application to yellow perch in Maryland could prove useful in the
production of a simulation yield model for the Maryland stock.
Its primary flaw is its need for large amounts of site-specific
data concerning consumption, metabolism, excretion, and, egestion.

B-16

This submodel requires the estimation of 17 parameters representing
these bodily functions.

Knight, IV. 1969. A formulation of the von Bertalanffy growth curve when
the growth rate is roughly constant. J. Fish. Res. Bd. Canada
26:3069-3072.

Abstract

The author contends that the parameters of any growth curve should
be a direct description of the graphical appearance of the data. For
growth that is even approximately linear this is not true of the von
Bertalanffy curve in its usual form (von Bertalanffy, Human Biol. 10:
181-213, 1938). On the above grounds, an alternate fo_rm-o-f-t Hevon
Bertalanffy curve for use in such instances is proposed.

Comment

This paper presents a linearization of the von Bertalanffy growth
equation.which adds a correction factor corresponding to the degree of
nonlinearity of the growth data. This approach adds little to the
conceptualization of a general growth submodel but does allow a more
exact fit to existing growth.data.

Pratt, D.M., and D.A. Campbell. 1956, Environmental factors affecting
growth in Venus mercenaria. Limol. Oceanogr. 1:2-17.

Abstract

The results of a five years' study of variations in quahog growth
rates in Narragansett Bay are summarized. Linear increments are inversely
proportional to initial length over the size range 35-70mm (greatest
aimension), whereas volumetric increments in this range are approxi-
mately constant. Growth in some parts of the Bay is three times as
fast as in others. Most of the year's growth occurs before mid-July,
and growth rates showed no significant variation from year to year.
Annual increments.in Narragansett Bay are considerably greater than at
Prince Edward Island, and they appear to fall in the same general range
as those reported for New Jersey and the south side of Cape Cod.

The effects of various environmental factors on growth have been
investigated over a three-year period. Growth was not appreciably
influenced by existing differences in current speed, dissolved oxygen
content, or salinity of the bottom water. Temperature imposes an
upper limit to the potential rate of growth, which is negligible below
100C and rises with increasing temperature at least to 23*C. Compari-
sons of growth rates with phytoplankton concentrations suggest that
small diatoms, either living or as detritus, are an important source

B-17

of quahog nutrition in these waters. Growth is retarded in sediments
with a high silt-clay content. This effect is discussed in relation
to the concomitant reduction in sediment permeability, the possible
accumulation of inhibitory substances, and the necessity for more
frequent clearing of the animals' filtering apparatus.

Comment

The paper presents a simple linear regression showing the effects
of temperature, sediment grain size, and phytoplankton concentration
on the growth rates of quahog clams. Its only foreseeable use for
@Iaryland fishery is to accent the need to account for environmental
variation in growth simulations, particularly for sedentary inverte-
brates.

Richards, F.J. 1959. A flexible growth function for empirical use.
J. Exp. Bot. 10:290-300.

Abstract

The application of an extended form of von Bertalanffy's growth
function to plant data is considered; the equation has considerable
flexibility, but is used only to supply an empirical fit. In order
to aid the biological analysis of such growth data as are capable of
representation by the function, general rate parameters are deduced
which are related in a simple manner to its constants.

Comment

This paper presents a generalized von Bertalanffy growth rela-
tionship and an excellent discussion of all the mathematical formula-
tions of growth (monomolecular, autocatalytic, and Gompertz). While
the presentation is primarily theoretical in nature, it removes many
of the restrictive assumptions of von Bertalanffy growth. It could
prove useful (but application may be difficult) for Maryland species.

Taylor, C.C. 1962. Growth equations with metabolic parameters. J. Cons.,
Cons. Int. Explor. Mer. 27:270-286.

Abstract

Growth in weight is the difference between anabolic and catabolic
processes. These are taken as proportional to length to the power of
a and b. Equations giving the length at any time, and also relating
length to the power of b-a at successive times are deduced. If b-a=l
these reduce to the Bertalanffy equation and the Ford-Walford plot.
1@ther values of a and b may produce parabolic growth, or an inflection
in curve of growth in length. Possible values of a and b based physio-
logical data are discussed, as are changes in the other growth parameters
I
with environmental factors., including food and temperature.

B-18

The methods are applied to data of several species, including
char, sturgeon, and sardine.

Comment

This paper presents extensions of the von Bertalanffy equation to
situations where length and weight are not related by a cubic exponent.
The approach is primarily theoretical (related to metabolic rates) and
the exponential value of the basic von Bertalanffy growth equation is
the difference between the exponential relationship of surface area to
length and the relationship of weight to length. This approach does
not appear to be useful or easily applied to Maryland fisheries.

Ursin, E. 1967. A mathematical model of some aspects of fish growth,
respiration, and mortality. J. Fish. Res. Bd. Canada 24:2355-2393.

Abstract

A simple metabolic model describing growth as the difference between
what enters the body and what leaves it, is elaborated assuping that
synthetic building-up processes (the anabolism) are consuming energy
supplied by processes of decomposition of the break-down (the catabolism).
This leads to partitioning total catabolism into two components, one
being a function of the rate of synthesis, another keeping the body
functioning independently of synthesis. The rate of synthesis is
described as a function of food taken, of the efficiencies of digestion
and energy conversion, and of the absorbing surface of the intestine.
Catabolic processes are supposed to be functions of the oxygen concentra-
tion in the water, the absorbing surface of the gills, and the rate of
oxygen transport. Both kinds of processes are made functions of temper-
ature in the way enzymatic processes usually are. Assuming that molecu-
lar interactions accidentally go wrong makes natural mortality, like
growth, a function of the rates of anabolic and catabolic processes
and body size.

Application of the model to data of length-at-age, food and oxygen
consumption, weight loss, gill area, and natural mortality indicates
that at least some of the main hypotheses cannot be rejected on avail-
able evidence.

Comment

This paper presents excellent submodels of growth and natural mor-
tality based on the bioenergetics of anabolism and catabolism. Unfor-
tunately, it cannot be practically employed for Maryland fisheries due
to the large number of parameters and amount of data necessary for the
determination of growth or natural mortality (81 parameters to determine
Jty). These submodels are theo-
the estimated growth and natural mortal L
retically comple-te biut practic'allytnusable.

B-19

W, I T-C 1). M. 197S. Relation betweeii egg size, growth, and natural mortality
of larval Ci.sh. 3. Fi.sh. Res-.13d. C,anada 32:2503-251.2.

Abstract

A set of density-dependent growth and survivorship equations is
C)
derived from evidence that the instantaneous death rate in the sea is
inversely proportional to particle size. The survivorship equation
reproduces several well-known phenomena observed in fish populations.
It predicts: 1) that winter and spring spawning species ought to pro-
duce larger eggs than summer spawners, 2) that it is advantageous for
species that spawn in batches to produce progressively smaller eggs
in spring and summer, and 3) that the death rate of a cohort of fish
should decrease continuously as the survivors grow and approach the
critical size.

The biological basis for the observed variation in the size of
pelagic fish eggs and larvae is thought to be due primarily to trophic
relations within the pelagic community. It is suggested from what is
known of the relative abundance and foraging capabilities of different
sized particles, that the survival rates of larval and juvenile fish
should increase as they grow and occupy a progressively higher position
in the food chain.

Comment

This paper presents a method of determining natural mortality
rates of larval fish from growth rates. The procedure requires
many parameters which are poorly defined biologically and appear to be
difficult to measure in situ. While this assessment of mortality
and growth appears bio-logically -reasonable, its application would
not be practical.

Yamaguchi, M. 1975. Estimating growth parameters from growth rate data.
Oecologia 20:321-332.

Abstract

The time interval over which growth rates are measured modified the
observed growth rates in non-linear growth curves. Growth rates obtained
from a sigmoid curve such as the logistic growth equation may appear as
if they were derived from the non-sigmoid von Bertalanffy growth equation
when the small stage is not represented in the hypothetical growth
observation. The inflection point of a sigmoid curve may be underesti-
mated in non-instantaneous growth rate data when they are plotted on a
graph against the initial sizes. This problem is significant for marine
macro-benthos., whose growth is likely to be sigmoid and initiates mostly
at microscopic sizes, when the popular von Bertalanffy growth equation
is fitted to the observed growth rate data. Even when the von Bertalanffy
growth equation appears to represent the observed growth rates adequately,

B-20

extrapolation of the equation t oward the smaller stage may require
an independent investigation.

Comment

This paper presents a method of determining growth parameters
from changes in individual size through time. It can be easily applied
and.may be useful in evaluating the growth of oysters and clams.

Mortality_Submodels (Fishing)

Brown, B.E., J.A. Brennan, and J.E. Palmer. 1979. Linear programming
simulations of the effects of bycatch of mixed species fisheries off
the northeastern coast of the United States. Fish. Bull. 76:851-860.

Abstract

We evaluated the results of using historic bycatch (incidental
catch) ratios in adjustin., fishing regulations by linear programming
techniques. We used both 1971 and 1973 bycatch ratios separately to
assess the sensitivity of the results to the reported changes in
bycatch ratios in estimating the total 1975 catch of countries fishing
in the northwest Atlantic. For 4 of the 11 countries for which data
were examined, the difference between the percentage of a country's
species total allowable catches (i.e., those catches allowed a country
by regulation) using the 1971 and 1973 bycatch ratios, was at least
20%. Only four countries were predicted to catch at least 80% of their
species total allowable catches. The predicted total catches of all
countries and all species was only 60% of the total species quotas.
The simulated directed fisheries constituted only 70% of the total catch
using 1971 bycatch ratios and only 73% using 1973 bycatch ratios.
Examination of the reported 1975 catches indicated that the total
allowable catches for herring were most frequently limiting a country's
catch. Except for USSR, the differences between reported and simulated
catches were less than 50 metric tons, with the difference less than
10 metric tons for 6 of the 11 countries. There was little difference
in reported versus simulated catches between the schemes using the
1971 and 1973 bycatch ratios.

Comment

This paper presents a statistical method of incorporating bycatch
Z>
mortality rates into fishing mortality rates. Their use of bycatch-
catch figures makes this procedure difficult to employ for Maryland
fisheries due to the probability that bycatch data for Maryland
fisheries do not exist.

B-21

Francis, R.C. 1974. Relationship of fishing mortality to natural mortality
at the level of maximum sustainable yield under the logistic stock
production model. J. Fish. Res. Bd. Canada 31:1539-1542.

Abstract

The often-used approximation that, for a stock of fish under
exploitation, the instantaneous fishing mortality rate equals the
instantaneous natural mortality rate at the point of the maximum
sustainable yield is examined with respect to its mathematical roots
and practical utility. Examples from two diverse fisheries are
utilized.

Comment

This paper does not present a submodel but simply discusses
the relationship of natural and fishing mortality at MSY for popu-
lations operating under the assumption of logistic growth.
It will not provide a useful submodel for this project.

Van Winkle, W., D.L. DeAngelis, and S.R. Blum. 1978. A density-dependent
function for fishing mortality rate and a method for determining ele-
ments of a Leslie matrix with density-dependent parameters. Trans.
Am. Fish. Soc. 107:395-401.

Abstract

A density-dependent function for the instantaneous fishing mortality
rate is presented. It is shown that this function may be readily incor-
porated into the age-specific probability of survival in a Leslie-matrix
population model.

A method is presented for indirectly determining the probability
of survival for age-class 0 of a fish population using a density-depen-
dent Leslie matrix. The method involves the two constraints that the
population be at equilibrium and that the index of absolute population
size in the density-dependent function be assigned a value. In addition,
given the probability of survival for age-class 0, it is shown that
the probability of survival through a selected life stage within age-
class 0 can be indirectly determined. Three problems in modeling a
fish population using a Leslie model are discussed in light of the
difficulties involved in modeling density dependence due to insufficient
information and lack of understanoding concerning density-dependent
phenomena.

Comment

This paper presents a methodology for decomposing instantaneous
fishing mortality into density-dependent and independent terms. lffiile
this decomposition is well based, information on the estimation of the
density- independent and dependent weighing factors is not included in the
presentation except in the special cases where F Dl or FD is equal to
zero. This formulation is too theoretical for applicatiRn to Maryland
species.
B-22

Young, W.D. 1975. An analysis of the effect of seasonal variability of
harvest on the estimate of exploitation rate. Trans. Am. Fish. Soc.
105:45-47.

Abstract

Five functional forms for the seasonal distribution of force of
fishing mortality were used in determining expectations of death from
fishing. Expectation of death from fishing calculated from E/F = A/Z
was compared to the actual expectation of death from fishing determined
by numerical integration. Bias results in the formula-calculated
expectation of death from fishing if the force of fishing mortality is
not a constant fraction of the force of total mortality. Bias is greater
when the force of fishing mortality is more asymmetrical. Bias is
positive when greater force of fishing mortality occurs early in the
year; negative bias occurs when force of fishing mortality is greater
later in the year. The magnitude of bias, for a given functional form
of force of fishing mortality, is a function of the relative size of
force of fishing mortality to force of total mortality..

Comment

This paper presents the addition of seasonal variation to fishing
mortality rates in a very theoretical framework. It assumed prior
knowledge of the functional form of fishing mortality and simply
investigates the consequences of altering that form. It appears
to be a better practice to subdivide the time scale into periods of
apparently equal fishing pressure than to employ this form of seasonal
variation.

Mortality Submodels (Natural)

Brousseau, D.J. 1978. Population'dynamics of the soft-shell clam, NVa
arenaria. Mar. Biol. 50:63-71.

Abstract

A life table was constructed for Mya arenaria from Gloucester,
Massachusetts, USA, based on schedules oT -age-specific fecundity and
mortality determined under natural conditions.- Mortality rates decrease
with size and age in this species, with the period of maximum mortality
occur-ring during the summer months. Mortality rates during the fall and
winter were considerably lower, perhaps due to the inactivity of natural
predators. The survivorship curve for M. arenaria approximates the
Type 3 curve of Deevey (1947). Mean liTe- expectancy is low in recently-
settled clam , peaks when the individual reaches 30.0 to 34.9 mm. (1 year

B-23

of age), and remains fairly high for most of the remainder of life.
The intrinsic rate of natural increase (r ) is very high: 4.74.
This enormous rate of potential increase Wx@ffset by high rates of
larval mortality in the plankton. Unlike the reproductive values of
most animals studied, those in M. arenaria peak late in life, well
after the known age of first reFroTu'ction. This is probably the
result of increased fecundity with age. The implications of this
work in the area of resource management are discussed.

Coment

This paper presents the use of life tables to determine the
intrinsic rate of natural increase (r). Size-specific natality and
mortality were converted to age-specific rates using the von Berta-
lanffy ago-size relationship. This procedure will be of little use
for this project.

Butler, S.A., and L.L. McDonald. 1979. A simulation study of a catch-
effort model for estimating mortality rates. Trans. Am. Fish. Soc.
108:353-357.

Abstract

The properties of the estimators of the instantaneous natural
mortality rate, M, and the instantaneous exploitation rate, q, are
investigated for a multiple regression model of catch-effort data
developed independently by D.G. Chapman and J.E. Paloheimo. It is
assumed that the units of effort are coded such that the values of
effort fall in an interval (a,b). Values of M and q are then con-
sidered which will give realistic probabilities of mortality over
two to 10 time periods. It is shown in a simulation study that the
approximation used to arrive at the regression model is accurate.
However., when sanpling error is taken into consideration, the esti-
mators of M and q do not simultaneously have acceptable properties
in the rang es anticipated in practice.

Comment

This paper presents an accepted methodology for computing instan-
taneous natural and fishing mortalities using catch/unit effort and
total effort data. A statistical and simlation analysis of this
submodel shows the variables to be covariant and a good estimate
of one provides a meaningless estimate of the other. The procedures
presented here are not directly applicable.

Marten, G.G. 1978. Calculating mortality rates and optimum yields from
length samples. J. Fish. Res. Bd. Canada 3S:197-201.

Abstract

An equation is derived for yield per recruit of a fishery (or
other exploited animal population) as a function of fishing intensity

.3-24

and age of first capture. The equation has the advantage that it
does not require explicit estimates of natural mortality or individual
growth rate parameters. Linear length growth is assumed until maxi-
mum size is'reached, and mortality parameters are expressed relative
to growth rate. Mortality parameters are estimated from average
length samples of separate populations experiencing different fishing
efforts in the same fishery. The equation may be used to compare
existing fishing efforts and age of first capture with optimal values.
Samples of the catfish, Bagrus docmac, from Lake Victoria (East Africa)
are used to illustrate the method.-

Comment

This paper presents a methodology for the determination of total
mortality and natural mortality from population length parameters
(length at time zero, asymptotic length,and mean length). The assump-
tions of equal growth and natural mortality used to determine natural
mortality from man lengths in two populations under different fishing
pressure are weak, but the method may provide useful estimates for
Maryland species where no other data exist.

Paloheimo J.E. 1961.- Studies on estimation of mortalities: I. Comparison
of a.method described by Beverton and Holt and a new linear formula.
J. Fish. Res. Bd. Canada 18:645-662.

Abstract

A method, described by Beverton (1954) and Beverton and Holt (1956
and 1957), giving estimates of the natural mortality rate, M, and the
catchability coefficient, q, from catch at age and.effort data, is
examined. This method requires 4 to 5 iterations to,arrive at the esti-
mates. rations unnecessary, and gives
We have derived approximate solutions for q and M in a closed
form. This makes the laborious ite
virtually the same values as arrived at by iterations.

The effectiveness of the iterative Beverton and Holt method is
evaluated by calculating q and M in 30 hypothetical examples. A new
and simple (linear formula) method for estimating q and M is derived.
Application of the new method to these 30 examples resulted in a 48%.
reduction of the standard deviation of q and a 45% reduction in that
of M. The new method is in part the same as one suggested by Gulland,
Beverton, and Holt (Beverton et al., MS, 1958; Holt, MS, 1959) to
arrive at initial values in their short-cut (iterative) method of esti-
mating the mortality rates. We show that these initial values are
actually better estimates than the final values arrived at by the
iteration.

Neither the Beverton and Holt method nor the linear formula give
necessarily unbiased estimates; the bias depends on the types of
variability in the data.

To arrive at non-biased, least-squares estimates would -require
ancillary information not normally available on the distributions of
the three variates: catch at age, effort, and catchability coefficient.

Comment

This paper presents a method of estimating natural mortality and
catchability which simplifies an earlier Beverton-Holt method. The
simplification requires primarily effort data and could be useful
for Maryland stocks.

Polgar, T.T. 1978. Striped bass icthyoplankton abundance, mortality, and
production estimation for the Potomac River population. Pages 109-
125 In W. Van Winkle (ed.), Assessing the Effects of Power-Plant-
Inducea-Mortality on Fish Populations. Sponsored by Oak Ridge
National Laboratory, Energy ResearcFand Development Administration
and Electric Power Research Institute.

Abstract

Methods are developed for estimating, from field survey data, the
mortality rate and production for each successive ichthyoplanktonic
stage. The abundance estimators used in the computation of these quan-
tities are also derived. An age-dependent, ichthyoplankton population
model is developed assuming either a uniform age distribution or an
exponential age distribution within each stage. Striped bass egg
and larval data from a 1974 ichthyoplankton survey in the Potomac River
are used in model computations. The various model estimates are evalu-
ated qualitatively, and the usefulness and limitations of the models
are discussed.

Comment

This paper and subsequent work by the author provide a framework
to determine natural mortality rates for fish populations exhibiting
variable yearclasses. This approach will prove to be very helpful
in this project.

Robson, D.S., and D.G. Chapman. 1961. Catch, curve and mortality rates.
Trans. Am. Fish. Soc. 90:181-189.

Abstract

The assumptions necessary to obtain a valid estimate of survival
rate from a sinale catch curve are discussed. An example of the best
estimate of survival rate and its variance is worked out for the case
that age is knoun exactly for the entire sample. A test for validity
of the model is illustrated. Methods of estimating the survival rate
are also given when some age groups are combined, when an age-length
key is used, and when only a segment of the catch curve is usable. A
table is provided to facilitate the estimation in this last ca-4e.

B-26

Comment
I The paper presents a simplistic method for determining natural
mortality rates from catch curves and age distributions. The deter-
mination requires several assumptions which may not be reasonable for
Maryland species (i.e., constant survival rate, constant yearclass
strength) but may prove useful in some cases.

Ursin, E. 1967. A mathematical model of some aspects of fish growth.,
respiration,, and mortality. J. Fish. Res. Bd. Canada 24:2355-2393.

Abstract

A simple metabolic model desribing growth as the difference
between what enters the body and what leaves it, is elaborated assum-
ing that synthetic building-up processes (the anabolism) are con-
suming energy supplied by processes of decomposition of the break-down
(the catabolism). This leads to partitioning total catabolism into tivo
components, one being a function of the rate of synthesis, another
keeping the body functioning independently of synthesis. The rate of
synthesis is described as a function of food taken, of the efficiencies
of digestion and energy conversion, and of the absorbing surface of
the intestine. Catabolic processes are suggested to be functions of
the oxygen concentration in the water, the absorbing surface of the gills,
and the rate of oxygen transport. Both kinds of processes are made
functions of temperature in the way enzymatic processes usually are.
Assuming that molecular interactions accidentally go wrong makes
natural mortality, like growth, a function of the rates of anabolic
and catabolic processes and.body size.

Application of the model to data of length-at-age, food and oxygen
consumption, weight loss, gill area, and natural mortality indicates
0
that at least some of the main hypotheses cannot be rejected on avail-
able evidence.

Comment

This paper presents excellent submodels of growth and natural
mortality based on the bioenergetics of anabolism and catabolism.
Unfortunately, it cannot be practically employed for Maryland fisheries
due to the large number of parameters and amount of data necessary for
the determination of zrowth or natural mortality (81 parameters to
estimate growth and natural mortality). These submodels are
theoretically complete but practically unusable.

Van Sickle, J. 1977. Mortality rates from size distributions. Oecologia
27:311-318.

B-27

Abstract

A population model explicitly describing the dynamics of an
arbitrary population size distribution is presented. One consequence
of the model is an equation for the exact shape of the size distribu-
tion of a stationary or steady-state population. The shape is expressed
as a function of size-specific mortality and growth rates. From the
equation, various mortality estimation formulas can be derived., two of
which are discussed in detail. One of the methods permits estimation
of size-specific mortality rates without the assumption of a theore-
tical arowth model.

Comment

This paper presents a useful and easily employed estimate of
natural mortality rates using only a knowledge of recruitment
and size-frequency distributions. The procedure may be useful for
the determination of natural mortality rates for oysters and clams.

Ware, D.M. 1975. Relation between egg size, growth, and natural mortality
of larval fish. J. Fish. Res. Bd. Canada 32:2503-.2512.

Abstract.

A set of density-dependent growth and survivorship equations is
derived from evidence that the instantaneous death rate in the sea is
inversely proportional to particle size. The survivorship equation
reproduces several well-known phenomena observed in fish populations.
It predicts: 1) that winter and spring spawning species ought to pro-
duce larger eggs than summer spawners, 2) that it is advantageous for
species that spawn in batches to produce progressively smaller eggs
in spring and summer, and 3) that the death rate of a cohort of fish
should decrease continuously as the survivors grow and approach.the
critical size.

Comment

This paper presents a method of determining natural mortality rates
of larval fish from growth rates. The procedure requires many parameters
which are poorly defined biologically and appear to be difficult to neasure
in situ. While this assessment of mortality and growth appears biologi-
cally reasonable,its application would not be practical.

B - 20"

Mortality Submodels (Total)

Ssentongo, G.W., and P.A. Larkin. 1973. Some simple methods of estimating
mortality rates of exploited fish populations. J. Fish. Res. Bd.
Canada 30:695-698.

Abstract

Some simple equations are derived from mortality functions that
enable estimation of the total mortality coefficient from the mean age
and the age of first capture, or the mean length and the length of
first capture, of fish in a catch.

Comment

This paper presents a methodology by which an estimate of total
mortality rate can be determined from mean age and age of first cap-
ture of mean length and length of first capture. Since it does not
partition total mortality (Z) into natural (ND and fishing (F) mortality
rates, the procedureis of little use to this project. Subsequent
works have expanded this procedure by partitioning total mortality into
its component parts.

Vaughan, D.S., and S.B. Saila. 1976. A method for determining mortality
rates using the Leslie matrix. Trans. Am. Fish. Soc. 105:380-383.

Abstract

The Leslie matrix algorithm has been utilized to estimate mortality
of a year class assuming an equilibrium population for a species. Under
this assumption an estimate of the mortality for the Otn year class of
the Atlantic bluefin tuna (Thunnus tjVnn@Ls thynnus) has been made indi-
cating about five survivors Trom 10 million eggs in the first year of
life. The mortality rates for later year classes were derived from
empirical data.

Comment

This paper presents a procedure for the calculation of total
mortality rates using age-specific fecundity and assuming an
equilibriun population level. This assumption realistically makes
the use of this procedure impractical for Maryland fisheries, but
in sow situations the method may prove useful.

B-29

Parameter Estimation (Specific Yield Models)

Fox, W.W., Jr. 1971. Random variability and parameter estimation for the
generalized production model. Fish. Bull. 69:569-580.

Abstract

Three alternative statistical models are proposed for estimating
the parameters of the generalized production model by the method of
least squares. A stochastic representation of the generalized production
model is constructed and simulation (or the Monte Carlo Method) is
employed to infer the effects of random variability on the variation
in catch. The use of residuals examination for selecting the appropri-
ate statistical model for least-squares estimation of the generalized
production model parameters is demonstrated for the yellowfin tuna
fishery in the eastern tropical Pacific Ocean. In both the simulation
and actual fishery, statistical Nbdel 3 --- assuming catch residual
variance is proportional to the catch squared --- best fulfills the
assizTtions of least-squares theory and should, therefore, provide the
best least-squares parameter estimates.

Comment -- Applicable model - Surplus production models

This paper presents-four easily utilizable methods of estimating
Schaefer, Pella-Tomlinson, and Fox surplus production model parameters
employing least-squares criteria. Two of these approaches do not meet
spveral of the assumptions concerning parameter resirluals and residual
variance.

FoxV W.W., Jr. 197S. Fitting the generalized stock production model by
least-squares and equilibrium approximation. Fish. Bull. 73:23-37.

Abstract

A least-squares method for fitting the gmeralized stock production
to fishery catch and fishing effort data which utilizes the equilibrium
approximation approach is described. A weighting procedure for provid-
ing improved estimates of equilibrium fishing effort and an estimator of
the catchability coefficient are developed. A computer program PRODFIT
for performing the calculations is presented. The utility and perfor-
mance of PRODFIT is illustrated with data from a simulated pandalid
shrimp population.

Comment -- Applicable mod el - Surplus production models

This paper adds parameter estimation methods of equilibtium aDprox-
imation to the least-squares criteria presented in an earlier paper. This
approach can be easily used for deriving the surplus production model
parameters a and @. Estimation procedures to determine q, catchability,
are also presented in an easily employable form. PRODFIT, a computer
program, is presented, which determines these parameters.

B-30

16 Var(I 0 1). , mcl L.J. Bledsoe. 1978. Parameter estimation for the Pella-
Tomlinson stock production model under non-equilibrium conditions.
Fish. Bull. 76:523-534.

Abstract

To estimate the parameters of the Pella-Tomlinson model, as restruc-
tured by Fletcher, we suggest a derivative-free version of the Levenberg-
Marquardt algorithm, along with an algorithm that locates starting values
for the iterative procedure. The iterative method of Levenberg-Marquardt
was applied to two versions of the restructured model: five parameters
were estimated in the first version and three in the second, the latter
preventing degeneracy of the model to exponential form. We discuss in
particular the causes of the degeneracies associated with previous appli-
cations of the model. Such faults lie, inherently, with the mathematical
indeterminacy of the system equations themselves, so that all nonlinear
estimation methods will tend to be inefficient in the absence of external
constraints. The effectiveness of the Levenberg-Marquardt method was
evaluated by Monte-Carlo simulation. As examples, we analyzed catch--
effort data from the yellowfin tuna fishery of the eastern Pacific and
catch-effort data from the Pacific halibut fishery (Area 2 of the Inter-
national Pacific Halibut Commission).

Comment -- Applicable model - Pella-Tomlinson surplus production model
(as altered by Fletcher)

This paper presents a methodology for estimating the five para-
meters necessary to fit the Pella-Tomlinson yield equation in non-
equilibrium conditions using a Levenberg-Marquardt algorithm. The authors
present a procedure for estimating the variability of these parameters
as well. These estimates are easily determined and in most instances
accurate for the Pella-Tomlinson yield determination.

Walter, G.G. 1975. Graphic al methods for estimating parameters in simple
C>
models of fisheries. J. Fish. Res. Bd. Canada 32:2163-2168.

Abstract

A-graphical method for calculating the coefficients for a Schaefer
model of a fishery is introduced. It involve's plotting catch per effort
vs effort data and then correcting the values for disequilibrium of the
fishery. A 'hypothetical and a realistic example are presented.
Comment Applicable models -.Grahm and Schaefer-surplus production.
models

This paper presents a simple methodology for the determination
of the parameters of the Schaefer surplus. production model. These esti-
mates are corrected according to the disequilibriun conditions of the
fishery. An estimate of catchability is also determined. These
estimates are easily obtainedand the data necessary for the estimates
are minimal.

B-31

Recruitment Submodels

Allen, K.R. 1968. Simplification of a method of computing recruitment rates.
J. Fish. Res. Bd. Canada 25:2701-2702.

Abstract

A simplification of a method for computing the proportion of new
recruits in each year-class in each year's catEh fro@i annual age
composition is presented.

Comment

This simplification is easily applicable to Maryland stocks for
which annual age compositions and fishing rates are known. The
method could provide important input information for non-equilibrium
surplus production models and simulation models.

Christensen, S.W., D.L. DeAngelis, and A.G. Clark. 1978. Development of
a stock-progeny model for assessing power plant effects on fish
populations. Pages 195-225 In W. Van Winkel (ed.) Assessing the
Effects of Power-Plant-induceCY-x)rtality on Fish Popuka@tions.
Sponsored by Oak Ridge National Laboratory, Energy Research
and Development Adminstration and Electric Power Research Institute.
Abstract

A multi-age-class model, based on simple but general biological
principles, is developed to assess the impact of power plants on fish
populations. The model is then parameterized in order to produce a
variety of stock-progeny relationships, assuming that the stock is
always at stable age distribution. The predicted response of the fish
stock to power plant cropping of young-of-the-year fish is investigated
for each of these stock-progeny relationships. In general, the sensi-
tivity of the equilibrium stock size to cropping is positively related
to the slope of the stock-progeny curve at the equilibrium point and,
to a lesser extent, negatively related to the slope of the curve at the
origin. In addition, the timing of power-plant-induced mortality in rela-
tion to the timing Of compensation is important. The maximum amount of
power-plant-induced mortality that can be tolerated by the stock can be
calculated from the slope of the curve at the origin. Application of the
model to specific cases will likely need to utilize time-series simula-
.tions in addition to the steady-state approach investigated here.

Comment

This paper presents a workable submodel for density-dependent re-
cruitment, but requires numerous parameters which will probably not
be available for Maryland species.

Ricker, W.E. 1954. Stock and recruitment. J. Fish. Res. Bd. Canada
11:559-623.

B-32

Abstract

Plotting net reproduction (reproductive potential of the adults
obtained) against the density of stock which produced them, for a
number of fish and invertebrate populations, gives a domed curve whose
apex lies above the line representing replacement reproduction. At
stock densities beyond the apex, reproduction declines either gradually
or abruptly. This decline gives a population a tendency to oscillate
in numbers; however, the oscillations are damped, not permanent, unless
reproduction decreases quite rapidly and there is not too much mixing
of generations in the breeding population. Removal of part of the adult
stock reduces the anplitude of oscillations that may be in progress and,
up to a point, increases reproduction.

Comment

This paper presents one of the most extensive treatments of
recruitment models. @,bst of the treatments are directed towards
salmon and are thus of little direct value for Maryland fisheries.
The paper @rimarily presents hypotheses with numerous examples, but
with little or no mathematical formulation. Regardless, this family
of recruitment curves will prove helpful in this project.

Yearclass Strength,Submodels

Stevens, D.E. 1977. Striped bass forone saxatilis) year class strength
in relation to river flow in tlie Sacra@ @nto-@San Joaquin estuary,
California. Trans. Am. Fish. Soc. 106:34-42.

Abstract

Striped bass, Morone saxatilis, abundance indices were developed
from two analyses oT-sportfishing party boat catch statistics for the
Sacramento-San Joaquin Estuary. These analyses cover the periods
1938-1954 and 1958-1972. The abundance indices provided evidence that
the size of the fishable population fluctuated by a factor of 3.7 during
the latter period and that river flows in the first.suTmer of life
affected recruitment during both periods.

Comment

This paper presents a simple statistical relationship between
freshwater discharge and young striped bass survival. It will probably
I
not be useful for determining yearclass strengths in 114aryl and*, striped
bass stocks.

B-33

Walter, G., and W.J. Hoagman. 1975. A method for estimating year class
strength from abundance data with application to the fishery of Green
Bay, Lake Michigan. Trans. Am. Fish. Soc. 104:245-255.

Abstract

A method of calculating indices of annual year class strength from
abundance data only is introduced. It involves a mathematical technique
based on difference equations. The method is applied to data for six
commercially harvested species of Green Bay, Lake Michigan. Estimated
year class strengths are then compared with relative abundance indices
of the parental spawning populations. Only for smelt and perch was
there a significant correlation between the two. The same species were
also the only ones for which the exponential relation of Ricker between
spawners and year class strengths could be established.

Comment

This paper presents a relatively complex submodel which predicts
yearclass strength from the biomass of exploitable stock, the biomass of
newly recruited exploitable stock, the hatching biomass, and pre- and post-
recruitment survival rates. It may prove useful for Maryland species
which exhibit variable yearclasses, but the data necessary for its
use probably do not exist.

B-34

APPENDIX C

An annotated bibliography of methods of data acquisition

for stock management models

C-1

Age Structure

White, M.L., and M.E. Chittenden. 1977. Age determination, reproduction
and population dynamics of the Atlantic croaker, Nticropogonias undulatus.
Fish. Bull. 7S:109-123.

Abstract

A validated scale method of age determination is described for the
Atlantic croaker, Micropogonias undulatus. Two age-classes were usually
observed, but only one was abundant. -Fle-an total lengths were 15S-165m
at age I and 270-280mm at age II based on three methods of growth esti-
mation. Fish matured near the end of their first year of life when they
were about 140-170mm total length. Spawning occurred from at least Sep-
tember through March, but there was a distinct peak about October. Soma-
tic weight-length relationships varied monthly, and changes appeared to
be associated with maturation and spawning. Somatic weight reached a
maximum in June, and the minimum was observed in'March. Maximum somatic
weight loss (24%) occurred in March, but no data were obtained from
December through February. In estuaries, age 0 croaker apparently
occupied soft-substrate habitats and older fish occurred near oyster reefs.
Life spans were only 1 or 2 yr, and the total annual mortality rate was
96%. The above life history pattern appears similar for croaker found
throughout the Carolinian Province. Contrasts are presented to illu-
strate differences in the life histories and population dynamics of
croaker found north and south of Cape Hatteras, N.C. A parallel is
drawn with apparently similar changes in the American shad, Alosa.
sapidissima, and the suggestion is made that changes in the population
dynamics of species that traverse the Cape Hatteras area may represent
a general phenomenon.

Growth

DuPaul, W.D., and J.D. McEachran. 1973. Age and growth of the butterfish,
Peprilus triacanthus, in the lower York River. Ches. Sci. 14:205-207.

Abstract

Age, rate of growth and the leng
,th-weight relationship of the
butterfish, Peprilus triacanthus, were determined. The age of speci-
mens collectecT in-September,--1-9'69 from the lower York River, Virginia,
was determined by counting rings in the otoliths. Four age groups
were represented in the sample: young-of-the-year fish (91-95mm),
year-old fish (98-139mm), two-year-old fish (142-187mm), and three-
year-old fish (174-200mm). The length-weight relationship for all
specimens is: log IV = -5.1852 + 3.2646 log L.
C,

C-2

Eldridge, P.J., W. Waltz, R.C. Gracy, and H.H. Hunt. 1976. Growth and
mortality rates of hatchery seed clams, Mercenaria mericenaria, in
protected trays in waters of South Carol'ina. Proc. 1Nat__._=e ifish.
Assoc. 66:13-20.

Abstract

Seed hard clams, Nlercenaria-mercenaria, were planted in trays at
densities.of 290, 580, and 869/m/- in three widely separated intertidal
areas in South Carolina. Survival of clams was similar at each site
although claw at Clark Sound experienced a lower survival rate.
Growth of clams planted at Clark Sound and Albergottie Creek was sig-
nificantly higher than those at Bull Bay. Growth occurred throughout
the year with the best growth experienced in summer and fall.

Feder, H.M., and A.J. Paul. 1974. Age, growth and size-weight relationships
of the soft-shell clam, Mya arenaria, in Prince William Sound, Alaska.
Proc. Nat. Shellfish. Assoc.7-4-:45-52.

Abstract

Soft-shell clamsY Mya arenaria, from Simpson Bay, William Sound,
Alaska, were examined. T-sin-glesample of 178 specimens was used to
determine the growth history of twelve year-classes by the annular
method. In Prince William Sound soft-shell clams reach a harvestable
size of 50 mm long in 6 or 7 years. Length-weight relationships are
considered. Dry meat weight (solids) averaged 18.8%.

Knudsen, E.E., and W.H. Herke. 1978. Growth rate of marked juvenile
Atlantic croakers, Ndcropogon undulatus., and length of stay in a
coastal marsh nursery in southwest Louisiana. Trans. Ain. Fish. Soc.
107:12-20.

Abstract

. To estimate the growth rate of the Atlantic croaker 113,670 juve-
niles were marked and released between 23 January and 23 'March 1975.
The experiment was divided into five tests; in each., the fish were
marked with a different fluorescent pigment. Recapture efforts pro-
duced 100 usable returns. Individual croakers appear to have had an
average maximum stay of l..8 months in the nursery. Growth rate was
estimated by the regression of length of recaptured.juveniles on time;
estimates ranged from 0.51 to 0.99mm/day, all of which were consider-
ably higher than most estimates in the iiterature. There was a trend
of increasing growth rate through the five tests. Length frequencies
of 159,381 croakers taken in a trap at the study area outlet could not
be interpreted in a reliable manner for growth rate estimates. Future
attempts to estimate young-of-the-year croaker growth rates should use
methods other than length frequency.

C-3

Ney, J.J., and L.L. Smith. 197S. First-year growth of the yellow perch,
Perca flavescens, in the Red Lakes, Minnesota. Trans. Am. Fish. Soc.
104:718-725.

Abstract

First-year growth of the yellow perch, Perca flavescens.(Mitchill),
was determined from 28,241 young-of-the-year-f-l-sT collected in the Red
Lakes during the summer (1 July-20 August) in 15 seasons and on back
calculations from I-annulus fish of the same year classes. Sexes did
not differ in growth rate. Geographic variations in first-season growth
occurred in 2 years but were not reflected at I-annulus. A consistent
seasonal growth pattern in all years was linear through the summer
sampling period. Growth rate after 20 August gradually declined and be-
came nearly asymptotic by the end of the season. Growth during midsummer
varied widely in different seasons, but length at I-annulus was -relatively
uniform (71-79mm) for most years as the result of apparent growth com-
pensation. Negative correlations of length on I July with midsummer
rate of growth and, consequently, with I-annulus length were noted.
Although high summer water temperature appeared to exert a positive
influence on growth, it did not obscure the compensatory phenomenon.
Growth compensation is probably not affected by predation, but may be
related to interaction of perch size and food availability.

Shelton, W.L., W.D. Davies, T.A. King, and T.J. Timmons. 1979. Variation
in the arowth of the initial year class of laraemouth bass in the West
0 Z>
Point Reservoir, Alabama and Georgia. Trans. Am. Fish. Soc. 108:142-149.

Abstract

West Point Reservoir, Alabama and Georgia, first reached full pool
in spring 1975. Growth within the initial year class of largemouth
bass (Nlicropterus salmoides) was highly variable. During the first
summer of impoundnient, length frequencies of the 1975 year class were
characterized by a single mode. However, there was an obvious condition
difference among individuals within the population. Generally, fish
longer than l7cm total length were in relatively good condition and
those 8-17cm long were in relatively poor condition. By fall (September
to October), one segment of the population had grown rapidly but the
other segment had grown little and a bimodal length-frequency distri-
bution was evident. A shortage of available prey for the smaller fish
was considered to be the cause of the growth disparity.

St. Pierre, R.A., and J. Davis. 1972. Age, growth,. and mortality of the
white perch, Morone americana, in the James and York rivers, Virginia.
Ches. Sci. 137=-28'1.

Abstract

More than 800 white perch, Morone americana, were collected from
each of two major tributaries of@_southe_r_n_Uie_sapeake Bay to compare

C-4

age structures, growth, and mortality rates. The James River is char-
acterized by heavy domestic and industrial pollution in several
tributaries and segments, whereas the York River is only slightly
polluted.
Maximum ages determined by scale analysis were 7 and 10 years for
males and females in both rivers. Yearling perch in the York River
appeared to have a significantly greater mean length than those from
the James but this difference is largely due to sampling bias. Both
sexes of white perch from the James River were significantly larger
than perch from the York River at ages II and III. However, mean
lengths of white perch from age groups IV and older were not signifi-
cantly different between rivers for either sex. In both rivers females
were significantly longer (to age V) and heavier (all ages) than males
of comparable age. Yearly growth increments were greatest early in
life. All year classes of white perch followed a similar growth pattern
during their first year in both estuaries. However, for later years
of lifeearlier year classes apparently had a greater mean length than
those from more recent years at comparable ages. Males were more abun-
dant than females for the first 3 years of life in both rivers; however,
females were present in significantly greater numbers in all older age
groups. Analysis of relative age frequcency suggested dominant year
classes in 1964 and 1965 in the James River, and a weak year class in
1968. No dominant year classes were apparent from the York River
collections.
Mortality rates were calculated from age frequency distributions.
Total annual mortality in the James River was about 69% for males after
age IV and for females after age VI. In the York River, malesat age
III and older die at a rate of 59%, whereas females older than age V
have an annual mortality of 57%.

Schwartz, F.J., and R. Jachotqski. 1965. The age, growth, and length-weight
Z>
relationship of -the Patuxent River, Maryland ictalurid white catfish,
Ictalurus catus. Ches. Sci. 6:226-237.

Abstract

The age and growth of 470 white catfish captured in 1963 from the
Patuxent River near Trueman Point, Maryland, was [sic] examined. Annual
growth increments varied 25-45mm. Oldest river specimens were 12 years.
The larrIgest knoim Maryland specimen from a freshwater millpond was
14 years old. Length back calculations at each age were possible by
the cu-Kved vertebral-fish length relationship formula Y =.63.21 + 89.64X
+ .01X4. The length-weight relationship for the same .specimens was
best exemplified by the log formula log Y = 1.9791 + .1689 log X. In
each case Patuxent specimens were shorter lived, shorter, and weighed
less than those from other parts of its natural or introduced range.

Winget R.R.,,.C.E. E-pifanio, T. Runnels, and P. Austin. 1976. Effects of
diet and temperature on growth and mortality of the blue crab, Calli-
n
tes sapidus, maintained in a recirculating culture system. Proc.
ec
,Va-tl. Shel-ji'i-sh. Assoc. 66:29-33.

C-5

Abstract

Blue crab growth parameters were measured over a sixty-day period
in a recirculating culture system, with each crab in physical isolation.
Dependent variables were molt interval, increase in carapace width per
molt, percent molt and mortality. No consistent growth differences were
detected in animals fed diets ranging from 26 to 75% protein content.
A temperature of 30*C generally increased molt frequency and percent of
animals molting compared to a temperature of 20*C. Increased temperature
appears to depress cuticle expansion and to decrease mortality.

Migration

Lund., W.A.,, and G.C. Maltezos. 1970. IMovements and migrations of the bluefish,
Pomatomus saltatrix, tagged in waters of New York and Southern New England.
Trans. Ain. Fish. $oc. 99:719-725.

Abstract

Bluefish were tagged in and near Long Island Sound between 1964 and
1969. Tag returas support the belief that there is a discrete northern
race of bluefish as more than 75% of the returns from fish at large more
than one season return to the general area of Long Island Sound. Small
bluefish move southward along the coast during late fall while adults,
fish over approximately 45 cm total length, have an inshore-offshore
migration.
Bluefish first arrive in the area when the water temperatures reach
12 to 15'C which is usually during iMay. The fish follow the warmer
water by entering the inner bays of Long Island or going to the western
end of Long Island Sound. Large numbers of bluefish arrive in the
general area during late July and August after spawning in offshore
waters. The fall migration takes place when the water temperature drops
to approximately 13 to 15'C.

Nicholson., W.R. 1971, Coastal movements of Atlantic menhaden as inferred
from changes inage and length distributions. Trans. Am. Fish. Soc.
100:708-716.

Abstract

Length frequency distributions of Atlantic menhaden, Brevoortia
tyraplus, plotted by age, month, and latitude, support the_H@F_othesis
_oT an-annual north-south movement. The majority of Atlantic menhaden,
wintering in offshore waters south of Cape Hatteras, North Carolina,
move northward in early spring. By about mid-June menhaden are distri-
buted in coastal waters from Florida to Maine, their age and size increas-
ing., from south to north. A slow northward movement of fish north of
False Cape, Virginia, continues throughout the summer. A southward
movement, beginning in early September north of Cape Cod, 'Massachusetts,
and involving all fish north of False Cape by November, culminates in

C-6

January when the majority of the population is again south of Cape
Hatteras.

Nicholson, W.R. 1978. %vements and population structure of Atlantic
menhaden indicated by tag returns. Estuaries 1:141-150.

Abstract

Over 968,000 adult Atlantic menhaden,, Brevoortia tyrannus, were
tagged from 1967 to 1969 and over 8S.,000 juvenile menhaa_enwer@ tagged
from 1969 to 1973. Recoveries of these tagged fish through 1975 pro-
vide direct evidence that Atlantic menhaden consist of a single popula-
tion that overwinters in offshore waters off the southeastern coast of
the United States, moves northward in spring and stratifies along the
coast by age and size during summer, and moves southward in late autumn.

Mortality (Natural or Fishing)

Coates, P.G., A.B. Howe, and A.E. Peterson. 1970. Analysis of winter
flounder tagging off Massachusetts, 1960-1965. IMass. Dept. Nat. Res.,
Div. Mar. Fish. 47 pp.

Abstract

From 1960 to 196S, 12,lSl winter flounder,Pseudopleuronectes
americanus OValbaum)., were tagged with Petersen tags at 21 locations
off Massac"hu etts. Returns through 1967 totaled 4,105 or 33.8% with
considerable variation between areas. The ratio of females to males
at tagging was 2.3. Movement was apparently related to water tenper-
ature C@ and was greatest south of Cape Cod. Intermingling of breeding
adults was more evident south of Cape Cod than north of the Cape.
Little mixing was evident between Georges Bank and inshore areas.
Returns during successive spawning seasons indicated some homing move-
ment. Growth rate varied with area and females grew larger than males
in all areas. Survival and mortality rates calculated from annual
returns show that flounder inshore south of Cape Cod were exploited
at a higher rate than those offshore. Returns showed negligible
foreign exploitation. Standing crop was greater south and east of
Cape Cod than on Georges Bank. Patterns of movement and other infor-
mation suggest the existence of three groups of winter flounder off
Massachusetts. Massachusetts trawling regulations and effects of
increases in winter flounder exploitation are discussed.

Dryfoos R.L., R.P. Ch eek, and R.L. Kroger. 1973. Preliminary analysis of
Atlantic menhaden, Brevoortia tyrannus, migration, population structure,
survival and exploitation 7-ates, and availability as indicated from
tag returns. Fish. Bull. 71:719-734.

C-7

Abstract

Over 1 million adult Atlantic menhaden, Brevoortia tyrannus, were
tagged from Loncy Island Sound to Florida between 1966 and 1969. Tag
recoveries indicate these fish migrated northward in spring and early
summer and southward in fall. As the fish grew older and larger, they
also migrated farther northward each spring. Calculation of rates of
interchange between fishing areas indicated that 21% of the recoveries
from fish released in Chesapeake Bay in 1967 and 1968 accounted for
72% of the catch of taacred fish 1 year later in New York and New Jersey.
Preliminary estimates of population parameters were made from tag-
recovery and catch data. Survival rates determined yearly from rate
of recoveries, however, varied due to fluctuations in availability.
Annual survival rates averaging 0.23 were calculated with Robson-Chapman
catch curve analysis and age composition of catch methods. From tag
recoveries, exploitation rate was extimated to be 50%, instantaneous
fishing mortality rate (F) was 0.95, and instantaneous natural mortality
(M) was 0.52. Tag returns also indicated that significant fluctuations
in availability of Atlantic menhaden occurred in Chesapeake Bay.

Henry, K.A. 1978. Estimating natural and fishing mortalities of chinook
salmon, Oncorhynchus tshauytscha, based on recoveries of marked.. fish.
Fish. Bull. 76:45-57.

Abstract

In this paper the method of calculating estimates of fishing mor-
tality (F) and natural mortality M occurring in the ocean for 1961
and 1962 brood Columbia River hatchery fall chinook salmon, Oncorhynchus
tshawytscha, based on assumed values of the proportion of fish th t
mature annually (m) and on recoveries of marked fish is demonstrated.
The advantages of this method over the method of assuming fixed
natural mortality rates and back calculating estimates is discussed.
It was possible to develop estimates of 1962 Spring Creek data up to
the fourth year of life and to compare these.estimates with values for
the 1961 brood whereas no estimates had been possible with the back
calculation method. Thus, estimates of Ml are slightly higher for the
1962 brood. A major difference between the two methods is that natural
mortality was assumed to be constant for the back calculation method
whereas estimates of natural mortality were obtained separately each
year using assumed proportions maturing. Thus, for the 1962 brood of
general marked fish, a M = 0.60 was used in the back calculation
method while estimates of Mi = 5.814, M2 = 0-510, M3 = 0.653, and
N4 = 0.727 were obtained by assuming varying proportions maturing.
A series of graphs are developed that permit a quick analysis of
any combination of proportions of fish maturing, fishing mortality,
and natural mortality and which clearly depict the relationship between
these various factors.

Houre, A.B., P.G. Coates, and D.E. Pierce. 1976. Winter flounder estuarine
year class abundance, mortality, and recruitment. Trans. Am. Fish. Soc.
105:647-657.

C-8

Abstract

Mark-recovery methodology, accounting for mortality, enabled annual
estimates of estuarine winter flounder (Pseud2@leuronectes americanus)
abundance at summer's end. Recruitment from the estuarine s@_stem to an
inshore-offshore otter trawl fishery occurred incompletely (73%) over
three age groups (II-IV). The release of tagged pre-recruits in two
consecutive years and subsequent return data, adjusted for non-reporting
of tag recoveries, yielded the following 1970 post-recruit, instantaneous
mortality parameters: total = 0.3570, fishing = 0.2445, natural = 0.1125
A comparison of total recruitment from southeastern 14assachusetts flounder
groups, derived from a population estimate and instantaneous total mor-
tality rate, with recruitment from the lVaquoit Bay-Eel Pond system
indicated that the latter constituted less than one percent of the total
required to maintain equilibrium catch.

King, T.A., W.D. Davies, and W.L. Shelton. 1979. Fishing and natural mortality:
Effects on the initial year class of largemouth bass in West Point Reser-
voir, Alabama and Georgia. Trans. Am. Fish. Soc. 108:1SO-lS5.

Abstract

Rates of growth, fishing mortality, and natural mortality of the
initial,(1975) year class of largemouth bass (Micr2pterus salmoides)
in 10.481-hectare West Point Reservoir were estimated over a 2-year
period by rotenone sampling, electrofishing, and tagging studies. The
estimate of an initial standing stock of 1,617 largemouth bass or
28.3 kg/hectare in August 1975 was reduced to 1.5 largemouth bass or
1.6 kg/hectare by August 1977. Instantaneous annual rates of growth,
fishing mortality, and natural mortality were 2.19, 1.17,, and 2.78,,
respectively, in the first year, and 0.65, 0.92, and 2.11 in the 'second.
Total yield from the 1975 year class was 21.4 kg/hectare, of which 95%
was harvested during the first year of impoundment, through August 1976 .
Natural and fishing mortality reduced the standing crop of 1975 large-
mouth bass by 88.2% (58.6% natural, 29.6% fishing) in the first year
and 86.2% (69.1% natural, 17.1% fishing) in the second. The abundant
initial year class of bass was responsible for creating a fishing "boom.
Population structure change brought about by high natural mortality and
fishing mortality resulted in a "bust" in largemouth bass fishing after
2 years of impoundment.

Lough, R.G. 1974. A re-evaluation of the combined effects of temperature
and salinity on survival and growth of mytilus edulis larvae using
response surface techniques. Proc. Natl. Shellfish. Pssoc. 64:73-76.

Abstract
Response surface techniques were used to critically examine the
combined effects of tenperature and salinity on late larval survival
and growth of Mytilus edulis using experimental data reported in the

C-9

literature. The range of conditions estimated for maximum survival
was found to be significantly different than those for maximum growth.
Temperature exerted a strong effect on both larval survival and growth,
while a temperature-salinity interaction effect was not significant.

Matlock, G.C., R.A. Marcello, and K. Strawn. 1975. Standard length-total
length relationships of Gulf menhaden, Brevoortia patronus Goode, Bay
anchovy, Anchoa mitchelli (Valenciennes), and AM@itic cro@ker, Dlicro-
pogon und@atu@s Tf1`n_n_a_eu__s_), from Galveston Bay. 'Trans. Am. Fish7_ -Soc.
104:408-409.

Abstract

Regression equations were developed between standard length CSQ
and total length (TL) for Brevoortia patronus, Anchoa mitchilli, and
Ndcr2@ogon undulatus taken from Ga ve;ton Bay--, Te-xas. The stand
length-total length onversion equation for B. patronus, was TL = 0.62044
+ 1.25323 SL, and for A. mitchilli it was Tt-7 0727-9-1-+ 1.26753 SL (for
fish 28-95 mm SQ; TL ;;-9.70548 + .17538 SL (for fish 102-159 mm SQ;
and TL = 19.88505 + 1.10952 SL (for fish 168-255 m SQ.

Oliver, J.D., G.F. Holeton, and K.E. Chua. 1979. Overwinter mortality of
fingerling, smallmouth bass in relation to size, relative energy stores
and environmental temperature. Trans. Am. Fish. Soc. 108:130-136.

Abstract

Hatchery-reared, 0 + age smallmouth (Ilicropterus dolomieui) of
55-107 mm total length were "wintered" from Sept-emFeTr 197S to Nilay 1976.
Temperature regimes were modelled on those of the natural environment
and final temperatures were 2, 4, and 60C. Final wintering temperature
did not noticeably influence mortality rates. Long fish survived the
period of low temperature better than did shorter ones. Body ratios
of dry weight/wet weight, lipid weight/dry weight, and ignitable weight/
dry weight all decreased during wintering, and the data indicated that
there may be critical percentages of dry weight/wet weight and ignitable
weight/dry weight below which these fish will die.

Otwell, W.S., and J.V. Merriner. 1975. Survival and growth of juvenile
striped bass, @Iorone saxatilis, in a factorial experiment with temper-
ature, salinit-yand -age. Trans. Am. Fish. Soc. 104:560-566.

Abstract

A 3n factorial experiment to evaluate the effects of an abrupt trans-
fer of striped bass, Morone saxatilis, from a closed system rearing
facility into a variety o experimental temperature and salinity com-
binations demonstrated a considerable hardiness of juvenile fish.
Striped bass less than two months old had over 80% survival during seven
days subsequent to acute introduction from the rearing facility into

C-10

temperatures of 180C or 240C and salinities of 4 ppt or 12 ppt.
Relative growth in these treatments was monitored, and the results
suggest a degree of flexibility for stocking programs into estuaries.

Youngs, W.D., and D.S. Robson. 197S. Estimating survival rates from tag
returns: Model tests and sample size determination. J. Fish. Res. Bd.
Canada 32:236S-2371.

Abstract

This paper brings to the attention of fishery biologists a method
of estimating survival and exploitation rates when a series of tag
recaptures from angler-killed fish is available; this model is appropri-
ate only for recaptures that are removed from the population. Tests
for the appropriateness of the model and methods for determining sample
size are presented. The test for the model is imperative since estimates
of survival would-be meaningless if data do not conform to model.

Stock Assessment

Austin, H.M., and C.R. Hickey. 1978. Predicting abundance of striped bass,
C@
Morone saxatilis, in New York from modal lengths. Fish. Bull. 76:467-473

Abstract

The abundance of cohorts for any given year class of striped bass,
Morone saxatilis, prior to their leaving Chesapeake Bay is inversely
related to tH-emodal length of fish in that year class 2 years later in
New York waters. The modal length of bass in their third year migrating
into the New York area is a reliable index of the abundance of that
year class. When back extrapolated modal lengths at the end of the
second year of life are considered for the dominant year classes in the
New York fishery (ages III-VI), a high degree of inverse correlation
is found between age II and modal length and reported landings sug-
gesting that this is an effective method of predicting the abundance
of the stock for the fishery.

Berggren, T.J., and J.T. Lieberman. 1978. Relative contribution of Hudson,
Chesapeake, and Roanoke striped bass, INIorone saxatilis, stocks to the
Atlantic coast fishery. Fish. Bull. 7F--335-345.

Abstract

Ikilorphological characters were used in discriminant analysis to
quantitatively estimate the relative contribution of striped bass,
Morone saxatilis, stocks from various estuaries to the striped bass
fishery alon-g-tSe Atlantic coast. Representative samples of the spawn-
ing stocks of the Hudson River, Chesapeake Bay system, and Roanoke
River were collected and counts and measurements were taken on each

C-11

specimen. Discriminant functions based on five morphological characters
correctly classified approximately 75% of the specimens. The effective-
ness of three types of estimates based on these functions in accurately
estimating stock proportions was investigated in a simulation study.
Results of the simulation study indicated which type of estimate was
least biased. A sampling design using geographical and temporal strata
was then employed to sample the Atlantic coastal fishery from Cape
Hatteras, N.C., to Maine. Observations for the morphological characters
were taken on collected fish and the resulting data entered into dis-
criminant functions obtained from spawning-stock collections. The speci-
mens were classified by area of origin and the.three types @f estimates
of relative contribution of the Hudson, Chesapeake, and Roanoke stocks
were obtained. Results indicated that the Chesapeake stock was the major
contributor to the Atlantic coastal striped bass fishery and the Hudson
and Roanoke stocks were minor contributors.

Buck, D.H., and C.F. Thoits. 1965. An evaluation of Petersen estimation
procedures employing seines in 1-acre ponds. J. Wildl. Manage.
29:598-621.

Abstract

Seines were used to make Petersen estimates of various size and age
components of single-species fish populations in 15 1-acre ponds at the
former IMcGraw Hydrobiological Laboratory near Dundee, Illinois. All
estimates were checked by draining censuses. Each of the five species
of fish involved was by itself in three of the ponds. Combinations of
fin-clip marks permitted a maximum of seven separate estimates for each
of the several size or age-groups of each species; together these con-
stituted a total of 274 individual estimates. Analyses applied to the
estimates and census data indicated that significant bias occurred most
often in the bluegill (Lepomis macrochirus) estimates (70 percent of 37),
least commonly among yellow perch (Perc-a-Ylavescens) estimates (13 per-
cent of 63), intermediately in smallmouth bass licropterus dolomieui)
and largemouth bass C@,Iicropterus salmoides) estimates716 percent S
and 18 percent of 61, respecFi@yely), and Brown bullhead (Ictalurus nebu-
losus) estimates (25 percent of 56). Although errors of estimate were
To-west for perch, 54 percent of separate estimates showed errors greater
than 10 percent; in bluegills, where errors of estimate were highest,
78 percent showed errors greater than 10 percent. Only among yellow
perch were marked fish commonly recaptured in proportions less than their
true abundance in the population. This resulted in more overestimations
than underestimations for perch. In all other species, underestimations
exceeded overestimations because of the recapture of disproportionately
large numbers of marked fish. These data were believed to indicate that
marking did not influence catchability. The cause of bias possibly dif-
fered with dif-ferent species'and was never conpletely assignable; however,
.bias may have been caused (1) by the greater susceptibility to capture
of certain segments of populations, and (2) by the existence of certain
groups of fishes living in areas of high vulnerability to which they
returned when marked and released. The variable quality of the estimates
based on data collected by seining demonstrates that such estimates may
be unreliable.

C-12

Kimura, D.K. 1976. Estimating the total number of marked fish present in
a catch. rrans. Am. Fish. Soc. 105:664-668.

Abstract

The simple ratio of the total number of marked fish present in a
catch is described and the variance estimates discussed. A maximum
likelihood estimate of the total number of marks present in a catch is
derived which combines both marks found in samples uid marks voluntarily
returned'from the non-sampled portion of the catch. 1he use of this
estimate is illustrated with data from the salmon sport fishery in Puget
Sound.

Kjelson, M.A. 1978. Estimating the size of juvenile fish populations in
southeastern coastal-plain estuaries. Pages 71-89 In W. Van Winkle
(ed.),. Assessing the I!@2act of Power-Plant-Induced M-ortality on Fish
Populati6ns. Sponsored by Oak Ridge National Laboratory, Energy
Research and Development Administration, and Electric Power Research
Institute.

Abstract

Understanding the ecological significance of man's activities upon
fishery resources requires information on the size of affected fish
stocks. The objective of this paper is to provide information to eval-
uate and plan sampling programs designed to obtain accurate and precise
estimates of fish abundance. Nursery habitats, as marsh-tidal creeks
and submerged grass beds, offer the optimal conditions for estimating
natural mortality rates for young-of-the-year fish in Atlantic and
Gulf of Mexico coast estuaries. The area-density method of abundance
estimation using quantitative gears is more feasible than either mark-
recapture or direct-count techniques. The blockage method provides the
most accurate estimates, while encircling devices enable highly mobile
species found in open water to be captured. Drop nets and lift nets
allow samples to be taken in obstructed sites, but trawls and seines are
the most economical gears. Replicate samples are necessary to improve
the precision of density required to improve the accuracy of density
estimates. Coefficients of variation for replicate trawl samples range
from 80 to 150%, while catch efficiencies for both trawls and seines
for many juvenile fishes range from approximately 30 to 70%.

Loesch, J.G., and D.S. Haven. 1973. Estimates of hard clam abundance from
hydraulic escalator samples by the Leslie method. Ches. Sci. 14:215-216.

Abstract

The feasibility of predicting hard clam abundance in a sample area
from subsamples collected with a hydraulic escalator was investigated.
Seven experimental plots, each about 1/2 acre in size, were sampled and
then completely harvested. The error for the difference between estimated
abundance and total c atch varied from 0 to 7.6 percent.

C-13

Messiah, S.N. 1975. Delineating spring and autumn herring populations in the
southern Gulf of St. Lawrence by discriminant function analysis. J. Fish.
Res. Bd. Canada 3Z:471-477.

Abstract

A discriminant function based on three variables, pectoral and
dorsal fin rays and gill rakers, was calculated for Atlantic herring
(Clupea harengu@ harengus) taken from spring- and autLum-spawning
concentrations in the southern Gulf of St. Lawrence and tested for
general applicability by classifying other herring of known origin.
C>
The function was then used to classify herring sampled from feeding
concentrations. Results showed that feeding herring (pre- and post-
spawning) comprised a mixture of spring and autumn herring populations
averaging 51.8 and 48.2%, respectively.
The observed heterogeneity of feeding concentrations, in contrast
to homogeneity of spawning concentrations, confirmed the hypothesis
that spring and autumn herring populations mix during feeding, though
they separate at the onset of spawning. Discriminant function analysis
is most useful for separating herring spawning groups during their
feeding season when overlap of maturation stages prevents separation
by maturation stage alone.

Rhodes, R.J., W.J. Keith, P.J. Eldridge, and V.G. Burrell. 1977. An
empirical evaluation of the Leslie-DeLury method applied to estimating
hard clam, Mercenaria mercenaTia, abundances in the Santee River estuary.,
South Carolln-a. Proc. Natl. Shellfish. Assoc. 67:44-52.

Abstract

This paper estimates the abundance of hard clams, Mercenaria
mercenaria, in the Santee River estuary based upon catch and ef t data
generateT-by hydraulic escalator clam harvesters between 1974 and 1976.
Using the Leslie method, catch per unit of standardized effort at each
time interval was rearessed on the cumulative catch. The resulting
regression equations had regression coefficients (estimates of catch-
ability) of .0006, .0014, and .0006 for the South Santee River, North
Santee River, and North Santee Bay, respectively. There were an esti-
mated 6.4 million, 5.0 million, and 10.7 million clams in the legal
harvestina areas of the South Santee River, North Santee River, and
North Santee Bay, respectively. The density of clams in the preferred
fishing areas varied between 18/m2 and 24/m4.
In this analysis, the Leslie-DeLury method had two major limitations:
first, the lack of effort estimates for specific locations, and second,
significant gear competition. It is suggested this method should be
considered only for supplementing designed, direct sampling.

C-14

APPENDIX D

A bibliography for life history data

on selected Maryland species

D-1

1. Schwartz, F.J., and R. Jachowski. 1965. The age, growth, and
length-weight relationship of the Patuxent River, Maryland icta-
lurid white catfish, Ictalurus catus. Ches. Sci. 6(4):_1216-1_31/.

2. Tsai, C., and G.R. Gibson, Jr. 1971. Fecundity of the yellow
perch, Perca flavescens Ititchill, in the Patuxent River, Maryland.
Gies. 8'cl12(4):270-284.

3. St. Pierre, R.A., and J. Davis. 1972. Age, growth, and mortality
of the white perch, Morone americana, in the James and York Rivers,
Virginia. Ches. Sci. 1_3T4_):272-_28l.

4. Norcross, J.J., S.L. Richardson, W.H. Massmann, and E.B. Joseph.
1974. Development of young bluefish (Pomatomus saltatrix) and
distribution of eggs and young in Virginian coastal waters.
Trans. Amer. Fish. Soc. 103(4):477-497.

S. White, M.L., and M.E. Chittenden, Jr. 1977. Age determination,
reproduction, and population dynamics of the Atlantic croaker,
Micropogoni undulatus. Fish. Bull. 75(l):109-123.

6. Winget, R.R., C.E. Epifanio,T,. Runnels, and P. Austin. 1976.
Effects of diet and temperature on growth and mortality of the
blue crab, Callinectes sapidus, maintained in a recirculatinc,
culture system. Proc. Natl. Sliellfish. Assoc. 66:29-33.

7. Sandifer, P.A. 1975. The role of pelagic larvae in recruitment to
populations of.adult decapod crustaceans in the York River estuary
and adjacent lower Chesapeake Bay, Virginia. Estuarine and Coastal
Mar. Sci. 3:269-279.

8. Whetstone, J.M., and A.G. Eversole. 1978. Predation on hard clams,
14ercenaria mercenaria, by mud crabs, Panopeu herbstii. Proc. Natl.
Shellfis-F- Assoc. 68:42-48.

9. Pratt, D.M., and D.A. CanTbell. 1956. Environmental factors affecting
growth in Venus mercenaria. Limnol. and Oceanog. 1:2-17.

10. Roosenburg, W. 1976. Can the oyster industry learn from livestock
breeders? Proc. Natl. Shellfish. Assoc. 66*-.95-99.

11. @Imcy, R.J. 1962. Life history of the yellow perch, Perca flavescens,
in estuarine waters of Severn River, a tributary of ChesapFake Bay-,
14aryland. Ches. Sci. 3:143-159.

12. Nelson, W.R., M.C. Ingham, and W.E. Schaaf. 1977. Larval transport
and year-class strength of Atlantic menhaden. Fish. Bull. 75(l):23-44.

D-2

13. Forney, J.L. 1971. Development of dominant year classes in a yellow
perch population. Trans. Amer. Fish. Soc. 100(4):739-749.

14. Pacheco, A.L. 1962. Movements of spot, Leiostomus xanthurus, in the
lower Chesapeake Bay. Ches. Sci. 3:256-257.

15. Pfitzenmeyer, H.T. 1962. Periods of spawning and setting of the soft-
shelled clam, Mya arenaria, at Solomons, Maryland. Ches. Sci. 3:114-120.

16. Krantz, G.E., and J.V. Chamberlin. 1978. Blue crab predation on
cultchless oyster spat. Proc. Natl. Shellfish. Assoc. 68:38-41.

17. Knudsen, E.E., and W.H. Herke. 1978. Growth rate of marked juvenile
Atlantic croakers) Micropogon undulatus, and length of stay in a
coastal marsh nursery in southwest Louisiana. Trans. Amer. Fish. Soc.
107(l):12-20.

18. Richards,- C.E. 1965. Availability patterns of marine fishes caught
by charter boats operating off Virginia's eastern shore, 1955-1962.
Ches. Sci. 6(2):96-108.

19. Otwell, W.S., and J.V. Merriner. 1975. Survival and growth of juvenile
striped bass, mOrone saxatilis, in a factorial experiment with tempera-
ture, salinity, and age. Trans. Amer. Fish. Soc. 104(3):560-566.

20. Kissil, G.W. 1974. Spawning of the anadromous alewife, Alosa pseudo-
harengus, in Bride Lake, Connecticut. Trans. Amer. Fish. Soc.
103(C2):312-317.

21. Kaneko, T., R.R. Colwell, and F. Hamons. 1975. Bacteriological studies
of Wicomico River soft-shell clam (Mya arenaria) mortalities. Ches. Sci.
16(1):3,13.

22. Nicholson, W.R. 1978. Movements and population structure of Atlantic
menhaden indicated by tag returns. Estuaries 1(3):141-150.

23. Neves, R.J., and L. Depres. 1979. The ocean migration of American
shad,, Alosa sapidissima, along the Atlantic coast. Fish. Bull.
77(l):199-212.

24. Nicholson, W.R. 1971. Coastal movements of Atlantic menhaden as
inferred from changes in age and length distributions. Trans. Amer.
Fish. Soc. 100(4):708-716.

25. Nelson W.R., and C.H. Walburg. 1977. Population dynamics of yellow
perch (Perca flavescens), sauger (Stizostedion canadense), and walleye
(S. vitreum vitreum) in four main stem Missouri River reservoirs.
J. Fish. Res. Bd. Canada 34:1748-1763.

26. Ney, J.J., and L.L. Smith, Jr. 1975. First-year growth of the yellow
perch, Perca flavescens, in the Red Lakes, Minnesota. Trans. Amer.
Fish. Soc. l04 (4):7l8-725.

D-3

27. Cooper, R.A. 1961. Early life history and spawning migration of
the alewife, Alosa pseudoharengus. M.S. Thesis, Univ. of Rhode
Island, Kingston, R.I.

28. Kendall, A.W., Jr., and L.A. Walford. 1979. Sources and distributions
of bluefish, Pomatomus saltatrix, larvae and juveniles off the east
coast of the United States. Fish. Bull. 77(l):213-227.

29. Markle, D.F. 1976. The seasonality of availability and movements
of fishes in the channel of the York River, Virginia. Ches. Sci.
17(l):5O-55.

30. Kendall, A.W., Jr., and J.W. Reintjes. 1975. Geographic and hydro-
graphic distribution of Atlantic menhaden eggs and larvae along the
middle Atlantic coast from R.V. Dolphin cruises, 1965-66. Fish. Bull.
73(2):317-335.

31. MacKenzie, C.L., Jr. 1977. Predation on hard clam (Mercenaria mer-
cenaria) populations. Trans. Amer. Fish. Soc. l06(6):53O-537.

32. Howe, A.B., P.G. Coates, and D.E. Pierce. 1976. Winter flounder
estuarine year-class abundance, mortality, and recruitment. Trans.
Amer. Fish. Soc. 105(6):647-657.

33. Havey, K.A. 1961. Restoration of anaqdromous alewives at Long Pond,
Maine. Trans. Amer. Fish. Soc. 90(3):281-1.86.

34. Feder, H.M., and A.J. Paul. 1974. Age, growth, and size-weight rela-
tionships of the soft-shell clam, Mya arenaria, in Prince William Sound,
Alaska. Proc. Natl. Shellfish. Assoc.64:45-52.

35. Frame, D.W. 1974. Feeding habits of young winter flounder (Pseudo-
pleuronectes americanus): prey availability and diversity. Trans.
Amer. Fish. Soc. lO3(2):261-269.

36. Eldridge P.J. W. Waltz, R.C. Gracy, and H.H. Hunt. 1976. Growth
and mortality rates of hatchery seed clams, Mercenaria mercenaria,
in protected trays in waters of South Carolina. Proc. Natl. She11fish.
Assoc. 66:13-20.

37. Dietrich, C.S., Jr. 1979. Fecundity of the Atlantic menhaden, Bre-
voortia tyrannus. Fish. Bull. 77(l):308-311.

38. Davis, J., J.V. Merriner, W.J. Hogman, R.A. St. Pierre, and W.L. Wilson.
1971. Annual Progress Report, Anadromous Fish Project. Proj. No.
Virginia AFC 7-1. Virginia Inst. of Mar. Sci., Gloucester Pt., Va. 100 pp.

39. Brousseau, D.J. 1978. Spawning cycle, fecundity, and recruitment in a
population of soft-shell clam, Mya arenaria, from Cape Ann, Massachu-
setts. Fish. Bull. 76(l):l55-166.

40. Berggren, T.J., and J.T. Lieberman. 1978. Relative contribution of
Hudson, Chesapeake, and Roanoke striped bass, Morone saxatilis, stocks
to the Atlantic coast fishery. Fish. Bull. 76(2):335-345.

D-4

41. Brazo, D.C., P.I. Tack, and C.R. Liston. 1975. Age, growth, and
fecundity of yellow perch, Perca flavescens, in Lake Michigan near
Ludington, Michigan. Trans. Amer. Fish. Soc. 104(4):726-730.

42. Brousseau, D.J. 1978. Population dynamics of the soft-shell clam,
Mva arenaria. @-,far. Biol. 50:63-71

43. Andrews, J.D., and J.L. Wood. 1967. Oyster mortality studies in
Virginia. VI. History and distribution of Ntinchinia nelsoni, a
pathogen of oysters, in Virginia. Ches. Sci. 8(l):1-13.

44. Leggett, W.C. 1977. Ocean migration rates of American shad (Alosa
C>
sapidissijna). J. Fish. Res. Bd. Canada 34:1422-1426.

45. Dryfoos, R.L., R.P. Cheek, and R.L. Kroger. 1973. Preliminary
analyses of Atlantic menhaden, Brevoortia tyrannus, migrations,
population structure, survival and exploitation r tes, and avail-
ability as indicated from tag returns. Fish. Bull. 71(3):719-734.

46. Cargo, D.G. 1958. The migration of adult female blue crabs, Calli-
nectes sapidus Rathbun, in Chincoteague Bay and adjacent waters.
Mar. Res. 16(3):180-191.

47. Henry, K.A. 1971. Atlantic menhaden (Brevoortia tyrannus) resource
and fishery-analysis of decline. U.S. Dept. of Coii;5erce; eattle,
IVA. NOAA Technical Report MFS SSRF-642. *32 pp.

48. Richkus, W.A. 1975. 'ilw1igratory behavior and growth of juvenile ana-
dromous alewives, Alosa pseudoharengj@, in a Rhode Island drainage.
Trans. Amer. Fish.793-c. 104(3):483-493.

49. Developmental Sciences, Inc. 1979. Maryland's Chesapeake Bay commer-
cial fisheries. Prepared for Energy and Coastal Zone Admin. IMDNR by
Developmental Sciences, Inc., Sagamore, Mass. Edited by M.M. Bundy
and J.B. Williams. 119 pp.

50. Marasco, R.J. 1972. The Chesapeake Bay fisheries: An economic
profile. Agricultural-Experiment Station, Univ. of Md., College
Park., Md. IMP No. 802. 203 pp.

51. Powell,, A.B. 1979. Personal communication. National Marine Fisheries
Service, BeaufortY North Carolina.

52. Jones, P.W.., F.D. Martin, and J.D. Hardy, Jr. 1978. Development of
Fishes of the Mid-Atlantic Bight, Vol. I. Biolouical Services Program,
Fish and M11-Idlife Service, U.S. Dept. of the Interior. 366 pp.
53. Hardy, J.D., Jr. 1978a. Devel2pment of Fishes of the Mid-Atlantic
II. Biological Services Program- Fish anU WildMe Ser-
Bight, Vol
vice, U.S. Dept. of the Interior. 458 pp.

54. Hardy, J.D., Jr. 1978b. Development of Fishes of the Mid-Atlantic
Bight, Vol. III. Biolooic-a-f -Services Program, Fish and Wildlife
C)
Service, U.S. Dept. of the Interior. 394 pp.

D-5

S5. Johnson, G.D. 1978. Development of Fishes of the Mid-Atlantic
Bight, Vol. IV. Biological Services Program, Fisri and WilTl-i7e-
Service, U.T.-Dept. of the Interior. 314 pp.
56. Martin, F.D., and G.E. DrewTy. 1978. Develo2ment of Fishes of the
Ndd-Atlantic Bight, Vol. VI. Biologic rvices Program, Fish and
ITildlife Service, U.S. Dept. of the Interior. 416 pp.

57. Smith, H.M. 1907. The Fishes of North Carolina, Vol. II. North
Carolina Geological and Economic Survey, Raleigh, N.C. 453 pp.
C>

58. Lippson, A.J., M.S. Haire, A.F. Holland, F. Jacobs, J.Jensen,
R.L. Moran-Johnson, T.T. Polgar, and W.A. Richkus. 1979. Environ-
mental Atlas of the Potomac Estuary. Prepared for Md. Dept-.-o-f-
Natural Resources, Power Plant Siting Program by-Martin Marietta
Corporation, Environmental Center. 280 pp.

59. Hildebrand, S.F., and W.C. Shroeder. 1928. Fishes of the Chesa-
peake Bay. U.S. Bur. Fish. Bull. Volume 53, Part 1. 388 pp.

60. Galtsoff, P.S. 1964. The American yster, Crassostr2a virginica
Gmelin. Fish. Bull. 64:1-480.

61. Setzler, E.1M., W.R. Boynton, K.V. Wood, H.H. Zion, L. Lubbers,
N.K. Mountford, P. Frere, L. Tucker, and J.A. Mihursky. 1979.
Synopsis of biological data on striped bass, Morone saxatilis
OValbaum). Cont. No. 802 (draft). Center for-Environmental and
Estuarine Studies, Univ. of 'L'Td. 209 pp.

62. Dawson, C.E. 1958. A study of the biology and life history of the
spot, Leiostomus xanthurus Lacepede, with special reference to South
Carolina. Contr. Bears Fluff Lab., Wadmalaw Island 28:1-48.

63. Lewis,, R.M., and R.R. Bonner, Jr. 1966. Fecundity of striped bass,
Roccus saxatilis (Walbaum). Trans. Amer. Fish. Soc. 95:328-331.

64. Jackson, H.W., and R.E. Tiller. 19S2. Preliminary observations on
spawning potential in striped bass (Roccus saxatilis . Md. Dept.
Res. and Ed. Publ. No. 93. 16 pp. -

65. Clark, J.R., and S.E. Smith. 1969. Migratory fish of the Hudson
estuary. In G.J. Lauer (ed.), Hudson River Ecology. N.Y. Dept. of
Envir. Con@_erv. 473 pp.

66. Van Engel, W.A. 1958. The blue crab and its fishery in Chesapeake
Bay. Comm. Fish Review 20(6):6-17.

67. Carriker, M.R. 1961. Interrelation of functional morphology, behavior,
and autecology in early stages of the bivalve, Mercenaria mercenaria.
J. E. Mitchell Soc. 77(2):168-241.

68. Meritt, D.W. 1977. Oyster spat set on natural cultch in the Maryland
portion of the Chesapeake Bay (1939-1975). Horn Point Environmental
Laboratories, Cambridge, Maryland. UINICEES Special Report No. 7. 30 pp.

D-6

69. Youna,, L.G. 1972. The survivorship'of Crassostrea virginica (Gmelin)
larvae of Clambank Creek, South Carolina. M.S. Thesis, Univ. of South
Carolina, Columbia, S.C. 48 pp.

Loesch, J.G., and W.A. Lund, Jr. 1977. A contribution to the life
history of the blueback herring, Alosa aestivalis. Trins. Amer. Fish.
Soc. 106(6):583-589.

71. Johnson, H.B., D.IV. Crocker, B.F. Holland, J.W. Gilliken, D.L. Taylor,
and M.IV. Street. 1978. Biology and management of mid-Atlantic ana-
dromous fishes under extended jurisdiction. Annual Report, Anadromous
Fish Project, 1978. North Carolina Dept. Nat. Res. and Comm. Dev., and
Va. Inst. Mar. Sci., Gloucester Point, Va. 175 pp.

72. Maurer, D., L. Watlin a, and R. Keck. 1971. The Delaware oyster indus-
try: A reality? Trans. Amer. Fish. Soc. 100(l):100-111.

73. Sulkin, S.D. 1973. Blue crab study in Chesapeake Bay, Maryland.
Natural Resources Institute, Univ. of Md., Ches. Biol. Lab., Solomons,
Md. Ref. No. 73-94. 82 pp.

74. Lippson, R.L. 1971. Blue crab study in Chesapeake Bay, Maryland.
Natural Resources Institute, Univ. of Md., Ches. Biol. Lab., Solomons,
Md. Ref. No. 71-9. 18 pp.

75. Carter, N., and H. Speir. 1979. Personal communication. Tidewater
Fisheries Tidewater Administration ' I
114d. Department of Natural Resources.

76. Hibbert, C.J. 1977. Growth and survivorship, in a tidal-flat population
of the bivalve,Mercenaria mercenaria, from Southhampton water. Mar. Biol.
44:71-76.

77. Wells, H.W. 1957. Abundance of the hard clam, Mercenaria mercenaria,
in relation to environmental factors. Ecology 38:123-10F - 1
78. Hibbert, C.J. 1977. Energy relations of the bivalve, '114ercenaria mercen.-
aria, on an intertidal mudflat. ?4ar. Biol. 44:77-84.

79. Talbot., G.B. 1954. Factors associated with fluctuation in abundance
of Hudson River shad. Fish. Bull.,56:373-413.

80. Woodin, S.A. 1976. Adult-larval interactions in dense infaunal assem-
blages: Patterns of abundance. J. Mar. Res. 34:25-41.

81. Krantz., G. 1979. Personal communication. Hom Point Environmental
Laboratory, University of Maryland.

82. Merriner, J.1'. 1976. Anadromous fishes of the Potomac estuary.In
IV. Mason and K. Flynn (eds.) The Potomac Estua@z: Trends and
Options. Interstate Commission on tNe -Potomac River Basin and Mary-
ldx_@Power Plant Siting Program. 140 pp.

D-7

83. Nelson, W.' 1979. Personal communication. Beaufort Laboratory,
National Marine Fisheries Service.

84. 'Scarlett., P. 1979. Personal commziication. New Jersev Division of
Fish, Shellfish and Wildlife.

8S. Massman,, III.H. 1963. Age and size composition of weakfish, Cynoscion
regalis, from pound nets in Chesapeake Bay, Virginia, 1954-19-58.
Ches. Sci. 4:43-51.

86. Perlmutter, A.W., W.S. Miller, and J.C. Poole. 1956. The weakfish,
Cynoscion regalis, in New York waters. N.Y. Fish and Game J. 3:1-43.
87. Merritt, D.W. 1977. Oyster spat on natural cultch in the Maryland
portion of the Chesapeake Bay kl939-1975). Center for Environmental
and Estuarine Studies, University of Maryland, Solomons, '111d. CEES
Species Report No. 7.

88. Mansueti, R.J. 1961. Movements, reproduction, and mortality of the
white perch, Roccus americanus, in the Patuxent estuary, Maryland.
Ches. Sci. 2:TT2---',-1-05.

89. Coates, P.G.9 A.B. Howe, and A.E. Peterson. 1970. Analysis of winter
flounder tagging off Massachusetts, 1960-1965. Div. of Mar. Fish.,
0 T
Dept. of Natural Resources of @-Iassachusetts.

90. Brousseau, D.J. 1979. Analysis of growth rate in Mya arenaria using
the von Bertalanffy equation. Mar. Biol. 51:[email protected]

91. Darnell, RJ4. 1959. Studies of the life history of the blue crab
(Callinectes sapidus Rathbun) in Louisiana waters. Trans. Amer. Fish.
Soc. 83:294-304.

92. Chittenden, M.E., Jr. 1975. Dynamics of American shad, Alosa sRidis-
sima, runs in the Delaware River. Fish. Bull. 73(3):487-494. '

93. Dame, R.F. 1971. The ecological energies of growth, respiration, and
assimilation in the intertidal American oyster, Crassostrea virginica.
(Gmelin). Ph.D. Thesis. Univ. of South Carolin-a-,-C-o-lU@E-ia, S.C.
81 pp.

94. Boone, J.G. 1979. Personal communication. Maryland Department of
Natural Resources, Annapolis, Maryland.

95. Early, S. 1979. Personal communication. 'Maryland Department of Natural
Resources, Annapolis, Maryland.

D-8

I
I
I
I
I
I
I
APPENDIX E I
A KEYWORD AND SPECIES INDEX TO APPENDICES A-D
I
I
I
- I
I
I
I
I
I
.. E-1 I
I

Age structure, A-9, B-2, B-3, C-2

Alosa aestivalis (blueback herring), D-7

Alosa pseudoharengus; (alewife), D-3, D-4, D-5

Alosa sapidissima (American shad), A-177, C-2, D-3, D-7, D-8

Anchoa mitchelli (bay anchovy), C-10

Autoregressive models, A-5O, A-125

Bagrus docmac (African catfish), A-117

Baleanoptera musculus; (blue whale), A-161

Baleanoptera physalis (Antarctic fin whale), A-6, A-37, A-161

Beverton-Holt yield-per-recruit models, A-10, A-19, A-20, A-31, A-38
A-51, A-84, A-92, A-117, A-168, A-169, A-191, B-5

Bioeconomic models, A-7, A-11, A-12, A-13, A-16, A-24, A-29, A-35
A-36, A-37, A-38, A-39, A-42, A-43, A-49, A-70, A-116, A-119
A-139, A-163, A-164, A-172, A-173, A-174

Bioenergetics models, A-73, A-99, A-113, A-179

Brevoortia patronus (Gulf menhaden), C-10

Brevoortia tyrannus; (Atlantic menhaden), A-122, A-154, B-6, B-7, C-6
C-7, C-8, D-2, D-3, D-4, D-5

By-catch, A-25

Callinectes sapidus (blue crab), C-5, D-2, D-3, D-5, D-7, D-8

Callorhinus ursinus; (fur seal), A-31, A-52

Carassius auratus (goldfish), A-120

Clupea harengus harengus (Atlantic herring), A-194, C-14

Cohort analysis, A-140, A-141, A-169, A-183, A-194, A-195

Coregonus; clupeaformis (whitefish), A-118, A-132

Crassostrea virginica (comon oyster), A-108, B-4, B-14, D-3, D-5
D-6 D-7 D-8

E-2

Decision models, A-41, A-109, A-116, A-137, A-138 A-142, A-181

Ecosystem models, A-10, A-32, A-133, A-134, A-144, A-159

Effort submodels, B-6, B-7, B-8, B-9

Ekman transport, A-122

Engraulis mordax (north ern anchovy), A-161

Environmental quality, A-5, A-32, A-77, A-186

Estuarine fisheries, A-70

Euthynnus pelamis (skipjack), A-161

Freshwater finfish communities, A-3, A-10, A-75, A-81, A-82, A-121, A-123

Gadus morhua (cod), A-105, A-110, A-161

Growth submodels, A-15, A-73, A-87, A-167, B-11, B-12, B-13, B-14, B-15
B-16, B-17, B-18, B-19, B-20, B-21, C-2, C-3, C-4, C-5

Hatchery, A-108, A-181

Hippoglossoides elassodon (flathead sole), B-14

Hippoglossoides platessoides (American plaice), A-113

Homarus americanus (American lobster), A-49, A-58, A-67, A-115, A-125

Ictalurus catus (white catfish), C-5, D-2

Ictalurus nebulosus (brown bullhead), C-12

Ictalurus punctatus; (channel catfish), A-120

Illex spp. (squid), A-168

Lake fisheries, A-3, A-23, A-28, A-34, A-40, A-75, A-81, A-82, A-118, A-121
A-122, A-123, A-132, A-144, A-152, A-180

E-3

Lasioderma serricorne (cigarette bettle), A-100

Leiostomus xanthurus (spot), D-3, D-6

Le2omis cyanellus (green sunfish), A-75

Lepomis macrochirus (bluegill), A-75, A-84, A-89, C-12

Lepomis microlophus (redear sunfish), A-75

Leslie matrix, A-17, A-33, A-46, A-47, A-77, A-100, A-101, A-102, A-127
A-184, B-22

Limanda ferruginea (yellowtail flounder), A-110, A-166, A-167
Loligo spp. (squid), A-168

Macoma balthica (Macoma clam), B-14

Maximum equilibrium yield MEY), A-12, A-13

Maximum sustainable yield (MSY), A-4, A-5, A-6, A-11, A-12, A-13, A-16
A-18, A-20, A-26, A-31, A-35, A-43, A-48, A-52, A-53, A-57, A-59
A-61, A-67, A-69, A-83 , A-84, A-97, A-115, A-119, A-128, A-136
A-154, A-155, A-156, A-157, A-158, A-162, A-169, A-187, A-188, A-189
A-190, A-195

Melanogramus aeglefinus (haddock), A-161

Mercenaria mercenaria (hard clam), B-17, C-3, C-13, C-14, D-2, D-4
D-6, D-7

Micropogonias undulatus (Atlantic croaker), C-2, C-3, C-10, D-2, D-3

Micropterus dolomieui (smallmouth bass), A-34, B-15, C-10, C-12

Micropterus salmoides (largemouth bass), A-34, A-126, A-127, B-3,
C-4, C-9, C-12

Migration, C-6, C-7

Morone americana (white perch), C-4, D-2, D-8

Morone saxatilis (striped bass), A-46, A-47, A-54, A-55, A-185, B-33
C-10, C-11, D-3, D-4, D-6

Morphoedaphic index (MEI), A-3, A-23, A-28, A-81, A-82, A-118, A-120,
A-121, A-123, A-132, A-144, A-152, A-180, A-193

Mortality submodels, A-15, A-61, A-73, A-84, B-S, B-21, B-22, B-23, B-24
B-25, B-26, B-27, B-28, B-29, C-7, C-8, C-9, C-10, C-11

E-4

Multinational fisheries, A-12

Multispecies simulation models, A-10, A-22, A-94, A-95, A-96, A-131
A-134, A-190

Mya arenaria (soft-shell clam), B-9, B-10, B-12, B-13, C-3, D-3, D-4,
D-5, D-8

Mytilus edulis (ribbed mussel), C-9

Network theory, A-106, A-107

North Sea fisheries, A-10, A-69

Oceanic fisheries, A-20, A-21, A-25, A-26, A-59, A-60, A-94, A-95, A-96
A-175

Oncorhynchus nerka (sockeye salmon), A-7, A-9, A-24, A-56, A-86, A-97, A-147

Oncorhynchus tshawytscha (chinook salmon), C-8

Optimal equilibrium yield (OEY), A-11

Optimal sustainable yield (OSY), A-31, A-35, A-43

Optimization techniques, A-S, A-42, A-76, A-79, A-91, A-92, A-109, A-128,
A-137, A-138, A-139, A-142, A-150, A-153

Pagophilus; groenlandicus (harp seal), A-104, A-195
Parlithodes camschatica (Alaskan king crab), A-26

Penaeus aztecus (brown shrimp), A-70, A-182

Peprilus triancanthus (butterfish), C-2
Perca flavescans (yellow perch), A-3, A-28, A-90, B-10, B-11, B-16, C-4
C-12, D-2, D-3, D-5

Pest management, A-151

Phoca vitulina (Dutch fur seal), A-62

Pomatomus saltatrix (bluefish), C-6, D-2, D-4

Predator-prey models, A-73
Pseudopleuronectes americanus (winter flounder), A-99, C-7, C-8, C-9, D-4
D-8

E-5

Rattus rattus (brown rat), A-101

Recreational fisheries, A-34, A-142, B-7, B-8, B-9

Resource allocation, A-56

Ricker yield-per-recruit models, A-15, A-31, A-51, A-84, A-85, A-103
A-105, A-110, A-122, A-135, A-145, A-146, A-147, A-154, A-169, A-178

Riverine fisheries, A-7, A-9, A-24, A-40, A-193

Salmo gairdneri (rainbow trout), A-120

Salmo salar (Atlantic salmon), B-13, B-14

Salvelinus fontinalis (brook trout) A-17

Salvelinus namaycus (lake trout), A-18, A-161

Sardinops caerlea (sardine), A-59, A-148

Sardinops sagnax (Pacific sardine) A-161

Scomber scombrus (mackerel), A-50, A-103, A-189

Shellfish, A-26, A-49, A-58, A-67, A-70, A-108, A-115, A-116, A-125
B-4, B-9, B-10, B-12, B-13, B-14, C-3, C-5, C-9, D-2, D-3, D-4
D-5, D-6, D-7, D-8

Simulation models, A-9, A-10, A-l5, A-21, A-22, A-26, A-30, A-33, A-46
A-47, A-53, A-54, A-55, A-60, A-62, A-63, A-70, A-74, A-80, A-86
A-87, A-89, A-90, A-94, A-95, A-96, A-97, A-98, A-103, A-104, A-106
A-107, A-109, A-111, A-113, A-114, A-122, A-124, A-126, A-127,
A-131) A-133, A-134, A-142, A-144, A-148, A-l5l, A-l59, A-160, A-161
A-165, A-166, A-167, A-168, A-169, A-170, A-171, A-174, A-176, A-179
A-185, A-186, A-190, A-192, A-194

Spatial models, A-26

Statistical models, A-3, A-14, A-23, A-28, A-44, A-50, A-58, A-65, A-81,
A-82, A-105, A-118, A-120, A-1121, A-122, A-123, A-125, A-132, A-144.
A-152, A-166, A-169, A-175, A-177, A-178, A-180, A-182, A-195

Stenella attenuata (bridled dolphin), A-171

Stizostedion canadense (sauger) , A-118, D-3

Stizostedion lucioperca (polish pikeperch) , A-23
Stizostedion vitreum glaucum (blue pike) , A-9

Stizostedion vitreum vitreum (walleye), A-3, A-28, A-90, A-118, A-160
B-16, D-3

E-6

Stochastic, A-45, A-76, A-79, A-104, A-107, A-114, A-148, A-160, A-178

Stock assessment, A-2, C-11, C-12, C-13, C-14

Stock-recruitment, A-6, A-7, A-44, A-61, A-169, B-10, B-32, B-33

Surplus production models, A-4, A-5, A-11, A-12, A-13, A-16, A-48, A-49
A-57, A-59, A-61, A-67, A-69, A-83, A-84, A-97, A-115, A-119, A-128
A-136, A-155, A-156, A-157, A-158, A-162, A-169, A-187, A-188, A-189
A-190, A-195, B-22, B-30, B-31
Tautogolabrus adspersus (cunner), A-77

Theoretical models, A-9, A-11, A-12, A-13, A-16, A-30, A-35, A-36, A-38
A-39, A-51, A-57, A-65, A-80, A-100, A-101, A-114, A-151, A-170,
A-171, A-176, A-187

Thunnus albacares (tuna), A-59, A-60, A-67, A-136, A-155, A-156, A-157
A-161

Thunnus obesus (bigeye tuna), A-161

Thunnus thynnus thynnus (Atlantic tuna), B-29

Tilapia spp., A-120, A-152

Time lags, A-30, A-37, A-115, A-187, A-189

Yield-per-recruit models, A-10, A-15, A-18, A-20, A-31, A-38, A-51, A-84,
A-85, A-92, A-103, A-105, A-110, A-117, A-122, A-135, A-145, A-146
A-147, A-154, A-168, A-169, A-178, A-191, B-5

E-7

I I
I
I
I
I
I
I
I
I
i
I
i
I
I
I
I
I
t-'-- - --L I
.-IIIIINIIIIIIIIEN
3 6668 14101 0118 1-