Final programmatic environmental impact statement, fish culture in floating net pens

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

Final Programmatic
Environmental Impact Statement

Fish Culture in Floating Net-Pens
Washinnton Department of Fisheries

Technical Appendices

January 1990

SCHOOL?

TECHNICAL APPENDICES

PLEASE NOTE:

The following technical appendices are intended to provide additional information on the
subject of fish farming. There has been no attempt to include information on all facets
of fish farming, nor has there been any attempt to evaluate the information presented
here. A determination has not been made concerning the applicability of the information
to the situation in Washington. Several comments on the Draft EIS requested additional
information on fish farming. When possible, the requested information was included in
the appendices. The purpose of these appendices is solely to provide information to aid
in the ongoing discussion of the fish farming industry. In addition to the appendices, a
list of sources of recent information on fish farming is included on the following page.
The following is a list of the titles of the appendices:

A Assessment and Prediction of the Effects of Salmon Fish Farm Culture on the
Benthic Community

B Modeling of Particulate Deposition Under Salmon Fish Farms

C Phytoplankton and Nutrient Studies Near Salmon Fish Farms at Squaxin Island,
Washington

D Infectious Diseases of Salmon in the Pacific Northwest

E The Economics of Salmon Farming

F Permits That May Be Required for Aquaculture Projects

G Viral Hemorrhagic Septicemia

H Norwegian and British Columbia Information

I Land-Based Tank Farms

J Legislation Authorizing the EIS

K Effect of Fish Farms on Surrounding Property Values

L Economic Aspects of Salmon Aquaculture

NOTE: Appendices A and E were completed under separate contracts with the
Department of Fisheries.

00
Li-BRARY
"TOA-A/CCEII
Q_-
1990 "013SON AVE.
C"A,(3- SC 2.9408-2623

I I I I 1 0 @

Property Of
NoAA Coastal Services center
Library

ADDITIONAL REFERENCES

Further information on fish farming can be obtained from the following sources:

Alaska Finfish Farming Task Force. 1989. Report to the Alaska legislature, draft for
public comment. 26p.

B.C. Ministry of Environment. 1988. Environmental monitoring program for fish farms.
Prepared by Waste Management Branch and Water Management Branch, British
Columbia Ministry of Environment. 77p.

B.C. Ministry of Environment. 1989. Environmental procedures and guidelines for
marine fish farms. Draft report. 41p.

Institute of Aquaculture, University of Stirling. 1988. The reduction of the impact of
fish farming on the natural marine environment. Prepared for the Nature
Conservancy Council: Scottish Headquarters: Edinburgh. 167p.

Institute for Environmental Studies, University of Washington. 1989. Focus on
aquaculture. The Northwest Environmental Journal. Vol. 5, No. 1.

International Council for the Exploration of the Sea. 1988. Cooperative research report
no. 154. Report of the ad hoc study group on "environmental impact of mariculture."
Copenhagen, Denmark. 83p.

Nature Conservancy Council. 1989. Fishfarming; and the safeguard of the natural marine
environment of Scotland. Nature Conservancy Council: Scottish Headquarters,
Edinburgh. 136p.

Standing Committee on Fisheries and Oceans. 1988. Aquaculture in Canada. Report
to the House of Commons. 129p.

APPENDIX A

ASSESSMENT AND PREDICTION
OF THE EFFECTS OF SALMON FISH FARM CULTURE
ON THE BENTHIC COMMUNITY

ASSESSMENT AND PREDICTION OF THE EFFECTS OF
SALMON NET-PEN CULTURE ON THE BENTHIC ENVIRONMENT

Donald P. Weston, Ph.D.

School of Oceanography, WB-10
University of Washington
Seattle, WA 98195

and

Richard J. Gowen, Ph.D.

Scottish Marine Biological Association
Dunstaffnage Marine Research Laboratory
PO Box 3
Oban, Argyll PA34 4AD
Scotland

Prepared for:

Washington Department of Fisheries

November, 1988

TABLE OF CONTENTS

Acknowledgements . . . . . . . . . . . . . . . . . . . . .
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . iv
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . 2
MATERIALS AND METHODS . . . . . . . . . . . . . . . . . . . 4
RESULTS
Clam Bay site . . . . . . . . . . . . . . . . . . . . . 10
Squaxin Island site . . . . . . . . . . . . . . . . . . 27
DISCUSSION
Clam Bay site . . . . . . . . . . . . . . . . . . . . . 48
Squaxin Island site . . . . . . . . . . . . . . . . . . 49
Assessment techniques . . . . . . . . . . . . . . . . . 50
Model evaluation . . . . . . . . . . . . . . . . . . . . 51
LITERATURE CITED . . . . . . . . . . . . . . . . . . . . . 60

ACKNOWLEDGMEWTS

The authors appreciate the assistance of many individuals
and organizations for providing information or logistical
support essential to completion of this project:

� The farm operators provided feed rate and production
information required for the analyses. The Squaxin
Island Tribe was particularly helpful by providing boats
and operators on several occasions. This work would not
have been possible without the cooperation and assistance
of tribal staff.

� The National Marine Fisheries Service assisted these
studies by providing boats for use in Clam Bay.

� The collection of samples in Clam Bay was made possible
by the help of Mr. Dwayne Karna, Mr. Dave Terpening and
other members of the EPA Region 10 dive team.

� The Washington Department of Fisheries provided patrol
boats for deployment and recovery of current meters.

� The prompt analytical services provided by Dr. Donald
McLusky of the University of Stirling are much
appreciated.

� Finally, the we are grateful to Mr. Ron Westley and Mr.
Eric Hurlburt of the Washington Department of Fisheries
for their support, patience and review of the draft
report.

ABSTRACT

Sediment chemistry and macrofaunal communities were
examined in the vicinity of two salmon mariculture facilities
in Puget Sound. The Clam Bay farm is a relatively large
operation containing 200 to 400 tons of salmon, and has been
in operation for about 13 years. Deposition of feed and feces
beneath the pens has created an area characterized by high
levels of organic carbon and nitrogen and depressed sediment
reduction-oxidation potentials. This area extends under the
pens and out to a distance of 15 to 60 m from the farm
perimeter. The macrofaunal community shows dramatic altera-
tions in this area including the disappearance of most species
characteristic of the natural community and high abundances of
nematodes and an opportunistic polychaete. More moderate
changes in the infaunal community extend at least 150 m from
the farm.
The Squaxin Island farm is a comparatively small facility
(20 to 40 tons of fish on site) and has been operating only
since early 1987. Unlike the Clam Bay site, there was little
effect of the farm on sediment chemistry, even directly under
the pens. The infaunal community shows evidence of disturb-
ance in an area extending from the pen perimeter out to a
distance of 6 m. Within this zone the community appears to
have been undergoing gradual change over the 18 months of farm
operation.
On the basis of these investigations, macrofaunal
community composition appears to be a more sensitive indicator
of benthic impacts than measurement of sediment chemical
parameters alone. Among the sediment parameters evaluated,
redox potential appears to be a valuable tool for rapid and
cost-effective impact assessment, at least in coarse-grained
sediments.
A model which predicts dispersion of feed and feces
from a farm site was tested for agreement with actual
conditions at the Clam Bay and Squaxin sites. The model was
found to be reliable to within a factor of two or less in
predicting the magnitude of organic loading to the seafloor.
At the Clam Bay site and, to a lesser extent, at Squaxin
Island the areas which the model predicted to receive the
greatest input of feed and feces were the same areas showing
the highest degree of sediment enrichment. On a broader
scale, the model appeared reliable in identifying the areal
extent of impact from net-pen culture. After tests at eight
farms the model has predicted enhanced carbon fluxes up to
70 m (and usually less than 30 m) from the pen perimeter.
These predictions are consistent with the results from this
study and other investigations. The model is useful in
identifying sites that would be clearly unsuitable for
culture or others where environmental impacts are likely to
be negligable. There are, however both inherent unknowns
and oversimplifying assumptions in the model, which should
be recognized to avoid indiscriminant application of the

model and misinterpretation of the results. The data base
of current velocity and direction on which model predictions
are based is rarely, if ever, available for siting
decisions, but would be valuable both to environmental
managers and farm operators.

ASSESSMENT AND PREDICTION OF THE EFFECTS OF

SALMON NET-PEN CULTURZ ON THE BENTHIC ENVIRONMENT

INTRODUCTION

Cultivation of salmon in estuaries and coastal embayments
generates substantial quantities of particulate organic wastes
consisting principally of feces and uningested feE!d. It has
been estimated that the production of 1 kg of salmon generates
0.5 to 0.7 kg of particulate waste (Weston, 1986; Gowen and
Bradbury, 1987). For large farms, which produce several
hundred tons of salmon annually, the quantity of particulate
waste settling to the seafloor is considerable. Accumulation
of organic waste on the sea bed has been found to alter some
aspects of sediment chemistry and benthic macrofaunal commun-
ity composition in the vicinity of salmonid culture opera-
tions. The changes in sediment chemistry include increases in
carbon, nitrogen and phosphate content (Hall and Holby, 1986),
depression in sediment reduction-oxidation potentials (Brown
gt al., 1987), and changes in the rates of nutrient cycling
(Kaspar et al., 1988). Shifts in the species composition and
relative abundances of benthic macrofauna have been reported,
occasionally with an azoic zone directly beneath the farm
(Pease, 1977; Brown et al., 1987). Waste accumulation on the
seafloor can have implications for the viability of the farm
itself. The release of hydrogen sulfide from anoxic sediments
and/or related changes in water quality can adversely effect
the health of the cultured fish (Arizono, 1979) and has been
the reason for the closure of some operations (Braaten et al.,
1983).
At the present time there is a rapid growth in the farm-
ing of salmon. The results of this growth will be the expan-
sion of existing farms, some to an annual production of 2,000
to 3,000 tons, and the establishment of multiple farms in
individual embayments. As development proceeds potential
pollution problems require careful scrutiny, and it has
recently been stated that there is a need for models which can

be used to predict ecological impact before establishment of a
farm (Rosenthal et al., 1987). The development of such
models, together with an understanding of the effects of
organic wastes on the benthos, is essential to ensure that
this form of mariculture does not cause broad ecological
change. An additional application of such model would be in
resource management, to ensure that farming is not conducted
in areas which, through ecological change, can not sustain the
long-term use of the site.
This paper describes an assessment of the intensity and
spatial extent of impact on the benthos of two farm sites in
Puget Sound, Washington. The sites chosen for study differ
greatly in physical conditions, farm size and duration of
operation, and potentially offer a broad range in the scale of
impact. The study was also intended to further test a
sedimentation model designed to predict the dispersal of solid
wastes from fish farms (Gowen et al, in press). A preliminary
test of the model (Gowen et al., 1988), using sediment redox
potentials as an indicator of organic pollution, suggested
that the model could be used to predict the spatial extent and
severity of ecological disturbance of the benthos and in the
selection of suitable sites.

MATERIALS AND METHODS

Field sampling and model verification was performed at
two salmon net-pen operations in Puget Sound. The Clam Bay
farm is a large operation with an annual production of 617
tons and containing 200 to 400 metric tons of salmon at any
given time. The farm has been in operation at the same site
continuously since the early 1970's. The Squaxin Island farm
holds 20 to 30 tons of salmon and has only been in operation
since 1987. There are 53 additional pens located 2,50-500 m
south of the Squaxin Island-farm. These pens are used to hold
juvenile salmon on a seasonal basis, and were largely unused
during the period of investigation. Physical characterisics
of each site and details of the Clam Bay and Squaxin farms are
given in Table 1.
Sampling stations were established at each farm at prede-
termined distances along 3 transect lines extending out from
the pens, varying in length from 30 to 165 m (Figure 1).
Field activities included deployment of current meters and
sediment traps, collection of sediments for chemistry and
grain size analysis, measurement of dissolved oxygen in near-
bottom waters, and sampling of benthic macrofauna.

Current measurements - Two Aanderaa current meters were
deployed at each of the farm sites, and set to record at 15
minute intervals for a period of 60 days. The meters were
placed 2.5 and 5.5 m above the seafloor at the Squaxin Island
and Clam Bay sites, respectively.

Sediment traps - Sediment traps were constructed of PVC pip-
ing, 15 cm in diameter and 45 cm in length. An array of three
traps were placed under the pens and at each station along
transects CB1 and SQ1. The distances between the sea bed and
the mouths of the traps were 0.5 and 5.5 m at the 115quaxin and
Clam Bay sites, respectively. Prior to deployment 250 g

Table 1
Physical characteristics of the farms
and surrounding environments

PARAMETER CLAM BAY SQUAXIN ISLAND

Depth of water 10 - 26 4.7 - 5.0
(m at mean lower low water)

Distance between bottom of
pens and sea bed 6 - 22 1.7 - 2.0
(m at mean lower low water)
Area of cages (m2) 14,560 1,184

Duration of operation (years) 13 1.5

Biomass held on farm 200 - 400 20 - 30
(metric tons)
Feed provided (kg-day-1) 2200 - 5900 400 - 600

mo@@5ester 0 '100 200
EPA Lob. Omm"
Scale in Meters

0 Transect
CBI

HHHHHHM
M
HHHHH!

Transect
Transect CB2
CB3
-480
Everett N

Seattle

GCOMO
-J- Transect
S03 0
I ympla 47
1230W 12 0 0 0
Transect
SQ2 9

Transect
S04

0 50 400

Scale in Meters

Figure 1. Locations of the two Puget Sound farm sites investi-
gated. Sampling stations are indicated by dots under
the pens and along three transects at each site.
Locations of the Aanderaa current meters indicated by
M symbol.

of reagent grade salt was added to each trap to retard micro-
bial activity and reduce loss of material by resuspension.
The traps were deployed for 9 days at the Clam Bay site and 15
days at Squaxin Island. Mercuric chloride was added to the
traps at the time of recovery to inhibit microbial activity
prior to analysis. The total amount of particulate material
retained in the traps was estimated by filtering duplicate
aliquots of the homogenized trap contents on to combusted and
preweighed 45 um silver filters. The material retained on the
filters was analyzed for carbon and nitrogen in a Carlo Erba
Model 1106 CHN microanalyzer following vapor phase acidifi-
cation to remove inorganic carbon (Hedges and Stern, 1984).

Sediment sampling - The upper 1 cm sediment stratum was
collected by SCUBA diver (Clam Bay) or Van Veen grab (Squaxin
Island). The sediment was analyzed for carbon and nitrogen
content in the same manner as the sediment trap material,
except that a Perkin Elmer 240 elemental analyzer was used.
Sediment redox potentials were measured in diver-collected
cores following the method of Pearson and Stanley (1979).
Grain size analysis was performed by dry sieving followed by
pipette analysis of the silt and clay fraction.

Dissolved oxygen - Near-bottom water samples were collected 5
to 10 cm above the sediment-water interface by SCUBA divers,
and analyzed by the Winkler titration method (Strickland and
Parsons, 1972).

Benthic macrofauna - Macrofauna samples were collected at each
farm site in connection with independent investigations. The
Squaxin samples were collected as part of the operator's
routine monitoring program which was scheduled so as to sample
concurrently with the present investigation. The Clam Bay
samples were collected as part of an EPA-funded study by the
University of Washington to examine the effects of organic
enrichment on benthic communities (Weston, 1988).

At the Squaxin Island farm macrofaunal samplE!S were
collected at each station along transect SQ1. Three 0.008-m2
cores were collected by diver at each station. The contents
of two of the cores were washed on a 1.0-mm screen, while the
third core was washed on stacked 0.5 and 1.0-mm screens. The
Clam Bay macrofauna samples were collected one year prior to
the present investigations at stations closely approximating
those of transect CB1. Three samples were collected at each
station using a 0.06-m2 spade corer, and washed on a 0.5 mm
screen sieve.

The predicted dispersal of feed and feces from the net-
pens was determined using the sedimentation model of Gowen
-et al. (in press). The model uses a value of 4 crft-sec-1 for
the settling velocity of feces (Warrer-Hansen, 1982), and a
settling velocity for uneaten feed that was determined in a
settling column using feed from the specific farm,. The area
occupied by the farm was divided into a grid of 1 m squares,
and waste production was assumed to be evenly distributed over
the farm area. Solid waste production (in units of g organic
carbon-m-2-hr-1) was calculated as a proportion of the feed
provided. The amount of waste feed was assumed to be 15%,
although the actual wastage varies greatly among farms and is
generally unquantified (see p. 54 for complete discussion).
The fecal production was estimated to be 30% of the ingested
feed (Penczak et al., 1982). Hourly values of current veloc-
ity and direction were used to calculate the horizontal
displacement of feed and feces in each array element using the
equations:

(D) x (V cos 0)

(D) x (V s in 0)

where I and J are the co-ordinates of the waste within the
array at the start of each hour. D (water depth under pens)
divided by U (settling velocity of feed and feces) determines
the time during which horizontal displacement takes place.
V cos 0 and V sin 0 provide the components of the horizontal
displacement of a particle. Thus, the above equations give
the X and Y co-ordinates of a particle on the sea bed in
relation to the position of the farm at the end of each hour.
The model was run using current data over one or more spring-
neap tidal cycles and integrated the dispersal and input of
organic carbon waste over this time period.

RESULTS

CLAM BAY SITE

Fourteen stations were sampled in the vicinity of the
Clam Bay farm site, ranging from directly under thE! net-pens
to a distance of 165 m from the perimeter of the complex
(Table 2). The bottom topography at the farm site was steeply
sloped, with water depths increasing to the north and east.
Water depths near the southwest corner of the cage complex
were 10 m at MLLW in comparison to 26 m at the northeast
corner. Sampling in the deeper areas was not possible by
SCUBA diving, thus all samples were collected at depths of 19
m or less. Seawater temperature and salinity of near bottom
waters were 130C and 30 ppt, respectively.
Sediments were primarily medium sands (0.25 - 0.5 mm
diameter) to the east and south of the pen complex with finer
sands (0.125 - 0.25 mm) to the northwest. Silts and clays
comprised less than 5% of the sediment throughout the study
area. Shell and gravel comprised a large fraction of the
sediment (19%) only at stations to the east of the farm site.

Water Currents

Two current meters were deployed near the Clam Bay farm
site. The first was positioned 100 m east of the net-pen
complex, 5.5 m above the seafloor and 23.5 to 27 m below the
surface (depending on tidal stage). The second was located
100 m northwest of the farm, 5i5 m above the seafloor and 9.5
to 13 m below the surface. Data collected during the period
July 6 through August 4, 1988 were analyzed and presented in
Figure 2. The current regimes showed pronounced differences
at the two sites as a result of either depth-related differ-
ences in current flow, or, more probably, local variations
attributable to the close proximity and curvature of the
shoreline.

Table 2
Clam Bay station summary

Station or Distance from Water depth
Transect net-nens (m) (m at MLLW) Substrate

A 0 15

Transect CB1
0 16
15 16
30 16
60 15
165 15 0.3% shell/gravel
95.7% sand
4.0% silt and clay

Transect CB2
0 13
15 14
30 15
60 19 19.4% shell/gravel
77.1% sand
3.5% silt and clay

Transect CB3
0 13
15 12
30 11 1.8% shell/gravel
98.2% sand
0% silt and clay
60 10

STA RT

100 m east
X@ of form N

`100 m northwest
of farm
1W START

k, 11 In
FINISH ly

@FINISH

Figure 2. Progressive vector diagrams based on the two current meter
records from,the Clam Bay site. X symbols indicate 24-hour
intervals.

The current data are presented in the form of progressive
vect@or diagrams. These diagrams were created by drawing a
vector for each of the 690 current observations such that the
oreintation of the vector corresponds to the direction of the
current and the vector's length corresponds to current
velocity. The vectors are then arranged in a head-to-tail
fashion with "start" and "finish" indicating the first and
last records, respectively. East of the farm site (Figure 2)
most of the vectors were oriented to the southeast, indicating
current flows primarily in this direction with little evidence
of tidal oscillation. The constancy factor at this site was
94.7%. (A constancy factor of 100% would indicate currents
consistently in one direction; a constancy factor of 0% would
indicate that currents flowed in all directions with equal
frequency). The mean current velocity over the period of
observation was 9 cm-sec-1, although velocities as high as 36
cm*sec-1 were recorded. These current data were collected to
model the dispersal of solid wastes sinking from the pens, so
the measurements were taken at a depth of about 25 m. Current
velocities at the depth of the net-pens (0-4 m) are likely to
be somewhat different.
Northwest of the farm site there was a strong tidal
influence with currents flowing alternately to the northwest
and to the south. The net current flow was to the west with a
relatively low constancy factor of 65.6%. Current velocities
were slightly lower than those recorded by the other meter.
Mean velocity was 6 cm-sec-l with a maximum of 31 cm-sec-1.

Sediment chemistry

Total organic carbon and total nitrogen in surficial
sediments exhibited similar patterns of enrichment throughout
the study area (Figures 3 and 4). The highest levels of
enrichment were found under the southern perimeter of the farm
site (CB3-0 m) and 15 m east of the site (CB2-15 m). To the
east of the farm site (CB2) the area of enrichment extended to

N '10-
Transect -0
C B 4 0-
r- 8-
0

0
HHHHH MD A 6'

77: 0
Transect 4.
Transect '9 CB2 0
C83
C83
12

0 15 iO @0
Distance from Farm (m)

io.

o 8-

.2 6, CBI C B 2

4-
0

@0 21
Y
T JA
165 @0 iO i5 -Pen; 0 15 30 60
Distancefrom Form(m)
Figure 3. Total organic carbon concentrations in sediments surrounding
the Clam Bay farm. Range bars are not shown if less than the
size of the mean symbol. Open circles show data collected
one year earlier as part of a separate investigation.

to
N

0 Transect
CB4 _0 0.8-

0 0L_ 0.6-

Z
0 -6 0.4-
O.-
Transect Transect 12 CB3
CB3 0 CB2 0.2.

0 15 30 60
Distance from Farm (m)

CTI 1.0,

0.8
0

10 0.6. CB1 CB2
Z
-6 0.4-
I... 0 0
12
0.2

0
i65 @0 3'0 1@ Net-Pens" 0 15 30 60
Distance from Farm (m)

Figure 4. Total nitrogen concentrations in sediments surrounding the
Clam Bay farm. Range bars are not shown if less than the
size of the mean symbol. Open circles show data collected
one year earlier as part of a separate investigation.

a distance of at least 30 m (nitrogen) to 60 m (carbon). To
the south enrichment was limited to within 15 m of the net-
pens. Northwest of the farm nitrogen showed a statistically
significant enrichment to a distance of 30 m (Kolmogorov-
Smirnov two-sample test among all possible sample pairs, a <
0.05), but carbon levels remained uniformly low throughout the
transect. The area northwest of the net-pens had also been
sampled in July 1987 as part of an independent study (Weston,
1988). The results of the current sampling were generally
similar to the results from the previous year except in close
proximity to the net-pens. Within 45 m of the farm. carbon and
nitrogen concentrations were substantially reduced relative to
their levels one year earlier. This reduction is not due to
stocking density since the total biomass of fish in the farm
was actually 20% greater during the present investigation than
one year earlier, but other factors such as small differences
in station location or sampling artifacts (diver vs. box
corer) may explain this apparent decrease.
The reduction-oxidation potential (Eh) is a quantitative
measure of the reducing or oxidizing intensity of sediments.
Positive Eh values are generally characteristic of sediments
which have a large grain size, are well oxygenated, and/or are
poor in organic matter. Negative Eh values are measured in
sediments which are rich in organic matter, consist. largely of
fine sediments, and/or are poorly oxygenated. Sed-iments
receiving high inputs of feed and feces from an aquaculture
facility would be expected to have more negative Ej., values
relative to background conditions assuming grain size is
comparable.
At the Clam Bay farm site trends in Eh values closely
mirrored gradients in total organic carbon and total nitrogen.
Background values were generally about 350 mv at the sediment
water interface and 250-300 mv at a depth of 4 cm in the
sediment column (Figure 5). With increasing proximity to the
pen site, Eh values were reduced throughout the sediment
column. The area of depressed potentials extended from 30 m

400- 0
N 3004 +
0 Transect E 0 +
C 8 4 . I
200, +

0 4-
4' 100-
till till I U111114t] 0
a- CB3
9 0 0 0-
0
Transect
Transect CB2 -100
C83

-200
0 15 iO @O
Distance from Form (m)

400.
0 0
300- + 0 +
E 0 0 0 0
200- +
+ +
+
2 100.
0 0 +
(L
X 0.

0
_i00 CBI CB2
-200
i@5 io so i5 e Net-Pens" U 10 30 60
Distance from Farm (m)

Figure 5. Reduction-oxidation potentials in sediments surrounding the
Clam Bay farm. Open circles represent potentials at the
IM9
LZ24

sediment-water interface; cross symbols represent potentials
at a depth of 4 cm in the sediment.

northwest of the net-pens, to 36 m east of the net--pens, to
15-30 m south. Reducing conditions at the sediment-water
interface were evident only 15 m east of the net-cage complex
and directly under the southern perimeter.

Dissolved oxygen

Dissolved oxygen concentrations at a height of 5 to 10 cm
above the sediment-water interface were uniformly about 8
mg-1-1 throughout the study area (Figure 6). The dissolved
oxygen sampling design was necessarily.less than ideal, since
the various stations were sampled over a 6 hr. period and it
is not possible to differentiate between upcurrent and down-
current sites. Nevertheless the data should.show if the
enriched sediments caused a dramatic depletion in dissolved
oxygen of the overlying water as has been observed elsewhere
(Brown et Al., 1987). No such depletion was evidE!nt, presum-
ably because of the high current velocities of thE! site.

Sediment traRs

Seven sediment trap arrays were deployed along transect
CB1, but because of limitations in bottom time and air supply,
the divers were only able to retrieve one array directly under
the net-pens (Station A) and the array at the farra perimeter
(CBI-O m). Duplicate samples from the three traps within each
array were analyzed, resulting in six estimates of deposition
rates at each station.
Directly under the net-pens the estimated sedimentation
rate was 52.1 kg dry wt.-m-2.yr-1 (range of six samples = 46.6
- 55.2). At the pen perimeter the sedimentation rate was 29.7
kg-m-2.yr-3- (range 27.8 - 30.8). The particulate material
collected directly under the net-pens was approximately twice
as enriched in organic carbon as that collected in the trap at
the farm perimeter (25.9% and 12.0%, respectively). The

0 Transect INO-
C84
8 0

X
6-
HIM
(D

0
Transect U)
Transect CB2 CB3
CB3 2

0 15 30 60
Distance from Form(m)

ko
12.

M 10-
E

0 6-
CBI CB2
75 4-
W

0
L

165 60 3b ;5 Net- Pens 0 Z 30 @O
Distance from Form (m)

Figure 6. Dissolved oxygen concentrations in near-bottom water
surrounding the Clam Bay farm site.

estimated flux of organic carbon to the seafloor was 13.3
kg C-m-2.yr-1 under the net-pens and 3.6 kg C-m-2.yr-1 at the
perimeter.

Macrofauna

Macrofauna were collected one year earlier at four
stations along transect CB1 at points close to but not
identical to those sampled during the present study (i.e.,
distances from the farm perimeter of 0, 45, 90, 150 and 450 m
vs. 0, 15, 30, 60 and 165 m in the present study). Extensive
data analysis is still in progress, so preliminary conclusions
are limited to assessment of abundance, biomass, species
richness and the density of indicator species.
Figure 7a illustrates typical qualitative changes in
species number, biomass and species abundance along a gradient
of organic enrichment (Pearson and Rosenberg, 1978). At low
levels of organic input, a transition zone develops in which
abundance, biomass and species richness gradually decrease
from levels typical of the unpolluted environment. In this
transition zone there may be a slight species richness and
biomass peak attributable to a phenomenon known as "biostim-
ulation". In this area the organic input provides a rich food
source, yet the rate of input is not so great that it inter-
feres with the mechanics of suspension feeding nor causes
serious oxygen depletion. At a somewhat higher rate of input,
total macrofaunal abundance attains a maximum value. Biomass
may also be slightly elevated, but the number of species is
very low. The increased abundance and biomass results from
the proliferation of a few opportunistic species. With still
higher rates of organic input there is a complete absence of
benthic macrofauna. The rate of organic input is so great
that oxygen levels in bottom waters and sediments decrease (or
sulfide levels increase) to such an extent that aerobic
organisms can not survive.

Opportunistic
Azoic Species Transition Reference
Conditions Dominate Zone Conditions

Species
X-1 %@____Biomass
Abundance

120-
5000- 150- .10

EiOO- species
4000- 0
W) CU C;
0 1E &
0.80- /----Biornass
0i0O_
3000- Z
01 JI
-.2*:! to 60-
co 4*
co W 1 J
0 C
E
2000- 6
C Abundance
40-
50-

4000- (0 20-

IF
0 45 @O 150 450
Distance from Farm (m)

Figure 7. Trends in species richness, biomass and macrofaunal
abundance along gradients of organic enrichment.
(A) Generalized trends based on Pearson and Rosenberg
(1978); (B) Data collected along transect CB1 at the
Clam Bay farm site. Inter-replicate variability is
not shown in order to simplify the graphical presentation.

Trends in species number, abundance and biOMELSs are shown
along transect CB1 for comparison with the ideal model (Figure
7b). Total macrofaunal abundance was elevated 4-fold at the
perimeter of the net-pen complex, and decreased to near back-
ground levels within 45 m. This peak in abundanCE! was due
almost entirely to the contributions of nematodes and the
polychaete Canitella cf. capitata. Areal species richness
increased consistently along the length of the trELnsect.
Biomass was much reduced to a distance of at least 45 m from
the pens, for despite the high density of individuals, the
organisms were relatively small. Moderate biomass levels were
found between 90 and 150 m. The highest biomass was observed
at 450 m from the pen site due to the appearance of several
large deep-burrowing organisms including bivalves, sipunculans
and echiurans.
Capitella cf. capitata is widely recognized as an
indicator of organic enrichment and has been found in the
vicinity of net-pens throughout the world (Kitamori, 1977;
Pease, 1977; Ervik et al., 1985; Brown, 1987). The species
was present in densities of over 12,000 indiv.-m-'! adjacent to
the net-pens, and remained in high densities up to 150 m or
more from the farm site (Figure 8).
The macrofaunal data from Clam Bay are generally consis-
tent with the ideal model of changes along an enrichment
gradient. No azoic conditions were observed, although no
samples were collected directly under the pen complex. The
data indicate dramatic community alterations beneath the
facility perimeter, including the disappearance of most
species characteristic of undisturbed Clam Bay habitats.
Moderate disturbance with gradually improving conditions was
evident between 45 and 150 m from the farm. Normal conditions
were reached at some point between 150 and 450 m from the

farm.

C@j
10000-
E

5000-
0

A-_
J6

0 45 90 150 450
Distance from Form (m)

Figure 8. Density of the opportunistic polychaete Capitella
cf. capitata along transect CB1 at the Clam Bay
farm site. Vertical bars represent range of the
three samples at each station.

Dispersion model

The sedimentation model was run based on the data of
Table 3. Standing stock and feeding rate were provided by the
farm operator, and based on monthly mean values of' these
parameters over the previous twelve months. The organic
carbon content of the feed was measured directly as was the
settling velocity of the feed pellets. Settling velocity of
feces was taken from the literature (Warrer-Hansort, 1982).
Literature values of feed wastage range from 1 to 30% (VKI,
1976; Penczak, gt Al., 1982; Braaten, gt al., 198-3; Gowen,
et al., 1985), and lacking a measurement specific to the Clam
Bay facility, a wastage of 15% was assumed arbitrELrily.
The dispersion model predicted that the area directly
under the net-pens would be subject to the greatest rate of
solid waste deposition (Figure 9). Areas to the north and
west of the farm should recieve very little feed and fecal
matter, with the vast majority of the material moving towards
the south and east. The area delimited by the 1 kg C-m-2.yr-1
isopleth extends 70 m from the farm at its most distant point.
There is extraordinary agreement between the rate of
organic carbon flux predicted by the model and thELt measured
in the sediment traps. Directly under the net-pens at the
trap location the predicted rate is 11.1 kg C-m-2.yr-1, in
comparison to the measured rate of 13.3 kg C-m-2.yr-l. At the
northwest corner of the farm site the predicted and measured
rates were 2.5 and 3.6-kg C-m-2-day-1, respectively.
The dispersion model predicts a deposition rate, and thus
a rigorous test of the model would also require rate measure-
ments, such as those obtained from the sediment traps. In the
strict sense, a static measurements such as organic carbon
concentration or redox potential can not be used to test the
model since no information is available on in sitzi post-
depositional processes. Nevertheless, since sediment trap
data were so limited, the assumption was made that- post-

Table 3
Data used in dispersion model at the Clam Bay farm site

Farm size: 280 m by 52 m with long axis oriented east-west

Depth of pens: 4 m

Standing stock: 352 metric tons
Feeding rate: 4409 kg-day-1

Organic carbon content of feed: 48%

Feed wastage: 15%
Settling velocity of feed: 10 cm-sec-1
Settling velocity of feces: 4 cm-sec-l

Water depth (approx. average of MLLW and MHHW along sampling
transects): 18 m

Current data: July 6 - August 4, 1988; meter located 100 m
east of farm site.

0 50 100
mom=
Scale in meters

2 2 2

- - - - - - - - - -

aZ@ Z- @x d.- --8

Figure 9. Model predictions of organic carbon loading to the seafloor
at the Clam Bay farm site. Contour units are kg C*M-2.yr-1.
Dots indicate locations of sampling stations used for model
verification.

depositonal processes (e.g., mineralization, resuspension)
were uniform throughout the study area, and a test was made of
the model's ability to identify areas of greatest organic
enrichment (Figure 10). There was a significant correlation
between predicted deposition rates and measurements of both
sediment organic carbon content and redox potential (Spearman
rank correlation, a >> 0.01). The model, therefore, performed
well in identifying those areas surrounding the Clam Bay farm
which experienced the greatest degree of organic enrichment.

SQUAXIN ISLAND SITE

Sixteen stations were sampled in the vicinity of the
Squaxin Island farm site (Table 4). Water depths in the area
were uniformly 5 m at mean lower low water. The bottom of the
net-pens were 2 m above the seafloor at this tidal stage and
approximately 6.5 m above the seafloor at mean higher high
water. Seawater temperatures were approximately 150C;
salinity was 30.5 ppt.
The substrate was principally silt with varying amounts
of sand (14 - 30%) and shell fragments (1 - 40%). Shell
debris was so dense at some sites that the collection of
undisturbed sediment cores was difficult or impossible. There
was no visible evidence of culture-related disturbance on the
sediment surface (e.g., Beggiatoa mats, feed or fecal
material).

Water Currents

Current meters were moored 60 m to the south and to the
north of the eastern end of the net-pen complex. Both were
positioned 2.5 m above the seafloor and 2.5 to 7 m below the
surface (depending on tidal stage), and left in place from
June 6 to August 5 1988.
Both meters showed strong north-south tidal oscillations,

8- 0 Under pens
0 CBI
A C62

C63

C
0

E 4-

0
z

000 0 a
28 -1 @O
Predicted deposition (kg C-rTf yr

300
0

euu"
0
0

4- 0
0 100.

C
4)
0
Q.

9
0
,a
(D
cr -100.

-200.
2 4 6 1'0
Pred icted d eposition kg C - rn-2. yr 4)

Figure 10. Comparisons of predicted carbon deposition based
on the sedimentation model with field measurements
of total organic carbon and redox potentials at
the Clam Bay farm site.
28

Table 4
Squaxin Island station summary

Station or Distance from Water depth
Transect net-ipens (m) (m at MLLW) Substrate

A 0 5
Transect SQ1 0 5 1.1% shell/gravel
14.3% sand
84.6% silt and clay

6 5 11.2% shell/gravel
13.6% sand
75.2% silt and clay

15 5 3.1% shell/gravel
16.9% sand
80.0% silt and clay

30 5 2.8% shell/gravel
18.0% sand
79.2% silt and clay

60 5 46.4% shell/gravel
32.2% sand
21.4% silt and clay

100 5

Transect SQ2
0 5
6 5
15 5
30 4 41.2% shell/gravel
29.8% sand
29.0% silt and clay

Transect SQ3
0 5
6 5
15 5
30 5
60 5 0.4% shell/gravel
17.3% sand
82.3% silt and clay

with the southerly flow predominating (Figure 11). A signifi-
cant westerly component was evident only in the meter located
south of the farm site. The constancy factors were 90.9 and
83.3% for the northern and southern locations, respectively.
Current velocities at the northern site averaged 6 cm-sed-1
over the 60 day deployment with a maximum recored velocity of
31 cm-sec-1. Current velocities of 15 to 18 cm-sec-l were
generally observed at least twice daily. Currents at the
southern meter location were slightly slower, averaging
7 cm-sec-l with a maximum recorded velocity of 23 cm-sec-l.

Sediment chemistry

Sediments at the Squaxin Island farm were much finer-
grained than at the Clam Bay facility, and thus had consid-
erably greater levels of organic carbon and total nitrogen.
Sediment carbon concentrations were typically 2 to 3% at the
Squaxin site in comparison to 0.3% in undisturbed areas of
Clam Bay. The Squaxin site also differed from Clara Bay in
that there were no gradients in sediment carbon or nitrogen
concentration that could be attributed to farm activities
(Figures 12 and 13). The degree of sediment enrichment was
more or less uniform throughout the study area, and the slight
,variations were unrelated to farm proximity. There was no
evidence of sediment enrichment above background levels even
directly under the net-cages. One sample collected at the
southern pen perimeter showed an organic carbon concentration
approximately triple typical values, but this high concentra-
tion was not reflected in the two other samples at the same

station.
Sediments at a depth of 4 cm (and probably much shal-
lower) were highly reducing due to the fine grain Size and
reduced porosity of the Squaxin Island sediments (Figure 14).
Reduction-oxidation potentials at the sediment-water interface
ranged from 50 to 425 mv. This extreme variability is prob-
ably a consequence in the sharp gradient of redox potentials

START

60 m north
of farm START

60 m south
of farm

FINISH FINISH
P4

)F I r@

Figure 11. Progressive vector diagrams based on the two current
meter records from the Squaxin Island site. X symbols
indicate 24-hour intervals.

6.
N
A -0-0
Transect
S03 C
0
4
Transect T 1
S02

0 2-
Transect -6
S04 4- SQ2
12

0 Iii io
Distance from Form (m)

0 2- 0
-6

SQ 3 SQ1
6'0 so 1@ 9 'N e 't-P e Ins 6 115 io 60 16o
Distancefrom Farm(m)

Figure 12. Total organic carbon concentrations in sediments surrounding
L

the Squaxin Island farm. Range bars not shown if the range
is less than the size of the mean symbol.

Transect 1.0-
S03 O-e
0.8-

In..t
S02 0.6-
Z
76

Transect
SQ4 0.2- SQ2

15 30
Distance from Form (m)

1.2-

- to -
0-0-

C
& 0.8-
0

Z 0.6-
-6 1
0.4- 0

0.2
SQ3 SQi
60 3b ib 'Ne 't'P 'en s 1@ 3b 6b 16o
Distance from Farm (m)

Figure 13. Total nitrogen concentrations in sediments surrounding the
Squaxin Island farm. Range bars not shown if the range is
less than the size of the mean symbol.

0
400.
N 0
Transact 0 300- 0
S03 0 E
-6200-
0 190 C
Transact a)
S02 0 100-
Q_

0-
Transact 4) SQ2
SQ4 Cr +
-100.

_200-
" 9 ib 30
0
D i sto nce f rom Fa rM (m)

400- 0
0 0
0
> 300- 0
E

_6 200. 0
.0-
C 0 0
20 i0o. 0
CL +
X
0 4-
0. +
+
-100. + +
_200- v SQ3 + + SQi
60 'N et -te 'ns 1@ 3 6b 1 0
Distance from Form W

Figure 14. Reduction-oxidation potentials in sediments surrounding the
Squaxin Island farm. Open circles represent potentials at
the sediment-water interface; cross symbols represent
potentials at a depth of 4 cm in the sediment.

in the upper few millimeters of the sediment column, and thus
the measured potential can vary widely depending upon whether
the probe is held a millimeter or two above or below the ill-
defined "interface". Variability of the redox potential at a
depth of 4 cm is probably due to the difficulty of collecting
an undisturbed core with the high concentration of shell
fragments, compounded by the difficulty of inserting the probe
into the core without encountering an obstruction and further
disturbing the core.
Redox potentials showed no pattern among the sampling
stations that could be attributed to the presence of the farm.
Unlike in Clam Bay, redox potentials were not consistently
lower near the net-cages, except perhaps along transect SQ3.

Dissolved oxygen

Dissolved oxygen concentrations at 5 to 10 cm above the
substrate were typically about 10 mg*1-1, and showed no
depression near the farm site (Figure 15). Two samples had an
unexpectedly high dissolved oxygen concentration indicative of
either analytical error or an abrupt gradient in dissolved
oxygen with distance above the seabed. The same qualifica-
tions expressed at Clam Bay regarding non-synoptic sampling
apply here as well, but it does appear that farm activities
are not depleting oxygen in the near-bottom waters.

Sediment traps

Seven sediment trap arrays were placed at the stations
along transect SQ1, but only four could be recovered after the
15-day deployment (Table 5). Results from the array directly
under the pens are problematic in that these traps captured
only half the material of the other traps, and the material
retained was comparatively low in organic carbon content.
Since the mouths of the traps were only 1.5 m below the bottom
of the net-pens at MLLW it was expected that a large amount of

Transect N 10.
N,
S03 0 C71
E
8-
4)
Transect (M
S02 X 6-
0
'0
> 4'
Transect
S04
9 SQ2

1@ io
Distance from Form(m)

12-

10.

E
8-

6-
0
V
0
a 4-
0
U)
An
2. SQ3 SQ1
6,0 iO 1 *5 "N "et "-P "en"s ;5 iO @O 460
Distancefrom Form(m)

Figure 15. Dissolved oxygen concentrations in near-bottom water
surrounding the Squaxin Island farm site.

Table 5
Sediment trap contents at the Squaxin Island farm
(Six measurements at most stations)

Distance Sedimentation Nitrogen Carbon Carbon
from pens ra@e content content flux
2.yr-1
W (kg-m- *yr-1) M (k M Ll

under pens 62.8 0.75 6.12 3.8
65.2 1.29 7.85 5.1
59.4 0.56 5.02 3.0
58.0 0.63 5.19 3.0
38.9 0.85 7.43 2.9
41.1 -0.99 7.62 3.1
MEANS 54.2 0.84 6.54 3.5

0 115.7 0.49 4.58 5.3
88.0 0.43 3.95 3.5
113.4 0.99 8.72 9.9
119.7 -0.66 6.23 7.5
MEANS 109.2 0.64 5.87 6.6
1

15 109.1 0.42 4.71 5.1
118.4 0.39 4.21 5.0
116.9 0.51 4.40 5.1
111.3 0.64 4.15 4.6
99.4 0.85 6.79 6.7
106.0 0.60 5.20 5.5
MEANS 110.2 0.57 4.91 5.3

30 112.4 0.47 3.95 4.4
98.9 0.35 2.13 2.1
106.9 0.40 3.70 4.0
108.7 0.48 5.22 5.7
109.2 0.34 2.57 2.8
107.7 -0.27 1.64 1.8
MEANS 107.3 0.39 3.20 3.5

organic-rich material would have been collected. It is
Possible that this array may have become entangled in its own
rope or in the nets during a very low tide, and turned on to
its side at some time prior to recovery. This was the only
array retrieved by a rope rather than by diver, and thus it is
not known if it was properly positioned at the time of

recovery.
On the basis of data from the other sediment traps, the
sedimentation rate was about 110 kg-m-2.yr-1 in the area from
the farm perimeter to a distance of 30 m. 'Lacking data from
greater than 30 m from the farm, it is difficult to establish
whether the measured sedimentation rate reflects natural
conditions or an enhanced sedimentation due to the presence of
the farm. The organic carbon and nitrogen content of the
trapped material was significantly different among traps
(Kruskal-Wallis one way analysis of variance, a < 0.05), and,
in fact, the traps nearest the pens contained solids with a
higher concentrations of both organic carbon and total
nitrogen.
In comparison to the Clam Bay site, sedimentation rates
at the Squaxin Island farm were two to three times greater (30
to 52 kg*m-2.yr-i vs. 107-to 110 kg*m-2,yr-l). The collected
material at Squaxin Island, however, contained only half the
carbon and nitrogen concentration of the Clam Bay material.
The fact that Clam Bay traps contained material with a
relatively high organic content (30 to 100 times greater than
backgr ound levels in surficial sediments) suggests that the
collected material was largely farm-derived. At the Squaxin
site the measured sedimentation rate was relatively high, yet
since the organic content of the trapped material was only
slightly greater than background concentrations in surficial
sediments (1.5 to 2-fold), it is likely that the bulk of the
collected material originated from natural sources.

Macrofauna

Macrofauna samples have been collected at most stations
along transect SQ1 as part of routine monitoring by the farm
operator. Baseline samples were collected in January 1987
after installation of the pens but prior to stocking with
fish. Additional samples were collected in August 1987,
January 1988, and August 1988. Intercomparisons among
sampling events is complicated by variations in sample design
between sampling periods and between replicates. In January
1987 each station was sampled with 3 cores of 5 cm diameter,
each of which was sieved through a 0.5-mm screen. In all
other sampling periods 10 cm diameter cores were used, two of
which were sieved on a 1.0-mm screen and one of which was
sieved on stacked 0.5 and 1.0-mm screens. Therefore, when
comparing among sampling periods it should be recognized that
the January 1987 sampling would tend to under-estimate species
richness and abundance because of the smaller area sampled,
but over-estimate these parameters because of the finer mesh
size. All data are expressed on a "per three sample basis"
which, in all but the January 1987 sampling, should be inter-
preted to include the material retained on a 1.0-mm screen in
two samples per station and on a 0.5-mm screen in one sample.
The baseline samples (1/87) and those collected after 6
months of operation (8/87) demonstrated that the five moni-
toring stations were comparable in species richness and faunal
composition and thus suitable for inter-station comparisons
(Figure 16). In January 1988 after 12 months of operation
there was a dramatic decrease in the species richness and
abundance at the station under the perimeter of the pens and,
to a lesser extent, at the station 6 m from the pens. This
same decrease near the pens relative to the other stations
along the transect was evident again after 18 months of
operation (8/88). At this time macrofauna present under the
pen perimeter included only the polychaetes Capitella cf.
capitata (4 individuals), Neyhtys cornuta fransiscana

20-

1U)

CL
E 8/88
u' 15-
........ ....... 8/87

,n 10-
(P
C
I/ae
-04/87

5-
CL

0 6 15 50 60
Distance from Form (m)
200-

,P 1 /87
J#

u) 150 -
CL 08/87
E

.E 100

%
%
A. 00
'--.w, 1/88
C %
........... 01 8/88
.0

50-

0 6 45 30 60
Distance from Form (m)
.0

Figure 16. Trends in species richness and macrofaunal abundance
along transect SQ1 at the Squaxin Island farm site.
Inter-replicate variability is not shown in order to
simplify the graphical presentation.

(3 indiv.) and Glycinde Pict (1 indiv.). The depauperate
fauna 0-6 m from the pens, an area which was formerly
comparable to the other sites, indicates a localized impact of
farm operation.
The distribution of Capitella cf. calpitata (Figure 17)
also suggests enrichment 0-6 m from the pen perimeter. No
C. capitata were found throughout the sampling transect during
the baseline sampling (1/87). In all subsequent sampling
periods, however, the species has been found in high densities
at the 6 m site and at much lower densities at the pen
perimeter and 15 m. Maximum density (945 individuals per
three replicates, or 1875 indiv-m-2) was obtained in the most
recent sampling (8/88).

Disipersion model

The dispersion model was run using the data of Table 6.
The standing stock value used was based on the biomass present
at the time of sampling. This value may seasonally vary by up
to 50% depending on stocking and marketing cycles. No data
were available on the actual amount of feed provided, so it
was assumed that the fish were fed at a rate of 2% of their
body weight per day, a rate typical of salmonid net-pen
culture and the target rate of the Squ*axin pen operators. All
other model parameters were determined as described for the
Clam Bay farm.
As a result of the shallow water depths, the sedimenta-
tion model predicted that the vast majority of feed and fecal
waste would reach the bottom directly under the pens or within
5 m of the farm perimeter (Figure 18). A maximum loading rate
of 14 kg C-m-2.yr-1 was predicted for most of the area under
the farm. Lesser rates of deposition were predicted to the
north and south, with essentially no accumulation of wastes to
the east and west. Loading rates of 1 kg C-m-2.yr-l or
greater were limited to the area 15 m north and 28 m south of
the farm site.

50-

CL
E 40

30- 8/88

1 /88

CL
20-

0 8 /8 7

10-

T I
0 6 0 50 60
Distance from Form (m

Figure 17. Density of the opportunistic polychaete Capitella
cf. canitata along transect SQ1 at the Squaxin Island
farm site. Inter-replicate variability is not shown
in order to simplify the graphical presentation.
8

Table 6
Data used in dispersion model at the
Squaxin Island farm site

Farm size: 74 m x 16 m with long axis oriented east-west

Depth of pens: 3 m

Standing stock: 20 metric tons
Feeding rate: 400 kg-day-1

Organic carbon content of feed: 50%

Feed wastage: 15%
Settling velocity of feed: 14 cm-sec-l
Settling velocity of feces: 4 cm-sec-l

Water depth (approx. average of MLLW and MHHW along sampling
transects): 7 m

Current data: June 6 - August 5, 1988; meter located 60 m
north of farm site.

0 10 20 30

Scale in meters

4 0 4

6 -6

'17

xz 11Z 1-Z 1-1 1-Z 1-.1 X@@ X-11 N, zzz 1-Z

4 4

Figure 18. Model predictions of organic carbon loading to the sea-
floor at the Squaxin Island farm site. Contour units are
kg C'm Z-yr-1. Dots indicate locations of sampling
stations used for model verification. Stations 60 and
100 m from farm are not shown.

The predicted carbon fluxes at 0, 15 and 30 m south of
the farm were 5.7, 2.5 and <1 kg C*m-2.yr-1, respectively.
Over the same distance actual rates of carbon flux as measured
by the sediment traps were 6.6, 5.3 and 3.5 kg C-m-2.yr-l. it
should be recognized, however, that the model predicts
deposition of solid wastes originating from farm activities,
whereas the sediment traps do not differentiate between
farm-derived particulates and deposition unrelated to the
culture operation. It was noted above that the bulk of the
material collected in the traps at the Squaxin Island site
probably originated from natural sources such as resuspension
of bottom sediments or terrigenous run-off. A true test of
the model would require that the contribution of the natural
sources be subtracted from the total material retained in the
traps. The strength of the natural "noise", however, is so
great compared to the pen "signal", that many more traps would
have to be deployed to discriminate between these two sediment
sources. The traps located farthest from the farm (60 and 100
m), which presumably would best represent the natural carbon
flux, were not recoverablet so the magnitude of carbon flux
typical for the area may be substantially less than the 3.5
kg C-m-2.yr-1 measured at at the 30 m trap. Model predictions
vary from actual measurements by a factor of about 2 or less,
depending on the particular trap and the assumed magnitude of
the natural carbon flux.
The model was further evaluated by comparing predictions
of deposition with sediment organic carbon concentration and
redox potential (Figure 19). It should again be recognized
that this comparison requires that the untested assumption be
made that post-depositional processes are similar at all
sampling sites. There was no significant relationship between
model predictions and sediment carbon content, although
predicted carbon flux was correlated with sediment redox
potential (Spearman rank correlation, a > 0.05). This correl-
ation was largely attributable to the influence of the 0 and 6
m stations along transect SQ-3, an area which the model

3.4 0 Under pens
OS01
3.2 0 A S02
oSQ3
3.0
0

2.8-

2.6-

0 2.4

2.2 0
0

2.0
ib i'p- W
-2 -4
Predicted deposition( KgC-m -yr

i0o

C
CD 0
t -i0o -

200
2 4 ib f2 14
Predicted deposition K g C - rn-2. yr-1

Figure 19. Comparisons of predicted carbon deposition based
on the sedimentation model with field measurements
of total organic carbon and redox potentials at
the Squaxin Island farm site.

predicted to be among the most enriched and which also was
found to have the lowest redox potentials.

DISCUSSION

CLAM BAY SITE

The various chemical parameters used to measure the
effects of the farm operation on the sediments of Clam Bay
showed good agreement, and provided a clear picture of the
areal extent of impact. The culture of fish at the Clam Bay
site has resulted in a measurable enrichment of the sediments
directly under the pens and to a distance of approximately
30 m from the perimeter of the pens. The exact distance of
impact depended upon the direction and the chemical parameter
used as indicator of impact, but in all cases varied from 15
to 60 m. Within this area deposition of feed and feces has
resulted in increased concentrations of total organic carbon
and total nitrogen. Degradation of these organic wastes has
depleted pore water oxygen, resulting in more negative
reduction-oxidation potentials. At a few sites reducing
conditions were found throughout the entire sediment column,
although at most stations oxidizing conditions persisted to a
depth of at least 4 cm. The enriched sediments did not
measurably decrease oxygen concentrations in the overlying
water at a height of 5 to 10 cm above the seafloor, although
this does not exclude the possibility that reduced dissolved
oxygen levels may be observed if measurements had been made on
a scale of millimeters rather than centimeters.
The biological data (available only to the northwest of
the farm) showed a severely disturbed community at the pen
perimeter. This assemblage was comprised almost entirely of
nematodes and Capitella cf. capitata. If the assumption could
be made that a similar fauna existed wherever organic carbon
concentrations were comparable (and given similar substrate
type), then such a community might also be found 15 to 30 m
south of the pens and at least 60 m to the east. Beyond this
zone of severe disturbance, moderate effects of the farm were

evident in communities 45 to 150 m from the site. This area
was characterized by reduced species richness, biomass, and
capitata densities in excess of 5000 indiv.-m-2.
All previous studies of the benthic effects of net-pen
farming have reported localized impacts comparable to those
found at Clam Bay. Brown et al. (1987) found normal condi-
tions appearing at some point between 15 and 120 m from a farm
in Scotland. In a survey of numerous Scottish farms, effects
such as depressed reduction-oxidation potentials and appear-
ance of Beggiatoa were commonly found up to 30 m from the farm
site (Earll et al., 1984). Doyle et al. (1984) found effects
extending 25 to 45 from a site in Ireland. The extent of
effects at the Clam Bay site is comparable to these earlier
studies based solely on sediment chemistry as a measure of
effect. The biological indicators of disturbance, however,
suggest effects extended to at least 150 m from the farm,
approximately five times the distance typically reported.

SQUAXIN SITE

The effects of the Squaxin Island site on the benthos
were more subtle and evident almost entirely on the basis of
the biological data alone. There was little or no evidence of
farm effects in the sediment organic carbon, nitrogen, redox,
and dissolved oxygen data. One of the three samples collected
below the southern pen perimeter showed a three-fold enrich-
ment in organic carbon, but since such enrichment was not
evident in the other samples at this site, the effects were
presumably very patchy. The best physical/chemical evidence
of farm impacts was provided by the elevated concentrations of
organic carbon and nitrogen in material collected by the
sediment traps nearest the pens. The macrofaunal data indi-
cated reduced species richness and/or abundance from the pen
perimeter to a distance of 6 m, and a peak in C. canitata
abundance at a distance of 6 m. C. capitata, an indicator of
enriched sediments, first appeared in the area 6 months after

culture began, and was increasingly abundant 12 and 18 months
after the initiation of culture. For comparison, Brown et al.
(1987) reported changes in sediment chemistry and appearance
of C. capitata 3 months after initiation of culture in a
Scottish loch. Mattson and Linden (1983) monitored benthic
conditons after installation of mussel longlines and found a
period-of 6 to 15 months were required for replacement of the
original fauna with an assemblage characteristic of enriched
conditions.
The limited measurable physical/chemical effects of
culture and the highly localized biological effects at the
Squaxin site are suprising given that the bottom of the pens
are only 2 m above the seafloor at low tide. There are two
possible explanations for this observation. First, the
current velocities at the Squaxin site are suprisingly high,
and only slightly less than near-bottom currents at Clam Bay.
The 15 to 18 cm-sec-l measured twice daily at the Squaxin site
may promote dispersal of the solid wastes, particularly if the
narrow distance between the nets and the seafloor promotes a
channeling effect and an acceleration of currents directly
under the pens. Secondly, the Squaxin pens have only been in
place for 18 months in comparison to approximately 13 years
for the Clam Bay farm. Benthic conditions at the Squaxin site
may continue to deteriorate with time, but a recent change in
farm operation may slow or halt this deterioration. The
operators of the Squaxin site have recently decided to curtail
aquaculture operations and use the pens primarily for delayed
release. The seasonal nature of use should minimize further

effects on the benthos.

ASSESSMENT TECHNIQUES

It should be noted that at both Clam Bay and the Squaxin
sites, the macroinfaunal data showed evidence of alteration in
areas where sediment chemistry data failed to show farm
effects. Biological data appeared to be a more sensitive

indicator of disturbance, and therefore suggest that chemical
information alone can not adequately define the extent of
benthic impacts from net-pen culture. The biota are certainly
better integrators of temporal variation, and they also
undoubtedly are responding to chemical and physical parameters
unmeasured in conventional surveys.
The reduction-oxidation measurements proved valuable,
particularly at the Clam Bay site where redox potentials
closely mirrored gradients in organic carbon and nitrogen.
Redox measurements have several advantages over carbon and
nitrogen analysis, most notably the fact that results are
obtained in the field immediately after sampling and there are
no analytical costs once the pH meter and redox electrode are
acquired. Redox measurements worked well in the sandy
sediments of Clam Bay where vertical gradients in sediment
redox potentials were gradual, but were more problematic in
the muddy sediments of the Squaxin sediments. In these fine-
grained sediments much of the vertical change in redox poten-
tials occurs in a thin veneer of sediments at the sediment-
water interface. Measured potentials (and the identification
of culture effects) become highly susceptible to minute
variations in the extent to which the probe is inserted.
Conventional redox probes are approximately 1 cm in diameter
and do not permit the fine-scale resolution necessary in muddy
sediments. Micro-electrodes are available, but are expensive
and subject to frequent breakage.

MODEL EVALUATION

The dispersion model was tested on its ability to
predict: 1) the absolute loading of particulate wastes; and
2) the relative loading among numerous sites. The former
evaluation was done by comparing the quantity of material
retained in the sediment traps with the quantity that the
model predicted would accumulate at the trap location. At the
Squaxin site natural deposition made discrimination of the pen

contribution difficult, but the predicted loading appeared to
differ from measured values by factors of 2 or less. In Clam
Bay the predictions were in good agreement with the measured
loadings, differing by factors of 1.2 and 1.4 in the two trap

arrays.
It should be recognized that there are inherent diffi-
culties of measuring actual carbon flux and there are many
potential artifacts associated with the use of sediment traps
(Butman, 1986). Nevertheless, if the sediment trap data can
be taken as an accurate representation of actual carbon flux,
then model errors of only 1 to 2-fold demonstrate remarkably
good predictive capability of the model.
The second test of the model was to compare predicted
loadings at all stations with measured values of sediment
carbon content and redox potential. At both farm sites the
model predictions of carbon flux showed a statistically
significant correlation with redox potential. Sediment carbon
content was correlated with model predictions only at the Clam
Bay site. The better model performance at Clam Bay than at
Squaxin Island was probably due to the greater amount of
sediment resuspension at the latter site. The sediment trap
data and the lack of pronounced physical/chemical gradients
with distance from the farm both suggest a high degree of
sediment transport and resuspension. Under such conditions it
is not suprising that the sites predicted to have the greatest
carbon input rate failed to show the highest levels of
sediment enrichment. This illustrates the shortcomings of
using static measurements (sediment carbon concentration) as a
test of rate measurements (model predictions of carbon flux).
The model predictions were correlated with sediment redox
potential at both of the sites examined, and with carbon
concentration at one of the two sites. The performance of the
model in Puget Sound is in general agreement with tests of the
model in Scotland where, out of six farms, the model
predictions correlated with measured redox potentials at all
six farms and with carbon at two farms (Gowen et al., 1988).

Ideally, the model should be capable not only of
predicting the extent of chemical change in sediments but also
the degree of biological disturbance. Such predictions are
much more difficult for the biological effects of a given rate
of carbon flux is likely to be habitat specific. Communities
of sandy substrates may not respond to a given flux in the
same manner as mud bottom communities. The model has not been
refined to the point where reliable biological predictions are
possible, but some preliminary observations have been made.
In model tests in Scotland, it was found that severely
disturbed sites, defined as containing four or fewer macro-
faunal species, had predicted loadings of at least 1.8 to 4 kg
C'm_2.yr-l (Gowen et al., 1988). At the Squaxin Island farm
four or less species were found where the predicted loading
was 5.7 kg C-m-2.yr-1 and undisturbed communities were present
where predicted loadings were 2.5 kg C-m-2.yr-1 or less. At
Clam Bay the fewest number of species (9) were found where the
predicted deposition was 2.5 kg C-m-2.yr-1, although more
moderate impacts were evident at lower rates of carbon flux.
other than order-of-magnitude approximations, the existing
data base is inadequate to determine a threshold carbon flux
beyond which biological effects are likely. Nevertheless,
such estimates may be possible with additional refinement of
the model and recognition of the habitat-dependence of
biological impacts.
The model has performed well in both Puget Sound and
Scotland, yet it has inherent limitations which should be
recognized:

1) As discussed earlier,. the model does not take into
consideration any post-depositional processes that may
occur, or differences in the rate of these processes
among sites. For example, the model only predicts the
point at which a settling particle impacts the bottom and
not any reuspension or,transport that may occur later.
Model predictions would be invalidated if the degree of

resuspension varied throughout the study area as might be
the case with differing substrate types or pen-related
alteration of current flow.

2) The model, as currently formulated, is incapable of
coping with variations in bottom topography. At the Clam
Bay site a single water depth (18 m) had to be assumed
despite the fact that water depths varied from 10 to 30 m
over the predicted area of depositon. The net effect is
that in shallow areas the model over-estimates the
lateral extent of depositon, and under-estimates the
lateral extent in deep areas. This shortcoming could be
remedied but would dramatically increase the computa-
tional requirements of the model.

3) An arbitrary assumption had to be made that feed wastage
at both farm sites was 15%. Puget Sound farm operators
typically claim a wastage factor of about 5%. The
magnitude of predicted organic carbon loading to the
bottom is dependent upon the degree of wastage assumed as
demonstrated by Figure 20. (Note also in this figure the
.depth-dependence of the loading with reference to the
limitation under point #2 above.) A reduction in the
wastage factor from the 15% employed in this analysis to
the 5% claimed by Puget Sound operators would result in a
decrease in predicted organic carbon loading of about
15%. In fact, there is probably no reliable estimate of
wastage in Puget Sound or throughout the industry in
general. Until there are reliable estimates of feed
wastage, neither this nor any other model will be able to
predict loading with a high degree of accuracy. The
wastage-dependence of loading shown in Figure 20 also
illustrates the effect that the operator's feeding
practices can have on waste production, and the environ-
mental benefits to be gained by reducing wastage.

-13-0

- 12-0
14 DINO
13. LOA =10-0

8.0

-6.0
7

4.
4.0
3.
%
DEPTII [M) 0 WA STA
FOO

Figure 20. The relationship between food wastage, depth and
organic loading to the sediment for a hypothetical
set of current data. Loading rates are given as
g C-m-2-day-l. (From Gowen et al., 1988).

4) The model requires the designation of a single settling
velocity for feed and a second settling velocity for
feces. As noted by Thomson (1986), however, the size and
density of particulates released from a net-pen are
likely to depend upon the species and size of the fish,
the type and pellet size of the feed, and the amount of
physical disturbance induced by either water current or
fish activity. In additon to these potential variables,
particulate wastes are not uniform in size or density,
and thus can not be adequately characterized by a single
settling velocity, or even the two velocities of the
model. Settling velocities of culture wastes are best
described by a frequency distribution, and the model
therefore requires that a major oversimplifying assump-
tion be made.

5) It is not possible to describe variations in the flow
field attributable to the presence of the pen structures,
and how these variations may influence depositon. If the
model is used for siting in a pre-development stage, the
installation of the pen structures may modify the
predicted magnitude and distribution of waste loadings.
The effect of pen installation on waste dispersion is,
however, likely to be small in most cases particularly
when the pens occupy a small proportion of the total

water column.

6) The hydrographic input to the model is based on current
records at a single point, and there is no allowance for
changes of the flow field along a particle's trajectory.
The current velocity and direction measured at the meter
mooring may not be representative of current regimes on
other sides of the farm complex or over the entire area
of deposition. At the Clam Bay farm the two current
meters gave very different pictures of current patterns.
The model was run using data from the eastern meter since

this site would be less subject to the complications of
bay geomorphology, but the model results would have been
somewhat different had data from the other meter been

used. The effects of lateral variation of currents on
model predictions are especially pronounced at Clam Bay
because of the atypically large size of the farm and
shoreline effects on current patterns. At the Squaxin
Island site and at other farms of small to moderate size
the effect of lateral variations in current patterns may
be negligable and ignoring these variations in the model
may be justified.

7) The model, as presently formulated, is incapable of
dealing with depth-related variation in current flow.
With increasing water depth, the probability increases
that currents measured at a single depth are unrepresen-
tative of the multiple current regimes a particle
encounters during settling to the seafloor. In the
present study the meters were positioned at a depth
approximating half the distance between the bottom of the
pens and the substrate. In some situations multiple
current meters at several depths may be necessary to
accurately predict waste dispersal.

8) The model has no mechanism by which to consider duration
of culture. The magnitudes of organic carbon loading
were comparable at the Squaxin and Clam Bay sites,.yet
the effects on sediment chemistry were much greater at
Clam Bay. This difference, which may be due to the short
period of time that the Squaxin site has been in use, can
not be incorporated into the model.

Despite the limitations of the model, it represents the
best means currently available to predict the magnitude and
extent of culture impact on the seafloor. The model has
performed well both in the present study and in previous tests

in Scotland, yet because of its limitations and the necessary
simplifying assumptions, the model should be used cautiously
as a predictive tool. As indicated by the probable high
degree of transport and/or resuspension at the Squaxin Island
site, consideration must be given to site-specific conditions
which violate model assumptions and therefore make predicted
loadings subject to error. on a fine scale (predicting carbon
flux at a specific point), model loading predictions appear to
be reliable to within a factor of two in most cases. While
this is generally adequate for site assessment, the potential
error should be recognized and compensated for in identifying
areas of potential impact.
On a broader scale (predicting the affected area), model
predictions appear very reliable, for at five sites examined
in Scotland, the model predicted that the 1 kg C-m-2.yr-1
isopleth would extend up to 15 to 30 m from the farm
perimeter. In the present study, the model predicted this
rate of loading up to 28 and 70 m from the Squaxin and Clam
Bay sites, respectively. Areal extents of impact of this
magnitude are consistent with observations from the present
study and other studies reported in the literature.
The current information collected as input to the model
is extremely useful both from the perspectives of environ-
mental protection and farm husbandry. The current meters
deployed in this study measured velocity and direction every
15 minutes for a period of 60 days. Such information is
invaluable not only to predict the dispersal of solid,wastes
but to determine mooring requirements, rates of water renewal
in the pens, and duration of slack water (which may be the
limiting water quality parameter in maintaining suitable
growing conditions).
The model is useful in condensing a massive data set of
current measurements into a single summary figure interpret-
able by non-specialists. It should therefore be helpful in

explaining probable impacts in public hearings and similar

forums.
The model is also helpful in identifying sites which
would be clearly unsuitable for culture. Given the current
state of knowledge it is not possible to define the impacts of
a given loading rate, however severe biological disturbance
has been observed in Scotland where predicted loadings were as
low as 1.8 kg C-m-2.yr-1. At some unquantified level below
this loading, development of a site is unlikely to have
significant effects on the benthos. At most farms examined in
Washington and Scotland, loading rates directly under the pens
range from 6 to 14 kg C-m-2.yr-1. Where loading rates are far
in excess of these values, the generation of hydrogen sulfide
by enriched sediments and the consequent effects on the
cultured fish themselves would be of serious concern.

LITERATURE CITED

Arizono, M., 1979. Disease control in mariculture, with
special reference to yellowtail culture. In G. Yamamoto
(ed.), Proceedings Seventh Japan-Soviet Joint Symposium
on Aquaculture, Tokai University, Japan, pp. 78-88.

Braaten, B., J. Aure, A. Ervik, and E. Boge. 1983. Pollution
problems in Norwegian fish farming. ICES, C.M.
1983/F:26. 11 pp.

Brown, J.R., R.J. Gowen and D.S. McLusky. 1987. The effect
of salmon farming on the benthos of a Sottish sea loch.
J. Exp. Mar. Biol. Ecol. 109:39-51.

Butman, C.A. 1986. Sediment trap biases in turbulent flows:
results from a laboratory flume study. J. Mar. Res.
44:645-693.

Doyle, J., M. Parker, T. Dunne, D. Baird and J. McArdle.
1984. The impact of blooms on mariculture in Ireland.
International Council for the Exploration of the Sea,
Special Meeting on the Causes, Dynamics and Effects of
Exceptional Marine Blooms and Related Events.
Copenhagen, 4-5 October 1984.

Earll, R.C., G. James, C. Lumb and R. Pagett. 1984. A report
on the effects of fish farming on the marine environment
of the Western Isles. Report to the Nature Conservancy
Council.

Ervik, A. P. Johannessen and J. Aure. 1985. Environmental
effects of marine Norwegian fish farms. ICES, C.M.
1985/F:37. 13 pp.

Gowen, R.J. and N.B. Bradbury. 1987. The ecological impact
of salmonid farming in coastal waters: a review.
Oceanogr. Mar. Biol. Ann. Rev. 25:563-575.

Gowen, R.J., N.B. Bradbury and J.R. Brown. 1985. The
ecological impact of salmon farming in Scottish coastal
waters: a preliminary appraisal. ICES, C.M., 1985/F:35.
6 pp.

Gowen, R.J., N.B. Bradbury and J.R. Brown. in press. The use
of simple models in assessing two of the interactions
between fish farming and the marine environment. Proc.
European Aquaculture Symposium.

Gowen, R., J. Brown, N. Bradbury and D.S. McLusky. 1988.
Investigations into benthic enrichment, hypernutri-
fication and eutrophication associated with mariculture
in Scottish coastal waters (1984-1988). Report to the
Highlands and Islands Development Board and others.

Hall, P. and 0. Holby. 1986. Environmental impact of a
marine fish cage culture. ICES, C.M. 1986/F:46. 19 pp.

Hedges, J.I. and J.H. Stern. 1984. Carbon and nitrogen
determinations of carbonate-containing solids. Limnol.
Oceanogr. 29:657-663.

Kaspar, H.F., G.H. Hall and A.J. Holland. 1988. Effects of
sea cage salmon farming on sediment nitrification and
dissimilatory nitrate reductions. Aquaculture 61.

Kitamori, R. 1977. (Changes in the species compositon
(principally benthic organisms)]. In Japanese Soc.
Scientific Fisheries (ed.), Senkai Yoshoku to Jika Osen
[Shallow-Sea Aquaculture and Self-Pollution], Suisan-gaku
Shirizu 21 (Fisheries Series 211. pp. 67-76. Koseisha
Koseikaku, Tokyo.

Mattson, J. and 0. Linden. 1983. Benthic macrofauna succes-
sion under mussels, Mytilus edulis L. (Bivalvia),
cultured on hanging long lines. Sarsia 68:97-102.

Pearson, T.H. and R. Rosenberg. 1978. Macrobenthic
succession in relation to organic enrichment and
pollution of the marine environment. Oceanogr. Mar.
Biol. Ann. Rev. 16:229-311.

Pearson, T.H. and S.O. Stanley. 1979. Comparative measure-
ments of the redox potentials of marine sediments as a
rapid menas of assessing the effect of organic pollution.
Mar. Biol. 53:371-379.

Pease, B.C. 1977. The effect of organic enrichment from a
salmon mariculture facility on the water quality and
benthic community of Henderson Inlet, Washington, Ph.D.
thesis, University of Washington. 145 pp.

Penczak, T., W. Galicka, M. Molinski, E. Kusto and M.
Zalewski. 1982. The enrichment of a mesotrophic lake by
carbon, phosphorus and nitrogen from the cage aquaculture
of rainbow trout, Salmo gairdneri. J. Appl. Ecol.
19:371-393.

Rosenthal, H., D.P. Weston, R.J. Gowen and E.A. Black. 1988.
Report of the ad-hoc study group on environmental impact
of mariculture. ICES, Cooperative Research Report 154,
Copenhagen, Denmark. 83 pp.

Strickland, J.H.D. and T.R. Parsons. 1972. A practical
handbook of seawater analysis. Bull. Fish Res. Board
Canada, No. 167, 31 pp.

Thomson, D.E. 1986. Determination of the effects of fish
size and feed pellet size on the settling characteristics
of rainbow trout (Salmo gairdneri) culture cleaning
wastes. M.S. Thesis, University of British Columbia,
Vancouver.

VKI. 1976. Vandkvalitetsinst. & Jydsk Teknologisk Inst.:
Forskellige driftsparameters indflytelse pa forureningen
fra dambrug. [The influence of different operational
parameters on the pollution from fish farms.] Rapport
til Teknologiradet. Horsholm, Denmark.

Warrer-Hansen, 1. 1982. Evaluation of matter discharged from
trout farming in Denmark. In J.S. Alabaster (ed.),
Report of the EIFAC Workshop on Fish-farm Effluents.
Silkeborg, Denmark, 26-28 May 1981. pp. 57-63. EIFAC
Tech. Pap 41.

Weston, D.P. 1986. The environmental effects of floating
mariculture in Puget Sound. School of Oceanography,
University of Washington, Seattle. 148 pp.

Weston, D.P. 1988. Measuring the effects of organic and
toxicant inputs on benthic communites. In Proceedings
First Annual Meeting on Puget Sound Research. Puget
Sound Water Quality Authority, Seattle, Washington.
pp. 552-566.

APPENDIX B

MODELING OF PARTICULATE DEPOSITION UNDER
SALMON FISH FARMS

MODELING OF PARTICULATE DEPOSITION
UNDER SALMON NET-PENS

by:

William P. Fox, P.E.

Parametrix, Inc.
P.O. Box 460
Sumner, WA 98390

for:

Washington Department of Fisheries, AX-11
115 General Administration Bldg.
Olympia, WA 98504

Job No. 35-1747-02

October 31, 1988

INTRODUCTION

one of the observed effects of some salmon net-pen facilities on the environment has
been organic enrichment of the underlying soils from the deposition of fish feces and
uneaten feed. Under separate report to the Department of Fisheries, Drs. Donald
Weston and Richard Gowen have examined deposition rates, chemical changes and
biological impacts under two net-pens sited in Puget Sound. Their evaluation included
a predictive deposition model developed by Gowen.

A separate modeling approach has been used in the evaluation of several proposed net-
pens recently by Parametrix, Inc. The purposes of this report are: (1) to run the
Parametrix model for the same sites modeled by Gowen and comment on the
comparative results, and (2) modify configuration, orientation and density of the pens and
evaluate the sensitivity of predicted deposition rates to these variables.

MODEL DESCRIPTION

The Parametrix model is a modification of a model developed by EPA (1982) to predict
the deposition of particulates from sewage treatment plant outfalls in coastal waters.
This model has already been applied to several proposed net-pens in the State of
Washington. The model relies on average speed and frequency along the principal and
minor axes to predict excursion distances from the pens and areal deposition rates for
settleable materials of distinct settling velocity. A sloping bottom can be accounted for
in the model. Comparison of the Parametrix model with Gowen's model will be saved
for the Discussion.

The Parametrix model includes evaluation of post-depositional processes related to decay
of the organic material. The organic material will decay as it accumulates on the sea
floor. When the rate of deposition matches the rate of biodegradation, a steady-state
accumulation of organic material will result. Decay of organic material will create an
oxygen demand in the bottom waters in the vicinity of the net-pens. The EPA methods
are used to predict the resulting steady D.O. depletion in the near-bottom waters.

RESULTS: CLAM BAY SITE

Currents Like Gowen's model, the Parametrix model is based on the current meter
located 100 m east of the farm site (#F2053). The current rose from this meter is
shown in Figure 1. Each of the current measurements has been catalogued into one of
eight 45* directional "bins." 'ne percentage of currents failing within each bin and the
average speed of those currents are indicated by the rose. The length of each rose petal
is proportional to the percentage of currents in that direction. The current rose for the
Clam Bay site used in the model indicates the predominant current direction falls witbin
the bin from 90' to 135' (ESE). Tbe average currents along the major and minor axes
of the net-pens used in the model are adjusted for the frequency and strength of currents

in each direction (for example, the fastest currents occur from 315' to 360' but are not
very frequent, thus are not weighted as high as the easterly currents).

Wasteload and Settling Velocities. The method of determining wasteload and settling
velocities used by Gowen is different than that used for this model. Ile wasteload
assumed for this model study is consistent with the wasteloads assumed in earlier
modeling efforts by Parametrix for net-pens in Puget Sound (Harding Creek, Discovery
Bay, and North Skagit Bay). The Gowen model is based on a carbon mass balance,
whereas the Parametrix model accounts for total solids. The results modeled herein are
converted to carbon for comparison with Gowen's results in the summary results, based
on 48 percent carbon content for feed, and 80 percent for feces.

The wasteload and settling velocities assumed for the Parametrix model are based on an
annual production of 617 metric tons, Gowen's published data for European farms, and
other European researchers (Gowen 1987). Lost feed would be about 10 percent of the
total feed. The data used in this modeling are tabulated below:

Waste Total Solids Total Carbon Settling
Component Loading Loadin Percen Veloci

Uneaten Feed 430 kg/d 206 kg/d 40% 10 cm/s
Large Feces 325 260 30 5
Small Feces 325 260 LO 2
Total 1,080 kg/d 726 kg/d 100%

By comparison, Gowen's model assumed 317 kg C/day lost feed and 540 kg C/day feces,
for a total loading of 857 kg C/day.

Model Runs. The model was run for the existing pen size and orientation and four
other configurations as summarized below:

Run Description

1 Existing pen size and orientation
2 Rotate existing pens 90'
3 3 round pens with same total surface area
4 Increase pen width by 50 %
7 Decrease fish production 50 %

The results of each model run are summarized below and in the Figures attached at the
end of this report (except Run 7 which would have the same areal coverage as Run 1,
with 50 percent of the deposition rate). The table below includes only the maximum
predicted impact, which would be concentrated directly under the pens. Total
accumulation is based on a steady-state decay of organic material. The model output is
provided at the end of this report for all runs.

Total Organic Total D.O.
Deposition Deposition Accumulation Depletion
Run kg/rr? /yr kg C/rr? /yr kg/n? mg/L

1 14.7 9.3 4.0 0.13
2 9.8 6.1 2.7 0.04
3 10.6 6.5 .2.9 0.08
4 10.8 6.9 2.9 0.11
7 7.4 4.6 2.0 0.07

RESULTS: SQUAXIN SITE

Currents The model runs for the Squaxin site are based on the meter #F2057 located
60 meters north of the existing pens. The current rose for this meter is shown in Figure
2. The rose reveals a predominant NNW by SSE current axis. The SSE component is
slightly less frequent, but much stronger than its counterpart. Current speeds used in the
model are handled similar to the Clam Bay site.

Wasteload and Settling Velocities, The Squaxin site is much smaller than the Clam Bay
site. The assumed wasteloads and settling velocities are tabulated below:

Waste Mass Settling
Component Loading Percent Velocity

Uneaten Feed 40 kg/day 40 01o 10 cm/sec
Large Feces 30 30 5
Small Feces 30 -30 2
Total 100 kg/day 100 'Yo

Model Runs. The model was run for the pen configurations as summarized below:

Run Description

5 Existing pen size and orientation
6 Rotate existing pens 90'

The results of each model run are summarized below and in the figures and printouts
attached at the end of this report.

Total Total D.O.
Deposition Accumulation Depletion
Run kg/m2/yr kg/rr? mg/L

5 11.9 3.3 0.02
6 17.2 2.7 0.05

DISCUSSION

Comparison with Gowen Model. The basic calculation in both models, horizontal
displacement of settleable particles, is governed by the same function of settling velocity
and current velocity. However, Gowen's model simulates individual particle trajectories
for each current velocity from the current meter data, whereas the Parametrix deals only
with averages. Accordingly, the Gowen model requires more input data and is capable
of a producing a larger number of sediment contours and more precise deposition
pattern.

Unlike the Parametrix model, Gowen's model does not consider sloping topography,
multiple settling velocities or post-depositional biological processes. However, Gowen's
model could certainly be modified to include these additional features. Although these
features are @esireable, the uncertainties regarding wasteload and post-depressional
processes exceed the precision achieved by these features.

Neither model accounts for dissolution, suspension or rp-suspension of particles by high
currents. However, a very detailed field investigation of the spatial distribution of
currents near each site would be necessary to predict impacts at remote locations if
material is resuspended. Resuspension and maintenance of fish feces in suspension under
turbulent conditions would be a valuable future research topic.

Comparison of Model Results. Gowen's model produces a more detailed map showing
deposition contours ranging from 0 to 12 kg C/ri?/yr for Clam Bay, with 1 kg C/n?/yr
intervals. The Parametrix model reveals only 3 contours, with an average of 9.3 kg
C/d/yr centered under the pens (Run 1). Gowen's assumed wasteload was greater than
that assumed for the Parametrix model. Given the uncertainties regarding wasteload
assumptions (discussed below) and the greater resolution possible with Gowen's model,
the results appear compatible.

As discussed in the model description, the Parametrix model accounts for the post-
depositional process of decay. The assumptions made are generally conservative, but may
be useful for management uses. The accumulation mass and thickness would be masked
in areas where natural deposition is also occurring. In each case modeled, the predicted
D.O. impact would be negligible, which is consistent with the observations of Weston.
The D.O. function in the model would be useful to flag potential problems, but could
not be relied on for an accurate prediction if oxygen problems are anticipated.

Sensitivi1y Analysis. The results of Runs 1 through 4 and 7 indicate the sensitivity of
deposition to various factors. Runs 1 through 4 represent an equal number of fish, or
wasteload. Run 7 reveals that the most effective way to reduce deposition is to reduce
the wasteload.

For a given wasteload, the greatest mitigation is achieved by orienting the pens
perpendicular to the predominant 'current direction (a reduction of 33 percent in this
case: see Runs 1 and 2). In this respect, the Squaxin site is already mitigated, as
evidenced by a like increase by rotating the pens by 90 degrees (see Runs 5 and 6). Of
course, engineering and navigation considerations may not allow this mitigation measure.

Separation of the pens into several pods may also mitigate deposition. Essentially, the
impact would be spread over a greater area. In this example, a reduction of 28 percent
would be achieved by the configuration chosen for Run 3. The reduction at any
proposed site would be site specific and a function of the size and separation of the
individual pens.

Run 4 increased the dimension perpendicular to the predominant current by 50 percent.
A reduction in deposition of approximately 25 percent is anticipated. Again, the
reduction obtained at any site would be specific to that site. An increase in the
dimension parallel to the predominant current would have little effect.

CONCLUSIONS

The primary conclusions of this analysis follow:

1. Gowen's model produces a deposition map of better resolution than the
Parametrix model. The Parametrix model crudely accounts for uniformly
sloping bottom topography and post-depositional decay, and can accept any
number of settling speeds or vertical variation of current speed.

2. Improvements could and should be made to both models regarding
suspension and resuspension of depositional materials.

3. Current speed and direction relative to the pens are critical to deposition.
Fecal material from the pens may remain suspended at some threshold
current speed. If current speeds are below this threshold level, pen
orientation to the axis of predominant currents is critical.

4. The greater resolution of Gowen's model or any other modeling
improvements (such as multiple settling velocities or complex topography)
are secondary to the need for more defined criteria regarding wasteload,
as discussed below.

The most settleable particles are the uneaten feed. However, lost feed ratios reported
in the literature range from 1 to over 30 percent of the applied feed. Run 7 revealed

the importance of the predicted wasteload in predicting deposition rates. Until more
accurate criteria for wasteload are developed, which seems unlikely, less subtle
improvements to the models are ineffective. The models may provide useful management
tools, particularly when comparing alternative sites and pen configurations, or establishing
best management practices to limit feed wastage.

REFERENCES

EPA. 1982. Revised Section 301(h) Technical Support Document. Prepared for U.S.
Environmental Protection Agency by Tetra Tech, Inc. EPA Pub. No. 430/9-82-
011.

Gowen, R.J. and N.B. Bradbury. 1987. The Ecological Impact of Salmonid Farming in
Coastal Waters: a Review. In: Oceanogr. Mar. Biol. Ann. Rev., 1987:25.

Weston, D.P. and R.J. Gowen. 1988. Assessment and Prediction of the Effects of
Salmon Fish Farm Culture on the Benthic Community. Report to the Washington
Department of Fisheries, Olympia, Washington.

315460-
7.2% 3odL45s
12.9 cm/**c 2.9%
8.2 cmisoc

4dL00-
2700-3150 4.9%
9.2% 7.7 CM/Sec
10.4 cm/soc

...........................................
...............
a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
::::::::: ::::::; : : : : : : : : : : :::: ;:::: : :::::::::::::::::: : 4:::::::: : ::::::::::::::::::::: :::::::!:::: : ::::: :::: :: :: ::: : : ::::;;;:: : : :: :a
::i::

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .:tt@ 4 ; ; ; 1 : : 1 t 1 t j ; ; :
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
225!-2700
...........
.....................................................................................................................
....................................
4.5% ..........

............................
8.5 cm/soc
..........
...............
................... ....................... ............... ............ ............. ....... ...
.......................
..............................
............................
..........
. . . .............. .............
. .. . .. . ...I....................................... .....................
..........

........... ............
.......................
........... ................. ..........................
................. ............ ;:::::::::::::::::::::::::::::::::::::.
................. ............
.........................................
...........
11.0 cm/sec
................. ...................
......................................
......................... ................. ..........
............................................
.............
.......... ................................
....................................................
..............................
...........
I Se-2250 13e- 1806
........................................
.......... ............. .....................
6.4% 18.7%
4.3 cmisec 9.0 cm/sec
.........................................
...................................
.....................
.................................

..................................
..........................
Ht .........................

Note: Current Rose Indicates Figure 1:
Percentage and Moan Current Rose For
Speed of Currents Within
Directional "Dine Length Clam Bay Site,
of Ross, Is Proportional Meter F2053
to Percentage.

316-3600
30-7%
cm/Goc

... . ......

...... .......
..... . ....

. . . . . . . . . .%.........@@@@@@@-..'.'@@...'..f

.......- -454
360

3-1 cmisec

270-3150

3.4 cmJ&*c 5.8 cm/see

..........

... . ........
225-270
3.7%
2.6 cmJw:c 90-1350

..... ........

.... .......

.... ......
190-2260
7.7%
4.3 calaac

. ....... . .

....... .......

... . .......
............

135-1800
27.31%
10.0 cm16*C

Note: Current Roe* Indicates
Percentage and Mean Figure 2:
Gpood of Currents Within Current Rose
Directional *Sine." Loingth For Squaxin Site,
of Roe* Is Proportional
to Percentage. Meter F2057

Not Pens (280 m x 50 m)

2 2
14.7 kqj m /yr) (7.1 kq C/m ./yr)

2 2
4.9 kg /0 /yr) (3.3 kg CIM /yr)
2.5 kglm 2/yr) (1.2 kq C/ M2/ y r)

Scale In Motors Run 1:
F--LJ-----] Clam BaY.
0 50 100 Existing Configuration

4.2 199/021
(2.0 Ito clyl,")

01.3 k9l 2
m
(0.6 Itoc a /Yr)

Not Pons
(50m X 280m)

9.8
4.7 kq C/m lyr)

Scal* In Motors Run 2:
FLJ---@ Clam Bay
0 50 100 Rotate 900

4.4 k91021yr
(2.1 kq C/021yr)

10.6 kilim2lyr
(5.1 kq C/M21yr)

1.4 kg/02/yr
(0.6 kq C/02/yr)

Not Pen
(3 x 77m Dia.)

seal* In Motors Run 3:
F-L-F--@ Clam Bay,
0 so 100 Round Pens

Not Pona
(280 in x 75 m) 2'0 k91021yf
(0.9 ko Cloglyr)

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

10A k91n2iVr
6-3 kgjfg2jyr (5-2 kv C1=21yr)
(2.5 ka CIM21yr)

Scale In Motors Run 4:
Clam Bay,
0 50 100 Double Wide Pens

11.9 ILgd,&2, F
Not Pons (6.7 Ito C/mil/yr)
(75 m x 17 M)

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

IILUJ I

L2 kols2lyr
(2.5 kg Cjm2lyr)

(0.9 ks Cla2lyr)

117@ Scale In Motors Run 5:
FLJ--@ Squaxin,
0 10 20 Existing Configuration

Not Pans
(17m x 76m)

17.2 kq /nFi-yr
(8-3 Its C/ie/yr)

8.3 kq Jlm@@
.0.0 kq C1 lvr)

3.0 ko wi'@Jyr
(1.4 kg C /O/yr)

Run 6:
Scale In Motors Squaxin Rotate 90*

0 10 20

APPENDIX C

PHYTOPLANKTON AND NUTRIENT STUDIES
NEAR SALMON FISH FARMS AT
SQUAXIN ISLAND, WASHINGTON

FINAL REPORT

PHYTOPLANKTON AND NUTRIENT STUDIES

NEAR SALMON NET-PENS

AT SQUAXIN ISLAND, WASHINGTON

John E. Rensel
Rensel Associates
Seattle, Washington

for

The Washington Department of Fisheries
and the
Technical Appendices of the
Programmatic Environmental Impact Statement:

Fish Culture in Floating Net-pens

September 11, 1989

Fisheries Scientist, American Fisheries society Certified

Present Address: College of Ocean and Fisheries Sciences,
University of Washington, WH-10
Seattle, Washington 98195

Table of Contents

List of figures ........................................ iii

List of tables ......................................... iii

Abstract ............................................... iv

INTRODUCTION ........................................... 1

Prior studies: primary production and near net-pens 2

Study site selection .............................. 3

Site description and hydrography .................. 6

Experimental design ............................... 9

METHODS ................................................ 9

RESULTS ................................................ 11

Experiment A ...................................... 12

Experiment B ...................................... 21

DISCUSSION ............................................. 23

Experiment A: effects-on phytoplankton ............ 23

Experiment B: effects on near-field water quality . 26

LITERATURE CITED ....................................... 30

APPENDICES

Appendix A-1. Productivity data from Squaxin Island
on May 25, 1988. 2 pp.

Appendix A-2. Productivity data from Squaxin Island
on June 21, 1988. 2 pp.

Appendix B-1. Summary of recent, near-field nutrient
studies at net-pen sites. 2pp.

List of Figures

Figure 1. Map of the Puget Sound region with Squaxin
Island net-pen site located in South Puget
Sound within oulined box . .......................... 4

Figure 2. Location and vicinity map of Squaxin Island net-
pens and surrounding area . ......................... 5

Figure 3. Dissolved nitrogen concentration before fish release
(top, Fig. 3a) and after most of the fish were
released (Fig. 3b) . ............................... 12

Figure 4. Chlorophyll a concentration before fish release
(top, Fig. 4a) and after most of the fish were
released (bottom, Fig. 4b) . ........................ 17

Figure 5. Primary productivity before fish release (top,
Fig. 5a) and after most of the fish were released
(bottom, Fig. 5b) . ................................. 19

List of Tables

Table 1. Sampling parameters and number of replicates per
depth . ............................... ............... 8

Table 2. Partial results of experiment A-1 from May 25, 1988..13

Table 3. Partial results of experiment A-2 from June 21,
1988 . .............................................. 14

Table 4. Total nitrogen and phosphorus concentration, and T:N
molar ratio . ....................................... 15

Table 5. Summary of cell counts from Peale Passage in cells
per milliliter . .................................... 20

Table 6. Results of experiment B from May 25, 1988 . ......... 22

Abstract

The effects of a salmon net-pen farm on dissolved nutrient
concentration, phytoplankton density and growth rate were
investigated in a shallow passage of southern Puget Sound, near
Squaxin Island, Washington. If background levels of dissolved
nitrogen were sufficiently low for long enough periods, excreted
nitrogen from the fish could enhance the growth of phytoplankton.
The net-pen complex was the largest in western Washington located
in surface waters that are depleted of dissolved nitrogen for at
least some period of the time. Accordingly, the site constituted
a "worst-available case" for net-pens in western Washington.

Two experiments were conducted. The first involved measurement
of phytoplankton density and growth rates at the farm site during
a period of maximum net-pen fish biomass and one month later
during similar tidal and weather condition, but after release of
60% of the fish. Monitoring of reference stations at both ends
of the passage, beyond the immediate area of the net-pens, was
conducted to assess source water conditions and provide a
comparison to the net-pen site.

The results of the first experiment suggest no consistant and
significant effect of the net-pens, however natural variation of
dissolved nitrogen concentrations confounded possible correlation
between phytoplankton density/growth rate and the net-pens or
reference stations. Moreover, only 2 of 12 samples were
collected when major dissolved nutrients could have been limiting
to phytoplankton growth. Therefore, most of the time
phytoplankton cells were not limited by the Ambient nitrogen
concentration and addition of nitrogen from the pens could not
have had a stimulating effect on their growth.

Although the timing and conditions were appropriate to maximize
the effects of the net-pens on phytoplankton, and some patterns
were observed, most of the statistical tests indicated that
phytoplankton growth rate did not significantly vary among
stations or times except during one monitoring period. The first
experiment further serves to illustrates the complexity of
monitoring phytoplankton in the field which involves a number of
potentially rapid fluctuating variables.

The second type of experiment involved near field monitoring of
nitrogen produced from the net-pens. During the period of
maximum fish biomass, minor increases in dissolved nitrogen
(NO3+N02+NH4+) were seen downstream of the pens during one tidal
period, but not during the next. Total ammonia was significantly
elevated within the pens compared to ambient concentrations, but
concentrations were well below the chronic exposure concentration
for salmonids and other sensitive coldwater fish. At a distance
of 30 m downstream approximately 80% of the ammonia had been was
in the form of nitrate, presumably oxidized through microbial
nitrification.

Introduction

Salmon reared in marine net-pens produce solid and dissolved wastes
including various forms of nitrogen. Dissolved nitrogen wastes from
salmon are as much as 60 to 90% ammonia, with lesser amounts of urea
and amino acids (Stickney 1979). Nitrogen is the nutrient most
likely to be limiting to the growth of marine phytoplankton.
Therefore, the potential for localized nutrient enrichment and
increased algal abundance near salmon farms exists, and will depend
mainly on the total size of the farms in the restricted water body
and existing hydrographic conditions (Gowen et al. 1985, Gowen and
Bradbury 1987).

Nutrient limitation of surface waters is a key consideration of the
State of Washington's Recommended Interim Guidelines for the
Management of Salmon Net-pen Culture in Puget Sound (SAIC 1986).
That document conservatively, but somewhat arbitrarily, designated
the nutrient limitation status of sub-areas of Puget Sound based on a
threshold of 0.1 mg/l, two and one-half times greater than one recent
literature value of 0.04 mg/l (0.6 uM, URS 1986a). As discussed
below, simple measurement of dissolved nitrogen concentrations is
inadequate to determine if nitrogen is adequate for algal growth, N:P
ratios also must be considered.

There are numerous difficulties in determining perturbations of
phytoplankton from fish-farms or from other of man's activities.
Foremost, as discussed above, once the net-pens are established,
there is usually no adequate means to establish baseline conditions.
Although nutrient uptake by phytoplankton may be rapid, there is a
lag time of up to a day or so between the addition of nutrients and a
measurable increase in phytoplankton biomass (Parsons et al. 1984).
Knowledge of local hydrodynamic processes of dispersion, such as
distance of tidal excursion, are required for interpretation of these
types of data. Simply monitoring upstream and downstream of a net-
pen farm site may reveal near-field nutrient effects, but may not be
adequate to monitor the relatively slow response of phytoplankton
populations to the increased nutrient concentrations. Monitoring of
phytoplankton abundance and dynamics in the field is also difficult
due to natural variations in time and space of the phytoplankton. A
number of discrete or loosely interacting measures of water column
ecology must be assessed, as no single measure provides all the
necessary information.

Nitrogenous wastes from net-pen reared fish or other sources are
unlikely to increase phytoplankton abundance in most of the main
channels of Puget Sound since nitrogen is already in abundance
(Collias and Lincoln 1977, Anderson et al. 1984, SAIC 1986).
Accordingly, this study focused on a worst-available-case of nutrient
enrichment from net-pens in what appeared to be nutrient-depleted
waters, at least during some tidal, seasonal and weather conditions.
To establish baseline conditions, the present study was conducted
before and after the release of large numbers of fish at a public

benefit, salmon rearing and release net-pen site. In addition to
having spatially separate reference areas, "before fish release"
water quality samples were used as the experimental data, and "after
fish release" monitoring served as a baseline.

There have been relatively few attempts to monitor nutrient
enrichment near salmon net-pens and even fewer studies of the effects
upon phytoplankton dynamics. Prior studies of nutrients from net-
pens strongly suggest there is little measurable effect beyond the
immediate area of the pens (near-field). While nutrient
concentrations are relatively easy to monitor, phytoplankton studies
(far-field second and third order effects) are more difficult to
conduct and have typically relied on measurement of chlorophyll -4
concentrations, the primary photosynthetic pigment in phytoplankton.

Prior Studies of Primary Productivity Near Net-pens

In one early study conducted in the Pacific northwest on this topic,
Pease (1977) surmised that no measurable impact on phytoplankton
populations occurred under worst-case conditions. The study area at
Henderson Inlet, Washington had limited circulation during summer
months, exacerbated by the net-pen site location within a shallow log
dumping and storage area. Intense dinoflagellates blooms occurred at
the site during the summer of 1974 which killed farmed salmon and
prawns (Rensel and Prentice 1980). The 1974 blooms, which appeared
to be exceptional in abundance, occurred throughout portions of
southern Puget Sound and also killed salmon in net-pens at Squaxin
Island. These conditions were not seen in the previous year at
either Henderson Inlet (Snyder et al. 1974) or in several previous or
one later year at Squaxin Island (Fraser 1976).

Pease (1977) found increased density of phytoplankton (chlorophyll A)
during summer months with increasing distance into the log rafting
area. The net-pens were located near the outer, seaward edge of the
log rafting area. Reference areas (controls) were located in the
main channel of Henderson Inlet, outside the log rafting area, and
further inside the log rafting area. From my analysis of his monthly
dissolved nitrate data (Table 5 and Fig. 8 of Pease 1977), it appears
that nitrogen-depleted conditions could have occurred only in July
when some of the samples had a concentration of less than 0.04 mg/1
dissolved nitrogen and adequate phosphate concentrations. A
generalized, inverse correlation between dissolved nutrient
concentrations (nitrate and orthophosphate) and phytoplankton density
was apparent over the entire summer period.

Phytoplankton standing crop during Pease's study was consistently
greater at an reference station inside the log rafting area and at
the net-pen site, compared to an outside, midchannel reference area
in the open water of Henderson Inlet. However, Pease (1977)
concluded that there were no abnormally high concentrations of
phytoplankton anywhere in the area, and that phytoplankton activity
throughout the inlet was unrelated to the net-pen rearing.

- 3 -

The second conclusion is possible, but neither supported or refuted
by his data, as the inside reference area was too close (a few
hundred feet) to the net-pens to be considered as separate and
unaffected. Since water currents are weak and often imperceptible at
the Henderson Inlet site, the inside log-rafting reference areas can
not be considered as separate from the net-pen site, with regard to
phytoplankton populations. These criticisms do not invalidate
Pease's conclusions regarding other water quality conditions, but
suggests that a greater standing stock of phytoplankton existed near
the net-pens and that the effect of location and nutrient impact of
the net-pens could not be sorted out given the experimental design.

Several years after removal of the Henderson Inlet facility Rensel
(unpublished data, 1988) found that chlorophyll p concentrations in
midchannel were about twice those found at the now vacant net-pen
site, opposite the condition that prevailed throughout Pease's year
long study (2.72 versus 1.71 ug Chl. _q, SD = 0.273 and 0.412
respectively, n = 6). However, these samples were taken too late in
the fall to be representative of optimum algal growing conditions and
indicated relatively low phytoplankton density at both areas.

Recent studies in Scotland (Gowen et al. 1988) focused on
phytoplankton density and growth rates in a restricted, fjordic sea-
loch that had slow water movement (maximum flow of 16 cm
sec 1) and a single, large salmon net-pen farm. Additionally, water
exchange into the 50 meter deep Loch Spelve is restricted by a 4
meter deep shallow sill. Study results indicate no measurable effect
of the farm on phytoplankton density, although localized
hypernutrification (elevated ammonia) was seasonally observed
immediately around the net-pen farm. Carbon-14 isotope productivity
data did not show any effect of the farm, althoughthe authors felt
that this portion of their study was based on insufficient data. In
spite of slow water flow near the net-pens, the residency time of
water was too brief to allow measurable increases in phytoplankton
density or growth rates.

Study Site Sglection

The criteria for selection of net-pen location for the present study
involved finding a net-pen site in western Washington that was
located in nutrient depleted waters, while still having relatively
large fish production. Based on the authors experience with these
facilities, the best site was located in Peale Passage, southern
Puget Sound. This site, located just east of Squaxin Island (Fig.
1), is operated as a cooperative Washington Department of Fisheries
and Squaxin Island Indian Tribal sponsored program. Coho
(Oncorhynchug kisutch) and other species of salmon have been reared
at the site since the early 1970's (STOWW 1974) and the 'nearby
beaches have abundant littleneck clam and planted oyster populations.
The Squaxin Island net-pens are presently the largest public benefit
facility in Washington state and produces substantial numbers of fish
for commercial and sport fisheries (Rensel et al. 1988).

200 2
U30W 40' 422* 49*

BELLINGHAM

04
-400
400

NACORT

VICTORIA

2& 20'

srRAIr oF .1uAN DE FvcA

480
8 .EVERETT
N N

OLYMPIC

PENINSULA
40# 40'
@TTLE

SQUAXIN ISLA
NET-PEN SITE
-Nil
201- 20'
''TACOMA

0 110 20 30
a .- . W=mj
47,* @A KILOMETERS 47*

20' 1230W 400 20, 4220
149

Figure 1. Map of the Puget Sound region with Squaxin Island
net-pen site located in South Puget Sound within
oulined box.

TAOOMA

iHELTOh 0
PEALE
Ik SOUAXIN ISLAND
PASSAGE NET PEN
SITE
INLEF

MET
G)
OLYMPIA
Vicinity Map

SOUAXIN
ISLAND HARTSTENE
NET-PEN$ ISLAND
#3

FLOOD

EBB

SOUTH
PEALE
PASSAGE

Figure 2. Location and vicinity map of Squaxin
SCAM IN FM Island net-pens and surrounding area.
r_1_F___1
0 1,wo um

6 -

There are three adjacent sets of net-pens, two for delayed release of
salmon and, in recent years, a third rearing facility to the north
operated by the tribe for normal commercial purposes (pen system 3 in
Fig. 2). All of the fish reared at the first two facilities are
destined for release into Puget Sound, and are held in the net-pens
for only a few months of the year after attaining smolt condition.
This final condition allowed for comparison of water column
conditions before and after fish release, during similar tidal
conditions discussed later.

Site DescriRtion and Hydrography

Peale Passage is a shallow tidal channel connecting Dana Passage on
the south and Pickering Passage to the north (Fig. 2). The Squaxin
Island Indian reservation forms the west boundary of Peale Passage,
Harstene Island the east boundary. The source waters for Peale
Passage are relatively well-mixed by strong tidal currents, although
only one year of sampling data were available for the Dana Passage
sampling station (unpublished WDOE water quality data DNA001). These
data indicate low dissolved nitrogen in surface waters occurred 15%
of the period April to November. Pickering Passage data show low
dissolved nitrogen in surface waters about 39% of the time (SAIC
1986).

Recent studies of circulation and nutrients in deep southern
Puget Sound (URS 1986b) indicate that additional source waters for
Peale Passage are the inlets at the west end of Dana Passage,
especially Budd and Eld Inlets, due to clockwise circulation of
surface and deeper waters in western Dana Passage. These inlets both
exhibit strong vertical stratification and nutrient limitation during
clement weather and undoubtedly influence Peale Passage surface
waters at times during calm weather.

An early study of hydrographic conditions in southern Puget Sound
measured vertical profiles of physicochemical parameters in Peale
Passage on a few occasions (Oclay 1959). Moring (1973) noted that
there was little background information concerning water quality in
the Peale Passage area. His studies provide some basic information
concerning conditions at the net-pen site. In subsequent years, fish
culturists of the Squaxin Island Tribe collected additional
information at the site that, combined with the earlier information,
is adequate to characterize the vicinity. These data indicate that
the area is well-mixed in the late fall to spring months but has a
gradual thermal gradient and very minor salinity gradient in the
clement weather periods of summer. There are no significant
freshwater sources in Peale passage and no sharp discontinuities of
water column characteristics.

Unpublished drogue (drift object) data collected for this study and
recording current meter data collected for a related study (Weston
and Gowen 1989) suggest that water passing through the net-pens does
not exit Peale Passage on a moderate tide. This situation
potentially could lead to an increased abundance of phytoplankton,

- 7 -

since phytoplankton will rapidly assimilate dissolved nitrogen during
periods of nitrogen depletion.

With an average depth of about 5.0 meters at mean lower low water
(MLLW), depths in the vicinity of the net-pen site are shallow
compared to other existing net-pens in Western Washington. Mean
water current velocity near the most northern set of net-pens is
about 6 to 7 cm sec-1 with a net directional flow to the south
(Weston and Gowen 1989). Currents diminish an undetermined amount in
the vicinity of the other two net-pens and may be affected by the
presence of a small cove that tends to slow water movement and form a
gyre, particularly on the flood tide (unpublished survey data of B.
Wood, Squaxin Island Tribe, 1982).

Although no historical nitrogen data were available from Peale
Passage, two days prior to the first sampling date of this study I
found dissolved nitrogen concentrations less than 0.04 mg/l, and a
dissolved nitrogen to phosphorus (N:P) ratio of about 1:1, indicative
of nitrogen limitation. Use of a single numerical value of dissolved
nitrogen may be misleading for representation of the actual threshold
of nutrient depletion (Welch 1980). Examination of both the
dissolved nutrient concentration and the N:P ratio, sequentially, is
more useful in determining if nitrogen depletion exists. Recent
studies in nearby Budd Inlet suggested that nutrient limitation in
surface waters occurred during summer months when the concentration
of dissolved nitrogen was less than 0.04 mg/l and dissolved molecular
N:P ratios were 5 to 1 or less (URS 1986a).

Table 1. Sampling parameters and number of replicates per depth.

14 Carbon Chloro- Dissolved Dissolved Secchi Temper.- Phytop. Total
isotope phyll a nutrients oxygen Disc Salinity counts N & P
------ ------ ------- ------- ----- ------- ------ ------

Experiment A-1: pens versus reference areas, May 25th, before fish release

Low Tide 3 1 1 1 1 1 1
High Tide 3 1 1 1 1 1 1 1

Experiment A-2: pens versus reference areas, June 21st, after fish release

Low Tide 3 3 3 1 1 1 1 1
High Tide 3 3 3 1 1 1 1 1

Experiment B: upstream and downstream of pens, May 25th

mid-flood - 3 1 1 - - 3
early ebb - 3 3 1 1

Codes: 14Carbon isotope primary productivity assessment, phyto. counts
microscope identification and enumeration of phytoplankton, total N
and P = total nitrogen and phosphorus (sum of dissolved and particulate)
Total N and P from center of the net-pens only on ebb tide, experiment B.
pH was sampled concurrent to collection of nutrient samples.

Experimental Desl'_qn

Two types of experiments were conducted. The goal of experiment A
was to investigate rates of primary productivity near the net-pens
and at reference areas, as measured by the uptake of radiocarbon
isotope C-14. Reference area stations were selected to be remote
enough from the net-pens to avoid having waters that had passed
through the net-pens on any single, moderate amplitude tide.
Sampling was conducted during morning and early afternoon hours that
coincided with the early and late portions of the flood tide,
essentially low and high tide. The experiment was conducted before
and after release of most of the delayed-release fish (experiments A-
1 and A-2, respectively), about a month apart, during similar tidal
exchange and timing. Both dates had relatively calm, warm weather
and were during the peak algal growing season in Puget Sound (Winter
et al. 1975, URS 1986a). Other direct measures of phytoplankton
density (chlorophyll a and species cell counts for relative
abundance) as well as indirect, surrogate measures (Secchi disk depth
and dissolved oxygen concentration) were made as time allowed.

During sampling in late May 1988, the three sets of net-pens had a
total of 118,600 kg of fish distributed 38% within system number 1,
45% in system number 2 and 17% in system number 3. Most of the fish
in system one and two were released in early June. During the second
sampling date in late June there was approximately 47,200 kg of fish
in the net-pens, 55% within system number three and 45% in system
two.

Experiment B was an assessment of water quality upstream and
downstream of the net-pens, similar to nutrient monitoring described
in the Interim Guidelines (SAIC 1986). The monitoring was conducted
at low and high tide in late May, before release of the fish. The
goal of experiment B was to monitor the near field effect of the net-
pens on nutrient and ammonia concentrations. Table 1 summarizes the
measurements and replication conducted for experiment A and B.

Worst-possible-case conditions were ensured by timing the experiments
during a period with: the greatest amount of fish in the pens
(experiment A-1) and with relatively small tidal exchange. The mean
tide for nearby Dofflemeyer Point is 10.4 feet, and the diurnal range
is 14.4 feet. Sampling on May 25 and June 21st was conducted on the
beginning and end of 6.3 and 4.9 foot flood tides, respectively.

Methods

Water velocity at the net-pen site was measured with a Scientific
Instruments Price Meter, fitted with an Swoffer optical sensor and
remote, digital readout unit. Surface drift sticks and 2 meter deep
drogues were used to monitor current direction immediately downstream
of the net-pens. Salinity and temperature were measured with a YSI
SCT-33 meter carefully calibrated to standard seawater solution. All
water samples were collected with a 2 liter Scott-Richards sampling

- 10 -

bottle. Dissolved oxygen was measured by a modified Winkler
titration method with an accuracy of 0.02 mg/l. pH was recorded in
the field using a VWR model 55 probe. Chlorophyll a samples were
collected by filtering 50 ml of water through Whatman GF/F filters.
Filters were folded, packaged and iced for analysis later the next
day.

Nutrient samples were collected in acid washed and sample-water
rinsed, polyethelene bottles, iced and frozen later the same day. No
filtering or acidification was conducted to avoid introduction of
broken cells and other artifacts, and to avoid destruction of nitrite
(APHA 1985). This is standard research methodology used for
dissolved nitrogen analysis of seawater samples at the University of
Washington.
Determination of dissolved nitrogen (defined as N03+NO@+NH4+) and
orthophosphate was conducted at the University of Washington Routine
Chemistry Laboratory using a technicon autoanalyzer. Dissolved
nutrient results were reported both as mg/l (ppm), for ease of
comparison to Weston (1986), as well as ug-at./l units (micromoles
also referred to as uM) for comparison to other of other studies
using such ratios. Identification and enumeration of larger
phytoplankton (greater than 5-10 microns) was conducted by an
experienced phycologist using sedimentation chambers to concentrate
samples and an inverted microscope (Unesco 1978).

Relative rates of phytoplankton production were estimated using a
modification of the carbon 14 productivity method (Steemann Nielsen
1952), using water from the same water bottle cast that provided
chlorophyll a, nutrients and other measures mentioned above.- Samples
were collected, during the morning and afternoon, from 2 m at three
sites: directly next to the net-pens and at the north and south
entrances to Peale Passage (Fig. 2). Triplicate 120 ml samples were
put into acid-cleaned 125 ml BOD bottles and transported in the dark
to the net-pens for incubation.
On May 25, 1988, 1.1 ml of the 14C stock (20 uCi/ml NaHC03 solution)
was added to each sample, and then 100 ul was immediately removed and
placed into a liquid scintillation cocktail containing phenethylamine
for total activity determination. The BOD bottles were incubated
next to the pens at 2 m depth within 20 minutes of sample water
collection. Care was taken to insure that the incubating samples
were never shaded by the pens. At the end of the incubation period
(5.67 hrs for the morning samples and 2.67 for the afternoon
samples), duplicate 20 ml aliquots were removed from each BOD bottle
and placed into a glass scintillation vial containing 1.8 ml of a 37%
formalin. Upon return to the laboratory, the formalin-killed
aliquots were filtered onto glass fiber filters (Whatman GF/F), fumed
over 12 N HCL for 15 seconds, and placed in scintillation vials.
Seven ml of liquid scintillation solution was added to the vial, and
shaken overnight. The samples were counted in a Beckman LS1800
liquid scintillation counter for 15 minutes. Carbon uptake was
calculated according to Strickland and Parsons, (1968). Reported
values are not corrected for time-zero controls or dark bottles.

On June 21, 1988, the same initial procedures were followed.
Additionally, single aliquots for time zero controls were removed and
placed into glass liquid scintillation vials containing 1 ml 6 N HCL.
At the end of sample bottle incubation, duplicate 5 ml aliquots were
removed from each BOD bottle and placed into glass scintillation
vials containing I ml of 6 N HCL. Upon returning to the laboratory,
8 ml of liquid scintillation solution was added to the vials, and
they were shaken overnight. The samples were counted in a Beckman
LS1800 liquid scintillation counter for 15 minutes. Reported values
were corrected for time zero controls but not for dark bottles.

Statistical analysis of hydrographic data utilized one way and two-
way analysis of variance and T-test procedures (Zar 1984).

Results

Experiment A

On sampling dates before and after the fish were released all
stations had weak thermal stratification with little difference
between the salinity at the surface and bottom (Tables 2 and 3). The
highest water temperatures usually occurred at the southern entry to
Peale Passage, not at the mid-channel areas near the net-pens. No
trends in dissolved oxygen concentrations were seen on either
sampling date. Secchi disk values (water transparency) were slightly
lower at the net-pen site on both sampling dates.

12 -

Nitro CN03+NG2+NH4)
Before ?11:@ Reloom Noy 25

Nitr"M (29/1)
LB LM Tide
L25 RgWIP TId.

LIS

LOS

a
L PmIs Pm& Obt PO L P"Is Pm&

After Fish Releam June 21

LB- LM Tlds
L25 RgHISP Tid.

LIS

LI-

L05

01 1 b-VVF"
L PmIe Pm& ftt Pa L Pwle
M 91

-177--Z

Figure 3. Dissolved nitrogen concentration before fish release
(top, Fig. 3a) and after most of the fish were released
(Fig. 3b). Standard deviation bars are omitted due to
the very small amount of variance.

- 13 -

Table 2. Partial results of experiment A-1 from May 25, 1988. Sampling
times noted for each tide.

Seccb! Dissolved Percent Dissolved
Depth Temp. Sal. Disk Oxygen oxygen Chl. a Nitrogen
Station (m) (C) PPT (a) mg/l Saturation ug/l mg/l
---- ---- ---- ---- ----- ------- -------- ----- ----------

LOW TIDE: 0940-1030 hours

South 0 13.5 28.9 5.8
Peale 2 13.5 29.0 10.8 124% 6.81 0.09
Passage 5 12.0 29.0
10 11.4 29.1

Net- 0 13.1 2%.0 3.6
Pen 2 12.2 29.0 8.7 98% 23.91 0.17
Site 5 11.4 29.0

North 0 12.1 28.9 4.0
Peale 2 12.1 28.9 10.3 116% 6.52 0.24
Passage 5 11.9 29.1

HIGH TIDE: 1305-1400 hours

South 0 15.1 28.1 4.5
Peale 2 13.5 28.4 13.7 158t 13.58 0.06
Passage 5 12.5 28.7
10 11.1 29.7 14.1 155% 15-25 0.70

Net- 0 13.1 29.0 3.5
Pen 2 12.2 29.0 12.1 134% 15.56 0.15
Site 5 11.4 29.0

North 0 13.6 29.0 3.9
Peale 2 12.9 29.0 10.4 119% 9.76 0.08
Passage 5 12.1 29.0

Table 3. Partial results of experiment A-2 from June 21, 1988. Standard
deviation shown in parenthesis. Sampling times shown for each tide.

Secchi Dissolved Oxygen Dissolved
Depth Temp. Sal. Disk Oxygen Saturation Chl. a Nitrogen
Station (a) (C) (PPT) (in) (mg/1) Percent (ug/1) (mg/1)
---- ---- ---- ---- ----- ------- -------- ----- ----------
LOW TIDE: 0830-0928 hours

South 0 14.9 29.2 3.9
Peale 2 13.8 29.1 9.6 116% 6.99 0.13
Passage 5 13.0 29.0 (0-082) (0.019)
10 12.6 29.0

Net- 0 15.1 29.3 3.8
Pen 2 14.9 29.0 12.0 143% 4.73 0.02
Site 5 14.1 28.8 (0.759) (0.004)

North 0 15.0 29.0 3.4
Peale 2 14.4 28.9 9.6 129% 6.84 0.10
Passage (0.799) (0.005)
5 13.8 28.7

HIGH TIDE: 1209-1310 hours

South 0 15.3 29.0 3.7
Peale 2 14.8 29.0 11.8 141% 3.95 0.01
Passage 5 13.5 29.0 (1.777) (0.001)
10 12.8 29.0 0.49

Net- 0 17.5 29.1 3.0
Pen 2 15.0 28.8 12.1 144% 4.20 0.07
Site 5 14.1 29.0 (0.950) (0.006)
9 13.8 28.9

North 0 16.4 28.5 3.5
Peale 2 15.2 28.5 11.0 131% 5.47 0.03
Passage 5 14.0 28.7 (1.800) (0.006)
10 13.6 28.8

15 -

Table 4. Dissolved nitrogen and phosphorus concentration, molecular
nitrogen to phosphate (N:P) ratios and probable nitrogen
limitation of diatom phytoplankton. Threshold of nitrogen
limitation for diatoms is about 0.04 mg/l dissolved nitrogen or
expressed in molecular value, about 0.65 ug-at/l. If dissolved
N was <0.65 ug-at/1 and N:P ratio < 5, nitrogen limitation was
probable. Phosphorus limitation could occur at N:P ratios >10-15.

Probable
Concentration in ug-at/1 X:P Ratio Nitrogen
Site Dissolved N Dissolvgd P (atomicl Limitation

Experiment A-1: before fish release on Nay 25, 1988.

Low Tide
south Peale no
net-pen site 3.09 1.20 2.6 no
north Peale 4.37 1.15 3.9 no

High Tide
south Peale 1.24 0.87 1.4 no
net-pen site 1.85 1.05 1.7 no
north Peale 1.24 0.87 1.4 no

Experiment A-2: after fish release on June 21, 1988.

Low Tide
south Peale 0.06 0.86 0.1 yes
net-pen site 3.21 1.07 3.0 no
north Peale 0.87 1.02 0.8 no

High Tide
south Peale 3.11 1.10 2.8 no
net-pen site 0.54 0.94 0.6 yes
north Peale 2.08 1.23 1.7 no

16 -

Nutrients: In general dissolved nitrogen concentrations (previously
defined as N03+NO2+NH4 were less after fish release than before at
all three sampling stations (Fig. 3, tables 2 and 3). Nutrient
limitation of diatoms growth was probably not in effect in any of the
samples collected before the fish were released, due to the
relatively high ambient dissolved nitrogen concentration (> 0.04 mg/l
threshold value for diatom growth limitation previously discussed,
table 4).

Large differences were noted between the concentration of dissolved
nitrogen at low and high tides for 4 of the 6 bar clusters in figure
3 and proved significant for the period after fish release (p <0.05).
Significance testing could not be acertained for the period before
fish release due to lack of sample replication, but there was likely
no difference at the net-pen site or at the south Peale station.
This judgement is based on the excellent precision normally seen in
nitrogen measurements conducted at the university's laboratory, as
found in the June samples discussed below.

It is important to note that the between tidal variation was
obviously unrelated to the net-pens, as it occurred at both the net-
pens and the reference stations, at least part of the time. Since
experiment A was predicated upon having nutrient limitation at all
times, this result complicates the later interpretation of of
radiocarbon productivity measurements. Changes in productivity could
have been related to fluctuating nutrient concentrations, masking any
effect that the net-pens would produce. There are, however, some
important bits of information to be gleaned from the rest of this
experiment.

There were some possible trends of nutrient concentration, such as a
a gradient of increasing dissolved nitrogen from south to north at
low tide before the fish release, but without further sampling
between stations, it is not possible to conclude that the trend was
real. Considered alone, dissolved nitrogen to phosphorus ratios
(table 4),were indicative of nitrogen limitation (Redfield 1958,
Ryther and Dunstad 1971, McCarthy 1980, URS 1986a), but the apparent
imbalance is totally overshadowed by the absolute concentrations of
nitrogen, which indicated no nitrogen limitation before fish release
and only partly after the fish release.

The precision of the measurements was very good since variation among
replicates nitrogen samples was very small. The variation among
replicates so small that it was graphically impossible to represent 5
of the 6 error bars in figure 3b.

Chlorophyll a: In general, increased chlorophyll -4 concentrations
were observed at all stations before fish release, matching the
greater concentration of dissolved nitrogen at that time (compare
Figs. 3 and 4). The concentration of chlorophyll o was much greater
than reference stations at the net-pens at low tide on May 25, but
only slightly greater by high tide (Fig. 4a). Since replicates were
not obtained, statistical significance is not known. However, the
concentration at low tide near the net-pens was nearly four times

- 17

Chloro"11 0
Before Fish RGIGUM NOY 25

US/I LAw TWO
3MSh Tide
20

L PSGIG Pam& met Pe L POGIG Pam

AftAr Fish Relecaft June 21

WS/I
25 Low TWO
Dip Tide
20

L PMIQ PM& L PM16 Pam
Figure 4. Chlorophyll A concentration before fish release (top-
Fig. 4a) and after most Of the fish were released
Al I FS

L7kjnE-7A-

(bottom, Fig. 4b).

- 18 -

greater than the reference areas and likely to be significantly
greater.

After most of the fish were released, on June 21 (Fig. 4b), there was
significantly less chlorophyll a (p <0.05) at the net-pen site,
compared to reference stations at low tide. However, by high tide no
statistical differences were apparent as suggested by the overlapping
standard deviation bars of that tide shown in figure 4b. These
results suggest that tidal related natural variation of phytoplankton
density occurred at the net-pen site.

Primary Productivity: Statistical analyses of the raw data
(appendices a-1 and A-2) indicated that there were significant
differences in primary productivity between the net-pen site and the
reference areas at low tide, but not at high tide (p < 0.05). Due to
methodological differences in processing the samples, there could be
no comparison of absolute values between sampling dates.

A similar general trends among stations and between tides was
apparent as seen in figures 5a and 5b. Increased productivity near
the net-pens during the morning low tide was noted on both dates.
Since there werer few statistical differences, this trend can only be
noted and not interpreted as significant. Note that standard error
bars were shown here, to save space, not standard deviation bars.
Standard error is the standard deviation divided by the square root
of the number of replicate measures. If standard deviation was shown
on figure 5b, there would be greater overlap of bars.

There were no statistical differences among stations or times of
sampling on June 21. Variance within replicate samples was fairly
high, as is common for radiocarbon data from the field. Some of the
replicate samples could have contained statistical outliers (see
replicate values in appendix table A-2, pens). However, this data
was not excluded from the analysis.

The results of the productivity experiment indicate significant
tidally related variation at the net-pens before the fish were
released, although rates of primary productivity were not
significantly different among stations. On there own, these data
suggest that the net-pens did not produce a significant effect, but
since dissolved nitrogen concentrations and chlorophyll a results
varied significantly among some of the stations and times, it is
impossible to conclusively state that there was no effect. This is
typical in all field studies, where correlative data is the norm.

Primary Productivity
Before Fish Releases May 25

C/ M a/ tr.
is -
14 L&u Tide
Hip Tide
Iz

L Peale Pose. WRt-Perw IL Peale Page.

Primary Productivity
After Fish Releases June 21

P9 C/ m m/ hr.
so -
El Low Tide
71D 22 High Tide
so

L Peale Pase. K Peale Poes.

Figure 5. Primary productivity before fish release (top.,
Fig. 5a) and after most of the fish were released
(bottom, Fig. 5b). Standard error bars are shown where
appropriate. N = 6 for each station.,

- 20

Cell Counts: A summary of the species assemblage present during
portions of the study is shown in table 5. On both sampling dates
diatoms were by far the dominate phytoplankton. During the May
sampling Coscinodiscus sm. and Thalassiosira spD., common chain
forming species were prevalent. By late June Chaetoceros, also a
very common diatom group, were most abundant. The Chaetoceros sm.
present were members of the subgenus Hyalochaete, thought not to be
directly responsible for fish kills in net-pens in the past as have
been members of the subgenus Phaeoceros (Gaines and Taylor 1986).
However, reports from Canada (Bell et al. 1974) and studies in
Washington State (Rensel et al. 1989), indicate that these species
may contribute to mortality of net-pen salmon when abundant. There
were not enough phytoplankton samples to make a rigorous comparison
between tide stage or to other environmental factors. However, the
slightly greater number of cells at the net-pens versus the reference
station on May 25 appeared to correlate with nutrient and chlorophyll
a values at high tide. The net-pen site had greater counts at high
tide of June 21, which correlated with increased nutrient
concentration, primary and specific productivity, but not with
chlorophyll a concentration.

Table 5. Summary of cell counts from 2 m depth in cells per ml.

BEFORE FISH RELEASE: Hay 25, 1350-1420 hr.

S. Peale Pen site N. Peale
Chaetoceros Hyalochaete spp. 51 0 not
Coscinodiscus ongusta-lineata 118 191 sampled
Thalasslosira �M. 24 16
total diatoms 231 278
total dinoflagellates 22 131
-------------------- ---- ----
total phytoplankton (c/ml) 536 664

AFTER FISH RELEASE: June 21, 1220-1310 hr.

Chaetoceros Hyalocbaete spp. 536 1,639 740
Skeletonema costatum 6 25 80
Misc. pennate diatoms 2 59 22
total diatoms 645 1,825 952
total dinoflagellates 28 52 52
-------------------- ---- ---- ----
total phytoplankton (c/ml) 806 2,069 1,109

- 21 -

Experiment B

During the morning flood tide, the concentration of dissolved
nitrogen immediately upstream of the net-pens averaged 0.07 mg/l
(table 6). The dissolved nitrogen (DN) was composed of nitrate (78%)
and ammonia (22%). Just downstream of the net-pens, the
concentration of dissolved nitrogen had doubled to 0.15 mg/l, which
was significantly greater than upstream (p <0.005). However, most of
the dissolved nitrogen was in the form of nitrate (N03 = 86% of DN),
not ammonia (NH4 + = 12% of DN). A similar picture emerges if the
data is viewed on a gram atom weight basis; the downstream nitrate
was about 2.5 times greater than the ambient concentration. There
was no statistical difference in the concentration of ammonia during
this tide (on either unit basis) from upstream to downstream, but
nitrate was significantly greater downstream (p <0.001).

Total ammonia concentration in the plume of the net-pen during the
morning flood averaged 0.0163 mg/l which is equivalent to an un-
ionized ammonia concentration of 0.00037 mg/l at pH 8 and temperature
13 C. The EPA (1986) chronic exposure criteria (four day average
concentration) for un-ionized ammonia under these conditions is 0.030
mg/l, about 80 times greater than the concentration observed
downstream from the net-pens.

During the early portions of the ebb tide, the concentration of total
dissolved nitrogen was significantly greater within the net-pens
compared to upstream values (p <0.001 for NH + and N03, p <0.005 for
N02). However, there was no significant dMerence in the upstream
and downstream values of total dissolved nitrogen. Within that
total, there was a significant increase in the concentration of
ammonia (table 7). The proportion of toxic, un-ionized ammonia in
the downstream area was about 90 times less than the EPA's (1986)
chronic exposure criteria.

The concentration of un-ionized ammonia within the center of the net-
pens during the ebb tide was 23 times greater than the upstream,
ambient concentration. However, this amounted to only about 10% of
the chronic exposure criteria mentioned above. The within-pen
results were from the widest set of net-pens (three pens abreast),
under maximum loading, and minimal current conditions. Ammonia
loading within other net-pen systems at Squaxin Island was probably
less due to the two abreast configuration which allows greater
dilution with surrounding waters.

22 -

Table 6. Results of experiment 8 from May 25, 1988. All date collected
from 2 meters depth where the water temperature was 12.2 C *. pH
remained at 8.0 throughout this sample collection.

Dissolved
Dissolved Oxygen Nitrogen** Total Un-ionized Current
Tidal Stage Oxygen Saturation (Mean & SD) Ammonia Ammonia Velocit
& station (mg/1) W (mg/1) (mg/1) (mg/1) (cm sec-@)
---------- --------- -------- -------- -------- -------- --------

MID-FLOOD: 1200-1300 hours

upstream: 11.7 132% 0.07 (0.020) 0.0155 0.00036 11.4

downs- 11.6 131% 0.15 (0.011) 0.0163 0.00037 5.2
stream

EARLY-EBB: 1500-1625 hours

upstream: 12.2 137% 0.03 (0.006) 0.0069 0.00014 8.1

within pens*** 10.4 117% 0.23 (0.006) 0.1606 0.00337 3.0

downstream 12.6 142% 0.03 (0.015) 0.0155 0.00033 7.9

Total net-pen salmon biomasss was 118,600 kg for the flood and 97,970
kg for the ebb, due to differing sampling location. Ebb tide samples
were collected downstream (south) of the larger WDF pens (number 2).

Dissolved nitrogen values represent the total of nitrate, nitrite and
ammonium. Nitrogen limitation for diatoms may occur below 0.04 mg/l.

within pens site was In center of the pen system (number 2, WDF) that
were 3 cages wide by 11 cages long with 52,640 kg of fish.

- 23 -

Table 7. Mmonia concentrations at the Squaxin Island net-pens during the upstreat/downstreas
analyses (experiment B). Distance to downstream sampling station was 30 m from facility number 3.
Percent change in concentration is relative to ambient, upstream concentration. chronic exposure
concentration is the four day average recommended by EPA (1986).

Total Mmonia Concentration (mg/1)
with standard deviation in parenthesis
Tidal Influencing --------------------------------------
Fish Biomass (Kg) Upstrea Within Pens Downstream

Flood tide 118,600 kg 0.0155 ---- 0.0163
(0.00686) (0.00202)

percent increase of total ammonia 5.01
percent increase of un-ionized ammonia 2.3%
percent of chronic toxicity concentration 1.2%

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

Ebb tide 97,970 kq 0.0069 0.1606 0.0155
(0.00038) (0.00337) (0.00262)

percent increase of total ammonia 2,337% 224%
percent increase of un-ionized ammonia 2.1% 2.1%
percent of chronic exposure concentration 11.2% 1.11

Discussion

EXperiment A: Effects on Phytoplankton

Phytoplankton abundance and the growth rate at the net-pens varied
significantly with the stage of the tidal cycle or the time of day.
The most striking feature of this analysis was the very similar
pattern of primary productivity seen before and after fish were
release from the pens (Fig. 5). There was a general trend toward
increased phytoplankton density (chlorophyll a) and C-14 productivity
at the net-pens during the morning ebb tide, with no significant
difference during the afternoon flood tide, regardless of the
sampling date and amount of fish held in the net-pens (compare figs.
5a and 5b).

Dissolved nitrogen to phosphorus ratios (table 4) indicated that
growth was not limited due to nutrient depletion during the period
before fish release, when greater phytoplankton density and
significantly greater primary productivity was observed at the net-
pens. Thus, the observed increased chlorophyll a during one tidal

24 -

period of maximum fish biomass was not the result of nutrient
production from the net-pens, but may be related to the natural
hydrodynamics of Peale Passage.

The reference areas used in experiment A were selected to be within
the same passage, but as remote as possible from the net-pens. These
areas were near the source waters of Pickering and Dana Passages,
which are both subject to fairly intense horizontal and vertical
mixing (URS 1986b) compared to the shallow Peale Passage area. It is
likely that vertical mixing, and accordingly, light limitation for
phytoplankton cells in both of the source areas is much greater than
in Peale Passage. By the time source waters bearing nutrients and
phytoplankton seed stock reach the middle portions of Peale Passage,
a day or longer has passed, allowing for significant growth to occur
in the water column, which is entirely within the euphotic zone.

Although similar trends in C-14 primary productivity were similar on
both sampling dates (figures 5a versus 5b), there was no statistical
difference among sites on the later sampling date, and only on one
tide of the earlier date. Further, the absolute values were
different between sampling dates due to methodological differences
and thus no comparison of primary production rates is possible
between sampling dates.

Dissolved nutrient concentrations appeared to be roughly correlated
with chlorophyll a values and were lower at all stations after most
of the salmon were released. As discussed previously, Pease (1977)
found an inverse relationship between the two throughout his study of
Henderson Inlet. No causal relationship between these factors at the
Squaxin Island net-pens is likely, due to the remoteness of the
unperturbed reference areas, the relatively small amount of dissolved
nitrogen contributed by the net-pens (theoretically about 33 kg/day
total dissolved nitrogen) and results of nutrient sampling over the
tidal cycle discussed below.

Most of the dissolved nitrogen concentrations at all stations during
this study were greater than the threshold of limitation for diatom
growth (0.04 mg/l; URS 1986). To show maximum effects, this study
should have been conducted when surface waters were nearly depleted
of dissolved nitrogen, if it ever occurs in the area, to test the
possibility that the pens could cause or sustain a phytoplankton
bloom. These results and pre-experiment nutrient sampling indicated
that dissolved nitrogen values at the pens fluctuated regularly,
above and below the 0.04 mg/l growth limitation threshold. This
value obtained for nutrient limitation of diatoms is not absolute and
should be used with caution because growth limitation also depends
upon the concentration of phytoplankton cells and organic matter in
the water, rates of remineralization of the organic matter to
dissolved inorganic nitrogen (Harris 1986), and the type and size of
phytoplankton (Redfield 1958, Ryther and Dunstad 1971, Eppley 1972,
McCarthy 1980, URS 1986a).

25 -

On May 25, during the period of maximum fish biomass, increasing
concentrations of dissolved nitrogen were seen at low tide on a South
to north, long-channel axis (Fig. 3a). This observation suggests
that nutrient-rich waters were entering from the north end of the
passage at that time. Due to the location of the net-pens relative
to the ebb tide flow (Fig. 2), it also suggests that nutrients from
the net-pens were not the major source of the higher concentrations
of dissolved nitrogen seen at the north end of the passage. Later
that day during the flood tide, a similar concentration of dissolved
nitrogen, about 0.07 mg/l (versus 0.09 mg/1), previously seen at the
south end of the passage was observed immediately upstream of the
net-pens in the results of the upstream/downstream analyses
(experiment B). If time and budgets allowed, additional sampling
between stations would have been useful to examine the trends more
closely.

The elevated level of chlorophyll A at the net-pens prior to fish
release and during the morning low tide (Fig. 4a; 24 ug/1), was
nearly 4 times that of the reference stations. This non-replicated
value was near the maximum values seen in two years of sampling in
central Puget Sound (e.g., Anderson et al. 1984), but was much lower
than summer values seen in nearby Henderson Inlet at control and
experiment stations by Pease (1977). By high tide', the concentration
of chlorophyll a had diminished at the Squaxin Island net-pen site.

After the fish release, there was significantly less chlorophyll a
at the net pens on the morning low tide, but no difference by the
afternoon. Compared to the period prior to fish release, and if one
disregarded other data collected in this study, this suggests an
effect due to the pens. However, the fact that dissolved nutrients
were not limiting to phytoplankton growth prior to fish release is a
more important factor and discounts any possible effect suggested by
the increased chlorophyll a.

Another factor to be accounted for is the biomass of fish stock on
hand before and after fish release. As previously mentioned, about
40% of the initial biomass (47,200 kg, both delayed release and fish
for commercial purposes) was still on hand during the June 21
sampling date. However, this amount of biomass is small relative to
the maximum amount of biomass that a typical two acre net-pen farm
could maintain (up to 250,000 kg, J. Lindberg, pers. comm. in Weston
1986). 1 would have preferred that more of the fish had been
released for the later sampling date, to provide a more
representative experimental control, but that was not possible.

These results suggest that tidal stage, time of day and ambient
nitrogen conditions were more important determinants of phytoplankton
conditions at the net-pens and that nutrients from the net-pens did
not produce a significant, measurable effect on the phytoplankton
production. If time and materials allowed, several sampling stations
midway between the pens and the reference areas would have been
useful to search for a gradient of effects. Nevertheless, given the
natural excess of total and dissolved nitrogen that existed
throughout the study area at the time of the study, no effects from

the pens were possible. Nitrogen may be limiting in the study area
later in the summer, but most of the salmon have generally been
released by that time.

Experiment B: Effects on Near Field Water Ouality: Ammonia

This experiment involved measurement of nutrient levels upstream and
downstream of the net-pen cages on May 25, when the maximum amount of
salmon was present in pens. Divergent dissolved nitrogen results
were seen between the two tidal stages monitored; the morning flood
showed significantly elevated levels of dissolved nitrogen
downstream, but only a very small, statistically insignificant
increase in ammonia concentration. Monitoring during the afternoon
ebb showed no significant difference in dissolved nitrogen
concentration, but within that measure, the total ammonia
concentration increased greatly as a percentage of upstream, ambient
levels. However, the maximum concentration of un-ionized ammonia was
only about 10 % of the EPA (1986) chronic exposure level for
11salmonids and other sensitive coldwater species" and far below acute
toxicity criteria.

The results of the flood tide analyses suggest that very rapid
nitrification of ammonia occurred in the plume of the net-pen,
consistent with general concepts of marine chemistry (Harris 1986).
Most of the dissolved nitrogen was in the form of nitrate (N03 = 86%
of dissolved N), not ammonia (NH4+ = 12% of dissolved N) a distance
of 30 m downstream of the third net-pen system.

The concentration of dissolved nitrogen was the same at the upstream
and downstream stations on the afternoon ebb (0.03 mg/1). Total
ammonia was about 52% of the dissolved nitrogen in the downstream
samples, compared with 70% inside the pens. In addition to
nitrification, reduced ammonia downstream was apparently due to
dilution during the ebb tide measurements.

The rate of ammonia nitrification can be approximated as follows,
using the flood tide data. Assuming that 70% of the dissolved
nitrogen within the pens was ammonia, as it was on the ebb, and given
the average current velocity of 8.3 cm sec-1, the mean distance from
the center of all three net-pen systems to the downstream sampling
location was 290 m. Accordingly, ammonia within the cages was
converted from 70% to 12% of the total dissolved nitrogen
concentration within about 1 hour. By then, the concentration of un-
ionized ammonia was far below the exposure criteria for sensitive
species such as salmonids (EPA 1986).

It has previously been conservatively assumed that all dissolved
nitrogen produced by the net-pen fish was in the form of ammonia or
urea, not nitrate (Weston 1986). While this is apparently true
within the pens, the results presented here substantiate that
nitrification converts the ammonia to nitrate over very short time
periods, rapid compared to the doubling time of phytoplankton
populations (minutes versus hours or days, respectively). The
dominant dissolved nitrogen compound measured immediately downstream

27 -

of the net-pens is nitrate. Nitrate is less preferred by
phytoplankton cells, and its uptake rate is slower than that of
ammonia.
The calculated production rate of dissolved nitrygen from the net-pen
should be at least 0.22 to 0.28 g kg_1 fish day- greater than the
ambient concentration. These values represent the sum of nitrate and
ammonia produced by salmon (SAIC, 1986 and Weston 1986,
respectively), multiplied by 0.87, the soluble fraction. Using the
greater figure to be conservative, on the flood tide there should
have been a net increase of 33.2 kg/day or 0.384 gram sec-1 nitrate
and ammonia in the plume of the net-pen (118,600 kg x 0.28 g divided
by 86,400 seconds per day). When spread over the average cross
sectional area of all the pens (93.6 m long x 4 m deep = 374 m2), and
dispersed with the current (0.083 m sec 1), this is equivalent to
0.012 mg/l greater than_the upstream, ambient concentrations (37.9 kg
divided by 31.04 m3 sec 1)@ The observed downstream concentration
was 0.08 mg/l, about six times greater than ambient dissolved
nitrogen, but less than an order of magnitude different. Variation
among sample replicates was very little, lending credibility to the
results. Using similar calculations for the ebb tide results, there
should have been a similar predicted increase of dissolved nitrogen
concentration downstream, but none was observed.

The lack of any measurable increase in dissolved nitrogen on one
tidal phase and the increase on the other could be due to tidal
hydrodynamics of the site and fish physiology. I have conducted
similar studies at several other locations and have found increased
concentrations of ammonia within or immediately downstream of the
every net-pen system monitored (appendix B), but total nitrogen
values have been more variable, sometimes even less than the upstream
values.

Measured increase of nutrients that exceed predicted concentrations
can be explained by several factors. The predicted increase was
based on literature from freshwater hatcheries, not from marine net-
pens. The Squaxin Island net-pen systems is mostly used for delayed
release of relatively small (<40 g) coho salmon that require more
rearing space per pound of fish than larger fish typically held in a
commercial grow out facility. Accordingly, there-is more netting and
floats for growth of fouling organism, that may contribute nutrients.
In addition, the net-pens are not removed and cleaned at the delayed
release facility during the rearing period. The nets in a delayed
release facility are installed in late winter and left in place until
late spring to early summer. By the initial sampling time of this
study (late spring), there was a considerable accumulation of
invertebrate and algal fouling organisms on the nets. This condition
is tolerated since the nets are removed in early summer for the
remainder of the year. To determine the contribution of nutrients
from floats and invertebrates growing on the nets, it would be useful
to measure nitrogen concentrations at the Squaxin Island net-pens
after the salmon were released, but before the nets are removed.

28 -

Finally, there is a inverse relationship between fish size and rate
of metabolite production. Clark et al. (1985) found rapidly
decreasing rate of ammonia production with increased fish size, about
a 50% reduction from the 10 gram to 200 gram mean weight. The
dominance of small fish at the Squaxin Island delayed release net-
pens would therefore result in proportionately greater nitrogen
discharge than a grow out facility with fish ranging in size from 20
grams to 10 kg or larger.

The lack of increased dissolved nitrogen downstream of the net-pens
on the ebb tide could be due to dilution or unknown and irregular
water motion. A one m deep drift stick placed immediately downstream
of the net-pens moved very little for 20 minutes prior to the ebb
tidal sample. From this observation, and numerous other observations
by the author at some other facilities during low velocity periods of
water movement, there appears to be an area immediately downstream of
some net-pen facilities that may temporarily exhibit lack of water
movement or anomalous patterns of water flow.

Other recent data collected to more accurately predict expected
dissolved nitrogen concentrations downstream of net-pen facilities
(appendix B) are insufficient in quantity at this time for regression
analyses. The following narrative describes some of the recent
studies.

Studies conducted near the world's largest net-pen facility, Domsea
Farms, Inc., found no measurable effect on downstream water quality
(D. Damkaer, KMFS, unpublished data cited in Weston 1986). Tidal
flows in that area are greater than at the Squaxin Island net-pen
site and annual production was approximately 8 times greater than at
Squaxin Island.

Milner-Rensel Associates (1986) found similar results for ammonia
production in a study of water quality near a relatively small net-
pen system in Port Angeles Harbor (see appendix B). The pens
contained 27,000 kg of fish and there was mean current velocity of 7
cm sec 1 during sampling. The concentration of ammonia increased
within the net-pens, but immediately downstream the ammonia was
converted to nitrate and diluted. The concentration of total ammonia
inside the net-pens at Port Angeles was 0.020 mg/l, compared to 0.007
mg/1 in the ambient, upstream water. The total ammonia concentration
had diminished 30 m downstream of the net-pens to 0.011 mg/l, or
0.004 mg/1 greater than the ambient, upstream value. Downstream
nitrate increased 0.049 mg/l over ambient, although there was a
fairly high variance within replicates. As the site was so near the
Strait of Juan de Fuca, total dissolved nitrogen was much higher than
at Squaxin Island and was 97% nitrate, the remainder being mostly
ammonia.

Two years later in Port Angeles Harbor there were 192,000 kg of fish
on hand with a current velocity of about 8 cm sec I (Rensel,
unpublished data). In spite of the relatively large size and minimal
currents during sampling, the concentration of toxic un-ionized

- 29 -

ammonia was less than 6% of the EPA (1986) four day chronic exposure
level for I'salmonids and other sensitive coldwater species".

Another upstream/downstream analysis was conducted at a very small
facility in the first year of operation at north Skagit Bay. The
results indicate extremely minor increases of ammonia (0.001 mg/1
increase in total ammonia downstream), and total dissolved nitrogen
levels actually decreased downstream (Rensel 1988).

- 30 -

Literature Cited

APHA. 1985. Standard Methods for the Examination of Water and
Wastewater. American Public Health Association, American Water
Works Association, Water Pollution Control Federation. sixteenth
edition. Washington D.C. 1,268 pp.

Anderson, J., A. Copping, T. Jagielo, J. Postel, W. Peterson, B.
Dumbauld, G. Heron, R. Hood, and M. Strom. 1984. Water Column
Ecology, Vol. III, Sec. 4. In: Q. J. Stober and K. K. Chew.
Renton Sewage Treatment Plant Project, University of Washington,
Seattle.

Bell, G. R., W. Griffioen, and 0. Kennedy. 1974. Mortalities of pen-
reared salmon associated with blooms of marine algae, pp. 58-60.
In: Proc. Northwest Fish Culture Conference, 25th Anniversary,
December 4-6, 1974, Seattle, WA.

Clark, E.R., J.P.-Harman and J.R.M. Forster. 1985. Production of
metabolic products by intensively farmed rainbow trout, Salmo
gairdneri Richardson. Journal of Fisheries Biology. 27:381-393.

Collias, E.E. and J.H. Lincoln. 1977. A study of the nutrients in
the main basin of Puget Sound. Report by the University of
Washington, Department of oceanography for the Municipality of
Metropolitan Seattle. M77-2. Seattle.

EPA. 1986. Quality criteria for water. U.S. Environmental
Protection Agency, Washington D.C.

Eppley, R.W. 1972. Temperature and phytoplankton growth in the sea.
Fisheries Bulletin 70:1063-85.

Fraser, J. 1976. Salmon pen culture at Squaxin Island. Final Report.
Squaxin island Tribe. Shelton, Washington. Contract No. 07-6-
01521 of Economic Development Agency. Washington D.C.

Gaines, G. and F. J. R. Taylor. 1986. A Mariculturist's Guide to
Potentially Harmful Marine Phytoplankton of the Pacific Coast of
North America. Information Report No. 10. Prepared for the
Marine Resources Section, Fisheries Branch, British Columbia
Ministry of Environment. 54 pp.

Gowen, R.J., N.B. Bradbury and J.R. Brown. 1985. The ecological
impact of salmon farming in Scottish coastal waters: a
preliminary appraisal. International Council for the
Exploration of the Sea. C.M. 1985/F:35 13 pp.

Gowen, R.J. and N.B. Bradbury. 1987. The ecological impact of
salmonid farming in coastal waters: a review. oceanography and
Marine Biology Annual Review 25:562-575.
Gowen, R., J. Brown, N. Bradbury and D.S. McLusky. 1988.
Investigations into benthic enrichment, hypernutrification and
eutrophication associated with mariculture in Scottish coastal

31 -

waters (1984 - 1988). Report prepared under contract to
Highlands & Islands Development Board, Crown Estates
Commissioners, Nature Conservancy Council, Countryside
Commission for Scotland, and the Scottish Salmon Growers
Association. 289 pp.

Harris, G.P. 1986. Phytoplankton Ecology, Structure, Function and
Fluctuation. Chapman and Hall. London.

McCarthy, J.J. 1980. Nitrogen, in I. Morris (ed.), The Physiological
Ecology of Phytoplankton, University of California Press,
Berkeley.

Milner-Rensel Associates. 1986. Aquatic conditions at the Seafarm of
Norway net-pen site in Port Angeles Harbor in April, 1986.
Prepared for Sea Farm of Norway, Inc. and the City of Port
Angeles Planning Department. 25 pp. and appendix.

Moring, J.R. 1973. Aspects of growth, and the effects of some
environmental factors on pen-reared chinook salmon, Ongorhynchus
tshawvtscha (Waulbaum) in Puget Sound, Washington. Dissertation.
University of Washington. Seattle. 222 pp.

Nishitani, L., G. Erickson, K.K. Chew. 1988. PSP research:
Implications for Puget Sound in Vol. 2, pps. 392-399,
Proceedings of the Puget Sound Water Quality Authority Research
Conference, Seattle.

Oclay, N. 1959. Oceanographic conditions near the head of Southern
Puget Sound August 1957 through September 1958. M.S. Thesis.
University of Washington, Seattle.

Pease, B.C. 1977. The effect of organic enrichment from a salmon
mariculture facility on the water quality and benthic community
of Henderson Inlet, Washington. Ph D. Dissertation. University
of Washington. Seattle.

Parsons, T.R., M. Takahashi and B. Redgrave. 1984. Biological
oceanographic processes. 4th edition. Pergamon Press. 330 pp.

Redfield, A.C. 1958. The biological control of chemical factors in
the environment. American Science 46:205-222.

Ren'sel, J. 1988. Environmental sampling at the American Aqua Foods
net-pen site near Lone Tree Point in north Skagit Bay,
Washington. Prepared by Rensel Associates for Pacific Aqua
Foods and the Washington Department of Natural Resources. 7 pp.
and 19 pp. appendix.

Rensel,J.E. and E.F. Prentice. 1980. Factors controlling growth and
survival of cultured spot prawn, Pandalus platyceros, in Puget
Sound, Washington. Fish. Bull., U.S. 78:781-788.

- 32 -

Rensel, J.E., R.P. Harris, T.T. Tynan. 1988. Fishery contribution
and spawning escapement of coho salmon reared in net-pens in
southern Puget Sound, Washington. North American Journal of
Fisheries Management 8:359-366.

Rensel, J.E., R.A. Horner, J.A. Postel. 1989. Phytoplankton blooms
in Puget Sound, Washington, and their affect on salmon
aquaculture: initial research. Northwest Environmental Journal
5:53-69.

Ryther, J.H. and W.M. Dunstad. 1971. Nitrogen, phosphorus and
eutrophication in the coastal marine environment. Science (New
York) 171:1008-1013.

SAIC (Science Applications International Corporation) 1986.
Recommended interim guidelines for the management of salmon net-
pen culture in Puget Sound. Prepared for the Washington
Department of Ecology in conjunction with the Department of
Fisheries, Agriculture and Natural Resources. 48 pp.

Snyder, B.P., A.J. Didier, Jr., and E.O. Salo. 1974. The culture of
salmon at Wilipa Bay, Gray's Harbor and Henderson Inlet in
southern Puget Sound. Final report for period Jan. 1, 1973 to
Feb. 15 1974. Fisheries Research Institute, University of
Washington, Unpublished Report. 211 pp.

Steemann Nielsen, E. 1952. The use of radioactive carbon (C-14) for
measuring organic production in the sea. J. Cons. Int. Explor.
Mer. 18:117-140.

Strickland, J.D.H. and T.R. Parsons. 1968. A Practical Handbook of
Seawater Analysis. Fisheries Research Board of Canada Bulletin
167, 311 pp.

Stickney, R.R. 1979. Principles of Warmwater Aquaculture. John
Wiley & Sons. New York. 375 pp.

Small Tribes of Western Washington (STOWW) 1974. Salmon pen culture
at Squaxin Island Indian Reservation. J. Fraser and G. Milner,
eds., Final Report for Economic Development Administration
Contract No. 07-6-09420-1. and the Bureau of Indian Affairs
Contract No. 14-20-0500-3517. Washington D.C. 68 pp.

United Nations Educational, scientific and Cultural organization
(Unesco) 1978. Phytoplankton Manual. A. Sornia, ed., Paris. 337
PP.

- 33

URS Co. 1986a. Southern Puget Sound water quality assessment study,
final report: Comprehensive circulation and water quality study
of Budd Inlet. Prepared for the Washington State Dept. of
Ecology. Seattle.

URS Co. 1986b. Southern Puget Sound water quality assessment study,
final report: Circulation and flushing in southern Puget Sound.
Prepared for the Washington State Dept. of Ecology. Seattle.

Welch, E.B. 1980. Ecological Effects of Waste Water. Cambridge
University Press. Cambridge. 337 pp.

Weston, D.P. 1986. The environmental effects of floating mariculture
in Puget Sound. School of Oceanography, College of Ocean and
Fisheries Sciences. Seattle, Washington. 148 pp.

Weston, D.P. and R. J. Gowen 1989. Assessment and prediction of the
effects of salmon net-pen culture on the benthic environment.
Prepared for the Washington Department of Fisheries,
Programmatic environmental impact statement, salmon aquaculture
in Puget Sound.

Winter, D.P., K. Banse and G.C. Anderson. 1975. The dynamics of
phytoplankton blooms in Puget Sound, a fjord in the northwestern
United States. Marine Biology 29:139-176.

Zar, J.H. 1984. Biostatistical Analysis. Prentice-Hall, Inc. New
Jersey. 718 pp.

Acknowledgement

The author wishes to thank the following individuals for their
assistance as noted. Mr. Richard Davis, University of Washington,
for C-14 work, field assistance and preliminary manuscript review.
Dr. Dale Kiefer, University of Southern California, and Dr. Donald
Weston, University of Washington for manuscript review. Ms. Jan
Downs, University of Washington for chlorophyll A analysis. Dr. Rita
Horner, University of Washington, for phytoplankton identification
and preliminary manuscript review. Mr. Richard Cromega for figure 1,
Ms. Beth Haskins for figure 2.

Appendix A
Appendix Table A-1. Productivity data from Squaxin Island
on May 25, 1988.

Total Activity
SAMPLE CPM H# DPM
LA1 45322 126.3 48385
LA2 40302 107.3 42667
LA3 41287 105.7 43688
LA4 41640 105.3 44056
LAS 43032 105 45525
LA6 37755 104.3 39934
LA7 44236 105.3 46803
LA8 41820 104.3 44234
LA9 40342 104 42667
LAIO 34420 105 36414
LA11 30763 105.7 32552
LA12 29903 104.7 31632
LB4 34385 105 36377
LB5 45706 103.3 48331
DA10 37593 104.7 39768
LBIO 42070 104.3 44499
L- 40516 107 42890
DA12 39540 103.7 41816
mean & sd 41791 4877

Sample Cpm H# dpm MgCIMA3 MgC/MA31h site mean st dev
N. Peale 11117 152.3 12098 35.09 6.19 7.37 2.36
1100 h 14557 141.7 15701 45.53 8.04
6355 144.3 6869 19.92 3.52
12779 144.7 13816 40.07 7.07
16809 149.3 18244 52.91 9.34
18191 143.7 19651 56.99 10.06
NP dark 4681 144.3 5059 14.67 2.59 2.49 0.13
4350 140.3 4687 13.59 2.40
S. Peale 6411 146.3 6940 20.13 3.55 3.96 0.64
1100 h 8969 140.0 9661 28.02 4.94
5904 144.0 6379 18.50 3.26
6374 141.7 6875 19.94 3.52
7269 144.7 7858 22.79 4.02
8044 142.3 8680 25.17 4.44
SP dark 5286 143.3 5709 16.56 2.92 2.77 0.21
4740 143.7 5121 14.85 2.62
Pens 24502 142.7 26448 76.70 13.54 14.10 3.40
1100 h 34093 143.3 36819 106.78 18.84
19775 143.0 21350 61.92 10.93
25980 155.3 28352 82.22 14.51
30384 146.7 32904 95.42 16.84
18019 144.3 19475 56.48 9.97
Pens dark 6009 141.0 6478 18.79 3.32 2.74 0.82
3918 139.7 4220 12.24 2.16

Appendix Table A-1 Continued
Pens 7065 140.7 7615 22.08 8.28 7.33 0.83
1420 h 5805 141.3 6259 18.15 6.81
5461 144.3 5902 17.12 6.42
6984 138.7 7516 21.80 8.17
5574 146.7 6036 17.50 6.56
6585 143.0 7110 20.62 7.73
Pens dark 5023 143.0 5423 15.73 5.90 5.98 0.11
5151 144.7 5569 16.15 6.06
N. Peale 8788 143.0 9488 27.52 10.32 9.83 2.3 5
1420 h 7746 141.3 8353 24.22 9.08
10884 142.7 11748 34.07 12.78
6095 142.7 6579 19.08 7.15
NP dark 5074 142.7 5477 15.88 5.96 7.57 2.29
7823 143.7 8451 24.51 9.19
S. Peale 9809 142.7 10588 30.71 11.52 12.59 3.90
1420 h 10953 144.0 11835 34.32 12.87
15027 145.0 16250 47.13 17.67
7061 142.7 7622 22.10 8.29
SP dark 8513 142.7 9189 26.65 9.99 10-63 0.90
9567 147.3 10366 30.06 11.27

Appendix Table A-2. Productivity data from Squaxin Island
on June 21, 1988.

Total Activity Time Zero
SAMPLE CPM H# DPM SAMPLE CPM H# DPM
I-Al 30709 97.7 32431 LA1 12914 187.3 14640
LA2 32320 97.3 34130 LA2 17286 188.0 19615
LA3 40114 96.7 42355 LA3 23496 191.3 26792
DA2 31354 97.7 33112 DA2 15559 190.0 17707
LA4 42788 100.0 45209 LA4 14053 191.3 16025
LAS 33342 98.0 35213 LAS 15352 191.0 17497
LA6 32321 97.7 34134 LA6 19196 191.3 21888
DA4 57645 97.3 60872 DA4 15914 190.0 18111
LA7 35327 96.7 37300 LA7 23220 190.7 26454
LAB 41635 97.3 43966 LAB 18071 189.7 20558
LA9 56040 92.3 59131 LA9 20634 189.0 23449
DA6 46998 97.0 49626 DA6 28962 188.7 32899
LA10 34203 98.3 36125 LA10 13760 189.7 15653
LA11 32854 97.0 34691 LA11 17595 189.7 20015
LA12 30044 97.0 31725 LA12 20718 190.7 23604
DAB 36880 99.3 38961 DAB 16791 190.3 19118
1-131 33806 98.7 35709 1-131 15999 190.7 18227
1-132 44646 98.0 47152 L82 23215 190.0 26420
L83 55483 96.7 58582 1-133 31380 191.0 35766
DA10 35420 97.0 37401 DA10 38891 187.7 44114
LB4 31400 96.7 33154 LB4 45703 187.0 51788
1-135 41051 95.7 43337 1-135 26202 188.3 29746
1-136 38103 90.3 40196 1-136 35817 190.7 40805
DA12 47776 111.0 50643 DA12 20863 189.7 23733

-nean & sd 41465 mean & sd - 25193
8848 9650
Inc Time 1 - 6.42
Inc Time 2 - 3.00
SAMPLE CPIVI H# DPIVI Mg C/MA3 Mg C/MA3/h site means st err
N. Peale 39926 190.7 45486 240.17 37.41 42.42 4.96
1020 h 35680 190.7 40649 182.93 28.49
40080 189.0 45547 240.91 37.52
42000 188.7 47709 266.48 41.51
43650 189.3 49626 289.18 45.04
52944 189.3 60193 414.24 64.52
dark 24044 189.5 27344 25.46 3.97 -6.09 10.06
14498 187.3 16435 -103.65 -16.14
S. Peale 47033 189.0 53449 334.43 52.09 31.07 5.86
1020 h 44343 189.3 50414 298.51 46.50
31438 190.0 35778 125.29 19.51
33980 188.3 38576 158.40 24.67
31870 190.0 36271 131.11 20.42
33273 188.7 37795 149.15 23.23
dark 31688 190.0 36063 128.65 20.04 10.94 9.10

Appendix Table A-2. Continued

23020 189.7 26187 11.77 1.83
Pens 30478 189.7 34671 112.18 17.47 67.39 16.82
1020 h 29487 189.3 33524 98.60 15.36
59108 190.0 67270 497.99 77.57
76650 187.3 86893 730-24 113.74
65200 189.0 74094 578.77 90.15
65085 189.7 74039 578.12 90.05
dark 31450 189.3 35756 125-02 19.47 22.09 2.62
33914 190.0 38597 158-65 24.71
N. Peale 29780 190.7 33927 103.37 34.46 51.66 16.51
1330 h 43497 191.3 49598 288.85 96.28
44550 188.7 50605 300-77 100.26
35031 188.7 39793 172.80 57.60
22695 189.7 25817 7.39 2.46
26412 188.3 29985 56.71 18.90
dark 25720 190.0 29271 48.27 16.09 0.40 15.69
18724 190.3 21319 -45.84 -15.28
S. Peale 25636 194.7 29384 49-60 16.53 65.47 18-93
1330 h 22595 188.3 25651 5.43 1.81
'48384 191.7 55204 355.19 118.40
41857 189.7 47615 265-38 88.46
43607 188.7 49534 288.08 96.03
38120 189.3 43339 214-77 71.59
dark 12257 190.0 13950 -133.07 -44.36 -22.69 21.67
21942 189.0 24935 -3.05 -1.02
Pens 25096 189.3 28532 39.52 13.17 31.80 12.34
1330 h 25007 189.0 28419 38.18 12.73
28747 190.0 32716 89.04 29.68
22293 187.7 25287 1.12 0.37
40063 189.0 45528 240-68 80.23
34303 190.0 39039 163-88 54.63
dark 21967 190.3 25011 -2.15 -0.72 15.55 16.26
29307 188.0 33256 95-44 31.81

Appendix B

A number of studies have been conducted in Washington state to
assess the nearfield effects of net-pen operation on nutrient and
dissolved oxygen concentration in marine waters. Several of
these studies have been conducted in accordance with methods
outlined in the State of Washington's Recommended Interim
Guidelines for the Management of Salmon Net-pen Culture in Puget
Sound (SAIC 1986). Other studies, conducted prior to the
guidelines, were essentially the same, although within net-pens
sampling stations were utilized instead of the the first
downstream sampling station.

All of the near field studies (Appendix Table B-1) have shown
increased concentrations of ammonia immediately downstream or
within the net-pens. Total ammonia values have increased from 3
to 2,327%. However, the amount of un-ionized ammonia increased
only a few percent of the four-day maximum chronic exposure level
recommended by EPA (1986), ranging from 0.9 to 11.2%. The
maximum increase (11.2% of the recommended chronic exposure
value) was found within, not downstream, of the largest of three
net-pen systems at Squaxin Island. These pens were configured to
have three pens wide by 10 pens long, causing significantly
reduced water flow within the center pens, where samples were
collected. As the concentration of any waste nutrient is greatly
dependent on tidal flow rate, and the samples were collected in a
worst case area at only a few cm per second flow, this appears to
be a worst-available-case analysis. To compensate for reduced
flows and possible reduced growth, fish culturists at that site
have reduced fish loading within the center pens.

Studies of nutrients in "upstream" and "downstream" waters
immediately adjacent to net-pens in Washington state have also
indicated that relatively rapid rates of nitrification occur,
i.e., ammonia (NH4+) is oxidized by microbial action to nitrate
(N03): A typical scenario involves elevated concentrations of
ammonia within the net-pens but a short distance downstream, the
ammonia has been converted to nitrate.

The data has been included here to illustrate the amount of
ammonia produced by a wide variety of pens, of differing size.
The results show that adverse effects are minimal, even at
relatively large facilities. The continued collection of this
data will eventually allow more precise calculation of the total
dissolved nitrogen produced from marine net-pen reared salmon,
the fouling organisms on the nets and floats and the fish
associated with the net-pen facility. Such data will allow more
precise calibration of numerical models that are designed to
assess the possible cummulative effect of salmon net-pens in
restricted embayments. Presently the only data available for
this purpose is from freshwater hatchery culture of relatively
small fish.

Appendix Table B-1. Summary of dissolved nitrogen, total and un-
ionized ammonia production from marine net-pen farms in
Washington state compared to maximum recommended four day
exposure concentration (EPA 1986). Percent change in
concentration is relative to upstream concentration.

Concentration (mg/1) & Percent Change
-----------------------------------
Site - Data Source & 30 meters
Instantaneous loading (Kg) Upstrea Within Pens Downgtream
PORT ANGELES HARBOR l/ 27,000 kg

Total dissolved nitrogen--> 0.832 0.882 0.887
Total Ammonia __> 0.0074 0.0201 0.0119
NH4 + percent increase relative to ambient -> 272% 62%
percent of ammonia that is toxic (NH3) --> 1.8% 1.8%
percent of chronic toxicity concentration -> 1.5% 0.9%

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

6 meters 30 meters
Upstream Downstream Downstream
SKAGIT BAy 2/ (Lone Tree Pt.) 4,300 kg

Total dissolved nitrogen--> 1.067 1.131 1.139
Total Ammonia __> 0.0277 0.0303 0.0287
NH4 + percent increase relative to ambient -> 9% 3%
percent of ammonia that is toxic (NH3) __> 0.3% 0.3%
percent of chronic toxicity concentration -> 1.9% 1.8%

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

SQUAXIN ISLAND 3/ (Peale Passage) 118,600 kg

6 meters 30 meters
Upstream Downstream Downstream

Total dissolved nitrogen--> 0.07 ---- 0.15
Total Ammonia __> 0.0155 ---- 0.0163
NH4 + percent increase relative to ambient -> 5%
percent of ammonia that is toxic (NH3) __> 2.3%
percent of chronic toxicity concentration -> 1.2%

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

Appendix Table B-1, continued

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

SQUAXIN ISLAND 4/ (Peale Passage) 97,970 kg

Concentration (mg/1) & Percent Change
-----------------------------------
30 meters
Upstream Within pens Downstream

Total dissolved nitrogen--> 0.03 0.23 0.03
Total Ammonia __> 0.0069 0.1606 0.0155
NH4 + percent increase relative to ambient ->2,327% 225%
percent of ammonia that is toxic NH3 __> 2.1% 2.1%
percent of chronic toxicity concentration -> 11.2% 1.1%

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

PORT ANGELES HARBOR 5/ 192,500 kg
6 meters 30 meters
Upstream Downstream Downstream

Total dissolved nitrogen--> 1.4869 1.6688 1.6207
Total Ammonia __> 0.0343 0.0873 0.0516
NH4 + percent increase relative to ambient -> 254% 150%
percent of ammonia that is toxic (NH3) __> 1.5% 1.5%
percent of chronic toxicity concentration -> 6.4% 3.8%

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

PORT ANGELES HARBOR 6/ 192,500 Jkg
6 meters 30 meters
Upstream Downstream Downstream

Total dissolved nitrogen--> 1.5862 1.5771 1.4992
Total Ammonia __> 0.0122 0.0642 0.0336
NH4 + percent increase relative to ambient -> 526% 275%
percent of ammonia that is toxic (NH3) __> 1.5% 1.5%
percent of chronic toxicity concentration -> 4.7% 2.4%

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

Data sources and downstream velocity: 1) Milner-Rensel
Associates 1986, 8.0 cm sec-1 @ 30m downstream; 2) Rensel 1988,
38.8 cm sec 1 @ 6m; 3) See main text, f Iood tide, 5.2 cm sec-1;
4) See main text, ebb tide, 7.9 cm sec- . 5) Rensel unpublished
on flood tide, 8.2 cm sec -1; 6) Rensel unpubished on ebb tide,
8.4 cm sec 1.

APPENDIX D

INFECTIOUS DISEASES OF SALMON
IN THE PACIFIC NORTHWEST

FISH DISEASES

Metazoan Parasites. External copepods (Lepegphtheiru salmonis and Caligu sp.) and
monogenean gill flukes Laminiscus sterkowli are the only metazoan parasites that have
been observed in sufficient intensities to be considered significant pathogens of net-pen
reared fish in Washington (Kent and Elston, 1987b; L W. Harrell, NMFS, Manchester,
WA, pers. comm).
Protozoans. Diseases due to marine protozoan parasites are common in net-pen reared
salmon. These include Parvicapsula sp. (Myxosporea: Myxozoa) which causes kidney
disease in pen-reared coho salmon (Hoffman 1984; Johnstone 1984, Kent and Elston
1987b) for which cod is the likely reservoir for infection (Johnstone, 1984); Paramoeb
12emaquidensis, a ubiquitous, normally free-living amoeba which infects gills (Kent et al.
1988b); an unidentified protozoan (rosette agent) which infects inflammatory cells of
maturing chinook salmon in net-pens (Elston et al. 1986; Harrell et al. 1986); and a
microsporidan protozoan which infects blood-forming cells of chinook and causes severe
anemia (Elston et al. 1987).

Freshwater protozoan pathogens may also be transmitted with fish when they are
introduced to net-pens. Kent and Elston (1987b) observed infections by a microsporidan
similar to Loma salmonae (Microspora) in the gills of coho salmon held in net-pens.
These infections were apparently contracted in freshwater. Ichiyobod (Costi ) is a
common flagellate protozoan pathogen in freshwater which can apparently survive and
cause disease in fish after seawater transfer (Ellis and Wooten 1978) and it has
occasionally been associated with gill disease in pen-reared salmon in Washington.

Bacteri . Renibacterium salmoninarum, the causative agent of bacterial kidney disease
of salmonids, is widespread in net-pen reared salmon in Washington State, as well as
British Columbia, and is a serious threat to the industry (Evelyn 1988). Salmonid fishes
are the primary hosts for this obligate pathogen but herring (Clul2e harenLyu ) and black
cod (Anol2lom fimbria) can be infected experimentally by injection of the bacterium
(rraxler and Bell in press). It is believed that the organism is not part of the normal
bacterial aquatic microflora (Austin and Austin 1987) and salmon are the likely reservoir
for infection (Fryer and Sanders 1981). The disease can be transmitted either
horizontally (from fish to fish) or vertically within eggs (Evelyn et al. 1984), and it is
often exacerbated after infected fish are transferred to seawater (Banner et al. 1983).
The bacterium can be detected in pen-reared salmon several months after transfer to
seawater and the disease can be transmitted to other salmon in adjacent net-pens (Evelyn
1988). The bacteria persist in wild fish in seawater (Banner et al. 1986) and it is
probable that wild brood stocks are a source of infections in some fish farms (Evelyn
1988).

It is difficult to treat fish with clinical disease so prevention is the most common control
method. Prevention strategies include screening brood stock and discarding eggs from
positive females and screening smolts prior to seawater introduction. Erythromycin
injection of females prior to spawning appears to induce high enough levels of the
antibiotic in eggs to reduce vertical transmission to the fry (Evelyn et al. 1986) and this
practice has been initiated at several fish farms.

Furunculosis, caused by the Gram-negative bacterium Aeromonas salmonicid , often
causes severe disease in freshwater fishes. Although the bacterium often originates in
freshwater, it can apparently survive and spread in seawater (Scott 1968), and it has
been recognized as a pathogen in seawater in Washington (Novotny 1978). As with
Renibacterium, epizootic disease in salmon with latent infections occurs after transfer to
seawater (Cox et al. 1986; Smith et al. 1982). Though the disease is most often
observed in salmonid fishes, it has also been reported in several non-salmonid marine
and freshwater species (Elliot and Shotts 1980; McCarthy 1975; Morrison et al. 1984).
Furunculosis is usually treated with oxytetracycline or Romet 30. There is active research
on an effective bath immersion vaccine, but this is not routinely used in production
facilities.

Vibriosis, caused by marine bacteria Of the genus Vibrio, is a cosmopolitan disease
infecting many fish species. It frequently occurs in net-pen reared fish in Washington
State (Novotny 1978) and British Columbia (Evelyn 1971). Though several Vibri spp.
have been incriminated as agents of disease in cultured fishes, only three species are well
documented pathogens of salmonid fishes; Y. anguillarum,.Y. ordalii and V. salmonicida,
(Egidius 1987). Only the former two species have been reported in fish from Washington
(Novotny 1978). Vibri anoillarum is ubiquitous in the marine environment (Muroga
et al. 1986) and can survive without a fish host for several months (Toranzo et al. 1982).
Therefore, V. anguillarum is considered a facultative pathogen and does not require a
fish host to survive in the marine environment (Muroga et al. 1986). Whereas V.
angLifflaruin. infects over 40 fish species and has been isolated from wild as well as
cultured fishes, V. ordalii has only been isolated from pen-reared salmon.

A newly identified Vibrio sp., Y. salmonicida is the causative agent of Hitra disease in
net-pen reared Atlantic salmon in Europe (Egidius et al. 1986; Wiik and Egidius 1986).
Vibrio salmonicida has not been detected in salmon reared in North America. No
Vibri spp. pathogenic to man have been associated with disease in salmon and human
health concerns with 'Vibrio spp. have been restricted to warm water aquaculture (Egidius
1987). Unlike R. salmoninarum, Vibrio infections usually occur only after seawater
transfer. Vibriosis is usually an acute systemic disease and fish which recover show
strong immunological protection against reinfection. Effective vaccines are commercially
available which protect fish from V.'angpillarum and V. ordalii infections and the disease
is prevented by vaccinating fish prior to seawater introduction.

Viruses. Though several viral diseases are important in salmon during their fresbwater
phase of development, none have been reported from salmon in seawater. Infectious
hematopoietic necrosis (IHN) is of most concern in the Pacific Northwest. This virus is
a persistent problem in fry and fingerling chinook salmon at several hatcheries.
Apparently only fingerlings and returning salmon in freshwater are infected with IHN and
it has not been isolated from fish in seawater. In vitro studies by Pietsch et al. (1977)
and Toranzo et al. (1982) indicate that the virus survives poorly in seawater.

Idiopathic Diseases. Kent and Elston (1987a) observed a condition similar to pancreas
disease in Atlantic salmon reared in Washington. 'Ihis disease has previously been
described in Atlantic salmon reared in Europe during their first year in seawater
(Munro et al. 1984; McVicar 1987). Fish become emaciated and histological examination
reveals diffuse necrosis and atrophy of the exocrine pancreas. The cause is unknown.
Researchers in Scotland have proposed various etiologies; Ferguson et al. (1986)
suggested that the condition may be related to vitamin E and selenium deficiencies,
whereas Munro et al. (1984) reported epizootiological evidence consistent with an
infectious etiology. If the cause of this disease is an infectious agent, it is of marine
origin with no link to freshwater or stock origin (McVicar 1987). An infectious etiology
is also indicated in a study by Ferguson (1986). Fish from the same egg lot were
transferred to two sites. Fish at the site where the disease was enzootic developed the
disease while fish transferred to a site with no history of the disease remained unaffected.

APPENDIX E

THE ECONOMICS OF SALMON FARMING

THE ECONOMICS OF SALMON FARMING:

ROBERT L. STOKES

REPORT TO THE WASHINGTON STATE
DEPARTMENT OF FISHERIES

OCTOBER 1988

EXECUTIVE SUMMARY

This report examines three economic issues arising from recent growth in Washington's
salmon farming industry. The first issue is potential gains in output, income and
employment to the economies of the state and of selected counties. The second is
impact on revenues and expenditures of state government. The third is implications for
real estate values of various (externally provided) assumptions concerning visual impacts
of salmon fanning facilities. The report concludes with a benefit-cost analysis of
hypothetical siting decisions.

The report examined neither the universe of policy issues elsewhere addressed in the EIS,
nor the subset of those issues amenable to economic analysis or comment. Hence, the
reader is referred other sections of the EIS for discussion of the effects of environmental
wasteloadings and fish disease; consequences for sport fishing, and marine recreation; and
economic effects of public perception concerning environmental quality. An article by
James A. Crutchfield (Appendix L) also provides a useftil overview of the entire salmon
farming issue from an economic as well as policy perspective.

Washington's salmon farming industry is a segment of the world's rapidly growing
mariculture industry. After some years of concentration on Pacific salmon (pan size
coho), industry interest has shifted to production of mature Atlantic salmon. Several sites
have been established in the past two years and many more are in various stages of
planning or application for permits.

The combination of favorable water temperatures, sheltered waters and infrastructure
make Washington's Puget Sound one of the prime sites in the U.S. for salmon farming.
Current operators compete favorably with Norwegian and Scottish producers for U.S.
markets, and industry leaders feel that the combination of domestic and Japanese
demand provides a market base for extensive future growth. The following are this
authors general conclusions concerning the economic consequences for Washington state
(and selected counties) of permitting or encouraging expansion of the fish farming
industry:

The economic impacts of such growth were determined by assuming that a
representative Atlantic salmon facility (1,000,000 lbs production, $5,000,000
revenue) was sited in each of Clallam, Jefferson, Kitsap, San Juan, and Skagit
Counties. Conclusions of that analysis were that the state economy would gain
(from all 5 sites) $38-$48 million in output,.$11-$21 million in household income
and 257-303 jobs. Average County impacts for a single site were output $5.8-
$6.8 million, household income $1.1-$2.1 million and 40-51 jobs.

� These economic impact results provided the basis for estimates of state fiscal
(revenue and expenditure) consequences. Depending on the economic impact
values used and method of relating economic impacts to fiscal consequences,
salmon farming would contribute $.36-$2.26 million to state revenues and $1.08-
$1.48 to state expenditures.

� Property values were examined by collecting and statistically analyzing 335 current
real estate listings and assessed property valuations. T'he average front footage
price of $409 had a standard deviation of $290, approximately half of which could
be accounted for by general location (County), land type (high/low bank), and
improvements (water, sewer, etc.). The remaining or "residual" price variation was
presumed to result, at least in part, from variations in visual aesthetic quality.

� Finally, benefit-cost and sensitivity analyses were used to relate gross economic
gains (household income) to potential losses (adverse property consequences).
Sixty-four benefit-cost ratios were calculated to reflect all combinations to data
input ranges and necessary subjective judgments, the latter including opportunity
costs of labor, interest rates, degree and geographic extent of adverse visual
impact, and interest rates. All ratios (including those most unfavorable to the fish
farming industry) exceed unity, suggesting net statewide economic gains from
salmon farming. Average r esults for all sensitivity calculations and results
calculated under assumptions favorable to the industry indicated substantial net
economic gains.

These results should, of course, be interpreted in terms of the limited scope of the study.
They suggest favorable balances between the benefits and costs calculated in the course
of accomplishing the three study tasks; regional input-output analysis, state fiscal analysis,
and property value analysis. Also, as noted, each result depends on assumed rather than
estimated adverse visual effects on property. Overall assessment of the economic
consequences of salmon farming for Washington must also include consideration of
numerous other economic issues, including, in particular, the economic implications of
issues addressed from an environmental or policy perspective in the main body of this
EIS.

TABLE OF CONTENTS

SECTION PAGE

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

INTRODUCTION 1

I. THE SALMON FARMING INDUSTRY 2

II. INPUT - OUTPUT THEORY 5

III. THE REPRESENTATIVE SALMON FARM 13

IV. REGIONAL ECONOMIC IMPACTS 18

V. STATE AND LOCAL FISCAL IMPACTS 28

VI. PROPERTY VALUES 40

VII. BENEFIT COST ANALYSIS 53

NOTES 61

REFERENCES 62

APPENDIX 1. PROPERTY VALUE SURVEY 63

LIST OF TABLES

NUMBER PAGE

------- ---

2.1 Clallam County Transaction Table 8

2.2 Clallam County Technical Coefficients 9

2.3 Clallam County Economic Impact 10
Illustration

2.4 Clallam County Impact Illustration 11
First Round Calculation

3.1 The Representative Atlantic Salmon 15
Farm: Revenues and Costs

3.2 The Representative Atlantic,Salmon 16
Farm: Regional Distribution of
Expenditure

4.1 Summary of Impacts 19

4.2 Detailed Impacts: 5,000,000 lb Atlantic 20
Salmon Farm: Washington State

4.3 Detailed Impacts: 1,000,000 lb Atlantic 21
Salmon Farm: Clallam County

4.4 Detailed Results: 1,000,000 lb Atlantic 22
Salmon Farm: Jefferson County

4.5 Detailed Results: 1,000,000 lb Atlantic 23
Salmon Farm: Kitsap County

4.6 Detailed Results: 1,000,000 lb Atlantic 24
Salmon Farm: San Juan County

4.7 Detailed Results: 1,000,000 lb Atlantic 25
Salmon Farm: Skagit County

4.8 Average County Impacts 27

5.1 Distribution of 1982 Washington State 29
Revenues and Expenditures

5.2 Statistics Used to Relate Regional 33
Economic Impacts to Washington State
Revenues and Expenditures

5.3 Washington State Economic and Fiscal Data 34
1970 - 1985

5.4 Algorithm for Calculating State Revenues 35
and Expenditures From Regional Economic
Impacts

5.5 Impacts of a $25,000,000 Salmon Farming 38
Industry on Washington State Revenues
and Expenditures

6.1 Puget Sound Waterfront Property 42
Front Footage Values

6.2 Puget Sound Waterfront Prop erty 43

6.3 Clallam County Waterfront Property 44

6.4 Jefferson County Waterfront Property 45

6.5 Kitsap County Waterfront Property 46

6.6 San Juan County Waterfront Property 47

6.7 Skagit County Waterfront Property 48

6.8 Regression Analysis of Puget Sound 51
Waterfront Property Values

7.1 Input Parameters to Sensitivity 54
Analysis

7.2 Sensitivity Analysis Algorithm and 59
Illustrative Calculation

7.3 Sensitivity Analysis Results 60

A.1 Puget Sound Waterfront Property Survey 63

LIST OF FIGURES

NUMBER PAGE

------- ----

2.1 Schematic Representation of 6
Input - Output Analysis

5.1 Schematic Description of Fiscal 30
Impact Calculation

INTRODUCTION

This report, prepared under contract with the Washington

State Department of Fisheries, examines selected economic aspects

of the Washington salmon farming industry. The stimulus for this

study is a general review of siting policy currently being

conducted by the Washington Department of Fisheries and other

state agencies.

Several of the issues referred to in recent salmon farm

siting decisions have been economic in nature. These include

potential gains in state and county income and employment, fiscal

impacts on state and local governments, and adverse consequences

for waterfront land values in the vicinity of proposed sites.

Each of these issues is addressed below. A brief history

and description of the salmon farming industry is provided in

Section I, followed by a discussion of regional input-output

theory (Section II) and discussion of the economic data used to

characterize salmon farming in input-output terms (Section III).

Section IV reports input-output results (state/county employment,

income, etc.) which are applied to the calculation of state

fiscal impacts in Section V. Section VI deals property values and

Section VII assembles the preceding economic data into a benefit-

cost model. sensitivity analysis.

I. THE SALMON FARMING INDUSTRY

Salmon farming is emerging in Washington State as part of a

rapidly growing world mariculture industry. Mariculture is the

cultivation of marine organisms for harvest, as distinguished

from capture fisheries, the harvest of naturally occurring fish,

and aquaculture (the cultivation of freshwater organisms).

Historically, mariculture has been a major world producer of

shellfish and finfish. Washington State has long participated in

mariculture as one of the primary US oyster producers.

Much of the science and technology which now underpins

private salmon mariculture was initially developed to support

Pacific Northwest public hatchery programs. In the past several

decades research supporting those hatchery programs has focused

on all five species of Pacific salmon, as well as on Atlantic

salmon. Early research on the hatchery production of Atlantic

salmon was done at the University of Washington US and Norwegian

biologists working in cooperation.

Transfer of Pacific salmon production technology to the

private sector was initiated by, among others, the Weyerhauser

Corporation in Oregon and Domsea Farms (Cambell Soup Co.) in

Washington. The first private production of Atlantic salmon

began in Norway in 1971, leading to a booming Atlantic salmon

farming industry in that nation, and later in Scotland as well.

In the Pacific Northwest, British Columbia has seen the

greatest growth in salmon farming. The first major British

Colombia site was licensed in 1973, with major production

beginning in about 1985. Today (1988) 128 sites produce 900

metric tons of chinook, coho, and Atlantic salmon, as well as
1
rainbow trout and arctic char.

After many years of relatively stable production (1.5 to

2.0 million lbs from 1979 to 1986) Washington State has become

the scene of increasing interest in Atlantic salmon pen culture.

There are now 15 sites operating in Puget Sound and North Puget

Sound Counties, as well as 17 in the permit cycle. Total 1987

production of 3.4 million pounds consisted primarily of coho, but
2
included 400 thousand pounds of Atlantic salmon.

Puget Sound is the largest potential salmon farming site in

the United States because of its desirable water temperature,

sheltered waters, and good economic infrastructure. Salmon

farms currently in production compete successfully with Norwegian

and Scottish imports on the US West Coast and in the Midwest.

Industry spokesmen suggest that production could expand

significanlty without saturating potential markets, particularly

if, for Japanese markets, Washington's geographic advantage over

Europe can be exploited.

II. INPUT - OUTPUT THEORY

Input-output theory is a widely used method of regional

economic analysis, appropriate to the issues addressed in this

report. Developed as a post World War II extension of Keynsian

national income analysis, input-output theory shares the Keynsian

assumption of an economy with slack producing capacity. In such

an economy output is, over moderate variations, determined by

aggregate demand. This characterization fits the situation of

flopen" regional economies, such as states and counties. In such

open economies, labor and other inputs are available at

relatively constant prices, and are employed in relatively fixed

proportions to produce goods and services demanded by regional

and external consumers (Richardson, 1972).

Typical evaluation applications of input-output theory

include determinations of the regional output, income and

employment implications of specific industrial facilities siting

decisions. Planning applications have included studies

anticipating the public infrastructure requirements of economic

growth, and more recently, studies assessing regional energy

requirements and environmental waste loads.

The general procedures by which input-output analysis

extracts estimates of regional output, income and employment from

externally provided estimates of final demand are traced in

figure 2.1. Economic information on the entire regional economy

Figure 2.1 Schematic Representation of Input - Output Analysis

TRANSACTIONS TABLE APPLICATION DATA
(REGIONAL ECONOMY) (SA.11-ION FAR14ING)
f

TECIRNICAL
COEFFICIENTS

APPLICATION
INPUT - OUTPUT SPECIFIC
CALCULATION COEFFICIENTS

STANDARD APPLICATION
INPUT - OUTPUT SPECIFIC
RESULTS RESULTS

(usually obtained from background studies) is recorded in a

transactions table, from which a table a technical coefficients

is computed. Data on the industry or facility under analysis is

then collected (as part of the specific application) and

introduced as a column vector of final demands. These final

demands are aggregated into the same catagories as those used in

the transactions table. Matrix multiplication of the final demand

vector and the technical coefficients table (and side

calculations as necessary), produce the desired results.

These procedures conform to specific accounting principles

and technological assumptions, which are illustrated with the

Clallam County input-output model. That model and structurally

identical ones for Washington State; and for Jefferson, Kitsap,

San Juan and Skagit Counties were developed for this study from

data found in the US Forest Service IMPLAN system. These models

necessarily reflect 1982 economic conditions. This is because

1982 is the effective year of IMPLAN, and is also the effective

year of the most recent complete census of manufactures. The

reader interested in the detailed mechanics of the input-output

procedures employed here is invited to trace the calculations

reported in tables 2.1 to 2.4 (U.S. Forest Service, 1988).

For the more general reader, we offer the following

observations on the appropriatness of input-ouput analysis to the

task a hand. An assumption crucial to the application of input-

output analysis is that the inputs and outputs of regional

industries can be varied in constant proportions, without

altering prices, encountering physical resource constraints, or

Tablc- 2.1 Clallam County Transac-tion Table
(thousand dollars;)

1 RgriculturQ, Fisheries. Fisheries 7 Retail Trade PH Value Added
2 Mi ni n-j 8 Finance, Inseirancr-. RQal E:.3t-a+.Q IMF I mplrt5
3 Construction 9 Servicer. PCE Per-@@;onal Consumption
I Manufacturing 10 FcedQr-al Gov4rnmont FD Final Demand
5 Transport, Comm, Public Utilities 11 State and Local Government r Total
6 Wholezalo Trade

1 2 :3 14 5 6 7 a 9 1.0 11 PGE' FD T

1 16 1 0 61 7?6 1 3, 1 2-39 3-113 1 2 ?95 5137.1 86C.0
2 0 0 1 7 0 0 ri 0 0 0 0 0 52 6--)
3 99 0 64 4403 1619 0 111 160.1 166.3 1 17F12 0 29650 -1-4322
1 56 0 200-3 57097 2 1 12 15 1.478 3136 1692 2 14.4 1639 311085 3?571S
5 221 2 993 851.1 7WIS 18 1'?8,3 586 9 400 0132 2461.4 57371
6 1 0 9 .49 3 0 1 a 15 0 0 35 1U6? 1181
? 23 0 1.930 G-qI 145 73 113 291 1 211 59-365
a is? 1 2.16 1310 555 1F1 1501 355A 29-45 3 31 3560,:; 5821? 10l 1?2
9 ill 2 Il -3 1 5690 V20 ?? 17,63 1512 6276 21 115 -5775--1 02031 1385 11
10 0 a 6 1.3 1? 0 3 190 11 12 1 1-15 :35? 756
11 11 0 -17 1012 255 21 111) 131 4 1 1 IS 1509 611,1914 68022
VA 3250 -jq 16616 12271130 2 -4 7 --". ? 786 43037 8IB22 ?3261 .458 63392 0 791? 138 110
IMP 1215 22 21053 173:399 20 I.-JO 2 -")'- q 9169 11525 4N.72. 2411 1608 -:013995 0 620261
00 T OF,60 62 .41'1;22 375?19 573111 1181 S9365 101172 138511 756 68022 130110 620261 1916511

Tablo 2.2 Clallan Courity Ter-hroical CoQfficients-
(thousand dollars)

I ftriculture, Fisheries, Fishr-ri,-:!-, ? Rrtaii rrad-;, VA Value fidded
2 Hi rd rig E, Firianci-,, Insurance, RQ,-%l Est.-Ai- PCE Pqrsonal Ccjr,5
3 Con5truction '3 Svr,.,i ces T Total
I Manufacturing 10 Fodcral GovrerrsmQrit
5 Trarcsport, Comm, Public Utilitie@, 11 St.&to and Local Goverrihh-nt
6 Wholozalc- TradQ

1 2 11 S 6 7 8 9 11 PCE

1 0.09--112 0.0000 0.001-1 0.13021 0 0 fir) 0.0009 FJOOO 0.002'.) 0.0025 0. Orl 11) .0000 0.1-10111:
2 0.0000 0.0021 0.0001 .00130 1).0000 0.0000 ff. 13000 O-OIjuO 0-0000 U. 01:101.) 0-01300 0.1-10011
3 0.0102 0.130M OXON 0.0117 fs.02!@-2 LI.00?1 A. 00711 0. 01-12 0.0120 O.OrJl;' - C1.02P.2 0.0000
-1 0.0065 0.0012 0. 01152 0.1520 0 . Oil) 3 V 0 . 0 1-,@; I I'l. 1)2-19 0.0038 0.0122 0. 0027 0.0006 cl. 003?
5 0.0258 (1.0-322 0.022-9 0.0227 0. 0.0-10-4 0.0301 0 - 0056 0.02:38 0. 0 116 0.0()Tn 0. 1:1 18b:
6 0.0001 0.0000 0.0002 0.0001 U. NO 1 0.0001 .0000 .0000 0.0001 0. ON I -01.11-A) I-1.1-joul
? 0.0026 It. 00 0 0.0-113 0. 00 17 fi. 0025 CAU15 0. OU 12 0.0011 0. 001-@ 1 0. CIO I S 0 . u 1) 0 '-;1 cl. 0189
a 0.0216 0.0180 0.013-49 0.0036 Ij. 009? cl. f) 1118 0.0253 0.0,1111 0.13213 0.00-15 0 . 0005 0. 138 I-J
9 0.0132 0. 0-3 19 0 .032-3 0.0151 0. 0300 0. 0652 0.0297 0.01.15 0. 0153 0 . 0'--' Is!:-, . 1) 0 1 -1., 0.0862
10 .0000 0.0000 fi. fifio I . 0000 0.0003 0.0002 0. 000 1 0.0018 0-0001 0.0161 .0000 0.000-@:I
11 A. 00 2 1. 0.0006 0.002? 1). 00-14 0.0016 1-1. 00 18 0.0013 0.0025 0.0010 U-000? 0.0036
1) A 0.3753 0. 55 11 0.3756 0 .'--;2U'I 0. 1 10 0.6659 0.7250 0. ?1307 0.5289 0 - 6058 0.939---i 0.0000
r 0.5099 0.619-4 0.5255 0 . 13.6-491 0. 9 IN 1-1.8-155 0.8991 0.6558 0. 6$11G 0. 9?F--q 0.21115

Table 2.3 Clallas CGuntq Econoiiic Impact illustration
(thousand dollars)

I Agriculture, Fisheries. Fishories- ? RG-tail Trade
2 Mining 8 Financ-re, Insurance. Real Estatq
3 Constructi on 'I Srervi cros
I Manufacturing 10 Fodcoral Government
5 Transport, Comm, Public Utilities 11 State and Local Government
6 Wholasalcp Trade VFI Valuo Rdded
T Trital

ROUND ------------------------- SECTOP ------------ -----------------------------------------------------------------------------------
1 2 3 14 5 6 7 a 9 10 1.1 VFI T

1 S20.16 S.00 S11.53 S19.31 S86.17 $0.25 $92-50, S162.?? S 180. 53 SO.72 S-9. 19 5368. 15 S98-4.15
2 $2.69 S.00 $13.52 $15.11 S29. 79 $0. 07 ;@ 19. 613' $13. 17 $49.13 SO.48 -';'@2. ? 1 $365. 0? $511.38
3 $1.08 S.00 $3.79 $5.66 $13.92 SO.ClIq 11w 1:3. 69 S3-1.15 $36.51 S 0 . L'2 $1.72 $1013.67 $2 15. -qE-
1 $0.413 S.00 $2.56 52.50 $5. 87 110 . 0 1 S5.26 10 . ?1, $12.02 SO.11 $0.61 $?Q.20 S1.10.3-4
5 $0.21 S.00 $0.98 S1.10 $2.82 $0. rI 1 $3.60 $6.55 $7.21 SO.05 SO.i;lq $23.69 $16. 'q?
6 SQ.09 S.00 $0.51 S0.51 $1.2-1 SAW 1. Z3 S2.43 S2.71 SO.02 &-0.11 S 13. Be S22. 76
7 SO.01 S.00 $0.20 50.23 SU. 513 $.CIO $0.71 $1.32 $1.15 $0.01 SO.O? $5. 29 $9.92
a SO.02 S.00 410.10 SO.11 50.26 $.CIO SQ.27 $0.53 SCI.59 S.00 $0.03 $2.81 S4.73
9 $0.01 $.Do $0.01 Sri .05 $0.12 5.00 $0. 11 $0.27 50.30 5.00 $0.01 $1.15 S2.10
10 S.00 S.00 $0. L12 50.02 So. 01:1 S.00 S13-06 SO . 11 $0.13 5.01) S0.01 S0.58 $0.99
T S24.75 $0.01 S-46.14 $74.60 5110.82 $0.39 S 11`12. 19 $261.08 $290.38 $1.62 $1.4.84 $951.49 51,938-30

Table 2.1 Clalla" County Ecr-%nomic Impact Illustratioki
First kaund Calculation

Fi r-st
Round > 300 0 0 250 0
Iriput
1 300 X 0.0532 # 0 X 0. 0000 + 0 X 0 . 0 Fj 1 -1 + 2 5 0 X 0.0021 + 251.) X@ 0.0000 + 0 X 0.0009
2 300 X 0.00011 + 0 X 0.0021. + 0 X 0. F100 I + 251.1 X 0.0000 + 2513 X 0.0000 + 0 X 1.) - OCICIO
3 300 X 0.0102 + 0 X 0.0076 + 0 X 0.0011 + 2 5 fj X 0.0117 + 251) X 0.0282 + 0 X 0.0071
.q 300 X O.OCIE65 + 0 X 0.0042 + 0 X. 0.0452 4 250 Y 0. 1520 + 25 fj 'A 0.00-57 + 0 'X, 0.0131
5 300 X 0.0250 + 0 X 0.0322 + 0 X 0 . 0'--, 2 1 + 250 X 0.022? + 250 X 0 - 1392 + 0 X 0.0-40-1
6 300 X 0.0001 + 0 x ri.13000 + 0 '-.' 0.0002 + 2 5 C, X 0.0001 + 250 '.-: 0 . 0 0 r) 1 + 0 X 0.0001
7 300 X 0.0026 + 0 X 0.000-S + 0 X 0. 011 @@ + 250 x O.0Ul? + 250 X 0-0025 + 0 w. 0. 00 15
8 300 x 0.02,16 + 0 x 0.0180 + 0 X 0.0019 + 250 X 0. 003.6 + @51) .1.0 0.009? + 0 X
9 300 x 0.0132 + 0 X 0.0319 + 0 X 0. 0`23 + 250 X 0. 0 151 + 250 X 0.0300 + 0 X 0 - OF, 5,2
1.0 3110 X 0. 0000 + 0 .4 0.0000 + U X 0.01-101 -+ 250 X 0.01)(10 + 250 '.-: 0.0003 + 0 X O-OC102
3.1 300 X 0. 00 1:3 + 0 X 0.0021. + 0 X 0. 0006 + 250 X 0.0027 + 250 X + 0 X 0. 00
VF1 300 X 0. 3?5.-j + 0 X rl. 55 11 + 0 X 0. -j?56 + 250 9 0.3268 4 250 X 0.1310 * 0 'X. U. 6E.513

Sector 7 8 1 L) 11 PCE Fir:st Round Ou
SQcrind Round I
Fi rst
Round > 0 0 125 0 11 1050.
Input
I. + 0 X 0.0000 f 0 X ri.002.-:i 125X 0. 0025 + 0 1%, 0.0019 + A X O.OOCIU 4 105.4 0.0018 = 20.16
2.. + 0 X 0.0000 + 0 X i 125X 0.0000 + 0 X 0.0000 + 0 0.00130 + 105X 0. 0000 = .00
3: + 0 X 0.0071 + 0 X C1.0-142 + 0.0120 + 0 X 0.0012 + 0 X 0.0262 + 185 X 0.0001) = 1-1.53
11, + 0 X 0. 0219 4 0 X C1.0030 f 1@*SX 0.0122 + 13 x 0.002? + rl X 0. 01106 + 185X 0.0037 = 19.31
5: + 0 X 0.0@01 + 0 X 0.005E. + 125X U.0208 + 0 N 0.0116 + 0 V 0.00?0 + 185X 0.0186 = 06. 17
E'. + 0 X 0.0001) + 0 X 0. 0000 + 1215X 0.0001 + 0 X 0.0001 + 0 X 0. 0000 + 18SX 0.0001 =
(1: + 0 X 0.0012 + 0 'A C1.0011 + 12SIX 0.0021 + 0 X 0.0015 + 0 Y, 0.013113 + 185@0 0. 0189 = 92.58
R .1 + 0 X 0.0253 + 0 X 0. 03111. + 125X 0.0213 + 0 X 0.00115 1 C, X l).U005 + 185X 0. 00 1:1 = 162'. 7?
4: + 0 X 0.029? + 0 x 0.0115 + 12sx 0.0-153 + U x U - 0355 + Cl x 0.001? + 0.0,162 180.33
10 -. + 0 9 0.0001 + 0 x 0.0010 + 12sx 0.0001 + 0 *% 0.010 + Cl x 0. 0000 4 0. 0003 0.?2
11. + 0 X 0 . 0 ri 181 + 0 X 0. 00 1.-:1 + 125X 0.0025 + 0 X 0.0010 + 0 X 0.000? + IOSX 0. 0036 9.11.)
VA. + 0 X 0.7251) + 0 X O.?80T + 125X 0.5289 + 0 X 0.6058 + 0 x 0.9391, + 105X 0.0000 368.15

otherwise changing base period economic and technological

conditions. These conditions are reasonably well satisfied in

this application. This because the current and proposed salmon

farming industry is small in relation to the economies of

Washington State and its coastal counties. One exception to

this expectation of constant technical coefficients is the

possibility that a growing salmon farming industry might lead to

import substitution. That is, instate suppliers may avail

themselves of opportunities to provide specialized inputs that

are now imported. This possibility is addressed by means of

sensitivity analysis (maximum and minimum impact calculations)

described in sections III and IV.

Another important assumption is that the accounting stance

of relevant decisionmakers coincides with the scope of the input-

output model. Such a coincidence would seem to exist for this

evaluation of state and county siting policy. Even within a

regional accounting stance, though, input-output results cannot,

without modification, be interpreted as net benefits in the

benefit-cost sense. This qualification is addressed in section

VII, where input-output results are given a net benefits

interpretation, but only after subtraction of opportunity costs.

III. THE REPRESENTATIVE SALMON FARM

To analyse the regional economic impacts of an activity like

salmon farming requires that it first be described in terms of

total revenue/expenditure and the allocation of each among sec-

tors of the regional economy. Of greatest importance to accurate

impact assessment is the division of revenues between export and

local sales, and of expenditures and net incomes between regional

and non regional recipients.

This step in the analysis was accomplished by interviews

with several salmon farming industry participants. These

individuals provided both information on their own operations, as

well as a general overview of what, in their view, the future

salmon farming industry would look like.

Production and financial profiles of several salmon farms

were obtained from their owners or managers. In some cases

intervewes provided exact revenues or expenditures, and in other

cases general planning factors, "Feed is $.90 per finished

pound.", "Sales cost is 5% of gross revenue.", etc. Confiden-

tiality commitments preclude identifying the firms supplying this

data, or given the fewness of firms, publication of specific

data in even masked form. Instead, the revenue/cost profile of a

representative operation (table 3.1) was built from the data

provided. By design, this profile describes the industry in

general, but no firm in particular. In addition, while estimates

of profit are sufficiently accurate for regional economic

analysis, they should not be considered reliable for investment

planning or other purposes.

Industry interviews also suggested that production will

include pan size and mature coho, chinook and Atlantic salmon,

but that mature (about 9 lb round weight) Atlantic salmon will

gain in significance over time. Hence, the profile described in

table 3.1 describes the production, revenues and expenditures of

a 1,000,000 finished pound Atlantic salmon production facility.

This production is assumed to sell entirely into out of state

markets at $5.00 per pound. The disbursement of the resulting

$5,000,000 in annual revenue is allocated among phases of

production (hatchery, fish farm, administration) and among input

types (feed, labor, etc).

The extent to which these inputs4'are supplied locally or

from outside the region is the primary determinant of the

facilities indirect and induced regional economic impact. The

local/import supply factors used here are reported in table 3.2.

For some inputs the pattern of regional/import supply was either

obvious, or could be reliably determined by interview. Labor is

necessarily supplied from within the state and, for the most

part, the subject county. Processing of fresh round fish is most

conveniently done nearby, certainly in Washington and most likely

in the subject county. Packaging materials, freight, and

brokerage services for fish shipped out of Seattle will most

likely by supplied by in-state, but not in-county firms.

Sources of other inputs are less certain, and may vary over

Table 3.1 Representative Atlantic Salmon Farm
Revenues and Costs (thousand dollars)

Production and Revenue

Finished pounds 1,000,000
Price $5.00
Gross revenue $5,000,000

Expenditures
------------------------------------------
Item Amount $/lb

---------------------- ------- -------
Hatchery: Eggs $300,000 $0.30
Hatchery: Labor $250,000 $0.25
Hatchery: Other $50,000 $0.05
Hatchery: Employment 8
Fish Farm: Labor $500,000 $0.50
Fish Farm: Other $450,000 $0.45
Fish Farm Employment 20
Feed $900,000 $0.90
Processing $250,000 $0.25
Packaging $100,000 $0.10
Freight $250,000 $0.25
Brokerage $250,000 $0.25
General Administration $250,000 $0.25
Administrative Employment 5
Debt Service $350,000 $0.35
Equity Return $1,100,000 $1.10
TOTAL $5,000,000 $5.00

Table 3.2 Representative Atlantic Salmon Farm
Regional Distribution of Expenditures

In County In State
----------------- -----------------
Max Min Max Min

-------- -------- -------- --------
Hatchery: Eggs 100% 0% 100% 50%
Hatchery: Labor 100% 100% 100% 100%
Hatchery: Other 50% 25% 75% 50%
Fish Farm: Labor 100% 100% 100% 100%
Fish Farm: Other 50% 25% 75% 50%
Feed 0% 0% 100% 50%
Processing 100% 100% 100% 100%
Packaging 0% 0% 100% 100%
Freight 0% 0% 100% 100%
Brokerage 0% 0% 100% 100%
General Administration 50% 25% 75% 50%
Debt Service 0% 0% 100% 0%
Equity Return 100% 0% 100% 0%

time with the size of the industry, and with other factors. The

following maximum and minimum local input factors were adopted

for these inputs. Currently, the preponderance of feed is

provided by Moore Clark of LaConner (Skagit County) Washington,

with some feed imported from Oregon. Washington feed supply is

set at 100 % to 50 %, a factor which will depend, among other

things, on the future competitive position of Washington

suppliers. Egg costs are set at either 100 % Washington and 100 %

county (local production) or 50 % Washington (external purchase).

other expenditures and administrative costs are set at 75% to 50%

Washington and 50 % to 25 % county supply. Debt service is set

at 100 % and 0 % Washington (in or out of state financing), and

equity return is set at 100 % Washington 100 % county (local

ownership) or 0 % Washington 0 % county (out of state ownership).

IV. REGIONAL ECONOMIC IMPACTS

Regional Economic impacts are computed by allocating the

expenditures of the representative salmon farm to appropriate

model catagories, and then performing the calculations described

in tables 2.3 and 2.4. All results vary depending on whether

minimum or maximum state/local supply assumptions are used.

County results vary further (although only slightly) as a result

of each counties different economic structure. Results are

reported in summary form in table 4.1, and in more detail in

tables 4.2 (statewide), table 4.3 (Clallam Co.), table 4.4

(Jefferson Co.), table 4.5 (Kitsap Co.). Table 4.6 (San Juan Co.)

and table 4.7 (Skagit Co.).

Statewide impacts were based on an industry expansion equal

to 5 of the representative salmon farms discussed in the

preceeding section. County impacts were based on one such

facility in each county. Because of the liniar nature of input-

output analysis, the computation of impacts for other industry

sizes (combination of facilities) can be accomplished by simple

multiplication of these results. That is 2 operations in a county

would have exactly twice the county impact of one, and 10

facilities would have twice the state impact of 5.

As described in table 4.1, a 5 million pound Atlantic salmon

farming industry (5 representative facilities) would contribute

between $38 and $48 million to state output, between $11 and

Table 4.1: Summary of Impacts: 5,000,000 lb Atlantic
Salmon Industry in Washington, and Average Results
for a 1,000,000 lb Facility in Clallam, Jefferson,
Kitsap, San Juan or Skagit County ($ thousands)

OUTPUT INCOME EMPLOYMENT

Maximum
State $48,395 $21,412 303
County $6,812 $2,748 51

Minimum
State $38,227 $10,615 257
County $5,775 $1,132 40

Table 4.2: Detailed Impacts: 5,000,000 lb Atlantic
Salmon Production: Washington State (thousand dollars)
Direct Total

------------------ ---------------
Local Supply Sector Maximum Minimum Maximum Minimum
--------------------------- -------- -------- ------ ------
Agriculture, Forestry, Fisheries $1,500 $750 $2,362 $1,255
Mining $0 $0 $42 $30
Construction $0 $0 $476 $274
Manufacturing $6,250 $4,000 $9,011 $5,682
Transport, Comm,.Utilities $3,125 $2,500 $4,897 $3,616
Wholesale Trade $1,250 $1,250 $1,299 $1,278
Retail Trade $0 $0 $682 $342
Finance, Insurance, Real Estate $1,750 $0 $3,968 $1,083
Services $938 $625 $3,202 $1,869
Federal Government $0 $0 $22 $10
State and Local Government $0 $0 $86 $48
Housholds $9,250 $3,750 $21,412 $10,615
Total Output $25,000 $25,000 $48,395 $38,227
Employment 161 161 303 257

Table 4.3: Detailed Results: 1,000,000 lb Atlantic
Salmon Farm: Clallam County (thousand dollars)
Direct Total

------------------ ---------------
Local Supply Sector Maximum Minimum Maximum Minimum
--------------------------- -------- -------- ------ ------
Agriculture, Forestry, Fisheries $300 $0 $325 $4
Mining $0 $0
Construction $36 $17
Manufacturing $250 $250 $325 $307
Transport, Comm, Utilities $250 $125 $391 $188
Wholesale Trade $0 $0 $0 $0
Retail Trade $142 $59
Finance, Insurance, Real Estate $0 $0 $261 $107
Services $125 $63 $415 $186
Federal Government $2 $1
State and Local Government $15 $7
Housholds $1,850 $750 $2,801 $1,171
Total Output $5,000 $5,000 $6,937 $5,858
Employment 32 32 52 41

Table 4.4: Detailed Results: 1,000,000 lb Atlantic
Salmon Farm: Jefferson County (thousand dollars)
Direct Total

------------------- ---------------
Local Supply Sector Maximum Minimum Maximum -Minimum
........................... -------- -------- ------ ------
Agriculture, Forestry, Fisheries $300 $0 $321 $6
Mining $0 $0
Construction $45 $16
Manufacturing $250 $250 $291 $281
Transport, Comm, Utilities $250 $125 $348 $172
Wholesale Trade $0 $0 $3 $1
Retail Trade $140 $58
Finance, Insurance, Real Estate $0 $0 $230 $95
Services $125 $63 $369 $166
Federal Government $1 $1
State and Local Government $16 $7
Housholds $1,850 $750 $2,786 $1,155
Total Output $5,000 $5,000 $6,776 $5,770
Employment 32 32 51 40

Table 4.5: Detailed Results: 1,000,000 lb Atlantic
Salmon Farm: Kitsap County (thousand dollars)
Direct Total

------------------ ---------------
Local Supply Sector Maximum Minimum Maximum Minimum
--------------------------- -------- -------- ------ ------
Agriculture, Forestry, Fisheries $300 $0 $310 $5
Mining $0 $0
Construction $14 $3
Manufacturing $250 $250 $286 $280
Transport, Comm, Utilities $250 $125 $309 $154
Wholesale Trade $0 $0 $1 $0
Retail Trade $84 $35
Finance, Insurance, Real Estate $0 $0 $188 $79
Services $125 $63 $304 $139
Federal Government $16 $7
State and Local Government $9 $4
Housholds $1,850 $750 $2,615 $1, 080
Total Output $5,000 $5,000 $6,361 $5,598
Employment 32 32 42 36

Table 4.6: Detailed Results: 1,000,000 lb Atlantic
Salmon Farm: San Juan County (thousand dollars)
Direct Total

------------------ ---------------
Local Supply Sector Maximum Minimum Maximum Minimum
--------------------------- -------- -------- ------ ------
Agriculture, Forestry, Fisheries $300 @$o $343 $8
Mining $2 $2
Construction $45 $19
Manufacturing $250 $250 $313 $284
Transport, Comm, Utilities $250 $125 $459 $218
Wholesale Trade $0 $0 $29 $12
Retail Trade $544 $225
Finance, Insurance, Real Estate $0 $0 $310 $127
Services $125 $63 $254 $119
Federal Government $1 $1
State and Local Government $35 $14
Housholds $1,850 $750 $2,957 $1,225
Total Output $5,000 $5,000 $7,518 $6,068
Employment 32 32 67 46

Table 4.7: Detailed Results: 1,000,000 lb Atlantic
Salmon Farm: Skagit County (thousand dollars)
Direct Total

------------------ ---------------
Local Supply Sector Maximum Minimum Maximum Minimum
--------------------------- -------- -------- ------ ------
Agriculture, Forestry, Fisheries $300 $0 $345 $10
Mining $0 $0
Construction $24 $10
Manufacturing $250 $250 $403 $325
Transport, Comm, Utilities $250 $125 $353 $172
Wholesale Trade $0 $0 $3 $1
Retail Trade $76 $30
Finance, Insurance, Real Estate $0 $0 $155 $60
Services $125 $63 $292 $130
Federal Government $0 $0
State and Local Government $7 $3
Housholds $1,850 $750 $2,581 $1,027
Total Output $5,000 $5,000 $6,466 $5,581
Employment 32 32 44 37

$21 million to houshold incomes, and would create between 257 and

303 jobs statewide. Averaging results obtained from individual

county models suggests that a single 1 million pound facility

would contribute between $5.8 and $6.13 million to county output,

between $1.1 and $2.7 million to county houshold income, and

would create 40 to 51 jobs within the county

Table 4.8: 1,000,000 lb Atlantic Salmon Farm, Average
County Results (thousand dollars)

Total Output Housholds Employment
---------------- ---------------- ----------------
County Maximum Minimum Maximum Minimum Maximum Minimum

---------- -------- -------- ------ ------ ------ ------
Clallam $6,937 $5,858 $2,801 $1,171 52 41

Jefferson $6,361 $5,598 $2,615 $1,080 42 36

Kitsap $6,776 $5,770 $2,786 $1,155 51 40

San Juan $7,518 $6,068 $2,957 $1,225 67 46

Skagit $6,466 $5,581 $2,581 $1,027 44 37

Average $6,812 $5,775 $2,748 $1,132 $51 $40

V. STATE AND LOCAL FISCAL IMPACTS

The foregoing input-output anal, lts provide a basis
ysis resu

for, the comprehensive assessment of how an expanded salmon

farming industry would effect state revenues and expenditures.

The required extensions to the input-output model can be

identified by examining the major items of revenue and

expenditure reported in table 5.1.

There, we see that 71 % of state revenue arises from 4

sources, sales taxes, gross receipts taxes, property taxes, and

federal grants, with the remaining 29 % qonsisting of

miscellaneous taxes and revenues. Similarily, 71 % of

expenditures are for education (all levels) and human resources

(including welfare), with the remaining 29 percent going to other

catagories including general government. The general government

catagory, however, includes debt service and pension

expendituresf many of which could properly be allocated by

function to education and human services as well.

Thus, obtaining a fiscal analysis from input-output results

involves relating 5 revenue types and 3 functional expenditure
catagories to the economic changes described by"the above input-

output analysis. Figure 5.1 illustrates, in general, what must

be done to accomplish this. The first steps, input-output

analysis to produce impacts on output, houshold income and

employment, have already been accomplished. This section is

Table 5.1 Distribution of 1982 Washington State
Revenues and Expenditures

REVENUE

Revenue or
Item 1982 $000 Expenditure
-------------------- --------- ---------------

General and Selective $1,901 30%
Sales Taxes

Gross Receipts Taxes $635 10%

Property and In Lieu $628 10%
Taxes

Other Taxes $319 5%

Federal Grants $1,356 21%

Other Revenue $1,550 24%

Total Revenue $6,389

EXPENDITURE

Education $2,751 44%

Human Resources $1,647 27%

Other Expenditures $1,796 29%

Total Expenditures $6,194

Figure 5.1 Schematic Description of Fiscal Impact Calculation

INPUT - OUTPUT ANALYSIS

AVERAGE AND

EMPLOYMENT VALUE ADDED '11ARGINAL

COEFFICIENTS

REVENUES AND EYPERDITURFS

devoted to the remaining steps; examining possible relationships

between those economic variables and revenues/expenditures,

choosing coefficients for calculation from among the examined

possibilities, and, finally, using sets of chosen coefficients to

calculate ranges of fiscal impacts.

The underlying (though not readily observable) economic

processes that relate economic expansion (or contraction) to

state revenue and cost are postulated to be as follows. Increases

in gross output (with houshold income as a proxy variable)

increase gross receipts taxes. Increases in houshold income

increase consumption which, in turn, increase sales tax revenue.

To the extent that both induce increases in taxable assets held

by business and consumers, they increase property taxes as well.

Depending on the their basis of application, other taxes and

revenues also rise.

Expenditures and federal grants are assumed to respond to

economic change in a somewhat more complex way. Input-output

estimated increases in employment represent jobs filled by some

combination of immigrants and unemployed current residents.

Increased labor force participation has little, if any, effect on

educational and general state expenditures, and may actually

reduce human resource expenditures. By contrast, increased

population will increase all three expenditure catagories;

education due to the children accompanying new immigrants, human

resources due to increased welfare and other case loads, and

general expenditures for similiar reasons. Federal grants will

also increase, to the extent that federal funding formulas

include general or targetpopulations.

Figure 5.1 describes the calculations required to implement

the above theory of fiscal effect. Included among the standard

input-output results are estimates of employment and houshold

income (value added) resulting from the siting of salmon farms.

The relationship between personal income and revenue is direct;

changes in houshold income effect sales taxes, gross receipts

taxes, property taxes, and other taxes and revenues. The degree

of effect is determined by the estimates reported in tables 5.2

(as estimated from table-5.3 data). The algorithm used to make

those calculations is reported in table 5.4. The relationship of

employment to expenditures and federal grants has two links;

employment to population, and population to enrollment. Each of

these is also estimated in table 5.2 and reflected in the

algorithm in table 5.4.

Two methods of relating economic change to fiscal magnitudes

are reported in table 5.2. In both cases estimates are based on

1970-1985 data, with financial magnitudes expressed in 1982

prices. The first method is that of average ratios. For example,

sales tax revenue averaged $.0078 per dollar of personal income,

population averaged 2.43 per employee, school enrollment .42 per

capita, and state educational expenditure $2,945 per school

child. The second method is that of marginal change, based on

regression analysis. For example, the marginal change in sales

tax revenue with respect to personal income was $.0014, the

change in population with respect to employment was 1.52, and the

Table 5.2 Statistics Used to Relata- Regional Economic Impacts to Washington
Stato Revenues and Expenditures

Rvr@,ragv Coefficient (B)
Ind,?p9rid-ant Variable Ml Dopendant Variablfi, (Y) Ratio (Y/X) (Y = R+B,'YiX) T Statistic R-2
--------------------- -------------------------- ------------ ------------------- ----------- ------
Pe@sonal Income Gross Receipts Taxes SCI.0078 SO.0014 3. 1.q 0.92
Personal Income Gen a Sel Sal-:os Tax $0.0331 50.0062 5.38 0.67
Personal Iricome Prop & In Lieu Tax $0.0071 S0.0006 10.38 C1. 88
Personal Income Other Taxes $0.0018 50.0010 3.58 0-18
Personal Income Other Revenue SO.016q S0.0052 2. 73 0.35
Employment Population 2.13 1.52 1-1.96 C1. 9.4
Population Federal Grants $311, SqO8 5.26 6. 66
Population Human Rpsources Exp S355 S,?r11 11.32 0.9
POPUlatiOn Other Expenditures S119 S?7.q 5.12 0.68
Population Enrol 1 merit 0.12
Enrollment Education Exp S2,1345

Table 5.3 Washington Statq Economic and Fiscal Data: 19?0 - 19S5 (1982

I Population (000) '? 6reneral and Selactivo Salo-- Tax (millionz. 13 Education (millionq)
2 Employment (000) 13 6ross Rev.-oipt:r Tax rmillions) 111 Human Res-ourcos (millions)
.3 Unemployment (000) 'I Propert4 and In-liou Exicos (millions) 15 Total Expendituros (millions)
I Childron (000) 10 Total Ta;4 (millinn:O 16 CPI
5 DSHS casipload (000) 11 Federal Grants (millions) 1? Other Taxd?z 10-9-8-7 (millions)
6 Personal IncomQ- (billions) 12 Total Pevonue (millionf.) le Othor Povonue 12-11-Irl (millions)
19 Other Expenditure 15-1-1-13 (millions)

1 2 -3 1 5 E. .? a 9 la 11 12 13 1-1 Is 16 17 is 19
?CI 31 B 1285 130 S 17 166.0 28.5, 1815.0 258.3 251.? 2291.6 896. -1 4076.8 19 10 - 7 997.7 4284. 8 0.925 166.2 890.8 13'?6. I
71 3,436 1259 1-12 805 200--I 29.8 163-3. 1 291.6 2-1S.0 2403. -? 859.2 3309.7 1831.5 956.-I 1107.2 0.965 226.3 6116.8 1319.4
72 :3430 129? 137 791 202.4 32.1 ISO-q. 1 276. 0 271.9 2-118. I 115?.? -4523.8 1832. 1 1 1'3-3. 0 2 I.OUO 206. -; 91?.6 2169.0
73 3111 136? 11? 788 20-?. 9 36.5 1686.2, 307.6 255.1 2505.9 1094.3 .1275.0. 1731.? 1.070.9 -4191.1 1.058 257.0 675.6 1388.5
?I '3509 1120 109 785 206.3 11. 1 1665.3 329.2 161.0 2103.3 112-I.? 1515.6 1857. 1 1 1'1'0. 5 +412.0 1. 1614 24? . a 1017.6 138-1.1
75 3560 1-112 1149 784 211.5 16 . 31 16861.1 3,35. -q 256. 5 2551.6 204-1. 8 12-22.8 1725.2 1007.4 4098-6 1.253 2?1.3 626.4 1286.1
116 3635 I-q3I 140 780 253.2 51.9 1816.0 315.7 123.9 2809.0 12fiel. 2 5263.1 2248.5 1351.1 5030.3 1.317 3133.1 1090.2 143C.-I
?'? 3? 15 1511 118 776 2-15.0 57.7 1?87.9 5-15.? 50?.2 3098.1 121-1. 1 19?6.2 2125.8 1277. ? -4755.9 1.393 257.3 66-4.0 1352.1
?a 3836 1681 121 769 2-11. 1 6?.:3 1981.2 59?.3 5:3? . 8 -.338-J. 1 17128. 6 5752.1 239?.3 I-q';S. .2 55,48.9 1. .490 [email protected] IO-qO. q 1656.3
79 3979 1806 131 "161 2140. 1 78.0 19 3 8 . -:; 6-19.1 591.5 31Mq. 3 12113.3 5274.2 2198.2 1 -S7 1 . 0 508?. 9 1.625 2F,2.,q 611.7 1518.?
80 -4132 1826 156 757 269.1 07. E. 1950.5 6115.6 608.8 33514.7 1531.2 6 113.3 2729.8 3.715.9 6217. 3 1.7'30 249. 7 1227.1 1901.6
81 1227? 11306 1SO 719 299.6 98.1 17 7 ? . 2' 669.4 631.2 3320.1 2 I 0 9 . 2 5626.1 2512.2 15?9.2 5719. 5 1.9115 2142.5 896.6 1658.1
82 426-4 1779 215 739 251.8 102.9 1901.0 635.0 679.0 3535.0 1--:'-55.5 6.175.0 2751.5 1646.5 61.9,1.0 2.060 320.0 1181.5 VS6.0
83 1285 1838 231 736 2q9.4 10?.? 2232.6 768.6 721.3 1060.2 130?.3 6148.2 2653.6 IS87.9 5973.6 2. 1@16 3.31 .7 ?80.7 1?32.1
81 1320 1859 19-4 ?ql 272.0 113.9 2-1?0. .3 756.1 706. E- 427-J.3 1-409.9 ?097.8 2868.3 1EI64.2 7 194. 6 2.2U-4 310.2 111-1.6 2162.1
85 -4381 191? 170 719 29?.9 126.0 22?5.2 780.2 606.3 1081.3 13,18.1 6?86.4 2?12.1 1782.1 6878.9 2.305 339.6 1357.0 2351.1

Table 5.4 Algorithm for Calculating State
Revenues and Expenditures From Regional Economic
Impacts

PERSONAL INCOME DRIVEN VARIABLES

ITEM AVERAGE MARGINAL

---------------------------- -------- --------
1. Gross Receip ts Taxes $0.0078 $0.0078
2. Gen & Selective Sales Tax $0.0331 $0.0331
3. Property & In Lieu Tax $0.0071 $0.0071
4. Other Taxes $0.0048 $0.0048
5.Other Revenue $0.0164 $0.0164
0.0692 0.0692

6. Employment >> Population 2.43 2.43

POPULATION DRIVEN VARIABLES

7. Federal Grants $317 $317
8. Human Resources Expenditure $355 $355
9. Other Expenditures $419 $419

10. Population >> Enrollment 0.42 0.42

ENROLLMENT DRIVEN VARIABLES

11. Education Expenditure $2,945 $2,945

Revenue = VA*(1+2+3+4+5) + E*6*7

Expenditure E*6*((8+9)+10*11)

Net Revenue - Expenditure

E Employment
VA = Value Added = Personal Income

change in human resource expenditure with respect to population

was $751.

Both the average ratio and marginal change method are

subject to error. For example, there is considerable evidence

that housholds maintain a reasonably stable standard of living

over the medium term, adjusting their savings (or dissavings)

rate as well as their consumption expenditures in the face of

short-term income changes. We would thus expect, (as is observed

in table 5.2) that marginal changes in consumption based taxes

will be less than average ratios. On. the expenditure side fixed

program costs will not vary in direct proportion to population or

caseload. Thus we again expect marginal effects to be less than

average ratios.

That this is not the case in table 5.2 points up the

principal defect of the marginal approach. Estimates of marginal

change, based on regression analysis can be, (as these estimates

undoubtedly are) biased by neglected changes occuring during the

period of estimation. Changes effecting revenue would include

alteration of tax rates and the basis of their application, both

of which have occured in the 1970 -1985 period., Most notably the

removal of the sales tax from food. On the expenditure side,

bias can result from legislated changes in the scale of state

programs and entitlement formulas, as well as changes in popula-

tion structure. Increased state level school funding, and

declines in fecundity between 1970 - 1985 illustrate sources of

potential bias in the marginal values reported in table 5.2.

The removal of these and other biases, if possible at all,

would require data gathering and statistical analysis beyond the

scope of this project. Hence, to give a range of possible fiscal

effects we use both average ratios and marginal values to

compute fiscal effects. An exception is the exclusive use of

average ratios to calculate changes in educational expenditure.

Estimated marginal values (regression coefficients) relating

children to population, and educational expenditures to children,

were both illogical negative values. The first undoubtedly

reflects declines in fecundity, and the second changes in the

state funding formula.

Table 5.4 describes the algorithm used to calculate fiscal

effects. To calculate taxes and other revenues, average and

marginal rates were multiplied by maximum and minimum estimates of

the statewide value added resulting from a 5,000,000 pound

salmon farming industry. Federal grant revenue, human resource

expenditures, and other expenditures were similarily calculated

by multiplying maximum and minimum employment estimates by the

product of population per employee and catagory expenditures per

capita. Education expenditures were calculated in a similar

manner, with the insertion of enrollment per capita.

Results, reported in table 5.5, indicate that annual state

revenues from a 5,000,000 pound industry could range from a high

of $2.26 million (maximum impact, average ratio) to a low of $.36

million (minimum impact, marginal value). Expenditures could

Table 5.5 Impacts of a $25,000,000 Salmon Farming Industry on
Washington State Revenues and Expenditures

Case Revenues Expenditures Net
---------------- ----- ------------ ----------- -------

Maximum Economic Impact

Average Calculation $2,257,105 $1,482,468 $774,637

Marginal Calculation $598,608 $1,273,621 ($675,013)

Minimum Economic Impact

Average Calculation $1,201,216 $1,256,711 ($55,494)

Marginal Calculation $362,981 $1,079,668 ($716,687)

range from a high of $1.48 million (maximum impact average ratio)

to a low of $1.08 million (minimum impact, marginal value).

VI. PROPERTY VALUES

The last empirical research task undertaken in this report

is to investigate the economic implications of assertions that

salmon farms will, due to negative visual aesthetics effects,

reduce adjacent waterfront property values.

It must be emphasised that it is the economic implications

of assertions about aesthetic loss and price decline that are

being examined, not the assertions themselves. Where the

existence of markets, or other circumstances permit the

observation of human behavior toward aesthetic resources, the

measurement of aesthetic values it theoretically possible and

occassionally attempted. However, employment here of the methods

used in such empirical inquiries, such as consumer surveys and

hedonic pricing, would require far more time and resources than

are currently available.

A simpler method is offered instead, which relies only on

publically available property value data and simple regression

analysis. The results of this analysis are, by means discussed

below and in the next section, combined with essentially

arbitrary judgments about the aesthetic effect of salmon farms.

The purpose of this exercise is to provide an analytical

framework within which the results of other research into (or

personal opinion concerning) aesthetic effects can be integrated

with other economic data to inform siting decisons.

The first step in implementing this approach was to collect

the types of data on waterfront property which are available from

real estate firms, multiple listing services, and county

assessors. Summary statistics on the 335 properties surveyed on

this basis are reported in tables 6.1 to 6.7. As indicated in

table 6.1, the average value of the 335 properties surveyed was

$409 per front foot, with a standard deviation of $209. The

lowest average value was in Clallam County $271, and the highest

was in San Juan County $506. This pattern, which coincides with

views of consulted realtors and assessors, results partly from

locational preference for the San Juan Islands, and partly from

the greater predominance of lower valued "high bank" waterfront

in Clallam County.

Among the classes of values obtained, market values (asked

or sold as reported by realtors and multiple listing services)

were on average $223 per front foot higher than assessed values,

$531 versus $303. This difference was also supported by the

experience of realtors and assessors. Current, full market value

is the legal standard for property assessment in Washington

State. However, the fewness of transactions in rural waterfront

areas often makes it difficult for assessors to keep values

current in times of price inflation.

Over the entire sample, values of low bank and no bank

property were, as expected, the highest of the three catagories,

$534 per front foot, versus $396 for medium bank and $312 for

high bank.

Finally, an index, called SCORE, was tabulated, as the sum

of listed property improvements (other than buildings) and other

Table 6.1 Puget Sound Waterfront Property Front F ootage Values

Assessed Market All

--------- ---------- -------
AVERAGE PRICE PER FRONT FOOT

Puget Sound $303 $531 $417
Clallam Co $223 $619 $271
Jefferson Co $300 $451 $428
Kitsap Co $437 $364 $425
San Juan Co 315 $614 $506
Skagit Co $305 $489 $381

AVERAGE PRICE PER
N FRONT FOOT

---- ---------------------------
All 335 $409

High Bank 116 $312
Medium Bank 50 $396
Low or No Bank 100 $534

Score
0 110 $324
1+ 205 $464
2+ 153 $476
3+ 101 $462
4+ 89 $458
5+ 31 $541

Table 6.2 Puget Sound Waterfront Property

Item Assessed Market All

--------- --------- --------- -------
Number 183 142 325
Front Footage
Maximum 3000 5348 5348
Average 408 198 316
Minimum 50 40 40
Acreage,
Maximum 206.00 264.00 264.00
Average 8.83 3.94 6.69
Minimum 0.33 0.17 0.17
Price
Maximum $2,110,000 $3,000,000 $3,000,000
Average $73,064 $140,223 $102,407
Minimum $1,700 $10,000 $1,700
Price per Front Foot
Maximum $1,665 $1,525 $1,665
Average $315 $531 $409
Minimum $9 $72 $9

Table 6.3 Clallam County Waterfront Property

Item Assessed Market All

--------- --------- --------- -------
Number of Prop 36 5 41
Front Footage
Maximum 1320 209 1320
Average 223 118 210
Minimum 62 80 62
Acreage
Maximum 11.71 2.00 11.71
Average 2.49 0.82 2.29
Minimum 0.50 0.17 0.17
Price
Maximum $72,600 $100,000 $100,000
Average $34,722 $60,800 $37,903
Minimum $9,340 $30,000 $9,340
Price per Front Foot
Maximum $450 $1,050 $1,050
Average $223 $619 $271
Minimum $55 -$144 $55

Table 6.4 Jefferson County Waterfront Property

Item Assessed Market All

--------- --------- --------- -------
Number of Prop 10 35 45
Front Footage
Maximum 320 400 400
Average 147 146 152
Minimum 70 60 60
Acreage
Maximum 20.00 7.11 20.00
Average 4.48 2.55 3.35
Minimum 0.89 0.17 0.17
Price
maximum $77,440 $105,000 $105,000
Average $41,429 $58,794 $56,024
Minimum $17,850 $17,500 $17,500
Price per Front Foot
maximum $360 $907 $907
Average $300 $451 $428
Minimum $142 $175 $142

Table 6.5 Kitsap County Waterfront Property

Item Assessed Market All

--------- --------- ---------- -------
Number of Prop 40 8 48
Front Footage
Maximum 540 330 540
Average 179 250 191
Minimum 50 80 50
Acreage
Maximum 11.14 5.03 11.14
Average 2.74 3.38 2.84
Minimum 0.44 0.30 0.30
Price
Maximum $156,500 $150,000 $156,500
Average $67,960 $73,000 $68,800
Minimum $7,600 $29,000 $7,600
Price per Front Foot
Maximum $802 $900 $900
Average $437 $364 $425
Minimum $91 $150 $91

Table 6.6 San Juan County Waterfront Property

Item Assessed Market All

--------- --------- --------- -------
Number of Prop 50 61 ill
Front Footage
Maximum 950 4500 4500
Average 304 339 323
Minimum 100 80 80
Acreage
Maximum 13.28 264.00 264.00
Average 3.75 10.38 7.39
Minimum 0.36 0.50 0.36
Price
Maximum $304,890 $2,000,000 $2,000,000
Average $97,711 $164,26A $94,648
Minimum $27,500 $28,500 $27,500
Price per Front Foot
Maximum $650 $1,525 $1,525
Average $375 $614 $506
Minimum $120 $72 $72

Table 6.7 Skagit County Waterfront Property

Item Assessed Market All

--------- --------- ---------- -------
Number of Prop 47 33 80
Front Footage
Maximum 3000 5348 5348
Average 910 558 765
Minimum 60 40 40
Acreage
Maximum 206.00 114-00 206.00
Average 25.21 13.00 20.17
Minimum 0.33 0.21 0.21
Price
Maximum .$2,110,000 $3,000,000 $3,000,000
Average $165,883 $1210,470 $184,275
Minimum 4,$1,700 $10,000 $1,700
Price per Front Foot
Maximum $1,665 $1,500 $1,665
Average $305 $489 $381
Minimum $9 $121 $9

positive features. For example, SCORE 3 might be availability to

the property of water, telephone and sewer. SCORE 2 might be an

access road and included tidelands. No effort was made to assign

relative value to the items that where added up to obtain the

variable SCORE. Also, of importance to subsequent discussion,

computation of SCORE was based entirely on features of the

property itself. Comments on the general area (near the golf

course, mountian view, etc.) were not counted. Average value

increased with the value of SCORE, from $324 for properties with

SCORE = 0, to $541 for those with SCORE >= 5.

The above summary statistics suggest a method of "backing

into" an estimate of the value of visual aesthetics. As

mentioned, the average value of sampled properties was $417 per

front foot, with a standard deviation of $290. This standard

deviation estimate suggests that, among all properties from which

the sample was drawn, about 68 out of any 100 should fall within

a price range of $417 +/- $290, or between $127 and $707.

The sample was drawn from areas throughout Puget Sound,

presumably including parcels overlooking a wide variety of visual

amenities, and disamenities. Thus, perceived differences in the

quality of nearby visual amenities must have given rise to at

least part of the reported variance in market value. Note,

however, that part of the variance in value can also be explained

by factors unrelated to view of the immediate area. Data in table

6.1 suggests that such non-aesthetic factors include county,

source of price information (assessor or realty firm), bank type

(high,low,medium) and SCORE (which by design reflected only the

degree of land improvement and/or positive features confined

within the parcel itself).

Multiple regression analysis is a statistical procedure in

which overall variance in a dependant variable (here price per

front foot) is either explained by a computed regression

equation, or assigned to the catagory of unexplained residual

variance. The summary statistic R-2 measures the proportion of

variance explained by the regression @equation, the statistic U -

R-2) then measures unexplained residual variance.

Multiple regression analysis was performed on the property

value data set reported in Appendix 1, with the results reported

in table 6.8. The regression equation R-2 of .52 suggests that 48

percent of the overall variance in price per front foot remains

unexplained.

To express this result in terms of price ranges, consider

the previously mentioned one standard deviation range around the

overall average price per front foot, $417 +/- $290, or $127 to

$707. If 76 percent of variance remains unexplained, then the

unexplained, one standard deviation. range is $417 +/- .48*$290,

or $278 to $556. Here we assume some portion of that variance

results from differences in visual. amenities adjacent to the

surveyed properties. In the next section we discuss how this

Table 6.8 Regression Analysis of Puget Sound Waterfront Property Values

Dependant Variable: PFF = Price Per Front Foot, R Square .52
Standard Deviation of PFF $290
Independant
Variables Definitions Coefficient T Statistic

------------- --------------- ------------ ------------
STEPWISE INCLUDED

Constant: 453.05 17.24

DPTl Dummy Variable for Asking Price 174.63 6.38
FF Front Footage Per Property -0.38 -12.00
PRICE Price Per Property 0.00 11.37
DC01 Dummy Variable for Clallam County 120.55 3.60
DBNK1 Dummy Variable for High Bank -167.61 -5.61
DBNK2 Dummy Variable for Medium Bank -177.81 -5.75
DPT2 Dummy Variable for Assesed Price 152.9 2.66
DC02 Dummy Variable for Jefferson County 64.78 2.21
453.05 17.24
STEPWISE EXCLUDED

DC03 Dummy Variable for Kitsap County 0.07 1.53
DC04 Dummy Variable for San Juan County -0.05 -0.94
ACRES Acres Per Property 0.04 0.52
SCORE Index of Property Improvements 0.08 1.72

variance range can be combined with the results of regional

input-output analysis to perform an overall benefit cost analysis

of salmon farm siting decisions.

VII.BENEFIT - COST ANkLYSIS OF SkLMON FkRM SITING DECISIONS

In this concluding section we organize the foregoing results

into a framework for evaluation (from a state economic

standpoint) of salmon farm siting decisions. Two preliminary

steps preceed development of an evaluation algorothim. The first

is to convert previously developed economic information into

comparable economic values. The second is to relate changes in

these values to the specific circumstances of salmon farm siting.

Each of these steps is accomplished by discussion of the

parameter ranges reported in table 7.1.

Regional economic benefits of salmon farming will accrue to

state or county residents during each year of the facilities

operation. kdverse visual effects, on the other hand, will cause

a one time reduction in property values when the facilities are

sited. However, any such reduction in capital value can be

expressed as the loss of an annual-income equivalent by use of an

appropriate interest rate. The economic logic behind making such

a conversion derives from the observation that a property owner

always has the option of selling his property and earning an

annual income from it, as determined by earnings on investments

available to him. That he does not sell, suggest that he places

at least this annual value on the utility or satisfaction derived

from the use of the land in recreational or residential use.

What interest rate should be chosen to reflect this actual

Table 7.1 Input Parameters to Sensitivity Analysis

PARAMETER
NAMES DESCRIPTIONS High Low
------ ------------------------------ ------- -------

RANGES

A = HOUSHOLD INCOME $21,412,000 $10,615,000
B = OPPORTUNITY COST OF CURRENT EARNINGS 50% 75%
C = VISUALLY EFFECTED MILES OF WATERFRONT 25 50
D = SAMPLE STANDARD DEVIATION $285 $285
F = % CHANGE IN AESTHETIC INDEX 10% 20%
G = INTEREST RATE 3% 8%

CONSTANTS

D = SAMPLE STANDARD DEVIATION $285.
G = EXPLAINED VARIANCE (RQ2) 0.24

or implicit annual income is a matter of considerable discussion

among economists. one point upon which they agree, though, is

that a "real" interest rate (i.e. financial rate less expected

inflation) should be used. Deduction of expected inflation is

necessary because the inflation premium in financial rates, which

only maintains initial capital value, provides no estimate of

actual net earnings. As reported in table 7.1, we adopt a real

interest rate range of 3% to 8% to convert property values into

annual equivalents comparable to regional economic benefits.

The different alternatives facing waterfront property owners

and individuals benefiting from regional economic expansion point

up the need for another conversion. Feasable non-

recreational/residental uses of rural waterfront property consist

primarily of agriculture and forestry, activities that would

support only a small fraction of prevailing market prices. Thus,

the waterfront property owner has no realistic alternative to

simply accepting any loss in value that results from diminished

visual asethetics.

By contrast, the houshold incomes earned due to local

economic expansion represent payments for labor and other factors

of production that have reasonably attractive alternatives. Most

workers employed on salmon farms, or in industries supporting

them, could find employment elsewere. For these otherwise

employable workers; incomes, working conditions or other values

achievable in alternative.employment comprise a significant share

of the value they place on their chosen employment. Alternative

value is not, however, likely to exceed value in the chosen

occupation, as in that case the rational worker would change

jobs.

Alternative value will also fall short of the value of

current employment, to the extent that there are costs (and

delays) in finding alternative employment, and to the extent that

some workers (such as the elderly or unskilled) lack viable

alternatives. We adopt an opportunity cost range of 50% to 75%

Ue an implicit net value of gross; income of 25% to 50%) to,

reflect the sum of all these differences between gross regional

income and opportunity cost.

We now address the task of -interpreting the preceeding

sections statistical analysis of property values in terms of lost

net economic value. For previously istated reasons, we begin by

restating our inability to determine, what, if any, negative

aesthetic effects can be attributed -to salmon farms. The purpose

of this report is to work out the economic implications of

independantly provided assessments of visual impact, not to
3
directly estimate these in economic or other terms.

Recall the conclusion of section IV, which suggests that the

front footage price of 68% (one standard deviation) of the

properties from which the sample was drawn should fall within a

range computed as follows:

Actual price regression calculated price unexplained

variance (.48) standard deviation ($290)

For the purposes of this sections benefit cost analysis we

posit an aesthetics index, ranging from zero to one, which

explains all of that otherwise unexplained variation. By this

formulation a property with a zero aesthetics index would fall at

the bottom of the one standard deviation range, ie its price

would be the regression calculated value less (.48*$290 = $139).

A property with an index of 1.0, would fall at the top of that

range (within which 68% of properties now fall), ie. as

calculated from the regression equation, plus $139. Were some

event to change a properties aesthetics index from one (best) to

zero (worst) the result would be a loss of $278 per front foot.

We assume that less than a 1 to 0 change in this aesthetics

index would result from siting salmon farming facilities.

Specifically our benefit cost calculations are based on a 10 % to

20 % range of reductions. These values are, as previously

mentioned, posited for illustration, rather than being offered

either as the results of this research, or as the judgments of

the author.

In addition to a judgment concerning the degree of aesthetic

loss (per effected front foot), we need a similar judgment

concerning the geographic extent (feet or miles of coastline)

over which that adverse effect will extend. Here we assume,

subject to the same qualifications as above, a range of 5 to 10

miles per site, or 25 to 50 miles of coastline for a 5 site

industry.

The final variable required by the benefit-cost algorithm is

gross benefit to the state from economic expansion. For this

purpose we enter the statewide maximum and minimum value added

estimates of $21.4 and $10.6 million.

The benefit-cost algorithm used to perform sensitivity

analysis over the above ranges is reported in table 7.2. Each

calculation compares maximum beneficiary willingness to pay

(numerator), with the minimum required to compensate loosers

(denominator). Maximum beneficiary willingness to pay in this

case is the statewide value added contributions of 5 salmon

farms, adjusted by an opportunity cost factor. Minimum

compensation of losers is the loss of waterfront property value,

calculated as discussed above.

A six variable, 64 case, sensitivity model was used to

calculate benefit cost ratios for all combinations of the input

parameters listed in table 7.1. Results are reported in table

7.3. For the input values and ranges adoped, all cases yield

benefit cost ratios in excess of unity. This suggests that,

under all circumstances and judgments represented by table 7.1

parameters, beneficiaries from salmon farm siting could more than

fully compensate loosers. The maximum ratio, resulting from the

most favorable combination of range variables, is 97.11. The

least favorable is 2.26. Finally, high and low range results are

calculated as the mean value of $21.06 +/- the standard

deviation of $19.34. These results range from 40.41 to 1.72.

Thus, under all parameters and parameter combinations

examined, siting 5 salmon farms would be in the states economic

interest, as this was defined above in terms of beneficiary

willingness to pay and amounts required to compensate loosers.

Table 7.2 Sensitivity Analysis Algorithm, and
Illustrative Calculation

BCR = ANNUAL BENEFITS/ ANNUAL COSTS = 2.26

ANNUAL BENEFITS = A*(J-B) = $2,653,750

ANNUAL COSTS = 5280*C*2*D*(I-G)*E*F = $1,175,962-

WHERE:
WHERE:
A = HOUSHOLD INCOME $10,615,000
B = OPPORTUNITY COST OF CURRENT EARNINGS 75%
C = VISUALLY EFFECTED MILES OF WATERFRONT 50
D = SAMPLE STANDARD DEVIATION $290
E = % CHANGE IN AESTHETIC INDEX 20%
F = INTEREST RATE 8%
G = EXPLAINED VARIANCE (RQ2) 0.52

Parameter values and ranges are reported in table 7.2.
Full sensitivity results in table 7.3.

Table 7.3 Sensitivity Analysis Results

F .>> 3% 3% 3% 3%
A (millions) $21 $21 $11 $11
B >> 50% 75% 50% 75%

c E D ----------------- BCR ----------------------

25 10% $290 97.11 48.5,5 48.14 24.07
25 10% $290 97.11 48.5,5 48.14 24.07
25 20% $290 48.55 24.28 24.07 12.04
25, 20% $290 48.55 24.2;8 24.07 12.04
50 10% $290 48.55 24.28 24.07 12.04
so 10% $290 48.55 24.28 24.07 12.04
50 20% $290 24.28 12.1.4 12.04 6.02
50 20% $290 24.'28 12.1.4 12.04 6.02

F >> 8% 8% 8% 8%
A (millions) $21 $211 $11 $11
B >> 50% 115% 50% 75%

c E D ----------------- BCR ----------------------

25 10% $290 36.42 18.2.1 18.05 9.03
25 10% $290 36.42 18.9.1 18.05 9.03
25 20% $290 18.21 9.10 9.03 4.51
25 20% $290 18.21 9.10 9.03 4.51
50 10% $290 18.21 9.10 9.03 4.51
50 10% $290 18.21 9.10 9.03 4.51
50 20% $290 9.10 4.55 4.51 2.26
50 20% $290 9.10 4.55 4.51 2.26

Average 21.06 High range 40.41
Standard 19.34 Low range 1.72
Maximum 97.11
Minimum 2.26

Notes

1. British Columbia data was provided by Jim Fraylick, British

Columbia Ministry of Agriculture and Fisheries.

2. Washington data was provided by Robert Hoyser, Washington

State Department of Natural Resources; and Eric Hurlburt,

Washington State Department of Fisheries.

3. Some data and an appraisers judgment concerning the degree

and geographic extent of adverse visual and market effect is

provided in Alpine Appraisers, 1988. The author of that document

concludes that "floating net pens have no effect on upland

property values in the areas studied (Peal Passage Mason County,

and Rich Passage, Kitsap County). Additionally, the appraiser

concludes that "the pens will have minimal, if any, visual impact

at distances over 2400 lineal feet.

References

Alpine Appraisal Service. "Influence of Floating Salmon Net

Pens on Residential Property Values." Report to the

Jamestown Clallam tribe, Sequim, Washington, August 30,

1988.

Richardson, Harry W. Invut Outiput and Recional Economics. (New

York: John Wiley and Sons, 1972).

U.S. Forest Service (Portland Oregon). IMPLAN Data, (Reports

provided to the Author).

Washington State, Office of Program Planning and Fiscal

Management. Pocket Data Guide. Olympia, WA 1983.

A.1 Puget Sound Waterfront Property Survey.
FRONT PRICE PPRICE
COUNTY SITE ACRES FEET PRICE MONT FTYPE BAM' FEATURES EWILOINGS SCORE SOURCE

Clallan Port Angeles 1.50 100 S45, 000 $150 assessed h uater.pavv 6?960 4 Clallan Cc Assessor
clallam Discovery Bay 1.00 136 S27.200 $200 assessed 0 Clalla" Cc Assessor
Clallan Port RngolQs 0.96 200 S35,000 S350 as-zossed h elec,mater.ph.pave -1 Clallam Cc Assessor
Clallan Discovery Say 1.13 136 S2?.200 S200 assessed 0 Cl &I I an Cc Assessor
clalla" Clalla" Bay 0.50 102 $9,340 S92 assessed I tdl el ec,wat. saw, ph,road 6 Clallan Co fissassor
Clallan W. of Port Angeles 11.26 1320 S721.600 SSS assessed erosion 0 Clallam Co Assessor
Clallan W. of Disc Bay 1.09 105 513,650 $130 assessed h elec,ph,access 295 3 Clalla" Cc Assessor
Clallan W. of Disc say 5.03 295 S60,670 S206 assessed h 13?5 0 Clalles" Cc Assessor
ClaLI14M W. of Disc Bay 5.03 2W S63,670 5250 assessed h elec,ph 1375 2 Clallan Cc Assessor
Clallan Cl al I a" Bay 0-50 IL21 $21,320 S201 assessed I tdl,.Plac,wat,s&e&.ph,road 48710 6 Clallars Co Assessor
Clallan Port Angeles 1.16 100 $45.000 $450 assessed h al ac, pave 2 Clallan Co Assessor
Cl al I an Clallam say 0-50 1.50 S30,150 $201 assessed I tdl , &I ec, &set. seas. ph, road 6 Clallam Co Assessor
Clallan Port Angeles 1.16 104 SAG,800 S150 assessed h el ec, seater, swepti c, ph. pave 69700 5 Clallam Cc Assessor
-_'Iallan Clallam Bay 0-50 62 S12.660 S;201 assessed I tdl,&Iec,&eat,sem,phroad 11820 6 Clallan Ca Assessor
Clallan Clallan say 0.50 63 S12.6W S201 assessed I tdI,elQ-r-,seat,s&s&,ph,ro&d 30LW30 5 Clallam Cc Assessor
'-Iallan W. of Port Angeles 9.05 560 S56,025 SIOO assessed h 0 Clalla" Co Assessor
lialla" Discovery Bay 1-22 200 $27.335 $137 assessed h j) Clallan Cc Assessor
clalla" M. of Disc Bay 5-03 2SS $31,835 $125 assessed h elec, no park 0 Clallam Cc Assessor
Clallan M. of Disc Bay 1.75 175 S;22.?50 5130 assessed h elec,ph 47S 2 Clallan Cc Assessor
clalla" M. of Disc say 1.10 100 S13,000 S130 assessed h elec.ph,accoss 3 Clallan Cc Assessor
Cl al 16" W. of Disc say 5.03 255 S61.6dW S212 assessed elec,ph,access 61" 1 Clallan Cc Assessor
clallam W- of Port Angeles 11-19 ;215 S21,500 $100 assessed h 0 Clallam Cc Assessor
Clallan Clallan Bay 0.50 63 S12,660 S201 assessed 1 elec.seater.sower.road 36270 A clalla" Cc Assessor
Clallan Discovery Bay 1-00 136 S22,000 S162 assessed 0 Clallam Cc FIssessor
Cl al I art Discovory Bay 1.09 1-36 S2?,200 S200 assessed 0 Clallam Cc Assessor
Clallan Discovery say 2-05 150 S52,500 5350 assessed 0 Clallam Cc Assessor
Malla" W. of Part Angeles 11.?l 1320 $72,600 $55 assessed po-m- terrain 0 Clallam Co Assessor
Clallan Port Angeles 1.18 90 S40.500 SISO assessed h e1&c.Matqr,Sept3LC,ph-P- 5 Clallan Cc Assessor
clallan Port fingeles; 1.44 100 $45,000 $150 assessed h elec,seater,septic.ph.paRlee 51920 5 Clallan Cc Assessor
Cl al 1 -an U- of Port Angeles 1.50 a3s S33,535 SIOO assessed h 0 Clalla" Co Assessor
clalla" Port Angeles 1.50 110 S:46. 750 S425 assessed h al ec. water. septi c' ph.pave 1350 5 Clallan Cc Assessor
z1alla" Di scoverg Bay 1-09 136 S2?,200 5200 assessed 0 Cl al I are Cc Assessor
Clallan M. of Disc Bay 1.23 100 S13,000 S130 assessed access,easersent 2 clalla" Cc Assessor
Clallars Port Angeles 1-3? 2-31 S561000 S3?3 assessed h 01 ec, water. --Aipp. ph, pave 6?"0 1 Clallan Cc Assessor
Clallan Di scovery Bay 1.00 200 S27.135 S136 assessed 0 Clallam Cc Assessor
clallan Clalla" Bay 0.50 99 S173,920 5-201 assessed 1 al ec, water. sawer, road 66780 4 Clallam Co Assessor
Clallan W. Ediz Nook 0.17 100 S50.000 $500 market h Buildable I Sea Ridge Realty
Ciallan The Place Beach 0.75 ADO S100.000 $1,000 market 1 lagoon 2 Sea Ridge, Realty
clallan Freshwater Bay 2.00 209 S30,000 5144 MWkQ-t h 0 Sea Ridge Realty
Clallam 4 Seasons Ranch 1.00 so 54H,000 S1,050 market I Drain field.planned dev,tdlnds I Sea Ridq& Realty
*:I al I an Straits 100 -Sqfj. 000 S400 market h Building lot I Sea Ridge Realty
Jefferson Ft. Flagler 0.99 100 S36.000 S360 as3Qssed h aasment 330-15 1 Jefferson Cc Assessor
Jefferson Ft. Flagler 1.91 200 ::-67,500 $338 assessed h 12125 A Jefferson Cc Assessor
)OffOrSOM Squamish Harbor 3-00 70 S17,950 S255 assessed 1 0 Jefferson Cc Assessor
Wferson Ft. Flagler 1.13 100 S:36.000 S360 assessed mater,dirt road 2 Jefferson Cc Assessor
Jefferson Ft. Flagler 1-93 2.10 S77,110 S323 assessed m poor access 21285 -1 Jefferson Cc fissassor
Jefferson Squarvish Harbor 5.00 110 S28.050 S-155 assessed 1 0 Jefferson Cc Assessor
Jefferson Ft. Flagler 0-92 100 $35,000 S350 assessed m 0 Jefferson Cc Assessor
Jefferson Ft. Flagler 5.04 120 11,43, 000 S358 assessed m 27030 -1 Jefferson Cc Assessor
Jefferson Squamish Harbor 20-00 320 $1-5,395 S142 assessed 1 0 Jefferson Cc Pessessor
Jefferson Squanish Harbor 5.00 1-110 S28,050 S255 assessed h 0 Jefferson Cc Assessor
Jefferson E. Hood Canal 0.61 100 SS8,000 5580 market hn septic,wator,voad 3 Jefferson Cc Realtor
Jof forson North Beach 0.26 30 SIM. 000 S500 narkot h water,ro-ad.sewar 4 Realty World
Jef forson Hiddle Pt. 5.02 121 S46,000 S390 market h po"'Or.road,need septic,need moll 2 Realty World
.Jefferson Squanish Harbor 1.11 105 5;45.000 5" market In sandy beach I Jefferson Cc Realtor
Jeff or --.on riots hats Bay 1.62 ]L'?O $10,500 S250 market he ti del ands.power,&aell 3 Realty World

A.1 Puget Sound Waterfront Property Survey

FRONT PRICE PPRICE
COUNTY SITE ACRES FEET PRICE FRONT FTYPE BRINK FEATURES BUILDIMGS SCORE SOURCE

Jefferson Oak Say 2.07 105 ST3,500 S567 nat-ket n ear-sownt I Jefferson Co Realtor
Jefferson harroustone 1. 2.32 148 S15,000 S301 "aricet Po6for I Jefferson Co Realtor
Jefferson Pt- Ludlou 0.17 60 S-1-3.530 5725 narket n matot-,marina mgnts 2 Jefferson Co Realtor
Jefferson Hats "ats Say 1.161 100 S?-q 1 000 S7-10 market septi c I Jefferson Co Realtor
Jefferson "adrrastone 1 4.96 300 S65,000 S217 market 0 Jefferson Co Realtor
Jefferson Toandas Pon. 2-63 1135 S".500 $269 nw-kot M power.fruxt troet; 2 Jefferson Co Realtor
Jefferson harroustone 1 3.03 1.50 S-", 000 $327 narket powor,septic 2 Jof fors4m Co Realtor
Jeffersort QuilconQ 2.70 100 S82,500 $825 narket n water, ti siber, beach 3 Jefferson Co Realtor
Jof forson harrastone 1 3.26 290 S105,000 S-162 narket 0 Jefferson Co Realtor
Jefferson Cape Goorge 0.57 1." S48,000 S322 narket h uator.poiwer.tal,draiin field I Realty World
Jefferson Oak Bay 0.95 100 S",500 $395 narkot 0 Jefferson Co Realtor
Jefferson E. Hood Canal 1.00 200 S%, 000 S560 narkpt I Mater.pomer 2 Jefferson Ca Realtor
Jofferson Hiddle Point 5.20 121 $52.000 S430 narkot h rcod.neod u4i.11. ewped septic 2 Realty World
Jefferson Admiralty Inlet 0.% 100 540,000 $100 narket h 0 Jefferson Cc. Real tor
Jofferson Hadlock 5-01 400 S92, 500 S231 "arkat ti dol ands, water, pomor 3 Jefferson Co Realtor
Jefferson E. Hood Canal 0.53 75 S68,000 $907 naw-ket- h eacsnont I Jefferson Co Realtor
Jefferson E. Hood Canal 2.05 lie S51,000 S132 narket M pork problem. 0 Jefferson Co Realtor-
Jefferson Hood Canal 2.63 185 $15,000 S-M3 market ti 091 ands. road, power. parks, 4 Realty World
Jefferson Strai ts 3.00 136 S49,WO S364 n-w-ket Moll I Jefferson Co Realtor
Jefferson harroustone 1. 0-86 100 S79.000 S750 ".w-kot I septi c I Jefferson Co Realtor
Jefferson Harroustone 1. 3.36 200 M,500 S498 "arkot P06"r I Jefferson Co Realtor
Jefferson Oak Head ?.11 200 Sfi% SOO S319 narket h po"Or I Jefferson Co Realtor
Jefferson "arroustone 1. 3.36 200 S99.500 S-198 narket ti del ands, mol I , power. tel I Realty World
Jefferson Ifiddle Pt. 5.30 121 S52, 500 5134 market h po*&or. road. need "oll.need septic. 2 Realty World
Jefferson rarboo Bay 1.23 100 S17,500 S IM narket n spring I Jefferson Co Realtor
0% Jefferson E. Hood Canal 3-47 200 $18,000 5210 nar-ka-t h power I Jefferson Co Realtor
4- Jefferson Gardiner 3.20 171 S?9.000' S462 narkot h pouor.ma-Ler,tinbor A Roaltq World
Jefferson Discovery Bay 2.52 100 57s, 000 S?50 market 1 0 Jefferson Cc. Realtor
Jefferson M ddl a Pt. 5.02 121 $46.000 S-380 narkot h powor. road. rwpod "oll.nood septic 2 Realty World
jef C-eer-e 0=40 90 S;4s_ @ 9 m_
55nn nm-k-at h rMumor @ ual 1 2 Jefferson Co Realtor
Kitsap, Poulsbo 2.70 175 $52,500 S300 assessed I.Q ipRx 3 Kitsap Co Assessor
Kitsap Fletcher Say 4.67 220 $107,350 SIBS assessed h BSP14FGX 62540 4 Kitsap Co Assessor
Kitsap Fletcher Bay 0.88 70 S-*3.390 S577 assessed le xxmxzxx 0 Kitsap Co Assessor
Kitsap Eglon 0.46 so S7,600 $152 assessed h XC-4"GX 0 Kitsap Co Assessor
Kitsap Eglon 0.56 60 5'9,120 $152 azsezsod h XSHWFGX 0 Kit-sap Co Assessor
Kitsap Colvos Passage 2.50 210 $42,000 S-200 assessed h 6SH14FGE 2 Kitsap Co Assessor
Kitsap Poulsbo 0.53 100 $30.000 S300 assessed le GXP14FFX 2 Kitsap Co Assessor
Kitsap Vinland 1-39 155 S109,910 S?02 assessed le IW.CPC)CW3X 90860 5 Kitsap Co Assessor
Kitsap Vinland 0.79 120 S%,240 $802 assessed le BX(PC)ICRGX 65360 5 Kitsap Co Assessor
Kitsap Fletcher Bay 1.97 160 S32,090 5576 acssessed le GX(PC)CFGE 39M 5 Kitsap Co Assessor
Kitsap Colvos Passage 5.07 330 550.160 $152 assessed h BCSWPUFGX 3 Kitsap Co Assessor
Kitsep Tokiu Pt. 5-97 S97.150 $229 assessed h GSPCFGX 2 Kitsap Co Assessor
Kitsap Vinland 2.36 S110.400 assessad le BX(PC.)CPIGX 41430 5 Kitsap Cc fls-s"assor
Kitsap Poulzbo 1.?? 205 S138,?80 S677 assessed le BXCPC)Cfm 31M S Kitsap Co Assessor
Kitsap Colvos Passage 5.09 112 $46,434 S32? asses5od le GxtMGE 2 Kitsap Co Assessor
Kitsap Vinland 3.914 2010 $100.100 5502 assessed le GX(PC)Cfi13E @M 10 5 Kitsep, Co Assessor
Kitsap Fletcher Bay 1.17 73 M,020 S726 assessed le BXCPC)CFGE 95230 6 Kitsap Co Assessor
Kit-sap Eglon 1-88 100 S9,100 $91 assessed h XSM14FGX 0 Kitsap Ca Assessor
Kitsap Eglan 2.02 170 S21,250 $125 assessed lo DXNXXXX I Ki tsap Co Assessor
Ki tsap Fletcher Bay 3.00 210 S156,500 S715 assezsed h BSU"CpXFGX 146150 5 Kitsap Co Assessor
Kitsap Fletcher Say 2.61 150 S101,050 S6-el -assessed le 6XCPM)XM4E 4 Kitsap Co Assessor
Kitsap Vinland 0.86 100 S80,200 SJ302 assessed I* WWCRGX 113380 4 Kitsap Co Assessor
Kitsap Tokiu Pt. 8.49 300 S101.980 SMO assassed lo 6Y(PC)CRGX 182700 4 Kitsap Co ftsossor
Kitsap Colvos Passage 3.26 200 3.%),000 S200 --m essed h U.S#MFGE I Kitsap Co Assessor
Ki tsap Colvos Passage 8.18 220 $77,440 S352 aLssessed le MCPC)CFGE I Kitsap Co Assessor
Ki'tsap Eglon 2.38 260 1,31.200 $120 &-5sessed h XSNWGX 0 Kitsap Co Rssessor

A.1 Puget Sound Waterfront Property Survey

FRONT PRICE PPRICE
CDUMTV 51TE ACRES FEET PRICE FRONT FTYPE SAW FEATURES BWLOINGS SCORE SOURCE
Ki t5ap Fletcher Day 1.61 39 5:59,260 S666 assessed h GS(PC)UFGE 69930 1 Kitsap Cc Rsse@ssor
Kitsap Tokiu Ft. 11-1.4 510 S102,310 S199 assessed le GGX I Kitsap Cc Assessor
Kitsap Colvos Passage 8.09 M S86,070 $302 assessed h US@MGE I Kxtsap Cc Assessor
Ki tsap Vi n1 and 0.66 100 S80.200 S802 assessed I* BXCPC)CfiGX 131180 5 Kitsap Cc Assessor
Ki tsap Poulsba 0.60 116 $;37.450 S323 assessed le GXCPC)CFRE 549100 5 Kitsap Cc Assessor
Kitsap Vinland 0.59 Ila SM,220 5.902 assessed le BXCPC)CFIGX 723;?Yj S Kitsap Cc Assessor
Ki tsap Colvos Passage 2.27 100 S20,200 S202 assessed h USNWGE I Kitsap Cc Rssossor
Kitsap Eglon 5.30 510 S7.3,500 S 111 a2ssessed h XSNWGX 0 Kitsap Cc Assessor
Kitsap Poulsba 0.50 197 S43.210 5652 assessed 19 OX(PC>C6r3X 5 Kitsap Co Assessor
Kitsap Poulsba 0-60 130 S91,260 S702 assessed I& ot*PC')CGGX 252SO 5 Kitsap Cc Assessor
Kitsap, Poul sba 1.30 100 S35,000 S350 azsassod le BXU31C)IcFftx 30356 5 Kitsap Cc Assessor
Ki t5ap Poulsba 1-11 115 SW,5530 1-5700 asse@sed le BXCPC@0C6M 102990 5 Kitsap Cc Assessor
Kitsap, Poulsba 0.14 190 S;36,000 S-400 assessed I* rv)WCFGX 2 Kitsap Cc Assessor
Kitsap Poulsbo 0.168 IGO S61,000 SVO azzo5sed le 6XCM)CFFIX 114510 4 Kitsap Cc Assessor
Kitsap Stavis Day Road 5.00 330 S419,500 S150 nairkot he phone, no alec: I Coldwall Banker
Kitsap Stavis Say Road 5.03 BL30 $62,500 S189 markot he pomor,phonQ 2 Coldwell Banker
Kitsap Stavis Day Road 5-00 330 S65,000 S197 market he *I ec. proone, no "tar 2 Cold"oll Banker
Kitsap Stavis Bay Ro-ad 5.00 330 S150,000 SISS market I power.phone,septic,mQll I ColdwQlI Banker
Kitsap SlLavis Say Road 5.00 3:10 $70,000 S212 market he elec,phone,no matec 2 Coldmall Banker
1(i tsap Big Beef Hairbor 0.50 U" 000 S279 market el&c,phone.needs swIl 2 Coldwall Banker
Kitsap Rich Passage 0.30 80 S7-2.000 $900 market I elec,phorw,mater 3 Coldmall Banker
Ki tsap Olympic View Road 1.23 163 S06,000 S529 markot he elec,phone. no mater 2 Ccl duel I Banker
San Juan San Juan M. 9.54 770 S301.850 S3% assessed to 59-40 -q San Juan Co Assessor
';4m Juan Sham S 1.38 310 S1113.050 53131 assess" 1 ti 601 ands 165M 1 San Juan Co AssQssor
0% San Ju&n Orcas 1.56 2 10 S790,510 $431 assessed m tidelands I San Juan Cc Assessor
Un San Juan Sham N 1-37 315 IM4.500 S300 assessed 1 69720 11 San Juan Cc Assessor
San Juan Lopez M 2.81 315 S91.500 $300 assessed h -1 San Juan Co Assessor
3an Juan Sham N 0.57 157 $5-1.950 S350 assessed m 212M 4 San Juan Co Assessor
San Juan Lopez E 0.43 270 S53.750 5125 assoss-ed n 0 San Juan Co Assessor
San Juan Sham S 4.9 7@W SISS,500 S-1'10 assessed h 11690 4 San Juan Cc Assessor
Gan Juan Orcas 3-02 310 $6-1,000 S200 assessed n 0 San Juan Cc Assessor
San Juan Lopez M 12.7 4010 S:90.010 S225 asswssed " 0 San Juan Cc Assessor
San Juan San Juan SM 1.91 213 S128.850 S605 assessed n ti del ands 7220 5 San Juan Co Assessor
San Juan Sham SE 1.18 100 $60,000 5600 assessod n 250-40 1 San Juan Cc Assessor
San Juan Orcas E 5-71 Z190 s9F"450 S333 aissessod h 16?0 0 San Juan Cc flssess@w
San Juan Sham N 2.83 300 11-90,000 S300 assessed m ?6910 1 San Juan Co Assessor
Lan Juan Sham SE 1.84 1w SO-1,400 5556 assessed n tidelands 13?00 5 San Juan Cc Assessor
San Juan Sham M 6.44 665 SIGO.,290 52-11 assessed 1 1280 0 San Juan Cc Assessor
Juan Lopez It 0-38 Ito S41.250 S375 assessed " ti del arpas 16750 5 San Juan Cc Assessor
San
San Juan Orcas E 5.09 360 S1,22,050 S311 arssessed m 121970 1 San Juan Co Assessor
San Juan Lopez H 12.29 T90 SIS-4.9% S263 assessed h 0 San Juan Cc Assessor
.1an Juan San Juan M. 2.OS 200 SIIQ,O(JC S550 a@ssessod h 0 San Juan Co Assessor
San Juan Lopez E 0.68 120 SGG@720 S556 assessed n ti del aras 12410 5 San Juan Cc Assessor
ian Juain Lopez E 2.5 1110 SIM,000 S600 assa=sad m 133?% 4 San Juan Cc Assessor
San Juen Lopez E 0.36 100 S;2?.SW S2?5 assessed " 0 San Juan Cc Assessor
Gan Juan Lopez S 9.75 1,10 S150.200 S341 assessed m 0 San Juan Co Assessor
San Juen Lopez HM 5.1 2190 S;7-1,550 S257 assessed h 2-60 0 San Juan Cc Assessor
San Juan Sham M 0-81 2100 S70,000 5350 azsossed m 1179,10 4 San Juan Co Assessor
San Juan Orcas SE -4.01 Z65 "3,550 S-342 assessed " 0 San Juan Cc Assessor
ian Juen Shajw S 2.7 950 S142,500 $150 assessed h 0 Son Juan Cc Assessor
San Ju,.,n Orcas; E 10.65 535 S151,650 S203 assessed h 0 San Juan Cc Assessor
San juen Orcas E 1-62 1!50 S?8.?50 S525 assessed h 40260 1 San Juan Cc Assessor
San Juain Lopez E 2.2 100 W5,000 S6SO assessed m ?1430 4 San Juan Cc Assessor
San Juan San Juan M. 1.0 160 SIO-4,000 SGSO a--3s9zsed n tidelands 813W 5 San Juan Co Assessor
San Juan Orcas SE 1q.02 250 S1.08.050 S132 assessed m 9226 A San Juan Cc Assessor
San juain Sulam N 2.1 5:10 @;G 1, 200 S120 assessed m 2550 0 San Juan Cc Assessor

A.1 Puget Sound Waterfront Property Survey

FRONT PRICE PPRICE
CGLJNTY SITE RCRES FEET PRICE FRONT FTYPE BRW FEATURES BUILDINGS SCORE SOURCE

San Juan San Juan M. 2.23 210 S105,000 S500 assessed h 192730 1 San Juan Co Rssessor
San Juan Orc4s 1.66 200 S86,200 S131 assessed h ti del arAis 1608-43 5 San Juan Co Rssessor
San Juan Sham N 7.31 135 S130,500 5300 assessed n 88810 1 San Juan Co A-ssessor
San Juan Orcas N 1.2 ISO S?5.000 $500 assessed 1 0 San Juan Co fissessor
San Juan Lopez E 0.83 135 S?5,060 S556 assessed 1 ti del ands I San Juan Co, ft:sses5or
San Juan Sham M. 13-28 2% 5 172. -100 $591 assessed 1 651-40 4 San Juan Co fissessor
San Juan Lopez M 1.99 120 S61,290 5511 assessed h 21690 1 San Juan Co fissessor
San Juan Orcas N 3.11 165 S80,880 SliO assessed 1 0 San Juan Co Assessor
San Juan Orcas E 5.1 -1k5 5 128. 760 $326 assossed h 992% Iq San Juan Co Assessor
San Juan Or-cAs M 0-166 2120 S33,000 S150 assessed h 0 San Juan Co fissessor
San Juan Lopez N 0.65 170 S52,050 5306 assessed h 156320 4 San Juan Co Rssessor
San Juan Lopez E 3.15 300 5112.500 S375 assessed 1 506110 1 San Juan Co Rssessor
San Juan Shaw M S.?? W- S166.500 $138 assessed m 0 San Juan Co fissessor
San Juan Sk-aw S 2.18 279 5113.000 5-W& &:;sessed I ti xiel ands 4310 4 San Juan Co fissessor
San Juan Orcas 1. q 355 S62,130 Sl?5 assessed " 0 San Juan Co fissessor
San Juan Orcas E 5.1 270 $9?.200 S360 assessed h 0 San Juan Co fissessor
San Juan "osquito Pass 5.00 367 $215,000 S586 market 1 0 Island CrA"tor Services
San Juan Karwka Bay 1.25 160 SIIO.000 5688 market 0 Island Computer Services
San Juan Orcas, Diamond Pt. 264.00 15W $2,000,000 S-q-q4 market hm 0 Dockside Property
San Juan Vacht Haven 0-70 1.10 $89.500 $639 market n pur,boach 2 Island Computer Services
San Juan Pew- Pt- 1.19 184 $112.500 $611 market pw,utr 2 Island Computer Services
San Juan Gr-i-ff in Bay 3-?8 265 S224.000 SO-45 "arket 0 Island Computer Services
San Juan Sunset Pt. 0.50 125 595.000 S760 market 0 Island Computer Services
San Juan Son Juan 0.50 130 S61,500 Sl% market 0 Island Computer Services
San Juan San Juan 2.30 240 S135.000 $396 market pur,wtr 2 Island Computer Services
San Juan Moil Day 0-711 100 S62,500 S62S market n pGAr.&&tr.-.spt 3 Island Computer Services
San Juan Cape San Juan 0-50 so $39.950 S199 market pr-c.pur,utr 3 Island Computer Services
San Juan Cape San Juan 0.50 ion S69.500 SE95 market pr-c,pwr,spt 3 Island Computer Services
San Juan San Juan 1.42 125 $120,000 S619 market I pmr,batr 2 Island Computer Services
Safi juar. @-Lal
e0_95 SM sns.wo !51@325 nouaq- . ma-ter. Dark. aravel beach S Oockzide Property
San Juan Dozidman Bay 5.10 1005 S 15?, SOO 5157 n-w-kat 0 Island Computer Services
San Juan Fr-iday Island 0.70 1010 -q6q,5OO S695 market 0 Island Computer Services
San Juan Garr-- son Baoj 56.00 200 S225,000 S 1. 121-; market ti mber,pw 2 Island Computer Services
San Juan University Hei0ts: 16.60 250 S139.900 S5&O n-w-ket O,Island Cottputar Services
San Juan Cattle Point 1.00 165 S65.000 $391 market pmr,&4tr 2 1 sl and Computor Services
San Juan VacJvt Haven 7-60 11?8 seq.5w S?2 market n P"r I I sl and Computer Servi ces;
San Juan SJ, Friday Harbor 3.00 q5O S185,000 5111 market In boach/cove 2 Dockside Property
San Juan San Juan 2-16.4 210 "5.000 5452 market pur,utr 2 Island Computer Services
San Juan San Juan 5.46 3110 5110.000 S-324 market pur I I sl and Computer Services
San Ju Roche Harbor 1.2s 160 5209,000 S1,300 market I tkuorago I Island Computer Services
San Juan University Heights 0.50 100 S58,500 S585 market P64r,64tr, 2 Island Computer Services
San Juan Stuart 1. 9.92 Aw S75,000 SIS6 market 0 Island Computer Services
San Juan Shaa4 2.00 225 5100.000 5411 market I pwr I Isl and Computer Service--
San Juan Mitir-hell Bay 5.15 625 $290.000 SIGA market 0 Island Computer Services
San Juan Uw-v--ott Bay 0.50 100 S72.500 S?25 market I pbar , w tr 2 Island Computer Services
San Juan Roche Harbor 2.00 175 S198.000 S1,131 market I Pur 1 Island Computer Services
San Juan Eaigle Cove 1.00 100 S57,500 55?5 nar-ket pur.ph 2 Island Computer Services
San Juan Cat-tl 9 Pt. 0.88 120 S65.000 5512 market pur.wtr 2 1 al and Computer Set-vices
San Juan 11inwal. Heights 0.50 I-qO S33.000 S:236 market pmr,utr, 2 Island Computer Set-vices
San Juan Daw-idson Head 0.50 100 S75.000 S?50 market utr I Island Computer Services
San Juan San Juan 1.00 1-30 $85,000 S654 market n 0 Island Computer Services
San Juan Moil Bay 0.62 200 S115.000 S1.150 market n pmr,utr 2 Island Computer Services
San Juan Sit ww-t 1. 1.00 112 S;28.500 S254 market In 0 Island Computer Services
San Juan Griffin Bay 2-25 I-IS S12-4.500 S859 market n pjwr I Island Computer Services
San Juan Opadnan Bay 5.30 780 S158.500 S203 market P14r I Island Computer Services
San Juan Capo San Juan 0.70 100 S79,000 5790 market dock,peal 2 Island Computer Services

A.1 Puget Sound Waterfront Property Survey
FRONT PRICE PPRICE
COUNTY SITE ACRES FEET PRICE FRONT FTYPE BFW, FEATURES BUILDINGS SCORE SOURCE

San Juan Orcas 0.50 100 S85.000 S850 narket pmr.batr 2 Island Computer Services
San Juan Eagle Cove 0.80 150 5n, 500 SS3O market h p&.er,utr 2 Island Computer S*rvicQs
San Juan San Juan 1.61 375 S210,000 S!%O market 1 0 Island Computer Services
San Juan Rocky Bay 3.11 240 S149,500 S623 market Pwr I Island Computer Services
San Juan Griffin f%ay 0.75 100 S09.500 S095 n-3rk-;-t pmr.cova 2 Island Conputeq- Services
San Juan Davidson Head 0-50 100 575.000 S750 narka-t pwr,setr 2 Island Computer Services
San Juan Kwwsah Hei ghts 1.00 200 S61.500 S3AS narkrot pr-c.pur.mtr,spt 4 Island Computer Services
San Juan Mescott Bay 0-50 wo 5;?6.500 S765 market pu-.mtr.:spt 3 Island Computer Services
:San Juan Reid Harbor 5.12 300 S65,000 5217 "arkot It dock I Island Computer Services
San Juan "scptt Bag 0.70 I'm S75.000 5536 market 1 0 Island Computer Services
San Juan San Juan 15.20 3630 S135,000 $355 market pwr I Island Computer Services
San Juan Universitej keights 0.50 IJOO S 120. 000 S1.200 market 1 0 Island Computer Services
San Juan San Juan 1 5.00 700 S75.000 $107 narket 0 Island Computer Services
San Juan San Juan ?19.25 360 S519.000 SI.525 market m pur I Island Compo tQw Services
San Juawt San Juan 0.81 185 S18S.000 S1,000 market n P-C I Island Computer Services
San Juan San Juan 3.97 6M S180.000 S300 market It pmr.wtr.spt. 3 Island Computer Services
'San Juan S. End 1.00 200 5;76.500 S383 market pur,setr 2 Island Computer Services
San Juan Moscott gag 5-00 755 S187,508 S219 mairket I bq;,ach I Island Computew Services
San Juan Griffin Bay 1.50 120 S99,506 S829 market " pwr I Island Computer Services
San Juan Eagle Pt. 6.17 375 S171,500 SIGS market h pur 0 satr 2 Island Computer Services
I utr.pur 2 Island Coisputer Services
...van Juan Li nestone 5.00 310 S217,000 $700 market
Skagi t E 6uamez Ch 9.6 7100 S31,300 S-95 assessed h road 312M 4 Skagit Co Assessor
Skagit Burroses Bag 5 Z" 535,500 $18 assessed h 0 Skagi t Cc Assessor
Skagit Padilla Bay 26 16W 590,000 S56 assessed h read I Skagi t Cc Assessor
S I Skagit Cc fissossor
Skagi t Padilla Paq 10.68 13aw S181900 S;15 assossed I road
0% Skagit Padilla Bag 1.5 200 S11?00 $13 assessed I seater 9000 1 Skagit Cc ftsessor
4 Skagit Ship Harbor 7-09 SOO V48.9w $61 assessed m road.mator.s@eptic 75900 1 Skagit Cc Assessor
Skagit Padilla Bay 'q?-?5 1600 536. 900 $23 assessed h &&ater.septic,road 3 Skagit Cc Assessor
4;kagi t Bellingham Say 3.?q 200 5;85.860 S429 assessed I seater.3eptic,road 3 Skagit Cc Assessor
Skagi t Rosario St 55-1 I-qw saa. 930 Sfil assessed h 0 Skagit Cc Assessor
Skagit Sinclair 1. 26-75 IGM S11-9,390 $75 assessed h 0 Skagit Cc Assessor
Skagit Sol I i "ham Ch 19- 11 3;20 S38.200 S119 assessed h road I Skagi t Cc Assessor
Skagi t Padilla Bay 35.8 19M S2,49,270 S131 asses!wd h road 1 Skagi t Co Assessor
Skagit Padilla Bay 314.?e low 52G.Ow S20 4354ssod 1 0 Skagit Cc Assessor
Skagit Rosari 0 St. .1 100 S45.800 S158 assessed h easnent,septi c, seater, road SIM -q Skagit Cc Assessor
Skagit Rosario St 0.38 IL 10 5:6(j. Sao S550 assessed I road,saptic.seater 372M 4 Skagit Cc Assessor
Skagit Bellingham Ch 2%.13 60 S". 900 S1,665 assessed I mater,road 2 Skagit Cc Assessor
Skagit Sinclair 1. 36.34 1300 S 131 assessed h road I Skagit Cc Assessor
Skagit Guenes Ch 114.52 410 S217.000 S-195 assessed 1 0 Skagit Co ft"zsor
Skagit Padilla Say 5-15 1IT50 SZi.210 $I? assessed h 0 Skagit Cc Assessor
Sk-agi t Ro--,W-i o St. so ifim ses.850 S54 435055fod 1 d Skagit Cc Assessor
Skagi t, Roswio St. 0-65 ?9 S13,100 SS49 assessod h road,"&tQr,svptic. 29600 4 Ska%t Cc Assessor
*agi t Rosario St. 0.33 163 S34.650 S550 as-sessed I eas"ent,septic,stater 41100 4 Skagit Cc Fissessor
Skagit Bel I i ngham Elay 2.1s Ijw S57,290 S573 assessed 1 water.septic,road 3 Skagit Cc fissossor
Skagi t. Sinclair 1. 30-8 1500 S77,WO S51 assessed h 0 Skagit Cc Assessor
Skagit Bellingham Smay 2.72 L*22 5-6-1.990 S533 assessed m road 930M 4 Skagi t Cc Assessor
Skagit Rosario St 0.39 2-50 S300.000 S1,200 assessed I soptic,"ater,road 3 Skagi t Cc Assessor
Skagit Bellingham Ch 14.26 17W -C-29. wo S17 assessed road I Skagit Cc Assessor
Skagit Sinclair 1. 9.02 GIGO S36,930 SSG assessed h road,mater,septic 3 Skagit Cc Assessor
Skagit E Guamws Ch 10.5 4010 531.500 S?19 assessed h septic.6&ater 2 Ska9z t Cc Assessor
Skagit Gue"es Ch 0.6 L30 $122,100 "39 assessed 1 septi c,"atQr-, road 67700 4 Skagx t Cc Assessor
Skagit Si ncl ai r 6 770 ---.-5,700 S98 assessed h road.water.septic 90800 1 Skagit Cc Assessor
Skagit Bellingham Ch 18-6 'rA- S-t6,WO S103 assessed h road I Skagi t Cc Assessor
Skagit Sinclair 1. 29.05 3800 S139,400 S16 &ssassod h road I Skagit Cc Assessor
Skagit Bel I i nghan Cb 26.41 11100 S106.1100 S?G assessed " road I Skagit Cc Assessor
Skagit Padilla Say 11 5100 51?,GW S35 assessed 1 0 Skagx t Cc Assossor

A.1 Puget Sound Waterfront Property Survey

FRONT PRI CE PPRI CE
COUNTV SIM ACRES FEET PRICE FRONT FrVPE ORNK FERWRES BUILDINGS SCORE SOURCE

Skagit Burrows Bay 206 1900 S2, 110,000 5 1, III assessed 1 road,watar,saptic 3 Skagit Co Rssossor
Skagit Bollingham Bay 1.35 66 S20,860 $316 assessed h road I Skagit Co Assessor
Skagit Si ncl ai r I . 1.4-19 I'm S35, -100 :;25 a3s"sed m 0 Skagit Co Assessor
Skagit Secret Harbor 10.75 11100 S26,800 S19 assessed h 0 Skagit Co Assezsor
Skagit Bellingba" Bay 3-1 -7,030 S37.-qOO S16 assessed I 0 Skagit Co Assessor
Skagit Padilla Bay 13.60 1-M S20,300 SIS assessed I 0 Skagit Co Assessor
Skagx t Guettas Ch 1?.15 7'913 5514,500 SGSI assessed I road I Skagit Co Assessor
Skagit Bol 1 i ngha" 8@ay 3-53 187 "2,090 S139 assessed h water,septic,road 2 1700 I Skagit Co Rsserssor
Skagit Burrows Bag 3.92 1.15 S99,000 SIM assessed h 0 Skagit Co Assessor
Skagit Ship Harbor ?.S 1100 $15,000 S38 assessed h mater,septic,road 3 Skagi t Co Assessor
Skagit Burrows My 53.6 1400 $06.510 S62 assessed I 0 Skagit Co Assessor
Skagit Ship Harbor 8-86 SW Sl?,700 $35 assessed h road I Skagit Co Assessor
Skagit M. Fidalgo 1.68 113? S210.000 $1,066 "arkot I 0 Caldwell Banker
Skagit Gus 1-81 500 $71,000 51-42 narket hm 0 Caldwell Banker
Skagit Coronet My 0.90 I." S60.000 $-II? markot 0 Skagit Co. "LS
Skagit Fidalgo 41.00 Z25 S99,000 S-"O narket 0 Skagit Co. 11LS
Skagi t E. Fidalgo 0.22 96 S65.000 56?? narkot he 0 Caldwell Baonkor
Skagit Allen 1. 20-00 660 5220,000 S-333 n-dw-ket h" 0 Cal duel I Banker
Skagit Allen Island 10-50 ?0O $135,000 $193 market 0 Skagit Co. "LS
Skagit M- Fidalgo 0-60 1.33 S74,500 S%0 narket he 0 Caldwell Banker
Skagit M.Fidalgo 1.23 135 S134,500 S72? narkot M 0 Col dmel I Banker
Skagit Sinclair 0.2-1 7S S10,000 S133 market hm 0 Caldwell Banker
Skagit Quiet Cove 0.50 100 S7?,OOO SM "arkot I road,mater,tidolarods 3 Fidalgo Realty
Skagit S. Gue"es 2.30 100 S819,000 St"O "ark-pt h 0 Caldwell Banker
Skagit HE Gueries 1.00 Z" S110,000 S'500 market M 0 Cal dual I Banker
CY% Skagit M. Fidalgo 0.55 ISO S-58,000 S38? market H 0 Caldwell Banker
Go Skagit Pear Tres Cowe, 8-00 ?00 S85,000 S 121 nark.;.t 0 Skagit Co- M-S
Skagit Fidalgo 20-00 1100 S200.000 $182 narket 0 Skagit Co. M-S
Skagit M. Fi dal go O-?8 184 $111.500 $622 narket hm 0 Caldwell Banker
Skagit 060104 Beach 0-10 100 52?.066 S170 "arki",
.d 2 Caldwell Banker
Skagi t Sinclair 4q.00 MO =4, 500 S 160 narket n 0 Caldwell Banker
Skagit Dec. Pass 0.69 ISO S75,000 S5300 market he tidelands I Coldmell 84nker
Skagi t Si ni I k Bamj 0.21 100 S65,000 $650 narkot h mater I Southside, Realty
Skagit M. Fidalgo 2-09 100 S78,500 S785 market hm 0 Coldmall Son or
Skagi t Fi d-A- go 8.15 ?b-0 51?7,000 S236 mark,&-t 0 Skagit Co. MLS
Skagit Sinclair E. 82.00 13:36 5200,000 5150 market h" 0 Coldmell Banker
Skagit Cypress 5.00 330 549,500 $150 mark-i-t h 0 Caldwell Banker
Skagit GLMMOS I. -4.00 -100 S136,500 S-311 market I beach 2 Fidalgo Realty
Skagit Skyline 0.25 :30 S80,000 S1,000 flarket I extensive Prep. --hwelopnont 2 Skyline Realty
Skagm t Deception Pass 0-30 130 575,000 S1,7? ro-arket h Mater I Southside Realty
Skagit Guenas 114-00 5-- IS $3,000,000 S561 r"rket ho timber 1 Caldwell Banker
Skagit Burroughs Ray 8.00 375- "25.000 S-?87 market h 0 Southside Realty
Skagit Fidalgo 86.00 22@39 S590,000 S-265 market 0 Skagit Co. MLS
Skagit Fidalgo 0.30 -10 S60,000 S1.500 narket 0 Skagit Co. ILS
Skagit Sinclair 0.22 75 510,000 -5 133 narkot h" 0 Coldmell Banker

RESPONSE TO COMMENTS

1. GENERAL

Public comments on the appendix titled "The Economics of Salmon Farming" are grouped
for response under appropriate headings of that document. Before turning to specific
discussion, though, a comment is in order concerning the scope of the economic study
that was defined by responsible Washington State officials and represented to the public
in the lead paragraph of the executive summary.

"The report examines.three economic issues arising from recent growth in Washington's
salmon farming industry. The first issue is potential gains in output, income, and
employment to the economies of the state and to selected counties. 'ne second is impact
on revenues and expenditures of state government, and the third is implications for real
estate values of various (externally provided) assumptions concerning visual impacts of
salmon farming facilities."

The report examined neither the universe of policy issues elsewhere addressed in the EIS,
nor the subset of those issues amenable to economic analysis or comment. Hence, the
reader is referred other sections of the EIS for discussion of the effects of
environmental wasteloadings and fish disease; consequences for sport and commercial
fishing, marine recreation; and economic effects of public perception concerning
environmental quality. An article by James A Crutchfield (Appendix L) provides an
overview of the Washington salmon farming issue from an economic as well as policy
perspective.

2. INPUT-OUTPUT ANALYSIS (Sections 11, 111,

Regional input-output analysis was conducted according to theoretical principles
articulated by Harry Richardson (Input- Output Analysis and Regional Economics,1972)
and empirically implemented in the US Forest Service Implant System. Both these works
are cited in the appendix and are generally familiar to practitioners of regional economics
in the Pacific Northwest. Results were expressed in terms of gross revenues (in total and
by sector), household incomes, and employment. Independent estimates were provided
for Washington State and each examined county.

One comment alluded to the need for independent replication of these results.
Crutchfield provides a partial basis for comparison. Crutchfield reported 7 - 10 direct
employees for a 500,000-pound facility, or 14 - 20 direct employees per million pounds
of production. 'ne representative (one million pounds sold at $5/lb) fish farm used to
calculated input-output results for this appendix assumed 20 fish farming employees.
Additionally, the representative facility assumed 8 employees in an associated hatchery,
and 5 administrative employees for the managing firm. It is unknown whether Crutchfield
included either of these components in his estimate. Crutchfield also estimated that
between 140 and 200 full-time jobs would directly or indirectly result from a 5 million
pound industry selling its product at $4/lb. The reported low range estimate of 257 jobs
best corresponds to Crutchfield's conclusions, by eliminating $1/lb of net profit from the
regional income account.

Responses to specific comments on input-output analysis are as follows.

1. Use of constant ratios (expenditures per dollar of revenue, etc) is standard
procedure in the input-output literature, as well as being reasonable in the current
situation where impacts are small relative to the magnitude of effected state and
county economies.

2. Independent county models do exist, as di-scussed above, those being derived
from the implant system.

3. State impacts were separately calculated from an independen t state model, not
aggregated from county results.

4. The local economic impacts of import substitution (replacing imports with locally
produced fish) are essentially the same as for export of the same volume and
value of product.

5. Whether hatchery location, and thus employment, occurs in the same county as
the fish farm will vary in the individual case, with effects on county but not state
results. While collocation was assumed in this assessment of overall industry
development, case specific information should be introduced in the evaluation of
specific sites. The same comment pertains to case specific variations from the
representative facility in terms of production volume and/or facility mix (hatchery,
farm, administrative unit)

3. FISCAL IMPACTS (Section V)

The analysis of fiscal impacts relied on the results of input-output analysis and published
data on five categories of state revenue and three categories of state expenditure. For
each category, fiscal factors were calculated that represented the relationship between
state revenues and costs on the one hand, and input-output results (output, income or
employment) on the other. Multiplication of fiscal factors by these input-output results
produced the reported state fiscal results. Local government fiscal impacts, as well as
site specific salmon farming costs, were too diverse and variable to permit similar
estimation.

The conclusion was ambiguous. That is, depending on the fiscal factors used, and the
input-output results to which they were applied, the Washington State government came
out ahead or behind on its own fiscal account.

Concerning lack of emphasis on the fiscal analysis that was done, the executive summary
reflects the ambiguous conclusion on cost account as follows:

"These economic impact results provided the basis for estimates of state fiscal
(revenue and expenditure) consequences. Depending on the economic impact
values used and the method of relating economic impact to fiscal consequences,
salmon farming would [annually] contribute $36 - $2.26 million to state revenues
and $1.08 - $1.48 to state expenditures."

A reading of this paragraph should adequately alert the reader to the reports conclusion
that, depending on method of calculation, the state government account comes out either
ahead or behind.

4. PROPERTY VALUE (Section VI)

DATA: Primary data on waterfront property was collected from county assessors, real
estate offices, and multiple listing services. There were 335 listings in total and at least
41 from each county. Descriptive summary tables indicated the range of variation in
front footage value between counties and property classifications (high/low bank, degree
of development). That data is useful only for its intended purpose and should not be
regarded as a general purpose data base for other purposes.

One commentor found Skagit County values different from her experience. I would need
to examine both sets of data to evaluate this difference.

STATISTICS: A multiple regression equation was estimated in order to isolate the
effects of known variables (county, bank type, degree of development) from residual
variance. The first step in determining impacts on property values was to assign all
residual variance to aesthetic quality. This procedure maximized salmon farming impacts,
relative to any apportionment of residual variance between aesthetic and other value
determining factors.

This simple statistical procedure for producing high range results was adopted over the
more sophisticated hedonic pricing approach. In other environmental resource evaluation
applications (such as sport fishery evaluation) hedonic pricing is used to directly
determine resource value impacts attributable to resource characteristics. An example
would be the use of angler success rates as a partial determinate of total angler day
values. Available financial resources and data fell far short of that required by the
hedonic pricing approach.

INTERPRETATION: The only direct information on the actual effect of salmon farms
on property values was a cited appraisers report (Appendix K) which concluded that
"floating net pens have no effect on upland property values in the areas studied [Peal
Passage, Mason County, and Rich Passage Kitsap County]." Assumed losses were
nevertheless included, as discussed below.

5. BENEFIT-COST ANALYSIS (Section VII)

Benefit-cost analysis was performed in terms of statewide annual gains and losses. These
were derived from the results of the foregoing estimation procedure by application of
factors reported in Table 7.2. One of these adjustments factors was the 8 % real interest
rate (financial rate less inflation) used to convert the asset value of waterfront property
to annual terms.

Salmon farming impacts on these asset values were included as costs, in spite of the
above assertion of no discernable effect. This was accomplished by introducing into the
benefit-cost analysis two additional factors reflecting the assumption that a defined quality
index would decline from 10 to 20 percent over 5 to 10 miles of shoreline per site. This
procedure was adopted to allow readers prepared. to assume adverse impact to readily
examine the economic implications of their assumptions. Considerable emphasis was given
to the fact that such reader provided assumptions were necessary to give meaning to this
procedure.

One commentor suggested that a better alternative to this quantitative approach would
have been to rely on qualitative judgment of all identified impacts. The main body of the
EIS, to which this appendix is supplementary, should provide the basis for such
judgement.

APPENDIX F

PERMITS THAT MAY BE REQUIRED FOR
AQUACULTURE PROJECTS

Permits which may be required for an aquaculture project.

Federal Permits Issuing AggnQ@
Section 10 Permit Army Corps of Engineers

Navigational Markings U.S. Coast Guard

Marine Mammal Protection Act National Marine Fisheries Service
Exemption

State Permits
Aquatic Land Lease Department of Natural Resources

Hydraulic Project Approval Department of Fisheries or Wildlife

Statement of Consistency with Department of Ecology
C7.1
Coastal Zone Management Act

Water Quality Certification Department of Ecology

Water Quality Standards Modification Department of Ecology

National Pollutant Discharge Department of Ecology
Elimination (NPDES) Permit

Aquacultural Identification of Department of Agriculture
Private Sector Products

Registration of Aquatic Farmers Department of Fisheries

Fish Disease Control Department of Fisheries

Shellfish Certification Department of Health

Finfish Import/Transfer Department of Fisheries

Local Pgrmits
Shoreline Substantial Development County or City

APPENDIX G

VIRAL HEMORRHAGIC SEPTICEMIA

Viral hemorrhagic septicemia (VHS), also known in Europe as Egtved disease (named
after a town in Denmark where the disease was first recognized), is an acute to chronic
disease, principally of rainbow trout, caused by a virus of the same name; i.e., VHSV.
There is much concern in Washington State and North America because of the isolation
of this virus here in 1988. Some people speculated that VHSV was introduced into
Washington as a result of aquaculture and sea-water net-pen activity with Atlantic
salmon. The scientific community has found no evidence to support this speculation.
This paper presents information about VHSV, how and where it was found in North
America and some suggestions as to the mode of introduction and potential impact.

BACKGROUND AND BIOLOGY OF VHSV

VHS is caused by a rhabdovirus. It occurs in continental Europe in the countries with
intensive salmonid culture to include Denmark, France, Germany, and Italy.
Observations of VHS have also been made in Poland, Czechoslovakia and is suspected
to be in Russia (Wolf 1988). The disease was observed in a trout farm in Norway in
the mid-1960s where rainbow trout had been imported from Denmark. The disease was
eradicated from the farm and has not reappeared in Norway (Hastein 1968 and personal
communication). VHS has never has observed in Finland or Great Britain.

The virus is very similar in its characteristics to a virus which does occur in North
America -infectious hematopoietic necrosis virus (IHNV). They both cause acute to
chronic mortality in rainbow trout with fry being the most seriously affected and having
the highest mortality. Species shown to be naturally infected by VHSV include rainbow
trout, brook trout, whitefish, grayling, and pike (Wolf, 1988; Rasmussen, 1965). While
researchers have been able to induce VHS in Atlantic salmon by an unnatural challenge
(interperitoneal injections) they have been unable to induce disease by a water-borne
challenge in the laboratory (Rasmussen 1965; deKinkelin and Castric 1982). In one
challenge, deKinkelin was able to demonstrate in the laboratory the presence of VHSV
in Atlantic salmon fry after exposure; however, the fish did not become diseased nor
were the Atlantics able to subsequently shed the virus and infect sentinel rainbow trout
in the same tank (deKinkelin and Castric 1982). VHSV has never reported to have
been found in hatchery or wild Atlantic salmon stocks even though extensive surveys and
certifications have been performed. Coho and chinook salmon have both been
demonstrated to be resistant to VHSV infection by both a water-borne challenge and
interperitoneal injections (deKinkelin et al. 1974; Ord 1976).

The manner in which viruses are isolated and broodstock are tested is also of interest.
For salmon and trout broodstocks in Washington or stocks outside Washington wishing
to enter the state, rigorous testing procedures are required. Samples of gonadal fluids,
as well as a kidney and spleen are taken from a statistically significant portion of the
population. The samples are assayed in a living tissue culture system using standard
methods (Amos 1985). Personnel and laboratories conducting these certifications are
inspected and approved by Washington Department of Fisheries personnel. Our staff and
the Olympia Fish Health Center (USFWS) were using these standard techniques when
they isolated VHSV in Washington state.

The known method by which VHSV is transmitted from fish to fish is via the water or
by ingesting infected material. This method of pathogen transmission is known as
horizontal transmission. This process also takes place with IHNV. Another method by
which virus may be transmitted is via the eggs or sex products. During spawning of
susceptible species (rainbow trout) VHSV and IH14V have been found to be present with
the sex products. When pathogens are transmitted from the parents to the offspring via
the eggs or sperm, this is referred to as vertical transmission. True vertical transmission
implies transmission of thepathogen within the eggs. This has never been demonstrated
to occur. We have observed a phenomenon with IHNV which is more appropriately
described as "egg-associated" virus transmission in which either through surface
contamination or possibly within the egg virus subsequently causes infection. These
observations were made on eggs incubated in well water so the assumption was made
that the known infected parents were the source, of the virus which infected the eggs.
The distinction between transmission on the egg or within the egg, is important as the
surface of the egg can be exposed to disinfectant while the inside of the egg cannot be
disinfected. Even though egg-associated transmission of IHNV has been observed, it is
not a common event and has been observed only in sockeye salmon and rainbow trout.
VHSV has never been observed as being egg-transmitted.

ISOIATION IN WASHINGTON STATE

Routine broodstock screening for virus in chinook salmon at Glenwood Springs (Orcas
Island) and coho salmon at the Makah National Fish Hatchery (Neah Bay) yielded
replicating agents which were identified to be VH:SV. This was a remarkable find in that
VHSV had never been found previously in North America. Furthermore, contrary to the
exisfing literature, VHSV had never been described in coho or chinook. As was
previously stated, researchers in Europe were unable to induce infections in chinook or
cobo.

As a result of those isolations, an action plan was put into effect by the Washington
Department of Fisheries. All fish and eggs at the affected hatcheries were destroyed and
disposed of in a sanitary manner. The facilities were completely disinfected. Our intent
was to eradicate VHSV. This was consistent with state and federal regulations and
policies. Subsequent surveys and live box testing of the watersheds failed to find virus.
Testing of fish in adjacent watersheds and also of feral fishes in the marine area failed
to produce VHS virus. Because of the concern that commercial net pens might have
been the source of the virus they were examined also. Consistent with ongoing testing
and viral certification of commercial broodstocks in Washington, they were all negative
for virus. In addition to testing, a thorough reiriew was made of introductions of fish
from Europe. We were unable to find documentation of introduction of fish from VHSV
endemic area into Washington. Since 1985 when commercial imports of fish came under
the Washington Department of Fisheries' jurisdiction, very few imports of eggs have come
to the state. These eggs have come from Norway and Finland, where VHSV is not
known to exist. Furthermore, the broodstock which provided the eggs were carefully
scrutinized. Records of the Washington Department of Fisheries and those maintained
by customs inspectors and USFWS inspectors are in agreement.

Virus inspections of 100% of the adult salmon returning to Glenwood Springs and the
Makah NFH as well as extensive screening of public and private salmon stocks failed to
isolate VHSV in 1989 broodstock with one exception to date (1/5/90). Coho salmon
adults returning to spawn to the Lummi Island Sea Ponds (saltwater rearing ponds
operated by the Lummi Tribe) were shown to be infected with VHSV. Only one pool
of samples was demonstrated to contain virus which likely represents only one but not
more than five individuals. As in 1988, this isolation was made from adults immediately
leaving the straits which again suggests that infections took place in the Pacific Ocean/
Puget Sound. Though WDF efforts to eradicate this virus from the Glenwood and
Makah facilities appears to have been successful, the source or opportunity for infection
seems to persist.

Yet to be resolved is the source or the reservoir for infection of VHSV in Washington
state. All the hatcheries are in proximity to the Straits of Juan de Fuca and all
hatcheries are very close to sea water. The data suggests that the adult salmon were
infected as they entered the hatcheries and were, therefore, infected in saltwater.
Potential sources of infection could be: (1) unknown carrier fish in the ocean, which are
circurnpolar in nature which came in contact with or were ingested by the salmon; (2)
introduction of carrier fish or animals in bilge water discharged off the Washington
Coast; (3) a condition which has existed in our salmon stocks for many years, but below
detection level; and (4) the legal or illegal introduction of fish or fish products into
Washington which, in turn, established a reservoir in some carrier animal in saltwater.

Many questions remain to be answered such as: How is our VHSV similar/different to
European strains? Does our isolate cause disease and if so, in what species? What is
the reservoir for the virus? Research to be conducted in 1990 will address these
questions.

SUMMARY

� VHSV was reported for the first time in North America in 1989 in coho and
chinook salmon adults in Washington state in 1988 broodstock.

� VHSV isolated in adult coho salmon in 1989 broodstock returning to Lummi Bay
Ponds, a new site.

� No disease or mortality was associated with the VHSV isolations in Washington
state.

� Extensive surveys failed to show the source of the infection.

� Infection of the adult salmon appears to have occurred in saltwater.

� No VHSV was found in fish from commercial net-pens.

� VHSV has never been reported to occur in Atlantic salmon.

VHSV has never been demonstrated to be transmitted via the eggs.
No evidence was found which indicated that import of eggs by public, private, or
Indian tribal entities was responsible for introducing VHSV.

REFERENCES

Amos, K.H., editor. 1985. Procedures for the detection and identification of certain fish
pathogens. 3rd ed. Fish Health Section, American Fisheries Society. Corvallis,
Oregon.

deKinkelin, P. and I Castric. 1982. An experimental study of the susceptibility of
Atlantic salmon fry, Salmo salar L, to viral haemorrhagic septicaemia. J. Fish Dis.
5:57-65.

deKinkelin, P., M. Le Berre, A. Meurillon, and M. Calmels. 1974. Septicemie
hemorragique virale: demonstration de 1'etat refractaire du saumon coho
(Oncorhynchus kisutch) et de la truite fario (salmo trutta). Bull. Fr. Piscic. 253:166-
176.

Hastein, T., G. Holt, and J. Krogsrud. 1968. Hemorrhagisk virusseptikemi (Egtvedsyke)
hos regnbueorret i Norge. Nord. Vet. Med. 20:708-711.

Ord, W. 1975. Resistance of chinook salmon (Oncorhynchus tschawytscha) fingerlings
experimentally infected with viral hemorrhagic septicemia virus. Bull. Fr. Piscic.
257:149-152.

Ord, W., M. Le Beffe, and P. deVinkelin. 1976. Viral hemorrhagic septicemia:
comparative susceptibility of rainbow trout (Salmo gairdneri) and hybrids (S. gairdneri
X Oncorhynchus kisutch) to experimental infection. I Fish Res. Board Can. 33:1205-
1208.

Rasmussen, C.J. 1965. A biological study of the Egtved disease (INUL). Ann. N.Y.
Acad. Sci. 126:427-460.

Wolf, K. 1988. Fish viruses and fish viral diseases. Cornell University Press, New York.

APPENDIX H

NORWEGIAN AND BRITISH COLUMBIA INFORMATION

ICES 1988 PAPER C.M. 1988/F:11

LENKA - A NATION-WIDE ANALYSIS OF THE SUITABILITY OF THE
NORWEGIAN COAST AND WATERCOURSES FOR AQUACULTURE.

A COASTAL ZONE MANAGEMENT PROGRAM

by
Tom N. Pedersen". Jan Aure3', Bjorn Berthelsen",

Siri Elvestad", Arne S. Ervik" and HAkon Kryvi2l.

1) 2)
Institute of Marine Research County Environmental
Division of Aquaculture Protection Department
C. Sundtsgt. 37 Walckendorffsgt. 7
N - 5004 Bergen N - 5000 Bergen

3) 4)
Institute of Marine Research Ministry of Environment
Box 1870 - Nordnes Department of Natural Resources
N 5034 Bergen P.b. 8013 Dep.
0030 Oslo 1

ABSTRACT

A coastal zone management program called LENKA was started in 1987
and is to be terminated in 1989. The aim is to make an efficient and
standardized tool for coastal zone planning, which, pertaining to law, is
the responsibility of the county and municipality. The program aims to be
beneficial for both the environment and for the fish farmers. Considera-
tion is taken to all important existing utilization and judicial aspects
connected to the Norwegian coastal waters. This is done by a systematic
collection of all available data. systemized in such a way that they are
available for future planning.

A model for the evaluation of the holding capacity primarily for cage
culture based on both oceanographical and topographical criteria is put
forth. The coast is divided into three categories of recipient based on
topography. A central clue in this model is the evaluation of indices for
the quantity of aquacultural activities (measured as organic deposits into
the recipient) one may have per square kilometer in differently categor-
ized recipients.

INTRODUCTIO14

Aquaculture in Norway is based on salmon and rainbow trout. The growth
of the industry has been rapid, with an almost two fold production

increase every second year. The total production this year is expected to

be about 80 000 metric tonnes, but the continued growth is expected to

be slower. Up to now. the limitation has mainly been on the number of

smolts available, this situation is now reversed. partly due to the libera-

tion of smolt production permits.

There is a keen interest in the potential of cultivating marine species,
especially halibut and cod. Much effort is put into solving the problems
of the rearing of juveniles, and this seems to be solved for cod and
turbot. Other species of interest are arctic char. wolf fish. eel and lump
fish. Some shellfish are being cultured. mostly blue mussels and oysters.
in addition to experiments on scallops. Also and some experiments on

ranching of lobster is being performed.

The main asset in Norway for this rapid growth in the aquaculture
industry has been the access to vast amounts of water of good quality,
both fresh water and salt water. Space and water quality was not a
limiting factor to begin with, but is becoming so now. So far, the only
measurement available in the assessment of holding capacity, is the

amounts of organic waste from mariculture.

There is a need for a planning tool. consisting of directions and know-
ledge, to aid the development in such a way so that a high productivity
is maintained at the same time as conflicts with fisheries, conservation
interests, leisure activities and other utilization is kept low. The tool will

have to be standardized and rational.

Both county and local municipality have the need for a plan on how to
utilize the marine resources. The county plan is a guiding one, the
judicial binding is not persistent before there exists a plan approved of
by the local municipality.

This paper, written by the expert group on marine environment, presents
the biological and oceanographical aspects of the project.

This is a description of an ongoing project where the guidelines are not
yet completed. As we believe that there is a considerable interest in

these matters, we find it appropriate to give some information on the

project at its present state.

THE PROJECT

The project is a cooperation of three ministries, the Ministry of Fish-
eries, the Ministry of Environment and the Ministry of Local Government

and Labour. Its name LENRA is a Norwegian abbreviation meaning: A

Nation-wide Analysis of the Suitability of the Norwegian Coast and

Watercourses for Aquaculture.

The project aims to :

S To contribute to a continued positive development and growth of
the aquaculture industry with minimal conflicts with other

utilizational and conservational interests.

* To contribute to the county and municipality planning in the

coastal areas and watercourses.

8 To contribute to the siting process of aquacultural activities.

The project is a planning tool, and not a plan in itself. Further, it does
not aim at the site as a working level, but handles larger areas as the

base unit, later referred to as LENKA zones.

Project organization

Figure 1 gives a schematic picture of the project organization.

The development of the working methods is done by the three expert
groups and the secretariat at the Ministry of Environment, while the
gathering of data. map work etc. is to be performed by the county
project organizations. The three expert groups are placed at the institu-
tions with the relevant competence. The group working with watercourses

Head of project

1E
3 expert Project Contact persons
groups working group to other
F ministries and
Marine environment other
Water courses 7 authorities
Maps and computing Secretariat

cou@ty project
11or2anization (17)

Figure 1: The LENKA - project organization. The head of the Project consists
of the Secretaries General from Lhe Ministries of Fisheries and
Environment. The project Workini; group has 3 members from the
Ministry of Fisheries, 3 from the Ministry of Environment and 1
member from the Ministry of Local Government and Labour. The
Secretariat is placed at the Ministry of Environment.

is placed at the Directorate for Nature Management. Trondheim. The
group working with maps and computing: is placed at the Norwegian
Hydrographic Service, Stavanger. The two latter's part of the project will
not be presented in this paper.

The group working with the aspects concerning the marine environment is
placed at the Institute of Marine Research, Bergen. In addition. the group
also has members from other institutions, such as the County Environ-
mental Protection Department, the Ministry of Environment and Nordland

College, Bodo.

The project has a total cost of 40 million NOK spread over three years.

THE MAIN WORKING PROCEDURE
71

C'
( A17

The main working procedure of the project is shown in figure 2 (next
page).

j
Coas@al zone
partitioning

Typification

Environ Existing on) Infra- g7al
ment exploitati structure S ci
areas

r,17@Z=ones of Fno@@
C3 Sorting procedure

Zones not to be
Decision of further 7 further ----- ON-
investigations investigated

furth@r
investigations
i +
Environ- Infi-a- d Special-
ment 'on) structure areas
_4 * .4
assessme I
9 Capacity nt

Need for further
investigations
7oais
p ;ta
ti
on

of f Dher__.o
.7ions

E -sting
xploi
Te tati
Cap;
ss
,7a

Figure 2 The main working procedure of the LENKA project.

ZONE PARTITIONING

In order to be able to deal with our 57 000 km long coast line in por-

tions of manageable size, a partitioning is necessary. The principle of the

partitioning is that each major water volume should be handled separate-
ly. Roughly, the coastal zone is divided into smaller areas (LENKA-
zones) each being either an archipelago, a fjord, a large sound or an

open fjord basin. The smaller areas will reflect the water bodies capacity
to handle the organic loadings received from both aquaculture and other

sources.

In order to separate the water volumes, the borders should to a large

extent as possible follow land. An exaTriple of how this looks like is
shown in appendix 1, where the partitioning of County Hordaland (where
Bergen is situated) is shown.

TYPIFICATION

Typification of zones is a registration of the environmental properties of
the area. In this project the aim is to collect and systemize the data that
already exist. For some parameters the data will be scarce. This is taken
into consideration, and follows the description of the area. The compiled
information will be transferred to maps,. the result being a visualized

presentation of the environmental properties of the marine environment.
Similarly, this is done for the other three main groups of parameters,
Existing exploitation. Infraatructure and Special areas.

The following parameters and their significance were used for the typi-
fication of the zones. These are the major environmental parameters that
have influence on the utilization of the coastal areas for aquaculture. We
would like to note that we do not consider any technological devise that
frees the farming installment from the marine environment surrounding it.
This is mere a question of economy, and will not be considered in this

project.

ENVIRONMENTAL PARAMETERS USED FOR TYPIFICATION OF ZONES

Pollution :

The point in this connection is that the contamination of the environment

effects the health or the marketability of the fish raised in these waters.

Also, we distinguish between two categories of pollution-, toxins as one
kind and organic loadings as another. Most important are the massive

outlets from industry and agriculture. Some areas are severely polluted by
heavy metals and toxins from specialized industries. In addition there are
several smaller sources of various kinds of pollution with a more or less

restricted effect on the marine environment.

Temperature :
When considering temperature conditions in Norwegian waters, low
temperatures is the main hindrance of aquacultural activities, though
there are some problems with too high summer temperatures in some
parts of the country. Of interest are the extreme temperatures occurring
within a time span of 5 - 6 years (our definition of frequent). Areas
reckoned as unsuitable for aquaculture have regular long periods, that is
6 weeks or more, with temperatures below zero centigrade. Measurements

ought to be taken at depths of 2 to 5 metres.

Ice cover :

Of interest are the areas covered with ice at least every five years.

Exposure :

The actual parameter here is wave height. though current velocity also is

part of the exposure problem. Current in itself only occurs as a problem
locally, but infers on the wave height. Suitable areas for cage culture is
where the wave height does not exceed 2 m. For wind to generate such a
wave height, a stretch of 10 km open water is needed. Here we would
like to add that the general development of the aquaculture industry in
Norway has been towards more robust cage constructions, with cage
systems being able to stand up to wave heights of up to both two and

three metres.

Depth conditions

The depth required under the cages is dependent on the current velocity

to ensure that the wastes from the farm is spread. Also this is a way to

avoid possible eruptions of hydrogen sulf[de gas from the sediment that
often accumulates under the cages to reach the fish in the cages. As a
general rule we have set 20 m depth to toe a minimum criterion for cage
culture, with the possibility of adjustments to current velocity.

Basins

A basin is a water volume restricted from the outer lying larger water

masses by a threshold. A basin is defined ets where the depth of the basin

is at least 10 m deeper than the threshold. This water volume is sensitive
to organic loadings, causing a possible disturbance of the oxygen balance.
All thresholds shallower than 50 m have been registered.

Salinity :

The influence of freshwater causes several problems for the fish farms. A
layer of brackish water on top of the salt water resulting in a strict
stratification, may cause severe fluctuations in salinity and also fluctua-
tions in temperature. As a limit for whem the influence of fresh waiter
becomes a problem, we have put the salinity measurement to 25 ppt.

Other main groups of parameters
Under the heading of existing !@xploitati:)n we list the followings para-

meters

- effects on settlement patterns

- open air recreation life

- port development

- fisheries

- shipping traffic

- other factors

Further, there is a separate heading called infrastructure, dealing with
the particular requirements which should be met for an aquaculture

enterprise to succeed. Main parameters are

- road development

- distribution of manufactured foed

- processing facilities

- health service and guiding sex-vice

- offal disposal systems.

The last heading is special areas. conditions that might conflict with

further development of aquacultural activities. Examples here are

- spawning grounds for important fisheries species

- reserves for coastal birds and marine mammals

- others.

A MODEL FOR CAPACITY ASSESSMENT

Some imperative reservations :
With capacity we mean holdinic capacity, which is - the maximum produc-
tion limited by a non trophic resource. Or put in a simpler way, what

quantity of aquacultural activity is possible in an area without there
being damage caused to the environment. This is measured as deteriora-
tion due to organic overloading causing eutrophication, oxygen depletion

a.s.o..

This method of capacity assessment of LENKA zones is based on the
emphasis of two main considerations :
1) the environmental impact from mariculture
2) the marine environment's impact on the cultured organism.

It is by no means possible to give exact values on what loadings from
mariculture are acceptable, that is, how much organic waste from maricul-

ture is possible without any negative influence on the surrounding
environment. Some general recommendations are given in the State Pollu-
tion Control Act, the entire aspects are being dealt with by the Ministry
of Environment and the State Pollution Control Authority. The total

environmental impact from fish farms will manifest themselves several
years after the farm has started production.

To be able to assess any capacity for aquaculture, one has to take into

consideration the contribution from all major sources of organic loadings.

Elements of the capacity assessment :
Many parameters affect the capacity assessment. Not only the above

mentioned parameters are of importance. The LENKA - project takes into
consideration the elements shown in figure 3. and the working procedure
is shown in figure 4. As is shown, there are two main aspect in the

Local distribution,

Conditions
economy a.s.o. Ohl.

Need for investments
Opt- Permit decisions (size and
and enterprizes distances), technology,
species a.s.o.
Other exploitational int.
Occupied Risk for expl.:
areas conflicts

capacity
vailable gross

Available Available
capacity for areal
org. ioadings capacity

Figure 3 Elements of the capacity assessment.

capacity assessment. One aspect is the evaluation of the capacity for
organic loadings in water body MENKA zone). This is done by treating
the zones properties as a recipient for organic loadings. The other aspect
is based on space. The water body. or more precise. parts of it, is
occupied by other activities as mentioned earlier. There exists a net area
available to aquacultural activities. One of these will set the limit to

aquacultural activities.

In our capacity assessment we have based the calculations on organic
loadings. and thereby neglected limitations set by factors such as risk for
spreading of diseases. use of chemicals and therapeutica etc. There exists
a veterinary regulation on distance betwe-en farms. this is set to 1 km.
Criteria as such may be altered as the knowledge increases.
'M ents I

ta tio nal int
isk for expl
0n fl icts

The recipient capacity

Classification of coastal areas within the zones

Classification is based on topography. This again reflects the water
exchange regime in the area, as well as being an indicator of the area's

Recipient Area
A-areas C-a7rpD
C
m
km 2 17km

Capacity for Unsuitable areas (envinomk.@
organic loadings
A-areas -Shallowareas
-Temperature
-Salinity
-Ice cover
Minus: -Exposed areas
Existing fishfanns and
other sources for org. loadin S
9) r-unsuitable areas (culture
-Pollution
I'Mailable Available Available
capacity r (c.@apacity for capacity for
org. loadii gs A rg. loadin s B (org. loadings @C
OLcupied areas

Net A I B I C D
areas

Choose Regulations on
the smallest size and distances
of available capacity
for organic loadings
areal
and available.
capacity for A-
B- and C Available A B C
areal capacity

Gross available capacity
A-capacity Either capacity for
+B-capacity organic loadings
+C-capacity or available areal
capacity
C-1, he n
or,
S
_T
-S
-it
E

Po 'u'ion

it for
D
ing
g.
B
71yd @Oc,'cupied jareas

Icily
ings

Availab
r I
@@a eal -a
ase
t
a -1 le ca
coa
Ire
@,u a' it liab Ile
C
fa th C,b
v. I
for
org. n
Id 'v,I
capacity rorA
BIndC

Figure 4 Calculation of gross available capacity.

property as a recipient. An exact classification with illustrations is given

in Appendix 2.

For each recipient category A, B and C there is given an index for how

much aquacultural activity is recommended. This is expressed as a certain

production in terms of tonnes per square kilometre. Here we would like
to emphasize what care was taken before these indices were given. The

procedure was as follows :

From empirical data we were able to extract the general statement on

how large production one could have in & specific area without it causing

damage to the environment. The effects: were investigated by sediment
fauna monitoring. These levels of production were converted to production
per square kilometre. Again, based on facts about Norwegian fish farms,
these values were converted to organic loadings. expressed as oxygen
consumption, total phosphorous and total nitrogen.

As a correction factor one would have to adjust these figures for other
major outlets of organic waste. At the moment we are, together with the
appropriate institutions, giving a simplified method for estimating the
impact on the marine recipient based on key figures ready available. The
capacity is calculated in terms of organic waste, and is therefore in-
dependent on the technology being used. New technologies resulting in
reduced outlets from the farms can easily be incorporated in the cal-

culations.

Further, when capacity is expressed as production as tonnes per kM2, this
sets restriction to the size of single farms and to the total activity in
larger areas. The values are not decided yet, but the capacity will be

expressed as the following

A - categorized areas : a maximum production per (4x4) kms. but not
more than a lesser specifies quantity at a single site. A site is defined as
occupying a minimum of 1 km2. Where the recipient conditions are par-
ticularly good, and the number of sites available is restricted. one may

exceed these recommendations.

B - categorized areas : similarly as above, there is given a maximum
production per (4x4) kmz, and a lesser one at a single site.

The capacity per (4x4) kml will be in the magnitude within one thousand
metric tonnes for both A- and B - categorized areas.

C - categorized areas : these areas are basins and silled fjords, and

special care should be taken in such areas. Aure & Stigebrandt have
developed a model for the calculation of oxygen consumption in silled
fjords (Aure and Stigebrandt 1988, Stigebrandt and Aure 1988). and this

can in turn be used as a method for the calculation of capacity in terms
of organic loadings. The calculations can be done given the hydrographic
data and topographical maps.

Where there is oxygen depletion in the basin water, aquacultural activities
are not recommended in silled fjords. In basins within archipelagos one

should ensure that the water in the deeper layers of the basin does not
suffer from oxygen depletion. This means that in an area categorized as a
C grade recipient, no aquacultural activities are recommended before one
has sufficient data so that damage to the environment is avoided.

This method is dependent on a monitoring and control scheme, and this
will have to be a perpet-191 process, In this way there is the possibility

of adjusting the proposed capacity assessment, and at there is possibility
of keeping an eye on what is happening to the environment. The monitor-

ing and control schemes are not established.

The &real capacity

Unsuitable areas :

Each LENKA - zone has a gross area divided into A, B and C type
recipients. Parts of these areas may be unsuitable for aquaculture, that
is. unsuitable for cage culture for as it is practiced in Norway. Unsuitable
areas consist of environmentally unfavorable areas from both natural

conditions and as a result of man's activities. The last case is mainly

pollution. and in this case pollution of toxicants that directly affect the
fish health and marketability.

The environmental parameters taken into account are : Shallow areas,

cold water. low salinity, ice cover and exposed areas.

In addition to these unsuitable areas there are certain areas that are

bound up by other activities. Such areas are

- area already occupied by existing aquacultural activities
- nature reserves and animal protection areas (both birds and sea
mammals)

- security zones for salmonid fish
In addition areas are occupied for military purposes and for ship naviga-

tion.

Having subtracted all these areas, one is left with a net areal capacity
which can be compared with the recipient capacity. The smallest of these
will set the limit. All these calculations will. be performed by computers
as all the information is to be tabulated ready for a for this purpose

constructed work sheet.

Finally, we would like to mention the work initiated to eliminate the

interactions between wild stocks of salmon and trout and farmed fish.

The possibility of affecting the genetics and spreading of diseases has
been much debated. There is now suggested temporary protection zones
for salmonids. with a supporting research program. Further information on
this is available on request.

In addition to the names and addresses in the author list, there are a few
more names to add. If anybody should have any particular interests, the
following persons may be contacted

MAPS AND COMPUTING WATERCOURSES

Asbiorn Hiksdal Oystein )Ubu
Norwegian Hydrographic Service Directorate of Nature Management
P.b. 60 Tungasle-fta 2
N - 4001 Stavanger N - 7047 Trondheim

PROTECTION ZONES FOR SALMONIDS

Bjorn Lindgren
Ministry of Environment
Department of Natural Resources
P.b. 8013 Dep.
0030 Oslo 1

REFERENCES

Aure, J. and Stigebrandt, A., 1988 : An investigation of 30 fjords in
More and Romsdal : The field program and general oceanogra-
phical conditions (in prep.. in Norwegian).

Berthelsen, B. and Pedersen, T. N. (eds.). 1987 : Zone partitioning and
numbering of the coastal areas. LENKA - method 2 1 - 10.
(In Norwegian).

LENKA - Secretariat (eds.) 1987 : Typification of coastal areas.
LENKA - method 5 : 1 - 42. (In Norwegian).

LENKA - Secretariat and Marine Environment Expert Group, 1988
Capacity Assessment of LENKA zones. LENKA - method 9.1 : 1
- 35. (In Norwegian).

Stigebrandt, A and Aure, J.. 1988 : On the influence of topographic
factors upon the oxygen consumption rate in sill fjors basins.
Submitted to : Estuarine, Coastal and Shelf Science 88 : 01.

Appendix I

SOONESAEU 'A

TIRE
r
FEW FE J0' E
HELLIS8V
j%
LV %A 0

vt%so

OSTYOYA
10

ZI
6; V.@-v r)
SAMNA'GEM-
Z
R. N VANO
, Zo ULLEMSVANO
b!

FJ.

14ARST IN kt

cs-
by I I
'A ""Gelt
1141F 10j

SLOT TEROY S T
.4 , ft
04ATMEIrJ

IIALSNO
J/0 1b AKRAF)DADEN

L Q 3Z
oo oooo

Figure 5 Example of zone partitioning for the County Hordaland.

Western Norway.

Append 2

Division of the coastal zone into smaller areas based on assumed water
exchange rate caused by topography.

A : Open coastal areas and large fjords where depth is larger than 50 m.

Ai Open coastal areas where depth is larger than 50 m.
Size and sills are not considered.

As Large fjords where :
Length of more than 10 km, and
No presence of sills".

B Other areas with good water exchange.

BiL Open, sill - free areas as A, (archipelagos) and large
fjords as As but where largest depth is less than 50 m.
Length above or less than 10 km.
Depth" is less than 50 m.
No presence of sills.

Be Smaller fjords, bays and inlets where
Length is less than 10 km.
No presence of sills.
Depth is greater than 50 m.

Be Large. silled fjords3) where :
I ength is greater than 10 km.
Presence of sills.
Depth may be more than 50 m.

C Small silled fjords an other siUed areas (archipelagos)
Length of fjord less than 10 km.
Presence of sills".
Depth may be more than 50 m.

Examples are shown on the sketch on the next page.

A silled area is defined as an area where the inside basin is at
least 10 m deeper than the sill. Sills down to 50 m are regi-
stered.

Fjords are reckoned as shallower than 50 m when more than 60
% of the area fulfills this criterion.

Fjords and other areas with several succeeding sills is reckoned
as a "new fjord" when the succeeding sill is shallower than the
preceding one.

In sounds and basins within archipelagos with several sills, the
deepest sill is reckoned as the main entrance to the basin.

..............
Ai
A 2 L,,[Okm
IPSOm To 7 50 01,
BI
-------------------------

17 to Ktn
TV 4
B3
--17:

4 10 It's
OT

V:A

Figure 6 Examples of division into categories A. B and C.
Legend D = depth. L length, Tv = threshold depth.

BRITISH COLUMBIA
AQUACULTURE LICENSING & REGULATIONS
September 27, 1989

AQUACULTURE LICENSING AND REGULATIONS - A SUMMARY

Rationale for Provincial Aquaculture L'

A large, completely new industry that is dependent on common
property resources cannot exist in a vacuum of government
involvement. Appropriate government intervention is needed to
protect the public interest, yet ensure that the economic
benefits of the aquaculture industry accrue to British Columbia.

The size and growth rate of aquaculture has threatened other
interest groups. Thig significant new industry needs some
regulation to ensure responsible growth and development and at
the same time Provide a comfort factor for groups that feel
threatened and would block further 9quaculture development,

Aquaculture is currently administered by six different agencies
in three levels of government (Appendix 1). There are two
categories of approvals needed by an aquaculture operator. The
first category is primarily approval to locate a facility and
includes Crown land tenures, navigation compliance and zoning
compliance. The second category is approval to operata a
facility. The latter has been issued by the Department of
Fisheries and Oceans (some marine sites) or Ministry of
Environment (freshwater sites). The Ministry of Agriculture
and Fisheries, as lead agency, proposes to consolidate and
reduce the operational licences to one as provided for in
recently signed agreements with both these agencies.
Approximately 740 sites are currently authorized for
aquaculture in British Columbia (Appendix 2).

Licensing options Considered

1. Status quo.

2. No licensing by any agency.

3. Consolidated Aguaculture Operation Licence Approved in
principle by Cabinet,

Use an aquaculture "operational* licence as a registration
and use regulations to facilitate orderly industry
development.

This option provides a balance between administrative
simplicity and government intervention. There are several
advantages to this approach:

1. Consolidates operations licensing within one agency,
reducing the total number of government agencies directly
licensing industry;

2. Establishes standard criteria for a licence and eliminates
inequities in the treatment of different components of the
industry that arise from multiple agency involvement;

3. Assists industry in obtaining operational financing.
Licensing is a legal tool for identifying persons who may
have a private property right in stock being cultivated;

4. Provides a framework to develop future controls, if
necessary, to limit or restrict practices that become
problematic;

5. Replaces the Federal Department of Fisheries and oceans and
the Ministry of Environment in their industry licensing
role while maintaining their input to the licensing process.

6. Establishes an equitable and efficient basis to determine
eligibility for and issuance of sales tax exemptions;

7. Provides a systematic means of identifying all aquaculture
operators for revenue and statistical purposes; and,

8. Establishes a uniform basis for identification of bona fide
aquaculturists for other regulatory purposes including:

- transportation and transplantation for cultured plants
and animals;

- purchase of therapeutants; and,

- purchase of surplus salmon eggs from the Federal
government.

Fees will be charged to recover costs of administering the
licensing system.

Fisheries Act and Draft Aquaculture R:egulations-

Relevant sections of the Fisheries A t and the current draft of
the Aquaculture Regulations form Appendix 3.

It is important that the regulations be read in the context of
the Fisheries Act R.S.B.C. upon which they are based and with
which they mesh. Appendix 3 is therefore organized in two
sections. The first is a compilation of those sections of the
Fisheries Act relevant to aquaculture licensing, together with
a commentary. Material from the Act is given in bold type,
while the commentary is shown in lighter type. This revision
includes changes which came into force with the proclamation of
specific sections of the Miscellaneous Statutes Amendment Act
(No.21, 1907. which occurred in June 1989.--and in the
Miscellaneous Statutes Amendment Act (Ng. 2). 1989. which was
passed in July and was proclaimed by Order in Council in
August 1989, The second section is the latest draft of the
Aquaculture Regulations with commentary and explanatory notes.
Proposed sections of the regulations are given in bold type and
the commentary is in lighter type.

This draft of the regulations, and particularly the commentary,
includes revisions made on the basis of comments made on
earlier drafts by aquaculture commodity groups, the Department
of Fisheries and Oceans, the Minister's Aquaculture Industry
Advisory Council, and from discussions with Legislative
Counsel, Ministry of Attorney General.

- 4 -

APPENDIX 1.
LICENSING AQUACULTURE

CURRENT SYSTEM PROPOSED SYSTEM

A. LOCATION APPROVALS

i. Ministry of Crown Lands Where operations are on Crown
land, aquaculture licensing
- issues leases and licences will be administratively linked
of occupation for aquatic to the Crown land application
land process to minimize industry's
paper burden.

ii. Federal Department of No change
Transport

- issues navigation
compliances for all marine
and freshwater aquaculture
operations where physical
structures do not impede
navigation

iii. Regional Districts, No change
Municipalities

- may control the location,
size, setbacks, etc., for
aquaculture operations
through zoning bylaws

B. OPERATIONAL APPROVALS

iv. Federal Department of The federal government will
Fisheries and Oceans Withdraw from licensing
aquaculture under a
- licenses salmon farms in federal-provincial agreement
freshwater and marine recently signed.
locations and invertebrate
species other than oysters

V. Ministry of Environment (NOE) MOE will withdraw from
permitting commercial
- issues permits for aquaculture activities under
freshwater fish farms and the recently signed agreement,
hatcheries but will continue to control
the holding of live fish for
purposes other than
aquaculture.

v i ministry of Agriculture and
Fisheries

- registers shellfish growers Aquaculture licences will
on Crown land replace this form of
registration.

issues Bona Fide The aquaculture licence will
Aquaculturists Certificates become a prerequisite for
for tax exemption these certificates.

vii. No agency New aquaculture licences will
apply to all operations of the
- some facilities, such as industry, including those
shellfish hatcheries and operations that currently are
some operations on private not responsible to any agency.
land, fall outside all
existing jurisdictions

APPENDIX 2
BREAKDOWN OF AQUACULTURE SITES
BY TYPE AND LOCATION OF SITESI

MARINE FRESHWATER
Existing Applications* Existing Appligations

TYPE

FINFISH

Hatcheries --- --- 44 N/A

Growout 211 266 62 N/A

SHELLFISH

Hatcheries 2 1 ---

Growout 414 109

MARINE PLANTS

Growout 6 3 ---

Subtotals 633 379 106 ---

TOTAL EXISTING 739

Total existing and applications = 1,118

Includes Investigative Permits and applications for all forms
of Crown land tenures.

Based on March 1989 data.

- 7 -

APPENDIX 3
Section 1

AQUACULTURE LICENSING

A. RELEVANT SECTIONS OF THE FISHERIES AC (R.S.B.C.) AND
COMMENTARY

PART 1 - GENERAL PROVISIONS

Interpretation

1. In this Act

"conservation officer* means a conservation officer
under the Wildlife Act.

Conservation officers may require aquaculture licenceholders to
produce, upon request, the records referred to in Section 20 of
the Act.

.fishm means the whole or any part of an aquatic
animal.

"Fish" include all marine, brackish water and freshwater
animals, whether vertebrates or invertebrates, and includes
finfish, shellfish and crustaceans.

laquaculture" means the growing and cultivation of
aquatic plants, as defined in Section 12, or fish, for
commercial purposes, in any water environment or in
man made containers of water, and includes the growing
and cultivation of shellfish on, in or under the
foreshore or in the water.

The growing of aquatic plants or animals for any non-commercial
purpose (e.g. for personal use), and the holding of live
aquatic plants and animals for research or display purposes, or
in restaurants and seafood wholesale establishments for resale,
do not constitute aquaculture and will not be regulated under
this Act. The Ministry of Environment will be issuing Live
Fish Permits for these purposes.

2(3) Every officer and constable of the provincial force as
defined in the Police Act, and every conservation
officer, is by virtue of his office an inspector of
fisheries under this Act and has power to act in that
capacity in every part of the Province.

2(5) An inspector under the Fish Inspection Act (Canada)
and a fishery officer or fishery guardian under the
Fisheries Act (Canada) is by virtue of his office an
inspector of fisheries under this Act.

- 8 -

By virtue of Section 2(3) and 2(5), every officer and constable
of the provincial force (the R.C.M.P.), inspector of fisheries
(provincial), fishery inspector (federal) and fishery officer
(federal) can require aquaculture licenceholders to produce the
records referred to in Section 20 of the Act.

PART 3 - LICENSING OF AQUACULTURE FACILITIES, FISH AND AQUATIC
PLANT PROCESSORS AND FISH BUYING STATIONS

Interpretation

12. In this Part

*aquatic plant" includes benthic and detached algae,
marine flowering plants, brown algae, red algae, green
algae and phytoplankton;

This definition includes all aquatic plants except freshwater
flowering plants.

"coastal waters' includes waters in the fishing zones
of Canada adjacent to British Columbia, all waters in
the territorial sea of Canada adjacent to British
Columbia and all internal waters of British Columbia.

This definition covers all brackish and marine waters, but may
not cover inland freshwater bodies. However, Section 13(4.1)
clarifies that anyone carrying out the business of aquaculture
*in the Province" or its coastal water must have an aquaculture
licence. Legal opinion is that "in the Province" includes all
inland freshwater bodies.

oestablishmento means a place, including a place used
for the business of aquaculture, where fish or aquatic
plants are handled, processed, graded, stored, grown
or cultivated.

Note: This new definition will be added with the proclamation
of the Miscellaneous Statutes-Amendment Act (1989).

Licence Required

13(4.1) No person shall carry on the business of aquaculture
at any location or facility in the Province or its
coastal waters unless he is the holder of a licence
issued for that purpose under this Part and has paid
the fee prescribed by the Lieutenant Governor in
Council.

This section provides the legislative prohibition against
carrying on an aquaculture business unless authorized by a

location-specific licence issued for that purpose. It also
allows the Lieutenant Governor in Council to establish a
licence fee.

Application for Licences

14. Every application for a licence under Section 13 shall
be made in writing to the minister, on a form to be
supplied by him, and on receipt of the application the
minister may issue a licence.

The minister determines the format and content of an
application. It is proposed that each application will include
a development plan with different plans for different types of
operations and species or species groups. Applications
involving Crown land will employ the same development plans as
used by the Ministry of Crown lands. Those based on private
land will be simplified.

For the initial round of licence issuance, we will be advising
all active aquaculture operations of their application
requirements. These will differ depending on whether or not a
Crown land tenure is involved, development plans exist and if
these accurately reflect the current status of the operation.
A non-refundable application fee will be required for all
applications.

Form of Licences

16. A licence under this Part shall set out

(a) the name and address of the licensee,
(b) subject to Section 15(2), the location of the
plant for which the licence is issued or the area
in which the licensed activity is to be carried
on, or both;
(c) the effective date and the term of the licence;
and,
(d) other terms and conditions as the minister
considers appropriate.

Subsection (d) was added with the proclamation of the
Miscellaneous Statutes Amendment Act (1989) in August 1989.

A standard or general set of terms and conditions will be
printed on the back of and apply to all aquaculture licences.
Additionally, one or more approved development plans will
constitute the specific terms and conditions of individual
aquaculture licences. Crown land based operations which
already have development plans approved by the Ministry of
Agriculture and Fisheries will not have to complete new plans

in the aquaculture licence application process. However,
operations based on private land will have to complete and have
approved development plans before aquaculture licences can be
issued.

Transfer of Licences

17. No licence issued under this; Part is transferable,
except that in the case of a change of ownership of
the plant the minister may agree to a transfer of the
licence to the new owner.

As a matter of policy, it is proposed that the minister apply
his discretionary power to refuse transfer of licences in
f avour of the issuance of a new licence to the new operator of
an aquaculture facility.

Suspension or Revocation of Licence

18(l) Where the holder of licence issued under this Part
violates any provisions of this Part or the
regulations or a condition of a licence, the minister,
after due investigation and hearing, if a hearing is
requested by the licensee, and on proof to his
satisfaction of the violation, may in addition to all
other penalties to which the licensee may be liable,
suspend the licence and all rights of the licensee for
a period the minister thinks fit, or he may revoke the
licence.

18(2) The minister shall preside at the hearing, and shall
have the same powers as the Supreme Court for
compelling the attendance of witnesses and of
examining them under oath, and compelling the
production and inspection of books, documents and
things.

Section 18(l) authorizes the minister to suspend, "for a period
the minister thinks fit*, or revoke an aquaculture licence
should the holder violate any provisions of Part 3 of the
Fisheries Act (R.S.B.C.), any regulation made under Part 3 of
the Act or any condition of the aquaculture licence. It also
clarifies that a licence holder may request a hearing before
licence suspension or revocation takes place. Section 18(2)
outlines the minister's powers in the conduct of such hearings.

Power to Refuse Further Licence

19. Where the licence held by any licensee has been
revoked, or where it is shoien to the satisfaction of

- 11 -

the minister that a licensee has violated any
provision of this Part or the regulations or condition
of a licence, or has conducted the business of his
establishment in contravention of the spirit and
intent of this Part, the minister may, in addition to
all other penalties to which the licensee may be
liable, refuse after that to issue a licence under
this Act to that licensee or to any person for the
establishment of that licensee.

This section outlines the power of the minister to refuse to
reissue a licence which has been revoked.

Records Kept by Licensees

20(l) A person holding a licence under this Part shall make
reports in the manner and form and at intervals
specified by the minister.

This section provides for the reporting of such information as
the minister may require (e.g. production, inventory,
productive capacity, employment) at such time as the minister
may require it. This information will be used to evaluate the
performance of industry as a whole and the compliance of
individual aquaculture licensees with the conditions of their
licences. This amended wording will come into force with the
proclamation of the Miscellaneous Statutes Amendment Act (1989).

Offence and Penalties

25(2) A person who to contravenes a provision of the Part, a
regulation made under this Part or a condition of a
licence issued under this Part commits an offence.

25(3) On conviction for contravention of section 13 (1), (2)
or (4.1), the penalty is a fine of not less than $500
and not more than $10,000.

25(4) on conviction for contravention of a provision of this
Part other than section 13(l), (2) or (4.1), the
penalty is a fine of not less than $100 and not more
than $2,000.

25(5) on conviction for contravention of
(a) a regulation made under this Part, or
M a condition of a licence issued under this Part,
the penalty is a fine of not more than $2,000.

25(6) Each day an establishment is operated 'in circumstances
that constitute an offence under subsection (2)
constitutes a separate offence.

- 12

25(7) In addition to other penalties or measures taken under
this Act or the regulations, all fish or fish products
or aquatic plants or aquatic plant products, whether
processed or not, on or aboult an establishment on or
after an offence occurs at that establishment, may be
seized by a Provincial constable as defined in the
Police Act or by an inspector of fisheries and, on the
direction of the minister, be forfeited to Her Majesty
and sold with the proceeds to be paid into the
consolidated revenue fund.

This section provides for fines of $500 - $10,000 for operating
an unlicensed aquaculture facility, lesser fines of $100 -
$2,000 for violation of other relevant sections of Part 3 of
the Fisheries Act (R.S.B.C.) and fines up to $2,000 for
violations of the regulations or conditions of a licence.
Additionally, Ticket Administr-4tion Regulations and Ticket
Information Fines Regulations under the Offence Act will, upon
amendments and at the direction of the ministry, be used to
issue tickets with "voluntary penalties" in the range of $50 -
$100 for offences rather than pressing for court summonses.

Further, Subsection 25(7) permits the seizure and sale by the
Crown of fish and aquatic plants in addition to other penalties.

Regulations

26(l) The Lieutenant Governor in Council may make
regulations.
26(2) Without limiting the generality of Subsection (1), the
Lieutenant Governor in Council may make regulations he
considers necessary or advisable:

(a) for safe and orderly aquaculture; and,
(b) for safe and orderly distribution of fish and
aquatic plants.

- 13

APPENDIX 3
Section 2

B. AQUACULTURE REGULATIONS

Interpretation

1. in this regulation

"Act" means the Fisheries Act;

flaquaculture licence" means the licence referred to in
section 13 (4.1) of the Act;

glaquaculture facility" means an establishment where the
business of aquaculture is carried on;

"attachment structure$' means mollusc shell, rope,
netting, tubes and other structures provided as
substrate for the attachment of aquatic plants and fish
for purposes of aquaculture;

"containment structure,$ means net cages, net pens,,
tanks, troughs, raceways, natural or manmade ponds,
trays and other structures used to contain aquatic
plants and fish for purposes of aquaculture;

Ufin fish's means fish of the classes Agnathal
Chondrichthyes and Osteichthyes grown by a holder;

"bolder" means the person to whom an aquaculture licence
is issued;

"location's means

(a) a contiguous area of land that is owned, leased,
or otherwise lawfully occupied by a person, and

(b) areas of land whether contiguous or not that are
occupied under a single

(i) lease# or

(ii) licence of occupation

granted under the Land Act,

"Contiguous" includes adjoining or abutting parcels of land,
ie. the boundaries must touch. More than two contiguous

14 -

parcels of land may be considered a location for the purpose
of this regulation. Parcels separated by any distance, no
matter how small, are not considered contiguous and will
require separate licences.

"Leased" in (a) above includes private upland leased from its
owner, as well as land under the jurisdiction of a Federal
Port Corporation or Harbour Commission and occupied under a
lease issued pursuant to the Port CorRoration Act (Canada) or
the Harbour Commission Act (Canada), respectively.

"Land" includes land under water. Lands occupied under
authority of the Park Act are also included as are Reserve
Lands under the Indian Act (Canada).

Paragraph (b) above is included because of our wish to
grandfather several operations that: have several parcels of
land covered by a single Crown land tenure, and which are
operated as a single production unit.

"manager" means the manager of aquaculture appointed
under section (6).

Separate licence for each location

2. No person shall carry on the business of aquaculture at a
location without first obtaining an aquaculture licence for
that location.

The culture of different species or groups of species and the
operation of several types of aquaculture facilities (eg.
hatchery and growout to food market size) may be covered in a
single aquaculture licence, provided they occur within the
location defined in the licence. In such cases, several
development plans will be attached to and become part of the
aquaculture licence. This is in keeping with existing policy
and procedure regarding Crown land tenures.

Application for licence and licence renewal

3. An applicant for an aquaculture licence shall make an
application to the minister under Section 14 of the Act and,
where the application is for a renewal of an aquaculture
licence, shall deliver it at least 60 days before the term of
the existing aquaculture licence expires.

It is our intention to combine the application for renewal
with the annual report required under section 20 of the Act.
This has been the practise of the Ministry of Environment

with regard to the issuance of renewals for its Commercial
Fish Culture Permits. This is expected to facilitate timely
submission of annual reports. The combined renewal
application-annual report form will be mailed to all
licensees at least three months before licence expiry.
Licensees will have at least one month to complete and
deliver the form to the address given in the instructions
which will accompany the form.

"Delivered", in the Interpretation Act, with reference to a
notice or other document, includes mail to or leave with a
person, or deposit in a person's mail box or receptacle at
the person's residence or place of business.

Failure to deliver the application for renewal at least 60
days before-licence eKRia will result in applicants havin
to reapply for a licence; ie.-to submit an application form
and fee in addition to the annual report form and licence
fee. Thus, there will be a dollar savings for those who
submit an aRplication at least 60 days before licence eXpiry.

Staff will have two months to review and validate annual
reports, determine eligibility for Bona Fide Aquaculturist
Certificates (BFAC) and prepare licence renewals and BFAC's
for mail out. Should government workload preclude renewing
licences before the expiry date, the Common Law "Doctrine of
Administrative Necessity" would ensure that holders would be
legally able to continue under the old licence.

Term of licence

4. An aquaculture licence is valid for a 12 mouth period
from the date on which it becomes effective.

While licences are valid for one year, all licences will not
have the same effective date. Licensing of existing
operations will be spread out over as much as a six month
period, with commodity groups (salmon, shellfish, trout)
being licensed within consecutive two month periods.

New operations will be licensed as the application reviews
are completed.

This will keep licence administration costs (hence
application and licence fees) to a minimum, since fewer staff
will be required to process the approximately 800
applications which are anticipated.

- 16 -

While licence renewal and annual report forms will be
submitted throughout the year, the requirement for production
of calendar year-based production statistics still exists.
Both needs will be met by amending the format and content of
the annual report forms. It is anticipated that most data
requests not directly related to regulatory requirements will
be deleted, particularly for fin fish farming operations.
Volume and value of product, by species, will be requested
for each month. The first reportwill cover more than a 12
month period. since it will have t) provide data from Januarv
1989 to the licence effective date. Additionally. the first
annual reports will cover the first: nine months of the
licence valid period. Thereafter, the annual report will
cover a 12 month period, including the last three months of
the licence valid period of the first licence and the initial
nine months of the first renewal licence. Applicants should
therefore ensure that their-records keep track of the volume
and value of product. by sRecies, on a month-by-month basis.
Statistics regarding productive'capacity (eg. net cage or
pond volume, meters of longlines) ZLnd livestock inventory
will be requested as of a partiCU12Lr date, probably December
31st each year.

Dealing in fish or aquatic plants

S.(1) A person shall not possess, buy, sell, introduce into
the Province or transplant within the Province, fish or
aquatic plants for the purpose of carrying on the business of
aquaculture unless the person is a holder or is acting on
behalf of a holder.

This subsection clarifies that only those persons having a
valid aquaculture licence, or their agents (including
employees or brokers acting on behalf of a holder) or
independent contractors, may possess, buy, sell, introduce
into or transplant within the Province, aquatic plants or
fish for purposes of carrying on the business of aquaculture.

If a licensed aquaculture facility is placed in receivership,
the Ministry of Agriculture and Fisheries will, as a matter
of policy, treat the Receiver-manager as the holder for the
duration of the licence term. Should the licence expire
before another person has secured the right to occupy the
location specified in the licence, the Receiver-manager will
have to apply for a licence renewal. Subsequently, the
Receiver-manager will have to request that the Minister
transfer the licence to a new person once that person has
secured the right to occupy the location specified in the
licence.

- 17 -

In the more infrequent event of foreclosure of an aquaculture
facility by a financial institution, the Ministry of
Agriculture and Fisheries will, as a matter of policy,
require that the holder and the financial institution apply
to the Minister for a transfer of the licence to the
foreclosing financial institution. Permission to transfer
the licence will not unreasonably be withheld.

Holders should note that this subsection does not authorize
the introduction into or transRlanation within the Province
of fish. An Import Permit og Transplant AyRroval issued
under the Fisheries Act (Canada) are the authorizing
implements for these activities. An aquaculture licence
will, however, become prerequisite to obtaining such
authorizations since it identifies persons as being
commercial aquaculturists.

A person in the business of transporting smolts or
live-hauling salmon to a processing plant would not, for
example, require an aquaculture'licence since the person's
business is transportation, not aquaculture as defined in the
Act. However, transport companies may not transfer live fish
from one fish farm to another or to a processing plant unless
authorized to do so by the Transplant Committee and must
conform to section 9 of these regulations. The Committee is
currently devising simple guidelines to accommodate this. By
completing and signing a Transfer Permit, a holder will, in
effect, transfer to a carrier the holders authority to
transport live fish. A separate Transfer Permit, showing
source, destination, species, carrier and pick up and
delivery dates must accompany each delivery of fish.

(2) Subsection (1) does not prevent a person who has taken
the fish or aquatic plants as collateral for a loan from
seizing or disposing of the fish or aquatic plants or
otherwise realizing on the person's interest in the fish or
aquatic plants to satisfy the obligations secured by them.

This subsection was added to ensure that persons could take
possession of and sell fish or aquatic plants which are given
as collateral for a loan, without those persons requiring an
aquaculture licence. Transplant Committee Approval would,
however, be necessary for any transfer of the seized fish to
a processing plant or another fish farm.

Manager of aquaculture

6. The minister may appoint a person in the Ministry of
Agriculture and Fisheries as manager of aquaculture.

- 18

The manager of aquaculture will be named to provide holders
with a key contact in the Ministry of Agriculture and
Fisheries. The manager is the person to whom holders must
report releases of fish or aquatic plants and the results of
any recapture attempts (see section 7).

Special proviso schedules attached to aquaculture licences
will identify other instances where holders must contact the
manager of aquaculture before or within a specified time
after certain actions are initiated. For example, fish farm
operators in the Sechelt Inlet system will be required to
advise the manager before, or within one day of, initiating
the relocation of netcages to designated "emergency
relocation areas", as provided for in the Sechelt Inlets
Coastal Strategy, in the event that heavy plankton blooms
threaten to kill their fish stock.

Release and escape

7.(1) No person shall release aquatic plants or fish to
fresh or tidal waters from an aquaculture facility or from
containment or attachment structuros in an aquaculture
facility unless authorized to do so by the terms or
conditions of an aquaculture licence.

This section prohibits the release into public waters of
aquatic plants or fish from an aquaculture facility unless
authorized to do so by a term or condition of an
aquaculture licence. The provision for such an authorization
is made because it is possible that deliberate release may
be desirable in certain very specific circumstances. For
example, it is already acceptable practise to release into
public waters, for stock enhancement purposes, salmonid
smolts raised in private sector hatcheries. Currently, the
actual release may be conducted by government personnel but
it is possible that commercial aquaculturists may effect
releases in the future. Holders must ensure that they have
obtained additional authorization-from the TransRlant
Committee or any other prescribed governmental authority
BEFORE effecting a release.

(2) A holder shall take reasonal)le precautions to prevent
the escape of aquatic plants and fish from the holderls
aquaculture facility and from contakinment and attachment
structures in the facility.

A holder is expected to apply existing methods and equipment
to prevent the escape of livestock,. Those found grossly
negligent would be subject to prosecution.

19 -

Reporting escape

8.(1) The holder, or person acting on behalf of the holder,
who discovers an escape or evidence suggesting an escape of
aquatic plants or fish from an attachment or containment
structure in the holderfs aquaculture facility shall report
the escape or evidence to the manager

(a) verbally, within 24 hours of the discovery, and
(b) in writing, within one week of the discovery, if
requested by the manager.

This section establishes both the requirement and the process
for reporting escapement of aquaculture livestock, including
fish, shellfish and aquatic plants. Holders will be
responsible for ensuring farm staff are aware of this
requirement and take the steps necessary to ensure that the
manager of aquaculture is notified within the time limits
specified above. This section does, however, make agents
(including employees) and independent contractors who are
operating the licensed facility for the holder responsible
for reporting escapes.

(2) A holder who recaptures or attempts to recapture
aquatic plants or fish that have escaped from an aquaculture
facility shall report in writing the results of the
recapture, or attempted recapture, to the manager within one
week of the recapture or attempted recapture.

It is recognized that aquaculture liverstock may be
intentionally (eg. by vandals) or accidentally (ie- due to
human error, equipment failure or such natural events as
severe storms or tsunamis) released.

BEFORE attempting to recapture fin fish which escape from
fish farms, holders MUST:

1. notify the District Fisheries officer of the federal
Pepartment of Fisheries and Oceans (DFO) of the-escape,
and

2. be issued a special Rermit by that Fisheries Office

DFO advises that it will reggire notification of any escape
within 24 hours of discovery.

It is understood that the Department will issue these permits
to particular vessels to effect the recapture. The vessels

could be owned by holders or their employees, or by
independent contractors to the holder. Guidelines for
issuance of these permits have yet to be established.

Where fin fish escapes occur into freshwater, it is
anticipated that the Ministry of Environment and the Ministry
of Agriculture and Fisheries will be involved in the permit
issuance process as well as in the guideline development
process.

It is acknowledged that bottom-cultured oysters and aquatic
plants could, under certain rather unusual circumstances,
"escape" from an aquaculture facility. Once beyond the
boundaries of the facility such livestock become, in any
practical sense, indistinguishable from wild stocks which are
managed by the Ministry of Agriculture and Fisheries.
Therefore, BEFORE a holder recaptures or attemRts to
recapture oysters or acruatic Rlants which are carried, by
such natural forces as heavy wave action and strong current
flow, beyond the boundaries of a"n aquaculture facility, the
holder MUST, in addition to notifvLng the manager of the
release, obtain an oyster harvesti,-ig permit issued under the
Fisheries Act Regulations ol: a licence issued under section
24 of the Act. These may be obtained from the Ministry of
Agriculture and Fisheries and are subject to payment of fees
as required by regulation.

Nothing in this section prevents a holder from retrieving
containment or attachment structures and the aquaculture
livestock contained therein or attached thereto, which have
broken free of their moorings, and resecuring these within
the boundary of the aquaculture facility.

Transportation

9. A person who transports aquatic plants or fish on, over
or through fresh or tidal waters shall take reasonable
precautions to prevent the escape of the plants or fish, as
the case may be.

This subsection requires any person who transports
aquaculture livestock to employ due diligence, that is use
available methods, equipment and surveillance, to prevent the
escape of the livestock being transported

Inspectors

10.(1) The minister may appoint any person as an aquaculture
inspector to investigate matters related to

21 -

(a) the conduct of the business of aquaculture, and

(b) compliance with the Act, this regulation and an
aquaculture licence and its conditions.

(2) An aquaculture inspector may enter an aquaculture
facility during normal business hours to investigate the
matters referred to in subsection (1) and no person shall
obstruct the inspector in the course of the inspector's
duties.

No person may obstruct the entry of an inspector to an
aquaculture facility during normal business hours, nor may
anyone obstruct an inspector as the inspector carries out
his/her duties. Aquaculture inspectors will be uniformed in
some way and will carry photographic identification cards.

(3) At the request of an aquaculture inspector, an
inspector of fisheries or a conservation officer, a holder
shall produce for inspection a record that is required to be
produced for inspection as a condition of an aquaculture
licence.

It will be a condition of all aquaculture licences that
holders keep records sufficient to allow an inspector to
determine whether or not the holder is complying with the
development plans which are part of the aquaculture licence.
Further, holders will, as a condition of licence, be
required to produce such records for inspection within 24
hours of an inspectors request.

Fees

11.(1) In Appendix I

"primary aquaculture product's means a fish or an aquatic
plant that is a product of aquaculture but does not
include a processed or manufactured product;

11production value'$ means the dollar value of sales of
primary aquaculture product in the previous licence
year, but where the terms and conditions of the
aquaculture licence contain a maximum volume of
production equivalent to a dollar value, it means that
dollar value.

This definition creates a parallel between the eligibility
criteria for Bona Fide Aquaculturist Certificates (BFAC)and
the criteria for distinguishing between larger and smaller

22 -

scale aquaculture operations. Since BFAC's will not be
issued for locations which produce less than $7,500 of
primary aquaculture product each year, the lesser fees
charged smaller scale operations are justified based on
reduced administrative work load.

(2) A person applyinq for a new aquaculturs licence, a
renewal of an aquaculture licence or an amendment of an
aquaculture licence shall pay the fee set out in Appendix 1.

(3) Subject to the Financial Administration Act,, the fee
for an application for a now aquaculture licence and the fee
for a licence amendment are not refundable.

- 23

APPENDIX I

Schedule of Fees

1. Application for initial licence $25
2. Licence amendment $50

3. Licence and licence renewal for

a. aquaculture facility on private land,
production value at least $7500 $100

b. aquaculture facility on private land,
production value less than $7500 $50

c. aquaculture facility on Crown land,
production value at leasp $7500

i. aquatic plants and fish other than
fin fish $150

ii. fin fish $200

d. aquaculture facility on Crown land,
production value less than $7500

i. aquatic plants and fish other than
fin fish $50

ii. fin fish $100

GENERAL TERXS OF AN AQUACEILTURE LICENCE

1. For the purpose of this licence

"Branch" means the Aquaculture and Commercial Fisheries
Branch of the Xinistry of Agriculture and Fisheries, and

"Development Plan" means a plan filed with and approved by
the Branch for the species and location specified on the face
of this licence.

2. The holder of an Aquaculture Licence shall

2(1) comply with the management and operating specifications
in each Development Plan;

2(2) apply for and have approved anendments to a Development
Plan before (a) increasing or decreasing production or
productive capacity by more than 20% from that currently
authorized or (b) changing the mode of operation currently
authorized;

2(3) culture or husband only those species authorized by this
licence, and only if importation and transplantation
authorizations have been obtained from all competent
governmental authorities;

2(4) take reasonable precautions, to prevent the escape of
aquatic plants or fish (a) if transporting aquatic plants or
fish on, over or through fresh or tidal waters, and (b) from
the holder's aquaculture facility and from containment and
attachment structures in the facility;

2(5) ensure that neither the holder nor any person acting on
-behalf of the holder deliberately releases fish or aquatic
plants from the holder's aquaculture facility, unless
authorized to do so by the terms ai,,id conditions of this
licence;

2(6) ensure that the holder or a person acting on behalf of
the holder who discovers an escape or evidence suggesting an
escape of aquatic plants or fish reports the escape or
evidence and the results of any recapture or recapture
attempt to the Xanager of Aquacultiare;

2(7) ensure that the aquatic plants and fish cultivated and
husbanded in the holder's aquaculture facility are given care
and attention consistent with their biological requirements
for good health and well being;

2(8) undertake at the holder's own expense, reasonable
husbandry practises necessary for (a) preventative predator
control and (b) prophylactic disease control and diagnostic
disease treatment, including that required by competent

governmental authorities;

2(9) keep records adequate to allow an Aquaculture Inspector,
an Inspector of Fisheries or a Conservation Officer to
determine if the holder is complying with the terms of this
licence including, but not limited to, those described in any
Development Plans;

2(10) make available to an Aquaculture Inspector, an
Inspector of Fisheries or a conservation Officer, the records
referred to in sub-paragraph 2(9) within 24 hours of a
request being made;

2(11) advise the Manager of Aquaculture within one week of
any change in the holder's (a) address, (b) telephone, radio
telephone or facsimile machine number, and (c) representative
(contact person) and that person's telephone, radio telephone
or facsimile machine number;

2(12) deliver to the Branch, in the form and at the interval
determined by the Minister, any information required to
determine compliance by the holder with the terms of this
licence, and any other information that the Branch requires
to evaluate trends and practises of-the aquaculture industry
as a whole;

2(13) apply for and possess a valid processing licence before
processing aquatic plants or fish within the location
specified on the face of this licence;

2(14) ensure that the aquaculture facility is operated in
accordance with standards established by the Branch;

2(15) comply with all laws, bylaws and orders of any
competent governmental authority which affects the
aquaculture facility described herein.

3. If the holder of this licence fails to perform any
obligations in this licence, the Minister may, in addition to
other penalties in the Fisheries Act (R.S.B.C.) and the
Aquaculture Regulations, suspend or cancel this licence and
refuse to reissue an aquaculture licence to that holder or to
any person for the establishment of that holder.

4. This licence is not transferable except with the written
permission of the Minister.

5. This licence does not abrogate, replace, or derogate from
any of the rights, powers or jurisdictions of the Province of
British Columbia or the Ministry of Agriculture and
Fisheries.

APPENDIX I

LAND-BASED TANK FARMS

A recent development in the culture of salmon is the rearing of fish on shore in large
tanks. Sea water is continuously pumped through the tanks or raceways and discharged
back into adjacent marine waters. Experimental culture of Atlantic salmon in Iceland
has demonstrated the feasibility of this culture method. However, wide-scale commercial
operations are just being initiated. Thus, the method must be still considered experimen-
tal, but one which may provide an alternative method of fish culture in some areas and
situations.

The primary advantage of tank farms to the fish grower is that he has much greater
control over the water and the fish culture environment. By selecting the depth of the
water source, the farmer can avoid noxious plankton, and have limited control of
temperature, salinity, and dissolved oxygen. He can also control flow rates through the
tanks to provide optimal growing conditions, and may add supplemental oxygen or air to
the water to allow higher stocking densities and lower disease risks. Other advantages
include the ability to work in any weather, avoidance of many of the potential conflicts
with other water users, and avoidance of predator problems. Tank farms also provide
the opportunity for treatment of the effluent in areas that may be sensitive to nutrient
enrichment.

Disadvantages of tank farms include the higher construction and operating costs to pump
water, Perhaps the greatest disadvantage is the limited availability of suitable sites,
which must be flat, near water, and close to sea level to minimize pumping require-
ments.

The following discussion briefly describes tank farms and the potential environmental
impacts of this culture method which, in some situations, may provide an alternative
method of fish culture to fish farms.

The primary features of a land-based system include:

� An intake pipeline located subtidally to provide a constant supply of high qual-
ity water

� A pump and delivery system to circulate water through the rearing tanks

� A series of upland rearing tanks and/or raceways (circular tanks up to 20 m [66
ft] in diameter and 3 m [10 ft] deep appear to be the preferred tank design).

Land-based sites low in elevation and near shoreline areas are preferred. Such locations
reduce the length of the intake system, and maintain pumping efficiency by limiting the
pumping head (the vertical height water must be pumped to supply the rearing ponds).
In addition, tank farms should be located in areas free of plankton blooms and near
relatively deep water where water can be drawn from below any blooms.

The physiological requirements of salmon reared in tank farms are the same as those of
salmon reared in fish farms. However, because of different rearing conditions and
economic considerations, there are notable differeaces in the two technologies in terms
of rearing densities and operation and maintenance procedures. These differences are
projected to have a significant affect on the qualily of the discharge.

The most notable difference between the two technologies may be the amount of feed
required for production of an equivalent amount of fish. Average food conversion ratio
(FCR) in net-pen facilities may vary from 1.5 to 1 (Hardy 1988, personal communica-
tion) to 2.0 to 1 (Weston 1986). Recent work suggests that a FCR of 1.2 to 1.0 or less
may be achieved (Asgard et al. 1988). This ratio accounts for conversion of feed to fish
flesh (dry pellets of 10% moisture compared to fish flesh of 70% moisture) as well as
loss of feed (a 0-20% loss of the feed depending on the site location, type of feed, and
rearing practices), loss of fish due to mortality, or other reasons. Since onshore tank
farms use circular tanks with controlled flow and oxygen conditions, proponents claim that
salmon are able to feed and convert fish feed more. efficiently to flesh than in fish farms
that are subject to variations in water velocity and existing oxygen conditions. As the
FCR improves (lower ratio), the amount of waste food and total solids loss drops
significantly.

Other positive aspects of land-based tank farming include the relatively high quality of
waste water that is a result of dilution by the large volume of water flow necessary in
the tanks. Since self-cleaning tanks may be designed, "shock loads" due to sudden
discharges of large amounts of organic waste during cleaning may be avoided. Stocks
of fish may be separated for disease isolation and treatment. Routine addition of oxygen
may improve dissolved oxygen levels relative to existing source water conditions. This
extra oxygen allows higher stocking densities and reduced incidence of disease.

Potential negative impacts of land-based tank farming include release of a more
concentrated effluent than fish farms. Because large volumes of flow are necessary for
land-based tank farms, the concentration of pollutants such as ammonia in the effluent
may be low, but not as low as that seen from fish farms. The National Pollution
Discharge Elimination System (NPDES) permit system administered by the Washington
Department of Ecology requires that effluent receiving waters must have active
hydrodynamics to allow dispersion of the solid and dissolved wastes. Due to salt content,
solids isolated from onshore tank farms are not readily disposable on land as fertilizer
or fill. Depending on its design and site, land-based tank farms may need to screen their
intakes to prevent fish from being taken up in the intake.

In general, land-based rearing of fish allows for tighter control over all phases of
outgrowing (growth to marketable size) compared to fish farms. In a land-based system,
water flow rates and dissolved oxygen concentrations (variables important to fish health)
can be adjusted depending upon the fish rearing requirements. In addition, fish reared
in tanks are easily observed and sampled. This accessibility to the fish helps develop
efficient feeding schedules, identify stress and disease, and aids in the treatment of fish
if a fish disease or parasite is identified.

On the other hand, land-based tank farms are more costly to construct and operate than
a fish farm system. Unlike fish farms, tank farms need intake and outlet structures, as
well as rearing ponds. Greater operational costs also occur with tank farms due to
maintenance of the rearing facilities and the cost of pumping and circulating rearing
water.

The successful operation of a land-based facility is, therefore, dependent upon efficient
management and close control over the rearing process. Compared to fish farm
operations, land-based facilities can potentially increase overall survival, improving harvest
rates, as well as improving feed conversion ratios which result in decreased feed costs.

While there are no operating land-based tank farms in Puget Sound, upland systems
have been proposed for Grays Harbor and Clallam Counties, and other areas of the
United States and Canada. Saltwater tank farms are successfully operating in western
Europe and Iceland. Upland tank farms are a new technology and have yet to be fully
proven economically. They appear, however, to offer an alternative means of growing
fish which may complement fish farming, and provide an alternative to fish farms in
situations where fish farms would otherwise be impossible.

Land-based tank farms must comply with all local, state and federal regulations pertaining
to fish farms, with the possible exception of the ArmyCorps of Engineers permits
concerning navigation. In addition, tank farms, as sources of point-source discharge, are
subject to permitting requirements under the National Pollutant Discharge Elimination
System.

The following discussion briefly summarizes the possible environmental impacts of upland
tank farms for the purpose of comparison with fish farm culture.

1. SEDIMENTATION

Tank farms can introduce sediment into the marine environment through discharge pipes
at an outfall. Unlike fish farms, the fish farmer can regulate the effluent from the
facility. Feces and excess feed frequently will settle to the bottom of the tanks where
they can be removed, or be collected in settling ponds.

Any sediment which is discharged from a tank farm would affect the marine environment
in a manner similar to the sediment deposited from a fish farm facility or similar
discharges. Unlike fish farms, which by their size provide a vast area for dispersal, tank
farm discharges are a point source which would concentrate sediment impacts without
adequate sediment removal or adequate dispersal of the discharged material.

A range of responses, similar to those described for fish farms, will occur at the effluent
outfall. Where effluent is rapidly and effectively dispersed, the effects will range from
local enrichment of the bottom community to no noticeable change. If dispersion is
minimal, the effects will be substantial, as all of the sedimentation will occur in a
concentrated area. Dispersion can be increased by placement of the discharge pipe in
areas of high current flow, and through the use of diffusers on the end of the pipe.

2. WATER QUALITY

The potential water quality impacts from land-b&,;ed tank farms will be like those from
floating fish farms. Because of the relatively small volumes of water in tank farms, the
dissolved oxygen concentration in the discharge water may be reduced. Data from 38
fresh water tank farms in Europe were used to calculate an average decrease in dissolved
oxygen of 1.6 mg/L through the facilities (Alabaster 1982). As a worse-case approxi-
mation, a decrease of 2 mg/L through a land-based tank farm, and an initial concentra-
tion of 6 mg/L would require a minimum dilution factor of about 10 to meet the state
standard (5.8 mg/L in this case). A dilution factor of 10 would likely be achieved in
close proximity to a land-based tank farm outfall. Tank farms also have the potential
to aerate or oxygenate water entering and leaving the tanks. Tlis can improve the
culture environment for the fish, as well as offse-t any oxygen demand from the fish or
the discharged nutrients.

Land-based tank farms are subject to the same nutrient enrichment considerations as
floating fish farms. That is, restricted embayments with nutrient sensitivity should be
avoided for both the good of the cultured fish due to dinoflagellate blooms, and for the
possible enrichment effect upon algae or phytoplankton in the discharge waters. The
three onshore tank farms proposed or being builtin Washington state are located in non-
nutrient sensitive waters. As in fish farms, about 70% of the nutrients are discharged
in solution. Retention time of water within the tank farms will generally be less than
two hours, and the water is actively moving in the tanks and pipes. Both the period of
time and movement of the water are not conducive for the development of optimum
algal growing conditions. Site characteristics, especially the physical oceanography, depth
of intake and discharge, and the density of water at both intake and discharge depths will
greatly influence the fate of discharged waste water.

Chemical usage in tank farms would generally be less than in fish farms, again because
the tank farm operator has much greater control over the culture enviromnent.
Antibiotic use would probably be less because the farmed fish will not be directly
exposed to disease carrying wild fish, and the controlled culture environment reduces the
probability of disease and permits easier control of any disease outbreaks. In addition,
it is likely that tank farms would not use large amounts of antifoulant materials. The
use of any chemical in tank farms would have impacts on the aquatic environment similar
to those described in Section 5.4, Chemicals.

3. FISH AND SHELLFISH

The primary impacts of fish farms on fish and shellfish are the possible smothering of
sessile (immobile) organisms below the farms, if -the farms are located in shallow, poorly
flushed areas, and the attraction of mobile fish and shellfish species to the site. Because
tank farms would have a much more concentrated discharge, the area of bottom affected
would be less than for fish farms, and the potential impact of sessile bottom-dwelling
organisms would be reduced. Construction of the. intake and outfall structures, however,
could destroy shellfish beds in the construction area. Because shellfish populations

usually occur in discrete beds, proper site selection can avoid significant impacts for
clams, oysters, geoducks, etc.

In the absence of a significant structure in the water, the attraction of fish and shellfish
to the site would be reduced or eliminated. Fish could be adversely affected if entrained
in the intake pipe; therefore, proper screening of the intake will be necessary. It will
also be necessary to avoid areas of intertidal herring and smelt spawning or important
habitats for other fish species. Because the location of all these habitats is unknown,
field observations will be necessary to determine which species use the area and if
important habitats would be affected.

'ne potential for fish to escape into the wild from tank farms is relatively remote
compared to fish farms. Even in the event of a major catastrophe (for example, a tank
ruptures during a large earthquake), most fish would be stranded on dry ground and die.
If fish were to escape, the impacts to wild populations would be the same as for fish
farms.

4. WILDLIFE

Construction of each land-based tank- farm could result in the loss of several acres of
upland habitat, depending upon the previous use of that land.

Most of the habitat loss would be due to construction of the rearing tanks and support
facilities such as operations buildings and new access roads. Clearing of vegetation
would remove habitat, and may result in losses or displacement of small vertebrates such
as mice, snakes, and frogs. Larger animals such as river otter, deer, raccoon, beaver,
and birds may temporarily avoid construction sites. Noise generated by farm construction
and operation may temporarily displace or disturb nearby wildlife. Consultation with fish
and wildlife agencies during permit review is required to avoid affecting the habitats of
any threatened or endangered species. Stretching netting over the top of upland facilities
is an effective technique for keeping predators away from the fish.

5. ODORS

Operation of a land-based tank farm facility would be less likely to produce odors than
would a floating fish farm facility because of the absence of nets and their associated
fouling organisms, and the availability of enclosed storage areas for food. Minor odors
could result from diesel engines used for emergency pumping during power outages, or
from trucks servicing the facility. All odor impacts would be occasional and intermittent.
As with floating fish farms, dead fish could create unpleasant odors if not removed from
the tanks and disposed of properly. Because tank farms are located on shore, they may
be closer to residents than fish farms. Consequently, any odors produced may have a
greater effect on these residents.

6. NOISE

Sources of noise at land-based tank farms would be similar for any small agricultural
or commercial activity. Urge pumps and compressors would be required for aeration
and pumping. These would be electrically powered and enclosed in buildings or located
below grade, and would probably produce little detectable noise off the farm property.

Land-based tank farms would be required to meet the relevant local and state noise
standards. In rural areas with low existing noise levels, noise levels meeting state
standards may be disturbing. In such areas, additional mufflers, sound enclosures, or
buffer zones could be used to minimize any disturbance of nearby residents.

7. UPLAND AND SHORELINE USE

Land-based tank farms have the same requirements for high quality water as floating
operations, and will provide ongoing monitoring of water quality. Other shoreline
activities adversely affecting water quality would harm. the fish culture operations. As in
the review of any proposed activity, new projects near the tank farm would be evaluated
for their effect on existing activities and their impacts on water quality and other
elements of the environment.

Activity levels associated with an upland tank farm will be similar to those of a small
farm or commercial facility. Increased vehicle traffic from employees travelling to and
from work, and from deliveries of food and other supplies and shipments of harvested
fish will occur. In some cases, the farms may also attract visitors and tourists. The
number of trips will depend upon the size of the facility and its proximity to suppliers.
In addition, land-based tank farms may have other commercial elements, such as fish
processing, which must be considered.

8. AESTHETICS

The extent of aesthetic impacts resulting from tEnk farms will vary depending on the
site, especially the existing activities and structures in the area, and the visibility of the
facility to outside observers. Highly visible tank farms may be perceived as visually
intrusive in rural or natural areas, yet be unobtnisive at sites surrounded by industrial
or commercial uses.

Aesthetic impacts in sensitive areas can be minimi:?,,ed by providing for adequate setbacks
from adjoining properties and by providing landscaping to visually shield land-based
facilities from nearby observers.

9. RECREATION

Recreational activities would not be impacted by land-based tank farms, except where
the facility displaced existing shore-based use. If the beach is privately owned,
recreational use by the public would be allowed only with the owner's permission. If the

beach is publicly owned, intake and discharge pipes could be buried to avoid any conflict
with existing use, except during construction.

10. LOCAL SERVICES

The impacts of land-based tank farms on local services are expected to be similar to the
impacts of floating fish farms and would not be significant. Tank farms would have more
demand on local services such as electricity, roads, and fire protection. They would also
be subject to local property and other taxes which currently do not apply to fish farms.

APPENDIX J

LEGISLATION AUTHORIZING THE EIS

WASHINGTON LAWS, 1987 Ist Ex. Sess. Ch. 7

General Fund Appropriation ..................... S 602,000
The appropriation in this section is subject to the following conditions
and limitations: $182,000 is provided solely for carrying out the Puget
Sound water quality plan.
NEW SECTION. Sec. 309. FOR THE PUGET SOUND WATER
QUALITY AUTHORITY
General Fund Appropriation ..................... S 2,910,000
Water Quality Account Appropriation ............. S 1,100,000
Total Appropriation .................. S 4,010,000
NEW SECTION. Sec. 310. FOR THE DEPARTMENT OF
FISHERIES
General Fund Appropriation----State ............. S 47,465,000
General Fund Appropriation-Federal ........... S 14,057,000
General Fund Appropriatioh-Private/Local ..... S 3,651,000
Aquatic Lands Enhancement Account Appro-
priation .................................. S 425,000
Total Appropriation .................. S 65,598,000
The appropriations in this section are subject to the following condi-
tions and limitations:
(1) $106,000 of the general fund tate appropriation is provided
solely for carrying out the Puget Sound water quality plan.
(2) $40,000 of the general fund----state appropriation is provided
solely for the purposes of reintroducing an early coho salmon run to the
Tilton river and Winston creeL
(3) S587,000 of the genend fund----state appropriation is provided
solely for implementing the titnber@ fish, and wildlife agreement. If Senate
Bill No. 5845 is not enacted by June 30, 1987, the amount provided in this
subsection shaH lapse.
(4) $150,000 of the general fund-state appropriation is provided
solely for sheMh enforcement on Hood Canal.
.(5) $150,000 of the aquatic Iands c- account appropriation
is provided solely for the preparation of an ecological impact statement on
thegWdel[inesforthe --m- __v no of salmon act pensin Pullet Sound.
(6) The department shall present to the natural resource committees of
the senate and house of sepiesentatives no later than February 1988 a re-
port on the department's watershed plan, with specific identification of the
benefits associated with the Queets hatchery and other Indian tribal
agreements.
(7) $194,000 of the general fund----state appropriation may be ex.
pended for additional food for the Deschutes hatchery.
(8) $400.000 of the general fawf---state appropriation is provided
solely for the purpose of a comy ehensive biological study conducted by the
department in conjunction with the University of Washington and Grays

127111

APPENDIX K

EFFECT OF FISH FARMS ON SURROUNDING PROPERTY VALUES

ALPINE APPRAISAL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIM. WASHINGTON 98382
(206) 683-7084

-REPORT FLOATING SALMON NET PENS

SITE #1: PEALE PASSAGE Mason County, WA.
Township 20 North, Range 2W, W.M.

DATES OF INSPECTION: August 10 and 11, 1988, February 15, 1989

PURPOSE AND FUNCTION OF REPORT: The purpose of this report is
to determine the effects, if any, of floating salmon net pens on
the surrounding upland property values. The function is to
provide information useful in siting floating salmon net pens.

CERTIFICATION AND LIMITING CONDITIONS: The Standard
Certification and Limiting Conditions are attached.

ALPINE APPRAISAL SERVICE
REAL ESTATE APPRAPSERS
150 S. 5TH AVE. SUITE 14 SEQUIM. WASHINGTON 98382
(206) 683-7084

AREA DESCRIPTION: Peale Passage is located between Squaxin
Island and Hartstene Island in Mason County, Washington. The
width of the passage varies from about, 500 feet at its north end
to about 4300 feet in the vicinity of the existing fish pens.
Squaxin Island, along the west side o4 Peale Passage, is an
Indian Reservation and is basically undeveloped. Access is
gained by boat-. Hartstene Island is t.o the east of Peale
Passage and is connected to the mainland by a bridge built in
1969. A majority of the residential development on the island
has taken place since 1969 as a result. of the accessibility and
the inflating property values of the mid to late 1970's in
western Washington. Hartstene Island's development is mainly
along the waterfront. The upland areats are for the most part
still used as forest land.

Because Squaxin Island is an Indian Reservation and Hartstene
Island has only been readily accessible for the past 20 years,*
much of the land along Peale Passage is still relatively
undeveloped.

Floating salmon net pens were first installed in Peale Passage
by the Squaxin Indian Tribe in 1971. Additional floating net
pens were put into operation by the Washington State Department
of Fisheries and the Squaxin Tribe in 1982 and 1986.

TOPOGRAPHICAL INFORMATION: The southwest portion of Hartstene
Island has very low lying terrain with a maximum of 277' feet of
e'levation in the area directly east o+ the net pens. I traveled
most of the accessible roads in the Southeast portion of
Hartstene Island and found no sites where the net pens were
visible other than along the waterfront. These waterfront sites
were all under 40 feet in elevation. The dense vegetation and
lack of upland development precludes seeing the water in Peale
Passage from other than along the shoreline.

Squaxin Island was not visited by this appraiser.

SIZE AND DESCRIPTION OF PENS: The most northerly pen complex is
69.5 feet x 320 feet and covers about one-half acre of water
surface. The middle pens are 329 feei: x 110 feet and cover
about .83 acres of water surface. The? most southerly pens are
69.5 feet by 320 feet and cover about one-half acre of water
surface. There are several barges anchored near-the pens which
serve as support structures for the pens. The elevations of
these structures vary from 12.5 feet I:o 25 feet above the water
surface.

-2-

ALPINE APPRAISAL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIM. WASHINGTON 98382
(206) 683-7084

VISIBILITY OF PENS FROM HARTSTENE ISLAND: The actual floating
net pens were not visible from any of the sites visited on
Hartstene Island. The orange anchor balls in the vicinity of
the pens were visi.ble and the support structures were visible.
From Hartstene Island it was difficult to tell if these support
structure were floating or built on-shore.

PROPERTY VALUES: Sales of both improved and unimproved real
estate on Hartstene Island, from which the fish pens might be
visible, were researched. Sales of similar properties from
other areas of Hartstene Island were also researched. Mason
County appraiser Darryl Cleveland provided information gathered
by the County Assessor's Office in their recent re-evaluation of
Hartstene Island. Several local real estate offices and
individual property owners having "For Sale' signs were
contacted to determine current asking prices far parcels in the
area bordering Peale Passage as well as other similar areas in
Mason County.

CONCLUSION: The appraiser found normal variations in front foot
values for waterfront lots based on the type of road access,
availability of utilities such as a water system, height a+ bank
at the waterfront, etc. The data gathered indicates that
properties having similar characteristics sold +or similar
amounts without regard to their location on Hartstene Island.

Three new homes are under construction at the present time on
the southwest side a+ Hartstene Island in the area nearest to
the.floating net pen sites. This further indicates the pens
have not inhibited the development of new homes in this area.

Property values based on sales history show a rapid appreciation
all over Hartstene Island in the mid 1970's. In the years
between 1983 and l9e6 property values decreased uniformly all
over Hartstene Island as they did generally in this part of
Washington.

In fact in 1987 the Mason County assessors office collected data
on sales of low bank, medium bank and high bank waterfront from
all areas a+ the island. As a result of this study the assessed
value per front foot a+ waterfront was lowered in all three
categories without regard to their location on the island.

-3-

ALPINE APPRA.ISA.18 SERVICE
REAL ESTATE APPRAISERS
ISO S. STH AVE. SUITE 14 SEQUIK WASHINGTON 98382
(206) 683-7084

After examining the comparable sales data from different areas
of Hartstene Island and similar waterfront parcels in Mason
County, it is the opinion of this appraiser that the Peale
Passage floating net pens have had no effect on property values
in this area.

It is also the opinion of the appraiser after visiting various
areas and taking photographs from these areas that there is no
visual impact, good or bad, from these pens.

-4-

ALPINE APPRAISAL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIM. WASHINGTON 98382
(206) 683-7084

SALES INFORMATION

RESEARCH AREAS ON PEALE PASSAGE - SW SIDE HARTSTENE ISLAND

CHAPMAN ROAD AREA - APPROXIMATELY 5400 LINEAL FEET TO PEN 12025

Lot 1, Sunset Acres PN 22014-50-00001
Community Water - Individual Septic System

1974 - $17,500 as unimproved waterfront lot.
1975 - Building Permit $18,000.
1975 - Added to Assessor Rolls in 1975 as 1.5 story 1593
sq. ft. home with 484 sq. ft. garage.
1984 - Sold for $105,000.

Lot 5, Sunset Acres
1976 Building Permit $30,000.
1976 Added to Assessor Rolls in 1976 as 1.5 story, 2048
sq. ft. home with 672 sq. ft. garage
1986 Sold for $125,000.

MAPLES ROAD AREA - APPROXIMATELY 3300 LINEAL FEET TO PEN 11284

Tract 3 Govt. Lot 2 and Tax 61-D and
Tract 3 of S.P. # 426 100 FF WF
1979 - $29,500 unimproved waterfront lot.
1980 - Building permit $70,590.
1980 - Added to Assessor Rolls in 1980 as a 3 story 1974
sq. foot home.
1983 - Sold for $135,000.

Tract 2 Govt. Lot 2 and Tax 61-C 100 FF WF
1988 New log home under construction at the present
time (photo).
1984 $55,000 unimproved waterfront lot.
1985 $63,500 unimproved waterfront lot.

-5-

ALPINE APPRAISAIL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIM, WASHINGTON 98382
(206) 683-7084

Tract 4 of Survey 6/5-6 135 FF WF
1981 - $65,575 unimproved waterfront lot.
1981 - $79,500 unimproved water-front lot.
1987 - $60,000 unimproved waterfront lot.
1987 - New home under construction.
Assessed value a+ improvements
Partially completed $679280
Assess. value of lot $60,700
Total $1279980

Tract 6 of Survey 6/5-6 110 FF WF
1981 - $67,500 unimproved waterfront lot.
1986 - New home 50% complete.

Tract 3 of SP #1200 Govt. Lot 5 105 FF WF
1983 - $42,500 unimproved watet-front lot.
1987 - $39,700 unimproved waterfront lot.

OLYMPIC VIEW TRACTS 4800-6900 LINEAL FEET TO PEN 11284

-6-

ALPINE APPRAISAL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIM. WASHINGTON 98382
(206) 683-7084

RESEARCH AREA ON CASE INLET -E. SIDE HARTSTENE ISLAND

E. SIDE HARTSTENE ISLAND - POINT WILSON
Plat (28 lots) Sec. 20 Twp. 20N, Rge 1W, W.M.

Tract 5 + Tax 1194-A and South 25' Tract 4 and Tax 1194-B-1
Home built 1946
125 FF WF Med. Bank (20 ft.+)
1977 - Sold for $33,500.

This plat consists of older cabins and homes, two or three new
homes and a few vacant parcels. There is similar sales activity
to the Peale Passage side.

PLAT OF ISLAND SHORES - Govt. Lots 2 and 31 Sec. 18, Twp 20N,
Rge IW W.M.

Tract 6 Island Shores 95 FF WF
Med. Bank (30 1- 501) Brushy with clearing.
1973 - $5,000 unimproved waterfront lot.
1983 - $34,000 unimproved waterfront lot.

Tract 7 Island Shores 100 FF WF
Med. Bank (30' - 50") Brushy - level
1982 - $33,500 unimproved waterfront lot.
1983 - New construction 1272 sq. ft., 1.5 story with
deck.

Tract 13 Island Shores 100 FF WF
Med. Bank (30' - 501)
1978 - $20,000 unimproved waterfront lot.
1981 - $429800 unimproved waterfront lot.

Tract 14 Island Shores 100 FF WF
Med.-Hi. Bank (301 - 501)
1971 - $15,000
1981 - $34,500

Tract 9 Island Shores Plat #2
1987 - $60,000 improved waterfront lot.

ALPINE APPRAISAL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUINI. WASHINGTON 98382
(206) 683-7084

REPORT - FLOATING SALMON NET PENS

SITE #2: RICH PASSAGE - Kitsap Counte, WA.
Township 24 North, Range 2E, W.M.

DATES OF INSPECTION: August 11, 17 and 18, 1988

PURPOSE AND FUNCTION OF REPORT: The purpose of this report is
to determine the effects, if any, of floating salmonnet pens on
the surrounding upland property values. The function is to
provide information useful in siting -floating salmon net pens.

CERTIFICATION AND LIMITING CONDITIONS: The standard
Certifications and Limiting Conditions are attached.

ALPINE APPRAISAL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIM. WASHINGTON 98382
(206) 683-7084

AREA DESCRIPTION: Rich Passage is located between the south end
of Bainbridge Island and the Manchester area on the Kitsap
Peninsula. It is the waterway used by the Seattle-Bremerton
ferries and the U.S. Navy Shipyards at Bremerton and Keyport.

A large portion of the south end of Bainbridge Island was the
Fort Ward Military reservation for many years. Today part of
the reservation is a Washington State Park and the balance was
sold by the U.S. Government to a private party. There is a row
of homes along the waterfront both east and west of the Fort
Ward area.

On the south side of Rich Passage the U.S. Government maintains
a naval reservation. A Washington State Park adjoins the
Reservation on its north boundary. The area north of the State
Park is known as Wautauga Beach and is a single family
residential area.

Other than along the waterfront, the upland areas on both the
north and south side of Rich Passage are largely undeveloped.

According to information from the Washington State Department of
Natural Resources Aquaculture Division, floating salmon net pens
were first placed in Rich Passage in June of 1971 by the
National Marine Fisheries Service. In March of 1972 Domsea Inc.
leased a large area on the south side of Rich Passage for
placement of net pens. Pens were placed an the north side of
Rich Passage by Domsea Inc. in 1974, by Domsea Inc. in 1979 and
by Passage Silver Inc. in 1987. Four of these pens are shown on
the aerial photo exhibit attached to this report. The fifth set
of pens was installed after these photos were taken and its
location an the exhibit is approximated from lease information
obtained from the Department of Natural Resources.

TOPOGRAPHICAL INFORMATION: The south end of Bainbridge Island
has a narrow level area along the waterfront, then a fairly
steep brush and tree covered bank that rises to about 100 feet
in elevation. The terrain then becomes a gradual slope to about
200 feet in elevation in most areas. The highest point is about
360 feet in elevation.

-2-

ALPINE APPRAISAIL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIN.I. WASHINGTON 98382
(206) 683-7084

The accessible areas on the south or Kitsap Peninsula side of
Rich Passage have a much more gradual slope up from the
waterfront with dense brush and tree cover in most areas.

Photographs were taken from a variety of elevations and
locations in an attempt to show the visual impact of the pens
from different elevations and distances. All photographs were
taken with a 50 MM lens.

SIZE AND DESCRIPTION OF PENS: The fcur floating net pens as
shown an the aerial photograph of Rich Passage scale as
follows: Pen number 9780 is approximately 200 feet by 850 feet;
Pen number 12584 is approximately 250 feet by 250 feet; Pen
number 10237 is approximately 250 feet by 500 feet; the pen an
the Environmental Protection Agency (EPA) dock is approximately
200 feet by 150 feet, the Passage Silver Pen was recently
installed and the only information regarding its size is the
Department a+ Natural Resources list showing it covers .41 acres
of water surface.

These floating net pens in this area are made of a variety of
materials ranging from wood to steel. Some of the pens are a
mixture of both. The Bremerton East Quadrangle Exhibit has
pictures taken of the pens which are attached to the EPA dock at
Manchester. Below it is an 85 MM photograph taken from the end
of the Domsea Inc. dock on the south end of Bainbridge Island.
It shows an area a+ the wood pens # '10237 that are attached to
the dock as well as Pen number 12584 near Orchard Rock which
appears to be a steel pen. The EPA dock and pens are visible in
the distance on the right hand edge of the photo as well as Pen
number 9780 behind and to the right of the red channel marker.

VISIBILITY OF PENS FROM DIFFERENT LA14D LOCATIONS AND ELEVATIONS
ARQUND RICH PASSAGEt A portion of the pens were'visible from
almost all the locations where there is a water view in this
area. At or near the shoreline (under@401 in elevation) the
floating net pens are hardly visible if over 2400 lineal feet
away. Within 2400 lineal feet and especially at elevations of
over 401 the pens are more visible. From an elevation of
approximately 90 feet and 1800 lineal feet away the Orchard Rock
Pen # 12584 is visible. The photos taken 6000 feet and 10,500'
feet distant were both taken at approximately 185 feet in
elevation. The floating net pens are a faint line at these
distances.

-3-

ALPINE APPRAISAL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIM. WASHINGTON 98382
(206) 683-7094

PROPERTY VALUES: The sales history of both improved and
unimproved real estate in the Wautauga Beach area was
researched. The floating net pens in Rich Passage are visible
from a portion of these properties and not visible from others.
A typical residence sale with a view of the pens was selected
and compared to other similar residence sales in Kitsap County.
I discussed the assessed values of properties in Kitsap County
with Ida Mae Ryen of the Kitsap County Assessor's Office. She
said the Rich Passage area was last valued in 1983 and is due
for a re-evaluation next year. She indicated that they would
examine every sale and check one against the other for any
impacts fr-om the pens. She also said this had been done in the
last valuation and so far no differences in value have been
evident.

CONCLUSION: After examining the comparable sales data from bc>th
improved and unimproved properties, some with a view of the
floating net pens and others with no view of the pens, it is the
opinion of this appraiser that the Rich Passage floating salmon
net pens have had no effect on property values in the two plats
at Wautauga Beach.

Based on observations of the floating net pens from a variety of
distances and elevations it is my opinion that pens over 2400
lineal feet distant are not visible enough to have any impact on
property values. At distances closer than 2400 feet they are
more visible. I could not locate any closed real estate
transactions of properties within this 2400 foot distance;
however, there is a new home on lot 10 in the plat of Sunset
Ridge an South Bainbridge Island. This home was added to the
Kitsap County Tax Rolls in 1988 at an assessed value of $28,380
+or the land and $94,230, for the improvements for a total of
$1229610.

I located two pending sales of waterfront properties within 800
lineal feet of dock pen number 10237 and 1500 lineal feet of
dock pen number 12584. The realtor who has the property listed
also indicated the 430 feet of waterfront adjoining the Domsea
dock and pen number 10237 was on the market. The pending sales
indicate to me that the existence of floating net pens has not
inhibited the development and sales of properties within 2400
lineal feet of the pens in this area.

-4-

ALPINE APPRAISAL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUINI. WASHINGTON 98382
(206) 683-7084

SALES INFORMA,rION

RESEARCH AREAS ON SOUTH SIDE OF RICH PASSAGE

Wautauga Beach Plat - Volume 5, Page 13, Gov"t. Lot 3 & 41
NE 1/4, Sec. 9, Twp. 24W, Rge. 2E, W.M.

5200 - 5700 lineal feet to Pen No.
12584 and No. 10237

SALES ACTIVITY ON LOTS WITH NO VIEW OF THE FLOATING NET PENS.

Lot 3 and Portion of Lot 4 plus Tidelands
2-23-77 - $48,000
6-20-88 - $124,000
Home was built in 1930; 1.0 + bsmt, remodeled 1978. Lot
has 100 FF WF.

Lot 10 -
10-29-74 $3250
5-28-76 $30,000
B-11-78 $35,000

Lot 15 -
11-5-82 $84,950
B-7-86 - $67,000

Lot 17 -
10-24-73 - $11,500
11-6-74 $20,000

Lot 4 -
2-4-88 - $72,000

SALES ACTIVITY ON LOTS WITH A VIEW OF THE PENS.

Lot 42 -
4-24-81 $100,000
10-1-87 $92,500
Mobile home placed on property in 1975 (cost $17,495)
1974 Commodore 24' x 601. 88.70 FT WF

Lot 31 -
1-9-79 $60,000
7-20-87 $81,455

jUPINE APPRAISAL SERVICE
REAL ESTATE APPRAISERS
ISO S. STH AVE. SUITE 14 SEQUIM, WASHINGTON 98382
(206) 683-7084

Lot 40 -
9-27-73 - $339000
1-27-84 - $989500
5-8-86 - $75,000
Home built 1946, remodeled 1971 and 1974 1, story.
88.70 FF WF

Lot 43
5-18-77 - $84,500
12-7-79 - $1361500
Home built 1940, remodeled 1963 and major remodel
19135-86 including swim pool, boat ramp, and marine
railway

Lot 35
4-25-77 - $599000
7-20-78 - $69,500

Lot 29
6-10-76 - $11,500
5-2-77 - $75,000

M.B. Crane's Waterfront - Addition to Manchester - Portion 13L
5, Sec. 9, Twp. 24N, Rge. 2E, W.M.

- 4700 - 5500 lineal feet to Pen No.
12584 and No. 10237

ALL LOTS IN THIS PLAT HAVE A VIEW OF THE FLOATING NET PENS

Lot 5
B-8-74 $239500
B-28-75 $27,500
4-12-79 $909000
9-26-86 $789000 - Divorce-property settlement
Home built in 1941, remodeled 1975 (interior) 60 FF WF

Lot 20
5-16-73 $8,500
1-8-75 $11,250
6-11-80 $117,000
Home built in 1975. Lot has 62 FF WF

Lot 21
9-12-73 - $31.,000
6-24-85 - $105,000
Home built in 1963, 1 story. Lot has 64 FF WF

-6-

ALPINE APPRAISAL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIK WASHINGTON 98382
(206) 6a3-7094

RESEARCH AREA ON N. SIDE OF RICH PASSAGE

HOMES WITH VIEW OF FLOATING NET PENS. PLAT OF SUNSET RIDGE9
VOL. 12,1 PG. 74.

Approximately 1000 lineal feet to Pen No. 12227; approximately
2000 lineal feet to Pen No. 12584.

Lot 10 + Tidelands - 94 FF WF
B-27-73 - $19.250
3-22-77 - $27,000
1977 - Quit claim deed - Divorce settlement
1988 - New home on tax rolls
$122,,610 assessed value.

Lot 9 + Tidelands - 97 FF WF
1985 New home on tax rolls, $106,490 assessed value.

AIPINE APPRAISA& SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIM. WASHINGTON 98382
(206) 683-7084

REPORT - FLOATING SALMON NET PENS
AND PROPOSED FLOATING SALMON NET PENS

SITE 3: SKAGIT BAY - Skagit County, WA.
Township 34 North, Range 2E, W.M.

DATES OF INSPECTION: October IS and 19, 1988, March 9, 1989

PURPOSE AND FUNCTION OF REPORT: The purpose of this report is
to determine the effects, if any, of the existing floating
salmon net pens on the surrounding upland property values. The
function is to provide information useful in siting floating
sal mon net pens.

CERTIFICATION AND LIMITING CONDITIONS: The standard
Certifications and Limiting Conditions are attached.

ALPINE APPRAISAL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIM. WASHINGTON 98382
(206) 683-7084

AREA DESCRIPTION: -Skagit Bay is located between Fidalgo Island
and the north end of Whidbey Island. A portion of the bay is in
Island County and a portion is in Skagit County. An existing
floating salmon net pen is located in Skagit County.

Hope Island and Skagit Island are located in Skagit Bay. These
two islands are both part of Deception Pass State Park.

The portion of Fidalgo Island bordering Skagit Bay an the east
is part of the Swinomish Indian Reservation. The reservation in
this area has a combination of lands in fee ownership, lands
leased by the tribe and tribal owned lands. All tidelands are
claimed by the tribe.

Residential development in this area of the Swinomish Indian
Reservation is mainly along the waterfront. The terrain here is
a gradual slope from beach level to about 300 feet in elevation
one-half mile inland.

Whidbey Island on the west side of Skagit Bay is a combination
of private ownership and Washington State Park ownership. The
portion in private ownership is sparsely populated due to the
steep terrain along the water.

The interior of both the Swinomish Indian Reservation and the
north end a+ Whidley Island are largely undeveloped.

Skagit Bay has an existing floating salmon net pen complex
approximately 1800 lineal feet north of Hope Island and 1200
lineal feet west of the Lone Tree Point. According to
information supplied by the Skagit Systems Cooperative,
installation of these pens was begun in May of 1987.

SIZE AND DESCRIPTION OF EXISTING FLOATING SALMON NET PENS: The
existing floating salmon net pen lease? # 12356 is approximately
100 feet by 480 feet and according to the Department of Natural
Resources lease covers .66 acres of water surface.

The pens appear-to be of steel constrLICtion when viewed from the
share. There is a supply building, 10 feet by 20 feet by 10
feet high, at the pen site as shown irk the photo exhibit.

LOCATION AND DESCRIPTION OF PROPOSED FLOATING SALMON NET PENS:
The Skagit System Cooperative is proposing to locate a 100 foot
by 480 foot complex of floating salmon net pens about 1800
lineal feet southeast of Hope Island and 4400 lineal feet due

-2-

ALPINE APPRAISAL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIM. WASHINGTON 98382
(206) 683-7084

west of Sunset Drive, a street in the Plat of Wagner's Hope
Island Addition. These pens will be about 7400 lineal feet from
the closest point on Whidbey Island and about 3100 feet
southwest of Snee-oosh Point, the nearest land on Fidalgo
Island. The pens are to be steel construction similar to the
steel pens shown in the lower left hand corner of the Rich
Passage exhibit. A small building, 10 feet by 25 feet by 10
feet high, will be located at the site of the pens to provide
protection for the employees and limited storage of food.

VISIBILITY OF PENS FROM DIFFERENT LAND LOCATIONS AND ELEVATIONS
AROUND SKAGIT BAY: Photographs (50 MM) were taken from
different areas as shown on the Anacortes South quadrangle map
exhibit. These areas were visited and photos taken of the
proposed location of a new fish pen complex and the existing
fish pens. Approximate distances from the pens and elevations,
at the site are noted near each picture.

PROPERTY VALUES: Sales of unimproved and improved real estate
along the east shore of Skagit Bay were researched for the
period of 1986, 19877 1988 and 1989. Two recent sales of homes
with a view of the existing net pens have recently been recorded
and a third sale is scheduled to close on April IQ), 1989. It is
my opinion that these homes sold at Fair Market Value.

CONCLUSION: The Skagit Bay area is presently providing the best
information an property values as sales are recorded on homes
less than one mile from an existing salmon net pen complex.

Sales data from the area indicates that two homes with a view of
the pens were recently sold at Fair Market Value. These homes
are approximately 3000 and 3900 feet from the pen s*ite.

There are several new homes under construction south of Snee-
oosh Point. These homes will be approximately 4100 lineal feet
east of the approved proposed pen complex.

After examining the comparable sales data from areas without a
view of the pens and the construction activity in the area of
the approved new pen site, it is the opinion of this appraiser
that the Skagit Bay existing floating net pens and proposed
floating net pens have not affected property values in this area
as as indicated by the recent market sales of homes at 1606 Snee-
oosh Road and 1575 Snee-oosh Road.

-3-

ALPINE APPRAISAJL SERVICE
REAL ESTATE APPRAISERS
150 S. STH AVE. SUITE 14 SEQUINI, WASHINGTON 98382
(206) 683-7084

SALES INFORMA"riON

RESEARCH AREAS ON SKAGIT BAY - S(J SIDE OF FIDALGO ISLAND

WAGNERS HOPE ISLAND ADDITION

Lot 6 & N. 20' of Lot 5, BI . 4 (1740 Gol den View Dr.)
4/86 $120,000,, Residence built 1958

Lot 31 BI 3 (1732 Sol den View Dr.)
11/86 $130,000, Residence built 1963

Lot 2 and Ptn. Lot 19 B1.4 (1746 Golden View Dr.)
5/88 $85,000, Residence built 1969

FAHLENS SNEE-OOSH TRACTS
Lot 4 (1668 Reef Point)
8/86 - $178,500, Residence built 1969

PLAT OF SNEE-OOSH
Lot 65 (712 Chilberg Ave.)
7/136 - $115,000, Residence built 1927

PLAT OF SHOREWOOD
Lot 9
1/87 - $42,500, Unimproved lot

PORTION OF GOV'T. LOT 1, Sec. 27, Twp. 34N, Rge. 2E, W.M.
1606 Snee-oosh Road
10/6/8B - $2849000, Residence built 1979,

PORTION OF GOV'T. LOT 1, Sec. 22, Twp. 34N,, Rge. 2E, W.M.
1575 Snee-oosh Road
10/6/88 - $73,500, Residence built 1967

-4-

ALPINE APPRAISA1 SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIM. WASHINGTON 98382
(206) 683-7084

RESEARCH AREA SIMILK BAY - FIDALGO ISLAND

SIMILK BEACH PLAT
Lot 23 & Ptn. Lot 24, Bl. 6 (624 Satterlee Rd.)
G/GS - $61,000, Residence built 1920

GIBRALTER AREA
Ptn. G.L. 6 Sec. 19, Twp. 34, Rge. 2E (500 Gibralter)
8/88 - $120,000, Residence built 1973

RESEARCH AREA SHELTER BAY - SE. SIDE OF FIDALGO ISLAND

SHELTER BAY PLAT
Lot 743, Shelter Bay #4 (743 Tillamuk Dr.)
8/88 - $130,000, Residence built 1972

Lot 457, Shelter Bay #3 (457 Klickitat Dr.)
8/88 - $123,000, Residence bui-It 1973

RESEARCH AREA SKYLINE - FIDALGO ISLAND

SKYLINE PLAT
Lot 85, Div. 8 (5104 Kingsway)
7/88 - $98,700, Residence built 1970

Lot 41,, Div. 11 (2205 Dover Drive)
7/88 - $125,000, Residence built 1980

-5-

ALPINE APPRAISAL SERVICE
j
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUINI. WASHINGTON 98382
(206) 683-7084

REPORT - FLOATING SALMON NET PENS

SITE #3: DISCOVERY BAY - Clallam CoLinty, WA.
Township 30 North, Range 2W, W.M.

DATES OF INSPECTION: August 19 & 20, 1988

PURPOSE AND FUNCTION OF REPORTz To compare Discovery Bay with
Peale Passage and Rich Passage. The i:Ltnction is to provide
information UsefLil in siting floating salmon net pens.

CERTIFICATION AND LIMITING CONDITIONS: The Standard
Certifications and Limiting Conditions are attached.

ALPINE APPRAISAL SERVICE
REAL ESTATE APPRAISERS
150 S. 5TH AVE. SUITE 14 SEQUIM. WASHINGTON 98382
(206) 683-7084

AREA DESCRIPTION: Discovery Bay is a large bay situated between
the Miller Peninsula and the Quimper Peninsula. A portion of
the bay is located in Jefferson County and a portion is in
Clallam County. The portion of the bay where the proposed
floating salmon net pens are located is in Clallam County.

Discovery Bay and Rich Passage are similar with areas of fairly
dense development, areas of scattered homes along the
waterfront, and some unimproved properties. Peale Passage is
relatively undeveloped by comparison with only 4 or 5 areas
where groups of waterfront homes are clustered near the end of a
road.

TOPOGRAPHICAL INFORMATION: Discovery Say has a much wider
variety of terrain than either Peale Passage or Rich Passage.
Discovery Bay has several no bank waterfront areas such as
Diamond Point, Beckett Point, and Gardiner. It also has many
medium to.high bank areas. Elevations near the waterfront range
from 11' to 600' feet above sea level.

SIZE AND DESCRIPTON OF PENS: It is my understanding that the
proposed Discovery Say pens will eventually be 100 feet by 1000
feet in area and are to be steel pens.

I do not know what materials were used in constructing the Peale
Passage Pens as they were not visible from any of the areas I
visited. The Rich Passage pens are constructed from a variety
of materials. Some were wood, others a combination of wood and
steel, and others were steel.

VISIBILITY OF PENS FROM DIFFERENT LAND LOCATIONS AND ELEVATIONS
AROUND DISCOVERY BAY: Photographs (50 MM) were taken from
different areas as shown an the attached Gardiner quadrangle map
exhibit. These areas were visited and photos taken of the
proposed fish pen location to make comparisons with the other
areas visited. Approximate distances from the proposed pens and
elevations at the site are noted near each picture.

CONCLUSION: After visiting all three study areas and evaluating
the information gathered in the field it is my opinion that
areas over 2400 lineal feet from the floating net pens will have
little visual impact and their property values will not be
adversely affected. Residential areas less than 2460 lineal
feet from the pens will have some Visual impact.

ALPINE APPRAISXL SERVICE
REAL ESTATE APPRAISERS
150 S. STH AVE. SUITE 14 SEQUIN. WASHINGTON 98382
(206) 683-7084

In Peale Passage and Rich Passage floating net pens were
originally located in areas with no residential development
within 2400 feet of the pens. In the past 3 years two new
waterfront homes with a view of the floating net pens were built
on the north side of Rich Passage and several waterfront parcels
adjoining the Domsea Dock and Pens in this same area are in the
process of being developed and sold. Because of the lack of
sales history for properties within this distance it is not
possible to make any direct value comparisons, however, the
building and development activity in the area indicates the
impacts have been minimal. This is consistent with my personal
experience as a real estate appraiser. Over the past B years, I
have appraised many waterfront and water view properties in
Jefferson and Clallam Counties. I have found that waterfront
and marine activities do not adverse]y affect upland and
waterfront property values.

APPENDIX L

ECONOMIC ASPECTS OF SALMON AQUACULTURE

The Northwest Environmental journal, 5:37-52, 1989
University of Washington, Seattle, Washington 98195

Economic Aspects of Salmon Aquaculture

James A. Crutchfield'

Scope and Purpose
Aquaculture, broadly defined, includes shellfish culture, ocean
ranching (i.e., the hatchery production of selected stocks which are
released to the ocean and harvested upon their returr), and pen-
rearing of various species of finfish. This paper focuses on the ex-
plosive and controversial growth in farming of salmon. Pen-rearing
of salmon is definitely where the action and the emotions are cen-
tered in Washington State, although controversy also has arisen over
the proposed rearing of edible seaweed (nori) and the expansion of
shellfish culture.
The policy issues involved in the disputes over pen-rearing in
Washington waters tend to be viewed as regional, but it is impossible
to assess them without considering similar issues elsewhere in the
world. Our local expertise and capital are drawn, in part, from abroad,
and farmed salmon now are standardized items in world trade.
The first part of this paper deals with the development and current
economic situation of salmon farming in the broader setting. The
second part discusses the potential role of Washington State in the
global market and a number of factors that give rise to some pe-
culiarly local problems. The term Puget Sound is used throughout to
mean the area from the Canadian border, south to Clallam County
on the Strait of Juan de Fuca.

World Salmon Farming: The Norwegian Experience
Pen-rearing is hardly a new toy in the fisheries world. Sophisti-
cated pond culture has been carried on in China and Southeast Asia
for centuries, and there are active brackish-water rearing enclosures
in Gaeta, Italy, which once served the elite of Roman society. Ex-
perimental and modest commercial production of pen-reared salm-
on goes back to the 1970s. Indeed, much of the best research on

Professor Emeritus, Department of Economics, Graduate School of Public
Affairs, and Institute for Marine Studies, % Institute for Marine Studies
(HF-05), University of Washington, Seattle, Washington 98195.

38 NORTHWEST ENVIRONMENTAL JOURNAL Vol. 5:1 1989 SALMON AQUACULTURE ECONOM[CS 39

rearing of salmonids has been carried on in Washington State for TABLE 1. World production of Atlantic and Pacific salmon (in metric tons), 1987-
decades by scientists at the University of Washington and the Na- 1990.
tional Marine Fisheries Service. Nevertheless, large-scale commer- Salmon Country 1987 1988 19892 19902
cial production under controlled conditions is a relatively new de-
velopment. Catfish in the United States, yellowtail and shrimp in Atlantic Canada 800 1,600 3,200 5,000
Japan, and shrimp in Latin America now support solidly established Faroe Islands 4,800 4,800 7,100 9,000
Iceland 800 1,800 2,500 5,000
industries. But nothing matches the excitement and the economic Ireland 2,200 4,500 6,600 10,100
impact of the explosive growth of pen-reared salmon, which fol- Norwayl 53,000 80,000 84,000 107,000
lowed years of limited success and experimentation in Washington, England 13,900 15,000 20,000 25,000
British Columbia, Chile, and elsewhere. United States 800 1,700 3,200 5,200
Pacific salmon (coho and chinook) remain the mainstay of pen Sub-total 76,300 109,400 139,400 166,300
operations in Japan, British Columbia, Chile, New Zealand, and the Pacific Canada 3,200 8,400 14,600 23,000
Chilel 1,700 3,500 15,400 17,000
state of Washington. In a dozen years, the output of farmed coho Japani 13,000 15,000 17,000 30,000
in Japan has risen to an estimated 15,000 metric tons (mt); Atkinson New Zealand 1,000 1,500 2,000 3,000
(1987) has forecast production of 30,000 mt by 1990. British Columbia United States 1,700 2,000 2,400 2,500
operations also are based largely on coho and chinook, and have Sub-total 20,600 30,400 51,400 75,500
grown very rapidly with relaxation of tight government restrictions. Total 96,900 139,800 190,800 241,800
Between 125 and 160 licensed farms presently are active. New Zea- I Figures for Norway, Japan, and Chile have been adjusted on the basis of more
land's farms are devoted to chinook, and Washington State's pen- recent reports.
rearing operations are producing coho and a small quantity of chi- Source: Cited in Anderson 1987. Norway continues to expand (from a report by
JETRO quoting data by National Marine Fisheries Service (NMFS)I. Suisan Tsushin,
nook. Tokyo. November 11, 1988. (Printed in Japanese.)
Except for Japan, however, change is in the wind. Washington's 2 Projected by above JETRO report quoting NMFS data.
largest coho farm was sold recently to Norwegian interests and will
st- : Ct lo A C salmon; --!! of the a-licatinne for nermitq nnw
IL11 L I., - rr --- ---
before the State are for Atlantics. Recent investments in Chile and unexpected problems-is expected to exceed 100,000 tons by 1990.
British Columbia by large European multinationals also will be de- This spectacular record has not been without its bumps and chuck-
voted to that species. Japanese firms, on the other hand, continue holes, however. The industry struggled in its early years to meet
to concentrate on coho in both production and imports, although color and size requirements for its dominant European market. In
some New Zealand chinooks and Norwegian Atlantics are now 1985, a temporary decline in prices caused some concern. In 1986
found in Japanese markets. At present, Atlantic salmon account for and 1987, production was curtailed by an outbreak of disease. In
about 78 percent of world production. 1988, a massive onset of growth of algae required physical removal
As might be expected in this fluid situation, estimates of total of many pens in Norway; though few fish were lost, the cost of
world production of pen-reared Atlantic and Pacific salmon from moving them out of danger was substantial. Scotland also has suf-
different sources vary widely. The current forecast figures in Table fered from algae blooms in recent years.
1 are believed to be reasonably reliable. All sources share a common The success of the Norwegians in pen-rearing Atlantic salmon
theme, however: Very large increases in the supply of pen-reared was not happenstance; it resulted from a combination of excellent
salmon will reach U.S., European, and Japanese markets in the next natural endowments, supportive government policy, and an ener-
few years. getic and well-organized industry (Lavin-Riely and Anderson 1986).
The real breakthroughs in rearing Atlantic salmon came first in Atlantic salmon can be "grown out" in pens from smolt to desirable
Norway, where a strong government program of research and de- market size (2-6 kg) in about 18 to 24 months, and are less susceptible
velopment grew out of the need for new economic opportunities to disease than Pacific species. Despite general depletion of Atlantic
in the depressed coastal fishing areas. Table 2 tells a tale of almost salmon after World War 11, Norway had a fairly good supply of
uninterrupted success. From a meager 3,500 metric tons in 1978, genetically-diverse wild stocks from which to draw. Its long coast-
production is projected at nearly 80,000 tons in 1989, and-barring line provided excellent water temperature regimes and an abun-

40 NORTHWEST ENVIRONMENTAL JOURNAL Vol. 5:1 1989 SALMON AQUACULTURE ECONOMICS 41
TABLE 2. Supply of farmed Atlantic salmon: Norway and other countries (in metric tered areas in Maine. The British Columbia Ministry of Agriculture
tons), 1978-1990. and Fisheries estimates that there are more than 700 potential sites
Other Percent of in British Columbia alone (DPA 1986).
Year countries Norway Total total The "internationalization" of salmon farming certainly would
1978 250 3,500 3,750 93.0 have occurred as a result of market forces alone. Another factor was
1979 470 4,150 5,598 89.9 the availability of suitable sites in many parts of the world. But the
1980 1,445 4,153 5,598 74.2 process has been accelerated by Norwegian fishery policies. (For an
1981 1,133 8,422 9,555 88.1 excellent summary of these policies, see Bjorndal 1988.) Briefly stat-
1982 3,527 10,265 13,792 74.4 ed, the government's approach to the new pen-rearing industry has
1983 2,839 17,016 19,855 85.7 been dominated by two goals: (1) to place farms in regions with
1984 4,552 22,300 26,852 83.1
1985 6,500 28,655 35,155 81.5 limited employment opportunities (perhaps resulting in less mi-
1986 10,600 38,000 48,600 78.2 gration to large urban areas); and (2) to gear the pace of production
1987 14,700 50,000 64,700 77.3 to growth in the market for pen-reared fish and, thus, stabilize prices
1988 N/A 80,000 98,000 at profitable levels. These objectives were reflected in tight restric-
1989* tions on the number of smolt-rearing and feed-out farms (where
1990* 40,000 107,000 147,000 72.8 hatchery-raised smolts that have been moved to rearing pens-
Source: The University of Stirling Institute for Retail Studies Market Reports, Volume usually in natural saltwater-are fed in the pens); limitations on
No. 2, pp. 42, 43. 1988 and 1990 estimates from trade sources.
Projected by the University of Stirling Institute for Retail Studies. farm size and multiple-site ownership; and insistence on industry-
wide pricing, quality control, and the provision of genetically-strong
eggs and smolt. The government also has subsidized the develop-
dance of sites relatively free Of pollution. There was little public ment of a multi-mode transportation system linking fish farms and
opposition to this new water use. Government aids, in the form of their suppliers (Mylchreest 1985).
scientific research, capital funds, subsidies, and quality control, were These policies have, in general, achieved their goals, but they also
used effectively. The industry rapidly developed marketing and spurred the export of Norwegian capital and expertise to other coun-
production strategies that made year-round supplies of high-quality tries. Experience and technological progress brought an awareness
fresh salmon widely available. Even if account is taken of the sub- that government restrictions were preventing the realization of
sidies, salmon farming has proved highly attractive to Norwegian economies of scale in the size of individual farms, operation of
investors. When 150 new permits were made available recently, multiple sites, and vertical integration of smolt production, feed-
there were more than 1,500 eager applicants. out, processing, and marketing. Many of Norway's farms are smaller
than the allowable 8,000 cubic meters. (The Norwegian government
The World joins In authorized a 12,000-cubic-meter limit but as of this writing, it has
not been implemented.) Several studies (e.g., Bjorndal 1988; Sal-
The boom has become international, although Norway's share of vanes 1986) indicate that operations several times that size would
the total production of Atlantic salmon was still over 70 percent in probably bring important reductions in unit costs. This, coupled
1987 (Table 2). The United Kingdom, Canada, Chile, Japan, Iceland, with the sobering realization that Norwegian salmon farmers would
Ireland, New Zealand, and the United States-together with five soon feel the pinch of stabilizing prices and increased foreign com-
other nations, several of them long active at low levels-now are petition, lead to a continuing effort to bring Norwegian skills and
expanding production rapidly. capital to locations free of these restrictions. In the U.S., Canada,
In addition, substantial quantities of farmed salmon may reach Chile, Scotland, and elsewhere, the dominant factor in salmon farm-
U.S. and European consumers from small 11 pockets" of good rearing ing has been the Norwegian presence. (For an excellent summary
sites that are favorably located near transportation and markets. For of Norwegian investment activity in the U.S. and Canada, see Parker
example, New Brunswick is producing Atlantic salmon for sale in 1988.)
the U.S. and is reported to have the potential for as much as 10,000 Norway provides an example of establishing quality controls
tons annually .. Substantial quantities may be forthcoming from shel- through a central agency (the Fish Farmers Sales Organization),

42 NORTHWEST ENVIRONMENTAL JOURNAL Vol. 5:1 1989 SALMON AQUACULTURE ECONOMICS 43

backed by government inspection and export certification. This ex- for sale in the imaginative plastic packaging developed for use in
ample has been followed, through government or industry action, retail chains and supermarkets.
by all major producing countries except the United States (Ringstad From the standpoint of the important European salmon-smoking
1986). Given the reliance on high and consistent quality as a device industry, the ability to buy raw materials, as needed, significantly
for market penetration, this type of control probably is essential to reduces holding costs (i.e., storage, freezer capacity, quality main-
prevent "free-riding" by less scrupulous producers and marketers. tenance, and interest charges).
In short, the U.S. development of pen-rearing potential must be Note the emphasis on continuity of supply of market-sized fish,
accompanied by measures to assure industry-wide adherence to high a problem never completely solved in efforts to build a strong market
quality standards. for pen-reared coho. Few farm operations, even with multiple sites,
can deliver farmed Atlantic salmon in fresh form through a full
year, but the growing diversity of sources in both northern and
Markets southern hemispheres means that the entire market can very nearly
meet that goal. Skillful use of the freezer, and care in subsequent
The phenomenal growth in farmed salmon production raises the handling, can bridge the remaining gap. For obvious reasons, mar-
two obvious questions of where it is going and how markets can keting of farmed fish is heaviest in the eight months when fresh
absorb such quantities. wild salmon are scarce or unavailable. It also should be noted that
The markets for salmon are highly segmented. The major markets some of the coho production is marketed at pan-size and competes
(U.S., Japan, United Kingdom, Germany, and France) distinguish with trout rather than other fresh salmon.
wild from farmed salmon; coho, chinook, and sockeye from Atlantic; In short, the growth in sales of farmed salmon to date reflects a
troll-caught from net-caught Pacific fish; fresh from frozen; and number of real marketing advantages over wild fish in the high
small from large fish. Cross-elasticities of demand (i.e., the sensi- quality fresh-fish-oriented segments in which they have concen-
tivity of demand in one segment to changes in prices in another) trated. The ability to maintain a high growth rate, however, will
are not identical, but are linked to some degree. Thus far, farmed require much broader penetration, probably with a wider variety
'ed into the rnarkets f0r. s-ahmo.- -1A. -
saimuit have I'M of end products. With wild salmon landings holding fairly steady
demand for fresh fish "out of season," and by edging into the Eu- at around 650,000 mt, farmed salmon will account for about 14
ropean smoked salmon markets traditionally served by wild fish or percent of total supply by 1990. The figure is much higher for wild
U.S. frozen fish. There is, contrary to one widely expressed view by Atlantics, Pacific chinook, and coho-the three species which com-
salmon farmers, both direct and indirect competition with wild fish, pete most directly. There is a big marketing job to be done in the
but farmed salmon clearly are making up the growth component. near future.
The "white tablecloth" restaurants are by far the most important
outlets for quality fresh salmon, and it is in this segment of the Industry Structure
market that farmed fish offer their greatest appeal. Restaurants can-
not afford to promote a high-priced item like salmon without a The basic structure of salmon farming is essentially that of a
guarantee of consistent supply. Unlike troll-caught chinook and marine feedlot operation. Selected salmon eggs are reared in a hatch-
coho, farmed Atlantics are available in uniform sizes, quantities, ery and raised to smolts of 35-50 g in a second stage. The smolts
and guaranteed freshness year-round. In particular, they fill the long then are moved to rearing pens, usually in natural saltwater envi-
gap left by the highly-seasonal availability of wild salmon. ronments, and grown to marketable size (about 2-6 kg). The fish
For the busy restaurateur, the ability to order by telephone to are harvested and bled on site for transportation to processors and
meet exact needs, rather than touring fish markets to examine in- then are frozen or shipped by air or truck to export destinations.
dividual lots, is an attractive feature that reduces costs and increases Subsequent distribution in the United States is handled through
menu flexibility. established fresh-fish marketing channels, with a small quantity sold
Uniformity of size and quality, and steady supply also may be the in frozen form. A rough estimate of price relationships in the U.S.
principal characteristics that make farmed salmon a prime candidate at various states in the process is shown in Table 3.

44 NORTHWEST ENVIRONMENTAL JOURNAL Vol. 5:1 1989 SALMON AQUACULTURE ECONOMICS 45
TABLE 3. Cost structure for fresh farmed salmon from Norway, in dollars (USD). Some of these threats will doubtless be reduced by the sheer
USD pressure of the marketplace, but always at some additional cost to
Price FOB Norway 6.56/kg producers. And the number of unknown and uncontrollable loss
Air freight ex Norway 1.74 factors will remain large enough to make pen-rearing a high risk
Margin of importer (7%-10%) 0.83 business.
Selling price of importer 9.13 Finally, the availability of high quality feed has been a recurrent
Overland freight 0.22 worry, particularly in newly developed salmon farming regions.
Margin of wholesaler Q0%) 1.87 The fish meal industry is geared to a huge demand for poultry and
Selling price to restaurant 11.22 animal feeds, which are not suitable for fish. In the intermediate
Source: Ringstad 1986. term, improved knowledge of nutritional requirements at each stage
of development (from egg to mature salmon) and the growth in the
size of the market for salmon feed will provide the market incentives
Biochemical Problems to meet the problem; in the interim, however, it remains an im-
portant concern in some areas.
The industry has struggled in its early phases with a number of
economic problems. Smolt production has been carried on as a sep- The Trend Toward Concentration
arate operation in most cases, and often has fallen short of demand
from the rapidly growing farm operations. The scramble for smolts It is not surprising that these unresolved biotechnical problems,
has, in turn, added to the problem of straying of escaped farm fish together with better access to capital and markets, have tended to
and genetic disruption of wild salmon stocks on which the Nor- push the salmon-farming industry toward fewer and larger units,
wegian industry is still dependent. Briefly, this results from mixing better able to reduce risks through multiple-site operations and to
of wild and farmed fish and the subsequent dilution of the complex tie successive input requirements (eggs, smolts, and feed) to con-
set of genetic characteristics that tailor each wild stock to the specific trolled sources through vertical integration. This trend has been
conditions of its river of origin. In Norway these problems have impeded by Norway's regulatory orientation, but is clearly evident
been addressed by a centralized governmental research program to in other countries such as Canada, Scotland, and Chile. A study of
develop brood stocks of appropriate diversity, as well as by careful British Columbia farms showed that about 60 percent of their pro-
control of imports. These are inherent problems that must be ad- duction came from 15 percent of the farms (DPA 1986).
dressed, by public or combined industry action, in any salmon farm- Geographic decentralization of the industry will doubtless con-
ing area. In Washington State the permitting process requires careful tinue, not only because of Norway's restrictive policies, but also
attention to the potential impact of salmon farmers on wild and because of high transport costs. Air freight to the U.S. from Norway
hatchery stocks. averages about $1.74/kg and up, which provides a substantial um-
A major factor affecting costs of pen-rearing is the need to control brella for farmers located closer to consuming centers. Chilean fish
disease and to improve survival rates. Like any other marine animal, are brought to the West Coast by one large U.S. firm in chartered
diseases become more troublesome as stocking density increases. Boeing 707s-an outlay that can only be absorbed because of south-
Drug treatment is expensive and raises questions about transfer of ern Chile's low wages and freedom from controversy about envi-
potentially toxic material to other marine organisms or to human ronmental problems. Canada, of course, has a clear advantage with
consumers. In Norway, the higher incidence of disease in older fish respect to transportation to U.S. markets.
has at times restricted the size to which salmon can be grown; this
runs counter to market preference for larger fish. Salmon Farming in Washington State
There is always the risk of adverse developments in the complex
marine environment on which pen-rearing depends. Changes in Why has Washington been so slow in joining the pen-rearing
salinity, temperature, algae blooms, and food web relations (not to boom? Washington has had some early experience in farming coho
mention oil spills, waste disposal, and other human insults) pose salmon-some of it marginally successful and some disastrous. A
contingent threats to salmon farms. great deal of sophisticated and practical research in rearing of sal-

46 NORTHWEST ENVIRONMENTAL IOURNAL Vol. 5:1 1989 SALMON AQUACULTURE ECONOMICS 47
monids has been carried on by the University of Washington and of scenic values for shoreline and upland property owners, and
the National Marine Fisheries Service. The state's potential sites restrictions on quality water-based recreational activities, particu-
have been investigated, largely by technically competent domestic larly boating. Unfortunately, there is no operative price mechanism
and Norwegian interests. Yet only 7-13 marine rearing operations which would help establish a schedule of best uses for these waters.
now are believed active, compared to some 125-160 in British Co- But the losses are certainly very real to those directly affected, as
lumbia. evidenced by the widespread opposition that has developed in the
Proponents of aquaculture are quick to point a finger at the slow greater Puget Sound region. Regional environmental quality is a
pace of governmental site certification and the failure of the state composite of many things, and it is vulnerable to the " nibbling"
to provide firm guidelines for site approval and farm operation. process that accompanies growth in population and industry. More-
There is some measure of truth in their position. For example, there over, environmental losses are all-too-frequently irreversible.
are definite conflicts between state agencies and local governments . It must be stressed that Puget Sound already is an intensively
that must be resolved. The concerned agencies-the Governor's of- utilized land/water system. Water transportation, commercial and
fice and the departments of Agriculture, Fisheries, and Natural Re- recreational fishermen, pleasure boaters, beachcombers, and shore-
sources-are strong supporters of rapid expansion of aquaculture. line residents all compete in varying degrees for use of these waters.
Local governments are more exposed to the -direct heat of constit- Virtually all of the desirable pen-rearing sites will put salmon farmers
uents who are concerned about negative effects and are generally in direct conflict with some other users.
opposed to anything more than token pen-rearing. They argue, Unlike other major sea-farming areas, e.g., the west coast of Nor-
again with considerable justification, that they pay most of the ex- way, Scotland's north coast and islands, and southern Chile, there
ternal costs and assume all of the environmental risks of salmon is no logical argument that salmon farming is needed in isolated
farming with extremely small economic benefits (see "Economic areas of Puget Sound where unemployment and labor immobility
Non-Issues," below). Coalitions of opposition groups originating at create serious social problems. The Puget Sound region is the most
the country level have become increasingly vocal at the state and prosperous in the state, and even the distressing condition of the
local levels. commercial salmon fishery is largely offset by the ready availability
it -,eems clear that the ratitious attitude of the state government of off-season employment. For most commercial fishermen. Wash-
toward action, as opposed to rhetoric, is a response to real public
concerns about the impact of a large pen-rearing industry on local ington offers a vari@ty of better jobs than unskilled manual labor
areas and on the state as a whole. Weston's studies (1987) and ex- on salmon pens.
perience in Norway confirm the likelihood of water quality deg-
radation in limited areas from fecal matter and unused feed. But Economic Non-Issues
how extensive or persistent these effects may be, particularly after The real issues in the controversies over site approvals have been
long periods of operation, remains an open question which can only obscured by a number of economically faulty arguments put forth
be answered on a site-by-site basis. If, for example, it turns out that by both sides. For example, proponents repeatedly have claimed
the pens must be moved every few years, the unpleasant possibility that a fully-developed Puget Sound salmon-rearing industry would
emerges of continuous struggles over certification of new sites. reduce the nation's serious international balance-of-payment prob-
I am not qualified to assess the significance of other biological lems. But in recent years, total salmon imports to the United States
impacts, e.g., transmission of diseases to native fish stocks, straying amounted to less than one-tenth of one percent of the $150 billion
and genetic damage, long-term effects of drugs and other chemicals deficit in the U.S. balance of trade.
on local biota and/or humans, and possible damage to birds or The promise of major increases in jobs and entrepreneurial op-
marine mammals. While it seems likely that the probability of cat- portunities is equally suspect. A recent study by the State Depart-
astrophic damage is small, the probability of zero damage is even ment of Trade and Economic Development (Inveen 1987) suggests
smaller. Whatever their level, biological impacts mean increased that primary employment in a typical Puget Sound pen-rearing
social costs-some that can be measured in dollars, others that can- operation would be eight to ten persons, with an average annual
not. wage of about $19,000-ranging from $14,500, to $30,000 for the
Probably the most important negative impact is the degradation manager. Capital investment required would be about $750,000-

48 NORTIIWEST ENVIRONMENTAL JOURNAL Vol. 5:1 1989 SALMON AQUACULTURE ECONOMICS 49

$1,000,000, and annual operating expenses about $1,400,000 (feed, portion (26 percent) felt that frozen farmed Atlantics compete di-
30 percent; labor, 14 percent; smolt, 12 percent; "other," 44 percent). rectly with frozen Pacifics, but this may change as more and more
Assuming eight additional jobs in secondary activities, the total imports come in frozen form from sources such as Chile and New
increase in employment from ten new salmon farms would be about Zealand.
160-200, and the number of new firms would be something less There is general agreement that exports of wild salmon to Europe
than ten: These are useful additions, to be sure, but not of major (about 10-15 percent, by value, of total U.S. production) will feel
significance. These figures are consistent with estimates of labor the impact of farmed salmon most severely. European smokers, tra-
inputs in the Norwegian industry (Bjorndal 1988). ditionally the major purchasers of frozen troll-caught chinook and
On the other side of the controversy, commercial fishermen have coho, have turned increasingly to Norwegian, Scottish, and Irish
resolutely opposed any increase in salmon farming for two reasons: farmed Atlantics. (This comment is based on an unpublished paper,
(1) the encroachment on fishing grounds or areas in which fish are "International Salmon Farming: Competition for U.S. Fishermen?"
transferred to buyers (where boats anchor during closed periods); by Stephen M. White for a fall 1986 class at the Institute for Marine
and (2) the adverse effects of farmed fish on market prices. The first Studies, University of Washington, Seattle.)
point is a legitimate one, but it could be met by identifying and For obvious reasons, the impact of farmed salmon on prices and
blocking out areas where pen-rearing activity would impact tradi- market shares for traditional fishermen is critically dependent- on
tional fishing activities. This would, however, further restrict the the long-run breadth and depth of the overall market for salmon.
number of sites that meet state guidelines. There is probably some validity to the aquaculturists' position that
The second point is more complex. Two questions are raised: year-round availability of salmon will boost demand for both farmed
Would farmed fish actually compete with wild fish in the market? and wild fish in both restaurant and retail markets. The excellent
if so, should traditional harvesters be shielded from such compe- job of quality control by marketers of farmed salmon may force
tition? To some extent competition in the fresh market is limited badly needed improvements in the handling of wild salmon, which
by timing. Wild salmon are available in fresh form only during a would add further strength. But a major, concerted promotional
"window" of about four months (and for Washington trollers and effort will be required if forecasts of 200,000-300,000 tons of farmed
net fishermen, the window is even narrower); imports of farmed salmon annually come to pass and the estimated minimum whole-
fish are heavier in the off-season. The boundaries are not that neat, sale prices of $3.50-$3.75 per pound are to be maintained.
however. Atlantic salmon can be found on restaurant menus year- From the standpoint of Washington State policy, perhaps the best
round, particularly in East Coast and Midwest markets. High quality response to the competition question is: So what? Consumer pref-
frozen wild fish, which normally filled the winter/spring gap, are erences ultimately determine the relative place of different salmon
directly competitive with farmed fish. If, as trade journals suggest, products in different markets, with price differentials serving as the
more and more farmed salmon will be frozen as production expands, allocative mechanism. This is the way private enterprise economies
competition with wild fish will broaden (Seafood Leader 1988). are supposed to work. Any effort to stifle production of a new or
Pen-reared coho, in smaller sizes, are considered more comparable better product for the benefit of an existing segment of industry
to farmed trout. Larger coho from Chile, however, are close substi- runs directly counter to the rules of the game, and would be doomed
tutes for wild fish, as are chinooks from farms in New Zealand and to failure in time. Moreover, the additional output of Washington
British Columbia. farmed salmon, even under the rosiest assumptions, would have
In aggregate terms, it is difficult to avoid the c 'onclusion that little or no measurable effect on prices that are determined by world-
farmed salmon must either moderate price increases or actually cut wide supply and demand.
real prices for domestically-harvested wild salmon. imports are much
greater than the supplies of troll-caught chinooks and coho, which Conclusions and a Look Forward
are most nearly comparable in quality. We conclude with words about roses and the thorns that go with
A recent study (Rogness and Lin 1986) offers partial confirmation them. First the good news:
of this conclusion. Seventy-nine percent of the wholesalers and
distributors responding to their survey felt that fresh farmed At- 1) If market limitations do not intrude, there are ample areas for
lantics were a direct substitute for fresh Pacific fish. A smaller pro- expansion of the physical production of salmon farming, even in

50 NORTHWEST ENVIRONMENTAL JOURNAL Vol. 5:1 1989 SALMON AQUACULTURE ECONOMICS 51

Norway where 700 pen-rearing operations are licensed. Chile, Brit- tracted, due to the long production cycle for salmon (Bjorndal 1988;
ish Columbia, and-if political opposition is overcome-Alaska have Ringstad 1986). Anderson (1988), in a recent analysis of demand for
hundreds of potential sites with suitable water and temperature farmed salmon, indicates that a supply of 125,000 mt could be moved
conditions and little or no competition for land or water. Others only at prices about 20 percent below present levels.
with growth potential include New Zealand, Ireland, Scotland, Ice- 2) Salmon farmers in some areas are beginning to feel the pinch
land, the Faeroe Islands (Denmark), and a few areas in New England as some previously "external" costs begin to fall on their shoulders.
and the Canadian maritime provinces. These include, for example, the increasing incidence of local pol-
2) Neither technical expertise nor capital appear to be limiting lution, which requires more stringent controls by the operators and,
factors over time. The former can be purchased, and the latter is in some cases, shifts in location. Adverse effects of some of the most
readily available from Norway and, increasingly, from multination- effective chemicals used to control disease and to retard fouling of
al corporations (e.g., Unilever and British Petroleum). pens may require use of less potent substitutes or location shifts.
3) Thus far, the export marketers of farmed salmon have concen- Governments cannot reasonably be expected to subsidize the in-
trated on dressed and whole fish in fresh form. New profit oppor- dustry indefinitely; eventually aquaculture must begin to contribute
tunities for exploitation of new consumer portions, preparations, to the necessary costs of research and management, and to pay full
and packaging have barely been touched. In addition to broader costs for all inputs.
consumer appeal, these developments will integrate the marketing 3) The combined effect of stable real prices and rising costs already
of farmed fish that have established food marketing channels, with has begun to cause concern about the financial condition of many
substantial savings in distribution costs. (The same avenues for im- marginal firms, even in Norway (Bjorndal 1988; Ringstad 1986). This
provement are also available to wild fish, of course.) has accelerated the search for new, low-cost production sites and
4) Shellfish culture and shrimp, salmon, and catfish rearing are will strengthen the move toward larger, integrated firms and the
by no means the end of the road for commercial aquaculture. Ex- linkage of fish farming to conglomerate international corporations.
perimental work is underway in controlled rearing of a variety of 4) Expansion of salmon farming in some areas (e.g.,.Washington)
other popular species-cod, halibut, sturgeon, Arctic char, and tur- has met determined opposition from property owners adjacent to
bot, to menti-n -nlv a fpw (Lavin-Riely and Anderson 1986). We the proposed pen sites, commercial fishermen, and some environ-
are far from the situation in animal husbandry, where centuries of mental groups. This resistance can be expected to continue, and
research, genetic modification, and field testing have identified the probably will prevent more than limited growth in areas close to
most promising species and further modified them for human use. major markets. Obviously, if the concerns about adverse environ-
Given the ubiquitous world need for more protein foods, there is mental effects turn out to be warranted, the restrictive pressure will
every reason to believe that frontiers in the cultivation of marine become much stronger.
animals will be pushed hard.
Now for the thorns: From the standpoint of all residents of the Puget Sound region,
there seems to be no reason to rush headlong into pen-rearing.
1) The trade literature reflects widespread uneasiness about the Fitting salmon farms into a heavily populated area with a wide array
timing and effect of the inevitable industry stabilization process. of water-dependent industries, recreational users, and shoreside
Markets do not expand indefinitely. As farmed salmon fill out their home owners will be a ticklish task. It will demand full public
present niches in the fish (and wider animal protein) markets, the review and evaluation of the state environmental impact statement,
relatively stable premium prices enjoyed since 1984 by farmers and detailed analysis of site-specific factors involved in each appli-
and distributors must soften, if production continues its headlong cation. Salmon farming is a legitimate claimant on Washington in-
growth. This "overrun" phenomenon, typical of new industries, shore waters, but it is only one of many. There is no apparent reason
may be deferred and its impact softened if new and much larger why it should be given special priority.
consumer groups can be reached (e.g., if high-volume chain retailers This cautious approach, essential if Puget Sound is to yield its
begin to push farmed salmon after only moderate declines in real greatest overall economic and social benefits, may delay the entrance
prices). If not, a painful shakeout period of considerably lower prices of Washington producers into the market. But what is lost? The
and bankruptcies can be expected. This period could be quite pro- market will still be there and, hopefully, growing. Farmed salmon

52 NORTHWEST ENVIRONMENTAL JOURNAL Vol. 5:1 The Northwest Environmental journal, 5:53-69, 1989
University of Washington, Seattle, Washington 98195
is an undifferentiated product, and no supplier is going to seal off
the market to new entrants. If Washington producers can meet prices
and quality standards of imported fish, they will sell readily. As Effects of Phytoplankton Blooms on Salmon
indicated above, the industry expects a fairly severe shakeout period
in the early 1990s, and it may be to the advantage of late entrants Aquaculture in Puget Sound, Washington:
to plan investment and marketing strategies after that process is Initial Research
complete.
John E. Rensel,l Rita A. Horner,' and James R. PosteJ3

References Introduction
Anderson, James C. 1988. Current and future market for salmon in the United Marine salmon farming throughout the world is expanding rap-
States. Aquaculturc International Congress and Exposition. Vancouver, B.C. In press. idly. The production of net-pen reared Atlantic salmon, Salmo salar,
Atkinson, Clinton. 1987. The fisheries and markets of Japan with special reference leads the expansion because this species commands a high market
to Salmon. Perspectives for Salmon Culture in Chile. Santiago de Chile: Fundacion value and is especially adaptable to culture conditions. Expansion
Chile. of Washington State salmon net-pen operations has slowly increased
Biorndai, Trond. 1989. The Norwegian aquaculture industry: Industrial structure since 1970 with 14 private sites operating as of late 1988.
and cost of production. Marine Policy Fall 1988, 122-141. In the early days of Washington State net-pen culture (1 970-1975),
DPA Group, Inc. 1986. Industrial organization of the B.C. salmon aquaculture industry:
Final report. Vancouver, British Columbia: Department of Fisheries and Oceans, losses of salmon were caused primarily by bacterial diseases, poor
September 1986. siting practices, and phytoplankton blooms. The first two problems
Inveen, Daniel. 1987. The aquaculture industry in Washington State: An economic over- have been lessened through the development of effective vaccines
view. Olympia: Washington State Department of Trade and Economic Development. and the move to deeper areas with stronger currents.
Lavin-Riely,PatriciaAnn,andjamesL.Anderson. 1986. ThestatusofAtlanticsalmon I Phytoplankton problems, however, persist and there has been no
aquaculture. Staff Paper 86-04. Kingston, Rhode Island: Department of Resource concerted effort to document or resolve them. In marine waters of
Economics, University of Rhode Island. western Washington, some net-pen systems have been removed due
My1chreest, Russel. 1985. Norway's Atlantic salmon aquaculture industry. Vancouver, to phytoplankton-induced losses of salmon. In 1987, phytoplankton
British Columbia: Regional Planning and Economics, Department of Fisheries and
Oceans. blooms in Washington were involved in the mortality of at least
Parker, Peggy. 1988. Salmon farm investors. Seafood Business. May/June 1988. 7(3): 250,000 Atlantic and Pacific salmon of all ages with monetary losses
78-83. over $0.5 million. Fish-farming industry officials in Washington
Ringstad, Karl. 1986. World markets for farmed salmon. Unpublished document pre- State presently consider this problem to be their number one re-
sented to Norwegian Trade Commission Fish and Farming Seminar, Seattle, Wash- search need.
ington, June 5, 1986. Besides threatening the current production of approximately 9,000
Rogness, Ronald V., and Biing-Hwan Lin. 1986. The marketing relationship between tons of Atlantic salmon per year in Washington State, the problem
Pacific and pen-raised salmon: A survey of U.S. seafood wholesalers. Alaska Sea Grant hampers expansion of the industry because it compounds the risks
Report 86-3. Fairbanks: University of Alaska. related to site development. Some private growers hesitate to con-
Salvanes, Kjell. 1986. An empirical analysis of economics of scale in the Norwegian fish
farming industry@ Discussion Paper No. 3, Institute of Fisheries Economics. Bergen: sider new sites because the costs of permit acquisition are high and
Norwegian School of Economics and Business Administration. untested sites may have unknown phytoplankton problems.
Seafood Leader. Fall 1988. International aquaculture review. Seattle: Waterfront Press
Company. 8(4):66-133.
Weston, Donald. 1987. The environmental effects of floating mariculture in Puget Sound. 'School of Fisheries (WH-10), University of Washington, Seattle, Wash-
Seattle: College of Ocean and Fisheries Science, University of Washington. ington 98195.
14211 N.E. 88th St., Seattle, Washington 98115.
3School of Oceanography (WB-10), University of Washington, Seattle,
Washington 98195.

366 -680
1.