Download Assessment of regional benthic impact of salmon mariculture within

ICES Journal of Marine Science, 58: 417–426. 2001
doi:10.1006/jmsc.2000.1039, available online at http://www.idealibrary.com on
Assessment of regional benthic impact of salmon mariculture
within the Letang Inlet, Bay of Fundy
G. Pohle, B. Frost, and R. Findlay
Pohle, G., Frost, B., and Findlay, R. 2001. Assessment of regional benthic impact of
salmon mariculture within the Letang Inlet, Bay of Fundy. – ICES Journal of Marine
Science, 58: 417–426.
Between 1994 and 1999, impact on the benthic fauna has been evaluated in two areas
with intensive salmon net-pen aquaculture (Lime Kiln Bay, Bliss Harbour). A third
embayment (Deadmans Harbour), which lacked significant aquaculture activity,
served as a reference site. Changes in benthic community structure were investigated
using multivariate, distributional, and univariate analyses. Such changes reflect
cumulative stress from various sources, including organic enrichment and chemical
therapeutants. Changes at the Deadmans Harbour reference site indicated a general
improvement in ecosystem quality during the study period. Lime Kiln Bay experienced
an increase in aquaculture activity during the first few years, followed by a complete
cessation in 1998. Analyses indicated increased biological stress on the benthic
community, suggesting major environmental alterations did take place in that bay.
Diversity declined significantly between 1994 and 1995, with a corresponding significant increase in sediment organic content. Analysis of indicator species corroborated
the enrichment trend. Sedimentary microbial biomass and organic matter concentrations decreased from 1997 to 1999, reflecting the cessation of fish farming, but the
benthos did not recover during this period. It is concluded that benthic impact in Lime
Kiln Bay persisted until 1999 even though organic loading decreased. Enrichment
levels in Bliss Harbour were elevated from the beginning of the study onwards and the
impact persisted throughout the study. The observed differences in community
structure within the three embayments were not attributable to differences in sediment
types, temperature, salinity, or water depth.
2001 International Council for the Exploration of the Sea
Key words: aquaculture, benthos, environmental impact, salmon.
Received 16 October 1999; accepted 15 March 2000.
G. Pohle and B. Frost: Huntsman Marine Science Centre, St. Andrews, NB, Canada
E5B 2L7 (tel: +1 506 529 1203; e-mail: [email protected]); R. Findlay: Department
of Microbiology, Miami University, Oxford, OH 45056, USA. Correspondence to
G. Pohle.
Introduction
The Letang Inlet in the Bay of Fundy was one of the first
sites of commercially successful marine fish farming
operations in North America. Fish farming leads to
organic pollution in the form of fish feces and excess
food (Hall et al., 1990; Stewart, 1997; Naylor et al.,
2000) that affect the surrounding benthos (Gowen and
Bradbury, 1987; Findlay et al., 1995). As the duration of
farming operations and the number of farm leases
increased, concern grew that the industry might alter the
benthos of the entire Letang region. Since 1991, aquaculture has become the dominant source of organic
wastes within the area (Strain et al., 1995; Strain, 1998).
The difficulty of accurately measuring benthic carbon
1054–3139/01/020417+10 $35.00/0
flux makes direct documentation of organic enrichment
problematic unless visible accumulation of wastes occurs
(but see Findlay et al., 1995). Organic enrichment has
been indirectly detected by elevated levels of volatile
organic solids (predominantly carbon) and bacterial
biomass, and by changes in sediment geochemistry
(Wildish et al., 1990; Johannessen et al., 1994; Findlay
and Watling, 1997a; Hargrave et al., 1997). Changes in
benthic macrofaunal community structure have also
been widely used to detect organic enrichment (Pearson
and Rosenberg, 1978). The effects on benthic community structure of organic loading originating from fish
farms are most pronounced under, and in the immediate
vicinity of, fish cages, but less so at increasing distances
from farming operations (Pohle et al., 1994).
2001 International Council for the Exploration of the Sea
418
G. Pohle et al.
The measurement of changes in the structure of
marine communities in combination with appropriate
environmental variables is widely used for the detection
and monitoring of human impact on the marine
environment (Pearson and Rosenberg, 1978; Warwick
et al., 1990; Agard et al., 1993). Among the different
biota, the soft-bottom macrobenthos is the most widely
utilized constituent for community structure analysis to
study environmental impacts (Warwick, 1993). Community and sediments (Hansen, 1994) represent an integrative measure of impact over time. Advantages of the use
of the bottom-dwelling faunal component include that it
is resident year round, is naturally abundant, diverse
and relatively easy to sample quantitatively, while also
being largely unexploited or intentionally managed.
Most importantly, this group is relatively immobile
and therefore useful in studying effects of pollutants
(Bilyard, 1987), such as organic enrichment, by responding through changes in population dynamics on a time
scale of weeks to years. The benthos can be also used in
the environmental impact assessment of chemical therapeutants, because many resident component species,
such as polychaete worms, may be sensitive to these
chemicals (Black et al., 1997).
Enrichment effects reflect the cumulative impact from
all contributing sources over large areas. Our objective
was to assess regional impacts on the benthic fauna
within two areas of intensive aquaculture in the Letang
Inlet over time, while monitoring a comparable area
devoid of farming activities.
Materials and methods
Sampling
For long-term, far-field monitoring, three sampling grids
were set up in the Letang region of the lower Bay of
Fundy (Figure 1): Lime Kiln Bay, representing an
established area of intensive aquaculture, Bliss Harbour,
with a somewhat more recently developed salmon farming industry, and Deadmans Harbour, serving as a
reference embayment without aquaculture activities
since December 1992.
To monitor for regional rather than site-specific
impact, sampling grids in Lime Kiln Bay and Bliss
Harbour were located so that the stations were at the
greatest possible distance, and more or less equidistant,
from cage operations. This resulted in any aquaculture
site being at least 200 m from the sampling grid. A
similar sampling grid was established in Deadmans
Harbour.
The original size of the 24-ha grids (400600 m)
sampled during the first year was reduced to 6–9 ha
(250–300150–300 m) in later years, based on sediment
characteristics within grids, which should be uniform to
avoid confounding effects. Only samples taken within
the reduced area were analysed. In December 1994,
1995, and 1997 and February 1999, 10–13 grab samples
were taken at each of the three area sites. However,
some samples from Lime Kiln Bay and Deadmans
Harbour were suspect because of sediment inconsistencies, or lacked sediment data, and were therefore disregarded in the analysis (up to four out of ten samples at
any given sampling period). Sampling stations were
originally randomly selected within each grid and resampled subsequently. Approximate water depth at
mean low tide ranged from 10 m to 16 m.
All samples were collected from the 13-m long RV
‘‘W. B. Scott’’, with a winch-operated, 0.04-m2 modified
Ponar grab sampler. Samples were sieved on-board
through 400-m screens, with the retained portion fixed
in 10% buffered formalin. Faunal components were
identified and enumerated in the laboratory using established procedures (Eleftheriou and Holme, 1984). Nontypical taxa and taxa that were not consistently retained
by the screen (e.g. nematodes) were excluded from the
analysis.
Characterization of sediments in terms of the silt/clay
fraction followed the methodology of Buchanan (1984).
Core subsamples, 5 cm in diameter and 7–8 cm deep,
were taken from grab samples at all stations. Sediment
subsamples were also taken from the top 2 cm of all grab
samples with cut-off 5-cc disposable syringes for the
determination of two enrichment measures: total microbial biomass was ascertained according to the phospholipid technique developed by Findlay et al. (1989),
while sediment organic carbon content was determined
by the surrogate variable percent volatile organic matter
obtained as weight loss by ignition, as proposed by
Wildish et al. (1990) and applied by Johannessen et al.
(1994).
Statistical analyses
Univariate and multivariate analyses were carried out
using PRIMER (Version 4, Plymouth Routines in
Multivariate Ecological research), with its various
subroutines. STATGRAPHICS PLUS for Windows
(Version 3.1, Manugistics Inc.) was used for plotting and
tests of significance of univariate measures. Tests of
significance were based on standard ANOVA for univariate measures, and on the ANOSIM equivalent for
k-dominance curves (Clarke, 1990) and multivariate
analyses (Clarke, 1993). Significant differences referred
to in the text were at the 5% level. Multiple Range Tests
were used to determine significant differences between
means. The data for total microbial biomass were
natural log-transformed prior to statistical analysis, and
evaluated for variance and normal distribution by
assessing residual plots and employing Cochran’s C,
Bartlett’s, and Hartley’s tests. For all data that did not
fully meet normality criteria (six out of 15 cases), the
Assessment of regional benthic impact of salmon mariculture
419
New Brunswick
N
Letang
45°10' N
1000 m
Scotch
Bay
ur
bo
ar
Letang
Peninsula
ng
H
a
et
L
45.05
Back
Bay Back
d
Li
ea
m
Frye
Island
eK
iln
Ba
y
Bay
ng
ta
ur
Le
Hilis
Island
Bliss
H
rbo
Blacks
Harbour
s
ad
n
ma
De
Ha
ad
De
ma
ns
He
ad
Harbour
d
an
s
lis
White
Head
Isl
Bay of Fundy
B
'
66.85
66°80' W
Figure 1. Map showing sampling areas (arrows) and sampling stations (dots) within the Bay of Fundy: Lime Kiln Bay, Bliss
Harbour and the Deadmans Harbour reference site. Hatched areas denote lease sites for aquaculture (actual net-pens occupy less
area; two sites in Lime Kiln Bay discontinued operations in 1998).
non-parametric Kruskal-Wallis test was employed, but
results did not differ from those based on ANOVA.
Community structure was analysed using three distinct statistical methods with different levels of sensitivity (Warwick and Clarke, 1991; Clarke and Warwick,
1994), including species-independent univariate indices,
a graphic descriptor method and species-dependent
multivariate techniques (Gray et al., 1988).
Univariate measures of community structure included
taxonomic richness, defined as the number of distinct
taxa per sample and the Shannon-Wiener diversity index
(H; Gray et al., 1992). Graphical descriptor techniques
can be thought of as an intermediate between univariate
summaries and full multivariate analysis (Clarke, 1990).
Unlike univariate measures, they extract information on
patterns of relative species abundance without reducing
that information to a single summary statistic; unlike
multivariate techniques, they extract universal features
of community structure which are not the function of
the specific taxa present, but related to levels of biological stress (Clarke and Warwick, 1994) that are visualized
as a graphical representation of impacts. We used
k-dominance curves, with cumulative ranked abundance
plotted against species rank, in decreasing order of their
importance. Species rank is log-transformed for a
smoothing effect on curves and to better visualize the
distribution of commoner species. More elevated curves
have lower diversity, supposedly representing more
impacted sites (Lambshead et al., 1983).
Multivariate methods were used to classify samples
showing similar community attributes into groups, indicating the degree of similarity or dissimilarity in species
composition at the same station over time (Clarke,
1993). Log transformations were needed to minimize
distortion when compressing multidimensional data.
The Bray–Curtis similarity coefficient (Bray and Curtis,
420
G. Pohle et al.
Table 1. Statistical comparison of sediment silt-clay fraction
(%) in samples taken at Lime Kiln Bay, Bliss Harbour and
Deadmans Harbour in 1994, 1995, and 1997 (no F-test was
significant at 0.05% level).
1994
1995
1997
(n=7)
79.4
1.5
(n=7)
83.6
1.0
(n=7)
88.3
1.7
(n=13)
79.6
1.1
(n=13)
84.9
0.8
(n=13)
88.1
1.2
(n=7)
78.7
1.5
(n=10)
82.8
0.9
(n=10)
84.4
1.4
(a)
1
2
2
2
4
4
Lime Kiln Bay
Mean
Standard error (pooled s)
Bliss Harbour
Mean
Standard error (pooled s)
Deadmans Harbour
Mean
Standard error (pooled s)
4
2
2
2
2 2
2
2
2
4
4
3
3 3
3
2
2
1
3
3
1
1
4
4
3
1
1
1
1957) was used to relate the overall similarity between
any samples by taking all the taxa into consideration.
Unlike other coefficients, it is not affected by joint
absences and is, therefore, sufficiently robust for marine
data. The resulting triangular array of similarity values,
or coefficients, between all pairs of samples was used to:
(1) discriminate samples from each other (ANOSIM
test); (2) cluster sites into groups that have similar
communities (hierarchical agglomerative clustering);
and (3) allow a gradation of sites to be represented
graphically where the relative distance apart of any
pair of samples is intended to reflect their relative
dissimilarity (Ordination). Here we present only plots
derived from non-metric multidimensional scaling
(MDS) because of theoretical advantages (Clarke and
Green, 1988) while also being empirically more robust
(Warwick et al., 1988). The extent to which sample
relations can be adequately represented in a twodimensional map is expressed by a ‘‘stress coefficient’’
statistic, here simply called stress (Clarke and Warwick,
1994).
Changes in community structure were linked to
environmental changes by statistical analysis of the
univariate measures. Findlay and Watling (1997b) established a direct linkage between benthic carbon flux
induced increases in benthic oxygen demand to the
benthic impacts associated with salmon net-pen aquaculture. As such, the parameters measured to examine
the issue of causality of changes in macrobenthic community structure included sedimentary microbial biomass and organic content. As macrobenthic community
structure is also controlled, in part, by sediment grain
size, the latter was included in the analysis.
Results
Particle size analysis (PSA) during the first three years
indicated a consistently high silt/clay fraction, ranging
from 70% to 95% (Table 1). During 1999, no data were
(b)
1
1
2
1
2
3
4
1 1
2
3
1
2 2 2
4 33
22
1
2 2
4 4
1
4
2
3 33 2 2 1
1
33 4
4
43
1
4
4
1
1
(c)
3
3
2
3
3
4
4 4
3
3
3
4
3
4
4 3
4
1
1
1
3
4
4
1
1
2
2
2
2
2
4
2
2
1
1
2
2
2
2
2
Figure 2. Two-dimensional MDS ordination plots of benthic
faunal communities based on samples from (a) Lime Kiln Bay,
(b) Bliss Harbour and (c) Deadmans Harbour taken during
1994 (1), 1995 (2), 1997 (3), and 1999 (4).
collected but there was no visual indication that any
major changes occurred in sediment characteristics.
Mean temperature and salinity during sampling periods
were 4.71.6C and 33.31.0 psu.
Multivariate analysis of community structure indicated that sets of samples for 1994 (1), 1995 (2) and 1997
Assessment of regional benthic impact of salmon mariculture
100
(a)
90
Cumulative % dominance
80
70
60
50
40
30
20
10
0
1
10
100
10
100
100
(b)
90
Cumulative % dominance
80
70
60
50
40
30
20
10
0
1
100
(c)
90
Cumulative % dominance
80
70
60
50
40
30
421
(3) differed significantly for all three areas (Figure 2),
with no significant shifts in population structure occurring between 1997 and 1999 (4). While samples from
Deadmans Harbour [Figure 2(c)] show overlap during
the first two and the last two years, all annual samples of
Lime Kiln Bay [Figure 2(a)] and Bliss Harbour [Figure
2(b)] are clearly separated during the first three years,
indicating more pronounced differences than at the
reference site.
K-dominance curves of samples from Lime Kiln Bay
[Figure 3(a)] show a statistically significant difference in
pattern, with increased elevation during all years compared to 1994, indicating a persistent reduction in diversity and thus an increased impact beginning in 1995.
During the following two years, apparently the impact
did not increase, because the curves largely overlap with
1995. The consistent elevation compared to 1994 suggests that conditions have not returned to those of 1994.
Bliss Harbour shows no indication of significant changes
in biological stress, with overlapping curves for all years
[Figure 3(b)]. The Deadmans Harbour reference site
showed a reduction in curve elevation, particularly in
1999, compared to 1994, but these differences were not
statistically significant [Figure 3(c)].
Univariate analyses for Lime Kiln Bay also provide
evidence of deteriorating conditions for the benthic fauna
in 1995 that persisted in later years, with both taxonomic
richness [Figure 4(a)] and the diversity index [Figure 4(b)]
declining significantly from 1994 to 1995, and remaining
low in 1997 and 1999. In concordance, organic matter
levels increased significantly in 1995, with a decreasing
trend thereafter [Figure 4(d)]. Microbial biomass [Figure
4(c)] did not change during the first two years, increased in
1997, and declined significantly between 1997 and 1999.
For Bliss Harbour, univariate community structure
indices [Figure 5(a),(b)] showed no profound changes
over the years. Environmental variables showed mixed
results, organic matter levels decreasing significantly in
1997 and remaining steady thereafter, while bacterial
levels dropped only slightly [Figure 5(c),(d)].
At the reference site, the number of taxa increased
significantly during the last two sampling periods, while
the diversity index did not change significantly over the
period [Figure 6(a),(b)]. However, the mean of the index
has increased since 1995, and was significantly higher in
1999 than in 1995. Bacterial biomass and organic matter
levels have steadily declined since 1995, with significantly lower values in 1999 than in 1994 [Figure
6(c),(d)].
20
10
0
1
10
Species rank
100
Figure 3. Mean k-dominance curves for amalgamated data of
replicate samples taken in Lime Kiln Bay (a) Bliss Harbour (b)
and Deadmans Harbour (c) by year. MDS stress values are: (a)
0.14; (b) 0.20; (c) 0.18. ——, Year 1 (1994); ——, Year 2
(1995); – – – –, Year 3 (1997); · · ·· · ·, Year 4 (1999).
422
G. Pohle et al.
70
3.5
(a)
(b)
3.4
65
3.3
60
3.2
Diversity (H' )
Number of taxa
55
50
45
40
3.1
3.0
2.9
2.8
2.7
35
2.5
25
2.4
240
8.9
(c)
(d)
8.4
200
% Organic matter
Bacterial biomass (nMol P g dry wt–1)
2.6
30
160
120
80
7.9
7.4
6.9
6.4
40
5.9
1994
1995 1997
Year
1999
1994
1995 1997
Year
1999
Figure 4. Univariate impact measures of Lime Kiln Bay for 1994, 1995, 1997, and 1999: (a) taxa richness; (b) Shannon–Wiener
Diversity index; (c) sedimentary microbial biomass; (d) percent organic matter. Data points and vertical lines represent means and
95% confidence intervals.
Discussion
The Letang Inlet is the primary area of mariculture
within the Bay of Fundy, producing the greatest proportion of reared salmon among aquaculture regions.
Within Lime Kiln Bay and Bliss Harbour, areas encompassing 3–4 km2 each, one can find a half dozen farming
operations with a combined annual production license
that increased from about 350 t in the 1980s to 700 t
during the late 1990s. In contrast, the roughly similarly
sized Deadmans Harbour historically contained a single
fish farming operation that was abandoned in December
1992. Two fishing weirs were the only other possible
sources of enrichment at the reference site. In the Letang
Inlet, potential sources of enrichment other than from
aquaculture include sewage outfall and fish processing.
However, since 1991 aquaculture represents the dominant source of organic wastes, primarily in the form of
excess fish feed and faeces, in the area as measured by
carbon, nitrogen, phosphorus and BOD inputs (Strain
et al., 1995; Strain, 1998).
In Lime Kiln Bay, a moderate mean current predominates that forms a clockwise eddy, whereas a mean
clockwise current of more or less the same magnitude
runs parallel to the shore at the Bliss Harbour site
(Trites and Petrie, 1995). Current data were not available for Deadmans Harbour, but from sampling experience mean flow is judged not to be greater than at the
other two sites. Findlay and Watling (1997b) have
demonstrated that the assimilative capacity of benthic
communities for carbon is greatly influenced by nearbottom current velocities. The similarity of current flows
Assessment of regional benthic impact of salmon mariculture
70
3.5
(a)
423
(b)
3.4
65
3.3
60
Diversity (H' )
Number of taxa
3.2
55
50
45
40
3.1
3.0
2.9
2.8
2.7
35
2.5
25
2.4
240
8.9
(c)
(d)
8.4
200
% Organic matter
Bacterial biomass (nMol P g dry wt–1)
2.6
30
160
120
80
7.9
7.4
6.9
6.4
40
5.9
1994
1995 1997
Year
1999
1994
1995 1997
Year
1999
Figure 5. Univariate impact measures of Bliss Harbour for 1994, 1995, 1997, and 1999: (a) taxa richness; (b) Shannon–Wiener
Diversity index; (c) sedimentary microbial biomass; (d) percent organic matter. See also Figure 3.
in the three embayments suggests that differences in
benthic carbon inputs – and not differences in assimilative capacity at each site – were responsible for any
observed changes in benthic community structure. All
three sampling sites are net depositional in nature, and
are predominated by fine bottom sediments. Mean
current speed in the experimental areas is less than
5 cm s 1 (Trites and Petrie, 1995).
Multivariate analysis showed that, with few exceptions, samples from individual sampling years formed
clusters (Figure 2), indicating a relatively uniform community structure within each area. Possible confounding
variables, such as water depth, temperature, salinity and
sediment particle size (Table 1) cannot account for the
observed changes in community structure from the
period 1994–1997 because they remained virtually
constant.
The k-dominance curves (Figure 3) for Lime Kiln Bay
displayed a significantly elevated curve for samples from
1995, suggesting a major impact on the community in
this year. A persistent biological stress is implied for
1997 and 1999 [Figure 3(a)], although the lower curves
in 1997 and 1999 might be interpreted as indicating some
improvement. Such changes were not apparent in Bliss
Harbour. While results for the Deadmans reference
site do not indicate major changes, the slightly lower
curves during later years imply some improvement in
environmental conditions.
Univariate diversity measures agree to a large extent
with findings of the graphical distribution analysis.
Although the fauna in Lime Kiln Bay (Figure 4) was
significantly less diverse in 1995, the indices suggest no
improvement in subsequent years while the graphical
analysis does. Like the k-dominance curves, univariate
424
G. Pohle et al.
70
3.5
(a)
(b)
3.4
65
3.3
60
Diversity (H' )
Number of taxa
3.2
55
50
45
40
3.1
3.0
2.9
2.8
2.7
35
2.5
25
2.4
240
8.9
(c)
(d)
8.4
200
% Organic matter
Bacterial biomass (nMol P g dry wt–1)
2.6
30
160
120
80
7.9
7.4
6.9
6.4
40
5.9
1994
1995 1997
Year
1999
1994
1995 1997
Year
1999
Figure 6. Univariate impact measures of Deadmans Harbour for 1994, 1995, 1997 and 1999: (a) taxa richness; (b) Shannon-Wiener
Diversity index; (c) sedimentary microbial biomass; (d) percent organic matter. See also Figure 3.
Table 2. Mean abundance of indicator species in grab samples taken at Lime Kiln Bay over the
five-year period.
Indicator species
1994
1995
1997
1999
Diplocirrus hirsutus
Crenella glandula
Astarte spp.
Photis macrocoxa
Diastylis spp.
Nucula proxima, N. delphinodonta
4.6
32.3
10.0
9.3
15.1
48.9
0.7
0.9
1.8
1.8
1.4
127.2
0.7
4.6
3.0
0.0
0.6
98.6
0.0
4.7
4.6
0.4
0.9
38.3
diversity measures for Bliss Harbour (Figure 5) show no
significant changes over the entire period, although there
is a trend of decreasing diversity over the last three
sampling periods. In Deadmans Harbour, there is some
indication of increased diversity over time (Figure 6), in
agreement with the graphical analysis.
Within Lime Kiln Bay, Pohle and Frost (1997) determined impact indicator species (Gray and Pearson,
Assessment of regional benthic impact of salmon mariculture
1982) based on data from a transect ranging from
stations next to cages to the middle of the bay. Table 2
provides mean abundance data for matching taxa for the
samples collected during the four years in the present
study. The results strongly support a major impact
between 1994 and 1995. The polychaete Diplocirrus
hirsutus, the bivalves Crenella decussata and Astarte
spp., the amphipod Photis macrocoxa and the
cumaceans Diastylis spp. all decreased dramatically in
mean abundance. The nut shells Nucula proxima and N.
delphinodonta, which, unlike the previous taxa, thrive in
a moderately enriched environment, became significantly more abundant after 1994. Results for 1997 and
1999 suggest in general some improvements compared to
1995 but not a return to conditions of 1994. No patterns
were evident for these indicator species at Bliss and
Deadmans Harbour.
The significant shift in community structure in Lime
Kiln Bay shown by multivariate analysis and indicator
species is interpreted as an impacted community with
decreased diversity. The shift correlates with environmental measures, particularly a significant increase in
sediment organic matter, providing evidence that
organic enrichment is the likely cause of the detrimental change in community structure (Pearson and
Rosenberg, 1978). Levels of organic matter and
microbes decreased during the last two sampling
periods, which is consistent with results from the graphical analysis. All analyses combined suggest detrimental
changes in community structure related to enrichment
within the area. The measured decline in organic loading
between 1997 and 1999 correlates with the cessation of
fish-farming operations in 1998 owing to an outbreak of
infectious salmon anemia in the Letang that peaked in
1997. However, a corresponding recovery of the benthic
macrofauna community had not taken place by 1999.
Analyses for Bliss Harbour were conflicting. While the
more sensitive species-dependent multivariate analysis
showed major changes in community structure along a
gradient during 1994–1997, the species-independent
graphical and univariate methods showed no significant
changes. One explanation may be that, unlike in Lime
Kiln Bay, concentrations of organic matter were already
elevated in Bliss Harbour at the onset [Figure 5(d)]. As
in Lime Kiln Bay, abiotic variables indicate decreased
organic loading during later years without appreciable
improvements in biotic variables. While concentrations
of organic matter in both areas have dropped, they are
still above 7%, which is considered to reflect enriched
conditions for the upper Letang (Wildish et al., 1977),
and are also higher than at the reference site. The
persistence of enrichment may explain the lack of
recovery in community structure in both aquaculture
regions.
At the Deadmans Harbour reference site (Figure 6),
organic content and microbial biomass decreased sig-
425
nificantly over the four years. This correlates with an
increased diversity according to univariate and graphical
analyses. All information taken together indicates that
the benthic community experienced improving environmental conditions. This may reflect recovery from
enrichment by the aquaculture site that was in operation
up to 1992, which, if true, suggests that improvements in
community structure may occur slowly over many years.
This would explain the lack of recovery in Lime Kiln
Bay and Bliss Harbour within a year after cessation of
aquaculture operations in those areas.
The consistent results obtained indicate that the monitoring system established and the techniques employed
are effective tools in detecting regional changes in
community structure in relation to fish-farming
activities.
Acknowledgements
Foremost we thank B. Clarke, Plymouth Marine Laboratory, UK, for his advice and assistance on statistics.
We thank the New Brunswick Department of Agriculture, Fisheries and Aquaculture for support and
acknowledge assistance by the crew of the ‘‘W. B.
Scott’’, in particular Jack Eldridge, Eldon Carter, and
Fred Purton. Lou and Tyler Van Guelpen, Fernando
Marques and Mark Burgess provided fieldwork assistance. Samples were processed by Mary Greenlaw. Art
MacIntyre helped with the preparation of figures.
References
Agard, J. B. R., Gobin, J., and Warwick, R. M. 1993. Analysis
of marine macrobenthic community structure in relation to
pollution, natural oil seepage and seasonal disturbance in
a tropical environment (Trinidad, West Indies). Marine
Ecology Progress Series, 92: 233–243.
Bilyard, G. R. 1987. The value of benthic infauna in marine
pollution monitoring studies. Marine Pollution Bulletin, 18:
581–585.
Black, K. D., Fleming, S., Nickell, T. D., and Pereira, M. F.
1997. The effects of ivermectin, used to control sea lice on
caged farmed salmonids, on infaunal polychaetes. ICES
Journal of Marine Science, 54: 276–279.
Bray, J. R., and Curtis, J. T. 1957. An ordination of the upland
forest communities of Southern Wisconsin. Ecological
Monographs, 27: 325–349.
Buchanan, J. B. 1984. Sediment analysis. In Methods for the
Study of Marine Benthos, pp. 41–65. Ed. by N. A. Holme,
and A. D. McIntyre. Blackwell Scientific Publications,
London. 387 pp.
Clarke, K. R. 1990. Comparisons of dominance curves. Journal
of Experimental Marine Biology and Ecology, 138: 143–157.
Clarke, K. R. 1993. Non-parametric multivariate analyses
of changes in community structure. Australian Journal of
Ecology, 18: 117–143.
Clarke, K. R., and Green, R. H. 1988. Statistical design and
analysis for a ‘‘biological effects’’ study. Marine Ecology
Progress Series, 46: 213–226.
426
G. Pohle et al.
Clarke, K. R., and Warwick, R. M. 1994. Change in marine
communities: an approach to statistical analysis and
interpretation. Natural Environment Research Council, UK.
144 pp.
Eleftheriou, A., and Holme, N. A. 1984. Macrofauna techniques. In Methods for the Study of Marine Benthos,
pp. 140–216. Ed. by N. A. Holme, and A. D. McIntyre.
Blackwell Scientific Publications, London. 387 pp.
Findlay, R. H., and Watling, L. 1997a. Seasonal variation in
sedimentary microbial community structure as a backdrop
for the detection of anthropogenic stress. In Molecular
Markers in Environmental Geochemistry, pp. 49–64. Ed. by
R. P. Eganhouse. ACS Press, Washington, DC.
Findlay, R. H., and Watling, L. 1997b. Prediction of benthic
impact for salmon net-pens based on the balance of benthic
oxygen supply and demand. Marine Ecology Progress Series,
155: 147–157.
Findlay, R. H., King, G. M., and Watling, L. 1989. Efficacy of
phospholipid analysis in determining microbial biomass in
sediments. Applications in Environmental Microbiology, 54:
2888–2893.
Findlay, R. H., Watling, L., and Mayer, L. M. 1995. Environmental impact of salmon net-pen culture on Maine marine
benthic communities: A case study. Estuaries, 18: 145–179.
Gowen, R. J., and Bradbury, N. B. 1987. The ecological impact
of salmonid farming in coastal waters: A review. Oceanography and Marine Biology Annual Review, 25: 563–575.
Gray, J. S., and Pearson, T. H. 1982. Objective selection of
sensitive species indicative of pollution-induced change in
benthic communities. I. Comparative methodology. Marine
Ecology Progress Series, 9: 111–119.
Gray, J. S., Aschan, M., Carr, M. R., Clarke, K. R., Green,
R. H., Pearson, T. H., Rosenberg, R., and Warwick, R. M.
1988. Analysis of community attributes of the benthic
macrofauna of Frierfjord/Langesundfjord and in a mesocosm experiment. Marine Ecology Progress Series, 46: 151–
165.
Gray, J. S., McIntyre, A. D., and Stirn, J. 1992. Manual of
methods in aquatic environment research. Part 11. Biological
assessment of marine pollution with particular reference to
benthos. FAO Fisheries Technical Paper, 324. 49 pp.
Hall, O. J., Anderson, L. G., Holby, O., Kollberg, S., and
Samuelsson, M.-O. 1990. Chemical fluxes and mass balances
in a marine fish cage farm. I. Carbon. Marine Ecology
Progress Series, 61: 61–73.
Hansen, P. K. 1994. Benthic impact of marine fish farming. In
Proceedings of the Canada–Norway Workshop on Environmental Impacts of Aquaculture, pp. 77–81. Ed. by A. Ervik,
P. K. Hansen, and V. Wennevik. Fisken og Havet, 13.
Hargrave, B. T., Phillips, G. A., Doucette, L. I., White, M. J.,
Milligan, T. G., Wildish, D. J., and Cranston, R. E. 1997.
Assessing benthic impacts of organic enrichment from
marine aquaculture. Water, Air and Soil Pollution, 99:
641–650.
Johannessen, P. J., Botnen, H. B., and Tvedten, Ø. F. 1994.
Macrobenthos: before, during and after a fish farm. Aquaculture and Fisheries Management, 25: 55–66.
Lambshead, P. J. D., Platt, H. M., and Shaw, K. M. 1983. The
detection of differences among assemblages of marine benthic species based on an assessment of dominance and
diversity. Journal of Natural History, 17: 859–874.
Naylor, R. L., Goldburg, R. J., Primavera, J. H., Kautsky, N.,
Beveridge, M. C. M., Clay, J., Folke, C., Lubchenco, J.,
Mooney, H., and Troell, M. 2000. Effect of aquaculture on
world fish supplies. Nature, 405: 1017–1024.
Pearson, T. H., and Rosenberg, R. 1978. Macrobenthic succession in relation to organic enrichment and pollution of the
marine environment. Oceanography and Marine Biology: an
Annual Review, 16: 229–311.
Pohle, G., and Frost, B. 1997. Establishment of standard
benthic monitoring sites to assess long-term ecological modification and provide predictive sequence of benthic community succession in the inner Bay of Fundy, New Brunswick.
Final report to the New Brunswick Department of Fisheries
and Aquaculture. 119 pp.
Pohle, G. W., Lim, S. S. L., and Frost, B. R. 1994. Benthic
monitoring of salmon aquaculture sites by the Huntsman
Marine Science Centre: effects of organic enrichment on
benthic macrofaunal communities in the lower Bay of Fundy,
Canada. In Proceedings of the Workshop on Ecological
Monitoring and Research in the Coastal Environment of
the Atlantic Maritime Ecozone, pp. 92–100. Ed. by B. M.
MacKinnon, and M. D. B. Burt. Environment Canada –
Atlantic Region, Occasional Report, 4.
Stewart, J. E. 1997. Environmental impacts of aquaculture.
World Aquaculture, 28: 47–52.
Strain, P. 1998. Letang Inlet: industry and water quality.
In Presentations from the Technical Session on Aquaculture/Environment Research Priorities. NBFDA/AECC
Workshop, January 1998.
Strain, P. M., Wildish, D. J., and Yeats, P. A. 1995. The
application of simple models of nutrient loading and oxygen
demand to the management of a marine tidal inlet. Marine
Pollution Bulletin, 30: 253–261.
Trites, R. W., and Petrie, L. 1995. Physical oceanographic
features of Letang Inlet including evaluation and results from
a numerical model. Canadian Technical Report of Hydrography and Ocean Sciences, 163: iv+55 pp.
Warwick, R. M. 1993. Environmental impact studies on marine
communities: pragmatical considerations. Australian Journal
of Ecology, 18: 63–80.
Warwick, R. M., and Clarke, K. R. 1991. A comparison of
some methods for analysing changes in benthic community
structure. Journal of the Marine Biological Association of
the United Kingdom, 71: 225–244.
Warwick, R. M., Carr, M. R., Clarke, K. R., Gee, J. M., and
Green, R. H. 1988. A mesocosm experiment on the effects of
hydrocarbon and copper pollution on a sublittoral softsediment meiobenthic community. Marine Ecology Progress
Series, 46: 181–191.
Warwick, R. M., Platt, H. M., Clarke, K. R., Agard, J., and
Gobin, J. 1990. Analysis of macrobenthic and meiobenthic
community structure in relation to pollution and disturbance
in Hamilton Harbour, Bermuda. Journal of Experimental
Marine Biolology and Ecology, 138: 119–142.
Wildish, D. J., Poole, N. J., and Kristmanson, D. D. 1977.
Temporal changes of sublittoral macrofauna in L’Etang Inlet
caused by sulfite pulp mill pollution. Fisheries and Marine
Service Technical Report, 718: 13 pp.
Wildish, D. J., Martin, J. L., Trites, R. W., and Saulnier, A. M.
1990. A proposal for environmental research and monitoring
of organic pollution caused by salmonid mariculture in the
Bay of Fundy. Canadian Technical Report of Fisheries and
Aquatic Sciences, 1724: iii+24 pp.