Marine Environmental Research 57 (2003) 103–120 www.elsevier.com/locate/marenvrev
Fish communities and life history attributes of English sole (Pleuronectes vetulus)in Vancouver Harbour
Colin Levings*, Stacey Ong Fisheries and Oceans Canada, Science Branch, West Vancouver Laboratory, 4160 Marine Drive, West Vancouver, British Columbia, Canada V7V 1N6
Abstract Data on demersal fish abundance, distribution, and spatial variation in community composition are given for Vancouver harbour and a far field reference station in outer Howe Sound. Flatfish (F. Pleuronectidae) were the dominant taxa in the trawl sampling, with the English sole (Pleuronectes vetulus) one of the most abundant species, especially in Port Moody Arm. Cluster and ordination analyses suggested a different community in Port Moody Arm relative to the outer harbour and the reference site. Length data from English sole suggested the Vancouver harbour fish may be from a different population relative to the far field reference station, with more juveniles in the harbour. Both male and female English sole were older and larger in Port Moody Arm and females were more common in this area. Growth rates of female English sole were slower at Port Moody and Indian Arm in comparison to the central harbour. Feeding habits of English sole were different at various parts of the harbour, with possible implications for con- taminant uptake. The diet of English sole was dominated by polychaetes in Port Moody Arm and by bivalve molluscs at the far field reference station. Fish from the middle and outer harbour fed on a mixture of polychaetes, bivalve molluscs, and crustaceans enabling multiple pathways for bioaccumulation of pollutants. # 2003 Published by Elsevier Ltd. Keywords: Fish; Feeding; Habitat; Life history; Sediments; Pollution tolerance; Chemical pollution; Community function
* Corresponding author. Fax: +1-604-666-3497. E-mail address: [email protected] (C. Levings).
0141-1136/$ - see front matter # 2003 Published by Elsevier Ltd. doi:10.1016/S0141-1136(03)00063-1 104 C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120
1. Introduction
The species composition of fish communities has been proposed as a key variable to assess the biological integrity of estuarine ecosystems (Deegan, Finn, Ayvazian, Ryder-Kieffer, & Buonaccorsi, 1997) and can also be used as a monitoring variable to detect changes in coastal water quality (Horn, 1979). In this paper, we describe the spatial changes and relative abundance of demersal fish in Vancouver harbour caught during the PICES Workshop and provide an analysis of fish community variation in this area. Demersal fish are bottom dwelling species which are in contact with the seafloor, and their biological characteristics are thought to reflect sediment conditions such as the presence of contaminants and changes in fish food inverte- brates. There are some baseline data available on demersal fish communities in Vancouver Harbour collected in 1985 (Goyette & Thomas, 1987), but in general the ecology of this taxocene is poorly documented in Vancouver Harbour. English sole (Pleuronectes vetulus) was identified as a dominant demersal species and was used as an indicator species in earlier work in the harbour by Goyette and Boyd (1989) and in several recent studies in Puget Sound (Collier, Johnson, Stehr, Meyers, & Stein, 1998). The physiological and health status of individual English sole was studied extensively by other investigators in the PICES Workshop (Bolton, Stehr, Boyd, Burrows, Tkalin, & Lishavskaya, 2003; Miller, Addison, & Bandiera, 2003; Stehr, Myers, Johnson, Spencer, & Stein, 2003). To complement their work, and to gain an insight into possible modes of contaminant effects on English sole basic information on length, weight, age, feeding, and growth was obtained from autopsied specimens.
2. Methods
2.1. Field sampling
A small otter trawl (mesh size in body/wing 38 mm, 3.2 mm in codend; width of opening estimated 4.9 m) was towed by the NOAA vessel HAROLDW. STRE- ETER at five stations, on a presumed pollution gradient from the inner to outer Vancouver Harbour and into Howe Sound (Levings, Stein, Stehr, & Samis, 2003). Station T-38 was considered representative of Port Moody Arm; Station T-48 of the Indian Arm location; T-11B of the central harbour; T-49 of the outer harbour; and T-50 of the far field reference location in Thornbrough Channel. Each station was sampled between three and seven times. The net was towed between 5 and 10 min, and sampled an estimated area of between 1643 and 8570 m2 in each trawl (Table 1). Coordinates of the trawl locations and other basic sampling data are given by Stehr and Horiguchi (2001). The total catch from each trawl was sorted by species, then enumerated by species and weight. The larger invertebrates such as Dungeness crab (Cancer magister), tan- ner crab (Chionocetes tanneri), anemone (Metridium spp.) and a few species of bivalve molluscs were also enumerated and weighed. Fish catch data were standardized to C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120 105
Table 1 Number of trawls, depth ranges, and estimated sample area for trawl samples in Vancouver harbour and Thornbrough Channel
Site name Station name No. of trawls Depth range (m) Sample area (m2)
Port Moody T-38 3 11–14 10896 Indian Arm T-48 3 26–30 15224 Central harbour T-11B 4 24–26 23642 Outer harbour T-49 6 30–45 38359 Thornbrough Channel T-50 3 55–73 12131
100 m2 using the sample areas shown in Table 1. When available, the total length (mm) of about 100 English sole from each station was obtained. We refer to these data as the general population length data. Data on weight, stomach content, and age were obtained for each English sole specimen autopsied by Stehr et al. (2003) for physiological condition and histopathology. The minimum size for the latter studies was 25 cm, the approximate length of sexual maturity for this species. After autopsy, the stomach was removed from each fish and preserved in 10.0% formalin. For aging, the right otolith was removed and placed in a glycerol-thymol mixture. The raw data from each trawl catch, fish ages, and stomach content data are presented elsewhere (Stehr & Horiguchi, 2001).
2.2. Laboratory methods
Contents of a random sample of ten English sole stomachs from each station were examined in the laboratory. A Wild M-5 Stereomicroscope was used to enumerate organisms, which were identified to the major group level. Ages were determined by the Fish Aging Unit, DFO Science Branch, Pacific Biological Station, Nanaimo. Condition factor was computed using Fulton’s K where K=wt/l3 105.
2.3. Statistical methods
Parametric tests, primarily ANOVA analyses were used to determine the sig- nificance of biological parameters such as abundance, length, and age distributions, with chi-square used for proportion data. A cluster analysis using the weighted pair group average linkage method (Statistica, Statsoft, Tulsa, OK) was used to investi- gate community structure using standardized numerical catch data from each of the five stations, averaged over all the replicate trawls. The latter data were also used in a principal component analysis. The Shannon index of species diversity (H0) and evenness (E) for each station was calculated using the formulae:
s 0 H ¼ -fg sum pilogpi i¼1 106 C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120
Where i is the proportion of the total sample belonging to ith species (average numerical data) and
E ¼ H0=ln S
Where ln S is the natural logarithm of the total number of species found at a station.
3. Results
3.1. Fish community data
Mean number of species obtained in each trawl ranged from 11 0.5 (Mean- SEM) at Port Moody Arm to 12.2 0.2 at the outer harbour (Fig. 1a). However based on the total number of species caught in the trawls at a particular site, the fish community at the central harbour was most diverse (19 species), and Port Moody Arm least diverse (12 species). The other stations showed intermediate values in number of species: Indian Arm, 16 species; the outer harbour, 17 species; and Thornbrough Channel, 17 species. Mean biomass ranged from 0.65 0.1 kg.100 m 2 (mean SEM) at Port Moody Arm to 0.15 0.1 kg.100 m 2 at the Central Harbour, and number of individuals 10 5.100 m 2 (mean SEM) to 2 0.3.100 m 2 at the latter two stations (Fig. 1b and c). Twenty-nine different fish species were obtained in the trawls (Tables 2 and 3, Appendix A). Flatfish (Pleuronectidae and Bothidae) were the most common spe- cies, especially English sole (P. vetulus), Starry flounder (Platichthys stellatus), Flat- head sole (Hippoglossoides elassodon), Dover sole (Microstomus pacificus), Rex sole (Errex zachirus), Slender sole (Lyopsetta exilis) and Rock sole (Pleuronectes bilinea- tus). Flatfish were the dominant taxa at the Port Moody Arm station accounting for more than 50% of the fish caught there. Other dominant species were the Pacific tomcod (Microgadus proximus) at Port Moody Arm, Indian Arm and the Central Harbour. The Blackbelly eelpout (Lycodes pacificus) was the most abundant species at the outer harbour. The latter species was the only dominant species that was not caught at Port Moody Arm (Table 2). Percentage species composition at the five sites differed significantly (P<0.05) after testing with w2. Species diversity ranged from about 1.4029 at Port Moody to 1.8261 at the outer harbour, and evenness showed a narrower range, from approxi- mately 0.5843 to 0.7337 (Fig. 2). Neither statistic showed significant statistical var- iation over the five sampling stations (P>0.05). Results of the cluster analysis showed that the fish community at Port Moody Arm was the least similar (3%) relative to the cluster of the other locations. Indian Arm and the central harbour were the most similar (36%), with the outer harbour joining at 28% and the combined cluster joined Port Moody Arm at the 19% level (Fig. 3). The principal component analysis (PCA) conducted on the same data showed somewhat different results. The first two components accounted for 87.76% of the C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120 107
Fig. 1. (a) Mean number of species from trawls at the five stations in Vancouver harbour and Thornbrough Channel. Stations are: T-38, Port Moody Arm; T-48, Indian Arm; T-11B, central harbour; T-49, outer harbour; T-50, Thornbrough Channel. (b) Mean number of fish (number 100 m 2) from trawls at the five stations in Vancouver harbour and Thornbrough Channel. Station identification as in (a). (c) Mean weight of fish (kg 100 m 2) from trawls at the five stations in Vancouver harbour and Thornbrough Channel. Station identification as in (a). variance, and the third an additional 10.79%. As shown in Fig. 4a, much of the variation was on principal component 2 (PC II), with Port Moody Arm, Indian Arm, the central harbour, and Thornbrough Channel grouping together and the outer harbour clearly separated. However principal component 3 (PC III) separated Thornbrough Channel from the above grouping (Fig. 4b). Pearson correlation analyses were used to investigate correlations between mean abundance of each fish species at the five stations and the PCA coordinates from the first three PCA axes. For PC I, correlations were statistically significant (P<0.05) for eulachon (Thaleichthys pacificus), blackbelly eelpout, rex sole, and slender sole. For PC II, correlations were only significant for the first two species. However these four species were more abundant at the outer harbour than any other station and their abundance there might well explain the distribution of stations on the two PCs. 108 C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120
Table 2 Percent abundance (numerical data) for fish species at the five stations sampled in Vancouver harbour and Thornbrough Channel
Species/station Port Moody Indian Central Outer Thornbrough Arm Arm harbour Harbour channel
Longfin smelt 1.7 3.1 1.2 0.0 0.0 Eulachon 0.0 0.0 0.0 <1.0 0.0 Herring 2.8 1.6 5.5 <1.0 0.0 Longnose skate 0.0 0.0 <1.0 0.0 0.0 Spiny dogfish 0.0 0.0 0.0 0.0 <1.0 Pacific hake 0.0 0.0 0.0 0.0 35.5 Walleye pollock 0.0 <1.0 0.0 0.0 0.0 Pacific tomcod 8.6 10.3 10.5 1.1 9.2 Shiner seaperch 4.1 1.0 1.0 <1.0 <1.0 Copper rockfish 0.0 0.0 0.0 0.0 <1.0 Tadpole sculpin 0.0 0.0 0.0 0.0 <1.0 Roughback sculpin 0.0 1.0 1.9 <1.0 0.0 Buffalo sculpin 0.0 0.0 <1.0 0.0 0.0 Staghorn sculpin 3.3 1.7 <1.0 <1.0 0.0 Sturgeon poacher 0.0 <1.0 1.0 <1.0 7.1 Midshipman 4.3 1.6 0.0 <1.0 <1.0 Whitespot greenling 0.0 <1.0 <1.0 0.0 0.0 Blackbelly eelpout 0.0 6.6 2.1 38.2 3.1 Flathead sole 1.0 14.7 1.0 14.1 1.6 Dover sole 13.1 0.0 0.0 1.3 <1.0 English sole 49.6 56.1 56.8 14.9 35.9 Rock sole 0.0 <1.0 6.2 1.5 <1.0 Slender sole 0.0 0.0 2.9 8.8 3.3 Starry flounder 9.6 1.0 4.5 5.7 0.0 Butter sole 0.0 0.0 <1.0 0.0 0.0 Rex sole 0.0 0.0 0.0 10.2 1.2 Sand sole 1.1 <1.0 2.6 1.3 0.0 Speckled sandab 0.0 0.0 0.0 0.0 <1.0 Pacific sandab 1.0 0.0 2.1 1.0 1.0
The coordinates of PC III were significantly correlated with six species: spiny dog- fish (Squalus acanthias), pacific hake (Merluccius productus), copper rockfish (Sebastes caurinus), tadpole sculpin (Psychrolutes paradoxus), sturgeon poacher (Agonus acipenserinus), and speckled sanddab (Citharichthys stigmaeus). Except for sturgeon poacher, all of the above species were absent from the three stations which grouped together (Port Moody Arm, Indian Arm, central harbour) when PC II and III were compared, but present at Thornbrough Channel, which was clearly sepa- rated in the PC plot.
3.2. English sole life history features
3.2.1. Abundance Mean abundance of English sole varied between the stations, ranging from about 6 1 fish.100 m 2 (Mean SEM) at Port Moody Arm to <1 1.100 m 2 at the C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120 109
Table 3 Mean abundance of fish species (number . 100 m 2 ) at the fiveStations sampled in Vancouver harbour
Species/station Port Moody Indian Central Outer Thornbrough Arm Arm harbour harbour Channel
Spiny dogfish 0.000 0.000 0.000 0.000 0.032 Longnose skate 0.000 0.000 0.005 0.000 0.000 Pacific herring 0.219 0.054 0.094 0.002 0.000 Longfin smelt 0.191 0.095 0.022 0.004 0.000 Eulachon 0.000 0.000 0.000 0.004 0.000 Pacific hake 0.000 0.000 0.000 0.000 2.921 Pacific tomcod 0.977 0.340 0.182 0.033 0.557 Walleye pollock 0.000 0.006 0.000 0.000 0.000 Blackbelly eelpout 0.000 0.239 0.045 1.411 0.215 Shiner perch 0.417 0.028 0.015 0.020 0.006 Copper rockfish 0.000 0.000 0.000 0.000 0.006 Whitespotted greenling 0.000 0.011 0.008 0.000 0.000 Roughback sculpin 0.000 0.027 0.033 0.002 0.000 Buffalo sculpin 0.000 0.000 0.004 0.000 0.000 Staghorn sculpin 0.340 0.065 0.005 0.014 0.000 Tadpole sculpin 0.000 0.000 0.000 0.000 0.020 Midshipman 0.524 0.055 0.000 0.014 0.006 Sturgeon poacher 0.000 0.021 0.012 0.000 0.368 Pacific sanddab 0.087 0.000 0.037 0.035 0.042 Speckled sanddab 0.000 0.000 0.000 0.000 0.081 Rex sole 0.000 0.000 0.000 0.362 0.149 Flathead sole 0.102 0.528 0.015 0.493 0.164 Butter sole 0.000 0.000 0.005 0.000 0.000 Rock sole 0.000 0.011 0.102 0.051 0.007 Slender sole 0.000 0.000 0.057 0.310 0.167 Dover sole 0.875 0.000 0.000 0.046 0.006 English sole 5.952 1.936 1.033 0.510 2.471 Starry flounder 0.649 0.043 0.082 0.208 0.000 Sand sole 0.105 0.027 0.047 0.056 0.000
Fig. 2. Shannon Index of diversity (H0) and Evenness (E) scores for species at the five stations in Vancouver harbour and Thornbrough Channel. Station identification as in Fig. 1a. 110 C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120
Fig. 3. Cluster analysis using numerical data on fish catch data from the five stations in Vancouver harbour and Thornbrough Channel. Station identification as in Fig. 1a.
Fig. 4. (a) Principal components plot for PC1 and PC2 using numerical data on fish catch data from the five stations in Vancouver harbour and Thornbrough Channel. Station identification as in Fig. 1a. (b) Principal components plot for PC2 and PC3 using numerical data on fish catch data from the five stations in Vancouver harbour and Thornbrough Channel. Station identification as in Fig. 1a. outer harbour (Tables 2 and 3). Biomass showed the same pattern, ranging from about 0.35 0.05 kg.100 m 2 (mean SEM) at Port Moody Arm to 0.05 0.7 kg.100 m 2 at the outer harbour. Abundance and biomass at Port Moody Arm were significantly higher (P<0.05) compared to the central harbour and the outer har- bour, but differences were not significant (P>0.05) when compared to Indian Arm and Thornbrough Channel.
3.2.2. Length, sex ratio, age, and growth Fig. 5 gives length frequency data for the general population of English sole. Mean lengths and standard errors at the various stations were as follows: Port C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120 111
Fig. 5. Length distribution of English sole from the general population at the five stations in Vancouver Harbour and Thornbrough Channel. Station identification as in Fig. 1a.
Moody Arm: 24 0.5 cm; Indian Arm: 21 0.4 cm; central harbour: 20 0.4 cm; outer harbour: 23 cm 0.4; Thornbrough Channel: 20 0.2 cm. Length differences between the five stations were statistically significant with ANOVA (P<0.05). Tukey post hoc tests showed that Port Moody Arm sole were significantly larger than those from Indian Arm, central harbour, and Thornbrough Channel; Indian Arm fish were smaller than sole from the outer harbour; central harbour fish were significantly smaller than those from the outer harbour; and the Thornbrough Channel fish were smaller than those from the outer harbour. Small numbers of juvenile English sole (<15 cm) were found at each of the harbour stations but not at Thornbrough Channel (Fig. 5). Table 4 gives data on mean length and age of the autopsied English sole caught at the five stations. Mean age for males ranged about 12 years at Port Moody Arm to 6 years at the central harbour, and mean lengths from 28.5 to 22.8 cm at the same stations. Mean age for females varied from 8 years at Port Moody Arm to 6 years at 112 C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120
Table 4 Age, length, and size data on autopsied English sole from the five stations Sampled in Vancouver harbour
Species/station Port Moody Indian Central Outer Thornbrough Arm Arm Harbour Harbour Channel
Mean age of males (years) (n) 12.1 (7) 9.2 (5) 6.3 (7) 6.4 (23) 8.2 (15) Mean length of males (cm) 28.5 25.4 22.8 24.2 23.4 Mean age of females (year) (n) 8.3 (21) 6.9 (24) 6.4 (23) 6.2 (19) 6.6 (12) Mean length of females (cm) 30.0 27.3 24.5 26.9 24.9 the outer harbour, and mean lengths from 30.0 cm at Port Moody Arm to 24.5 cm at the central harbour. Mean lengths and ages were significantly different (ANOVA, P<0.05) for both sexes when data were compared across stations. However com- parisons between stations using Tukey’s test using data for males showed that for both age and sex differences were significantly different (P<0.05) only when Port Moody Arm was contrasted with the central harbour, outer harbour, and Thorn- brough Channel. Female age data were significantly higher (P<0.05) (8 years vs. 6 years) when Port Moody Arm was contrasted with the central harbour and outer harbour. With female length data, fish from Port Moody Arm were significantly longer (P<0.05) when data were contrasted with those from the central harbour and Indian Arm. However autopsied English sole from the central harbour were sig- nificantly smaller (P<0.05) from Indian Arm and the outer harbour. Female English sole were more common (w2, P<0.05) from the autopsied fish from Port Moody Arm, Indian Arm, and the central harbour) relative to the outer harbour and Thornbrough Channel stations (Tables 5 and 6).
Table 5 Distribution of male and female fish for autopsied English sole from the five stations sampled in Van- couver harbour
Site/Sex Port Moody Indian Central Outer Thornbrough Arm Arm harbour harbour Channel
Percent Female 75 86 76 45 44 Percent Male 25 14 24 55 56 Number of Fish 28 28 30 42 27
Table 6 Fulton’s condition factor K, (mean SEM) (number of fish) for autopsied female and male English sole from five stations in Vancouver harbour
Station Female English sole Male English sole
Port Moody Arm 0.787 0.01 (21) 0.726 0.02 (7) Indian Arm 0.821 0.02 (24) 0.808 0.04 (4) Central harbour 0.853 0.01 (23) 0.838 0.02 (7) Outer harbour 0.827 0.02 (19) 0.844 0.01 (23) Thornbrough Channel 0.758 0.02 (12) 0.716 0.02 (15) C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120 113
English sole growth was estimated by the slope of the regression line between age (x) and log 10 length (y) for female sole. Because of small sample sizes, correlation coeffi- cients for the regression lines were statistically significant (P<0.05) only for female sole from Port Moody Arm, Indian Arm and the central harbour (Fig. 6). Differences in the slope coefficients for the three stations were statistically significant (P<0.05). Judging from the slope coefficients, fish at the central harbour grew faster relative to fish from Indian Arm and Port Moody (0.0177, 0.0099, and 0.0127, respectively).
3.2.3. Condition factor and feeding habits Condition factors were lower for autopsied male and female English sole when data from Port Moody Arm were compared with those from the central harbour, and data from Thornbrough Channel were compared with those from the central and outer harbour. Condition factors for females were significantly lower (P<0.05) at Thorn- brough Channel in comparison to those from Indian Arm. A similar pattern was shown when data from male English sole from Port Moody Arm were compared with those from the outer harbour. All other comparisons were not significant (P>0.05). In ranked order, polychaetes, bivalve molluscs, foraminifera, amphipods, and unidentified crustaceans were represented fairly equally in English sole stomachs in Indian Arm, the central harbour, and the outer harbour (Fig. 7). At Port Moody Arm and Thornbrough Channel, polychaetes and bivalve molluscs were the domi- nant taxa in the stomachs. The mean number of organisms per stomach ranged from about 48 10.fish 1 (Mean SEM)) at the Port Moody Arm to 22 9.fish 1 at the central harbour.
Fig. 6. Relationship between age and length (Log10) for autopsied English sole from Port Moody (T-38), Indian Arm (T-48) and the central harbour (T-11B). 114 C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120
Fig. 7. Percentage (numerical data) of various taxa found in the stomachs of autopsied English sole at the five stations in Vancouver Harbour and Thornbrough Channel. Station identification as in Fig. 1a.
4. Discussion
4.1. Value of fish community data as an index of environmental effects
The fish catch data clearly show that different fish communities were found in Port Moody Arm, but further data are needed to determine if the variation is owing to environmental degradation and community disruption as was concluded with the benthic invertebrate community analyses (Belan, 2003; Je, Belan, Levings, & Koo, 2003). In the principal component analyses, stations from Thornbrough Channel and the outer harbour clearly separated from Port Moody Arm, central harbour, and Indian Arm stations. Depth differences between Thornbrough Channel (55–73 m) and the outer harbour (30–45 m) and the other stations (all <30 m) could explain some of the variation, since the abundance of deep water species such as Pacific hake and Rex sole were significantly correlated with the principal components. The cluster analyses separated Port Moody Arm from the other stations, possibly C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120 115 because English sole dominated the catches there. However, there are few data on tolerance of this species to acute or chronic pollution from organic contaminants relative to other fish species in the harbour community. Water temperature may also be a stressor for English sole in the Port Moody Arm area, since available data indicate summer temperatures can reach 16–18 C in the summer (Levings et al., 2003). In laboratory experiments, growth rates of English sole at 18 C were reduced 28–47% relative to those at 15 C(Williams & Caldwell, 1978). Levels of organic contaminants such as PAHs and pesticides were higher in sediments at the Port Moody Arm relative to other sites (Bolton et al., 2003) but it is not known if these levels were sufficient to cause direct mortality. English sole show an enhanced tolerance to low oxygen environments (Boese, 1988), but there are no data on bottom water oxygen in Port Moody Arm. None of the fish species caught in the survey was known pollution indicators. However, to our knowledge there are no marine fish species widely used as indicators, unlike the situation in some stream habitats where dominance of species such as carp (Cyprinus carpio) suggest degraded water quality (Karr, Fausch, Angermeier, Yant, & Schlosser, 1986). Finally, number of species, species diversity and evenness were about the same at each of the stations, suggesting there was no major shift in community structure in the harbour.
4.2. Comparison with previous surveys and other harbours
Detailed comparisons of our data with previous trawl surveys in Vancouver har- bour are difficult because of differences in methods used, especially mesh size. However, an earlier investigation in 1985 (Goyette & Thomas, 1987) showed that, as in our surveys, English sole were the dominant flatfish in the Port Moody Arm- central harbour area, suggesting the bottom fish community has not changed dra- matically over the past 14 years. There was no evidence from any station suggesting that the harbour’s demersal fish community was modified by the introduction of exotic fish species, as was found in San Francisco Bay (Meng, Moyle, & Herbold, 1994), or by the dominance of fish species tolerant of pollution, as described by Wong (1980) for Hong Kong Harbour. An application of an Index of Biological Integrity (IBI) to the fish community, as was done in Hamilton Harbour (Smokor- owski, Stoneman, Cairns, Minns, Randall, & Valere, 1998), would require seasonal data because of the migratory nature of many species in Vancouver harbour. As noted by Horn (1979), seasonal differences in fish communities have major influ- ences on diversity measures and need to be considered when using these metrics to indicate environmental stress. Deegan et al. (1997) developed an IBI for a Massa- chusetts estuary to test for qualitative differences in fish communities at seagrass habitats at nine sites in two different estuaries. They used eight metrics (total num- ber of species, dominance, fish abundance, number of nursery species, number of estuarine spawning species, number of resident species, proportion of benthic-asso- ciated species, and proportion abnormal or diseased). Only three of these metrics (total number of species, dominance, and fish abundance) were applicable to the fish community in Vancouver, so if this methodology were to be applied it is likely a specific, and redesigned, IBI would have to be developed. 116 C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120
4.3. Migration and exposure of English sole to contaminants in sediments
The residence time of English sole in Vancouver harbour, and hence their expo- sure to contaminants in bottom sediments, is not known. However since con- taminants were bioaccumulating in English sole by directly uptake from water or sediment, or indirectly by food, their exposure time must have been sufficient to result in higher body burdens of some chemicals (Bolton et al., 2003; Stehr et al., 2003). A notable example was the higher concentrations of HCB in liver composites from Indian Arm, Lonsdale Quay, and Port Moody (Bolton et al., 2003). Ideally, to use English sole as a ‘‘sentinel’’ species for contaminant monitoring, data are needed on the year-round residency of this species in particular habitats. Ultrasonic tagging experiments conducted by Nelson and Levings (1995) on starry flounder (Platichyths stellatus) in the Fraser estuary showed that this species moved at least 0.5 km each day. Movement in the estuary was directional or somewhat random depending on life cycle. A somewhat similar situation might be expected with English sole, but investigations are needed to confirm migration patterns. Different patterns of spawning migration to deeper water might be expected from fish in Port Moody Arm relative to fish in the outer harbour. The former could move to deeper water in Indian Arm, whereas the latter might migrate to deeper parts of the Strait of Georgia, or even to US waters in northern Puget Sound. Forrester and Ketchen (1963) found that English sole tagged off the Fraser River estuary were recovered in northern Puget Sound. Miller et al. (in press) found a positive association between the expression of the detoxification enzyme CYP1A in English sole and low molecular weight aromatic hydrocarbons and PCB levels in all the harbour sites but not in fish from Howe Sound. There was some evidence from our study from the general population length data suggesting the Thornbrough Channel English sole have different population characteristics compared to Vancouver harbour. If they were genetically different, this might be a confounding effect on CYP1A response. A few smaller fish (<15 cm; likely <2 years old) were found at all the harbour stations; this length group was missing at Thornbrough Channel. Since younger English sole are typically found in shallower water (Toole, Barnhart, & Onuf, 1987) this may have been a depth effect.
4.4. Age, growth and sex ratio
Data from the general population as well as the autopsied fish showed that Eng- lish sole were larger, grew more slowly, and were 2–3 years older in Port Moody Arm relative to the other harbour sites (Table 4). It is not known if this was a growth reduction effect from exposure to contaminants in sediments at Port Moody, or a possible density-dependent effect since the English sole population density was higher there relative to the other locations. Further data are needed to increase sample sizes, particularly younger fish, to confirm the possible growth effects we found (Fig. 6). Another explanation might be an indirect effect of the shrimp fishery in the outer harbour and Thornbrough Channel (Levings et al., 2003), where English sole have likely been killed in by-catch for many years, possibly leading to a reduc- C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120 117 tion in the number of older fish. All fish caught in our survey were much smaller than expected from the Strait of Georgia fish, in comparison to data from the last assessment 36 years ago by Forrester and Ketchen (1963). At that time 7–9 years old fish then ranged from 36–40 cm. English sole of those ages in our survey were between 24 and 29 cm. Differences in sampling methodology may be a factor since Forrester and Ketchen’s (1963) data came from the commercial fishery, which uses larger nets. Larger fish may have been able to escape from the small trawl used in our work. The predominance of female fish at Port Moody may also be due to escapement of larger fish, as female English sole grow larger than males (Toole et al., 1987). Becker (1988) found that sediment type influenced sex ratios in English sole, with higher proportions of males found on finer sediments. However our data sug- gest an opposite effect, with more females found on the muddy sediments at Port Moody Arm, Indian Arm, and the central harbour. With the exception of Thorn- brough Channel, all our samples were located on muddy sediments. Feminization of English sole owing to contaminants is another possibility (Arukwe, 2001).
4.5. Feeding in relation to available food and condition factor
It is possible there are several different different food pathways for contaminant uptake by English sole in the harbour, but further data on pollutant concentrations in food items are needed to confirm this hypothesis. English sole seemed to be capitalizing on more abundant benthic food at Port Moody Arm and Thornbrough Channel, since more polychaetes and bivalve molluscs, respectively, were found in stomachs at these two stations relative to other sampling locations. Stomach con- tents at other stations included more of a mix of polychaetes, molluscs, crustaceans, and Foraminifera, with the latter taxa likely taken in with sediment. The food items at particular stations may have been a function of availability, since benthic sam- pling (Je et al., in press) showed that polychaetes and bivalve molluscs were the most abundant invertebrates at the above two stations. Subtidal bivalves collected during the Workshop such as the Milky Venus (Compsomyax subdiaphana), showed mod- erately high levels of TBT (Horiguchi, 2001). This species is potentially available as food for English sole and it is likely other fish food invertebrates were also con- taminated at stations where pollutants were found in sediments. Condition factors (CF) for both male and female autopsied English sole were lower at Port Moody Arm and Thornbrough Channel relative to stations in the outer and central harbour. However, it is well known that CF changes with spawn- ing condition and more data on timing of spawning are needed to investigate the possibility that these changes are owing to contaminant effects.
5. Conclusions
Effects of a presumed pollution gradient in Vancouver harbour on bottom fish communities were equivocal with the limited data obtained in the Workshop survey. Since a depth gradient accompanied the presumed pollution gradient, control for 118 C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120 depth and fishing effects was compromised because of logistic considerations. How- ever other investigators in the Workshop found that body burdens of organic con- taminants in English sole were significantly higher in the inner harbour and this pattern matched the distribution of pollutants in sediments. Further data on seaso- nal changes and migration patterns would be helpful to explain the processes involved in the bioaccumulation, as well as bioassay experiments to test the toxicity of sediment contaminants to indicator species including English sole. The seasonal change data would also be useful for development of an index of biological integrity or another community metric to assess marine environmental quality. Life history data from autopsied English sole showed that older, larger fish dominated in the Port Moody Arm and female fish were also more abundant in this area. Condition factors for English sole were lower at the Port Moody Arm stations and a far field reference site relative to central and outer harbour stations. However, since the harbour English sole populations may have different spawning timings, seasonal changes in condition factor also need further investigation. In Port Moody Arm the fish appeared to be feeding opportunistically on the increased population of poly- chaete worms at this location. Multiple food pathways are likely involved in con- taminant bioaccumulation in English sole.
Acknowledgements
Thanks are extended to Dan Lomax for his key role in operating the STREETER during trawling. We also sincerely thank all the participants for their assistance during field work on the STREETER, and to Carla Stehr for enabling us to work with autopsied English sole specimens. We are also most grateful to staff of the fish aging unit at the DFO Pacific Biological Station in Nanaimo for age determination. Dr. Ricardo Scrosati assisted with the principal component analysis. The DFO Science Branch, Pacific Region, provided funding for stomach content identification and statistical analyses.
Appendix. Scientific and common names of fish caught in the PICES trawl surveys
Common name Scientific name Eulachon Thaleichthys pacificus Longfin smelt Spirinchus thaleichthys Pacific herring Clupea harengus pallasi Longnose skate Raja rhina Spiny dogfish Squalas acanthias Pacific hake Merluccius productus Walleye pollock Theragra chalcogramma Pacific tomcod Microgadus proximus C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120 119
Shiner perch Cymatogaster aggregata Copper rockfish Sebastes caurinus Tadpole sculpin Psychrolutes paradoxus Roughback sculpin Chitonotus pugetensis Buffalo sculpin Enophrys bison Staghorn sculpin Leptocottus armatus Sturgeon poacher Podothecus acipenserinus Midshipman Porichthys notatus Whitespotted greenling Hexagrammos stelleri Blackbelly eelpout Lycodes pacificus Flathead sole Hippoglossoides elassodon Dover sole Microstomus pacificus English sole Pleuronectes vetulus Rock sole Lepidopsetta bilineata Slender sole Lyopsetta exilis Starry flounder Platichthys stellatus Butter sole Pleuronectes isolepis Rex sole Glyptocephalus zachirus Sand sole Psettichthys melanostictus Speckled sanddab Citharichthys stigmaeus Pacific sanddab Citharichthys sordidus
References
Arukwe, A. (2001). Cellular and molecular responses to endocrine modulators and the impact on fish reproduction. Marine Pollution Bulletin, 42, 643–655. Becker, D. S. (1988). Relationship between sediment character and sex segregation in English sole, Parophrys vetulus. Fishery Bulletin (US), 86, 517–524. Belan, T. A. (2003). Marine environmental quality assessment using polychaete taxocene characteristics in Vancouver Harbour. Marine Environmental Research, 57, 89–101. Boese, B. L. (1988). Hypoxia-induced respiratory changes in English sole (Parophrys vetulus Girard). Comparative Biochemistry and Physiology, 89A, 257–260. Bolton, J. L., Stehr, C. M., Boyd, D. T., Burrows, D. G., Tkalin, A. V. & Lishavskaya, T. S. (2003). Organic and trace metal contaminants in sediments and English sole tissues from Vancouver Harbour, Canada. Marine Environmental Research, 57, 19–36. Collier, T. K., Johnson, L. L., Stehr, C. M., Meyers, M. S., & Stein, J. E. (1998). A comprehensive assessment of the impacts of contaminants on fish from an urban waterway. Marine Environmental Research, 46, 243–247. Deegan, L. A., Finn, J. T., Ayvazian, S. G., Ryder-Kieffer, C. A., & Buonaccorsi, J. (1997). Development and validation of an Estuarine Biotic Integrity Index. Estuaries, 20, 601–617. Forrester, C. R., & Ketchen, K. S. (1963). A review of the Strait of Georgia Trawl Fishery. Bulletin of the Fisheries Research Board of Canada, 139, 78. Goyette, D., & Boyd, J. (1989). Distribution and environmental impact of selected benthic contaminants in Vancouver Harbour, British Columbia, 1985–1987. Environment Canada Regional Program Report, 89(02), 99. Goyette, D., & Thomas, M. (1987). Vancouver harbour benthic environmental quality studies May 1985 to September 1986. Relative species abundance and distribution—trawl catch. Environment Canada, Regional Data Report, 87(03), 62. 120 C. Levings, S. Ong / Marine Environmental Research 57 (2003) 103–120
Horiguchi, T. (2001). Tributyl tin in all molluscs sampled from Vancouver Harbour and Victoria. In C. M. Stehr, & T. Horiguchi (Eds.), Environmental assessment of Vancouver Harbour: data report for the PICES practical workshop. PICES scientific report no. 16 (pp. 206). Sidney BC, Canada: North Pacific Marine Science Organization (PICES). Horn, M. H. (1979). Diel and seasonal variation in abundance and diversity of shallow-water fish popu- lations in Morro Bay, California. Fishery Bulletin (US), 78, 759–770. Je, J.-G., Belan, T., Levings, C. D., & Koo, B. J. (2003). Changes in benthic communities along a presumed pollution gradient in Vancouver Harbour. Marine Environmental Research, 57, 121–135. Karr, J. R., Fausch, K. D., Angermeier, P. L., Yant, P. R., & Schlosser, I. J. (1986). Assessing biological integrity in running waters: a method and its rationale. Illinois Natural History Survey Special Publica- tion, 5, 28. Levings, C. D., Stein, J. E., Stehr, C., & Samis, S. C. (2003). Introduction to the PICES Practical Work- shop: Objectives, Overview of the Study Area, and Projects conducted by the participants. Marine Environmental Research, 57, 3–18. Meng, L., Moyle, P. B., & Herbold, B. (1994). Changes in abundance and distribution of native and introduced fishes of Suisun Marsh. Transactions of the American Fisheries Society, 123, 498–507. Miller, K. A., Addison, R. F., & Bandiera, S. M. (2003). Hepatic CYP1A levels and ERODactivity in English sole: biomonitoring of marine contaminants in Vancouver Harbour. Marine Environmental Research, 57, 37–54. Nelson, T. C., & Levings, C. D. (1995). Executive summary. In Studies of the starry flounder in the Fraser River 1992–1994 (pp. 1–17). Fraser River Estuary Management Program Technical Report WQM 95- 02. Smokorowski, K. E., Stoneman, M. G., Cairns, V. W., Minns, C. K., Randall, R. G., & Valere, B. (1998). Trends in the nearshore fish community of Hamilton Harbour, 1988 to 1997, as measured using an index of biotic integrity. Canadian Technical Report of Fisheries and Aquatic Science, 2230, 97. Stehr, C. M., & Horiguchi, T. (2001). Environmental assessment of Vancouver Harbour: data report for the PICES practical workshop. PICES scientific report No. 16. Sidney BC, Canada: North Pacific Marine Science Organization (PICES). Stehr, C. M., Myers, M. S., Johnson, L. L., Spencer, S., & Stein, J. E. (2003). Toxicopathic liver lesions in English sole and chemical contaminant exposure in Vancouver Harbour, Canada. Marine Environ- mental Research, 57, 55–74. Toole, C. L., Barnhart, R. A., & Onuf, C. P. (1987). Habitat suitability index models: juvenile English sole. US Fish and Wildlife Service Biological Report, 82(10.133), 27. Williams, S. F., & Caldwell, R. S. (1978). Growth, food conversion and survival of O-group English sole (Parophrys vetulus Girard) at five temperatures and five rations. Aquaculture, 15, 129–139. Wong, K. C. (1980). A preliminary survey of the benthic fishes of Tolo Harbour. In B. S. Morton, & C. K. Tseng (Eds.), Proc. First International Marine Biological Workshop: The marine flora and fauna of Hong Kong and Southern China, Hong Kong, 1980 (pp. 921–933). Hong Kong: Hong Kong University Press.