2007 NORTHEASTERN NATURALIST 14(2):251–268 Relationships Between Fish Assemblage Structure and Selected Environmental Factors in Maryland’s Coastal Bays Joseph W. Love1,* and Eric B. May1 Abstract - We surveyed little-known ray-finned fish assemblages from Maryland’s coastal bays in order to establish species-habitat relationships for common species. From 1996–1999, 25 sites were sampled monthly with otter trawls in the coastal bays of Maryland. Anchoa mitchilli (bay anchovy) constituted nearly 50% of the catch for each year, and species composition was largely similar across years, with some differences likely related to variation in recruitment. For example, Clupea harengus (Atlantic herring) was particularly abundant during 1996 and 1999 following their spawning season. We used canonical correspondence analysis to determine how assemblages were related to temperature, dissolved oxygen (DO), salinity, and landuse variables during summer (June–September) and throughout the rest of the year. A gradient correlated with temperature and DO significantly structured assemblages throughout most of the year; during summer, the proportion of wetland habitat was important. We demonstrate that environmental gradients important for structuring fish assemblages differ between summer and non-summer months and there is a general shift in habitat use during summer from the lower estuary to other areas of the coastal bays. Our data also provide support for earlier observations that temperature was the major factor influencing changes in fish assemblage structure in the coastal bays. Our results point to better characterization of fish habitats in order to effectively manage coastal ecosystems of Maryland. Introduction Habitats necessary for spawning, foraging, and growth of federally managed species are defined as essential fish habitat (EFH) by the 1996 reauthorization of the Magnuson-Stevens Fishery Conservation and Management Act. When habitat requirements are met, species abundance will be greatest (Brown et al. 1995, Grinnell 1917, Hutchinson 1957), which potentially reflects EFH. Managing EFH requires an understanding of how fishes are distributed within their habitat, and how they respond to changes in their environment. Such changes may include dredging (Aldridge 2000, Koenig et al. 2000, Slacum et al. 2000), urbanization (Tong 2001), and deteriorating water quality (Kramer 1987, Tong 2001), which are all factors that currently affect fish habitat within the coastal bays of Maryland. Coastal embayments of Maryland are estuarine, with measurable input from freshwater sources such as the St. Martin’s River, which is the largest river system draining the coastal bay watershed. The coastal habitats of 1NOAA Living Marine Resources Cooperative Science Center, University of Maryland Eastern Shore, Princess Anne, MD 21853. *Corresponding author - jlove@umes.edu. 252 Northeastern Naturalist Vol. 14, No. 2 Maryland are productive nursery habitats for juvenile fishes, and provide a prey base for adult fishes that seasonally enter the watershed. Seasonal changes in salinity, dissolved oxygen (DO), and temperature are common and species able to persist year-round in estuaries tend to be euryhaline and tolerate a wide range of temperature and DO. Land-use patterns also affect aquatic environments (Goudswaard et al. 2002, Tong 2001). During summer, habitats near urban and agricultural settings become hypoxic (MDDNR 2004). Wetland habitats may buffer against eutrophication (Chescheir et al. 1991), thereby buffering against hypoxic zones that adversely affect aquatic life. Over 100 species of juvenile fishes and many adult fishes are seasonally common throughout the coastal bay watershed (Casey et al. 1995, 2001; Schwartz 1961, 1964). Nearly 100 species have been collected from Chincoteague Bay, VA (Richards and Castagna 1970), which is the largest of the coastal bays. Pomatomus saltatrix (bluefish) and Paralichthys dentatus (summer flounder) are among many federally or state-managed fin-fish species, emphasizing the importance of documenting patterns of fish distribution and habitat use in the coastal bays. The forage-fish index and many coastal bay fish populations have declined notably over the past 20 years in Maryland (Casey et al. 2001), despite a sizeable increase in submerged aquatic vegetation (SAV) (MDDNR 2004) that may serve as complex underwater refugia (Heck and Orth 1980, Hovel and Lipcius 2001). Species that occupy higher trophic levels, such as sandbar sharks, may rely on forage fish and crabs in the lower portions of Chincoteague Bay (Medved et al. 1985), possibly explaining declines in the abundance of top predators. For this project, we explored distributions of ray-finned fishes across important water-quality and land-use gradients from the coastal bays of Maryland. We predicted that water quality, the amount of SAV, and neighboring land-use patterns affect the distribution of fish species. In particular, we focused on the distribution of abundant species, flounders, and species that are indicators of long-term trends of coastal bay fish community structure (Brevoortia tyrannus [Atlantic menhaden], Leiostomus xanthurus [spot], Menidia menidia [Atlantic silverside], and Anchoa mitchilli [bay anchovy]). Methods Study sites Maryland’s coastal bay watershed includes five large, partially isolated bays that can be accessed through either a northern or southern inlet (Fig. 1). The northern inlet (Ocean City Inlet [OCI]) was created fewer than 80 years ago and likely facilitates dispersal for some species into the northern coastal bays (Schwartz 1961, 1964). The southern inlet is also an important dispersal corridor for species inhabiting the largest of the bays, Chincoteague Bay (Schwartz 1961). 2007 J.W. Love and E.B. May 253 We analyzed Slacum et al.’s (2000) data for 604 different collections from sites sampled monthly for water quality and finfish distribution from April 1996–December 1999 (Fig. 1; Table 1). To randomly choose sites for sampling, Slacum et al. (2000) divided the coastal bays into a series of 30- x 30-m2 cells using a geographic information system (GIS) and randomly selected cells each year to sample. In 1996, sites characterized by 50% cover of submerged aquatic vegetation (SAV) were purposely avoided because of their high value as recruiting areas. Table 1. The number of sampled sites for each year of this study for Chincoteague Bay (CB), Sinepuxent Bay (SB), Newport Bay (NB), Isle of Wight (IW), and Assawoman Bay (AB). CB SB NB IW AB 1996 2 4 1 3 6 1997 5 2 1 5 2 1998 7 4 2 6 2 1999 10 4 1 10 0 Figure 1. Map of surveyed sites (1996–1999) from the coastal bays of Maryland ( see boxed inset). Each site was sampled each month, with the exception of the first year when sampling began in April. The Ocean City and Chincoteague inlets are labeled with black arrows. 254 Northeastern Naturalist Vol. 14, No. 2 Habitat measurements Beginning in April 1996, sites were measured monthly for water quality. Dissolved oxygen (DO), temperature, salinity, pH, turbidity, and depth were measured using a Hydrolab H20 and Surveyor 3 (Hydrolab Corporation, Loveland, CO). DO and pH measurements were taken near the substrate. The coefficient of variation (CV) was calculated for each variable (Sokal and Rohlf 1994) and converted to a percentage. We transformed all environmental measurements by log10 prior to analyses, except pH which is measured on a logarithmic scale. We measured land use as the proportion of land-use type within a 3-km radius of each site. A variety of radii were initially tried, but 3 km was chosen based on the minimum area necessary to capture some land type near the majority of sites. We used the buffer tool of ArcView (Version 3.3, Environmental Systems Research Institute, Inc., Redlands, CA) to determine the proportional area of four land types (wetland, forest, agriculture, urban) within the 3-km radius of each site. These characterizations were generalized from spatial data provided by the Maryland-Delaware-New Jersey GAP analysis program (US Geological Survey 1999). The proportion of each land type (Xi) was calculated as: n Xi = xi /􀂙xi, i = 1 where xi is the area of land type i is divided by the sum of all xi’s. The resulting value is the proportion of land type i for all land. Because sites differed with respect to the proportion of water surrounding them and distance from shore, we also included distance of each site to the closest shoreline for analyses. Distances were transformed by log10 before analyses. The amount of SAV within three kilometers of each site was determined similarly to that above. We obtained spatial data for SAV coverage from the Virginia Institute of Marine Science (2001). Spatial data provided an accurate measure of SAV distribution in the coastal bays. Aerial photography (1:24,000 scale) was used to determine the perimeter of SAV beds, and area was calculated from perimeter. The area of SAV was transformed by log10 before analyses. Fish sampling Fishes were sampled at each site using a 4.9-m semi-balloon otter trawl (19-mm body mesh and 4.8-mm inner-liner, cod end mesh) that was towed for six minutes at depths usually greater than one meter. Two 6-minute trawls were conducted at each site, and data from both trawls were combined for each sampling location. All specimens were identified and counted in the field, but voucher specimens were preserved in 10% Formalin for later verification. The first 25 individuals of each species were measured for total length. All data were entered into spreadsheets twice by different people and cross-checked for errors. We analyzed data for ray-finned fishes (Class: Actinopterygii) and excluded elasmobranchs in the analysis, even though 2007 J.W. Love and E.B. May 255 they were sampled by Slacum et al. (2000) and are ecologically important (Medved et al. 1985). Our goal was to report habitat relationships for rayfinned fishes that are both important to recreational fisheries of Maryland and are long-term indicators of ecosystem change. We also excluded rare species, which were defined as those occupying 􀂔 1% of the sample at a site and those occurring across 􀂔 1% of the sites (i.e., low abundance and restricted distribution). These species were: Elops saurus Linnaeus (ladyfish), Gobiosoma ginsburgi Hildebrand and Schroeder (seaboard goby), Hyporhamphus meeki Banford and Collette (halfbeak), Mugil cephalus Linnaeus (striped mullet), M. curema Valenciennes (white mullet), Rachycentron canadum (Linnaeus) (cobia), Scomberomorus cavalla (Cuvier) (king mackerel), Sphyraena guachancho Cuvier (Guaguanche), Trachinotus falcatus (Linnaeus) (Permit), and Trichiurus lepturus Linnaeus (Atlantic cutlassfish). Abundances were square-root transformed prior to analyses (McCune and Grace 2002). Data analysis To relate species distributions to measured environmental and land-use variables, we used canonical correspondence analysis (CCA) (ter Braak 1986). We used the resulting ordination plot to explore how species composition and number differed along gradients of water quality and land use. The analysis related a site x species matrix to a site x environmental variable matrix. The CCA ordinates species and site scores along canonical axes that are constrained by the environmental variables. Axes were scaled to optimize the representation of species scores in ordination space. Site scores were re-scaled with mean = 0 and variance = 1. Thus, the distances among species scores best approximates the relationships among each other and to the environmental variables plotted in the ordination diagram (McCune and Grace 2002). Scores were calculated using weighted averaging of scores derived from community data. We analyzed community and environmental data collected during summer separately from those collected during non-summer. The summer months (June–September) were chosen based on seasonally high temperatures and low DO, which we observed from 5–10 years of data collected by the Maryland Department of Natural Resources across sites in the coastal bays of Maryland (http://mddnr.chesapeakebay.net/eyesonthebay/ index.cfm). For each CCA analysis, multiple regression methods were used to determine the relationship between environmental gradients and species distributions. Environmental variables strongly correlated with each gradient (or orthogonal axis) of the CCA were plotted as vectors in a joint-plot with species scores. The correlation of a species score that represents its optimal abundance to the environmental gradient was determined using Pearson Correlation coefficients (r). We used a Monte Carlo randomization routine to determine the significance of the eigenvalues and species-environment correlations. Eigenvalues 256 Northeastern Naturalist Vol. 14, No. 2 indicate the proportion of total variance explained by each CCA axis. Species- environment correlations are the correlation between canonical scores derived from species data and those derived from linear combinations of the environmental variables. The Monte Carlo procedure used 1000 randomizations of the data to test the null hypothesis of no relationship between the community and environmental matrices. The number of randomizations that yielded a higher or equal eigenvalue for each CCA axis than the observed eigenvalue was calculated to yield the probability of committing a Type 1 error. Ordination analyses and Monte Carlo randomizations were performed using PC-ORD (Version 4; McCune and Mefford 1999; MjM Software Design, Gleneden Beach, OR). Results Fishes were collected from shallow (0.6 m) to moderately deep water sites (9.3 m), but at an average of 1.90–1.92 m (Table 1). Sites varied by distance from land (21–4233 m), and on average, agriculture and forest were the dominate land types near sampled sites (both 26%). Most sites were also near SAV (> 93%), including from 1149 m2 to 78,816 m2 of SAV. All year long, depth and turbidity varied more than other environmental variables (Table 2). The least variable environmental measurement was pH (CV = 2–3%). Seventy-two ray-finned fish species were collected from 604 trawls over the study period. Only 11 species had an average abundance across years that was greater than 1 (Appendix 1). Of these 11 species, bay anchovy was Table 2. Water quality and landscape data recorded from the coastal bays of Maryland. Landscape data include distance of the site to land, type of land as a proportion, and the area of submerged aquatic vegetation (SAV). The means, coefficients of variation (CV), and range of values were calculated for all sampled sites. Water-quality data were collected monthly and averages, Cvs and ranges are provided for June–September and October–May. October–May June–September a. Water quality Mean CV Range Mean CV Range pH 7.89 2% 7.32– 8.73 7.80 3% 6.99– 8.48 Depth (m) 1.90 62% 0.60– 9.30 1.92 66% 0.60– 7.00 Temperature (ºC) 11.97 44% 0.10–25.04 24.68 12% 17.29–30.38 Salinity (ppt) 25.19 13% 9.10–30.70 26.91 9% 12.70–30.50 DO (mg/L) 9.05 19% 2.30–15.85 6.71 17% 0.53–10.34 Turbidity (NTU) 10.64 81% 0.90–67.90 17.90 58% 1.00–61.60 b. Landscape Mean CV Range Distance (m) 1006 95% 21–4233 Agriculture 0.26 100% 0–0.84 Urban 0.10 140% 0–0.55 Forest 0.26 81% 0–0.71 Wetland 0.23 91% 0–1.00 SAV (m) 23,630 81% 0–78,816 2007 J.W. Love and E.B. May 257 the most abundant for all 4 years of the study. Atlantic menhaden, spot, and Atlantic silverside were also among the most abundant. We rejected the null hypothesis of no relationship between the community and environmental data sets for summer and the rest of the year (Table 3). Species-environment correlations and the amount of total variance in the community data explained by CCA axes were low. Three environmental gradients explained from 5–8% of the variance in species data for both time periods (Table 3). We represented the distribution of the 11 most abundant species with respect to environmental gradients using ordination plots (Figs. 2, 3). Environmental variables important for summer fish assemblages differed from those for the rest of the year. For summer, the first environmental gradient (CCA axis one) was correlated with wetland habitat (r = -0.45) and salinity (r = 0.57) and it represented a spatial gradient from OCI (higher salinity, low proportion of wetland) to the lower estuary (lower salinity, high proportion of wetland). Wetland areas were correlated with the co-occurrence of spot (r = -0.34) and bay anchovy (r = -0.43; Fig. 2). Weakfish and Atlantic menhaden were also positively associated with wetland habitat (Fig. 2). Distributions of deepwater species such as Etropus microstomus (smallmouth flounder) and Scophthalmus aquosus (windowpane) were correlated with deeper, saltier water (r = 0.42 and 0.43, respectively). The second axis was correlated with pH (r = -0.56), but no species showed a strong correlation with this gradient (r < 0.25, for all). Alosa aestivalis (blueback herring) showed a positive association with SAV (Fig. 2B), which was correlated to the third axis (r = 0.77). Species distributions were structured by water quality variables during the rest of the year (Fig. 3), indicating a general shift away from lower parts of the estuary (near wetlands) or SAV habitat in summer. A gradient of temperature (r = -0.86) and DO (r = 0.58) was related to the distributions of most species (Fig. 3), including Atlantic silverside that were abundant when DO was high throughout the year (Figs. 2, 3). Spot, Atlantic menhaden, and bay anchovy occurred in different habitats than Atlantic silverside and tended to be abundant where and when water was warmer. Smallmouth Table 3. Summary statistics for 3 canonical axes derived from a canonical correspondence analysis (CCA) of fish assemblage data collected monthly from June–September and October– May from sites within the coastal bays of Maryland (1996–1999). The eigenvalue is the proportion of total variance in the community data set explained by the axis (i.e., % explained). The Pearson correlation (Spp-Env) is the correlation between the species matrix and a species matrix constrained by environmental variables. We tested the null hypothesis of no linear relationship between matrices (Spp-Env) and whether the observed eigenvalue was greater than that expected from chance using Monte Carlo methods (** indicates P < 0.01). June–September October–May CCA Axis Axis 1 Axis 2 Axis 3 Axis 1 Axis 2 Axis 3 Eigenvalue 0.18** 0.12** 0.10** 0.29** 0.14** 0.12** % explained 3.70 2.40 2.10 3.10 1.40 1.20 Spp-Env 0.69** 0.56** 0.54** 0.77** 0.60** 0.60** 258 Northeastern Naturalist Vol. 14, No. 2 Figure 2. Canonical correspondence analysis of fish assemblage and environmental data collected from sites surveyed June–September in the coastal bays of Maryland (1996–1999). Scores for the 11 most abundant species are plotted with labeled habitat vectors. Species abbreviations are: weakfish (We), Atlantic croaker (AC), bay anchovy (BA), spot (Sp), Atlantic menhaden (AM), silver perch (SP), blueback herring (BH), summer flounder (SF), striped anchovy (SA), Atlantic herring (AH) and Atlantic silverside (AS). The position of a species’ score in relation to habitat vectors indicates whether it has a higher- or lower-than-average optimal abundance on that environmental variable. Ordination scores are plotted for axes 1 and 2 (A) and axes 1 and 3 (B). 2007 J.W. Love and E.B. May 259 Figure 3. Canonical correspondence analysis of fish assemblage and environmental data collected from sites surveyed October–May in the coastal bays of Maryland (1996–1999). Scores for the 11 most abundant species are plotted with labeled habitat vectors. Species abbreviations are: weakfish (We), Atlantic croaker (AC), bay anchovy (BA), spot (Sp), Atlantic menhaden (AM), silver perch (SP), blueback herring (BH), summer flounder (SF), striped anchovy (SA), Atlantic herring (AH) and Atlantic silverside (AS). The position of a species’ score in relation to habitat vectors indicates whether it has a higher- or lower-than-average optimal abundance on that environmental variable. Ordination scores are plotted for axes 1 and 2 (A) and axes 1 and 3 (B). 260 Northeastern Naturalist Vol. 14, No. 2 and summer flounder were negatively correlated with the temperature/DO gradient (r = -0.31 and r = -0.36, respectively), indicating that they resided in warmer habitats with lower DO. The second and third axes were correlated with salinity (r = 0.64) and depth (r = -0.58) or turbidity (r = -0.67), respectively. Atlantic croaker was optimally abundant at high salinities. Discussion Coastal estuaries are influenced by seasonal environmental changes that are predictable across years, leading to repeatable and seasonal patterns of immigration and fish-assemblage structure (Witting et al. 1999) that are largely explained by temperature variation throughout the year (Desmond et al. 2002). In addition, our research indicates that 1) environmental gradients structuring fish assemblages differ between non-summer and summer months, and 2) habitat use by many abundant species changes from the lower estuary during summer to other areas of coastal bays for the rest of the year. While the only measured variables that structured distributions of species during the non-summer months were related to water quality, SAV and wetland areas played more important roles in explaining distributions of fishes during summer. The level of variability in water quality was much higher during the non-summer months than during summer. Water temperature in the coastal bays of Maryland ranged slightly over 10 °C during the non-summer months versus 2 °C in summer. The influence of water temperature on fish assemblages supports earlier observations that many species moved into and out of the coastal bays depending on water temperature (Schwartz 1961). In addition to dispersal, abundance may differ along temperature gradients because of lower survivorship at low temperatures (Kraus and Musick 2001, Malloy and Targett 1991). Dissolved oxygen (DO) was inversely related to temperature. Low DO levels can cause physiological stress that results in aerial or aquatic surface respiration (Kramer 1987, Love and Rees 2002) and, in severe cases, death (Love and Rees 2002, Shepard 1955). Some schooling fishes, such as Atlantic menhaden, can substantially reduce DO availability in enclosed waters (Oviatt et al. 1972), resulting in their own death (Smith 1999) Dissolved oxygen remained low throughout summer, but was rarely severely hypoxic (< 2.0 mg/L). Salinity gradients also structured fish distributions during the non-summer months. Monaco et al. (1998) showed that many estuarine fishes from the mid-Atlantic region responded strongly to salinity differences. Salinity fluctuations can greatly influence development of larvae (Holliday and Blaxter 1961), osmoregulation (Haney 1999), and age-specific movement (Able et al. 2001). Wetlands and SAV habitat played a larger role in explaining distributions of species during summer than the rest of the year. Atlantic menhaden, spot, and bay anchovy were abundant near wetlands during summer. Wetlands may act as biological filters for agricultural nutrients (Chescheir et al. 1991), providing a “processing center” for an otherwise agriculturally dominated 2007 J.W. Love and E.B. May 261 landscape. Wetlands can also be associated greater prey availability for juvenile fishes (Fisher and Willis 2000) and refugia from predators, leading to their distinct fish assemblages (Jude and Pappas 1992). Spot and Atlantic menhaden as well as many other species listed here may utilize wetland habitats as nursery areas during summer. The size range of specimens for each of these species included that expected for juveniles. Unfortunately, we do not have enough data to generate reliable length-frequencies for summer or to explore the size classes found near wetlands. Little total variance in species distributions was extracted by the canonical axes, and species-environment correlations were lower than those found for other studies (Bhat 2004, O’Connell et al. 2004). These two results have interesting implications for management of fishery habitat in the coastal bays of Maryland. Traditionally measured environmental variables may not adequately characterize habitats for fishes. A likely problem affecting many exploratory studies is the choice of environmental predictors. While we considered traditional water-quality variables and included estimates of land use, we did not consider human disturbance, substrate diversity and fine-scale densities of SAV, or prey abundance. Tidal changes may also influence fish distributions (Jaafar et al. 2004), but we included water depth to help account for environmental variation related to tidal height. Human modification of demersal habitats strongly affected changes in fish community structure in the Lake Pontchartrain estuary (O’Connell et al. 2004). While a spatial gradient of human disturbance exists in the coastal bays (Slacum et al. 2000), this element was not rigorously quantified as part of Slacum et al.’s study. Substrate diversity and the local density of SAV may strongly affect fish distribution (Casey et al. 2001, Orth and Moore 1983, Raposa and Oviatt 2000). Using a coarse estimate of SAV density, we found that some species were associated with SAV beds during summer (e.g., blueback herring), but this was not generally the case for our study. Others have noted that greater fish diversity and biomass are found in SAV beds than in areas dominated by macroalgae (Snodgrass 1992), and species collected in this study, such as Aluterus schoepfi (orange filefish) may frequent SAV (Murdy et al. 1997). We suggest that more detailed resolution of local substrate composition, including density of SAV, may be more useful for better defining specieshabitat relationships in the coastal bays. Substrates in the coastal bays are mostly sandy, with the percentages of silt and clay representing less than 40% in Chincoteague, Assawoman, and Isle of Wight Bays (Wells et al. 1994). In Chincoteague Bay, silt is most common in the channels. Silt is also abundant in Newport Bay, while sand dominates Sinepuxent Bay. These strong differences in sediment type throughout the bays likely affected fish distributions, especially for benthic or demersal species (Li and Gelwick 2005). In addition to substrate characterization, prey distribution should also be considered because growth and survivorship of larvae or juveniles is enhanced when their needs are met by the environment (e.g., prey base; Cushing and Horwood 1994). 262 Northeastern Naturalist Vol. 14, No. 2 Species-environment correlations may also be low if species are abundant in poor quality habitat when persisting from a source of immigrants (i.e., source-sink dynamics; Magoulick and Kobza 2003, Pulliam 1988) or if poor-quality habitats are purposely selected by species (Donovan and Thompson 2001). Species may also be pushed into poor-quality habitat because high-quality habitat has been destroyed. As a result, these processes may confound investigations into EFH. To date, the aforementioned processes have not been explored as mechanisms to explain patterns of species abundance in the coastal bays. Our results elucidate some very complex and not well-understood relationships between species and their coastal bay environment. We emphasize the importance of wetlands for protection of EFH, but also show that much remains to be learned. Developing this understanding contributes to a greater knowledge of resource needs and trophic interactions of coastal fishes. Such knowledge is essential for multispecies approaches of ecosystem management (Christensen and Pauly 2004), which are rapidly gaining popularity in federal agencies (Zabel et al. 2003). No studies have established the necessary foundation for such management in the coastal bays of Maryland. In the years ahead, this approach may be the best hope of maintaining healthy, dynamic, and diverse coastal fish communities. Acknowledgments We gratefully acknowledge the efforts of Mr. Ward Slacum in collecting and identifying most of the species collected during this project. We also thank reviewers of this manuscript for their helpful comments. This research was funded by the Coastal Zone Management Program of the Maryland Department of Natural Resources. Literature Cited Able, K.W., D.M. Nemerson, R. Bush, and P. Light. 2001. Spatial variation in Delaware Bay (USA) marsh creek fish assemblages. Estuaries 24:441–452. Aldridge, D.C. 2000. The impacts of dredging and weed cutting on a population of freshwater mussels (Bivalvia: Unionidae). Biological Conservation 95:247–257. Baht, A. 2004. Patterns in the distribution of freshwater fishes in rivers of Central Western Ghats, India, and their associations with environmental gradients. Hydrobiologia 529:83–97. Brown, J.H., D.W. Mehlman, and G.C. Stevens. 1995. Spatial variation in abundance. Ecology 76:2028–2043. Casey, J.F., S.B. Doctor, and A.E. Wesche. 1995. Investigation of Maryland’s Atlantic Ocean and coastal bay finfish stocks. 1995. Maryland Department of Natural Resources, Fisheries Service Biological Monitoring and Analysis, Annapolis, MD. Casey, J.F., S.B. Doctor, and A.E. Wesche. 2001. Investigation of Maryland’s Atlantic Ocean and Coastal Bay Finfish Stocks. Federal Aid Project No. F-50-R. Maryland Department of Natural Resources Fisheries Service, Annapolis, MD. 28 pp. Chescheir, G.M., J.W. Gilliam, R.W. Skaggs, and R.G. Broadhead. 1991. Nutrient and sediment removal in forested wetlands receiving pumped agricultural drainage water. Wetlands 11:87–103. 2007 J.W. Love and E.B. May 263 Christensen, V., and D. Pauly. 2004. Placing fisheries in their ecosystem context: An introduction. Ecological Modeling 172:103–107. Cushing, D.H., and J.W. Horwood. 1994. The growth and death of fish larvae. Journal of Plankton Research 16:291–300. Desmond, J.S., D.H. Deutschman, and J.B. Zedler. 2002. Spatial and temporal variation in estuarine fish and invertebrate assemblages: Analysis of an 11-year data set. Estuaries 25:552–569. Donovan, T.M., and F.R. Thompson III. 2001. Modeling the ecological trap hypothesis: A habitat and demographic analysis for migrant songbirds. Ecological Applications 11:871–882. Fisher, S.J., and D.W. Willis. 2000. Seasonal dynamics of aquatic fauna and habitat parameters in a perched upper Missouri River wetland. Wetlands 20:470–478. Goudswaard, K., F. Witte, and L.J. Chapman. 2002. Decline of the African lungfish (Protopterus aethiopicus) in Lake Victoria (East Africa). African Journal of Ecology 40:42–52. Grinnell, J. 1917. The niche-relationships of the California Thrasher. Auk 34:427–433. Haney, D.C. 1999. Osmoregulation in the sheepshead minnow, Cyprinodon variegatus: Influence of a fluctuating salinity regime. Estuaries 22:1071–1077. Heck, K.L., Jr., and R.J. Orth. 1980. Seagrass habitats: The roles of habitat complexity, competition, and predation in structuring associated fish and motile macroinvertebrate assemblages. Pp. 449–464. In V.S. Kennedy (Ed.). Estuarine Perspectives. Academic Press, New York, NY. Holliday, F.G.T., and J.H.S. Blaxter. 1961. The effects of salinity on herring after metamorphosis. Journal of Marine Biology Association, UK 41:37–48. Hovel, K.A., and R.N. Lipcius. 2001. Habitat fragmentation in a seagrass landscape: Patch size and complexity control blue crab survival. Ecology 82:1814–1829. Hutchinson, R.E. 1957. Concluding remarks. Cold Spring Harbor. Symposium on Quantitative Biology 22:415–427. Jaafar, Z., S. Hajisamae, L. Chou, and Y. Yatiman. 2004. Community structure of coastal fishes in relation to heavily impacted human-modified habitats. Hydrobiologia 511:113–123. Jude, D.J., and J. Pappas. 1992. Fish utilization of Great Lakes coastal wetlands. Journal of Great Lakes Research 18:651–672. Koenig, C.C., F.C. Coleman, C.B. Grimes, G.R. Fitzhugh, K.M. Scanlon, C.T. Gledhill, and M. Grace. 2000. Protection of fish spawning habitat for the conservation of warm-temperate reef-fish fisheries of shelf-edge reefs of Florida. Bulletin of Marine Science 66:592–616. Kramer, D.L. 1987. Dissolved oxygen and fish behavior. Environmental Biology of Fishes 18:81–92. Kraus, R.T., and J.A. Musick. 2001. A brief interpretation of summer flounder, Paralichthys dentatus, movements and stock structure with new tagging data on juveniles. Marine Fisheries Review 63:1–6. Li, R.Y., and F.P. Gelwick. 2005. The relationship of environmental factors to spatial and temporal variation of fish assemblages in a floodplain river in Texas, USA. Ecology of Freshwater Fish 14:319–330. Love, J.W., and B.B. Rees. 2002. Seasonal differences in hypoxia tolerance in gulf killifish, Fundulus grandis (Fundulidae). Environmental Biology of Fishes 63:103–115. 264 Northeastern Naturalist Vol. 14, No. 2 Magoulick, D.D., and R.M. Kobza. 2003. The role of refugia for fishes during drought: A review and synthesis. Freshwater Biology 48:1186–1198. Malloy, K.D., and T.E. Targett. 1991. Feeding, growth, and survival of juvenile summer flounder Paralicthys dentatus: Experimental analysis of the effects of temperature and salinity. Marine Ecology Progress Series 72:213–223. McCune, B., and J.B. Grace. 2002. Analysis of Ecological Communities. MjM Software Design, Gleneden Beach, OR. 300 pp. McCune, B., and M.J. Mefford. 1999. PC-ORD. Multivariate Analysis of Ecological Data, Version 4. MjM Software Design, Gleneden Beach, OR. Maryland Coastal Bays Program (MCBP). 1999. Maryland Coastal Bays Ecosystem Health Assessment 2004. Berlin, MD. Maryland Department of Natural Resources (MDDNR). 2004. State of the Maryland Coastal Bays. Annapolis, MD. 44 pp. Medved, R.J., C.E. Stillwell, and J.J. Casey. 1985. Stomach contents of young sandbar sharks, Carcharhinus plumbeus, in Chincoteague Bay, Virginia. Fishery Bulletin 83:395–402. Mnaya, B., and E. Wolanski. 2002. Water circulation and fish larvae recruitment in papyrus wetlands, Rubondo Island, Lake Victoria. Wetlands Ecology and Management 10:133–143. Monaco, M.E., S.B. Weisberg, and T.A. Lowery. 1998. Summer habitat affinities of estuarine fish in US mid-Atlantic coastal systems. Fisheries Management and Ecology 5:161–171. Murdy, E.O., R.S. Birdsong, and J.A. Musick. 1997. Fishes of Chesapeake Bay. Smithsonian Institution Press, Washington, DC. 324 pp. O’Connell, M.T., R.C. Cashner, and C.S. Schieble. 2004. Fish assemblage stability over fifty years in the Lake Pontchartrain estuary: Comparisons among habitats using canonical correspondence analysis. Estuaries 27:807–817. Orth, R.J., and K.A. Moore. 1983. Chesapeake Bay: An unprecedented decline in submerged aquatic vegetation. Science 222:51–53. Oviatt, C.A., A.L. Gall, and S.W. Nixon. 1972. Environmental effects of Atlantic menhaden on surrounding waters. Chesapeake Science 13:321–323. Pulliam, H.R. 1988. Sources, sinks, and population regulation. American Naturalist 132:652–661. Raposa, K.B., and C.A. Oviatt. 2000. The influence of contiguous shoreline type, distance from shore, and vegetation biomass on nekton community structure in eelgrass beds. Estuaries 23:46–55. Richards, C.E., and M. Castagna. 1970. Marine fishes of Virgina’s eastern shore (inlet and marsh, seaside waters). Chesapeake Science 11:235–248. Schwartz, F.J. 1961. Fishes of Chincoteague and Sinepuxent Bays. American Midland Naturalist 65:384–408. Schwartz, F.J. 1964. Fishes of Isle of Wight and Assawoman Bays near Ocean City, Maryland. Chesapeake Science 5:172–193. Shepard, M.P. 1955. Resistance and tolerance of young speckled trout (Salvelinus fontinalis) to oxygen lack, with special reference to low oxygen acclimation. Journal of the Fisheries Research Board of Canada 12:387–446. Slacum, H.W., W.J. Jones, S. Ruhl, and E.B. May. 2000. Ecological evaluation of proposed coastal bay spoil-deposit habitat restoration sites. Maryland Cooperative Fish and Wildlife Research Unit, University of Maryland Eastern Shore, Princess Anne, MD. 281 pp. 2007 J.W. Love and E.B. May 265 Smith, J.W. 1999. A large fish kill of Atlantic menhaden, Brevoortia tyrannus, on the North Carolina coast. Journal of the Elisha Mitchell Scientific Society 115:157–163. Snodgrass, J.W. 1992. Comparison of fishes occurring in alga and seagrass habitats on the east coast of Florida. Northeastern Gulf Science 12:119–128. Sokal, R.R., and F.J. Rohlf. 1994. Biometry. W.H. Freeman, 3rd Edition. New York, NY. 880 pp. ter Braak, C.J.F. 1986. Canonical correspondence analysis: A new eigenvector technique for multivariate direct gradient analysis. Ecology 67:1167–1179. Tong, S.T.Y. 2001. An integrated exploratory approach to examining the relationships of environmental stressors and fish responses. Journal of Aquatic Ecosystem Stress and Recovery 9:1–19. US Geological Survey (USGS). 1999. Maryland-Delaware-New Jersey GAP analysis program. Sioux Falls, SD. Available online at http://gapanalysis.nbii.gov/ portal/server.pt. Virginia Institute of Marine Science (VIMS). 2001. Chesapeake Bay submerged aquatic vegetation. Remote-sensing image. Available online at http:www.msgic.state.md.us/techtool. Data accessed in fall 2004. Wells, D.V., R.D. Conkwright, and J. Park. 1997. Geochemistry and geophysical framework of the shallow sediments of Assawoman Bay and Isle of Wight Bay in Maryland. Coastal and Estuarine Geology Open File Report No. 15. Maryland Geological Survey, Baltimore, MD. Witting, D.A., K.W. Able, and M.P. Fahay. 1999. Larval fishes of a middle Atlantic bight estuary: Assemblage structure and temporal stability. Canadian Journal of Fisheries and Aquatic Sciences 56:222–230. Zabel, R.W., C.J. Harvey, S.L. Katz, T.P. Good, and P.S. Levin. 2003. Ecologically sustainable yield. American Scientist 91:150–157. 266 Northeastern Naturalist Vol. 14, No. 2 Appendix 1. Summary data for species collected monthly in the coastal bays of Maryland (1996–1999). Each species’ abundance is represented by the percent of the total catch for that year. The species’ range of total length (mm), average total length (TL), and percent variation in length (%CV) across years are also given. Abundance Size Species (common name) 1996 1997 1998 1999 Range TL (mm) %CV Anchoa mitchilli Valenciennes (bay anchovy) 47.4926 48.9730 76.1857 52.1660 8–184 56.23 33.20 Clupea harengus Linnaeus (Atlantic herring) 27.6880 10.7271 9.1307 14.7069 22–333 59.98 82.74 Menidia menidia (Linnaeus) (Atlantic silverside) 8.0861 0.8863 0.8331 3.5906 21–138 81.96 24.62 Bairdiella chrysoura (Lacepède) (silver perch) 4.7498 3.8154 2.9639 1.3749 11–226 89.60 52.84 Alosa aestivalis (Mitchill) (blueback herring) 2.4424 0.3038 0.0398 0.2598 37–258 124.11 38.42 Cynoscion regalis (Bloch and Schneider) (weakfish) 1.6993 3.4085 0.5400 0.5471 17–497 101.13 57.34 Anchoa hepsetus (Linnaeus) (striped anchovy) 1.6785 0.4097 0.1394 0.0584 35–132 73.87 24.72 Micropogonias undulatus (Linnaeus) (Atlantic croaker) 1.3096 1.6011 3.5680 2.3645 8–331 82.03 87.03 Brevoortia tyrannus Latrobe (Atlantic menhaden) 0.9354 0.1003 0.2954 1.4187 28–280 66.79 52.41 Leiostomus xanthurus Lacepède (spot) 0.8211 25.7295 0.8696 2.4434 13–262 123.91 35.04 Paralichthys dentatus (Linnaeus) (summer flounder) 0.4573 0.6995 0.7800 1.7982 14–572 171.89 56.38 Gobiosoma bosc (Lacepède) (naked goby) 0.4417 0.1059 0.1029 0.3620 17–67 40.69 4.08 Syngnathus fuscus Storer (northern pipefish) 0.4261 0.2341 0.2522 0.4350 60–258 155.62 24.00 Sphoeroides maculatus (Bloch and Schneider) (northern puffer) 0.3430 0.1324 0.0797 1.0889 20–271 125.45 54.57 Pseudopleuronectes americanus Walbaum (winter flounder) 0.1559 0.4013 0.3219 0.7210 9–386 80.42 68.79 Pomatomus saltatrix (Linnaeus) (bluefish) 0.1247 0.0111 0.0166 0.0175 42–570 138.58 61.38 Etropus microstomus (Gill) (smallmouth flounder) 0.1143 0.1380 0.1593 1.2699 23–525 78.68 31.25 Lucania parva (Baird and Girard) (rainwater killifish) 0.0987 0.1366 0.0697 0.1606 17–67 35.99 0.21 Hippocampus erectus Perry (lined seahorse) 0.0935 0.0446 0.0797 0.1547 62–184 106.07 0.25 Apeltes quadracus (Mitchill) (fourspine stickleback) 0.0883 0.3512 0.3419 0.6393 10–60 43.57 0.19 Prionotus carolinus (Linnaeus) (northern searobin) 0.0831 0.1449 0.0896 0.6510 15–224 111.26 0.35 Scophthalmus aquosus (Mitchill) (windowpane) 0.0831 0.1324 0.0398 0.1752 36–314 157.59 0.48 2007 J.W. Love and E.B. May 267 Abundance Size Species (common name) 1996 1997 1998 1999 Range TL (mm) %CV Opsanus tau (Linnaeus) (oyster toadfish) 0.0572 0.0530 0.0365 0.2131 21–239 110.82 0.41 Peprilus triacanthus (Peck) (butterfish) 0.0572 0.0223 0.1029 0.0175 13–162 60.73 0.63 Menticirrhus saxatilis (Bloch and Schneider) (northern kingfish) 0.0520 0.0139 0.0100 0.0788 22–246 117.93 0.59 Trinectes maculatus (Bloch and Schneider) (hogchoker) 0.0468 0.0766 0.0929 0.2832 27–202 109.92 0.32 Syngnathus floridae (Jordan and Gilbert) (dusky pipefish) 0.0364 0.1533 0.4248 0.1372 33–185 123.15 0.22 Centropristis striata (Linnaeus) (black seabass) 0.0312 0.2174 0.0631 0.6977 34–265 141.22 0.30 Alosa pseudoharengus (Wilson) (alewife) 0.312 0.1226 0.0166 0.3386 98–159 125.31 0.12 Fundulus heteroclitus (Linnaeus) (mummichog) 0.0312 0.0334 0.0133 0.0117 24–100 59.99 0.25 Microgobius thalassinus (Jordan and Gilbert) (green goby) 0.0260 0.1449 0.0531 0.1022 21–57 39.75 0.24 Chasmodes bosquianus (Lacepède) (striped blenny) 0.0260 0.0362 0.0266 0.0234 23–105 66.00 0.30 Urophycis regia (Walbaum) (spotted hake) 0.0208 0.1031 0.5012 0.2948 34–284 115.65 0.44 Cyprinodon variegatus Lacepède (sheepshead minnow) 0.0208 0.0920 0.0100 0.0117 20–55 38.11 0.18 Pogonias chromis (Linnaeus) (black drum) 0.0208 0.0084 0.0498 0.0117 57–253 174.88 0.29 Selene vomer (Linnaeus) (lookdown) 0.0208 0.0000 0.0000 0.0117 48–135 80.19 0.30 Stenotomus chrysops (Linnaeus) (scup) 0.0156 0.0056 0.0199 0.0234 89–192 137.29 0.20 Astroscopus guttatus Abbott (northern stargazer) 0.0156 0.0028 0.0000 0.0117 25–334 109.50 0.90 Prionotus evolans (Linnaeus) (striped searobin) 0.0104 0.0376 0.0100 0.0525 49–231 119.14 0.32 Strongylura marina (Walbaum) (Atlantic needlefish) 0.0104 0.0056 0.0199 0.0000 50–503 215.31 0.52 Anguilla rostrata (Lesueur) (American eel) 0.0052 0.0418 0.0199 0.0642 45–789 221.11 0.78 Orthopristis chrysoptera (Linnaeus) (pigfish) 0.0052 0.0279 0.0066 0.1693 28–231 105.20 0.44 Synodus foetens (Linnaeus) (inshore lizardfish) 0.0052 0.0167 0.0697 0.2014 38–282 158.53 0.32 Spyhraena borealis DeKay (northern sennet) 0.0052 0.0056 0.0000 0.0000 27–91 43.60 0.61 Morone saxatilis (Walbaum) (striped bass) 0.0052 0.0028 0.0066 0.0000 129–437 263.43 0.33 Monocanthus hipsidus (Linnaeus) (planehead filefish) 0.0052 0.0028 0.0000 0.0000 81–106 96.25 0.11 Conger oceanicus (Mitchill) (conger eel) 0.0052 0.0000 0.0498 0.0058 81–406 250.69 0.50 Gobiesox strumosus Cope (skilletfish) 0.0052 0.0000 0.0199 0.0117 30–66 50.100 0.24 268 Northeastern Naturalist Vol. 14, No. 2 Abundance Size Species (common name) 1996 1997 1998 1999 Range TL (mm) %CV Chilomycterus schoepfii (Walbaum) (striped burrfish) 0.0052 0.0000 0.0133 0.0292 43–236 153.03 0.30 Mycteroperca microlepis (Goode and Bean) (gag) 0.0052 0.0000 0.0000 0.0000 73–157 128.29 0.27 Opsanus tau (Linnaeus) (oyster toadfish) 0.0572 0.0530 0.0365 0.2131 21–239 110.82 0.41 Scorpaena plumieri Bloch (spotted scorpionfish) 0.0052 0.0000 0.0000 0.0029 50–60 55.00 0.13 Gasterosteus aculeatus Linnaeus (threespine stickleback) 0.0000 0.0808 0.0066 0.0000 27–72 61.12 0.14 Alosa sapidissima (Wilson) (American shad) 0.0000 0.0530 0.2522 0.1168 98–159 125.31 0.12 Ophidion marginatum (DeKay) (striped cusk eel) 0.0000 0.0432 0.0100 0.0642 77–239 152.89 0.29 Symphurus plagiusa (Linnaeus) (blackcheeked tonguefish) 0.0000 0.0362 0.1029 0.3766 33–168 96.58 0.37 Morone americana (Gmelin) (white perch) 0.0000 0.0195 0.0199 0.0029 57–315 156.57 0.33 Scomberomorus maculatus (Mitchill) (Spanish mackerel) 0.0000 0.0139 0.0166 0.0000 45–120 73.73 0.29 Tautoga onitis (Linnaeus) (Tautog) 0.0000 0.0139 0.0033 0.0117 30–315 114.03 0.59 Hypsoblennius hentz (Lesueur) (feather blenny) 0.0000 0.0084 0.0000 0.0438 34–123 85.23 0.24 Chaetodon ocellatus Bloch (spotfin butterflyfish) 0.0000 0.0056 0.0033 0.0058 20–85 52.18 0.40 Pollachius virens (Linnaeus) (pollock) 0.0000 0.0056 0.0000 0.0058 37–44 41.63 0.05 Archosargus probatocephalus (Walbaum) (sheepshead) 0.0000 0.0028 0.0033 0.0000 41–138 95.25 0.43 Aluterus schoepfii (Walbaum) (orange filefish) 0.0000 0.0028 0.0000 0.0000 87–259 165.75 0.36 Fistularia tabacaria Linnaeus (bluespotted cornetfish) 0.0000 0.0014 0.0100 0.0175 165–503 350.54 0.28 Eucinostomus gula (Quoy and Gaimard) (silver jenny) 0.0000 0.0000 0.0100 0.0175 39–172 66.94 0.54 Caranx hippos (Linnaeus) (crevalle jack) 0.0000 0.0000 0.0066 0.0000 38–116 73.63 0.41 Lagodon rhomboides (Linnaeus) (pinfish) 0.0000 0.0000 0.0066 0.1051 31–220 101.00 0.69 Myrophis punctatus Lütken (speckled worm eel) 0.0000 0.0000 0.0066 0.0000 82–88 85.00 0.05 Chaetodipterus faber (Broussonet) (Atlantic spadefish) 0.0000 0.0000 0.0033 0.0000 103 103.00 0.00 Eucinostomus argenteus Baird and Girard (spotfin mojarra) 0.0000 0.0000 0.0033 0.0000 39–172 66.94 0.54 Urophycis chuss Walbaum (red hake) 0.0000 0.0000 0.0000 0.0175 45–190 71.75 0.55 Diplodus holbrookii (Bean) (spottail pinfish) 0.0000 0.0000 0.0000 0.0058 156–157 156.50 0.01 Total Catch 19,243 35,881 30,129 34,296