Southeastern Naturalist 623 C.S. Andrews, L.E. Miranda, and R. Kroger 22001155 SOUTHEASTERN NATURALIST 1V4o(4l.) :1642,3 N–6o3. 44 Fish and Water Quality in the Forested Wetlands Adjacent to an Oxbow Lake Caroline S. Andrews1, Leandro E. Miranda2,*, and Robert Kroger1 Abstract - Forested wetlands represent some of the most distinct environments in the Lower Mississippi Alluvial Valley. Depending on season, water in forested wetlands can be warm, stagnant, and oxygen-depleted, yet may support high fish diversity. Fish assemblages in forested wetlands are not well studied because of difficulties in sampling heavily structured environments. During the April–July period, we surveyed and compared the water quality and assemblages of small fish in a margin wetland (forested fringe along a lake shore), contiguous wetland (forested wetland adjacent to a lake), and the open water of an oxbow lake. Dissolved-oxygen levels measured hourly 0.5 m below the surface were higher in the open water than in either of the forested wetlands. Despite reduced water quality, fish-species richness and catch rates estimated with light traps were greater in the forested wetlands than in the open water. The forested wetlands supported large numbers of fish and unique fish assemblages that included some rare species, likely because of their structural complexity. Programs developed to refine agricultural practices, preserve riparian zones, and restore lakes should include guidance to protect and reestablish forest ed wetlands. Introduction Forested wetlands represent some of the most distinct environments in the Lower Mississippi Alluvial Valley (LMAV). These wetlands have been described as mysterious and primordial (Mitsch and Gosselink 2007), supporting a diversity of unique aquatic species adapted to survive under seasonally risky water-quality conditions (Hoover and Killgore 1998, Wharton et al. 1982). Depending on the season, water in forested wetlands can be warm, stagnant, and oxygen-depleted (Conner and Buford 1998). Yet, in regions with high fish-species diversity such as the LMAV, forested wetlands reportedly include a high percentage of the total species count (Baker et al. 1991, Hoover and Killgore 1998). Three types of forested wetlands are often associated with oxbow lakes in the LMAV. Margin wetlands, also identified as fringe wetlands, are transitional areas between lakes and upland areas, and of variable width depending on bank topography and lake age (Snodgrass and Burger 2001). The natural levee of oxbow lakes may produce a littoral zone that allows only a narrow margin of wetland. However, ridge and swale formations found in large meanders may provide a wider marginal wetland on the swale side of oxbow lakes and access to additional depressional lands inside the meander bend (Hodges 1997). Terminus wetlands develop over 1Department of Wildlife, Fisheries, and Aquaculture, Mississippi State University, Mississippi State, MS 39762. 2US Geological Survey, Mississippi Cooperative Fish and Wildlife Research Unit, Mississippi State University, Mississippi State, MS 39762. *Corresponding author - smiranda@usgs.gov. Manuscript Editor: Quenton Tuckett Southeastern Naturalist C.S. Andrews, L.E. Miranda, and R. Kroger 2015 Vol. 14, No. 4 624 sediment plugs in one or both ends of oxbow lakes (Mitsch and Gosselink 2007). The shallower depth provided by sediment plugs facilitates development of these sites into forested terminus wetlands, and over time the forests advance towards the meander bend as lakes slowly lose depth due to sedimentation (Westlake et al. 2009). Contiguous wetlands abut lakes and may represent ancient channels or other depressions in forested or vegetative succession. These 3 types of forested wetlands are characterized by a mix of Taxodium distichum (L.) Rich. (Bald Cypress), Nyssa spp. (tupelos), Salix spp. (willows), Betula nigra L. (River Birch), Populus deltoides Bartram ex Marshall (Cottonwood), Quercus lyrata Walter (Overcup Oak), Carya aquatica (Michx. f.) (Water Hickory), and others depending on wetland age and hydrology (Hodges 1997, Shepard et al. 1998). Water quality and fish communities in forested wetlands are not well studied relative to other aquatic systems, likely due to the difficulties of sampling in these heavily structured environments. It is generally assumed that forested wetlands adjacent to waterbodies enrich fish assemblages and may provide temporary nursery habitats for many fish species (Jude and Pappas 1992, Ward et al. 1999). Killgore and Baker (1996) studied fish assemblages in a river channel and in adjacent forested wetlands. They reported that species richness was similar among these environments due to hydraulic mixing, but that some families were more common and overall abundances were higher in forested wetlands. Adams et al. (2007) reported that backwaters contiguous to a stream, intermittent floodplain wetlands, and isolated floodplain wetlands had distinct fish-assemblages—contiguous backwaters supported open-water species, isolated backwaters supported typical swamp species, and intermittent floodplain wetlands supported a mixture of species. Nevertheless, no equivalent information is available for fish assemblages that develop in forested wetlands adjacent to oxbow lakes. This information is needed to inform decisions about protection and restoration of the wetlands surrounding the hundreds of oxbow lakes that occur in the LMAV. We examined assemblages of small fish and water-quality conditions in margin wetlands, contiguous wetlands, and the open water of an oxbow lake. Specifically, our objective was to determine if the 2 forested wetlands accommodated unique fish assemblages relative to the open water. Water-quality conditions often play a major role in structuring fish assemblages, and water-quality conditions reportedly can be harsh in wetlands (Verhoeven et al. 2006). Thus, to aid in interpreting potential differences in fish assemblages among wetlands and open-water environments, a secondary objective was to investigate selected water-quality variables to determine if they differed temporally and spatially among environments. We included a temporal component because seasonal variation in water quality and seasonal-habitat needs of juvenile and other small fish may be important factors in structuring fish assemblages. We hypothesised that fish-taxa composition would differ among environments and that these differences would correspond to habitat preferences shaped by water quality. We further hypothesized that these differences would vary temporally over the course of the spring and early summer as water quality changed over the reproductive period. Southeastern Naturalist 625 C.S. Andrews, L.E. Miranda, and R. Kroger 2015 Vol. 14, No. 4 Study Site Blue Lake is an oxbow lake in west-central Mississippi. The 21-ha lake is shallow (mean depth = 1.3 m, maximum depth = 4.4 m), 3.5 km long, and permanently connected at its upper end to Gayden Brake, a 334-ha cypress-tupelo contiguous wetland with a mean depth of 0.6 m and a maximum depth of 1.1 m. Blue Lake and Gayden Brake reportedly are prehistoric channels of the Ohio River, which presently flows 400 km north of the study site (Fisk 1944). The connection between Blue Lake and Gayden Brake represents a unique transition zone that includes an open-water environment with mean and maximum depths of 2.6 m and 3.0 m, respectively, and a margin wetland with mean and maximum depths of 0.4 m and 1.0 m, respectively (Fig. 1). Precipitation in nearby Greenwood, MS during January– July 2012 (77 cm) was nearly normal (79 cm); thus, water levels in the study area were typical. Average Secchi-disk visibility in Blue Lake is 60 cm (Andrews 2013). Methods Water-quality assessment We used a Eureka Manta Multiprobe (Eureka Environmental, Austin, TX) to determine diel temperature and dissolved oxygen (DO) trends. We deployed a multiprobe 0.5 m below the surface in each environment to record water-quality Figure 1. Blue Lake and Gayden Brake, in the Bear Creek watersh ed, MS. The lower inset shows approximate location of light-traps fished along transects in (1) open water, (2) margin wetlands, and (3) contiguous wetlands. Each transect was ~100 m long. Southeastern Naturalist C.S. Andrews, L.E. Miranda, and R. Kroger 2015 Vol. 14, No. 4 626 parameters hourly at a stationary location (Fig. 1). The probes recorded complete diel water-quality cycles over 95 days from early April through mid-July 2012. We collected the multiprobes weekly and downloaded the data. Water-quality trends were represented by temperature means and maxima and DO mean and minima computed from complete 24-hour diel cycles. Fish collection We assessed trends in small-fish assemblages with light traps fished in fixedstation transects within each of the 3 environments. Light traps are selective for fish attracted to light and target primarily larval and juvenile stages, or adults of small fish-species (Hickford and Schiel 1999). We based our light-trap design on a modified quatrefoil pattern, ~25 cm x 25 cm x 30 cm with a 7-cm-wide vertical entrance covered with a 1-mm-mesh screen; a funnel-collection assembly as described by Secor et al. (1992) facilitated fish removal. The light source was a solar-powered high-intensity LED light (Gyekis et al. 2006). We suspended 6 light traps along each transect, with the lower end positioned approximately in line with the Manta multiprobes. We placed traps ~20–25 m apart, a distance reported to attenuate light intensity so that trap collections are independent (Fisher and Bellwood 2002). We deployed the traps from early April through mid-July approximately weekly at nearly the same locations to ensure standardized sampling. The sampling period represented spring and early summer, a time span during which most local fish species start and finish spawning (Ross 2001). We set all traps at dusk and retrieved them ~40 h later at dawn, allowing for 2 nightly cycles. Upon collection, we preserved individual-trap contents in 10% formalin until identification with taxonomic keys (Wallus and Simon 2008). We pooled the catches from the 6 traps deployed in each transect by sampling date to reduce statistical complications associated with excessive zero counts. Statistical analysis We analyzed water-quality and fish data to assess patterns across the 3 environments. We assessed each data matrix with a permutation multivariate analysis of covariance (perMANCOVA; 1000 permutations) with either temperature (mean and maxima), dissolved oxygen (mean and minima), or fish counts (multiple taxa) as the response variables; environment type as the predictor variable; day-of-year (DOY) as a covariate; and an interaction term between environment type and DOY. We included DOY to account for temporal variation across the sampling season and the interaction term to detect potential differences in response among environments as the season progressed. Variables were loge(x + 1)-transformed as needed to meet assumptions of linearity or distribution. We analyzed the water-quality data using a Euclidean-similarity coefficient and the fish data using a Bray-Curtis coefficient. Analyses were conducted with PRIMER version 6 (Clarke and Gorley 2006) and the PERMANOVA+ add-on package (Anderson et al. 2008). If the perMANCOVA identified a statistically significant (P ≤ 0.05) DOY effect on fish counts, we examined cumulative fish catch of selected taxa over DOY. We expected that cumulative catches would exhibit a sigmoidal response over time, Southeastern Naturalist 627 C.S. Andrews, L.E. Miranda, and R. Kroger 2015 Vol. 14, No. 4 with a relatively flat curve initially as few fish were caught early in the year, a rising curve as catch rates increased through mid-season, and a flattening curve once catch rates slowed or stopped later in the season. The sigmoidal pattern was fit with a logistic model as: y = bo + b1·DOY + b2·(DOY × environment) (1) and y = loge [(1 / Pcum) - 1], (2) where Pcum = cumulative frequency of catch expressed as a proportion (0 - 1), b0 = intercept of model, b1 = slope of model for DOY, and b2 = slope modification for DOY according to environment. In equation (1), b0 / - (b1 + b2) represents the inflection point in the sigmoid curve given in days. The inflection point reflects the DOY when 50% of the fish had been collected. A significant b2 suggests that inflection points occurred at different DOY depending on environment. If so, a t-test determined which paired environments differed. The logistic analyses were run using the GLM procedure (SAS 2008). Results Water-quality assessment The water-quality data overlapped with the light-trap samples but did not always coincide, because the water-quality equipment could be deployed for longer time periods. Due to battery issues associated with multiprobe deployment, environments were unequally sampled, with 20 cycles in the open-water environment, 45 cycles in the margin wetland, and 30 cycles in the contiguous wetland. The perMANCOVA detected spatial (F = 4.8, P < 0.01), temporal (F = 80.1, P < 0.01), and interaction (F = 5.1, P < 0.01) effects on temperature and DO descriptors. These results suggested that there were differences among environments, conditions changed over time, and trends differed among all or some environments. In general, the open-water and contiguous wetlands followed similar temporal trends in mean and maximum temperature and in mean DO (Fig. 2). Although the trends were similar, temperature was consistently higher in the contiguous wetland than the open water (~1.5–2 oC warmer), and mean DO was consistently lower (~1.8 mg/L lower). The margin wetland followed a different trend in that it had a higher mean and maximum temperature and mean DO early in the sampling season, but values for these parameters were lower than those of other the 2 environments late in the sampling season (Fig. 2). Minimum DO decreased over the sampling season at different rates in all environments, with the margin wetland exhibiting the fastest decrease and reaching levels below 2 mg/L in early June, about 1 month sooner than the contiguous wetland. By the time the study was concluded on DOY 196, DO level remained above 2 mg/L only in the open water. Fish collection Fish catches in the light traps included 9 families, 14 taxa, and 1940 fish ranging in total length (TL) from 7 to 80 mm (mean = 31 mm). Catch rates were Southeastern Naturalist C.S. Andrews, L.E. Miranda, and R. Kroger 2015 Vol. 14, No. 4 628 highest in the margin wetland, followed by the contiguous wetland and then the open-water environment (Table 1). Taxa richness followed the same pattern. The most numerous individuals represented 3 taxa: Lepomis spp. (sunfish), Micropterus salmoides (Largemouth Bass), and Labidesthes sicculus (Brook Silverside). Nevertheless, the biggest differences among environments were not in the catch rates of the most abundant species, but in the infrequent catches of uncommon species (Table 1). The perMANCOVA detected spatial (F = 1.8, P = 0.05), temporal (F = 8.2, P less than 0.01), and interaction (F = 2.6, P < 0.01) effects on catches of the 14 taxa. The interaction effect indicated that spatial differences occurred, but the differences depended on time. These differences were inspected for the most common taxa with equation (1). Sunfish inflection points were DOY 155, 162, and 156 in open water, margin wetland, and contiguous wetland, respectively, and were not significantly Figure 2. Diel water-quality parameters in 3 habitats of Blue Lake, MS, 5 April–14 July 2012 (DOY 96–196). Southeastern Naturalist 629 C.S. Andrews, L.E. Miranda, and R. Kroger 2015 Vol. 14, No. 4 different (P = 0.87). Largemouth Bass inflection points were DOY 109, 136, and 121 in open water, margin wetland, and contiguous wetland, respectively, and were significantly different (P < 0.01). Brook Silverside inflection points were DOY 162, 183, and 165 in open water, margin, and contiguous wetland, respectively, and were significantly different (P < 0.01). Across these 3 taxa, inflection points were consistently earliest in open water and latest in the margin wetland, earlier for Largemouth Bass and latest for Brook Silverside, and least-well defined (i.e., highest standard error) for sunfish (Fig. 3). Assessment of temporal trends of other taxa was not possible because catches were low and sporadic. Table 1. Catch rate (fish/6 light-traps) of 14 taxa captured in 3 environments of Blue Lake, MS, with passive light-traps during 13 sampling periods, April–July 2012. TL range = 7–80 mm (mean = 31). Margin Contiguous Taxa Common name Open water wetland wetland Lepomis spp. sunfishes 15.7 30.2 28.8 Micropterus salmoides (Lacepède) Largemouth Bass 13.4 40.4 18.5 Labidesthes sicculus (Cope) Brook Silverside 17.3 11.8 23.3 Gambusia affinis (Baird and Girard) Western Mosquitofish 0.7 1.8 4.6 Pomoxis spp. Crappies 1.6 0.9 0.9 Fundulus chrysotus (Günther) Golden Topminnow - 2.1 0.5 Percidae Darters - 1.8 0.2 Notemigonus crysoleucas (Mitchill) Golden Shiner - 0.5 0.2 Opsopoeodus emiliae Hay Pugnose Minnow - - 0.5 Catostomidae Suckers - 0.2 - Centrarchus macropterus (Lacepède) Flier - 0.2 - Elassoma zonatum Jordan Banded Pigmy Sunfish - 0.2 - Ictaluridae Catfishes - 0.2 - Lepisosteidae Gars - - 0.2 All 48.7 90.3 77.7 Figure 3. The DOY when 50% of the fish had been collected during the DOY 96– 196 sampling period. Values are identified according to open water, margin wetland, and contiguous wetland, and according to 3 taxa. Error bars represent ± 1 SE. Southeastern Naturalist C.S. Andrews, L.E. Miranda, and R. Kroger 2015 Vol. 14, No. 4 630 Discussion The 3 distinct environments we studied showed differences in water-quality conditions and fish assemblages. The contiguous wetland was the warmest and least oxygenated of the 3 environments, but had greater catch-rates compared to the cooler, better-oxygenated open-water environment. While it seems counterintuitive that heavily shaded water in the forest was warmer than open water, we suspect that protection from recurring cold fronts in winter and spring, and reduced depth and wind in summer kept the contiguous wetland warmer. The margin wetland exhibited the slowest rise in temperature and the steepest decline in oxygen, yet also showed the highest relative fish abundance and taxa ric hness. As expected, DO was highest in the open water and, for most of the season, lowest in the contiguous wetland. The open water was deeper and possibly less affected by sediment respiration than the contiguous wetland. Moreover, besides receiving less sunlight due to increased canopy cover, the contiguous wetland accumulated leaf and woody debris generated by the forest, creating additional oxygen demand via microbial respiration (Sharitz and Mitsch 1993). We expected that oxygen conditions in the margin wetland would be intermediate because mixing with the open water would temper the effects of the forest. Nevertheless, oxygen conditions decreased faster in the margin wetland over the course of the season than in the other environments. Over time, the margin wetland developed dense mats of emergent aquatic macrophytes (principally Alternanthera philoxeroides (Mart.) Griseb. [Alligator Weed]), which tend to reduce DO and increase diel DO fluctuations (Miranda and Hodges 2000). Unlike the contiguous wetland, the margin wetland had more light that allowed development of emergent aquatic macrophytes. By July, mean DO in the margin wetland had dropped to 2 mg/L, a level generally considered a lower threshold for warm-water fish (Breitbur g et al. 2009). Water-quality conditions may have different effects on the fish assemblages in the 2 types of forested wetlands. Fish inhabiting the contiguous wetland may find it more difficult to escape occasional hypoxia, whereas those in the margin wetland have easier access to open water where water quality is higher. This effect is analogous to that reported in large stands of aquatic macrophytes (Miranda et al. 2000). On one hand, the plants provide abundant invertebrate foods, cover, and an environment with few or no large predators. On the other hand, water-chemistry conditions can be precarious, and fish at these sites run the risk of being trapped in unsuitable microhabitats during the daily reshuffling of physical and chemical conditions. Activity changes, such as increased energy allocation to ventilation, may nullify the benefits of an abundant food supply. Also, surfacing for respiration at the air–water interface and other movements forced by adverse water-quality conditions may increase the chance of predation by other fish or by avian predators. Shallow wetlands, like large stands of aquatic macrophytes, provide food and cover but can have negative water-quality consequences. By summer, when DO levels drop to potentially dangerous levels, we speculate that only specialized species can remain in contiguous wetlands, but fish may exhibit daily movements in and out of margin wetlands. Southeastern Naturalist 631 C.S. Andrews, L.E. Miranda, and R. Kroger 2015 Vol. 14, No. 4 Inflection points in sigmoid cumulative curves provided insight into distribution of common fish over time. Of the 3 most-common taxa trapped, we collected Largemouth Bass earliest, sunfish next, and Brook Silverside soon after, following an expected temperature-related sequence in hatching (Ross 2001). For each of these 3 taxa, 50% of the catch was attained earliest in open water and latest in margin wetland. Small fish recruited into these environments over the spring and summer may be disproportionately recruited into forested wetlands. This pattern is particularly true of the margin wetland given that emergent macrophytes that develop over the growing season potentially attract small fish seeking shelter from predation. We do not know if these 3 generalist taxa remained in the forested wetlands after water quality was reduced because as they grew they were no longer captured by the light traps; however, we hypothesize that many of them exited the wetlands in summer after water-quality conditions declined. Additional research is needed to track latesummer and fall use of these wetlands by fish. We predicted that forested wetlands would attract more small individuals and more small species than the open-water habitats. As expected, in comparison to the open water, catch was 1.9 times higher in the margin wetland and 1.6 times higher in the contiguous wetland than in the open water. Species counts were 2.4 times higher in the margin wetland than in the open water and 2.0 times higher in the contiguous wetland than in the open water. Relative to species composition, both forested wetlands contributed rare species to the overall collection. We collected several species including Opsopoeodus emiliae (Pugnose Minnow), Centrarchus macropterus (Flier), Elassoma zonatum (Banded Pygmy Sunfish), and Fundulus chrysotus (Golden Topminnow) only in the forested wetlands, and there only infrequently. These species comprise a distinct wetland assemblage in the LMAV, are typically only found in shallow semi-isolated waterbodies, and their distribution in the LMAV has been impacted disproportionally by wetland clearing and habitat degradation. These species occur in the margin and contiguous wetlands for several reasons including their adaptation to reduced water quality and their propensity to exploit structurally complex habitats. Other than these rare species, fish assemblages in the open water and wetlands were similar. The results of our study demonstrated that habitat diversity provided by forested wetlands associated with oxbow lakes contribute to fish-species diversity in at least 2 ways. First, these unique habitats promote rare species that show habitat specificity. Second, they contribute to temporal diversity in environmental conditions that help sustain habitat-generalist species. Although some of the species we collected in forested wetlands are now considered rare, they might have once been more common. Originally, the LMAV included nearly 10 million hectares of forested wetlands, but towards the end of the 20th century there were only about 2 million hectares left (MacDonald et al. 1979). The remaining wetlands have experienced accelerated sedimentation rates from adjacent agricultural operations and increased risk of dewatering during drought because the water table has been substantially lowered by groundwater pumping in sections of the LMAV (Kingsbury et al. 2014). Programs directed at refining agricultural practices, preserving riparian zones, and Southeastern Naturalist C.S. Andrews, L.E. Miranda, and R. Kroger 2015 Vol. 14, No. 4 632 restoring lakes should consider expanding their conservation efforts to include protection and reestablishment of forested wetlands. A limitation of our study was that it represented a single lake/wetlands system. This lack of treatment replication limits the applicability of our conclusions. However, it is not unreasonable to hypothesize that margin wetlands in oxbow lakes in general may support more fish of all life stages and higher species richness because margin wetlands represent a large edge-zone that includes a mixture of fish species adapted to both open water and nearshore structure. When water quality deteriorates in margin wetlands, fish may find it easier to temporarily escape into open water. Conversely, whereas contiguous wetlands with their extensive structural complexity have the potential to support large numbers of fish and unique fish assemblages, they may not always do so because fish may avoid them seasonally to avoid entrapment during periods with inadequate water quality. The extent to which forested wetlands interact with oxbow lakes to shape overall fish assemblages is likely to be time- and site-specific depending on factors such size and depth of the lake and of its associated wetlands, forest type and maturity stage, distribution in the landscape, and inter-annual differences in hydrology and temperature. Additional replication is needed to improve understanding of the temporal and spatial interplay between water quality, fish assemblages, and forested wetlands in the LMAV. Acknowledgments We thank D. Devries and E. Dibble for lending us light traps. B. Botti, C. Shoemaker, D. Goetz, E. Mower, L. Kaczka, and M. Arnold provided field support. J. Killgore, D. Faust, and 2 anonymous referees provided helpful reviews. This work was supported in part by the US Army Corps of Engineers. Specimen collections were authorized under Mississippi State University’s Institutional Animal Care and Use Committee protocol number 08–034. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the US Government. Literature Cited Adams, S.R., B.S. Williams, M.D. Schroeder, and R.L. Clark. 2007. Abundance and distribution of fishes in riparian wetlands of the Arkansas River. Arkansas Game and Fish Commission special report, Little Rock, AR, 108 pp. Anderson, M.J., R.N. Gorley, and K.R. Clarke. 2008. PERMANOVA for PRIMER: Guide to Software and Statistical Methods. PRIMER-E, Plymouth, UK. 214 pp. Andrews, C.S. 2013. Floodplain lake assessment and fish-assemblage dynamics in the Mississippi Alluvial Valley. M.Sc. Thesis. Mississippi State University, Mississippi State, MS. 74 pp. Baker, J.A., K.J. Killgore, and R.L. Kasul. 1991. Aquatic habitats and fish communities in the lower Mississippi River. Aquatic Sciences 3:313–356. Breitburg, D.L., D.W. Hondorp, L.A. Davias, and R.J. Diaz. 2009. Hypoxia, nitrogen, and fisheries: Integrating effects across local and global landscapes. Annual Review of Marine Science 1:329–349. Clarke, K.R., and R.N. Gorley. 2006. PRIMER v6: User’s Manual/Tutorial. PRIMER-E, Plymouth, UK. 190 pp. Southeastern Naturalist 633 C.S. Andrews, L.E. Miranda, and R. Kroger 2015 Vol. 14, No. 4 Conner, W.H., and M.A. Buford. 1998. Southern deepwater swamps. Pp. 261–289, In M.G. Messina and W.H. Conner (Eds.). Southern Forested Wetlands: Ecology and Management. CRC Press, Boca Raton, FL. 616 pp. Fisher, R., and D.R. Bellwood. 2002. A light-trap design for stratum-specific sampling of reef-fish larvae. Journal of Experimental Marine Biology and Ecology 269:27–37. Fisk, H.N. 1944. Geological investigation of the alluvial valley of the lower Mississippi River. US Army Corps of Engineers, Mississippi River Commission, Vicksburg, MS. 78 pp. Gyekis, K.F., M.J. Cooper, and D.G. Uzarski. 2006. A high-intensity LED light source for larval fish and aquatic invertebrate floating quatrefoil light traps. Journal of Freshwater Ecology 21:621–626. Hickford, M.J.H., and D.R. Schiel. 1999. Evaluation of the performance of light traps for sampling fish larvae in inshore temperate waters. Marine Ecology Progress Series 186:293–302. Hodges, J.D. 1997. Development and ecology of bottomland-hardwood sites. Forest Ecology and Management 90:117–125. Hoover, J.J., and K.J. Killgore. 1998. Fish communities. Pp. 237–260, In M.G. Messina and W.H. Conner (Eds.). Southern Forested Wetlands: Ecology and Management. CRC Press, Boca Raton, FL. 616 pp. Jude, D.J., and J. Pappas. 1992. Fish utilization of Great Lakes coastal wetlands. Journal of Great Lakes Research 18:651–672. Killgore, K.J., and J.A. Baker. 1996. Patterns of larval-fish abundance in a bottomland hardwood wetland. Wetlands 16:288–295. Kingsbury, J.A., J.R.B. Barlow, B.G. Katz, H.L. Welch, R.W. Tollett, and L.S. Fahlquist. 2014. The quality of our Nation’s waters: Water quality in the Mississippi embayment– Texas coastal-uplands aquifer system and Mississippi River Valley alluvial aquifer, south-central United States, 1994–2008 US Geological Survey Circular 1356. 72 pp. Available online at http://dx.doi.org/10.3133/cir1356. Accessed 21 February 2015. MacDonald, P.O., W.E. Frayer, and J.K. Clauser. 1979. Documentation, chronology, and future projections of bottomland-hardwood habitat loss in the Lower Mississippi Alluvial Plain. Volume I: Basic report. US Fish and Wildlife Service, Ecological Services, Vicksburg, MS. 133 pp. Miranda, L.E., and K.B. Hodges. 2000. Role of aquatic-vegetation coverage on hypoxia and sunfish abundance in bays of a eutrophic reservoir. Hydrobiologia 427:51–57. Miranda, L.E., M.M. Driscoll, and M.S. Allen. 2000. Transient physicochemical microhabitats allow fish survival in inhospitable aquatic-plant stands. Freshwater Biology 44:617–628. Mitsch, W.J., and J.G. Gosselink. 2007. Wetlands. 4th Edition. Wiley, Hoboken, NJ. 600 pp. Ross, S.T. 2001. Inland Fishes of Mississippi. University Press of Mississippi, Jackson, MS. 624 pp. SAS Institute, Inc. 2008. Statistical Analysis System, Version 9.2. Cary, NC. 1292 pp. Secor, D.H., J.M. Dean, and J.H. Hansbarger. 1992. Modification of the quatrefoil light-trap for use in hatchery ponds. The Progressive Fish-Culturist 54:202–205. Sharitz, R.R., and W.J. Mitsch. 1993. Southern floodplain forests. Pp. 311–371, In W.H. Martin, S.G. Boyce, and A.C. Echternacht (Eds.). Biodiversity of the Southeast United States/Lowland Terrestrial Communities. Wiley, New York, NY. 528 pp. Shepard, J.P., S.J. Brady, N.D. Cost, and C.G. Storrs. 1998. Classification and inventory. Pp. 3–28, In M.G. Messina and W.H. Conner (Eds.). Southern Forested Wetlands: Ecology and Management. CRC Press, Boca Raton, FL. 616 pp. Southeastern Naturalist C.S. Andrews, L.E. Miranda, and R. Kroger 2015 Vol. 14, No. 4 634 Snodgrass, J.W., and J. Burger. 2001. Management of wetland-fish populations: Population maintenance and control. Pp. 357–386, In R.B. Rader, D.P. Batzer, and S.A. Wissinger (Eds.). Bioassessment and Management of North American Freshwater Wetlands. Wiley, New York, NY. 480 pp. Verhoeven, J.T.A., B. Arheimer, C. Yin, and M.M. Hefting. 2006. Regional and global concerns over wetlands and water quality. Trends in Ecology and Evolution 21:96–103. Wallus, R., and T.P. Simon. 2008. Reproductive Biology and Early Life-history of Fishes in the Ohio River Drainage: Elassomatidae and Centrarchidae. CRC Press, Boca Raton, FL. 472 pp. Ward, J.V., K. Tockner, and F. Schiemer. 1999. Biodiversity of floodplain river ecosystems: Ecotones and connectivity. Regulated Rivers: Research and Management 15:125–139. Westlake, D.F., J. Kvet, and A. Szczepanski. 2009. The Production Ecology of Wetlands: The IBP Synthesis. University Press, Cambridge, UK. 592 pp. Wharton, C.H., W.M. Kitchens, E.C. Pendleton, and T.W. Sipe. 1982. The ecology of bottomland- hardwood swamps of the southeast: A community profile. US Fish and Wildlife Service, Biological Service Program, FWS/OBS-81/37, Washington, DC. 133 pp.