1. Conclusions:  Hg levels in Gambusia did not follow a consistent pattern with hydroperiod.  Wild and caged fish results were similar.  Little difference in diets between hydroperiods.  Hg levels in fish across the landscape seem to be mediated by site characteristics, including availability and temperature. Conclusions:  Most Everglades fishes were omnivorous; few species were strict herbivores or piscivores.  Hg levels varied within and among taxa.  Trophic positions estimated from gut and stable-isotope analyses were similar.  Hg bioaccumulation was positively correlated with the trophic positions (diet) of aquatic animals. Accumulation and Fate of Mercury in an Everglades Aquatic Food Web William F. Loftus USGS - BRD, Everglades N. P. Field Station 40001 State Road 9336, Homestead, FL 33034 Joel C. Trexler Department of Biological Sciences Florida International University, Miami, FL 33199Acknowledgements Support provided by: Florida Dept. of Environmental Protection, Dr. Tom Atkeson; Florida International University; U. S. National Park Service; USGS-Biological Resources Division General Study Conclusions:  Hg is widespread in this Everglades aquatic food web.  Levels measured in ENP animals are very similar to those from northern Everglades.  Levels from Everglades can potentially cause health effects in fish and wildlife based on lab and field studies elsewhere. Applications: Data used in Bioaccumulation in Aquatic Systems Simulation model (BASS). Risk assessment for top-level predators. Food-web construction and analysis. Hypothesis generation for Phase II studies. Results and Discussion:  Most Everglades fishes have generalistic diets. Large native species feed mainly on invertebrates and fishes; smaller native fishes are primarily omnivorous. Few species are strict herbivores/detritivores (Table1, Figure 2).  The trophic scores/positions of the fishes based on prey volumes from guts agreed well with the ordination patterns from Nonmetric-multidimensional scaling (NMDS) (Figure 3) (Stress = 0.153, r 2 = 0.887). General Study Objectives Describe the trophic patterns of the fish community at high water in pond and spikerush habitats; Relate Total Hg in fishes and invertebrates to their trophic positions; Relate hydroperiod to Hg in the eastern mosquitofish PROBLEM: Mercury (Hg) is a neurotoxin to which the developing central nervous system of animals is particularly vulnerable. US FDA and EPA Hg Limits: 0.5-1.0 ppm: limited consumption but not by pregnant women or children >1.0 ppm: no consumption These Hg levels are exceeded by many Everglades fishes. This study was proposed to answer a series of questions posed by a multi-agency task force convened to address the mercury problem in the Everglades. These questions included understanding the extent of contamination in aquatic biota, the pathways by which Hg moves through the aquatic food web, and the processes that affect bioaccumulation. Rationale:  Differences in fish and invertebrate densities between short- and long-hydroperiod marshes may result from food-web differences.  Differing food resources may result in differences in mercury accumulation.  If Gambusia experience different levels of Hg, chronic exposure may result in tolerant genotypes. Transplanting of fish across hydroperiods may show differences in survival or growth. Objectives:  Measure total Hg levels in wild Gambusia from 3 pairs of short- and long- hydroperiod marshes across seasons (Figure 7).  Estimate Hg concentrations, Hg uptake, growth rates, and survival of caged Gambusia from those marshes to compare to wild-fish patterns.  Examine diet differences among hydroperiods from gut and stable-isotope analyses with relation to Hg patterns. PART 2 Table 2. Mean Hg + 1SE (n) levels in selected fishes and invertebrates from Site 1-L. Species (Trophic class)ng/g Invertebrates Tettigoniid species (1)2.03 + 2.46 (9) Planorbella duryi (1) 14.21 + 2.61 (9) Chironomid larvae (1.5)38.03 + 3.35 (22) Cladocera (1.5)46.35 + 6.45 (10) Procambarus fallax (1.5)64.33 + 3.28 (24) Pelocoris femoratus (2) 95.98 + 7.26 (22) Palaemonetes paludosus (2)186.31 + 12.22 (41) Dolomedes sp.(2)258.04 + 58.01 (9) Small Fishes Jordanella floridae (1)84.63 + 4.68 (25) Poecilia latipinna (1)105.47 + 14.36 (14) Heterandria formosa (2)258.37 + 12.59 (20) Fundulus chrysotus (3)253.81 + 28.35 (34) Lucania goodei (3)313.56 + 27.48 (20) Belonesox belizanus (5)500.95 + 51.82 (11) Large Fishes Erimyzon sucetta (2)119.44 + 14.95 (16) Lepomis microlophus (3)244.68 + 15.80 (27) Lepomis macrochirus (4)478.48 + 68.64 (14) Ameiurus natalis (4)535.57 + 40.62 (18) Micropterus salmoides (5)967.47 + 43.21 (24) Lepisosteus platyrhincus (5)1160.72 + 95.24 (10) Freshwater InvertebratesFreshwater Fishes Figure 6a,b. Relationship of δ 15 N values and transformed Hg levels for selected invertebrates (a, left), and fishes (b, right). Figure 5. Plot of stable-isotope signatures for Nitrogen and Carbon for common aquatic plants and animals from Site 1-L. Figure 2. Trophic classes of representative species based on prey volumes in guts. Jordanella Poecilia Tilapia Notropis Erimyzon Lucania Gambusia Ameiurus Lepomis Chaenobryttus Lepisosteus Belonesox Micropterus Table 1. Sample of the aggregated volumetric diet proportion data for fishes at high water PART 1 Methods:  Fishes collected from three habitats at high and low water in a long-hydroperiod wetland at Site 1-L (Figures 1) in Shark Slough, Everglades National Park. Only high-water samples used in Hg correlations.  ~ 4,000 specimens of 26 native fishes from 1977-1981, supplemented by 6 introduced species during 1995-1997, collected with rotenone, electrofishing, nets, and angling.  Small size of food items required the pooling of like taxa from each size-class of fish for volumetric analysis.Volumes measured to 0.001ml with a blood-sedimentation tube.  For Hg Total all samples collected with clean techniques.  Invertebrate samples were composited, while large fish were homogenized and subsampled.  Whole-body samples of animals were digested with acid in sealed ampules in an autoclave.  Total Hg measured with atomic-fluorescence spectrometer, after generating callibration curves and including Standard Reference Materials as checks for each run.  Trophic Classification Fish trophic scores were calculated by summing the products of the dietary contributions of the 34 prey types and their trophic levels (Adams et al. 1983). Scores were used to group fishes into five discrete trophic classes (guilds): 1 = Mainly Herbivorous, >50% plant material; 2 = Omnivorous, 25-50% plant material; 3 = Omnivorous, <25% Plant material; 4 = Omnivorous, mainly animal prey; 5 = Predominantly carnivorous, fish and decapods. Objectives:  Examine trophic relationships in the Everglades fish community to identify patterns of resource use by a temperate fauna in a subtropical, seasonal wetland.  Group species into trophic groups based on volumetric contributions of prey items and corroborate with stable-isotope analysis.  Measure total Hg in water, soils, plants, invertebrates, and fishes.  Correlate Hg levels with trophic positions of invertebrates and fishes based on gut and stable-isotope analyses. Figure 1: Three major aquatic habitats in Shark Slough, ENP. Hypothesis: Assuming that direct uptake is constant across taxa, Hg Herbivores < Hg Omnivores < Hg Carnivores Ex: © Noel Burkhead P. Latipinna Trophic class = 1 G. holbrooki Trophic class = 3 L. platyrhincus Trophic class = 5 METHYLMERCURY IN FLOATING PERIPHYTON ALL CYCLES, 1995 & 1996 (USEPA – Stober et al. 1998) 1-S 1-L 2-L 2-S 3-L 3-S mg/kg Figure 10. Map showing the various levels of Hg in periphyton in S. Florida. Figure 3. Relationship of fish diets in multivariate space from NMDS. Proximity of species in the figure is directly related to the similarity of their diets. Major prey types have been added as markers to help the reader. Trophic Class Figure 4. Figure 4. Mean sqr-transformed Hg levels (SD) for the five trophic classes of fishes. Hg vs. Trophic Class  The trophic scores and total mercury levels of the fishes were positively correlated (Spearman’s r = 0.706, P< 0.001, n=39), as were those of the invertebrates (Pearson’s r = 0.749, p < 0.001). There were significant differences in square-root- transformed total Hg concentrations among the species of fishes (ANOVA, F 31,515 = 52.035, P < 0.001). The five classes differed significantly in mean Hg levels (ANOVA, F, 4, 24 = 9.17, P < 0.0005) (Figure 4).  Nitrogen and carbon stable-isotope signatures were measured for aquatic animals and plants at Site 1-L (Figure 5). Those of invertebrates were highly correlated with Hg levels (Spearman’s r= 0.745, P = 0.02, n=10)(Figure 6a, Table 2). Fishes had the same result (Pearson's r = 0.806, P < 0.001, n=21) (Figure 6b, Table 2). Results and Discussion:  Mean concentrations varied from 38.4 ng g -1 in 3-S, June 1997 to 265.07 ng g -1 at 1-L, Spring, 1996. Hydroperiod significantly affected Hg concentrations and explained the greatest amount of variation in the model, but there were interaction effects among marsh hydroperiod, fish size, site, and time-of year (Figure 8).  Uptake rates in caged fish ranged from 0.25-3.61 ng g -1 d -1, and interactions occurred among time-of-year, site and hydroperiod. The patterns of uptake were similar for wild and caged Gambusia (Figures 8 and 9). Figure 9. Mercury uptake in caged Gambusia over 8 sample periods. Incomplete data  Survival normally exceeded 80%.  Growth varied with time of year, site and hydroperiod, but greatest growth occurred in the short-hydroperiod marshes.  The results of these studies suggest that dietary bioaccumulation determines mercury in Everglades aquatic animals, and that, although hydroperiod plays a role in mercury uptake, its effect varies with season and specific location (Figure 10). Methods:  Each month, a minimum of 30 specimens of wild fish were collected from each site for total Hg analysis.  As a control, long- and short-hydroperiod stock mosquitofish were captive- raised at ENP to be stocked into field cages.  Possible Hg-tolerance differences in the populations were examined by transplanting mosquitofish originating from short-hydroperiod stocks into cages at long-hydroperiod locations, and visa versa. Five cages at each site were stocked with progeny from the same hydroperiod type, and 5 alternate cages stocked with progeny from the other category (Figure 7).  At the trial’s end, stocked caged fish were recovered and wild fish were collected from the same site. All were measured and weighed (for growth analysis) and analyzed for Hg. Figure 8. Mercury uptake in wild Gambusia over 8 sample periods. Incomplete data LS Mean Differences [sqr Hg(ng g -1 )]  Uptake rates were generally greatest at Site 1 (North SRS), and lowest at Site 3 (TS).  At site 2, fish from 2-L usually had higher Hg uptake rates than at 2-S, but the effect of hydroperiod was generally reversed at the other two locations. Figure 7. Locations of sampling sites and cage study design. 3S 3L 2S 2L 1S 1L