MERCURY CONTAMINATION IN FLORIDA PANTHERS
EXECUTIVE SUMMARY
Mercury was identified as an important contaminant in free-ranging panthers, raccoons, otters, and alligators but not in bobcats in southern Florida. Those animals with relatively high levels of mercury were found in the Shark River Slough of the Everglades National Park, Water Conservation Area 3A, and adjacent wetlands. Mercury toxicosis may have been responsible for at least one panther death in the Everglades National Park and is strongly implicated in two others since 1989.

There were significant differences in levels of mercury in panthers when compared by geographical location and age. Average levels of mercury were greatest in panthers from the eastern portion of the range, particularly from the Shark River Slough area, and lowest values were noted in panthers from north of Alligator Alley. The mean liver mercury level for the younger group of panthers (less than 8 years old) living in the eastern range was significantly higher than that from the western range. When only the western group was considered, older animals had significantly higher liver mercury levels than did younger ones. The liver mercury burden was much higher among older animals living in the Fakahatchee Strand State Preserve than the single older animal living north of that area.

Females with elevated mercury had poorer reproductive success than those with low mercury levels. However, concomitant nutritional stress associated with their prey base probably also contributed to the poor reproductive performance of the females in the Fakahatchee Strand State Preserve but apparently not in the Everglades National Park.

The most probable source of mercury contamination in panthers is via the food chain. The panthers north of Alligator Alley had the lowest levels of mercury and fed primarily on white-tailed deer and feral hogs. Although nothing is known about tissue mercury levels in the hog, mercury levels have been shown to be low in deer tissues from southern Florida. Panthers with elevated levels of mercury occur where they consume mercury-contaminated non-ungulate prey as part or all of their diet (raccoon is probably the primary source of mercury). Mercury levels in panthers living in the Fakahatchee Strand State Preserve have dropped significantly (P<0.01) since the fall of 1987 when land management actions were initiated to enhance deer density in that area.

Chronic exposure to mercury, resulting in mortality and lowered reproductive success, may be a significant factor responsible for lower than expected population densities of panthers in large portions of their range and is likely contributing to the extinction of this endangered mammal.

Recommendations concerning continued monitoring and additional research are presented.

MERCURY CONTAMINATION IN FLORIDA PANTHERS
INTRODUCTION
Mercury contamination of freshwater fish and alligators (Alligator mississippiensis) in various watersheds throughout the State of Florida has recently become a human health issue (Delany et Al. 1988, Hord et al. 1990). Health advisories have been issued to curtail the consumption of largemouth bass (Micropterus salmoides) in southern Florida and the entire Everglades watershed has been closed to the commercial harvest of alligators due to elevated mercury residues in their flesh.

Mercury is accumulated and concentrated in the aquatic food chain and the highest levels occur in the longer lived species at the upper trophic levels (Clarkson and Marsh 1982, Eisler 1987). The Florida panther (Felis concolor coryi) is a top terrestrial mammalian carnivore in the southern Florida ecosystem. White-tailed deer (Odocoileus virginianus) and feral hog (Sus scrofa) are the preferred prey, but, in some areas, panthers also consume small mammals, i.e. raccoon (Procyon lotor), armadillo (Dasypus novemcinctus), and rabbit (Sylvilagus sp.), as a significant part of their diet (Maehr et al., 1990, Roelke et al., 1986). Additionally, panthers in the Everglades National Parks (ENP) have been documented to consume alligators (O. Bass, ENP, pers. comm.).

We first became aware of mercury contamination in Florida panthers when a 3- to 4-year old radio-instrumented female (FP#27) died in the ENP during July, 1989. Her carcass was retrieved 24-36 hours postmortem. both gross and histopathologic examination were unremarkable, although brain tissue was too autolyzed for a definitive examination. The relatively good condition of the carcass and lack of significant pathologic findings coupled with concern over the loss of a prime breeding-age female (at the time one of only three in the ENP) prompted a more extensive examination with selected tissues analyzed for pesticides, PCB's, and heavy metals. The only contaminant found to be present in relatively high levels was mercury. The liver contained 110 parts per million (ppm) (wt. wet) of mercury. Death due to mercury toxicosis was reported in feral domestic cats with liver concentrations of 37-145 ppm (Harada and Smith, 1974).

This report presents the magnitude and distribution of mercury levels in various tissues from free-ranging Florida panthers sampled over the past 13 years. Mercury concentrations in selected tissues from white-tailed deer, raccoon, bobcat, otter, and alligator are also reported.

Mercury in Southern Florida ecosystems: Historical Perspective

Mercury contamination of Florida's ecosystems has not been studied intensively until recently. Early in the 1970's, an ENP survey resulted in measured mercury concentrations ranging from 0.05 ppm in fish to 2.62 ppm in white ibis (Eudocimus albus) (Ogden et al. 1973). The U.S. Fish and Wildlife Service (FWS) conducted a study in 1986 of contaminant concentrations in selected fish and bird species inhabiting the Loxahatchee National Wildlife Refuge (LNWR). Mercury concentrations in anhinga (Anhinga anhinga) livers ranged from 0.42 to 2.72 ppm and largemouth bass whole body mercury concentrations ranged from 0.56 to 1.05 ppm. Largemouth bass and bullheads (Ictalurus spp.) were again collected from ENP and LNWR in April and May 1989, respectively. Fillets from ENP bass had levels ranging from 0.26 to 3.53 ppm, and averaged 0.96 ppm (Loftus, 1990). Yellow bullhead (I. natalis) fillets from ENP were similar to the ENP bass with a mean concentration of 0.91 ppm (range: 0.64-1.17). At LNWR, the mean concentration in largemouth bass fillets was 0.81 ppm (range: 0.35-2.13). Concentrations in brown bullheads (I. Nebulosus) from LNWR averaged 0.51 ppm. As a result, fishery advisories were posted at both ENP and LNWR.

The Florida Game and Fresh Water Fish Commission has collected fish samples from a number of areas since 1989 in the Shark river Slough north of Everglades National Park. Largemouth bass tissue collected from the south end of Lake Okeechobee had a mean mercury concentration of 0.22 ppm (range: 0.20-0.30). Bass samples collected from Canal L-20, south of Lake Okeechobee, had a mean of 0.39 ppm (range 0.22-0.59). Samples of bass from sites within the agricultural area on Canals L-18 and L-23 had averaged 0.93 and 0.73 ppm, respectively. Mercury concentrations increased sharply in bass tissue sampled from sites south of the Everglades Agricultural Area into Conservation areas 2A and 3 with bass containing as high as 4.4 ppm. Concentrations averaged 2.47 ppm (range: 1.10-3.40) in bass from Canal L-35B, 2.7 ppm (range 2.3-3.1) from Canal C-123 and 2.17 ppm (range: 1.04-4.40) from Canal L-67A.

During the fall of 1989, additional fish, both freshwater and marine species, were collected from ENP and analyzed (Loftus, 1990). Mercury concentrations ranged from a low of 0.056 ppm in a gray snapper (Lutijanus griseus) to a high of 1.90 ppm in an oscar (Astronotus ocellatus) (R. Pennington, USFWS, pers. comm.). Mean concentrations and species analyzed were: oscar=1.28 ppm, blue gill (Lepomis macrochirus)=1.01 ppm, Crevalle jack (Caranx hippos)=0.29 ppm, spotted sea trout (Cynoscion nubulosus)=0.29 ppm, and gray snapper (Lutjanus griseus)=0.15 ppm. Ten largemouth bass were also collected from Big Cypress National Preserve (BCNP). Mercury levels in fillets from these fish ranged from 0.51 to 2.91 ppm with a mean of 1.01 ppm. The FWS sampled fish from several refuges, including LNWR and Florida Panther National Wildlife Refuge, the following year. Mercury concentrations detected in at least some fish from each of the refuges sampled exceed the Florida HRS Health Advisory for once/week consumption (0.5 to 1.5 ppm) (Brim and Facemire, USFWS, unpublished data).

None of these studies, however, addressed the source of mercury contamination in Florida's aquatic ecosystems or the impact that bio-magnification of methylmercury (the most toxic organic form of mercury) through the food chain may be having on top predators.

Mierle (1990) stated that atmospheric combustion of pollutants are the primary source of mercury contamination in lake fish in Canada. He determined that at least 57 percent and perhaps, as much as 70 percent of the total mercury input into a central Ontario (Canada) lake was in the form of wet deposition. An atmospheric source may be an important source to consider in Florida as well. For example, it is estimated that approximately nine tons of mercury were emitted into the atmosphere by solid waste incinerators in Florida during 1989 (T. Rogers, Florida DER, pers. comm.) and more incinerators are coming on line yearly. It is known that plants take up mercury from the soil (Schacklette 19780), and Simons (1991) estimated that over 10 tons of mercury could be released into the atmosphere annually from burning and processing sugar cane if mercury concentrations in sugar cane average 0.5 ppm. However, this figure was based only on mercury content of the cane (0.003 to 1.12 ppm), and did not consider the possibility of mercury on the external surface of the cane as a result of atmospheric deposition. The residence time for atmospheric mercury, of which some 25 to 30 percent is of anthropogenic origin, has been estimated by Clarkson et al. (1984) to be somewhere between 6 and 90 days.

The oxidation of natural pea soils of south Florida has previously been hypothesized as a major source of mercury contamination (USFWS 1989). Simons (1991) noted that the South Florida Water Management District found mercury concentrations in peat ranging from 0.1 to 0.3 ppm. Assuming (1) that on the average, 1.12 inches of peat is lost to oxidation annually, (2) that the average mercury content of peat is 0.2 ppm, (3) that oxidation occurs uniformly over the approximately 430,000 acres of peat soils in south Florida, and (4) that methylation of mercury is occurring in peat as a result of natural biological activity associated with oxidation, then approximately 8.2 tons of methylmercury could be released from peat deposits annually (Simons 1991). However, mercury levels in sediment cores from lakes and estuaries have increased twofold to fivefold since pre-cultural times (Eisler 1987), and that same phenomenon is likely true of peat sediments. Thus, it is likely that mercury concentrations in the remaining peat will increase over time due to wet deposition of atmospheric mercury.

MATERIALS AND METHODS
Panther tissue collections

Two-hundred two tissue samples from 52 (29 male and 23 female) free-ranging Florida panthers were collected opportunistically between 1978 and 1991. Hair and whole blood from living panther (n=43) were sampled during routine capture on one or more occasions for radio-telemetry studies or when removed from the wild for rehabilitation (Roelke 1990). Liver kidney, muscle, brain, hair, and blood were collected at necropsy from dead panthers (n=26). A limited number of fecal samples were examined (n=7) from panthers which had whole blood samples available. Nineteen of the panthers were sampled both when live captured and when necropsied later. All panthers were collected in southern peninsular Florida. Forty-one individuals were sampled from the ENP (Dade and Monroe Counties). One other individual was sampled outside this range in Palm Beach County. For discussions in this paper, panther locations within the Big Cypress Swamp and Everglades physiographic region of southern Florida are described using the following designations (roughly from northwest to southeast): (1) land north of Alligator Alley/I-75 including the Florida Panther National Wildlife Refuge, Bear Island Unit of the Big Cypress National Park (BCNP) and adjacent private ranches to the north and east (NA); (2) the Fakahatchee Strand State Preserve (FS); (3) Raccoon Point (RP) area in eastern BCNP (now part of Corn Dance Unit); (4) Shark River Slough (SS) area of ENP (includes east Everglades (EE)); and (5) Long Pine Key/Hole-in-the-Donut/Taylor Slough (LPK) area of ENP (Figure 1).

Muscle and liver tissues were collected from four to seven raccoons captured in 1990-1991 at each of nine sites (n=48) within respective panther habitat. The nine collection sites included those described for the panther 1) NA, 2) FS, 3) RP, plus these additional sites: 4) Loop Road (LR) Unit of BCNP; 5) Water Conservation Area 3A (3A); 6) Shark River Slough (SS); 7) Long Pine Key (LPK); 8) Taylor Slough (TS); and 9) Flamingo Visitor Center (FL), ENP (Figure 1) (TS was considered separately from LPK for raccoons but not for panthers). The raccoons were captured in April to May 1990 with the exception of the TS, FL, and three of five LPK raccoons which were captured in the spring of 1991. The FL raccoons were collected as they fed at dumpsters and may not truly reflect mercury contamination in that part of ENP due to their unnatural diet.

Liver tissue from 23 road- and hunter-killed bobcats was collected from areas similar to the panther collection sites, with the addition of Card Sound Road southeast of ENP (Figure 1). Most of the bobcats were collected in 1984-85. All 20 otters sampled were collected as road-killed animals, and therefore, the location descriptions may vary slightly from above (Figure 1); i.e., NA included animals killed along SR 29 north of I-75; a "far west" (FW) sample included animals from SR 41 near the Collier Seminole State Park on the far western edge of panther range; FS included animals killed on SR 29 south of I-75 as well as those killed on SR 41 from the western Fakahatchee Strand State Preserve boundary east to Turner River Road; the Oasis (OA) Ranger Station sample included road-kills along SR 41 from Monroe Station east to 50-mile-Bend; and the ENP sample was from Chikika State Park on the east side of the ENP/east Everglades. The otters were collected during 1984-1985 and 1989-1991. Muscle samples from road-killed alligators (n=4) were collected in the NA and OA areas in June 1990. Five additional alligators were live-captured in February through April 1991 in SS and surgically biopsied to obtain muscle tissue.

Tissue analysis

All samples were stored at either -10 degrees C or -75 degrees C until submitted for analysis. The primary laboratory used was the Patuxent Analytical Control Facility (US Fish and Wildlife Service) in Laurel, Maryland. Tissue samples homogenates were digested under reflex in sulfuric and nitric acids (Monk 1961). Total mercury(inorganic and organic mercury combined) concentrations were determined by cold vapor atomic absorption spectrophotometry (Hatch and Ott, 1968) using a Spectro Products mercury analyzer equipped with a Varian VGA-76 vapor generation accessory. A limited number of samples were similarly analyzed by Brooks Rand, Inc. in Seattle, Washington and the Florida Game and Fresh Water Fish Commission, Fisheries Research Laboratory in Eustis, Florida. Split samples were used to validate results received from all laboratories. Muscle and liver tissue samples from other carnivores (raccoons, bobcats, otters and alligators) were analyzed at the University of Florida, Gainesville, Florida, using similar techniques.

Mehtylmercury analyses were performed by Brooks Rand, Inc. Tissues were digested in KOH/methanol to release the methylmercury. The methylmercury was then ethylated, subjected to cryogenic gas chromatography, and quantified by a cold vapor atomic fluorescence detector (Bloom 1989).

Data analysis

All tissues results in this paper are reported in ppm of mercury on a wet weight basis. The data for mercury concentrations in tissue appeared to fit a log-normal distribution, so they were log-transformed prior to data analysis. There fore, all reported means are geometric means (GM) rather than arithmetic. When comparisons involved more than two means, one-way analysis of variance (ANOVA) and the Student-Newman-Kuels multiple range test were used to detect differences; otherwise, differences between means were detected using Student's "t" test of the Paired-sign test as appropriate. Linear regression analysis was used to develop predictive models for mercury concentrations in liver and blood. The significance of correlations in regression analyses was calculated using ANOVA.

Mercury levels in alternative blood products (saline washed red cells or whole blood clots) were used in a limited number of cases (n=5) to calculate the whole blood mercury (HgWB CALC) concentration for panthers lacking archived EDTA or heparinized whole blood 6. There was no significant difference (p>0.5) between the means of paired samples of whole blood and calculated whole blood mercury levels based on either red blood cells or clots, therefore, mercury results from whole blood and HgWB CALC were used interchangeably as "blood" in analyses and discussion in this paper. Calculated results from post-mortem serum could not be validated directly with paired blood samples and were only utilized to compare with hair and liver samples in a separate regression.

RESULTS

Mercury levels in panther tissues

All panther tissues examined had detectable levels of mercury, with some considered elevated (Appendix A). Table 1 shows the concentration of mercury in various panther tissues from differing geographical locations. Panthers with the highest mercury levels in all tissues (except kidney) lived in the eastern portion of the panthers' range in and around SS (E. Monroe County and W. Dada County), while panthers with the lowest levels lived in the western portion of the range north of Alligator Alley (NE Collier County and SW Hendry County).

Methylmercury composition

A limited number of samples were submitted to determine the percentage of total mercury that was methylmercury. The percentages of total mercury found to be methylmercury in panther whole blood (n=5), blood clot (n=4) and hair (n=4) were 84.8, 85.8, 99.8, respectively, whereas percentages in liver (n=4) and kidney (n=2) were 9.2 and 23.4, respectively. Methylmercury is the neurotoxic form of mercury, therefore it may be desirable to examine additional tissue samples to more fully understand the partitioning and potential excretion routes of mercury forms within the panther.

Liver levels

The mercury values for liver samples from dead panthers and their approximate geographic locations are presented in Figure 2. There were differences when these data were examined by location and age of the panther. The mean liver mercury level for panthers less than eight years old (young) living in southeastern Florida was significantly higher than similarly aged animals located in southwestern Florida (p=0.024, geometric mean [GM] east=25.8 ppm, GM west=0.304 ppm). If only the western group is considered, animals over eight years of age (old) had significantly higher liver levels than young ones (p=0.029, GM old=14.6 ppm, young=0.304 ppm). The mercury burden was much higher in old animals living in the FS (19-20 ppm) than in the one older panther in NA (7.8 ppm). Only one old panther has been sampled in the eastern portion of the panthers' range, and she had 35.0 ppm in her liver (this value was not included in the above statistical analysis).

Predictive models

We utilized hair to develop a model for predicting mercury liver concentrations in living animals. There was a significant, positive correlation (R=0.89, p<0.001) between mercury in hair and liver collected from dead panthers (Figure 3). This also has been reported for bobcats and raccoons (Cumbie 1975). Using the regression formula y=0.5143 * x^1.0425, it is theoretically possible to predict liver mercury concentration from concentrations in hair collected antemortem. We do not know the lower threshold for mercury toxicosis in free-ranging panthers, but if we use the level documented for domestic cats (i.e. 35 ppm), the model would predict that panthers with hair mercury levels of >57.3 (+/- 95 percent C.I.) are in the toxic range. The more conservative "10 percent rule" level (10 percent of the mean toxic level as suggested by Dr. W. Buck, pers. comm.) would predict that panthers with hair levels of >12.57 ppm are at risk.

Clinical toxicologic data exist for blood from the domestic cat (Charbonneau et al., 1974). From these data, one may predict mercury body burdens and potential toxicosis expected at different blood concentrations. Figure 4 demonstrates a significant, positive correlation (R=0.75, p<0.001) between mercury in blood and hair for the Florida panther. Therefore, given a hair sample, it might be possible to predict clinically relevant blood (and possibly liver) mercury concentrations. the relationship between blood and liver mercury concentration in the panther was established by examining post-mortem blood/hair and blood/liver correlations (Figures 5 and 6), both of which are positive and significant; R=0.98, p<0.001 and R=0.78, p<0.01, respectively.

A significant, positive relationship was observed between the blood mercury level and the fecal concentration (R=0.78, p<0.05) (Figure 7). This comparison was examined to determine if feces could be used as a possible means of monitoring mercury exposure and/or excretion in free-ranging panthers without the necessity of immobilization. It is not know how much of the ingested mercury bound in the hair of the prey animal is absorbed by the panther nor how much is passed in the feces. Likewise, nothing is known about the rate of excretion of previously ingested mercury from contaminated prey. With additional analysis of fecal mercury it may be possible to obtain a gross measure of exposure to mercury.

Figures 8,9, and 10 demonstrate the relationship of mercury in liver compared to three other organ systems; muscle, kidney, and brain, respectively. With these regressions, it is possible (with limited sample type) to predict potential toxic levels for selected panthers.

DISTRIBUTION OF MERCURY IN PANTHERS BY LOCATION

Mean mercury concentrations in blood and hair from living and dead panthers were utilized to assess the geographical distribution of mercury contamination (Figures 11 and 12). The pattern of distribution by location for both hair and blood is similar to that seen in the liver (Figure 2). The highest levels occurred in panthers in SS (GM: hair=55/532 ppm, GM: blood=1.986 ppm). The lowest levels were found in panthers living to the northwest in NA (GM: hair=1.77 ppm, GM: blood=0.089 ppm). Both the hair and blood mercury levels from NA were significantly lower than from SS, LPK, and FS (p<0.01). Further, mercury levels in hair from SS panthers were significantly higher than from all other panthers (p<0.01), whereas blood of SS animals was not different from those in FS, but differed from LPK (p<0.05) panthers. Since there were only two samples from RP, those data could not be included in the analysis of variance, but the mean for both hair and blood for RP samples fell between FS and SS.

Southwestern Florida

Panthers living in the FS between 1985 and 1991 had significantly higher hair and blood mercury concentrations than did the panthers directly to the north across State Road 84 (now Interstate Highway I-75) collected during that same time period (FS GM: hair=7.18 ppm, GM: blood=0.384 ppm; NA GM: hair=1.17 ppm, GM: blood=0.089 ppm) (Fig. 1 and Table 1).

Panther studies in the early-mid 1980's demonstrated that those animals living in the FS were generally underweight and anemic, had poor reproductive success, and consumed numerous raccoons and armadillos (Roelke 1986). At the time, the poor physical condition of the panthers was considered to be primarily a nutritional problem. These data, coupled with field observations indicating a very low deer density, prompted the Game and Fresh Water Fish Commission and Department of Natural Resources to implement management actions in the fall of 1987 to increase density and availability of ungulate prey for panthers in the Fs. These actions included utilization of fire as a habitat management tool, creation of experimental food-plots, salt-licks and feeders for deer, enhancement of law enforcement efforts to curtail illegal killing of deer, and closure of the area to hunting of deer and hogs.

There was a significant drop in mercury levels in the FS panthers coincidental with the above actions. Figure 13 displays the comparison of the mean mercury concentration in panther blood from FS and NA before and after the fall of 1987. There was a significant difference (p<0.001) in blood mercury values between the FS and NA during the early period (FS GM: 0.78 ppm, n=4 individuals/7 samples; NA GM: 0.074 ppm, n=3 individuals/samples), while the values during the later period were not significantly different (FS GM: 0.14 ppm vs. NA GM: 0.12 ppm; n=4 individuals/5 samples and 28 individuals/39 samples respectively). Further, there was no significant difference in blood values between the two time periods within the NA, while a significant difference (p<0.005) did exist within the FS (before=GM: 0/78 ppm, 4 individuals/7 samples; after GM: 0.14 ppm, n=4 individuals/5 samples.)

A comparison of the individual blood and hair mercury values for sampled FS panthers is presented in Figure 14. It should be noted that the range of respective blood and hair values for the 4 panthers sampled in the early period (blood 0.5-1.7 ppm; hair 18-20 ppm) did not overlap any of the values of the 5 panthers sampled in the later period (blood 0.046-0.363; hair 0.6-5.8 ppm). The blood and hair values for one female, FP#09, present throughout the entire period are highlighted in Figure 14. Since 1987, FP#09 has gained 13 percent in body weight (10 lbs) and has experienced a 77.2 percent drop in blood mercury levels (1985=0.630 ppm, 1987=0.598, 1988=0.363 ppm, and 1990=0.140 ppm). Additionally, FP#09's two surviving 1990 offspring sampled in 1991 had considerable lower blood mercury values than did her only other documented surviving offspring born in 1985 (0.140 ppm and 0.170 vs. 0.749 ppm, respectively).

These data suggest that over the past 4 years there has been a marked reduction in the intake of mercury by panthers in the FS. Investigations of the FS deer herd during this same period also suggest that deer numbers increased in the area (McGown 1991). It is possible that the management actions may have improved the prey base and, therefore, had a positive effect on panthers in the FS.

REPRODUCTIVE SUCCESS

Mercury passes through the placenta and is concentrated in the fetus at levels equal to or higher than those of the mother in domestic cats (Khera 1974). Additionally, nervous tissue of the fetus is far more sensitive to the effects of mercury than is the adult nervous tissue. Even low mercury levels in the dam have resulted in profound neuronal disarray in the offspring when the exposure occurred during critical early stages of development (T. Clarkson, U of Rochester, pers. comm.). The disruption of normal fetal development has been documented to cause abortions, stillbirths, congenital anomalies, and behavioral changes resulting in early neonatal death (Khera, 1974). Mercury is also excreted in the milk so that mercury exposure to surviving neonates continues during the suckling period. While it has not been possible to examine any Florida panther neonates to determine if any of these developmental problems exist, it has been determined that, like the domestic cat and other mammals (Khera, 1974), there is a significant positive correlation between the mercury levels in the dams and their surviving dependent kittens (6 months or older) (blood mercury R=0.59, p<0.05) (Figure 16). The regression formula predicts that dams with low blood mercury concentrations will have offspring with low levels (e.g., dam=0.015 ppm - offspring=0.06 ppm) and, conversely, female with elevated mercury levels will have offspring with increased levels (e.g., dam =1.0 ppm - offspring=0.4 ppm).

Our data indicate that mercury contamination, acting independently or in conjunction with poor nutrition, may have affected reproductive success in certain female Florida panthers. Figure 15 displays the average number of surviving young (> 6 mons.) per female-year grouped by the mother's blood mercury level at the approximate time of pregnancy. Females with blood mercury values >0.5 ppm had significantly fewer (p<0.01) surviving offspring (mean=0.167 kittens/female-year) than females with blood mercury values <0.25 ppm (mean=1.46 kittens/female-year).

While generally poor nutrition would otherwise explain lower reproductive success in females, it must be considered that mercury contamination may complicate the effect of poor nutrition on reproduction and, in certain cases, may be a primary factor responsible for lower reproductive success. In the absence of an adequate ungulate prey base such as deer and hogs, the panther will take less desirable prey such as raccoons, which may have lower nutritional value as well as elevated mercury concentrations. Females living in the FSSP had high mercury burdens and poor reproductive performance, also were typically in poor physical condition suggesting nutritional deprivation (Roelke 1986). The affect of mercury contamination and poor nutrition on reproductive success is further confounded in that starvation may mobilize mercury stored in the muscle tissue.

However, 3 females in the ENP and eastern Everglades, which were in otherwise good physical condition, also had high blood mercury values (i.e., #14 = 1.25 ppm, #23 = 0.75 ppm, #27 =1.7 ppm) and experienced reproductive failures and/or death with neonatal dependent kittens. This would suggest that, at least in these cases, sub-clinical mercury toxicosis may have been a factor.


POTENTIAL SOURCES OF MERCURY FOR THE PANTHER

Panthers with the highest levels of tissue mercury generally were those panthers that lived in areas where non-ungulates were frequently consumed; primarily raccoons (Figure 17), armadillos, rabbits, and for two adult male ENP panthers, alligators. Panthers in NA had the lowest levels of tissue mercury and fed primarily on white-tailed deer and feral hogs that occur at relatively high densities on these largely private lands. Although little is known about tissue mercury levels in the hog, mercury concentrations in livers of south Florida deer (n=119) were all less than 1 ppm (Figure 18, data is for dry wt. values) (D. Forrester and S. Sundlof, unpublished data). As mercury is known to bio-accumulate in the food chain (Eisler 1987), the geographic pattern of mercury distribution in the panther led to the hypothesis that the most likely source of contamination for panthers was via the food chain and was most probably associated with more aquatic prey. To test this hypothesis, four other species of carnivores, both prey and non-prey, were examined across the breadth of south Florida to 1) determine which species might be able to deliver mercury to the panthers and 2) determine if there were regional differences that might explain the variation in mercury levels seen among panther sampled in different areas.

Relationship of mercury in raccoon tissue to panther tissue
Mercury in the muscle and liver of raccoons varied significantly between the different watersheds and habitats across the panther's range in south Florida (Figures 19 and 20). The overall pattern of distribution of mercury concentration in raccoon muscle is very similar to that in the blood of panthers living in the same locations (Figures 1, 11, and 13). The highest values for both species occur in SS and adjacent land whereas the lowest levels for both occur in NA. The processes of mercury distribution and concentration in the raccoon as it relates to its food source is unknown at this time.

It appears that panthers on an otherwise good nutritional plane (such as most of those in the ENP and RP) that infrequently consume heavily contaminated prey items such as raccoons, ma accumulate potentially toxic levels of mercury in spite of a predominantly ungulate diet. For example, FP#38, who was in excellent condition, had elevated mercury levels when sampled in the spring of 1990 and, yet has appeared to kill only deer since that time. This female makes periodic forays into 3A where muscle tissue of raccoons has almost three times as much mercury as in RP (Figure 19). It is possible that she accumulates mercury by opportunistically consuming an occasional raccoon or other mercury contaminated prey items in addition to an otherwise deer diet.

Potential toxic effect of consuming raccoons

Considerable information exists in the domestic cat literature regarding toxicity experienced from consuming differing levels of mercury for varying lengths of time. Table 3 presents a summary of dosages and the cumulative time for clinical effects to manifest in domestic cats (Charbonneau et al, 1976, Buck et al., 1987). A tentative model for panther toxicosis was generated by extrapolating from the above literature based on the amount of mercury contained in raccoon muscle from different sample locations (Table 4). This model allows adult female panthers (approx. 75 lb.) to consume approximately 6 lbs. of muscle from the average raccoon per feeding. as a worst case scenario, the model sets the consumption frequency at one raccoon per day. From this rate of mercury consumption the mg of mercury/kg of panther/day was calculated. This "dosage" rate was compared to that reported in various experimental feeding trials using the domestic cat to project time to clinical effect. The model does not consider the mercury contained in the raccoon hair, hide, and organs that might be absorbed along with the flesh.

The locations with the shortest projected time-interval for the occurrence of clinical toxicosis includes the same area where panthers have been identified as having the highest concentrations of mercury, i.e., SS and east Everglades (Table 4) and is the area where FP#27 died of presumptive mercury toxicosis. It is interesting to note that all areas south of Alligator Alley have a projected exposure interval prior to clinical effects of < or = 60 weeks. Whereas the time interval for the area sampled north of the Alley is longer than any of the available domestic cat data can project, > or = 2 years. This modeled distribution of toxic food sources was consistent with the distribution of mercury levels observed in all examined tissues.

Non-raccoon sources of mercury

Another potential food source of mercury for panthers in the ENP is alligator. The adult male FP#16 (ENP) has killed numerous alligators estimated to be 4-5 ft. in length (O. Bass, pers. comm.). In addition, alligator scutes were found in the stomach and intestine of FP#39 at necropsy. Both of these males lived in and/or traversed the SS and had the highest blood mercury values recorded for any living panther (up to 3.4 ppm) and hair mercury values that approached that of FP#27 (up to 100 ppm). Muscle from alligators collected in SS had a mean mercury concentration of GM 2.90. Similar to the raccoon, alligator muscle from SS had higher levels than those in adjacent areas (Table 2, Figure 22). There also appeared to be a linear relationship between the length of the alligator (age) and the amount of mercury in the muscle (Figure 23), but even a small (less than 3 ft.) alligators had elevated levels of muscle mercury (1.66 ppm).

We examined a non-aquatic foraging terrestrial carnivore, the bobcat, to further evaluate the bio-accumulation of mercury within the more aquatic food web. Published reports of bobcat food habits (Maehr and Brady, 1986) indicate that they feed on herbivores, primarily cotton rats and marsh rabbits. It was predicted that because of dietary characteristics the bobcat, like the deer, would have low mercury levels. Of the 23 bobcat livers examined, all except on had liver values less than 1 ppm (Table 2, Figure 24).

The otter, on the other hand, is an obligate aquatic food chain forager, consuming primarily fish and crustaceans. The pattern of regional mercury contamination in the otter is similar to that of the raccoon and panther, supporting the hypothesis that the highest level of environmental mercury contamination is associated with the more aquatic trophic levels in the Everglades system (Table 2, Figure 25).

Other environmental consideration: drought

The drought in the Everglades ecosystem (1989-1991) may also have affected the amount of mercury accumulating in panthers. Comparison of mean hair and blood mercury values for LPK panthers prior to the drought (Dec. 1986-June 1988) vs. the two subsequent years (July 1989-Dec. 1990) show an increase in mercury values in hair (n=5) from 6.353 ppm to 33.10 ppm (n=2) and in blood from 0.184 ppm (n=10) to 0.671 ppm (n=3). However, these differences were not statistically significant due to the high variance associated with the small post drought sample size. Mercury levels in otter liver tissue collected from the eastern edge of the Big Cypress National Preserve in 1989-1991 increased 280 percent over liver collected in 1984-85 (p<0.05) (Figure 25).

DISCUSSION

Chronic exposure to mercury, resulting in mortality and possibly lowered reproductive success, is a likely factor responsible for lower than expected population densities of panther in large portions of their range and may be contributing to the extinction of this endangered mammal. Mercury has been identified in this study as being particularly high in those areas associated with the historic Everglades drainage from Lake Okeechobee; 3A, the SS, and LR, with lower, but yet still significant, levels on adjacent lands.

While the density of large, ungulate prey species (deer and hogs) may be a contributing factor, the presence of mercury-contaminated small prey (raccoons, alligators, and possibly otters) is more likely the primary factor determining the levels of mercury which result in panthers. As prime panther habitat, largely in private ownership, continues to diminish, the ability to maintain optimal densities of uncontaminated prey, to increase the numbers and health of panthers where they occur and to recolonize areas of public land on which panthers are now functionally extinct will have great bearing on our efforts to recover and maintain free-ranging panthers in southern Florida. The significance and potential long term detrimental effects of mercury, no matter the source, in certain portions of the wild panther population should not be minimized. with so few Florida panthers living in southern Florida, every factor that results in the depression of reproductive fitness and/or increased mortality will jeopardize the continued existence of the panther in Florida.

CONCLUSIONS

  1. Mercury contamination of Florida panthers has been documented with the highest levels observed in animals examined in the southern everglades portion of their range and lowest levels observed in animals inhabiting the more upland, pine flatwood and mixed hardwood hammock habitats in the northwest portion of their range.
  2. Mercury contamination likely was responsible for the death of female panther No. 27 in 1989 and may have contributed to the deaths of female panther No. 14 in 1991 and male panther No. 39 in 1990; all of these animals inhabited the southern everglades
  3. Mercury contamination may be contributing to lower reproductive performance of panther in those areas where elevated body burdens of mercury have been documented.
  4. Nercury appears to bio-accumulate through aquatic, carnivorous links in the panther food web; the primary dietary source of mercury for panthers appears to be raccoons.
  5. Mercury contamination in raccoons is documented to occur in a distributional pattern similar to that of Florida panthers with highest body burdens observed from the southern everglades region.
  6. Elevated body burdens of mercury have been documented in Florida panthers whose diets appeared to comprise a higher relative occurrence of raccoons and/or included raccoons from areas likely to be more highly contaminated with mercury.
  7. Chronic exposure of Florida panthers to mercury appears to be compromising their relative health and possibly productivity, especially in the everglades portions of their ranges; mercury must be considered a threat to the continuing existence of this already endangered taxon.
  8. Management practices that encourage greater availability of ungulate prey (i.e., white-tailed deer) may result in reduced dietary exposure to mercury of certain panthers.
  9. Although various management actions may be taken to reduce exposure of certain panthers to mercury (i.e., prey management, land management, removal of individual panthers from contaminated areas) the ultimate need is to understand the environmental processes responsible for the availability of mercury in the areas and aquatic food web affecting panthers.

RECOMMENDATIONS

  1. All Florida panthers recovered at death should be evaluated for mercury levels in all major tissues including liver, kidney, muscle, brain, whole blood, and hair.
  2. All living Florida panthers handled as study animals should be evaluated for mercury in whole blood and hair upon initial capture and re-evaluated upon routine subsequent captures.
  3. Any study animal that is determined to have a relatively elevated body burden of mercury upon initial capture, or subsequent captures, should be monitored intensively and re-evaluated as follows. Panthers with blood mercury values of 0.5 ppm to <= 1.0 ppm should be recaptured and analyzed on a yearly basis, while those with mercury levels 1.0 ppm or greater should be re-evaluate within 4-6 months. Those individuals found with mercury levels of 1.0 ppm or greater should immediately be evaluated by the technical staff of the FPIC to determine any emergency actions which may be appropriate and which may include translocation or removal of that individual from the wild. (Three of 4 panthers that have had blood mercury levels above 1.0 ppm died within 3-17 months of documentation of elevated levels, with mercury likely responsible for or contributing to their deaths.
  4. All study animals that are due for routine re-capture and instrumentation should be captured as early during the winter capture period as possible to allow time for laboratory analysis and re-capture and re-evaluation prior to the end of seasonal capture activities, if appropriate.
  5. Raccoons should be collected on an annual basis to evaluate trends in the availability of mercury to panthers; six each should be collected from the FSSP, Corn Dance Unit/BCNP, and three sites in the ENP (Shark River Slough, East Everglades, and Long Pine Key).
  6. Current and optional management practices which may affect prey resources should be evaluated for all public lands occupied by panthers and appropriate recommendations for action identified annually in revised participation schedules.
  7. The Florida panther habitat preservation plan currently scheduled for completion in Spring 1992 should be aggressively implemented to preserve sufficient habitat, space, and non-contaminated prey resources to assure the continued existence of the panther in south Florida.
  8. Strong support should be given to the Governor's Mercury Subcommittee proposal for studies required to understand the processes responsible for accelerated availability of mercury in Florida aquatic environments and to develop appropriate solutions for reducing levels of mercury exposure in those areas.
  9. The nature and extent of mercury contamination in the ENP should be further investigated and a strategy developed for the safe re-establishment of panthers to that area.
ACKNOWLEDGEMENTS

We wish to thank the veterinary assistants, filed biologists, and technicians who assisted with the panther captures and/or collection and processing of samples: Andrena J. Anderson, Oron L. Bass Jr., Robert C. Belden, Deborah K. Jansen, Nicola Keeling, Mrnie Lamm, E. Darrell Land, David S. Maehr, Roy T. McBride, J. Walter McCown, Steve H. Parker, John Roboski, Jayde C. Roof, Toni K. Ruth, and Jeff Wentworth. We wish to thank Ted R. Lange of the FGFWFC fisheries Laboratory in Eustis, Florida for conducting some preliminary analyses for this study. Special thanks go to William Buck, DVM University of Illinois, Champaign, Illinois and Thomas Clarkson, Ph. D., University of Rochester, Rochester, New York, for their contributions; and to Joan Forrester Collier and Terri Steele for their perseverance in typing the numerous drafts and the final document. We also thank Jim Brady, Brad Gruver, and Tom Logan for editorial review. This project was funded in part by federal grant-in-aid funds administered through the U.S. Fish and Wildlife Service under Section 6 of the Endangered Species Act of 1973 (PL No. 93-205), the Florida Panther Research and Management Trust Fund, and the Nongame Wildlife Trust Fund.

LITERATURE CITED

  1. Albanus, L., L. Frankenburg, C. Grant, V. von Haartman, A. Jernelov, G. Nordberg, M. Rydalv, A. schutz, and S. Skersving. 1972. Toxicity for cats for methylmercury in contaminated fish from Swedish lakes and of methylmercury hydroxide added to fish. Environmental Research 5 p:425-442.
  2. Bigler, W.J., J.H. Jenkins, P.M. Cumbie, G.L. Hoff, and E.C. Prather. 1975. Wildlife and environmental health: Raccoons as indicators of zoonoses and pollutants in the southern United States. J. Amer. Vet. Med. Assoc 167(1):592-597.
  3. Bloom, N.S. 1989. Determination of picogram levels of methylmercury by aqueous phase ethylation, followed by cyrogenic gas chromatography with cold vapor atomic fluorescence detection. Can. 3. Fish Aq. Sci. Vol. 46(7):1131-1140.
  4. Charbonneau, S.M., I.C. Munro, E.A. Nera, R.F. Willes, T. Kuiper-Goodman, F. Iverson, C.A. Moodie, D.R. Stoltz, F.A.J. Armstrong, J.R. Uthe, and H.C. Grice, 1974. Subacute toxicity of methylmercury in the adult cat. Toxic Appl. Pharm. 27:569-581
  5. Clarkson, T.W., R. Hamada and L. Amin-Zaki. 1984. Mercury. Pp. 285-309 in J. O. Nriagu (ed.). Changing metal cycles and human health. Springer - Verlag, Berlin.
  6. Clarkson, T.W. and D.O. March. 1982. Mercury toxicity in man. Pages 549-568 in A.S. Prasad (ed.). Clinical, biochemical, and nutrititional apsects of trace elements, Vol. 6. Alan R. Liss, Inc., New York.
  7. Cumbie, P.M. 1975. Mercury in hair of bobcats and raccoons. J. Wildl. Manage. 39(2):419-425.
  8. Delany, M.R., J.U. Bell, and S.F. Sundlof. 1988. Concentrations of contaminants in muscle of American alligator in Florida. J. Wildl. Dis. 24:62-66.
  9. Eisler, R. 1987. Mercury hazards to fish, wildlife, and invertebrates: a synoptic review. U.S. Fish Wild. Serv. Biol. Rep. 85(1.10). 90 pp.
  10. Hatch, W.R., and W.L. Ott, 1968. Determination of submicrogram quantities of mercury by atomic absorption spectrophotometry. Anal. Chem. 40:2085-2087.
  11. Harada, M. and A.M. Smith. 1974. Minamata disease: A medical report. In Mercury pp. 180-192.
  12. Hord, L.J., M.J. Jennings, A. Brunell. 1990. Mercury Contamination of Florida alligators. 229-240 pp. In Crocodiles. Proceedings of the 10th Working Meeting of the Crocodile Specialist Group, IUCN--The World Conservation Union, Gland, Switzerland. Volume 1. ISBN 2-8327-0022-1.xvi + 354 p.
  13. Houpt, K.A., L.A. Essick, and E.B. Shaw. 1988. A tuna fish diet influences cat behavior. J. Toxicology and Environmental Health, 24:161-172.
  14. Khera, K.S. 1974. Teratogenic effects of methylmercury in the cat: Note on the use of this species as a model for teratogeneticity studies. Teratology 8:293-304.
  15. Lange, T., H. Royals and L. Connor. 1991. Preliminary results examining the relationship between mercury in largemouth bass and physical and chemical lake characteristics. (unpubl. MS).
  16. Loftus, W. 1990. IN Proceedings of workshop on mercury contamination in Florida: Impacts and soutions. Center for biomedical and toxicological research and waste management. Florida State University, Tallahassee, Fl.
  17. Maehr, D.S., and J.R. Brady. 1986. Food habits of bobcats in Florida. J. Mamm. 67(1):133-138.
  18. Maehr, D.S., R.C. Belden, E.D. Land, and L. Wilkins. 1990. Food habits of panthers in southwestern Florida. J. Wildl. Manage. 54(3):420-423.
  19. McCown, J.W. 1991. Big cypress deer/panther relationships: deer herd health and reproduction. Final performanced report, endangered species project E-1 II-E-5a. Florida Game and Fresh Water Fish Commission. 85 pp.
  20. Mierle, G. 1990. Aqueous inputs of mercury to precambrian shield lakes in. Ontario. Environ. Toxicol. Chem. 9:843-851
  21. Monk, H.E. 1961. Recommended methods of analysis of pesticide residues in food stuffs. Report by the Joint Mercury Residue Panel. Anal. Chem. 86:608-614.
  22. NAS. 1978. An assessment of mercury in the environment. Natl. Acad. Sci., Washington, DC. 185 pp.
  23. Ogden, J.C., W.B. Robertson, G.E. Davis, and T.W. Schmidt. 1973. Polychlorinated biphenyls and heavy metals in upper food chain levels, Everglades National Park and vicinity. South Florida Environmental Rept. No. DI-SFEP-74-16.
  24. Roelke, M.E. 1990. Florida Panther Biomedical Investigation: health and reproduction. Final performance report, Endangered Species Project E-I II-E-6 7506. Florida Game and Fresh Water Fish Commission. 176 pp.
  25. Roelke, M.E., E.R. Jacobson, G.V. Kollias, and d.J. Forrester. 1986. Medical management and biomedical findings on the Florida panther (Felis concolor coryi). Annual performance report, Endangered Species Project E-1-10 7506. Florida Game and Fresh Water Fish Commission. 65 pp.
  26. Schacklette, H.T. 1980. Elements in fruits and vegetables from areas of commercial production in the conterminous United States. U.S. Geol. Surv. Prof. Paper 1178.
  27. Simons, J.N. 1991. Mercury in the Everglades: What is the role of agriculture? Florida Nat. Spring 1991:7-9.
  28. SFWMD. 1990. Surface water improvement and management paln for the Everglades. South Florida Water Management District.
  29. Takeuchi, T., F.M. D'Itri, P.V. Fischer, C.S. Annett, and M. Okabe. 1977. The outbreak of Minamata disease (methyl mercury poisoning) in cats on northwestern Ontario reserves. Environ. Res. 13:215-228.
  30. United States Fish and Wildlife Service. 1989. Status report on mercury contamination in Florida panthers. Prepared for the Florida Panther Interagency Committee. 44pp.
  31. Winfrey, M.R. and J.W.M. Rudd. 1990. Environmental factors affecting the formation of methylmercury in low pH lakes. Environ. Toxicol. Chem. 9:853-869.