Davis Pond freshwater prediversion biomonitoring study: freshwater fisheries and eagles

In January 2001, the construction of the Davis Pond freshwater diversion structure was completed by the U.S. Army Corps of Engineers. The diversion of freshwater from the Mississippi River is intended to mitigate saltwater intrusion from the Gulf of Mexico and to lessen the concomitant loss of wetland areas. In addition to the freshwater inflow, Barataria Bay basin would receive nutrients, increased flows of sediments, and water-borne and sediment-bound compounds. The purpose of this biomonitoring study was, therefore, to serve as a baseline for prediversion concentrations of selected contaminants in bald eagle (Haliaeetus leucocephalus) nestlings (hereafter referred to as eaglets), representative freshwater fish, and bivalves. Samples were collected from January through June 2001. Two similarly designed postdiversion studies, as described in the biological monitoring program, are planned.

Active bald eagle nests targeted for sampling eaglet blood (n = 6) were generally located southwest and south of the diversion structure. The designated sites for aquatic animal sampling were at Lake Salvador, at Lake Cataouatche, at Bayou Couba, and along the Mississippi River. Aquatic animals representative of eagle prey were collected. Fish were from three different trophic levels and have varying feeding strategies and life histories. These included herbivorous striped mullet (Mugil cephalus), omnivorous blue catfish (Ictalurus furcatus), and carnivorous largemouth bass (Micropterus salmoides). Three individuals per species were collected at each of the four sampling sites. Freshwater Atlantic rangia clams (Rangia cuneata) were collected at the downstream marsh sites, and zebra mussels (Dreissena spp.) were collected on the Mississippi River.

The U.S. Geological Survey (USGS) Biomonitoring of Environmental Status and Trends (BEST) protocols served as guides for fish sampling and health assessments. Fish are useful for monitoring aquatic ecosystems because they accumulate pesticides and other contaminants. Biomarker data on individual fish, generated at the USGS National Wetlands Research Center (Lafayette, La.), included percent white blood cells in whole blood, spleen weight to body weight ratio, liver weight to body weight ratio, condition factor, splenic macrophage aggregates, and liver microsomal 7-ethoxyresorufin-o-deethylase (EROD) activity. Fish age was estimated by comparing total lengths with values from the same species in the Southeast United States as determined from the literature. Contaminant analyses were coordinated by the U.S. Fish and Wildlife Service (USFWS) Analytical Control Facility (Laurel, Md.), where residues of organochlorine (OC) pesticides, total polychlorinated biphenyls (PCBs), polycyclic aromatic hydrocarbons (PAHs), aliphatic hydrocarbons (AHs), and trace elements were determined. The organic contaminant data were generated at the Mississippi State University Chemical Lab (Mississippi State, Miss.), and the inorganic contaminant data were generated by the Texas A&M University Geochemical and Environmental Research Group (College Station, Tex.). Statistical tests were performed to assess relationships among contaminants, fish age, fish species, and collection sites.

Trends in interspecific differences among fish in concentrations of contaminants were noted. Striped mullet (hereafter mullet) frequently displayed the highest chemical concentrations. Levels of contaminants were generally higher in samples obtained from the Mississippi River than in those collected from the diversion area and were higher in mussels and clams (hereafter bivalves) than in fish. Because the Mississippi River sampling site for mullet and largemouth bass was downriver of the structure and south of New Orleans and the catfish site was upriver, the downriver data may not be directly reflective of the results from the receiving waters at the Davis Pond structure. Compared to the Caernarvon freshwater prediversion study in 1990 that assessed possible influx of contaminants with the freshwater diversion, contaminant levels in fishes and bivalves in this study were generally lower, yet three nontoxic inorganic elements in Davis Pond fish samples exhibited ranges of concentrations that were more than two times higher than did those from Caernarvon. Levels in bivalves were different between diversions but about equal in the numbers of trace elements showing high levels per location. Contaminant values were compared to those listed in various literature and agency sources, both regional and national, including the National Contaminant Biomonitoring Program (NCBP), in which the 85th percentile and above represents what is considered to be an elevated contaminant concentration and cause for concern. 

Generally, bivalves were at the high end of their ranges for both organic and inorganic contaminants. In this study, OCs were detectable in 67 percent of fish from the Mississippi River site, ranging from 0.15 to 1.09 μg/g wet weight (ww) or fresh weight (fw), and in 11 percent of the fish from the marsh sites, ranging from 0.06 to 0.612 μg/g ww. Bivalves from the Mississippi River had OC levels of 0.096 μg/g ww, whereas none were detectable in bivalves at the marsh sites. In this study, p,p’-dichlorodiphenyldichloroethylene (p,p’-DDE) (a biodegradation product of DDT [dichlorodiphenyl trichloroethane]) and total PCBs were the most frequently detected OCs and were primarily from the Mississippi River. For total OC content, using adjusted least squares means, some significant interactions were noted between fish species and sites. PAHs were detected in aquatic animals at all sites (range of 0.017–17.534 μg/g ww), as were AHs (range of 0.423–4.549 μg/g ww); the highest levels of PAHs and AHs were found in bivalves from the Mississippi River. When analysis of variance (α = 0.05) was performed with data from aquatic animals, there were only two significant relationships between PAHs, AHs, and OCs between species, site, and age or the interaction among these variables. There was an interaction between fish species and n-decane (an AH) in that mullet and largemouth bass had significantly higher levels than did catfish (P= 0.0175).

When general linear means were used to investigate associations of inorganic contaminants among fish species, site, and age or any interactions among these variables, no significant results were noted for arsenic, cadmium, lead, beryllium, boron, molybdenum, or nickel. The range of mercury in fish in this study was 0.04–0.14 μg/g ww (0.14– 0.48 μg/g dry weight [dw]), with the most elevated levels detected in predatory largemouth bass at the sampling point farthest downstream from the structure and within the marsh area. Mercury was positively correlated with fish age (P= 0.0152), where levels were estimated to increase 0.0253 parts per million (ppm) dw per year. In the Mississippi River, catfish showed significantly higher levels of mercury than did mullet or largemouth bass (P= 0.00167).

Among fish species, mullet displayed the highest levels in fish of aluminum, barium, manganese, and iron, all considered to have low toxicity in hydrologic systems. An interaction between fish and site was seen with aluminum (P= 0.0031), where concentrations in mullet were significantly higher in the Mississippi River than at the other sites, as was also seen with barium (P= 0.0009), chromium (P= <0.0001), manganese (P= 0.0004), strontium (P= 0.0074), vanadium (P= 0.0156), and zinc (P= 0.0059). For iron (P= 0.0.0001), mullet and largemouth bass at both the Mississippi River and Lake Salvador showed higher levels than did catfish, and these two species showed higher levels at two of the four sites. An interaction between fish and site was also seen with chromium (P= <0.0001) in that concentrations in mullet were significantly higher in the Mississippi River than at the other sites, as was also seen with strontium (P= 0.0074), vanadium (P= 0.0156), and zinc (P= 0.0059), metals for which deleterious effects have been demonstrated in other ecosystems. The NCBP program lists the 85th percentile for zinc at 34.2 μg/g fw (117.9 μg/g dw). In the Davis Pond prediversion biomonitoring study (hereafter the current study), one fish (MUL31RIVER, fish ID 8) showed values higher than that (125.4 μg/g dw or 37.54 μg/g ww), and the Mississippi River bivalve sample (MUSSRIVER) had a value of 140 μg/g dw (41.2 μg/g ww).

In the current study, approximately 86 percent of the fish had measurable selenium levels, yet none reached the 85th percentile. The 85th percentile for selenium from the NCBP was 0.73 μg/g ww. Significantly higher levels of selenium were seen in mullet than in largemouth bass and catfish (P= 0.0023). The NCBP 85th percentile for lead is 0.22 μg/g ww (0.76 μg/g dw). In the current study, the range of concentrations of lead was as much as 18.3 ppm dw (MUL31RIVER, fish ID 8), with the three most elevated values (range of 3.46–5.31 μg/g ww) coming from mullet from the Mississippi River.

Biomarker data are measurable and directly reflect the condition of the animal, and measuring more than one biomarker in an individual increases confidence in health assessments. In the current study, biomarkers included macrophage aggregates (MAs), liver (hepatosomatic index [HSI]) and spleen (splenosomatic index [SSI]) weight to body weight ratios, percent white blood cells (WBCs) in whole blood, and condition factor. Few significant differences were noted with any of the biomarkers between sites, and there were no relationships between species and sites. For improved use of biomarker assessments, an increase in fish sample size would be useful for postdiversion sampling, as would comparisons of fish of the same sex and reproductive condition.

During the current study, success for eagle nests in the diversion area and reference sites was similar as determined by numbers of nestlings fledged. When temperatures were below average during winter 2000, nests in both regions similarly failed. At each nest, the primary evidence of food items was small mammals. Eaglets (n = 6) generally appeared healthy, and whole blood concentrations of organic contaminants exceeded detection limits with three incidences of p,p’-DDE (0.002–0.006 μg/L ww) and one incidence of oxychlordane (0.002 μg/L ww). The levels of p,p’-DDE were well below those that have been inversely correlated with productivity and success rates of nesting bald eagles on a regional scale. The low values found in the whole blood samples for OC pesticides and PCBs were even lower when corrected for plasma volume. Aluminum values were 3.66 and 5.75 μg/L in two samples, zinc ranged from 5.21 to 6.77 μg/L ww in six samples, and silicon ranged from 1.7 to 4.6 μg/L in four samples. Selenium was detectable in each bird with the range at 0.332–0.566 μg/L ww, and strontium ranged from 0.0581 to 0.0975 μg/L ww. Mercury was detectable in blood samples from each bird and ranged from 0.0254 to 0.0845 μg/L ww, whereas lead was detectable in four samples and ranged from 0.0042 to 0.0136 μg/L ww. Although no detectable levels of total PCBs were found (also correlated with decreased reproductive productivity), 70 percent of the aquatic animals from the Mississippi River contained total PCBs (range 0.13–0.79 μg/L), whereas only about 7 percent of the aquatic animals sampled from the marsh area contained PCBs.

Suggestions for postdiversion sampling include lowering the analytical detection limit for some metals, sampling aquatic animals over the course of a single season, obtaining a higher sample number of mature fish of one species (for example, blue catfish) within a range of total lengths for biomarker analyses, obtaining otoliths for estimating fish ages, assessing dioxins in eaglet blood, examining triazines in water, and obtaining all Mississippi River fish samples as close to the Davis Pond structure intake as possible. Because contaminants found in blood of eaglets reflect their prey species and because of the contaminant levels found in fish in the current study, eaglets may not be consuming primarily these species; therefore, obtaining juvenile nutria (Myocastor coypus) or turtle species for contaminant analyses might be considered, as well as collecting greater blood volume and using plasma to measure OCs and PCBs. Data obtained postdiversion will be compared with prediversion data to monitor changes.