Lead Poisoning in Bald Eagles Admitted to Wildlife Rehabilitation Facilities in Iowa, 2004–2014

Lead poisoning is a serious wildlife health issue causing adverse effects for upland birds, water birds, raptors, scavengers, and waterfowl. More than 120 species of birds, including bald eagles Haliaeetus leucocephalus, have been documented with lead poisoning (Pain et al. 2009; Tranel and Kimmel 2009). The prevalence of lead poisoning in bald eagles has been researched using various sampling methods. Elevated lead levels have been found in bald eagles admitted to wildlife rehabilitation facilities (Jacobson et al. 1977; Redig 1979; Kramer and Redig 1997; Neumann 2009; Stauber et al. 2010; Cruz-Martinez et al. 2012); in blood samples from wild-trapped, free-flying bald eagles (Harmata 2011; Bedrosian et al. 2012); and in bald eagle tissue samples from carcasses found in the field (Russell and Franson 2014; Warner et al. 2014). Lead affects multiple tissues, and clinical signs of lead toxicosis in raptors include ataxia, lime-green feces, vomiting, seizures, paresis or paralysis of the wings and legs, amaurosis (impaired vision), and death (Gilsleider and Oehme 1982; Samour and Naldo 2002). Clinical lead levels alone could result in death (Golden et al. 2016). Fatal lead toxicosis exposures are common, but subclinical toxicosis is also important as these birds may be more susceptible to mortality or morbidity due to decreased performance levels (Redig 1979).

Lead exposure in wildlife can come from direct and indirect sources. Waterfowl and upland birds directly consume lead shot or fishing weights, mistaking them for seeds or grit (Anderson et al. 2000; Franson et al. 2009). Raptors or scavengers inadvertently consume lead from animal carcasses or gut piles containing lead shot or lead shrapnel. Previous studies have demonstrated the presence of lead in animal carcasses and in gut piles resulting from hunting activities (Hunt et al. 2006; Pauli and Buskirk 2007; Neumann 2009; Grund et al. 2010; Warner et al. 2014). Lead ammunition embedded in these food sources has been found to be a major source of lead exposure for bald eagles (Nelson et al. 1989; Elliott et al. 1992; Tranel and Kimmel 2009; Harmata 2011; Cruz-Martinez et al. 2012; Golden et al. 2016).

Iowa's Midwinter Bald Eagle Survey and Audubon's Christmas Bird Count show that Iowa is a significant wintering area for bald eagles, with thousands of bald eagles arriving in November and staying until March (IDNR 2014; National Audubon Society 2010). Hunting and trapping seasons for game species coincide with this influx of bald eagles to the state. A variety of harvest methods and ammunition types are allowed in Iowa during these seasons, including lead muzzleloader bullets, rifle bullets, shotgun slugs, and shot. Nonlead and nontoxic ammunition is only required for waterfowl hunting and on some specific public and private hunting areas in Iowa (IDNR 2016–2017). Shotgun slugs are used to harvest the majority of white-tailed deer Odocoileus virginianus in Iowa (IDNR 2014). This makes the analysis of lead poisoning in Iowa bald eagles an important addition to the data gathered from other states where rifle ammunition is used most commonly to harvest big game.

This retrospective study analyzes 11 y (2004–2014) of bald eagle data from four independent, state and federally permitted wildlife rehabilitation centers (hereafter rehabilitation centers) in Iowa. These four rehabilitation facilities admit bald eagles from across the state of Iowa; Black Hawk Wildlife Rehabilitation Project in Black Hawk County, MacBride Raptor Project now The Raptor Advocacy Research and Education Group in Johnson County, Saving Our Avian Resources in Carroll County, and Wildlife Care Clinic in Story County. Our objectives were to 1) determine the prevalence of subclinical and clinical lead levels in bald eagles admitted to wildlife rehabilitators in Iowa, 2) determine the relationship between lead levels and recorded information (age, season, and outcome), and 3) examine necropsy results to determine potential sources of lead exposure.

From January 1, 2004, to December 31, 2014, four rehabilitation centers in Iowa admitted 322 bald eagles with a variety of different health issues. Most entry data points included the year, month, and county of collection (Data S1, Supplemental Material). The centers used the Clinical Wildlife Health Initiative (CWHI 2010) criteria as presented in the Wildlife Incident Log/Database and Online Network medical record to categorize bald eagle data (WILD-ONe, Wildlife Center of Virginian, Waynesboro, VA). The four centers screened 273 individuals for lead toxicity by using antemortem blood collection (n = 209), postmortem liver tissue samples (n = 59), or bone tissue samples (n = 5). Rehabilitators obtained liver or bone samples from bald eagles that died before being able to collect a blood sample. This study includes only one sample per bald eagle (i.e., no animals had multiple samples submitted for analysis). Age classification of bald eagles based on plumage characteristics includes; immature (<1 y old), juvenile (1–5 y old), or adult (>5 y old).

Rehabilitators collected blood samples from 209 bald eagles by venipuncture of the basilar vein on the wing following manual restraint. Collecting centers either processed blood samples immediately using the portable ESA LeadCare System (Magellan Diagnostics, North Billerica, MA) or placed blood samples in heparinized microtainers and sent them on ice to Iowa State University Diagnostic Laboratory (Ames, IA) or Antech (Irvine, CA) for laboratory analysis. The LeadCare blood testing system is a rapid, on-site, portable electrochemical assay for the quantitative measurement of lead levels in blood samples. The technology used by the LeadCare blood testing system is derived from that used in the ESA anodic stripping voltammetry method to measure blood lead concentrations. At minimum, 50 μL of blood is placed directly into heparin coated microhematocrit tubes. The blood is then mixed with 250 μL of a dilute hydrochloric acid solution in water (0.34 M). This solution is analyzed for final blood lead levels by using the LeadCare detecting system. The lower and upper limits of detection for this method are 0.033 and 0.65 parts per million (ppm), respectively. Samples that were sent to Iowa State University Diagnostic Laboratory and Antech were analyzed within 24–48 h of sample collection. Antech uses the LeadCare blood testing system for analysis of lead levels. Atomic absorption spectrometry using a graphite furnace with Zeeman correction is used by Iowa State University Diagnostic Laboratory. For this process, blood is diluted with Triton X-100 and quantitated by inserting values on a standard curve. The lower and upper limits of detection for this method of analysis are 0.01 and 1 ppm, respectively. Additional dilutions were performed to extend the upper range of data.

Collected liver (n = 59) and bone (n = 5) tissue were first placed in a freezer at −17.8°C (0°F) for no more than 6 mo. Samples were then sent to Iowa State University Diagnostic Laboratory on ice where the samples were thawed. Lead extraction from liver occurred by placing 1 g of homogenized sample in nitric acid. Bone samples were ashed (measurements in dry weight). Flame atomic absorption spectrometer using a graphite furnace with Zeeman correction then determined lead levels in both the liver and bone samples. The lower and upper limits of detection for this method of analysis are 0.1 and 10 ppm, respectively. Dilutions were performed as necessary for samples outside of the upper range.

We used Tobit regression to predict log-scale blood lead concentrations from associations with age (immature, juvenile, or adult), outcome (died, euthanized, or released), season, year, and radiographic findings. We defined seasons as spring (February–May), summer (June–September), and autumn (October–January). Before beginning Tobit analyses, we identified outliers and rejected them via Horn's algorithm and Tukey's interquartile fences (Geffory et al. 2009; Friedrichs et al. 2012). Tobit regression accounts for left-censoring of observations below detection limit as well as for right-censoring of those observations above 0.65 ppm obtained from the LeadCare detecting system. We chose a lognormal distribution because conditional plots of blood lead concentrations suggested a right-skewed distribution. In considering how to model interactions among age, outcome, and season, we fit the model separately with all possible interaction combinations and then selected the model with the lowest Akaike's Information Criterion (AIC) as the most informative for this analysis. We assessed maximum likelihood estimates by using the SAS (SAS Institute Inc., Cary, NC) procedure PROC LIFEREG.

We built a logistic regression model to assess log-scale blood lead concentration as a predictor for the probability of a bald eagle having a negative rehabilitative outcome (dying or being euthanized) by using the SAS procedure PROC LOGISTIC. We treated observations below the 0.01 ppm detection limit as 0.01 ppm, and we treated LeadCare observations of >0.65 ppm as 0.65 ppm; however, because right censoring to 0.65 ppm understated the true blood concentration, we added an indicator predictor variable that was equal to 1 if the observation was right censored and 0 otherwise. Other predictors included bald eagle age, season, year, and radiographic findings. Due to imbalances in the dataset, most interaction effects among predictors lead to data separation; thus, we could only consider the interaction between bald eagle age and log-scale blood lead concentration, and we did not use AIC model selection.

Veterinarians performed radiographs on 169 bald eagles and assessed them for the presence of metallic opacities. Bald eagles admitted dead on arrival, humanely euthanized due to injury severity, or those that died during rehabilitation efforts were necropsied to determine the source of the metallic opacities. During necropsy, rehabilitators and veterinarians identified metallic opacities by recognizable shape, malleability, and color. We used a commercially available bullet and shot sizing template to determine the shot size and bullet caliber of ammunition found during necropsy. Shot and shrapnel were tested with a magnet to establish whether the items were potentially lead (nonmagnetic) or steel (magnetic). Eagle radiographs with metallic opacities were compared to radiographs of game animals harvested with a known ammunition type to assess objects' similarity in shape and opacity.

Ninety-five (45%) of 209 bald eagles had subclinical-to-clinical levels of lead present in the bloodstream at time of collection. Two hundred and six (99%) bald eagles had nonzero values for lead in their blood, and only 3 (1%) had blood lead levels below detectable limits. Thirty-six (61%) of 59 submitted liver samples had subclinical to clinical levels of lead present. Five (100%) of five bone samples submitted for analysis contained subclinical to clinical lead levels. All 64 tissue samples had some degree of lead present (Table 1).

Of the 273 bald eagles tested for lead, 136 (50%) had clinical (n = 99) or subclinical (n = 37) lead levels and 137 (50%) had background or zero lead levels. The majority 115 (85%) of the lead-affected group had no other injuries, only exhibiting symptoms of lead exposure. Twenty-one (15%) of the lead-affected group sustained other injuries. Only 15 (11%) of the lead-affect group had a positive rehabilitative outcome and were released. Comparatively, 55 (40%) of the nonlead-affected group were released (Table 2).

Bald eagles categorized with subclinical or clinical lead levels account for half of the lead-tested bald eagles seen at wildlife rehabilitation centers in Iowa and were admitted from across the state (Figure 1). Of the 99 counties in Iowa, 62 counties had at least one bald eagle found with subclinical or clinical lead levels. Four counties had more than five bald eagles with lead exposure: Carroll (six), Dubuque (seven), Linn (six), and Marion (seven).

Across models predicting log-scale blood lead concentrations, the Tobit regression model with the lowest AIC (by 2.98 vs. the next model) identified a season–outcome interaction but no interactions with age. Thus, the final model featured fixed effects for age, season, outcome, season–outcome, year, and radiographic findings. Adult bald eagles had higher blood lead concentrations than immature and juvenile birds; bald eagles with positive radiographs for metallic opacities in the digestive tract had higher blood lead concentrations than those with negative results or those for which no radiograph was collected. Although season and outcome findings must be interpreted in the context of the season–outcome interaction, in general, blood lead concentrations decreased across outcomes as ordered: died – euthanized – released, whereas seasonal concentrations decreased from autumn – spring – summer (Table 3).

In the logistic regression model, log-scale blood lead concentrations and the indicator variable for right censoring of log blood lead concentrations were both significant predictors (α = 0.05) of negative outcomes (died or euthanized); no other explanatory predictors in the model were significant (Table 4). The estimated increase in the log odds of a negative rehabilitative outcome associated with a 1-unit rise in log-scale blood lead concentration was 0.44 with standard error 0.15 when averaged across all age groups. The odds ratio of a negative rehabilitative outcome comparing right-censored observations to noncensored observations with blood lead concentrations of 0.65 ppm was estimated at 4.50 with 95% Wald confidence limits from 1.33 to 15.2. The significance of the right-censoring indicator term reflects the dual facts that higher lead concentrations were associated with negative rehabilitative outcome, and censored samples had higher lead concentrations than reflected by the 0.65 ppm assigned to them in the analysis. Forty-three of 181 LeadCare samples were at or above this level.

One hundred sixty-nine bald eagles received whole body radiographs that were interpreted by rehabilitators and veterinarians. One hundred twenty-eight (76%) of the radiographs were negative for metallic opacities. Of these 128 bald eagles, 118 were tested for lead levels. Twenty-two (13%) bald eagles had radiographs with metallic opacities outside of the digestive tract (consistent with ammunition shrapnel or shot, indicative of gunshot). Of these 22 bald eagles, 14 were tested for lead levels. Nineteen (11%) bald eagles had radiographs with metallic opacities in the digestive tract. Of these 19 bald eagles, 16 were tested for lead levels (Table 5).

Two of the 14 bald eagles that were gunshot and tested for lead levels were released: one with subclinical lead levels and one with background lead levels. Two of the 16 bald eagles with metallic opacities in their digestive tract and tested for lead levels were released. One of these released birds showed shrapnel in its digestive tract in the radiograph and had subclinical lead levels. The other bald eagle released, egested a deer hair pellet on admission that was the subject of examination to confirm shrapnel found on radiograph and had background lead levels.

Necropsies were performed on 12 of the 22 bald eagles with radiographic evidence consistent with gunshot wounds, finding five (42%) with ammunition shrapnel (indicative of being shot with a lead rifle bullet), five (42%) with lead shot, and two (16%) with steel shot. Necropsies were performed on 17 of 19 bald eagles that had radiographic evidence of metallic opacities in the digestive tract. Necropsies found 11 (65%) with ammunition shrapnel, 5 (29%) with lead shot, and 1 (6%) with an expanded copper-jacketed lead-core 0.22 mag bullet in the digestive tracts (Figures 2–4).

The data gathered by wildlife rehabilitators in Iowa on subclinical and clinical lead exposure in bald eagles are consistent with datasets from many other rehabilitation centers. The debilitating effects of lead exposure have become a significant cause for admission to rehabilitation centers for bald eagles across the United States (Harris and Sleeman 2007; Stauber et al. 2010; Cruz-Martinez et al. 2012). Assessing this information is useful in attempting to monitor, limit, or prevent risk factors in the conservation efforts of bald eagles. Analysis of this bald eagle morbidity and mortality dataset adds valuable information on lead poisoning in Iowa, an area where bald eagles seasonally congregate (IDNR 2014). It also highlights several clinically important aspects that are beneficial for rehabilitators handling bald eagles with potential lead exposure.

Radiographic evidence is key in determining the cause of lead exposure in bald eagles. It is important to note that 51% of the bald eagles that were negative for metallic opacities in the digestive tract on radiographs still had subclinical to clinical lead levels. One possible explanation for this is these bald eagles had already digested or passed the lead source through the gastrointestinal tract or the metallic pieces were ejected in a pellet. Only 11% of bald eagles had radiographs positive for metallic opacities in the digestive tract. This group had higher blood lead concentrations than those with negative results and 94% had subclinical-to-clinical lead levels. Other studies have found a similar proportion of bald eagles with radiographic evidence of metallic opacities in their digestive tract and reported that this correlates with elevated blood or tissue lead levels (Kramer and Redig 1997; Cruz-Martinez et al. 2012; Franson and Russell 2014). Necropsy is needed to identify metallic opacities. Shot and bullet findings indicate an ammunition source. Shrapnel findings could be further tested to confirm the implication of bullet or slug fragments.

Lead retained embedded in tissue does not, in most cases, cause lead poisoning (DeMartini et al. 2001). Lead fragments in soft tissue often become encapsulated by fibrous tissue and are effectively inert. However, in rare situations, lead associated with synovial membranes and other blood vessel dense tissue can result in elevated lead levels (DeMartini et al. 2001). Franson and Russell (2014) suggest that elevated liver lead levels in gunshot cases may reflect contamination of the liver tissue from ammunition fragments from the gunshot itself vs. the bald eagle ingesting lead. The bald eagles in this study with shrapnel or shot present outside the digestive tract (gunshot) had the same percentage of subclinical-to-clinical lead levels in liver and blood samples as the entire dataset (50%). These results and sampling methods measuring lead from blood and liver sources suggest that bald eagles in this study with gunshot wounds likely had elevated lead levels due to ingested lead.

Adult bald eagles had higher lead concentrations than immature or juvenile bald eagles, consistent with data on rehabilitating bald eagles (Cruz-Martinez et al. 2012) and data from bald eagles found dead (Franson and Russell 2014). Lead accumulates in many body tissues, with deposition in bone being a site of chronic storage (Franson et al. 2009). This lead accumulation could possibly explain the higher levels of lead found in adult birds. Because lead mimics calcium in body systems, growth and embryo and egg formation are causes for an increase in calcium demand and a possible increase in available lead uptake. This would suggest that growing nestlings and prenesting adult female bald eagles would be the most lead-susceptible groups. Understanding which age group and sex of bald eagles is most susceptible to lead accumulation and how this might impact productivity would be important topics for future research.

Lead concentrations were highest across all age classes during the autumn (October–January), which coincides with the hunting and trapping seasons in Iowa. This finding is consistent with datasets from other rehabilitation centers in various geographical areas, samples from free-flying bald eagles, and assessments of bald eagles found dead (Kramer and Redig 1997; Harris and Sleeman 2007; Neumann 2009; Redig et al. 2009; Stauber et al. 2010; Bedrosian et al. 2012; Cruz-Martinez et al. 2012; Warner et al. 2014). Other research has shown increased availability of carcasses and gut piles within this season, supporting ammunition as a potential source of lead exposure (Nixon et al. 2001; Hunt et al. 2009; Neumann 2009; Stauber et al. 2010; Warner et al. 2014). Potential management strategies to prevent lead exposure may benefit from focusing efforts on decreasing exposure sources during October–January.

Wildlife rehabilitators routinely collect blood samples from bald eagles admitted. Our logistic regression model demonstrates that blood lead concentrations can be used as a predictor of patient outcome. Prioritizing blood collection for lead analysis as a first line diagnostic modality in bald eagles can help direct case management and allocation of rehabilitation funds.

In 1991, the U.S. Fish and Wildlife Service issued a rule requiring the use of nontoxic shot for all waterfowl hunting to prevent lead toxicosis in the then endangered bald eagle (U.S. Endangered Species Act [ESA 1973, as amended]; Anderson 1992). Although this regulation has decreased waterfowl deaths from lead poisoning (Anderson et al. 2000), the incidence of lead poisoning in bald eagles has not declined significantly (Kramer and Redig 1997). From this 11-y review of data, bald eagles categorized with subclinical or clinical lead levels account for half of the lead-tested bald eagles seen at wildlife rehabilitation centers in Iowa. Radiograph and necropsy findings of shrapnel, shot, and a rifle bullet in eagle digestive tracts suggest there are more sources of lead exposure for bald eagles. These sources include lead ammunition still used for upland small game and large game hunting. By working together, natural resource agencies and organizations, industry, and stakeholders can reduce the amount of lead toxicosis encountered in bald eagles and other wildlife species.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.

Data S1. Data collected on bald eagles Haliaeetus leucocephalus admitted to four Iowa wildlife rehabilitators, 2004–2014.

Found at DOI: http://dx.doi.org/10.3996/122015-JFWM-124.S1 (90 KB XLS).

Reference S1. Friend M, Franson J, Ciganovich E. 1999. Field manual of wildlife diseases. Chapter 43. U.S. Department of the Interior, U.S. Geological Survey, Washington, D.C.

Found at DOI: http://dx.doi.org/10.3996/122015-JFWM-124.S2 (3703 KB PDF); also available at http://www.nwhc.usgs.gov/publications/field_manual/field_manual_of_wildlife_diseases.pdf (May 2017).

Reference S2. Iowa Department of Natural Resources. 2014. Trends in Iowa wildlife populations and harvest 2013.

Found at DOI: http://dx.doi.org/10.3996/122015-JFWM-124.S3 (5809 KB PDF); also available at http://www.iowadnr.gov/Portals/idnr/uploads/Hunting/2013_logbook.pdf (May 2017).

Reference S3. Iowa Department of Natural Resources. 2016–2017. Iowa hunting and trapping regulations.

Found at DOI: http://dx.doi.org/10.3996/122015-JFWM-124.S4 (2666 KB PDF); also available at http://www.iowadnr.gov/Portals/idnr/uploads/Hunting/huntingregs.pdf (May 2017).

We thank S. Ensley and his toxicology team for insight into this issue and many years of service of lead testing; W.R. Clark for editorial contributions; and R. Struve, G. Riordan, and R. Dirks for assistance with necropsies. We also thank the many volunteers that rescue and care for wildlife. The Journal of Fish and Wildlife Management reviewers and the Associate Editor were very helpful in the writing process.

Any use of trade, product, website, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.