Analysis of Toxic and Essential Elements in the Blood and Feathers of Humboldt Penguins (Spheniscus humboldti) at Punta San Juan, Peru Open Access

The Punta San Juan Reserve (PSJ; 15°22′S, 75°12′W) protects the largest rookery of endangered Humboldt Penguins (Spheniscus humboldti) in Peru with 3,000–5,000 individuals, accounting for 25–50% of the entire Peruvian population and 10% of the global population (BirdLife International 2017). Conservation at PSJ relies on population health data, particularly in the face of ongoing industrial development in the adjacent town of San Juan de Marcona. Previous research at PSJ characterized organic pollutants in S. humboldti but did not investigate other toxicants (Adkesson et al. 2018).

Limited data exist on elemental exposure in S. humboldti (Smith et al. 2008; Celis et al. 2014). Other studies, primarily in the Antarctic and subantarctic islands, have established penguins as viable bioindicators of inorganic environmental pollutants. Review of these studies highlights the need for data from penguins at lower latitudes, as well as research on blood levels, in addition to feathers and excreta (Espejo et al. 2018). Our objective was to assess 55 elements in the blood and feathers of Humboldt Penguins at PSJ.

We collected samples from 29 adult penguins in May 2009. Penguins were removed from nest sites following previously described collection methods (Adkesson et al. 2018). A veterinarian determined that animals were in normal health on the basis of an examination, hematology, and plasma biochemistry parameters. A maximum of 24 mL of blood was collected from the jugular vein and placed into royal blue trace element tubes (no additive; Vacutainer, BD, Franklin Lakes, New Jersey, USA) for serum elemental analysis and cryovials (Nunc, Thermo Fisher Scientific, Rochester, New York, USA) for whole blood element analysis. Samples were stored on ice for up to 6 h until centrifugation (1,132 × G for 10 min). Resultant serum and whole blood samples were then frozen to –20 C, placed in liquid nitrogen for exportation, and subsequently maintained at –70 C until laboratory analysis. Whole blood samples from two animals were lost. Feathers (1–2 g; 20–40 feathers) were manually plucked from the dorsal scapular region and placed into plastic bags. Analyses were performed at the Illinois Sustainable Technology Center, University of Illinois at Urbana-Champaign.

In the laboratory, feather samples were repeatedly rinsed with deionized water, rinsed with acetone, and allowed to air dry. All samples were prepared by a nitric acid and microwave-assisted digestion procedure following USEPA Method 3051 (USEPA 1998). Samples were completely dissolved into solution under these conditions. Filtration was unnecessary; therefore, reported results are for total metals.

Mercury (Hg) analysis by cold vapor atomic fluorescence was performed per USEPA Method 245.7 (USEPA 2005). Analysis of other elements was conducted by using a VG Elemental PQ ExCell (Thermo Fisher Scientific, Waltham, Massachusetts, USA) inductively coupled plasma mass spectrometer with detection limits (DLs) listed in Table 1. Both instruments were calibrated daily by using commercial reference materials (SPEX Certi- Prep, Metuchen, New Jersey, USA).

Three control blanks were prepared to detect introduction of any contaminants. Deionized water was aliquoted into the cryovials prior to travel to PSJ, at the time of blood collection, and at the time of sample centrifugation. Quality control measures during analysis included digestion blanks, reagent blank spikes, analytic duplicates, and spiked samples. The Hg digestion blanks were low relative to the sample concentrations. The analytic duplicates reproduced well and yielded relative differences ranging from 0.3% to 15%. The analytic spikes and reagent blank spikes recovered well within a range of 76 to 103%, with the exception of two laboratory control standards that were biased low, one at 62% and the other at 65%.

Data were tested for normality by using the Shapiro-Wilks and Anderson-Darling tests. Homogeneity of variances among groups was tested by using the Levene test, and correlations were determined with the Pearson method. Effect of gender, nesting location, and interactions on elemental concentration were determined by using repeated measures analysis of variance. Analyses were done by using SYSTAT 13.0 (Systat Software, San Jose, California, USA) and SAS 9.1.3 (SAS Institute Inc., Cary, North Carolina, USA). Values below the DL were proxied as the DL, thereby estimating a maximum concentration (hence, risk) in that penguin.

Aluminum (Al), calcium, iron (Fe), Hg, potassium (K), magnesium (Mg), sodium, and zinc (Zn) were detected in serum (Table 2). These, plus selenium and titanium, were present in whole blood (Table 3). Of the quantifiable elements in common between whole blood and serum, only Mg concentrations were correlated (r=–0.38, P=0.05).

Fourteen elements were quantifiable in feathers (Table 4), including all the elements found in serum plus arsenic (As), boron, barium, copper (Cu), manganese, and titanium. Arsenic (6.3 mg/kg) was present in only one penguin. Of the elements above DLs in both feathers and whole blood, Mg (r=0.50, P=0.008), Hg (r=0.49, P=0.01), and Zn (r=0.53, P=0.004) were correlated between the two tissues.

With the exception of Hg in feathers (t=–2.97, P=0.007), there was no significant difference between sexes in Al, calcium, Cu, Mg, Hg, K, sodium, or Zn concentrations in any sample type with enough quantifiable values for comparison (t=–1.64 to t=1.75, P=0.091 to P=0.95). The average Hg concentration in feathers from females (1.11 mg/kg) was 19% lower than in males (1.37 mg/kg).

Human population growth near coastlines often leads to environmental contamination. Waste materials from the largest open-pit Fe ore mine in Peru, an expanding Cu mine, and sewage and refuse from San Juan de Marcona are deposited in locations as close as 4 km from PSJ. Although mines are classified on the basis of their predominant products, they typically release a complex mixture of elements to the downstream environment (Dudka and Adriano 1997). Such perturbations can have profound effects on ecologic diversity and the health of marine wildlife, though many of these toxic elements of concern were not detected in this study, notably cadmium, chromium, lead, nickel, and silver.

Mercury is a noteworthy contaminant in marine environments due to its persistence and biomagnification throughout the food chain. A number of studies have examined Hg in penguin tissues, but there are few published accounts of blood concentrations. Our maximum serum and whole blood Hg concentrations are one to three orders of magnitude lower than mean blood Hg concentrations reported in Adélie Penguins (Pygoscelis adeliae; Honda et al. 1986). Our levels are also lower than that associated with decreased reproductive productivity in Common Loons (Gavia immer; Burgess and Meyer 2008), supporting an observed lack of reproductive impairment in Humboldt Penguins at PSJ.

Feather concentrations of trace elements reflect circulating blood concentrations during feather formation (Burger et al. 2008) and can be used as a surrogate for blood for some elements. Penguins are known to transfer part of their Hg load into feathers during molt (Metcheva et al. 2006). Peak molting for Humboldt Penguins at PSJ occurs in January–February, approximately 3 mo before our samples were collected. Following molt, Humboldt Penguins undergo a period of hyperphagy in preparation for breeding, during which time there would be a presumed net accumulation of Hg until egg laying. The lower concentration of Hg in the feathers of females in this study was consistent with the maternal transfer of Hg into eggs reported in other penguin species (Becker et al. 2002). The maximum Hg concentration in Humboldt Penguin feathers (1.8 µg/g dry weight) was well below that associated with flight feather asymmetry in Common Loons (40 µg/g Hg fresh weight; Evers et al. 2008) and below the threshold associated with reproductive effects (5.0 µg/g) in various species (Burger and Gochfield 1997). Mercury concentrations in the feathers of Humboldt Penguins at PSJ were, on average, similar to that reported in seabirds collected along the Peruvian coast over 30 yr ago (Gochfeld 1980).

Our DLs for the toxic elements As, cadmium, chromium, lead, and selenium in feathers were high (5.0 mg/kg) relative to the means reported for elements in the feathers of other penguin species (Metcheva et al. 2006; Jerez et al. 2011). Because all but one value for these elements (As, 6.3 mg/kg) were below DLs, comparison of our results to these studies is only possible in a broad sense. It is unknown if the absence of detectable concentrations is an artifact of the high DLs or a true lack of exposure. Our mean concentrations of Cu and Zn were also within or below the ranges reported for Antarctic penguin species.

Of the essential elements in common with other studies, Humboldt Penguins in our study had notably higher levels of Fe, K, Mg, and manganese. The mean Fe concentration in the feathers of Humboldt Penguins was much higher than reported for adult penguins in other studies (Metcheva et al. 2006; Jerez et al. 2011). Discharge from a large, 670 km² open-pit mine that produces >7 million metric tons of Fe ore per year occurs 13 km to the north of PSJ, where ore is also loaded for shipment. Mine proximity represents a plausible explanation for the observed Fe levels, although Fe can change markedly with age and molt (Ghebremeskel et al. 1989). Feather Fe concentration may be a useful geographic marker for PSJ, as trace element concentrations can be used to discern geographic locations of exposure (Norris et al. 2007).

Coastal development and regional population growth near PSJ highlight the need for reference data. Our results indicate breeding Humboldt Penguins at PSJ did not appear to be exposed to contaminants at levels detrimental to health and reproduction. However, measurable concentrations of Al, As, Cu, Hg, and Zn underscore the need for monitoring to protect this critical population, particularly in the face of continued coastal development in the region.

Research was authorized under Peruvian permit 131-2009-AG-DGFFS-DGEFFS, US Convention on International Trade in Endangered Species (CITES) permit 09US205977/9, and Peruvian CITES permit 000353. Funding was provided by the Saint Louis Zoo's Field Research for Conservation Fund and WildCare Institute, the Chicago Zoological Society, and Chicago Board of Trade Endangered Species Fund. The Department of Comparative Biosciences, University of Illinois at Urbana-Champaign, contributed in-kind services. We thank Patricia Majluf, Val Beasley, Marco Cardeña, Santiago de la Puente, Franco Garcia, Paulo Guerrero, Maria Jose Ganoza, Alonso Bussalleu, and Michael Macek for their support of this project. In addition, we thank Nandakishore Rajagopalan, Gerald Bargren, and Christie Teausant for technical assistance and laboratory support with chemical analysis.