Seasonal Differences in Contaminant Accumulation in Neotropical Migrant and Resident Songbirds
For many years, it has been hypothesized that Neotropical migrants breeding in the United States and Canada accumulate organochlorine pesticides (OCPs) while on their wintering grounds in Latin America. We investigated the seasonal accumulation of persistent organic pollutant (POPs) in migrant and resident passerines in Texas, Yucatán, and Costa Rica collected during the fall, winter, and spring from 2011 to 2013. A total of 153 birds were collected, and all contained detectable levels of polychlorinated biphenyls (PCBs), polybrominated diphenyl ethers (PBDEs), and OCPs with dichlorodiphenyldichloroethylene (DDE) being the most predominant pesticide. OCPs and PCBs were the predominant contaminants, accounting for ≥80 % of the total POPs burden, whereas PBDEs accounted for ≤16 %. Only spring migrants from Texas had significantly greater DDE concentrations (64.6 ng/g dry weight [dw]) than migrants collected in Costa Rica (23.2 ng/g dw). Resident birds in Texas had significantly greater levels of DDE (121 ng/g dw) and ΣPBDEs (34.8 ng/g dw) compared with residents in Yucatán and Costa Rica. For ΣPCBs, resident birds from Costa Rica had significantly lower concentrations (9.60 ng/g dw) compared with their migrant counterparts (43.7 ng/g dw) and residents from Texas (48.3 ng/g dw) and the Yucatán (32.1 ng/g dw). Migrant and resident passerines had similar congener profiles for PCBs and PBDEs suggesting similar exposure and retention of these contaminants. No significant accumulation of DDE was observed in migrants while on their wintering grounds. Relatively high concentrations of PBDEs in resident birds from Costa Rica warrant future studies of PBDE contamination in Latin America.
Neotropical migratory passerines are those songbird species that breed in North America (United States and Canada) during the summer months and migrate south in the fall to their wintering grounds (Caribbean, Mexico, Central and South America). Due to their annual movements, these long distance migrants may be exposed to a wide range of pollutants over large geographic areas. Some migratory songbirds have suffered significant population declines over the past several decades; explanations for these have included habitat loss, parasitism, climate change, and environmental pollutants (Robbins et al. 1989; Robinson and Wilcove 1994; DeGraaf and Rappole 1995; Both et al. 2006). However, the relationship between environmental contaminants and population declines of Neotropical migratory songbirds is poorly understood (Finch and Martin 1995; Gard et al. 1993). In contrast, population declines for larger avian species and dichlorodiphenyltrichloroethane (DDT) are well documented in the literature. The primary breakdown product of DDT, dichlorodiphenyldichloroethylene (DDE), has been shown to cause eggshell thinning, thus critically affecting reproduction and hatching success (Peakall 1993; Fry 1995; King et al. 2003). Because most Neotropical migrants spend the majority of their annual cycle either migrating or on their wintering grounds (Newton 2010), it is critical to understand how contaminants in these areas may impact individuals and populations during migration. Migratory behavior can play an important role in the accumulation of contaminants and their distribution within the body (Van Velzen et al. 1972; Henny et al. 1982). Several studies have shown that there are seasonal differences in body burdens for migratory birds; there have also been observed differences in the levels and specific contaminant profiles between migrants and residents (Tanabe et al. 1998; Kunisue et al. 2003; Yogui and Sericano 2009; Seegar et al. 2015). Migration is energetically costly, and birds rely primarily on fat stores to fuel their flights. Lipophilic contaminants especially can thus become mobilized and redistributed to different organs, thus posing additional risk to migrating birds that may already be under physiological stress (Scollon et al. 2012).In addition, contaminant exposure in songbirds has been shown to alter migratory behavior by disrupting orientation capabilities and impairing molt completion, which are both critical for successful migration (Flahr et al. 2015).
The use of many persistent organic pollutants (POPs), such as organochlorine pesticides (OCPs) and polychlorinated biphenyls (PCBs), was banned in North America in the early 1970s, which lead to significant decreases of these contaminants in passerines and other avian species (Johnston 1974; Olsson et al. 2000; Henny et al. 2009b; Mora et al. 2016). However, due to their persistence in the environment, detectable levels are still observed in many passerine species and have the potential to cause adverse health effects (McCarty and Secord 1999a; Neigh et al. 2007; DeLeon et al. 2013). In addition, polybrominated diphenyl ethers (PBDEs), a class of flame retardants, are relatively new emergent contaminants of concern with increasing temporal trends seen in environmental media and biota (Shaw and Kannan 2009). Encouragingly, restrictions and bans on the use and production of certain PBDE formulations in developing countries have led to observable decreases in the levels of flame retardants in wildlife and the environment (Law et al. 2014). However, the persistence of these contaminants, as well as their ability to undergo long-range transport, has led to their ubiquity in all environmental media, humans, and wildlife (Jones and De Voogt 1999).
A long-held hypothesis is that migratory songbirds accumulate OCPs, in particular DDT and its degradation products, while on their wintering grounds in Latin America where it was used in agriculture and for malarial control and not banned until the late 1980s and early 1990s (Von Duszeln 1991; Albert 1996; Díaz-Barriga et al. 2003). Although an earlier study by Henny et al. (1982) showed that migrating Peregrine falcons (Falco peregrinus) accumulated chlorinated pesticides while on their wintering grounds, later studies with Black-crowned night-herons (Nycticorax nycticorax) suggested that populations wintering in Mexico had limited exposure to OCPs compared with those wintering in the Southwestern United States (Henny and Blus 1986). A recent article by Mora et al. (2016) examined temporal and latitudinal contaminant trends in birds from North America. Results from the analyses showed decreasing concentrations of DDE with increasing latitude for migrant passerines. The investigators suggested that these trends could be due to possible hotspots of DDE contamination in the Southwestern United States and Mexico. In contrast, the current literature shows relatively low levels of POP residues for resident passerines in Latin America and low exposure in migrants while on their wintering grounds (Fyfe et al. 1990; Harper et al. 1996; Capparella et al. 2003; Klemens et al. 2003; Mora 2008). Given these findings, our study aims to help determine where and when Neotropical migrants are at greatest risk of exposure to contaminants. This information is important for understanding the broader impacts of pollutants on individuals and populations. Our main objective was to examine seasonal variation in contaminant accumulation of POPs, including OCPs, PCBs, and PBDEs, in migratory and resident songbirds from locations in North and Central America. Due to their year-round residency, resident species are good indicators of local contamination levels and can be used to determine pesticide acquisition in migrants.
Study Areas, Sample Collection, and Preparation
Species information, lipid, and moisture content for passerines collected in Texas, Yucatán, and Costa Rica from 2011 to 2013
Whole-body homogenate samples were analyzed for OCPs, PCBs, and PBDEs at the Geochemical and Environmental Research Group (GERG), Texas AandM University, College Station, Texas, USA+. The following pesticides were measured: DDT isomers (o,p’-DDE, p,p’-DDE, o,p’-DDD, p,p’-DDD, o,p’-DDT, p,p’-DDT), tetrachlorobenzene 1,2,4,5, tetrachlorobenzene 1,2,3,4, pentachlorobenzene, hexachlorobenzene (HCB), α-HCH, β-HCH, γ–HCH, δ-HCH, heptachlor, heptachlor epoxide, oxychlordane, α-chlordane, γ-chlordane, cis- and trans-nonachlor, aldrin, dieldrin, endrin, pentachloroanisole, chlorpyrifos, mirex, and endosulfan I and II. The following PCB congeners were quantified: 1, 7, 8, 15, 16/32, 18, 22, 24, 25, 147 26, 28, 29, 31, 33, 39, 40, 41/64, 44, 45, 46, 47/48, 148 48, 49, 52, 53, 60/56, 63, 66, 67, 69, 70, 72, 74, 77, 149 81, 82, 83, 84, 85, 87, 92, 95/80, 97, 99 101, 105, 150 107/108/144, 110, 114, 118/108/149, 119, 126, 128, 151 129, 130, 135, 136, 138, 141, 146, 149, 151, 153, 152 156/171/202, 158, 166, 167, 169, 170, 171/202, 172, 153 174, 175, 177, 178, 180, 183, 185, 187/182/159, 189, 154 191, 193, 194, 195, 196, 197, 199, 200, 201, 205, 155 206, 207, and 209. Congeners that coeluted during chemical analysis are grouped together by slashes. A total of 38 PBDE congeners (i.e., 1, 2, 3, 7, 8/11, 10, 12, 13, 15, 17, 25, 28, 30, 32, 33, 35, 37, 47, 49, 66, 71, 75, 77, 85, 99, 100, 116, 118, 119, 126, 138, 153, 154, 155, 166, 181, and 183) were investigated. Total PCBs (ΣPCBs) and PBDEs (ΣPBDEs) were the sums of all detectable congeners in a sample.
Chemical analysis followed methods described by GERG standard operating procedures. Briefly, for each sample approximately 1 g of homogenated tissue was mixed with Hydromatrix or anhydrous sodium sulfate and solvent extracted with methylene chloride using an ASE 200 accelerated solvent extractor (Dionex Corp., Sunnyvale, CA). Before extraction, samples, reference material, and blanks were spiked with surrogate standards for both chlorinated and brominated compounds. Samples were then concentrated and cleaned with silica/alumina chromatography columns using 200 mL of 1:1 pentane/methylene chloride mix as the eluent. High-performance liquid chromatography (HPLC) was used to further purify the samples. Finally, extracts were concentrated to 100 μL in hexane. The internal standards tetrachloro-meta-xylene and PCB 103 were added before gas chromatographic analysis. Procedural blanks, sample duplicates, and matrix spikes were run with every 20 batch of samples for quality control. The average percent recovery of target analytes in the spiked samples were, with a few exceptions, within the method of acceptable range (40 to 120 %), and the relative percent differences (RDP) between the duplicates were <15 % for all compounds. The method detection limits (MDLs) were compound- and sample-specific and based on sample weight. The mean MDLs were 0.52 ± 0.05 ng/g dw (range 0.44 to 0.86), 0.52 ± 0.06 ng/g dw (range 0.25 to 0.86), and 0.51 ± 0.06 ng/g dw (range 0.25 to 0.86) for OCs, PCBs, and PBDEs, respectively. Standard reference materials (i.e., SRM2947A and SRM1944) were analyzed for certified PCB and PBDE congeners and OCPs for comparison. Quantitative analysis for OCPs, PCBs, and PBDEs was accomplished using an Agilent 6890 N Gas Chromatograph Agilent Technologies, USA coupled to a low-resolution 5975C inert mass selective detector in the selected ion monitoring (GC/MSD-SIM) and a 30 m × 0.25-mm i.d. fused silica capillary column with DB-5MS (J&W Scientific Co., USA) bonded phase for organochlorine analytes. Automatic splitless injections of 2 μL were introduced into the column using helium as carrier gas at 1 mL min−1. For PCBs, the oven temperature was programmed from 75 °C (3-min hold) to 150 °C at a rate of 15 °C min−1 to 260 °C at a rate of 2 °C min−1 and then to 300 °C (1-minute hold) at a rate of 20 °C min−1. For OCs, the oven temperature was programmed from 100 °C to 200 °C at a rate of 10 °C min−1 and then to 265 °C (4-min hold). Injector temperature was held at 270 °C and the detector temperature at 310 °C. For PBDEs, the oven temperature was programmed from 130 °C (1-min hold) to 154 °C at a rate of 12 °C min−1, then ramped at 2 °C min−1 to 210 °C, and finally ramped at 3 °C min−1 to 300 °C (5-min hold).
An aliquot of extract was used to calculate the percent extractable organic material, hereafter referred to as “lipids,” by gravimetric analysis. Percent moisture was determined by weighing homogenized samples before and after freeze-drying. Percent moisture was calculated as the percent difference between the wet weight and dry weight (Table 1). Contaminant concentrations on a lipid-weight (lw) basis were significantly and negatively correlated with lipid content; therefore, final concentrations of contaminants are expressed in ng/g on a dry-weight (dw) basis.
Statistical analyses were performed using JMP software program (SAS Institute 2015, Inc. Cary, NC). Samples with concentrations at or lower than the MDL were assigned a value equal to half the MDL. Data were not normally distributed (Shapiro–Wilk test, p > 0.05) and were log-transformed to meet the assumptions of normality and homogeneity. One-way analysis of variance with post hoc test (Tukey–Kramer HSD) was used to test for differences between migrant and resident birds as well as among locations. The significance level for all tests was set at p < 0.05.
Of the 29 OCPs analyzed, only HCB, trans-nanochlor, mirex, and p,p’–DDE (hereafter, DDE) were detected at greater than the MDL in ≥50 % of the samples (Table 1). DDE was detected in 100 % of all samples; low levels of HCB were detected in 91 % of samples; and mirex was detected in 77 % of all samples. Low levels of trans-nanochlor were detected in 59 % of all samples; however, this was only in songbirds from Texas. Pentachlorobenzene, heptachlor epoxide, cis-nonachlor, and α- and γ-chlordane were also detected, but they were generally below the MDL. Resident birds from Texas had levels of oxychlordane ranging from 1.29 to 195 ng/g dw. The metabolite DDE was the most predominant contaminant of the OCPs accounting for 34 % to 94 % of the total OCP burden. Nine birds also contained p,p–DDD above the MDL: 6 were from Texas, and one individual was from Costa Rica.
Concentrations (geometric mean and range ng/g dw) of POPs for migrant and resident passerines collected in Texas, Yucatán, and Costa Rica from 2011 to 2013
Mean ΣPCBs concentrations for migrant birds collected in Texas, Yucatán, and Costa Rica ranged from 43.7 to 79.7 ng/g dw and for resident birds from 9.60 to 48.3 ng/g dw (Table 2). For ΣPCBs, no significant differences in contaminant levels were observed for migrants. In contrast, concentrations of ΣPCBs in resident birds from Costa Rica were significantly lower (p ≤ 0.01) compared with levels in all other resident groups (Table 2). In general, migrant birds had greater levels of ΣPCB concentrations compared with their resident counterparts, but only migrants from Costa Rica had significantly greater concentrations than resident species (p = 0.0007; Table 2).
Mean ΣPBDEs concentrations for migrant birds collected in Texas, Yucatán, and Costa Rica ranged from 8.71 to 19.4 ng/g dw and for resident birds from 6.0 to 34.9 ng/g dw (Table 2). One single male Carolina wren (Thryothorus ludovicianus) collected in Texas during spring 2013 had a ΣPBDE concentration of 4207 ng/g dw and had approximately 9 times greater concentrations than the next highest concentration; therefore, this value was not included in the statistical analysis or the calculation of summary statistics for Table 2. For ΣPBDEs, only resident species showed significant differences in body burdens. Resident birds collected in Texas had greater concentrations of ΣPBDEs compared with resident birds from both Yucatán (p ≤ 0.005) and Costa Rica (p ≤ 0.002; Table 2).
Spatial Variation of Contaminant Profiles for Migrants and Residents
The data presented in this study did not support the hypothesis that migratory songbirds accumulate contaminants while residing on their wintering grounds. Furthermore, residents collected in Yucatán and Costa Rica had the lowest concentrations of DDE, thus indicating low levels of contamination and unlikely pesticide acquisition by migrants during the wintering period in these areas. These findings are in agreement with other studies that have examined contaminant accumulation in Neotropical migrants (Harper et al. 1996; Mora 2008). For example, Harper et al. (1996) investigated OCP levels in Neotropical migrants and found no significant differences between hatch year (HY) and after hatch year (AHY) birds, which would suggest exposure and accumulation of OCPs on their breeding grounds or by way of maternal transfer. In addition, other researchers have reported relatively low pesticide contaminant levels in passerines from Central and South America (Fyfe et al. 1990; Capparella et al. 2003; Klemens et al. 2003).
We detected relatively low levels of POPs in migratory and resident songbirds (Table 2). The concentrations of DDE, ΣPCBs, and ΣPBDEs in this study were also relatively lower than those reported in other recent ecotoxicology studies (Mora 2008, 2012); however, the observed concentration differences for passerine species within Texas may be attributed to sampling location or species and diet differences (Table S-1). All migrant passerines collected in this study were warbler species, which are mainly terrestrial insect gleaners, whereas Mora et al. (2012) examined POP concentrations in cliff swallows (Petrochelidon pyrrhonota) from the Rio Grande geographic area of Texas. Cliff swallows feed primarily on aquatic insects, such as odonates—which are carnivorous insects—and may be feeding at a greater trophic level (De Graaf et al. 1985; Brown et al. 1995). In this study, the sampling locations in Yucatán were predominantly rural areas not near traditional agriculture, whereas Mora (2008) sampled birds near agricultural areas in western Mexico, which may explain the greater concentrations of DDE reported in the study compared with values reported here (see Supplementary Table S-1).
Present findings also showed that migrants collected in Yucatán had slightly, yet not significantly, greater concentrations of DDE compared with resident species (Table 2). These results are noteworthy given that they are not in agreement with other studies of passerines from Southeastern Mexico. Herrera-Herrera et al. (2013) observed that resident songbirds had significantly greater concentrations of DDT compared with migrants (Table S-1). Levels of ΣPCBs in the migrants we collected in Yucatán (27.2 ng/g ww) were greater than previously reported concentrations in passerine birds from Southeastern Mexico (7.60 ng/g ww) (Herrera-Herrera et al. 2015). Levels of resident species, however, were comparable (9.10 ng/g ww; Table S-1).
In the present study, ΣPBDE concentrations for resident birds from Costa Rica ranged from 1.40 to 194 ng/g dw (Table 2); this is a tenth-fold greater difference compared with migrants from Costa Rica (range 4.30 to 18.2 ng/g dw; Table 2). However, this reported difference could be attributed to a smaller sample size of migrants collected. In addition, the ΣPBDE range for these Costa Rican residents was also greater than the range for migrants collected in Texas during the fall (5 to 96.3 ng/g dw) as well as migrants from Yucatán (2.10 to 62.9 ng/g dw; Table 2). These findings are noteworthy because, in general, birds from North America have greater PBDE burdens compared with birds from other regions, which is likely due to greater demand for PBDE containing products (Chen and Hale 2010). PBDE contamination in passerines has been shown to be highly correlated to distance from industrialized and urban areas (Sun et al. 2012; Tang et al. 2015). Findings in the present study show that birds from non-industrialized areas can also accumulate relatively high concentrations of PBDE flame retardants.
Migrant and resident passerines had comparable congener profiles for PCBs and PBDEs suggesting similar acquisition and retention of these contaminants. PCB congener profiles presented in this study are also similar to those reported from studies in China on other passerine species (Yu et al. 2014) and aquatic birds from Texas, South America, and India (Mora 1996; Senthilkumar et al. 1999; Colabuono et al. 2012). In addition, the pattern of dominant PCB congeners observed in present songbirds is consistent with observations for other insectivorous passerines (Winter and Streit 1992; Dauwe et al. 2003; Yu et al. 2014) and aquatic birds from south India (Senthilkumar et al. 1999). It should be noted that resident birds from Yucatán and Costa Rica had greater contributions of the less chlorinated congeners (3, 4, and 5 chlorines) compared with resident species in Texas suggesting differences in exposure to various PCB Aroclor mixtures. The PBDE congener profiles in the present study for both migrant and resident birds are consistent with penta-BDE technical mixture patterns (La Guardia et al. 2006), and congener profiles reported for other passerines from Texas and China (Mora et al. 2012; Tang et al. 2015).
The low mean concentrations of PBDEs and PCBs observed in birds from Costa Rica and low PBDE concentrations in birds from Yucatán during this study are similar to the observations made on other environmental media (soil and ambient air) in developing regions (Mochungong and Zhu 2015). Low atmospheric concentrations of PCBs and PBDEs for the Yucatán peninsula and Costa Rica have been reported in the literature (Shen et al. 2006). In addition, Daly et al. (2007) reported low levels of DDT metabolites in soils from Costa Rica. The absence of DDT and low levels of DDE in birds sampled from the Yucatán peninsula and Costa Rica in this study may be attributed in part to the climatic condition of the tropics, which favor volatilization and degradation of OCPs (Carvalho et al. 1994; Wania and Mackay 1993). Low levels could also be attributed to low historic use of DDT in the areas selected as sampling sites in Costa Rica and Mexico.
Levels for DDE, ΣPCBs, and ΣPBDEs in the present study were lower than those reported to cause adverse health effects in relation to behavior, reproduction, and development seen in other passerines (McCarty and Secord 1999b; Neigh et al. 2007; DeLeon et al. 2013; Eng et al. 2014) and raptors (Newton 1988; Henny et al. 2009a). However, one male Carolina wren from Texas collected during the spring had a concentration that exceeded levels known to cause reproductive impairments in other birds (Henny et al. 2009a). Exposure and accumulation of environmental contaminants is dependent on a large variety of factors including historic use, varying life histories, environmental factors, and physiochemical properties of the contaminant (Jones and de Voogt 1999; Smith et al. 2007). A limitation of this study is the large species variability of the samples collected. Although target species were selected based on the likelihood of presence for each sampling location, not all species were collected at each of the sites nor in the same proportion. In addition, even within a relatively small geographic area, there may be much heterogeneity of contaminant residues in the soil affecting accumulation by songbirds, thus further complicating the inferences that can be made from exposure and accumulation studies (Reynolds et al. 2001).
In the present study, relatively low, but detectable, levels of POPs were observed in migrant and resident songbirds from Texas, Mexico, and Costa Rica. No significant accumulation of DDE was observed in migrants while residing on their wintering grounds. In addition, low levels of DDE observed in resident birds from Yucatán and Costa Rica also suggest limited exposure in these areas. Results from this study are in agreement with others showing relatively low levels of OCPs observed in passerines from Latin America and no significant contaminant accumulation on wintering grounds. Furthermore, no variation in PBDEs or accumulation of PCBs was observed in migrants throughout the migration period. However, industrial contaminant burdens of PCBs and PBDEs were greater for migrant and resident birds breeding in more northern latitudes compared with resident species in Yucatán and Costa Rica. However, some resident songbirds from Costa Rica had individual concentrations of ΣPBDEs similar to values seen in the migrant and resident species from developed regions in North America. Concentrations of individual contaminants were relatively low and unlikely to cause adverse health effects. The limited availability of data on POPs in passerines and other wildlife for Latin America makes effective comparisons difficult; future studies should focus on increasing monitoring studies in resident birds for these areas.
We appreciate the assistance from the personnel of the Gulf Coast Bird Observatory, Universidad Autónoma de Yucatán, La Selva Biological Research Station, University of Georgia Athens Costa Rica Station, Eugenio Gonzalez from the Texas A&dM University Soltis Center, and private-property owners for providing access to properties as well as logistical and field support. Texas A&M University and the Consejo Nacional de Ciencia y Tecnolgia grant program and the Alfred P. Sloan Foundation provided funding support for this work. We also thank the anonymous reviewers for helpful comments on our manuscript.