- Top of page
- MATERIALS AND METHODS
Mercury (Hg) depuration into growing feathers is a well-studied phenomenon in waterbirds. Although the kinetics of Hg excretion in relation to molt and diet has been studied extensively, the relationship between the individual nutritional condition of nestlings and dietary Hg accumulation has not been investigated. In the present study, a body-condition index (BCI) and nutritional condition index (NCI) for nestlings of two waterbird species occupying different trophic positions on the aquatic food web were determined and used to predict Hg accumulation through diet. Candidate models consisting of these indices and nestling age were compared using Akaike's information criterion corrected for small sample sizes. For both species, the top-performing model contained the sole parameter of nutritional condition index (NCI). The relationship between Hg and NCI was stronger in the species foraging higher on the trophic web, which experienced higher rates of Hg depuration into feathers. Models containing BCI could not be discounted (AICc < 2) for one of the species and the utility of this index is discussed. Environ. Toxicol. Chem. 2012; 31: 1143–1148. © 2012 SETAC
- Top of page
- MATERIALS AND METHODS
Monitoring mercury (Hg) accumulation in waterbirds is a commonly used conservation tool 1–4, and it has been established that feathers from nestlings are particularly useful for indicating Hg content of a local foraging habitat 5–9. This is due largely to the limited movement of adults around a central breeding location 10. Mercury accumulated through diet during the nestling phase is eliminated from the body into growing feathers and should represent the Hg content of the prey items in the habitat surrounding the nest. Because of the ephemeral nature of prey items in estuarine systems and individual foraging habits of adults, however, a study investigating the interaction between the nutritional condition of developing waterbird nestlings and Hg accumulated through diet would strengthen our use of feathers as biomarkers. A dietary study would also strengthen our understanding of the dynamics of terrestrial consumers foraging in aquatic food webs. One might expect nestlings fed prey items that are higher on the trophic web would maintain a better nutritional condition at the expense of accumulating more Hg than a nestling fed a diet of low-trophic level prey items. Such a tradeoff has been observed in piscivorous waterbirds, but the relationship has not been investigated in species foraging primarily on invertebrate prey 11.
Ptilochronology is an accurate, inexpensive technique that uses growth bars, or alternating patterns in deposition of pigment over a 24-h period into growing feathers to determine the nutritional condition of an individual 12, 13. Limited dietary nutrition will result in slower feather growth and thus in narrower growth bars 14. Empirical studies using ptilochronology have established the utility of growth bar width as a biomarker to determine habitat and diet quality 15–18. When coupled with Hg analysis, ptilochronology can further our understanding of the relationship between nutritional condition and Hg accumulation in waterbird nestlings.
The goal of the present study was to apply novel biomarkers in waterbird species to predict Hg accumulation through diet originating from a local foraging habitat. Specifically, our aims were to demonstrate whether a measure of individual nutritional or body condition can predict Hg accumulation in two waterbird species occupying different ends of a foraging habitat-use spectrum and to demonstrate the utility of these biomarkers at two locations along the East Coast of the United States experiencing different regimes with respect to disturbance.
- Top of page
- MATERIALS AND METHODS
No difference existed in any of the measurements for either species by state based on a general linear model ANOVA (glossy ibis: NCI: F = 2.05, p = 0.12; BCI: F = 1.38, p = 0.26; Hg: F = 1.59, p = 0.21; double-crested cormorant: NCI: F = 0.87, p = 0.36; Hg: F = 0.00, p = 0.96; culmen was used as a covariate to control for nestling age). There was high repeatability in feather measurements based on intraclass correlation (r = 0.92).
Cormorant feathers collected from New York were taken from the scapular region, whereas feathers collected in Virginia were the same as those collected from glossy ibis (P1). While Hg deposition into growing feathers is variable by feather tract, mercury analysis did not reveal a difference in Hg (t = −0.73, p = 0.47, df = 19) nor NCI (t = 0.90, p = 0.38, df = 16) between scapulars and primary remiges of the same bird 25.
Nutritional and body condition as indicators of Hg
Three candidate models were chosen to explain variation in log_Hg content (model goodness-of-fit = 0.10) (Table 2). The most parsimonious model (ωi = 0.25) contained the single parameter NCI, whereas the second best-performing model (ωi = 0.23) included the parameters NCI and BCI. The last model considered (ωi = 0.21) contained the single parameter BCI. The relative importance of NCI at predicting variation in feather log_Hg content was greater than BCI (parameter likelihood values: 0.48, 0.44, respectively, evidence ratio value = 1.19). There was a weak negative relationship between average NCI and BCI and log_Hg (Fig. 1).
Table 2. Top-ranked models predicting mercury content of feathers using AIC analysis adjusted for small sample sizes (AICc)a
|GLIB|| || || || || || |
| ||NCIb||2||−129.12||0||0.25|| |
| ||NCI + BCIb||3||−129.32||0.09||0.23||1.05|
| ||NCI + log_culmen||3||−127.12||2.29||0.08||3.15|
| ||BCI + log_culmen||3||−127.11||2.30||0.08||3.16|
|DCCO|| || || || || || |
| ||NCIb||2||34.32||0||0.51|| |
| ||NCI + culmen||3||36.93||2.61||0.14||3.68|
| ||NCI + BCI||3||37.04||2.72||0.13||3.89|
| ||BCI + culmen||3||43.15||8.83||0.01||82.50|
Figure 1. Relationships of log_THg content and (A) nutritional condition index (NCI, mm/day) of nestling glossy ibis (Plegadis falcinellus) feathers and (B) body-condition index (BCI). Nestlings in better nutritional and body condition tended to have less mercury (Hg) burden.
Download figure to PowerPoint
Only one model was chosen to explain variation in log_Hg content (model goodness-of-fit: 0.46). As with glossy ibis, the most parsimonious model (ωi = 0.51) contained the single parameter NCI (Table 2). Nutritional condition index declined with increasing Hg (Fig. 2).
Figure 2. Relationship between total mercury (THg) content and nutritional condition index (NCI, mm/day) of nestling double-crested cormorants (Phalacrocorax auritus).
Download figure to PowerPoint
For glossy ibis nestlings, the majority of prey items in the regurgitant were from the orders Coleoptera (Dysticidae), Diptera (Tabanidae and Muscidae) and Xiphosurida (Limulidae). From diet samples, it was clear that some adults foraged in freshwater systems and selected aquatic macroinvertebrates and amphibians, whereas others foraged primarily in saltwater systems, choosing to forage on eggs of fish and horseshoe crabs. Double-crested cormorant nestlings were fed all marine-derived prey items and predominantly fish species (Fundulidae, Batrachoididae, and Atherinopsidae) and grass shrimp (Palaemonetes spp.) (Table 3).
Table 3. Frequency and percent biomass (PB) of prey items fed to glossy ibis (GLIB, n = 20) and double-crested cormorant (DCCO, n = 10) nestlingsa
| Unidentified eggs||22||4.21||0||0|
| Opsanus tao||0||0||4||5.80|
| Menidia menidia||0||0||2||2.90|
| Palaemonetes spp.||0||0||21||30.43|
| Uca pugnax||1||0.19||0||0|
| Gemma gemma||3||0.57||0||0|
| Littorina irrorata||6||1.14||0||0|
| Belostoma spp.||2||0.38||0||0|
| Aeshna spp.||6||1.14||0||0|
- Top of page
- MATERIALS AND METHODS
The fact that Hg content of feathers did not differ significantly between colonies for the same species suggests that bioavailable Hg may not differ greatly, even though the New York colony resides in a more Hg-contaminated environment. A sufficient dietary overlap may also exist in waterbird foraging, regardless of geographic locale, to make dietary Hg content as much a function of trophic level of prey as it is of geographic variability. Although the total amount of Hg in the New York metro region may be higher, the bioavailable portion may be equally high in the state of Virginia, where large shallow estuaries present ideal locations for the conversion of Hg into a bioavailable form 32.
Many studies have investigated the depuration of Hg into rapidly growing nestling waterbird feathers 33, 34; however, no studies have combined this analysis with ptilochronology to determine the interaction between individual nutritional condition and Hg accumulation. Although the results of the present study suggest that NCI serves as an accurate predictor of the Hg load that nestling waterbirds experience through diet, the fact that individuals in better nutritional condition tend to have lower Hg burdens seems counterintuitive for piscivorous species and requires further investigation. Individual nestlings with a higher Hg burden yet lower NCI may be consuming high trophic level prey items that do not supply adequate energy content to promote feather growth. A nestling glossy ibis that is fed a larger percentage of predatory insects (Odonata, Araneidae) will likely accumulate more Hg than an individual that is fed prey items that forage low on the aquatic food web yet supply high levels of energy. This is illustrated in the present study by individuals that preyed on horseshoe crab eggs, which are high-energy diet items but are lower on the trophic web than predacious insects. The phenomenon of higher rates of mercury accumulation in birds foraging on insects rather than fish has been documented by Critol et al. 35. A study investigating the relationship between energy content, trophic position, and Hg content of commonly consumed prey items would strengthen the present study.
Shortly after fledging, blood and tissue concentrations of Hg increase as feather growth ceases, and this excretory pathway no longer exists 34, 36, 37. The depuration of Hg into growing feathers therefore represents a highly efficient method of reducing body load of Hg during critical periods of development. In the present study, the first 10 d of feather growth were analyzed, but diet samples were obtained during the period of feather collection, when nestlings were between 25 and 30 d old. Therefore, the diet sample collected may not accurately reflect the diet nestlings were fed during the first 10 d of feather growth. A comprehensive study of dietary changes over the course of the nestling phase is needed to account for this source of potential variation. However, because adult foraging behavior dictates nestling diet, it seems safe to assume that a nestling fed a more nutritious diet during the first 10 d of feather growth would be maintained on this diet for the duration of the nestling phase.
Models incorporating the BCI of the individual could not be discounted for glossy ibis in the present study and suggest that the structural size of developing nestlings may be a useful predictor of Hg content for this species. Immature rats dosed with mercury demonstrated inhibited growth in proximal tibia 38. When fed ad libitum diets containing selenium and methylmercury, adult mallards (Anas platyrhynchos) produced embryos with small or malformed wings and legs 39 and stunted growth has been documented in the embryos of waterbird species exposed to methylmercury 40. Traditional body condition indices are calculated as the residuals of body mass on a structural skeletal component (tarsus, wing cord) or as mass gain over the nestling period 41. These indices could be misleading in the present study for two reasons: First, nestlings were weighed once during the study and at different times of day. Because nestling mass varies with time of day and feeding schedule, this single mass measurement would lend itself poorly to comparison 42. Second, metrics of developmental health, such as nestling growth-rate are useful for detecting certain contaminants; however, because nestlings depurate Hg into actively growing feathers, catabolism of body tissue may not take place until after this excretory pathway ceases 34. Because feather growth does not cease until after fledging, use of traditional mass-based growth-rate of nestlings may not be an adequate biomonitoring tool for detecting high levels of Hg in local food sources. Further, a single feather collection is less invasive than multiple trips into a breeding colony to obtain data for growth-rate calculations and represents a more efficient biomarker that causes fewer disturbances. In glossy ibis, the parameter BCI was the third best-performing model and parameter likelihood values suggest that BCI was only strengthened as a predictive tool when combined with NCI. Because the goal of the present study was to identify a biomarker capable of effectively predicting Hg content in nestlings as a proxy for local habitat conditions, body condition indices, whether based on mass gain or bone growth, may be insufficient predictors of Hg in pre-fledged chicks. This is supported by the fact that BCI was not found to be a predictive tool in double-crested cormorants, which is of particular interest considering the much higher Hg accumulation in this species. Despite these shortcomings, the novel BCI used in the present study may be a useful biomarker for predicting contaminants that are known to stunt growth yet for which no known efficient excretory pathway exists, such as cadmium 43, 44.
While ptilochronology has not been applied as a biomarker in waterbird studies, its use could greatly enhance our understanding of local habitat quality and the interactions between dietary items and individual nutritional condition. In addition, the capacity to reduce observer-caused disturbance in species prone to nest abandonment could increase the attractiveness of the technique to conservation and management agencies. A single feather can serve as a dietary record for the entirety of the nestling period without the need to enter breeding colonies repeatedly. The present study demonstrates that the nutritional condition of the individual is not only tied to energy content of diet, but also can be used to gain insight into the dietary contaminant loads derived from diet. When applied appropriately, there is a promising future of ptilochronology as a biomarker in waterbirds.