Abstract Mercury is a pervasive pollutant that readily enters the food web, accumulates within organisms and across trophic levels, and poses a significant threat to long-lived top predators, such as many seabirds. Since mercury can be deposited in feathers and therefore excreted through moult, feather analysis offers a minimally invasive way to assess mercury contamination levels in birds. However, it remains largely unclear whether mercury levels vary among feather types within individuals, how feather mercury levels relate to those in blood, and whether all levels vary with individual characteristics, such as sex and age, in a similar way. To address these issues, we collected dorsal, ventral, primary and tail feathers from > 200 common terns ( Sterna hirundo ) from a breeding population in the north of Germany, of which a subset was blood sampled as well. Additionally, we collected multiple dorsal feather samples from 313 individuals across years (i.e. longitudinally), to assess whether age effects were likely to reflect within-individual processes. Mercury levels were significantly correlated across feather types within individuals, with those in dorsal and ventral feathers, both grown in the wintering area, also correlating with mercury levels in blood sampled during the breeding season. Males and females did not differ in their feather mercury levels, but older birds exhibited higher mercury levels in all feather types except for tail feathers. Unique analysis of the longitudinally collected dorsal feathers showed that although the within-individual estimate for age was not significant itself, it was very similar to the among-individual age effect, suggesting that within-individual accumulation of mercury during ageing is a likely explanation for older birds having higher mercury levels in most of their feathers. We suggest that body feathers are most suitable for characterising the mercury levels of seabirds in a little-invasive way, but that being able to account for the age of the sampled individuals is important.
Bertram et al. (Tue,) studied this question.