ABSTRACT
BACKGROUND: DNA-based studies have demonstrated that avian genetic mating systems vary widely, with many species deviating from long-assumed monogamy by practicing extra-pair paternity and conspecific brood parasitism. Colonially breeding waterbirds provide interesting models in which to investigate this question because they show nesting habits proposed to promote alternative reproductive strategies. However, little is known about the genetic mating systems of this group of birds, mainly due to difficulties in obtaining genetic data from incubating adults at nests that are necessary for conducting conventional parentage studies. Here, we inferred kinship patterns among offspring in broods of three co-distributed waterbird species, Wood Stork (Mycteria americana), Roseate Spoonbill (Platalea ajaja) and Great Egret (Ardea alba egretta), to investigate genetic mating system in the absence of parental data. RESULTS: Multi-step analyses combining estimates of relatedness coefficients, formulation of relationship-hypotheses, significance testing of alternative hypotheses, and maximum-likelihood sibship reconstruction techniques revealed evidence that alternative reproductive strategies may be present in natural populations of Wood Storks and Roseate Spoonbills, whereas relatedness of co-nestlings diagnosed in the Great Egrets did not deviate from a hypothesis of genetic monogamy. Specifically, under this analytical framework, inferred kinship relationships revealed that Great Egret nests contained full-sibling nestlings (100%), with the Roseate Spoonbill (RS) and Wood Stork (WS) exhibiting proportions of half-siblings (RS: 5%) and/or unrelated nestlings (RS: 24%; WS: 70%), patterns consistent with extra-pair paternity and conspecific brood parasitism, respectively. CONCLUSIONS: We provide evidence that genetic monogamy occurs in Brazilian natural breeding colonies of the Great Egret, but is not the sole reproductive strategy employed by the Wood Stork and the Roseate Spoonbill. In fact, extra-pair paternity and conspecific brood parasitism were common in the latter two species, with a combined frequency of 7.5% and 11.3% in Roseate Spoonbill and Wood Stork, respectively. Although geographically co-distributed, differences among these species may be due to variation in their life histories. From a methodological standpoint, the approach implemented here, although not free from limitations, can have broad application for analyzing systems with limited genealogical information and/or in studying similarly challenging organisms in which obtaining genetic data on complete families is problematic.
