Integrating molecular techniques with field methods in studies of social behavior: a revolution results

INTRODUCTION

When social behavior is defined as encompassing interactions between individuals, in the sense of Tinbergen (1953), then the purview of this topic is wide. It includes mating behavior as well as family and group interactions, cooperation, competition, and the evolution of societies. Molecular methods have the potential to enrich studies over this entire spectrum, but their application so far has been limited to particular taxa and questions. Most studies are of birds, mammals, and Hymenoptera (ants, bees, and wasps), and address either parentage or genetic relatedness. If the results of such studies were mere confirmation and refinement of what observations would predict, then we might expect a relatively short list of publications with titles similar to a recent one: "Behavior predicts genetic structure in a wild primate group" (Altmann et al. 1996). This is not the case, however. The application of molecular methods has yielded startling revelations about mating systems and social groups.

In this review I give an overview of recent publications, emphasizing work that has challenged accepted theory and suggested new avenues for research. I first consider how molecular methods have affected the study of mating systems in species for which mating associations are prolonged. Molecular methods have enlivened the study of monogamy by showing that individuals may mate outside of their pair bond. I also describe how molecular methods have so changed our understanding of other mating systems that underlying theory must be reevaluated. Second, I consider how molecular methods have enriched the study of sociality by detailing how reproductive success is partitioned among individuals in groups. Finally, I describe how molecular methods have been applied to eusocial insects to estimate genetic relatedness among colony members and to identify mechanisms generating variance in relatedness such as number of mates per queen.

MATING SYSTEMS AND GENETIC MATING SYSTEMS

The study of mating systems focuses on ways individuals obtain mates, the number of individuals with which they mate, how long mates stay together, and allocation of parental care (although many would argue that this last feature is a separate question). These behaviors are often analyzed in the context of how they affect lifetime reproductive success (e.g., Emlen and Oring 1977, Davies 1991). Mating patterns often correlate with ecological factors; for example, monogamy and biparental care may be responses to food limitation so severe that a single parent could not succeed. Mating systems have often been described from behavioral data alone, but molecular techniques can reveal unexpected patterns of gene transmission (pedigree connection) resulting from the diverse behavioral tactics that individuals employ. The "genetic mating system" is a description of which individuals are reproducing and with whom, as distinct from the "social mating system" based on observable associations between individuals (these social mating systems are used as subheadings below). Accurate parentage assignment permits determination of the genetic payoff for observed behavioral strategies, and lifetime reproductive success.

Monogamy

Monogamy is the nearly exclusive association of one male with one female for one or more breeding attempts; it is characteristic of birds, but rare in other taxa (Mock and Fujioka 1990). The recent explosion in use of molecular markers has shown that this apparently simple mating system frequently involves more than just a single pair of birds. In monogamous species, observations have been frequently documented of pair-bonded individuals copulating with individuals other than their mates (see reviews in Birkhead 1987, Westneat et al. 1990), but the genetic consequence of these "extra-pair copulations" (EPCs) were little known before the widespread use of molecular methods. Earlier work, using estimates of heritability, experimental manipulation, plumage markers, and allozymes had shown that extra-pair copulations could lead to extra-pair fertilizations (EPFs) (Westneat et al. 1990). Two important consequences of EPFs are that males often invest parental care in chicks to which they are unrelated and that they can achieve reproductive success outside of the pair bond.

Molecular markers are the easiest and most accurate way to assess the consequences of matings observed between nonpair mates because behavioral observations can dramatically underestimate or overestimate the frequency of EPFs. For example, in Indigo Buntings, Passerina cyanea, only 3.3% of observed copulations were EPCs, yet 27-42% of young resulted from EPFs (Westneat 1987a, b). In other cases, the frequency of EPCs overestimates the frequency of EPFs. In a study of Willow Warblers, Phylloscopus trochilus, 13% of copulations were EPCs, but none of 120 offspring were determined to have resulted from EPFs using DNA fingerprints (Gyllensten et al. 1990). In yet other studies of avian populations, EPCs accurately predict EPFs, such as in the Shag, Phalacrocorax aristotelis (Graves et al. 1991).