Amphisexual parental behavior of a terrestrial breeding frog Eleutherodactylus johnstonei in Guyana

Parental care in Eleutherodactylus johnstonei, a terrestrially breeding frog, was uniparental, but care was provided by either sex. I tested Maynard Smith's evolutionarily stable strategy, continuous breeding model III, by examining social and ecological conditions under which the sexes provided parental care in Georgetown, Guyana, South America. During the long wet season most males chosen by females as mates provided parental care. No males that acquired females through satellite behavior and displacement competition provided care. Females provided parental care in most cases when males failed to do so, and some returned to nests for 72 h, even after being repeatedly driven away by attending males. Females initiated maternal care up to 72 h after oviposition following temporary removal of males. Male-biased operational sex ratios were significantly different for wet and dry months, and no females provided care when monthly rainfall was <200 mm. Additionally, operational sex ratio multiply regressed on paternal and maternal care and rainfall indicated that paternal and maternal care were significantly correlated with operational sex ratio. Rainfall was not related to operational sex ratio. During the dryer months only males provided care. Males provided less care during wetter months, when more gravid females were available as mates. Significant mortality of eggs and embryos occurred in clutches with no parental care, and females and males did not differ in the ability to protect progeny. Although there was support for Maynard Smith's model of evolution of amphisexual care from biparental care, it was difficult to ascertain clearly whether amphisexual parental behavior evolved from either biparental behavior or no care. This uncertainty ensues because the behavior of females and males and development of progeny may have become coadapted after initial evolution of care in E. johnstomei.