Decline in abundance of a Costa Rican poison-dart frog (Oophaga pumilio) following rapid loss of an understory plant (Araceae: Dieffenbachia sp.) used for tadpole rearing

Background/Question/Methods

As in many tropical habitats, amphibian populations in the rain forest of the La Selva Biological Station in Costa Rica have been declining in recent decades.  An apparent exception to this trend has been the stable population of the poison-dart frog Oophaga pumilio.  Because amphibian populations can decline rapidly, however, continual monitoring is necessary to ensure long-term viability.  We surveyed La Selva populations of O. pumilio and of the plants used by the frogs to rear tadpoles in 1998 and again in 2010.  In the 1998 survey, we sampled 50 plots of 100 m² located both in primary forest and in previously disturbed secondary forest (including abandoned agroforestry plantations).  We systematically searched the litter in each plot and counted the number of adult frogs.  We also counted all small pools of water that the frogs could use for rearing tadpoles.  The same areas were resampled in 2010, when we used calling O. pumilio males as an index of frog population size.  As in 1998, we also quantified the abundance of potential tadpole-rearing locations within plots. 

Results/Conclusions

There were 4.9 times more frogs in secondary forest than in primary forest in 1998; a similar pattern also has been observed in multiple other published studies from La Selva.  The large majority of tadpole-rearing sites in 1998 were in the leaf axils of Dieffenbachia sp., a large understory herb.  Dieffenbachia plants were much more abundant in secondary forest than in primary forest.  The results were much different 12 years later.  In 2010, there was no difference in frog abundance between primary and secondary forest.  Apparently the frog density in secondary forest had fallen to the lower level previously observed in primary forest.  There also was a dramatic decline in the population of Dieffenbachia between 1998 and 2010.  Dieffenbachia had completely disappeared from almost all locations where it was previously abundant, particularly in secondary forests.  We propose that the rapid loss of Dieffenbachia tadpole-rearing sites caused the attendant decline in O. pumilo populations in secondary forest.  A likely cause of the Dieffenbachia loss is consumption by collared peccaries (Pecari tajacu), which have increased in abundance in the last decade.  Thus O. pumilio could have declined as an indirect effect of the increase of a Dieffenbachia herbivore.