Homing in the Rubí Poison Frog Andinobates bombetes (Dendrobatidae)
Homing is a behavior in which an animal returns to a specific place after they have moved or migrated to a distant place. In anurans, most of our knowledge about homing comes from studies in temperate-region species with nocturnal activity and reproduction associated with ponds. Recently, studies with poison frogs (Dendrobatidae) have increased our understanding about homing in tropical frogs with diurnal activity, and that do not breed in large ponds. The Rubí Poison Frog Andinobates bombetes offers a good opportunity to further increase the knowledge of behavioral ecology of homing in anurans because some natural history traits in this species differ from those exhibited in most poison frogs in which homing ability has been studied. For instance, A. bombetes have a smaller body size and use phytotelmata in bromeliads for tadpole development while others use terrestrial pools. To quantify the homing ability and the factors influencing it in A. bombetes, we performed translocation experiments of individuals at distances between 5 and 90 m outside their territory in a forest remnant located in the department of Quindío, Central Andes of Colombia. In this study, we included a large sample size of females, which is important because homing studies with poison frogs has been almost exclusively studied in territorial males. Of 104 displaced individuals, 39 returned to their territory. The probability of homing in A. bombetes was negatively related to the translocation distance, but was unrelated to body size and sex. Apparently, this species has a limited homing ability when compared to most poison frogs studied so far except for O. pumilio, which seems more similar in body size and resources used for reproduction. Overall, homing ability appears to be widely shared in the family Dendrobatidae, Andinobates being the fourth genus of this family for which homing ability has been corroborated experimentally.
Image of an adult male Rubí Poison Frog Andinobates bombetes (Dendrobatidae) carrying three tadpoles on its back.
Aerial image and geographic location of the study area (El Placer, municipality of Filandia, Quindío) in the Central Andes of Colombia, South America. The red line indicates the approximate location and direction of the sampling plot in the forest remnant in which translocation experiments with the Rubí Poison Frog A. bombetes were performed.
Example of the variation in patterns of ventral coloration in A. bombetes used for identifying specific individuals. Illustrations by M. J. Tovar-Gil.
Number of individuals of A. bombetes translocated (red bars) and the number of these individuals that were then recaptured (black bars) based on distances of the experimental translocations.
Predicted probability of homing based on translocation distance, body size, and sex (36 males, 23 females) of the frog Andinobates bombetes (A). In this plot, the variation in the probability of homing at the same translocation distance is due to body size differences among individuals. Given that body size and sex were unrelated to homing probability, we performed a new analysis without including such predictive/independent variables (B). A similar inverse relationship between translocation distance and probability of homing was obtained when we included in the analysis 45 additional unsexed individuals (C). The values on the y-axis (probability of homing) for the three plots are calculated using a log-odds mathematical approach based on the actual recorded values for a binary dependent variable (i.e., 1 = frog returned, 0 = frog did not return).
Contributor Notes