Ineffectiveness of Call Surveys for Estimating Population Size in a Widespread Neotropical Frog, Oophaga pumilio
Call surveys offer a valuable method to monitor anuran populations attributable to their temporal breeding habits and close association with water. Many temperate locations have adopted citizen science programs to monitor local anuran populations using call surveys and road transects. These surveys, however, are not commonly conducted in the tropics. I tested use of call surveys for estimating population density of a small terrestrial poison frog, Oophaga pumilio, in six different populations in Bocas del Toro, Panama. By conducting three-minute call surveys, and searching for all individual frogs in a 10 m radius of the survey point, I directly compared the number of calling males to the observed number of frogs in a given area. I found call density to be a poor predictor of population density. Despite there being differences in population densities, the lack of a relationship between the number of calls and population density highlights the limited use for call surveys for terrestrial, territorial species. Although call surveys may be useful in some taxa for general abundance estimates, this study clearly demonstrates a startling deficiency of call surveys for anuran monitoring and highlights the need for species-specific analysis to further explore the utility of this method.Abstract
Map of the different sampling locations for six different populations in Bocas del Toro, Panama. Populations were 1) Isla Colon, 2) Loma Partida, 3) Pastores, 4) Popa North, 5) San Cristobal, and 6) Solarte. Five surveys were done on each island. Surveys were at least 100 m apart. Because of restriction of suitable habitat on Solarte, two surveys were conducted approximately 1.5 km away from the other three. Compared to the first site, this second site had a high abundance of frogs.
Effects plots of the final mixed-effects model for (A) log10-transformed relative humidity and (B) population density in relation to standardized call rate. Call rates were standardized by dividing the observed call density by the average male snout–vent length of the representative population to remove potential impacts body size has on calling intensity.
Relationship of the proportion of males and population density for Isla Colon, San Cristobal, Solarte, Loma Partida, and Pastores. Proportion of males was the total number of apparent males (identified by gular patch) to total number of frogs found in each survey plot. Regressing male density against population density yielded a negative relationship (R2-value of 0.324; P = 0.0029), possibly suggesting alternative reproductive strategies (i.e., acoustic niche partitioning, satellite males) in Oophaga pumilio.
Contributor Notes
Present address: Department of Biology, University of Mississippi, Box 1848, Oxford, Mississippi 38655, USA; E-mail: jlawrenc@go.olemiss.edu
