1. Recent theoretical and empirical research has shown that habitat isolation can have strong effects on patterns of abundance and population dynamics of animals. Effects of habitat patch isolation may be particularly strong in taxa such as amphibians, which may have limited dispersal abilities and yet rely on breeding habitats that are variable in space and time.
2. We manipulated the distribution of replicated artificial ponds in order to determine the effects of breeding pond isolation on the spatial and temporal dynamics of pond use by the tungara frog (Physalaemus pustulosus Lynch), on Barro Colorado Island (BCI), Panama. Our principal objectives were to determine whether pond isolation affects if, when, and how much a pond will be used, and to determine the spatial scale over which isolation effects are likely to be important.
3. We used three sets of experimental ponds: ponds placed at distances from 1 to 200 m from active natural breeding sites; ponds placed at these same distances from recently abandoned breeding sites; and ponds placed among different terrestrial habitat types and at least 200 m from any known breeding sites (‘sampling ponds’).
4. We found that time to pond use increased with pond isolation, and that ponds near active sites were used more quickly than ponds near abandoned sites. The number of egg masses per pond decreased with pond isolation and was significantly lower near abandoned sites. Ponds near active breeding sites were more likely to be used by tungara frogs than both ponds near abandoned sites and sampling ponds.
5. Used sampling ponds were clumped at spatial scales less than 200 m, and mark–recapture data suggest that tungara frogs routinely move between ponds over these distances.
6. We conclude that breeding pond isolation is a primary determinant of the temporal and spatial dynamics of pond use by tungara frogs, and can result in spatial patterns in pond use that are independent of pond quality. These isolation effects are, however, unrelated to metapopulation processes, as most turnover in pond use is probably a result of dispersal and breeding site selection by individuals, and not resulting from the extinction of local populations.