Print PageDo the controls of population dynamics vary when examined at different spatial scales? If local controls vary across a landscape inhabited by a regional population, controls at the regional level may overwhelm local dynamics and have a greater impact on long-term persistence. In particular, predators commonly have a larger dispersal distance than prey and may stabilize local population dynamics by invading and swamping local influences. And yet, few studies attempt to separate local and landscape controls of population dynamics. I report a study of the controls of population dynamics of two species of small fish, bluefin killifish and least killifish, in a large, spatially complex, wetland that experiences annual fluctuation in habitat size. This fluctuation of water level forces small fish and their predators to seek hydrological refuge to avoid stranding. I analyzed 14-year time series of density data from 11 study sites scattered over the southern half of the Florida Everglades, USA, to test for density dependence. Fish density was estimated by use of 1-m2 throw traps and was accompanied by estimates of relative density of large piscivorous fish (catch per unit effort) made with airboat-mounted electrofishing. Hydrological parameters indicating the days since the local site was most recently reflooded following a drought and water depth at the time of sampling were included in models evaluated. I compared these results to models of data from the most recent 6 years sampled with a more extensive spatial coverage, but less frequent temporal sampling (4 samples per year for the long-term data and 1 sample per year for the spatially extensive data).
Results/Conclusions
Analyzing local population dynamics, I found that the strength of negative feedback on population growth depended on the return frequency of droughts, with more frequently drying sites displaying stronger intraspecific density feedback. I hypothesized that drying concentrated fish and elevated local biotic interactions, contrary to conventional predictions that disturbed populations would experience less density-dependent control. Bluefin killifish became denser as water level receded, while least killifish displayed the opposite pattern. The more spatially extensive analysis indicated a greater role for predators in controlling population dynamics than the analysis of local population dynamics and the hydrological signal was reduced, probably because these data were limited to wet-season samples. This analysis supports the idea that landscape controls may be different than local controls and may provide more information about population persistence, particularly in disturbed or fragmented landscapes.
