Application of a foraging model to estimate fish movement rates from trap data

Background/Question/Methods      We hypothesize that long-hydroperiod wetlands are source populations of fishes and short-hydroperiod wetlands are sink populations, therefore creating different movement patterns between the two hydrological regimes.  Unfortunately, traditional methods for monitoring fish movement, such as telemetry and mark-recapture, are not readily applicable to small species.  Drift fences, a passive sampling device similar to a fyke net, give an estimate of encounter rate by capturing moving fishes over a given sampling time.  Our drift fences used four arms (12 m in length) radiating in an x pattern, with minnow traps placed in the intersection of the arms to capture fishes from four compass directions.  We examined the use of a foraging model to estimate fish movement rates by treating drift fences as a sit-and-wait predator and with an independent estimate of density obtained from throw trap data.  We modeled four years of density and encounter data from long-hydroperiod sites in Everglades National Park to estimate seasonal patterns of movement.  Using underwater videos, we estimated bias from trap avoidance and escape from the minnow traps and incorporated a correction factor to our movement rate estimates. The Everglades is a seasonally flooded wetland, where hydrology dictates fish distribution patterns. 

Results/Conclusions      Movement rate decreased during the later stages of the dry season (February and April), followed by a sharp increase throughout the wet season (July through December).  Several of our study sites were located in dry-season refuges, which acted as recipient sites for fishes escaping short-hydroperiod marshes during dry-season conditions.  In future work, we will study the effects of hydrology in a short-hydroperiod (sink population) on fish movement patterns, where we expect to see two peaks of movement at the start of the wet and dry seasons.