干旱区生态环境知识资源中心

1. Field studies of darkling beetles (Coleoptera: Tenebrionidae: Eleodes spp.) in a semi-arid grassland were coupled with simulation modelling to determine how habitat-specific movements of individuals influenced rates of capture in pitfall traps and estimates of population size.

2. The relationships among individual movement, trap encounter and trap design were examined using three geometric arrangements of pitfall traps: a transect (a line), a rectangular grid (a filled plane) and a Sierpinski gasket (a partially filled triangle).

3. Trap arrays were located in areas differing in vegetation structure. For each trap array, we compared mark-recapture data on the most abundant beetle (Eleodes obsoleta) with rates of capture and estimates of population size predicted by simulations.

4. Simulations of movement pathways were generated by random draws from empirical distributions of movement distances and turning angles, and trap encounter rates within trap arrays were determined from numerous simulated pathways. Population size was estimated by the simulated number of individuals required to approximate the number of captures in the field.

5. Densities obtained from simulations and mark-recapture were similar in areas dominated by grass and bare ground, but simulated densities were greater than mark-recapture estimates in areas with more cactus and shrub cover. Mark-recapture may underestimate actual densities in cactus-shrub areas because beetles have small net displacements and thus low trap-encounter rates.

6. A field validation of the simulation model was conducted by comparing simulated and observed capture rates in trap arrays with and without drift fences. The model correctly predicted the direction of change in capture rates, but the magnitude of the predictions differed from the observed in two of four trap arrays located in different habitats.

7. Numbers of captures were also simulated for a wide range of movements using a series of correlated random walks. Capture probability was related to net displacement through an exponential saturation function with parameters that varied according to trap geometry and the spatial confinement of individuals to the trap area. The capture probabilities obtained from simulations of field data on beetles were consistent with those derived from theoretical walks.