Estimation of dispersal distance by mark-recapture experiments using traps: correction of bias caused by the artificial removal by traps

Although in mark-recapture experiments traps are useful to estimate the dispersal distance of organisms, they cause a dilemma that may be called a kind of Heisenberg effect: a large number of traps should be placed to yield a precise estimate of mean dispersal distance, while these traps shorten the mean dispersal distance itself by intercepting organisms that should have dispersed for further distances. We propose a procedure to solve this dilemma by placing traps uniformly in a lattice pattern, and by assuming a random movement and a constant rate of settlement for organisms. We applied this procedure to estimate the dispersal distance of the sugarcane wireworm Melanotus okinawensis Ohira (Coleoptera: Elateridae). The estimated mean dispersal distance was 143.8 m. Through the use of a conventional method of estimation, the mean dispersal distance was estimated to be 118.1 m. Thus, it was shown that the conventional estimate of dispersal distance was 18% smaller than the corrected estimate in our experiment.