Measuring dispersal and detecting departures from a random walk model in a grasshopper hybrid zone

1. The grasshopper species Chorthippus brunneus and C. jacobsi form a complex mosaic hybrid zone in northern Spain. Two mark-release-recapture studies were carried out near the centre of the zone in order to make direct estimates of lifetime dispersal. 2. A model framework based on a simple random walk in homogeneous habitat was extended to include the estimation of philopatry and flying propensity. Each model was compared with the real data, correcting for spatial and temporal biases in the data sets. 3. All four data sets (males and females at each site) deviated significantly from a random walk. Three of the data sets showed strong philopatry and three had a long dispersal tail, indicating a low propensity to move further than predicted by the random walk model. 4. Neighbourhood size estimates were 76 and 227 for the two sites. These estimates may underestimate effective population size, which could be increased by the long tail to the dispersal function. The random walk model overestimates lifetime dispersal and hence the minimum spatial scale of adaptation. 5. Best estimates of lifetime dispersal distance of 7-33 m per generation were considerably lower than a previous indirect estimate of 1344 m per generation. This discrepancy could be influenced by prezygotic isolation, an inherent by-product of mosaic hybrid zone structure.