Age-dependent dispersal and habitat choice in black-tailed godwits (<i>Limosa l. limosa</i>) across a mosaic of tradinional and modern grassland habitats

Whether to disperse, and where to, are two of the most prominent decisions in an individual ’ s life, with major consequences for reproductive success. We studied natal and breeding dispersal in the monogamous black-tailed godwit Limosa limosa limosa in the Netherlands, where they breed in agricultural grasslands. The majority of these grasslands recently changed from wet herb-rich meadows into well-drained grassland monocultures, on which godwits have a lower reproductive success. Here we examine habitat selection with a multistate mark – recapture analysis. Habitat transition probabilities between meadows and monocultures were estimated on the basis of 1810 marked chicks and 531 adults during seven years in a 8500 ha study area. Young and adult godwits may differ in habitat selection because: 1) adults may have gained experience from previous nest success where to settle, 2) younger individuals may find it harder to compete for the best territories. Both young and adults moved at a higher rate from the predominant monocultures to meadows than the other way around, thus actively selecting the habitat with better quality. However, dispersal distance of adults was not affected by previous nest success. The average dispersal distance from place of birth of godwits breeding for the first time was ten times larger than that of adult godwits. That godwits breeding in their second calendar year arrived and laid at similar dates and were equally able to select territories in areas with high breeding densities, suggests that young birds were not competitively inferior to adults. Although on monocultures reproduction is insufficient to maintain constant populations, birds sometimes moved from meadows to monocultures. This explains why even after 30 years of land-use intensification, godwits still breed in low-quality habitat. The adjustment to changing habitat conditions at the population level appears to be a slow process