Stopover Strategies in Passerine Bird Migration: A Simulation Study

Abstract

Long-distance bird migration consists of a series of stopovers (for refuelling) and flights, with flights taking little time compared to stopovers. Therefore, it has been hypothesized that birds minimize the total time taken for migration through efficient stopover behaviour. Current optimality models for stopover include (1) the fixed expectation rule and (2) the global update rule. These rules maximize the speed of migration by determining the optimal departure fuel load for a given fuel deposition rate. We were interested in simple behavioural rules approaching the stopover behaviour of real birds and how these rules compare to the time minimizing models above with respect to the total time taken for migration. The simple strategies were to stay at a site (1) until a fixed fuel load was reached or (2) for a constant number of days. We simulated migration of small nocturnal passerine birds across an environment of continuously distributed but variable fuel deposition rates, and investigated the influence of different stopover strategies on the duration of migration. Staying for a constant number of days at each stopover site, irrespective of the fuel deposition rate, resulted in only slightly longer than minimum values for migration duration. Additionally, the constant stopover duration, e.g. 10 days, may change by a day or two (per stopover) without having a large effect on total migration duration. There is therefore a possibility that real birds may be close to optimal migration speed without the need for very complex behaviour. When assessing the sensitivity of migration duration to factors other than stopover duration, we found that flight costs, search and settling time, mean fuel deposition rate and the accuracy in the choice of flight direction were the factors with the largest influence. Our results suggest that migrating birds can approximate optimal stopover duration relatively easy with a simple rule, and that other factors, e.g. those above, are more relevant for travel time.