A guide to the provision of refuges for waders: An analysis of 30 years of ringing data from the Wash, England

1. It has been proposed to establish roost refuges to limit the disturbance to internationally important over-wintering populations of waders on the Wash. 2. Waders have been caught regularly on the Wash since 1959. By 1993, 24 576 oystercatchers, 4125 grey plovers, 38 041 knots, 96 801 dunlins and 11 729 redshanks had been ringed at 85 roost sites around the Wash. 3. Analysis of the movements between roosts provides a means of determining the relative mobility of waders End the effect of spacing on the availability of the proposed refuges to waders during normal roost movements. This analysis does not take into account either movements between feeding sites nor between roosting and feeding sites. 4. The Wash was divided into five sections according to observed wader movements between roosts and feeding grounds. At least 79% of inter-roost movements of oystercatchers, grey plovers, knots, dunlins and redshanks within a year were restricted to a single section. Juvenile waders changed section more frequently than adults. Waders were almost as site faithful between-year as within-year. 5. A model, based on the variation in retrap rate with distance of the three least mobile wader species within a section of the Wash, was used to describe wader dispersal between roosts. The model was then used to estimate the effect of inter-refuge distance on the proportion of waders reaching at least one refuge during normal roost movements. 6. Between roosts waders dispersed over short distances within a section of the Wash. Placing refuges 2, 4 and 7 km apart for grey plovers, 2.5, 5 and 10 km apart for dunlins within-year, and 3.5, 5.5 and 9.5 km apart for redshanks put the refuges within reach of 90, 75 and 50% of the respective populations of these species during their normal roost movements. To validate the sectional approach, the analysis was repeated using roosting movements of dunlins over another section and over the whole Wash; the results were similar to those obtained in the original analysis. 7. The implications of wader dispersal for refuge creation are discussed.