Towards developing thresholds for waterbirds that take into account turnover

摘要: To attain international importance and thus protection as a Ramsar site or Special Protection Area (SPA) wetland must either “regularly” support at least 20,000 waterbirds seabirds, 1% of the individuals population species subspecies waterbird. In most cases, sites have been designated by using maxima individual counts. These counts will underestimate volume (i.e. total number) birds passing through if turnover occurs. Using count data, observations individually marked survival recruitment mark-recapture models, we present three different methods (V1, V2 & V3) implemented in StopOver Duration Analysis SODA program (Choquet Pradel 2007) for estimating site. We use simulated data to determine their performance both biased unbiased data. Specifically, tested whether estimates were where following parameters varied: proportion marked, daily resighting rate, timing arrival, transients population, heterogeneity rates some with high low rate), variation arrival stopover time error. With relatively simple dataset (single no biases), had little effect on reliability resulting V1 V3. Estimates from always overestimated. The major factor that caused small positive bias V3 was probability. Lower probabilities estimates. Resighting more likely be seen than others) substantial all estimators. Transience stopping over shorter V3, but strong negative V2.Transience seemed reduce due when present. trap-dependent models those allow differential recapture rates) showed promise estimate observed there moderate amount individuals’ rates. V1 performed well under scenarios varying duration error introduced. consistently (see Table 4.1) The method highest precision; it is recommend volume. It important goodness fit tests are used biases appropriate Program SODA. Although estimates, care taken setting up study bias. eight ways ensuring reduced during collection Practical deal discussed. Recommendations section 4.2 further details) to: (i) Count same reading colour rings; (ii) approximately one-third length stay interval, e.g. thought ten days passage then every 5 days; (iii) aim resight > 30 period, although preferably more; obtain far possible representative samples being studied; (iv) marking species, number included,