Validating the use of constant effort mist-netting to monitor avian populations

Nur, Nadav; Geupel, Geoffrey

The four objectives of this report are: to examine mist-net measures of adult survival that rely on state-of-the-art statistical analysis, to examine the assumption that birds caught are a random sample of those alive, to determine whether recapture probability of birds in subsequent years is high, and to make recommendations for the conduct of constant-effort mist-netting. Constant-effort mist-netting has been proposed as a means of obtaining information on demographic processes of avian populations. To determine whether data gathered from a standardized regime of mist-netting yields valid results regarding adult survival, to determine the probability of capture and recapture of individuals, and to determine whether these probabilities might be biased by one or more factors, the Point Reyes Bird Observatory analyzed data on Wrentits gathered over eleven years. On the study site, at Palomarin Field station, Point Reyes National Seashore, Wrentits were mist-netted, nests of all breeding birds were located and their young banded, and the adults were individually color-banded and their territories determined. Capture probability was strongly related to distance from the nest to the nearest mist-net, for both juveniles and adults. Breeding Wrentits were seldom caught more than 200m from their nest, but juveniles were regularly caught to 700m and beyond. The catchment area for juveniles likely extends to several kilometers. The probability a juvenile was caught declined with fledging date. Juveniles were almost always first caught 20 to 60 days after fledging. Most adults caught were non-breeders. Non-breeders were very transient; consequently recapture probability in the following year was very low (estimated to be 4%, using the statistical program SURGE). In contrast, recapture probability was high for breeders (71%). There was no evidence for net-avoidance. Most breeders were caught three times or more during the same season, most non-breeders were never recaptured. Age, reproductive success and sex did not influence capture probability of breeders; the earliest breeding Wrentits were least likely to be caught. Analyzing capture/recapture data from mist-nets with SURGE and differentiating breeders and non-breeders, yielded a survival estimate of 57.4% for breeders, a value close to those obtained from re-sightings of color-banded birds in a subsequent year (59% to 64%). However, pooling data from breeders and non-breeders, yielded a survival estimate of 31% and recapture probability of 38%. The aberrant value for survival is due to non-breeders (with very low recapture probability) being lumped with breeders (whose recapture probability was high). We present recommendations for further work and to guide monitoring programs. In particular, we stress the importance of attempting to distinguish breeders from non-breeders if one wishes to infer survival from capture/recapture data. We suggest that restricting the analysis to individuals caught more than once in the same season may be helpful in this regard.