Power to Detect Trends in Waterbird Abundance with Call-Response Surveys

We broadcast calls of grebes, rails, and bitterns, and passively counted other species, along 36 survey routes in Maine wetlands during 1989-92. Our objectives were to determine the magnitude and temporal variation in waterbird call-response rates, and to estimate, through computer simulation, the statistical power of variously configured monitoring programs. Average call-response rates on survey mini-routes ranged from 2.8 to 6.0 responses per 10-station mini-route, and temporal variation in counts was relatively high (CV in all species >80%). A power analysis, performed by coupling Monte Carlo methods with linear regression, indicated that changes in call-response rates of 10% would be difficult to detect over a 10-year monitoring period in any species. Changes of 25%, however, could be detected reliably (P > 0.90) by surveying 40-80 mini-routes on 2-3 occasions annually. Passive counts at call-response stations yielded useful abundance information on many wetland species currently undersampled by existing avian monitoring schemes. We conclude that call-response surveys for selected species, coupled with passive counts of other species, are a logistically feasible and statistically defensible means of monitoring trends in regional populations of wetland birds.