Point counting has been widely accepted as a method for monitoring trends in bird populations. Using a rigorously standardized protocol at 210 counting stations at the San Joaquin Experimental Range, Madera Co., California, we have been studying sources of variability in point counts of birds. Vegetation types in the study area have not changed during the 11 years of the study, so annual variability in counts must reflect some normal range of variation associated with other factors. Here we show that counts of breeding species varied markedly from year to year, we suspect primarily in response to annual variation in precipitation and secondarily to variation in temperature. With ample data, practitioners can examine various periods of time for creating running averages to smooth background levels of annual variation in counts, thus establishing a baseline abundance for comparative purposes. We have explored running averages in our data, based on 2- to 7-year periods, with the objective of attaining a coefficient of variation of 10% or less in mean annual counts. Twenty-two of 34 species met this criterion, based on 7-year running averages. Linear and exponential extrapolations from observed data suggested that 11-22 years would be needed to obtain baseline data for all breeding species in our dataset.
|
