Assessing Rates of Increase from Trend Data

Wildlife researchers often use trend data to estimate rates of increase. Consequently, we simulated the growth of populations of 3 species of mammals to assess the accuracy and precision of methods that estimate rates of change from population trend data. Our results indicate that the record of trend of an individual population may provide a relatively unbiased estimate of the underlying rate of change, but it does not appear to contain the information needed to make statements about the likely behavior of similar populations, or to compare trends in different populations. Projections of future trend of an individual population may be best obtained by simple linear regression on logarithms of counts. Confidence limits on such projections were satisfactory for short-term projections in some examples we provide.