Number of censuses required for demographic estimation of effective population size

Document Type


Publication Date



Adequate population viability assessment may require estimation of effective population size (Ne). But failure to take into account the effect of temporal fluctuation in population size (FPS) on N(e) may routinely lead to unrealistically optimistic viability assessments. We thus evaluate a technique that accounts for the effect of FPS on N(e). Using time series of annual counts of 48 free-ranging animal populations, we show that N(e) is dependent on timescale: as more census records are incorporated, estimates of FPS tend to increase, and thus estimates of N(e) tend to decrease. Estimates based on, say, 10 annual counts tend to grossly underestimate the influence of FPS on estimates of long-term N(e), often by more than 100%. Moreover, a newly derived expression for the confidence intervals of N(e)/N (in which N is population size) reveals that estimates of N(e) based on only a few annual counts are quite unreliable. Our work thus emphasizes the need for long-term population monitoring and provides a framework for interpreting N(e) estimates calculated from limited census data.

Publication Title

Conservation Biology