Extensive theoretical work on demographic Allee effects has led to the latent assumption that they are ubiquitous in natural populations, yet current empirical support for this phenomenon is sparse. We extended previous single-taxon analyses to evaluate the empirical support for demographic Allee effects in the per capita population growth, rate of 1198 natural populations spanning all major taxa. For each population, we quantified the empirical support for five population growth models: no growth (random walk); exponential growth, with and without an Allee effect; and logistic growth, with and without an Allee effect. We used two metrics to quantify empirical support, information-theoretic and Bayesian strength of evidence, and observed top-rank frequency. The Ricker logistic model was both the most supported and most frequently top-ranked model, followed by random walk. Allee models had a combined relative support of 12.0% but were top-ranked in only 1.1 % of the time series. Accounting for local climate variation and measurement error caused the loss of topranked Allee models, although the latter also increased their relative support. The 13 time series exhibiting Allee models were shorter and less variable than other time series, although only three were non-trending. Time series containing observations at low abundance were not more likely and did not show higher support for Allee effect models. We conclude that there is relatively high potential for demographic Allee effects in these 1198 time series but comparatively few observed cases, perhaps due to the influences of climate and measurement error.
- Population dynamics
- demographic Allee effect
- random walk
- Bayesian information criterion (bic)
- Empirical support
- Akaike information criterion (aic)
- Demographic allee effect