Of power and despair in cetacean conservation: estimation and detection of trend in abundance with noisy and short time-series.

Matthieu Authier,Anita Gilles,Jérôme Spitz,Anders Galatius

doi:10.7717/peerj.9436

Abstract

Many conservation instruments rely on detecting and estimating a population decline in a target species to take action. Trend estimation is difficult because of small sample size and relatively large uncertainty in abundance/density estimates of many wild populations of animals. Focusing on cetaceans, we performed a prospective analysis to estimate power, type-I, sign (type-S) and magnitude (type-M) error rates of detecting a decline in short time-series of abundance estimates with different signal-to-noise ratio. We contrasted results from both unregularized (classical) and regularized approaches. The latter allows to incorporate prior information when estimating a trend. Power to detect a statistically significant estimates was in general lower than 80%, except for large declines. The unregularized approach (status quo) had inflated type-I error rates and gave biased (either over- or under-) estimates of a trend. The regularized approach with a weakly-informative prior offered the best trade-off in terms of bias, statistical power, type-I, type-S and type-M error rates and confidence interval coverage. To facilitate timely conservation decisions, we recommend to use the regularized approach with a weakly-informative prior in the detection and estimation of trend with short and noisy time-series of abundance estimates.

Highlights

Ecologists have long strived for power, often of the statistical kind (Gerrodette, 1987; Link & Hatfield, 1990; Thomas, 1996; Seavy & Reynolds, 2007; White, 2018)
The unregularized approach had a Type-I error rate of at least 10% when the significance level was set to 5%; and a Type-I error rate of at least 30% when the significance level was set to 20%
The regularized approach with a weakly-informative prior had a Type-I error rate less than 10% when significance was set at 5%, and close to 5% for coefficient of variation (CV) less

Summary

Introduction

Ecologists have long strived for power, often of the statistical kind (Gerrodette, 1987; Link & Hatfield, 1990; Thomas, 1996; Seavy & Reynolds, 2007; White, 2018). The issue of low statistical power to detect change in time-series of population abundance estimates arose early on (Anganuzzi, 1993), with obvious, and sometimes dire, consequences for applied conservation. Of power and despair in cetacean conservation: estimation and detection of trend in abundance with noisy and short time-series. While blaming statistical power for the vaquita’s quiet vanishing out of the Anthropocene would be excessive (see Bessesen (2018) for an overview of the vaquita case), we think that it illustrates how ecologists may have painted themselves into a corner in their insistence for statistical ‘orthodoxy’ inherited from the uneasy wedding of Fisherian (statistical significance) and Neyman–Pearsonian (type-I and type-II errors) philosophies (Hubbard & Bayarri, 2003; Christensen, 2005)

Methods

Results

Discussion

Conclusion