The use and abuse of population viability analysis

Abstract

A recent study by Brook et al. empirically tested the performance of population viability analysis (PVA) using data from 21 populations across a wide range of species. The study concluded that PVAs are good at predicting the future dynamics of populations. We suggest that this conclusion is a result of a bias in the studies that Brook et al. included in their analyses. We present arguments that PVAs can only be accurate at predicting extinction probabilities if data are extensive and reliable, and if the distribution of vital rates between individuals and years can be assumed stationary in the future, or if any changes can be accurately predicted. In particular, we note that although catastrophes are likely to have precipitated many extinctions, estimates of the probability of catastrophes are unreliable.Population viability analysis (PVA) is a modelling tool that estimates the future size and risk of extinction for populations of organisms 1,2 . PVA works by using life-history or population growth-rate data to parameterize a population model that is then used to project dynamics and estimate future population size and structure 3 . User-friendly PVA software packages allow conservation managers to predict future population sizes and risks of extinction for any population they choose 3 . Because of this ease of application of PVAs, it is important to determine and understand the limits to their predictive accuracy 1,4-6 . Brook et al. have tested the predictive accuracy of PVA using data from many populations and conclude that PVA is not a useless tool, and that it should not be dispensed with in favour of alternative untested methods. Do PVAs work?The predictive accuracy of a PVA will depend on the purpose to which it is being applied. In practice, there has been a range of alternative uses. PVAs can be used to: (1) predict the future size of a population 1,5,6 ; (2) estimate the probability of a population going extinct over a given time 5 ; (3) assess which of a suite of management or conservation strategies is likely to maximize the probability of a population persisting 7 ; and (4) explore the consequences of different assumptions on population dynamics for small populations 8 . In reality, only the predictive accuracy of the first two cases is estimable, as there are rarely sufficient replicate populations from which to collect data to determine whether the comparative predictions of the third use are accurate, and the fourth use has not generated testable predictions.There are two ways that the predictive accuracy of PVAs can be assessed. The first approach is to use historical data and, at a point in the future, predict the population size and compare this to what actually happened. To avoid circularity, the data used to parameterize the model should not include data from the time-period over which predictions are made. The whole population exists as a suite of spatially structured local populations) can then be compared to the distribution of predicted population sizes from the model projections...

Population resilience of the Mediterranean monk seal Monachus monachus at Cabo Blanco peninsula

“…Despite the high probability of extinction due to demographic and environmental stochasticity (Lande 1993, Courchamp et al 1999, Coulson et al 2001, Berec et al 2007, Melbourne & Hastings 2008, small populations maintain the potential for fast growth once conditions improve. The demographic parameters responsible for recovery are often unknown.…”

“…Small populations are at greater risk of extinction than large ones as a result of environmental variation or demographic stochasticity and processes, such as the Allee effect (Lande 1993, Mangel & Tier 1994, Courchamp et al 1999, Coulson et al 2001, Gerber & Hilborn 2001, Berec et al 2007, Good et al 2008, Melbourne & Hastings 2008. These general conclusions raise concerns for the conservation of small populations.…”

Repeatability and Reproducibility of Population Viability Analysis (PVA) and the Implications for Threatened Species Management

“…We wanted to test PVA models with the best data so we focused our data collection on species that included well-known keystone species (e.g., wolves), species involved in tourism (e.g., whale sharks), or species involved in subsistence or commercial hunting (e.g., dugongs)(n = 148), to maximize the potential for repeating these models. The demographic data on these species tends to be more extensive and as a result, PVA models and the subsequent population predictions are more robust (Brook et al, 2000;Coulson et al, 2001;Gordon et al, 2004).…”

European quaternary refugia: a factor in large carnivore extinction?

“…Retrospective tests using data from long-term studies of vertebrate populations suggests that its predictions are 'surprisingly accurate' at least for well-studied populations (Brook et al, 2000). Even critics of such approaches appear happy that PVAs are suitable for 'exploring theoretically the implications of model assumptions on extinction probabilities' (Coulson et al, 2001). This is the way we have used VORTEX in the current paper.…”