Note: I’ve just rehashed this post (30/03/2011) because the paper is now available online (see comment stream). Stay tuned for the media release next week. – CJAB
I’ve been more or less underground for the last 3 weeks. It has been a wonderful break (mostly) from the normally hectic pace of academic life. Thanks for all those who remain despite the recent silence.
But I’m back now with a post about a paper we’ve just had accepted in Frontiers in Ecology and Environment. In my opinion it’s a leap forward in how we measure relative threat risk among species, despite some criticism.
I’ve written in past posts about the ‘magic’ minimum number of individuals that should be in a population to reduce the chance of extinction from random events. The so-called ‘minimum viable population (MVP) size’ is basically the abundance of a (connected) population below which random events take over from factors causing sustained declines (Caughley’s distinction between the ‘declining’ and ‘small’ population paradigms).
Up until the last few years, the MVP size was considered to be a population- or species-specific value, and it required very detailed demographic, genetic and biogeographical data to estimate – not something that biologists tend to have at their fingertips for most high-risk species. However, several papers published by our group (Minimum viable population size and global extinction risk are unrelated, Minimum viable population size: a meta-analysis of 30 years of published estimates and Pragmatic population viability targets in a rapidly changing world) have shown that there is in fact little variation in this number among the best-studied species; both demographic and genetic data support a number of around 5000 to avoid crossing the deadly threshold.
Now the fourth paper in this series has just been accepted (sorry, no link yet, but I’ll let you all know as soon as it is available), and it was organised and led by Reuben Clements, and co-written by me, Barry Brook and Bill Laurance.
The idea is fairly simple and it somewhat amazes me that it hasn’t been implemented before. The SAFE (Species Ability to Forestall Extinction) index is simply the distance a population is (in terms of abundance) from its MVP. In the absence of a species-specific value, we used the 5000-individual threshold. Thus,
SAFE = log10(N) – log10(MVPt)
where N = the species’ population estimate and MVPt = an empirically supported threshold MVP target (e.g., 5000).
Of course, when there is uncertainty in the MVP or N estimates, then these can be used to create confidence bounds on the SAFE.
We went a step further than just providing the formula – we showed that the SAFE is a better predictor of a mammal species’ IUCN Red List category (95 species) than percentage range loss (à la Ceballos & Ehrlich 2005).
Now, I’ll leave all the nitty-gritty detail to the paper itself (when it is available online), but I’ll leave with the final suggestion we make – the SAFE should provide a relative scale of threat risk for species within the same Red List category. We’re not suggesting that the Red List categorisations be abandoned – we just think that not all Critically Endangered species are equal, and some empirical metric of relative risk should be available. The SAFE does this.
Keen to hear your opinions. If you want a pre-print, just email Reuben and he’ll be able to send you one soon.
Brook, B.W., Traill, L.W., & Bradshaw, C.J.A. (2006). Minimum viable population sizes and global extinction risk are unrelated Ecology Letters, 9 (4), 375-382 DOI: 10.1111/j.1461-0248.2006.00883.x
Traill, L.W., Bradshaw, C.J.A., & Brook, B.W. (2007). Minimum viable population size: A meta-analysis of 30 years of published estimates Biological Conservation, 139 (1-2), 159-166 DOI: 10.1016/j.biocon.2007.06.011
Traill, L.W., Brook, B.W., Frankham, R., & Bradshaw, C.J.A. (2010). Pragmatic population viability targets in a rapidly changing world Biological Conservation, 143 (1), 28-34 DOI: 10.1016/j.biocon.2009.09.001