How fast are we losing species anyway?

28 03 2011

© W. Laurance

I’ve indicated over the last few weeks on Twitter that a group of us were recently awarded funding from the Australian Centre for Ecological Synthesis and Analysis – ACEAS – (much like the US version of the same thing – NCEAS) to run a series of analytical workshops to estimate, with a little more precision and less bias than has been done previously, the extinction rates of today’s biota relative to deep-time extinctions.

So what’s the issue? The Earth’s impressive diversity of life has experienced at least five mass extinction events over geological time. Species’ extinctions have kept pace with evolution, with more than 99 % of all species that have ever existed now gone (Bradshaw & Brook 2009). Despite general consensus that biodiversity has entered the sixth mass extinction event because of human-driven degradation of the planet, estimated extinction rates remain highly imprecise (from 100s to 10000s times background rates). This arises partly because the total number of species is unknown for many groups, and most extinctions go unnoticed.

So how are we going to improve on our highly imprecise estimates? One way is to look at the species-area relationship (SAR), which to estimate extinction requires one to extrapolate back to the origin in taxon- and region-specific SARs (e.g., with a time series of deforestation, one can estimate how many species would have been lost if we know how species diversity changes in relation to habitat area).

The problem with SARs is that they are generally poorly parameterised, uncertain, and until rather recently, have failed to take the composition and configuration of remaining habitat types post-deforestation into account (solved cleverly though by Lian Pin Koh and colleagues in a few recent papers; Koh et al. 2010; Koh & Ghazoul 2010). One remaining issue – and I have to play a little cloak-and-dagger here because of an existing media embargo – is that the sampling used to construct an SAR doesn’t necessarily mean the the same statistical structure is maintained when extrapolating back down the curve. A little bird I know has told me that the problem has been solved and will be published shortly in Nature. Thankfully, that little bird is part of our ACEAS Working Group.

So why is this of any real interest to conservation per se, other than a more precise chronicling of the now full-blown Anthropocene Sixth Mass Extinction?

Species-area relationship models have been instrumental in estimating extinction rates but, like correlative species distribution models, they have been haunted by the assumptions of out-dated statistical techniques. Recent developments and those we propose include the influences of edge effects and matrix habitats. In the same way that metapopulation species distribution models better predict area of occurrence and thus extinction threat for individual species, new species-area relationship models promise to provide more ecologically realistic estimates of extinction rates: knowledge that will have a profound impact on policy.

More accurate and precise models of extinction rates allow policy makers, scientists and conservationists to explore a whole variety of other questions concerning land-use options. Whereas the conventional species-area relationship could be used to predict extinctions based only on the amount of forest loss, better models such as updated matrix-calibrated species-area relationships over a broader range of species could be used to explore the consequences of changes over entire landscapes, including not just the amount of forest but also the quality of the resultant land uses that now comprises the matrix. Not only will stakeholders be able to make more accurate predictions of biodiversity loss from land-use change, they could also use our models to predict potential biodiversity enhancements from improvements in the quality of the matrix. This is a major step forward because we could potentially prescribe measures to increase biodiversity.

The key to dealing with the conflicts of interest between environmental conservation and development in many emerging economies is the ability to evaluate trade-offs among different land-use and development options. This will be another major outcome of our modelling development. Conservationists will be able to use our tools to explore questions such as the trade-offs in biodiversity loss between converting forested landscapes to extensive but wildlife-friendly agriculture, versus converting a portion of the forest to intensive production systems. These practical questions cannot still be easily addressed with the conventional species-area relationship estimates of extinction.

I should mention finally that we’re not focussing only on SARs – we’ll be looking at a lot of species occurrence/disappearance datasets as well and relate these better to deep-time extinctions (see Regan et al. 2001 for an example).

The group is, I must say, world-class, and I’m proud to call them my colleagues (some I’ve worked with before, others not much yet). They include (in alphabetical order):

Barry Brook, Mark Burgman, Stephen Gregory, Fangliang He, Chris Johnson, Lian Pin Koh, Tien Ming Lee, Bill Laurance, Andy LoweÇagan Sekercioglu, Navjot Sodhi and Nigel Stork.

Our first workshop will be in early July, and I suspect we’ll have something concrete published by the end of the year. Stay tuned!

CJA Bradshaw



4 responses

9 07 2011
Life, death and Linneaus «

[…] blogged back in March that a group of us were awarded a fully funded series of workshops to address that question by the […]


26 05 2011
simon batterbury

Nigel Stork is moving to Griffith U. effective mid July 2011, after a few years at Univ of Melbourne.


28 03 2011
Tom Keen

Sounds like a fascinating and exciting project. Best of luck with it. I look forward to reading the results, and sincerely hope that it does ultimately have the profound impact on policy that you discussed. Something certainly needs to!


30 03 2011

Thanks, Tom. Will be fun and I think productive. Our challenge is certainly to make the results as policy-relevant as possible.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: