The loss of indirect interactions leads to cascading extinctions of carnivores

Sanders et al. The loss of indirect interactions leads to cascading extinctions  of carnivores Ecology Letters (2013) 16: 664–669

The population-consumption-environment nexus. Apparently. From

Broad bean aphids on a broad bean. From Wikipedia.

Regan Early

Regan Early

I chose this paper because I’m moving from biogeography and species distribution modelling into experimental research into the fundamental processes underlying species distributions. Well, really I want to do both at the same time, and an experimental system like this seems ideal to ask how biotic interactions might affect species geographic distributions.

My first instinct is that these experiments seem so elegant, and simple, that I don’t quite understand why they haven’t been done long ago. Perhaps someone with a stronger background in experimental ecology can help me out here.

The authors convinced me that this paper is important because carnivore extinctions are important. First carnivores are disproportionately likely to go extinct – their (usually) larger body size and low abundance make it particularly hard for them to wait out disturbances as can species lower down the food web. Second, carnivore extinctions can exert complex effects down the food web, which bounce back and result in even more complex effects back up the food web, i.e. on other top predators. To test these ideas out, the authors made toy communities consisting of a food plant, three aphid species, and three parasitoid wasp species, each of which specialized on a particular aphid species. In three sets of these communities one of the species of wasps was culled, and in the fourth set of communities all species were allowed to remain. What happened was that for two of the wasp species, their removal led to reduced persistence of one or more of the other wasp species. So we might assume that the reduction of one carnivore caused its aphid prey to grow in numbers, outcompeting the aphids on which another wasp species relies, and causing the bounce-back extinction effect at high trophic levels, right? It sounds like a text book example. Except that’s not what happened. Only one of the three aphid species showed a clear decline following the extinction of a wasp predating their competitor species, and no aphid species went extinct. The competitor species did increase in numbers though, and it seems that this confused the predators of the other aphids. The wasps were now so overwhelmed by the weird smells and evasive manoeuvers of all the aphids that they couldn’t eat, that their spidery-sense couldn’t find the aphids they did want to eat (this is a metaphor – I am not actually confusing my arthropods).

For me, the wider upshot of this is that as communities lose species, we should expect ripple effects that are really hard to predict using standard techniques of population growth and per capita attack rates. In this case extinction seemed to be mediated by anti-predator behaviour. To my knowledge, we don’t have a good framework for predicting which species might be susceptible to these effects. The authors argue that it should be common for attack rates to be reduced when predators are faced with and confused by too much prey that they can’t actually eat . But they don’t cite a wealth of evidence for this. It seems like a cross-taxonomic review of this phenomenon is in order, as a step towards developing tools to quantitatively predict its effects. I’m certainly prepared to be convinced that indirectly-driven secondary extinctions deserve our attention. This is particularly the case in marine systems where top predators are regularly harvested and prey species display complex behavioural adjustments when their predators disappear.

Finally all predators here were specialists and, as the authors point out, a generalist predator might help smooth out big differences in population densities of different prey species. They argue that this means the loss of biodiversity will increase susceptibility to extinction cascades. This may well be correct, but the argument seems a bit disingenuous – retaining a few generalist species could prevent secondary extinctions. Happily, for biodiversity’s sake, complexity does seem to make biological systems more resilient (Loreau, Naeem et al. 2001, Reusch, Ehlers et al. 2005). The challenge will be to evaluate why and how by approximating this phenomenon with microcosms, where more complexity means more noise and less signal.

Lynsey McInnes

Lynsey McInnes

What a neat paper. I’ve said it before, I’m a big fan of micro/mesocosm experiments. They seem so elegant and useful, a step up from computer simulations, but easier to manage than field experiments. I agree with Regan that it is surprising this kind of thing has not been done before; seems like an important, yet tractable question. That the authors find it is more complicated than simply host extinction that causes the ripple effects of extinction on other parasitoid species underlines we have a long way to go before being able to robustly predict the effects of single extinctions on ecological communities.

The authors acknowledge that a generalist predator might smooth out these effects, and I’d be interested to see if they do. I imagine there is only so much a generalist can smooth and extinctions of specialists (of which there are many) will eventually impact communities in the way demonstrated here. Seems like there is plenty of studies showing this kind of effect (that diversity itself aids resilience).

I guess what I’d like to see now is to go one step back from an experiment/question like this one and see an attempt to work out how to predict which predators or prey are vulnerable to extinction ( here the authors have identified members of higher trophic levels, but surely not all members?). The next step would be to see whether an extinction can be corrected for by a generalist or other specialist and for how long/how broadly. When does a community or at least a specific trophic level become so homogenised that it falls apart? How much does diversity per se matter? I know this has been studied before, but am not sure how general the conclusions have been or whether specific traits or scenarios have been identified as more or less likely to lead to collapse. An approach such an Laughlin’s axes of functional traits might be helpful in this regard.

I had it hammered home to melast week that a field element to all ecology is necessary for it to be meaningful. I’ve chosen to take this to mean that an exchange between theory, models, field data and experiments is necessary for the best kinds of ecological insights. I take this paper as a good example of this, clear question, well tested, conclusions related to but distinct to expectations and with plenty of fodor for future investigations. Oh yes, and with an applied angle too. Nice.


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 )

Google+ photo

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

Connecting to %s

%d bloggers like this: