Friday, March 18, 2011

The regional community, maximum entropy, and other ideas in ecology

Looking through my feed of community ecology papers this month, I couldn’t help but notice that while most tested well-established concepts–density-dependence, niche partitioning, metacommunities, competition, dispersal limitation–there was also–as I suppose is usually true–a subset of papers championing newer, less established ideas.

For example, the article “Applying a regional community concept to forest birds of eastern North America” by Robert Ricklefs, furthers the regional community concept he introduced in 2008. Ricklefs is uncomfortable with how ecologists typically define local communities – i.e as spatially and ecologically discreet entities – and the predominant focus in community ecology on local coexistence. He argues that communities make sense as entities only at a larger scale, taking into account that local communities are not isolated, but instead interact as a function of overlapping ranges and species dispersal. In this paper he applies this concept to Breeding Bird Survey data to examine the distribution and abundance of birds in eastern NA.

Partel, Szava-Kovats, and Zobel are also critical of the predominant focus on local diversity. In their paper “Dark diversity: shedding light on absent species”, they pitch the idea of “dark diversity” as a valid diversity metric. Dark diversity accounts for the number of species which belong to the species pool for a particular habitat in a region but are not actually present in a local community of that habitat type. The resulting value can be used to calculate a dimensionless ratio of local to dark diversity, suitable for comparison of diversity components in dissimilar regions.

Lastly, in “A strong test of a maximum entropy model of trait-based community assembly”, Shipley et al. further test Shipley’s model of Entropy Maximization, using it to predict the composition of communities in the South African fynbos. The model predicts community composition (species identity and relative abundances) through an assumption of random assembly (or entropy maximization) within environmental constraints on species traits.

New ideas are a constant in ecology, but they face stiff competition in an already crowded field. The possible mechanisms of local coexistence, for example, are already a long list. What determines which of these–or any–ideas become entrenched in ecology? The likelihood of a concept becoming established must be a complex function relying on a cost-benefit analysis–what does applying this idea cost compared to the gain in understanding it produces?–further adjusted by intangible variables like timing and the skill and prestige of an idea’s advocate. After all, some ideas require decades to establish properly, requiring changes in the theoretical climate or technical capabilities, for example, neutral theory or spatial ecology. Others seem to catch on immediately. Philosophers have written more cogently on how scientific ideas change and paradigms shift, but as participants in the process, we have a rather unique perspective. After all, as scientists we play an active role in driving these shifts in thought and action. You might argue that the merit of the ecological ideas that become established are as much a reflection on those who accept and institute them, as on those who propose them.

Tuesday, March 8, 2011

awesome infographic by Caroline Tucker
(click to go large)

There exists a problem in science so complicated that decades of work have yet to solve it. Its causes and consequences make some of the toughest questions in complex analysis or astrophysics look like child’s play. And yet when we consider this problem, the conclusion is immediately obvious and simple: it should not exist.

I am talking about the fact that today, in 2011, female scientists are punished solely because they are female scientists.

In theory, this problem doesn’t exist anymore. Multiple waves of feminism should have chipped away whatever glass ceilings once capped our ivory towers. Women are receiving more PhDs than men in many fields and they are earning such a high proportion of bachelor’s degrees that we may have to rethink that name.

But in the last several weeks, some disturbing realities have resurfaced in the science media. At the end of January, Nature reported that women earn fewer scholarly awards than they should, based on the proportion of their respective fields that they represent. That same week, Science published a graph showing the number of European Research Council grants awarded to women in its last funding round – 9.4%.

Stats like these are nothing new; they pop up all the time. What is new, however, is the article that followed in Proceedings of the National Academy of Sciences a few days later. It turns out that there is no longer much evidence for overt discrimination against women applying for jobs or grants in quantitative fields. Instead, disparities in available resources are causing many of the differences between women and men’s scientific careers.

Yes, there are discrepancies in publication acceptance rates and grants, but the authors attribute these to factors like women occupying more positions at teaching-intensive schools rather than research institutions. When they compared men and women with similar resources, the biases disappeared, or in some cases, favoured women. (If you don’t want to read the whole article, there’s a nice summary of it here.)

Ok great, the science community isn’t explicitly discriminating against women. But this leads us to a much more troubling conclusion; the culprits are actually deeply engrained societal expectations and constraints that likely extend well beyond the sciences, and certainly beyond the scope of this blog post, though a few of them are highlighted in this thoughtful opinion piece.

Here’s what I will say: it’s not written in our DNA. How many times have you heard lines like, “Men and women are just different, they always will be, our brains aren’t wired the same”? This kind of just so statement is rarely backed up with evidence. For a good debunking of these misconceptions, check out two new books, reviewed here.

Now it’s possible that I, as a young male grad student, do not hold the most valuable two cents on these issues. I could keep rambling about things that I don’t fully understand, but my perspective is limited, and I think maybe the most constructive thing to do at this point would be to hear about other people’s ideas and experiences in the comments section below. So I’m cutting this short and leaving it incomplete in favour of a more open forum. In particular, it occurred to me that we in the ecology and evolution community have a unique opportunity to shed light on many gender issues. The PNAS article focuses on the underrepresentation of women in math-intensive fields, but comparing mathy fields to less-mathy fields entails a lot of confounding factors. In ecology and evolution, however, we cover the whole spectrum, from the completely mathless and descriptive, to the suspender-wearing, calculator-toting quants. We generally all come from relatively similar biology backgrounds, eliminating many of those confounding factors, and it would be great to hear how you all think these issues play out in E & E. So go for it blogosphere, do your thing.