Papers on biodiversity are not my regular science reading fare and the reason I found my self reading the article “Initial community evenness favours functionality under selective stress” by Wittebolle et al. in Nature from last week isn’t very important. But I did find myself reading about the “biodiversity-stability relationship” in a microbial ecosystem, so rather than let all that effort go to waste I decided to write about it here. Like most things that wind up on this page, there is an extra little twist at the end, this time where I come out against biodiversity, but to get there you’ll have to wade through a description of the paper first (or you could just skip to the end, but then you wouldn’t know what I was talking about).

The paper is about biodiversity, but exactly what is biodiversity (other than that we know it’s “good”)? Biodiversity for most people means species richness, that is the existence of many different kinds of species co-existing in an ecological community. But that’s not the only possible measure. Another is “evenness,” a measure of how equal or unequal are the sizes of different kinds of communities. Does one or a few species dominate? Or are there roughly even contributions of many different species? There are probably other measures as well, but just considering these two you can see how complicated this can get. The authors of this paper wanted to see what effect evenness had on the ability of a community to withstand stress. They used as their model system a mixture of denitrifying micro-organisms.

First let’s discuss denitrification. Micro-organisms (usually bacteria) are important in taking more complex molecules and breaking them down into simpler ones in a process we call decomposition. Decomposition of organic materials is what turns discarded refuse into compost and turns us into “dust” (as in “dust to dust”). The complex molecules we are made of (proteins, starches, etc.) took energy to produce, just as it takes energy to build a house out of bricks (you have to lift the bricks and pile them up on top of each other, just as smaller building blocks like amino acids are piled up into proteins or the constituents of amino acids, one of which is gas nitrogen, are made by connecting other atoms. We get the energy to make our own body’s complex materials by burning food. Our food, in turn, is made of other complex molecules and we extract energy from them by breaking them down, like harnessing the energy stored in a building’s structure by letting it fall down in a controlled manner. At some point there is an end to the chain as some organism (usually a plant) uses the energy of the sun to take the simplest building blocks and hook them together into the more complex ones that make up a plant. All along the way we can start the cycle again by taking a dead plant or animal and breaking it down into the simplest building blocks again. That’s decomposition, and without it we would be over our heads in dead plants and animals. Denitrifying bacteria mainly live in the soil (and marine ecosystems) and take more complicated nitrogen-containing organic materials and break them down into nitrogen gas.

What this research group did is harvest some 18 different kinds of denitrifying bacteria from natural sources and grow them together in a community, each community composed of varying proportions of each species. In other words, they kept the richness of species constant (18 different species) but varied the evenness. They then measured how well this community of many different species was able to go about its job of denitrifying. This was done by incubating each different mixture in an array of little wells in a plastic plate. Some of the plates containing this range of mixtures were left alone, some were stressed by lowering the temperature and some were stressed by exposing them to increased salt. All these combinations of evenness and stress were measured for the ability of the microbial community to denitrify a test solution (of nitrite).

That’s the basic set-up. How did they measure “evenness”? They borrowed a measure first used in economics to measure income inequality. To get this measure you start out with something called a Lorenz curve by economists. A Lorenz curve is the kind of thing that gives cumulative income distribution versus the different proportions of the population, as in “the bottom 80% of the population accounts for only 20% of the total income of society.” This kind of curve shows up in a lot of different places. I’ve used it for some of my work and people evaluating diagnostic tests use a variant they call an R-O-C curve (for Receiver Operating Characteristic; the origin is from studies of radar during World War II). If the income (or in this case, distribution of evenness) is perfectly equitable the Lorenz curve looks like a straight line from bottom left to top right of a square (in other words a diagonal). The more unequal the more it is bowed up, connecting the two corners but facing concave down. By measuring the area between the diagonal and the curve you can measure the departure from an equal distribution, leading to something called the Gini coefficient. Here is a figure from the paper showing the Lorenz curves for the many different combinations of evenness from the experiment:

This figure shows lots of different Lorenz curves. You can see that the experimental communities are well distributed over departures from the diagonal, i.e., have a wide range of Gini coefficients.

At this point there are some statistical details which I was going to explain before this post got over long, so I’ll briefly summarize. Lack of evenness (e.g., having one species dominate the rest) was somewhat bad for overall ability of the community to denitrify, while inequality didn’t seem to affect the ability to cope with temperature stress. Rather than this showing a good thing, it is probably indicative of the across-the-board effect of temperature. The whole community did badly. If you drop a nuke on the city both the rich and the poor get fried. But salt stress was different. First, it was selective, seeming to affect some species in the mix more than others. Second, the effect of size inequality was quite evident. The more uneven the distribution the worse the community coped with salt stress. Their conclusion?

Biodiversity protects ecosystems against declines in their functionality and allows for adaptation to changing conditions, because the coexistence of many species provides a greater guarantee that some will back up a given function when others fail. Within the frame of this ‘insurance hypothesis’, two aspects are important: (1) functional redundancy, in the sense of there being multiple species for each functional group, and (2) the relative abundances among these redundant species. At lower levels of species richness, the functionality of the ecosystem decreases. In this research, all communities had the same degree of richness; hence, the importance of evenness for functional stability was isolated. Our results demonstrate that a community must have an even distribution among its functional redundant members if it is to respond rapidly to selective stress. In fact, when an ecosystem function in a highly uneven community depends strongly on the dominant species, the functional stability is endangered by environmental fluctuations6. Even under non-stressed conditions, high initial evenness is desirable for good functionality. Moreover, natural and anthropogenic activities influence the relative abundances more than the richness of species, and this has important consequences for ecosystems long before a species is threatened by extinction. In conclusion, the existence of a highly diverse community, where redundant species may offer equivalent contributions to a specific function, may lead to higher functional stability during environmental fluctuations. This implies that changes in community evenness should warrant increased attention in biodiversity surveys. (Wittebolle et al., Nature (cites omitted) [doi:10.1038/nature07840; subs. req'd])

In other words, the redundancy of having lots of different species doing roughly the same thing (denitrification) makes it less likely that anything selectively bad for one species will hurt the overall enterprise. This isn’t as true for things that affect all the species more or less equally, but these and other data suggest that trying to eliminate an entire community that way isn’t that easy. So these results make sense. Biodiversity is “good.”

So why did I say that I would have a perverse twist on this? Because not all microbial communities are “good,” at least good from our selfish point of view. There are complex microbial communities called biofilms that cause lots of health problems, from urinary tract and middle ear infections to gum disease. Because they have the resources of a community they are much more difficult to get rid of when they do cause trouble. This paper suggests that we might have better luck dealing with them if we use selective agents that will increase the unevenness in the biofilm, not just try to nuke the biofilm with a braod spectrum antibiotic.

Probably more than you wanted to know about this stuff. But I like science that crosses boundaries, in this case economics, sociology, ecology and clinical medicine.