A few months ago, I was sent a paper to review for a journal (better not name authors or journal title … if you are that interested, you will probably be able to work these out from the information I am about to give you). I read the paper, and sent back my recommendation: reject. So, imagine my surprise when I saw the same paper appear, little-altered from the version I had refereed, in the very same journal. Of course, a single peer reviewer does not have the ultimate say on whether or not a paper should be accepted but, in this case, I had queried both the experimental question and some aspects of data analysis, both of which seemed quite fundamental. In the final version, the question is the same, the weaknesses in the dataset remain and the original analyses are now shored-up by some heavy-duty stats which go a small way to allaying some of my concerns.
The paper looks at how the diversity of diatom assemblages varies in different types of rivers, and suggests that this might be a useful property to include in ecological status assessments. Most ecologists would probably agree that the diversity of organisms ought to be a property that we ought to consider when assessing the condition of a site but the use of diatom diversity has a troubled history. For some types of pollution (heavy metals, for example), there is a clear relationship, with fewer species being recorded as the level of the pressure increases. For other types of pollution (nutrient and organic pollution for example), results have been much more ambiguous. This may be partly due to the measures of diversity that we use, but it also reflects a more fundamental problem: the focus on diatoms alone is very artificial. We have had success when we use diatoms to indicate, indirectly, the level of chemical pressures but it is much harder to infer fundamental ecological processes when you have digested away all the algae that share the habitat with the diatoms.
My concerns about this paper were prompted in part by some analyses I had done with Dean DeNicola of Slippery Rock University, Pennsylvania, which had shown us that the diversity of diatoms alone had very little relationship to the diversity of all the algae present in a dataset of samples from the littoral zone of lakes. The paper is currently in press in Freshwater Science, so I could not quote chapter and verse in my review. As levels of pressure change, it is possible not just that interactions amongst the diatoms will change (and, thus, influence their diversity) but also that interactions between diatoms and other algae will change (which will not necessarily be reflected in a diversity measure based on diatoms alone).
This is a manifestation of a much broader problem: diatomists form a very insular and self-referential community. Methods have evolved that focus on the (dead) diatom shell, not on the living organism, with a prevailing belief that the ability to recognise many species from the shapes and patterns of these shells trumps the insights that come from looking at the whole community of benthic algae. Diatomists review each other’s papers and generally accept these assumptions wholesale. As a result, the community, particularly in Europe, is drifting further and further away from mainstream functional ecology. It is something that I have talked about before and will need to return to again if we are to start having any serious influence on the decision-making process.
DeNicola, D. & Kelly, M.G. (2014). The role of benthic algal communities in ecological assessment of lakes. Freshwater Science (in press).
Kelly, M.G. (2013). The semiotics of slime: visual representation of phytobenthos as an aid to understanding ecological status. Freshwater Reviews 5: 105119.