I should have been in the Lake District last week on a sampling trip. It was not lockdown that stopped me but the weather. There was rain at the weekend and the river levels went up. I then watched the hydrographs slowly sink back but not quite to a safe level for wading before Storm Christoph blew through and the river levels went up steeply. After two days of almost constant rain, the temperature dropped, then the rain turned to snow which is great news for me because snow on the fells means that the river levels start dropping, even if it is fiendishly cold. As I write, the levels are just about back to conditions that permit safe wading, albeit with air temperatures below freezing. We should get out early next week. Meanwhile, I am left thinking about the lakes and streams rather than actually experiencing their chilly reality.

My ruminations took me back to Ennerdale Water and along themes that I explored in a post from 2017 (“Lost in detail?”). Once again, we are looking at the genus Brachysira in this lake but this time, rather than pick up on the identity of a single species, I want to reflect on just how many different species of Brachysira seem to inhabit the littoral habitat of one lake. This also links with two other recent posts: The stream eats itself … and Curried diatoms?.

As was the case for the 2017 post, the slide was used in the UK/Ireland diatom ring test, which means that it was scrutinised by more analysts than just myself. All of us, however, agreed that several species of Brachysira were present with, in one case, eight different species being recorded by a single analyst. I’ve included illustrations of all of them and the first comment that anyone who is not familiar with diatom taxonomy will make is, almost certainly, that they don’t look particularly different from one another. There are differences but these mostly lie just at the edge of the resolution of the light microscope. More pertinently, those analysts who do not have light microscopes with the very highest specification will be unlikely to discern these differences. Whereas 25 years ago, you could probably do a perfectly adequate analysis of a sample such as this with a microscope equipped with a 100x achromatic objective with a numerical objective of 1.4, now you need a plan achromatic objective with a numerical objective of 1.25 and a differential interference contrast condenser and you will still struggle to differentiate some of these species.

Several questions arise. One is that if there is so much diversity discernible (albeit only just) with a light microscope (i.e. “semi-cryptic variation”), how much more are we missing because it is not discernible with a light microscope (“cryptic variation”)? For some freshwater diatom genera we have been able to walk away from taxonomy’s traditional reliance on morphology and approach the problem from a different perspective. Mostly, in this age, this different perspective comes from molecular genetics and in some cases (e.g. Fragilaria, Nitzschia, Sellaphora) it suggests that there really is variation beyond that readily discernible with light microscopes. In others (e.g. Gomphonema) the evidence is less clear cut and it is even possible that taxonomist’s enthusiasm may have run ahead of the actuality. Unfortunately, we don’t have any detailed molecular studies of Brachysira with which to test the ideas of species limits developed with morphological approaches.

A second question is whether we need all this information in order to draw the ecological insights that catchment managers need in order to decide what measures, if any, are needed to keep the water body in a healthy state. The autecological information that we have suggests that Brachysira species are mostly associated with healthy, low nutrient, circumneutral water bodies. Where there are differences, it is primarily due to the hardness of the water rather than to the levels of human-derived pressures. The very fact that we see several species in one lake is evidence that their preferences for these conditions, in any case, overlap. So there is little evidence that different species of Brachysira “indicate” starkly different environments.

As I suggested in Curried diatoms?, however, there may be some value in knowing that there are several species present but for reasons that are not associated with rather naïve associations between species and chemical measurements. As the present generation of methods for assessing lake or river health are tied to these tired notions of indicator values there is probably not a lot of extra “signal” to be squeezed from splitting Brachysira further.

Having said that water hardness plays a greater role in determining the distribution of Brachysira species than water quality, one of the oddities of this sample was that there were small numbers of two species associated with hard water (B. neglectissima and B. vitrea). There is absolutely no limestone or any other rocks that might create hard water conditions within the Ennerdale catchment, so why are these species here? I have found B. vitrea, especially, in small numbers in other soft water areas so these are not isolated occurrences. Do they represent genuine viable populations hanging on at the extreme edge of their range (and, one assumes, competing for resources with physiologically “fitter” Brachysira species) or are they “noise”, carried into the catchment by wind or on the feet of mammalian vectors such as, er…, myself? The latter has implications: firstly, it lends support to the controversial “everything is everywhere, environment selects” hypothesis (that I’ve mentioned in the past but never fully explained … I’ll do so in a post in the near future) and it also casts doubt on some of the claims of metabarcoding enthusiasts that they can extract more ecological information because the sequencing depth that is now possible exceeds that of traditional ecological analyses. If what we hope is “signal” is, in fact, “noise”, then that depth is as likely to confuse or even mislead as it is to inform better decision making.

Studies of the microscopic world are only ever as good as the technology available. But, equally important, studies of microbial diversity are also complicated by the observer’s expectations. How different do two cells have to be before you treat them as separate species? Not only did Hustedt not have such good microscopes as are now available when he wrote his Flora in 1930, but he also assumed that diatom species concepts were much broader than we now know to be the case. And the final twist is that the ability of anyone to match what s/he sees with the images in the literature. Hustedt used a single drawing to illustrate “Anomoensis exilis” (what we would now call Brachysira microcephala) in his Flora whereas Bryan Kennedy used 96 photographs, taken with differential interference contrast lighting, along with 31 scanning electron micrographs in his 2017 paper highlighting, in the process, four different morphotypes. Whether these represent different species is yet to be determined. My point is that “seeing” depends partly on how light stimulates the optical nerve (which depends on the quality of equipment we have), but more on how that raw signal is processed by our brains.