A month ago, I wrote about the problems I had identifying Gomphonema species in the River Ehen (see “Diatoms and Dinosaurs”). My work on the River Ehen happens in brief, intense bursts before I have to move on to other jobs and, once in a while, I need to sit back and take some time to revisit some of my older samples in order to compare their composition with that of my most recent collections. The first picture below is a sample from March 2013, the lower comes from a location about 2.5 km downstream in April 2014. In both cases, the larger cells correspond most closely to Gomphonema gracile whilst the smaller ones look like members of the G. parvulum complex. However, when you arrange the cells in order of diminishing size, there seems to be a continuum between the long “G. gracile” cells and the shorter “G. parvulum” cells. This seems to confirm what Dawn Rose and Eileen Cox were saying in the paper to which I referred in my earlier post. You may remember, too, that I commented on a similar phenomenon occurring in populations of Hannaea arcus from the River Ehen (see “Diminishing with age …”).
Gomphonema gracile / parvulum from the River Ehen, Mill, March 2013.
Gomphonema gracile / parvulum from the River Ehen, Ox-bow, April 2014.
The phenomenon becomes more interesting when I compare the dimensions of Gomphonema cells in the two samples. First, I measured 25 cells in two samples from exactly the same location in the River Ehen but collected 15 months apart. In December 2013, the median length was 32 µm (max: 51; min: 21 µm) whilst in April 2014 it was 25 µm (max: 38; min: 17). This confirmed what I suspected as I had photographed the diatoms: the smaller diatoms were more abundant in the April 2014 sample compared with December 2013. Note that the largest cells encountered are of a similar size, but there is a longer “tail” of small cells, as we might expect if we were observing the natural cycle of size reduction.
Next, I compared the dimensions of two samples collected on the same date but 2.5 km apart. This time, cells at the Mill (the upstream site) were generally larger (median: 32; max: 39; min: 23) than those at the Oxbow. What I think may be happening here is that the upstream populations are, to some extent, “seeding” populations further downstream. Some of the cells will be washed out of the biofilms and a few of these may become lodged into biofilms further downstream where they, in turn, will establish and, in time, divide. So far, we only have the scraps of evidence from the Ehen plus some data for a different species in the River Wear to support this hypothesis but it is worth exploring in more detail at some point.
Valve length for 25 cells of Gomphonema gracile / parvulum from River Ehen, oxbow, in December 2013 and April 2014.
Valve length for 25 cells of Gomphonema gracile / parvulum from River Ehen, Mill and Oxbow, in April 2014.
There is another twist to this tail because the consensus in the literature is that “Gomphonema gracile” is more sensitive to pollution than “Gomphonema parvulum”. My experience is that any reference to the ecology of diatoms should be treated with caution, as it is often laced with hearsay and tradition rather than underpinned by hard evidence. However, a quick look at datasets that I have suggested that this difference in preferences might be the case, despite the morphological continuum. The 90th percentile of alkalinity measurements for samples which contained “G. gracile” was 58 mg L-1 CaCO3, whereas for samples containing “G. parvulum” it was 175 mg L-1 CaCO3. An outlier of G. gracile was recorded at 224 mg L-1 CaCO3 but this record represented a single valve (i.e. half a dead diatom). I have put both species names in inverted commas because of the taxonomic and nomenclatural confusion that makes all of our records somewhat suspect. A similar situation occurs with phosphorus, with “G. gracile” being associated with sites with lower phosphorus concentrations than “G. parvulum”. The 90th percentile of reactive P measurements for samples with “G. gracile” was 28 µg L-1, whereas for “G. parvlum” it was 173 µg L-1.
I think that there are two possible explanations for these patterns. The first is that the species that is more likely to occur downstream will be associated with higher alkalinity and nutrients, because this reflects the general longitudinal change along rivers. However, another possibility is that environmental preferences change during the cell cycle. There is extensive evidence from ecotoxicology which demonstrates that young life stages of many species are much more sensitive to a whole range of stressors than mature stages. There is, to the best of my knowledge, no similar evidence for diatoms, but it is conceivable that the initial cells, for example, may be more sensitive than cells later in the life cycle. This, then, could also work to restrict the distribution of “G. gracile” to cleaner stretches of river, and prevent reproduction in the enriched conditions that “Gomphonema parvulum” can tolerate.
Far more questions than answers. As I alluded in Diatoms and Dinosaurs, there is an unwritten assumption amongst many diatom taxonomists that the silica frustule contains all that we need to know in order to understand diatom taxonomy. A number of taxonomists are challenging this assumption with their work now, but there are still many traditionally-minded taxonomists out there whose horizons do not extend beyond the frustule. Yet there are also several diatom genera, not just Gomphonema, where silica-based taxonomy seems to generate more heat than light. I do wonder how many more of these problems might resolve themselves if people took the time to stand back and consider the life cycle in a more rigorous manner.