Quantifying our ignorance …


I am fairly sure that I am not a popular person after my latest choice of slide for the “ring test”, the regular calibration exercise that UK and Irish diatomists perform.   I had noticed a few taxa that we had not seen in previous ring tests in a sample I collected during my visit to the Shetland Islands back in May 2019 (see “Hyperepiphytes in the Shetland Islands”) but, on closer examination, the sample proved to be both highly diverse and very challenging.  The seven experienced analysts who provide the benchmark analyses for the ring test found, between them, over 150 different species: some we could name with confidence, but others we could match to no published description.  Amongst those was the species of Achnanthidium photographed below.   It might be Achnanthidium digitatum or possibly A. ertzii but, then again, it does not quite match the characteristics of either of these so, once again, we have left it unnamed (you can find the original descriptions of both these species in the reference list).

According to Algaebase there are 116 species of Achnanthidium that are currently accepted but descriptions of these are scattered through the literature so it is really hard to be confident that you have found a new species during a routine survey.  This is particularly the case when we only have light microscopical analyses with which to work, as the small size of Achnanthidium species means that you really need a scanning electron microscope to see the fine details clearly.  This, however, assumes that the pool of unnamed Achnanthidium species is finite and that the 116 species on Algaebase is a significant proportion of the total number of Achnanthidium species.  A recent study by Eveline Pinseel and colleagues based on samples from Arctic regions offers hints that there is still plenty of diversity within the genus that cannot be linked to named species

This may, however, be a naïve assumption.   My colleague Maria Kahlert, who works in Sweden, comments that she is quite happy looking at samples that I send her from polluted sites in the UK as she can name most of the species (Achnanthidium and otherwise) from her own experience.   It is the samples from pristine habitats that fox her because so many of the forms are different to anything she has encountered in Sweden.  We have, in other words, a neat reversal of the opening line of Anna Karenina (“All happy families are alike, each unhappy family is unhappy in its own way”), with very high beta and gamma diversity of diatoms (probably other microalgae too) as a characteristic of regions with low population density (see “Baffled by the benthos (2)”).  We often miss this in our enthusiasm to fit all that we see down the microscope to published descriptions, but when we take time to look hard, that diversity – and those differences between sites – start to mount up.


The unknown Achnanthidium species from Petta Water, Mainland, Shetland Islands (pictured at the top of the post).  Scale bar: 10 micrometres (= 1/100th of a millimetre).   Photographs: Lydia King

Let’s think of this as an ecological experiment to understand the diversity of Achnanthidium, following the capture-mark-capture approach.   Capture-mark-recapture is a technique used by ecologists to assess the size of a population.   As it is rarely possible to count all individuals, a portion of the population is collected, marked (a dab of paint on a snail’s back, for example) and released.   Some time later, the population is sampled again, and the proportion of those that bear the mark in this second sample is used as an indicator of the proportion of the population captured by the original sample.   Though devised for population biology, some have used the same principles to understand diversity in other contexts too so might it work as a means of understanding the yet-to-be discovered diversity of diatoms?

What we have in the scattered taxonomic literature is a record of all the Achnanthidium species that have been “captured” (i.e. observed) and “marked” (i.e. described) by taxonomists.   Suppose we now go some locations not previously visited by taxonomists, take some new samples and see 1) how many different forms of Achanthidium we can see and b) how many of these are “recaptured” (i.e. forms that align with previously described species).   Or, thinking about the problem in a different way, the number of named species could be compared with the number of distinct “operational taxonomic units” (“OTUs”) detected by metabarcoding.   More relevantly, how many extra OTUs are added when more lakes and streams are added to the dataset?   There are well-established methods for deriving “rarefaction curves” that might be useful in understanding regional diversity of diatoms, and modifications of “capture-mark-recapture” have been used to understand taxonomic diversity in palaeobiolgoical contexts, so why not in contemporary ecology too?

The Shetland Islands would make an ideal test ground for such a study as they are geologically-diverse habitats providing the types of conditions where Achnanthidium species thrive (low population density and agricultural intensity.   The diatoms of the region were studied about 40 years ago by my late mentor John Carter and although one of his samples yielded the type material for Achnanthidium caledonicum there have been so many developments in Achnanthidum taxonomy subsequently that this archipelago represents a tabula rasa for a modern taxonomist.   Its many remote lochs and streams offer the setting for a natural experiment which sets out, to put it bluntly, to quantify our ignorance.


Achnanthidium caledonicum from Loch Osgaig, Highland Region, Scotland.   Originally described as Achnanthes microcephala f. scotica Carter & Bailey-Watts 1981 (Scale bar: 10 micrometres (= 100th of a millimetre).  Photographs: Lydia King.


Carter J. R., Bailey-Watts A. E. (1981). A taxonomic study of diatoms from standing freshwaters in Shetland. Nova Hedwigia. 33: 513-630.

Pinseel, E., Vanormelingen, P., Hamilton, P. B., Vyverman, W., Van de Vijver, B., & Kopalova, K. (2017). Molecular and morphological characterization of the Achnanthidium minutissimum complex (Bacillariophyta) in Petuniabukta (Spitsbergen, High Arctic) including the description of A. digitatum sp. nov. European Journal of Phycology 52: 264-280. https://doi.org/10.1080/09670262.2017.1283540

Van der Vijver, B., Jarlman, A., Lange-Bertalot, H., Mertens, A., de Haan, M. & Ector, L. (2011).  Four new European Achnanthidium species (Bacillariophyceae).  Algological Studies 136/137: 193-210.

Liow, L.H. & Nichols, J.D. (2010). Estimating Rates and Probabilities of Origination and Extinction Using Taxonomic Occurrence Data: Capture-Mark-Recapture (CMR) Approaches.  The Paleontological Society Papers 16: 81-94).

This week’s other highlights:

Wrote this whilst listening to: Sheku Kanneh-Mason’s recording of Elgar’s Cello Concerto.   Taking me back to his performance at the proms on a warm evening last summer.

Cultural highlight: Sam Mendes’ film 1917 which, coincidentally, uses the River Tees (as featured sporadically in this blog) as one of its locations

Currently reading: I have just finished Good Economics for Hard Times by Abhijit V. Banerjee and Esther Duflo, which I mentioned a couple of weeks ago.  It left me with the feeling that, had both Boris Johnson and Jerermy Corbyn read it and taken on its messages, the election campaign and the UK political landscape might have been very different.

Culinary highlight: OK Diner on the southbound side of the A1 near Grantham.  Felt like we were walking into the opening scene from Pulp Fiction (the one where Tim Roth jumps up onto a table and attempts to rob all the customers).   Escaped with wallet intact.


The devil lies in the detail …

Our latest ring test* slide took us on a vicarious journey to the beautiful River Don in Aberdeenshire.  Maybe because I have been doing this job for so long, but the quality of the landscape was clear to me as I peered through my microscope 500 kilometres away: the range of diatoms that I could see would not have thrived anywhere with more than the lightest touch from humankind.

One of the clues for me lay in some of the smallest diatoms on the slide.   It took some discussion amongst my fellow experts, but we eventually came up with a list of five different species of Achnanthidium (all illustrated below) which, together, constituted about a third of all the diatoms on the slide (admittedly, because they are small, they constitute rather less than a third of the total volume of diatoms, but that is another story ….).   The mere presence of several Achnanthidium species is, in my experience, usually a sign of high habitat quality (see “Baffled by the benthos (2)”) but unravelling the identities of the different species with a light microscope is challenging.

Achnanthidium-minutissimum-Medwin_WaterAchnanthidium minutissimum from Medwin Water, Scotland. Photographs from the Diatom Flora of Britain and Ireland by Ingrid Jüttner.  Scale bar: 10 micrometres (= 1/100thof a millimetre). 


Achnanthidium pyrenaicum from the River Don, Towie, Aberdeenshire.  Photographs by Lydia King.  Scale bar: 10 micrometres (= 1/100thof a millimetre). 

The genus Achnanthidium is a good example of the delicate co-existence between “identification” and “taxonomy” in the world of diatoms.   Individuals from this genus are usually small so anyone using a light microscope for routine analyses will be working right at the optical limits of their equipment whilst anyone with a serious interest in taxonomy will depend upon a scanning electron microscope for the insights needed for critical differentiation between species.

This divergence between the working methods of “identifiers” and “taxonomists” means that it is rarely possible to name every individual of Achnanthidium with complete confidence.  The ones that present clearly in valve view (i.e. face-up) can mostly be assigned to a species based on features we can see with a light microscope, but it is not always straightforward for those seen in girdle view (i.e. on their side) or which are partly obscured by other diatoms or extraneous matter on the slide.   In this example from the River Don, we also noticed that smaller individuals of A. gracillimum lost their characteristic rostrate/sub-capitate ends and were, as a result, not easy to differentiate from A. pyrenaicum.


Achnanthidium gracillimum from the River Don, Towie, Aberdeenshire.  Photographs by Lydia King.   Scale bar: 10 micrometres (= 1/100thof a millimetre). 

What continues to mystify me is why so many closely-related species can live in such close proximity. It is Achnanthidium that prompt this question here, but other genera display similar tendencies (see “When is a diatom like a London bus?”).  And this immediately generates another question: why are more people not asking this question of diatoms and, indeed, microscopic algae in general?

The answer to that question falls into two parts. The first is that understanding the precise ecological requirements of microscopic algae is not a trivial task, and assumes that you are able to get several closely-related species to live in culture (which, itself, assumes you know the precise ecological requirements of each … you see the problem?).   There is, as a result, a tendency to avoid experimental approaches and, instead, look for how species associate with likely environmental variables in datasets collected from sites exhibiting strong gradients of conditions.   However, this assumes that the forces that drive the differentiation between species work at the same scale at which we sample (see “Our patchwork heritage …” for more on this).

Underlying this, however, is a deeply-held belief, dating back at least forty years, that the niches of freshwater diatoms are determined primarily by the chemistry of the overlying water.   This is a dogma that has served us well when using diatoms for understanding the effects of environmental pollution but which is, ultimately, a limitation when trying to explain why we found five separate Achnanthidium species in a single sample, all exposed to the same stream water.


Achnanthidium lineare (first three images from the left) and A. affine (two images on the right) from River Don, Towie, Aberdeenshire.  Photographs by Lydia King.  Scale bar: 10 micrometres (= 1/100thof a millimetre). 

I will go one step further: this dogma is so deeply held that referees rarely challenge the weak evidence that is produced to demonstrate different responses to environmental conditions between closely-related species.  There are certainly variations in environmental preferences between Achnanthidium species, but these are best expressed as trends rather than unambiguous differences and I have never seen such trends subject to rigorous statistical testing.

I blame better microscopes: greater magnification and resolution has revealed such a baffling amount of diversity that all the energy of bright diatomists is absorbed unravelling this rather than trying to explain what it all means (see “The meaning of … nothing”).  If we were bumbling along with the quality of equipment that Hustedt depended upon, then maybe we would be cheerfully lumping all these forms together and focussing on functional ecology instead.   Maybe.

* see “Reaching a half century” for more about the ring test scheme

The desmid dilemma …


The second location we investigated during the Quekett Microscopy Club / British Phycological Society weekend was Cogra Moss, a small reservoir about four kilometres north of Ennerdale Water.  It is also  about a kilometre or so west of Lampleugh Green where I was staring mournfully at my flat tire whilst the advance party, unaware of my predicament, was out collecting samples.  They must have missed me by a matter of minutes.

As at Moss Dub they found some promising locations around the margins and, in the small tarn at the north-east corner, some patches of Sphagnum from which desmids could be squeezed, plus some floating vegetation.   Once again, I’ve illustrated some of those that we encountered, with a warning that this is a limited selection of the more photogenic ones and we’ve sent samples off to David Williamson for a more comprehensive analysis.   And, once again, the sheer diversity of desmids in the sample is a source of wonderment.   How can one small lake support so many variations on a one type of alga?  I’ve speculated on such issues in the past, drawing on G.E. Hutchinson’s “Paradox of the Plankton” (see “Baffled by the benthos (1)”).   In that post I suggested that it is partly a matter of scale and perception and, in this case, I suspect that the desmids we see in a Sphagnum squeezing are adapted to a wide range of microhabitats.  That means that the desmids would have had a three-dimensional arrangement within the Sphagnum whilst it is in situ but this is lost when we drag a handful of moss from the lake margin and squeeze it into a pot.


Desmids from Cogra Moss: a. Eurastrum crassum (length: 140 – 180 micrometres; width: 75 – 92 micrometres); Netrium digitus (length: 130 – 390 micrometres; width: 40 – 82 micrometres); Closterium kuetzingii (length: 300 – 550 micrometres) and Pleurotaenium trabecula (length: 277 – 600 micrometres; 22 -46 micrometres).  The photograph of Cogra Moss at the top of the post is by Judy Johns.


More desmids (and other algae) from Cogra Moss: e. Micrasterias thomasiana (length: 200 – 288 micrometres; breadth: 170 – 269 micrometres); Tetmemonus laevis(length: 67. 5 – 123 micrometres; breadth: 20 – 31.5 micrometres);
g. Schroederia setigera (85 – 200 micrometres long); h. Gonatozygon monotaenium (length: 90-327 micrometres; width: 6.2-12.5 micrometres); i. Staurodesmus extensus (width: 42-50 micrometres, including spines); j. Cylindrocystis gracile (length: 20 – 80 micrometres; width: 11 – 15 micrometres).

But, coming at this issue of desmid diversity from another direction, the term “desmid” is about as particular as the term “mammal”, insofar all belong to the same Class.   In “The big pictures …” I described how desmids were related to other green algae (acknowledging, in the process, that the term “green algae” is, itself, outdated).   This listed five separate families of desmids: four in the order Desmidales and one in the Zygnemetales (I’ve listed the examples from this and the previous post in the table below).   Think laterally and translate this level of organisation to the landscape around Cogra Moss and Ennerdale: the forests contain red squirrels (Rodentia), foxes (Carnivora) and deer (Artiodactlya) and there are otters in the River Ehen (another Carnivora but in the family Mustelidae rather than Canidae).   If we can appreciate how different mammals can interact within a landscape, then we should be able to apply the same principles on a much finer scale to organisms that are five orders of magnitude smaller.   It’s the principle behind fractals, but applied to biological  diversity rather than to geometry.

Earlier in the year, I published a paper with two colleagues that tried to explain how the way we study the microbial world can shape and, in many cases, impede our understanding (it’s open-access, so click on the link below if you want to read it).  We illustrated this with pictures that tried to demonstrate how microscopic algae interact with other organisms.  These included host plants, in the case of epiphytic algae, but also the protozoans that feed on them.  Most of our examples were diatoms, and there was a reasonable literature on which we could draw.  Curiously, I’ve never come across papers that provide this contextual information for desmids. Perhaps I just don’t look in the right places.   If it is out there and I’ve missed it, please do let me know.


Kelly, M. G., King, L., & Yallop, M. L. (2019). As trees walking: the pros and cons of partial sight in the analysis of stream biofilms. Plant Ecology and Evolution152: 120-130.

Organisation of the class Conjugatophyceae with examples encountered in Moss Dub and Cogra Moss.

Order / Family Examples
     Closteriaceae Closterium
     Desmidaceae Desmidium, Euastrum, Pleurotaenium, Staurodesmus, Tememorus
     Gonzatozygaceae Gonatozygon
     Peniaceae No examples in these posts, but see “Desmid Diversity” for illustrations of representatives in Kelly Hall and Long Moss tarns.
     Mesotaeniaceae Cylindrocystis
     Zygnemetaceae Not desmids: Filamentous algae including Mougeotia, Spirogyra and Zygnema – examples from Ennerdale area are described in several other posts,

Besotted by beavers …


The beaver dam at Dubh Loch, Knapdale, Argyll, photographed in April 2013.

Beavers are back in the news. An article in today’s Independent reports that Natural England have granted a licence to Devon Wildlife Trust to manage a population of beavers that had escaped from a wildlife park.   This overturned a decision by the Department for Environment, Food and Rural Affairs which argued that beavers are an invasive non-native species and that the Devon population should, therefore, be trapped and returned to a zoo. As I wrote in 2013 (see “In pursuit of beavers …”), beavers do challenge our preconceptions of what is “natural” and I can also report two other recent studies that show how beavers can change both the structure and function of freshwater ecosystems.

The first is a study co-authored by my sometime colleague Nigel Willby on the Scottish Beaver Trial at Knapdale in Argyll (the locations I wrote about last year) which showed changes in the aquatic vegetation due to direct grazing by the beavers, as well as by the water level changes brought about by dam-building. A few species (Schoenoplectus lacustris and Cladium mariscus in particular) decreased due to grazing but the site where the water level rose due to dam building saw an increase in both the diversity and heterogeneity of aquatic vegetation. This, in turn, had beneficial effects on invertebrate diversity in the area.   Or, more accurately, the long-term decline and eventual extinction of beavers due to human pressures altered the balance of plants and animals and the Knapdale trials are giving us an insight into the natural state of the area 200-300 years ago. Paradoxically, our ideas of “natural” are largely shaped by our own expectations, as I commented in a post about Himalayan Balsam last year (see “The future is pink …”)

The second study looks at the wider consequences of beaver activities.   Beaver dams create large areas of standing waters in places that once would have been terrestrial or semi-terrestrial habitats with a stream running through them.   The sediments of these standing waters do not have as much access to oxygen compared to their pre-beaver condition and, consequently, organisms capable of anaerobic respiration can thrive. Respiration in the presence of oxygen produces carbon dioxide as an end-product; however, anaerobic respiration can produce methane which is a far more potent greenhouse gas than carbon dioxide.   A team from the University of Saskatoon in Canada have done some calculations and shown that the resurgence of beavers worldwide is now contributing 200 times more methane to the atmosphere than was the case in 1900.

Like all broad-scale modelling studies, the calculations are based on several assumptions and extrapolations, meaning that we need to treat the figures with caution. However, it is useful because it broadens our understanding of how beavers change landscapes from the visually obvious (less club rush, for example) to the more subtle.   And, once again, it challenges our current understandings. The reality is that beaver-mediated methane production underwent a “blip” between the 17th and 20th centuries and is now returning, slowly, to values that our medieval ancestors (those with gas chromatographs, at least) would have recognised.   Even today, greenhouse gas budgets are dominated by natural sources; though this is no cause for complacency: the absence of beavers in effect created a little extra “headroom” into which we pumped the by-products of our addiction to fossil fuels.


Whitfield, C.J., Baulch, H.M., Chun, K.P. & Westbrook, C.J. (2014). Beaver-mediated methane emission: the effect of population growth in Eurasia and the Americas. Ambio 44: 7-15.

Willby, N.W., Perfect, C. & Law, A. (2014). The Scottish Beaver Trial: Monitoring of aquatic vegetation and associated features of the Knapdale lochs 2008-2013, final report.   Scottish Natural Heritage Commissioned Report No. 688.