The democratisation of stream ecology?


About a year ago, I reported on the development of a rapid assessment method for assessing the condition of rivers using algae that are visible to the naked eye (see “Ecological assessment in the fast lane …“).   Last week, our first paper on this subject was published in Science of the Total Environment, a journal which also requires authors to submit a “graphical abstract” (“a single, concise, pictorial and visual summary of the main findings of the article”, according to the Elsevier website). I’ve put our graphical abstract at the head of this post to spare you, dear readers, of the need to read the article.

On the left we have a river then, moving to the right, there are two options: the upper image shows diatoms, followed by a stopwatch indicating the passage of time, and finally a block of five colours which are supposed to indicate the five ecological status classes specified by the Water Framework Directive (blue= high; green = good; yellow = moderate, orange = poor; red = bad). Underneath, we have a picture which is supposed to indicate visible algal growths (green filaments and a crust of Hildenbrandia), followed by a stopwatch indicating the passage of rather less time than is the case for diatoms, leading, again, to the five ecological status classes.   It is, in brief, a concise pictorial and visual summary of why you should not assume that someone with a Fine Art degree can do something that is, strictly speaking, the specialism of someone with a degree in Graphic Design. I make no excuses.

I am, to be honest, rather proud of this paper, largely because it is a piece of “bottom up” science, driven by the needs of end-users rather than, as is often the case, by an agenda set by the high-ups.   It is also a piece of science that eschews the trend to confuse complicated methods with intellectual sophistication.   Unfortunately, this did not endear us to one of the referees.   Only after suggesting, diplomatically, to the editor that perhaps peer review required the reviewer to actually read the paper properly before writing down all of his prejudices, did we scrape through and get our manuscript accepted.

The big change in RAPPER since I last wrote about it, is that we now have two “maybe at risk” categories.   The first (“maybe (1)”) represents situations where there are genuinely ambiguous combinations of algae present at the site, suggesting that there may be some enrichment, but not so much that the sensitive species have disappeared.   The second (“maybe (2)”) represents situations where we did not get a strong signal either suggesting good conditions or enrichment. This happens sometimes, for example when the stream is strongly shaded or (in a few cases) where the dominant algal genus has a very broad ecological tolerance.   In earlier versions, we referred to this latter case as “no data” but, on reflection, it represents a different type of uncertainty and there are usually good reasons why few algae are present at a site.


Graph comparing the TDI (Trophic Diatom Index) scores for sites classified by RAPPER as not at risk, maybe at risk or at risk.   There is a significant difference between TDI scores for the four categories. From Kelly et al. (2016).

One other reason why I am proud of this work is that it is a step towards reconnecting science with non-technical water users. I wrote about this last June (see “So what?”).   I have been concerned for some time that the science of assessing the state of rivers has become so complicated that it now means little to the average angler or to a man walking his dog along the bank.   RAPPER focusses on visible growths of algae, and so there is a more direct link between the outcome of the assessment and what these people actually see.   The “at risk” category, for example, largely consists of sites with a blanket covering of green algal growths on the stream bed and links into some work in the USA which demonstrated how such growths impact upon public opinion of the health of streams and rivers.

This, in turn, raises some further questions about whether the method can be used by people with an interest in the state of rivers but who are not specialist biologists.   The problem is that not all green algal growths necessarily indicate bad conditions (although persistent growths are usually indicative of a problem of sometime). This, in turn, means that the identity of the algae needs to be checked.   At the moment, everyone who has used RAPPER has access to a good microscope.   How well can the method be used by someone with just a hand-held microscope (which rarely exceed 100x magnification)?   That’s the next question that we need to address.   I have been working with a student to explore the potential for modifying the method so that it can be used for “citizen science”, and have some encouraging results.   I’ve also had some useful discussions with a river trust in the north of England and am hoping to meet with angler’s groups too.   That would be a step towards making ecological assessment “a priesthood of all believers” rather than the exclusive preserve of nerds …


Kelly, M.G., Krokowski, J. & Harding, J.PC. (2016).   RAPPER: a new method for rapid assessment of macroalgae as a complement to diatom-based assessment of ecological status. Science of the Total Environment

Suplee, M.W., Watson, V., Teply, M., McKee, H., 2009. How green is too green? Public opinion of what constitutes undesirable algae levels in streams. J. American Water Resources Association 45: 123–140.


Ecological assessment in the fast lane …

Fifteen months ago, I reported on a new approach to rapid assessment of streams and rivers that I was developing with some colleagues (see “A RAPPER in da Lake District …”). Since then, we’ve continued developing and testing the method and, I am pleased to say, the results are very encouraging.

The idea behind RAPPER (“Rapid Assessment of PeriPhyton Ecology in Rivers”) is that a biologist can make a quick examination of the composition of larger algae which can then be used to assess the ecological status of the stream.   This involves recording the presence of different types of algae and, if necessary, checking the identity under a microscope. At the moment, this check needs to be done back in the laboratory, but it should also be possible to do this with a field microscope, which means that results will be available almost immediately.   The method allows sites to be classified as “high or good status” (which are both acceptable conditions according to the Water Framework Directive) or “moderate, poor or bad status” (which means that their condition needs to be improved). The former category is defined by the presence of algae which we know to be sensitive to pollution; the latter by high abundances of algae that we know are tolerant to pollution.

We managed to achieve agreement between RAPPER results and outcomes from more labour-intensive assessments in 72% of cases, which compares favourably with other situations where I have compared different types of algal-based assessments.   Of the remaining sites, 17% either had no appropriate indicator species or low abundances of tolerant species, whilst the remaining 11% had both sensitive and tolerant species present.   After some more analyses, we decided that these sites were probably in an intermediate state of risk.

The sensitivity of the RAPPER classification is highlighted by the charts below. The first shows the relationship between the three classes identified by RAPPER and the accompanying Trophic Diatom Index (TDI) value.   The first three classes show a nice clear separation, with low TDI values associated with the high and good status sites and higher TDI values associated with the more impacted sites. The final class (4_no_data), those for which we had no reliable indication of status, spanned a wide range of TDI values.   There are a number of possible reasons why these sites could not be classified, including a lack of suitable substrata for the algae to grow or heavy shade.  Maybe some extra refinements to RAPPER will mean that we can obtain more reliable classifications but it is important that we do no lose sight of the principle of “rapid assessment” in the process.   Better, perhaps, to reliably classify most sites, in order to free resources to look at those sites that are problematic in more detail?


Box-and-whisker plot showing difference in average TDI values for sites classified by RAPPER as high or good status (“1_HG”), maybe at risk of eutrophication (“2_both”) or moderate, poor or bad status (“3_MPB”) along with sites that could not be classified (“4_no_data”)

The second graph shows the relationship between the RAPPER classifications and soluble phosphorus and a similar pattern emerges, with sites classified as high or good status associated with low phosphorus concentrations, and those classified as impacted (3_MPB) associated with higher values.   Results for total phosphorus, total oxidised nitrogen and ammonia-nitrogen show similar trends, all of which suggest that the method is classifying sites correctly.


Box-and-whisker plot showing difference in median for soluble P for sites classified by RAPPER as high or good status (“1_HG”), maybe at risk of eutrophication (“2_both”) or moderate, poor or bad status (“3_MPB”) along with sites that could not be classified (“4_no_data”).

I’m particularly enthusiastic about RAPPER because it the outcome of convergent thinking between myself and colleagues in two separate agencies dealing with variants of the same need to cover a lot of ground with limited resources. The cost of full-scale ecological assessments is such that the sampling network is relatively thin.   This means that a diagnosis of a problem needs to be followed up by an investigation of the source of that problem in order to develop an appropriate programme of measures to address this. The idea of RAPPER and other rapid assessment approaches is not to replace the official methods, but to enable the operational biologists to collect information from a number of sites in a catchment that would then allow them to use resources more intensively.   RAPPER is, in other words, is part of an ecological “triage”.   And, with the public spending squeeze set to continue, being able to prioritise resources effectively continues to be one of our biggest challenges.