Species richness – what is it good for?

Species richness is the iconic measure of biodiversity. It is simple to interpret* and it is one of the most commonly measured metrics in ecology. From the early beginnings of ecology Darwin, Wallace and von Humbolt noted the striking differences in the number of species found in different places and ecologists are still fascinated by it . However, over the last few months I have begun to question how useful it is for applied research.

These thoughts were probably triggered by two recent papers that suggested that there have not been systematic losses of species richness in plant and animal communities in the last century. This result is startling. Up until the publication of these papers it had been widely assumed that the global extinction of species led to a reduction in species richness at a local scale (although some papers had previously hinted there may be little change in local richness). However, this doesn’t appear to be the case, at least where change in ecosystems is not a result of dramatic changes in land-use. One of these studies found that the species communities investigated had changed in composition over time even if there hadn’t been a reduction in the total number of species. This change in communities could well have resulted in the loss of species considered important for maintaining ecosystem services, such as so-called ecosystem engineers, or species of perceived conservation importance. It is also possible that widespread species may be moving in to colonise areas forcing out rarer species.  However, just by looking at changes in species richness we don’t gain any insight into what species are in these communities.

This begs the question, just what is species richness good for, in the context of conservation and applied ecology? A question I have asked in various different forms on Twitter recently…

 

 

You can probably tell that I am somewhat sceptical about the value of species richness as a metric, even though I have used it in plenty of my own work. However, there are some good reasons for measuring it and like so many things in conservation whether or not you think it is a valuable metric largely depends on what you think conservation is for.

Firstly, there is the utilitarian argument for species richness being important. Countless experiments have suggested that increased species richness leads to increased provision of some ecosystem services, an increase in the number of services that can be provided in any one location and an increase in the stability of provision of these ecosystem services. In addition there have been some studies that show that people value diverse systems with lots of species more than those with less and get more aesthetic appreciation from them. Of course every single one of these statements is contentious, and some have more support than others. However, they represent the broad arguments I have heard in favour of species richness and I largely agree with these points.

While there are sometimes good reasons for using species richness there are also situations in which the logic for the use of species richness as a metric is flawed. One clear problem with it is that it doesn’t tell you anything about the characteristics of the species counted such as their functional traits or how common/widespread they are. When the aim is to conserve species that are most at threat of local/global extinction, as is often the case, measures of rarity are important. An example that I have encountered a lot recently is that related to the debates surrounding land-sharing/land-sparing strategies to protect biodiversity (see my previous posts on the topic here and here). A number of studies suggest that as the proportion of land that is used for low intensity agriculture increases the species richness of an area remains stable until a threshold is passed at which species richness declines. This would suggest that low intensity farming systems retain a lot of their conservation value. However, when you look at the same issue taking into account the identity of the species you see that the populations of many specialist species decline rapidly and the populations of generalists increase, thereby limiting the change in species richness**. If we consider species that are specialists to be of greater conservation concern than generalist species then species richness is clearly not very useful as a metric in this context. The same is true of any study looking at land-use change.

In addition if you want an idea of how a change in land-use might influence ecosystem functions and services using a measure of functional diversity related to their traits is probably more useful than a measure of species richness. This approach is more mechanistic than simply using species richness as a large body of work has already tried to identify which traits link to particular processes and functions. In an ideal world doing this allows you to infer how land use may influence ecosystem processes. Doing this will give us a more general idea of the consequences of changes in biodiversity for human well-being.

I’m sure there are countless other arguments for or against the use of species richness in applied research, so feel free to comment below. At the moment I feel that people measure it because it is relatively easy to measure and because everyone else does. We should think why it might be useful to know about species richness changes rather than just blindly measuring it. So next time you think it would be a useful thing to measure, ask yourself why.


 

*I know richness is not that simple to interpret because it depends on exactly how it has been measured, whether it is rarefied by area or number of individuals etc but I couldn’t be bothered to go into that. Not here, not now.

**I realise we’re getting dangerously close to talking about the intermediate disturbance hypothesis here, but again I’m going to ignore the problems of that for the time-being.

 

Advertisements

22 thoughts on “Species richness – what is it good for?

  1. An interesting and nuanced discussion piece, Phillip. I agree, there are times when species richness is less than useful as a metric and needs to be considered in relation to other measures of biodiversity, including functional traits, species interactions, genetic diversity, etc. It all very much depends on the questions being asked.

    In my opinion, though, species diversity measures (Simpson’s Index, Shannon-Wiener, etc.) are even less useful in conservation as they are so hard to interpret and compare across sites. Personally I find them non-intuitive: give me species richness any day 🙂

    1. Hi Jeff. I’m glad you liked the piece, the thoughts have been knocking around in my head for a bit and I thought it was about time I got them down on paper, even if it didn’t quite come out as I wanted.

      I think this came out of a frustration with many of the papers I read equating greater species richness with greater conservation value and I think this isn’t always true. People inherently value different species differently so a hypothetical species poor community made up entirely of colourful individuals may be of greater aesthetic value than a species rich but largely dull individuals.

      Also I am with you on the other metrics. I stay away from them entirely, except for similarity indices which can be pretty useful.

    2. Species richness compared between sites within the same habitat can be informative. However species richness compared between sites across different habitats provides a less than straight-forward index for biodiversity conservation purposes. In such instances where multiple habitats are of interest across local, regional or even macro-ecological scales, perhaps its more intuitive to consider species richness relative to the habitat specific species pool (i.e. measuring a community completeness index, sensu Pärtel and colleagues (2013. Folia Geobotanica, 48: 307-317).

  2. Interesting piece, but I still think richness is an irreplaceable metric in the applied ecologist’s toolbox. We should, however, not have unreasonable expectations about its utility!

    To use a simple medical analogy, species richness is equivalent to using temperature to diagnose an illness. Any mother, regardless of her medical training, can place her hand on her child’s forehead and feel that something isn’t quite right. However, while knowing how to measure temperature is useful, it doesn’t turn you into a doctor. A trained expert can use all sorts of other metrics (blood pressure, blood smears, biopsies etc) to narrow down the diagnosis.

    Species richness is the same. It’s easy and convenient to measure and interpret… as a first step. But eventually an expert needs to apply more elaborate and sophisticated techniques.

    Some other comments:
    1) For a review of the usefulness of species richness, see:
    Fleishman, Noss & Noon. 2006. Utility and limitations of species richness for conservation planning. Ecological Indicators, 6, 543-553.
    http://www.sciencedirect.com/science/article/pii/S1470160X05000592

    2) You linked to the two papers that failed to show any long-term changes in species richness. While I haven’t read Vellend’s PNAS plant study in much detail, I have strong opinions about Dornelas’ animal paper in Science.

    Basically, I think they didn’t find any trends in richness because their data set is made up (almost exclusively) of localities where you are less likely to expect changes in richness.

    First, they didn’t have many sampling localities in developing countries, especially in the tropics (only 7 of the 37 terrestrial sites were in the tropics). Most of South America, Africa and east Asia were not represented, despite these areas being most affected by biodiversity loss.

    Second, they limited their study to sampling localities with reliable time-series data. This meant that any localities that were completely destroyed before ecologists had the chance to compile a time-series dataset were automatically excluded from their analysis.

    Don’t get me wrong, their study is still seriously impressive: 174 independent time-series, all analysed using a suite a techniques that compared empirical patterns to random expectations. It’s a mind-boggling achievement. However, I don’t think we should be blind to its flaws.

    1. Hey Falko, good to hear from you with a well thought out comment as ever.

      I think I generally agree with you, that species richness can be useful but as Brian says further down it is something of a blunt tool. I’m just a bit tired of seeing it used as the holy grail of metrics.

      Regarding the Fleishman paper I have seen it, I just wanted to voice my concerns on the issue. They make many more points than I have here but I agree with them that species richness is most useful when used along with some other “diagnostic tool” as you might put it.

      Finally I agree that both the Dornelas and Vellend papers have issues that I glossed over. Of course they had limited sampling in developing countries, and in that way they were reflective of the biases of ecology as a whole. I would guess as you say that changes in these locations is likely to be more dramatic. For example the Craigie et al study that found massive declines in mammal populations in protected areas in Africa really shocked me when I first saw it (http://www.sciencedirect.com/science/article/pii/S0006320710002739) and I would imagine that there would have been changes in richness in these locations too. In addition the problem of not sampling time series that were disrupted by dramatic change, such as land-use change, obviously biased the dataset towards showing less dramatic impacts on richness.

      I think that Chris Thomas probably summed it up well in an opinion piece in PNAS (http://www.pnas.org/content/110/48/19187.full.pdf+html) when he said that local richness is almost certainly declining when you include sites that have been subject to agricultural conversion etc. However, it is still pretty remarkable that in sites that don’t fit this description changes in richness show little pattern. I am interested to see where the general discussion of this goes as there are sure to be other meta-analyses that challenge these findings.

      1. Yes, the Craigie paper was a shock to me too!

        They found that mammal populations are declining in protected areas throughout western and eastern Africa, but are relatively stable in southern Africa. I have little doubt that the situation is even worse outside of protected areas.

        When Craigie’s paper came out, I correlated the population declines to the Failed State Index and found quite a strong association. This suggests that conservation success may be linked to the strength of governance.

        Of course, it is probably more likely the result of different baselines. For instance, during the 19th century the animal hide industry in southern Africa caused drastic declines in mammal. So when they started keeping records in protected areas, the populations in southern Africa may have already been depleted.

        The same is true for WWF’s Living Planet index. In 2012 they showed that the index (which is a composite of population trends for several key species) improved by 5% in developed countries, but declined by 44% in developing countries. Again, this is probably due to different baselines, not more effective conservation.

  3. Hi Phil,

    I agree with your points and also believe that Species richness has many flaws, which can be crucial if not taken into consideration. Nevertheless it is the most widely used diversity measurement and especially in macroecological studies or studies relevant to policy makers we often ether (1) have no other available data or (2) it is simply conventional to report a increasing/decreasing number of species. To me it sometimes feel like a p-value. You don’t want to use it, but you somehow feel obligated to report it anyway…
    The underlying drivers of SR (rarity in numbers, habitat size, distribution, evolutionary origin, …) should be critically discussed more often in many ecological studies. Here you mentioned the Phenomena of common species getting more common and rare species becoming more rare. This has been coined “Biotic Homogenization” by a number of people and is getting increasingly popular in the last years. Have for instance a look at Vincent Devictors publications, who has published a lot about this issue both in local- and large-scale studies.

    In my opinion another clearly under-appreciated diversity measurement is Phylogenetic diversity. Not only does it summarize the trait space within a community. Through its evolutionary view it also covers physiological or biochemical traits that are often not considered in estimates of functional diversity, but may be important for the understanding of community assembly and the relationship between diversity and ecosystem functions. The increasing availability of sequences in freely-accessible databases and some later studies (http://birdtree.org/) make these kind of calculations even easier.

    M.

    1. Hey Martin,

      I kind of agree from the policy perspective, like you said it depends largely on what data you have available. I think the kind of indicator used in the UK in which changes in species abundances are calculated (such as those used by BTO etc) are much more meaningful. But I also realise the UK is an outlier in that it is a small place, with lots of people happy to go and count birds/butterflies/plants as a hobby!

      I know a bit about biotic homogenisation (I should do some of my analyses for my new job are looking at it :\) from the work of my friend Sally Keith (http://rspb.royalsocietypublishing.org/content/early/2009/07/20/rspb.2009.0938.full) but I’ll look at bit more at Devictor’s papers when I get the chance.

      On your final point, I have to admit I probably undervalue phylogenetic diversity. I have always been put off by the relative difficulty in actually calculating it and have warned off a student I am helping with an MSc project from using for this very reason. I might be mistaken, and it might be fantastically easy now. Do you know of any good resources to get a taster for what needs to be done to calculate it?

      1. Good points and thank you for the Commentary of Thomas in your other reply. This is pretty interesting and I think I am gonna include this in my thesis.
        And I believed that Sal only does Corals and marine stuff (at least from what I hear). Interesting.

        PD certainly is awesome if it can be done. A successful application often depends on the branch depth (species/genera/family) and if data is available. For instance for birds, mammals, plants and trees there are a lot of phylogeny’s already available as far as I know, but only at the genera level at most. There is even more data scarcity when it comes to reptiles or amphibia. We just have to wait for the next big-shoot science publication of a general phylogeny of these classes.
        For a general easy-to-read overview have a look at Winter and Devictors work from 2012: http://www.sciencedirect.com/science/article/pii/S0169534712002881
        A working example with phylogenic distance can be found in Winter et al. 2009 in PNAS where they used both SR and PD (http://www.pnas.org/content/106/51/21721.full.pdf&embedded=true), but there are probably other more recent examples as well.
        In R there are some pretty awesome packages already. Namely ape, picante and geneland and many many more. Especially Picante makes it incredible easy to calculate phylogenetic metrics –> http://picante.r-forge.r-project.org/picante-intro.pdf
        But remember the GIGO principle 😀

  4. A few points

    1) Richness is just a metric. It’s well defined. I don’t see any argument to say “richness is flawed”. What can be flawed is the use of richness to say something else.

    2) Did richness go down at global scale and remain constant at local scale? This is again a well-defined question that doesn’t leave much room for interpretation. And I see scientific value in answering this question. In general, I find it plausible that human impacts affect richness more strongly at the gamma than at the alpha scale, considering human activities increase dispersal, and ecological theory suggests that niches that become empty locally can be filled by other species if there is enough dispersal capacity. But it may be true that we still need more data to be sure what’s going on.

    3) Does richness itself have any value, and should we protect or plan for it? Or turning the question around, which is the diversity measure that we should conserve? I’m not sure whether there is a definite answer to this, in the same way as there can be no definite answer to the question of what values a society should optimize. It’s a political question. What are the targets we have for conservation? If that is clear, we may have a look whether they correlate with richness. I thought Maier, D. S. (2012) What’s So Good about Biodiversity?: A Call for Better Reasoning about Nature’s Value. Springer made a quite interesting attempt to look at this question from a philosophical viewpoint.

    1. Hi Florian. Thanks for the comment.

      1. I agree, I don’t think the metric is flawed. What I don’t like is peoples use of it that implies it is the most important metric of biodiversity. Sometimes it is, sometimes it isn’t. It depends on what your question is.

      2. I agree, these studies are dramatic but we need more data to work out what is going on, particularly in undersampled regions such as the tropics and developing countries where change may be expected to be more dramatic.

      3. I was kind of trying to get at this in my post, but you have put it better than I did. I agree, I’m not sure there is a definite answer to your questions. As you say it really depends on what society values and how that translates into political will. Regarding the Mair book, I have heard it mentioned a few times and feel like I should give it a look. I’ll put it on my to-read list.

  5. I just came across this piece. I am increasingly thinking you are right. Richness might be our bluntest, least informative possible metric. That might upset the apple cart a bit if that is in fact what the community is discovering, but that is what science is about.

    Looking forward to reading more of your blog!

    1. Hi Brian.

      I agree and really liked your dynamic ecology post on whether scientists need to present a united front. I generally agree with you that disagreement is an important part of the process. Without it we stay with the status quo and science doesn’t advance. it cuts both ways though, and if people start presenting data that goes against what we ‘believe’ we should do our best not to disregard it however hard that might be.

  6. Good post Philip. I see a lot of conservation-oriented studies which draw highly questionable conclusions from using species richness (and as Jeff points out above, other simplistic diversity metrics such as Shannon indices etc. are no better for most conservation purposes).

    If a study finds a large change in species richness, then it’s likely that there is something going on that conservationists should be interested in. No problem with that interpretation.

    The problem is when people interpret no change (or only a small change) in species richness to mean no change in conservation status. We can wipe out a rare habitat and replace it with something banal and see little change in local species richness. Similarly, we can wipe out most of a species’ population and still have it register in our indicators as “present”. But it would be wrong to interpret those results as reasons not to be concerned about the changes that have happened. Species richness is simply not a sensitive enough indicator to tell us about most of the changes that conservationists are concerned about. It would not have detected most of the declines documented by Craigie et al., for example.

    I don’t buy the arguments about ecosystem services or aesthetic appreciation. If those are the outcomes of concern, better to use more direct ways of measuring them.

    1. Hi Ben, nice to hear from you.

      I agree with you on the point of no change in richness being interpreted as no change in biodiversity. This is something that I have come across a lot in the literature on logging where people have argued in the past (e.g. Putz et al 2012) that a small change of ~10% or so means that these forests retain much of their value for conservation. While I think conservation of logged forests is important we still haven’t had that many studies that say what is going on with biodiversity apart from species richness.

      I’m not sure I buy entirely the ecosystem service arguments for measuring species richness, but I was just pointing out that they are valid reasons why people might consider richness to be important. One of the main problems with ecosystem service research (as I have found to my cost) is that measuring services is usually hard, and so people don’t do it or they use a proxy. However, that is a discussion for another time I think.

  7. At the risk of blowing my own trumpet, here’s an example of a study with a more nuanced view of species richness from a conservation perspective. It’s from the PhD work of one of my former students, Sam Tarrant, who was looking at plants and pollinators on restored landfill sites compared to nearby nature reserves. . The headline findings are that (1) landfill sites and nature reserves are very similar in their per-survey species richness of pollinators and the plants they visit; and (2) landfill sites and nature reserves have very similar assemblages of plants; but (3) assemblages of pollinators are rather different in composition across the two types of sites.

    So richness tells us one thing but assemblage structure tells us something else!

    Here’s the reference:

    Tarrant, S., Ollerton, J., Rahman, L. Md., Griffin, J. & McCollin, D. (2013) Grassland restoration on landfill sites in the East Midlands, UK: an evaluation of floral resources and pollinating insects. Restoration Ecology 21: 560–568

    And here’s a link to the paper:

    http://oldweb.northampton.ac.uk/aps/env/lbrg/journals/papers/tarrant-2012-pollinators-on-landfill.pdf

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s