Previous | Next | Scaling problems in biodiversity assessment
Topic: Scaling problems in species diversity measures
Conf: Scaling problems in biodiversity assessment, Msg: 6721
From: James Mallet (j.mallet@ucl.ac.uk)
Date: 18/03/2005 10:58 AM

Scaling problems in species diversity measures James Mallet james j.mallet@ucl.ac.uk Scaling problems in species diversity measures
Jim Mallet, UCL London

Studies of biodiversity and biodiversity conservation efforts seem to find it hard to get away from the use of species richness as a metric in comparisons of biodiversity in different areas or different taxonomic or functional groups.

It seems to me that the problem is language. We need words such as "species" to express concepts. However, these concepts may be fuzzy, little different from criteria for overlapping clusters of things. Problems crop up when users of words such as "species" begin to read too much into the concept, and start to perceive the species as a "real thing".

Now of course, I don't mean that robins and blackbirds are not distinct, or that field guides don't work; what I mean is that species are not real in the sense of being connected together evolutionarily, so that there is actually no a-priori, or god-given location in the biodiversity hierarchy where the term "species" must logically apply. It is more like a cake: you cut it where it is convenient to make bite-sized chunks.

Beginning around 1890, until around 1980, a steady consensus had been building among animal biologists (at least) to adopt a more inclusive "polytypic concept" or "biological species concept". If there was some doubt whether two divergent populations, if overlapping, would remain distinct, that is they were suspected or observed to form intermediate or "hybrid" populations, they would be placed often as subspecies within a large Rassenkreis or polytypic species.

Unfortunately, instead of justifying this by a practical argument of consensus or international agreement, Theodosius Dobzhansky and especially the late Ernst Mayr, justified the polytypic species on the grounds that reproductive isolation was the "real" essence of species, so that the absence of blending in overlap was the true, underlying reality of species, rather than merely a cause of species distinctions in overlap, and a convenient stopping place in taxonomy. Nonetheless, this worked fine from the 1940s to about 1980, when Hennigian and other ideas became more prevalent.

Quite suddenly, in the last part of the last century, fixed morphological or genetic differences began to be used to distinguish species, with proponents arguing that the phylogeny is, if you like, the "real" basis of species, instead of reproductive isolation.

We are now in a situation where conservationists, biodiversity specialists, and other users of taxonomy wouldn't be able to tell how many species there are, even if all the taxa were known. There has been recent exponential growth in the numbers of "species" of groups such as primates, while groups that have received less recent taxonomic attention, such as the carnivores, have not increased at all (Isaac et al. 2004, Trends in Ecology & Evolution 19(9): 464-469). Opinions differ as to how much difference this species concept change will make; some estimates give as little as two-fold change, but my belief, partly based on my knowledge of butterfly taxonomy, is that each of the current species can very likely be divided still further, so that an order of magnitude won't be improbable.

In the changes in species richness in vertebrates, there have been relatively few descriptions of new taxa; mostly what has happened is that taxa formerly considered as subspecies have been elevated to species status. The new species are usually although not always) related to the older more inclusive species via hierarchical links. The supporters of the biological species concept and supporters of the more finely-divided phylogenetic concepts both agree on the evolutionary explanation for hierarchy, but merely disagree on the precise level of "real" species.

This seems a strange situation to be in. There are scaling issues here, and it seems likely that the phylogenetic species diversity and the biological species diversity will be approximately, although not perfectly correlated, as has been found in other biodiversity scaling issues, and perhaps it doesn't matter too much which one we use so long as everyone agrees to use the same criteria in comparisons. (But this should be very much under study, as underscored by Jack Lennon's and Jean-Luc Solandt's contributions in this series). The problem is that we can't do this; different groups, and organisms on different continents are currently classified differently.

I don't know what the answer is, but it does seem to me that it might be useful to come to an international agreement about species level and stick with it. The trouble is that there really are a lot of people who feel that their particular reality of species is the correct one and that using some other criterion would be wrong, and would misinform in biodiversity studies. My own personal preference is for a more inclusive, polytypic style species, because it makes sense in two areas:

(1) what speciation is -- it doesn't mean much more than "evolution" if it simply means fixing a new genetic or morphological marker in a separated or partially separated population.

(2) local, alpha-diversity at the species level -- is only enhanced by the ability of two populations to overlap; it doesn't increase just because of genetic divergence across a taxon's range.

Furthermore, there might often be jobs for the more finely-divided phylogenetic species; but these needs could easily be accommodated by the use of the existing subspecies rank in Linnean nomenclature. It would just need acceptance that sometimes there will be paraphyletic polytypic species that have budded off local species that are more closely related one or more of the parent's subspecies -- of course this is anathema to phylogeneticists.

But of course there are exceptions and intergrades here as well. For example, in spite of the fact that we know they hybridize and blend together freely in zones of contact, I have yet to find any modern herpetologist who would designate Bombina bombina and Bombina variegata as subspecies within a species (even among supporters of the biological species concept). So we are always going to have taxonomic anomalies, that may differ in different groups even if we had a strong international agreement to iron them out.

So, even if we had a perfect world where all nations agreed to use the same type of species concept (which I strongly doubt will happen any time soon!), we therefore would still need a second approach:

Hendry et al. (2000, Conserv. Genet. 1, 67-76) have suggested doing away with species in conservation altogether, and using genetic metrics to express biodiversity. While I understand this argument, we are a word-based culture, and words like "species" can be useful even though they do not express underlying realities. I can't see any reason why we can't continue to use the word "species", more carefully, and I strongly doubt that the word will go away in conservation, and especially in the public mind, any time soon.

Instead, we ourselves have to become more educated, and we have to educate all those politicians and the public better, about the fuzzy nature of our most widely-used metric of biodiversity, the species. In our comparisons and counts of biodiversity, we must make this clear at every opportunity, and try to make sure that our analyses in macroecology and conservation science are robust to precise estimates of species numbers. If we fail to emphasise to users the lack of certainty of this particular biodiversity metric, errors will certainly be made.