Cryptic dinosaur diversity, ‘real taxon’ counts, curse of the nomina dubia, and the holy grail: matrix-assisted laser desorption ionization

The unthinkable has happened. After more than three months of work, the 20,000-word, 100-page review manuscript on British dinosaurs has finally been submitted (by the way, 20,000 words is horribly close to the 30,000 words required as minimum for a phd science thesis). Well, actually, it hasn’t been submitted, but it’s now out of my hands and will be submitted within the next few days, so for the purposes of this blog post let’s just pretend that it has been submitted. I learnt a lot from the project, and it also got me thinking about some issues that extend beyond the subject of British dinosaurs.

Here’s what’s bothering me. Of 108 named British dinosaur taxa (representing the valid species and the nomina dubia), over 50% are nomina dubia (that is, based on remains that lack autapomorphies [unique, diagnostic characters]). It was only natural that in the discussion section, we (= Naish & Martill) made a few comments on British dinosaur diversity: on how the taxonomic spread compares with global dinosaur diversity, and on how many species we have for each group. The contention we hold (in the present version of the MS: this might change after review) is that the ‘real diversity’ preserved in the fossil record is higher than the diversity count indicated only by the valid taxa. How so?

Why valid species aren’t the only species we should count

Consider that, rather than being useless, indeterminate objects that do not really reflect biological entities, some nomina dubia surely represent real taxa, it’s just that we lack the data to diagnose them by autapomorphies. Many nomina dubia come from geological units where there are no named valid taxa. While a nomen dubium might lack unique characters, if it’s not definitely referable to anything else, then surely it really is a valid taxon? There’s already a precedent for recognising taxa that lack autapomorphies: they’re called metataxa, and I think that at least some nomina dubia should be regarded as such. As soon as they are, the ‘real taxon’ count goes up. In the case of British dinosaurs, it does so by 20%.

Then there are unnamed taxa: specimens that can be identified to a clade but remain unnamed because, like nomina dubia, they lack autapomorphies. In many cases, however, they represent the only member of their group known from the relevant unit. In the British record, examples of such are the Bexhill diplodocid, Wessex Formation rebbachisaurid (see previous post on Wealden sauropods) and a large theropod from Stonesfield that is not the same thing as Megalosaurus. While such specimens shouldn’t be given binomials (pending the discovery of better material), they should be regarded as countable taxa. Inclusion of them also therefore ups the ‘real taxon’ count, and in the case of British dinosaurs, that count now increases by a further c. 22%.

These observations mean that diversity counts for dinosaurs (and other fossil groups) are necessarily under-estimates, as they only account for named valid species. But, if anything, dinosaur diversity studies have been ultra-conservative, and have not only ignored this cryptic diversity, they have also down-played even the counts of named valid taxa. In a recent talk on dinosaur diversity, Mike Benton (2005) argued that dinosaurs were more taxonomically inflated than most other animal groups. Across living and fossil taxa, it has been shown that about 20% of all ‘taxa’ eventually turn out to be synonymous with others (Alroy 2002), but for dinosaurs Mike argued that the figure was more like 50%. So while dinosaur workers are patting themselves on their backs for being in a golden age of taxonomic discovery (the number of named dinosaurs has increased 70% since 1990), Mike’s contention is that most of this is over-zealous and/or over-enthusiastic. But, hold on, exactly who says that as many as 50% of dinosaur taxa are junior synonyms? I’m not going to name names here, or go into this in any depth (it would involve detailed discussion of specific groups and their taxonomic histories), but I am not convinced that this view is valid.

‘Un-lumping’ in the extant tetrapod fauna, and the recognition of cryptic diversity

It is very easy to be critical of those systematists that we term ‘splitters’: those who erect new species and/or genera based on the smallest of differences. Mostly the criticisms are justified, as true splitters really do use the tiniest, tiniest little differences to erect their new species/genera, and subsequent work usually shows that these ‘species’/’genera’ fall within the range of individual, ontogenetic or sexual variation of a previously named taxon. Fair enough, I’ve collapsed over-split taxa myself (Martill & Naish, in press [on azhdarchoid pterosaurs]). But there is less criticism of those systematists termed ‘lumpers’: those who synonymise species and/or genera and are prepared to accept considerable variation within their concept of species and genera. And it seems that, with dinosaurs, the lumping may actually be more laissez-faire than the splitting: in the relevant cases, the workers concerned have synonymised taxa because of gestalt similarities. NOT because detailed, careful work has shown that the taxa really can be regarded as conspecific or congeneric.

If dinosaurs really do get lumped more than they already are, it strikes me that the trend within dinosaur systematics is in direct opposition to that occurring among extant tetrapods. No, the number of recognised taxa is not going down because of increasing rates of synonymy, it is going up as specialists, working across various tetrapod groups, are resurrecting taxa from synonymy, or are discovering cryptic diversity and thereby coining new species and genera.

On the first point (resurrection of taxa from synonymy), it seems that the 20th century trend of synonymising the many taxa erected by Victorian biologists was often more laissez-faire and over-zealous than was the splitting that occurred before. Indeed the standard taxonomic compilations that exist for extant tetrapods all now seem guilty of lumping on a massive, uncritical scale. Judging by statements made in the literature, this appears to be the case for big mammals (see previous post on giraffes) and birds more than for other groups. It’s easy to observe this in action on bird systematics: most specialists agree that certain ‘standard’ reference works (e.g. Peters 1951) were just too heavy-handed with their lumping. Judging by the ‘un-lumping’ that’s gone on so far, there is every indication that the count of c. 9000 extant bird species usually given in the literature is an under-estimate.

On the second point (the discovery of cryptic diversity), DNA-based studies on extant species seem – as a generalisation – to mostly discover that groups include more diversity than was thought before, not less. For a few good, specific examples of this sort of thing, see Wüster & Thorpe (1994), Mayer & von Helversen (2001), Glaw et al. (2001), Glaw & Vences (2002a, b), Vallan et al. (2003), Parra-Olea et al. (2005) and Olsson et al. (2005). Hand-in-hand with this is the debate over the status of ‘subspecies’, a complex issue that I’ll avoid here. I agree with those who argue that the subspecies concept is not useful, and that definable taxonomic units are the same thing as species (Zink 2004). Again, when such taxonomic units are transformed into species, diversity counts go up. For specific examples see Collins (1991) [where 55 North American amphibians and reptiles were raised to species rank]; Sibley & Monroe (1991) [where the number of extant bird species was upped from c. 9000 to 9672]; Robertson & Nunn (1998) [where 10 albatross taxa were raised to species rank]; and Alström et al. (2003) [where 16 wagtail taxa are noted as ‘new’ species if the phylogenetic species concept is applied].

On birds again, one study has even claimed that the recognition of all of this cryptic diversity ‘would lead to a doubling of the currently recognized species diversity of birds’ such that there might be 20,000 extant species (Zink 1996). Indeed it is widely acknowledged by those who have criticised both lumping, and the subspecies concept, that lumping is potentially harmful both to diversity studies and to conservation. It essentially downplays true potential diversity and diverts interest, research and conservation concern, plus most subspecies were named for convenience anyway and don’t reflect real evolutionary units (Zink 2004).

In view of both the discovery of cryptic diversity and the transformation of ‘subspecies’, and keeping in mind that a significant percentage of extant species are all but indistinguishable osteologically, it is a safe assumption that any measure of diversity among fossil species is an under-estimate. Ralph Molnar (1990) made this point some time ago, but I’ve yet to see a study on dinosaur diversity that acknowledges or refers to his point. I realise that I’m on dangerous ground in employing this argument, as it is simply not possible to examine fossil taxa at the same level of detail. The new data on extant diversity mostly comes from genetics, and there isn’t too much data of that sort available when we’re thinking of Mesozoic animals. So this is sort of an apples and oranges situation, and claims that ‘fossil species and/or genera are over-lumped’ come down to nothing more than arm-wavy assertions or speculations. Given confounds like ontogeny and individual variation, and then the really really cool stuff like phenotypic plasticity and resource polymorphism, any claims that fossil taxa are, or are not, synonymous are by nature subjective, and actually unknowable. Right?

A holy grail for palaeodiversity?

So wouldn’t it be nice if there were some sort of holy grail. Some magical method, akin to mtDNA analysis, that might allow us to work out whether two fossil specimens were or were not members of the same species or genus? Well here’s the big news: there is. Sort of.

Osteocalcin is a mineral-associated protein, exclusive to vertebrates, that occurs within bone matrix. It is in fact the second most abundant protein within bone, and only a tiny amount of bone is needed to extract it. After extraction and purification, the protein is sequenced with matrix-assisted laser desorption ionisation mass spectrometry (MALDI-MS). Here’s the big news for the world of palaeontology: (1) osteocalcin sequences appear to be genus-specific, and (2) osteocalcin remains stable and detectable in old fossil bone and, while to date it’s only been applied to Pleistocene fossils, no-one yet knows how far back it can be detected. Presumably this is because no-one has yet really looked (please tell me if you know otherwise). I have been reliably informed by a geochemist than there is no theoretical reason why osteocalcin can’t be detected in bone that is tens or millions, or hundred of millions, of years old.

To date two studies have applied MALDI-MS to fossil samples, and both produced positive, highly encouraging results. Nielsen-Marsh et al. (2002) looked at fossil bison sequences, and Nielsen-Marsh et al. (2005) looked at sequences in Neanderthals. In both cases, the fossil taxon proved to have osteocalcin sequences that were, firstly, identical to those of its congeneric living relative (B. bison and H. sapiens respectively), and, secondly, distinctly different from those of related genera (Bos in the case of Bison, and Pan, Gorilla and Pongo in the case of Homo). The sequences cannot, it seems, allow species to be differentiated, but they can resolve generic membership.

Let’s make a big assumption and imagine that osteocalcin sequences can be recovered from (say) Mesozoic dinosaur taxa. Unfortunately, we’re not going to be able to resolve synonymy at the specific level, nor discover cryptic species in the same way that DNA analysis can. But we would be able to resolve matters of generic membership: whether two genera are really synonymous for example (e.g. Tyrannosaurus and Tarbosaurus), and whether a nomen dubium represents something distinct, or is just a non-diagnostic remnant of another taxon. Such knowledge would have a huge impact on diversity studies. I think this has the potential to be something pretty big, IF osteocalcin really does prove as resistant, accessible and widespread among fossils as has been implied.

Where do we go from here, and are we on the brink of a revolution in palaeodiversity studies?

PS - for the latest news on Tetrapod Zoology do go here.

Refs - -

Alroy, J. 2002. How many named species are valid? Proceedings of the National Academy of Sciences 99, 3706-3711.

Alström, P., Mild, K. & Zetterström, B. 2003. Pipits and Wagtails of Europe, Asia and North America. Christopher Helm (London), pp. 496.

Benton, M. J. 2005. The discovery pattern of dinosaurs … and how many more species are to be found? In Barrett, P. M. (ed) 53rd Symposium of Vertebrate Palaeontology and Comparative Antomy, Abstracts. The Natural History Museum (London), p. 8.

Collins, J. T. 1991. Viewpoint: a new taxonomic arrangement for some North American amphibians and reptiles. Herpetological Review 22 (2), 42-43.

Glaw, F. Vences, M. 2002. A new cryptic frog species of the Mantidactylus boulengeri group with a divergent vocal sac structure. Amphibia-Reptilia 23, 293-304.

- . & Vences, M. 2002. A new sibling species of the anuran subgenus Blommersia from Madagascar (Amphibia: Mantellidae: Mantidactylus) and its molecular phylogenetic relationships. Herpetological Journal 12, 11-20.

- ., Vences, M., Andreone, F. & Vallan, D. 2001. Revision of the Boophis majori group (Amphibia: Mantellidae) from Madagascar, with descriptions of five new species. Zoological Journal of the Linnean Society 133, 495-529.

Mayer, F. & von Helversen, O. 2001. Cryptic diversity in European bats. Proceedings of the Royal Society of London B 268, 1825-1832.

Molnar, R. E. 1990. Variation in theory and in theropods. In Carpenter, K. & Currie, P. J. (eds) Dinosaur Systematics: Approaches and Perspectives (Cambridge University Press, Cambridge), pp. 71-79.

Nielsen-Marsh, C. M., Ostrom, P. H., Gandhi, H., Shapiro, B., Cooper, A., Hauschka, P. V. & Collins, M. J. 2002. Sequence preservation of osteocalcin protein and mitochondrial DNA in bison bones older than 55 ka. Geology 30, 1099-1102.

- ., Richards, M. P., Hauschka, P. V., Thomas-Oates, J. E., Trinkaus, E., Pettitt, P. B., Karavanic, I., Poinar, H. & Collins, M. J. 2005. Osteocalcin protein sequences of Neanderthals and modern primates. Proceedings of the National Academy of Sciences 102, 4409-4413.

Olsson, U., Alström, P., Ericson, P. G. P. & Sundberg, P. 2005. Non-monophyletic taxa and cryptic species – evidence from a molecular phylogeny of leaf-warblers (Phylloscopus, Aves). Molecular Phylogenetics and Evolution 36, 261-276.

Parra-Olea, G., Garcia-Paris, M., Papenfuss, T. J. & Wake, D. B. 2005. Systematics of the Pseudoeurycea bellii (Caudata: Plethodontidae) species complex. Herpetologica 61, 145-158.

Peters, J. L. 1951. Check-List of birds of the world. Volume VII. Museum of Comparative Zoology, Cambridge, Massachusetts.

Robertson, C. J. R. & Nunn, G. B. 1998. Towards a new taxonomy for albatrosses. In Robertson, G. & Gales, R. (eds) Albatross Biology and Conservation. Surrey Beatty (Sydney), pp. 13-19.

Sibley, C. G. & Monroe, B. L. 1991. Distribution and Taxonomy of Birds of the World. Yale University Press, New Haven.

Vallan, D., Vences, M. & Glaw, F. 2003. Two new species of the Boophis mandraka complex (Anura, Mantellidae) from the Andasibe region in eastern Madagascar. Amphibia-Reptilia 24, 305-319.

Wüster, W. & Thorpe, R. S. 1994. Naja siamensis, a cryptic species of venomous snake revealed by mtDNA sequencing. Experientia 50, 75-79.

Zink, R. M. 1996. Bird species diversity. Nature 381, 566.

- . 2004. The role of subspecies in obscuring avian biological diversity and misleading conservation policy. Proceedings of the Royal Society of London B 271, 561-564.