Such expressions as that famous one of Linnæus, and which we often meet with in a more or less concealed form, that the characters do not make the genus, but that the genus gives the characters, seem to imply that something more is included in our classification, than mere resemblance. I believe that something more is included; and that propinquity of descent,—the only known cause of the similarity of organic beings,—is the bond, hidden as it is by various degrees of modification, which is partially revealed to us by our classifications (Darwin, 1859, p. 413f).

Saturday, 19 September 2009

(Not) Lars (Brundin)

Nestling, possibly unread by many, in the first comment on the announcement of the downloadable copy of Systematics and Biogeography, is an offering from Professor Lars (of whom about we know absolutely nothing). At first we thought it a jape, one of our friends or colleagues trying to tease us. But no! (“Blimey!!”, that was Williams [he’s British]; “Cricky”, that was Ebach [he’s Australian]). We read the text closely and could see that the well thought out and reasoned commentary was, indeed, real – a series of penetrating observations on the state of systematics today – and, of course, the follies of the past. We were humbled in its presence – but unwilling to let it pass and, because we are both very humble people, have decided to bring it to wider attention. Read on and enjoy, and consider what might be the follies of present day systematics and systematists.

Lars said...
I can't see why cladists (remember this term was first used by Ernst Mayr to distinguish them from other schools) consider themselves "revolutionaries". Cladistics is only efficient when one is dealing with morphology, and this kind of information is so terribly biased by subjectivity, that it should never be used as phylogenetic inferential data.

Molecular data, on the other hand, is much less biased (bias comes from sequencing errors, lateral gene transfers and poorly chosen alignment parameters) and should reflect the correct evolutionary relationships if correctly analyzed. With the advance of genetic barcoding, the Cladistic methods has become obsolete.

Morphology can be interpreted in many ways by different authors, and given the infinity of manners the same information can be scored, it becomes not much more than an exercise of subjectivity.

Cladistics is traditionally viewed by serious molecular biologists as a sectarian (almost a religious cult / secret society!) branch of evolutionary research that claims to possess the most efficient and and best logic to propose hypotheses on organismal relationships.

Unfortunately, this is not true. Cladists are often narrow-minded and do not accept their "parsimony" method is much more prone to LBA artifact than those usually referred to as "Phenetics", such as Neighbor-Joining and Maximum-Likelihood. This last is the pinnacle of evolutionary inference, since it uses both raw nucleotide data AND genetic distance in its calculations, thus using all data explanatory power.

Hence, considering the high subjectivity implied in morphologic matrices, the risk of LBA bias, and the suboptimal use of data explanatory power, Cladistics should be avoided.

This makes Phenetics, and not Cladistics, the actual revolution!


Martin Brazeau said...

Well, I'm glad that's settled! Can we proceed to throw in the towel on comparative developmental biology, too? After all, since morphology is so subjective how can we be sure of anything in developmental biology? I'm guessing the solution to this problem will be 'genes', too?

Lars said...

As far as I can see, morphology has its role as a descriptive tool. This includes morphology applied to developmental biology too. However, when one intents are to assess the evolutionary relationships among species, it requires data which can fulfill an objective inferential framework, or the error will be enormous.

To this, genes suffice.

But genes are data. They need to be analyzed according to robust methods which can provide an account on error measurement and that can handle all possible information, regardless on the researcher's subjectivity on homologies. This is where Maximum-Likelihood algorithms enter.

In the context of Paleontology, however, DNA is rarely available, thus morphology can be used as a second-choice kind of data. But how can it be scored? Morphometric characters are indicated, as they can be converted to distance matrices.

Obviously, morphological features are associated to phenotype evolution, and hold information on organismal natural history. But the amount of convergences and the subjectivity on descriptions is too large to be overlooked. Thus, only a few well-known and potentially inclusive characters can be used with confidence. But keep in mind that different morphological information will produce different trees. This is another major drawback of morphology. But it seems cladists will never agree!
For this matter, you should read the paper by Scotland et al. (2003)*.

Genetic barcoding provides an efficient tool for taxonomists, since they can almost perfectly be used to identify known species and to estimate the occurrence of organisms yet unknown to science. Remember that morphology can be treacherous, as ecotypes and intraspecific polymorphism do exist! Furthermore, genetic barcoding, although more expensive than simple morphology-based taxonomy, can significantly address the issue of taxonomic impediment.

*SCOTLAND, R. W., R. G. OLMSTEAD, AND J. R. BENNETT. 2003. Phylogeny
reconstruction: the role of morphology. Syst. Biol. 52:539-548.

Wills Flowers said...

Genetic barcoding, now that it is being used on a wide scale, is showing just as many pitfalls and confusing pictures of evolution as, well, morphology. With paraphyletic and polyphyletic gene trees widespread (Funk & Omland 2003), classical morphology stops looking so "treacherous". Every caution Lars cites for morphology also applies to genes. And despite some initial propaganda, even partisans of barcoding no longer claim that barcodes will do anything significant to address the taxonomic impediment.
For some recent publications (excluding the outright polemics for or against barcoding) see also Little & Stevenson 2007, Meyer & Paulay 2005, and Song et al. 2008.

Funk, D.J. and K. E. Omland. 2003. SPECIES-LEVEL PARAPHYLY AND POLYPHYLY:Frequency, Causes, and Consequences, withInsights from Animal Mitochondrial DNA. Annu. Rev. Ecol. Evol. Syst. 2003. 34:397–423.

Little,D.P. and D.W. Stevenson. 2007. A comparison of algorithms for the identification of specimens using DNA barcodes: examples from gymnosperms. Cladistics 23 (2007) 1–21.

Meyer CP, Paulay G. 2005. DNA barcoding: Error rates based on comprehensive sampling. PLoS Biol 3(12): e422.

Song, H., J.E. Buhay,, M.F. Whiting, M.F., and Keith A. Crandall. 2008. Many species in one: DNA barcoding overestimates
the number of species when nuclear mitochondrial pseudogenes are coamplified. PNAS 105: 13486–13491.

Lars said...

You should read the paper by Packer et al., 2009.

It is a magnificent work that explains in detail and reasonable argumentation how DNA barcoding overcame morphology.

Whilst morphologic descriptions may be necessary as scientific information on species, it is completely irrelevant for taxonomic purposes. DNA barcoding can address this task faster and with more accuracy, without the need of cumbersome taxonomic keys and decades of work.

The paper can be downloaded from: