[Taxacom] Describing genera without molecular phyolgies

[Richard Zander]

Lynn:

Gaps are, of course, the first things taxonomists look for as signals that species and higher taxa are evolutionarily both distinct and self-coherent. That is, that the subunits of each are closely related through descent and preserve their particular direction(s) of evolution.  

A somewhat better criterion for taxon coherency is the idea that there is an empirical taxonomic unit higher than species in which such coherency is generated.  In my opinion, that is the idea that there may be one or more descendants of some generalized progenitor. Each descendant is identical (except for minor anagenetic changes) to the progenitor plus at least two additional traits that make it a different species (one trait may just be a rogue mutation). Progenitor-descendant relationships mostly without trait reversals are iron linkages because they are second order Markov chains, one supports another and the second supports a third. This is why traits of such a linkage may be assigned one Shannon informational bit for each trait, and all the traits can be added to result in a very high Bayesian posterior probability. 

Thus a coherent radiative genus consisting of well nested descendants excludes other genera, including, say, a one-species genus. The one species genus may be a remnant of an ancient lineage. Removing morphological traits that are considered highly adaptive in other related genera, i.e., are found in specialized descendants, leaves conservative traits that are at least hints of a long-lost lineage. Such an intellectual adventure is suppressed by cladistics tree-thinking, that is, sets of species each evolving into two smaller sets of species, which is not a model of evolution.

Rather than wondering where are the gaps in difficult groups, one might search for a progenitor (a modern species with the same traits as a putative ancestor) and see if other species can be nested around it as adaptive or nearly neutral descendants. Radiation does not need to be adaptive, which is now pretty much accepted. Modern genera, I think, have many radiative genera in them and no one has looked for them as empirical taxa higher than species. At least I have demonstrated these in my own publications.

Consider a large group you are familiar with. Can you split it into two or more radiative genera on the basis of expressed traits (i.e., those actually involved in evolution)? I bet you can. Publish. The rationale is available to counter the squawks of phylogenetic reviewers and editors.

Richard Zander



-----Original Message-----
From: Lynn Raw <lynn at afriherp.org> 
Sent: Sunday, May 24, 2020 3:45 AM
To: Richard Zander <Richard.Zander at mobot.org>
Cc: JF Mate <aphodiinaemate at gmail.com>; Richard Pyle <deepreef at bishopmuseum.org>; Taxacom <taxacom at mailman.nhm.ku.edu>
Subject: Re: [Taxacom] Describing genera without molecular phyolgies

Richard, I did qualify my statement with the further reference to Mayr that a genus is a single species or monophyletic group which is separated from other taxa by a decided (morphological) gap which is in inverse ratio to the size of the taxon. What I meant by arbitrary was the decision as to where that gap lies. Depending on where you decide to make the division everything within the genus can be monophyletic, even though there are smaller monophyletic groups within it. The smallest of these might be a single species, subspecies or population but we usually choose natural groups with more distinctive differences to set the boundaries.

> On 24 May 2020, at 00:51, Richard Zander via Taxacom <taxacom at mailman.nhm.ku.edu> wrote:
> 
> Lynn Raw wrote that a genus is purely an arbitrary concept. No it isn't, that's a sad meme. Those groups of species we intuitively identify in nature are actually clusters of descendants radiating away from some generalized progenitor. I've tested this purely empirical concept of genus, and it works as an explanation. See my recent stuff on ResearchGate.
>