[Taxacom] Geophylogeny

[John Grehan]

Courtesy of Bob Mesibov I was able to start reading Kidd's paper. Below
are my responses to the introduction section. As this is quite long and
tedious for those not interested in biogeography I will comment on other
sections separately. While in sympathy with some of the goals expressed,
I found myself wondering what purpose it served as it seems to present a
revisionist history of biogeography that pretends that certain things
have never happened. It seems that Kidd will now make the dramatic foray
into the future and no doubt justify a nice beg technically complex
grant application.

I appreciate that most on this list will probably not agree with me (and
I already know I am intemperate, curmudgeonly etc etc), so please hold
off on expressing exasperation, but counter points on anything below
will be welcome.

John Grehan

Introduction

"the MoL will stimulate the generation of new scientific hypotheses
through the juxtaposition of previously disparate information"

Sounds a bit like a political slogan


"Comparative analyses of interspecific biogeographical pattern will
facilitate the identification of the geographical and ecological context
surrounding the evolution of specific traits, speciose radiations,
patterns of extinction, and community assemblage"

Really? Note reference to "will" as if such analyses do not currently
exist.

"Hypotheses regarding the relative roles of dispersal and vicariance in
speciation will be interpretable in a geographically explicit
framework."

As if there is none already.

"Similarly, analyses of the spatiotemporal history of genetic variation
will link changing landscapes with the frequency and intensity of past
population bottlenecks and expansions."

More slogans

"In addition to the scientific rewards, wider societal benefits can be
envisaged through the explicit presentation of the intimate connections
between all life in space and time and between life and the history of
the planet through an everyday visual metaphor, the geographical map."

"Querying the MoL would allow a student to discover whether a species
seen in their locality arrived a few thousand years ago with the
Holocene warming or evolved in situ several millions of years ago."

This is just tripe. A map will not tell you that. The age of a taxon is
derived from other information other than the map itself.

"They could quickly locate where its closest relatives live and access
information explaining why such closely related species are found in
different places."

The irony of this is that this method already exists in the form of
panbiogeographic mapping.

"The MoL could show the location of fossils, the ranges of related
species, and the distribution of genetic variation from which histories
are inferred."

Already demonstrated in panbiogeography.

"More than 163 million species observations are currently available
through the Global Biodiversity Information Facility
(http://www2.gbif.org/annual report 2008.pdf), 814,000 fossil records
are served by the Paleobiology database (http://www .paleodb.org/), and
the International Union for Conservation of Nature (IUCN) provides range
maps for all mammals (http://www.iucnredlist.org)."

These resources are well and good, but are dependant on how much control
there is over the identification of the taxa in question.

"the Paleoclimate Modelling Intercomparison Project II provides 18
climate models covering the last 21,000 years
(http://pmip2.lsce.ipsl.fr/), whereas Paleomap (http://www.scotese.com/)
provides continental plate reconstructions over the last 1.1 billion
years."

Of course these are all just theories, not empirical facts. Even
geologists refer to their models are 'cartoons'.

"Synthesis of this diverse and disparate information will require an
informatics infrastructure in which both primary data and derived
information are interoperable and can be combined through common
geographical, temporal, taxonomic, and phylogenetic referencing"

This is just fluff. Synthesis does not come from simply dumping stuff
together, but through the conceptual framework being applied. What is
the conceptual framework being applied to these resources?

"With the help of expert knowledge, ranges can be inferred directly from
species, gene, and fossil observations; the IUCN mammal range database
was created in this manner."

Duh. Ranges are inferred from where they occur now and as fossils. Not
sure about what genes are supposed to tell here.

"Alternatively, ranges can be estimated algorithmically using minimum
convex polygons"

Here the author finally begins to be explicit about his map concept
which is tied to the concept of a range map (in contrast to
panbiogeography that uses localities connected by a line graph).

"or historically inferred from phylogenetic (Ronquist 1997; Ree et al.
2005) or phylogeographic
(Vogel et al. 1999; Avise 2000; Lemmon A. and Lemmon E.M. 2008)
pattern."

Whatever that is supposed to mean.

"Inference of historical ranges is likely to remain problematic as
direct evidence is often limited and inference algorithms lack realism"

So why bother?

"However, confidence in reconstructions can be increased if alternative,
preferably independent, approaches generate consistent inferences"

In other words it might all be crap, but we will do it anyway -
especially as we have so many good technical resources that we just have
to use (I've seen the same style of argument used in phylogeny)

"Where alternative historical scenarios exist, methods can be applied
that compare the likelihood of competing hypotheses."

Yes one can dream up all sorts of things.

"a recent maximum likelihood approach (Ree et al. 2005; Ree and Smith
2008) allows competing historical scenarios, which differ in a variety
of constraints which might include area connectivity through time or the
location of fossils, to be ranked and likely geographical pathways
identified."

In other words you can subordinate your theorizing according to some
imagined theory of likelihood with reference to imagined "geographically
pathways".

"As long as the explanatory premises that underlie reconstructions are
necessary conditions and are explicitly stated, reconstructions can be
falsified when one of the premises is rejected (Andersson 1996)."

Perhaps whether this has any real meaning will be explained later in the
paper.

"More traditional historical biogeographical approaches that seek
congruent patterns between multiple phylogenies (for reviews, see Crisci
et al. 2003; Morrone 2009) will also provide valuable information."

A meaningless statement.

"Collaboration to build the MoL can only enhance mutual understanding of
these different methods and has the potential to stimulate the
development of new methods that combine the strengths of individual
approaches."

As the MoL has not yet been characterized one might wonder.