[Taxacom] Bubble science

[John Grehan]

Science in a bubble

 In these days of mass communications and electronic accessibility of
information one would think that the present world of research would be
much advanced over what prevailed 20 or even 10 years ago. Then it was much
more difficult to track down all pertinent information and publications and
it took a lot more time and digging (literally digging through books and
journals at times). Now there are electronic resources that would reduce
that time and effort and improve efficiency of coverage (although I find I
still miss items).

 All those improvements are nothing if they are not sought, and in some
sciences they definitely are not. Primate biogeography and systematics is
one such area, where researchers live in science bubbles that float away by
their denial of anything other. This is amply illustrated in a recent paper
by Anne Yoder (2013) titled “The lemur revolution starts now: The genomic
coming of age for a non-model organism” (Molecular Phylogenetics and
Evolution 66: 442–452).

 In the abstract she states “New technologies for gathering and analyzing

genomic data will allow investigators to build upon what can now be
considered a nearly-known phylogeny of the Lemuriformes in order to ask
innovative questions about the evolutionary mechanisms that generate and
maintain the extraordinary breadth and depth of biological diversity within
this remarkable clade of primates.”

 Her thesis is that the phylogeny of the Lemuriformes is “nearly-known”
through genomic data (sequence analysis). The tensions in her thesis begin
early in the introduction where she states

 “Although the first decade or so of this revolution relied upon indirect
measures of genetic distance such as DNA–DNA hybridization, numerous
breakthroughs in our understanding of evolutionary relationships were
achieved, such as the (very controversial, at the time) finding that
chimpanzees are more closely related to humans than to gorillas (Sibley and
Ahlquist, 1984).”

 She does mention that there was some opposition to the chimpanzee model:

 “Several authors took exception to these results in particular, both in
terms of the obvious incongruence with the morphological details shared by
chimpanzees and gorillas (reviewed in Holmquist et al., 1988), but also due
to various subtleties of statistical analysis (Farris, 1985; Templeton,
1985).”

 This borders on the truly bizarre, that it is worth mentioning
morphological ‘details’ shared by chimpanzees and gorillas, but never mind
the existence of those shared by humans and orangutans. Bubble science at
work.

 Of course, in the mind of Yoder (and pretty much all of her
contemporaries) the truth of phylogeny lies in DNA sequences alone:

“The dispute was for many settled decisively by Felsenstein (1987) who
employed a maximum likelihood mixed model analysis of variance method to
show that there was indeed significant support for the human-chimp clade
contained within the DNA–DNA hybridization data published by Sibley and
Ahlquist, 1984.”

 Her principle focus is on the origins of Madagascan lemurs. She notes that:

“…in the early to mid-1980s, nearly all primate classifications (Fleagle,
1988; Schwartz, 1986; Szalay and Delson, 1979) placed one of the lemuriform
groups, the dwarf and mouse lemurs (family, Cheirogaleidae), into the
Lorisiformes due to their shared and otherwise unique condition of the
cranial blood supply (Cartmill, 1975; Szalay and Katz, 1973).”

 The biogeographic consequences of these relationships are seen as:

“In both cases, that of the dwarf lemurs and of the aye–aye, a paraphyletic
Lemuriformes would necessitate at least two crossings of the Mozambique
Channel (Yoder, 1996; Yoder et al., 1996a).”

 But of course only if one assumes dispersal or constructs a divergence
model requiring such crossings.

 Under the section on lemur biogeography Yoder states:

 “In order to ask how lemurs arrived in Madagascar, however, we first need
to understand when.”

 “Given that Madagascar has been surrounded by an oceanic barrier for at
least 88 my, we must conclude that dispersal, not vicariance, would have
been the mode of their arrival.”

 We must? Must? Oh yes, Yoder has said so and so shall it be. And as to how
they arrived, one must surely appeal to the only remaining implausible
alternative:

 “With these terrestrial routes rejected, we are left with one remaining
alternative: implausible as it may seem (Stankiewicz et al., 2006), lemurs
must have dispersed via rafting across a formidable oceanic barrier,
perhaps aided by an ancestral capacity for heterothermia (Kappeler, 2000).”

 Must have, must have, and must have. If it is said enough times it will be
so.

 To Yoder, timing is everything, but divergence estimates (from sequence
analysis) are variable:

 Estimated ages range from the late-Cretaceous (Arnason et al., 2008), to
the early- to middle-Paleocene (Perelman et al., 2011; Roos et al., 2004;
Yoder et al., 2003, 1996a; Yoder and Yang, 2004) to the early- to
middle-Eocene (Dos Reis et al., 2012; Yoder et al., 1996a).

 With all this variation Yoder asks:

 But how do we decide which estimate is best supported? She notes that
there are a variety of problems with rate estimation and the “problematic
nature” of the fossil record. She concludes with the view that these
problems will be overcome in the technological advances of phylogenomics
and in the end:

 “we can begin to explore what might have been the specific adaptive
advantage that allowed lemurs to endure what must have been a treacherous
journey from Africa to Madagascar, and potentially, the genetic advantages
that allowed them to cement their survival and diversification upon
arrival.”

 Yoder’s paper reads to me as a case of a world view that is established by
leaving out what might otherwise expose the internal tensions. The
possibility that morphogenetic evidence may represent an alternative model
is dismissed out of hand. And the problematic nature (never really
specified) of fossil calibration is never identified, never mind that the
traditional use of fossil calibration in divergence estimates consistently
misrepresent minimal ages as actual or maximal.


Most glaring is the total absence of the vicariance model based on
molecular evidence presented in Heads (2012) and (2009). Now perhaps Yoder
may be forgiven for having missed the book as it may not have been reviewed
in Molecular Phylogenetics and Evolution, but missing the 2009 article
borders on the fantastic. If Yoder failed to do any basic literature search
on the web, then surely one of her many colleagues would have brought it to
light. But perhaps not? This is the trouble with science living in a bubble.


John Grehan