[Taxacom] Annual biogeographers' Exam Q6

[John Grehan]

Annual Biogeographers' Exam Question 6

What did Trewick and Wallis (2001) get wrong about the Alpine Fault? (bonus
point – give an example of a transcurrent fault correlation).

The idea that rocks and living communities have both undergone lateral
displacement along the Alpine fault was criticized by Wallis and Trewick
(2001). They concluded that distribution data can only be used to frame
evolutionary and biogeographic hypotheses, not to test them. Yet many
authors, including Charles Darwin, have acknowledged the significance of
distribution, and have accepted it as evidence in its own right. In an 1845
letter to Hooker, Darwin referred to”…that grand subject, that almost
keystone of the laws of creation, Geographical Distribution” (Burckhardt
and Smith, 1987: 139), and he used distributional data both in framing and
in testing hypotheses. In any case, Wallis and Trewick (2001) concluded
their critique of the Alpine fault hypothesis of community disruption by
accepting that it could explain the distribution of some taxa.

In another study, Trewick and Wallis (2001) examined a selection of
widespread South Island invertebrates in order to investigate the possible
effects of fault displacement. Rather than testing the idea with detailed
mapping (this would require field work in mountainous areas with difficult
access), they used a molecular clock approach. They found that most of the
clades are older than the Pleistocene. This indicates that glacial
extirpation has *not* been responsible for the biotic gap, and the authors
accepted that taxa have survived the Pleistocene glaciations more or less *in
situ*, perhaps on nonglaciated ridge above glaciated basins. They argued
against the Alpine fault hypothesis, as they found that clades are not old
enough, but most of the fault movement has been more recent than they
assumed (Heads and Craw, 2004).  In addition, Trewick and Wallis (2001)
relied on the standard insect rate (2.3% divergence per million years)
criticized elsewhere (Heads, 2012a), and the clades are probably older than
they indicated. In a later paper, Wallis and Trewick (2009) admitted that
the Alpine fault hypothesis was a “old suggestion,” but they concluded that
“Despite its appeal there are numerous problems” (p. 3562). The only
problem they mentioned is that the disjunct groups have been estimated to
be too young, but they dates relied on flawed calibrations.

Trewick and Wallis suggested that the Alpine fault disjunction was caused
by greater uplift and consequent extinction in the central areas, but
transcurrent movement also occurred with uplift, and extinction does not
account for the presence of sister groups or endemism in the gap. Trewick
and Wallis (20001) also found that in many cases populations in Nelson and
Otago are not sister groups, and they suggested that this is incompatible
with the fault hypothesis. Yet there is no need for groups that were
adjacent before fault movement, for example, in Nelson and Otago, to be
sister-groups, and so they will not be sister groups after fault movement.
The fault hypothesis is compatible with a range of different distributions,
and the distribution of clades after fault movement depends on heir prior
distribution. For example, taxa that were widespread before fault movement
will be widespread after.

The Alpine fault pattern, as with all common biogeographic patterns, is
seen in many taxa with different ecologies and phylogenetic branch lengths
(degrees of differentiation). This suggests that the distribution pattern
is not determined by ecology and that branch length in the different groups
is not determined simply by time.