? Page 2 ? Biogeograf?a 2 ? July, 2007 ?
L E A D  A R T I C L E
An alternative Gondwana: Biota links South America,
New Zealand and Australia
BY DENNIS MCCARTHY, MALTE C. EBACH, JUAN J. MORRONE AND LYNNE R. PARENTI
Ever since Eduard Suess first theorized its existence in
an effort to explain the distributions of Glossopteris,
Gondwana has been a biological or paleontological
phenomenon. Today, it remains an elegant symbol of
the intricate relationship between geological and bio-
logical history. Some early 20th century paleontolo-
gists, such as Hermann von Ihering, believed that ?only
the history of life on the Earth enables one to grasp
the geographical transformations of the past? (Ihering
in Wegener 1966, vii, original italics). Ihering?s senti-
ments are echoed by L?on Croizat?s ?Earth and life
evolve together? (Croizat 1964, p. iv), which trans-
formed 20th century biogeography. Gondwana was
discovered by biogeographers, biologists and paleont-
ologists, who identified repeated distributions across
impermeable barriers such as oceans and mountain
chains. Land bridges became an all-purpose, theoreti-
cal ?escape clause? to explain trans-Atlantic distribu-
tions of terrestrial and freshwater organisms.
Throughout most of the twentieth century, geolo-
gists dominated the study of Earth history, which in
turn, rigidly molded our views of historical biogeogra-
phy. Geologists looked at the surface of the Earth from
a static perspective. The Earth provided an inert and
ever-stationary landscape over which life randomly
dispersed. The hypothesis of a planet Earth that did
nothing more than rotate and tumble around the sun
was not open to any further speculation. The end of
the apparent battle between Vulcanists and Neptun-
ists saw an end to geologists at war (see Hallam 1983).
Geology today, as then, is a highly organized disci-
pline that frowns upon unconventional interpretations,
especially those supported by non-geological evidence.
Perhaps this is why biogeographers, a highly disorga-
nized, multidisciplinary group of biologists and geog-
raphers, continue to challenge geological interpreta-
tions of the Earth. Taxic distributions do not match
current palaeoreconstructions of the Earth. Geologists
may dismiss this simply as ad hoc dispersal, flora and
fauna trekking across the globe in a random manner.
So great is this belief that many biologists, not accus-
tomed to challenging geological hypotheses, quickly
accept this view of the world.
Despite the conventional interpretations, there are
indeed patterns: the same, repeated distributions ex-
hibited by a great number of unrelated organisms. It is
these patterns that caught the attention of Alfred
Wegener, Alexander Du Toit and Dwight Taylor. Many
like Wegener did not see such patterns as evidence that
would necessarily engage the geological establishment
at the time. Instead they sought geological or geophysi-
cal evidence, which at the time was difficult to come by
or required equipment that was not available or was
financially crippling. The reconstructions of a large land-
mass called Gondwana that existed 65 million years ago
was dismissed universally as unrealistic. Ironically
enough, when the static Earth view was overthrown, it
was that same reconstruction that would immediately
become accepted as ?the model? for geology. Why
would a reconstruction that was ridiculed for over 100
years suddenly be accepted without question? Where
were the alternatives? Why has everyone favored that
particular continental drift reconstruction? The answer
lies in our world view.
Everyone who owns a map of the world will see
Africa placed proudly in the middle and Alaska and
eastern Russia framing the world on each side (Figure
1). Australia is usually crammed in at the bottom right
with New Zealand in tow. The single largest part of
the Earth, the Pacific ocean, is mercilessly cut in half,
reducing it both in apparent size and relevance. There
are few large countries in the middle of the Pacific.
This was the map that Wegener and many other sci-
entists grew up with, and it is from this map that
Gondwana, in its present reconstruction, was con-
ceived. East Africa was placed against the west coast
of South America. Madagascar was connected to In-
dia to form Sclater?s (1864) Lemuria. Land bridges lit-
erally paved the way in reconstructing Gondwana by
indicating the most parsimonious routes that a land
mass could take. But what if we were to take a differ-
ent look at the world ?by using a map which has the
Solomon Islands placed at its centre?
Unlike most other global maps, Figure 2 features a
? Biogeograf?a 2 ? July, 2007 ? Page 3 ?
whole Pacific. Larger nations have been inconveniently
placed to either side of the Pacific and Indian oceans.
The Americas are squeezed to one side and Europe
and Africa have been pushed to the other. In the middle
is the big blue Pacific Ocean with a date line straight
down its middle. If we are fortunate to own a geo-
logical map with the same projection we see that the
Australian plate is immense, with its eastern edge ex-
tending well beyond New Zealand (Figure 3). Wegener
did not own such a map but certainly should have
been aware of the taxic distributional patterns across
the Pacific. Many Australian and New Zealand taxa
share close relationships to South American taxa. Many
Southeast Asian taxa exhibit a close relationship with
Mesoamerican taxa, and North East Asian taxa appear
closely allied with Western North American taxa. Had
Wegener and others used a Pacific-centered map they
may have also seen that the Australian and Asian con-
tinents fit neatly into the continental margins of South
and North America, respectively. If such maps had been
conventional, perhaps a different reconstruction of
Gondawana would have resulted, one that had no
Fig. 1. World map centred on the Atlantic Ocean. Fig. 2. Alternative map featuring a whole Pacific Ocean.
Fig. 3. Map of the tectonic plates of the world.
Fig. 4. Alternative reconstruction of the paleoconti-
nent Gondwana.
Panthalassa, a pre-Pacific super-ocean stuck between
the east coast of Australia and western coast of South
America, but instead expanded the current Indian
Ocean between the Western Australian coast, east-
ern Africa, Madagascar and India ?situated like a gi-
ant Tethys or ?Lemurian? Ocean (Figure 4). Such an
alternate Gondwana has rarely been considered de-
spite the hoariness and notoriety of trans-Pacific dis-
junctions. Instead, land bridges, island hopping, sea-
mount hopping, and long distance dispersal have long
been favored as explanations for such problematic
distributions. Viewing the world from a Pacific per-
spective 100 years ago may well have started wide-
ranging endorsement for this alternative Gondwana
?a reconstruction that would not be suggested until
the 1970s (see Carey 1976).
The difference between the speculative
and the confirmed in plate tectonics
One of the most significant scientific advances in the
? Page 4 ? Biogeograf?a 2 ? July, 2007 ?
twentieth century was the theorization and empirical
validation of plate tectonics. Evidence supporting plate
tectonics includes geodetic data (GPS, VLBI, etc.) that
have confirmed motion of plates, matching most plate
tectonic predictions to a high degree of accuracy; seis-
mic tomography showing cool, dense slabs penetrat-
ing as much as 660 km into the Earth at trenches;
magnetic anomaly data and confirmation of seafloor
spreading, etc. Researchers have also gathered a con-
siderable amount of evidence supporting numerous
aspects of the theoretical paleomap of Pangaea first
put forth by Wegener (1912). This includes matching
continental outlines and geological ties linking regions
across the Atlantic and Indian Oceans, juvenile age
(<200 ma) of the Atlantic and Indian seafloor; identi-
cal or closely related, poor-dispersing taxa, both fossil
and extant, linking South Africa with South America
and India with Madagascar. This latter evidence pro-
vides the simplest and most compelling case for
Pangaea/Gondwana, and its probity seems even more
obvious today than when Wegener first highlighted it
nearly a century ago. For example, the fit of South
America into Africa and the fossil taxa that these re-
gions share ?like the small aquatic reptile Mesosaurus?
usually appear in the introductions to the subject of
plate tectonics in grade-school text books.
Still, a few aspects of the classical Pangaean/Gond-
wanan reconstructions have not enjoyed similar em-
pirical support and have been repeatedly scrutinized
and challenged. For example, the classical depictions
of Pangaea (e.g., Barron et al. 1981) show the South-
ern Asian regions of Lhasa, Qiangtang, Sibumasu,
Tarim, West Burma, North China, South China, and
Indochina as part of Laurasia, separated from the
Gondwanan regions of India and Australia by a wide
Tethys Ocean. A few researchers, however, have chal-
lenged this depiction, noting both geological and bio-
geographical links that would seem to contradict the
view of such a vast separation (Shields 1979; Carey
1988, pp. 158-161). Since that time, it has now be-
come conventional that these presumed Laurasian re-
gions actually began in the Paleozoic as northern
Gondwanan terranes, juxtaposed with India and Aus-
tralia. Thus, while a few biogeographers noted that
many of the closely-related and apparently trans-Tethys
taxa shared by these regions (trilobites, brachiopods,
corals and stromatoporoids, nautiloids, gastropods,
conodonts, etc.) would have had a difficult time cross-
ing the full extent of an ocean, it has become widely
accepted that these taxa did not in fact cross an ocean
(e.g., Metcalfe 1998 and references therein). Likewise,
Briggs (2003) has also recently challenged the con-
vention view of Late Cretaceous India as an isolated
micro-continent in the middle of the vast Tethys Ocean
due to poor-dispersing taxa shared by India, Mada-
gascar and Southern Asia. These taxa include iguanid
lizards, boid snakes, crocodilians, two families of frogs,
a mammal, freshwater ostracods, and charophyte al-
gae (see McCarthy 2005 and references therein.) Once
again, another hypothetical oceanic gap between a
number of closely-related fossil taxa has been removed
from paleomaps.
These recent shifts in paleomaps should give bio-
geographers pause whenever they are confronted with
paleo-oceanic disjunctions, particularly involving ter-
restrial or shallow marine taxa. It may also be benefi-
cial for researchers to understand precisely why paleo-
maps are still subject to extreme theoretical shifts in
the locations and extent of marine gaps ?despite the
overall success of plate tectonics.
As shown earlier, evidence for certain aspects of
Pangaea ?like the connection of South America and
Africa? follows from independent analyses of a num-
ber of facts from a number of disciplines. This is posi-
tive, existing evidence from biogeography and geol-
ogy that all support the same conclusion. Theories re-
garding the extent and location of ancient seafloor,
however, have the peculiar problem that no seafloor
in the world is older than 200 ma, and the majority of
it formed in the last 60 ma. Thus, the extent and place-
ment of most Early Tertiary or Mesozoic marine gaps
is merely speculative. For example, Briggs has no theo-
retical concerns in shrinking the Late Cretaceous Tethys
gap between northern India and southern Asia because
all of the Tethys has been subducted anyway. Certainly,
we have marine deposits throughout northern India
and southern Asia, requiring the existence of an epi-
continental Tethys seaway throughout much of the
Mesozoic. But evidence confirming the actual extent
or duration of the north-south oceanic gap between
Gondwanan and Laurasian regions simply does not
exist. Biogeographical patterns, at this point, can be
used to test various theories of the extent, timing, and
location of marine gaps.
This conservative approach to paleomaps, in which
an effort is made to distinguish between the specu-
lative and the implied, is used here with respect to
Southern Trans-Pacific distributions. This will ensure
that we do not dismiss repeated examples of pat-
terns of disjunctions, all strongly supporting vicar-
iance, simply because such a conclusion conflicts with
certain conventional paleocartographic assumptions
about the location and extent of ancient and now-
vanished seafloor.
The biotic links among Australia,
New Zealand, and South America
Southerly trans-Pacific disjunctions are among the most
conspicuous and notorious of all known distributional
patterns. In The theory of island biogeography, Mac-
? Biogeograf?a 2 ? July, 2007 ? Page 5 ?
Arthur and Wilson (1967) provided four examples of
major questions typically considered by biogeogra-
phers. One of the four was: ?How can we account for
the phylogenetic similarities of the biotas of southern
South America and New Zealand?? (MacArthur and
Wilson 1967, p. 5). In a now classic paper on biogeog-
raphy, Lars Brundin stated that ?Among the problems
raised by the distribution of plants and animals in the
southern hemisphere, there is none which takes a more
central position and is more stimulating to the imagi-
nation than the problem of transantarctic [trans-Pa-
cific] relationships? (Brundin 1966, p. 46). Brundin
speculated that by 1966 the vast literature on the sub-
ject of southern disjunctions ?would probably by far
exceed 1000 titles? (Brundin 1966, p. 49).
Significant studies of trans-Pacific disjunctions had
already begun by the middle of the nineteenth cen-
tury, and Joseph Hooker, focusing on plants, was per-
haps the first to rule out dispersal:
?Enough is here given to show that many of the
peculiarities of each of the three great areas of land in
the southern latitudes are representative ones, effect-
ing a botanical relationship as strong as that which
prevails throughout the lands within the Arctic and
Northern Temperate zones, and which is not to be
accounted for by any theory of transport or variation,
but which is agreeable to the hypothesis of all being
members of a once more extensive flora, which has
been broken up by geological and climatic causes?
(Hooker 1853-55, p. xxxvi).
Brundin (1966) also quoted this remarkable descrip-
tion of Southern Hemisphere vicariance and recounted
many of Hooker?s figures on trans-Pacific flora. For
example, Hooker counted 76 genera shared by South
America and New Zealand, 17 of which did not live in
Australia. Padding Hooker?s list, Brundin added trans-
Pacific disjuncts from the Diptera, among which
Brundin estimated the existence of no fewer than 50
trans-Pacific groups that implied direct land connec-
tions. Brundin, writing at the advent of the continen-
tal drift revolution, used the distributions to argue for
Antarctica as a migration route, hence the phrase
?transantarctic distributions?.
A number of recent papers that have also focused
on southern trans-Pacific disjunctions have only
strengthened the trans-Pacific bond. Sequeira and
Farrell (2001) provided a table of twenty-four disjunct
groups of Gondwanan insects at various taxonomic
levels, from family to genus. All twenty-four live in both
Australia and Central/South America, and twenty of
the twenty-four live exclusively in these Gondwanan
regions. The other four also live in South Africa. This
should seem remarkable given that in conventional
reconstructions Australia and South America are on
opposite sides of Gondwana with South Africa in be-
tween them.
Heads (1999) provided a detailed analysis of the
multiple trans-Pacific disjunctions among the plant-
genus Abrotanella and its relatives, concluding that
vicariance was the only reasonable explanation for the
distribution and that the sister taxa linked a number
of trans-Pacific regions, including northern New Guin-
ea-western USA-Chile, East Tasmania-Chile, and Stew-
art Island-southern Chile. Cranston and Edward (1999)
studied two sister group relationships among the in-
sect genus Botryocladius (Diptera: Chironomidae), link-
ing Tasmania to Chile. They also concluded the distri-
butions supported vicariance. Shields (1998) studied
Triassic fossil flora from Argentina and Queensland,
Australia, noting the regions share so many identical
or closely related fossil plants that the distributions
are most reasonably explained by vicariance.
Other taxa linking Australia with South America in-
clude meiolaniid turtles and monotreme mammals.
Other taxa linking New Zealand with South America
include the lichen Fuscopannaria subimmixta, the ev-
ergreen Laurelia novae-zelandiae (New Zealand) and
its Chilean sister, Laureliopsis philippiana, and the flat
oyster Ostrea chilensis. New Zealand?s lizard-like Tuat-
ara is a living fossil, an example of the last remaining
sphenodon ?and its most recent fossil relatives are the
Late Cretaceous sphenodontians of Patagonia (Apeste-
guia and Novas 2003). The freshwater fish, Galaxias,
and the beech-like trees, Nothofagus, both boast mul-
tiple trans-Pacific sisters linking all three regions ?South
America, New Zealand and Australia. McCarthy (2003)
noted many of these distributions conformed to a very
precise pattern, linking specific regions of Eastern
Australia and New Zealand with specific regions of
South America, again implicating vicariance.
More recent studies have likewise strengthened the
biotic bonds across the Southern Pacific. Giribet and
Edgecombe (2006), for example, constructed a cla-
dogram for the stone centipede Paralamyctes using
morphological and molecular data. They noted that
Paralamyctes was a great candidate for cladistic analy-
ses because it occurs on all Gondwanan continents
except Antarctica, has limited dispersal capability, and
has very narrow distributional limits. Because of the
centipede?s distribution, Giribet and Edgecombe em-
ployed narrow areas of endemism, or ?microareas,?
such as Chile and North New South Wales ?rather than
the broader continental regions typically used. Their
analysis uncovered a repeated pattern of southern
South America (Australia, New Zealand) in their sub-
trees. Relevant to the discussion here, Giribet and
Edgecombe concluded:
?The microarea approach does not prohibit ?reas-
sembling? continental areas to explore alternative pat-
terns. For example, it remains possible to retrieve a
repeated pattern of (southern South America (Austra-
lia + New Zealand)) in our subtrees?, corresponding
? Page 6 ? Biogeograf?a 2 ? July, 2007 ?
to the so-called ?plant southern pattern? of Sanmart?n
and Ronquist (2004)? (Giribet and Edgecombe 2006,
p. 76).
Sanmart?n and Ronquist (2004) used parsimony-
based tree fitting in conjunction with permutation tests
in order to uncover biogeographical patterns among
Gondwanan taxa in what they described as ?the larg-
est biogeographic analysis of the Southern Hemisphere
attempted so far,? involving 54 animal and 19 plant
phylogenies (1393 terminals). Significantly, they  over-
came sampling bias by providing results in terms of a
ratio of dispersal events ?that we would have seen if
all areas had been equally well represented in the data
set? (Sanmart?n and Ronquist 2004, p. 230). Figure 5,
from Sanmart?n and Ronquist (2004), display their re-
sults, detailing frequencies and directions of dispersal
(as inferred by Tree-fitter) of terrestrial plants and ani-
mals superimposed on PT polar views of the Southern
Hemisphere in the Late Cretaceous-Early Tertiary. The
frequencies are proportional to arrow thickness. Note
that the large arrows pointing into Antarctica (from
South America and Australia) indicate taxa that have
moved between South America and Australia. Antarc-
tica was not part of the analysis, as ?none of the stud-
ied groups are now present there?. The arrows point-
ing to Antarctica merely reflect the theoretical as-
sumption of Sanmart?n and Ronquist that all the taxa
shared by Australia and South America had to move
through Antarctica. Also, while the one Antarctic ar-
row appears to be pointing from Western Australia,
Australia is treated as a single biogeographic unit,
and the arrow is not meant to designate any particu-
lar part of Australia. Many of the Australian taxa in
the data set of Sanmart?n and Ronquist were refer-
enced by McCarthy (2003), and all of those taxa oc-
cur in eastern Australia.
The figures are significant because they display in
a visually intuitive way the ease of dispersal (or range
expansion) between many pairs of Gondwanan re-
gions. And as is clear from a glance, New Zealand,
Australia, and South America share more ancestral taxa
than any other pairs of Gondwanan regions studied
by Sanmart?n and Ronquist. For example, although
Madagascar was directly attached to Africa in the Early
Cretaceous and has never been separated by more than
400 km from its continental neighbor, plants and ani-
mals have managed to move with a greater facility
between New Zealand and South America than be-
tween Madagascar and Africa. Or to put it another
way, even though New Zealand and South America
have allegedly never been connected and are currently
separated by the most significant oceanic barrier on
the globe, the plants and animals of New Zealand more
closely resemble that of South America than Madag-
ascan taxa resemble that of Africa. Unflappable is the
biogeographer who is not astounded by that fact.
Fig. 5. Maps from Sanmart?n and Ronquist (2004, figs.
9, 10), illustrating the frequency and direction of dis-
persal of terrestrial animals and plants in the southern
hemisphere.
Thus, the mystery that Hooker faced in the middle of
the nineteenth century still taunts us today: What is
the explanation for these trans-Pacific biotic links?
? Biogeograf?a 2 ? July, 2007 ? Page 7 ?
Solutions
Essentially all researchers who have provided wide-rang-
ing biogeographical analyses of South America-Austra-
lia/New Zealand have concluded that many of the bi-
otic links among the regions are indicative of vicariance
(Hooker 1853-55; Croizat 1964; Brundin 1964; Matile
1990; Crisci et al. 1991; Shields 1998; Heads 1999;
Glasby 1999; Cracraft 2001; Cranston 2003; McCarthy
2003, 2005). Even Sanmart?n and Ronquist (2004), who
posit long distance dispersal for the biogeographical
pattern displayed by plants, agree that the animal links
are best explained by vicariance. Apparently, the only
disagreement involves the question of the vicariant
event, that is, the researchers disagree about the geo-
logical cause that led to the disjunctions.
Matile (1990), Shields (1998), Glasby (1999), Glasby
and ?lvarez (1999), Cranston (2003) and McCarthy
(2003, 2005) agree that the most parsimonious expla-
nation for these distributions is a past juxtaposition of
the vicariant regions, New Zealand, Australia and South
America. Followers of panbiogeography (Croizat 1964;
Heads 1999) appear to support something not too dis-
similar ?the juxtaposition, at the very least, of the dis-
junct terranes upon which the taxa are found and a
Mesozoic (or at the very least a late) birth of the Pacific.
This conclusion is consistent with all other analy-
ses of similar biogeographical patterns across other
marine barriers. In fact, in nearly every other case, when
researchers conclude that the great number of biotic
ties currently shared by pairs of trans-oceanic regions
is the result of vicariance, they invariably mean that
these currently separated regions used to be juxta-
posed. All geologists and biogeographers now agree
that the fossil taxa shared by South America and Af-
rica, on opposites sides of the Atlantic, or shared by
India and Madagascar, separated by the Indian Ocean,
are evidence the regions were once united. As noted
earlier, since the paper by Metcalfe (1998), it is now
convention that the fossil taxa shared by East Asian
regions and Australia in the Paleozoic, classically de-
picted as separated by a vast Tethys, are evidence that
the regions were together. Moreover, Briggs (2003)
has now argued that in the Late Cretaceous India and
South Asia, classically depicted as being separated
by a vast Tethys, were actually in proximity ?a view
that is likely to become convention soon. As shown,
the South Pacific regions Australia, New Zealand and
South America flaunt an even greater number of
trans-oceanic disjunctions of poor dispersing taxa,
providing even more compelling evidence for juxta-
position. The researchers who have closed the gap
between these trans-Pacific regions use the same ar-
guments and follow the same line of reasoning as
those who have closed the trans-Atlantic, trans-In-
dian, and trans-Tethys disjunctions.
Those committed to conventional paleomaps, how-
ever, must seek alternative explanations, which is to
say, they must make an arbitrary distinction between
vicariant signals across the Pacific and those across
other oceans. For the plant ties, for example, Sanmart?n
and Ronquist suggest long-distance, trans-Pacific ?con-
cordant dispersal,? perhaps using Antarctica as a step-
ping stone. This theory of coordinated-jump-dispersal
across the full breadth of an ocean, first employed by
followers of continental fixity to explain the trans-At-
lantic and trans-Indian Ocean disjunctions, has a con-
spicuous problem (Figure 6). New Zealand is currently
8000 km away from South America and 2700 km away
from Antarctica at its closest point. Australia is an-
other 2000 km west of New Zealand. (For the sake of
comparison, the distance across the Atlantic between
Brazil and West Africa is 3000 km). Yet despite these
great distances, a great number of the taxa alleged to
have jump-dispersed the Pacific are restricted to a very
narrow range of regions and are absent from all other
Southern Pacific Ocean islands and other nearby conti-
Fig. 6. Distance from South America to Australia in
Gondwana and its comparison with the current width
of Eurasia.
? Page 8 ? Biogeograf?a 2 ? July, 2007 ?
nental locations. If these taxa can really jump-disperse
such vast distances why would they not appear on any
other oceanic islands or any other continents?
The common sense notion that taxa that can jump
disperse an ocean should be wide ranging often ap-
pears in many geological primers that discuss Alfred
Wegener?s simple argument that a number of trans-
Atlantic disjunctions suggest a closed Atlantic Ocean.
For example, the following quote appears in a parent/
teacher guide for introducing the concepts of plate
tectonics on an educational website (Robinson 2005):
?Firstly, [Wegener] was able to show that fossils of a
small reptile called Mesosaurus were found only in
South Africa and Brazil. These two places are now sepa-
rated by the Atlantic Ocean, so how could Mesosaurus
have lived in both locations? The chance of it crossing
the ocean was very small and, if it had been able to
cross the sea, why hadn?t it been found in other places
as well? To Wegener it seemed that the only sensible
explanation was that South Africa and Brazil had been
joined together at the time when Mesosaurus lived?
(Robinson 2005, http://www.geography-site.co.uk/
pages/physical/earth/tect.html).
 Another educational website, ?Geology Basics?
from Carleton College (Daum and Savina 2000), stated
this idea more simply:
?While it is remotely possible that Mesosaurus could
swim the ocean and simply chose to live in eastern South
America and west Africa exclusively, it is much more
likely that these two areas represent what was a con-
tinuous range several hundred million years ago, when
the reptile lived and the two continents were joined?
(Daum and Savina 2000, http://www.acad.carleton.edu/
curricular/GEOL/classes/geo120/geology_ basics.htm).
The problem, as noted by these elementary websites,
is not simply the assumption that so many poor dis-
persing taxa have managed to jump disperse an ocean;
the major difficulty is that these taxa never ended up
anywhere else. And why would this argument be any
less compelling for disjunctions across the Pacific than
the Atlantic? This question seems particularly trouble-
some given that the Pacific is nearly three times wider,
the Pacific has a far greater number of oceanic islands
to support would-be crossers, and trans-Pacific disjunc-
tions involve a far greater number of taxa.
The theory of full ocean jump-dispersal of a wide
variety of taxa exclusively located in a very limited num-
ber of narrow regions not only wars upon reason, it
also conflicts with all known empirical data. The Pitcairn
Island group, for example, is only half way across the
Pacific. Yet a wide ranging study of all 114 species of
flora on the Pitcairn group confirms that it does not
exclusively share any plant with South America. Or to
put it another way, every plant taxon that has man-
aged to disperse between South America and Pitcairn
has also managed to colonize other islands significantly
closer ?such as Easter Island or the Cook Islands. If a
plant taxon can send propagules 4800 km to Pitcairn, it
can send significantly more 2400 km, 1200 km, 600 km,
etc. New Zealand is 8000 km away from South America.
Yet very few of the full-Pacific crossers that have been
the subject of such controversy managed to reach any-
where else. Even using Antarctica as a stepping stone
would not help resolve this problem as it still requires a
number of significant dispersal events.
Moreover, concordant cross-ocean dispersal cannot
explain all of the disjuncts, and Sanmart?n and Ronquist
(2004), like Brundin (1966) and Cracraft (2001), accept
that many of the distributions are indicative of vicariance
(Sanmart?n and Ronquist specifically posit vicariance for
animals.) But as noted conventional paleomaps demand
that these researchers interpret the vicariant signal in a
unique manner. Noting the current paleomaps never
place the regions together, they contend that the
vicariant explanation is consistent with the regions
merely being part of the same multi-continental land
mass. That is, they argue that South America, Australia
and New Zealand were ?connected through Antarctica.?
But it should be noted that South America and Africa
are also currently part of the same land mass, ?con-
nected through? Eurasia and North America. When the
Bering or Thulean Bridge is above sea level, terrestrial
taxa could, at least, theoretically move between South
America and Africa without ever crossing a marine bar-
rier. Indeed, this explanation was proffered by those
favoring continental fixism for the South African and
South American biotic ties.
The hypothesis of cross-Antarctica migration suf-
fers a similar problem as well as many others. Antarc-
tica, for example, is not a small continent ?and the
distance from the northern tip of the Antarctic penin-
sula to Victoria Land, the Antarctic region that was
attached to Southern Australia, is ~ 5200 km, which
is a little more than the width of North America from
say, New York to San Francisco. This leads to two sig-
nificant problems:
1. The trans-Pacific disjuncts are latitudinally strati-
fied and include purely Neotropical taxa that link with
taxa from Northeast Australia, New Guinea, New
Caledonia and Fiji.
2. The trans-Pacific disjuncts are longitudinally
stratified, predominantly linking western South Amer-
ica with eastern Australia, a pattern not immediately
explainable by trans-Antarctic migration.
While trans-Antarctic migration would help solve
the problem of eliminating major marine gaps, the
range expansion and contraction required would still
be imposing. As shown in McCarthy (2003), many of
the South American sister taxa of Tasmanian taxa live
in south central Chile, restricted between 35 N and 42
N. The distance between 35 N and the southern tip of
Tierra del Fuego is more than 2000 km and the entire
? Biogeograf?a 2 ? July, 2007 ? Page 9 ?
width of Antarctica, as noted, is 5200 km, requiring
more than 7200 km of range expansion for taxa with
very narrow distributions. And this represents the
smallest range expansion required by Australian-South
American disjunctions.
For comparison, consider the biogeographical data
from a corresponding Northern Hemisphere analysis
(e.g., Sanmart?n et al. 2001). Unlike the Southern Hemi-
sphere disjunctions, none of the Northern Hemisphere
patterns is problematic or surprising. Instead, regions
that are closest together (like Western and Eastern
North America or Western Eurasia and Eastern Eurasia)
share the most closely related taxa, and regions that
are further apart share the fewest (like Western North
America and Western Eurasia). Note also that Eastern
North America and Eastern Eurasia, are considered
?disjunct? and share relatively few sister taxa despite
the fact that (1) the Bering Land Bridge has been avail-
able for crossing far more recently than Antarctica;
and (2) these regions are connected by less continen-
tal material than South Central Chile and Tasmania on
conventional paleomaps. The same argument may be
made for Western North America and Europe and their
connection via the Thulean bridge.
Given the multi-continent range expansion required
by the trans-Antarctic explanation, a problem arises
for taxa that have a particularly narrow range and/or
exclude Australia or New Zealand. New Zealand?s liz-
ard-like tuatara, for example, is the only extant sphe-
nodon, and its most recent fossil relatives occur, not
in Australia, but in the Late Cretaceous of Patagonia
(Apesteguia and Novas 2003). Another problem arises
for groups that boast multiple trans-pacific disjunc-
tions ?as with the freshwater fish Galaxias, the midge
Botryocladius, or the southern beeches Nothofagus.
This is problematic because the assumption of vicariant
-induced speciation assumes uninterrupted gene-flow
across three continents ?South America, Antarctica,
Australia? until the rifting of Australia from Antarc-
tica. Yet, multiple trans-Pacific sister taxa from the same
genus requires allopartic speciation confined narrowly
to one of the regions first. In other words, taxa origi-
nally divided by some unknown local barrier in, say,
Chile must then all find no barriers to either dispersal
or gene flow throughout Chile, all of Antarctica and
into Australia. But are there any other pairs of regions
so widely separated that exclusively share so many sis-
ter taxa ?or exclusively share multiple sister taxa within
the same group? In comparison, according to the San-
mart?n and Ronquist dataset, none of the groups
shared by Eastern North America and Eastern Asia,
regions as disjunct as Chile and Tasmania on paleo-
maps, contains more than one sister relationship.
These trans-Antarctic relationships become even
more bewildering when all of these problems occur
within the same genus ?as with the plant genus
Abrotanella. Abrotanella forsteroides from eastern
Tasmania is the sister taxon to a Chilean Abrotanella?
not to the Western Tasmania species. Thus, the most
recent common ancestor to this disjunct group must
either have migrated across Antarctica or must have
been transported across the full width of the Pacific
and Tasman Sea while either avoiding or disappearing
from New Zealand, and all the rest of Australia includ-
ing Western Tasmania. Within the same genus, we also
find A. muscosa, which is endemic to Stewart Island,
and A. fertilis from New Zealand, which are most
closely related, not to each other, but to other Chilean
species, again requiring extraordinary jaunts, whether
through Antarctica or across the Pacific, between Chile
and very specific Western Pacific locations, while avoid-
ing all the habitable regions significantly closer, such
as all intervening ocean islands or Australia to the west.
But the most persistent doubts must follow from the
effort to use Antarctica to explain tropical trans-Pa-
cific links. Sanmart?n and Ronquist define ?southern
South America? as ?the southern temperate region
(Argentina, Chile, Paraguay, Bolivia, and the Falklands
and Antarctic islands), and north-central South America
west of the Andes.? ?Northern South America,? in con-
trast, refers to ?north-central South America east of the
Andes (Brazil, Colombia and Venezuela).? Thus, ?south-
ern South America? really describes essentially all of
western South America and likely extends into the North-
ern Hemisphere perhaps as far north as ?northern South
America.? ?Northern South America? may have been
just as easily named ?eastern South America.?
On the other side of the Pacific, Australia is treated
as a single biogeographical unit, from the southern
tip of Tasmania to the northern most points of Austra-
lia. Thus, some of the taxa that link ?southern South
America? with ?Australia? in their analysis more spe-
cifically link taxa exclusively found in tropical South
America and/or exclusively found in tropical Australasia.
For example, part of their data included distributions
of Keroplatidae of Matile (1990), but a number of these
taxa have exclusively tropical trans-Pacific distributions
and Matile considered them as evidence that tropical
South America was once juxtaposed with north-east
Australia, New Caledonia, and parts of Indonesia (See
Matile 1990 or McCarthy 2003). Sanmart?n and Ron-
quist also reference some of P. Cranston?s analyses of
the Chironomidae. In 1999, Cranston noted that one
trans-Pacific disjunct Chironomidae genus, Nandeva,
found in Brazil and south central Chile, has not been
discovered south of 16 S in Australia, ?despite conti-
nent-wide exuvial sampling? (Cranston 1999, p. 298).
Cranston (1999) also noted that another trans-Pacific
genus, Fissimentum, occurs from Indonesia to South-
ern Australia but is restricted to the Neotropics in South
America. Cracraft (1999 2001) records many tropical
trans-Pacific disjunctions among birds, including the
? Page 10 ? Biogeograf?a 2 ? July, 2007 ?
most primitive relatives of chickens and pheasants ?
the megapodes from Australasia and the guans and
curassows from tropical America. Cracraft also notes
that the weakly flying sun bittern (Eurypyga helias),
from tropical South and Central America, is the clos-
est relative of the flightless kagu of New Caledonia
and two extinct flightless species (Apterornis) from
New Zealand. One of the more notorious trans-Pacific
disjunctions involves the astonishing fact that igua-
nas, a predominantly Neotropical taxon, live on Fiji-
Tonga, while no other iguanas occur anywhere in the
West or Central Pacific. Recent phylogenetic analyses
hypothesize that both Brachylophus (Fiji and Tonga)
and Dipsosaurus (California) occupy basal positions
among Iguanidae (Sites et al. 1996.)
From the equator in South America, west of the
Andes to the southern tip of Tierra del Fuego is more
than 6300 km. Across the full width of Antarctica, as
has been noted, is more than 5200 km. And from Tas-
mania to Queensland (20 S) is more than 2500 km. This
totals more than 14000 km of multi-continental range
expansion and then significant range contraction to
only the tropical regions in one or both areas. To give
a comparison, the entire width of Eurasia from the
Western coast of Portugal to the Eastern coast of Ja-
pan (at 40 N) is a little less than 10,600 km. Thus, while
Northern Peru and Northeast Australia are ?connected
through Antarctica? on conventional paleomaps, these
regions would have been separated by more continen-
tal material than are Portugal and Japan. Yet, Portugal
and Japan, of course, do not share an unusual number
of sister taxa found nowhere else in the world. And if
they did, it might be considered one of the world?s
greatest biogeographical mysteries.
Moreover, Sanmart?n and Ronquist note, there has
been remarkably little biotic exchange between North-
ern and Southern South America ?and the small num-
ber of taxa that they do share appear to have been
infiltrations from the South:
?...the biotas of northern South America and south-
ern South America do not appear to be closely related.
Southern South America appears more closely related
to the austral landmasses (Australia, New Zealand) than
to northern South America in the plant and all animal
area cladograms. The frequency of trans-American dis-
persal between northern and southern South America
is significantly lower than that between northern South
America and the Holarctic in the animal and insect
data sets? This result confirms previous biogeographic
results suggesting that the South American biota is
formed by a northern tropical component and a south-
ern temperate component, each with different bio-
geographic affinities (Crisci et al. 1991; Amorim and
Tozoni 1994; Lopretto and Morrone 1998)? (Sanmart?n
and Ronquist 2004, p. 239).
Sanmart?n and Ronquist suggest the tropical for-
ests of the Amazonian Basin likely provided ?a long-
lasting, powerful barrier to dispersal,? and note that
?[d]ispersal of South American animals was signifi-
cantly more frequent from the southern to the north-
ern region than in the other direction?? (Sanmart?n
and Ronquist 2004, p. 239). But if Neotropical taxa
have found it difficult to move into Southern South
America how reasonable is it to use Antarctica to ex-
plain tropical trans-Pacific disjunctions?
Still another problem, as noted, is that the trans-
Pacific taxa are longitudinally stratified. If Antarctica
served as the biotic corridor between South America
and Australia, one would expect to find all of Southern
South America, particularly south of 42 degrees S, link-
ing with all of Southern Australia ?and that neither the
eastern or western coastal regions of either South
America or Australia should be particularly favored. As
the conventional paleomaps in Figure 5 a and b show,
the eastern coast of South America is just as close to
the Antarctic Peninsula as the western coast of South
America. For example, the Antarctic Peninsula is much
closer to regions in eastern Brazil than to, say, Bolivia or
Peru. The same point stands for Western Australia. Taxa
that have migrated across the full extent of Antarctica
should also include links that exclude Eastern Austra-
lia and Western South America. But why are there few
if any taxa exclusively shared by Western Australia with
Eastern South America? Trans-Antarctic migration can-
not effectively explain why the disjunct taxa appear to
link very specific regions along Eastern Australia with
specific regions along Western South America.
The geological facts
Given all the difficulties generated by attempts to ex-
plain disjuncts assuming conventional reconstructions,
one would assume that the geological characteristics
of the seafloor between Australasia and South America
overwhelmingly support this view, and that the cur-
rent distributional explanations, like the cross-ocean
rafting of the Fijian iguanas, are hypotheses of last
resort. Yet the South Pacific seafloor data confirm that
essentially all of it is less than 83 my old ?and the vast
majority of it is less than 40 my old. This is consistent
with the fact that Australia/New Zealand and South
America share so many more closely related extant taxa
than Africa and Madagascar or Africa and South
America. Based on crustal age of the seafloor, we know,
as Sanmart?n and Ronquist note, that Madagascar
rifted from East Africa 121 MYA and that Northern
South America rifted from Western African regions
about 100 MYA. This same reasoning based on crustal
age data would imply South America and Australasia
remained in proximity until a much later date. The bio-
geographical links predict the difference in ages of
? Biogeograf?a 2 ? July, 2007 ? Page 11 ?
seafloor between the various Gondwanan regions, with
the strong trans-Pacific links supporting the juvenile
age of the South Pacific.
Conventional paleomaps that indicate a significant
separation of these trans-Pacific regions during the
Mesozoic assume the existence of enormous ocean-
wide tracts of hypothetical seafloor crust that have
since been subducted. But all evidence for precisely
how much of this seafloor actually existed has been
forever lost at trenches. As noted earlier, Briggs (2003)
and Metcalfe (1998) have now juxtaposed regions that
were previously thought to have been separated by the
Tethys. The reason their views are geologically viable is
because the ancient Tethys has now been completely
subducted. Regardless of what used to be convention-
ally drawn on paleomaps, no one really knows how
large the marine gap was between any of the circum-
Tethys regions. Briggs (2003) and Metcalfe (1998) have
both narrowed or closed these gaps due, either in whole
or in part, to problematic trans-Tethys disjunctions.
The same reasoning applies in the Pacific. Not only
has all pre-Pacific seafloor been subducted, the dis-
junct regions actually share many more closely related
taxa than do any other disjunct regions now known
to have been juxtaposed ?such as Africa-South Am-
erica, India-Madagascar, or the amphi-Tethys regions
brought together by Briggs (2003) and Metcalfe
(1998). Like Wegener, du Toit, Briggs, and Metcalfe,
those biogeographers who argue for the juxtaposition
of the trans-Pacific regions are merely contesting well
known drawings ?not actual geological facts.
The regions on opposite sides of the Pacific ?par-
ticularly Australia and South America, fit together
neatly. Reuniting the matching outlines closes the bi-
otic disjunctions and brings together many trans-Pa-
cific sister taxa. The trans-Pacific regions also share
geological links. For example, geological correlations
between north-east Australia and Mexico involving
Grenville-type Proterozoic basement material have led
some geologists to juxtapose these trans-Pacific regions
at that time. The resulting configuration is now known
known as ?AUSMEX.? We also find Proterozoic Gren-
villian material in Southern Peru (Martignole et al.
2005), showing a correlation between these regions
and eastern Australia. Also, not only did the Carib-
bean and Ontong Java Plateaus, just northeast of New
Guinea, form at the same time, with eruption peaks
at 120 ma and again at 90 ma, they also are remark-
ably similar in chemical and isotopic composition
(Arndt and Weis 2002).
Reconstructions that juxtapose Australia and New
Zealand with South America predict that we should
find matching trans-Pacific outlines, geological corre-
lations, and a great number of disjunct sister taxa. In
contrast, those working from conventional paleomaps
often express some measure of surprise at the trans-
Pacific relationships. For example, Anderson (2000, p.
202) noted that his molecular DNA analyses of loliginid
squid, which lack teleplanic larvae, ?suggest S(epio-
teuthis) lessoniana [Indonesia] and S(epioteuthis) se-
piodea (Caribbean) ?a species found on the other side
of the planet from S. lessoniana? are sister species.?
About the occurrence of iguanas on Fiji-Tonga, Pregill
and Steadman (2004, p. 18) wrote that ?long-distance
dispersal from the Americas, incredible as that seems,
is the only viable hypothesis that explains their pres-
ence in the southwest Pacific....? In a discussion of the
multiple trans-Pacific disjunctions among Abrotanella,
Swenson and Bremer (1997, p. 505) wrote that ?pos-
sible long distance dispersal events of Abrotanella are
hard to envision, however.? Phrases like ?incredible as
that seems,? ?hard to envision,? and ?on the other
side of the planet? help underscore that extraordinari-
ness of these distributions and seem to imply that the
researchers understand that their own biogeographi-
cal hypotheses may seem improbable. These improb-
abilities continue to multiply with every disjunction ?
and as noted the regions share hundreds of identical
or closely related taxa.
Those who place the Southern Pacific continental
regions in proximity in the Mesozoic are doing noth-
ing more than letting the biogeographical facts speak
for themselves. They are letting the data determine
their conclusions ?regardless of challenges to conven-
tion. The situation is reminiscent of a legendary story
in physics involving Paul Dirac and the development
of the Dirac Equation, an equation that requires the
existence of a particle very much like the electron but
that is opposite in sign. This consequence of the Dirac
equation was famously confirmed in 1932 by the dis-
covery of the positron, but Dirac, himself, had never
formally predicted its existence. When later asked why
he did not underscore this unconventional but natu-
ral consequence of his equation and boldly and for-
mally predict the existence of the positron, Dirac re-
sponded flatly, ?pure cowardice.?
At some point, biogeographers are going to have
to stop acting like Dirac and admit frankly and aloud
what their analyses so conspicuously demand.
References
Amorim DS, Tozoni SHS. 1994. Phylogenetic and biogeographic
analysis of the Anisopodoidea (Diptera, Bibionomorpha), with an
area cladogram for intercontinental relationships. Revista Brasileira
de Entomologia 38: 517-543.
Anderson FE. 2000. Phylogeny and historical biogeography of the
loliginid squids (Mollusca: Cephalopoda) based on mitochondrial
DNA sequence data. Molecular Phylogenetics and Evolution 15:
191-214.
Arndt N, Weis D. 2002. Oceanic plateaus as windows to the Earth?s
? Page 12 ? Biogeograf?a 2 ? July, 2007 ?
Interior: An ODP Success story. Joides Journal, Special Issue 28: 79-84.
Briggs JC. 2003. The biogeographic and tectonic history of India.
Journal of Biogeography 30: 381-388.
Brundin L. 1966. Transantarctic relationships and their significance,
as evidenced by the chironimid midges. Kungliga Svenska Vetens-
kapsakademiens Handlingar 11: 1-472.
Carey SW. 1976. The expanding Earth. Elsevier, Amsterdam.
Carey SW. 1988. Theories of the Earth and Universe: A history of
dogma in earth sciences. Stanford, Stanford University Press.
Cracraft J. 1999. Regional and global patterns of biodiversity loss and
conservation capacity: Predicting future trends and identifying needs.
The Living Planet in Crisis: Biodiversity Science and Policy. Cracraft J,
Grifo F eds. pp. 139-172. Columbia University Press, New York.
Cracraft J. 2001. Avian evolution, Gondwana biogeography and
the Cretaceous-Tertiary mass extinction event. Proceedings of the
Royal Society of London, Biological Series, 268, 459-469.
Cranston PS. 1999. Two unusual Chironomini (Diptera: Chirono-
midae) from Australian rainforest streams: One new genus and a
Neotropical genus new for the region. Australian Journal of Ento-
mology 38: 291-299.
Cranston PS. 2003. Biogeographic patterns in the evolution of
Diptera. The evolutionary biology of flies. Wiegmann BM, Yeates
DK eds. Columbia University Press, New York.
Crisci JV, Cigliano MM, Morrone JJ, Roig-Ju?ent S. 1991. His-
torical biogeography of southern South America. Systematic Zool-
ogy 40: 152-171.
Croizat L. 1964. Space, time, form: The biological synthesis. Pub-
lished by the author, Caracas.
Daum C, Savina M. 2000. Geology basics: Rocks and minerals,
plate tectonics, climate history, surface water and groundwater.
Agriculture and the American Midwest: Literature and the Envi-
ronment, 30 Nov. 2000.
Giribet G, Edgecombe GD. 2006. Conflict between data sets and
phylogeny of centipedes: An analysis based on seven genes and
morphology. Proceedings of the Royal Society, B, 273: 531-538.
Glasby CJ. 1999. The Namanereidinae (Polychaeta; Nereididae),
Part 2, cladistic biogeography. Records of the Australian Museum,
Supplement 25: 131-144.
Glasby CJ, ?lvarez B. 1999. Distribution patterns and biogeo-
graphic analysis of Austral Polychaeta (Annelida). Journal of Bio-
geography 26: 507-533.
Hallam A. 1983. Great geological controversies. Oxford University
Press, New York.
Heads M. 1999. Vicariance biogeography and terrane tectonics in
the South Pacific; analysis of the genus Abrotanella (Compositae).
Biological Journal of the Linnean Society 67: 391-432.
Hooker JD. 1853-55. The Botany of the Antarctic Voyage, Part 2:
Flora Novae-Zelandiae. 2 Vols. Lovell Reeve, London.
Lopretto EC, Morrone JJ. 1998. Anaspidacea, Bathynellacea
(Syncarida), generalised tracks, and the biogeographical relation-
ships of South America. Zoologica Scripta 27: 311-318.
MacArthur RH, Wilson EO. 1967. The theory of island biogeogra-
phy. Princeton University Press, Princeton.
Martignole J, Stevenson R, Martelat JE. 2005. A Grenvillian an-
orthosite-mangerite-charnockite-granite suite in the basement of
the Andes: The Ilo AMCG site (southern Peru), ISAG, Barcelona,
Extended Abstracts. pp. 481-484.
Matile L. 1990. Recherches sur la syst?matique et l??volution des
Keroplatidae (Diptera, Mycetophiloidea). M?moires du Museum
National d?Histoire Naturelle, s?rie A, 148: 1-682.
McCarthy D. 2003. The trans-Pacific zipper effect: Disjunct sister
taxa and matching geological outlines that link the Pacific mar-
gins. Journal of Biogeography 30: 1545-1561.
McCarthy D. 2005. Biogeographical and geological evidence for a
smaller, completely-enclosed Pacific Basin in the Late Cretaceous.
Journal of Biogeography 32: 2161-2177
Metcalfe I. 1998. Palaeozoic and Mesozoic geological evolution
of the SE Asian region: multidisciplinary constraints and implica-
tions for biogeography. Biogeography and Geological Evolution
of SE Asia. Hall R, Holloway JD eds. pp. 25-41. Backhuys Publish-
ers, Amsterdam.
Pregill GK, Steadman DW. 2004. South Pacific iguanas: Human
impacts and a new species. Journal of Herpetology 38: 15-21.
Sanmart?n I, Enghoff H, Ronquist F. 2001. Patterns of animal
dispersal, vicariance and diversification in the Holarctic. Biological
Journal of the Linnean Society 73: 345-390.
Sanmart?n I, Ronquist F. 2004. Southern Hemisphere biogeogra-
phy inferred by event-based models: plant versus animal patterns.
Systematic Biology 53: 216-243.
Sclater PL. 1864. The mammals of Madagascar. The Quarterly Jour-
nal of Science 1: 213-219.
Shields O. 1979. Evidence for initial opening of the Pacific Ocean
in the Jurassic. Palaeogeography, Palaeoclimatology, Palaeoecology
26: 181-220.
Shields O. 1998. Upper Triassic Pacific vicariance as a test of geo-
logical theories. Journal of Biogeography 25: 203-211.
Swenson U, Bremer K. 1997. Pacific biogeography of the
Asteraceae genus Abrotanella (Senecioneae, Blennospermatinae).
Systematic Botany 22: 493-508.
Wegener A. 1966. The origin of continents and oceans. Dover Publica-
tions, New York.
Dennis McCarthy
Buffalo Museum of Science, 1020 Humboldt Park-
way, Buffalo, NY 14211-1293, USA.
Malte C. Ebach
Botanic Garden and Botanical Museum Berlin-
Dahlem, Freie Universit?t Berlin, K?nigin-Luise-Str.
6-8, D-14191 Berlin, Germany.
Juan J. Morrone
Museo de Zoolog?a ?Alfonso L. Herrera?, Departa-
mento de Biolog?a Evolutiva, Facultad de Ciencias,
UNAM, 04510 Mexico, D.F., Mexico.
Lynne R. Parenti
Division of Fishes, National Museum of Natural His-
tory, Smithsonian Institution, PO Box 37012, MRC
159, Washington, D.C. 20013-7012, USA.