Agriculture Reference
In-Depth Information
islands of the Mid-Atlantic Ridge. There is ample evidence that other fauna have colonized the
Lesser Antilles from South America and from the Greater Antilles and other islands to the north
(Savage 1982; Roughgarden 1990; Humphries and Parenti 1999), so it would not be particularly
surprising to find another group of organisms doing the same. Another region with many islands,
a mix of geological histories, and partly known or unknown earthworm faunas is Southeast Asia.
There, the possibility of dispersal of earthworms from the Australian region or East Asia into the
extensive archipelagoes is found.
The general pattern of distribution of earthworms in the Lesser Antilles has been presented
elsewhere (Fragoso et al. 1995). No earthworms other than peregrine species were found in the
southern Lesser Antilles or the small volcanic islands north of Guadeloupe. Many new species
endemic to their islands were found on Guadeloupe, Dominica, Martinique, and St. Lucia (James,
unpublished data). Based on the number of genera and morphologically homogeneous groups of
species within genera, at least nine successful dispersal events would be required to establish the
current earthworm fauna of those four islands. If so, this is a very dense cluster of dispersal events
in the middle of the archipelago, farthest from sources of colonization. However, earthworms have
failed to colonize the Lesser Antilles by over-water dispersal from nearby land masses populated
with indigenous earthworms. There is no evidence that the South American earthworm fauna has
spread northward into the Lesser Antilles, or that the elements of the earthworm fauna of the Greater
Antilles, particularly Puerto Rico and the Virgin Islands, have dispersed to the east and south.
The GuadeloupeÏSt. Lucia axis poses a challenge to the conclusions that otherwise could be
reached easily from the data from other islands. Without recourse to an analytical approach, the
challenge cannot be met because hypotheses of dispersal are logically unfalsifiable. Any distribution
could arise by postfragmentation dispersal if sufficient complexity of dispersal history is allowed.
In this case, progress will come only from testing the hypothesis that the distributions are the result
of vicariance because it is logically conceivable that such a hypothesis could be rejected.
If, for the time being, it can be accepted that over-water dispersal can be ignored as a factor
in earthworm distribution to islands, or between any two land areas separated by saltwater, then
earthworms provide nearly ideal indicators of past land-area connections. Their evolutionary history
should mirror these past connections because a severing of the land connection isolates populations
of species. From the perspective of vicariance biogeography, these isolating or vicariating events
are the primary factors of interest in the history of the flora and fauna of land areas. Dispersal is
seen as a source of noise in the data, analogous to homoplasious character evolution (Sober 1988).
Vicariance biogeography represents a system of formulating hypotheses that predict organismal
distribution patterns based on underlying historical models (Wiley 1988). The underlying historical
models have to do with creation of barriers to genetic exchange. Any species, earthworms included,
could be affected by the fragmentation of a range because of climatic, geologic, or other processes.
Thus, it should be possible to make a very good map relating the branch points on an earthworm
cladogram and the separations of the land masses on which earthworm taxa occur. In the jargon
of vicariance biogeography, one replaces the terminal taxa on the phylogenetic tree with the names
of collection locations to make an area cladogram. A geological model provides a predicted area
cladogram, which can then be compared with a biologically derived one. The success of this
approach depends heavily on a very good phylogenetic analysis, on having the data set that is as
free as possible from dispersal-induced noise, and on a proper choice of area units.
Hausdorf (2002) argued for the use of biotic elements, rather than the more traditional areas
of endemism, as the units of biogeographical analysis. An area of endemism is a region characterized
by a number of taxa unique to the area. Hausdorf defined a biotic element as Ña group of taxa
whose ranges are significantly more similar to each other than to those of taxa of other such groups.Ò
The difficulty in defining areas of endemism arises from dispersal: ÑThe delimitation of areas of
endemism is not problematic when species originate by vicariance and there is no dispersalÒ
(Hausdorf 2002). The biotic element concept is a general rule, encompassing a range of natural
possibilities, including sharply delimited areas of endemism inhabited by nearly dispersal-free taxa.
