Geography Reference
In-Depth Information
• Vicariance events have a null cost of 0. Speciation is assumed to be
by vicariance.
• Duplication events receive a null cost of 0. This implies sympatric spe-
ciation.
• Dispersal events receive a cost of 1 per area unit added to a distribu-
tion.
• Extinction events receive a cost of 1 per area unit deleted from a dis-
tribution.
Ronquist (1997b) developed a modification of this method, called con-
strained dispersal-vicariance analysis (DIVA), which distinguishes between
random dispersals (those that imply that the taxon passes through a barrier)
and predictable dispersals (those that occur when a barrier disappears). In
constrained DIVA, the cost matrix is constructed according to the following
rules:
• Vicariance events receive a benefit value of -1.
• Duplication events receive a null cost of 0.
• Extinction events receive a cost of 1.
• Random dispersal events receive a cost of 1.
• Predictable dispersal events receive a benefit value of -1.
Ronquist (1998) maintained that dispersal-vicariance models are over-
simplified. However, he predicted that the three-dimensional cost matrix
framework is powerful enough to allow optimization of any conceivable
biogeographic model regardless of its complexity.
Algorithm
The algorithm consists of the following steps (Biswas and Pawar
2006; Ronquist 1997a):
1. Obtain the taxonomic cladograms of the taxa distributed in the areas ana-
lyzed.
2. Replace the terminal taxa from the taxonomic cladograms with the areas
inhabited by them, to obtain taxon-area cladograms.
3. Construct a cost matrix assigning values to vicariance, duplication, dis-
persal, and extinction events.
4. Obtain the general area cladogram that optimizes these values.
