title.jpg 4.39 K

Spatial Analysis of Intraspecific Diversity: A Point-centred Approach
object.jpg 1.39 K
Diversity indices can be useful in the identification of in situ conservation areas or the location of areas most worthy of collecting trips, and, in conjunction with additional GIS analyses, may show potential in the prediction of new areas of high diversity. On the other hand areas of low diversity might help us to better define environmental factors relating to endemic species.

method.jpg 1.41 K
The majority of diversity measures are generally aimed at species diversity and based on the principal of discretely countable units, the most famous ones being the Shannon-Weaver index defined as SW = t Pt ln (Pt), where Pt is the value characterizing the accession (e. g. cluster number).

When used for spatial analysis, these indices are generally calculated on the basis of a grid associating a value with each grid square. Instead of a grid, we use a circle centered on each accession point, compute the diversity index based on all the points within the circle and assign the result index to the center point. We refined the approach by giving more weight to a point closer to the center according to a distance weighted function. An external surface (e. g. a Digital Elevation Model) can also be used to modulate the weighting function, for to separate accessions

from two valleys divided by a high mountain range. We used data from field studies of Pouteria sapota in Central America for testing the model. Six quantitative descriptors from 227 individuals were used as trial data and clustered into 22 groups. A program in C has been written to implement this method within the the UNIX environment. A larger radius allows more data to be included in the calculations for a given point (and improved statistics), but then spatial specificity can be lost. The way diversity indices vary with the influence radius gives insights on the nature and origin of diversity. result.jpg 1.67 K
A larger radius allows more data to be included in the calculations for a given point (and improved statistics), but then spatial specificity can be lost. The way diversity indices vary with the influence radius gives insights on the nature and origin of diversity. On the figures we find areas of high diversity in Nicaragua and along the pacific coast of Central America, and low diversity areas along the Atlantic coast in Costa Rica. Pouteria sapota does not naturally spread easily, so its diversity is probably more caused by human factors such as migration. The Pacific coast of Central America has been the first colonized, while the Atlantic coast has lacked transportation infrastructure until the 1940s.



Distribution of Shannon-Weaver diversity of Pouteria Zapota
Above:
0.25 degree radius
Left:
1 degree radius

address.jpg 4.33 K
Contact : Grégoire Leclerc, William Bell, Andrew Nelson, GIS Unit,
International Centre for Tropical Agriculture (CIAT), A.A. 6713, Cali, Colombia.
Phone: +1 415 833 6625 or +57 2 445 0089 Fax: +1 415 833 6626.
E-mail: g.leclerc@cgnet.com | http://www.ciat.cgiar.org/
Mikkel Grum, IPGRI Regional Office for Sub-Saharan Africa c/o ICRAF, PO Box 30677, Nairobi, Kenya