Factor analysis of tree distribution patterns of six forest reserves in Ashanti Region.
Loading...
Date
July, 2010
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This study demonstrates aspects of three recognized principles, which are important in phytosociological interpretations using multivariate analysis techniques. First, when a restricted geographic area is analyzed (Ashanti region, in this case), species with different total ranges may be correlated. Secondly, when the utilized data are on a relatively course scale, microhabitat preferences are not reflected in the plant association found. Third, species may have different ecological characteristics in different portion of their range. In view of this, factor analysis was used to elucidate general distribution patterns for hundred tree species and some environmental variables for six of the selected forest reserves in the ecological zone of Ashanti Region. The three factors combined various tree species and environmental variables into groups and reflect general patterns that were meaningful in relation to the ecological characteristics of the various plants included. The general factors group together species from six forest reserves in Ashanti Region namely Onyimsu forest reserve, Bobiri forest reserve, Mirasa Hills, Prakaw forest reserve, Pra Anum forest reserve and Onuem Bepo forest reserve. Each of the three factors indicates a group of species and/or environmental variables which is superficially explained first by distribution and then by environmental variability, past history, competitive ability and habitat preference. There are a number of tree species that do not fit into these general patterns, but instead show independent patterns of distribution. The species positively associated with factor I are all characteristics of Onyimsu and Bobiri forest. Factor I loadings were influenced by considering distributions only for a relatively small area in contrast to conducting the analysis on all the reserves in Ghana. Factor III is the most clearly defined of all the factors. The environmental variables with the highest correlations are useful descriptors in differentiating the Onuem Bepo forest from the others. There are other variables with higher loadings than rainfall variables in factor II and factor III and also a temperature variable shows a higher correlation than rainfall variable.
This does not mean that precipitation (rainfall) is not important, but suggests that it alone may have less effect than other variables. Moisture directly available to tree species is of course, governed by other environmental variables such as topographic relief and the water retaining characteristics of soils in a particular area.
Description
A thesis submitted to the Department of Mathematics, Kwame Nkrumah University of Science and Technology, In partial fulfilment of the requirements for the degree of Master of Philosophy in Mathematics.