Abstract
Thirty stands on the slopes of calcareous hills around Karachi and its vicinity were sampled by point-centered quarter method. Sampling was restricted to trees, shrubs and perennial herbs. Soil samples were collected from each stand and were analyzed physically and chemically. The vegetational and environmental (soil) data sets were subjected to various multivariate data exploratory techniques including principal component analysis (PCA), correspondence analysis (CA), detrended correspondence analysis (DCA), and canonical correspondence analysis (CCA). The results of the different techniques were compared. Each technique provided useful, often unique, information pertaining to the ecosystem. In general, information on compositional variation was helpful" in the interpretation of the vegetational dynamics of the hills. Canonical correspondence analysis was able to expose the underlying environmental gradients associated with the vegetational variation. The merits and demerits of the techniques are discussed.
