Monthly statistics of fish yield, water and soil samples were collected for two years. The data were analysed for computation of diversity indices, fish population dynamics and impact of physico chemical complexes on fish yield of the lake.

Diversity encompasses two different concepts of variety and variability; viz., richness and evenness. These two concepts, in theory, can be applied across a hierarchy of scales, from genetic diversity through to ecosystem diversity (Burton, et al., 1992). In the present study, diversity was measured by the number of species (species richness) and by using the two indices namely Shannon-Weaver [H'] and Simpson's [D] Indices. These are given by -

1. Shannon-Weaver index of diversity (Shannon, and Weaver, 1949; Wolda, H. 1983)

   H' = - pi 1n pi

   where pi is the proportion of the i^th species in the sample. The samples were collected from a large fish community, in which, the total number of species is known.

2. Simpson's index of diversity (Simpson, 1949) D = 1 - (p1² )

Physico-Chemical Analyses:
The physical and chemical parameters of the water and soil have been analysed respectively through water and soil sampling at fortnightly intervals. During 24 months field investigations, thirteen independent variables were estimated: water temperature (WT), turbidity (t), water pH (WpH), dissolved oxygen (DO), free carbon dioxide (FCO₂), total alkalinity (TA), water conductivity (WC), soil temperature (ST), soil pH (SpH), soil organic carbon (SOC), soil phosphorus (SP), soil potassium (SK) and aquatic macrophytic biomass (AMB). In order to look at the relationship between these variables and fish yield, regression/correlation analyses were carried out. Twenty five different equations (both linear and quadratic relationships) were looked at, and the best relationship is selected based on strong correlation coefficient and least percentage of standard error of Y estimate (dependent variable). Finally, the step-wise forward selection method was used in a linear multiple regression of fish yield (FY) against the independent variables. The stepwise procedure made no allowance for the possibility that a variable entered as significant might become insignificant on the addition of further variables into equation, and a variable entered into regression equation and subsequently removed (after addition of several other variables) would indicate that the variable is significant in reducing the variation. The final regression thus indicates only significant (p <= 0.05) variables.