Malakar et al.

The forward selection procedure for the CCA resulted in the retention of eight

variablesassignificantcontributorstovariationintheordination.Thefirstordination

axis accounted for 44.83% of the variance of the species data, whereas the second

axis accounted for 18.62% of this variance (Table 3). Species abundances were

significantly correlated with the environmental factors (P=0.001 along axes one

and two, Monte Carlo test with 1,000 permutations). Flow, temperature and TDS

were positively correlated with first ordination axis, whereas depth and turbidity

were negatively correlated (Figure 1). Conductivity was positively correlated with

second ordination axis.The species environmentcorrelationsof each axis were 0.90

(axis 1) and 0.64 (axis 2). Each significant environmental factor increased along a

vector away from the origin with its length being a measure of the magnitude. O.

pabda, O. pabo and O. bimaculatus were separated significantly from each other

in habitat space with respect to the significant habitat characteristics. O. pabda was

associated with the sites with fast water with shallow depth, high dissolve solids

with higher water temperature conditions. O. bimaculatus was associated with the

sites containing deep waters with slow water current and turbid water whereas,

O. pabo was associated themselves with the sites containing higher conductivity.

Figure 1. Canonical correspondence analysis showing correlation between species

composition and environmental variables.

ID: Opbo = Ompok pabo ; Opbd = Ompok pabda ; Obm = Ompok bimaculatus .

Localfishassemblagesarestructuredbyenvironmentalinfluencesatmultiplespatial

scales (Schlosser, 1982, 1987; Marsh-Matthews and Matthews, 2000). Species in

the present study displayed considerable differences in habitat associations. The

most important environmental variables for the assemblage of Ompok species in

140

International Journal of Bioinformatics and Biological Sciences : v.2 n.3&4. Dec. 2014