S. Ravichandran et al., ECOREGIONS FOR DESCRIBING WATER-QUALITY PATTERNS IN TAMIRAPARANI BASIN, SOUTH-INDIA, Journal of hydrology, 178(1-4), 1996, pp. 257-276
A regional approach which was developed for managing environmental res
ources, known as an 'ecoregional concept', was found useful for descri
bing spatial water-quality patterns. Ecoregions, when defined within t
he framework of a river basin concept, provided a satisfactory explana
tion of water-quality variations occurring in the Tamiraparani basin,
South India. The methodology consists of principal component analysis
(PCA) of 23 features of the geological, geomorphological, basin morpho
metry and land-use aspects of the Tamiraparani basin defined in terms
of 63 micro-basins. The PCA scores calculated on five components were
used to cluster the micro-basins into groups based on a similarity mea
sure. The groups identified in the analysis were traced on the drainag
e map to delineate nine ecoregions. A water-quality survey of the iden
tified ecoregions was carried out: The PH, EC, DO, TDS, major ions and
nutrients were estimated in 278 water samples. A PCA of the water-qua
lity data revealed that three processes appear to be particularly impo
rtant for water quality in this basin: the geological origin of ionic
richness variables, nutrient leaching from agricultural operations, an
d the carbonate system. The spatial ability of ecoregions to account f
or regional variations in water quality was compared using two existin
g classification methods (hydrological and limnological). Water sample
s were grouped in terms of ecoregional, hydrological and limnological
classifications, based on their location in the basin. visual examinat
ion of the box plots of water-quality variables showed that ecoregions
have less within-region variation with statistically significant diff
erences between the group means than either hydrological or limnologic
al classification. Discriminant analysis displayed the relatively bett
er ability of ecoregions to account for spatial water-quality variatio
ns than the other groups in the space defined by the first two discrim
inant functions representing the ionic richness and nutrient variabili
ty of the water samples.