A total of 140 diatom taxa were identified across sites, 61 of them reaching a relative abundance of over 5% in at least one site. Annexure 1 provides the checklist of diatoms. The species compositions were dominated by Gomphonema gandhii Karthick and Kociolek, Achnanthidium minutissimum Kützing, Achnanthidium sp., Gomphonema sp., Gomphonema parvulum Kützing, Nitzschia palea (Kützing) W.Smith, Nitzschia frustulum (Kützing) Grunow var. frustulum, Navicula sp., Navicula cryptocephala Kützing, Cyclostephanos sp., Cymbella sp., Eolimna subminuscula (Manguin) Moser Lange-Bertalot and Metzeltin, Sellaphora pupula (Kützing) Mereschkowksy, Eunotia minor (Kützing) Grunow in Van Heurck, Nitzschia amphibian Grunow f. amphibia, Cyclotella meneghiniana Kützing, Gomphonema difformum Karthick and Kociolek, Navicula rostellataKützing, Cocconeis placentula Ehrenberg var. euglypta (Ehr.) Grunow, Brachysira sp., Stauroneis sp., Encyonema minutum (Hilse in Rabh.) D.G. Mann, Cyclotella sp. and Nitzschia sp. The species composition contains cosmopolitan to possible Western Ghats endemic species and in general species from oligotrophy to highly eutrophic condition were also observed. The current study also documents some of the species for the first time in Western Ghats and many new species descriptions are underway. Waters were circumneutral throughout the whole study area (Table 4), with certain tendency towards alkalinity in the streams drained from agriculture and urban catchment. The highest ionic and nutrient values correspond to the agriculture catchment dominated streams, particularly in the leeward side of the mountains. Oxygenation was generally close to saturation; the lowest values are due to wastewater water inflows in few localities. The most oligotrophic sites were located in mountain watercourses, while downstream sites were generally more polluted, becoming eutrophic in condition. The detailed water chemistry variables are presented in Table 5.

The results of correlation performed between diatom indices and water chemistry variables are presented in the table 6. It is observed that significant correlations exist to varying degrees between most of the diatom indices and water chemistry variables. Diatom indices IPS, EPI and SID showed correlation with more number of water chemistry variables when compared to the other indices. TDI and IPS are negatively correlated with pH, EC, TDS, alkalinity, calcium, magnesium, sodium and potassium.  Percent pollution tolerant diatoms were positively correlated with most of the ionic variables. None of the indices were correlated with water temperature. The first four axis of CCA explains 70.1% variance of species-environment relation and the ordination plot reveals two distinct clusters of species.

The species composition contains cosmopolitan to possible Western Ghats endemic species and in general species from oligotrophy to highly eutrophic condition were also observed. Among the species observed in this study, two species were possibly endemic to Western Ghats (G. gandhii, G. difformum and few other species yet to identify). In few sites these species were very dominant reaching more than 80% of the total assemblages. The remaining dominant taxa were cosmopolitan and well documented in international literatures (Krammer and Lange Bertalot, 1986-1991). It is important to note that the indices that were developed and tested in European rivers, lacks Western Ghats endemic taxa. Most sites were oligo-mesotrophic and only a few of the streams were eutrophic. The differences in the water quality of these rivers were reflected in the values for the diatom indices, by the relative abundances of indicators of trophic/saprobic stage and by different types of diatom community. Classification of stream sites according to the IPS score is given in Table 6.

Table 4: Summary of the Canonical correspondence analysis for the stream sites from Central Western Ghats.

Table 5: Summary table of water chemistry variables in 45 sites of CWG streams.

Table 6:  Pearson correlation coefficients between measured water chemistry variables and diatom index scores in 45 sites of CWG streams

Note: WT-Water Temperature, EC-Electric Conductivity, TDS-Total Dissolved Solids, ALK-Alkalinity, Cl-Chlorides, Ha-Total Hardness, Ca-Calcium hardness, Mg-Magnesium hardness, Na-Sodium, K-Potassium. Diatom Indices: SLA-Sládeček’s index, DESCY-Descy’s pollution metric, SHE-Steinberg and Schiefele trophic metric, WAT-Watanabe index, TDI-Tropical diatom Index, GENRE-Generic Diatom Index, CEE-Comission for economical community Index, IPS-Specific Pollution Sensitivity Metric, IBD-Biological diatom index, IDAP-IndiceDiatomique Artois Picardie, EPI-D-Eutrophication/pollution Index, IDP-Pampean Diatom Index,%PT-Percentage Tolerant).

In general, the diatom indices show significant correlations to water quality. The correlations obtained in the present study are comparable to those demonstrated by Taylor et al., (2007c) in South Africa and by Kwandrans et al., (1998), Prygiel and Coste (1993) and Prygiel et al., (1999) in Europe. Significant correlations emphasize that diatom indices can be used to reflect changes in general water quality (Table 6). No correlation of temperature with any of the indices observed that may be due to differing temperature regime in tropical when compared to temperate streams. Similar observation has been recorded by Taylor et al., (2007) from South African rivers. Canonical correspondence analysis (Figure 22) demonstrates that certain widely distributed taxa have similar ecological characteristics in widely separated geographic areas. Species commonly associated with poor water quality in Europe e.g.,Eolimna subminuscula Lange-Bertalot, Nitzschia palea (Kützing) W. Smith, Sellaphora pupula (Kützing) Mereschkowksy, Gomphonema parvulum (Kützing) ordinate on the right hand side of the CCA together with elevated levels of ionic and nutrients. Taxa typical of cleaner, less impacted waters ordinate out on the left hand side of the diagram e.g., Gomphonema difformum Karthick and Kociolek. However Gomphonema gandhii Karthick and Kociolek seems to have a wider ecological tolerance when compared to its morphologically related species. Achananthidium minutissimum group from Western Ghats streams contains morphologically three distinct taxa with wide ecological preferences. Despite the reevaluation of this genus multiple times (Lange-Bertalot and Krammer, 1989, Krammer and Lange-Bertalot, 1991, Potapova and Hamilton, 2007) there are still major gaps in taxonomy and ecology apart from non-inclusion of specimens from tropical rivers. The similar problem holds good for some of the other genus like Gomphonema. This analysis helps to demonstrate that the widely distributed species encountered in the streams of Western Ghats which are not identical only with morphology also have similar environmental tolerances G. gandhii and G. difformum are one among the dominant taxa in this data set but not included in any of the index calculations and their omission could result in an under or overestimation of the index scores. Taylor et al., (2007d) cautioned about the associated problems with usage of European indices in South African rivers. These suggest that European diatom indices can be used in India provided indices address the issues concerned with ecology of endemic species. Hence, the list of taxa included in the indices needs to be adapted according to the study region with giving more importance to the local endemic flora which encourages taxonomic and ecological studies in tropics. The structure of benthic diatom communities and the use of diatom indices yield good result in water quality monitoring in India. It is also evident from the current study that many widely distributed diatom species have similar environmental tolerances to those recorded for these species in Europe and elsewhere. However, the occurrence of possible endemic species will necessitate the work towards diatom index unique to India. In the mean time considerable amount of importance should be paid for taxonomic and ecological understanding of endemic flora for development of improved biomonitoring program.
LULC Analysis: LULC showed considerable variability among catchments, with forest/vegetation land cover as a dominant class (mean = 64.36%, range = 0.13–95.45%), followed by agriculture/cultivation area (mean = 24.27% range = 2.55-63.63%), among the 24 catchments. The mean percentage area covered by open scrub/barren land, water bodies, built up and other classes were 6.65±9.16 (0.16-31.48), 1.68±1.76 (0.01-5.61), 0.47±0.4 (0.06-1.52), and 2.58±2.51 (0.07-9.58) respectively. LULC analysis shows that natural vegetation is poor towards the leeward side of the mountains (eastern region), due to intense anthropogenic activities. This region has more of agriculture, open scrub/barren land, and built-up area. In the entire study region the class forest/vegetation covers predominantly moist deciduous type, with small isolated patches of semi-evergreen vegetation in the eastern region and the western region (windward side) with rugged hilly terrain and heavier rainfall (~5000 mm) having characteristic evergreen to semi-evergreen forests. The detailed LULC for each catchment is given in the Table 7 and the land cover images are given in Figure 23. The dendrogram of sites based on LULC obtained by Ward's method is shown in Figure 18a. Three well differentiated clusters can be seen, with forest cover decreasing and agriculture/cultivable land cover increasing from top to bottom. The third cluster from top includes sites SAN, KAL, MAN, SAP and GUN, which are characterized with high amount of agricultural activities (>50%). The group in center is characterized by more forest cover (>50%) with moderate amount of agricultural land. Another top most group is dominated by forest land cover of more than 80%.

Water Chemistry: Water chemistry was characterized by high spatial variability in nutrients and ionic strength among the 24 sites. Nutrients like nitrates showed a many fold variability from 0.07 to 4.24 mgL-1. Sites in the mountain range and wind ward side were characterized by low nutrient and ionic variables; where as the sites on the lee ward sides of the mountains were characterized with high nutrients, ions and low dissolved oxygen levels. The dendrogram of sites based on the water chemistry is shown in Figure 25. Two well differentiated clusters can be seen, with sites containing moderate to low levels of nutrient and ions in the top cluster and two severely polluted sites at bottom. The top cluster further divide in to four sub cluster based on the ionic and nutrient values. Clustering of most of the sites in cluster based on water chemistry variables are similar to the cluster of sites based on the land cover variables. A PCA bi-plot of water quality variables and LULC for all sample sites is given in Figure 26. The two-dimensional bi-plot describes 65% of the variation in data, where 52% displayed on the first axis and 13% is displayed on the second axis. Among water chemistry variables, ionic variables were positively correlated with first axis and among the LULC variables percentage agriculture and scrub land cover were positively related to the first axis. Sites with more than 50% of agriculture land cover were separated from other sites on the PC2 axis indicating trends in water quality may be related to land use. Agriculture dominated sites were placed due to the higher conductivity, ionic and nitrates levels relative to the forest dominated sites, which are characterized by low ionic and nutrient in nature.

Diatom Assemblages: Among the 113 taxa the most common and dominant diatom taxa are Eolimna subminuscula, Achnanthidium sp., Navicula sp., Nitzschia palea, Gomphonema parvulum, Gomphonema sp., Gomphonema gandhii, Achnanthidium minutissima and Cyclostephanos sp. The list of species with their abundance is presented in the Appendix 1. Species richness varied from 4 to 29 with an average of 15. Shannon-Wiener diversity varied from 0.71 to 2.94 with an average of 1.76. According to the pH classification, diatom assemblages were characterized by a high proportion of neutrophilous diatom species (64.62%) followed by alcaliphilous species (26.64%). Salinity classification based on the diatom species assemblages infer the fresh to brackish water species were the dominant form with 86.16% followed by brackish to freshwater (7.84%) and exclusively freshwater (5.3%) flora.

Nitrogen autotrophic taxa, which tolerate elevated concentrations of organically bound nitrogen, were dominant with 53.31%.  Species which require 100% oxygen saturation were prevailing community with 42.98% followed by low level (30% oxygen saturation) oxygen requirement species by 29.08%. The composition of diatom community with respect to saprobity in the order or oligosaprobous, β-mesosaprobous, α-mesosaprobous, α-meso-/polysaprobous and polysaprobous were 7.8%, 46.09%, 10.58%, 26.56% and 8.97% respectively. The species occurs in the eutraphentic and oligo to eutraphentic were equally dominant with respect to the trophic state explained by diatoms.