Discussion

The data analysis of 106 samples collected from urban water bodies highlights the role of environmental drivers in structuring diatom species assemblages, conforming to the earlier studies (Taylor et al. 2007a). Elevated ionic level is proportional to the dissolved major elements (cations and anions like calcium, chlorides, sodium and magnesium) that have influenced the aquatic biota, sedimentation process and oxygen levels which adversely affected the availability of microhabitats (Leira & Cantonati 2008). Ions accumulate with the sustained inflow of sewage and industrial effluents (human activities) or originate from mineral sources such as erosion of rocks and soils (natural activities). The increased sedimentation rates imbalances the biological and chemical oxygen demand in water and thus becomes responsible for species shift. These are the characteristics of polluted water bodies which resemble the wetlands of the current study region (Table 1). In the current study, wetlands in urban/ populated pockets showed higher nutrient and ionic load, which is evident in DCCA analysis. Natural organic detritus/ organic waste from wastewater, agriculture and urban runoff attributed to higher BOD and COD. It also explained diminishing DO levels that are lethal for most fishes, many aquatic insects and amphibians (Kar et al. 2006). High chloride was due to point source (sewage inputs), road deicing salts and weathering (Müller & Gächter 2012) and are directly related to urbanization and increasing human density (Brönmark & Hansson 2005).

Diatom composition and distributional pattern across wetlands were comparable to environmental variables. Species composition comprised of cosmopolitan and endemic species of the genus Gomphonema, Nitzschia and Surirella (Fig. 2). Among 43 wetlands, 10 wetlands recorded the presence of endemic diatoms with relative abundance ranging from 5 to 65% across habitats. Most dominant endemic diatoms include G. dharwarensis Gandhi, G. spiculoides Gandhi, Surirella sp., N. williamsii Alakananda et al. and N. taylorii Alakananda et al. A recent study (Karthick & Kociolek 2012, Karthick et al. 2012) reported high abundance of many endemic diatoms such as Gomphonema gandhii Karthick et al., Gomphonema diformum, Gomphonema tamilensis Karthick & Kociolek from this region. Further, these studies emphasized the need for understanding the ecological profile of endemic species and its role in biomonitoring programs in the Indian subcontinent. For example, G. dharwarensis grows abundantly in low nutrient and highly aerated waters and hence forms a potential indicator species of oligotrophic condition. Further, many unidentified species of Achnanthidium and Gomphonema which occurred in this study are in the process of taxonomic description along with the study of their ecological preference which will be helpful for future biomonitoring work in this part of the continent. Most of the new diatoms occurring in low to moderately polluted waters from this region have higher ecological importance, and also play a vital role in classifying wetlands either based on already available indices or in formulating a new index for peninsular India. This ensures the importance of endemic diatoms in tropical urban waters in bioassessment programs and eventually helps in the conservation of these highly threatened fragile wetlands of peninsular India.

Species such as Rophalodia gibba, Surirella sp.and Cymbella sp. are substrate specific and grow abundantly on sediment (epipelic) and plant (epiphytic) substrates. Among all others, 90% of species showed no substrate specificity because of less competition for growth of pollution tolerant species in high electrolyte waters (Liancourt et al. 2005). This could be attributed to the abundant availability of submerged macrophytes in most of the wetlands unlike in multi-substrate environments (rivers and streams). Recent studies in India recorded diatoms from epiphytic and epilithic habitats of rivers and streams, explaining substrate specific diatom assemblages (Humane et al. 2010, Karthick et al. 2011). Based on the availability pattern, recent standard protocol suggested the use of epilithic diatoms for monitoring of lotic ecosystems such as rivers and streams (Kelly et al. 1998, Karthick et al. 2010) and epiphytic diatoms for monitoring lentic ecosystems such as wetlands (Stenger-Kovács et al. 2007, Cejudo-Figueiras et al. 2010). Although epipelic diatoms are used in most studies, it is implemented in exploring paleo-limnological characters i.e., past water quality of lakes (Wang et al. 2012). Collection of all available substrates is recommended for better understanding of species-environment relationships (Winter and Duthie 2000, Alakananda et al. 2011). Again, there is a lack in understanding of complex autecology (physical substrate and chemical habitat). The relationship between increasing pollution and substrate type would substantiate the need for further investigation of species autecology in urban region.

DCCA was performed to analyze the relationship of diatom community to water chemistry and results showed a gradient of conductivity and organic load from rural wetlands to urban wetlands and to sewage fed wetlands (Fig.3). Subsequently, a gradual shift from sensitive species (left side of the plot) to pollution tolerant species (right side of the plot) highlights dissimilar diatom composition in rural, suburban and urban wetlands. The increasing populations in urban pockets showed a proportional increase in degradation of surrounding wetlands due to a high amount of untreated domestic sewage ending up in wetlands. The rapid unplanned urbanisation (Ramachandra et al. 2012) and its impact on water bodies was evident from DCCA analyses through diatom composition. However, Sankey, Venkateshpura and Ullalu wetlands are relatively oligotrophic despite being located in urban pockets due to the recent endeavor by the government in restoring these lakes and letting only treated water into these water bodies. DCCA tri-plot demonstrates that organic matter and phosphates are determining factors for assessing the absence of sensitive species and prolific growth of tolerant species in polluted waters and hence they have been aptly used as surrogate indicators of urban development in the earlier studies (Walker & Pan 2006, Walsh & Wepener 2009). TWINSPAN (Fig. 4), shows oligotrophic status with the presence of Achnanthidium sp., Gomphonema sp. and G. gracile (group A-C); mesotrophic status with C. atomus and Fragilaria ulna (group D & E); andmeso-eutrophic status by N. palea and C. meneghiniana (group F & G). TWINSPAN in Walker and Pan (2006) demonstrated that the diatom composition consistently differ between urban and rural sites, particularly in terms of conductivity and organic pollution. DCCA in other studies from Europe, North America and South America explained the correlation of ions such as potassium, chlorides and calcium with changes in species assemblage (De Fabricus et al. 2003, Jüttner et al. 2010), due to the intense anthropogenic activities. Factors such as canopy cover, water velocity, light penetration and organic matter affect sensitive diatoms in rivers but in slow flowing/ stagnant environments like wetlands, ion concentrations influence diatom assemblages, mainly of tolerant species. Abundant population of N. palea is observed in high nitrates and silt conditions, similar to the earlier scores (Van Dam et al. 1994) in European waters. Along with N. palea, other species of Nitzschia inhabit a majority of wetlands and ecological profiling of these may be very helpful in future for wetland monitoring. These pollution tolerant groups of species also reflect an intense land use change in the catchment such as the conversion of an area under vegetation into an urban pocket (Bere & Tundusi 2011). Wetlands of ecological importance are ranked from most polluted to least polluted based on marked factors (phosphorus, BOD and COD) in DCCA.