Abstract

Ants’ form one of the very dominant arthropods with more than 10,000 described species. Their presence, more often their dominance in extremes of climates and a variety of habitats is overwhelming. Ants reveal the status of the ecosystems and are considered as good biological indicators. This necessitates spatial distribution analyses along with mapping of its habitats. This helps to understand ant geography and also to determine the kind of the stress levied on the ecosystem. This research is being carried out in the Sharavathi river basin (13°43′24.96 N to 14°11′57.48 N latitude and 74°40′58.44 E to 75°18′34.92 E longitude) in the central Western Ghats of Karnataka state, India.

Multi spectral data with 23.5 m spatial resolution acquired from Indian Remote Sensing (IRS) Satellite was used to determine the land cover and land use in the study area. The application of vegetation indices to delineate vegetative and non-vegetative areas determined 70% of the study area under vegetation cover. False color composite was generated with IRS (Indian remote Sensing) data (0.52 – 0.59μ m, 0.62 – 0.68μm and 0.77 – 0.86μm), which helped in the selection of training polygons. A numerical description of the spectral attributes of each land use type was developed (supervised classification) using Maximum Likelihood Classifier. This also helped in identifying 14 landscape elements (LSE’s). To analyze the spatial distribution of ants, a converging sampling strategy was adopted in four radial directions. Along each direction at every four km, three samples were laid at a 200 m distance, resulting in a total of seventy-eight 30 x 30 m quadrants, each of those were marked using a GPS (Global Positioning System). Sampling was carried out in representative LSE’s (seven LSE’s) in varying replicates to determine the ant fauna by using pitfall traps, leaf litter collections, bait traps, and visual collections. The ants were then identified and the pooled data of various sampling techniques (pitfall traps, leaf litter, etc.) were quantified to compute Land Scape Element wise species richness and composition. Data analysis results reveal that ant species richness increased where a mosaic of habitats (more diverse habitats) were present. Moist deciduous forests are the most ant rich habitats and evergreen forests have the highest ant species per plot richness. Species such as Anoplolepis longipes are present in habitats (except evergreen and semi-evergreen forests) that are closer to human settlements indicating human interference with the ecosystems, while, species such as Harpegnathos saltator is present only in undisturbed systems. Polyrhachis mayri was found only in highly undisturbed semi evergreen forests. Forest patches with small breaks in canopy covers provide the specific niches required for Pachycondyla rufipes. Arboreal ants as Oceophylla smaragdina and Polyrhachis species are present in heterogeneous forest patches but are totally absent in monocultures (like plantations, etc.). However, the niches in Acacia plantations (72% of the sampled Acacia plantations) harbour the specialist predator Diacamma rugosm. Scrub jungles are deprived of all species of Leptogenys. This reveals the intra and inter linkages of landscape elements with species distribution, which is essential for conservation of endemic, rare and endangered species of flora and fauna. This endeavor demonstrates the application of the spatial analyses tools such as GIS, GPS and remote sensing data in habitat mapping and spatial distribution analyses of biodiversity. These exercises help in evolving the appropriate conservation and restoration strategies for the sustainable management of ecosystem.