Habitat modeling of non-migratory waterfowl in Paya Indah Wetland, Malaysia

Waterfowl are the most conspicuous components of freshwater wetlands ecosystem and their presence or absence may indicate the ecological conditions of the wetlands. Paya Indah Wetlands (PIW) reserve is one of Malaysia‟s premier ecotourism parks, covering approximately 3100 hectares in the State of Selangor. Lesser Whistling Duck (LWD) (Dendrocygna javanica) and cotton pygmy goose (CPG) (Nettapus coromandelianus) are the two only non-migratory waterfowl species from among the approximately 122 bird species which exist in the area. Both species have small populations in the area and are supposed to be recluse. Understanding species-habitat relationships and processes and mechanisms of which habitat balance species density are central themes of our study.We demonstrate how modern technology and field survey can be used to develop ecological baseline data including landuse/landcover, water depth, water quality, waterlevel fluctuation, and normalized difference vegetation index (NDVI). Landuse/landcover classifications were applied by visual interpretation. NDVI was extracted based on red and near infra-red 2 bands. The Topo to Raster method was used for interpolation of water depths. Water-quality index and water-level fluctuation of lakes were interpolated across lakes using the inverse-distance weighted method. Qualitative and quantitative accuracy assessment of our classification was promising and clearly illustrated that World View-2 imagery can yield fast and reasonably precise identification of ecosystem characteristics for ecological baseline data. On other part of the study we monitored both waterfowl species in PIW by Point-Count methods. We marked 48 observation points in our study site. We determined ducks location based on the standpoint location, distance to ducks and initial bearing. We demonstrate how a useavailability design can be applied to fit habitat suitability for LWD and CPG by MaxEnt. We fit 10 a priori models based on biological relevant combinations of explanatory variables and ranked them based on AICc for both species. The results of the top models show that water depth and water-level fluctuation play a vital role in determining the quality and quantity of habitats for lesser whistling duck and cotton pygmy goose. The influence of NDVI and proportion of spike rush and water lily indicated that the quantity and quality of those particular vegetation types increased suitability of a particular habitat. On the last part of the study, we worked on hypothesis that distribution and density-based models should reveal environmental conditions which functionally relate to distribution and density. We applied Resource Selection Functions (RSF) to understand habitat selection of LWD and CPG by employing binomial generalized linear models for presence/absent data and zero-altered negative binomial for count data. The results show that LWD and CPG prefer similar habitats. Fine scale measures of habitat quality influenced habitat selection and species distribution most, while a combination of anthropogenic activities and habitat quality were the best predictors of selection at a density level. Our habitat suitability models identified important areas and selection models enlighten habitats that need more attention. Furthermore, density model provides progressively more information for conservation. Our modeling approaches provide a baseline explanation of waterfowl habitat selection which could be extrapolated to Malaysia and even South East Asia.

Preview Text FH 2014 18 - IR.pdf Download (2MB) | Preview

Actions (login required)

View Item