Habitat suitability assessment for the endangered Nilgiri Laughingthrush: A multiple logistic regression approach

Bombay Natural History Society, 400 023, Mumbai, India; Department of Wildlife Ecology and Conservation, University of Florida, Gainesville, FL, USA; Indian Institute of Remote Sensing, Department of Space, Government of India, 248 001, Dehradun, India; Present address: Centre for Biodiversity Studies, Baba Ghulam Shah Badshah University, Rajouri, India
Current science (Impact Factor: 0.83). 06/2008; 94(11):1487-1494.

ABSTRACT Application of remote sensing and Geographic Informa-tion System (GIS) tools has assumed an increasingly important role in conservation biology and wildlife management by providing means for modelling poten-tial distributions of species and their habitats, unlike the conventional ground surveys. We present here a predic-tive model of habitat suitability for the Nilgiri Laugh-ingthrush, Garrulax cachinnans based on a synergistic use of field surveys and digitally processed satellite im-agery combined with features mapped using GIS data layers. Collateral data were created in a GIS framework based on ground surveys comprising layers such as land-use, measures of proximity to likely features of distur-bance and a digital terrain model. Multiple binomial logistic regression approach was used for modelling, and the model performance was assessed by the area under the receiver operating characteristics (ROC) curve. About 320 km 2 , 25.12% of the area of the Nilgiris considered for modelling was predicted to be suitable for the Nilgiri Laughingthrush. The area under the ROC curve was found to be 0.984 ± 0.003 (R 2 : 0.93 at P < 0.0001), implying a highly effective model. The as-sessed suitable habitat was highly fragmented and comprised of 1352 patches (natural as well as man-made) distributed all over the study area. The smallest suitable patch identified by the model was 400 m 2 and the largest patch 17.65 km 2 . Also, ca. 92% of all patches were smaller than 0.5 km 2 . We presume that some suit-able habitat patches may be unoccupied due to strong fidelity of the species to shola (montane wet temperate forest) patches, low colonization rates, or large inter-fragment distances. Also, larger fragments might serve as source or 'exporters' of surplus individuals to main-tain sink populations throughout the rest of the range. We discuss the implication of habitat fragmentation and narrow geographical range and anthropogenic pressure for the conservation of the Nilgiri Laughing-thrush.

  • Ecological Modelling - ECOL MODEL. 01/2011; 15(1):1-3.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Application of remote sensing and Geographic Information System (GIS) as a tool has assumed an immense significance in habitat suitability modelling for various wildlife species and now a days these are widely used in conservation biology and wildlife management.This paper evaluates habitat suitability for tigers (Panthera tigris tigris) in Chandoli National Park, India. This research provides information about potential areas that can be declared as Tiger Reserve, a federal designation of protection. Habitat evaluation was completed in the Chandoli National Park (17° 04′ 00″ N to 17° 19′ 54″ N and 73° 40′ 43″ E to 73° 53′ 09″ E) using habitat suitability index (HSI). Remotely sensed data of satellite IRS-P6, LISS-III of 25th February 2005 was procured from National Remote Sensing Agency, India. The satellite imagery data was digitally processed and collateral data were generated from topographic maps in a GIS framework. Various layers of different variables such as landuse land cover, forest density, measures of proximity to disturbances and water resources and a digital terrain model were created based on ground truthing. These layers, GPS location of animal's presence and “binomial multiple logistic regression (BMLR)” techniques were integrated in a GIS environment for the HSI modelling.Results indicate that approximately 160.48 km2 (55%) of the forest of Chandoli National Park is highly suitable for herbivores population, and 176.52 km2 (50%) of the Park is suitable for tigers. Therefore, this study concludes that forest areas within the Chandoli National Park are appropriate for consideration as Tiger Reserve.Furthermore, the identification and mapping of wildlife corridors and other forest patches lying in between neighboring protected area like Koyana sanctuary, suggests that these region of Sahyadri range of Western Ghats should be included into the proposed Tiger Reserve. This will be a benefit for tigers, as tigers require larger and least disturbed forest habitats for establishing and maintaining their high-density population. Fortunately, in May 2008 the Government of India declared Chandoli NP as tiger reserve. Thus declaration of Chandoli National Park as a Tiger Reserve reflects that the habitat suitability model developed by us for tiger was not only statistically sound but also has potential to be considered while taking any decision regarding wildlife habitat management.
    Ecological Modelling 12/2009; 220(24):3621–3629. · 2.07 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Los Planes de Ordenación de Recursos Forestales (PORFs) contemplan como un objetivo fundamental el mantenimiento de la biodiversidad a distintos niveles (poblaciones, hábitats, ecosistemas y paisajes). En este sentido, resulta necesario poner a punto y contrastar métodos para, utilizando como base la información existente, establecer modelos de disponibilidad de hábitats, relaciones hábitat ocupado/hábitat disponible y modelos de conectividad y/o fragmentación para determinadas especies a distintas escalas. En el presente trabajo se propone un ejemplo de esta metodología aplicada a dos especies de rapaces forestales, el abejero europeo (Pernis apivorus) y el gavilán (Accipiter nisus). La zona de estudio comprende un sector de 30.000 km2 del centro-este peninsular con las comarcas forestales de Urbión y Alto Tajo como límites septentrional y meridional respectivamente. Los datos de presencia de las especies proceden del Atlas de las Aves Nidificantes de España, mientras que los correspondientes a variables ambientales (climáticas, topográficas, etc) proceden de distintas fuentes digitales. El análisis de los datos y las representaciones de los resultados se realiza mediante Estatistica y ArcGIS.
    5º Congreso Forestal Español, Avila; 09/2009

Full-text (3 Sources)

Available from
Jun 4, 2014