Summary of Abstract Submission

Abstract Submission No. IO50-09-0038PresentationPoster

APPLICATION OF MAXENT MODEL FOR EXPLORING INDIAN MACKEREL (RASTRELLIGER KANAGURTA) SPECIES DISTRIBUTION IN THE NORTHWESTERN ARABIAN SEA.

Dhyey Bhatpuria*1, H.U. Solanki1, Prakash Chauhan1

1 Space Applications Centre, ISRO, Ahmedabad, Gujarat, India

ABSTRACT :

In the present study, we aim to understand the usage of the maximum entropy method (MaxEnt) to explore the spatial distribution of Indian Mackerel, a commercially important fish species. Understanding of the spatial distribution of marine fishes is a preceding step for effective management and conservation. Spatial species distribution models along with remotely sensed data facilitates a novel opportunity to map and model marine species distributions. Presence-only data of 1998-2004 for the months of January-March and December were used for the study. The dataset was split into training and testing sets with a ratio of 80:20. Satellite derived chl-a concentration (Chl-a), sea surface temperature (SST), sea surface height anomaly (SSHA) and photosynthetically available radiation (PAR) were the environmental parameters considered apart from physical parameters like depth. The model was evaluated using the area under the curve (AUC). Model predictions for the months of February, December and March were higher (AUC > 0.8) indicating model¿s ability to discriminate suitable versus unsuitable areas for the species. Month wise potential species distribution maps could serve as a base for stock assessment surveys, fish prediction modeling, interspecies dependence studies and other ecological studies. MaxEnt model performed as an efficient tool for identifying potential habitat for Indian mackerel.