Spatio-temporal analysis and forecasting of fish production in Bangladesh: A comparative study of pond and river using GIS and ARIMA model

The identification and monitoring of changes in land use patterns for pond construction compared to the existing river area and fish production from ponds and rivers by spatiotemporal analysis of Bangladesh, at large, is unknown. This study aims to identify aquaculture clusters developed over the last two decades (2001-2020) in a spatiotemporal approach by applying geographic information systems (GIS) and forecast pond area, pond, and river production for the next 10 years using the autoregressive integrated moving average (ARIMA) model based on secondary data collected from the Department of Fisheries (DoF). The results show that a high volume of pond area and production is concentrated in north-west, North-central, south-west, and south-east regions of Bangladesh. River production is concentrated in south-central and south-east regions of Bangladesh, especially in Bola, Barishal, Patuakhali, Borguna, Chandpur, and Lakshmipur districts. Over the last 19 years, pond area has increased by 52.35%, aquaculture (pond) production has increased by 199%, and capture fisheries (river) has increased by 131%. Although river production has grown, it is still six times less than pond production. In south-central Bangladesh, river fish production showed a reciprocal relationship with aquaculture production. The ARIMA model forecasts that land acquisition through pond construction will increase to 3.13% of Bangladesh's total area by 2030, up from 1.71% in 2001. The pond area will be increased by approximately 75% between 2001 and 2030. In 2030, the production of aquaculture and capture fisheries will be 2.88 million MT and 0.6 million MT, respectively. Our study recommends some important strategies for policymakers to formulate a region-based approach and promote commercial upgrading of aquaculture production with modern technologies, followed by limiting land acquisition.