Bhavani Shankar*
Department of Agricultural and Food Economics, University of Reading, P.O. Box 237, Reading, RG6 6AR, UK
Ashley Halls
Renewable Resources Assessment Group, Department of Environmental Science and Technology, Imperial College of Science, Technology and Medicine, Prince Consort Road, London, SW7 2AZ, UK
Julian Barr
ITAD Ltd, Ditchling Common, Ditchling, Hassocks, West Sussex, BN6 8SG, UK
Bangladesh; Fisheries; Floodplains; Rice; Irrigation; Water.
Crop-Soil-Water Management
Water management
Data and study site The simulations presented in this study are based upon a model calibrated to extensive hydrological and biological data collected from a section of the Pabna Irrigation and Rural Development Project (PIRDP) in Bangladesh. The data, covering the hydrological year 1995–1996, were collected during the course of a research project ‘Fisheries dynamics of modified floodplains in Southern Asia’ (Hoggarth and Halls, 1997) funded by the Department for International Development, UK. Located to the south of the lower Atrai basin in the Northwest region of Bangladesh, the PIRDP is a large flood control, drainage and irrigation (FCDI) project. Almost 160 km of embankments protect an area of about 184,000 hectares from the flooding of the Atrai river to the north, the Brahmaputra to the east and the Ganges to the south (FAP 17, 1994). The data used in this research were collected from a study site based in the Southwest corner of the PIRDP scheme, at the confluence of the Jamuna and the Padma. Detailed biological data were collected using catch/effort, length frequency and mark-recapture surveys, as well as data on morphological details of local dry-season water bodies and hydrological parameters such as sluice gate management practices. The area is in Agro-Ecological Zone 12, the Low Ganges River Floodplain, and has been historically prone to deep flooding from the Jamuna and the Padma. Although medium-high lands occupy the largest area, the extent of medium-low, low and very low lands is substantial. In accordance with its flooding status and elevation make-up, the PIRDP has historically been one of the major deepwater rice areas of Bangladesh. However, the construction of the FCDI structure has encouraged the expansion of winter rice cultivation in the area. As in other parts of the country, Boro cultivation is now omnipresent during the dry season at the study site. A local socio-economic study conducted by FAP 17 (1994) confirms that water abstraction from DSWB for Boro irrigation is widespread in the area. Thus it provides a good setting for our proposed effort to quantify the abstraction-fish production tradeoff. 5.3 Methodology In order to assess the effects of surface water abstraction on the fishery, typical irrigation schedules were established, and corresponding volumes of water were ‘removed’ from the baseline distribution of water levels observed in the DSWB at the study site. The FPFMODEL is based on weekly water levels, and hence the irrigation schedules and abstraction were expressed on a weekly basis. Three alternate Boro irrigation schedules were put together. Multiple schedules were necessitated by the wide variation in Boro water applications in Bangladesh. The application rates and their temporal distribution through the Rabi season were established on the basis of previous literature (Biswas and Mandal, 1993; Mandal and Dutta, 1995; Sufian, 2001). Additionally, experts on rice production and irrigation from the Bangladesh Agricultural Research Council (BARC) were consulted in preparing the schedules. The developed schedules are presented in Table 4. In all cases, land preparation is assumed to begin in the second week of December, continuing on to the first week of January. Transplanting is assumed to occur in the second week of January, and harvest in the second week of April. It is assumed that there is no rainfall during the Boro growing season. Obviously, factors such as the number of applications and the amount of water in each application are subject to wide variation. The developed schedules attempt to present an ‘average’ or ‘typical’ picture, with consideration given to widely observed local farmer practices, such as reductions in water applications in weeks seven and eight after transplanting, to promote tillering. With the schedules established, the FPFMODEL was used to simulate the effect on PIRDP fish catches of these extractions. The 1995–1996 dry-season water levels were chosen as the baseline. Then the water volumes in the PIRDP dry-season waterbodies were reduced by the abstracted amounts, and the FPFMODEL was used to generate predictions of numbers of recruits, yield per recruit and other biological parameters, and of course, yield itself. The simulations (for each irrigation schedule) proceeded by assuming one hectare of Boro irrigated by abstraction, and then two, three, and so on until the abstraction levels caused recruitment failure.
Int. J. Water, Vol. 3, No. 1, 2005
Journal