Management of soil and water resources for higher productivity of coastal saline ricelands in Bangladesh

In the topsoil (0-15 cm), salinity in the dry season varied from 4.0 to 9.0 dS/m in moderately saline soils at Mirzapur and from 5.0 to 12.0 dS/m in saline soils at Barodanga. In wet season (mid-July to mid-December), the corresponding figures were from 1.5 to 2.5 dS/m and 2.0 to 3.0 dS/m, respectively. In deeper soil layers (15-100 cm), salinity levels mostly remained below 4.0 dS/m, the critical salinity level for most crops. Dry season cropping significantly reduced topsoil (0-15 cm) salinity at both research sites. In fallow lands, topsoil salinity rose up to about 9.0 and 12.0 dS/m at Mirzapur and Barodanga, respectively. Overall peak salinity in cropped lands was 38 percent lower than that in fallow lands at both locations. Peak salinity in plowed cropped lands at Barodanga was about 40 percent less than that in non-plowed cropped lands. Moreover, topsoil salinity in cropped lands came down to 4.0 dS/m or less at least one month earlier than fallow lands. Such a low salinity level opened up an opportunity to grow modern rice varieties (whether salt-tolerant or not) in the wet season, by using rainfall, much earlier than the present cropping schedule and offered a yield advantage of about 1.5 t/ha over the present productive level. Slightly saline groundwater was used to irrigate dry season crops. Of all the crops grown only sesame responded positively to irrigation. Yields were not significantly different for any crop between irrigated and rainfed conditions mostly because of about 60 mm rainfall occurred at the pod formation stage of crops, a week after irrigation. Although groundwater was slightly saline, it reduced topsoil salinity when used to irrigate dry season crops. Therefore groundwater can be used to irrigate profitable non-rice crops in the dry season. The marginal rate of return indicated that sesame and mungbean were highly profitable in saline soil environments, but mungbean grown under rainfed conditions maximized farmer's income from dry season cropping. The deepest water table was within 1.0-1.5 m below ground surface at both locations. The aquifer was fully recharged and the water table remained either above or close to the soil surface from the last week of June to the first week of December. Rainfall amount in 1996 was similar to long-term average rainfall recorded at the Khulna meteorological station. Moreover, the aquifer was fully recharged by the end of June, leaving more rainy months to contribute to groundwater recharge. Multiple linear and non-linear regression models were developed to predict topsoil salinity (ECe) and dry season moisture contents (MC) of fallow lands for both moderately saline (MS) and saline soils (SS). The independent variables, average daily rainfall (RF), and evaporation (EV) between two samplings were used for the salinity model. Total rainfall (TRF), total evaporation (TEV), and average water table depth (WTD) below ground surface were used for the soil moisture model. The linear and non-linear salinity models showed almost similar prediction of topsoil salinity. But soil moisture was better prdicted by the non-linear model. The recommended salinity and moisture models are presented