Soil salinity and acidity : spatial variabil[it]y and effects on rice production in West Africa's mangrove zone

In the mangrove environment of West Africa, high spatial and temporal variability of soil constraints (salinity and acidity) to rice production is a problem for the transfer and adoption of new agronomic techniques, for land use planning, and for soil and water management. Recently, several National and International Agricultural Centers have initiated research programs to characterize environments where their newly developed technologies have to be applied. However, the mangrove agro- ecosystems in West Africa have not been characterized in a detailed way. Most of the soils in this environment are potential or actual saline acid sulfate soils. The spatial and temporal variability of soil salinity and acidity in these coastal lowlands results from complex interactions between climate, coastal morphology, river hydrology, vegetation, landform and tidal flooding. Diagnosing the occurrence of both potential and actual acid sulfate soils is the first step in land use planning for such areas. But to cope with the intricacies of these soils, understanding the processes of soil salinization and acidification at different scales should be formalized to properly characterize mangrove environments.The main objectives of this thesis were: 1) to give a comprehensive characterization framework for the West African mangrove environments with emphasis on the possibilities of and constraints for rice cultivation; 2) to determine the various causal factors for soil salinization and acidification; 3) to test whether temporal variability of soil chemistry is sufficient to provide a time window of minimum stress during the rice growing period; 4) to relate the response of rice to improved agronomic practices in specific environments and to provide a means to characterize specific rice growing locations, and 5) to test rice varietal responses to saline and acid sods under different agronomic practices and to relate yields and yield components to the nutrient contents in leaves, in order to diagnose physiological disorders.First, a multi-scale approach was developed involving a range from Macro to Micro level based on the pre-conditions of acid sulfate soil formation. The main factors for classification are climate and coastal morphology at Macro scale; hydrology, physiography and vegetation complexes at Meso level; and topography (catena), vegetation species, tidal flooding and sedimentation rate at Micro level. Information from previous process-based studies on acid sulfate soil formation and data from secondary sources were used. Different environments were then distinguished and their characteristics were summarized by ecological zone. Constraints to rice production and potentials for agricultural development were matched with environmental conservation issues.To determine the significance of the causal factors developed in the multiple scale approach, 12 sites were selected along 4 river basins in West Africa, vz. from north to south the Gambia, the Casamance (Senegal), the Geba (Guinea Bissau) and the Great Scarcies (Sierra Leone). Along each river basin 3 sites were selected based on distance from the river mouth. Within a site a strip of land perpendicular to the river was selected for intensive grid sampling (40 by 20 m). Soil samples were taken at each grid node during the dry season of 1991. The relation between causal factors and soil salinization and acidification was determined at Macro and Meso levels by nested ANOVA and yielded a classification of the study area in main ecoregions and sub-environments within watershed. At a detailed scale, geostatistics were applied and zones within catena were defined in terms of their main soil characteristics. A nested statistical approach and geostatistics were used complementarily to disentangle the complexity of the causes of soil salinization and acidification.Temporal variability was studied by monitoring soil solution chemistry at each main landscape unit within the catena. Since the production of rice critically depends on the lowering of salinity and acidity by natural flooding during rainy season, time windows during which soil limitations are minimal were defined and matched with rice varietal duration.The response of rice to different improvement techniques were tested by means of a network of trials in the 1991 and 1992 rainy seasons. The residual effects of lime and phosphate rock (applied in 1991) during 1992 was also evaluated. Lime dressing (2 t ha <sup>-1</SUP>) was found effective whenever dissolved Ca and Mg in the soil were low, and had a clear residual effect in the year after application. Application of phosphate rock did not seem to be effective in general. For iron toxicity, the molar fraction of Fe and (Ca + Mg) in soil solution and in flag leaves were found to be more relevant for diagnosing physiological disorders than the absolute Fe content in the soil solution and in rice flag leaves at panicle initiation.In the 1993 rainy season, rice varietal behavior under different improvement techniques within the main soil limitations in the mangrove environment was tested. Differences in yield and yield components and element contents in flag leaves at panicle initiation were observed between varieties in saline and less saline acid soils. Multiple correlation between rice yields, yield components, element contents in flag leaves at panicle initiation was found to be an effective diagnostic tool for assessing physiological disorders.The approach used in this study provides a logical framework to describe mangrove environments. The multiple-scale can assist in identifying the information required to cater for the needs of various decision-makers and land use planners. It also provides a key to develop technology packages for intensified and sustainable use. It can be used for the extrapolation of site-specific information to geographically different areas, with similar characteristics.