Conservation Agriculture: Its effects on crop and soil in rice-based cropping systems in Bangladesh

Intensive rice-based cropping systems in the Eastern Indo-Gangetic Plains (Eastern India and Bangladesh) have played a pivotal role in increasing food security in that region but sustainability of these cropping systems is under threat. Conservation agriculture (CA) ‒ cropping systems based on minimum soil disturbance, crop residue retention and suitable crop rotations ‒ has been proposed to address these challenges but there has been limited research on its effects on crop productivity and soil properties in Bangladesh. This thesis examines the effects of implementing minimum soil disturbance and increased crop residue retention on soil properties and crop performance over three years in two rice-based rotations. Two field trials were conducted during 2010-2013 in contrasting triple-cropping rotations (with crop number in parentheses): 1. Legume-dominated rotation ‒ lentil (1, 4 and 7)-mungbean (2 and 5)-monsoon rice (3 and 6) in an Alluvial soil region; and 2. Cereal-dominated rotation ‒ wheat (1, 4 and 7)-mungbean (2 and 5)-monsoon rice (3 and 6) in the High Barind Tract (HBT) region of north-west Bangladesh. There were three tillage treatments in main plots ‒ strip tillage (ST), bed planting (BP) and conventional tillage (CT). Sub-plots comprised two levels of residue ‒ high residue (HR) and low residue (LR). Puddled transplanted rice was applied in CT and unpuddled transplanted rice in ST and BP. This thesis focuses on soil properties and the growth and yield of the cool-dry season crops in each year, namely lentil on the Alluvial soil and wheat on the HBT soil. During the first two growing seasons treatment effects on soil properties and crop performance were marginal but became clearly apparent in the third year. In the legume-dominated system, grain yield of lentil was 15 % higher in HR than LR averaged across all tillage types in Year 2. In Year 3, the yield of lentil was higher by 23 % in ST and 18 % in BP compared with CT. In the cereal-dominated system, grain yield was not affected by tillage and residue treatments in Year 1. However, in Year 2, grain yield of wheat was depressed by 39 % in BP due to poor crop establishment. In Year 3, the yield of wheat was greater by 9 % in ST and 7 % in BP than CT; wheat yield of HR was 3 % higher compared to LR. The soil water content (SWC) increased and bulk density (BD) and penetration resistance (PR) decreased in surface soil (0-5 cm) with ST and at 5-10 cm and 10-15 cm soil depth with BP, compared to CT. The retention of more intact residue left between the plant rows conserved more SWC and lowered the soil BD and PR of surface soil under ST. Implementation of ST and BP with HR treatment gradually improved soil physical properties and alleviated puddling effects that characterise current practices (CT and LR) in rice-based systems. Such improvements are probably due to increases in soil organic carbon (SOC) and total nitrogen (TN) with ST and BP. Greater root growth under BP was not associated with increased grain yield. However, the overall improvement in soil surface conditions and greater root growth at depth may have allowed extraction of water and nutrients from a larger soil volume in ST resulting in a gradual increase in crop productivity over time. After 2.5 years in both legume- and cereal-dominated rotations, the SOC concentrations, SOC-stocks and labile C fraction (water soluble carbon ‒ WSC) at 0-7.5 cm soil depth were greater in ST than CT. By contrast, the SOC concentrations and storage, and WSC increased at 7.5-15 cm soil depth in BP compared to CT and ST. Soil C losses through the emission of CO2 were greater in CT than ST and BP. The relative efficacy of tillage in storing SOC was in the order of ST>BP>CT. High residue retention increased SOC concentrations, SOC storage, WSC and CO2 emission from soil. In the cereal-dominated rotation, ST sequestered 0.44-0.20 Mg C/ha annually while CT caused 0.41-0.66 Mg C/ha loss at 0-15 cm soil depth. In contrast to the legume-dominated rotation, neither CT nor ST sequestered SOC but ST reduced the loss by 0.40 Mg C/ha annually compared to CT. Based on the C balance, it is estimated that annual carbon inputs of 3.8 Mg C/ha under ST and 6 Mg C/ha under CT condition in the legume-dominated system, and 1.0 Mg C/ha under ST and 7.7 Mg C/ha under CT condition in the cereal-dominated system, would be required to maintain SOC at the antecedent level. In the present study, ST and HR treatment increased TN, N-stocks, total soluble nitrogen and potentially mineralizable nitrogen (PMN) in the surface soil (0-7.5 cm) as compared to CT and LR at the end of Crop 7. In ST and HR, the lower mineral N (NH₄-N and NO₃-N) and larger PMN indicated the greater immobilization or less mineralization of N, or both, and restricted the potential losses of N. Retention of HR resulted in positive N balance while LR caused a negative N balance. Regardless of treatment variation, the soil TN, N-stocks and available N were greater in the cereal-dominated cropping system than in the legume-dominated system, probably due to carry-over of higher N fertilizer rates applied to the cereal crop, and greater above- and below-ground biomass. The changes of soil TN due to residue were only apparent in legume-dominated system. The greater input derived from nitrogenous residue of mungbean and lentil may account for the positive effects of HR in the legume-dominated system. Application of ST and HR has potential for increasing carbon sequestration and N accumulation while reducing N losses, hence improving soil properties and thereby crop growth and yields, within 2-3 years in rice-based systems of Bangladesh. However, further studies are required over a longer time period to evaluate the performance of unpuddled rice rotated with ST non-rice crops with a range of residue retention levels under different soil, climatic, and socio-economic conditions in the eastern Indo-Gangetic Plains