Field Experimental and Modelling Assessment of The Effects of Different Water Management and N-Fertilizer Levels on Co2 Emissions, Net Ecosystem Carbon Balance and Rice Production

Abstract

Accurate measurement of carbon dioxide ( CO 2 ) emissions and water H 2 O ) fluxes is crucial for reducing global greenhouse gas (GHG) emissions and mitigating climate change, especially in agricultural land systems. To explore effective ways to improve rice production in Benin while protecting the environment, an adaptive, low cost, open source data logger has been developed to measure CO 2 , evapotranspiration ( ET )), and estimate the net carbon (C) balance from irrigated lowland rice cultivation in Benin. Laboratory tests and field validation on different Non Dispersive Infrared NDIR sensors showed that the K30 Fr and SCD30 sensors are the most accurate compared to the Li COR 850 IRGA. The field application aims to identify the best farming management practices that improve water and nitrogen use efficiency, grain yield, and the net ecosystem carbon balance while keeping CO 2 emissions low. It was established at Koussin Lele , southern Benin, using a split plot design where the main plot is different water management, and the subplot i s different level of nitrogen fertilizer. The water management i ncluded continuous flooding (CF) and two levels of alternate wetting and drying at 15 cm and 25 cm thresholds (AWD 15 and AWD 25 ). The different levels of nitrogen fertilizer were: no nitrogen, N 90 kg ha 1 , and N120 kg ha 1 . To measure gaseous seasonal C emissions and soil CO 2 respiration, the NFT NSS closed chamber system is used, and results showed that both net ecosystem exchange, the net C balance, and evaporation were higher under AWD 15 and C F even though all the treatments are sinks of carbon. The heterotrophic respiration was higher under AWD 25 . In terms of water productivity, grain yield, and N uptake, results showed that the treatment AWD 15 with N 90 kg ha 1 is the best. Consequently, this treatment should be promoted to reduce costs resulting from fertilizer inputs, conserve yield, and combat water scarcity. In addition, the calibration of the DSSAT CERES rice model was successful in simulating growth and yield, and the results illustrated that the model is a useful tool for evaluating alternative management options to maintain yields while saving water. As a recommendation, for further study, it could be interesting to add manure fertilizer and assess CH4 emissions from each treatment. In perspective, this study will help to make decisions regarding the C balance and the optimization of water and nitrogen inputs for sustainable irrigated rice.