Abstract :
[en] In recent decades, climate change has become the major constraint for agricultural production in Cambodia and leads to future concern of food security. Improving effective irrigation management, especially for non-rice crops that use less water is urgently needed.
The aim of this thesis is to contribute the development of vegetable production during the dry season in Cambodia. The specific objective of the research is to optimise irrigation water use at the on-farm level. We focused on developing a methodology for i) characterising soil hydraulic properties for crop models and ii) exploring the best irrigation scenarios for vegetable irrigation. Two growing season experiments with lettuce were conducted during 2016 and 2017 in five farm fields in the Chrey Bak catchment, Kampong Chhnnag Province, Cambodia.
Two approaches were tested to achieve irrigation water saving. Firstly, a method using a soil water model, HYDRUS-1D, was used to inversely estimate the van Genuchten soil hydraulic functions in the unsaturated zone. The five experimental fields had different soil textures, loamy sand, sand and loam, and the field data (i.e., irrigation amounts, weather, lettuce growth data, soil permeability, etc.) were collected and measured to feed the given model. To generate the soil parameters, the objective functions were generated using inverse data such as measured soil moisture dynamics and soil water retention curves, using a combination of a soil moisture sensor, 10HS, and soil potential sensor, MPS-2, for a 30 minute time step. Our analysis showed that the inverse modelling successfully estimated the soil water retention curve and soil water dynamic with reasonable accuracy when compared to the observed values. However, uncertainties of the simulation and data measurement were observed, especially for the SWRC in dry and wet conditions. For the second study approach, to explore irrigation water saving the AquaCrop model was used to simulate irrigation scheduling under water stressed conditions.
In the second study approach, to explore the irrigation water saving, the water driven model, AquaCrop was selected for simulating irrigation scheduling under water-stressed conditions. The crop growth parameters for lettuce were calibrated using field data from the 2017 experimental growing season from two fields that have sand and loam soil textures. In the calibration process the measured crop growth data were used, e.g. canopy cover and above ground biomass collected over a 3 day time step, taking into account the other factors of irrigation management such as drip irrigation and mulching. Then, the method for optimal irrigation scheduling was described. Two main categories of irrigation scheduling for deficit irrigation were developed. The first varied thresholds of readily available water content (RAW) for the stop irrigation point under different no water stress and stress conditions. The second decreased deficit irrigations below field capacity (FC). The results of the calibration for crop growth were quite satisfactory. A primer set of adjusted lettuce parameters was obtained. The results highlighted limitations of the model in defining heat stress and root depth of the vegetable. Primarily, analysis results of irrigation scheduling scenarios show the capabilities of the model to identify the optimum water saving alternatives under limited water conditions.
Overall, this PhD thesis opens perspectives for improving irrigation management for increased crop productivity in Cambodia
