Over-irrigation of Coffea canephora in the Central Highlands of Vietnam revisited : Simulation of soil moisture dynamics in Rhodic Ferralsols

An irrigation study was carried out to evaluate farmers’ irrigation practices on Rhodic Ferralsols which dominate the Dak Lak basalt plateau of Central Vietnam, where Robusta coffee has dramatically expanded over the last years. The experiment aimed at testing crop response on three irrigation application depths; 100 mm (advised dose), 80 and 60 mm. Initially the one-dimensional WAVE 2.1 model was evaluated for its prediction capabilities of soil water dynamics under tropical conditions. The model was calibrated with an experimental dataset of 2001, validated with an independent dataset of 2002, and verified with statistical indicators such as average error, root mean square error, model efficiency and coefficient of residual mass. The results validated the simulation with an overall model efficiency that increased towards the deeper soil layers. Subsequently the model was used to evaluate the issue of over-irrigation and to verify the water requirement for Coffea canephora. Therefore soil water stress was related to the physiology of coffee blossoming. Seven to eight days of optimal root water extraction after an irrigation event is an absolute minimum for optimal flower set. The lowest simulated irrigation depth of 60 mm indicated no water stress for 13 days after irrigation. Data collected in situ confirmed this. No significant differences in blossom set, yield and harvest quality were observed between the three treatments. A reduction of the actually advised application depth to 60 mm whenever soil water content drops below 30 vol.% (on average every 20–25 days) can hence be advised. This goes hand in hand with a reduction in deep percolation losses of 40%.