Calibration and validation of AquaCrop for irrigated peanut (Arachis hypogaea) in lowland rice systems of southern Laos

There are opportunities to enhance rural household incomes and improve nutritional status through the diversification of smallholder agricultural systems from wet season rice monoculture to the dryseason field and horticultural crop production e.g. maize, mung bean, soybean and peanut. In the semi-arid,rice-growing lowlands of southern Laos, there are a number of physiochemical soil constraints that prevent the successful production of non-rice cultivars in these systems which are further exacerbated by limited water supply. As a first assessment of yield potential, crop modelling offers a relatively inexpensive alternative to time-consuming, and often costly, field trials. However, in less developed regions, modelling is often restricted by limited technical capacity and fragmented datasets; thus, the use of complex models that require equallycomplex datasets and skills may not be locally appropriate and/or available. Therefore, the primary objective of this study was to calibrate and test the ability of the FAO AquaCrop model, which has comparatively simple input requirements and is relatively easy to use, to simulate peanut (Arachis hypogaea) production in common lowland rice soils under different irrigation regimes. In 2015/16, a field trial was conducted at Phone Ngam Rice Research Centre at Pakse in Champasak province, southern Laos to collate a robust agronomic dataset that could be used to assess the use of AquaCrop as a tool to adequately model peanut production in these environments. Results showed that the model performance in soils commonly found in the rice-based lowlandsystems was good under well-watered conditions when simulating canopy cover (%) and aboveground biomass (t ha-1) (nRMSE = 24.6 % and 30.6 t ha-1, respectively). However, performance in water-limited conditions wasmoderate to poor (nRMSE = 27.0 % and 35.1 t ha-1; and 44.2 % and 38.8 t ha-1, respectively). Furthermore, themodel was unable to simulate soil water with any degree of reliability given the limited lateral waterdistribution across the root zone, as evidenced by lack of soil water tension sensor response to irrigation andrainfall events, particularly once flowering had occurred. In addition to increased plant water use, the lack oflateral movement can be attributed to the physical characteristics of these hard-setting, sandy soils that areknown to collapse soon after cultivation. Therefore, further work is required to thoroughly evaluate the use ofAquaCrop to simulate soil water in relation to peanut production in the rice-growing lowland soils of the region.Subsequent experimentation to improve lateral water movement in the root zone, including: (i) the use oforganic soil ameliorants; and (ii) irrigation system design may assist in the improved calibration of the soilwater component of the model under these conditions. Lastly, as well as being the first assessment of peanutproduction in Laos, this paper also presents the first reported evaluation of AquaCrop use for the simulatedproduction of peanut (Arachis hypogaea), more widely.