Capturing cultivar differences in water-limited yield potential of potato in the Netherlands via experimentation and modelling

Climate change leads to more frequent droughts. In recent years, potato production in the Netherlands suffered, particularly on sandy soils. Farmers need decision-support for crop management and adaptation to climate change and variability, but currently crop models do not capture cultivar differences and are therefore not suited for on-farm decision support. Recently, new WOFOST parameter sets for potential yield of potato in the Netherlands were calibrated (ten Den et al., in review). These parameter sets allow for cultivar-specific estimation of potential yields; of the seven parameters five are correlated with cultivar earliness. With the old parameter set (derived in the 1993) the potential yield in 2019 was 12.6 t DM ha-1 while with the new parameter sets the potential yields ranged from 16.3 to 18.5 t DM ha-1 (66 to 83 t ha-1 in FM). In a next step we will calibrate water-limited potential yields of modern potato cultivars using WOFOST, in combination with SWAP.
Two experimental fields were set up, one on a sandy soil near Vredepeel and one on a clayey soil near Lelystad. Throughout the season data was collected on crop phenology and growth, e.g. leaf biomass, stem biomass, and tuber biomass as well as on soil conditions such as groundwater level, soil moisture content (pF), and soil temperatures. Two irrigation treatments were used to allow for assessment of both a potential and water-limited production situation. The optimally irrigated plants were irrigated to maintain a target pF value of 2.4 while the drought stressed plants had a pF target of 3. The experiment was repeated over two years: the 2020 data were used for calibration and the 2019 data for model validation. The difference in water supply between the two irrigation treatments for Lelystad was 121 mm (2019) and 73 mm (2020). In Vredepeel the difference was 148 mm (2019) and 139 mm (2020). Five different commercial potato cultivars were used, with for Lelystad the cultivars Innovator, Fontane, and Markies and for Vredepeel Premiere, Fontane, and Festien. Drought stress in 2020 was more severe in Vredepeel than in Lelystad with an average 21% yield reduction in Vredepeel and 6.9% in Lelystad. There was no noticeable yield reduction due to drought in 2019 in Lelystad and a reduction of 13% in Vredepeel. There was large variation between the three replications of the trial which made that only the yield reduction for the early cultivar Premiere in Vredepeel in 2020 was significant. There was also a large within treatment variation in measured LAI and leaf biomass. For Lelystad there were few differences in LAI and leaf biomass between the irrigation treatments. In Vredepeel the strongest effect was for Premiere in which the drought stressed plants had a lower LAI from July onwards for both the 2019 and 2020 seasons. Drought stress had little to no effect on stem biomass, plant height, or light interception.
For modelling water-limited production the models WOFOST and SWAP-WOFOST are used.