Analyzing potato response to irrigation and nitrogen regimes in a sub-tropical environment using SUBSTOR-Potato model

•Performance of SUBSTOR-Potato model was exceedingly well for a range of irrigation and N regimes in sub-tropical environment of north-west India.•Potential tuber yield was high for October 1 planting that was reduced for later (October 16 and October 31) plantings.•Tuber yield and ET-based water productivity were greater on a sandy loam than a loamy sand soil.•Greater initial soil water status reduced post-sown irrigation and N needs than with lower initial soil water for realizing comparable tuber yields.•Initial soil mineral-N also affected fertilizer N and irrigation regime to achieve a given yield.

Increasing water scarcity and rising fertilizer cost warrant their judicious use in the world agriculture. It implies that management interventions should synchronize crop needs with soil supplies of these resources in the root zone to reduce probability of water and nutrient stresses and yield loss. This study examines the applicability of SUBSTOR-Potato model in analyzing potato response to irrigation and nitrogen regimes in a semi-arid sub-tropical environment of Punjab. In order to evaluate the model, database was generated by establishing a field experiment on potato (cv. Kufri Badshah) planted in mid-October on a sandy loam soil. Combinations of three irrigation regimes viz., irrigation water (IW) to pan evaporation (Ep) ratios of 0.8, 1.4 and 2.0 and four N rates viz., 0, 135, 180 and 225 kg ha−1 were randomized in a split-plot design with three replications. Half dose of N for a given rate was applied at planting and the rest half with a common irrigation 4 weeks after planting. The crop was managed for pest control by adopting local recommendations and harvested in end-January.Irrigation and fertilizer N had significant effects on tuber fresh yield, water use and N uptake. Performance of the SUBSTOR-Potato model was reasonable as indicated by close agreement of simulated crop phenology, biomass, water use, tuber yield and N uptake with the measured data. The normalized root mean square of deviations (RMSD) between simulated and measured values for harvest-time dry biomass and fresh tuber yield was 7.3 and 12.6%, while normalized RSMD for seasonal water use (ET + drainage) and total (tuber + haulm) N uptake was 12.4 and 19%. Simulation of tuber yield for independent data (2008–2009) was as good as for calibration data (2010–2011) giving confidence in the model. Scenario analysis based on historical weather data showed that mean potential tuber yield was 50.8 t ha−1 for October 1 planting that reduced to 41.8 and 37.8 t ha−1 for later (October 16 and October 31) plantings. The analysis also demonstrated that yield and ET-based water productivity (WPET) were greater on a sandy loam than a loamy sand soil for comparable irrigation and N regime. Interaction effects of irrigation and N on yield and ET indicated that irrigation response was greater in the presence of N. Greater initial soil water status reduced irrigation and N needs than with lower initial soil water for comparable tuber yield. Initial soil mineral-N also affected N and irrigation regime to achieve a given yield.