Analyzing crop yield gaps and their causes using cropping systems modelling–A case study of the Punjab rice-wheat system, Pakistan

• Published in: Field crops research 2019-02, Vol.232, p.119-130
• Main Authors: Khaliq, Tasneem, Gaydon, Donald S., Ahmad, Mobin-ud-Din, Cheema, M.J.M., Gull, Umair
• Format: Article
• Language: English
• Subjects: APSIM; Crop modelling; Land and water productivity; Sowing dates; Sustainable intensification; Yield gaps
• ISSN: 0378-4290; 1872-6852
• DOI: 10.1016/j.fcr.2018.12.010

•Field experiments, regional farmer data and modelling were used in analysis.•Farmers currently achieve only around 36–67% of potential yields for rice and wheat.•Nitrogen deficiency was found to be the primary driver for the large yield gaps.•Further work to understand farmers’ reasons for low N-rates is required. Pakistan faces significant future challenges to feed a growing population, with 47% of the population currently categorised as food insecure. The rice-wheat (RW) zone of Punjab is the major aromatic rice producing area of Pakistan. Here the rice-wheat cropping rotation is practiced over a variety of soils under a range of agronomic and irrigation practices. It is important to quantify and understand the existing crop yield gaps and associated factors in this zone to identify opportunities for gap-reduction leading to increased land and water productivity and sustainable water use in Pakistan. The APSIM model was parameterised for local soils and climate, and then calibrated for rice and wheat growth, phenology and yields using experimental data sets. The model calibration was then validated using 5-year (2009–2014) farmer grain yield records from each of the highest (Gujranwala) and lowest (Narowal) performing districts in Punjab, before being used to simulate long-term (34 years) rice-wheat yields for current farmer practices and also for potential (no water or N-stress) production, in both districts. The revealed yield gaps were further assessed as a function of sowing date. Observed and simulated farmer R–W yields for Gujranwala were greater than those for Narowal; however potential yields were similar at both sites (for both rice and wheat). Farmers currently achieve only around 36% (Narowal) to 67% (Gujranwala) of potential yield for rice, and 48–56% of potential for wheat (Narowal and Gujranwala respectively). The modelling analysis of abiotic stresses under which the farmers’ crops were grown revealed that nitrogen (N) deficiency was the primary driver for the large yield gaps at both sites, for both crops, with low topsoil carbon (0.19% Narowal; 0.49% Gujranwala). The simulated farmer crops were not significantly constrained by water limitations under the current N application rates, or by high/low temperatures. Gaining a better understanding of farmers’ reasons for low N-rates is important – it could be related to (non-modelled) soil micronutrient constraints or risk aversion associated with irrigation water supply reliability. Current