Ensuring sustainable crop production when yield gaps are small: A data-driven integrated assessment for wheat farms in Northwest India

Northwest India achieved remarkable wheat productivity gains during the past decades. However, this has been accompanied by increasing input levels and intensive production practices, raising questions about the economic and environmental sustainability of current cropping systems. A multicriteria integrated assessment is required for wheat farms in the region to understand the scope for cleaner wheat production in the future. Production practices from irrigated wheat fields (n = 3928) were evaluated for multiple sustainability indicators, namely yield gap, nitrogen (N)-use efficiency, profitability, and greenhouse gas emissions. Stochastic frontier analysis was combined with simulated potential yield (Yp) data to identify the causes of wheat yield gaps in the region. N-use efficiency was estimated by calculating the partial factor productivity of N, profitability was computed based on reported input-output amounts and prices, and greenhouse gas emissions were quantified using the Mitigation Options Tool (MOT). These indicators were subjected to a multicriteria assessment using the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) under different scenarios (i.e., different weights for different indicators). For each scenario, farmers’ fields were classified as most efficient, efficient, less efficient, and least efficient, and random forest was used to identify the most important management practices governing the field classification. Wheat yield gaps were small (25–30 % of Yp or 2.4 t ha⁻¹) and mostly attributed to the technology yield gap (ca. 20 % of Yp or 1.5 t ha⁻¹). Ranking and grouping the farmers’ fields in the scenario with equal weights for all indicators revealed that at least 25 % of the fields had very high greenhouse gas emissions (>1500 kg CO₂-eq ha⁻¹) at a productivity level of < 4.5 t ha⁻¹, and that it is possible to produce wheat sustainably without compromising yields in Northwest India, as indicated by the performance of the most efficient fields. Tillage intensity and N application rates can be adjusted for least efficient fields (<10 % least efficient fields adopting zero tillage vs >80 % most efficient fields adopting zero tillage) to achieve an overall objective of higher yield, lower greenhouse gas emissions, more profit and higher N-use efficiency, whereas residue retention and tillage intensity would need to be prioritized for minimizing greenhouse gas emissions. For the most efficient fields the decrease in greenhouse gas emissions was always associated with a decline in yield level. The most important management practices governing the field classification included the crop establishment method used for the previous rice crop, the number of tillage operations, residue retention, and the N fertilizer rate for wheat. The study provides a data-driven approach to screen trade-offs between performance indicators and to identify the management practices that can deliver sustainable and cleaner crop production in the future.