Agriculture`s impact on water–energy balance varies across climates

Masoud Zaerpour, Shadi Hatami, André S. Ballarin, Simon Michael Papalexiou, Alain Pietroniro, and Ali Nazemi

PNAS; March 17, 2025; 122 (12) e2410521122; https://doi.org/10.1073/pnas.2410521122

Significance

Agriculture plays a key role in global food security, intricately tied to water resources for crop growth. However, navigating the interplay between agriculture and water availability poses challenges, especially during the Anthropocene, where traditional perspectives often overlook agriculture’s impacts on the water cycle. Understanding and integrating agriculture’s influence on water dynamics becomes imperative in addressing contemporary challenges. Our study highlights the contrasting impacts of agricultural activities across temperate and snowy climates. In temperate catchments, agriculture weakens the precipitation-streamflow (P-Q) relationship, contributing to precipitation-driven deviations from the water–energy balance, while in snowy catchments, agricultural activities strengthen the P-Q relationship. These findings offer insights for shaping effective water management strategies, ensuring food security, and promoting sustainable development globally.

Abstract

Agriculture is a cornerstone of global food production, accounting for a substantial portion of water withdrawals worldwide. As the world’s population grows, so does the demand for water in agriculture, leading to alterations in regional water–energy balances. We present an approach to identify the influence of agriculture on the water–energy balance using empirical data. We explore the departure from the Budyko curve for catchments with agricultural expansion and their associations with changes in the water–energy balance using a causal discovery algorithm. Analyzing data from 1,342 catchments across three Köppen-Geiger climate classes—temperate, snowy, and others—from 1980 to 2014, we show that temperate and snowy catchments, which account for over 90% of stations, exhibit distinct patterns. Cropland percentage (CL%) emerges as the dominant factor, explaining 47 and 37% of the variance in deviations from the Budyko curve in temperate and snowy catchments, respectively. In temperate catchments, CL% shows a strong negative correlation with precipitation-streamflow (P-Q) causal strength (Spearman ρ=−0.75), suggesting that cropland exacerbates precipitation-driven deviations. A moderate negative correlation with aridity-streamflow (AR-Q) causal strength (ρ=−0.42) indicates additional influences of cropland through aridity-driven interactions. In snowy catchments, CL% is similarly influential, with a positive correlation with P-Q causal strength (ρ=0.51). However, the negative correlation with AR-Q causal strength (ρ=−0.45) underscores the role of aridity as a secondary driver. While vegetation and precipitation seasonality also contribute to the deviations, their impacts are comparatively lower. These findings underscore the need for inclusion of agricultural activities in changing water–energy balance to secure future water supplies.

See https://www.pnas.org/doi/10.1073/pnas.2410521122

Figure 1

Quantifying the causal link strength between main components of water–energy balance including streamflow, precipitation, and AR. The strength of these causal links is assessed through the MCI test statistic, where MCI values range from –1 to 1. Panels (A) and (B) illustrate the strength of causal links between streamflow-precipitation (P-Q) in the United States and Great Britain (GB), respectively. Panels (C) and (D) depict the strength of causal links between streamflow-aridity (AR-Q) in the United States and GB, respectively. White dots indicate catchments where the causal link strengths are not statistically significant.