ENSO influence on corn and soybean yields as a base of an early warning system for agriculture in Córdoba, Argentina

Abstract

Reducing uncertainty and productive risk increase the resilience of agricultural systems and food security of a region. While crop productivity strongly depends on weather conditions during the growing season, El Niño Southern Oscillation (ENSO) phenomenon is the main driver of global interannual climate variability and affects crop production in many regions. The ENSO, Pacific Decadal Oscillation (PDO) and Indian Ocean Dipole (IOD) impact on corn and soybean yields between 1973 and 2017 was compared in Córdoba Province, Argentina, looking for an early signal to support agricultural activity. The analysis suggests that not only ENSO signal (SOI and SST El Niño 3.4) but also PDO anomaly (PDOA) link better than IOD with productivity for both crops, as well as the Southern Oscillation Index of August and September (SOIas) is a proper yield predictor. An inverse relationship was determined between SOIas and yields anomaly, more spread territorially for soybeans since the linear functions are significant (p < 0.05) in 9 of the 12 administrative areas and, with a slightly lower significance (p < 0.1), in the remaining too. On the other hand, the low R2 values indicate little capacity to explain the yield interannual variability from the SOIas signal that for corn, has a departmental range between 2 % and 19 % and, for soybeans, slightly higher between 7 % and 24 %. While the mean difference test of yield anomaly between different ENSO phases (El Niño, La Niña and Neutral) supports the more general indifference of both extreme phases with the neutral cases, it reaches a significant character in 8 departments when the cold phase is compared with the warm. The seasonal rainfall regime has also a reduced capacity to explain productive variability. However, this information is useful to realize that both the wettest (El Niño) and the driest years (La Niña) fail to differentiate from those with normal rainfall. This confusion was verified by a combinatory frequency analysis of rainfall and productivity data related to the ENSO phases. Because ENSO allows anticipating the rainfall condition and the productive potential of the main rain-fed crops in Córdoba, Argentina, it constitutes appropriate information to reduce uncertainty and the risk level in agricultural production. Therefore, although the ENSO predictive capacity is limited and more knowledge about the influence of other climatic variability sources like PDO is needed, its incorporation into a climate monitoring protocol must conforms to the basis of an early warning system for a sustainable agriculture in the region.