ARS | Publication request: Dynamics of mean-variance-skewness of cumulative crop yield impact temporal yield variance

Submitted to: International Journal of Agronomy
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: May 24, 2011
Publication Date: July 31, 2011
Citation: Jaradat, A.A. 2011. Dynamics of mean-variance-skewness of cumulative crop yield impact temporal yield variance. International Journal of Agronomy. DOI: 10.1155/2011/426582.

Interpretive Summary: Cropping systems can influence the variation and stability of crop yields in long-term experiments. Production risks associated with cropping systems influence farmers’ decisions to adopt a new management practice or a production system; therefore, it is imperative to compare systems’ performance, and not only the effects of single factors within systems. Yield data from a long-term experiment was used to compare the performance of two distinct cropping systems. The experiment consisted of 16 combinations, two each of cropping system (conventional and organic), crop rotations (2-yr of corn-soybean and 4-yr of corn-soybean-wheat-alfalfa), tillage (conventional and strip), and fertility (with and without nitrogen) factors. Cropping systems differed significantly as to their yield distribution and the probability of producing a certain level of yield over years. The 16 factor combinations were clustered into three distinct groups. The first and third clusters represented subsystems of conventional and organic, respectively; whereas, the second was composed of a mixture of conventional- and organic-subsystems. This multi-level analysis provided statistical evidences indicating that differences in management among cropping systems contributed differently to the goal of achieving the yield potential of each cropping system. Results of the study may help farmers, agronomists, and crop consultants identify novel components of cropping systems that can be implemented on the farm with less external inputs and may result in reducing temporal variation of crop yields.

Technical Abstract: Production risk associated with cropping systems influences farmers’ decisions to adopt a new management practice or a production system. Cumulative yield (CY), temporal yield variance (TYV) and coefficient of variation (CV) were used to assess the risk associated with adopting combinations of new management practices in a long-term field experiment with conventional (CNV) and organic (ORG) cropping systems. The mean-variance-skewness (M-V-S) of CY proved to be a reliable indicator of yield stability in both cropping systems. Using partial least squares (PLS) regression, CY accounted for 71% of total variation of TYV in CNV, 45% in ORG, 62% in CNV-subsystems, 45% in ORG-subsystems, and 33% in mixed subsystems. Using CY to predict TYV eliminated the confounding effects of the differential magnitude of the means caused by differences among crops, years, management factors and their interactions; however, it resulted in different validation PLS models fit (Q2) and residual means squares error (RMSE) values. The RMSE in the validation PLS models provided quantitative estimates of the deviation from measured TYV; whereas, the multi-level statistical analyses helped identify stable and productive subsystems within each of the CNV and ORG cropping systems.