Back

Yield losses associated with peanut smut incidence in Argentina: a quantitative synthesis across field studies

Cazon, L. I.; Gonzalez, N. R.; Del Ponte, E. M.; Costa de Carvalho, A. C.; Asinari, F.; Camiletti, B. X.; Paredes, J. A.

2026-08-11 plant biology
10.64898/2026.08.11.744131 bioRxiv
Show abstract

Peanut smut, caused by Thecaphora frezzii, is an important constraint to peanut production in Argentina, but quantitative estimates of yield losses across environments remain limited. We quantified the relationship between disease incidence and kernel yield using 922 observations from 26 field studies conducted in Cordoba, Argentina, between 2021 and 2025. Study-specific incidence-yield relationships were analyzed using linear regression, random-effects meta-analysis, and linear mixed-effects models. Peanut smut incidence was consistently associated with yield reduction across studies. The estimated damage coefficient ranged from 24.2 to 28.7 kg ha-{superscript 1} per 1% increase in disease incidence, corresponding to a relative yield reduction of 0.74-0.87% of attainable yield. In contrast, attainable yield varied markedly among studies, ranging from 1,370 to 5,409 kg ha-{superscript 1}. Although an exploratory segmented analysis suggested a breakpoint near 12% incidence, subsequent moderator analyses, study- specific regressions, and normalized response curves provided no evidence of a biologically meaningful change in the damage coefficient across incidence or yield classes. These results indicate that differences among environments were primarily associated with attainable yield rather than with changes in the magnitude of disease-associated yield loss. The resulting damage function provides a quantitative basis for yield-loss assessment and disease management in peanut.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.