Identification of resistance to ramulosis caused by Colletotrichum gossypii var. cephalosporioides in cotton advanced breeding lines and monitoring of ramulosis disease by RGB-image analysis

Abstract

Cotton growing regions in South America are affected by Colletotrichum gossypii var. cephalosporioides (Cgc). The most severe epidemics provokes considerable yield reductions linked to meristem necrosis, oversprouting, excessive branching and stunting (Figure 1). The Sinu Valley is a major cotton producer in Colombia and is heavily affected by this disease. Rainfall was identified as the main driver of ramulosis development in the Sinu Valley prevailing in crops planted at the beginning of the main rainy season (Figure 2). Fifty five advanced breeding lines (ABLs) were assessed by ramulosis field resistance. Nine ABLs exhibited high levels of partial resistance (< 10% of plants exhibiting oversprouting). With the aim to optimize disease assessing accuracy and breeding efforts for ramulosis resistance, we had evaluated the use of red-green-blue (RGB) based indices for automated assessment of ramulosis disease. Eleven cultivars exhibiting contrasting level of ramulosis resistance were grown and photographed at different phenological stages. RGB indices extracted by Breedpix software from these plot images were compared with visual assessment of plant disease severity. The RGB indices Hue, Saturation, b, and v measured ten weeks after planting (boll opening) were correlated with accumulated disease severity and oversprouting (estimated as the area under the disease progress stairs). Oversprouting exhibited the higher correlation coefficients (r = 0.60, -0.65, -0.65, -0.60 and 0.54, P < 0.001). Additionally, destructive sampling across phenological development showed that green area (GA) has a positive correlation with total fresh biomass, leaf area index, leaf fresh biomass and green cover (GC) (r = 0.65, 0.60, 0.70 and 0.95, P < 0.001). RGB-based indices are appropriate predictors of cotton growth and ramulosis severity and a cost effective tool for cotton phenotyping based on automation of RGB-images assessment and affordable cost of RGB-cameras Más...