6 .2 ǀ Breeding strategies
Iron toxicity in the field is rarely uniform across the larger fields needed to evaluate breeding populations and the severity of stress may vary from year to year. Where such environmental variation reduces the reliability of phenotyping, marker assisted selection rather than conventional phenotypic selection should be the breeding method of choice. Markers associated with aspects of Fe toxicity tolerance exist, identified through genome-wide associations studies (GWAS) (Matthus et al. 2015; Pawar et al. 2021; Melandri et al., 2021) or through bi-parental QTL mapping populations (Dufey et al. 2015; Rasheed et al., 2020 and references therein). To our knowledge none of these studies identified loci with large enough effects to be utilized in applied breeding.
Additionally, QTL mapping for Fe toxicity has been done in the lowland NERICA (NEw RICe for Africa) hybrids of O. glaberrima andO. sativa with the aim to identify markers linked to O. glaberrima -derived tolerance loci. Two varieties (NERICA-L19 and NERICA-L43) known to be highly tolerant of Fe toxicity (Dramé et al., 2011; Ndjiondjop et al., 2018) were used to develop mapping populations with recurrent parent IR64, but again, only minor QTLs were identified (Melandri et al., 2021).
Based on QTL mapping, GWAS and transcriptome analysis, it appears that tolerance of Fe toxicity in rice is a complex trait controlled by multiple genes distributed throughout the genome (Diop et al., 2020; Melandri et al., 2021; Wairich et al., 2021). In order to successfully identify markers of utility in rice breeding, it may be necessary to break down the complex tolerance response into component traits based on the tolerance mechanisms outlined above, and to develop tailored screening methods for such individual tolerance mechanisms.
It should be noted that these should take the complexities of below-ground plant-soil interactions into account rather than focusing predominantly on the visible Fe toxicity symptoms in leaves. Nozoe, Agbisit, Fukuta, Rodriguez & Yanagihara (2004) compared selections from the same breeding population made either in the field (for grain yield) or in nutrient solution (for leaf bronzing) and showed that the yield advantage under Fe toxicity of best lines selected in the field was 45% or more over the sensitive parent IR64, whereas best lines selected in nutrient solution provided no gain in grain yield in the field.