Morphological diversity in a barley composite cross-derived population evolved under low-input conditions and its relationship with molecular diversity: indications for breeding

Abstract

SUMMARYIn order to reduce the environmental impact of agriculture and improve the resilience and sustainability of our food systems, there is an increasing interest in shifting from the present agricultural systems, which are characterized by high external inputs, to low-input productive systems characterized by high resilience and sustainability. Purposely developed varieties are needed for the latter. With the rapid disappearance of landraces, heterogeneous populations such as composite cross populations (CCPs) or line mixtures, developed through evolutionary plant breeding, could be the ideal source of breeding material for the development of new cultivars for low-input productive systems. Parental lines of CCPs should be selected among old breeding lines, varieties or landraces because the specific characteristics required for low-input or organic farming systems might have been lost during selection of modern varieties. In the current scenario of renewed interest in evolutionary plant breeding, the evolution of diversity in heterogeneous populations needs to be better investigated to maximize the advantages that can be obtained by their utilization.The present paper reports on the analysis of 88 barley plants chosen randomly from a CCP, namely AUT DBA (where AUT indicates autumn sowing and DBA is the acronym of the former Department) that was multiplied for 13 years under a low-input management system without any conscious human selection, aiming to investigate the morphological diversity still existing in the population and its potential value as source of breeding material for low-input/organic agriculture and understanding the traits that contributed to the adaptive success of certain groups of individuals.Eighteen plant and spike morphological traits were analysed using bi-dimensional spatial analysis, cluster analysis, non-parametric tests and multivariate approaches. Low lodging and loose smut damage were observed in the CCP where several individuals were superior to the best control for at least one of the four yield-related traits, namely spike weight, number of seeds per spike, weight of seeds in a spike and grain weight. Three morphological clusters were identified using cluster analysis. Clusters 2 and 3 grouped the largest number of CCP individuals which, compared with those in cluster 1, were characterized mainly by heavier spikes with higher seed number, taller culms and early flowering. Interestingly, the plant architecture of all the controls was different from that of the most frequent genotypes in the CCP, showing that low-input systems may require a plant architecture different from the one usually considered as the most suitable for high-input systems. Taking advantage of results from Raggi et al. (2015), phenotypic data were also analysed according to individual genetic group assignment. Results suggest that plant height at the beginning of stem elongation, and days to heading, together with traits related to culm and leaf morphology, could have played a significant role in determining the success of plants from genetic group D, which is the group most represented in the CCP.According to the wide range of morphological diversity existing in the AUT DBA and the high percentage of lines that show favourable combinations of different traits, this population could be a useful gene-pool from which to select lines for breeding activities. Even though further use of the CCP for breeding purposes may be limited by its possible evolution, there are different ways of manipulating the CCP to counteract the undesirable changes without great economic and/or technical efforts. The high number of multi-locus genotypes and the evolutionary responses observed in AUT DBA show that the prediction that phenotypic micro-evolution in natural systems may be limited by low genetic variances in harsh environments and low selection pressure in good environments is not necessarily true for low-input systems.