Introduction
When carefully combined, crop mixtures such as intercrops and variety mixtures are found to enhance agronomic performance and ecosystem services such as improved resource use efficiency, pest and pathogen management and yield stability. These benefits rely on a range of physiological, morphological and phenological characteristics of the component partners in interaction with the growing environment and management. For some of these traits, similarity among mixing components is desirable (mainly traits conferring good performance under agronomic management, such as early establishment, simultaneous ripening and high yield), whereas complementarity in other traits may confer advantages such as enhanced resource uptake and improved weed suppression. In addition, some component-specific traits enable mixing partners to facilitate the growth and performance of others, including structural support against lodging, switching of nutrient sources, and weed suppression. While most traits in the former category were already selected in traditional sole crop breeding, traits in the other two categories are exclusive to crop mixtures. Hence, ideotypes for sole cropping may not perform well in crop mixtures, and the need for developing specific genotypes of crops for intercropping has long been recognized (e.g. Finlay, 1976).
2 Workshop format
The cross-disciplinary workshop will feature flash talks from representatives of breeding, ecology and agronomy. Facilitated discussions will address key challenges for the practical implementation of modern breeding programs for crop mixtures.
3 Discussion topics
3.1 Identifying relevant breeding targets
Some plant traits that optimise agronomic performance and environmental benefits of crop mixtures have been identified and more are under investigation. Generally, useful traits for components of a crop mixture are those that optimize complementarity and facilitation (Costanzo & Barberi, 2014). To be discussed: Is the performance of all components equally important –should crop improvement focus more on specific components? Is it useful to identify mixture-ideotypes for specific crop mixtures?
3.2 Identifying operational breeding methods
A number of methods for crop mixtures breeding have been applied. Each method having a number of pending questions and uncertainties.
(i) Statistical evaluation of genotypic performance in (a range of) mixtures (e.g. Federer, 2012). This includes several attempts to estimate the general (and specific) mixing ability of genotypes, for example using diallel and nested mixing designs. To be discussed: Are these designs sufficiently efficient in the search for general mixing ability? Is it realistic to generalize mixing ability beyond the set of tested combinations? Can genetic markers associated with general mixing ability be identified?
(ii) Evolutionary plant breeding (Döring et al., 2011) and within-mixture breeding. Selection for ‘mixable’ genotypes may be more efficient in mixtures than in sole crops (Harper, 1967; Finlay, 1975; Zimmermann et al., 1984; Zuppinger-Dingley et al., 2014). To be discussed: What could feasible and efficient programmes for within-mixture breeding look like? E.g. is it feasible to carry out selection within mixtures with representatives (testers) of a wider companion gene pool (cf. Davis & Woolley, 1993)? At which stages of selection - early (e.g. Barot et al., 2017)?, How important is the effect of neighbour phenotype relative to the effect of not growing with conspecifics? How to avoid selecting merely for competitive ability?
(iii) A trait-based approach, selecting for specific plant functional traits that improve performance in species mixtures (Brooker et al., 2015). Recent research has focused on the potential for devising mixture ideotypes, assembly rules and mixture breeding design (e. g. Gaba et al., 2015; Litrico and Violle, 2015). Being inspired partly by the findings of the widely successful large-scale biodiversity experiments (e. g. Scherber et al., 2010), these trait-based approaches aim to embrace plant community complexity at levels beyond the assignment of general genotypic traits for mixing ability and competitiveness. To be discussed: Can we identify clearly what traits are more important? Should traits and outcomes not typically measured in sole crop breeding programmes be included?
3.3 Managing genotype-environment interactions
Correlation of performance under mixed and sole crop conditions is often low (Francis, 1986), suggesting that trait information from sole crop trials is not always useful. The phenotypic plasticity that enables plants to adapt to a range of growing conditions also shapes their ecological niche in the community (Berg and Ellers, 2010) and ultimately governs the yield stability of agricultural crops (Lazzaro et al., 2017). Significant levels of crop trait plasticity are reported on a regular basis, also in response to mixing (Zuh et al., 2016). (G × G) × E × M interactions in crop mixtures therefore poses an additional level of complexity. To be discussed: Assuming that plasticity levels differ among traits, species, ‘neighbours’ and environments, which traits categories are more plastic in modern temperate crop varieties and hence more variable? Does trait plasticity tend to reinforce or reduce niche overlap in crop mixtures – and to what extent does this influence crop mixture synergy and the prediction of optimal combinations?
3.4 Market opportunities and challenges
Carefully combined crop mixtures and ‘mixture-ready’ cultivars with specific trait combinations could present a novel market opportunity for breeders and seed sellers. To be discussed: Is this attractive for actors along the value chain? Is convergence possible, or is there inherent conflict between rapid short-term gains from sole crop breeding and breeding for crop mixtures, which may need more long-term efforts?