Stabilization of yield in plant genotype mixtures through compensation rather than complementation

Abstract

Background and aims: Plant genotypic mixtures have the potential to increase yield stability in variable, often unpredictable environments, yet knowledge of the specific mechanisms underlying enhanced yield stability remains limited. Field studies are constrained by environmental conditions which cannot be fully controlled and thus reproduced. A suitable model system would allow reproducible experiments on processes operating within crop genetic mixtures.

Methods: Phenotypically dissimilar genotypes of Arabidopsis thaliana were grown in monocultures and mixtures under high levels of competition for abiotic resources. Seed production, flowering time and rosette size were recorded.

Key results: Mixtures achieved high yield stability across environments through compensatory interactions. Compensation was greatest when plants were under high levels of heat and nutrient stress. Competitive ability and mixture performance were predictable from above-ground phenotypic traits even though below-ground competition appeared to be more intense.

Conclusions: This study indicates that the mixing ability of plant genotypes can be predicted from their phenotypes expressed in a range of relevant environments, and implies that a phenotypic screen of genotypes could improve the selection of suitable components of genotypic mixtures in agriculture intended to be resilient to environmental stress.

Keywords: Arabidopsis thaliana; compensation; experimental ecology; genotype mixtures; model-to-crop translational research; plant competition; resistance to environmental stress; variety mixtures; yield stability.

Fig. 1.

Mean seed yields of arabidopsis genotypes grown in mixtures and in monoculture in the four-way mixture experiments conducted during the autumn and winter seasons, and the summer season. n = 1880. Error bars = 95 % confidence interval.

Fig. 2.

Mean seed production per plant of each genotype grown in monoculture or mixtures for each four-way mixture experiment. n = 1880. Error bars = 95 % confidence interval.

Fig. 3.

Relative yield (mixture yield/monoculture yield) for each arabidopsis genotype under high and low nutrient treatment in a four-way mixture experiment conducted during (A) the autumn and winter (n = 1260) and (B) the summer (n = 620). (C) Relative yields for eight genotypes in the pair-wise interaction experiment. Competitive groups of genotypes increase from left to right on the graph. n = 639. Error bars = 95 % confidence interval.

Fig. 4.

(A) Standard deviation of the mean seed mass produced per tray (block) in a four-way mixture experiment. (B) Mean plant yield in genotypic monoculture and the four-way mixture averaged over the entire experiment. n = 1880. Error bars = 95 % confidence interval.

Fig. 5.

Mean seed production of focal arabidopsis plants from four competitive groups (1 = least competitive, 4 = most competitive) under three competition treatments (above-ground competition only, above- and below-ground competition, single plant) in a pair-wise interaction experiment. n = 639. Error bars = 95 % confidence interval.

Fig. 6.

Mean seed production of focal arabidopsis plants grown with plants of four competitive groups (1 = least competitive, 4 = most competitive) in a pair-wise interaction experiment, (A) when competition was restricted to above-ground only, (B) when competition occurred both above- and below-ground. For competitive groups, see Table 1. n = 639. Error bars = 95 % confidence interval.