Single-gene resolution of diversity-driven overyielding in plant genotype mixtures

The study addresses why and how plant diversity increases community productivity (overyielding), focusing on identifying causal drivers of positive biodiversity–ecosystem functioning (BEF) relationships. Traditional trait-based approaches infer niche complementarity from phenotypic trait differences but face challenges including trait covariation, uncertain causality, and potentially many small, interacting trait differences. Positive BEF effects occur both among species and within species; mixtures of genotypes can overyield relative to monocultures. The authors propose a gene-centered strategy in Arabidopsis thaliana to dissect the genetic basis of overyielding, leveraging crosses and recombinant lines to manipulate allelic combinations and establish causal links from genetic differences to community-level outcomes. They hypothesize that allelic variation underlying niche complementarity—potentially shaped by adaptation to soil chemistry gradients—can be resolved to specific loci and genes.

Prior work shows that plant diversity often enhances productivity across ecosystems (grasslands, forests, croplands). Trait-based ecology links diversity to functional trait variation and niche complementarity, yet identifying which traits drive BEF remains difficult due to trait trade-offs, environmental correlations, and multidimensional niche differences. Intraspecific genetic diversity can also increase productivity, though results vary. Recent community genetics studies demonstrate that genetic variation can scale up to influence community and ecosystem properties, including cases where single genes have disproportionately large ecological effects (e.g., keystone genes). Arabidopsis root-growth regulator BRX has documented adaptive variation along soil pH gradients, influencing root architecture and competition, suggesting edaphic adaptation as a potential axis for niche complementarity.

Overview: The authors combined (i) an initial screen for overyielding among Arabidopsis accession pairs, (ii) a QTL mapping experiment using a competition half-diallel of recombinant inbred lines (RILs) from Sav-0 × Uk-1, (iii) an association study re-analysis linking mixture performance to SNP differences within the QTL interval, and (iv) functional assays testing protein activity and root growth responses across substrate pH. Screen for overyielding: Ten accession pairs (nine with available RIL populations plus one additional pair) were grown as monocultures and two-genotype mixtures across multiple soils and pot sizes, in replicated blocks. Mixture overyielding was calculated as the deviation from mean monoculture yields; a meta-analysis across experiments assessed significance. Mixtures of Sav-0 and Uk-1 consistently showed higher biomass than expected. QTL mapping via competition half-diallel: A half-diallel included 20 genotypes (18 Sav-0 × Uk-1 RILs plus the two parents), combining all 190 pairwise mixtures and 20 monocultures. The design was replicated across four temporal blocks (total 920 sown pots; 808 analyzed after exclusions) and two substrates (sand-rich vs peaty mixes). Aboveground dry biomass per pot was measured. General combining ability (GCA) and specific combining ability (SCA) were estimated; SCA was standardized by mean community biomass per substrate to account for differences and ensure normality. Whole-genome resequencing generated high-density genotype maps for the 18 RILs. Marker regression tested whether SCA variation associated with allelic diversity (bi-allelic vs mono-allelic) at genomic regions. A LOD threshold of 3 (-log10 p-value) was used based on simulations. Additive partitioning: Using least-squares means adjusted for block effects, mixture overyielding was partitioned into selection effects (SE) and complementarity effects (CE) following Loreau and Hector, and contrasted between bi-allelic and mono-allelic mixtures at the mapped QTL. Association study re-analysis: A previously published factorial competition experiment (10 testers: Sav-0, Uk-1, Col-0, Sf-2, St-0, C24, Sha, Bay-0, Ler-1, Cvi-0; × 98 accessions) with two blocks and all monocultures was re-analyzed. SCAs were computed after adjusting for a quadratic relation of pot biomass with mean GCA. Marker regressions tested SNP-by-diversity associations within the chromosome 2 QTL interval; multiple testing was controlled (Bonferroni). Functional assays: AtSUC8 coding sequences from Sav-0, Uk-1 (and reference Col-0) were Sanger-sequenced to identify nonsynonymous polymorphisms. SUC8 variants were expressed in Xenopus laevis oocytes; 14C-sucrose uptake was measured after 1 h to assess transporter activity. Root plasticity assays used 80 Sav-0 × Uk-1 RILs grown on agarose MS media at pH ≈4.8 vs ≈6.8 with 1% sucrose; primary root length was measured at days 4 and 7. Effects of alleles at BRX and AtSUC8 on log2-fold root length difference (pH 4.8 vs 6.8) were analyzed by ANOVA. Growth conditions, soil mixtures, sterilization, genotyping (CAPS marker for AtSUC8 EcoRV site), genome alignments (BWA), variant calling (SAMtools), and data/statistical pipelines (glht contrasts; t-tests; ANOVA) followed detailed Methods.

- Sav-0 × Uk-1 mixtures overyielded consistently across conditions: average +5.6% biomass (range 0–12%) vs expectation; meta-analysis P = 0.012. - In the half-diallel (190 mixtures), average overyielding was 2.8% (two-sided t-test, t189 = 4.74, P < 0.001). - A single major-effect QTL on chromosome 2 explained increased SCA when mixtures were bi-allelic at this locus. Bi-allelic mixtures had SCA higher by 2.8% (±0.8% s.e.m.) than mono-allelic mixtures (contrast t187 = 3.53, P < 0.001). The QTL interval was narrowed to ~178 kb. - Additive partitioning showed overyielding driven primarily by complementarity effects (CE): CE increased in bi-allelic vs mono-allelic mixtures (t187 = 2.57, P = 0.01); no evidence for selection effects (SE) (t187 = 0.201, P = 0.84). Average CE was 24.7 mg (~3.1% of mean community biomass). - Association mapping within the QTL interval identified a single SNP in the AtSUC8 coding region significantly associated with positive diversity effects (t947 = 4.1; P = 5 × 10−5; Bonferroni-adjusted P = 0.007; standardized effect size = 3.2%). - Sequence variation: Sav-0 AtSUC8 carries 3 amino acid changes (one non-conservative); Uk-1 carries 11 changes (six non-conservative), including K320T and R472G. - Functional validation: Xenopus oocyte assays showed SUC8Sav-0 conferred efficient sucrose uptake, whereas SUC8Uk-1 had significantly lower import after 1 h compared to Sav-0 and above mock. - Root plasticity: Genotypes with AtSUC8Uk-1 exhibited significantly smaller root length reduction at low pH (pH 4.8 vs 6.8) than those with AtSUC8Sav-0 (ANOVA F1,74 = 5.8; P = 0.02). BRX allele did not affect relative sensitivity to low pH in this assay. - BRX genetic variation, despite known adaptive significance on acidic soils, was not associated with mixture overyielding in this experiment.

The study demonstrates that diversity-driven overyielding in Arabidopsis genotype mixtures can be largely explained by allelic diversity at a single locus, with AtSUC8 emerging as the key candidate gene. The mapping and association evidence, combined with functional assays, indicate that allelic differences in a proton–sucrose symporter expressed in root tissues alter transporter activity and modulate root growth sensitivity to substrate acidity. These differences likely promote niche complementarity belowground, either via differential exploitation of spatial pH heterogeneity or altered root foraging and reduced intergenotypic root competition. The results show that community-level properties can have relatively simple genetic underpinnings, suggesting a productive strategy: identify causal genetic differences first, then infer the responsible functional traits. This gene-centered approach complements trait-based ecology, helps link ecological outcomes to evolutionary drivers (e.g., local adaptation along edaphic gradients), and offers a framework for designing productive and stable crop variety mixtures.

A gene-centric dissection of BEF in Arabidopsis mixtures resolved a substantial portion of overyielding to a single QTL on chromosome 2, implicating AtSUC8. Allelic diversity at this locus increases mixture performance predominantly through complementarity, likely mediated by pH-dependent root physiological differences. Functional assays confirmed allele-specific transporter activity and root growth responses. The work establishes that complex mixture effects can stem from simple, identifiable genetic causes and advocates for integrating genetic mapping with ecological experiments to uncover mechanisms of overyielding. Future research should pinpoint the causal polymorphism(s) at AtSUC8, delineate the mechanistic links between transporter function, root system architecture, and resource partitioning, test generality across environments and species, and translate these insights to design high-performance, diverse crop mixtures under field conditions.

- Causality at the nucleotide level is not proven; the associated SNP in AtSUC8 may be in tight linkage disequilibrium with the true causal variant. - Experiments were conducted in controlled greenhouse/pot conditions, which may not capture field heterogeneity and biotic interactions; generalizability requires field validation. - The QTL mapping used a subset of 18 RILs plus parents; while high mapping resolution was achieved, sampling of recombination events and genotypic space was limited. - Heterogeneous inbred families at AtSUC8 could not be isolated, limiting fine-mapping via residual heterozygosity. - Overlapping adaptive loci (e.g., BRX) did not show effects on overyielding here; reasons for context dependence remain unresolved. - Root function–overyielding linkage is inferred from assays and associations; direct demonstration of soil pH heterogeneity driving niche partitioning in mixtures was not performed.