We predict that modern pathogen resistance alleles may trade off with the maintenance of symbiotic function because plant responses to pathogens and beneficial microbes proceed via similar molecular pathways, as explored in Box 4.Plants could adapt to avoid some of the trade-offs we describe. Plants could evolve more sensitive adaptive plasticity to express robust symbiosis traits if symbiotic resources are in short supply and to reduce investment in symbiosis if the rewards of symbiosis are freely available in the environment . For example, in the model symbiosis between legumes and rhizobia, plants adjust root nodule formation and nitrogen fixation. In autoregulation of nodulation, rhizobial infection activates the transcriptional regulator, NIN, triggering systemic long-range signaling between roots and shoots to inhibit further nodule formation. Furthermore, high soil nitrate levels activate a related NIN-Like Protein, NLP, which triggers the same long-range signaling to inhibit nodule formation and nitrogen fixation.Plant species differ with respect to the plasticity of symbiosis traits. Negative genetic correlations between the ability to benefit from symbiosis in low-resource environments and the ability to divest from symbiosis under high-resource conditions could hinder the evolution of adaptively plastic symbiosis in crops, but this remains to be tested.The small effective population sizes and low diversity of domesticated plant populations can result in stochastic increases in the number and frequency of deleterious genetic variants. Deleterious genetic variants can accumulate stochastically, despite selection against them if: the small effective population sizes common in domesticated plants reduce the efficacy of selection relative to genetic drift, allowing for fixation of mildly deleterious mutations ; strong artificial selection results in ‘genetic draft’, whereby deleterious mutations hitchhike to high frequency because they are linked to genes that fix under strong artificial selection; domesticated plants experience ‘expansion load’, wherein deleterious mutations reach high frequencies via drift after a demographic bottleneck; inbreeding during domestication decreases the efficacy of recombination in breaking up linkage between beneficial and deleterious loci,macetas 30 litros reducing the chance of beneficial alleles moving into a genomic background with fewer linked deleterious alleles.
Disruption of symbiosis functions in crops can occur as a demographic consequence of domestication, despite selection against deleterious variants at symbiosis loci. While deleterious alleles are more common in crops than in their wild relatives, we lack tests of genetic costs for symbiosis traits. Research has begun to annotate and map symbiosis loci in plant genomes , yet the impact of genetic costs on symbiosis function remain unknown. Linkage disequilibrium tends to be more extensive in crops than in their wild relatives. Higher LD in inbred crops reduces the probability that deleterious symbiosis alleles are purged when linked to beneficial alleles and increases the probability that beneficial symbiosis alleles linked to strongly deleterious alleles will be lost. We predict that deleterious symbiosis alleles will be enriched near loci of agronomic importance that have been subject to selective sweeps, and in regions of low effective recombination rate, as occurs for deleterious variants in soybean and sunflower, although there is mixed evidence for maize. If some deleterious symbiosis alleles have hitchhiked with a linked beneficial allele, as occurs for one in ten deleterious alleles in maize, these deleterious symbiosis alleles will exhibit signals of positive selection.For example, a deleterious allele linked to artificially selected semidwarfed stature in rice resulted in severe drought sensitivity in modern dwarfed rice cultivars. Structural variants, such as deletions, insertions, duplications, inversions, and translocations, have a critical role in domestication evolution, yet their impacts on symbiotic function in crops are unexplored. The extent to which the stochastic forces we highlight disrupt symbiosis traits remains a frontier in plant science. Since these forces arestronger in populations with smaller effective population sizes, their negative impacts on symbiosis function are likely exacerbated in crops compared with wild crop relatives.The Selection Relaxation Hypothesis predicts that some traits critical for plant fitness in the wild experience relaxed selection in agriculture.
Traits that do not contribute to success in agriculture will stocastically accumulate deleterious genetic variants because these mutations are not removed by artificial selection . In contrast to genetic costs, relaxed selection results when artificial selection on a trait is weaker than natural selection in the wild, irrespective of demography. As neutral traits, alleles that disrupt symbiotic function will accumulate at a rate approaching the mutation rate because artificial selection no longer purges them, and they increase in frequency via drift and hitchhiking. Selection relaxation can be tested experimentally. Under selection relaxation, robust symbiosis traits confer no fitness advantage over disrupted symbiosis traits in agricultural conditions, yet robust symbiosis traits are advantageous under less luxuriant conditions. This predicts indistinguishable genotypic selection gradients for symbiosis traits among plant genotypes that differ in symbiotic function under agricultural conditions, but are under selection to maintain robust symbiosis traits elsewhere. If artificial selection is relaxed on symbiosis traits, but remains strong on other traits in domesticated plants, we predict that domesticated, but not wild progenitor, plants will exhibit an elevated ratio of nonsynonymous to synonymous substitutions in symbiosis genes, and that nonsynonymous substitutions will reach higher frequencies and induce more radical amino acid changes to symbiosis loci, than in other functional regions of the genome. The selection relaxation hypothesis predicts that a lack of artificial selection on symbiosis traits in the fertile, high-nutrient, low-stress environment of agriculture results in the disruption of symbiosis function in crops. Consistent with this hypothesis, diverse crop taxa have evolved a reduced ability to associate with mycorrhizae under high levels of phosphorus fertilization, which is assumed to relax selection on phosphorous uptake. Agricultural soils can harbor low densities of microbial symbionts due to the disruptive impacts of tillage, rotational planting, chemical inputs, or crop rotation patterns, which could further relax selection on symbiosis traits in plants . These examples are consistent with, but do not directly test, the selection relaxation hypothesis. Traits that no longer confer a benefit under intensive agriculture could include the ability to preferentially acquire nutrients from superior symbionts because resources may be freely available from fertilizers, or the ability to enhance defense from antagonists based on microbial symbiosis, because pesticides may reduce selection due to these pests. The selection relaxation hypothesis predicts that germplasm bred under more intensive agricultural conditions will exhibit greater symbiosis trait disruption than germplasm bred under less intensively managed conditions.
We note that not all traits decay under relaxed selection: symbiosis traits might not decay during domestication if new or secondary functions result in high yield, if symbiosis traits are positively genetically correlated with traits that result in high yield, or if crops retain gene flow with wild populations where the trait is under selection for function.Identifying pathways of symbiosis disruption will inform strategies to maximize the benefits of symbiosis in crops. Under the genetic costs and relaxed selection hypotheses, symbiosis traits are degraded stochastically and there is potential to increase yield or symbiotic function by the introgression of desirable symbiosis loci from wild or related lineages into crops, similar to the introgression of desirable disease resistance loci from wild congeners into crops. Genomic regions with fewer deleterious variants have introgressed into maize from wild populations, and resistance to a fungal pathogen in wheat appears to result from introgression with wild populations. Symbiosis function could be improved in a similar manner. Hybrid production of seed by crossing distinct breeding pools could allow complementation of deleterious variants if different breeding pools have deleterious alleles at different loci. In soybean and barley , ~40% of deleterious variants are private to individual cultivars and could be purged from breeding populations. In regions with low effective recombination, it is difficult for breeding programs to purge deleterious alleles,macetas 5 litros so targeted introgression, hybridization, and gene editing could be useful in these regions. These approaches could precipitate the development of more symbiotically robust crop cultivars better able to thrive under less resource intensive methods of sustainable agriculture. Symbiosis traits that degrade because of trade-offs are more difficult to overcome because they suggest that the genetic variants most strongly favored in modern agriculture are in conflict with robust symbiosis. Characterizing trade-offs between symbiosis traits and agriculturally significant traits requires measuring yield traits and symbiosis traits under selection and finding negative trait covariances. The genetic basis for these trade-offs can reflect either fundamental biophysical constraints, or genetic linkage between loci under opposing selection regimes. Evolutionary trade-offs due to fundamental constraints are difficult to overcome because possible paths to minimize trade-offs across the adaptive landscape have likely been unsuccessfully tested by selection over evolutionary time. Due to these constraints, variants that increase plant investment in symbiosis will be detrimental to overall agricultural performance. However, trade-offs that are driven by genetic linkage can be alleviated by breeding programs that select for recombination events that break up such linkage. Trade-offs can also be alleviated by gene duplication, or by tissue-specific or ontogenetic stage-specific gene regulation. Finally, trade-offs between yield and symbiosis function could be accentuated only under specific agricultural conditions; under altered conditions, such trade-offs might be reduced. The three genetic scenarios we highlight would motivate contrasting crop breeding strategies. Under relaxed selection and genetic costs models, symbiosis traits are disrupted stochastically , so restoring these traits to all varietal backgrounds would be beneficial. However, under the trade-off model, there would be costs to restoring symbiosis traits to elite varieties designed to perform in high input conditions, while there could be strong benefits for varieties grown in low input conditions. Given these dependent effects, targeted breeding programs for distinct high versus low input agricultural conditions would be beneficial.Efforts to maximize symbiosis benefits in crop plants have focused on the characteristics of microbial strains and consortia within symbiotic inocula.
However, optimizing the symbiosis traits of the crop plants themselves should have a bigger impact on improving symbiotic benefits in agriculture. To understand the plant symbiosis traits that evolved during domestication, we must understand how trade-offs, relaxed selection, and genetic costs in domesticated plants impact symbiotic regulation. This will be key to success in increasing the benefits of microbial symbiosis in crop germplasm and designing synthetic symbioses that hold great promise in modern agriculture. For example, understanding trade-offs incurred by symbiosis traits and any ways in which symbiosis traits are more robust in the wild relatives of crops could aid breeders in minimizing the costs of symbiosis in current efforts to develop a novel capacity for symbiotic nitrogen fixation in nonleguminous crops, such as maize. Research on symbiosis trait disruption in crops will also serve as a model system for understanding fundamental patterns and mechanisms of the evolution of reduced investment in microbial symbiosis by host organisms, which appears to be a common evolutionary pathway .A more complete characterization of symbiosis function in crops is needed to inform diverse aspects of agriculture . Few modifications to genetically engineered crops are tested for impacts on symbiosis, yet alterations of immune function or phytohormone production can impact symbiosis with beneficial microbes. It is critical to understand whether key symbiosis traits trade off with disease resistance, or whether variants with deleterious symbiotic effects are fixed in crops, to successfully design a robust symbiosis, potentially via genome editing. Other agricultural practices, such as grafting perennial crops onto the root stock of wild relatives to confer resistance to pathogens and abiotic stress, are rarely assessed for their impact on symbiosis. New studies should analyze whether more robust symbiosis could contribute to the benefits of such practices. Alleles or genotypes that improve symbiosis outcomes for crops should be targeted via marker assisted breeding programs. One goal would be to develop generalist cultivars, able to benefit from symbiosis with indigenous microbes in diverse locations. Nodule cysteine-rich peptides, which modulate plant benefits from rhizobia, could be fruitful targets for engineering legume–rhizobia partnerships with improved agricultural properties. Moreover, feral crop populations could harbor superior symbiosis alleles with low negative epistatic effects in the genomes of crops due to their close relatedness. Selection for symbiosis function might be stronger on feral crops than on crops in modern agriculture, but these populations can face similar genetic costs to domestication, such as reduced genetic diversity relative to wild progenitor species. In addition, weeds are unintentionally domesticated plants: studying whether weeds experience symbiosis trait disruption could compliment work in crops because, while weeds adapt to the same agricultural conditions as crops, they are not subject to the same targets of artificial selection.