Plant breeding has advanced through genomics and predictive models, yet the plant-associated microbiome remains largely excluded from plant breeding trials. Microbial communities strongly influence nutrient acquisition, stress tolerance, and disease resistance, shaping key agronomic traits. We argue that neglecting microbiome variation biases heritability and G×E estimates, constraining genetic gains. Integrating microbiome information into breeding trials offers a feasible, scalable path toward more predictive, resilient, and sustainable crop improvement.The microbiome as a missing component of plant breeding trialsPlant breeding has been developed rapidly over the past decades, supported by advances in genomics, phenomics, and increasingly sophisticated statistical tools. These innovations have substantially improved our ability to quantify genetic variation, dissect complex traits, and accelerate selection through predictive models. Yet despite these advances, the plant-associated microbiome remains largely neglected by plant breeding pipelines1. It is well-known that microbes stimulate plant growth, increase nutrient-use efficiency, and increase plant resilience2. The plant–microbe interactions influence all physiological processes relevant to agricultural productivity, from nutrient acquisition to stress resilience and disease suppression. Thus, the exclusion of microbiome information in plant breeding is becoming a significant blind spot that risks constraining genetic gains. Here, we argue that the time has come for plant breeders to systematically incorporate the microbiome into breeding trials (Fig. 1). This perspective is consistent with recent comprehensive syntheses highlighting that microbiome effects on plant performance are already well established, while their systematic integration into breeding frameworks remains limited3. Building on established literature on plant–microbiome interactions1,4, this perspective proposes microbiome profiling as a comparative and explanatory tool.Fig. 1: Conceptual comparison between classical plant breeding and microbiome-guided breeding.The alternative text for this image may have been generated using AI.Full size imageIn classical breeding, plant phenotype is determined by the interaction between genotype (G), environment (E), and their interaction (G × E). In microbiome-guided breeding, the rhizosphere microbiome is explicitly integrated as an additional driver of plant phenotype.Microbial variation, heritability, and genotype × environment interactionsIn general, plant breeding programs apply multi-environment trials to quantify genotype performance and stability, i.e., the interactions genotype × environment (G × E) that determine phenotypic stability and adaptability across diverse conditions5. These multi-environment trials are designed to capture environmental heterogeneity, management differences, and soil physicochemical variation, all of which are recognized sources of variance that influence trait expression. Yet microbial variation, despite its well-known effects on plant growth and function, remains unmeasured, being considered an uncontrolled factor. This is paradoxical because root-associated microbial communities regulate key biological processes, including root development, nutrient uptake, hormone balance, and pathogen defense2. They are not an uncontrolled factor but actively shape the expression of many traits that breeders select, such as biomass accumulation, grain yield, nitrogen-use efficiency, and stress tolerance. Neglecting the microbiome can misestimate heritability, misinterpret G × E interactions, and discard genotypes that perform well only under supportive microbiomes.Treating microbial variation as an uncontrolled factor has several consequences. First, it may lead to biased estimates of heritability if part of the phenotypic variance attributed to the environment is, in fact, microbiome-driven. Second, failing to account for the microbiome can distort G × E estimates, obscuring true genetic signals or inflating apparent environmental effects. Third, breeders may unintentionally discard genotypes that perform well only when paired with compatible or beneficial microbial communities. Conversely, some genotypes may be selected because they perform well in microbially favorable environments but fail to express their potential in microbiome-poor soils. In all these cases, plant genetic gain is limited not by plant biology per se but by a lack of information about the microbial partners that modulate plant performance.Root traits as heritable drivers of microbiome assemblyWe argue that one reason for including the microbiome in plant breeding is that plants differ in their ability to recruit beneficial microbes6, a process mediated by heritable traits such as root architecture (for instance, root hair number and tip abundance) and root exudation patterns5. These traits create distinct rhizosphere chemical and physical environments, which in turn drive microbial assembly. Root exudates, including sugars, amino acids, organic acids, flavonoids, and phytohormones, act as selective filters that enrich specific microbial taxa. Root architectural traits, such as lateral root density, root hair length, or root tip abundance, change oxygen gradients, moisture distribution, and habitat complexity, all of which influence microbial colonization. Because many of these root traits are genetically encoded and heritable6, the ability to recruit and maintain beneficial microbiomes is itself heritable. Indeed, recent studies have shown consistent genotype-specific microbial signatures, providing evidence of microbiome heritability across breeding generations7. Recent large-scale and multi-environment analyses further demonstrate that these genotype-associated microbiome signatures are strongly modulated by environmental context and temporal dynamics, reinforcing the need for cautious interpretation in breeding applications8.Microbiome heritability and the extended plant phenotypeSince the microbiome is heritable, this should be considered part of the extended plant phenotype. Indeed, a recent study suggests that the plant microbiome is, to some extent, heritable, meaning that microbial effects should be considered part of the plant’s extended phenotype9. This supports the proposition by Mueller and Linksvayer10, who introduced the term “microbiability,” indicating H² and h² for broad- and narrow-sense genetic heritability and using mH² and mh² for microbiome-related contributions. To avoid conflating genetic and microbial variance, Week et al.9 referred to these microbial effects as “microbial transmissibility,” which captures microbial contributions to host phenotypes. This recognition of microbial transmissibility supports evolutionary frameworks treating hosts and their microbiomes as integrated units shaped by interacting genetic and ecological processes.From microbiome profiling to practical breeding applicationsRecent studies have emphasized that amplicon-based approaches are subject to methodological biases and limited predictive power in complex breeding scenarios, particularly across environments3,8. In this context, combining broad, trait-based selection strategies (e.g., root architectural traits influencing microbial recruitment) with more targeted, function- or taxa-focused microbiome approaches allows breeders to balance generalizable, high-throughput screening with crop- and trait-specific microbiome optimization. It is known that microbiome composition is strongly influenced by spatiotemporal dynamics, environmental context, and host–environment interactions, which can constrain the predictive value of single-time-point microbiome surveys.Importantly, microbiome-informed breeding does not imply a single universal strategy applicable to all crops. Broad approaches, such as selecting for root architectural traits that influence microbial recruitment, offer a scalable and high-throughput entry point, particularly for large breeding populations. Similar principles have been discussed in previous syntheses and empirical studies emphasizing host genetic control of microbiome assembly and the use of plant traits to steer microbial recruitment3,11. However, more targeted strategies may be better suited for specific crops or agronomic goals, particularly when specific microbial functions or taxa are known to enhance key traits. For example, targeted recruitment or management of nitrogen-fixing or phosphorus-solubilizing bacteria can improve nutrient-use efficiency12, while selecting genotypes that preferentially associate with disease-suppressive microbial consortia can enhance resistance to soil-borne pathogens13. In addition, enrichment of drought-associated microbial guilds may contribute to improved stress tolerance under arid conditions14. Likewise, microbiome profiling using amplicon sequencing can provide different layers of information, ranging from the relative abundance of individual microbial taxa and functional groups to proportional shifts among dominant lineages and changes in the core community structure.This includes a practical dimension, where microbiome-informed breeding could accelerate the development of microbial consortia with high field efficacy. Despite the proliferation of microbial-based agricultural inputs, inconsistent performance remains a major bottleneck. This inconsistency has been widely reported in the literature and reflects, in part, methodological constraints as well as ecological variability in plant–microbiome associations. One reason is that microbial inoculants interact differently with different plant genotypes. Incorporating microbiome profiling into breeding trials would allow breeders to identify genotypes that consistently benefit from specific microbial taxa or consortia15. Here, microbiome profiling is proposed as a comparative and explanatory tool rather than an idealized or bias-free method. This would facilitate the design of plant–microbe combinations optimized for specific environments, moving us toward a systems-level approach where plant genetics and microbial inputs are co-optimized rather than developed independently. Crucially, adding microbiome metrics to breeding programs does not meaningfully increase complexity. Rather, it adds explanatory power to existing trial designs and reduces unexplained variance.Microbiome-informed breeding in the context of climate changeAs climate change intensifies, the case for microbiome-informed breeding becomes even stronger. Many traits essential for climate resilience, such as drought tolerance, heat resilience, salinity resistance, and nutrient-use efficiency, are mediated, at least in part, by microbial partners. Microbes can enhance root growth under drought, produce osmoprotectants, modulate hormone signaling, and suppress pathogens that often intensify under climate stress. If breeding programs aim to produce crops that remain productive under increasingly variable conditions, microbial contributions must be explicitly measured, managed, and incorporated into predictive models.ConclusionsHistorically, plant breeding has adapted to new biological insights. From Mendelian genetics to quantitative genetics, from molecular markers to genomic selection and machine learning, each technological shift has redefined what is possible. The microbiome represents the next frontier, an ecological dimension of plant phenotype that is measurable, heritable, and exploitable for crop improvement (Morris and Bohannan6,7). Integrating microbiome information into plant breeding trials is a necessary step toward more predictive, stable, and sustainable breeding outcomes. Crucially, adding microbiome metrics to breeding programs does not meaningfully increase complexity. Rather, it adds explanatory power to existing trial designs and reduces unexplained variance, while still requiring careful interpretation given the strong environmental and temporal modulation of plant–microbiome associations.Data availabilityNo datasets were generated or analysed during the current study.ReferencesDwivedi, S. L. et al. Exploitation of rhizosphere microbiome biodiversity in plant breeding. Trends Plant Sci. 30, 1033–1045 (2025).Article CAS PubMed Google Scholar Sharma, N., Mahawar, L., Mishra, A. & Albrectsen, B. R. Microbial contributions to plant growth and stress tolerance: Mechanisms for sustainable plant production. Plant Stress 17, 100966 (2025).Article CAS Google Scholar Escudero-Martinez, C. & Bulgarelli, D. Engineering the crop microbiota through host genetics. Ann. Rev. Phytopathol. 61, 257–277 (2023).Article CAS Google Scholar Zhao, T. et al. Harnessing microbiome-plant synergies: microbiome-interactive traits enhance plant growth and support sustainable agriculture. npj Sustain. Agric. 3, 50 (2025).Article CAS Google Scholar Mullualem, D. et al. Genotype-by-environment interaction and stability analysis of grain yield of bread wheat (Triticum aestivum L.) genotypes using AMMI and GGE biplot analyses. Heliyon 14, e32918 (2024).Article Google Scholar Araujo, A. S. F., Pereira, A. P. A., de Medeiros, E. V. & Mendes, L. W. Root architecture and the rhizosphere microbiome: Shaping sustainable agriculture. Plant Sci. 359, 112599 (2025).Article CAS PubMed Google Scholar Morris, A. H. & Bohannan, B. J. M. Estimates of microbiome heritability across hosts. Nat. Microbiol. 9, 3110–3119 (2024).Article CAS PubMed Google Scholar Cernava, T. Coming of age for Microbiome gene breeding in plants. Nat. Commun. 15, 6623 (2024).Article CAS PubMed PubMed Central Google Scholar Week, B. et al. Quantitative genetics of microbiome-mediated traits. Evolution 79, 2487–2502 (2025).Article CAS PubMed PubMed Central Google Scholar Mueller, U. G. & Linksvayer, T. A. Microbiome breeding: conceptual and practical issues. Trends Microbiol. 30, 997–1011 (2022).Article CAS PubMed Google Scholar Zhang, J., Liu, W., Bu, J., Lin, Y. & Bai, Y. Host genetics regulate the plant microbiome. Curr. Opin. Microbiol. 72, 102268 (2023).Article CAS PubMed Google Scholar Liu-Xu, L. et al. Harnessing green helpers: nitrogen-fixing bacteria and other beneficial microorganisms in plant–microbe interactions for sustainable agriculture. Horticulturae 10, 621 (2024).Article Google Scholar Trivedi, P. et al. Keystone microbial taxa regulate the invasion of a fungal pathogen in agro-ecosystems. Soil Biol. Biochem. 111, 10–14 (2017).Article CAS Google Scholar Ait-El-Mokhtar, M., Meddich, A. & Baslam, M. Plant-microbiome interactions under drought—insights from the molecular machinist’s toolbox. Front. Sustain. Food Syst. 7, 1253735 (2023).Article Google Scholar Nadarajah, K. & Abdul Rahman, N. S. N. The microbial connection to sustainable agriculture. Plants 12, 2307 (2023).Article CAS PubMed PubMed Central Google Scholar Download referencesAcknowledgementsNo funding was received for this research.Author informationAuthors and AffiliationsSoil Microbial Ecology Group, Agricultural Science Center; Federal University of Piauí, Teresina, BrazilAdemir Sergio Ferreira Araujo & Romario Martins CostaDepartment of Plant Science, Agricultural Science Center; Federal University of Piauí, Teresina, BrazilAngela Celis de Almeida LopesAuthorsAdemir Sergio Ferreira AraujoView author publicationsSearch author on:PubMed Google ScholarRomario Martins CostaView author publicationsSearch author on:PubMed Google ScholarAngela Celis de Almeida LopesView author publicationsSearch author on:PubMed Google ScholarContributionsA.S.F.A, R.M.C., and A.C.A.L. wrote and revised the manuscript. All authors read and approved the manuscript.Corresponding authorsCorrespondence to Ademir Sergio Ferreira Araujo or Angela Celis de Almeida Lopes.Ethics declarationsCompeting interestsThe authors declare no competing interests.Additional informationPublisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Rights and permissionsOpen Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.Reprints and permissionsAbout this article