Species-level studies do not upscale to community-wide plant-soil feedbacks

Species-level studies do not upscale to community-wide plant-soil feedbacks | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Species-level studies do not upscale to community-wide plant-soil feedbacks XINMIN Lu, Mengyue Li, Yan Sun, Chunqiang Wei, Lunlun Gao, Evan Siemann, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6100233/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Soil legacy effects of plants (i.e., plant-soil feedback, PSF) are key drivers of the maintenance of biodiversity and alien plant invasion. While most research (88.9% of 460 experiments) has focused on PSFs of single-species, we showed that 4.6 (on average) herbaceous species co-occur and likely to interact belowground and collectively shaping plant-soil interactions across ~2000 0.25 m2 plots (only 0.03% hosting a single species) in a field survey of herbaceous plant communities in East China. However, can species-level PSFs directly translate to community-level PSFs remains untested. We experimentally showed that rhizosphere fungal communities of three species were significantly influenced by neighbouring species diversity, rather than just focal species itself. A two-phase PSF experiment further showed that, when the soil was conditioned by more than four species, community-level PSFs were disproportionately influenced by specific species rather than being a simple additive contribution from single-species PSFs. This inequality was more pronounced with increasing species richness. These findings highlight the importance of prioritizing research on community-level PSFs, particularly in species-rich ecosystems, which may reshape our understanding on plant-soil interactions and their role in shaping plant community dynamics. Biological sciences/Plant sciences/Plant ecology Biological sciences/Ecology Figures Figure 1 Figure 2 Figure 3 Main Text Plant-soil feedback (PSF, also known as soil legacy effect of plants) refers to the process by which living plants and their litter modify soil communities—including microbes and animals—as well as soil abiotic properties, ultimately influencing the growth and performance of subsequent plants 1-3 . PSFs can be negative, neutral, or positive, depending on factors such as the balance between pathogens, mutualists, and decomposers 4 , as well as the phylogenetic relationships between preceding and succeeding plant species 5 . To date, PSFs is acknowledged as a key driver of alien species invasion 6,7 , the maintenance of biodiversity, ecosystem functioning 8-10 , and ecosystem responses to climate change 11,12 . The current understanding of PSFs is largely based on experiments consisting of two phases: conditioning and feedback phases 1,10,13 . In the conditioning phase, a single species is typically grown to modify biotic and abiotic soil properties 10,14 . During the feedback (response) phase, the same or different species (hereafter: responding species), or entire plant communities (hereafter: responding community) are grown in the conditioned soil. Soil feedback effects of a single species (hereafter: species-level PSFs) are then assessed by comparing the performance of the responding species or communities in sterilized versus non-sterilized conditioned soils or in conspecific versus heterospecific conditioned soils 15 . The estimated PSFs (including pairwise PSFs) are used to predict species coexistence, plant range shift, and community dynamics 1,9-11 . As a result, our current understanding of PSFs and the predictions regarding their effects on plant species coexistence and community composition are predominantly based on findings derived from species-level studies. However, these studies often fail to account for the fact that plant species typically coexist in mixtures, forming complex belowground interaction networks 16,17 and collectively conditioning soil abiotic and biotic properties, thereby influencing the dynamics of subsequent plant communities 18 . This process hereafter referred to as community-level PSFs. In plant mixtures, neighbouring plants can influence the soil communities associated with the roots of a focal plant (hereafter referred to as rhizosphere communities) through several mechanisms (Fig. 1A). First, neighbouring plants may cause spillover or dilution of pathogens and mutualists, such as arbuscular mycorrhizal fungi (AMF) 19-21 . Second, heterospecific neighbours can modify the rhizosphere chemistry of a focal plant by releasing their own chemicals or altering the focal plant’s root exudates 22,23 , thereby reshaping the focal plant’s rhizosphere communities 24,25 . For instance, the invasive plant Centaurea stoebe in North America changed the diversity and composition of rhizosphere fungal communities associated with neighbouring native plants by releasing allelopathic chemicals 26 . Third, heterospecific neighbours can influence the abiotic properties of the focal plant’s rhizosphere soil, such as nitrogen content and pH 27,28 . As a result, soil communities conditioned by plant mixtures, and in turn their PSFs, may not simply be the sum of those conditioned by individual component species. Research further suggests that species-rich or genetically diverse plant communities can exert either stronger 29 or weaker 30-33 negative PSFs on subsequent plant species compared to monocultures, with effects potentially varying based on plant or genetic richness. These findings raise a pivotal but yet unexplored question: Can the insights from species-level PSF studies reliably predict community-level PSFs? To address this question, we first performed a systematic review of the PSF literatures to examine how PSFs are experimentally tested in relation to the number of conditioning species. We identified experimental studies that met two criteria: 1) inclusion of both conditioning and feedback phases and 2) reporting the number of conditioning species per experimental unit (e.g., pot). Among the 460 PSF experiments (from 451 papers, Table S1) that met these criteria, 88.9% conditioned soils using a single plant species per experimental unit (Fig. 1B). Second, we conducted a regional-scale field survey of herbaceous plant communities within 1,997 plots (0.5 m × 0.5 m) across 442 grassland sites, covering an area of 780,717 km 2 in East China (Fig. 1C). We aimed to assess the richness of herbaceous plant species that likely to interact belowground under natural field conditions. Co-occurring herbs and grasses in grassland ecosystems can interact belowground at a distance of up to 0.8 m 26,34 . The number of species per plot ranged from one to 13 (average= 4.6), with the highest frequency observed at four species. Notably, only 0.03% of the plots contained a single species, while 46.7% contained more than four species (Fig. 1D). This finding indicates that plant individuals were often surrounded and may interact belowground with diverse heterospecific individuals, thereby collectively affecting subsequent plants via PSFs in nature. The literature survey and the field survey revealed a significant mismatch between current PSF experimental studies and the complexity of plant-soil interactions in natural ecosystems. To bridge this knowledge gap, we first conducted a greenhouse experiment to test how neighboring species richness influences the rhizosphere fungal communities of three plant species (Fig. S1). In this experiment, an individual plant (hereafter referred to as the focal plant) of one of the three species was surrounded by eight individuals from one, two, or four species in each experimental pot filled with topsoil collected from abandoned farmland near the Huazhong Agricultural University campus. There were 300 pots in total. Five months later, we characterized fungal communities within the rhizosphere soil (that strongly adheres to plant roots) of each focal plant. We specifically focused on fungal communities within the rhizosphere soil of focal plants, as these communities play a critical role in regulating plant growth, nutrient dynamics, and PSFs 35,36 . Additionally, plant legacy effects on soil fungi are likely to persist longer than those on bacteria 3,37 . Next, we performed a typical two-phase PSF experiment in the same greenhouse. In the conditioning phase, the same field-collected topsoil was homogenized and conditioned with living plants during the growing season and root litter during the non-growing season, as He, et al. performed 2 . We tested five levels of conditioning species richness (one, two, four, eight, and twelve species) randomly selected from a species pool of 28 plant species (including those used in the other greenhouse experiment, Fig. S2, Table S2). Plant aboveground biomass accumulated similarly during the conditioning phase across different species richness levels (F 4, 275 = 0.30, P = 0.876). In the feedback phase, we tested the effects of these conditioned soils on a six-species plant community, maintaining identical species composition across replicates. Neighbouring species richness affects focal species rhizosphere fungal community The Shannon diversity and inverse Simpson diversity of rhizosphere putative pathogenic fungal communities and AMF communities were significantly affected by focal species identity, neighbouring plant richness, and their interaction (Table S3). Importantly, the diversity of these fungal communities either increased or decreased with neighbouring plant richness, depending on the focal species identity and functional groups of fungi involved (Fig. 2A, 2B, 2E & 2F, Table S4). Focal species identity explained 64.9% -70.5% and 1.4% - 2.0% of the variation in the two variables for pathogenic fungal communities and AMF fungal communities, respectively (Fig. 2D & 2H, Table S5). Meanwhile, neighbouring plant richness and its interaction with focal species identity jointly explained 3.0%-4.5% and 7.2%-7.5% of the variation in the two variables for pathogenic fungal communities and AMF fungal communities, respectively (Fig. 2D & 2H, Table S6). Consistently, neighbouring plant richness and its interaction with focal species identity significantly influenced the composition of rhizosphere overall fungal communities, pathogenic fungal communities (Fig. 2C), and AMF fungal communities (Fig. 2G). These factors accounted for 4.1% - 4.8% of the compositional variation in these communities, while their composition was primarily shaped by the focal species identity, which explained approximately 10.7% - 24.5% of the variation (Fig. 2D & 2H, Table S5). Species-rich community-PSFs were unpredictable In our PSF experiment, we investigated whether higher species richness in the conditioning phase leads to predictable changes in the biomass of the responding plant community using three null models—additive, multiplicative, and dominative scenarios. These models evaluated whether community biomass aligned with expectations based on single-species effects (i.e. when soil is conditioned by a single species). At the single-conditioning species level, the biomass of the responding community varied based on the identity of the conditioning species, with the effects ranging from neutral to strongly negative (Fig. 3A, Table S6), in comparison to the biomass of the responding community when grown in sterilized soil. Notably, responding community biomass changed significantly at richness levels of four or more conditioning species, deviating substantially from null model predictions (Fig. 3B, Table S7). These deviations suggest the presence of synergistic interactions among conditioning species within the same communities, particularly when comparing observed responding community biomass to predictions from additive and multiplicative models. Incorporating conditioning species identity or effect size into the models greatly improved explanatory power, increasing the explained variance from 2.0% to 23.1% or 24.6% (Fig. 3C). These findings indicate that PSFs of species-rich conditioning communities cannot be reliably predicted based on single conditioning species effects alone, highlighting the complexity of plant-soil interactions in diverse communities. Discussion Over the past decades, a rising number of studies have conducted species-level PSF experiments and developed statistical models to predict species coexistence and community dynamics in natural ecosystems 10,13,38,39 , significantly advancing the field of community ecology and plant sciences. However, most of these studies 40,41 assume that species-level PSFs act additively, and thus they simply pooled equal amounts of soils conditioned by monocultures of component plant species when testing community-level PSFs. Our study challenges this assumption by revealing that PSFs in species-rich communities (e.g., richness above four in this study) were not simply the sum of PSFs by component species, but were instead dominated by PSFs of specific species. Additionally, including conditioning plant species identity further improved model accuracy. These findings suggest that within plant communities: 1) different plant species contribute unequally to community-level PSFs; and 2) the likelihood of including plant species with disproportionately large impacts on community-level PSFs increases with rising conditioning species richness, aligning with sampling effects observed in biodiversity-ecosystem functioning experiments 42,43 . Therefore, species-level PSFs may fail to reliably predict community-PSFs for 46.7% of the surveyed 0.5 m × 0.5 m plots that contain more than four species in this study. This failure rate could be even higher, as belowground interactions among co-occurring plants could extend beyond 0.5 m in grasslands 26 and further in forests 44 . These findings emphasize that species-level PSFs alone are insufficient for predicting the PSFs of species-rich communities in nature. We found that rhizosphere fungal communities, particularly putative pathogenic and mycorrhizal fungi, were interactively shaped by focal species identity and the richness of neighbouring species. First, focal species identity is the primary factor shaping their rhizosphere fungal composition, aligning with other studies 45,46 reporting distinct rhizosphere communities among plant species grown in monocultures. Second, the presence of heterospecific neighbours substantially modified the rhizosphere communities of the focal species, as found in other studies 20,34 . Notably, the rhizosphere fungal communities of the three focal species responded in a species-specific manner to the richness of neighbouring species, reflecting the role of high-order plant interactions (i.e., interactions between multiple plants 47 ) in regulating the assembly of rhizosphere microbes. Our findings suggest that within a community, certain species may disproportionately shape soil fungal communities and community-level PSFs. While it remains unclear whether this applies to other soil organisms (e.g., bacteria or nematodes), these results offer a potential explanation for the disproportional high contribution of specific species to the community-level PSFs (Fig. 3B), particularly when the richness of conditioning species exceeds four. The observed PSFs were less likely to be affected by soil abiotic properties. This is because there was no difference in plant biomass across different richness levels during the conditioning phase, suggesting the depletion of soil nutrients was likely similar. In conclusion, our literature review and comprehensive field survey reveal a critical limitation in current PSF experimental studies: the neglect of high-order plant interactions and their influence on plant-soil interactions in natural ecosystems. The two experiments further revealed that diverse plants can profoundly affect soil microbial assembly in the rhizosphere of neighbouring species, resulting in unequal contributions of different conditioning species to soil microbial communities and community-level PSFs. As a result, PSFs in species-rich communities cannot be accurately predicted by simply extrapolating from species-level PSFs or assuming additive effects. Consequently, we may have underestimated or overestimated the role of soil biota in maintaining plant diversity and driving alien plant invasion within species-rich plant communities. Therefore, our findings underscore the urgent need to shift the focus of plant-soil interaction research from individual plant species to whole plant communities, particularly in studies addressing plant interactions and community dynamics. Technically, we propose that future PSF experiments should incorporate a conditioning phase in which soil is cultivated by multi-species assemblages that mirror natural communities, followed by a feedback phase to test impacts of conditioned soil on individual plant species or plant communities. Declarations Data availability The data have been deposited to https://doi.org/10.6084/m9.figshare.28395707. DNA sequences have been deposited to the National Center for Biotechnology Information (NCBI, accession number PRJNA1206692). Code availability All code used to complete analyses for the manuscripts is available at the following link: https://doi.org/10.6084/m9.figshare.28395707. Data analyses were conducted and were visualizations generated in R (version 4.2.2). Acknowledgements We would like to thank Yifan He, Biao Zhu, Wei Chen, and Yanli Zhan for their assistance in the experiment, Ragan M. Callaway, Leho Tedersoo, Marina Semchenko, Mark van Kleunen, Bernhard Schmid, Heinz Müller-Schärer for comments on the manuscript. Author Contributions X.L. conceived and designed the study. M.Y. conducted the literature review and the experiments. C.W., and L.G. performed the field survey. Y. S., and M.Y. performed statistical analysis. X.L., Y.S., & N.X wrote the draft of the manuscript. All authors contributed to revisions and gave final approval for publication. Competing interests . The authors declare no competing interest. Funding This study was financially supported by the National Natural Science Foundation of China (32171585, 32371749 & 32201438). References Bever, J. D., Westover, K. M. & Antonovics, J. Incorporating the soil community into plant population dynamics: the utility of the feedback approach. J. Ecol. 85 , 561-573 (1997). He, Y. et al. Leaf litter presence in the non-growing season prolongs plant legacy effects on soil fungal communities and succeeding plant growth. J. Ecol. 111 , 1997-2009 (2023). Heinen, R. et al. Plant community composition steers grassland vegetation via soil legacy effects. Ecol. Lett. 23 , 973-982 (2020). van der Putten, W. H., Bradford, M. A., Pernilla Brinkman, E., van de Voorde, T. F. J. & Veen, G. F. Where, when and how plant-soil feedback matters in a changing world. Funct. Ecol. 30 , 1109-1121 (2016). Wandrag, E. M. et al. Phylogenetic signals and predictability in plant-soil feedbacks. New Phytol. 228 , 1440-1449 (2020). Callaway, R. M., Thelen, G. C., Rodriguez, A. & Holben, W. E. Soil biota and exotic plant invasion. Nature 427 , 731-733 (2004). Klironomos, J. N. Feedback with soil biota contributes to plant rarity and invasiveness in communities. Nature 417 , 67-70 (2002). Thakur, M. P. et al. Plant-soil feedbacks and temporal dynamics of plant diversity-productivity relationships. Trends Ecol. Evol. 36 , 651-661 (2021). Teste, F. P. et al. Plant-soil feedback and the maintenance of diversity in Mediterranean-climate shrublands. Science 355 , 173-176 (2017). Yan, X., Levine, J. M. & Kandlikar, G. S. A quantitative synthesis of soil microbial effects on plant species coexistence. Proc. Natl. Acad. Sci. USA 119 , e2122088119 (2022). Van Nuland, M. E., Bailey, J. K. & Schweitzer, J. A. Divergent plant-soil feedbacks could alter future elevation ranges and ecosystem dynamics. Nat. Ecol. Evol. 1 , 0150 (2017). Dudenhöffer, J.-H., Luecke, N. C. & Crawford, K. M. Changes in precipitation patterns can destabilize plant species coexistence via changes in plant-soil feedback. Nat. Ecol. Evol. 6 , 546-554 (2022). van der Putten, W. H. et al. Plant-soil feedbacks: the past, the present and future challenges. J. Ecol. 101 , 265-276 (2013). de Vries, F., Lau, J., Hawkes, C. & Semchenko, M. Plant-soil feedback under drought: does history shape the future? Trends Ecol. Evol. 38 , 708-718 (2023). Pernilla Brinkman, E., Van der Putten, W. H., Bakker, E.-J. & Verhoeven, K. J. F. Plant-soil feedback: experimental approaches, statistical analyses and ecological interpretations. J. Ecol. 98 , 1063-1073 (2010). Toju, H., Guimarães, P. R., Olesen, J. M. & Thompson, J. N. Assembly of complex plant-fungus networks. Nat. Commun. 5 , 5273 (2014). Toju, H., Guimarães, P. R., Olesen, J. M. & Thompson, J. N. Below-ground plant-fungus network topology is not congruent with above-ground plant-animal network topology. Sci. Adv. 1 , e1500291 (2015). Kardol, P., Martijn Bezemer, T. & van der Putten, W. H. Temporal variation in plant-soil feedback controls succession. Ecol. Lett. 9 , 1080-1088 (2006). Mony, C., Gaudu, V., Ricono, C., Jambon, O. & Vandenkoornhuyse, P. Plant neighbours shape fungal assemblages associated with plant roots: a new understanding of niche-partitioning in plant communities. Funct. Ecol. 35 , 1768-1782 (2021). Hu, J. et al. Neighbourhood effect of weeds on wheat root endospheric mycobiota. J. Ecol. 111 , 994-1008 (2023). Keesing, F., Holt, R. D. & Ostfeld, R. S. Effects of species diversity on disease risk. Ecol. Lett. 9 , 485-498 (2006). Weinhold, A. et al. Tree species richness differentially affects the chemical composition of leaves, roots and root exudates in four subtropical tree species. J. Ecol. 110 , 97-116 (2022). Turner, S. C. & Schweitzer, J. A. Plant neighbors differentially alter a focal species' biotic interactions through changes to resource allocation. Ecology 105 , e4395 (2024). Zhalnina, K. et al. Dynamic root exudate chemistry and microbial substrate preferences drive patterns in rhizosphere microbial community assembly. Nat. Microbiol. 3 , 470-480 (2018). Delory, B. M., Callaway, R. M. & Semchenko, M. A trait-based framework linking the soil metabolome to plant-soil feedbacks. New Phytol. 241 , 1910-1921 (2024). Broz, A. K., Manter, D. K. & Vivanco, J. M. Soil fungal abundance and diversity: another victim of the invasive plant Centaurea maculosa . ISME J. 1 , 763-765 (2007). Li, L. et al. Diversity enhances agricultural productivity via rhizosphere phosphorus facilitation on phosphorus-deficient soils. Proc. Natl. Acad. Sci. USA 104 , 11192-11196 (2007). Sun, Y. et al. Plant intraspecific competition and growth stage alter carbon and nitrogen mineralization in the rhizosphere. Plant Cell Environ. 44 , 1231-1242 (2021). Cheng, C. et al. Genotype diversity enhances invasion resistance of native plants via soil biotic feedbacks. Ecol. Lett. 27 , e14384 (2024). Yang, L., Maron, J. L. & Callaway, R. M. Inhibitory effects of soil biota are ameliorated by high plant diversity. Oecologia 179 , 519-525 (2015). Guerrero-Ramírez, N. R., Reich, P. B., Wagg, C., Ciobanu, M. & Eisenhauer, N. Diversity-dependent plant-soil feedbacks underlie long-term plan diversity effects on primary productivity. Ecosphere 10 , e02704 (2019). Ren, H. et al. Diversity-conditioned soil strengthens plant diversity-productivity relationship. J. Appl. Ecol. 60 , 1881-1892 (2023). Forero, L. E., Kulmatiski, A., Grenzer, J. & Norton, J. M. Plant-soil feedbacks help explain biodiversity-productivity relationships. Commun. Biol. 4 , 789 (2021). Gao, L. et al. Latitudinal variation in the diversity and composition of various organisms associated with an exotic plant: the role of climate and plant invasion. New Phytol. 231 , 1559-1569 (2021). Wu, S. et al. Soil organic matter dynamics mediated by arbuscular mycorrhizal fungi-an updated conceptual framework. New Phytol. 242 , 1417-1425 (2024). Banerjee, S. & van der Heijden, M. G. A. Soil microbiomes and one health. Nat. Rev. Microbiol. 21 , 6-20 (2023). Hannula, S. E. et al. Persistence of plant-mediated microbial soil legacy effects in soil and inside roots. Nat. Commun. 12 , 5686 (2021). Eppinga, M. B. et al. Frequency-dependent feedback constrains plant community coexistence. Nat. Ecol. Evol. 2 , 1403-1407 (2018). Bever, J. D. Soil community feedback and the coexistence of competitors: conceptual frameworks and empirical tests. New Phytol. 157 , 465-473 (2003). Zhang, Z., Liu, Y., Brunel, C. & van Kleunen, M. Soil-microorganism-mediated invasional meltdown in plants. Nat. Ecol. Evol. 4 , 1612-1621 (2020). Zhang, Z., Liu, Y., Brunel, C. & van Kleunen, M. Evidence for Elton's diversity-invasibility hypothesis from belowground. Ecology 101 , e03187 (2020). Wardle, D. A. Is "sampling effect" a problem for experiments investigating biodiversity-ecosystem function relationships? Oikos 87 , 403-407 (1999). Loreau, M. & Hector, A. Partitioning selection and complementarity in biodiversity experiments. Nature 412 , 72-76 (2001). Klein, T., Siegwolf, R. T. W. & Körner, C. Belowground carbon trade among tall trees in a temperate forest. Science 352 , 342-344 (2016). Fitzpatrick, C. R. et al. Assembly and ecological function of the root microbiome across angiosperm plant species. Proc. Natl. Acad. Sci. USA 115 , E1157-E1165 (2018). Sweeney, C. J., de Vries, F. T., van Dongen, B. E. & Bardgett, R. D. Root traits explain rhizosphere fungal community composition among temperate grassland plant species. New Phytol. 229 , 1492-1507 (2021). Levine, J. M., Bascompte, J., Adler, P. B. & Allesina, S. Beyond pairwise mechanisms of species coexistence in complex communities. Nature 546 , 56-64 (2017). Methods Literature synthesis We performed a literature search on the 3rd of December 2024 in Web of Science with the search string: ‘soil feedback’ or ‘soil legacy effect’ or ‘soil effect’ and ‘plant’, restricted to articles written in English. This search yielded 1,602 articles. We then performed an additional search using the terms: ‘plant-soil feedback’ and ‘meta’, resulting in 25 meta-analysis articles. Subsequently, we pooled all 1,602 articles and all the references cited in these meta-analysis articles 10,48-71 , removing duplicates, and obtained a total of 2,502 articles. We selected articles from the 2,502 articles based on the following criteria: 1) experimental studies only, excluding synthesis papers and meta-analyses; 2) the experiment contains both conditioning (i.e., soils conditioned by plants or plant communities) and feedback phases; and 3) the article clearly described the number of conditioning species for each experimental unit (e.g., pot or plot). Studies investigating the legacy effects of plant communities in the field without species-level information were excluded, as were studies examining the effects of genetic diversity within a single species on PSFs. In total, 460 experiments from 451 articles met these criteria and were included in our data synthesis (Table S1). Field survey To explore the number of herbaceous plant species that are likely to interact belowground in nature, we performed a comprehensive field survey between October and November over four years, 2017 (49 sites, 245 plots), 2019 (128 sites, 389 plots), 2020 (97sites, 503 plots) and 2021 (168 sites, 840 plots), covering a latitudinal range from 21.5°N to 36.7°N, for a total of 442 randomly selected sites (Fig. 1C) and 1,977 plots (0.5 m × 0.5 m). These sites were in natural ecosystems (e.g., river banks) or abandoned farmlands. Species richness did not differ between natural ecosystems and farmlands (linear model: χ 2 =1.98, P =0.1593). At each site, we randomly established three to five 0.5 m × 0.5 m quadrats (≥ 2 m apart), at which spatial scale co-occurring herbaceous species are likely to interact with each other belowground in grasslands 26,34 and thus jointly shaping community-level PSFs. No big trees (height > 1 m) within 10 meters. We morphologically identified all plant species in situ and recorded the number of species within each plot, as in Gao, et al. 34 . Neighbouring effect experiment: effects of neighbouring plant richness on the rhizosphere fungal communities of three species To explore how neighbouring species richness affects the rhizosphere fungal communities of a plant within plant communities, we performed a greenhouse experiment on the campus of Huazhong Agricultural University (HZAU) in Wuhan, China (30°47′ N, 110°35′ E). The average minimum and maximum temperatures are 1.0 °C in January and 32.9 °C in July, and the mean annual precipitation is 1,316.0 mm (www.nmc.cn). We focus on the rhizosphere communities of Setaria viridis (L.) P. Beauv., Celosia argentea L., and Senna occidentalis (L.) Link (hereafter referred to as focal species). In addition, we chose Chenopodium album L. ( Amaranthaceae), Perilla frutescens (L.) Britton (Lamiaceae), Sigesbeckia orientalis L. (Asteraceae) and Urena lobata L. (Malvaceae) as neighbouring species to perform this experiment. All these species frequently co-occur in the natural fields. In October 2021, for each plant species, we collected seeds from three sites (> 10 km apart from each other) in Wuhan, pooled, and stored the seeds at 4 °C and low relative humidity (< 30%). This experiment is a factorial, randomized block design and was performed in a greenhouse at the HZAU campus. In March 2022, we germinated the seeds of these species by sowing surface-sterilized seeds in trays filled with a mixture of gamma-irradiated (dose > 25 kGy) peat and sand (2:1 by volume). The trays were placed randomly in a greenhouse on the HZAU campus. Seeds were surface sterilized by immersing in 70% ethanol for 30 seconds, followed by rinsing in demineralized water twice. Germinated seedlings were watered every other day. At the same time, we collected topsoil (0-10 cm depth) from a recently abandoned field near the HZAU campus. All the seven species had not occurred in the field before. In a lab, we sieved and homogenized the soil (2 mm), and then potted soil into 8 L experimental pots (diameter: 27 cm, height: 18 cm). In late April 2022, an individual of a focal plant species (hereafter: focal plant) was planted at the centre of each experimental pot. Considering the potential effects of distance on the focal plants' rhizosphere fungi, we planted three co-occur individuals (either the same as the focal species or three of one of four other species: C. album , P. frutescens , S. orientalis or U. lobate ) 5 cm away from the focal plant (Fig. S1). To test the effects of neighbouring species richness on a focal plant’s rhizosphere fungi, eight neighbouring plants planted 10 cm away from the focal plant were from 1 (8 each), 2 (4 each) or 4 (2 each) species from a pool of the focal plant species plus the four other species (Fig. S1). There were 25, 50 and 25 replicates, respectively, of the 1, 2 and 4 species diversity levels for each focal species. In total, we planted all the combinations for each focal plant species, resulting in 300 pots (3 focal species × 100 neighbouring richness treatments). Similar-sized seedlings were selected for each species to ensure seedlings were planted at the same growth stage (i.e., similar height). Seedlings that died within one week after transplanting were replaced by new ones. The 300 experimental pots were randomly placed in the greenhouse and repositioned (40 cm apart) monthly. Each pot was watered with 500 mL water every other day with glass beakers to avoid cross-contamination through water splash. Five months later, for each focal individual, we gently shook the roots for 10 min to remove the bulk soil and then sampled the soil that still adhered to the taproot surface (rhizosphere soil) with a brush. After that, we put the rhizosphere soil into a 2 mL sterile centrifuge tube and stored the samples at -80℃ for DNA extraction. In 300 pots, 20 focal plants died. Fungi DNA sequencing We sequenced the rhizosphere fungal communities of each focal plant, as described in Gao, et al. 34 . Soil microbial DNA was extracted using cetyltrimethylammonium bromide method 72 . DNA quality, concentration, and purification were checked by 1% agarose gel electrophoresis. The fungal ITS1 region was amplified with primers ITS1-F (5’-CTTGGTCATTTAGAGGAAGTAA-3’) and ITS2 (5’-GCTGCGTTCTTCATCGATGC-3’) 73,74 . Polymerase chain reaction (PCR) was carried out with 15 µL of Phusion® High-Fidelity PCR Master Mix (New England Biolabs), 2 µM of forward and reverse primers, and 10 ng of template DNA. The PCR amplification included an initial denaturation step at 98℃ for 1 min, followed by 30 cycles of denaturation at 98℃ (10 s), annealing at 50℃ (30 s), extension a 72℃ (30 s), and a final extension at 72°C (5 min). PCR products were checked on a 2% agarose gel to determine amplification success and the relative intensity of bands. Multiple PCR samples were pooled together. Pooled samples were extracted and purified using the Qiagen Gel Extraction Kit (Qiagen, Germany). DNA libraries were prepared using TruSeq® DNA PCR-Free Sample Preparation Kit for Illumina (New England Biolabs, USA) following the manufacturer’s protocol, and PE250 sequencing was performed on a NovaSeq 6000 Sequencing System (Illumina, Inc. San Diego, USA) in Novogene Company. Strict quality control steps were applied to the sequencing data. Briefly, after the barcode and primer sequences were trimmed from reads, paired-end reads were merged using FLASH (version 1.2.7) 75 , and the joint pairs were quality filtered in QIIME (version 1.9.l) 76 , following the standard method. All chimeric sequences were removed independently of the Unite database (v8.2) using the UCHIME algorithm 77 . OTUs were categorized at 97% sequence similarity using the UPARSE pipeline (version 7.0.1001), and taxonomic assignment for fungi was performed using the Unite database (v8.2) via the BLAST method in QIIME (Version 1.9.1). In total, there were 21,883,668 reads of 11,851 fungal OTUs, and the number of OTU reads for each sample ranged from 46,985 to 90,589. Fungal ITS1 sequences were blasted against the GenBank nucleotide database. Subsequently, we assigned fungi to plant pathogens or arbuscular mycorrhizal fungi (AMF) based on the FUNGuild database 78 . As many species in the Fusarium genus are major plant pathogens 79 , we assigned OTUs belonging to this genus as plant pathogens. Plant-soil feedback (PSF) experiment: the effects of conditioning species richness on PSFs To explore how the identity and richness of conditioning plant species affect plant-soil feedbacks (PSFs), we performed a two-stage PSF experiment from 2021 to 2022 in the same greenhouse as the ‘Neighbouring effects experiment’. In October 2020, we collected seeds from 28 common plant species belonging to nine families, of which 25 are annuals and three are perennials (which cannot survive the cold winters in Wuhan) (Table S2). All the plant species used in the ‘neighbouring effect experiment’ were used in this experiment. For each species, we collected seeds from three sites (> 10 km apart), pooled and stored at 4 °C and low relative humidity (< 30%). Conditioning phase Based on the field survey (Fig. 1D, plant richness at plot level), we established a plant richness gradient of 1, 2, 4, 6, or 12 species and constructed plant communities randomly using the 28 plant species (Fig. S2). To conduct the experiment, we germinated the seeds of these species with the same method as in the ‘Neighbouring effect experiment’, and collected topsoil (0-10 cm depth) from the same abandoned field at the end of June 2021 in a greenhouse. None of the 28 species had occurred in the field before. In another greenhouse (the one where the ‘neighbouring effect experiment’ was performed), we sieved (2 mm mesh), homogenized, and potted the soil into 8 L experimental pots. In late July 2021, we planted similarly sized seedlings (~ 8 cm in height) into experimental pots at a density of 12 individuals per pot (~5 cm apart). Within each pot, the species were randomly positioned, ensuring that the number of individuals was equal for each species (e.g., for two species, six individuals per species). Monocultures of each species were repeated five times (28 species × 5 replicates), and each richness level had 36 replicates (4 richness levels × 36 replicates), resulting in a total of 284 pots. The experimental pots were spaced 40 cm apart and randomly repositioned in the greenhouse monthly. Plants were watered with 500 mL/pot of water every other day throughout the trial. In November 2021, for each pot, we harvested and weighed each species' aboveground biomass (after drying at 80 °C for 48 hours) separately. Plant roots were left in the soil to decompose (almost all plant roots decomposed before the onset of the feedback experiment) and reinforce the effects of living individuals in the non-growing season, as in He, et al. 2 . To minimize potential air-mediated influences between neighbouring pots, all pots were sealed with tinfoil, with nine one-cm-diameter holes on top for watering. Each pot received 500 mL of water monthly during the winter (November to February) and weekly during the spring (March to April). Feedback phase In mid-April 2022, we began the feedback experiment in the same greenhouse. In this phase, we used an identical plant community as the responding community (Fig. S1), composed of Paspalum conjugatum P.J.Bergius (Poaceae), Setaria viridis (L.) P. Beauv. (Poaceae), Achyranthes aspera L. (Amaranthaceae), Amaranthus hybridus L. (Amaranthaceae), Senna tora (L.) Roxb. (Fabaceae) and Senna occidentalis (L.) Link (Fabaceae). We mixed 2 L of conditioned soil from each of the 284 conditioning pots with 6 L of gamma-irradiated field-collected topsoil from the same source as in the conditioning phase. The mixtures were then potted into 8 L experimental pots. Additionally, five experimental pots were filled with only sterilized soil as a control (Fig. S2). This resulted in a total of 289 experimental pots, which were randomly repositioned in the greenhouse (40 cm apart) every month. In April, we planted similarly-sized seedlings of the six plant species in each pot (two individuals per species). The seedlings of the six species were germinated in a greenhouse with the same methods as in the conditioning phase. In each pot, plant individuals were randomly positioned, with a 5 cm spacing between them. Dead seedlings were replaced within one week. Plants were watered with 500 mL water/pot every other day throughout the trial. Four months later, for each pot, we harvested and weighed (after drying at 80 °C for 48 hours) the aboveground biomass of each species separately. Because the roots of the plants were severely intertwined, we collected, washed and weighed (after drying at 80 °C for 48 hours) the entire root mass for each pot. Data analysis All the data analyses were performed with R software (version 4.2.2). For multiple-variable analysis, P values were adjusted using the Bonferroni correction method 80 . Effect of neighbouring plant richness on the rhizosphere fungal communities of the three focal species Before data analyses, we removed the OTUs with a maximum of 10 reads across all samples or observed in fewer than three samples to eliminate potential sequencing artifacts 81 . We then rarefied the reads to 46,985 reads per sample, and log 2 (x+1)-transformed the OTU-abundance data to improve the normality of the residues. We chose Shannon diversity and the inverse Simpson index as alpha diversity indexes. We calculated these variables for the overall, putative pathogenic fungal and AMF fungi from rarefied counts separately with the vegan package 82 . We performed linear mixed models (LMMs) to explore the dependence of these variables on the fixed factors ‘the identity of focal species’ (FS), ‘the richness of neighbouring species’ (NR), their interaction (FS:NR), and the random factor ‘co-occur plat species type (COT, same vs. differ)’. We included ‘COT’ as a random factor in all analyses because neither this factor nor its interactions with other variables were statistically significant for the overall and AMF fungal communities (adjusted P -values ³ 0.2 in all cases, Table S8). Moreover, as this factor and its interactions with others are not central to our research focus, this method allows us to concentrate on the main variables of interest by taking into account variability related to ‘co-occur plat species type’ without explicitly modelling its effects or interactions. These analyses were performed with the glmmTMB package, which uses maximum likelihood to estimate model parameters with a Gaussian distribution 83 . We checked the normality of the residuals in all models using QQ-plots. We quantified the individual contributions of each factor in explaining the variation in rhizosphere fungal diversity using the MuMIn package 84 . We performed redundancy analysis (RDA) to explore the effects of focal species identity, the richness of neighbouring species and their interactions on the composition of rhizosphere fungal communities (DESeq2 normalized OTU abundance), using the vegan package. After that, we performed hierarchical partitioning analyses to calculate the individual contributions of each factor in explaining the compositional variation in rhizosphere fungal communities. The above analyses were performed for the overall fungal community, the putative pathogenic and AMF fungal communities separately. Whether community-level PSFs be predicted by PSFs of composing species Since some plant species in the responding plant community were also present during the conditioning phase, we first examined whether their presence and interactions with species richness in the conditioning phase affected the total biomass of the responding community. To assess this, we fitted a linear model with the total biomass of the responding community as the response variable and species richness, individual species presence, and their interactions as predictors. Residual diagnostics were checked using QQ plots, and statistical significance was evaluated using ANOVA. All analyses were statistically non-significant ( P -values ≥ 0.592 in all cases, Table S9). We thus removed the potential confounding effects of species co-occurrence between the conditioning and responding phases in subsequent analyses. In addition, we tested whether the biomass of responding community is strongly affected by the phylogenetic distance between conditioning species or communities and the responding community. To do this, we generated a phylogenic tree of the 28 plant species based on the ITS, matK and rbcL sequences downloaded from NCBI Genbank, with the maximum likelihood method in IQ-tree software 85 . Amborella trichopoda was selected as the outgroup , as it is the unique sister species to all other extant angiosperms 86 . We estimated UniFrac distance between the conditioning and responding communities. A linear model revealed that the biomass of the responding community was weekly (r = 0.14, p = 0.016) correlated with the uniFrac distance between the conditioning and responding communities. Given this weak correlation, phylogenetic distances between the conditioning and responding communities was not considered in the subsequent analyses. We quantified the mean and 95% confidence intervals (CIs) of biomass of the responding community for each conditioning species or community using a non-parametric bootstrap (1,000 iterations for each species). We quantified the PSFs as the biomass of the responding communities in conditioned soils compared to those in sterilized soils. For each conditioning plant community, we summed the PSFs of each composing species based on three null assumptions: additive (summing effects), multiplicative (combining proportional changes), and dominative (choosing the strongest effect), following the protocol in Rillig, et al. 87 . After that, we compared the predicted value to the observed PSFs across plant communities of differing richness. The goal was to check if any of these assumptions could predict community-PSFs based on PSFs of composing species. If not, it would suggest that community-level PSFs are unpredictable due to higher-order interactions. To investigate how the plant community richness explains variability (R 2 ) in the community-level PSFs, and to assess the influence of incorporating the identity of composing species and effect sizes on predictability, we modelled community-level PSFs using species richness as a baseline predictor. The model was then refined by adding the identity of composing species as an explanatory variable. Alternatively, effect sizes derived from predicted community-PSFs under the three assumptions (additive, multiplicative, and dominative) were included in place of the species identity, providing expected outcomes for community-level PSFs under each assumption. We performed random forest models with cforest in the party packages to yield unbiased variable importance and more reliable predictions due to its use of unbiased trees 88 . The model used 1,000 trees and specified five as the number of variables randomly selected at each node. Bootstrap resampling was applied to estimate 95% CIs, and the proportion of variance explained by the mode (R 2 ) was estimated by the postResample function in the caret packages. References 48 Busetto, R., Breschi, V. & Formentin, S. Meta-learning for model-reference data-driven control. Automatica 172 , 112006 (2025). 49 Jiang, F. et al. Global patterns and drivers of plant-soil microbe interactions. Ecol. Lett. 27 , e14364 (2024). 50 Hager, H. A., Gailis, M. & Newman, J. A. Allelopathic effects of Epichloe fungal endophytes: experiment and meta-analysis. Plant Soil 488 , 217-232 (2023). 51 Bates, S. E., Wandrag, E. M. & Duncan, R. P. Calculating the uncertainty associated with log response ratios in plant-soil feedback studies. Plant Ecol. 221 , 829-836 (2020). 52 Hassan, K., Dastogeer, K. M. G., Carrillo, Y. & Nielsen, U. N. Climate change-driven shifts in plant-soil feedbacks: a meta-analysis. Ecol. Process 11 , 64 (2022). 53 Suding, K. N. et al. Consequences of plant-soil feedbacks in invasion. J. Ecol. 101 , 298-308 (2013). 54 Veresoglou, S. D., Li, G. C., Chen, J. & Johnson, D. Direction of plant-soil feedback determines plant responses to drought. Glob. Change Biol. 28 , 3995-3997 (2022). 55 Reinhart, K. O. et al. Globally, plant-soil feedbacks are weak predictors of plant abundance. Glob. Change Biol. 11 , 1756-1768 (2021). 56 Kutakova, E., Herben, T. & Munzbergova, Z. Heterospecific plant-soil feedback and its relationship to plant traits, species relatedness, and co-occurrence in natural communities. Oecologia 187 , 679-688 (2018). 57 Wang, H.-R., Zhang, J.-M., Zhao, X.-Y. & Feng, F. N limit as a switch node between positive and negative plant-soil feedback: a meta-analysis based on the covariant diagnosis of plant growth and soil factors. Ecotoxicol. Environ. Saf. 237 , 113557 (2022). 58 Waring, B. G. et al. Pervasive and strong effects of plants on soil chemistry: a meta-analysis of individual plant 'Zinke' effects. Proc. R. Soc. B: Biol. Sci. 282 , 91-98 (2015). 59 Brinkman, E. P., Van der Putten, W. H., Bakker, E.-J. & Verhoeven, K. J. F. Plant-soil feedback: experimental approaches, statistical analyses and ecological interpretations. J. Ecol. 98 , 1063-1073 (2010). 60 De Long, J. R. et al. Plant-soil feedback: incorporating untested influential drivers and reconciling terminology. Plant Soil 485 , 7-43 (2023). 61 Meisner, A. et al. Plant-soil feedbacks of exotic plant species across life forms: a meta-analysis. Biol. Invasions 16 , 2551-2561 (2014). 62 Beals, K. K. et al. Predicting plant-soil feedback in the field: meta-analysis reveals that competition and environmental stress differentially influence PSF. Front. Ecol. Evol. 8 , 191 (2020). 63 Mehrabi, Z. & Tuck, S. L. Relatedness is a poor predictor of negative plant-soil feedbacks. New Phytol. 205 , 1071-1075 (2015). 64 Lekberg, Y. et al. Relative importance of competition and plant-soil feedback, their synergy, context dependency and implications for coexistence. Ecol. Lett. 21 , 1268-1281 (2018). 65 Refsland, T. K. et al. Synthesis of plant-soil feedback effects on eastern North American tree species: implications for climate-adaptive forestry. Front. Ecol. Evol. 11 , 1073724 (2023). 66 Comita, L. S. et al. Testing predictions of the Janzen-Connell hypothesis: a meta-analysis of experimental evidence for distance- and density-dependent seed and seedling survival. J. Ecol. 102 , 845-856 (2014). 67 Cheng, C., Gundale, M. J., Li, B. & Wu, J. Deciphering the drivers of plant-soil feedbacks and their context-dependence: a meta-analysis. Plant Soil 10.1007/s11104-024-06922-1 (2024). 68 Kulmatiski, A., Beard, K. H., Stevens, J. R. & Cobbold, S. M. Plant-soil feedbacks: a meta-analytical review. Ecol. Lett. 11 , 980-992 (2008). 69 Xi, N. et al. Relationships between plant-soil feedbacks and functional traits. J. Ecol. 109 , 3411-3423 (2021). 70 Allen, W. J., Sapsford, S. J. & Dickie, I. A. Soil sample pooling generates no consistent inference bias: a meta-analysis of 71 plant-soil feedback experiments. New Phytol. 231 , 1308-1315 (2021). 71 Crawford, K. M. et al. When and where plant-soil feedback may promote plant coexistence: a meta-analysis. Ecol. Lett. 22 , 1274-1284 (2019). 72 Doyle, J. J. & Doyle, J. L. A rapid DNA isolation procedure for small amounts of fresh leaf tissue. Phytochem. Bull. 19 , 11-15 (1987). 73 White, T. J., Bruns, T., Lee, S. & Taylor, J. in PCR Protocols https://doi.org/10.1016/B978-0-12-372180-8.50042-1 (eds Michael A. Innis, David H. Gelfand, John J. Sninsky, & Thomas J. White) 315-322 (Academic Press, 1990). 74 Gardes, M. & Bruns, T. D. ITS primers with enhanced specificity for basidiomycetes--application to the identification of mycorrhizae and rusts. Mol. Ecol. 2 , 113-118 (1993). 75 Magoč, T. & Salzberg, S. L. FLASH: fast length adjustment of short reads to improve genome assemblies. Bioinformatics 27 , 2957-2963 (2011). 76 Caporaso, J. G. et al. QIIME allows analysis of high-throughput community sequencing data. Nat. Methods 7 , 335-336 (2010). 77 Edgar, R. C., Haas, B. J., Clemente, J. C., Quince, C. & Knight, R. UCHIME improves sensitivity and speed of chimera detection. Bioinformatics 27 , 2194-2200 (2011). 78 Nguyen, N. H. et al. FUNGuild: an open annotation tool for parsing fungal community datasets by ecological guild. Funct. Ecol. 20 , 241-248 (2016). 79 Ma, L.-J. et al. Fusarium pathogenomics. Annu. Rev. Microbiol. 67 , 399-416 (2013). 80 Simes, R. J. & Simes, B. R. J. An improved Bonferroni procedure for multiple tests of significance. Biometrika 73 , 751-754 (1986). 81 Gomez-Alvarez, V., Teal, T. K. & Schmidt, T. M. Systematic artifacts in metagenomes from complex microbial communities. ISME J. 3 , 1314-1317 (2009). 82 Oksanen, J. et al. vegan: community ecology package. The Comprehensive R Archive Network doi: 0.32614/CRAN.package.vegan (2024). 83 Bates, D., Mächler, M., Bolker, B. & Walker, S. Fitting linear mixed-effects models using lme4. J. Stat. Softw. 67 , 1-48 (2015). 84 Bartoń, K. MuMIn: Multi-Model Inference. R package version 1.48.4 https://CRAN.R-project.org/package=MuMIn (2024). 85 Nguyen, L. T., Schmidt, H. A., von Haeseler, A. & Minh, B. Q. IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. Mol. Biol. Evol. 32 , 268-274 (2015). 86 Mathews, S. & Donoghue, M. J. The root of angiosperm phylogeny inferred from duplicate phytochrome genes. Science 286 , 947-950 (1999). 87 Rillig, M. C. et al. The role of multiple global change factors in driving soil functions and microbial biodiversity. Science 366 , 886-890 (2019). 88 Strobl, C., Boulesteix, A.-L., Kneib, T., Augustin, T. & Zeileis, A. Conditional variable importance for random forests. BMC Bioinform. 9 , 307 (2008). Additional Declarations There is NO Competing Interest. Supplementary Files Supportinginformationluli0223.docx Supporting tables and figures Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6100233","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":423487205,"identity":"87a0c6a5-cce8-4768-aa1a-c6a7d297fac9","order_by":0,"name":"XINMIN Lu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBUlEQVRIie3NMUvDQBjG8TccnMvZri8U7Fd4SyAoav0qOQJmkBTBpVOJCHHSWcGvIBwIwfEkg0tL14BLsjvo4lTQOzs5XMjocH+OG4778QD4fP+wAQBrkI72AIJLHeTbV+wi3BzaPz8NDcl7E8D5RyXt554Ez2KqiaWP15VZeV6MSbOXNwHTmZOIlWrviGflUhqy5BOleXIoILlwkp0bFSKJrNSWFCJQWkQjAVrmLsJ21WhDmEbr1hI8UXr41U3MCiJRHNW/KyTNCu8mYlUSUjwpa7Miizi5r3h48ECJkwwxe2pw8z2O1mnbfBaL49vXq7Z+n0+d5G+xvZi9qNd/n8/n8zn6AUvmWh8VHWu6AAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0001-8776-7869","institution":"Huazhong Agricultural University","correspondingAuthor":true,"prefix":"","firstName":"XINMIN","middleName":"","lastName":"Lu","suffix":""},{"id":423487206,"identity":"465655fd-1afc-426d-9b8f-693ae7035f1d","order_by":1,"name":"Mengyue Li","email":"","orcid":"","institution":"Huazhong Agricultural University","correspondingAuthor":false,"prefix":"","firstName":"Mengyue","middleName":"","lastName":"Li","suffix":""},{"id":423487207,"identity":"173f66dd-4d9a-46d6-a51d-fb8d6018df29","order_by":2,"name":"Yan Sun","email":"","orcid":"https://orcid.org/0000-0002-6439-266X","institution":"Huazhong Agicultural University","correspondingAuthor":false,"prefix":"","firstName":"Yan","middleName":"","lastName":"Sun","suffix":""},{"id":423487208,"identity":"10baaf65-39b3-41c2-a76d-6dab436505fe","order_by":3,"name":"Chunqiang Wei","email":"","orcid":"","institution":"Central China Normal University","correspondingAuthor":false,"prefix":"","firstName":"Chunqiang","middleName":"","lastName":"Wei","suffix":""},{"id":423487209,"identity":"a8a77c69-b8f0-4eed-886a-69e6fe6e0524","order_by":4,"name":"Lunlun Gao","email":"","orcid":"","institution":"Huazhong Agicultural University","correspondingAuthor":false,"prefix":"","firstName":"Lunlun","middleName":"","lastName":"Gao","suffix":""},{"id":423487210,"identity":"355aaeb5-7009-47f9-9dd9-4ff95a42fbe0","order_by":5,"name":"Evan Siemann","email":"","orcid":"","institution":"Rice University","correspondingAuthor":false,"prefix":"","firstName":"Evan","middleName":"","lastName":"Siemann","suffix":""},{"id":423487211,"identity":"5c62eb5c-0dea-4297-afd9-4f1cebf8845f","order_by":6,"name":"Paul Kardol","email":"","orcid":"https://orcid.org/0000-0001-7065-3435","institution":"Swedish University of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Paul","middleName":"","lastName":"Kardol","suffix":""},{"id":423487212,"identity":"1e2f9b72-c5a1-44a9-b46e-dc2d8895d8f2","order_by":7,"name":"Nianxun Xi","email":"","orcid":"https://orcid.org/0000-0002-1711-3875","institution":"Hainan University","correspondingAuthor":false,"prefix":"","firstName":"Nianxun","middleName":"","lastName":"Xi","suffix":""}],"badges":[],"createdAt":"2025-02-25 00:15:29","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6100233/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6100233/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":77705713,"identity":"7138edef-6796-4582-9f6f-cbad22b69ab0","added_by":"auto","created_at":"2025-03-04 12:00:29","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":156420,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eA mismatch between plant-soil feedback (PSF) experiments and the complexity of plant-soil interactions in nature. \u003c/strong\u003e(A) A conceptual diagram illustrating how neighbouring plants influence a focal plant’s rhizosphere communities. (B) The rapid increase in PSF studies over the past 30 years and the distribution of the maximum number of conditioning plant species per experimental unit (MaxNo_Csp) across these studies, with colours representing different MaxNo_Csp values. (C) Locations of the 442 sites for the field survey of herbaceous plant communities across East China, where plant species richness and composition were recorded. Dot size corresponds to the number of plots at each site. (D) The distribution of species richness in surveyed plots (0.5 m × 0.5 m). Panel (A) was created with BioRender.com.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6100233/v1/8bd0cc8b2c5a322e1a7eca53.png"},{"id":77704618,"identity":"47cd4744-91a5-4b10-acc1-adadf74272d9","added_by":"auto","created_at":"2025-03-04 11:52:30","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":102909,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEffects of neighbouring plant richness on the composition and diversity of rhizosphere fungal communities for three plant species. \u003c/strong\u003eShannon diversity (A \u0026amp; E) and Inverse Simpson diversity (B \u0026amp; F) of the putative pathogenic and arbuscular mycorrhizal fungal (AMF)\u003cstrong\u003e \u003c/strong\u003ecommunities. Redundancy analysis (RDA) examines the effect of neighbouring plant species richness on the composition of the fungal community for the three plant species (C \u0026amp; G). The percentage contribution of explained variance of Shannon and Inverse Simpson diversity and fungal community composition is explained by focal species identity, neighbouring plant species richness, and their interaction; the percentage displayed on each bar represents the total variance explained by the model (D \u0026amp; H; details in Supporting Information Table S6). In (A, B, E \u0026amp; F), solid lines indicate significant relationships (\u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003eadjust\u003c/em\u003e\u003c/sub\u003e \u0026lt; 0.05), while dashed lines indicate non-significant relationships (\u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003eadjust\u003c/em\u003e\u003c/sub\u003e ³ 0.05) between fungal diversity and neighbouring plant richness. In (C \u0026amp; G), ellipses represent 95% confidence intervals for different levels of neighbouring plant richness. Abbreviations for focal plant species: Sv: \u003cem\u003eSetaria\u0026nbsp;viridis\u003c/em\u003e; Car\u003cem\u003e: Celosia\u0026nbsp;argentea; \u003c/em\u003eSoc: \u003cem\u003eSenna\u0026nbsp;occidentalis\u003c/em\u003e.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6100233/v1/f9a580d30bd460f4639a8048.png"},{"id":77704616,"identity":"38497f46-f4e3-4e6f-9631-844bac804c39","added_by":"auto","created_at":"2025-03-04 11:52:29","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":103712,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEffects of conditioning species richness on the biomass of the responding plant community via plant-soil feedbacks (PSFs). \u003c/strong\u003e(A) Soil-mediated effects of single (Sor-Cs on the x-axis) and differing richness levels (1, 2, 4, 6, and 12 species) of conditioning species on the biomass of the responding plant community in the PSF experiment. (B) Predicted effects of species richness based on three different assumptions for combining multiple effect sizes: additive, multiplicative, and dominative. (C) Variability is explained by species richness alone (grey), with the addition of species identity (purple), or with effect size information (predicted value based on three assumptions; brown). In A, replicates are displayed as smaller dots and visualized using density ridgeline plots. Horizontal dashed lines indicate the biomass of responding community grown in sterilized soil (control), and error bars represent ± 95% confidence intervals. Different letters along the x-axis denote differing plant species (For detailed information of plant species, please see Table S2).\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6100233/v1/749591cc7f191fa1dd1ff4d0.png"},{"id":78892482,"identity":"a4db4bb4-166c-45b8-8f6c-916537955948","added_by":"auto","created_at":"2025-03-20 11:18:38","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1410058,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6100233/v1/fc4f23df-f1e2-418f-aa79-1c1e1a05800a.pdf"},{"id":77705717,"identity":"3140177a-7bf1-4d7e-98a2-cf0921eff8bc","added_by":"auto","created_at":"2025-03-04 12:00:30","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":763788,"visible":true,"origin":"","legend":"Supporting tables and figures","description":"","filename":"Supportinginformationluli0223.docx","url":"https://assets-eu.researchsquare.com/files/rs-6100233/v1/f4be37bb50bb4b2b8acdaec9.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Species-level studies do not upscale to community-wide plant-soil feedbacks","fulltext":[{"header":"Main Text","content":"\u003cp\u003ePlant-soil feedback (PSF, also known as soil legacy effect of plants) refers to the process by which living plants and their litter modify soil communities\u0026mdash;including microbes and animals\u0026mdash;as well as soil abiotic properties, ultimately influencing the growth and performance of subsequent plants \u003csup\u003e1-3\u003c/sup\u003e. PSFs can be negative, neutral, or positive, depending on factors such as the balance between pathogens, mutualists, and decomposers \u003csup\u003e4\u003c/sup\u003e, as well as the phylogenetic relationships between preceding and succeeding plant species \u003csup\u003e5\u003c/sup\u003e. To date, PSFs is acknowledged as a key driver of alien species invasion \u003csup\u003e6,7\u003c/sup\u003e, the maintenance of biodiversity, ecosystem functioning \u003csup\u003e8-10\u003c/sup\u003e, and ecosystem responses to climate change \u003csup\u003e11,12\u003c/sup\u003e. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe current understanding of PSFs is largely based on experiments consisting of two phases: conditioning and feedback phases \u003csup\u003e1,10,13\u003c/sup\u003e. In the conditioning phase, a single species is typically grown to modify biotic and abiotic soil properties \u003csup\u003e10,14\u003c/sup\u003e. During the feedback (response) phase, the same or different species (hereafter: responding species), or entire plant communities (hereafter: responding community) are grown in the conditioned soil. Soil feedback effects of a single species (hereafter: species-level PSFs) are then assessed by comparing the performance of the responding species or communities in sterilized versus non-sterilized conditioned soils or in conspecific versus heterospecific conditioned soils \u003csup\u003e15\u003c/sup\u003e. The estimated PSFs (including pairwise PSFs) are used to predict species coexistence, plant range shift, and community dynamics \u003csup\u003e1,9-11\u003c/sup\u003e. As a result, our current understanding of PSFs and the predictions regarding their effects on plant species coexistence and community composition are predominantly based on findings derived from species-level studies. However, these studies often fail to account for the fact that plant species typically coexist in mixtures, forming complex belowground interaction networks \u003csup\u003e16,17\u003c/sup\u003e and collectively conditioning soil abiotic and biotic properties, thereby influencing the dynamics of subsequent plant communities \u003csup\u003e18\u003c/sup\u003e. This process hereafter referred to as community-level PSFs.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn plant mixtures, neighbouring plants can influence the soil communities associated with the roots of a focal plant (hereafter referred to as rhizosphere communities) through several mechanisms (Fig. 1A). First, neighbouring plants may cause spillover or dilution of pathogens and mutualists, such as arbuscular mycorrhizal fungi (AMF) \u003csup\u003e19-21\u003c/sup\u003e. Second, heterospecific neighbours can modify the rhizosphere chemistry of a focal plant by releasing their own chemicals or altering the focal plant\u0026rsquo;s root exudates \u003csup\u003e22,23\u003c/sup\u003e, thereby reshaping the focal plant\u0026rsquo;s rhizosphere communities \u003csup\u003e24,25\u003c/sup\u003e. For instance, the invasive plant \u003cem\u003eCentaurea stoebe\u003c/em\u003e in North America changed the diversity and composition of rhizosphere fungal communities associated with neighbouring native plants by releasing allelopathic chemicals \u003csup\u003e26\u003c/sup\u003e. Third, heterospecific neighbours can influence the abiotic properties of the focal plant\u0026rsquo;s rhizosphere soil, such as nitrogen content and pH \u003csup\u003e27,28\u003c/sup\u003e. As a result, soil communities conditioned by plant mixtures, and in turn their PSFs, may not simply be the sum of those conditioned by individual component species. Research further suggests that species-rich or genetically diverse plant communities can exert either stronger \u003csup\u003e29\u003c/sup\u003e or weaker \u003csup\u003e30-33\u003c/sup\u003e negative PSFs on subsequent plant species compared to monocultures, with effects potentially varying based on plant or genetic richness. These findings raise a pivotal but yet unexplored question: Can the insights from species-level PSF studies reliably predict community-level PSFs?\u003c/p\u003e\n\u003cp\u003eTo address this question, we first performed a systematic review of the PSF literatures to examine how PSFs are experimentally tested in relation to the number of conditioning species. We identified experimental studies that met two criteria: 1) inclusion of both conditioning and feedback phases and 2) reporting the number of conditioning species per experimental unit (e.g., pot). Among the 460 PSF experiments (from 451 papers, Table S1) that met these criteria, 88.9% conditioned soils using a single plant species per experimental unit (Fig. 1B).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSecond, we conducted\u0026nbsp;a\u0026nbsp;regional-scale\u0026nbsp;field survey\u0026nbsp;of\u0026nbsp;herbaceous plant communities\u0026nbsp;within\u0026nbsp;1,997 plots\u0026nbsp;(0.5 m\u0026nbsp;\u0026times;\u0026nbsp;0.5 m)\u0026nbsp;across\u0026nbsp;442\u0026nbsp;grassland\u0026nbsp;sites,\u0026nbsp;covering an area of\u0026nbsp;780,717\u0026nbsp;km\u003csup\u003e2\u003c/sup\u003e in East China (Fig. 1C). We aimed to assess the richness of herbaceous plant species that likely to interact belowground under natural field conditions. Co-occurring herbs and grasses in grassland ecosystems can interact belowground at a distance of up to 0.8 m \u003csup\u003e26,34\u003c/sup\u003e.\u0026nbsp;The number of species per plot ranged from\u0026nbsp;one\u0026nbsp;to 13 (average= 4.6), with the highest frequency observed\u0026nbsp;at\u0026nbsp;four species.\u0026nbsp;Notably, only 0.03% of the plots contained a single species, while 46.7%\u0026nbsp;contained more than four species\u0026nbsp;(Fig. 1D). This finding indicates that plant individuals were often surrounded and may interact belowground with diverse heterospecific individuals, thereby collectively affecting subsequent plants via PSFs in nature.\u0026nbsp;The literature survey and the field survey\u0026nbsp;revealed\u0026nbsp;a significant\u0026nbsp;mismatch between current PSF experimental studies and the complexity of plant-soil interactions in natural ecosystems.\u003c/p\u003e\n\u003cp\u003eTo bridge this knowledge gap, we first conducted a greenhouse experiment to test how neighboring species richness influences the rhizosphere fungal communities of three plant species (Fig. S1). In this experiment, an individual plant (hereafter referred to as the focal plant) of one of the three species was surrounded by eight individuals from one, two, or four species in each experimental pot filled with topsoil collected from abandoned farmland near the Huazhong Agricultural University campus. There were 300 pots in total. Five months later, we characterized fungal communities within the rhizosphere soil (that strongly adheres to plant roots) of each focal plant. We specifically focused on fungal communities within the rhizosphere soil of focal plants, as these communities play a critical role in regulating plant growth, nutrient dynamics, and PSFs \u003csup\u003e35,36\u003c/sup\u003e. Additionally, plant legacy effects on soil fungi are likely to persist longer than those on bacteria \u003csup\u003e3,37\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eNext,\u0026nbsp;we performed a typical two-phase PSF experiment\u0026nbsp;in the same greenhouse. In the conditioning phase,\u0026nbsp;the same\u0026nbsp;field-collected topsoil was homogenized and conditioned with living plants during the growing season and root litter during the non-growing season, as He, et al.\u003cem\u003e\u0026nbsp;\u003c/em\u003eperformed \u003csup\u003e2\u003c/sup\u003e. We tested five levels of\u0026nbsp;conditioning\u0026nbsp;species richness (one, two, four, eight,\u0026nbsp;and\u0026nbsp;twelve\u0026nbsp;species) randomly selected from a\u0026nbsp;species\u0026nbsp;pool of 28\u0026nbsp;plant\u0026nbsp;species (including those used in the other greenhouse experiment, Fig. S2,\u0026nbsp;Table S2).\u0026nbsp;Plant aboveground biomass accumulated similarly during the conditioning phase across different species richness levels (F\u003csub\u003e4, 275\u0026nbsp;\u003c/sub\u003e= 0.30, \u003cem\u003eP\u0026nbsp;\u003c/em\u003e= 0.876). In the feedback phase, we tested the effects of these conditioned soils on a six-species plant community, maintaining identical species composition across replicates.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNeighbouring species richness affects focal species rhizosphere fungal community \u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe Shannon diversity and inverse Simpson diversity of rhizosphere putative pathogenic fungal communities and AMF communities were significantly affected by\u0026nbsp;focal species identity, neighbouring plant richness, and their interaction\u0026nbsp;(Table S3). Importantly, the diversity of these fungal communities either increased or decreased with neighbouring plant richness, depending on the focal species identity and functional groups of fungi\u0026nbsp;involved\u0026nbsp;(Fig.\u0026nbsp;2A, 2B,\u0026nbsp;2E \u0026amp; 2F,\u0026nbsp;Table S4). Focal species identity\u0026nbsp;explained 64.9% -70.5% and 1.4% - 2.0% of the variation in the two variables for pathogenic fungal communities and AMF fungal communities, respectively\u0026nbsp;(Fig. 2D \u0026amp; 2H,\u0026nbsp;Table S5). Meanwhile, neighbouring plant richness and its interaction with focal species identity jointly explained 3.0%-4.5% and 7.2%-7.5% of the variation in the two variables for pathogenic fungal communities and AMF fungal communities, respectively (Fig. 2D \u0026amp; 2H, Table S6). Consistently, neighbouring plant richness and its interaction with focal species identity significantly influenced the composition of rhizosphere overall fungal communities, pathogenic fungal communities (Fig. 2C), and AMF fungal communities (Fig. 2G). These factors accounted for 4.1% - 4.8% of the compositional variation in these communities, while their composition was primarily shaped by the focal species identity, which explained approximately 10.7% - 24.5% of the variation (Fig. 2D \u0026amp; 2H,\u0026nbsp;Table S5).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSpecies-rich community-PSFs were unpredictable\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn our PSF experiment, we investigated whether higher species richness in the conditioning phase leads to predictable changes in the biomass of the responding plant community using three null models\u0026mdash;additive, multiplicative, and dominative scenarios. These models evaluated whether community biomass aligned with expectations based on single-species effects (i.e. when soil is conditioned by a single species). At the single-conditioning species level, the biomass of the responding community varied based on the identity of the conditioning species, with the effects ranging from neutral to strongly negative (Fig. 3A, Table S6), in comparison to the biomass of the responding community when grown in sterilized soil. Notably, responding community biomass changed significantly at richness levels of four or more conditioning species, deviating substantially from null model predictions (Fig. 3B, Table S7). These deviations suggest the presence of synergistic interactions among conditioning species within the same communities, particularly when comparing observed responding community biomass to predictions from additive and multiplicative models. Incorporating conditioning species identity or effect size into the models greatly improved explanatory power, increasing the explained variance from 2.0% to 23.1% or 24.6% (Fig. 3C). These findings indicate that PSFs of species-rich conditioning communities cannot be reliably predicted based on single conditioning species effects alone, highlighting the complexity of plant-soil interactions in diverse communities.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOver the past decades,\u0026nbsp;a rising number of\u0026nbsp;studies have\u0026nbsp;conducted\u0026nbsp;species-level PSF\u0026nbsp;experiments\u0026nbsp;and developed statistical models to predict species coexistence and community dynamics in natural\u0026nbsp;ecosystems\u0026nbsp;\u003csup\u003e10,13,38,39\u003c/sup\u003e, significantly advancing\u0026nbsp;the field of\u0026nbsp;community ecology\u0026nbsp;and plant sciences. However, most\u0026nbsp;of these studies\u0026nbsp;\u003csup\u003e40,41\u003c/sup\u003e assume that species-level PSFs act additively, and thus they simply pooled equal amounts of soils conditioned by monocultures of component plant species when testing community-level PSFs. Our\u0026nbsp;study\u0026nbsp;challenges this assumption by revealing that PSFs in\u0026nbsp;species-rich communities (e.g., richness above four in this study)\u0026nbsp;were not simply the sum of PSFs by component species, but were instead dominated by PSFs of specific species. Additionally, including\u0026nbsp;conditioning\u0026nbsp;plant species\u0026nbsp;identity further improved model accuracy. These findings suggest that within plant communities: 1) different plant species contribute unequally to community-level PSFs; and 2) the likelihood of including plant species with disproportionately large impacts on community-level PSFs\u0026nbsp;increases with\u0026nbsp;rising conditioning\u0026nbsp;species richness, aligning with sampling effects observed in biodiversity-ecosystem functioning experiments\u0026nbsp;\u003csup\u003e42,43\u003c/sup\u003e.\u0026nbsp;Therefore,\u0026nbsp;species-level PSFs\u0026nbsp;may fail to reliably predict community-PSFs for 46.7%\u0026nbsp;of the\u0026nbsp;surveyed\u0026nbsp;0.5 m \u0026times; 0.5 m plots\u0026nbsp;that contain more than four species in this study.\u0026nbsp;This failure rate\u0026nbsp;could\u0026nbsp;be\u0026nbsp;even\u0026nbsp;higher, as\u0026nbsp;belowground interactions among co-occurring plants could extend beyond 0.5 m\u0026nbsp;in grasslands\u0026nbsp;\u003csup\u003e26\u003c/sup\u003e and further in forests\u0026nbsp;\u003csup\u003e44\u003c/sup\u003e.\u0026nbsp;These findings emphasize that species-level PSFs alone are insufficient for predicting the PSFs of species-rich communities in nature.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe found that rhizosphere fungal communities,\u0026nbsp;particularly\u0026nbsp;putative pathogenic and mycorrhizal fungi, were interactively shaped by focal species\u0026nbsp;identity\u0026nbsp;and\u0026nbsp;the\u0026nbsp;richness of neighbouring species.\u0026nbsp;First,\u0026nbsp;focal\u0026nbsp;species\u0026nbsp;identity\u0026nbsp;is the primary factor shaping their rhizosphere\u0026nbsp;fungal composition, aligning with other studies\u0026nbsp;\u003csup\u003e45,46\u003c/sup\u003e reporting distinct rhizosphere communities among plant species\u0026nbsp;grown in monocultures. Second,\u0026nbsp;the presence of heterospecific neighbours substantially modified\u0026nbsp;the rhizosphere communities of the focal species, as found in other studies\u0026nbsp;\u003csup\u003e20,34\u003c/sup\u003e.\u0026nbsp;Notably,\u0026nbsp;the rhizosphere fungal communities of the three\u0026nbsp;focal species responded\u0026nbsp;in a\u0026nbsp;species-specific\u0026nbsp;manner\u0026nbsp;to the richness of neighbouring species,\u0026nbsp;reflecting the role of high-order plant interactions\u0026nbsp;(i.e., interactions between multiple plants\u0026nbsp;\u003csup\u003e47\u003c/sup\u003e)\u0026nbsp;in regulating the\u0026nbsp;assembly of rhizosphere microbes. Our findings suggest that within a community,\u0026nbsp;certain\u0026nbsp;species may disproportionately shape soil fungal communities and community-level PSFs. While it remains unclear whether this applies to other soil\u0026nbsp;organisms\u0026nbsp;(e.g., bacteria or nematodes), these results offer a potential explanation for the\u0026nbsp;disproportional high contribution of specific species to the\u0026nbsp;community-level PSFs\u0026nbsp;(Fig. 3B), particularly when the richness of conditioning species exceeds four.\u0026nbsp;The observed PSFs were less likely to be affected by soil abiotic properties. This is because there was no difference in plant\u0026nbsp;biomass\u0026nbsp;across\u0026nbsp;different richness\u0026nbsp;levels during\u0026nbsp;the conditioning phase,\u0026nbsp;suggesting the depletion of soil nutrients was likely similar. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn conclusion, our literature review and comprehensive field survey reveal a critical limitation in current PSF experimental studies: the neglect of high-order plant interactions and their influence on plant-soil interactions in natural ecosystems. The two experiments further revealed that diverse plants can profoundly affect soil microbial assembly in the rhizosphere of neighbouring species, resulting in unequal contributions of different conditioning species to soil microbial communities and community-level PSFs. As a result, PSFs in species-rich communities cannot be accurately predicted by simply extrapolating from species-level PSFs or assuming additive effects. Consequently, we may have underestimated or overestimated the role of soil biota in maintaining plant diversity and driving alien plant invasion within species-rich plant communities. Therefore, our findings underscore the urgent need to shift the focus of plant-soil interaction research from individual plant species to whole plant communities, particularly in studies addressing plant interactions and community dynamics. Technically, we propose that future PSF experiments should incorporate a conditioning phase in which soil is cultivated by multi-species assemblages that mirror natural communities, followed by a feedback phase to test impacts of conditioned soil on individual plant species or plant communities.\u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data have been deposited to https://doi.org/10.6084/m9.figshare.28395707. DNA sequences have been deposited to the National Center for Biotechnology Information (NCBI, accession number PRJNA1206692).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCode availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll code used to complete analyses for the manuscripts is available at the following link: https://doi.org/10.6084/m9.figshare.28395707. \u0026nbsp;Data analyses were conducted and were visualizations generated in R (version 4.2.2).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to thank Yifan He, Biao Zhu, Wei Chen, and Yanli Zhan for their assistance in the experiment, Ragan M. Callaway, Leho Tedersoo, Marina Semchenko, Mark van Kleunen, Bernhard Schmid, Heinz M\u0026uuml;ller-Sch\u0026auml;rer for comments on the manuscript. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eX.L. conceived and designed the study. M.Y. conducted the literature review and the experiments. C.W., and L.G. performed the field survey. Y. S., and M.Y. performed statistical analysis. X.L., Y.S., \u0026amp; N.X wrote the draft of the manuscript. All authors contributed to revisions and gave final approval for publication.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interest.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis study was financially supported by the National Natural Science Foundation of China (32171585, 32371749 \u0026amp; 32201438).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBever, J. D., Westover, K. M. \u0026amp; Antonovics, J. Incorporating the soil community into plant population dynamics: the utility of the feedback approach. \u003cem\u003eJ. Ecol.\u003c/em\u003e \u003cstrong\u003e85\u003c/strong\u003e, 561-573 (1997).\u003c/li\u003e\n\u003cli\u003eHe, Y.\u003cem\u003e et al.\u003c/em\u003e Leaf litter presence in the non-growing season prolongs plant legacy effects on soil fungal communities and succeeding plant growth. \u003cem\u003eJ. Ecol.\u003c/em\u003e \u003cstrong\u003e111\u003c/strong\u003e, 1997-2009 (2023).\u003c/li\u003e\n\u003cli\u003eHeinen, R.\u003cem\u003e et al.\u003c/em\u003e Plant community composition steers grassland vegetation via soil legacy effects. \u003cem\u003eEcol. Lett.\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 973-982 (2020).\u003c/li\u003e\n\u003cli\u003evan der Putten, W. H., Bradford, M. A., Pernilla Brinkman, E., van de Voorde, T. F. J. \u0026amp; Veen, G. F. Where, when and how plant-soil feedback matters in a changing world. \u003cem\u003eFunct. Ecol.\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 1109-1121 (2016).\u003c/li\u003e\n\u003cli\u003eWandrag, E. M.\u003cem\u003e et al.\u003c/em\u003e Phylogenetic signals and predictability in plant-soil feedbacks. \u003cem\u003eNew Phytol.\u003c/em\u003e \u003cstrong\u003e228\u003c/strong\u003e, 1440-1449 (2020).\u003c/li\u003e\n\u003cli\u003eCallaway, R. M., Thelen, G. C., Rodriguez, A. \u0026amp; Holben, W. E. Soil biota and exotic plant invasion. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e427\u003c/strong\u003e, 731-733 (2004).\u003c/li\u003e\n\u003cli\u003eKlironomos, J. N. Feedback with soil biota contributes to plant rarity and invasiveness in communities. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e417\u003c/strong\u003e, 67-70 (2002).\u003c/li\u003e\n\u003cli\u003eThakur, M. P.\u003cem\u003e et al.\u003c/em\u003e Plant-soil feedbacks and temporal dynamics of plant diversity-productivity relationships. \u003cem\u003eTrends Ecol. Evol.\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 651-661 (2021).\u003c/li\u003e\n\u003cli\u003eTeste, F. P.\u003cem\u003e et al.\u003c/em\u003e Plant-soil feedback and the maintenance of diversity in Mediterranean-climate shrublands. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e355\u003c/strong\u003e, 173-176 (2017).\u003c/li\u003e\n\u003cli\u003eYan, X., Levine, J. M. \u0026amp; Kandlikar, G. S. A quantitative synthesis of soil microbial effects on plant species coexistence. \u003cem\u003eProc. Natl. Acad. Sci. USA\u003c/em\u003e \u003cstrong\u003e119\u003c/strong\u003e, e2122088119 (2022).\u003c/li\u003e\n\u003cli\u003eVan Nuland, M. E., Bailey, J. K. \u0026amp; Schweitzer, J. A. Divergent plant-soil feedbacks could alter future elevation ranges and ecosystem dynamics. \u003cem\u003eNat. Ecol. Evol.\u003c/em\u003e \u003cstrong\u003e1\u003c/strong\u003e, 0150 (2017).\u003c/li\u003e\n\u003cli\u003eDudenh\u0026ouml;ffer, J.-H., Luecke, N. C. \u0026amp; Crawford, K. M. Changes in precipitation patterns can destabilize plant species coexistence via changes in plant-soil feedback. \u003cem\u003eNat. Ecol. Evol.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 546-554 (2022).\u003c/li\u003e\n\u003cli\u003evan der Putten, W. H.\u003cem\u003e et al.\u003c/em\u003e Plant-soil feedbacks: the past, the present and future challenges. \u003cem\u003eJ. Ecol.\u003c/em\u003e \u003cstrong\u003e101\u003c/strong\u003e, 265-276 (2013).\u003c/li\u003e\n\u003cli\u003ede Vries, F., Lau, J., Hawkes, C. \u0026amp; Semchenko, M. Plant-soil feedback under drought: does history shape the future? \u003cem\u003eTrends Ecol. Evol.\u003c/em\u003e \u003cstrong\u003e38\u003c/strong\u003e, 708-718 (2023).\u003c/li\u003e\n\u003cli\u003ePernilla Brinkman, E., Van der Putten, W. H., Bakker, E.-J. \u0026amp; Verhoeven, K. J. F. Plant-soil feedback: experimental approaches, statistical analyses and ecological interpretations. \u003cem\u003eJ. Ecol.\u003c/em\u003e \u003cstrong\u003e98\u003c/strong\u003e, 1063-1073 (2010).\u003c/li\u003e\n\u003cli\u003eToju, H., Guimar\u0026atilde;es, P. R., Olesen, J. M. \u0026amp; Thompson, J. N. Assembly of complex plant-fungus networks. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 5273 (2014).\u003c/li\u003e\n\u003cli\u003eToju, H., Guimar\u0026atilde;es, P. R., Olesen, J. M. \u0026amp; Thompson, J. N. Below-ground plant-fungus network topology is not congruent with above-ground plant-animal network topology. \u003cem\u003eSci. Adv.\u003c/em\u003e \u003cstrong\u003e1\u003c/strong\u003e, e1500291 (2015).\u003c/li\u003e\n\u003cli\u003eKardol, P., Martijn Bezemer, T. \u0026amp; van der Putten, W. H. Temporal variation in plant-soil feedback controls succession. \u003cem\u003eEcol. Lett.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 1080-1088 (2006).\u003c/li\u003e\n\u003cli\u003eMony, C., Gaudu, V., Ricono, C., Jambon, O. \u0026amp; Vandenkoornhuyse, P. Plant neighbours shape fungal assemblages associated with plant roots: a new understanding of niche-partitioning in plant communities. \u003cem\u003eFunct. Ecol.\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 1768-1782 (2021).\u003c/li\u003e\n\u003cli\u003eHu, J.\u003cem\u003e et al.\u003c/em\u003e Neighbourhood effect of weeds on wheat root endospheric mycobiota. \u003cem\u003eJ. Ecol.\u003c/em\u003e \u003cstrong\u003e111\u003c/strong\u003e, 994-1008 (2023).\u003c/li\u003e\n\u003cli\u003eKeesing, F., Holt, R. D. \u0026amp; Ostfeld, R. S. Effects of species diversity on disease risk. \u003cem\u003eEcol. Lett.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 485-498 (2006).\u003c/li\u003e\n\u003cli\u003eWeinhold, A.\u003cem\u003e et al.\u003c/em\u003e Tree species richness differentially affects the chemical composition of leaves, roots and root exudates in four subtropical tree species. \u003cem\u003eJ. Ecol.\u003c/em\u003e \u003cstrong\u003e110\u003c/strong\u003e, 97-116 (2022).\u003c/li\u003e\n\u003cli\u003eTurner, S. C. \u0026amp; Schweitzer, J. A. Plant neighbors differentially alter a focal species\u0026apos; biotic interactions through changes to resource allocation. \u003cem\u003eEcology\u003c/em\u003e \u003cstrong\u003e105\u003c/strong\u003e, e4395 (2024).\u003c/li\u003e\n\u003cli\u003eZhalnina, K.\u003cem\u003e et al.\u003c/em\u003e Dynamic root exudate chemistry and microbial substrate preferences drive patterns in rhizosphere microbial community assembly. \u003cem\u003eNat. Microbiol.\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 470-480 (2018).\u003c/li\u003e\n\u003cli\u003eDelory, B. M., Callaway, R. M. \u0026amp; Semchenko, M. A trait-based framework linking the soil metabolome to plant-soil feedbacks. \u003cem\u003eNew Phytol.\u003c/em\u003e \u003cstrong\u003e241\u003c/strong\u003e, 1910-1921 (2024).\u003c/li\u003e\n\u003cli\u003eBroz, A. K., Manter, D. K. \u0026amp; Vivanco, J. M. Soil fungal abundance and diversity: another victim of the invasive plant \u003cem\u003eCentaurea maculosa\u003c/em\u003e. \u003cem\u003eISME J.\u003c/em\u003e \u003cstrong\u003e1\u003c/strong\u003e, 763-765 (2007).\u003c/li\u003e\n\u003cli\u003eLi, L.\u003cem\u003e et al.\u003c/em\u003e Diversity enhances agricultural productivity via rhizosphere phosphorus facilitation on phosphorus-deficient soils. \u003cem\u003eProc. Natl. Acad. Sci. USA\u003c/em\u003e \u003cstrong\u003e104\u003c/strong\u003e, 11192-11196 (2007).\u003c/li\u003e\n\u003cli\u003eSun, Y.\u003cem\u003e et al.\u003c/em\u003e Plant intraspecific competition and growth stage alter carbon and nitrogen mineralization in the rhizosphere. \u003cem\u003ePlant Cell Environ.\u003c/em\u003e \u003cstrong\u003e44\u003c/strong\u003e, 1231-1242 (2021).\u003c/li\u003e\n\u003cli\u003eCheng, C.\u003cem\u003e et al.\u003c/em\u003e Genotype diversity enhances invasion resistance of native plants via soil biotic feedbacks. \u003cem\u003eEcol. Lett.\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, e14384 (2024).\u003c/li\u003e\n\u003cli\u003eYang, L., Maron, J. L. \u0026amp; Callaway, R. M. Inhibitory effects of soil biota are ameliorated by high plant diversity. \u003cem\u003eOecologia\u003c/em\u003e \u003cstrong\u003e179\u003c/strong\u003e, 519-525 (2015).\u003c/li\u003e\n\u003cli\u003eGuerrero-Ram\u0026iacute;rez, N. R., Reich, P. B., Wagg, C., Ciobanu, M. \u0026amp; Eisenhauer, N. Diversity-dependent plant-soil feedbacks underlie long-term plan diversity effects on primary productivity. \u003cem\u003eEcosphere\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, e02704 (2019).\u003c/li\u003e\n\u003cli\u003eRen, H.\u003cem\u003e et al.\u003c/em\u003e Diversity-conditioned soil strengthens plant diversity-productivity relationship. \u003cem\u003eJ. Appl. Ecol.\u003c/em\u003e \u003cstrong\u003e60\u003c/strong\u003e, 1881-1892 (2023).\u003c/li\u003e\n\u003cli\u003eForero, L. E., Kulmatiski, A., Grenzer, J. \u0026amp; Norton, J. M. Plant-soil feedbacks help explain biodiversity-productivity relationships. \u003cem\u003eCommun. Biol.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 789 (2021).\u003c/li\u003e\n\u003cli\u003eGao, L.\u003cem\u003e et al.\u003c/em\u003e Latitudinal variation in the diversity and composition of various organisms associated with an exotic plant: the role of climate and plant invasion. \u003cem\u003eNew Phytol.\u003c/em\u003e \u003cstrong\u003e231\u003c/strong\u003e, 1559-1569 (2021).\u003c/li\u003e\n\u003cli\u003eWu, S.\u003cem\u003e et al.\u003c/em\u003e Soil organic matter dynamics mediated by arbuscular mycorrhizal fungi-an updated conceptual framework. \u003cem\u003eNew Phytol.\u003c/em\u003e \u003cstrong\u003e242\u003c/strong\u003e, 1417-1425 (2024).\u003c/li\u003e\n\u003cli\u003eBanerjee, S. \u0026amp; van der Heijden, M. G. A. Soil microbiomes and one health. \u003cem\u003eNat. Rev. Microbiol.\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 6-20 (2023).\u003c/li\u003e\n\u003cli\u003eHannula, S. E.\u003cem\u003e et al.\u003c/em\u003e Persistence of plant-mediated microbial soil legacy effects in soil and inside roots. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 5686 (2021).\u003c/li\u003e\n\u003cli\u003eEppinga, M. B.\u003cem\u003e et al.\u003c/em\u003e Frequency-dependent feedback constrains plant community coexistence. \u003cem\u003eNat. Ecol. Evol.\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 1403-1407 (2018).\u003c/li\u003e\n\u003cli\u003eBever, J. D. Soil community feedback and the coexistence of competitors: conceptual frameworks and empirical tests. \u003cem\u003eNew Phytol.\u003c/em\u003e \u003cstrong\u003e157\u003c/strong\u003e, 465-473 (2003).\u003c/li\u003e\n\u003cli\u003eZhang, Z., Liu, Y., Brunel, C. \u0026amp; van Kleunen, M. Soil-microorganism-mediated invasional meltdown in plants. \u003cem\u003eNat. Ecol. Evol.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 1612-1621 (2020).\u003c/li\u003e\n\u003cli\u003eZhang, Z., Liu, Y., Brunel, C. \u0026amp; van Kleunen, M. Evidence for Elton\u0026apos;s diversity-invasibility hypothesis from belowground. \u003cem\u003eEcology\u003c/em\u003e \u003cstrong\u003e101\u003c/strong\u003e, e03187 (2020).\u003c/li\u003e\n\u003cli\u003eWardle, D. A. Is \u0026quot;sampling effect\u0026quot; a problem for experiments investigating biodiversity-ecosystem function relationships? \u003cem\u003eOikos\u003c/em\u003e \u003cstrong\u003e87\u003c/strong\u003e, 403-407 (1999).\u003c/li\u003e\n\u003cli\u003eLoreau, M. \u0026amp; Hector, A. Partitioning selection and complementarity in biodiversity experiments. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e412\u003c/strong\u003e, 72-76 (2001).\u003c/li\u003e\n\u003cli\u003eKlein, T., Siegwolf, R. T. W. \u0026amp; K\u0026ouml;rner, C. Belowground carbon trade among tall trees in a temperate forest. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e352\u003c/strong\u003e, 342-344 (2016).\u003c/li\u003e\n\u003cli\u003eFitzpatrick, C. R.\u003cem\u003e et al.\u003c/em\u003e Assembly and ecological function of the root microbiome across angiosperm plant species. \u003cem\u003eProc. Natl. Acad. Sci. USA\u003c/em\u003e \u003cstrong\u003e115\u003c/strong\u003e, E1157-E1165 (2018).\u003c/li\u003e\n\u003cli\u003eSweeney, C. J., de Vries, F. T., van Dongen, B. E. \u0026amp; Bardgett, R. D. Root traits explain rhizosphere fungal community composition among temperate grassland plant species. \u003cem\u003eNew Phytol.\u003c/em\u003e \u003cstrong\u003e229\u003c/strong\u003e, 1492-1507 (2021).\u003c/li\u003e\n\u003cli\u003eLevine, J. M., Bascompte, J., Adler, P. B. \u0026amp; Allesina, S. Beyond pairwise mechanisms of species coexistence in complex communities. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e546\u003c/strong\u003e, 56-64 (2017).\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eLiterature synthesis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe performed a literature search on the 3rd of December 2024 in Web of Science with the search string: \u0026lsquo;soil feedback\u0026rsquo; or \u0026lsquo;soil legacy effect\u0026rsquo; or \u0026lsquo;soil effect\u0026rsquo; and \u0026lsquo;plant\u0026rsquo;, restricted to articles written in English. This search yielded 1,602 articles. We then performed an additional search using the terms: \u0026lsquo;plant-soil feedback\u0026rsquo; and \u0026lsquo;meta\u0026rsquo;, resulting in 25 meta-analysis articles. Subsequently, we pooled all 1,602 articles and all the references cited in these meta-analysis articles \u003csup\u003e10,48-71\u003c/sup\u003e, removing duplicates, and obtained a total of 2,502 articles.\u003c/p\u003e\n\u003cp\u003eWe selected articles from the 2,502 articles based on the following criteria: 1) experimental studies only, excluding synthesis papers and meta-analyses; 2) the experiment contains both conditioning (i.e., soils conditioned by plants or plant communities) and feedback phases; and 3)\u0026nbsp;the article clearly described the number of conditioning species for each experimental unit (e.g., pot or plot).\u0026nbsp;Studies investigating the legacy effects of plant communities in the field without species-level information were excluded, as were studies examining the effects of genetic diversity within a single species on PSFs. In total, 460 experiments from 451 articles met these criteria and were included in our data synthesis (Table S1).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eField survey\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo explore the number of herbaceous plant species that are likely to interact belowground in nature, we performed a comprehensive field survey between October and November over four years, 2017 (49 sites, 245 plots), 2019 (128 sites, 389 plots), 2020 (97sites, 503 plots) and 2021 (168 sites, 840 plots), covering a latitudinal range from 21.5\u0026deg;N to 36.7\u0026deg;N, for a total of 442 randomly selected sites (Fig. 1C) and 1,977 plots (0.5 m \u0026times; 0.5 m). These sites were in natural ecosystems (e.g., river banks) or abandoned farmlands. Species richness did not differ between natural ecosystems and farmlands (linear model:\u0026nbsp;\u0026chi;\u003csup\u003e2\u003c/sup\u003e=1.98, \u003cem\u003eP\u003c/em\u003e=0.1593). At each site, we randomly established three to five 0.5 m \u0026times; 0.5 m quadrats (\u0026ge; 2 m apart), at which spatial scale co-occurring herbaceous species are likely to interact with each other belowground in grasslands\u0026nbsp;\u003csup\u003e26,34\u003c/sup\u003e and thus jointly shaping community-level PSFs. No big trees (height \u0026gt; 1 m) within 10 meters. We morphologically identified all plant species in situ and recorded the number of species within each plot, as in Gao, et al.\u0026nbsp;\u003csup\u003e34\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNeighbouring effect experiment: effects of neighbouring plant richness on the rhizosphere fungal communities of three species\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo explore how neighbouring species richness affects the rhizosphere fungal communities of a plant within plant communities, we performed a greenhouse experiment on the campus of\u0026nbsp;Huazhong Agricultural University (HZAU) in Wuhan, China\u0026nbsp;(30\u0026deg;47\u0026prime; N, 110\u0026deg;35\u0026prime; E).\u0026nbsp;The average minimum and maximum temperatures are 1.0 \u0026deg;C\u0026nbsp;in January and 32.9 \u0026deg;C in July, and\u0026nbsp;the\u0026nbsp;mean annual precipitation is 1,316.0 mm (www.nmc.cn).\u0026nbsp;We focus on the rhizosphere communities of \u003cem\u003eSetaria viridis\u003c/em\u003e (L.) P. Beauv., \u003cem\u003eCelosia argentea\u0026nbsp;\u003c/em\u003eL., and \u003cem\u003eSenna occidentalis\u003c/em\u003e (L.) Link (hereafter referred to as focal species). In addition, we chose \u003cem\u003eChenopodium album\u003c/em\u003e L. \u003cem\u003e(\u003c/em\u003eAmaranthaceae), \u003cem\u003ePerilla frutescens\u003c/em\u003e (L.) Britton\u003cem\u003e\u0026nbsp;\u003c/em\u003e(Lamiaceae), \u003cem\u003eSigesbeckia orientalis\u003c/em\u003e L. (Asteraceae) and \u003cem\u003eUrena lobata\u003c/em\u003e L. (Malvaceae) as neighbouring species to perform this experiment. All these species frequently co-occur in the natural fields.\u003c/p\u003e\n\u003cp\u003eIn October 2021, for each plant species, we collected seeds from three sites (\u0026gt; 10 km apart from each other) in Wuhan, pooled, and stored the seeds at 4 \u0026deg;C and low relative humidity (\u0026lt; 30%). This experiment is a factorial, randomized block design and was performed in a greenhouse at the HZAU campus. In March 2022,\u0026nbsp;we germinated the seeds of these species by sowing surface-sterilized seeds in trays filled with a mixture of gamma-irradiated (dose \u0026gt; 25 kGy) peat and sand (2:1 by volume). The trays were placed randomly in a greenhouse on the HZAU campus. Seeds were surface sterilized by immersing in 70% ethanol for 30 seconds, followed by rinsing in demineralized water twice. Germinated seedlings were watered every other day. At the same time, we collected topsoil (0-10 cm depth) from a recently abandoned field near the HZAU campus. All the seven species had not occurred in the field before. In a lab, we sieved and homogenized the soil (2 mm), and then potted soil into 8 L experimental pots (diameter: 27 cm, height: 18 cm).\u003c/p\u003e\n\u003cp\u003eIn late April 2022, an individual of a focal plant species (hereafter: focal plant) was planted at the centre of each experimental pot. Considering the potential effects of distance on the focal plants\u0026apos; rhizosphere fungi, we planted three co-occur individuals (either the same\u0026nbsp;as the focal species or three of one of four other species: \u003cem\u003eC. album\u003c/em\u003e, \u003cem\u003eP. frutescens\u003c/em\u003e, \u003cem\u003eS. orientalis\u003c/em\u003e or \u003cem\u003eU. lobate\u003c/em\u003e)\u0026nbsp;5 cm away from the focal plant (Fig. S1). To test the effects of neighbouring species richness on a focal plant\u0026rsquo;s rhizosphere fungi,\u0026nbsp;eight neighbouring plants planted 10 cm away from the focal plant were from 1 (8 each), 2 (4 each) or 4 (2 each) species from a pool of the focal plant species plus the four other species (Fig. S1).\u0026nbsp;There were 25, 50 and 25 replicates, respectively, of the 1, 2 and 4 species diversity levels for each focal species. In total, we planted all the combinations for each focal plant species, resulting in 300 pots (3 focal species\u0026thinsp;\u0026times;\u0026thinsp;100 neighbouring richness treatments).\u0026nbsp;Similar-sized seedlings were selected for each species to ensure seedlings were planted at the same growth stage (i.e., similar height). Seedlings that died within one week after transplanting were replaced by new ones. The 300 experimental\u0026nbsp;pots were randomly placed in the greenhouse and repositioned (40 cm apart) monthly. Each pot was watered with 500\u0026thinsp;mL water every other day with glass beakers to avoid cross-contamination through water splash.\u003c/p\u003e\n\u003cp\u003eFive months later, for each focal individual, we gently shook the roots for 10 min to remove the bulk soil and then sampled the soil that still adhered to the taproot surface (rhizosphere soil) with a brush. After that, we put the rhizosphere soil into a 2 mL sterile centrifuge tube and stored the samples at -80℃ for DNA extraction. In 300 pots, 20 focal plants died.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eFungi DNA sequencing\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe sequenced the rhizosphere fungal communities of each focal plant, as described in\u0026nbsp;Gao,\u0026nbsp;et al.\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003csup\u003e34\u003c/sup\u003e. Soil microbial DNA was extracted using cetyltrimethylammonium bromide method \u003csup\u003e72\u003c/sup\u003e. DNA quality, concentration, and purification were checked by 1% agarose gel electrophoresis. The fungal ITS1 region was amplified with primers ITS1-F (5\u0026rsquo;-CTTGGTCATTTAGAGGAAGTAA-3\u0026rsquo;) and ITS2 (5\u0026rsquo;-GCTGCGTTCTTCATCGATGC-3\u0026rsquo;) \u003csup\u003e73,74\u003c/sup\u003e. Polymerase chain reaction (PCR) was carried out with 15 \u0026micro;L of Phusion\u0026reg; High-Fidelity PCR Master Mix (New England Biolabs),\u0026nbsp;2 \u0026micro;M of forward and reverse primers, and 10 ng of template DNA.\u0026nbsp;The PCR amplification included an initial denaturation step at 98℃ for 1 min, followed by 30 cycles of denaturation at 98℃ (10 s), annealing at 50℃ (30 s), extension a 72℃ (30 s), and a final extension at 72\u0026deg;C (5 min). PCR products were checked on a 2% agarose gel to determine amplification success and the relative intensity of bands. Multiple PCR samples were pooled together. Pooled samples were extracted and purified using the Qiagen Gel Extraction Kit (Qiagen, Germany). DNA libraries were prepared using TruSeq\u0026reg; DNA PCR-Free Sample Preparation Kit for Illumina\u0026nbsp;(New England Biolabs, USA)\u0026nbsp;following the manufacturer\u0026rsquo;s protocol, and PE250 sequencing was performed on a NovaSeq 6000 Sequencing System\u0026nbsp;(Illumina, Inc. San Diego, USA) in Novogene Company.\u003c/p\u003e\n\u003cp\u003eStrict quality control steps were applied to the sequencing data. Briefly, after the barcode and primer sequences were trimmed from reads, paired-end reads were merged using FLASH (version 1.2.7) \u003csup\u003e75\u003c/sup\u003e, and the joint pairs were quality filtered in QIIME (version 1.9.l) \u003csup\u003e76\u003c/sup\u003e, following the standard method. All chimeric sequences were removed independently of the Unite database (v8.2) using the UCHIME algorithm \u003csup\u003e77\u003c/sup\u003e. OTUs were categorized at 97% sequence similarity using the UPARSE pipeline (version 7.0.1001), and taxonomic assignment for fungi was performed using the Unite database\u0026nbsp;(v8.2) via the BLAST method in QIIME (Version 1.9.1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn total, there were 21,883,668 reads of 11,851 fungal OTUs, and the number of OTU reads for each sample ranged from 46,985 to 90,589. Fungal ITS1 sequences were blasted against the GenBank nucleotide database. Subsequently, we assigned fungi to plant pathogens or arbuscular mycorrhizal fungi (AMF) based on the FUNGuild database \u003csup\u003e78\u003c/sup\u003e. As many species in the \u003cem\u003eFusarium\u003c/em\u003e genus are major plant pathogens \u003csup\u003e79\u003c/sup\u003e, we assigned OTUs belonging to this genus as plant pathogens.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003ePlant-soil feedback (PSF) experiment: the effects of conditioning species richness on PSFs\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo explore how the identity and richness of conditioning plant species affect plant-soil feedbacks (PSFs), we performed a two-stage PSF experiment from 2021 to 2022 in the same greenhouse as the \u0026lsquo;Neighbouring effects experiment\u0026rsquo;. In October 2020, we collected seeds from 28 common plant species belonging to nine families, of which 25 are annuals and three are perennials (which cannot survive the cold winters in Wuhan) (Table S2). All the plant species used in the \u0026lsquo;neighbouring effect experiment\u0026rsquo; were used in this experiment. For each species, we collected seeds from three sites (\u0026gt; 10 km apart), pooled and stored at 4 \u0026deg;C and low relative humidity (\u0026lt; 30%).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eConditioning phase\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on the field survey (Fig. 1D, plant richness at plot level), we established a plant richness gradient of 1, 2, 4, 6, or 12 species and constructed plant communities randomly using the 28 plant species (Fig. S2). To conduct the experiment, we germinated the seeds of these species with the same method as in the\u0026nbsp;\u0026lsquo;Neighbouring effect experiment\u0026rsquo;, and collected topsoil (0-10 cm depth) from the same abandoned field at the end of June 2021 in a greenhouse.\u0026nbsp;None of the 28 species had occurred in the field before. In another greenhouse (the one where the \u0026lsquo;neighbouring effect experiment\u0026rsquo; was performed), we sieved (2 mm mesh), homogenized, and potted the soil into 8 L experimental pots.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;In late July 2021, we planted similarly sized seedlings (~ 8 cm in height) into experimental pots at a density of 12 individuals per pot (~5 cm apart). Within each pot, the species were randomly positioned, ensuring that the number of individuals was equal for each species (e.g., for two species, six individuals per species). Monocultures of each species were repeated five times (28 species \u0026times; 5 replicates), and each richness level had 36 replicates (4 richness levels \u0026times; 36 replicates), resulting in a total of 284 pots. The experimental pots were spaced 40 cm apart and randomly repositioned in the greenhouse monthly. Plants were watered with 500 mL/pot of water every other day throughout the trial.\u003c/p\u003e\n\u003cp\u003eIn November 2021, for each pot, we harvested and weighed each species\u0026apos; aboveground biomass (after drying at 80 \u0026deg;C for 48 hours) separately. Plant roots were left in the soil to decompose (almost all plant roots decomposed before the onset of the feedback experiment) and reinforce the effects of living individuals in the non-growing season, as in He, et al.\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e. To minimize potential air-mediated influences between neighbouring pots, all pots were sealed with tinfoil, with nine one-cm-diameter holes on top for watering. Each pot received 500 mL of water monthly during the winter (November to February) and weekly during the spring (March to April). \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eFeedback phase\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn mid-April 2022, we began the feedback experiment in the same greenhouse. In this phase, we used an identical plant community as the responding community\u0026nbsp;(Fig. S1), composed of\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003cem\u003ePaspalum\u0026nbsp;conjugatum\u003c/em\u003e P.J.Bergius (Poaceae),\u0026nbsp;\u003cem\u003eSetaria viridis\u0026nbsp;\u003c/em\u003e(L.) P. Beauv. (Poaceae),\u003cem\u003e\u0026nbsp;Achyranthes aspera\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003eL. (Amaranthaceae),\u003cem\u003e\u0026nbsp;Amaranthus hybridus\u0026nbsp;\u003c/em\u003eL. (Amaranthaceae), \u003cem\u003eSenna tora\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e(L.) Roxb.\u0026nbsp;(Fabaceae) and \u003cem\u003eSenna occidentalis\u0026nbsp;\u003c/em\u003e(L.) Link\u0026nbsp;(Fabaceae). We mixed 2 L of conditioned soil from each of the 284 conditioning pots with 6 L of gamma-irradiated field-collected topsoil from the same source as in the conditioning phase. The mixtures were then potted into 8 L experimental pots. Additionally, five experimental pots were filled with only sterilized soil\u0026nbsp;as a control (Fig. S2). This resulted in a total of 289 experimental pots, which were\u0026nbsp;randomly repositioned\u0026nbsp;in the greenhouse (40 cm apart)\u0026nbsp;every month.\u003c/p\u003e\n\u003cp\u003eIn April, we planted similarly-sized seedlings of the six plant species in each pot (two individuals per species). The seedlings of the six species were germinated in a greenhouse with the same methods as in the conditioning phase. In each pot, plant individuals were randomly positioned, with a 5 cm spacing between them. Dead seedlings were replaced within one week. Plants were watered with 500 mL water/pot every other day throughout the trial. Four months later, for each pot, we harvested and weighed (after drying at 80\u0026nbsp;\u0026deg;C\u0026nbsp;for 48 hours) the aboveground biomass of each species separately. Because the roots of the plants were severely intertwined, we collected, washed and weighed (after drying at 80\u0026nbsp;\u0026deg;C\u0026nbsp;for 48 hours) the entire root mass for each pot.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll the data analyses were performed with R software (version\u0026nbsp;4.2.2). For multiple-variable analysis, \u003cem\u003eP\u003c/em\u003e values were adjusted using the Bonferroni correction method \u003csup\u003e80\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eEffect of neighbouring plant richness on the rhizosphere fungal communities of the three focal species\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBefore data analyses, we removed the OTUs with a maximum of 10 reads across all samples or observed in fewer than three samples to eliminate potential sequencing artifacts \u003csup\u003e81\u003c/sup\u003e. We then rarefied the reads to\u0026nbsp;46,985 reads\u0026nbsp;per sample, and\u0026nbsp;log\u003csub\u003e2\u003c/sub\u003e(x+1)-transformed the OTU-abundance data to improve the normality of the residues.\u003c/p\u003e\n\u003cp\u003eWe chose Shannon diversity and the inverse Simpson index as alpha diversity indexes. We calculated these variables for the overall, putative pathogenic fungal and AMF fungi from rarefied counts separately with the \u003cem\u003evegan\u003c/em\u003e package \u003csup\u003e82\u003c/sup\u003e. We performed linear mixed models (LMMs) to explore the dependence of these variables on the fixed factors \u0026lsquo;the identity of focal species\u0026rsquo; (FS), \u0026lsquo;the richness of neighbouring species\u0026rsquo; (NR), their interaction (FS:NR), and the random factor \u0026lsquo;co-occur plat species type (COT, same vs. differ)\u0026rsquo;. We included \u0026lsquo;COT\u0026rsquo; as a random factor in all analyses because neither this factor nor its interactions with other variables were statistically significant for the overall and AMF fungal communities (adjusted \u003cem\u003eP\u003c/em\u003e-values\u0026nbsp;\u0026sup3;\u0026nbsp;0.2 in all cases, Table S8). Moreover, as this factor and its interactions with others are not central to our research focus, this method allows us to concentrate on the main variables of interest by taking into account variability related to \u0026lsquo;co-occur plat species type\u0026rsquo; without explicitly modelling its effects or interactions. These analyses were performed with the glmmTMB package, which uses maximum likelihood to estimate model parameters with a Gaussian distribution \u003csup\u003e83\u003c/sup\u003e. We checked the normality of the residuals in all models using QQ-plots. We quantified\u0026nbsp;the individual contributions of\u0026nbsp;each factor\u0026nbsp;in explaining the variation in rhizosphere fungal diversity using the\u0026nbsp;MuMIn package\u0026nbsp;\u003csup\u003e84\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eWe performed redundancy analysis (RDA) to explore the effects of focal species identity, the richness of neighbouring species and their interactions on the composition of rhizosphere fungal communities (DESeq2\u0026nbsp;normalized OTU abundance), using the \u003cem\u003evegan\u003c/em\u003e package. After that, we performed\u0026nbsp;hierarchical partitioning analyses to calculate the individual contributions of each factor in explaining the compositional variation in rhizosphere fungal communities. The above analyses were performed for the overall fungal community, the putative pathogenic and AMF fungal communities separately.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eWhether community-level PSFs be predicted by PSFs of composing species\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSince some plant species in the responding plant community were also present during the conditioning phase, we first examined whether their presence and interactions with species richness in the conditioning phase affected the total biomass of the responding community. To assess this, we fitted a linear model with the total biomass of the responding community as the response variable and species richness, individual species presence, and their interactions as predictors. Residual diagnostics were checked using QQ plots, and statistical significance was evaluated using ANOVA. All analyses were statistically non-significant (\u003cem\u003eP\u003c/em\u003e-values \u0026ge; 0.592 in all cases, Table S9). We thus removed the potential confounding effects of species co-occurrence between the conditioning and responding phases in subsequent analyses.\u003c/p\u003e\n\u003cp\u003eIn addition,\u0026nbsp;we tested whether the biomass of responding community is strongly affected by the phylogenetic distance between conditioning species or communities and the responding community. To do this, we generated a phylogenic tree of the 28 plant species based on the ITS, matK and rbcL sequences downloaded from NCBI Genbank, with the maximum likelihood method in IQ-tree software \u003csup\u003e85\u003c/sup\u003e. \u003cem\u003eAmborella trichopoda\u0026nbsp;\u003c/em\u003ewas selected\u003cem\u003e\u0026nbsp;\u003c/em\u003eas the outgroup\u003cem\u003e,\u0026nbsp;\u003c/em\u003eas it is the unique sister species to all other extant angiosperms \u003csup\u003e86\u003c/sup\u003e.\u003cem\u003e\u0026nbsp;\u003c/em\u003eWe estimated\u003cem\u003e\u0026nbsp;\u003c/em\u003eUniFrac distance between the conditioning and responding communities. A linear model revealed that the biomass of the responding community was weekly (r = 0.14, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= 0.016) correlated with the uniFrac distance between the conditioning and responding communities. Given this weak correlation, phylogenetic distances between the conditioning and responding communities was not considered in the subsequent analyses.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe quantified the mean and 95% confidence intervals (CIs) of biomass of the responding community for each conditioning species or community using a non-parametric bootstrap (1,000 iterations for each species). We quantified the PSFs as the biomass of the responding communities in conditioned soils compared to those in sterilized soils. For each conditioning plant community, we summed the PSFs of each composing species based on three null assumptions: additive (summing effects), multiplicative (combining proportional changes), and dominative (choosing the strongest effect), following the protocol in Rillig, et al. \u003csup\u003e87\u003c/sup\u003e.\u0026nbsp;After that, we compared the predicted value to the observed PSFs across plant communities of differing richness. The goal was to check if any of these assumptions could predict community-PSFs based on PSFs of composing species. If not, it would suggest that community-level PSFs are unpredictable due to higher-order interactions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo investigate how the plant community richness explains variability (R\u003csup\u003e2\u003c/sup\u003e) in the community-level PSFs, and to assess the influence of incorporating the identity of composing species and effect sizes on predictability, we modelled community-level PSFs using species richness as a baseline predictor.\u0026nbsp;The model was then refined by adding the identity of composing species as an explanatory variable. Alternatively, effect sizes derived from predicted community-PSFs under the three assumptions (additive, multiplicative, and dominative) were included in place of the species identity, providing expected outcomes for community-level PSFs under each assumption. We performed random forest models with \u003cem\u003ecforest\u0026nbsp;\u003c/em\u003ein the \u003cem\u003eparty\u0026nbsp;\u003c/em\u003epackages to yield unbiased variable importance and more reliable predictions due to its use of unbiased trees \u003csup\u003e88\u003c/sup\u003e. The model used 1,000 trees and specified five as the number of variables randomly selected at each node. Bootstrap resampling was applied to estimate 95% CIs, and the proportion of variance explained by the mode (R\u003csup\u003e2\u003c/sup\u003e) was estimated by the \u003cem\u003epostResample\u0026nbsp;\u003c/em\u003efunction in the \u003cem\u003ecaret\u0026nbsp;\u003c/em\u003epackages.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eReferences\u003c/p\u003e\n\u003cp\u003e48\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Busetto, R., Breschi, V. \u0026amp; Formentin, S. Meta-learning for model-reference data-driven control. \u003cem\u003eAutomatica\u003c/em\u003e \u003cstrong\u003e172\u003c/strong\u003e, 112006 (2025).\u003c/p\u003e\n\u003cp\u003e49\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Jiang, F.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Global patterns and drivers of plant-soil microbe interactions. \u003cem\u003eEcol. Lett.\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, e14364 (2024).\u003c/p\u003e\n\u003cp\u003e50\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Hager, H. A., Gailis, M. \u0026amp; Newman, J. A. Allelopathic effects of \u003cem\u003eEpichloe\u003c/em\u003e fungal endophytes: experiment and meta-analysis. \u003cem\u003ePlant Soil\u003c/em\u003e \u003cstrong\u003e488\u003c/strong\u003e, 217-232 (2023).\u003c/p\u003e\n\u003cp\u003e51\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Bates, S. E., Wandrag, E. M. \u0026amp; Duncan, R. P. Calculating the uncertainty associated with log response ratios in plant-soil feedback studies. \u003cem\u003ePlant Ecol.\u003c/em\u003e \u003cstrong\u003e221\u003c/strong\u003e, 829-836 (2020).\u003c/p\u003e\n\u003cp\u003e52\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Hassan, K., Dastogeer, K. M. G., Carrillo, Y. \u0026amp; Nielsen, U. N. Climate change-driven shifts in plant-soil feedbacks: a meta-analysis. \u003cem\u003eEcol. Process\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 64 (2022).\u003c/p\u003e\n\u003cp\u003e53\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Suding, K. N.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Consequences of plant-soil feedbacks in invasion. \u003cem\u003eJ. Ecol.\u003c/em\u003e \u003cstrong\u003e101\u003c/strong\u003e, 298-308 (2013).\u003c/p\u003e\n\u003cp\u003e54\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Veresoglou, S. D., Li, G. C., Chen, J. \u0026amp; Johnson, D. Direction of plant-soil feedback determines plant responses to drought. \u003cem\u003eGlob. Change Biol.\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 3995-3997 (2022).\u003c/p\u003e\n\u003cp\u003e55\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Reinhart, K. O.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Globally, plant-soil feedbacks are weak predictors of plant abundance. \u003cem\u003eGlob. Change Biol.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 1756-1768 (2021).\u003c/p\u003e\n\u003cp\u003e56\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Kutakova, E., Herben, T. \u0026amp; Munzbergova, Z. Heterospecific plant-soil feedback and its relationship to plant traits, species relatedness, and co-occurrence in natural communities. \u003cem\u003eOecologia\u003c/em\u003e \u003cstrong\u003e187\u003c/strong\u003e, 679-688 (2018).\u003c/p\u003e\n\u003cp\u003e57\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Wang, H.-R., Zhang, J.-M., Zhao, X.-Y. \u0026amp; Feng, F. N limit as a switch node between positive and negative plant-soil feedback: a meta-analysis based on the covariant diagnosis of plant growth and soil factors. \u003cem\u003eEcotoxicol. Environ. Saf.\u003c/em\u003e \u003cstrong\u003e237\u003c/strong\u003e, 113557 (2022).\u003c/p\u003e\n\u003cp\u003e58\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Waring, B. G.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Pervasive and strong effects of plants on soil chemistry: a meta-analysis of individual plant \u0026apos;Zinke\u0026apos; effects. \u003cem\u003eProc. R. Soc. B: Biol. Sci.\u003c/em\u003e \u003cstrong\u003e282\u003c/strong\u003e, 91-98 (2015).\u003c/p\u003e\n\u003cp\u003e59\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Brinkman, E. P., Van der Putten, W. H., Bakker, E.-J. \u0026amp; Verhoeven, K. J. F. Plant-soil feedback: experimental approaches, statistical analyses and ecological interpretations. \u003cem\u003eJ. Ecol.\u003c/em\u003e \u003cstrong\u003e98\u003c/strong\u003e, 1063-1073 (2010).\u003c/p\u003e\n\u003cp\u003e60\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;De Long, J. R.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Plant-soil feedback: incorporating untested influential drivers and reconciling terminology. \u003cem\u003ePlant Soil\u003c/em\u003e \u003cstrong\u003e485\u003c/strong\u003e, 7-43 (2023).\u003c/p\u003e\n\u003cp\u003e61\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Meisner, A.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Plant-soil feedbacks of exotic plant species across life forms: a meta-analysis. \u003cem\u003eBiol. Invasions\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 2551-2561 (2014).\u003c/p\u003e\n\u003cp\u003e62\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Beals, K. K.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Predicting plant-soil feedback in the field: meta-analysis reveals that competition and environmental stress differentially influence PSF. \u003cem\u003eFront. Ecol. Evol.\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 191 (2020).\u003c/p\u003e\n\u003cp\u003e63\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Mehrabi, Z. \u0026amp; Tuck, S. L. Relatedness is a poor predictor of negative plant-soil feedbacks. \u003cem\u003eNew Phytol.\u003c/em\u003e \u003cstrong\u003e205\u003c/strong\u003e, 1071-1075 (2015).\u003c/p\u003e\n\u003cp\u003e64\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Lekberg, Y.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Relative importance of competition and plant-soil feedback, their synergy, context dependency and implications for coexistence. \u003cem\u003eEcol. Lett.\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 1268-1281 (2018).\u003c/p\u003e\n\u003cp\u003e65\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Refsland, T. K.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Synthesis of plant-soil feedback effects on eastern North American tree species: implications for climate-adaptive forestry. \u003cem\u003eFront. Ecol. Evol.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 1073724 (2023).\u003c/p\u003e\n\u003cp\u003e66\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Comita, L. S.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Testing predictions of the Janzen-Connell hypothesis: a meta-analysis of experimental evidence for distance- and density-dependent seed and seedling survival. \u003cem\u003eJ. Ecol.\u003c/em\u003e \u003cstrong\u003e102\u003c/strong\u003e, 845-856 (2014).\u003c/p\u003e\n\u003cp\u003e67\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Cheng, C., Gundale, M. J., Li, B. \u0026amp; Wu, J. Deciphering the drivers of plant-soil feedbacks and their context-dependence: a meta-analysis. \u003cem\u003ePlant Soil\u003c/em\u003e 10.1007/s11104-024-06922-1 (2024).\u003c/p\u003e\n\u003cp\u003e68\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Kulmatiski, A., Beard, K. H., Stevens, J. R. \u0026amp; Cobbold, S. M. Plant-soil feedbacks: a meta-analytical review. \u003cem\u003eEcol. Lett.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 980-992 (2008).\u003c/p\u003e\n\u003cp\u003e69\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Xi, N.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Relationships between plant-soil feedbacks and functional traits. \u003cem\u003eJ. Ecol.\u003c/em\u003e \u003cstrong\u003e109\u003c/strong\u003e, 3411-3423 (2021).\u003c/p\u003e\n\u003cp\u003e70\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Allen, W. J., Sapsford, S. J. \u0026amp; Dickie, I. A. Soil sample pooling generates no consistent inference bias: a meta-analysis of 71 plant-soil feedback experiments. \u003cem\u003eNew Phytol.\u003c/em\u003e \u003cstrong\u003e231\u003c/strong\u003e, 1308-1315 (2021).\u003c/p\u003e\n\u003cp\u003e71\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Crawford, K. M.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e When and where plant-soil feedback may promote plant coexistence: a meta-analysis. \u003cem\u003eEcol. Lett.\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 1274-1284 (2019).\u003c/p\u003e\n\u003cp\u003e72\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Doyle, J. J. \u0026amp; Doyle, J. L. A rapid DNA isolation procedure for small amounts of fresh leaf tissue. \u003cem\u003ePhytochem. Bull.\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 11-15 (1987).\u003c/p\u003e\n\u003cp\u003e73\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;White, T. J., Bruns, T., Lee, S. \u0026amp; Taylor, J. in \u003cem\u003ePCR Protocols\u003c/em\u003e\u0026nbsp; \u003ca href=\"https://doi.org/10.1016/B978-0-12-372180-8.50042-1\"\u003ehttps://doi.org/10.1016/B978-0-12-372180-8.50042-1\u003c/a\u003e\u0026nbsp; (eds Michael A. Innis, David H. Gelfand, John J. Sninsky, \u0026amp; Thomas J. White) \u0026nbsp; 315-322 (Academic Press, 1990).\u003c/p\u003e\n\u003cp\u003e74\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Gardes, M. \u0026amp; Bruns, T. D. ITS primers with enhanced specificity for basidiomycetes--application to the identification of mycorrhizae and rusts. \u003cem\u003eMol. Ecol.\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 113-118 (1993).\u003c/p\u003e\n\u003cp\u003e75\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Magoč, T. \u0026amp; Salzberg, S. L. FLASH: fast length adjustment of short reads to improve genome assemblies. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 2957-2963 (2011).\u003c/p\u003e\n\u003cp\u003e76\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Caporaso, J. G.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e QIIME allows analysis of high-throughput community sequencing data. \u003cem\u003eNat. Methods\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 335-336 (2010).\u003c/p\u003e\n\u003cp\u003e77\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Edgar, R. C., Haas, B. J., Clemente, J. C., Quince, C. \u0026amp; Knight, R. UCHIME improves sensitivity and speed of chimera detection. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 2194-2200 (2011).\u003c/p\u003e\n\u003cp\u003e78\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Nguyen, N. H.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e FUNGuild: an open annotation tool for parsing fungal community datasets by ecological guild. \u003cem\u003eFunct. Ecol.\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 241-248 (2016).\u003c/p\u003e\n\u003cp\u003e79\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Ma, L.-J.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e Fusarium pathogenomics. \u003cem\u003eAnnu. Rev. Microbiol.\u003c/em\u003e \u003cstrong\u003e67\u003c/strong\u003e, 399-416 (2013).\u003c/p\u003e\n\u003cp\u003e80\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Simes, R. J. \u0026amp; Simes, B. R. J. An improved Bonferroni procedure for multiple tests of significance. \u003cem\u003eBiometrika\u003c/em\u003e \u003cstrong\u003e73\u003c/strong\u003e, 751-754 (1986).\u003c/p\u003e\n\u003cp\u003e81\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Gomez-Alvarez, V., Teal, T. K. \u0026amp; Schmidt, T. M. Systematic artifacts in metagenomes from complex microbial communities. \u003cem\u003eISME J.\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 1314-1317 (2009).\u003c/p\u003e\n\u003cp\u003e82\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Oksanen, J.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e vegan: community ecology package. \u003cem\u003eThe Comprehensive R Archive Network\u003c/em\u003e doi: 0.32614/CRAN.package.vegan (2024).\u003c/p\u003e\n\u003cp\u003e83\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Bates, D., M\u0026auml;chler, M., Bolker, B. \u0026amp; Walker, S. Fitting linear mixed-effects models using lme4. \u003cem\u003eJ. Stat. Softw.\u003c/em\u003e \u003cstrong\u003e67\u003c/strong\u003e, 1-48 (2015).\u003c/p\u003e\n\u003cp\u003e84\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Bartoń, K. MuMIn: Multi-Model Inference. \u003cem\u003eR package version 1.48.4\u003c/em\u003e https://CRAN.R-project.org/package=MuMIn (2024).\u003c/p\u003e\n\u003cp\u003e85\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Nguyen, L. T., Schmidt, H. A., von Haeseler, A. \u0026amp; Minh, B. Q. IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. \u003cem\u003eMol. Biol. Evol.\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, 268-274 (2015).\u003c/p\u003e\n\u003cp\u003e86\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Mathews, S. \u0026amp; Donoghue, M. J. The root of angiosperm phylogeny inferred from duplicate phytochrome genes. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e286\u003c/strong\u003e, 947-950 (1999).\u003c/p\u003e\n\u003cp\u003e87\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Rillig, M. C.\u003cem\u003e\u0026nbsp;et al.\u003c/em\u003e The role of multiple global change factors in driving soil functions and microbial biodiversity. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e366\u003c/strong\u003e, 886-890 (2019).\u003c/p\u003e\n\u003cp\u003e88\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Strobl, C., Boulesteix, A.-L., Kneib, T., Augustin, T. \u0026amp; Zeileis, A. Conditional variable importance for random forests. \u003cem\u003eBMC Bioinform.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 307 (2008).\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6100233/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6100233/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Soil legacy effects of plants (i.e., plant-soil feedback, PSF) are key drivers of the maintenance of biodiversity and alien plant invasion. While most research (88.9% of 460 experiments) has focused on PSFs of single-species, we showed that 4.6 (on average) herbaceous species co-occur and likely to interact belowground and collectively shaping plant-soil interactions across ~2000 0.25 m2 plots (only 0.03% hosting a single species) in a field survey of herbaceous plant communities in East China. However, can species-level PSFs directly translate to community-level PSFs remains untested. We experimentally showed that rhizosphere fungal communities of three species were significantly influenced by neighbouring species diversity, rather than just focal species itself. A two-phase PSF experiment further showed that, when the soil was conditioned by more than four species, community-level PSFs were disproportionately influenced by specific species rather than being a simple additive contribution from single-species PSFs. This inequality was more pronounced with increasing species richness. These findings highlight the importance of prioritizing research on community-level PSFs, particularly in species-rich ecosystems, which may reshape our understanding on plant-soil interactions and their role in shaping plant community dynamics.","manuscriptTitle":"Species-level studies do not upscale to community-wide plant-soil feedbacks","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-03-04 11:52:25","doi":"10.21203/rs.3.rs-6100233/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"3460e1a6-a0cc-4dad-be37-cd8835d335af","owner":[],"postedDate":"March 4th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":45129716,"name":"Biological sciences/Plant sciences/Plant ecology"},{"id":45129717,"name":"Biological sciences/Ecology"}],"tags":[],"updatedAt":"2025-03-20T11:10:31+00:00","versionOfRecord":[],"versionCreatedAt":"2025-03-04 11:52:25","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6100233","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6100233","identity":"rs-6100233","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}