Roots With Larger Specic Root Length and C: N Ratio Sustain More Complex Nematode Community

Aims Roots bridge above and belowground systems, and play a pivotal role in structuring root-associated organisms via inuencing food resources and habitat conditions. Most studies focused on the relationships between plant identity and root-associated organisms, however, little is known about how root traits affect nematode communities within the rhizosphere. Methods We investigated the relationships between root traits of four plant species and nematode diversity, community structure and trophic complexity in an ex-arable eld. Results While the relative abundance of herbivorous nematodes was negatively associated with specic root length (SRL), specic root area (SRA), root length density (RLD) and root C: N ratio, free-living nematodes were positively affected by these traits, implying a multifaceted effect of root traits on root-associated organisms. Importantly, we found that ner root systems promoted the complexity of the nematode community, by increasing the relative abundance of high trophic-level nematodes (i.e., omnivores and predators) and enhancing nematode diversity. Conclusion Our ndings suggest that root traits could be reliable indicators of soil community structure and interactions, and provide new insights into soil biodiversity and functional maintenance.

Here, we tested the responses of nematode community composition and structure to changes in plant species with diverse root traits (including root biomass, root diameter, speci c root length, speci c root area, root length density, root tissue density and root C: N ratio) in an ex-arable eld. We hypothesized that (1) roots with greater biomass could recruit more nematodes, whereas roots with lower N content or greater speci c root length would suppress herbivores; and (2) roots with greater speci c root length and area would enhance relative abundance of omnivores and predators through improving habitat conditions, thus facilitating the diversity and complexity of nematode community.

Materials And Methods
Experimental site description and plant species collection A total of 20 species were identi ed by their morphological features (Table S1) The plant samples were collected according to following criteria that (1) ve patches of per species were selected as separately replicated sampling sites at the study site; and (2) each replicated patch of per plant species were distanced from each other by over 10 m; and (3) harvested well-developed individuals in the center of target species patches to exclude the confounding effects of other adjacent species. After excavation with spade (the depth was beyond 30 cm to ensure root system integrity) and gently shaking away loose soil from roots, samples were stored in labeled polyethylene bags respectively and were conveyed to laboratory and kept at 4°C straight away for further analysis.

Root-traits measurement
Before measuring root traits, rhizosphere soil samples (directly adhered to the root system) were removed into clean polyethylene bags after obtained through brushing and mixing. Root samples were washed clean in running water to remove soil adhered. All root samples were scanned with scanner (EPSON LA2400 Scanner) and measured with WINRHIZO software to get the values of root length, surface area, volume, and diameter. Then, plant roots were dried to a constant weight at 60℃ to determine root biomass (g). Root length or surface area production per unit dry mass were described in terms of speci c root length (SRL; m g −1 ) or speci c root length separately (SRA; cm 2 g −1 ), root length density (RLD; m cm −3 ) calculated as the ratio between root total length and volume, and root tissue density (RTD; g cm −3 ) representing the fraction of vascular tissue in a single root segment was calculated as root dry mass divided by volume (Cornelissen et al. 2003; Pérez-Harguindeguy et al. 2013). Ball-milled samples were analyzed by potassium dichromate oxidation-ferrous sulphate titration and the Kjeldahl digestion with sulfuric acid and hydrogen peroxide to analyze root carbon content and root nitrogen content (Sparks et al. 1996).

Assessment of nematode community
Rhizosphere soil nematodes were extracted from 50g fresh soil with modi ed Baermann method followed by sugar centrifugal otation (Liu et al. 2008). Nematodes were identi ed to genus with a light microscope 400× magni cation (150 randomly-selected individuals per sample or all individuals in the condition of fewer than 150) after counted. Meanwhile, all nematodes identi ed were divided into bacterivores, fungivores, herbivores and omnivore-predators following Yeates et al. (1993). The abundance of total nematodes was expressed as individuals per 1 g dry weight soil.
We calculated indices to evaluate nematode community structure. First, the ratio between fungivores and bacterivores was calculated as FF/BF which could re ect the relative importance of energy channels in soil (Bardgett 2010;Wasilewska 1997). Also, the ratio between omnivore-predators and herbivores (OP/H) was used to indicate nematode community complexity, because the prevalence of omnivores and predators is advantage to balance distribution and increasing interactions in nematode community via grazing. (Thakur et al. 2014;Wasilewska 1997). Furthermore, the Maturity Index (MI) was calculated with the weighted mean of the individual colonizer-persister values (ranged from c-p 1 to c-p 5) of free-living nematodes as follows: where v(i) is the c-p value of taxon i and f(i) is the frequency of that taxon in a sample (Bongers 1990). The MI was used as an indicator of environmental disturbance and soil food web complexity, which high value re ected stable and complex conditions (Bongers and Bongers 1998; Siebert et al. 2020). The number of taxa was calculated to demonstrate nematode richness. Finally, nematode networks were presented to indicate the linkage density of taxa within nematode community.

Data analysis
All analyses were performed in R Version 4.0.5 (R Core Team, 2020). The relationships between nematode indices and comprehensive root traits were measured using simple regression analysis based on R 2 values using the function lm in R package base. Signi cant difference was determined at levels of p < 0.05, p < 0.01, and p < 0.001. Nematode association networks for each plant species were created by rst creating a network meta-matrix using the standardized statistics of nematodes according to identi cation. Nematode network analyses were carried out using "Vegan" and "igraph" packages (Csardi and Nepusz, 2005; Oksanen et al., 2020). We removed poorly represented nematode taxa to facilitate the determination of the core community, and analyzed as following criteria: (1) taxa relative abundance > 0.1%; (2) presented in 3 samples at least; (3) Spearman's correlation coe cient (r) > 0.6; and (4) signi cant difference was determined at levels of <0.05. In order to describe the topology of the networks, we calculated average node connectivity, average path length, diameter, cumulative degree distribution, clustering coe cient and modularity. Visualization of networks were presented with the interactive platform gephi (Bastian et al., 2009).

Root traits across plant species
At the growth stage, root biomass of S. viridis was signi cantly larger than other plant species, but lower than other plant species at the maturity stage (p < 0.05, Fig. 1a). P. americana presented the highest value of root diameter compared to other plant species (p < 0.05, Fig. 1b). Besides, SRL, SRA, RLD and root C: N ratio showed consistent and decreasing trends in the order of S. viridis, X. sibiricum, P. perfoliatum and P. americana at both growth and maturity stage (p < 0.05, Fig. 1c-e and Fig. 1g).

Soil nematode community
Total nematode individuals differed signi cantly among plant species at the maturity stage (p < 0.05; Fig. 2a), but were similar at the growth stage (p > 0.05; Fig. 2a). S. viridis, which had the highest values of SRL, SRA, RLD and root C: N ratio had signi cantly higher relative abundance of bacterivores and fungivores (p < 0.05; Fig. 2c-d), but lower relative abundance of herbivores than other plant species (p < 0.05; Fig. 2b). Besides, the value of FF/BF under S. viridis treatment was signi cantly higher than that of other plant species at the maturity stage (p < 0.05; Fig. 2f). Increase of SRL, SRA, RLD and root C: N ratio shifted nematode community towards a more richness community, as indicated by higher nematode taxa richness (p < 0.05; Fig. 1 and Fig. 2g). S. viridis and X. sibiricum presented signi cant higher values of MI and OP/H than other treatments at the growth stage, whereas S. viridis and P. perfoliatum showed higher values of MI and OP/H at the maturity stage, as indicated in Fig. 2h and Fig. 2i. Generally, plant species had signi cant effect on nematode community composition (Fig. 2).

Relations of nematode community composition with root traits
Total nematode abundance was negatively correlated with SRL (R= -0.57, p < 0.05; Table 1) and RLD (R= -0.63, p < 0.05) at the maturity stage. Besides, root biomass presented signi cantly relationships with relative abundance of herbivores, bacterivores and fungivores (p < 0.05), but such relationships were inconsistent at growth and maturity stage. Root traits including SRL, SRA, RLD and root C: N ratio were consistently and negatively related to the relative abundance of herbivores respectively, but positively related to the relative abundance of microbivorous nematodes (p < 0.05). Table 1 Relations between nematode abundance and root traits. The R values followed with *, **, and *** indicate p < 0.05, p < 0.01, and p < 0.001, respectively. Biomass-root biomass, Diameter-root diameter, SRLspeci c root length, SRA-speci c root area, RLD-root length density, RTD-root tissue density, C: N-root C: N ratio

Network complexity of nematode communities
The complexity of nematode communities was evaluated within taxa of identi ed nematode using a correlation network (Fig. 4). Generally, the numbers of network average nodes, edges declined in a particular order of treatments (S. viridis followed with X. sibiricum, P. perfoliatum and P. americana) which suggested the decrease of nematode community complexity, because greater network complexity often depended upon the number of taxa (nodes) and the linkage density (the number of edges) within the network (Table S2). In this study, a trait-based approach was used to explore the relationships between root systems and nematode communities in an ex-arable eld grown with four pioneer plant species. The results showed that root traits signi cantly affected nematode community composition and structure, with ner and greater C: N root systems supporting greater species richness and community complexity than in thicker and lower C:

Discussion
N root systems. These ndings highlight that root traits could be used to predict soil nematode communities, and especially, would be an indicator for soil food-web structure without soil sampling when root traits are known for many plant species.
Although the total nematode individuals were largely unaffected by root biomass, which disproved our rst hypothesis, our results suggested that the variation of root biomass would change the functional composition of nematode communities. The insigni cant association between total nematode individuals and root biomass may attribute to different responses of nematodes in different trophic groups (De Deyn et al. 2004; Wardle et al. 2003). Interestingly, the relationships between the relative abundance of nematode (herbivores or microbivores) and root biomass were distinct at two stages. One possible explanation for this contrasting result may due to the senescence of the annual species (such as S. viridis) and continued growing of perennial species (such as P. americana) at the maturity stage, resulting in the discrepant tendency of root biomass among plant species between two stages. Such results emphasized the limitation with regard to the disentanglement of plant-soil interaction with considering root biomass only, and also, highlight for more dynamic investigations.
Our results demonstrated that the relative abundance of herbivores was negatively related to root C: N ratio, which was consistent with previous studies (Gough et al. 2021; Viketoft 2008). Given that root C: N ratio re ects structural defenses (e.g., the concentration of lignin) related to the relative investment of C and N at the tissue level (Ma et al. 2018), greater C: N might prevent certain herbivores from penetrating cells of roots and thereby restricting survival (Ramachandran et al. 2020;Yeates 1999). These results con rmed that both quantity and quality of food resources were potential drivers of community composition within the rhizosphere (De Deyn et al. 2004). More importantly, the negative relationships between the relative abundance of herbivores and root traits including speci c root length, speci c root area and root length density suggest a multifaceted control of root traits on root-associated microorganisms. Roots with greater values of speci c root length were associated with greater nutrient uptake capacity and high turnover rates as well as shorter longevity (Roumet et al. 2006), which reduced survival time of herbivores. This result is inconsistent with a study conducted in a restoration chronosequence in a mixed-grass prairie (Ot nowski and Coffey 2020). Such discrepancies among these studies suggest the complex relationships between root traits and rhizosphere herbivore communities (Neher 2010), which call for the need to conduct more studies considering root exudates, biota-interactions and spatiotemporal heterogeneity (Sun et al. 2021; Thakur and Geisen 2019; Xiong et al. 2021). For example, root exudates, which were not considered in the current study, are thought to affect the performance of nematodes, and have been shown to attract or repel root herbivores (Sikder and Vestergard 2020).
Speci c root length, speci c root area, root length density were positively related to the relative abundance of free-living nematodes in this study. In particular, a greater speci c root area likely provides more niches for reproduction in root microbes that may cascade up to affect microbivorous nematodes and consumers in higher trophic levels (Bukovinszky et al. 2008). These results illustrate that different effects of multiple root traits on distinct trophic groups of soil micro-organisms, and further supports the idea that plants could affect the nematode abundance through different pathways (e.g., rhizosphere and litter pathways) (Zhang et al. 2019).
Consistent with our second hypothesis that roots with greater speci c root length and area could result in greater complexity of nematode communities, we showed that the increases of complexity corresponded with increasing the relative abundance of high-trophic nematode (i.e., omnivores and predators) and nematode diversity. Finer root systems enhanced soil porosity in the processes of pushing soil particles aside for growth (Bardgett et

Conclusions And Outlook
Although there was limited sample size and trait assemblage in this study, our ndings revealed that plants with different root traits exert contrasting effects on distinct nematode trophic groups. In particular, denser and ner root systems increased the community complexity of nematode by providing more available micro-habitats, which increased in relative abundance of higher trophic level nematodes (omnivores and predators) and nematode diversity. While our current work builds a bridge between plant roots and nematode communities from a trait perspective, these relationships were obtained from a single-site experiment and only based on four species, and hence requires further exploration across a wider range of species and environmental conditions. Based on our current ndings and limitations, furthermore, a broader generalization of the presented ndings requires further studies with greater sample numbers, from multiple sites, integrate multi-dimensional root trait variations, across different climatic conditions and accounting for coordination between plant roots and soil biota. This will help us to better understand the effects of plant communities on belowground biodiversity and functioning in a changing world.  Table S2 with repeated measures ANOVA</p>  Table S2 with repeated measures ANOVA</p> Figure 3 <p>Separate relationships between ecological indices of soil nematode community and root traits at two stages (represented by different colours). R<sup>2</sup> is the coe cient of determination, and signi cance levels are <sup>*</sup><em>p</em> &lt;0.05, <sup>**</sup><em>p</em> &lt; 0.01 and <sup>***</sup><em>p</em> &lt; 0.001. R<sup>2 </sup>in black indicates correlation coe cient, which calculated by combined value of two stages for the consistent tendency of growth stage and maturity stage. Linear trend lines were only plotted for groups that got signi cance levels</p>  Table S1</p> Supplementary Files This is a list of supplementary les associated with this preprint. Click to download. SupplementaryMaterial.docx