Nematode biomass changes along an elevational gradient are trophic group dependent but independent of body size

Aboveground, large and higher trophic‐level organisms often respond more strongly to environmental changes than small and lower trophic‐level organisms. However, whether this trophic or size‐dependent sensitivity also applies to the most abundant animals, microscopic soil‐borne nematodes, remains largely unknown. Here, we sampled an altitudinal transect across the Tibetan Plateau and applied a community‐weighted mean (CWM) approach to test how differences in climatic and edaphic properties affect nematode CWM biomass at the level of community, trophic group and taxon mean biomass within trophic groups. We found that climatic and edaphic properties, particularly soil water‐related properties, positively affected nematode CWM biomass, with no overall impact of altitude on nematode CWM biomass. Higher trophic‐level omnivorous and predatory nematodes responded more strongly to climatic and edaphic properties, particularly to temperature, soil pH, and soil water content than lower trophic‐level bacterivorous and fungivorous nematodes. However, these differences were likely not (only) driven by size, as we did not observe significant interactions between climatic and edaphic properties and mean biomasses within trophic groups. Together, our research implies a stronger, size‐independent trophic sensitivity of higher trophic‐level nematodes compared with lower trophic‐level ones. Therefore, our findings provide new insights into the mechanisms underlying nematode body size structure in alpine grasslands and highlight that traits independent of size need to be found to explain increased sensitivity of higher trophic‐level nematodes to climatic and edaphic properties, which might affect soil functioning.


| INTRODUC TI ON
Grasslands cover around 40% of the Earth's terrestrial surface (Petermann & Buzhdygan, 2021). These habitats, especially in high alpine regions, are key biodiversity hotspots (Myers et al., 2000;Wang et al., 2022). One of the largest alpine grassland systems is located on the Tibetan Plateau, with a size of approximately 2.5 million km 2 and an average elevation above 4000 m (Qiu, 2008). The immense diversity in Tibetan grasslands can be explained by harsh, elevation-based shifts in environmental conditions, such as temperature, precipitation, and soil carbon content that increase available niche space through temporal fluctuations (Kuang & Jiao, 2016;Yang et al., 2008). Such elevational gradients are often used to detect the ecological response of biota to changes in their physicochemical surroundings, such as those associated with climate change (Korner, 2007). However, most studies investigating the impacts of climate change-related factors along altitudinal gradients have focused on aboveground organisms (Gobbi et al., 2006;McCain, 2007), neglecting the even more diverse and abundant soil organisms Guerra et al., 2020).
Soil biodiversity is dominated by microbes and small metazoans that drive many ecosystem functions, such as soil carbon cycling and plant performance (Bardgett & van der Putten, 2014;Crowther et al., 2019). In recent years, research has led to a better understanding of the main climatic and edaphic properties that affect the structure and diversity of soil organismal communities (Delgado-Baquerizo et al., 2016;Oliverio et al., 2020;Potapov et al., 2023;Tedersoo et al., 2014), knowledge also gained for the most prevalent soil metazoans: nematodes . Nematodes occupy all trophic levels in soil food webs as bacterivores, fungivores, herbivores, omnivores, and predators (Neher, 2010;Yeates et al., 1993). As such, nematodes catalyze nutrient cycling and can both positively and negatively affect the plant performance (Wilschut & Geisen, 2021). The major climatic and edaphic properties that stimulate nematode abundances and alter nematode community composition are related to soil water and organic carbon (Nielsen et al., 2014;van den Hoogen et al., 2019), while increased temperature and higher soil pH reduce nematode diversity and density (Zhao et al., 2017). As these and other climatic and edaphic properties change with elevation, nematode abundances, particularly that of fungivores, herbivores, and omnivores, as well as nematode diversity have been reported to decrease with elevation (Afzal et al., 2021;Kashyap et al., 2022;Liu et al., 2019).
Other studies found that nematode abundance and functional diversity increase along the elevation (Kergunteuil et al., 2016;Kouser et al., 2021), suggesting that patterns of nematode communities across elevation are not always the same.
Nearly, all this knowledge on nematode community variation along spatial and environmental gradients has focused on compositional data based on non-quantitative molecular tools or abundance-based data of individual taxa or functional groups (Quist et al., 2016;Xiong et al., 2020). However, data on species richness, abundance, and composition may have limited utility for linking soil (nematode) biodiversity data with their actual ecological role in ecosystem processes (Shade, 2017), as differences in body size and biomass can easily compensate for profound variations in organismal abundances (Blackburn & Gaston, 1994;Norkko et al., 2013).
In fact, nematodes can range in size by up to three orders of magnitude, and profound size differences prevail even within nematode trophic groups (Verschoor et al., 2001). For example, among bacterivores, the genus Plectus has a body size five times larger than that of Acrobeloides, while, among predators, Clarkus has a body size three times larger than Mylonchulus (Mulder & Vonk, 2011;Zhao et al., 2015). To determine size-related differences in complex nematode communities in soil, measurements of body size, such as with a trait-based community-weighted mean (CWM) approach are needed (Lavorel et al., 2008;Sechi et al., 2017). Yet, due to the profound workload of identifying and measuring at least dozens of single nematode individuals per sample, the CWM approach has only rarely been applied in ecological studies focusing on nematodes, with few exceptions, as reported by (Andriuzzi et al., 2020;Liu et al., 2015).
Body size also is a sensitive trait to environmental disturbance and global change (Gardner et al., 2011;Sheridan & Bickford, 2011). In soils, studies on collembola revealed that biomass or size-based approaches can provide a comprehensive understanding of not only body size distribution but also functional changes in the community (Turnbull & Lindo, 2015). Across different organismal groups in soil, including nematodes, body size differences have been shown to determine ecological processes, like community assembly (Luan et al., 2020). Changes in nematode body size have been applied as a measure to assess changes in soil and benthic food webs as indicators for habitat quality (George & Lindo, 2015;Tita et al., 1999). Typically, large and higher trophiclevel organisms are more susceptible to changes than small and lower trophic-level organisms-the trophic sensitivity hypothesis (Vasseur & McCann, 2005;Voigt et al., 2003). This concept also holds for nematodes, as often higher trophic-level omnivores and predators are more vulnerable to changes in soil, including agricultural management, than lower trophic-level bacterivores and fungivores (Bongers, 1999). These effects can be explained by longer generation times and lower fecundity of larger, higher trophic-level nematodes compared with smaller, lower trophic-level nematodes (Bongers & Bongers, 1998;Ferris et al., 2001). Individual nematode species can also adapt to environmental factors, by increasing in body size with higher resource availability (Andriuzzi & Wall, 2018), while decreasing in response to increasing temperature (Simmons et al., 2009). Yet, we lack an understanding of how size and consequently biomass of nematodes at the community level, especially within the same trophic group, respond to alternations in climatic and edaphic properties.
Here, we sampled an altitudinal transect consisting of 59 sites across Tibetan grasslands to investigate how climatic and edaphic properties impact soil nematode biomass at the levels of the whole community, trophic groups, and genus level.
We tested four hypotheses: 1. Nematode CWM biomass decreases toward higher altitudinal sites.

Nematode CWM biomass is affected by climatic and edaphic
properties, increasing with higher precipitation and soil organic carbon (SOC) and reducing with higher temperature. 3. Higher trophic-level nematodes (omnivores and predators) respond more strongly to climatic and edaphic properties than lower trophic-level nematodes (bacterivores and fungivores).
4. Within nematode feeding groups, larger taxa respond more strongly to climatic and edaphic properties than smaller taxa.

| Aboveground plant community and soil samples
For each site, we selected five plots at approximately 50 m apart. In each plot (1 m 2 ) we morphologically identified plants to species level, then calculated each site's plant species richness (PSR) based on the cover. Shoots were cut at the soil surface, and dry weight was determined after oven drying at 65°C for 48 h. Plant cover (PCV) was calculated as the ratio of the shaded area to the total area in each plot.
Five soil cores were collected in each plot (3.8 cm in diameter and 20 cm in depth) and taken to the laboratory. After removing stones and larger roots, we sieved soil through a 2-mm mesh. Five soil cores were thoroughly mixed and divided into two subsamples. One subsample was used to measure soil physiochemical properties, and the other subsample was used to extract the nematode community.

| Soil physicochemical properties measurements
Soil water content (SWC) was determined from fresh soil heated at 105°C for 48 h (Barrett et al., 2008). The remaining soil was airdried, avoiding direct sunlight and sieved through a 0.15 mm mesh to measure other soil properties. Soil pH was measured in a 1:2.5 soildeionized water slurry using a pH meter (PHSJ-3F, Shanghai INESA Scientific Instrument Co., Ltd; Wang et al., 2019). Soil total nitrogen (TN) and phosphorus (TP) were both estimated after digestion using concentrated H 2 SO 4 at 375°C for 3 h and 45 min, following the protocol (Mehlich, 1984). Soil available nitrogen (including ammonium and nitrate; AN), available phosphorus (AP) and SOC were determined as described by (Chen et al., 2021).

| Nematode community extraction and identification
The soil samples were stored at 4°C before nematode extraction within 7 days. We extracted nematodes from 50 g of fresh soil using Baermann wet funnels (Hooper et al., 2005). We counted nematodes in entire samples and identified the first 150 individuals to the genus level using a light microscope (Olympus CX31, 100×-400× magnification) according to the Nemaplex database (http://nemap lex.ucdav is.edu/) and (Bongers & Bongers, 1998). All taxa were assigned to F I G U R E 1 Distributions of sampling sites along geographic distance on the Tibetan Plateau. five feeding groups (bacterivores, fungivores, herbivores, omnivores, and predators; Yeates et al., 1993).

| Nematode body size calculation
For each identified nematode, we measured length (L) and greatest diameter (D) to calculate the nematode body size V = (L × D 2 )/1.7 according to Ferris (2010). Nematode biomass was calculated by the formula W = L × D 2 /(1.6 × 10 6 ; Andrássy, 1956), where W is the nematode fresh weight per individual. The CWM biomass of each was calculated according to (Ricotta & Moretti, 2011), using the formula: where P i is the relative abundance of the genus i (i = 1, 2… S), and x i is the mean biomass for genus i.

| Statistical analysis
All statistical analyses were computed in R software, version 4.0.3 (R Core Team, 2020), and all figures were produced using the gg-plot2 package (Wickham, 2016). We used site mean value averaged from five plots to perform all analyses. The nematode CWM biomass met the assumptions of normal error distribution and variance homogeneity. Due to the low numbers of omnivorous and predatory nematodes, we combined them in our analyses. General linear and polynomial regressions were performed using the dplyr package (Wickham et al., 2022) and ggpmisc package (Pedro, 2021) to analyze the response of nematode CWM biomass to altitude, latitude, longitude, MAT, MAP, Aug_Tem, Aug_Pre, PCV, PSR, pH, AN, AP TN, TP, SOC, and SWC. We calculated the variance inflation factors (VIFs) in the car package (Fox & Weisberg, 2019) and dropped some variables with VIFs greater than 7. The following variables were used in our global model: To obtain the main predictors of variation in nematode CWM biomass, a Random Forest analysis (Breiman, 2001) was conducted, the random forest analysis can be applied to small sample sizes and does not require predictor selection (Cutler et al., 2007). The relative importance (increase in mean square error percentage) was estimated by random forest statistics for each tree that was averaged over a forest (500 trees). The significance of the random forest models and each variable was assessed by 999 permutations. Random forest was computed using the packages rfPermute (Archer, 2022) and randomForest (Liaw & Wiener, 2002).
To test whether higher trophic-level and lower trophic-level nematodes responded differently to environmental characteristics, we combined the bacterivores and fungivores as the lower trophic-level group and combined omnivores and predators as the higher trophiclevel group. To meet the assumptions of normal error distribution and variance homogeneity, we performed log(x + 1)-transformation for CWM biomass. For each of the selected environmental characteristics, we ran a separate regression with log(x + 1)-transformed CWM biomass as the response factor and trophic-level group, environmental characteristic and their two-way interaction as fixed factors. To test if larger taxa respond more strongly than smaller taxa across the trophic groups, we first created a Spearman's rank correlation matrix using the corrplot package (Wei et al., 2021). We applied Spearman's correlation coefficient (ρ)

| Trophic-level patterns
The correlation between CWM biomass and climatic and edaphic  Aug_Tem × Trophic level, F 1,114 = 5.29, p < .05; Figure S2d), while CWM biomass of lower trophic levels did not show such responses.
SOC appeared to be more positively related to CWM biomass of higher trophic-level nematodes than to that of lower trophic-level nematodes (SOC × Trophic level, F 1,114 = 3.58, p = .06; Figure 3f). Additionally, higher trophic-level CWM biomass increased with MAP, whereas lower trophic-level CWM biomass decreased with MAP (MAP × Trophic level, F 1,114 = 4.02, p < .05; Figure S2c).

| Taxon-specific patterns
Within trophic groups, nematode taxa responded in taxon-specific ways to climatic and edaphic properties ( Figure S4). However, there was no overall significant effect of the standardized Spearman correlation coefficient of climatic and edaphic properties on nematode taxon biomass within trophic groups ( Figure 4). Moreover, the relationship between taxon biomass and the standardized Spearman correlation coefficient did not differ between trophic groups (all Group × Biomass interactions: p > .1; Figure 4).

| DISCUSS ION
While aboveground animals often show size-and trophic groupbased sensitivity to environmental changes, it has so far remained unknown whether the same applies to belowground animal communities. Using a CWM approach to investigate average nematode biomass (CWM biomass) responses to climatic and edaphic properties along a 1300 km transect across the Tibetan Plateau, we show that monthly precipitation and SWC were the most important drivers shaping the overall nematode CWM biomass. We demonstrate that higher trophic-level nematodes show stronger changes in body size in response to a number of climatic and edaphic properties than lower trophic-level nematodes. These effects were likely independent of size, as size differences within trophic groups did not correlate with changes in climatic and edaphic properties. Therefore, our research suggests that changes in climatic and edaphic properties may affect body size structure within nematode communities, with possible consequences for ecosystem functioning.
We found no evidence for a reduction in nematode CWM biomass with increasing altitude. This finding contradicted Hypothesis 1 and previous studies showing declines in nematode abundance  and higher trophic-level nematodes with increasing altitude (Afzal et al., 2021), which should lead to a reduction in nematode CWM biomass. Therefore, our findings are more in line with studies showing increases in the abundance and diversity of soil nematodes with elevation (Kergunteuil et al., 2016;Kouser et al., 2021). The contrasting patterns among our study and previous ones might be explained by differences in the altitudinal range Andriuzzi & Wall, 2018), as well as nematode abundance, and change nematode functional group composition (Nielsen et al., 2014;van den Hoogen et al., 2019). Mechanistically, the positive links between CWM biomass and increased precipitation can be explained by more favorable conditions and habitable space for nematodes. In fact, nematodes are water-bound (Wallace, 1968), while SOC provides the growth basis for microorganisms that bacterivorous and fungivorous nematodes prey upon (Wilschut & Geisen, 2021). The increase in CWM biomass with increased SWC and SOC might originate from a more complex food web that supports a higher share of largebodied, higher trophic-level nematodes (Verschoor et al., 2001). Unlike soil carbon and water content, temperature did not affect nematode CWM biomass, which partly contradicts hypothesis 2.
This assumption showing negative temperature effects on nematode body size was largely based on studies focusing on individual or dominant taxa in nematode communities (Simmons et al., 2009).
The different responses to temperature in our community approach is likely explained by the large spatial range covered where tem- CWM biomass of higher trophic-level nematodes generally responded more strongly to climatic and edaphic properties (PCV, pH, SOC, SWC, and sampling month temperature) than CWM biomass of lower trophic-level nematodes, providing support for hypothesis 3, and the concept behind the trophic sensitivity hypothesis (Vasseur & McCann, 2005;Voigt et al., 2003). These findings can be explained by the increased resource availability, here indicated by SOC and PCV, that support root feeders, microbes and their bacterivorous and fungivorous nematode consumers and, subsequently, higher trophic-level nematodes (Andriuzzi & Wall, 2018;Zhang et al., 2022). Therefore, more complex food webs establish that can host more large-bodied soil nematodes (Andriuzzi & Wall, 2018;Verschoor et al., 2001). In turn, microhabitat conditions driven by, for example, pH, monthly temperature and SWC affect higher trophic-level nematodes more than lower trophic-level nematodes, suggesting that higher trophic-level nematodes have a narrower niche space than lower trophic-level nematodes (Andriuzzi et al., 2020;Gibert & DeLong, 2014). Higher trophic-level nematodes contain more sensitive taxa with high c-p (denoting the ratio of colonizers to persisters) values (Bongers, 1999). In fact, these differences in traits led to nematode-based soil indices to assess soil quality including susceptibility to change (e.g., climatic and especially management) and have proven informative (Du Preez et al., 2022)-which we here confirm.
Despite the suggested strong body size-related response underlying potential trophic-level effects (hypothesis 3), we did not observe any size-related responses to climatic and edaphic properties within trophic groups. This finding rejects hypothesis 4 and suggests that body size is not or only partly underlying differential responses of nematodes to changes in climatic and edaphic properties. Body size might only matter under specific environmental conditions rather than across large spatial scales (Stuber et al., 2018).
Differences to other groups, including collembola and vertebrates, for which larger taxa responded more strongly to external change than smaller taxa (Capdevila et al., 2022;Thakur et al., 2023), might be explained by the high similarity of nematode taxa in size and shape, especially within trophic groups (Bongers & Bongers, 1998).
However, as the phylum Nematoda is nearly 500 million years old (Rota-Stabelli et al., 2013), species in each trophic group can be phylogenetically highly diverse (Blaxter et al., 1998) and vary in many other traits than just size (Du Preez et al., 2022). Other traits than body size are therefore at least partially underlying the patterns observed across trophic groups, which we assume to be linked to different feeding styles. Further studies are needed to determine size-independent traits that impact nematode responsiveness to climatic and edaphic changes.

| CON CLUS ION
In summary, our research identified the climatic and edaphic properties that shape nematode body size structures across grasslands on the Tibetan Plateau. We demonstrate that sampling month precipitation and SWC were the most important drivers in shaping nematode CWM biomass. Higher trophic-level nematode CWM biomass responded more to changes in climatic and edaphic properties than lower trophic-level nematodes. However, these effects were at least partly decoupled from size differences, as no changes in taxon responsiveness linked to size within trophic groups were

ACK N OWLED G M ENTS
We thank Alejandro Berlinches de Gea, Arne Schwelm for their comments on the earlier version of the manuscript. We also thank the editor and the two anonymous referees for providing constructive comments on our manuscript. We thank Dr. Wei Qi, Dr. Peng Jia, Dr.