Coexistence of Tree Species Promotes the Similarity of the Elementome in Soil Profiles

All living beings are composed of various atoms of chemical elements, and the use of biogeochemical approaches to ecology enables an understanding of the ecological processes connecting organisms with environments. Based on the biogeochemical niche (BN) hypothesis, this study aimed to unravel the responses of the elementomes of the soil to different species of trees. The field experiment was conducted in Chongli, Zhangjiakou City, Hebei Province, China, and soil bioelements were analyzed at three soil layers (A, B, and C horizon) in four forests (Picea asperata (PA), Larix principis-rupprechtii (LP), Betula platyphylla (BP), and Betula platyphylla - Larix principis-rupprechtii (L-B) mixed forest). Based on soil C:N:P stoichiometry, soil elementomes of 11 bioelements (C, N, P, O, S, K, Ca, Na, Mg, Mn, and Cr) were analyzed by principal component analysis (PCA), and elementoms distance (ED) were calculated to investigate differences between horizons and forests. We found that soil elementomes differed in different soil layers and that ED between B and C horizons were larger than that between A and B. Moreover, differences in soil elementomes were smaller for tree species that often coexist compared to those that rarely live together. Our results suggest that tree species coexistence promotes the similarity of soil elementomes and may create soil condition that sustains their coexistence. In conclusion, these findings provide new knowledge on biogeochemical processes driving ecological interactions between organisms and their environments, contributing to a better understanding of terrestrial trophic ecology.


Introduction
All living beings are composed of various atoms of chemical elements, although these atoms form remarkably diverse molecules with various functions (Filipiak and Filipiak 2022). Some of these elements (i.e. carbon, hydrogen, and oxygen) are necessary to constitute the skeleton of organic molecules. Others play key roles in the structures of biomolecules (e.g. nitrogen, phosphorus, sulphur), maintenance of ionic balance (e.g. calcium, sodium, magnesium), and catalytic functions (e.g. iron, manganese) (Fraústo da Silva and Williams 2001). These atoms circulate endlessly through the food web and are ceaselessly incorporated into molecules, this process is known as biogeochemical cycling (Filipiak et al. 2021). Because no element functions in isolation, it is necessary to discover elemental linkages in biology (Elser and Hassett 1994;Redfield 1960). The use of biogeochemical approaches to ecology (Bianchi 2021), namely, ionomics and ecological stoichiometry, enables an understanding of the ecological processes connecting organisms with environments and the underlying biodiversity (Elser et al. 2000a;Elser et al. 2000b;Filipiak and Filipiak 2022;Sardans et al. 2015).
The ratios of abundant bioelements (C, N, and P) have been useful in exploring key ecological (Sterner et al. 1992), evolutionary (Kay et al. 2005), and eco-evolutionary processes (Jeyasingh et al. 2014). However, an elementome based on C, N, and P only may miss key information that could be offered by additional elements (Kaspari 2021). Several ecological stoichiometry studies in recent decades have included elements beyond C, N, and P as other elements are also involved in key functions in organisms. For instance, the 1 3 compositions and ratios of K are related to drought resistance, Mg to the light environment, and K, Ca, Mg, and S to the levels of N and S deposition (Rivas-Ubach et al. 2012;Sardans et al. 2016;Sardans et al. 2011). Similarly, elements other than N, P, and K may be limiting in some terrestrial ecosystems and may cause community-level effects (Sardans et al. 2021a). For instance, concentrations of N, P, K, Ca, Mg, S, Fe, Si, and other microelements influence macrophyte community structure in several lakes on the Yunnan Plateau, China (Xing et al. 2015). Limitation of elements other than N, P, and K in food may compromise the fitness of herbivores (Filipiak and Weiner 2017;Filipiak 2016). In line with the multidimensional stoichiometric niche framework, introducing additional elements in the analysis of the elementome can effectively improve the prediction of ecosystem functioning (Hofmann et al. 2021;Huang et al. 2019) and should be considered as an entirety in ecological stoichiometric studies (Filipiak 2018;Zhang et al. 2022a;Zhang et al. 2022b).
In recent years, ionomic approaches have emerged that correlate ecosystem function with bioelements (mainly mineral sources). Plant ionomics is the examination of the relationships between the distribution of mineral elements in plants and the concentration of nutrients and trace elements in the soil, as well as the relationship with functional characteristics and gene expression in plants (Baxter et al. 2012;Huang and Salt 2016). Several studies have reported that disruption of a single gene can greatly affect the concentration of one or more bioelements, allowing the identification of genes that control the uptake, storage, and use of soil nutrients and trace elements by plants (Baxter et al. 2012;Lowry et al. 2012). Ionomic studies have shown that descendants evolved to mitigate elemental imbalances evident in the ancestor in response to limiting conditions (Jeyasingh et al. 2023). These results further allow identifying the effects of the dynamic interactions among genetics, traits, and the environment (Buescher et al. 2010;Jeyasingh et al. 2023;Lowry et al. 2012).
Another recent approach, the "biogeochemical niche" (BN) concept, which is characterized by a particular elementome defined as the content of all (or at least most) bioelements has been proposed (Penuelas et al. 2019). It is effective in measuring ecological stoichiometry and can be regarded as an extension of the ecological niche concept. In contrast, ionomics aims to account for shifts in ionic composition by identifying specific genes and their role in changing environmental conditions, while the BN hypothesis is based on a more general idea linked to the overall direct and indirect "genetic control" of bio-elements (Sardans et al. 2021a). Peñuelas et al. (2019) investigated tree species in a holmoak evergreen Mediterranean forest distribution and found that species with more overlapping ecological niches had greater differences in their BNs. Plants growing in diverse communities tended to change their elemental compositions to either reduce or enhance N and P concentrations depending on the species compared to monocultures (Dehuang et al. 2020;Guiz et al. 2018). Fernández-Martínez et al. (2021) further revealed that pairwise differences in elementomes between species were large as the possibility of coexistence increased. Based on these empirical approaches, it was hypothesized that "at equilibrium, coexisting species tend to have distinct elementomes to minimize competitive pressure" (Sardans et al. 2021a). This BN hypothesis suggested that each species would have a specific need for certain bioelement to avoid nutritional competition with other species. In a forest, the fierce competition between species under similar soil properties would result in remarkable differences in the plant elementome with their likelihood of coexistence (Bai et al. 2019;Fernández-Martínez et al. 2021). As with the genome, the soil elementomes are defined as the concentration of elements and can be represented the state of soil development. Considering that coexisting species can adjust ecological strategies by competing for soil resources, it can be further hypothesized that the similarity of the soil elementome leads to tree species elementome segregation for coexisting species in competition.
In this study, we assessed the soil stoichiometry of different soil horizons in four different forests, i.e. Picea asperata, Larix principis-rupprechtii, Betula platyphylla, and a mixture of Betula platyphylla-Larix principis-rupprechtii in Hebei Province, China. The objectives of this study were (1) determining the distribution of C:N:P ratio and concentrations of bioelements C, N, O, S, P, K, Ca, Na, Mg, Mn and Cr in soil, and (2) investigating the soil profiles of different horizons and forests types and their relationships with soil elementomes. We predict that (1) soil elementomes differed in different soil layers, and were mainly affected by biological processes, (2) coexistence of tree species in different forests reduced the differences in the soil elementome across sites.

Study Area
The study area is located in Chongli District, Zhangjiakou City, Hebei Province, P.R. China. The latitudes and longitudes of Chongli are 40°47′ N to 41°17′ N and 114°17′ E to 115°34′ E (Fig. 1). The altitude extends from 814 to 2174 m. The climate is classified as continental monsoon with average annual temperatures of 3.7-19°C and annual precipitation of 483.3 mm. Eighty percent of the territory in Chongli is mountainous, and the forest coverage rate reaches 67% in 2021 The main tree species are Picea asperata, Larix principis-rupprechtii, and Betula platyphylla, among which Picea asperata and Larix principis-rupprechtii are artificially planted.

Experimental Design
In July 2019, four different forests of Picea asperata (PA), Larix principis-rupprechtii (LP), Betula platyphylla (BP), and the mixed forest of Betula platyphylla and Larix principis-rupprechtii (B-L) were selected in the study area. The four forests were all on mountainous slopes, and in each forest, three quadrate subplots (20 m × 20 m) were uniformly arranged from the bottom to the top of the slope. One sampling point was set in the center of each subplot.
The steps of sample collection were as follows: first, the surface coverages of litter and other sundry were removed from the sampling points; second, a vertical soil profile of approximately 1 m depth was dug using shovels, and the soil profile was found to be three soil formation layers A (surface horizon, darker color), B (granular, prismatic structure), and C (bottom layer, unconsolidated earth material) horizons according to the soil textures; finally, in each soil layer, two samples (each approximately 100 cm 3 ) were collected with a ring knife to analyze the soil physicochemical properties and bioelements. Samples were put into sealed bags and brought to the laboratory for analysis.
Soil pH was measured using a pH meter (type:PHS-3Cby China) with the soil and water ratio as 2.5:1. Soil bulk density (BD) was measured using the ring knife method. Soil organic matter (SOM) was determined by the external heating method of potassium dichromate and concentrated sulfuric acid. Purging and trapping techniques were used to determine O, N, and S concentration by an elemental analyzer (type: Elementar Vario Macro cube by Germany). The total concentrations of several nutrients (P, K, Ca, Na, Mg, Mn, and Cr) in soils were determined by inductively coupled plasma-optical emission spectrometry (Sparks et al. 1996) (type: Agilent 5110 ICP-OES by the USA).
Litter samples were collected simultaneously with soil in July 2019. Only the existing intact litter that had not been decomposed in the sample plot were collected as test samples to ensure the consistency of freshness of litter. Soil and other impurities were removed in the laboratory and air-dried to a constant weight of 65 °C. Litterbags with specifications of 20 × 20 cm and an aperture of 0.01 mm (excluding the influence of soil fauna) were loaded with 20 g. Litterbags were stitched together with nylon mesh with corresponding apertures and were placed in the corresponding plot in January 2020 and decomposed under natural conditions. In mid-June, early August, mid-September, and early November 2020, three litterbags were randomly taken from each sampling site and brought back to the laboratory. The litter was dried and weighed to measure the decomposition rate.

Fig. 1 Study area and soil sampling plots
where L is litter decomposition rate (%); W t is the mass of litter at t decomposition time (g); W 0 is the initial mass of litter (g). Litter net decomposition rate: where L n is litter net decomposition rate (%); L ct is the litter decomposition rate at t decomposition time, L c(t − 1) is the litter decomposition rate of the last sampling time up to time t. All data in this study were described by the mean and standard deviation. SOM and elemental concentrations were described by mass content, the values of C:N, C:P, and N:P were molar ratios. A significance level of p<0.05 was specified in this study. Analysis of variance (ANOVA) and Least significant difference (LSD) at a 5% level of significance were used to compare the difference among horizon and forest. PCA was performed on elemental concentrations to estimate elementomes. Elementome Euclidean distance (ED) was used to quantitatively indicate the difference between elementomes (Fernández-Martínez 2021).
All statistical analyses mentioned above were implemented using SPSS 25.0 (IBM, Armonk, New York, NY, USA). Related graphs were drawn by Origin 2021b (Hampton, MA, USA).

Soil Chemical and Physical Properties and C, N, P Stoichiometry of Different Forests
The C:N ratios did not differ significantly between the PA, LP, and BP forests during the three horizons (Table 1) and (2) L n (%) = L ct − L c (t−1) were significantly lower in the soil A and B horizons than in the C horizon of the B-L mixed forest. All forests showed no significant differences in the C:N of the C horizon. A lower C:N ratio was observed in PA surface soil than that in BP and B-L mixed forests. As the soil depth increased, the C:P and N:P ratios decreased. BP forest obtained the highest C:P and N:P ratios. There were no significant differences in the C:P and N:P ratios between the A and B horizons of the other three forests. PA forests had the lowest ratios of C:P and N:P at the C horizon. Overall, the SOM in broad-leafed forests (BP) was higher than that in coniferous forests (PA and LP). In addition, the soil C:N, C:P and N:P ratios were higher in the B-L mixed forest than in the Larix principisrupprechtii monoculture, indicating that mixed forest can effectively enhance soil organic matter quality in Larix principis-rupprechtii forest.
The net decomposition rate of litter during the decomposition process was calculated (Fig. 2). During litter decomposition, the net decomposition rates of litter gradually increased from 0 to 210 d and gradually decreased after 210 d in both forest types. The cumulative decomposition rate of litter with different forests was BP (51.62%) > B-L (42.64%) > PA (32.81%) > LP (25.41%), and the cumulative litter decomposition rate in the B-L was significantly higher than that in the LP (p < 0.05), indicating that litter quality is an essential factor affecting the litter decomposition.

Soil Elementome Distribution from PCA
The distribution of elementomes analyzed by PCA method is shown in Fig. 3. Three principal components were able to explain a total of 86.05% of the variance. According to the results, loading values and explained variance were mapped to each component after PCA. PC1 accounted for 52.46% of the total variance and was significantly correlated to C, N, O, S, and P contents. Thus, PC1 remarkably described the biological elements, i.e. C, N, O, S, and P, which are indispensable nutrients for the growth and development of all plants in forest ecosystems.
PC2 explained 22.17% of the variance in the original data, with K, Ca, Na, and Mg having the major loadings. These elements are nutrient cations that are subjected to biological activity and chemical activity to maintain their normal growth.
Accounting for 11.42% of the variance, PC3 substantially described the contents of Mn and Cr in the study area. This component can be described as soil bedrock which is the main influencer of these elements.

Soil Elementome Differences between Horizons and Forests
Mean ± SE of the scores of A, B, and C individual horizons in the PC1-PC3 components of the PCA was calculated ( Table 2). The elementome distances (ED) between horizons in four forests (Picea asperata, Larix principis-rupprechtii, Betula platyphylla, and Betula platyphylla -Larix principisrupprechtii mixed forest) in this study were calculated and shown in Fig. 4. In all four forests, soil elementomes differed in different soil layers. The elementome distances (ED) between the B and C horizons were larger than the ED between A and B horizons (Fig. 4), among which ED BC accounted for 61~91% of the entire soil profile. In comparison with ED BC , the proportion of ED AB was as low as 9~39%, which showed a larger difference in the bottom two horizons.
We found that soils of the different forest had different elementomes. In mixed forests of Betula platyphylla and Larix principis-rupprechtii, the soil elementomes were higher than those in pure forests (Fig. 5). Among all forests, Picea asperata had the lowest soil elementomes. Successive plantation planting can degrade forest soil fertility, and nutrient accumulation can be effectively increased by mixed needle and broad-leaved planting.
Based on the forest survey, we can obtain the distribution and coexistence situation of tree species. We used score distances for PC1 and PC2 of the PCA to describe differences in soil elementomes and found that species rarely living together show larger differences in soil elementomes than those that frequently coexist. The highest elementome distance (ED) value, 1.69, appeared between Picea asperata and Betula platyphylla, and the lowest ED value, 0.53, appeared between Picea asperata and Larix principis-rupprechtii.

Discussion
The proportional relationship between C, N, and P is an important indicator of soil nutrient status ). The C:N and C:P ratios of soil determined the decomposition of SOM, whereas the N:P ratio reflected the element restriction of the ecosystem (Hui et al. 2021). In this study, we found that the C:N, C:P, and N:P ratios in the broadleaved forest were higher than that in the coniferous forest. This may have been due to plants affected by soil organic C and TN contents through root exudates and litter decomposition process (Luo et al. 2016;Tao et al. 2020). According to the net decomposition rate of litter during the decomposition process, the litter decomposition rate of Betula platyphylla was significantly higher than that of pure Larix principisrupprechtii. Besides, soil C:N and N:P ratios in the B-L mixed forest were higher than in the LP monoculture.
Mixed forests are known to have the potential to increase ecosystem C storage and soil nutrient protection (Fichtner et al. 2018), which was mainly affected by the species identity, functional traits (e.g. complementary shade-tolerance or root systems), and climatic conditions (Blaško et al. 2020). In mixed forests, chemical differences in litter, the transfer of nutrients and secondary metabolites between litter, and variations in the microhabitat of decomposers led to the accelerated decomposition of mixed litters (Gartner and Cardon 2004;Song et al. 2010). Our study provided evidence that stands conversion from BP to B-L mixed culture substantially improved soil quality. Numerous studies have shown that soil C, N, and P contents and stoichiometry showed a decreasing trend from surface to deeper layers (Wang et al. 2022;Zhang et al. 2019). In our results, the C:N:P ratios and the soil C:N:P ratios of Picea asperata, Larix principis-rupprechtii, Betula platyphylla, and Betula platyphylla -Larix principis-rupprechtii mixed forest decreased from 137:11:1, 127:10:1, 167:13:1, and 123:9:1 in the A horizon to 54:4:1, 98:7:1, 149:11:1, and 115:8:1 in the C horizon respectively, consistent with some previous reports such as Bing et al. (2015) reported that C:N:P ratios decreased from 343:16:1 in the A horizon to 63:3:1 in the C horizon.
Compared to the average N:P and C:N ratios in China (13.83 and 8.43), all four forests had lower soil N:P ratios (4.32-12.55) and higher C:N ratios (12.12-13.50) (Tian et al. 2010). In general, when N:P was less than 14:1, plant growth was more restricted by N; when N:P was higher than 16:1, plant productivity was more restricted by P; and when N:P was in the middle, plant growth was restricted by both nitrogen and phosphorus (Olde Venterink et al. 2003). Our  However, measuring a small subset of elements to understand any biological process is bound to ignore a substantial proportion of underlying mechanisms (Jeyasingh et al. 2023). Several studies in recent decades have highlighted the importance of incorporating more elements to characterize the stoichiometric niches of soil animal taxa (Zhang et al. 2022a). Soil C, N, O, S, and P are major structural components in living organisms and also participate in many biochemical organisms (Penuelas et al. 2019). In addition, K, Ca, Na, and Mg are essential elements for a variety of organisms. Therefore, the changes in the main mineral elements in the ecosystem and the mechanism of their recycling are important contents of the primary succession theory since they represent the main functional process of an ecosystem and determine its pattern. Healthy ecosystems depend heavily on the normal circulation of mineral elements that are related to their stability and sustainability (Diaz et al. 2015;Reich and Oleksyn 2004). In addition, manganese (Mn) and chromium (Cr) are also required for normal plant growth and development which cannot be decomposed by soil microorganisms, so they are easy to accumulate. Nevertheless, excessive concentrations of Mn and Cr would be detrimental to plant growth. (Shuyu et al. 2021;Zemunik et al. 2020). Such a high-dimensional perspective of stoichiometric niches should be adopted in research to understand the ecological interactions between organisms and their environments.
The adsorption, analysis, decomposition, and aggregation of various elements in soil constitute the biogeochemical cycle of the soil environment (Chen et al. 2014). Soil biological processes (e.g., plant residue decomposing) and soil physicochemical processes also control the nutrient accumulation and release from organic matter decomposing in mineral soils (Manzoni et al. 2010). Biological and chemical processes take place throughout the soil profile, interacting at a wide range of temporal scales and together driving the elemental cycle of the soil profile (Kirkby 2018). The biological process is mainly composed of two parts: biological residue is decomposed into inorganic compounds via humification and mineralization by soil microorganisms, and living organisms absorb soil elements. In geological processes, leaching and diagenesis fix soil elements into bedrock while weathering releases them (Waroszewski et al. 2018). A previous study indicated that the degree of control of soil nutrient availability by geochemical and biological processes is inconsistent (Luo et al. 2016). The elementomes in this study were significantly different between soil layers, in particular, a more significant difference in the bottom two horizons. Our results show that the biological cycle was more vigorous than the chemical cycle, and soil elementomes were more affected by the biological activity rather than the bedrock, consistent with some previous reports such as Pistocchi et al. (2018) reported that biological processes dominate phosphorus dynamics in organic horizons of temperate forest soils. Organisms played an important role in soil ecosystem balance and stability as the most active factors in soil formation.
Trees can shape the proportions of elements in the soil (Zederer et al. 2017) and can create a soil environment that enhances their ability to compete and thus increases their fitness (Cools et al. 2014). The nutrient content of tree species determined leaf-fall decomposition, nutrient return, and nutrient release into the soil in forests, affecting soil fertility. In many studies, differences in litter lignin and nutrient content were found to influence microbial decomposition; i.e, litters with higher lignin forms decomposed slower, which subsequently affected the soil elements of the forest floor (Hansson et al. 2011;Hobbie et al. 2006;Lovett et al. 2002;Vesterdal et al. 2012). Each species generated soil conditions that reflected the environmental conditions where it dominated, at a local level, with its life history and nutritional strategies (Perez-Ramos and Marañón 2011; Vivanco and Austin 2008). According to Aponte et al. (2013), tree species-induced variations in soil conditions created positive feedback through niche partitioning that enabled the coexistence of tree species. Species were unique genetic pools and products of long-term evolutionary processes. The genotypic elements shaped coexistence and accounted for a large part of foliar element composition (Sardans et al. 2021a).
Tree species-induced variations in soil nutrient contents influenced elementomes, enabled the separation of biogeochemical niches, and maintained their coexistence. According to our results, differences in soil elementomes were minor for tree species that often coexist. Numerous studies have demonstrated that elementomes differed more for coexisting species and individuals than for noncoexisting ones . Additionally, there was evidence that species would compete for resources under similar soil elementomes, causing niche partitioning (Loreau and de Mazancourt 2013) with the likelihood of coexistence. The evolution and bioelemental composition of ecosystems were bidirectional because nutrient supply could affect evolutionary processes and the effects of evolution on nutrient supply (Durston and El-Sabaawi 2017). We can understand the processes underlying species shifts in bioelemental composition by studying their responses to environmental changes (Leal et al. 2017;Yamamichi et al. 2015) and, therefore, the effects of organisms on ecosystem functioning and services (Leal et al. 2017). In this way, elementomes constitute a quantifiable tool for detecting, quantifying, and understanding the mechanisms and processes underlying community evolution and species turnover (Penuelas et al. 2019). Under global change, the study of ecosystem functioning should be based on an elementomes approach.

Conclusion
In this study, we investigated the variations of bioelemental composition between different soil layers, as well as the responses of the elementomes of the soil to different species of trees. Our results suggest that elementome distances between the B and C horizons were larger than that between A and B horizons. This suggests that soil elementome are mainly affected by biological processes. Our results showed that the coexistence of tree species promotes the similarity of the elementome. These findings provide implications for the understanding of the processes underlying species shifts in soil bioelemental composition and the responses of organisms to environmental changes and, in turn, the effects of organisms on ecosystem functioning and services. In conclusion, the study of elementome could provide new knowledge on biogeochemical processes driving ecological interactions between organisms and their environments, contributing to a better understanding of terrestrial trophic ecology.