Statistical analysis
The total number of soil nematodes in each trophic group was converted to the total number of individuals per 100 g of dry soil. The following indices were used to determine nematode diversity as described by Zhao et al. (2014):
Shannon-Wiener diversity index (H’ ),\(H^{{}^{\prime}}=-\sum{\text{Pi}\ln\text{Pi}}\) (1)
Pielou evenness index (J ), \(J=\frac{H^{\prime}}{H^{\prime}\max}\) (2)
\(H^{{}^{\prime}}max=lnS\) (3)
Simpson dominance index, \(\lambda=\sum\)Pi 2 (4)
Margalef richness index, \(SR=\frac{S-1}{\ln N}\) (5)
where pi is the proportion of the i th species (i = 1, 2, 3… S), S is the number of nematode categories in one soil sample, and N is the total number of individuals of all nematode categories.
Two maturity indices (MI25 and PPI) were also calculated to represent the diversity of nematode life history based on colonizer-persister (c-p) values; c-p values range from 1 to 5, i.e., from colonizer (r -strategy) to persister (K -strategy) (Bongers, 1990). MI25 is a maturity index for free-living nematodes (c-p value = 2, 3, 4, and 5) except plant-feeding nematodes and juveniles, and PPI is a maturity index for plant-feeding nematodes (Bongers, 1990). These indices are calculated as follows:
\(MI25=\sum{{v(i)}^{cp2-5}\times{f(i)}^{cp2-5}}\) (6)
\(PPI=\sum{{v(i)}^{cp1-5}\times{f(i)}^{cp1-5}}\) (7)
where v (i ) is the c-p value of the taxon, andf (i ) is the proportional abundance of this taxon relative to the total abundance of free-living nematodes (for MI25) or plant-feeding nematodes (for PPI) in one soil sample.
The functional diversity of nematodes in each soil sample was determined by calculating the enrichment index (EI), structure index (SI), basal index (BI), and channel index (CI). According to Ferris et al. (2001), the EI value is the expected response of the opportunistic Ba1 (bacterial-feeding nematodes with c-p = 1) and Fu2 (fungal-feeding nematodes with c-p = 2) to abundant resources; the SI value indicates the conditions of the soil food web. BI and CI values are used to evaluate the effects of soil management methods on nematode communities (Berkelmans et al. 2003). These indices were calculated as follows:
\(EI=100\times(\frac{e}{b+e})\) (8)
\(SI=100\times(\frac{s}{b+s})\) (9)
\(BI=100\times(\frac{b}{b+e+s})\) (10)
\(CI=100\times(\frac{\text{ef}}{eb+ef})\) (11)
where e is the enrichment of Ba1 and Fu2; b is the basal component of Ba2 and Fu2; s is the structural component containing the guilds of Ba3-5, Fu3-5, Om3-5, and Pr2-5; ef is a product of the abundance × weight value of the Fu2 guild; and ebis a product of the abundance × weight value of the Ba1 guild.
Two-way ANOVAs were conducted to assess the effects of two main factors (N addition and sampling depth, N and D) and their interactions (N*D) on all dependent variables. One-way ANOVAs were used to compare the four treatments. If ANOVAs were significant, means were compared with an LSD test at P< 0.05. SPSS Statistics 19.0 was used for statistical analyses (SPSS Inc., Chicago, USA). Structural equation modeling (SEM) was performed by “lavaan” package in R (Rosseel, 2012), to analyze hypothetical pathways that may explain how N loading affected different trophic groups of soil nematodes. Before we employed the model, we assessed model identification according to t -Rule, which the number of nonredundant elements in the covariance matrix of the observed variables must be greater than or equal to the number of unknown parameters (https://stat.utexas.edu/software-faqs/lisrel/146-training/software/655-lisrel-assessing-model-identification). The response variables (showed in the Supplementary Table 1) for soil total C, total N and total P; for NO3-N, NH4+-N and available P were indicated by principal component scores (both PC1 scores were employed). The co-occurrence networks of different nematode genera were used to determine the stability of their coexistence under different N levels; these networks were constructed using the “wgcna” package in R based on the Spearman correlation matrix (Langfelder and Horvath, 2012).