Effect of coupled reduced irrigation and nitrogen fertilizer on soil mite community composition in a wheat field

Abstract Groundwater and nitrogen fertilizer overuse severely threatens crop productions; thus, current ecological agriculture requires low irrigation and nitrogen fertilizer inputs. The effects of combined reduced irrigation and nitrogen fertilizer addition on soil organism (e.g., mite) community and biodiversity remain poorly understood. We analyzed soil mite community composition, wheat grain yield, and soil characteristics in a 10‐year manipulation experiment with two levels of irrigation (reduced and conventional irrigation) and five nitrogen fertilizer levels (0, 70, 140, 210, and 280 kg N/ha). Reduced irrigation (20% reduction, from 280 to 220 mm) and nitrogen fertilizer (25% reduction, from 280 to 210 kg N/ha) addition did not significantly influence soil mite community and wheat yield. The relative abundances of fungivores and predators showed negative quadratic relationships with wheat yield, while that of plant parasites showed a positive relationship. The relationships between soil mite trophic groups and wheat yield revealed that we can evaluate the impacts of reduced irrigation and nitrogen fertilizer addition from the perspective of soil fauna. Soil mite community composition was altered by soil abiotic factors prior to reduced irrigation and nitrogen fertilizer addition. Overall, moderate reductions of irrigation and nitrogen fertilizer may not threaten to soil mite community and diversity or decrease crop production; in contrast, such reductions will benefit mite community development and the sustainable agriculture.

Furthermore, the composition of the mite community can be affected by soil environmental characteristics (e.g., soil organic matter, pH, and available phosphorus) via changes in exogenous inputs (Cao et al., 2011;Melguizo-Ruiz et al., 2017;Wang, Tan, Fan, Ruan, & Zheng, 2015). Therefore, monitoring soil mites responses may help us better understand the effects of different artificial disturbance on agroecosystems.
The North China Plain (NCP) is one of the major agricultural regions in China and provides approximately 80% of the nation's wheat (NBS, 2018). From 1978 to 2018, as a result of intensive agricultural management (e.g., mechanization, irrigation, and mineral fertilization), the annual wheat yield increased by 148% (from 54 to 134 million tons) and the average grain yield per unit area increased by 198% (from 1.85 to 5.48 t/ha) in China (NBS, 2018). However, side effects following increased crop yields include groundwater overuse and excessive nitrogen application. For example, in the NCP, annual precipitation is uneven and insufficient during the wheat grown season (Meng et al., 2012). To achieve high grain yields, additional water has been applied, which has resulted in a rapid decline in the groundwater table at an average rate of approximately 1 m per year in recent years (IGSNRR, 2016) and restricted sustainable agricultural development (Zhang et al., 2004). In addition to water shortages, excessive N application has also caused a series of environmental problems, including increased greenhouse gas emission, soil acidification, and biodiversity loss (Bardgett & Wardle, 2010;Guo et al., 2010;Tan et al., 2017). The amount of fertilizer applied is far greater than the wheat growth demand (325 kg N/ha; Ju et al., 2009), while the recommended nitrogen rate for wheat in the winter wheat-summer maize rotation systems in the NCP is 203 to 230 kg/ha (Liu et al., 2016). Therefore, there is an urgent need to identify adequate combinations of reduced irrigation and nitrogen application to ensure sufficient yield with relatively lower environmental costs for the government and researchers (Chen et al., 2014;Ju et al., 2009).
Water and nitrogen are two of limiting factors of crop growth, and changes in the levels of water and nitrogen applied to soil could influence ecosystem function and organisms community composition in agroecosystems (Liang et al., 2009;Paungfoo-Lonhienne et al., 2015;Zhou, Wang, Zheng, Jiang, & Luo, 2017). Soil organisms (including microorganisms, nematodes, collembolas, and mites) play crucial roles in nutrients cycling, crop growth, and ecosystem processes. Therefore, it is important to understand the different response of soil organisms to water and nitrogen management in order to predict their sensitivity to agricultural production measures.
Changes in water and nitrogen fertilizer addition may affect soil organism communities through different mechanisms. Although many studies had been conducted to evaluate the effects of water and nitrogen fertilizer addition on soil organism community, many variations and inconsistencies can be found in the literature (Melguizo-Ruiz et al., 2017;Sun et al., 2017). For example, elevated water addition increased soil fungi and decreased soil bacteria, while increased nitrogen fertilizer addition had the opposite effects because water addition led to significant changes in soil characteristics (Ma et al., 2016). Moreover, the effects of water and nitrogen fertilizer addition on the soil nematode community were complex and lack consistency. For instance, water addition enhanced the abundance and species richness of soil nematodes, but nitrogen fertilizer addition reduced their generic richness (Song et al., 2016), and the effect of different levels of nitrogen fertilizer addition on soil nematode also fluctuated (Chen et al., 2019;Song et al., 2015). In addition, although several studies have evaluated the combined effects of water and nitrogen fertilizer addition on soil organisms , few studies have researched the response of soil mites to changes in water and nitrogen fertilizer application, with the exception of two studies in forest ecosystem (Lindberg & Persson, 2004;Melguizo-Ruiz et al., 2017).
To investigate the hypothesis that potential changes in soil physicochemical properties caused by reduced irrigation and nitrogen fertilizer addition affect mite community composition and then the wheat yield, a long-term field manipulative experiment with reduced irrigation and nitrogen fertilizer addition treatments was conducted in a wheat farmland ecosystem. Monitoring soil mites responses to reduced irrigation and nitrogen fertilizer addition may help us better understand the effects of different artificial disturbance on agroecosystems. The objectives of the present study were (a) to evaluate the changes in wheat yield and mite community composition under reduced irrigation and nitrogen fertilizer addition treatments, (b) to identify the main drivers of the soil mite community (e.g., reduced irrigation addition, reduced nitrogen fertilizer addition, or soil physicochemical properties), and (c) to examine how potential changes in mite community composition affect wheat yield.

| Experimental setup
A field experiment was conducted at the Yucheng Comprehensive Experimental Station in southwestern Shandong Province, China (36°57′N, 116°36′E, 20 m elevation). The experimental site is located in a typical temperate semiarid zone with a mean annual temperature of 13.3°C and an annual rainfall of 559.8 mm according to long-term observations (1998Zhao, Li, Ai, Li, & Gu, 2017).
The soil is classified as a Calcaric Fluvisols according to the FAO-Uneson system and contains 12.9% sand, 65.1% silt, and 22.0% clay.
At the beginning of the experiment, the surface soil (0-20 cm) had a pH of 8.3 (H 2 O 1:2.5) as well as 12.2 g/kg soil organic matter (SOM) and 0.5 g/kg total nitrogen (TN).
The experiment followed a split-plot design with irrigation and nitrogen fertilizer addition in a wheat and maize rotation system established in 2005. There were ten treatments with the combinations of two levels of irrigation addition and five levels of nitrogen fertilizer, with three replicates for each treatment. The plots were hydrologically isolated with partition walls filled with cement to prevent unexpected species moving between the individual plots.
The main plots received the conventional irrigation (CIR) and reduced irrigation (RIR) treatments (80% of CIR), and subplots were randomly assigned to five levels of nitrogen fertilizer addition (0, 70, 140, 210, and 280 kg N/ha in the grown season, denoted N0, N1, N2, N3, and N4, respectively). The CIR and N4 treatments were set according to the typical applications used by local farmers. Irrigation was applied four times over the wheat growing season, and nitrogen fertilizer was added in the form of urea, which was applied prior to wheat sowing (50% of the total dose) and at the jointing stage (50% of the total dose). For all treatments, a total of 120 kg/ha P 2 O 5 (as Ca(H 2 PO 4 ) 2 ) and 80 kg/ha K 2 O (as K 2 SO 4 ) were applied in each plot before sowing.

| Sampling and measurement of grain yields and environmental parameters
Soil samples were collected from each plot before wheat harvest in 2014 and 2015. In each subplot, nine soil cores (2.5 cm diameter) were collected from a depth of 0-20 cm and combined. The soil samples were divided into two subsamples: one was used to extract soil mites, and the other was sieved through a 2-mm sieve and air-dried for soil environmental parameters analysis. Wheat yield was evaluated on the day of harvest. Wheat grains were harvested from three areas measuring 1 m 2 (1 m × 1 m) in each plot and oven-dried at 60°C, after which the dry weight was determined. Soil environmental parameters were measured and have been described elsewhere (unpublished). Briefly, the soil organic matter (SOM) content was not significantly influenced by reduced irrigation, nitrogen fertilizer addition, or their interaction (generalized linear mixed models [GLMMs], p > .05). In contrast, the available nitrogen content (AN) and carbon to N ratio (CN ratio) presented an initial increase followed by a decrease in response to nitrogen fertilizer addition under CIR addition, and the same trend was observed for total nitrogen (TN) content under RIR addition. Furthermore, the soil water content (SWC) decreased from 16.1 ± 0.7% to 10.3 ± 0.9% in the reduced irrigation and nitrogen fertilizer addition treatments. Soil pH decreased significantly in response to the interaction effect of elevated irrigation and nitrogen fertilizer addition, but the decrease was small.

| Statistical analysis
Data analyses were conducted using R software (2018, version 3.5.1, R Foundation for Statistical Computing, https ://www.rproje ct.org/). The data were log-transformed prior to analyses to improve normality. The effects of reduced irrigation and nitrogen fertilizer addition on response variables (wheat yield, soil mite abundance, and soil mite community composition) were examined

| Correlations between reduced irrigation addition, reduced nitrogen fertilizer addition or mites and wheat yields
Changes in nitrogen fertilizer addition significantly influenced wheat yields (Figure 1a,b; p < .05), whereas reduced irrigation addition and the interaction between reduced irrigation and nitrogen fertilizer addition showed no effects on wheat yield (Figure 1a,b; p > .05).
Compared with N4, the reduced nitrogen fertilizer treatments (N3, N2, N1, and N0) changed wheat yield by −83.9% to 3.4% in 2014 and −80.4% to 16.6% in 2015. In addition, the wheat yields were similar in the N2, N3, and N4 application treatments, with the highest production in N3 treatments, while yield was significantly lower in the N1 and N0 application treatments than in the treatments with the other three levels of nitrogen fertilizer addition. Reduced irrigation and nitrogen fertilizer addition and their interactions did not alter the total abundances of soil mites (Figure 1c,d) or the relative abundances of mite trophic groups (Figure 2). Wheat yield exhibited quadratic relationships with the relative abundance of all mite trophic groups except omnivores ( Figure 3). Specifically, an inverted bell-shaped relationship was observed between the yield and the relative abundance of plant parasites (y = 0.022x 2 + 0.294x + 1.150; R 2 = .125, p = .023), a bell-shaped relationship between yield and the fungivore relative abundance (y = −0.170x 2 + 1.478x−0.062; R 2 = .132, p = .018), and a bell-shaped relationship between yield and predator relative abundance (y = −0.224x 2 + 0.900x + 2.340; R 2 = .140, p = .014), respectively (Figure 3a). We also found a significantly negative linear relationship between the relative abundances of plant parasites and predators (y = −0.406x + 5.095; R 2 = .136, p = .004; Figure 3b).

| Community composition of soil mite under reduced irrigation and nitrogen fertilizer addition treatments
The reduced irrigation and nitrogen fertilizer addition treatments significantly altered the relative abundances of some mite taxa (Table 1) (Table 1).

| Relationship between soil environmental parameters and mite community composition
According to the Mantel test and partial Mantel test, the soil mite community composition showed significant correlations with F I G U R E 1 Changes in wheat yields (a and b) and total soil mite population (c and d) under reduced irrigation and nitrogen fertilizer addition treatments (CIR, conventional irrigation; RIR, reduced irrigation; N0, N1, N2, N3, and N4, N fertilizer addition were 0, 70, 140, 210, and 280 kg N/ha, respectively; *, ** The significant differences at p < .05 and p < .01, respectively; ns, no significant differences; The different letters indicate significant differences at the same irrigation level [p < .05]) sampling year and soil environmental parameters but no correlations with reduced irrigation addition or reduced nitrogen fertilizer application (Table 2). Redundancy analysis (RDA) revealed that the variation in the soil mite community composition was significantly correlated with soil environmental parameters based on Monte Carlo test (F = 2.479, p = .002; Figure 6). The RDA ordination plot clearly demonstrated the relationships between soil environmental parameters and soil mite community composition, which had eigenvalues of 0.159 for the first axis and 0.022 for the second axis. The vectors showed that SOM, TN, and the CN ratio play a great role than SWC, pH, and AN in driving soil mite community composition. Stepwise regression further indicated that the soil mite community composition was mainly explained by SOM, TN, and the CN ratio (Table 3). In particular, the relative abundances of Epilohmania (Epilohmanniidae), Stigmaeidae, Laelapidae, and Digamasellidae were correlated with only the CN ratio, and the relative abundance of Ceratozetes (Ceratozetidae) was correlated with only TN. The relative abundances of Truncopes (Oribatulidae), Trhypochthonius (Trhypochthoniidae), and total nymphs could be explain by the SOM, TN, and the CN ratio, and the relative abundance of Bella was correlated with TN and the CN ratio. Furthermore, the relative abundance of Tetranychidae was correlated with SWC, pH, and the CN ratio, while that of Scutacaridae was correlated with both the SWC and the CN ratio.

| D ISCUSS I ON
Our results showed that wheat yields were not significantly decreased by moderately reduced nitrogen fertilizer addition (280-140 kg N/ha at increments of 70 kg N/ha), and maximum yields were obtained with 210 kg N/ha except under conventional irrigation treatments in 2014. The effect of reduced nitrogen fertilizer addition on wheat yields was similar to that in a previous data analysis in which the maximum yield was achieved at approximately 210 kg N/ ha in NCP (Chen et al., 2014;Liu et al., 2016). This result suggests that an appropriate decline in nitrogen fertilizer addition will not reduce the wheat grain yield but will increase nitrogen partial factor productivity and reduce environmental hazards (Chen et al., 2014;Ju et al., 2009). The reduced irrigation treatments (20% lower than conventional irrigation addition) did not have a more significant negative effect on wheat grain production than the conventional irrigation treatments. The wheat yields in the reduced irrigation treatments were the same (2014) or higher (2015) than those in the conventional irrigation treatments. Similar results were obtained in previous studies in which reduced irrigation addition saved water use and advanced the wheat flowering date, which was helpful to prevent agricultural disaster (e.g., dry hot wind) and increasing yield (Meng et al., 2012;Wang et al., 2016Wang et al., , 2018. According to the abovementioned positive effects of reduced irrigation and nitrogen fertilizer addition on grain yield, our study demonstrated the response of soil mite community composition to reduced irrigation and nitrogen fertilizer addition treatments in two continuous sampling years in a long-term field trial. The reduced irrigation treatments increased the relative abundances of Acaridae and Laelapidae, suggesting that soil water content was F I G U R E 2 The relative abundance of the four soil mite trophic groups under reduced irrigation and nitrogen addition treatments in 2014 (a) and 2015 (b), respectively (CIR, conventional irrigation; RIR, reduced irrigation; N0, N1, N2, N3, and N4, N fertilizer addition were 0, 70, 140, 210, and 280 kg N/ha, respectively) F I G U R E 3 The correlations between the relative abundance of four soil mite trophic groups (fungivore, plant parasite, omnivore, and predator) and the wheat yields for reduced irrigation and nitrogen fertilizer addition treatments  important in affecting soil mite abundance (Olear & Blair, 1999).
This finding was consistent with that a previous study demonstrated that mites were less abundant in irrigated plots (Olear & Blair, 1999). In contrast, the results of a short-term experiment showed that irrigation had no effects on mite abundance (Liu, Li, Liu, & Yang, 2017;Pressler, Foster, Moore, & Cotrufo, 2017), whereas in other treatments with increased water addition in grassland ecosystems, there were significant positive effects on mite communities (Chikoski, Ferguson, & Meyer, 2006;Wu et al., 2014). The inconsistent impacts of irrigation addition on soil mites suggested that the responses of mite to changes in irrigation addition are independent of taxa and the duration and vegetation type in the experiment (Blankinship, Niklaus, & Hungate, 2011). Our results also revealed that reduced irrigation and nitrogen fertilizer addition had interactive effects on the relative abundances of Microdispidae, Allosuctobelba (Suctobelbidae), Bella (Damaeidae), and Tetranychidae. A previous study found that low-dose nitrogen fertilizer in combination with irrigation had slightly positive effects on mites (Lindberg & Persson, 2004). Few studies have investigated the response of soil mite community to combined irrigation and nitrogen fertilizer addition. Therefore, more relevant studies are needed to evaluate the specific response of soil mites to reduced irrigation and nitrogen fertilizer addition in various ecosystems.
In the present study, soil mite community structure was grouped separately by sampling year, and the mite abundance was lower in 2014 than in 2015. We suspected that the decline was correlated with the lack of resources available in soil (e.g., debris and exudates from wheat) to mites in wheat field. The growth of wheat is influenced by multiple factors, such as precipitation, soil fertility, and climate disasters (Telfer et al., 2018;Zhang, Wang, & Niu, 2019 the soil mite community (Lemanski & Scheu, 2014;Strickland, Wickings, & Bradford, 2012).
The impacts of reduced irrigation and nitrogen fertilizer addition on soil mite taxa distribution were presented by heatmaps in this study. The distribution of soil mites was inconsistent between 2014 and 2015, which suggested that the soil mite community may not be directly influenced by the irrigation and nitrogen fertilizer addition. Multiple studies have shown that soil properties shape the soil mite community structure (Cao et al., 2011;Nielsen, Osler, Campbell, Burslem, & Van Der Wal, 2010;Wang et al., 2015). Our results revealed that soil environmental parameters (e.g., SOM and TN) significantly influenced soil mite community composition. RDA analysis and stepwise regression analysis were used to explore the influences of soil properties and to reveal the factors that affected the soil mite taxa. We found that SOM, TN, and the CN ratio strongly influenced the soil mite community, an effect that was previously explained by supplying of food resources (Taylor & Wolters, 2005;Wickings & Grandy, 2011). This finding is consistent with previous studies in which SOM explained a significant part of the variation in  Note: Bold font indicates the significant mantel score (r).
F I G U R E 6 Redundancy analysis of soil properties and soil mite community composition (AN, available nitrogen; CN, carbon and nitrogen ratio; SOM, soil organic matter; SWC, soil water content; TN, total nitrogen) soil mite community distribution (Gao, He, Zhang, Liu, & Wu, 2014).
Reduced irrigation and nitrogen fertilizer addition had no effects on the soil properties and soil mite community in this study, whereas the SOM, TN, and the CN ratio had significant effects on the relative abundance of soil mite taxa, both individually and in combination.
Based on previous studies Wu et al., 2014), we suspect that reduced irrigation and nitrogen fertilizer addition primarily changed soil mite community by affecting SOM, TN, and the CN ratio. To clarify the influence of reduced irrigation and nitrogen fertilizer addition on soil mite community composition, more relevant studies should be conducted in the future.
We explored the potential correlations between soil mite trophic groups and wheat yields, which revealed the interactive relationships between the mite community composition and wheat yields.
Soil mite community could be influenced by plant growth, as plants provide habitats and/or resources for soil organism through inputs of chemical compounds and organic matters (Bardgett & Wardle, 2010). In this study, the mite trophic groups had significant nonlinear relationships with wheat yields, suggesting that the effects of soil mites on crops may through indirect pathways (Dirilgen et al., 2016;Ito, 1971;Maraun et al., 2014). Previous studies showed that fungivorous mites and parasitic mites contributed to the decomposition of organic matter in the soil from roots and other organic sources for soil nutrients releasing and activation (Siepel & Maaskamp, 1994), and the feeding activity of fungivores plays a key role in increasing soil aggregate stability and soil formation (Maaß et al., 2015). The effects of soil mite trophic groups on wheat yields maybe mainly indirectly change soil nutrients or regulate other soil organism community (Koehler, 1999;Maraun et al., 2014). Additional, plant growth was influenced by management practices, biotic and abiotic factors, the effects of soil mite on wheat yields are also lack of relevant researches. Therefore, it is necessary to conduct more studies to evaluate the effects of soil mite and combined other soil organism on wheat yields.

| CON CLUS IONS
This study explored the effects of the combined reduction in irrigation and nitrogen fertilizer addition on wheat grain yield and the soil mite community composition in a long-term field trial. The results indicated that reasonable irrigation and nitrogen fertilizer reduction will not decrease the wheat yields or impact the soil mite community composition, which were 20% and 25% lower, respectively, than those under conventional irrigation and nitrogen fertilizer addition.
Wheat yields were affected by changes in the relative abundances Abbreviations: CN, carbon and nitrogen ratio; SOM, soil organic matter; SWC, soil water content; TN, total nitrogen. TA B L E 3 Variables responsible for the changes in soil mite community composition of fungivores, plant parasites, and predators. Soil mite community taxa were more directly responsive to changes in soil environmental parameters than to irrigation and nitrogen fertilizer addition. Our results suggest that the impacts of reduced irrigation and nitrogen addition on soil mite taxa may occur indirectly through changes in soil environmental parameters.

ACK N OWLED G M ENTS
We are grateful to Dr. Yongsheng Zhang and the staffs of the Yucheng Experimental Station, Chinese Academy of Sciences for their valuable help during the experiments. We thank Dr. Guo Zhang for his helpful advices to the article writing.

CO N FLI C T O F I NTE R E S T
None declared.

AUTH O R CO NTR I B UTI O N S
CYZ identified soil mite, analyzed the data, and wrote the paper. FG and FOY designed and oversaw the study. FOY, CYZ, and XHL did field sampling. ZOY, JHM, and FHZ assisted with the field work.

DATA AVA I L A B I L I T Y S TAT E M E N T
We agree to make our data publicly available in a relevant reposi-