Interactive effects of nitrogen and water addition on soil microbial metabolic limitation in 1 temperate desert shrublands

24 Aims Soil microbes play critical roles in regulating the turnover of soil organic carbon (SOC) 25 and nutrients, and microbial metabolic limitation should draw more attention in desert 26 ecosystems. However, soil extracellular enzymes activity (EEA) response and microbial 27 metabolic limitation to atmospheric N deposition and increased precipitation in desert- 28 shrubland are still poorly understood. 29 Methods The study examined the effects of long-term (9 year) N and water additions (i.e., 5 g 30 N m −2 yr −1 , 30% ambient precipitation increase and their combination) on EEAs and soil 31 microbial resource limitation, as well as explored their controlling factors in the Gurbantunggut 32 Desert in northwestern China. 33 Results The results showed that N and water additions significantly enhanced soil EEAs and 34 considerably aggravated microbial phosphorous (P) limitation. Water addition and the N-water 35 combination addition alleviated carbon (C) limitation, but N addition alone strengthened 36 microbial C limitation. The interaction of N and water additions relieved the negative impact 37 of N addition on soil microbial C limitation, and positively aggravated microbial P limitation. 38 Soil microbial C limitation was primarily driven by soil moisture and organic C concentration, 39 while the soil microbial N/P limitation was chiefly controlled by soil water and available P 40 contents. Conclusions The influences biogeochemical processes may be altered by their concurrent occurrence. Overall, these 43 findings highlight water availability is more effective at modifying microbial metabolisms than N accumulation in desert ecosystems. Altogether, this may help to predict how terrestrial C and nutrient flow could be induced by global change factors..

Each pan was erected and constructed at a slight angle with an area of 1.9 m × 1 m. In total, 18 149 precipitation pans were set up covering equally 30% of the total area of each plot. Precipitation 150 was collected in plastic containers after each rainfall incident, and then sprinkled onto the 151 corresponding plot by hand during the late afternoon or early morning (Huang et al., 2015b). 152 Therefore, the W and NW plots received 30% additional natural precipitation, the amount of 153 increase that is currently predicted for northern China (Liu et al., 2010). N fertilizer was applied 154 twice a year in early April and July in the liquid form of NH4NO3 with a total input of 5 g N 155 m −2 every year, which was diluted in 15 L of distilled water and evenly sprayed onto the 156 corresponding plots. Distilled water with the same amount of N fertilizer was also applied to 157 the CK and W treatments. The experiment began in September 2010, while N fertilizer and the 158 water addition treatments were applied yearly staring at the beginning of 2011.

159
Soil sample collection and related soil physicochemical measurements.

160
After 9 years of continuous experimental treatment, 10 soil cores (0-10 cm) from each 161 plot were collected and randomly mixed to form a composite sample after removing the plant 162 debris layer using a 2-mm sieve in August 2019. The soil samples were taken into a portable 163 cooler box and transported to the laboratory for future measurements. Then, the soil samples 164 were divided into two parts: one subsample was air dried at room temperature for physio-165 chemical analysis, while the other subsample was used for microbial analyses at 4 °C. 166 Furthermore, soil samples were collected from each quadrat to measure the soil bulk density 167 and soil moisture.

168
About 100 g of fresh soil was used to determine soil moisture content (SM, %), which was 169 oven-drying at 105 °C for 48 h. Soil bulk density (BD, g cm -3 ) were determine by the ring 170 cutting method. Soil pH was estimated in a 5:1 ratio of CO2-free water to soil with a pH meter 171 (PHS-3C pH acidometer, China). Soil organic carbon (SOC) content were analyzed with the 8 H2SO4-K2Cr2O7 by calorific boiling and combustion method (Nelson and Sommers, 1982). Soil 173 total nitrogen (STN) content was determined by the Kjeldahl method (Bremner, 1996). 174 Dissolved organic carbon (DOC) was measured by wet-oxidation method was extracted and 175 measured a TOC analyzer. The contents of NO3 --N and NH4 + -N (Soil available N, AN) were 176 measured by a flow injection analyzer. Molybdenum blue method with an ultraviolet 177 spectrophotometer (Hitachi UV2300) to determine the soil total P (TP) and available P (AP) 178 content, which were extracted with H2SO4-HClO4 and sodium bicarbonate, respectively (Olsen
The microbial N/P limitation (vector angles).

225
(°) = ( 2(x, y)) (7)  the strongest effect by 15.0% compared with the ambient N. NW addition significantly reduced 286 the enzyme C:P ratio ratio compared to the N addition, but there were no significant differences 287 between the W and NW additions ( Fig. 2A). N addition significantly increased the enzyme C:P 288 ratio ratio, but W and NW additions significantly decreased the enzyme C:P ratio ratio ( Fig.   289 2B). N, W and NW additions significantly reduced the enzyme N:P ratio ratio by 10.1%, 13.3%, 290 13 and 18.7%, respectively. N and W additions showed no significant interaction effects (P > 0.10) 291 on enzyme C:N ratio, but exerted significant interactive effects on the of enzyme C:P and N:P 292 ratios. Furthermore, there were significant relationships between the changes in the C-acquiring 293 enzymes and N-acquiring enzymes (R 2 =0.61, P =0.008, Fig. 2D), C-acquiring enzymes and P-294 acquiring enzymes (R 2 =0.76, P < 0.001, Fig. 2E), and ratios of N-acquiring enzymes and P-295 acquiring enzymes (R 2 =0.77, P < 0.001, Fig. 2F). significantly decreased microbial C limitation compared to that of W addition alone, but it had 306 exerted significant higher on microbial C limitatio compared with that of W addition (Fig. 2B).

307
N and W addition significantly and interactively affected the microbial vector angles (Fig.3C).

308
The vector angle was significantly higher under N, W and NW additions than the ambient by 309 18.1%, 26.5%, and 39.7%, respectively, and the vector angles were all > 45°, suggesting that 310 the soil microbial metabolism was limited by soil P. NW addition demonstrated the largest 311 microbial P limitation (55.8±0.5°), and significantly increased microbial P limitation as 312 compared to the N a or W additions alone (Fig.3C).

313
Furthermore, N-induced increases in C and P limitations were lower in the watered plots 14 limitation in the watered plots (Fig. 4A). The water-induced declines in C limitation were 316 greater when fertilizer was applied, and the water-induced decreases in P limitation were lower 317 in both N and ambient N plots (Fig. 4B). These results suggest that water may alleviate the C 318 induced by N enrichment, while N and water addition in combination may decrease P limitation 319 than their alone. In addition, the microbial C limitation was negatively correlated with microbial 320 P limitation in response to N and water additions (Fig.3D, P<0.05). The VPA revealed that the soil nutrients (25.8%) and physical properties (24.7%) 326 explained much a substantially greater portion of the variation in microbial C limitation than 327 the nutrient ratios (10.4%) or microbial properties (6.2%) (Fig. 5A). The physical properties 328 and soil nutrients showed strong effects on soil microbial N/P limitation (Fig. 5B). The stepwise 329 regression models detected that SM, SOC and N/P were the most influential factor on microbial 330 C limitation (Table 3), and SM, SAP, AN/SAP and MBC were the most important factors 331 affecting the soil microbial N/P limitation (Table 3).

332
The SEM analysis demonstrated that soil moisture and SOC directly and indirectly 333 affected microbial C limitation, and N/P exhibited direct effects on microbial C limitation (Fig.   334 6A). SM, SAP, AN/SAP, and MBC directly and indirectly affected on microbial N/P limitation, 335 and MBC exhibited direct effects on microbial N/P limitation (Fig. 6B). SM and SOC exerted 336 significantly negative direct effects on microbial C limitation (Fig. 6A), while SM and SAP 337 significantly positively, and directly affected microbial N/P limitation (Fig. 6B). SM was the 338 common significant driver of microbial C and N/P limitation. The correlation analysis also 339 revealed that microbial C limitation were negatively correlated with SM and SOC but positively correlated with N/P. The microbial N/P limitation were positively correlated with SM and 341 AN/SAP but negatively correlated with SAP and MBC (Fig. S1). correlations between soil pH and N-acquiring enzyme activities were also recorded (Fig. S1).

359
In contrast to N addition, water addition caused a decline of the soil N-acquiring enzyme Water addition exhibited a greater impact on inducing soil EEAs (except specific C-367 acquiring enzyme activities, Fig.1), due to the water-stress being more severe than the N 368 limitation in the desert environment. Moreover, it was demonstrated that N and water additions 369 in combination significantly improved soil EEAs as compared to N addition alone (Fig. 1), as addition masked the positive effects by N deposition in C-acquiring enzyme activities (Fig. 1A), 377 offset the negative effects on the N-acquiring enzyme (Fig. 1B), and enhanced the positive 378 effects on P-acquiring enzyme (Fig. 1C)   ( Fig.4), the present study found that N and W additions in combination have the greatest and 440 interactive effects on microbial P limitation (Fig. 3C). Several mechanisms may explain the  . In general, the study found that soil moisture was the 502 most common driver of microbial C and P limitations, because water availability is the primary 503 limiting factor on microbial metabolic activity in desert ecosystems.

505
This study revealed that N and water additions significantly enhanced EEAs. N addition 506 aggravated microbial C and P limitations, whereas water addition was found to intensify 507 microbial P limitations. N and water additions showed interactive effects on microbial 508 metabolic limitation. The interactions of N and W could offset the impacts of N deposition and was found to decrease soil microbial C limitation while strengthening the effects of N and water 510 addition alone, resulting in an increased P limitation. Moreover, soil microbial C limitation was 511 chiefly driven by soil moisture and SOC in the desert shrublands, whereas soil microbial N/P