Responses of Soil C-, N-, and P-Hydrolyzing Enzyme Activities to N and P Addition in an Evergreen Broad-Leaved Forest in Southwest China

Human activities-mediated input of nitrogen (N) and phosphorus (P) to ecosystem may signicantly affect soil hydrolyzing enzyme activities (Hy-EAs). However, the mechanisms underlying the responses of soil Hy-EAs to change in N and P availability remains unclear. Here, a two-year eld N and P addition experiment was conducted in a subtropical evergreen broad-leaved forest to elucidate the effects of N addition, P addition, and NP co-additions on soil Hy-EAs and biochemistry properties. The invertase, cellulase, and acid phosphatase activities were increased in N treatment but reduced in P treatment. The urease activity was reduced in N treatment but did not alter in P treatment. NP treatment signicantly increased the invertase and cellulase activities. Furthermore, the cellulase activity was positively correlated with soil organic carbon concentration. The acid phosphatase activity was negatively correlated with microbial biomass carbon (MBC), total P, and available P concentrations. Whereas the urease activity was not strongly dependent on total N concentrations, but positively correlated with soil and MBC. These Hy-EAs were signicantly correlated with C-to-P and N-to-P ratios, while no signicantly correlation with ratio.


Introduction
Nitrogen (N) and phosphorus (P) are the essential elements of the biology, and generally regarded as the two limiting factors for the net primary productivity of the forest ecosystem (Elser et al. 2007; Zhang et al. 2015). However, with the intensi cation of human activities (e.g., the large consumption of fossil fuels and fertilizers), the input rates of N and P in the global terrestrial ecosystem have been greatly increased and still maintained an upward trend (Vitousek et al. 1997; Galloway 2000Galloway , 2004; Du et al. 2016), but compared to the anthropogenic N input, the amount of P input is far less (Galloway et al. 2004; Mahowald et al. 2008). The imbalance of N and P input may transform the forest ecosystem that was originally restricted by N into a state of N saturation, enhance the dependence of forest primary productivity on P (Burns et al. 2013; Zhu et al. 2016), and greatly affects the availability of carbon (C) and nutrients in the soil , especially in tropical/subtropical forests ecosystems which are generally limited by P (Cleveland et al. 2013; Du et al. 2016).
Soil hydrolytic enzyme activities (Hy-EAs) re ects the relationship between plant and microbial nutrient demand and soil nutrient supply (Sinsabaugh et al. 2008;Hill et al. 2012), and has been shown to be closely related to pH, microbial biomass, C, N, and P concentrations and their stoichiometric characteristics (Rodríguez-Loinaz et al. 2008;Sinsabaugh et al. 2008;Burns et al. 2013). Changes in soil pH, microbial activity and N and P availability caused by human-induced N and P input not only have a direct impact on soil enzyme activities (Waring et al. 2014;Shi et al. 2018), but also may regulate enzyme activities indirectly by changing the microbial allocation to soil enzymes production and the abundance of speci c enzyme-producing microorganisms (Sinsabaugh and Moorhead 1994;Allison et al. 2008; Yokoyama et al. 2017). However, the response of soil Hy-EAs to N and/or P inputs are complex and still uncertain.
In recent years, the effect of N addition or deposition on soil enzymes activities has been widely studied (Sinsabaugh et Yokoyama et al. 2017) effects of P addition on C-, N-, and P-Hy EAs have been reported. These different results indicate that the effect of P addition on soil C-, N-, and P-Hy EAs are site-speci c. More research are needed to enhance our understanding of the underlying mechanism change in soil P availability affect hydrolytic enzymes activities.
Here, we conducted a two-year eld N and P addition experiment in a natural evergreen secondary broadleaved forest in Southwest China. In this study, we investigated the response of four soil Hy-EAs, i.e., invertase, cellulase, urease and acid phosphatase to N and/or P addition. We also simultaneously measured the soil pH, microbial biomass C, and nutrient (soil organic C, total N, total P and available P) concentrations. Our objective is to evaluate how the N and/or P addition affect soil C-, N-, and P-Hy EAs in subtropical evergreen broad-leaved forests, and to explore soil hydrolytic enzymes activities correlations with the soil MBC, pH, nutrient concentrations, and stoichiometric characteristics under the N and/or P addition. According to the production of soil enzymes follows the resource allocation theory (Allison and Vitousek 2005;Allison et al. 2007), enzyme producers tend to synthesize and secrete enzymes related to the decomposition of elements that limit their own growth. The N and/or P addition may increase the availability of N and/or P in the soil, but will increase soil microbial demands for C. Therefore, we hypothesized that: (1) N treatment would increase the soil invertase, cellulase and acid phosphatase activities, while decrease the soil urease activity. (2) P treatment would increase the soil invertase, cellulase and urease activities, while decrease soil acid phosphatase activity. (3) NP treatment would increase the soil invertase and cellulase activities, while decrease the soil urease and acid phosphatase activities.

Study site
The study was conducted in the Bi Feng Gorge Science spot, Ya'an City, Sichuan Province, China (103°00′25″E, 30°04′ 06″N, and 970 m a.s.l.). The region is located in a subtropical moist forest zone with a monsoon climate, mean annual precipitation is approximately 1658 mm (average from 1971 to 2017), which has distinct dry and wet seasons. About 60% of the rainfall from the wet season (June to August) and only 4.4% from dry season (December to February), the mean annual temperature is 16.4˚C (Zhou et al. 2018). More details on study sites were described in Wei et al. (2020) .

Experimental design
The experiment was established with a complete randomized block design. Three replicate blocks were established in September 2017 and each block was separated by 20 meters. In every block, four plots (5 m× 5 m) were established and separated by a 2-m-wide buffer zone. With reference to the levels of N and P addition in the previous study (Jing et al. 2016;Zhou et al. 2017a), four treatments were established in this study: control (Ctr: no N or P addition), N addition (N, 100 kg N ha −1 year −1 ), P addition (P, 50 kg P ha −1 year −1 ), and N and P co-addition treatment (NP, 100 kg N ha −1 year −1 +50 kg P ha −1 year −1 ). NH 4 NO 3 and NaH 2 PO 4 solutions were used as fertilizers, and were sprayed below the canopy monthly from October 2017. During each application, the supplies were dissolved in 2 L of water and sprayed onto the forest oor using a hand-held sprayer. The control plots received 24 L of water without supply each year, this equivalent to an annual precipitation increase of 0.96 mm, which was negligible (Wei et al. 2020).

Field sampling
Surface soil (0-20cm) samples were collected using an auger (20 cm in height, 9 cm inside diameter) in April, July and October 2018, January, April, July and October 2019 as well as January 2020. Three cores were randomly collected from each plots, the stones and coarse roots in cores were removed, thoroughly mixed and sieved using a 2-mm mesh sieve, then divided into two subsamples. One subsample was stored at 4 ℃ and used to measure the soil hydrolase activity (measure within one week), the other was air-dried at ambient temperature for the chemical properties measurements.

Soil enzyme activities and chemical properties
The assays of the enzyme activities were based on the colorimetric determination of the products released by the enzymes (Guan 1986 invertase and cellulase were expressed by determined the released reducing sugars. Urease (EC 3.5.1.5) activity was determined using 10% urea as the substrate under standard conditions (24h at 37℃), based on the number of enzymatic products to analyze the activity. Acid phosphatase (EC 3.1.3.2) activity was measured using 0.5% disodium phenyl phosphate as the substrate and incubated at 37℃ for 2 h. Enzyme activities were expressed as the micro molar of speci c products produced by soil enzymes in unit time unit dry soil mass (μmol·g -1 ·h -1 / nmol·g -1 ·h -1 ).
Soil pH was determined by a glass electrode in a soil/water (1:2.5) suspension, soil organic C (SOC) was determined using dichromate oxidation and titration with ferrous ammonium sulfate, the total N (TN) was determined by Kjeldahl digestion, and the total P was determined by molybdenum-antimony colorimetry (Allen et al. 1974). The soil available P (AP) was extracted using 0.72 mol L −1 sodium acetate anhydrous and 0.52 mol L −1 acetic acid and analyzed calorimetrically using a spectrophotometer (UV2550, Shimadzu, Japan). Soil microbial biomass carbon (MBC) and were measured using the chloroform fumigation extraction technique via the total C analyzer (Wu et al. 1990). .

Statistical analyses
All variables were rst tested for a normal distribution and homogeneity of variance using the Shapiro-Wilk test and Levene's test, respectively. The linear mixed effects models employing the restricted maximum likelihood estimation method were used to determine the main effects of the N or/and P addition, sampling time and their interactions on the four soil Hy-EAs (invertase, cellulase, urease, and acid phosphatase), chemical properties (SOC, TN, TP and AP concentrations, and pH) and MBC. The N addition, P addition, and sampling time were designated as xed effects, and the plots nested in blocks as random effects. After establishing the general linear model, the Bonferroni adjustment method was used to the con dence intervals and signi cance values to account for multiple comparisons. Linear regression analysis was applied to evaluate the relationship between soil C-, N-, and P-Hy EAs and soil MBC, pH, SOC, TN, TP (or AP) concentrations and their stoichiometric proportion. Prior to linear regression analysis, parameters were Ln-transformed to maintain the same order of magnitude. All statistical analyses were performed in SPSS 20.0 (SPSS Inc., USA).

Soil hydrolytic enzymes
During the 2 years experiment, soil invertase and acid phosphatase activities did not show signi cant seasonal dynamic (Fig.1a, d). The activities of the soil cellulase and urease were seasonally dynamic, the maximum activities were observed in July and October, respectively; and the minimum occurred in January and April, respectively (Fig.1b, c). The linear mixed model showed that the soil invertase, cellulase, urease, and acid phosphatase activities were signi cantly affected by the main effect of N addition, P addition and sampling time (P < 0.05, Table 1). Notably, the interaction effect between N and/or P addition and sampling time signi cantly affected soil invertase, urease, and acid phosphatase activities (P < 0.05), while did not signi cantly affect soil cellulase activity (P > 0.05, Table 1).
Multiple comparisons showed that, compared with the control, the mean activities of soil invertase and cellulase were signi cantly increased by 27.5% and 25.2% in the N treatments, respectively; and NP treatment signi cantly increased the mean activities of invertase and cellulase by 30.9% and 20.3%, respectively (P < 0.05, Fig.1a, b). On the contrary, P treatment signi cantly decreased the mean activities of invertase and cellulase by 10.1% and 13.6%, respectively (P < 0.05, Fig.1a, b). There was no signi cant difference in the mean activity of urease under the P and NP treatments (Fig.1c), but signi cantly decreased by 11.2% in N treatment (P < 0.05, Fig.1c). Notably, we found that short-term (12 months) N treatment promoted urease activity, but it was reduced with continuous N addition (Fig.1c). The mean activity of soil acid phosphatase was not signi cantly affected by the NP treatment (Fig.1d), but signi cantly increased by 37.8% in N treatment and decreased by 18.4% in P treatment (P < 0.05, Fig.1d).

Discussion
Effects of N addition on soil hydrolytic enzymes In agreement with our rst hypothesis and some previous studies (Wang et al. 2008;Heuck et al. 2018;Jia et al. 2020), N treatment signi cantly increased soil invertase and cellulase mean activities (Fig.1a, b), indicating that N addition had a positive effect on the process of soil invertase and cellulase participating in soil unstable C metabolism (Buchkowski et al. 2015;Sun et al. 2016). The increase of soil invertase and cellulase activities may be related to soil limited target nutrients, that is, N addition increases the N availability relative to C in the soil (Chen et al. , 2018b). Using the excessive N resources, plant roots and microorganisms can synthesize C-hydrolyzing enzymes (such as invertase and cellulase) to accelerate soil labile C mineralization Heuck et al. 2018). Notably, the increase of soil total C:P and N:P ratios after N addition maybe also promote soil invertase and cellulase activities, because we found signi cant correlation between soil invertase activity with total N:P ratio (Fig.2c), and between soil cellulase activity with C:P ratio (Fig.2d). These results support the premise that soil stoichiometric characteristics plays an important role in the regulation of enzymes activities (Fanin et al. 2015; Wang et al. 2020b). However, we found N treatment signi cantly increased SOC concentration in our study ( Table 2), indicating that the increase of soil invertase and cellulase activities did not reduce the accumulation of soil SOC. The increased of SOC concentration could be results from the N-induced suppression of soil microbial biomass (Table 2)  ). In addition, we found signi cant correlation between soil cellulase activities with SOC concentration (Fig.2a), this means that the increase of cellulase activity in soil may be related to the increase of substrate utilization (Pancholy and Rice 1973; Katsalirou et al. 2010). Therefore, we believe that although N addition increased SOC concentration, the soil microorganisms growth may also be limited by the C availability, and the increase of soil C-Hy EAs (such as invertase and cellulase) maybe promote the soil unstable C mineralization to improve the availability of C and thus to feed back the growth of microorganisms.
As expected, we found that N treatment signi cantly decreased the soil urease mean activity in this study (Fig.1c), which in line with some previous N addition experiments (Ajwa et al. 1999; Kang and Lee 2005; Feng et al. 2020). The inhibition effect of N addition on soil urease activity may be explained by the following possible mechanisms. On the one hand, we found that N addition reduced microbial the biomass and soil pH ( Table 2), given that signi cantly positively correlations between soil urease activity with soil pH (Fig.2f) and MBC concentration (Fig.2g), N addition may lead to a decrease in urease activity via N-induced reduction of microbial biomass and the acidi cation (Table 2) (Fig.1c), positive effect and no respond also were observed during the rst year (Fig.1c). Which mean that N addition may suppressed the soil urease activity by reducing its production, while the process required a certain time to make the urease activity at a lower level (Lloyd and Sheaffe 1973;Burns et al. 2013;Nannipieri et al. 2018). This maybe well explained that there was no signi cant negative correlation between soil urease activity and TN concentration (Fig.S1d) Supporting our rst hypothesis, we found that soil acid phosphatase activity increased in N treatment (Fig.1d) Tie et al. 2020b). In this study, N treatment signi cantly increased soil total C:P and N:P ratios (Table 2), indicating the increase of soil C and N concentrations led to the lack of P in the soil (Tian et al. 2010;Zechmeister-Boltenstern et al. 2011). We also found that soil acid phosphatase activity was positively correlated with total C:P (Fig.2j) and total N:P (Fig.2k) ratios, meaning that with the increase of soil C and N concentrations or the decrease of soil P concentrations, soil acid phosphatase activity will increase. Therefore, the increase of soil acid phosphatase activity may be the result of excessive N resources being used by microbial and plants for the synthesis of phosphatase to enhancing the release of inorganic P from organic P compounds (

Effects of P addition soil hydrolytic enzymes
In agreement with our second hypothesis, we found that P treatment signi cantly decreased soil invertase and cellulase mean activities (Fig.1a, b). This result was consistent with the results found in other subtropical forests, where C-Hy EAs were strongly inhibited by P addition ( invertase and cellulase activities. This was also supported by our results that soil invertase and cellulase activities had signi cant positive correlation with soil total C:P ratio (Fig.2b). Third, P addition may reduce in C-hydrolyzing enzymes abundance in soil. On the one hand, P addition may reduce the gene abundance of enzymes that degrade labile C substrates in the soil and decreased the expression of Chydrolyzing enzymes (Yao et al. 2018). On the other hand, the changes in microbial composition caused by P addition may also reduce soil C-hydrolyzing enzymes abundance. Previous studies have shown that soil MBC had no signi cant response to P addition in subtropical broad-leaved forests, but greatly increased the relative abundance of symbiotic fungi (Liu et al. 2012), which may reduce the production of C-hydrolyzing enzymes in the soil, because fungi acquire C from host plants rather than producing C degrading enzymes (Gartner et al. 2012). In addition, in this study, we also found that the addition of P signi cantly reduced the SOC concentration (Table 2), which may be caused by the addition of P to increase the microbial activity in the soil, resulting in a higher rapid of soil organic matter transformation (Liu et al. 2012;Wang et al. 2020c). Therefore, it is not surprising that SOC concentration decreases even if C-Hy EAs is inhibited under P addition. This result may imply that the increase in soil P availability will have a negative impact on soil C sequestration compared with the positive effects of N addition (Reay et al. 2008;Stiles et al. 2017).
Contrary to our expectation, there was no signi cant effect of P treatment on soil urease mean activity in this study (Fig.1c), meaning that maybe there was no close relationship between soil N demand and urease activity after P addition (Olander and Vitousek 2000). Consistent with our results, P addition had no signi cant effect on soil urease and other N-hydrolyzing enzymes activities, as indicated by a recent meta-analyze (Xiao et al. 2018) and some experiments (Wang et al. 2020a;Zhang et al. 2020). We speculated that the neutral effect of P addition on soil urease activity observed in this study may be related to the N status of our forest ecosystem. In our study region, the annual average N deposition (wet) reached 95 kg ha -1 (Xu et al. 2013), the higher N deposition may provide su cient available N for the forest ecosystem in this region (Zhou et al. 2017a). When there was enough N to meet the needs of plants and microorganisms, the production of microbial N-hydrolyzing enzymes will return to the constituent level (Chróst 1991), and soil N-Hy EAs may not change with P addition. In addition, since N is abundant and short-term P addition may not affect the soil N supply (Zhu et al. 2015), a short-term P addition might prevent us from detecting statistically signi cant changes in soil urease activity. From this, compared with the strict control of phosphatase activity by N addition, the regulation of Nhydrolyzing enzymes by P addition is much weaker in our experimental forest.
In our study, we found P treatment signi cantly decreased soil acid phosphatase mean activity (Fig.1d), which was consistent with our second hypothesis and the general pattern in most terrestrial ecosystems that P addition have inhibitory effect on soil phosphatase activities ( In our study, we also found that P treatment signi cantly reduced the soil total C:P ratio (Table 2). Compared with the higher soil total C:P ratio, the lower soil total C:P ratio after P addition indicated that increased availability of P in the soil, plant and soil microorganisms growth were not restricted by P (Cleveland and Townsend 2006; Mooshammer et al. 2012), the competition between plants and soil microorganisms for P therefore was weakened, thus reducing the microbial dependence on organic P mineralization as the main source of P (Rodríguez and Fraga 1999; Wang et al. 2007b). That is, when the supply of P is su cient, microorganisms will reduce the investment in phosphatase and the rate of P mineralization will decrease(Marklein and Houlton 2012; Yokoyama et al. 2017). This was also supported by our results that soil AP and TP concentrations had a very signi cant negative correlation with soil acid phosphatase activity (Fig.2h, i).
Effects of NP co-addition soil hydrolytic enzymes In our study, NP treatment signi cantly increased soil invertase and cellulase mean activities (Fig.1 a-b), which is consistent with our third hypothesis. We also found that NP treatment signi cantly increased soil pH and MBC (Table 2). Therefore, we speculated that the increase in soil invertase and cellulases activities caused by NP treatment might be the result of the increased demand for availability C by soil biological growth (Cleveland et al. 2002;Li et al. 2015). However, unexpectedly, we found NP treatment did not signi cantly affect the SOC concentration (Table 2), in our study, possible explanation could be the positive effects of N addition on SOC sequestration counteracted by the negative effects of P addition on SOC sequestration, indicating that the SOC mineralization was limited by soil P availability, and P addition could mitigate the effects of N on soil C sequestration in this system (Bradford et al. 2008;Street et al. 2018). In contrast to our third hypothesis that NP treatment would inhibit soil urease and acid phosphatase activities, we found that NP treatment did not signi cantly change soil urease and acid phosphatase mean activities in our study (Fig.1c, d). The results showed that that under the combination N and P addition, the addition of P might counteract the inhibition of N availability on soil urease activity and the promotion of acid phosphatase activity in subtropical forest soil. We also found that soil N and P concentrations and their stoichiometric ratios were not signi cantly affected by NP treatment (Table 2), which may explain the minor response of soil urease and phosphatase activities to NP treatment in this study. Consistent with our results, recent meta-analyses also indicated that NP co-addition signi cantly increased soil C-Hy EAs but the effect on soil N-and P-Hy EAs were not signi cantly (Xiao et al. 2018).
Different from the effect of alone N or P addition, the effect of NP co-addition on soil hydrolytic enzymes activities may also be the result of the interaction effect of N and P, because there is a widespread close coupling relationship between N and P in the ecosystem ( Indeed, our results also showed that the interaction of N and P addition had a signi cant effect on soil invertase, urease, and acid phosphatase activities (Table 1), the interactions found in this study may be attributed to the opposing responses of alone N and P addition on soil C-, P-, and N-Hy EAs, and further research is needed.

Conclusions
In summary, in our study, N addition increased soil invertase, cellulose, and acid phosphatase activities, but decreased soil urease activity. In contrast, P addition decreased soil invertase, cellulase, and acid phosphatase activities, while not affect soil urease activity. The responses of N and P addition on soil hydrolase activity may be related to the change of soil nutrient availability and microbial biomass.
Overall, in our studied subtropical evergreen broad-leaved forest, ongoing N deposition may increase soil C sequestration, but will result in the limitation of soil C and P availability. Although the exogenous P input may reduce the soil C sequestration, while alleviates the limitation of soil C and P availability caused by the long-term of N deposition.

Declarations
The authors declare that they have no known competing nancial interests or personal relationships that could have appeared to in uence the work reported in this paper.