Science of the Total Environment 515–516 (2015) 83–91

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Phosphatase activity in relation to key litter and soil properties in mature subtropical forests in China Enqing Hou a,b, Chengrong Chen b,⁎, Dazhi Wen a,⁎, Xian Liu b a b

Key Laboratory of Vegetation Restoration and Management of Degraded Ecosystems, South China Botanical Garden, Chinese Academy of Sciences, Guangzhou 510650, China Griffith School of Environment, Griffith University, Nathan, Qld 4111, Australia

H I G H L I G H T S • The potential activities of AcPME and PDE were positively correlated in all horizons. • AcPME activity was significantly lower in the L than the F/H horizon while PDE activity was comparable between them. • The relationships between phosphatase activities and key edaphic properties vary with horizon.

a r t i c l e

i n f o

Article history: Received 8 October 2014 Received in revised form 12 February 2015 Accepted 12 February 2015 Available online xxxx Editor: Charlotte Poschenrieder Keywords: Acid phosphomonoesterase Phosphodiesterase Microbial biomass C Fine root biomass C:N:P stoichiometry Mature forest

a b s t r a c t Phosphatase-mediated phosphorus (P) mineralization is one of the critical processes in biogeochemical cycling of P and determines soil P availability in forest ecosystems; however, the regulation of soil phosphatase activity remains elusive. This study investigated the potential extracellular activities of acid phosphomonoesterase (AcPME) and phosphodiesterase (PDE) and how they were related to key edaphic properties in the L horizon (undecomposed litter) and F/H horizon (fermented and humified litter) and the underlying mineral soil at the 0–15 cm depth in eight mature subtropical forests in China. AcPME activity decreased significantly in the order of F/H horizon N L horizon N mineral soil horizon, while the order for PDE activity was L horizon = F/H horizon N mineral soil horizon. AcPME (X axis) and PDE (Y axis) activities were positively correlated in all horizons with significantly higher slope in the L and F/H horizons than in the mineral soil horizon. Both AcPME and PDE activities were positively related to microbial biomass C, moisture content and water-holding capacity in the L horizon, and were positively related to soil C:P, N:P and C:N ratios and fine root (diameter ≤ 2 mm) biomass in the mineral soil horizon. Both enzyme activities were also interactively affected by forest and horizon, partly due to the interactive effect of forest and horizon on microbial biomass. Our results suggest that modulator(s) of the potential extracellular activity of phosphatases vary with horizon, depending on the relative C, P and water availability of the horizon. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Human-induced elevation of atmospheric CO2 concentration and increases in nitrogen (N) deposition have led to a strong imbalance with phosphorus (P), conferring an increasingly important role of P availability in the function of natural ecosystems (Peñuelas et al., 2012). Phosphatase-mediated P mineralization is one of the critical processes in P biogeochemical cycling and determines P availability in forest ecosystems (Marklein and Houlton, 2012; Olander and Vitousek, 2000) as well as other natural ecosystems (Marklein and Houlton, 2012; Turner and Haygarth, 2005). Identification of the key

⁎ Corresponding authors. E-mail addresses: c.chen@griffith.edu.au (C. Chen), [email protected] (D. Wen).

http://dx.doi.org/10.1016/j.scitotenv.2015.02.044 0048-9697/© 2015 Elsevier B.V. All rights reserved.

environmental factors regulating phosphatase activity is critical for advancing our understanding of the dynamics and availability of P in forest ecosystems (Henry, 2012; Nannipieri et al., 2011). Phosphatases (e.g. acid phosphomonoesterase) have been found in both plants and microbes, with an exception that alkaline phosphomonoesterase has only been found in microbes (Nannipieri et al., 2011). Of the phosphatases present in soil, phosphomonoesterases are the most widely studied; usually, acid phosphomonoesterase prevails in acidic soils whereas alkaline phosphomonoesterase prevails in alkaline soil (Nannipieri et al., 2011). Phosphomonoesterases act on a range of low molecular weight P compounds with monoester bonds (namely phosphomonoesters), including mononucleotides, sugar phosphatases, and polyphosphates, and can also catalyze the hydrolysis of low-order inositol phosphatases (Nannipieri et al., 2011; Turner and Haygarth, 2005). Phosphodiesterases are a group of enzymes

84

E. Hou et al. / Science of the Total Environment 515–516 (2015) 83–91

involving the hydrolysis of P compounds with diester bonds (namely phosphodiesters) such as phospholipids and nucleic acids, which constitute the majority of the fresh organic P inputs to soil (Bunemann, 2008; Koukol et al., 2006; Turner and Haygarth, 2005). Hydrolysis of a phosphodiester is initiated by phosphodiesterase to release a phosphomonoester, which must then be hydrolyzed by phosphomonoesterase to release a free phosphate for biological uptake (Bunemann, 2008; Turner and Haygarth, 2005). The production of phosphatase is mainly regulated by the demand for P by organisms and environmental P availability (Olander and Vitousek, 2000; Sinsabaugh and Follstad Shah, 2012), in addition to other edaphic factors (e.g. C and N availability) (Allison and Vitousek, 2005; Marklein and Houlton, 2012; Naples and Fisk, 2010). Soil phosphatase activity is usually inversely related to soil P availability as a result of negative feedback by soil available P to the production and activity of phosphatase (Olander and Vitousek, 2000), but may be positively related to the amount of soil extractable organic P due to the enhancement of phosphatase activity by the availability of hydrolyzable organic P sources (Tarafdar and Claassen, 1988). Soil phosphatase activity is usually reported to be positively related to soil C and N availability (Allison and Vitousek, 2005; Marklein and Houlton, 2012), mainly because the production and release of phosphatase are significant C- and N-consuming processes (Allison and Vitousek, 2005). After its release into soil, phosphatase is subject to degradation and stabilization (Nannipieri et al., 2011; Sinsabaugh and Follstad Shah, 2012) and is affected by a number of edaphic factors such as soil organic matter (Zornoza et al., 2007, 2007), contents of clay (Nannipieri et al., 2011) and moisture (Henry, 2012), and soil pH (Turner, 2010). Since extracellular phosphatase activity can be regulated by multiple environmental factors, the relationships between phosphatase activity and key litter and soil properties vary widely in different terrestrial ecosystems (Marklein and Houlton, 2012; Sinsabaugh et al., 2008) in relation to site background C, N, and P availability (Allison and Vitousek, 2005; Olander and Vitousek, 2000), variation in organic matter (or C) content with soil depth (Chen et al., 2000; Sinsabaugh et al., 2008), or organism composition, which affects nutrient assimilation and allocation (Naples and Fisk, 2010). Moreover, the extracellular activities of different phosphatases may be dissimilarly regulated by the same edaphic factors (e.g. soil total organic P) (Turner, 2010; Turner and Haygarth, 2005), probably as a result of their different functions in P mineralization (Turner and Haygarth, 2005) and/or their different vulnerabilities to degradation and stabilization (Turner, 2010). Recent studies suggest that soil P availability is a significant limiting factor in the growth of both plant and soil microorganisms in natural subtropical forests in China (Hou et al., 2012; Huang et al., 2013). Phosphatase-mediated P mineralization is considered to be one of the significant processes governing soil P availability in these forest ecosystems (Huang et al., 2013). However, what drives soil phosphatase activity in these forests remains elusive. This study sampled forest floor litters (separated into L and F/H horizons) and the underlying soil at the 0–15 cm depth from eight subtropical forests in China. The potential extracellular activities of acid phosphomonoesterase and phosphodiesterase and contents of total C, total N, total P, microbial biomass C, and moisture were determined. Soil extractable inorganic P and organic P fractions and fine root biomass in the 0–15 cm soil horizon were also determined. The main objective was to identify the edaphic properties that could significantly modulate AcPME and PDE activities in litter and soil in mature subtropical forests using a relationship analysis approach. In particular, we were interested in whether and how phosphatase activities might be related to C:N:P stoichiometry in litter and soil, as an increasing number of studies have suggested that soil C and N availability can have impacts on soil P availability by influencing the soil phosphatase activity (Allison and Vitousek, 2005; Marklein and Houlton, 2012).

2. Materials and methods 2.1. Site description The study was conducted in the Dinghushan Biosphere Reserve, located in the middle of Guangdong Province in China (112°31′ E to 112°34′ E, 23°09′ N to 23°12′ N). The Reserve covers an area of 1155 ha and has a typical subtropical monsoon climate. More than 1800 plant species have been identified and documented across the Reserve. Mean annual temperature at the site is about 21 °C, and mean annual precipitation is around 1900 mm. Nearly 80% of the precipitation falls in the wet season (from April to September) and 20% in the dry season (from October to March). Elevation ranges from 10 to 1000 m above sea level. Surface forest soils have developed from Devonian sandstone and shale during the Holocene (Shen et al., 2001), and are classified as Ferralsols according to the FAO classification. Table 1 shows the basic information of eight selected forests, which are within a straight-line distance of 4 km from each other in Dinghushan, China. They represent five major types of forests and cover a large range of topographic conditions in south China. All sites have been protected well since their establishment, except for the pine forest site which suffered frequent harvest of litter and understory by local residents between the 1930s and early 1990s (Mo et al., 1995). Harvest of the understory (mainly seedlings of broadleaved species and herbs) has impeded the succession of pine forest to pine and broadleaved mixed forest at this site (Mo et al., 1995). 2.2. Sampling and sample preparation Litters and soils were sampled from eight selected forests in October 2010, at the start of dry season. At each site, four subplots (20 × 20 m2) were randomly set up with a distance of at least 10 m between them. In each subplot, three sampling areas (20 × 20 cm2) were randomly selected with the constraints that they were 1–2 m away from the nearest tree (diameter at breast height ≥ 5 cm) and at least 5 m from their nearest neighbor. Fine litter (senesced branches, bark, flowers and fruits with diameters ≤ 1 cm, and leaf litter) within the sampling area were fully collected and separated into the L horizon (undecomposed litter) and F/H horizon (mixture of partly decomposed litter and amorphous humus) in the field. After the litters were collected, a soil profile was excavated in the bare soil area. Soil at 0–15 cm depth was sampled by three successive cutting rings (height: 5 cm; volume: 100 cm3) from top to bottom (5 cm depth per cutting ring). The three litter samples from the same layer of each subplot were bulked together as one composite sample (one composite L horizon sample and one composite F/H horizon sample per subplot). Nine soils (three cutting rings per area × three areas) from each subplot were bulked together to form one composite soil sample. Both litter and soil samples were stored at 4 °C in the refrigerator within 4 h after sampling. For each litter sample, fresh weight was measured and then the sample was mixed well. A subsample was oven-dried at 65 °C to constant weight and weighed for the determination of moisture content, after which the oven-dried subsample was ground for the determination of total C, total N, and total P. The rest of the fresh litter was cut into small pieces and then stored at 4 °C prior to the determination of microbial biomass C, enzyme activities, and water-holding capacity. For each soil sample, fresh weight was measured, and then the sample was mixed well and sieved to pass through a 4-mm mesh, with stones of diameter N4 mm picked out and weighed. A first subsample of the sieved soil was oven-dried at 105 °C to constant weight for the determination of moisture content. A second subsample was stored at 4 °C for the determination of microbial biomass C, enzyme activities, and water-holding capacity. A third subsample was weighed and air-dried for 2 weeks prior to grinding and determination of soil chemical properties. The remaining sieved fresh soil was weighed and used for the calculation of fine root biomass. In general, living fine roots

E. Hou et al. / Science of the Total Environment 515–516 (2015) 83–91

85

Table 1 Basic site characteristics of eight study forests in Dinghushan, China. Forest

Approximate stand age (year)

Aspect

Slope

Altitude (m)

Major tree species

Disturbed pine forest (PF) Pine and broadleaved mixed forest 1 (PBM1) Pine and broadleaved mixed forest 2 (PBM2) Pine and broadleaved mixed forest 3 (PBM3) Ravine evergreen broadleaved forest 1 (REB1) Ravine evergreen broadleaved forest 2 (REB2) Monsoon evergreen broadleaved forest (MEB) Mountainous evergreen broadleaved forest (MTEB)

80 80 80 80 300 300 400 100

SE SE NE NE E W NE NE

25° 30° 40° 25° 20–25° 30–40° 30° 30°

50–150 50–200 150–200 300–350 200–250 100–150 250–300 550–600

Pinus massoniana, Schima superba, and Euodia lepta P. massoniana, S. superba, and Castanea henryi P. massoniana, S. superba, and C. henryi P. massoniana, S. superba, and C. henryi Aporosa yunnanensis, Cryptocarya concinna, and Ormosia fordiana Gironniera subaequalis, Sterculia lanceolata, and Caryota ochlandra S. superba, C. henryi, and C. concinna Engelhardia roxburghiana, Rhododendron henryi, and Machilus breviflora

(distinguished from the dead fine roots by their colors; diameter ≤ 2 mm) retained on a 0.6 mm screen were collected. The collected fine roots were rinsed with distilled water and then oven-dried at 65 °C to constant weight. 2.3. Analytical methods Microbial biomass C of both litters and soils was determined by a fumigation–extraction method (moist weight to volume ratios of 1:25 and 1:5 for litter and soil samples, respectively) (Vance et al., 1987). Total C and total N were determined using an Isoprime isotope ratio mass spectrometer with a Eurovector elemental analyzer (IsoprimeEuro EA 3000). Total P was measured using nitric acid/perchloric acid digestion followed by colorimetric analysis (Murphy and Riley, 1962) using a UV1800 (Shimadzu, Japan). Potential acid phosphomonoesterase (AcPME, EC 3.1.3.2) activity was assayed by the standard method of Tabatabai and Bremner (1969) using p-nitrophenyl phosphate (Sigma N4645) as substrate at a temperature of 37 °C. Potential phosphodiesterase (PDE, EC 3.1.4.1) activity was assayed by the standard method of Browman and Tabatabai (1978) using bis(p-nitrophenyl) phosphate (Sigma N3002) as substrate at a temperature of 37 °C. Potential activities of both enzymes were expressed as micromoles of p-nitrophenol produced per gram of soil or litter per hour (μmol pNP g−1 h−1) at 37 °C. Soil P fractions were extracted using sequential fractionation mainly according to McDowell and Condron (2000). Briefly, 3.0 g of air-dried soil was sequentially extracted by 30 ml of 1.0 M NH4Cl (defined as soluble P), 30 ml of 0.1 M NH4F (aluminum-associated P), 30 ml of 0.1 M NaOH (iron-associated P), and 30 ml of 0.5 M H2SO4 (calcium-associated P). Inorganic P concentration in the extracts of 0.1 M NH4F (Al–Pi), 0.1 M NaOH (Fe–Pi), and 0.5 M H2SO4 (Ca–Pi) was directly measured using the molybdate blue method (Murphy and Riley, 1962). Inorganic P concentration in the 1.0 M NH4Cl extract (soluble Pi) was below the detection limit of the molybdate blue method and was therefore reanalyzed using the malachite green method (Ohno and Zibilske, 1991). Organic P in the 0.1 M NH4F (Al–Po) and 0.1 M NaOH (Fe–Po) extracts was calculated as the difference between total P determined after persulfate digestion (Ormaza-González and Statham, 1996) and inorganic P. The soil P fraction data are cited from Hou et al. (2014). All data presented in this study are based on an oven-dried weight (65 °C for plant materials and 105 °C for soil). 2.4. Statistical analyses Litter and soil C:N:P ratios were all calculated on a mass basis. Differences in bulk density, fine root biomass and P fractions in the mineral soil horizon among forests were examined by Tukey's hsd post hoc test after one-way analysis of variance (ANOVA). Enzyme activities and other edaphic properties were determined for all three horizons and whether they were significantly affected by forest type, horizon, or their interaction were examined by two-way ANOVAs with forest and horizon as fixed factors. As forest × horizon interaction was

statistically significant (P b 0.05) for all of these edaphic properties (Supplementary data), differences in them among forests were further explored by Tukey's hsd post hoc test after one-way ANOVA for each horizon, and differences in them among horizons were further tested by Friedman's 2-way ANOVA by ranks both for each forest and for all forests. Principal component analysis (PCA) with varimax rotation was performed for each horizon to explore the relationships among selected edaphic properties and was also performed to explore the relationships among shifts in selected edaphic properties from the L horizon to the F/H horizon (calculated as the ratio of L horizon value to F/H horizon value). All the data were tested for normal distribution with the Kolmogorov–Smirnov test before statistical analyses. All above analyses were performed using SPSS 19.0 for Windows. Relationships between soil AcPME and PDE activities and soil C:P and N:P ratios were also plotted and fitted by exponential growth equations using Sigmaplot 12.0. Comparison of slopes of the relations between AcPME and PDE activities among horizons was performed using reduced major axis (RMA) regression using SMATR 2.0 (Falster et al., 2006). 3. Results 3.1. Comparison among forests and horizons Main effects of forest and horizon and their interaction are all statistically significant for the selected parameters: AcPME and PDE activities, moisture content, water-holding capacity, microbial biomass C, litter biomass, total C, total N and total P and C:N:P ratios (P b 0.05, twoway ANOVA; Supplementary data). In the L horizon, AcPME and PDE activities were both significantly higher in the PF and REB1 than in the other forests; similar trends were also observed for microbial biomass C and moisture content (Table 2). Water-holding capacity was significantly higher in the REB1 than in the other forests (Table 2). Total N and total P were both highest in the REB1 and lowest in the PBM1 or PF; the opposite was generally true for total C, and C:N, C:P and N:P ratios (Table 3). In the F/H horizon, AcPME activity was significantly higher in the PBM2 than in the MEB; PDE activity was highest in the MTEB and lowest in the PBM1, PBM2, and MEB; microbial biomass C was significantly higher in the REB2 than in all other forests; moisture content, litter biomass and water-holding capacity were all highest in the PF among forests (Table 2). Total N and total P were significantly higher in the REB2 and PBM2 than in the PBM3; total C and C:N, C:P and N:P ratios were all lowest in the REB1 among forests (Table 3). In the 0–15 cm mineral soil horizon, AcPME activity was highest in the PBM3 and MTEB among forests; a similar trend was also observed for PDE activity (Table 3). Microbial biomass C, moisture content, water-holding capacity, total C, total N, total P, and N:P ratio all tended to be lower in the PF than in the other forests with most of the differences statistically significant (P b 0.05; Tables 2 and 3), while the opposite was true for bulk density (Table 4). Fine root biomass was significantly higher in the MTEB than in the PF and MEB (Table 4). All soil P fractions revealed generally similar differences among forests as

86

E. Hou et al. / Science of the Total Environment 515–516 (2015) 83–91

Table 2 Comparison of phosphatase activities and some other edaphic properties among forests and horizons.a Forest

AcPMEb (μmol p-NP g−1 h−1)

PDEb (μmol p-NP g−1 h−1)

Microbial biomass C (mg/g)

Moisture (g/g)

Water-holding capacity (g/g)

L horizon PF PBM1 PBM2 PBM3 REB1 REB2 MEB MTEB Mean

70.0 (16.8)aA 17.0 (4.6)bB 16.9 (4.8)bB 30.6 (6.6)bB 66.2 (18.3)aA 26.2 (11.7)bA 23.3 (5.4)bB 26.0 (5.6)bB 34.5 (22.3)B

3.7 (0.9)aA 0.9 (0.1)bA 0.9 (0.2)bA 1.4 (0.4)bA 3.4 (1.2)aA 1.5 (0.8)bA 1.8 (0.3)bA 1.8 (0.2)bA 1.9 (1.1)A

25.9 (3.4)aA 9.4 (1.5)bcA 5.2 (0.7)cB 4.9 (1.9)cA 24.9 (10.3)aA 10.8 (2.1)abcA 22.0 (5.4)abA 14 (13.6)abcA 14.7 (10.0)A

0.41 (0.1.)abB 0.18 (0.02)cC 0.19 (0.03)cC 0.30 (0.04)bcB 0.58 (0.19)aB 0.28 (0.04)bcC 0.29 (0.02)bcC 0.32 (0.05)bcB 0.32 (0.14)B

3.6 (0.7)bB 2.4 (0.3)bB 2.7 (0.3)bB 2.5 (0.3)bA 5.3 (0.9)aA 3.5 (0.6)bA 3.2 (1.0)bA 2.9 (0.6)bA 3.3 (1.1)A

F/H horizon PF PBM1 PBM2 PBM3 REB1 REB2 MEB MTEB Mean

60.5 (18.9)abB 49.8 (10.5)abA 68.9 (27.8)aA 57.2 (8.4)abA 46.1 (12.2)abA 40.3 (5.1)abA 33.8 (7.4)bA 65.1 (13.8)abA 52.7 (17.4)A

1.2 (0.3)bcB 0.7 (0.1)cA 0.9 (0.4)cA 1.9 (0.4)abA 1.2 (0.5)abB 1.0 (0.5)bcA 0.8 (0.2)cB 2.3 (0.7)aA 1.3 (0.7)A

4.8 (0.7)bB 5.9 (3.2)bA 12.8 (1.4)bA 6.6 (2.2)bA 13.5 (5.6)bA 23.2 (6.9)aA 10.7 (1.9)bB 11.0 (4.5)bA 11.1 (6.6)A

1.63 (0.29)aA 0.77 (0.21)bA 0.88 (0.22)bA 0.69 (0.03)bA 0.95 (0.08)bA 0.90 (0.06)bA 0.83 (0.24)bA 0.70 (0.11)bA 0.92 (0.33)A

5.6 (1.2)aA 3.5 (1.1)abcA 4.8 (1.1)abA 2.4 (0.5)cA 3.6 (0.9)abcB 2.6 (0.4)bcA 2.9 (1.4)bcA 3.4 (1.0)abcA 3.6 (1.4)A

0.18 (0.03)cC 0.31 (0.01)bB 0.32 (0.03)bB 0.35 (0.04)abB 0.32 (0.01)bC 0.39 (0.05)abB 0.41 (0.05)aB 0.35 (0.04)abB 0.33 (0.07)B

0.46 (0.13)cC 0.63 (0.02)abcC 0.62 (0.08)abcC 0.69 (0.08)abB 0.60 (0.12)bcC 0.77 (0.06)abB 0.83 (0.10)aB 0.74 (0.07)abB 0.67 (0.13)B

0–15 cm mineral soil horizon PF 5.1 (1.2)bC PBM1 11.4 (1.2)bC PBM2 10.3 (1.8)bC PBM3 26.2 (9.0)aB REB1 6.1 (2.3)bB REB2 7.6 (3.9)bB MEB 9.7 (3.9)bC MTEB 26.3 (6.2)aB Mean 12.8 (9.0)C

0.11 (0.04)cC 0.21 (0.02)abcB 0.17 (0.01)bcB 0.37 (0.18)abB 0.12 (0.06)cC 0.14 (0.06)bcB 0.20 (0.08)abcC 0.43 (0.16)aB 0.22 (0.14)B

0.4 (0.2)bC 0.9 (0.2)abB 1.0 (0.2)abC 0.9 (0.3)abB 1.1 (0.3)aB 1.1 (0.3)aB 1.2 (0.3)aC 1.2 (0.2)aB 1.0 (0.3)B

Litter biomass (Mg/ha) 4.2 (0.6)aB 3.4 (0.2)abB 3.6 (1.0)abB 2.3 (0.7)bcdB 1.4 (0.4)dB 1.3 (0.2)dB 2.9 (0.5)bcB 1.8 (0.4)cdB 2.6 (1.1)B

38.5 (8.7)aA 16.6 (3.0)bA 11.3 (4.5)bcA 12.2 (1.1)bcA 11.2 (5.2)bcA 5.2 (1.2)cA 7.6 (1.5)bcA 7.3 (2.9)bcA 13.7 (10.8)A

Different lowercase letters after data of the same horizon in the same column indicate significant (P b 0.05) difference in the property among forests. Different uppercase letters after data of the same forest in the same column indicate significant (P b 0.05) difference in the property among horizons. a Data are mean (SD), n = 4 for each forest and 32 for the mean. Full names of the eight forests are listed in Table 1. b AcPME: acid phosphomonoesterase activity and PDE: phosphodiesterase activity.

total P, with lowest values in the PF and highest values in the REB1 or REB2 (Table 4). Comparison among horizons showed that the decreasing order is F/H horizon N L horizon N mineral soil horizon for AcPME activity (Table 2) and total P (Table 3); L horizon = F/H horizon N mineral soil horizon for PDE activity, microbial biomass C and water-holding capacity (Table 2); F/H horizon N L horizon = mineral soil horizon for moisture content and litter biomass (no mineral soil horizon for litter biomass) (Table 2); and L horizon N F/H horizon N mineral soil horizon for total C, total N, and C:N, C:P and N:P ratios (Table 3). Significant forest × horizon interactions on all selected parameters except moisture content were mainly because of the dissimilar differences between the L and the F/H horizons in different forests (Tables 2 and 3 and Supplementary data). From the L horizon to the F/H horizon, AcPME activity decreased in the PF but increased in the PBMs, MEB and MTEB and did not change significantly in the REB1 and REB2; PDE activity decreased in the PF, REB1, and MEB but did not change significantly in the other five forests; microbial biomass C decreased in the PF but increased in the PBM2 and did not change significantly in the other six forests (Table 2). PCA on the shifts in selected parameters from the L horizon to the F/H horizon extracted three PCs with eigenvalues N1, which together accounted for 81.1% of the total variation (Table 5). The PC2 (accounted for 24.6% of the total variation) was strongly (factor loading score N 0.7) positively correlated with the shifts in AcPME and PDE activities, and microbial biomass C (Table 5). 3.2. Relationship analyses There was a positive correlation between AcPME and PDE activities in all horizons, with stronger correlations in the mineral soil and

L horizons than in the F/H horizon (Table 6 and Fig. 1). The RMA slope of the relationship between AcPME (X axis) and PDE (Y axis) activities was significantly higher in the L (0.051) and F/H (0.037) horizons than in the mineral soil horizon (0.015) (Table 6). PCAs showed that AcPME and PDE activities had similar relations with selected edaphic properties in each horizon; however, the pattern of relations between them and selected edaphic properties varied with horizons (Table 7). For the L horizon, three PCs had eigenvalues N1, which together accounted for 82.3% of the total variation (Table 7). The PC1 (accounted for 32.3%) was positively dominated by AcPME and PDE activities and was also strongly positively correlated with microbial biomass C, moisture content, and water-holding capacity (Table 7). For the F/H horizon, four PCs had eigenvalues N 1, which together accounted for 87.2% of the total variation (Table 7). The PC4 (accounted for 14.4%) was strongly positively correlated with AcPME and PDE activities but was not correlated significantly with any of other edaphic properties (P N 0.05; Table 7). For the mineral soil horizon, all soil P fractions were not correlated significantly with AcPME or PDE activity (P N 0.05; data not shown) and thus were not included in the mineral soil horizon PCA. Two PCs had eigenvalues N 1 in the mineral soil horizon PCA, which together accounted for 78.6% of the total variation (Table 7). The PC1 (accounted for 42.5%) was strongly positively correlated with microbial biomass C, water-holding capacity, and moisture, total C, total N, and total P contents and strongly negatively correlated with bulk density (Table 7). The PC2 (accounted for 36.1%) was strongly positively correlated with AcPME and PDE activities, and C:P, N:P, and C:N ratios (Table 7), and was also moderately positively correlated with fine root biomass (factor loading score 0.68). Plots of C:P and N:P ratios (as X axis) against AcPME and PDE activities (as Y axis) in the mineral soil horizon showed that these relationships were

E. Hou et al. / Science of the Total Environment 515–516 (2015) 83–91

87

Table 3 Comparison of C, N and P measures among forests and horizons.a Forest

Total C (mg/g)

Total N (mg/g)

Total P (mg/g)

C:N ratio

C:P ratio

L horizon PF PBM1 PBM2 PBM3 REB1 REB2 MEB MTEB Mean

495 (8)abA 499 (5)aA 486 (7)abA 456 (15)bcdA 427 (12)cdA 422 (36)eA 466 (21)abcA 481 (19)abA 467 (32)A

12.7 (0.6)bB 13.7 (1.8)bA 16.3 (0.5)abA 15.2 (1.2)bA 20.7 (3.2)aA 15.7 (2.3)abA 17.0 (4.2)abA 17.1 (1.8)abA 16.1 (3.1)A

0.30 (0.03)cdB 0.24 (0.04)dB 0.38 (0.03)bcdB 0.32 (0.01)bcdA 0.77 (0.11)aA 0.49 (0.11)bB 0.43 (0.14)bcA 0.39 (0.05)bcdA 0.41 (0.17)B

39 (2)aA 37 (5)abA 30 (1)bcA 30 (2)bcA 21 (3)dA 27 (4)cdA 29 (7)bcdA 28 (4)bcdA 30 (6)A

1648 (194)abA 2103 (322)aA 1290 (131)bcA 1420 (70)bA 562 (71)dA 893 (175)cdA 1181 (347)bcA 1240 (182)bcA 1292 (478)A

42 (4)abcA 57 (2)aA 43 (3)abcA 47 (3)abA 27 (4)cA 34 (11)bcA 44 (20)abcA 44 (5)abcA 42 (11)A

F/H horizon PF PBM1 PBM2 PBM3 REB1 REB2 MEB MTEB Mean

398 (29)aB 310 (74)abB 396 (49)aB 269 (15)abB 235 (32)bB 393 (65)aA 334 (31)abB 391 (98)aB 341 (78)B

14.9 (1)abA 11.4 (1.8)bcB 17.1 (2.1)aA 9.4 (0.8)cB 12.6 (2.4)bcB 14.7 (1.4)abA 12.9 (2.1)abcA 12.8 (2.7)abcB 13.2 (2.8)B

0.36 (0.04)cdA 0.35 (0.03)cdA 0.50 (0.05)bA 0.30 (0.03)dA 0.65 (0.06)aA 0.53 (0.08)bA 0.45 (0.04)bcA 0.34 (0.05)cdA 0.43 (0.12)A

27 (1)abB 27 (3)abB 23 (2)abB 29 (3)aA 19 (2)bB 27 (3)abA 26 (5)abA 31 (7)aA 26 (5)B

1123 (90)abB 878 (171)abB 793 (130)abcB 917 (107)abB 361 (52)cB 748 (108)bcA 751 (127)bcA 1214 (463)aA 848 (302)B

42 (3)aA 32 (4)abcB 34 (4)abB 32 (6)abcB 19 (3)cB 28 (2)bcA 29 (4)abcB 39 (13)abA 32 (9)B

0.6 (0.2)bC 1.4 (0.2)aC 1.7 (0.2)aB 1.4 (0.2)abC 1.7 (0.3)aC 2.1 (0.5)aB 2.2 (0.5)aB 1.9 (0.3)aC 1.6 (0.6)C

0.12 (0.01)dC 0.17 (0.01)cdC 0.25 (0.01)bC 0.13 (0.01)dB 0.31 (0.04)aB 0.26 (0.03)abC 0.27 (0.02)abB 0.22 (0.03)bcB 0.22 (0.07)C

15 (1)cC 20 (1)aC 18 (1)abC 17 (2)bcB 13 (1)dC 15 (1)cB 15 (1)cB 17 (0)bB 16 (2)C

80 (19)deC 167 (15)bcC 124 (11)bcdC 177 (34)aC 69 (8)eC 126 (26)bcdB 119 (22)cdB 153 (12)abcB 127 (41)C

0–15 cm mineral soil horizon PF 10 (3)bC PBM1 29 (4)aC PBM2 31 (4)aC PBM3 23 (4)aC REB1 22 (4)abC REB2 33 (9)aB MEB 33 (8)aC MTEB 34 (6)aC Mean 27 (9)C

N:P ratio

5.3 (1.1)cB 8.5 (0.7)abC 7.0 (0.7)bcC 10.7 (1.3)aC 5.4 (0.4)cC 8.2 (1.3)bB 8.0 (1.3)bC 8.8 (0.7)abB 7.7 (1.9)C

Different lowercase letters after data of the same horizon in the same column indicate significant (P b 0.05) difference in the property among forests. Different uppercase letters after data of the same forest in the same column indicate significant (P b 0.05) difference in the property among horizons. a Data are mean (SD), n = 4 for each forest and 32 for the mean. Full names of the eight forests are listed in Table 1.

better depicted by exponential growth curves than by linear curves (Fig. 2). 4. Discussion 4.1. Comparison of AcPME and PDE activities among horizons and forests AcPME activity in the L (16.9–70.0 μmol p-NP g−1 h−1, same unit for the below enzyme activity) and F/H (33.8–68.9) horizons were generally comparable to the values reported in a Douglas-fir forest in Canada (mixture of L horizon to H horizon 35.5; Pang and Kolenko, 1986) and a pine plantation in New Zealand (L horizon 21.8, F horizon 74.4; Chen et al., 2000) using similar analytic procedure. PDE activity in the L (0.9–3.7) and F/H (0.7–2.3) horizons covered the values in the pine plantation in New Zealand (L horizon 1.2, F horizon 2.3; Chen et al.,

2000). Mineral soil AcPME activity (5.1–26.3) was comparable to the previous reports in bulk soils in the study area (5.6–21.4; Huang et al., 2012a, 2013) and in a Jarrah forest in south-western Australia (4.7–29.7; Grierson and Adams, 2000) but tended to be higher than in a temperate woodland in northern China (1.8; Zhang et al., 2012), a Karri forest in Australia (0.7–9.4; Adams, 1992), and five natural forests in Spain (2.1–3.6; Zornoza et al., 2007). Shift in the activity of phosphatase as well as other extracellular enzymes along soil profile has been demonstrated by some previous studies, however the pattern has been diverse as related to the dissimilar shifts in microbial community in these studies (Chen et al., 2000; Šnajdr et al., 2008). Shifts in AcPME and PDE activities from the L horizon to the F/H horizon were partly explained by the shift in microbial biomass C. However, significant increase in AcPME activity but decreases, though not statistically significant, in PDE activity and microbial

Table 4 Comparison of fine root biomass, bulk density and P fractions in the 0–15 cm mineral soil horizon among forests.a Forest

Bulk density (g/cm3)

Fine root biomass (Mg/ha)

Soluble Pi fractionb (mg/kg)

Al–Pi fractionb (mg/kg)

Al–Po fractionb (mg/kg)

Fe–Pi fractionb (mg/kg)

Fe–Po fractionb (mg/kg)

Ca–Pi fractionb (mg/kg)

PF PBM1 PBM2 PBM3 REB1 REB2 MEB MTEB

1.35 (0.05)a 0.92 (0.06)bc 1.02 (0.14)bc 0.98 (0.08)bc 1.09 (0.08)b 0.89 (0.08)c 0.84 (0.07)c 0.88 (0.05)c

0.7 (0.5)b 1.7 (0.5)ab 1.7 (0.4)ab 1.5 (0.7)ab 0.8 (0.4)ab 1.2 (0.5)ab 0.8 (0.2)b 2.3 (1.2)a

0.04 (0.03)b 0.21 (0.1)ab 0.20 (0.14)ab 0.10 (0.02)ab 0.27 (0.10)a 0.25 (0.15)ab 0.27 (0.10)ab 0.16 (0.08)ab

1.3 (1.4)a 1.2 (0.4)a 1.6 (1.3)a 1.3 (0.8)a 2.5 (0.9)a 3.4 (0.7)a 2.9 (1.3)a 1.4 (1.0)a

3.4 (1.7)c 10.9 (2.6)ab 11.8 (1.3)ab 6.7 (2.6)bc 15.0 (2.9)a 14.5 (6.4)a 13.0 (1.2)ab 8.9 (2.4)abc

10.7 (4.0)b 13.0 (1.8)b 15.1 (2.3)b 11.6 (2.3)b 49.5 (16.6)a 25.8 (4.8)b 23.8 (2.3)b 17.6 (4.2)b

17.7 (1.8)e 41.0 (4.7)cd 50.2 (7.9)cd 33.7 (1.2)de 75.9 (13.7)ab 90.2 (10.5)a 77.0 (7.1)ab 57.3 (12.6)bc

2.6 (0.6)c 3.5 (0.1)c 5.4 (0.5)b 2.7 (0.3)c 8.7 (1.1)a 6.5 (1.0)b 5.9 (0.2)b 5.2 (1.0)b

Different letters after data in the same column indicate significant difference in the property among forests (P b 0.05). a Data are mean (SD), n = 4. Full names of the eight forests are listed in Table 1. b Soil P fraction data are cited from Hou et al. (2014).

88

E. Hou et al. / Science of the Total Environment 515–516 (2015) 83–91

Table 5 Principal component loadings for shifts in selected parameters from the L horizon to the F/ H horizon (calculated as the ratio of L horizon value to F/H horizon value).a

L/FH-AcPMEb L/FH-PDEb L/FH-microbial biomass C L/FH-moisture content L/FH-water-holding capacity L/FH-total C L/FH-total N L/FH-total P L/FH-C:N ratio L/FH-C:P ratio L/FH-N:P ratio % of variance a b

PC1

PC2

PC3

0.37 0.05 −0.12 0.78 0.73 0.24 −0.76 0.89 −0.76 −0.52 −0.08 32.3

0.82 0.90 0.86 0.27 0.27 0.18 0.24 0.04 0.43 0.06 −0.23 24.6

0.04 −0.02 −0.08 0.14 0.01 0.87 −0.57 −0.22 0.22 0.83 0.88 24.2

Loading scores N0.7 are shown in bold. AcPME: acid phosphomonoesterase activity; PDE: phosphodiesterase activity.

biomass C from the L horizon to the F/H horizon suggest that microbial community is undertaking different enzymatic processes due to the different decomposition state of litter. Similarly, an increase in the ratio of AcPME activity to PDE activity from L horizon (18.5) to F horizon (31.8) has been demonstrated by Chen et al. (2000). A possible explanation is that the F/H or F horizon has undergone decomposition and potentially has a higher ratio of phosphomonoesters to phosphodiesters than the L horizon (Gressel et al., 1996), which might stimulate microbial community to produce AcPME (Tarafdar and Claassen, 1988; Turner and Haygarth, 2005). This might also explain the overall larger decrease in PDE activity (35.2%) than in microbial biomass C (24.4%) from the L horizon to the F/H horizon. Similarly, smaller ratio of PDE activity to AcPME activity in the mineral soil than the L and F/H horizons might be also related to the typically lower ratio of phosphodiesters to phosphomonoesters in mineral soil than in litter (Gressel et al., 1996; Turner et al., 2014). The ratio of PDE to AcPME derived from plant fine roots might differ from that derived from soil microbes, potentially also causing a difference in the PDE to AcPME ratio between mineral soil and litter. Consistent with the previous studies (Chen et al., 2000; Šnajdr et al., 2008), both AcPME and PDE activities were significantly lower in the mineral soil than the L and F/H horizons, perhaps a combined result of the lower microbial biomass C, C and N availability, and readily available organic P in the mineral soil. For the L horizon, highest AcPME and PDE activities in the REB1 and PF can be partly explained by the relatively high microbial biomass C and moisture content in these forests. High L horizon total P (assumed to be mainly in organic form) might be also responsible for the high phosphatase activities in the REB1, as labile organic P could stimulate phosphatase activity via its substrate effect (Tarafdar and Claassen,

Table 6 Relationship between acid phosphomonoesterase (X axis) and phosphodiesterase (Y axis) activities in the L, F/H, and 0–15 cm mineral soil horizons.a Horizon

Mean L horizon F/H horizon Mineral soil horizon a

Interceptb

Slope c

0.051a 0.037a 0.015b

Low CI

High CI

Mean

Low CI

High CI

0.042 0.027 0.014

0.062 0.052 0.018

0.16 −0.71 0.02

−0.24 −1.41 −0.01

0.57 −0.01 0.05

r2

P

0.73 0.20 0.87

b0.001 0.010 b0.001

The fitting equation is: Y = Slope ∗ (X) + Intercept; n = 32; unit: μmol p-nitrophenol g−1 h−1. Reduced major axis intercepts and slopes are shown, as well as the correspondent lower and upper 95% confidence interval (CI), and r2. b Intercepts were not standardized to a common slope, and thus are not contrasted among horizons. c Significant differences (P b 0.05) in slopes among horizons are shown by the different letters.

Fig. 1. Relationships between acid phosphomonoesterase (AcPME) and phosphodiesterase (PDE) activities in the L, F/H, and 0–15 cm mineral soil horizons. The details of these relations using reduced major axis (RMA) regressions are presented in Table 6.

1988) and increasing P availability is likely to promote litter microbial biomass and activity in the study area (Hou et al., 2012; Liu et al., 2012). Relatively high moisture content in the REB1 was likely due to its high water-holding capacity, while relatively high moisture content in the PF might be attributable to both high water-holding capacity and large litter biomass in this forest. In the PF, long-term harvest practice possibly affected phosphatase activities in the L and also other horizons via its potential influences on the composition of microbial and plant communities (Henry, 2012; Mo et al., 1995). Higher mineral soil AcPME and PDE activities in the high-elevated (MTEB) rather than the low-elevated (REB1, REB2, and MEB) broadleaved forests was inconsistent with a recent study in northern China in which soil acid phosphatase activity was lower at high elevations (1280–2360 m a.s.l.) than at the low elevations (540–750 m a.s.l.) (Xu et al., 2015). This inconsistency was probably because soil C and N availability (P data not shown) decreased significantly with altitude in the study of Xu et al. (2015) while they did not change in this study. Relatively low soil P availability and relatively high soil C:P ratio probably contribute to the relatively high soil AcPME and PDE activities in the MTEB, as discussed in Subsection 4.2. Positive relationships between shifts in AcPME and PDE activities and microbial biomass C from the L horizon to the F/H horizon indicate that forest × horizon interaction on AcPME and PDE activities were partly due to the interaction on microbial biomass. The measured activity of enzymes can be also influenced by their association with surface reactive materials (e.g. organic acid) (Nannipieri et al., 2011; Rao et al., 2000), therefore the possibly differing extent of litter decomposition and C compounds composition among forests might also contribute to the forest × horizon interaction on AcPME and PDE activities. Moreover, plant species and decomposition state might interactively act on the composition of microbial community and thus on the extracellular activity of enzymes in litter and soil (Ushio et al., 2010a; Ushio et al., 2010b). As the composition of microbial community was not determined, possible influence of microbial community composition on enzyme activities could not be evaluated by the present study. 4.2. Relating phosphatase activities to key edaphic properties in different horizons Similar relationships of AcPME and PDE activities with selected edaphic properties suggest generally similar regulatory mechanisms for them. Different relationship patterns among horizons were probably related to the relative position in the field and the extent of decomposition of different horizons. For example, as revealed in some previous studies (De Boois, 1974; Keith et al., 2010), moisture content was lower in the L horizon compared to in the F/H horizon due to its direct

E. Hou et al. / Science of the Total Environment 515–516 (2015) 83–91

89

Table 7 Principal component loadings for phosphatase activities and some other edaphic properties in the L, F/H, and 0–15 cm mineral soil horizons.a Variable

AcPMEb PDEb MBCb Moisture WHCb Total C Total N Total P C:N ratio C:P ratio N:P ratio Litter biomass Fine root biomass Bulk density % of variance a b

L horizon PCA

F/H horizon PCA

Soil horizon PCA

PC1

PC2

PC3

PC1

PC2

PC3

PC4

PC1

PC2

0.90 0.91 0.77 0.77 0.76 −0.02 0.16 0.45 −0.02 −0.27 −0.41 0.08

0.01 0.03 −0.12 −0.36 −0.48 0.75 −0.20 −0.73 0.49 0.82 0.84 0.78

0.11 0.12 −0.08 0.10 0.21 −0.33 0.96 0.40 −0.86 −0.42 0.15 −0.26

0.10 0.27 −0.22 −0.04 0.03 0.68 0.00 −0.77 0.87 0.93 0.74 0.16

0.31 −0.25 −0.38 0.88 0.82 0.24 0.48 −0.03 −0.28 0.15 0.41 0.86

0.16 −0.09 0.79 0.09 0.32 0.66 0.82 0.54 −0.11 0.02 0.08 −0.31

0.82 0.81 −0.02 −0.02 0.27 0.07 0.15 −0.19 −0.06 0.29 0.40 −0.08

0.17 0.20 0.88 0.89 0.82 0.86 0.98 0.78 −0.12 0.24 0.40

0.86 0.84 0.04 0.16 0.25 0.37 0.09 −0.48 0.75 0.92 0.83

0.07 −0.82 42.5

0.68 −0.45 36.1

32.3

31.0

19.1

28.2

25.4

19.1

14.4

Loading scores N0.7 are shown in bold. Loading scores N0.35 correlated significantly with the principal component (P b 0.05, Pearson correlation method). AcPME: acid phosphomonoesterase activity; PDE: phosphodiesterase activity; MBC: microbial biomass C; and WHC: water holding capacity.

exposure to the environment. Therefore, moisture content could be a significant regulating factor of microbial biomass C and phosphatase activities in the L horizon but not in the F/H horizon. It is notable that whether and how moisture content may regulate the extracellular activity of enzymes in litter or soil in field may depend on the magnitude of and the time since the last rainfall, site topographic conditions, etc. (Henry, 2012; Keith et al., 2010). Carbon:P ratio was markedly lower in the mineral soil than the L and F/H horizons, therefore it could be a significant regulator of microbial biomass C and phosphatase activities in the mineral soil horizon but not in the L or F/H horizon. In the F/H horizon, AcPME and PDE activities were weakly correlated and both were weakly or insignificantly related to the selected edaphic properties. Much poorer correlations among extracellular enzyme activities in H layer litter than in L layer litter were also found in a Quercus petraea forest in the Czech Republic (Šnajdr et al., 2008). The reason is not clear but it is proposed that it is related to the heterogeneous character of the F/H

or H horizon, which includes a mixture of litters that vary considerably in the extent of decomposition. In the mineral soil, the activity of neither AcPME nor PDE was correlated with the extractable inorganic P (e.g. soluble Pi) or organic P (e.g. Al–Po) fraction, implying that the activations of both soil AcPME and PDE were neither substrate dependent nor end-product dependent (Kitayama, 2013). Instead, both soil AcPME and PDE activities were positively regulated by soil C:P, N:P and C:N ratios. These results suggest a stoichiometric control of AcPME and PDE activities in mineral soil, and demonstrate the benefits of using a stoichiometric approach to ascertain the key edaphic properties controlling the activities of phosphatase enzymes. Stoichiometric control of soil phosphatase activity supports the hypothetical model of resource allocation to extracellular enzyme production (Sinsabaugh and Moorhead, 1994; Sinsabaugh and Follstad Shah, 2012). The model suggests that microbial communities can maximize their productivity by optimizing their allocation of

Fig. 2. Acid phosphomonoesterase (AcPME) and phosphodiesterase (PDE) activities as a function of soil C:P and N:P ratios in the 0–15 cm mineral soil horizon: AcPME vs. C:P ratio (a); AcPME vs. N:P ratio (b); PDE vs. C:P ratio (c); and PDE vs. N:P ratio (d). Solid lines are exponential growth curves with the coefficient of determination (R2) and probability (P) shown.

90

E. Hou et al. / Science of the Total Environment 515–516 (2015) 83–91

resources (e.g. C, N, and P) in the production of extracellular enzymes acquiring a relatively scarce resource (Sinsabaugh and Moorhead, 1994). If P is a relatively scarce resource, plants and/or soil microorganisms may invest more relatively abundant C and N to produce and excrete enzymes involved in P cycling (Allison and Vitousek, 2005; Marklein and Houlton, 2012; Olander and Vitousek, 2000). Exponential, rather than linear, increases in soil AcPME and PDE activities with soil C:P and N:P ratios demonstrate the high response of soil AcPME and PDE activities to small increases in soil C:P and N:P ratios. Microbial biomass C – representing the biomass of a whole microbial community – in the mineral soil was simultaneously positively regulated by soil C, N, P and water availability, while the activities of phosphatase-producing microbes might be negatively influenced by soil P availability (Olander and Vitousek, 2000; Spiers and McGill, 1979). This might at least partly explain the lack of significant relationships between microbial biomass C and phosphatase activities in the mineral soil. Mycorrhizal fungi are usually proposed to contribute significantly to the extracellular activity of phosphatase in terrestrial ecosystems, particularly in P-limited environments (Koukol et al., 2006; Nannipieri et al., 2011). Mycorrhizal fungi were not investigated in this study; however, positive correlations between soil AcPME and PDE activities and fine root biomass imply that mycorrhizae may contribute significantly to the extracellular activities of AcPME and PDE in the studied soils, which needs to be explored in future studies. 5. Conclusions Our results have shown that the potential extracellular activities of both AcPME and PDE differed significantly among forests and horizons and could also be affected by the forest × horizon interaction. Main effects of forest on phosphatase activities were related to different moisture content among forests in the L horizon and different C:P, N:P and C:N ratios and fine root biomass among forests in the mineral soil horizon. Main effects of horizon were likely due to the potential shift in P chemical forms during the decomposition of litter or organic matter. Forest × horizon interaction on phosphatase activities was related to their interaction on microbial biomass. Dissimilar patterns of relationships between phosphatase activities and selected edaphic properties among horizons were related to the relative position in the field and the extent of decomposition of different horizons. Positive relationships between soil C:P and N:P ratios and AcPME and PDE activities in the mineral soil support the model of resource allocation to extracellular enzyme production and demonstrate the benefits of using a stoichiometric approach to ascertain the key environmental factors controlling the potential extracellular activities of phosphatase enzymes. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.scitotenv.2015.02.044. Acknowledgments We thank two anonymous reviewers for their helpful comments, Marijke Heenan for language improvement, and Yu Wang for her laboratory assistance. This study was jointly supported by the National Natural Science Foundation of China (Nos. 41401326 and 31070409), the Strategic Priority Research Program “Climate Change: Carbon Budget and Relevant Issues” of the Chinese Academy of Sciences (No. XDA05050205), and the Australian Research Council (FT0990547). References Adams, M.A., 1992. Phosphatase activity and phosphorus fractions in karri (Eucalyptus diversicolor F. Muell.) forest soils. Biol. Fertil. Soils 14, 200–204. Allison, S.D., Vitousek, P.M., 2005. Responses of extracellular enzymes to simple and complex nutrient inputs. Soil Biol. Biochem. 37, 937–944. Browman, M., Tabatabai, M., 1978. Phosphodiesterase activity of soils. Soil Sci. Soc. Am. J. 42, 284–290.

Bunemann, E.K., 2008. Enzyme additions as a tool to assess the potential bioavailability of organically bound nutrients. Soil Biol. Biochem. 40, 2116–2129. Chen, C.R., Condron, L., Davis, M.R., Sherlock, R.R., 2000. Effects of afforestation on phosphorus dynamics and biological properties in a New Zealand grassland soil. Plant Soil 220, 151–163. De Boois, H.M., 1974. Measurement of seasonal variations in the oxygen uptake of various litter layers of an oak forest. Plant Soil 40, 545–555. Falster, D., Warton, D., Wright, I., 2006. SMATR: standardised major axis tests and routines, version 2.0. URL, http://www.bio.mq.edu.au/ecology/SMATR (Last accessed on September 30, 2014). Gressel, N., McColl, J.G., Preston, C.M., Newman, R.H., Powers, R.F., 1996. Linkages between phosphorus transformations and carbon decomposition in a forest soil. Biogeochemistry 33, 97–123. Grierson, P., Adams, M., 2000. Plant species affect acid phosphatase, ergosterol and microbial P in a jarrah (Eucalyptus marginata Donn ex Sm.) forest in south-western Australia. Soil Biol. Biochem. 32, 1817–1827. Henry, H.A.L., 2012. Soil extracellular enzyme dynamics in a changing climate. Soil Biol. Biochem. 47, 53–59. Hou, E.Q., Chen, C.R., McGroddy, M.E., Wen, D.Z., 2012. Nutrient limitation on ecosystem productivity and processes of mature and old-growth subtropical forests in China. PLoS One 7, e52071. Hou, E.Q., Chen, C.R., Wen, D.Z., Liu, X., 2014. Relationships of phosphorus fractions to organic carbon content in surface soils in mature subtropical forests, Dinghushan, China. Soil Res. 52, 55–63. Huang, W.J., Zhang, D.Q., Li, Y.L., Lu, X.K., Zhang, W., Huang, J., et al., 2012. Responses of soil acid phosphomonoesterase activity to simulated nitrogen deposition in three forests of subtropical China. Pedosphere 22, 698–706. Huang, W.J., Liu, J.X., Wang, Y.P., Zhou, G.Y., Han, T.F., Li, Y., 2013. Increasing phosphorus limitation along three successional forests in southern China. Plant Soil 364, 181–191. Keith, D., Johnson, E., Valeo, C., 2010. Moisture cycles of the forest floor organic layer (F and H layers) during drying. Water Resour. Res. 46, W07529. Kitayama, K., 2013. The activities of soil and root acid phosphatase in the nine tropical rain forests that differ in phosphorus availability on Mount Kinabalu, Borneo. Plant Soil 367, 215–224. Koukol, O., Novák, F., Hrabal, R., Vosátka, M., 2006. Saprotrophic fungi transform organic phosphorus from spruce needle litter. Soil Biol. Biochem. 38, 3372–3379. Liu, L., Gundersen, P., Zhang, T., Mo, J.M., 2012. Effects of phosphorus addition on soil microbial biomass and community composition in three forest types in tropical China. Soil Biol. Biochem. 44, 31–38. Marklein, A.R., Houlton, B.Z., 2012. Nitrogen inputs accelerate phosphorus cycling rates across a wide variety of terrestrial ecosystems. New Phytol. 193, 696–704. McDowell, R.W., Condron, L.M., 2000. Chemical nature and potential mobility of phosphorus in fertilized grassland soils. Nutr. Cycl. Agroecosyst. 57, 225–233. Mo, J.M., Brown, S., Lenart, M., Kong, G.H., 1995. Nutrient dynamics of a human-impacted pine forest in a MAB reserve of subtropical China. Biotropica 27, 290–304. Murphy, J., Riley, J.P., 1962. A modified single solution method for the determination of phosphate in natural waters. Anal. Chim. Acta 27, 31–36. Nannipieri, P., Giagnoni, L., Landi, L., Renella, G., 2011. Role of phosphatase enzymes in soil. In: Bünemann, E., Oberson, A., Frossard, E. (Eds.), Phosphorus in Action. Springer Berlin Heidelberg, pp. 215–243. Naples, B.K., Fisk, M.C., 2010. Belowground insights into nutrient limitation in northern hardwood forests. Biogeochemistry 97, 109–121. Ohno, T., Zibilske, L.M., 1991. Determination of low concentrations of phosphorus in soil extracts using malachite green. Soil Sci. Soc. Am. J. 55, 892–895. Olander, L.P., Vitousek, P.M., 2000. Regulation of soil phosphatase and chitinase activity by N and P availability. Biogeochemistry 49, 175–191. Ormaza-González, F.I., Statham, P.J., 1996. A comparison of methods for the determination of dissolved and particulate phosphorus in natural waters. Water Res. 30, 2739–2747. Pang, P.C.K., Kolenko, H., 1986. Phosphomonoesterase activity in forest soils. Soil Biol. Biochem. 18, 35–39. Peñuelas, J., Sardans, J., Rivas, A., Janssens, I.A., 2012. The human induced imbalance between C, N and P in earth's life-system. Glob. Chang. Biol. 18, 3–6. Rao, M.A., Violante, A., Gianfreda, L., 2000. Interaction of acid phosphatase with clays, organic molecules and organo-mineral complexes: kinetics and stability. Soil Biol. Biochem. 32, 1007–1014. Shen, C.D., Yi, W.X., Sun, Y.M., et al., 2001. Distribution of 14C and 13C in forest soils of the Dinghushan Biosphere Reserve. Radiocarbon 43, 671–678. Sinsabaugh, R.L., Follstad Shah, J.J., 2012. Ecoenzymatic stoichiometry and ecological theory. Annu. Rev. Ecol. Evol. Syst. 43, 313–343. Sinsabaugh, R., Moorhead, D., 1994. Resource allocation to extracellular enzyme production: a model for nitrogen and phosphorus control of litter decomposition. Soil Biol. Biochem. 26, 1305–1311. Sinsabaugh, R.L., Lauber, C.L., Weintraub, M.N., Ahmed, B., Allison, S.D., Crenshaw, C., et al., 2008. Stoichiometry of soil enzyme activity at global scale. Ecol. Lett. 11, 1252–1264. Šnajdr, J., Valášková, V., Merhautová, Vr., Herinková, J., Cajthaml, T., Baldrian, P., 2008. Spatial variability of enzyme activities and microbial biomass in the upper layers of Quercus petraea forest soil. Soil Biol. Biochem. 40, 2068–2075. Spiers, G., McGill, W., 1979. Effects of phosphorus addition and energy supply on acid phosphatase production and activity in soils. Soil Biol. Biochem. 11, 3–8. Tabatabai, M.A., Bremner, J.M., 1969. Use of p-nitrophenyl phosphate for assay of soil phosphatase activity. Soil Biol. Biochem. 1, 301–307. Tarafdar, J., Claassen, N., 1988. Organic phosphorus compounds as a phosphorus source for higher plants through the activity of phosphatases produced by plant roots and microorganisms. Biol. Fertel. Soils 5, 308–312.

E. Hou et al. / Science of the Total Environment 515–516 (2015) 83–91 Turner, B.L., 2010. Variation in pH optima of hydrolytic enzyme activities in tropical rain forest soils. Appl. Environ. Microbiol. 76, 6485–6493. Turner, B.L., Haygarth, P.M., 2005. Phosphatase activity in temperate pasture soils: potential regulation of labile organic phosphorus turnover by phosphodiesterase activity. Sci. Total Environ. 344, 27–36. Turner, B.L., Wells, A., Condron, L.M., 2014. Soil organic phosphorus transformations along a coastal dune chronosequence under New Zealand temperate rain forest. Biogeochemistry 121, 595–611. Ushio, M., Kitayama, K., Balser, T.C., 2010a. Tree species-mediated spatial patchiness of the composition of microbial community and physicochemical properties in the topsoils of a tropical montane forest. Soil Biol. Biochem. 42, 1588–1595. Ushio, M., Kitayama, K., Balser, T.C., 2010b. Tree species effects on soil enzyme activities through effects on soil physicochemical and microbial properties in a tropical montane forest on Mt. Kinabalu, Borneo. Pedobiologia 53, 227–233.

91

Vance, E.D., Brookes, P.C., Jenkinson, D.S., 1987. An extraction method for measuring soil microbial biomass C. Soil Biol. Biochem. 19, 703–707. Xu, Z., Yu, G., Zhang, X., Ge, J., He, N., Wang, Q., et al., 2015. The variations in soil microbial communities, enzyme activities and their relationships with soil organic matter decomposition along the northern slope of Changbai Mountain. Appl. Soil Ecol. 86, 19–29. Zhang, A., Chen, Z., Zhang, G., Chen, L., Wu, Z., 2012. Soil phosphorus composition determined by 31P NMR spectroscopy and relative phosphatase activities influenced by land use. Eur. J. Soil Biol. 52, 73–77. Zornoza, R., Mataix-Solera, J., Guerrero, C., Arcenegui, V., Mayoral, A., Morales, J., et al., 2007. Soil properties under natural forest in the Alicante Province of Spain. Geoderma 142, 334–341.