Influence of soil factors on the soil enzyme inhibition by Cd

Abstract

A comparative study was conducted on the toxicity of Cd to alkaline phosphatase activity (ALP) and dehydrogenase activity (DHA) in 18 top soils with contrasting soil properties representative of 14 major soil types in China. Soil pH and carbonate content, soil organic matter, and cation exchange capacity (CEC) largely affected the Cd toxicity on two enzyme activities; with the soil pH having only minor effect on the median ecological dose values based on total Cd concentrations (ED_{50 T}). The values of ED_{50 T}/ED_{50 W} (based on water-soluble Cd content) of alkaline phosphatase and dehydrogenase were strongly influenced by pH and CEC contents, which explained up to 71% of the variation for alkaline phosphatase, 82% of the variation for dehydrogenase, and also were significantly correlated with the parameter K_F derived from Freundlich adsorption isotherms. This study suggests that the values of ED_{50 T}/ED_{50 W} could be useful to evaluate the buffer capacity of soils which protects soil enzymes from harmful effects of heavy metal.

Keywords:

• cadmium
• alkaline phosphatase
• dehydrogenase
• ecotoxicity
• soil properties

Introduction

Anthropogenic activities have increased the concentration of heavy metals in the environment, producing toxic effects on agro-ecosystems and natural ecosystems, which is one of the major environmental problems in many parts of the world. Also, there has been a growing public concern about the potential accumulation of trace elements in agricultural soils in China owing to the rapid industrialization and urbanization and increasing reliance on chemicals in the last two decades (Luo et al. 2009). It has been estimated that contaminated soils result in a loss of more than 1.0 × 10⁷ tons of food supplies in China annually (Wei & Chen 2001). Increasing inputs of Cd to agricultural soil in China as well as many other World's areas is of particular concern, and it is expected that the Cd concentration in soils will exceed the current environmental standard (0.3 mg kg⁻¹ for soil with pH 6.5–7.5) [State Environmental Protection Administration of China (SEPAC) 2006] in the next 50 years at the current Cd input rate of 0.004 mg kg⁻¹ yr⁻¹ (Luo et al. 2009). Cadmium has no known biological function and poses serious health risk due to plant absorption and spreading in the food chain through animal grazing and a direct threat to humans via uptake into rice (Ros et al. 2009). Cd is one of the most studied heavy metals in terms of its toxicity to biota and their beneficial processes in soil (Vig et al. 2003; Giller et al. 2009; Karaca et al. 2010).

Determination of the effects of Cd on soil biological activity often yields conflicting results. Bååth (1989) summarized data from both laboratory and field studies in terms of the highest metal concentration in the soil where no effect (HNOEC) was found and the lowest metal concentration where an effect (LOEC) was observed and concluded that for Cd the HNOEC and LOEC values differed from 100 to 1000 folds among the various soils. Giller et al. (1998) reported the highest amount of metal added that did not produce an effect (HNOED) and the lowest amount added that resulted in a decrease (LOAED) in the rate of base respiration for Cd in laboratory studies also varied between 100 and 1000 folds.

The ED₅₀ values for Cd reported in the vast literature on the topic lie in a wide concentration range. For example, ED₅₀ values in the range 90–5555 mg kg⁻¹ were reported for dehydrogenase activity (DHA) (Welp 1999; Moreno et al. 2001; Gao, Zhou, Mao, Zhi, Shi 2010; Gao, Zhou, Mao, Zhi, Zhang, et al. 2010), from 30 to 7240 mg kg⁻¹ for urease activity (Doelman & Haanstra 1986; Welp 1999; Moreno et al. 2001, 2003), from 14 to 9870 mg kg⁻¹ for acid phosphatase activity (Doelman & Haanstra 1989; Moreno et al. 2003; Renella et al. 2003; Gao, Zhou, Mao, Zhi, Shi 2010; Gao, Zhou, Mao, Zhi, Zhang, et al. 2010), from 4 to 5485 mg kg⁻¹ for alkaline phosphatase activity (ALP) (Doelman & Haanstra 1989; Renella et al. 2003), from 357 to 4975 mg kg⁻¹ for β-glucosidase activity (Moreno et al. 2003), from 43 to 8677 mg kg⁻¹ for N-a-benzoyl-L-argininamide protease activity (Moreno et al. 2003), from 121 to 3192 mg kg⁻¹ for arylsulfatase activity (Haanstra & Doelman 1991), from 77 to 33,333 mg kg⁻¹ for adenosine triphosphate (ATP; Moreno et al. 2001), and from 85 to 1270 mg kg⁻¹ for Fe (III) reduction (Welp & Brümmer 1997).

It is important to underline that different ED₅₀ values are obtained by using different dose response models, as discussed by Speir et al. (1995, 1999). While the inhibition of soil enzyme activity should be related to the soluble meal fraction (e.g. water soluble, diethylenetriaminepentaacetic acid-extractable), valid fittings of soil enzyme activity in dose response models have only been obtained by plotting the soil enzyme activity in function of total heavy metal concentrations independently on the used model (Moreno et al. 2001; Renella et al. 2003). This may be explained by the fact that heavy metal speciation in soil depends on several soil properties such as soil organic matter (SOM) content, cation exchange capacity (CEC), soil pH value, metal ionic properties, and time. For example, it has been reported that Cu and Cr toxicity to soil microorganisms increased with increasing of SOM and CEC and decreased with increasing of soil pH values (D'Ascoli et al. 2006), and similar results were reported by Welp and Brümmer (1997) on the toxicity thresholds for Ni and Hg in various German soils. Cd solubility in soil is mainly controlled by soil pH and sorption onto clay particles (McBride 1989) and by carbonate content and quality (Renella et al. 2004). However, an in depth analysis of the responses of soil enzyme activity to Cd solubility in dose response experiments has not been published so far.

The objective of the present study was to describe the response of soil enzyme activity to Cd solubility in various Chinese soils using dose response experiments. The importance of main soil properties in modeling of enzyme responses to Cd is also discussed.

Materials and methods

Soil samples and handling

A total of 18 surface (0–15 cm depth) soils representative of the major soil types in China were collected from agricultural soils of 18 provinces in China (Table 1). These soils cover a wide range in soil properties expected to affect the bioavailability and toxicity of Cd in soils. The soils were air dried and sieved (mesh size, 1 mm) prior to use.

Table 1. Selected soil properties for the 18 soil samples tested.

Samples	Soil type	pH	OM (g kg^−1)	CaCO3 (g kg^−1)	CEC (cmol kg)	Clay %	Total Cd (mg kg^−1)
Hunan	Red soil (Argi-Udic Ferrosols)	4.90	15.52	0.00	10.85	42.91	0.178
Liaoning	Brown soil (Hapli-Udic Argosols)	5.74	25.84	0.00	12.19	17.32	0.159
Chongqing	Purple soil (Dystric Purpli-Udic Cambosols)	5.74	17.48	0.00	21.34	24.96	0.183
Yunnan	Latosolic red soil (Hapli-Udic Ferralosols)	5.92	34.26	0.00	11.10	27.52	0.259
Jiangxi	Red soil (Argi-Udic Ferrosols)	6.01	11.69	0.00	8.70	36.51	0.144
Anhui	Yellow brown soil (Ferri-Udic Argosols)	6.25	20.04	0.00	19.08	16.84	0.107
Heilongjiang	Black soil (Hapli-Udic Isohumosols)	6.27	35.69	0.00	28.59	19.33	0.183
Jilin	Black soil (Hapli-Udic Isohumosols)	6.82	32.85	0.00	31.11	30.18	0.125
Jiangsu	Paddy soil (Gleyi-Stagnic Anthrosols)	6.93	47.69	0.00	26.20	45.94	0.155
Shaanxi	Lou soil (Earth-cumuli-Orthic Anthrosols)	7.90	16.49	35.60	22.37	26.01	0.239
Hebei	Fluvo-aquic soil (Ochri-Aquic Cambosols)	7.98	8.57	17.62	8.12	10.50	0.189
Henan	Fluvo-aquic soil (Ochri-Aquic Cambosols)	8.07	17.79	27.50	16.01	18.18	0.190
Xinjiang	Gray desert soil (Calci-Orthic Aridosols)	8.12	19.43	15.06	25.25	9.57	0.165
Shanxi	Cinnamon soil (Hapli-Ustic Argosols)	8.24	23.17	25.15	16.80	17.74	0.196
Tianjin	Coastal saline soil (Marinic Aqui-Orthic Halosols)	8.29	22.02	53.57	24.67	7.59	0.186
Gansu	Anthropogenic-alluvial soil (Typic Siltigi-Orthic Anthrosols)	8.37	19.27	38.51	11.23	6.66	0.200
Shandong	Fluvo-aquic soil (Ochri-Aquic Cambosols)	8.65	11.84	31.69	13.09	17.11	0.184
Inner Mongolia	Chestnut colored soil (Typic Calci-Ustic Isohumosols)	8.80	16.30	11.51	11.61	10.51	0.211

OM, organic matter; CEC, cation exchange capacity.

Soil organic matter (OM) was determined by the K₂Cr₂O₇–H₂SO₄ oxidation method. Soil calcium carbonate (CaCO₃) was determined from pressure increases after addition of HCl to the soil in closed containers. Soil texture was measured by pipette method. CEC was analyzed by 1 mol l⁻¹ ammonium acetate exchange method. Soil pH was measured in 1:5 soil:water (Lu 1999). Total Cd was analyzed by Inductively Coupled Plasma Mass Spectrometry (ICP-MS; X-7, Thermo-Elemental, USA) after digestion with a mixture of hydrofluoric (HF) and perchloric (HClO₄) acid by microwave digestion system with 15 min.

Experimental design

Air-dried soils were spiked with an aqueous stock solution of CdSO₄ to result in final Cd concentrations of 0, 0.6, 5, 25, 50, 100, 200, 300, and 500 mg kg⁻¹ soil (soil:CdSO₄ solution ratio, 1:1). The Cd solution was added dropwise to moisten the whole soil sample. After 30 min of equilibration, the enzyme activities in the soils were measured according to Acosta-Martínez and Tabatabai (2001).

Soil analyses

ALP activity was determined using 5 g air-dried soil mixed with 5 ml solution of CdSO₄, after 30 min mixed with five drops of toluene, 10 ml of disodium phenyl phosphate solution, and 10 ml of Borax sodium hydroxide buffer (pH 9.4). The suspensions were stationarily incubated for 24 h at 37 °C. The samples were then filtered and the filtrate was colored with 0.25 ammonia–ammonium chloride buffer at pH 9.6, 0.5 ml of 2% 4-aminoantipyrine, and 8% potassium ferrocyanide, the phenol released was determined colorimetrically in a spectrophotometer at 510 nm. Results were expressed as µg phenol·g⁻¹ soil·h⁻¹(Guan et al. 1987).

Soil DHA was assayed according to Burns (1978). Briefly, 3 g soil sample was mixed with 3 ml solution of CdSO₄, after 30 min, added to 0.03 g calcium carbonate, was mixed with 0.5 ml of 3% (w/v) solution of triphenyl tetrazolium chloride as the substrate and 1.25 ml of distilled water. The soil slurries were mixed thoroughly and incubated at 37 °C for 24 h. The triphenyl formazan (TPF) produced by the DHA was extracted with methanol and the activity were expressed as µg TPF TPF·g⁻¹ soil·h⁻¹. All results are the means of three replicates.

Analysis of Cd solubility: the water-soluble Cd was determined on 5 g air-dried soil mixed with 5 ml of Cd aqueous solution into 50-ml centrifuge tubes. The soil suspensions were mixed for 30 min, and then added with 20 ml of distilled water and shaken end-over-end for 2 h. After centrifugation (15 min, 4000 × g) the supernatant was filtered through membrane filter paper and stabilized with HNO₃. The supernatant was decanted into clean test tubes and kept at 4 °C until Cd concentration measurement. The Cd concentrations were measured by atomic absorption spectrophotometer and Inductively Coupled Plasma Mass Spectrometry (ICP-MS; X-7, Thermo-Elemental, USA).

Data modeling

A Michaelis Menten kinetic approach was used to model inhibition of the biological properties by Cd. The mathematical models are as follows:

(1)

(2)

and the reasons for their choice are outlined in Speir et al. (1995). The constants a, b, and c are always positive, with b > a. In each of the biological and biochemical analyses described earlier, a single substrate concentration was used or assumed for the ALP and DHA. Where data fit model 1, it describes full inhibition, that is, fully competitive, fully non-competitive, etc. For model 2, the equation describes partial inhibition, that is, partially competitive, partially non-competitive, etc. The ecological dose (ED) values can be derived graphically or by statistical means.

Meanwhile, the logarithm of the added concentrations as independent variables was described using a logistic response model (Doelman & Haanstra 1986). This relationship is expressed using the following equation:

(3)

The parameter Y is the observed enzyme activity at the logarithm base 10 of the heavy metal concentration x and c is the calculated initial (uninhibited) level of enzyme activity, determined by the enzyme activity of the control sample. The parameter b is a slope parameter indicating the inhibition rate. The parameter a is the logarithm of the concentration at which the enzyme activity is half the uninhibited level. Logarithm of the heavy metal concentration of the blanks could not be calculated, so the all data of heavy metal concentration were added to 1 (e.g. blanks was substituted by 1 mg kg⁻¹). Consequently, a value of one (1) was subtracted from the ED value calculated from model 3.

Mean results of the triplicate measurements at each inhibitor concentration were used as the raw data for the models. Statistical analyses were performed using SPSS 18.0 and Origin 8.5 software.

Results

Soil physico-chemical properties and Cd solubility

Soil pH ranged from 4.90 at Hunan to 8.80 at Inner Mongolia, with six of the soils having a pH below 6.5 (Table 1). The OM content of the soils varied a great deal (8.57–47.7 g kg⁻¹) and the average of OM content was 22.0 g kg⁻¹. The maximum OM was 4.8 % at Jiangsu. Nine out of 18 soils (pH > 7) had free CaCO₃ (the mean of 28.56 g kg⁻¹). Clay content ranged from 6.7% to 46% and CEC ranged from 8.12 to 31.1 cmol kg⁻¹. Background metal concentrations for Cd ranged from 2.22 to 167 µg kg⁻¹.

The water-soluble Cd concentrations in soils increased with an increase in total Cd of contaminated soil in all samples (Figure 1). Except Heilongjiang, water-soluble Cd was higher in acid soils (pH<6.5) than other soils. In these soils water extracted a large percentage of the added Cd, which ranged from 3.36% (16.8 mg kg⁻¹) for Chongqing soil to 16.5% (82.6 mg kg⁻¹) for Hunan soil contaminated with Cd at 500 mg kg⁻¹. The other neutral and alkaline soils and Heilongjiang had low water-soluble Cd content and the values were less than 0.6%, except Hebei and Gansu soil samples. Water-soluble Cd in 500 mg kg⁻¹ treatments were negative related to soil pH and CEC content, and the correlation coefficients were −0.73 (n = 18, p < 0.01) and −0.51 (n = 18, p < 0.05).

Figure 1. Concentrations of water-soluble cadmium in Cd-contaminated soils (mg kg⁻¹ soil).

Effect of Cd on soil ALP and DHA

Toxicity data were expressed using two different measures of metal content in soil (i.e. total and water soluble) as a direct comparison between the different fractions and to improve Cd toxicity predictions across sites. The ED₅₀ values determined in function of total and water-soluble Cd fractions are shown in Table 2. Models 2 and 3 described well ALP and DHA activities, respectively, with a coefficient of determination (R²) higher than 0.8, suggesting a high degree of fitting with these models.

Table 2. ED₅₀ values of alkaline phosphatase activity (ALP) and dehydrogenase activity (DHA) in different soils.

	Water-soluble Cd				Total Cd
	ALP Equation (2)		DHA Equation (3)		ALP Equation (2)		DHA Equation (3)
Samples	ED50 W^a	R^2	ED50 W	R^2	ED50 T	R^2	ED50 T	R^2
Hunan	0.37	0.789	6.97	0.957	5.9	0.823	104.4	0.974
Liaoning	0.20	0.986	1.69	0.977	14.8	0.971	105.9	0.973
Chongqing	0.47	0.923	1.15	0.969	34.7	0.934	66.9	0.969
Yunnan	1.28	0.873	13.15	0.956	52.6	0.871	211.3	0.963
Jiangxi	0.75	0.924	1.67	0.995	37.0	0.888	77.7	0.995
Anhui	1.34	0.754	2.77	0.995	120.9	0.809	173.6	0.963
Heilongjiang	1.93	0.956	0.82	0.907	698.8	0.980	177.2	0.956
Jilin	1.72^b	0.936	1.67	0.953	682.7	0.984	467.8	0.860
Jiangsu	0.82^b	0.97	0.51	0.973	642.5	0.989	171.6	0.995
Shaanxi	0.40	0.939	0.43	0.926	199.7	0.995	213.8	0.955
Hebei	0.17	0.991	0.20	0.965	61.3	0.996	59.3	0.965
Henan	0.22	0.952	0.39	0.904	58.8	0.991	144.9	0.970
Xinjiang	0.34^b	0.976	0.37	0.926	147.9	0.993	128.4	0.905
Shanxi	0.25	0.959	0.40	0.943	55.5	0.997	146.2	0.988
Tianjin	0.20	0.921	0.49	0.934	75.4	0.991	336.3	0.978
Gansu	0.26	0.984	1.11	0.978	78.8	0.989	164.2	0.974
Shandong	0.19	0.986	0.25	0.969	28.9	0.998	44.7	0.977
Inner Mongolia	1.21	0.928	1.67	0.949	78.3	0.982	201.8	0.956

^a ED_{50 T} or ED_{50 W} values are expressed in total or water-soluble Cd (mg/kg soil).

^b ED₅₀ values were calculated using Equation (1).

A large variance of ED₅₀ values were found among soils irrespective of the metal fraction used. ED₅₀ values for the effects of Cd (water soluble) on ALP varied from 0.17 to 1.93 mg kg⁻¹ and for DHA ranged from 0.2 to 13.5 mg kg⁻¹. Using total concentrations of Cd in soil, ED₅₀ values varied from 6 to 699 mg kg⁻¹ for ALP and from 45 to 468 mg kg⁻¹ for DHA. Soil solution-based ED₅₀ values for ALP varied approximately 10-fold less than total-based ED₅₀ values across a variety of soils. But the variation of soil solution-based toxicity values for DHA was 6-fold larger than the corresponding variation of values based on total metal concentrations.

The water extractable Cd-based ED₅₀ values showed an increase with increasing OM content for ALP (r = 0.538, p < 0.05), whereas for DHA the ED₅₀ values decreased with an increase in soil solution pH (r = −0.533, p < 0.05). When expressed on a total metal concentration, there was a strong relationship between ED₅₀ value of ALP and OM and CEC content (ED₅₀/OM, r = 0.748, P < 0.001; ED₅₀/CEC, r = 0.763, p < 0.001); the ED₅₀ value of DHA was also strongly correlated with CEC content (r = 0.600, p < 0.001). No remarkable correlation was found between water extractable Cd-based and total Cd-based ED₅₀ values of DHA, while there was a strong relationship for ALP (r = 0.675, p < 0.001).

The ratios of total to water solution-based ED₅₀ of two enzymes were remarkably correlated with soil pH (ALP: ED_{50 T}/ED_{50 W}, r = 0.583, p < 0.05; DHA: ED_{50 T}/ED_{50 W}, r = 0.742, p < 0.001) and CEC (ALP: ED_{50 T}/ED_{50 W}, r = 0.617, p < 0.001; DHA: ED_{50 T}/ED_{50 W}, r = 0.531, p < 0.05).

Correlation between soil properties and ED₅₀ values

To a large extent, Cd toxicity thresholds can be explained by soil properties (all data except pH were log-transformed). Results of multiple linear regressions (stepwise forward) for ED₅₀ and soil properties are presented in Table 3.

Table 3. Multiple linear regressions between Cd toxicity threshold values and soil properties.

Regression equation		R^2	Rig
ALP	logED50 W = 1.016 logOM − 1.642	0.272	0.027
	logED50 T = 2.116 logCEC − 0.663	0.503	0.001
	log(ED50 T/ED50 W) = 0.039 logCEC + 1.534	0.381	0.006
	log(ED50 T/ED50 W) = 0.039 logCEC + 0.215 pH + 0.002	0.708	<0.001
DHA	logED50 W = −0.28 pH + 1.994	0.497	0.001
	logED50 T = 0.879 logOM + 1.004	0.411	0.004
	Log(ED50 T/ED50 W) = 0.297 pH + 0.031	0.552	<0.001
	Log(ED50 T/ED50 W) = 0.294 pH + 0.035 logCEC − 0.556	0.82	<0.001

ALP, alkaline phosphatase activity; DHA, dehydrogenase activity.

For ALP, soil OM and CEC were the most influential factors affecting Cd toxicity across soils. OM was the best predictor for water-soluble Cd toxicity thresholds, which explains up to 25% of the variation. The CEC content of the soil was the best predictor for the toxicity thresholds based on total Cd, which explains 50% of the variation in ED₅₀ values. The ED₅₀ values of DHA could be modeled by regression models using pH and OM. In contrast to ALP, soil pH explained to a large extent (50%) the variation in DHA ED₅₀ values based on water-soluble Cd. The OM generally explained most of the variation (up to 41%) in toxicity thresholds of added Cd among soils.

Ratios of toxicity thresholds based on total to water-soluble Cd were explained by soil properties (Table 3). The pH and the CEC content of the soil were the best predictors for the ratio of toxicity thresholds for ALP and DHA assays (up to 71% of the variation for ALP, and 82% of the variation for DHA).

In order to find the reason for why the ratios of toxicity thresholds had significant correlation with soil properties than toxicity thresholds based on total or water-soluble Cd, we did further work on sorption mechanisms. Sorption isotherms helps to better understand the relationship between the sorbed and soluble fraction of Cd. Wang (2012) reported that the partitioning of Cd between the solid and the solution phase can be fitted by the Freundlich equation in these soils (R² > 0.92). The relationship between the ratios of total to water solution-based ED₅₀ values of ALP and DHA and the Freundlich sorption constants K_F (in logs) of the metals tested is shown in Figures 2 and 3. Values of ED_{50 T}/ED_{50 W} (in logs) ranged from 1 to 3 and increasing logK_F should increase the ratios of toxicity thresholds. The variations of the ratios of toxicity thresholds using logK_F are described by the following equations:

(4)

(5)

Figure 2. Relationship between the ratios of total to water solution-based ED₅₀ values of alkaline phosphatase activity (ALP) and the Freundlich sorption constants (K_F) of the metals tested.

Figure 3. Relationship between the ratios of total to water solution-based ED₅₀ values of dehydrogenase activity (DHA) and the Freundlich sorption constants (K_F) of the metals tested.

Discussion

The water-soluble Cd was greater in acidic soils than in neutral and alkaline soils (Table 1). In the Heilongjiang soil which had a pH value of 6.5 and higher values of SOM and clay contents and CEC, the water-soluble Cd was 32 times lower than in other soils (Table 1). The water-soluble Cd fraction in soil can be considered to be the forms capable of inhibiting the soil enzyme activity (Vig et al. 2003). The water-soluble Cd fraction accounted for 1.7% and 1.8% of total Cd for Hebei and Gansu soils, respectively, which could be due to their low CEC and clay contents. Moreno et al. (2001) reported that the water-soluble Cd content was higher in sandy acidic soils than in clay alkaline soils. The low Cd water solubility in the Hebei and Gansu soils could be also due to their carbonate contents. Rapid Cd insolublization has been reported (Renella et al. 2004) and changes in the Cd water-soluble fractions depending on the soil management have been reported (Vig et al. 2003).

The water-extracted fractions of Cd are generally very low in field soils. Thus, some smelter soils had very low water extractable Cd with a mean value of only 0.13% of the soil total Cd (Ivask et al. 2010) and only about 0.1% of Cd was extracted from some other polluted soils (Jopony & Young 1994). François et al. (2004) reported on average 0.5% water extractability of Cd while Naidu et al. (1994) showed that Cd was barely extracted from soils with water in soils at pH above 6.5. By comparison, it is shown that the concentration of soil solution Cd in freshly contaminated alkaline soils were lower than acid soils, even being at the same level of field studies. Precipitation of Cd as carbonates and hydroxides could occur in alkaline soil due to the presence of carbonates and alkalinity.

The effect of heavy metals on different soil enzymatic activities has been well studied. The mode of action of heavy metals varies from one enzyme to another, and the effects on enzymes also depend on the heavy metal itself (Moreno et al. 2003). The ED_{50 T} values obtained for ALP and DHA in this study were comparable to those previously reported (Welp & Brümmer 1997; Welp 1999; Moreno et al. 2001). Renella et al. (2003) observed that in the acidic soil the ALP was more sensitive to Cd than the acid phosphatase activity whereas in the alkaline soils the opposite behavior was observed, and suggested that a stronger inhibition of phosphatase activity by Cd metals at alkaline than acidic pH values could be due to higher reactivity of deprotonated amino acids. The stronger enzyme activity inhibition by Cd under sub-acidic and alkaline conditions could be also due to the formation of hydroxo–metal complexes, with the monovalent Cd(OH)⁺ being the predominant Cd species interacting with enzymes and soil microorganisms.

Both the ED_{50 T} and ED_{50 W} values increased with increasing of SOM and the ED_{50 T} values increased with CEC. It is well known that the reaction of metal ions with organic ligands influences their solubility and bonding capacity in soil (Welp & Brümmer 1997). Differences in ED₅₀ values in soils with different sorptive capacity have been reported (Speir et al. 1995). Oorts et al. (2006) reported that total Cu toxicity thresholds on nitrification potential, glucose-induced respiration, and C-mineralization of plant residues increased with increasing SOM content and CEC, and Ni toxicity thresholds increased with increasing CEC and clay content for the same soil microbial endpoints. For substrate-induced nitrification in 12 Australian field trials, soil pH was the most influential factor affecting Cu and Zn toxicity across soils, explaining 73% and 55% of the variation in EC₅₀ values, respectively (Broos et al. 2007). Welp and Brümmer (1997) reported that the EC₁₀ values for Cd increased with increasing concentration of dissolved organic C. However, the above mentioned microbial activities are mainly intracellular and therefore more similar to the soil DHA than to the ALP. In dose response soil toxicity experiments, the soil properties generally explain less than 50% of the variation of the toxicity thresholds based on the measurement of total or water-soluble metal concentrations, this due to the fact that too large difference of soil properties in present study.

In spite of the fact that ED₅₀ of ALP and DHA had different variances (wide concentration range), the values of ED_{50 T}/ED_{50 W} of the two enzymes exhibited same properties. The stronger a metal is sorbed by the soil, the bigger is the difference between the total and water solution-based ED₅₀. The sorption and solubility data fitted with the Freundlich sorption constants K_F (log scale) and were remarkably correlated with the ED_{50 T}/ED_{50 W} ratio values of both ALP and DHA, confirming that soil sorptive capacity is a suitable predictor Cd toxicity thresholds for enzymatic and microbial activity in soil (Welp 1999). In this sense we propose that the ED_{50 T}/ED_{50 W} ratio values can be used to evaluate the mitigation capacity of soils toward metal inhibition of soil enzymes activity. Relationships such as those developed in this study will permit China environmental regulation to move from single-value national soil quality guidelines to soil-specific quality guidelines and permit soil-specific risk assessments to be undertaken.

Funding

This work was supported by the National Hi-tech Research and Development Program of China [grant number 2012AA101402] and the Fundamental Research Funds for Scientific Research Innovation Key Projects for the Northwest A&F University [grant number ZD2013012] and the Special Fund for Agro-scientific Research in the Public Interest [grant number 200903015].

Additional information

Funding

Funding: This work was supported by the National Hi-tech Research and Development Program of China [grant number 2012AA101402] and the Fundamental Research Funds for Scientific Research Innovation Key Projects for the Northwest A&F University [grant number ZD2013012] and the Special Fund for Agro-scientific Research in the Public Interest [grant number 200903015].