Is the computed speciation of copper in a wide range of Chinese soils reliable?

Figures & data

Table 1. Soil solid-phase properties.

No.	Site	Location (longitude, latitude)	pH (1:5)	EC (μS cm^−1)	CEC (cmol^+ kg^−1)	SOC (%)	Clay (%, <2 μm)	Total Cu (mg kg^-1)
S1	Haikou	19°55′N,111°29′E	4.93	111	4.77	1.51	66	50.5
S2	Qiyang	26°45′N,111°52′E	5.31	74	5.52	0.87	46	22.2
S3	Hailun	47°28′N, 126°57′E	6.54	153	29.5	3.03	40	18.3
S4	Jiaxing	30º77′N,120°76′E	6.70	159	21.4	1.42	39	31.3
S5	Hangzhou	30°26′N,120º25′E	6.80	203	14.9	2.46	41	47.2
S6	Chongqing	30°26′N,106°26′E	7.12	71	26.7	0.99	27	10.7
S7	Guangzhou	23°10′N, 113°18′E	7.20	137	10.0	1.47	25	10.8
S8	Lingshan	39°55′N, 116°8′E	7.48	93	25.4	4.28	20	14.2
S9	Hulunber	46°03′N, 122°03′E	7.66	888	24.1	2.66	37	14.7
S10	Gongzhuling	42º40′N, 124º88′E	7.82	147	29.1	2.17	45	21.4
S11	Shijiazhuang	38°03′N, 114°26′E	8.19	302	16.8	1.00	10	16.5
S12	Urumchi	43º95′N, 87º46′E	8.72	227	14.8	0.87	25	26.5
S13	Yangling	34°19′N,108°0′E	8.83	83	14.4	0.62	16	21.5
S14	Langfang	39°31′N, 116°44′E	8.84	5.7	10.5	0.6	21	9.51
S15	Zhengzhou	34º47′N,112º40′E	8.86	109	13.3	1.57	28	13.4
S16	Zhangye	38°56′N,100°27′E	8.86	152	13.1	1.02	20	27.6
S17	Dezhou	37°20′N, 116°29′E	8.90	112	14.3	0.69	18	15.0

Table 2. Soil pore water properties.

	Total Cu (mg kg^−1)	pH	EC (μS cm^−1)	DOC (mg L^−1)	Soluble cation (mmol L^−1)				Total Cu (μmol L^−1)	Log (free Cu^2+ activity) (mol L^-1)	%free Cu^2+ activity	Log (Kd) (L kg^−1)
					Ca^2+	Mg^2+	K^+	Na^+
Minimum	30.8	4.39	1148	102	2.10	0.63	0.07	0.48	1.10	−5.30	0.001	2.10
Quartile 1	151	6.89	2683	180	11.04	1.73	0.21	1.08	5.83	−7.80	0.006	2.50
Median	301	7.42	3367	246	14.46	4.83	0.38	2.13	10.2	−8.48	0.043	2.64
Mean	450	7.11	4816	299	17.32	6.77	0.68	11.4	14.1	−8.54	2.73	2.69
Quartile 3	674	7.80	5091	350	21.47	6.79	1.10	6.14	19.7	−9.69	0.435	2.89
Maximum	1485	8.07	17260	765	42.68	37.50	2.58	120	48.9	−11.5	39.8	3.31

Figure 1. Relationships between (a) soluble Cu in pore water and total Cu in soil. (b) soluble Cu and DOC in pore water.

Note: The solid line represents regression line.

Table 3. The regression equations between Cu speciation and soil or soil pore-water properties.

No.	Regression equations	r^2	N
1	log (soluble Cu) = 0.515 log (total Cu in soil) + 1.223 Log (DOC) – 2.976	0.63	34
2	Log (Kd) = 0.475 Log (total Cu in soil) – 0.666 Log (DOC) + 3.124	0.28	34
3	Log (Kd) = −0.452 Log (soluble Cu) + 0.808 Log (CEC) + 2.179	0.42	34
4	Log (Kd) = 0.617 log (total Cu in soil) –1.391 log (DOC) + 1.545 log (CEC) + 2.549 for soils with pH >7.20	0.88	20
5	Log (free Cu^2+ activity) = −1.171 pH + 1.639 log (total Cu in soil) – 4.303	0.65	34
6	Log (free Cu^2+ activity) = −1.310 pH + 1.60 log (soluble Cu) – 0.783	0.65	34
7	Log (%free Cu^2+ activity) = −1.216 pH + 7.134	0.64	34

Figure 2. Comparison of WHAM VI calculated free Cu²⁺ ion and measured free Cu²⁺ ion in 34 samples of soil pore water. Free Cu²⁺ activity was predicted with WHAM VI using either (a) 65% AFA and default log K_CuFA, (b) adjusted AFA rate from 10% to 125% and default log K_CuFA, (c) 65% AFA and adjusted log K_CuFA from 1.0 to 2.3.

Note: The thin solid line represents 1:1 line and dot line represents 1 order of magnitude either side of the 1:1 line. The thick solid line represents regression line.

Table 4. Summary of the parameterization of WHAM 6 and NICA-Donnan parameters based on goodness-of-fit of measured free Cu²⁺ activity data.

WHAM 6 parameters		Model fit indices
log KCuFA	ΔLK2	%AFA	RMSE	Slope	r^2
Default values
2.1	2.34	65	1.23	0.44	0.54
2.1	2.34	30	1.06	0.43	0.53
2.1	2.34	20	1.18	0.43	0.55
2.1	2.34	10–30	0.84	0.63	0.81
2.1	2.34	10–65	0.42	0.83	0.94
2.1	2.34	10–125	0.23	0.92	0.98
Adjusting binding constant
1.8	2.34	65	0.84	0.59	0.72
1.0	2.34	65	0.77	0.71	0.74
1.0–2.3	2.34	65	0.49	0.89	0.91
Decreasing site heterogeneity
2.1	1.80	65	1.04	0.49	0.64
2.1	1.40	65	0.96	0.53	0.67
NICA-Donnan parameters			Model fit indices
log KCuFA1	log KCuFA2	%AFA	RMSE	Slope	r^2
Default values
0.26	8.24	65	1.34	0.72	0.78
0.26	8.24	30	0.78	0.67	0.77
0.26	8.24	20	0.79	0.64	0.76
0.26	8.24	15–30	0.61	0.78	0.84
0.26	8.24	15–65	0.32	0.94	0.96
Decreasing binding constant
0.26	8.00	65	1.14	0.64	0.55
0.26	7.50	65	1.09	0.56	0.50

Note:

The goodness-of-fit was evaluated by RMSE of free Cu²⁺ activity, and the slope of the linear regression line of measured v. predicted free Cu²⁺ activity. The best fit is shown in bold.

Figure 3. The proportions of Cu species calculated by WHAM 6 and measured free Cu²⁺ activity (log mol L⁻¹) in soil pore water for 34 samples at toxicity thresholds of EC_x (x ranged from 5 to 90). A lower and higher inhibition concentrations of Cu to barley root elongation were adopted for each soil, respectively.

Figure 4. Comparison of NICA-Donnan calculated free Cu²⁺ ion and measured free Cu²⁺ ion in the 34 samples of soil pore water. Free Cu²⁺ activity was predicted with NICA-Donnan using either (a) 65% AFA and default log K_CuFA2, (b) adjusted AFA rate from 15% to 65% and default log K_CuFA2.

Note: The thin solid line represents 1:1 line and dot line represents 1 order of magnitude either side of the 1:1 line. The thick solid line represents regression line.