Copper micronutrient fixation kinetics and interactions with soil constituents in semi-arid alkaline soils

ABSTRACT

This study examined the fixation pattern and kinetics of plant-available [diethylene triamine pentaacetic acid (DTPA)-extractable] copper (Cu), as well as basic soil properties that influence Cu availability in selected semi-arid soils. Soil samples from six different series were used and data obtained from Cu extraction experiments fitted to various kinetic models. Soils were also characterized for a suite of chemical and physical properties. The majority (80%) of the plant-available Cu fixed over the experimental period of 90 d occurred within the first 14 d. The amount of plant-available Cu fixed within the first 14 d tended to be influenced by the combination of organic matter (OM) and pH. The total amount of Cu fixed at the end of the experimental period of 90 d was influenced by pH and a combination of pH and calcium carbonate. The fixation of plant-available Cu over the experimental period was better described by the power function model [R² = 0.90, Standard Error (SE) = 0.099] but poorly by the other models (R²: 0.58 to 0.59), while reactions within the first 35 d were better described by the second-order model (R² = 0.98, SE = 0.008), suggesting a different fixation pattern. Findings from this study provide a basis for a more mechanistic approach to evaluating and comparing the fixation of Cu micronutrient compounds in these semi-arid soils for more scientific management decision making.

KEY WORDS:

• micronutrients
• diethylene triamine pentaacetic acid
• kinetic models
• soil fertility
• West Texas

1. Introduction

The environmental fate of micronutrients and their availability to plants are controlled by a variety of soil variables such as pH, organic matter (OM), calcium carbonate (CaCO₃) content, cation exchange capacity (CEC) and soil texture (Sharma et al. 2004; Ammari and Mengel 2006; Buekers et al. 2007; Wu et al. 2010; Najafi-Ghiri et al. 2013). Likewise, a number of soil reactions such as complexation with organic and inorganic ligands, ion exchange, adsorption and desorption processes, precipitation and dissolution of solids, and acid-base equilibria may affect the fate of micronutrients (Evangelou 1998; Sparks 2003; Sun et al. 2015). Thus, the chemistry of micronutrients could vary depending on soil characteristics, most of which vary temporally and spatially.

Copper (Cu) is one of the micronutrients, notable for its role in enzyme activation in plants (Havlin et al. 2013; Adhikari et al. 2016). It is also a known heavy metal and thus could be of environmental and health concern (Gaetke and Chow 2003; Udeigwe et al. 2015). Soil-solution Cu and plant-available portions are mainly controlled by soil pH, as solubility tends to decrease with increasing pH. Copper interaction with OM functional groups such as the carboxyl and phenol to form inner sphere complexes, a reaction that affects its plant available pool and environmental fate, has also been well documented (Evangelou 1998; Sparks 2003). Likewise, Cu interaction with other soil colloids such as clay minerals (McBride 1989; Sparks 2003; Hizal and Apak 2006) and oxides of Fe and Al (Karthikeyan et al. 1997; Evangelou 1998; Sparks 2003) has also been shown to reduce the amount of plant-available Cu. Competition among micronutrients resulting in antagonism is also widely noted (Havlin et al. 2013; Bindraban et al. 2015); for instance copper absorption by plants has been shown to be limited by elevated zinc (Zn), iron (Fe) and phosphorus (P) concentrations in the soil. In relation to soil physical properties, Cu deficiency is often more pronounced in coarse-textured soils (Sharma et al. 2002; Sidhu and Sharma 2010; Havlin et al. 2013).

As a result of the aforementioned interactions of Cu with soil constituents, total Cu concentration is not often a good indicator of the plant-available portion as a result of chemical processes that lead to the fixation of micronutrients in soils. Micronutrient fixation, a process that occurs over time, leads to the gradual reduction of the plant-available portion. The reduction of plant-available Cu will tend to follow an exponential decay pattern (Follett and Lindsay 1971), suggesting that its fixation in soils can be described by kinetic equations (such as zero-, first- and/or second-order equations, etc.). Kinetic parameters such as reaction rate constants obtained can be used for comparisons among micronutrients, and among soil types, as well as for evaluating various blended micronutrient products. Such information on the kinetics of micronutrient fixation will help to further understand nutrient dynamics and also to provide groundwork for the development and/or improvement of micronutrient management practices in agricultural systems. Unfortunately, some previous studies (Kuo and Mikkelsen 1980; Manouchehri et al. 2006; Reyhanitabar and Gilkes 2010; Abbas and Salem 2011) on kinetics of micronutrients in soil systems are challenged by their experimental conditions such as sample size, reaction times, sampling intervals, study duration (1 to 30 d), etc., which largely limited the applications of findings, particularly to field settings.

This study was partly prompted by the insufficient information on micronutrient chemistry in the semi-arid to arid regions and also due to the obvious fact that the heterogeneity of soils makes the application of findings from one soil type and region to another often misleading. Despite the agronomic significance of the Southern High Plains (SHP) of the United States, an area that produces over 40% of total US cotton (USDA-NASS 2014), there is still limited information on the chemistry of micronutrients in soils of this region.

An extensive literature search revealed that the questions of how much of applied plant-available Cu will be present at a specific time, what the reaction rates and mechanism of Cu fixation are, and how these could compare to those of other micronutrients in semi-arid soils remain largely unanswered. This research was therefore conducted to (1) examine the fixation pattern and kinetics of plant-available [diethylene triamine pentaacetic acid (DTPA)-extractable] Cu in semi-arid alkaline soils of the SHP, and (2) examine soil properties that influence Cu availability in these semi-arid soils. Findings will help us to further understand the chemistry of Cu in the semi-arid alkaline soils of the SHP and could be extended to other arid and semi-arid regions of the world with similar soil type for improved micronutrient management.

2. Material and methods

2.1. Soil description and sampling

Soil samples were collected from six different production sites in West Texas at 0–15 cm and 15–30 cm depths, covering five agriculturally important soil series [Acuff (O), Amarillo (A), Pullman (P), Pyron (S) and Mansker (M)] in the SHP of the US for a total of 12 soil samples (Table 1). The sites where these soils were collected fall within a semi-arid climate with a 30-year average annual precipitation of approximately 470 mm and a mean annual temperature of ~16°C (NOAA 2014). Soils of interest to this study were identified using the Web Soil Survey of the Natural Resources Conservation Services. At each representative site, soil samples were randomly collected from 12 to 15 spots and combined to get a composite sample. In some cases, the sampling was restricted to a defined area to avoid crossing into a different soil series. Sampling was conducted using a digging spade, marked to show the designated depths. Approximately 10 kg of each soil type was collected.

Table 1. Soil classification, identification and selected properties of the studied semi-arid alkaline soils of the Southern High Plains, USA.

Series (classification)	Sample	Sampling	pH	EC	OM	CaCO3	Clay	Sand	Silt	Textural
	identification	location		dS m^−1	%					class*
Pyron (Fine, mixed, superactive, thermic Typic Argiustolls)	Sa	32.7211°N,	7.52	0.22	1.56	2.34	32.8	39.9	27.3	CL
	Sb	100.8386°W	8.02	0.24	1.57	4.16	33	38.7	28.3	CL
Acuff (Fine-loamy, mixed, superactive, thermic Aridic Paleustolls)	Oa	33.8403°N,	8.39	0.35	1.31	4.68	29.7	49.1	21.3	SCL
	Ob	101.6999°W	8.36	0.27	1.43	4.55	37.6	50.7	11.7	SC
Amarillo (Fine-loamy, mixed, superactive, thermic Aridic Paleustalfs)	A1a	33.5935°N,	8.18	0.23	1.7	1.95	26	61.1	12.9	SCL
	A1b	101.9058°W	8.24	0.21	1.35	1.17	24.7	61.7	13.6	SCL
	A2a	33.6058°N,	8.32	0.17	0.59	0.46	19.4	72.4	8.2	SL
	A2b	101.9073°W	8.3	0.17	0.75	0.65	23.2	65.7	11.1	SCL
Pullman (fine, mixed, superactive, thermic Torrertic Paleustolls)	Pa	34.05901°N,	8.07	0.24	1.05	2.41	17.8	74.9	7.4	SL
	Pb	101.4773°W	8.35	0.22	0.93	5.59	24	64.2	11.8	SCL
Mansker (coarse-loamy, carbonatic, thermic Calcidic Paleustolls)	Ma	34.1261°N,	8.12	0.27	1.63	0.98	27.9	56.5	15.6	SCL
	Mb	101.5899°W	8.2	0.25	1.24	0.13	41.9	40.9	17.2	C

EC, electrical conductivity; OM, organic matter; CaCO₃, calcium carbonate; SCL: sandy clay loam; SC: sandy clay; SL: sandy loam; CL: clay loam; C: clay; a = 0–15 cm; b = 15–30 cm.

2.2. Sample preparation

Collected soil samples were thoroughly mixed and a representative portion of each placed in a 4-L plastic pot. Sorghum (Sorghum bicolor) was grown over a period of 35 d in a greenhouse, an optional practice aimed at further depleting the original nutrient level of the soils prior to treatment application. After this practice, the soils were air-dried, thoroughly mixed, ground, sieved with a 2-mm sieve and stored in plastic bags at approximately room temperature (23°C).

2.3. Soil characterization

Soil samples were analyzed for pH, electrical conductivity (EC), OM, percent CaCO₃, particle size, exchangeable cations [calcium (Ca), magnesium (Mg), sodium (Na), and potassium (K)], total elements [aluminum (Al), boron (B), Cu, Fe, K, Mg, manganese (Mn), molybdenum (Mo), P and lead (Pb)], Mehlich 3-P, and DTPA-extractable (plant-available) micronutrients (Cu, Fe, Mn and Zn) following approved routine analytical methods.

Soil pH_1:2 and EC_1:2 were measured at a 1:2 soil/water ratio using the applicable methods described by Sparks et al. (1996). Soil OM was estimated using the loss on ignition method following the procedure by Nelson and Sommers (1996). Percent CaCO₃ was determined using tensimeter methods 4E and 4E1 of the United States Department of Agriculture–Natural Resources Conservation Services Soil Survey Investigation Report (Soil Survey Staff 2014). Soil particle size was determined using the modified hydrometer method as described by Gee and Bauder (1986). Soil-test P was determined using the Mehlich-3 procedure (Mehlich 1984). Total elemental analysis was conducted using the DigiPREP Digestion System (DigiPREP MS, SCP SCIENCE, Quebec, Canada) following the US Environmental Protection Agency (USEPA) Method 3050B (USEPA 1996). Initial plant-available micronutrients were determined by DTPA extraction following the procedure by Lindsay and Norvell (1978). Concentration of elements in extracts were measured using inductively coupled plasma-optical emission spectroscopy (ICP-OES) (iCAP 7400, Thermo Scientific, Waltham, MA). All analyses were conducted in duplicate.

2.4. Treatment and extraction procedure

Approximately 250 g of soil sample from each series was treated with 80 mL of copper(II) sulfate pentahydrate (CuSO₄.5H₂O) solution prepared to deliver 10 mg Cu kg⁻¹ soil. Thus, there were a total of 12 soil samples (six sampling sites by two depths) for 12 treatments, each replicated twice. All treated samples were kept in an open space in the laboratory and subjected to the same conditions at room temperature of approximately 23°C. Subsamples were taken from each treated soil at approximately 2, 5, 8, 14, 21, 28, 35, 49, 63, 77 and 90 d and analyzed for plant-available micronutrients using the DTPA-extraction technique. After each subsampling, the remaining soil samples were wetted to approximately 50% of the field capacity to initiate a wetting and drying cycle and to facilitate chemical reactions in the soil. Within the first 7 d, the soils were wetted after each subsampling event, and then wetted at 7-d intervals afterward.

The preparation of the DTPA extractant and the extraction procedure followed the method described by Lindsay and Norvell (1978). In brief, 10 g of air-dried soil sample was placed in a 50-mL plastic tube, and 20 mL of DTPA-extracting solution was added. Tubes were placed on a reciprocal shaker for 2 h at room temperature and 180 oscillations min⁻¹. After shaking, samples were centrifuged for 10 min at 4000 rpm, and the solutions filtered into 16-mm boroglass tubes using a Whatman No. 2 filter paper. All filtrates were analyzed for Cu using the ICP-OES (iCAP 7400, Thermo Scientific, Waltham, MA) following USEPA Method 200.7 (USEPA-ICP Users Group 1982). All analyses were conducted in duplicate. Instrument calibration was conducted using standard reference materials and evaluated using second source standards from a different vendor. Check samples were inserted after every 20–25 samples. Relative percentage differences (RPD) between duplicates were also examined, and 10% was set as the acceptance standard.

2.5. Statistical analyses

All statistical analyses were performed using the Statistical Analysis Software (SAS 9.4, SAS Institute, Cary, NC). Where applicable, differences among means were examined using PROC GLM and mean comparison conducted using Fisher’s least significance difference at α level of 0.05. Single and multiple linear regression analyses between available nutrients and soil properties were examined using the PROC REG procedure. Experimental data obtained from the kinetic studies were fitted to selected kinetic models to examine model fit and derive the needed parameters using the PROC NLIN procedure.

3. Results and discussion

3.1. Soil characteristics

Selected chemical and physical properties of the soils are presented in Table 1. Soil pH values ranged from 7.52 (Pullman-P_a) to 8.39 (Pyron-S_a) with an average pH of 8.17 indicating that the soils are alkaline. Soil pH values showed no clear trend with depth. Soil OM ranged from 0.59% (Amarillo 1-A1_a) to 1.7% (Acuff-O_a), and also showed no clear trend with depth. The average OM content of 1.26% observed in the examined soil is typical of the soils of the SHP (Udeigwe et al. 2015). The average soil EC_1:2 of 0.24 dS m⁻¹ suggests there are no apparent salinity or sodicity hazards in these agricultural soils. CaCO₃ content varied widely from 0.13 to 5.59%, with an average concentration of 2.42%. The majority of the differences in properties observed among these agricultural soils could be ascribed to their geological differences and also to the variation in management practices. The average sand and clay contents of approximately 56 and 28%, respectively, indicate the soils are predominantly sandy, with the exception of the Pullman and the subsurface soil of the Mansker which are clay loam and clay, respectively (Table 1).

The estimated total and plant-available concentrations of elements are summarized in Table 2. The total concentrations of elements such as Al, Ca, Cu, Fe, Mn, Zn and P were within typical background levels found in most soils (Adriano 2001; Kabata-Pendias 2010). Interactions among soil constituents, which often influence nutrient chemistry, were also examined using regression analyses. Some of these interactions were worth noting since they were significant, despite the small sample number (n) of 12 used. Positive and significant relationships were observed between soil pH and each of total Al, Fe and P (R, 0.59 to 0.66; P, 0.03 to 0.04). Notably, soil clay content was also significantly correlated with total Ca (R = 0.64, P < 0.03) and total Cu (R = 0.73, P < 0.01). Among the elements, total Fe was positively correlated with each of total Al, Cu and Mn (R, 0.62 to 0.99; P, 0.001 to 0.04). A positive correlation between Fe (particularly in the form of Fe (III) oxide) and other heavy metals such as Cu is often used to distinguish between natural levels of these elements from anthropogenic sources, with more correlation with Fe often suggesting more lithogenic influences (Presley et al. 1992; Li et al. 2000; Eze et al. 2010).

Table 2. Soil total elemental content and plant-available estimates for the studied semi-arid alkaline soils of the Southern High Plains, USA.

		Total						Plant-available estimate
Series	Sample ID	Al	Ca	Cu	Fe	Mn	Zn	CuDTPA	FeDTPA	MnDTPA	ZnDTPA	PM3
		– – –– – –– – –– – –– – – – – – – –mg kg^−1– – – – – – – – –– –––						– – –– – – – – – mg kg^−1 – – – – – – – – –––
Pyron	Sa	21305	19377	7.9	14188	190	43.2	0.63	4.13	3.76	0.40	24.3
	Sb	22359	19031	7.7	14471	183	43.6	0.68	5.20	4.07	0.32	13.2
Acuff	Oa	18429	8769	6.1	12164	158	39.8	0.54	3.29	4.31	0.09	7.83
	Ob	18424	6551	6.2	11997	138	42.9	0.53	4.00	3.55	0.13	3.89
Amarillo 1	A1a	14835	1775	4.9	10596	149	38.9	0.39	3.37	4.67	0.24	15.7
	A1b	16631	1830	5.3	11601	164	43.6	0.48	4.46	5.57	0.17	8.95
Amarillo 2	A2a	17187	1823	5.5	11946	170	44.5	0.97	9.73	4.53	0.96	123
	A2b	13823	4307	5.6	9623	144	41.6	0.54	3.32	3.25	0.20	58.7
Pullman	Pa	11571	7252	6.1	8191	128	42.8	0.59	3.79	4.97	0.56	28.1
	Pb	12943	14433	6.0	8964	130	41.6	0.85	4.97	4.29	0.17	11.6
Mansker	Ma	13808	21008	6.1	9552	135	51.5	0.99	6.83	5.92	0.20	33.4
	Mb	16840	11584	7.1	11856	202	59.1	0.89	3.14	7.09	0.19	8.93
Mean		16513	9812	6.0	11262	158	44.0	0.67	4.69	4.67	0.30	28.1
SD		3156	6874	0.88	1869	23.6	5.33	0.20	1.91	1.09	0.24	33.5

Al, aluminum; Ca, calcium; Cu, copper; Fe, iron; Mn, manganese; Zn, zinc. DTPA: diethylene triamine pentaacetic acid; M3: Mehlich 3; SD: standard deviation.

The initial background levels of available nutrients are shown in Table 2. Mean concentrations of estimated plant-available Fe, Mn, Cu, and Zn were 4.69, 4.67, 0.67 and 0.30 mg kg⁻¹, respectively. These concentrations are typical of those found in agricultural soils with insufficient levels of these nutrients (Havlin et al. 2013). Soil test P varied widely, ranging from 8.93 to 123 mg kg⁻¹. The high level of P (average of 90.0 mg kg⁻¹) found in the Amarillo 1 soil is likely as a result of P input through fertilization.

3.2. Short- and long-term copper fixation examination and relationship with soil constituents

The amount of plant-available Cu fixed within the first 14 d (designated short-term fixation) and the total amount fixed at the end of the experimental period of 90 d (long-term fixation) were individually examined and further related to the measured soil properties to determine whether these were controlled by different group of soil variables. It is important to note that only days 14 (short term) and 90 (long term) were selected for this purpose and the percentage estimates were approximated using the differences between the amounts of plant-available Cu at days 2 and 14, and days 2 and 90, respectively. For simplicity, the results are presented as the averages for all soils (Table 3), because individual soil examination showed no justifiable differences or pattern among the soils that worth focusing on. Furthermore, the increase in numbers of samples used will also enhance the statistical evaluation of the data. The amount of Cu fixed after 14 and 90 d in the selected semi-arid alkaline soils at each depth and in both depths are summarized in Table 3. The findings indicated that the majority (approximately 80%) of the available Cu was tied up into less plant-available forms within the first 14 d. The amount of Cu fixed in the first 14 d was higher in the subsurface soil compared to the surface, although not statistically different (P > 0.05), and this can be partly attributed to differences in the content of clay, which creates more sorptive surfaces for Cu, partly decreasing its solubility and availability to plants (Martı́nez and Motto 2000; Veli and Alyüz 2007).

Table 3. Average percentage (with standard deviation) of added copper (Cu) fixed after 14 and 90 d in the selected semi-arid alkaline soils of the Southern High Plains, USA.

	% Fixed after
Depth (cm)	14 d	90 d
0–15^c	40.7 (25.2) a	43.8 (21.9) a
15–30^c	31.4 (13.8) a	46.3 (19.6) a
All^d	36.0 (19.5)	45.1 (20.7)

c: n = 6; d: n = 12. Mean values in a column with the same lowercase letter are not statistically different (Fisher’s Least Significant Difference at α level of 0.05).

The influence of soil properties on Cu availability was also examined. Similarly, short-term fixation (i.e., the amount of available Cu fixed after 14 d) was examined separately from long-term fixation (i.e., the total amount fixed after 90 d) Table 4). The purpose was to examine whether a separate group of soil variables were responsible for Cu fixation at these stages. Findings presented in Table 4 suggest that Cu fixation that occurred within the first 14 d (short term) in these semi-arid alkaline soils was not significantly related to any of the measured soil properties when examined individually, but showed a mild trend with the combination of OM and pH in a multiple regression (R² = 0.42). The total amount of Cu fixed after 90 d (long term) showed a mild positive trend with pH (R² = 0.32), and a weaker (and non-significant) trend with OM and CaCO₃ (R² = 0.18 and 0.20, respectively). An increasing fixation of Cu with pH is expected because the solubility of metals generally decreases with soil alkalinity due to likely interactions with CaCO₃ (if calcareous) (Martı́nez and Motto 2000) and the formation of insoluble metal hydroxides (Sposito 1989; Sparks 2003). Multiple regression analysis revealed that the combination of CaCO₃ and pH significantly explained the variability (R² = 0.66, P < 0.02) associated with fixation of plant-available Cu in these semi-arid soils within the experimental period of 90 d. The complexation of metal ions such as Cu by soil OM has been widely documented (Sparks 2003; Da Rosa Couto et al. 2015). Furthermore, metal-carbonate interaction, a process that can affect the availability of metal micronutrients such as Cu, is also widely documented (Sposito et al. 1982; Tack and Verloo 1995; McLaren and Clucas 2001). Because CaCO₃ is likely the dominant carbonate form present in these semi-arid soils, a relationship between available Cu and CaCO₃ is somewhat expected, although not significant in this study. The findings here suggest that Cu fixation in the short and long terms may be influenced by different pools of soil properties.

Table 4. Coefficient of determination (R²) of relationships between percentage of plant-available copper (Cu) fixed after 14 and 90 d and selected soil properties of the semi-arid soils of the Southern High Plains, USA (n = 12).

	Plant-available Cu fixed after:
	14 d	90 d
Soil property	R^2	R^2
Clay	0.00	0.02
CaCO3	0.03	0.18
OM	0.12	0.2
pH	0.12	0.32
EC	0.06	0.08
PTotal	0.06	0.12
OM + pH	0.43	-
clay + OM	0.20	-
pH + EC	0.17	-
CaCO3 + pH	-	0.66^a
pH + EC	-	0.44
OM + PTotal	-	0.42

OM: organic matter; CaCO₃, calcium carbonate; EC: electrical conductivity, P_Total: total phosphorus; a: significant at P < 0.05.

3.3. Kinetics of copper fixation

Copper fixation kinetics in these semi-arid soils were examined by fitting the data obtained from the kinetic experiment to various kinetic models. A number of kinetic models commonly applied in the field of soil science (Table 5) were selected based on findings in the literature (Dang et al. 1994; Reyhanitabar and Gilkes 2010) and the experimental conditions of this particular study. Evaluations were made among the six soils as well as within depths (0–15 cm and 15–30 cm). Coefficient of determination (R²) and standard error (SE) were used as criteria for evaluating best fit among the models (Dang et al. 1994; Reyhanitabar and Gilkes 2010). A better fit to the zero-order model implies that the rate of reaction does not depend on the concentration of the reactant (Cu); first-order implies that the rate of reaction is dependent on the concentration of only one reactant (e.g., Cu); and second-order implies that the reaction rate depends on the concentration of two reactants (e.g., Cu and another soil constituent) (Evangelou 1998; Sparks 2003).

Table 5. Kinetic models used for the study of copper (Cu) fixation in the semi-arid alkaline soils of the Southern High Plains, USA.

Kinetic model	Equation	Parameter
Zero order	qt = q0 – k0t	k0, zero-order rate constant (mg kg^−1 d^−1)
First order	ln qt = ln q0 – k1t	k1, first-order rate constant (d^−1)
Second order	1/qt = 1/q0 + k2t	k2, second-order rate constant [(mg kg^−1)^−1 d^−1]
Power function	qt = at^b	a, initial reaction magnitude constant [(mg kg^−1 (d^−1)^b)], and b, reaction rate constant

q₀ and q_t are the amount of micronutrient at time zero and t, respectively.

The soils were examined individually and evaluations also made within depths (0–15 cm and 15–30 cm). However, the depth examination showed no findings worth focusing on; thus, the results presented here are the averages of both depths. This approach also enabled the use of more data points, thus enhancing the statistical evaluation of the data. The data points from the first 35 d were further examined separately because of an observed discontinuity in slope after the first 35 d, suggesting a possible difference in Cu fixation mechanisms in the short (35 d) and long (90 d) terms.

Coefficients of determination (R²) of kinetic models used in describing Cu fixation in each of the semi-arid soils of the Southern High Plains over the experimental period of 90 and 35 d are presented in Table 6. It is evident that the strengths of the models vary among the soils, with Amarillo 2, Pullman and Mansker more poorly described (Table 6). Additional studies may be needed to elucidate the poor fits observed in these soils. Although there were no drastic differences in soil properties that could confidently be attributed to this, the original average concentrations of estimated plant-available (DTPA) Cu, CaCO₃ and OM in the poor fit soils (Amarillo 2, Pullman, Mansker) compared to the other soils (Acuff, Pyron, and Amarillo 1) were 0.54 vs. 0.81 mg kg⁻¹, 3.14 vs. 1.70%, and 1.49 vs. 1.03%, respectively. The generally lower average DTPA Cu and higher CaCO₃ and OM contents in the poor-fit soils could imply that the available Cu was quickly tied up by these other soil constituents, thus not following the gradual decay pattern observed in the other soils. The individual examination of the Cu fixation kinetics for each soil did not show justifiable reasons to focus our discussion on the comparison of soils as earlier intended. Thus, for further examination of Cu fixation, average data points for all soils were used. This approach also helped to enhance the statistical reliability of the findings since more data points were employed. The data points were averaged across soils and fitted to the various kinetic models to examine the pattern of Cu fixation in the semi-arid soils of the SHP. Figure 1 shows Cu data fitted to zero-, first-, second-order and power function models, where qt represents the amount of DTPA extractable (plant-available) Cu remaining at time t, in d. Across the soils, Cu fixation was poorly described by the zero-, first- and second-order models (R²: 0.58 to 0.59), with the power function model (R² = 0.90, SE = 0.099) having the best fit. The better fit to the power function model obtained here indicates that the fixation of plant-available Cu is somewhat not linear over the experimental duration of 90 d, an indication of a more complex fixation pattern.

Figure 1. Diethylene triamine pentaacetic acid (DTPA)-extractable copper (Cu) over the long term (90 d) fitted to (a) zero order, (b) first order, (c) second order and (d) power function models [qt = amount remaining at time t, (mg kg⁻¹); error bars are for standard errors computed from 12 data points (six soils by two depths)].

Table 6. Summary of coefficients of determination (R²) of kinetic models used for describing micronutrient fixation for experimental periods of 90 and 35 d in the semi-arid soils of the Southern High Plains, USA.

	Zero	First	Second	Power
90 d
Pyron	0.60	0.69	0.81	0.68
Acuff	0.77	0.77	0.81	0.71
Amarillo 1	0.77	0.72	0.81	0.85
Amarillo 2	0.09	0.05	0.02	0.22
Pullman	0.04	0.01	0.03	0.11
Mansker	0.24	0.13	0.04	0.28
35 d
Pyron	0.66	0.75	0.80	0.58
Acuff	0.84	0.82	0.77	0.67
Amarillo 1	0.74	0.82	0.88	0.96
Amarillo 2	0.40	0.44	0.45	0.63
Pullman	0.31	0.24	0.16	0.40
Mansker	0.25	0.13	0.04	0.28

However, within the first 35 d, it was evident that fixation of available Cu was better described by the second-order model (R² = 0.98, SE = 0.008). The findings here suggest that the reaction rate governing the fixation of availability of Cu within the first 35 d most likely depends on the concentration of other soil constituents, e.g. another soil element, OM, etc. As evidenced from Fig. 1, there appears to be a more rapid and somewhat linear decrease in the amount of available Cu within the first 35 d, after which the lines tended to level out, implying that majority of the available Cu will be fixed within the first 35 d in these semi-arid soil. This is supported by the higher reaction rate constant of 0.281 for 35 d (data not shown) vs. 0.215 for the 90-d experimental period (Fig. 1), obtained from the power function models. This initial rapid decrease followed by a decrease at a slower rate has also been reported by Ma et al. (2006). These findings highlight the significance of timing in Cu management in these soils.

Limited comparisons could be made between the findings from this study and the few other studies on micronutrient kinetics in soils (Dang et al. 1994; Reyhanitabar and Gilkes 2010). These studies were conducted within shorter reaction times of 1 to 10 d under laboratory conditions, where soil shaking is often employed to enhance reactions. Their experimental conditions limit the direct application of findings to field settings.

3.4. Application of findings to copper management

The reactions leading to the fixation of Cu in the examined semi-arid soils tended to follow the power function model, an indication of a more complex fixation pattern. Evidence gathered from the short- and long-term Cu fixation examinations strongly suggests the importance of timing for Cu management in these soils, as the majority of the available Cu was fixed within the first 14 d. The reaction rate constants obtained from this study could be used to approximate how much added Cu micronutrient will be available at a specific point in time, although this can significantly vary among soils. A notable application of the findings from this study will be for the comparison of the fixation pattern of Cu to those of other micronutrients within these semi-arid soils. Likewise, the reaction rate constants obtained from these soils can also be compared to those obtained for Cu in other soils. Of interest, applications developed from this study provide a basis for a more mechanistic approach to the evaluation and comparison of commercial Cu micronutrient products to conventional products by the examination of their fixation patterns and kinetic parameters. A compilation of the reaction rate constants derived for different chelated Cu compounds can be compared among themselves or against conventional (non-chelated) compounds, and this information utilized in making Cu management decisions for these semi-arid soils. A similar approach could also be applied to the other micronutrient elements such as Fe, Mn and Zn.

4. Conclusions

The reduction of plant-available Cu in the studied semi-arid soils followed the power function model, suggesting a more complex fixation pattern. However, in the short term (35 d) Cu fixation was better described by the second-order model, suggesting a reaction rate that does not depend only on the concentration of Cu. Findings strongly highlight the importance of timing in Cu management in these soils. The reaction rate constants obtained from this study could be used for the comparison of Cu fixation pattern to those of other micronutrients within these soils, as well as for comparisons among soils. This will also provide a basis for evaluating and comparing different Cu micronutrient blends, for more scientific decision-making. This is the first known study to probe Cu fixation kinetics in these semi-arid soils, and thus it provides important background information for future micronutrient modelling studies.

Acknowledgment

The authors wish to thank the College of Agricultural Sciences and Natural Resources, Texas Tech University, for providing the research enhancement award that partly funded this study.