Next Article in Journal
Both Multi-Segment Light Intensity and Extended Photoperiod Lighting Strategies, with the Same Daily Light Integral, Promoted Lactuca sativa L. Growth and Photosynthesis
Previous Article in Journal
Effects of Nitrogen Management on Biomass Production and Dry Matter Distribution of Processing Tomato Cropped in Southern Italy
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Agronomy
Volume 9
Issue 12
10.3390/agronomy9120856
agronomy-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Prediction Models for Bioavailability of Mn, Cu, Zn, Ni and Pb in Soils of Republic of Serbia

by
Zoran Dinić
*,
Jelena Maksimović
,
Aleksandra Stanojković-Sebić
and
Radmila Pivić
Institute of Soil Science, Teodora Drajzera 7, 11000 Belgrade, Serbia
*
Author to whom correspondence should be addressed.
Agronomy 2019, 9(12), 856; https://doi.org/10.3390/agronomy9120856
Submission received: 4 November 2019 / Revised: 2 December 2019 / Accepted: 4 December 2019 / Published: 6 December 2019
Browse Figure
Versions Notes

Abstract

:
The bioavailability of trace elements (TEs) is one of the major factors for successful plant production and environmental protection. The aim of this study was to determine the extent to which TEs are bioavailable and which of the basic soil parameters affect bioavailability. The survey included agricultural soil samples taken from 240 locations on the territory of the Republic of Serbia, where the soil analytics were carried out. On the basis of the analyzed data the prediction models were derived based on the Freundlich model, showing the dependence between trace elements (TEs) extracted using the DTPA buffer solution in relation to the trace elements extracted using an aqua regia, the organic matter content (SOM), the clay fractions content, and soil pH. On one part of the samples, the prediction models were separated on the basis of a suspension for determining the pH in H2O and 1M KCl. The model was applied for the following TEs: Mn, Ni, Pb, Zn, and Cu. The content of the pseudo total forms of TEs statistically significantly influenced the bioavailability of TEs in all prediction models for all studied elements. The pH value statistically significantly affected the bioavailability of Ni, Mn, Pb, and Cu also in all prediction models. The impact of SOM and clay varied depending on the model and TEs. Multiple linear regression showed that the prediction models for TE Cu (R2 = 0.763–0.848) were the most reliable and that the bioavailability of Cu was significantly influenced by all the studied soil parameters except clay. Reliable results were also shown by the prediction models for TE Pb, but the values of the determination coefficient and investigated parameters that influenced the bioavailability varied depending on the model. The derived models for TE Mn, Zn, and Ni were less reliable (R2 is approximately 50% or less), and the effect of the tested parameters on bioavailability varied depending on the model and TEs.

1. Introduction

Soil contains the main source of trace elements (micronutrients and non-essential elements) for plants [1]. There are two main transfer pathways that are significant for the evaluation of trace elements (TEs) risk in relation to living organisms: the soil-plant pathway, such as fodder and food crop, which introduces TEs into the food chain, and/or phytotoxicity and direct uptake through ingesting or inhalation by humans and animals [2].
The TEs bioavailability is considered to be the most decisive problem in both agricultural and environmental studies [1]. Before the TEs become bioavailable to plants and other biota soil, they have to be mobile [3]. Plants are known to have mechanisms that regulate the uptake of TEs under conditions of deficiency or excess in soil. As the mechanisms by which the exclusion of TEs in excess are much weaker than those in deficit, the excess of TEs presents greater stress for the plant [1]. The content of TEs in soil, in addition to its geochemical origin, has been increasing in recent decades through anthropogenic impact, intensive agricultural production, and industrialization. The mobility of TEs in soil is influenced by many factors, including the most important such as soil pH, SOM, and chemical speciation, while the clay content and redox potential are considered to be less important [3].
According to Hough [4], copper (Cu) is considered to be an essential TE for human life and health, as well as an essential TE in plants. It is a constituent of several “key” enzymes, playing an important role in physiological processes [5]. Cu in soil originates from the parent substrate and anthropogenic sources [6]. Agricultural practice also contributes to the inputs of Cu to soil. Copper is present in some agrochemicals, animal feed, as well as in the co-products, such as livestock manure (in some cases as a result of feeding comprising Cu), compost, and the sludge from wastewater [4].
Manganese (Mn) is an essential TEs in plant nutrition and has several important functions [7,8]. The most important function of manganese in plants is related to the processes of oxidation and reduction [9]. The total Mn content in the soil is practically derived from the parent substrate, and as a deficiency element for some plants, it is added through fertilizers [6].
According to Adriano [7], trace amounts of nickel (Ni) in plants are required for certain biological processes, although its toxicity has been recorded at elevated levels. However, Kabata-Pendias and Szteke [9] noted that there is no evidence of an essential role of Ni in plant metabolism, but there are several suggestions that it might be needed for plants. Recent studies of Wenzel et al. [10] confirmed that Ni is essential for plants. In addition to the parent substrate, Ni enters the soil anthropogenically [6]. In most soils, Ni is present at moderate levels, being the fifth most abundant element on earth [11].
Zinc (Zn) is an essential TE for plant nutrition but required in only low concentrations, unlike the major nutrient elements that are ordinarily supplied in mixed fertilizers [7]. It is deficient in the soil but also phytotoxic due to industrial contamination [12]. High concentrations of Zn are toxic to both plants and soil microorganisms and may have harmful effects on soil fertility and yield of crops [13].
Lead (Pb) is known and classified as a carcinogenic and toxic TE for most living organisms at higher exposure. Unlike most TEs that occur in soil, there is no demonstrated biological need for Pb [14].
The content of TEs in soils should be determined in order to identify the contaminated soils, and on soils where the deficiency of TEs is found, compensate them for successful crop production [6]. Based on the literature review of many authors [4,5,7,8,9,10,12,13], it is considered that most of the TEs studied, in addition to their essential roles in plant metabolism, also act phytotoxically at certain levels in given soil conditions, so it is important to determine their soil content and bioavailability to the cultivated plants. Bioavailability presents the part of a trace element that is or has been available to cultivated plants [5].
Most countries still utilize the total soil TEs content as a simple index of hazard in contaminated soils in their current legislation. Nevertheless, total TEs load takes no account of soil properties that modify the bioavailability of TEs pollutants in contaminated soils [15]. The Dutch have recognized the problem of the TEs behavior in soil and, based on years of research, adopted in their legislation that the determination of target and remediation values for TEs in soil is carried out with correction depending on the content of clay and organic matter [16].
The soil test procedure, which includes DTPA-TEA or diethylenetriaminepentaacetic acid-triethanolamine, is among the more widely used techniques in the identification of soils with inadequate levels of available Cu, Mn, and Zn [17]. Kashem and Singh [18] signify TEs removed by the chelating agent DTPA as the plant-available fractions.
In both agricultural and environmental studies predicting the mobility in soil and bioavailability of TEs has been a “hot topic” for years [1]. The prediction of TEs solubility uses the regression models as an efficient tool, where most of the models consist of pH, OM, and total content of TEs [19].
The Freundlich isotherm is most often used to describe the adsorption of TEs in soil depending on soil parameters [7]. Many researchers have applied the principle of Freundlich isotherm by examining the content of TEs in a plant depending on individual soil parameters [20,21,22,23]. In their research, de Souza Braz et al. [24], Buchter et al. [25], Sauvea et al. [26], and Sauvea et al. [27] applied the Freundlich isotherm and determined the dependence of TEs mobility on certain soil parameters. Kashem and Singh [18] used stepwise multiple-regression procedures to determine the relationship of the mobile, immobile, and DTPA fraction of TE (Ni, Cd, and Zn) with organic matter, clay, and dust. Due to the importance of determining TEDTPA quantities (referred to earlier as plant-available fractions) and not having a complete prediction model that will include the most important soil parameters as factors affecting the mobility of TEs and pseudo total TEs forms (TEAR), we propose to test the following prediction model:
log (TEDTPA) = A + b*pH + c*log(SOM) + d*log(Clay) + e*log(TEAR).
The prediction model presented by this study is based on the Freundlich model showing the dependence of TEs extracted with DTPA buffer solution—available forms (TEDTPA) in relation to TEs extracted by aqua regia—pseudo total forms (TEAR), organic matter content (SOM), clay fraction content, and soil pH. Soil parameters—pH, SOM and clay, recognized by many authors as highly influential on the mobility of TEs, can be routinely determined in laboratories [21]; thus, their inclusion in the prediction model intended to serve a larger number of users for future predictions of TEs behavior in soil.
After determining the physico-chemical parameters and the concentration of TEs in DTPA and TEs in aqua regia, the aim of this study was to determine which of the tested independent variables affects the bioavailability of TEs (dependent variable). The research will define the prediction models.
Due to the different R2 values in studies of Gavriloaiei [28], and Thomas [29], and the unequal difference between the pH of water suspension and 1M KCl suspension in determining the pH of the soil, we decided to perform separate prediction models on one part of the samples for both solvents. The coefficient of determination of the prediction models obtained (for pH in water suspension and 1M KCl suspension) would indicate whether there is a difference in the choice of solvents for the future formation of prediction models and interpretation of the results.

2. Materials and Methods

2.1. Soil Sampling and Preparation

Samples of the soil material were taken from 240 sites belonging to agricultural soil on the territory of the Republic of Serbia, and one composite sample was taken in the period 2010–2018. The areas covered by this survey are: Municipality of Veliko Gradište, Mačva and Toplica districts, a section of the E75 highway through Serbia from Belgrade to Preševo, and the area of Radočelo Mountain. Samples were taken in disturbed conditions from each site, and a composite sample, consisting of 20–25 individual samples, was taken using a soil sampling probe from a depth of 0–30 cm. After sampling, samples were prepared for analyses in accordance with SRPS ISO 11464 [30].

2.2. Soil Analysis

Soil pH was determined potentiometrically (in water, in 1M KCl 1:5 v/v) according to SRPS ISO 10390 [31]; SOM content was determined by dry combustion according to SRPS ISO 10694 [32]; clay content was determined by the pipette method (modified International B method) [33]; TEs content was determined by extraction with DTPA buffer solution according to SRPS ISO 14870 [34], and detection using inductively coupled plasma-atomic emission spectrometry (ICP-AES) according to ISO 22036 [35], and also by extraction with aqua regia according to SRPS ISO 11466 [36] and detection using inductively coupled plasma-atomic emission spectrometry (ICP-AES) according to ISO 22036 [35]. For quality control in the analyses, every tenth sample was analyzed in duplicate using certified reference materials, as follows: ERM-CC141 loam soil (TEs), NCS ZC73005 soil (SOM), and Eurosoil 7 (SOM).

2.3. Statistical Analysis

To model the dependence of soil TEDTPA (γi) in relation to pH (χ1), SOM (χ2), clay (χ3) and soil TEAR (χ4), a multiple linear regression model with logarithmic values of all input variables was used as follows:
log(γi) = β0 + β1log(χ1i) + β2 log(χ2i) + β3 log(χ3i) + β4 log(χ4i) + εi.
For ease of the Freundlich models, the values of these parameters are logarithmic, except for the pH value, which is itself a negative logarithm of the H+ ion concentration. The dependence of the regression coefficients was determined at the 5% level. The criterion for evaluating the suitability of the prediction model is represented by the coefficient of determination (R2).
Descriptive statistics and multiple linear regression models were obtained from the SPSS (IBM SPSS Statistics 25, Armonk, New York, USA) and Microsoft Excel 2016 (Redmond, Washington, USA) statistical programs.
A prediction model based on the pH value obtained in 1M KCl solution (model 1) was determined on a bulk of 240 samples. For obtaining the prediction models, especially for the pH value in 1M KCl suspension (model 2) and water suspension (model 3), a bulk of 204 soil samples was used (the bulk of samples was reduced by samples from the section of the Belgrade–Preševo highway E75).

3. Results and Discussion

3.1. Physicochemical Properties of the Test Soil

The studied soils are very different in terms of the analyzed chemical and physical properties, which is to be expected since there are a lot of samples and different localities (Figure 1). The results of the analyzed soil samples for the bulk of 240 samples (model 1) are presented in Table 1.
A total of 83% of the tested samples had a highly acidic to acidic soil reaction for pH in 1M KCl; 41% of the samples had high to extremely high SOM content; 66% of samples had medium and high content of clay fraction [39]. Criteria for interpreting the contents of TEAR, as well as the assessment of soil pollution, are the MPL (maximum permitted levels) values, prescribed by the ordinance of the Republic of Serbia. In 45 samples, Ni content was above MPL, in five samples, Pb content was above MPL, and in 2 samples, Cu content was above MPL. The Zn content in all samples was within the permitted limits. The content of TEDTPA (interpreted on the basis of classification published by Barrett at al. [38]) showed that the high values of available Mn were determined in 196 soil samples, high values of available Cu—in 216 samples, and high values of available Zn—in 21 soil samples. The obtained values of TEs in this research are comparable with the results obtained for the part of Central Serbia. The average contents of Ni, Cu, and Pb are within the limit values for the soil of Central Serbia region, while the average Zn content is slightly higher and closer to the Zn content specific to the soils of Vojvodina [40].
The results of the analyzed soil samples for models 2 and 3 for the bulk of 204 samples (Table 2) showed that 84% of the tested samples had a very acidic to acidic soil reaction for pH in 1M KCl, 79%—very acidic to slightly acidic for pH in H2O, 38% of the samples had high to extremely high SOM content, while 74% of the samples had medium and high content of clay fraction [39].
The TEAR content showed that the content of Ni was above the MPL in 30 soil samples, the content of Cu in 1 sample, while the contents of Zn and Pb were within the allowed limits. The TEDTPA content in 161 soil samples showed high values of available Mn, in 182 samples—high values of available Cu, and in 11 soil samples—high values of available Zn.

3.2. Prediction Models

By statistical analysis, linear regression models were obtained and the coefficients of determination were determined. Statistically significant regression coefficients at the 5% level are bolded (Table 3, Table 4, and Table 5). Prediction reliability is related to the coefficient of determination (R2).
Prediction model 1 for TEs Mn, Ni, Pb, Zn, and Cu is shown in Table 3. Reliability of prediction models for all TEs tested ranged from 0.25 to 0.76. The most reliable prediction model was for Cu (R2 = 0.76).
Prediction models 2 and 3 are based on the bulk of 204 soil samples for TEs Mn, Ni, Pb, Zn, and Cu, and are displayed in Table 4 and Table 5. Reliability of the prediction model based on pH in 1M KCl (model 2) for all TEs tested ranged from 0.38 to 0.85. The most reliable prediction model was for Cu (R2 = 0.85). The reliability of this model covers a larger bulk of samples (9%) in relation to the model 1.
The reliability of the prediction model based on pH in H2O (model 3) for all TEs tested ranged from 0.38 to 0.85. As with models 1 and 2, the most reliable prediction model was for Cu (R2 = 0.85), and the values are almost the same as in model 2.
The most reliable prediction model for TEDTPA Mn is model 3 (for the bulk of 204 samples and pH value in H2O). Considering all three models, it can be concluded that the difference between the coefficient of determination for the tested element is a maximum of 4%. Unlike models 2 and 3, in model 1, in addition to pH and TEAR, the content of the clay fraction also has a statistically significant effect.
The reliability of the prediction model for TEDTPA Ni in models 2 and 3 is significantly higher (13%) than in model 1. In all three models, there are statistically significant effects of the same parameters, including pH, clay, and TEAR Ni.
The prediction model reliability for TEDTPA Pb is 22% more reliable in models 2 and 3 (R2 is the same for models pH in H2O and pH in KCl) compared to model 1. Unlike model 1, in models 2 and 3, in addition to pH and TEAR, the SOM content also had a statistically significant effect.
For TEDTPA Zn, the more reliable prediction model is model 1 (6%). In all three models, there are statistically significant effects of the same parameters, including clay and TEAR Zn.
The reliability of the prediction model for TEDTPA Cu is significantly higher in models 2 and 3 (8%) than in model 1. The same parameters—pH, SOM, and TEAR Cu, have a statistically significant effect in all three models.
Certain authors have indicated that by dividing the samples based on pH, SOM type, and soil type [41], it is possible to obtain more reliable prediction models (with higher R2).
By comparing the prediction models 2 and 3 (between pH in KCl and pH in H2O) for Pb and Zn, the models have the same coefficient of reliability, while for Mn, Ni, and Cu, model 3 is slightly more accurate. Based on the obtained results, it can be ascertained that for the prediction models projection, it does not matter from which suspension the pH value of soil samples is determined. A similar conclusion was reached by Gavriloaiei [28], who emphasized a strong correlation between pH in H2O and in 1M KCl (R2 = 0.98). In the studies of Gandois et al. [42], and Zhang et al. [43], prediction models were derived based on different soil pH values (acidic, neutral, and alkaline soil reaction). Depending on the pH value, the models showed different coefficients of determination and variables that statistically significantly affect the bioavailability of TEs.

3.2.1. Manganese

In all prediction models, the pH value statistically significantly affects the bioavailability of Mn. Sungur et al. [44] confirmed the statistical significance of the pH values effect on Mn mobility by statistically processing the results of sequential analysis in order to determine the relationship between soil properties and TEs. Adriano [7] stated that pH had the greatest effect on the availability of Mn, which led to the conclusion, among other things, that Mn is generally considered to be one of the most important toxic TEs in acid soils. Clay has a statistically significant effect in prediction model 1, as confirmed in the studies of Mahmoudabadi et al. [45] and Rékási and Filep [46]. Mahmoudabadi et al. [45] obtained the statistical significance of SOM effects on Mn bioavailability, while our model did not confirm such an impact. The bioavailability of Mn is statistically significantly influenced by the pseudo total forms, as confirmed by Behera and Shukla [47].

3.2.2. Nickel

The bioavailability of Ni in all prediction models is statistically significantly influenced by pH, as confirmed by the results of other studies [24,25,45]. Clay also statistically significantly affects the bioavailability of Ni, as in many previous studies [25,45,46,48,49]. In some prediction models of Zhang et al. [43], in cases where soils were either acid or alkaline, the clay had a statistically significant effect on Ni bioavailability. SOM had no influence in any model on Ni bioavailability. Unlike our models, other studies have shown that SOM had a statistically significant effect on Ni mobility [26,42]. The pseudo total forms have a statistically significant effect on Ni bioavailability in the present study, as well as in studies of other authors [26,42,43]. According to Kabata-Pendias and Mukherjee [5], the forms of Ni in soils are diverse and range from highly mobile to ones that have no reactivity at all. The same authors reported that the behavior and phytoavailability of Ni are controlled by several soil properties, particularly clay and SOM contents, as well as soil reaction.

3.2.3. Lead

In all prediction models, the pH value has a statistically significant effect on the Pb bioavailability, which was also confirmed in previous studies [24,50,51]. In models 2 and 3, our research results show that SOM affects the bioavailability of Pb, which is consistent with the results of other authors [27,45,50,51]. Jahiruddin et al. [50], concluded that clay had a statistically significant effect on Pb mobility, which is not the case with any prediction model in our study. Pb mobility is statistically significantly influenced by the pseudo total forms, as confirmed by Vries et al. [52]. Jopony and Young [53] reported an evident significant linear relationship, which covers a very wide range of Pb total and pH values. In the results of McBride et al. [54], SOM was the only variable, which showed statistical significance on the bioavailability of Pb.

3.2.4. Zinc

In all prediction models, the clay content statistically significantly affects the bioavailability of Zn, as shown previously [24,50]. The pH value in our study does not affect the bioavailability of Zn, unlike the results obtained in studies of Buchter et al. [25], Jahiruddin et al. [50], Sauvé et al. [26], Sauvé et al. [27], Ivezić et al. [51], Mahmoudabadi et al. [45], and Rutkowska et al. [13]. SOM does not have a statistically significant effect on Zn bioavailability, as confirmed by the results of other authors [13,24,52], and unconfirmed by Jahiruddin et al. [50], Sauvé et al. [26], Dragović et al. [55], and Mahmoudabadi et al. [45], that concluded that there was an influence. Zn bioavailability is statistically significantly influenced by TEAR, as confirmed by Vries et al. [52] and Rutkowska et al. [13]. McBride et al. [54] obtained pH and total Zn dependence in their studies. The multiple regressions in previous studies [43], showed that soil pH and total soil Ni contents were the most important parameters in the prediction of the soluble Ni concentrations with the adjusted determination coefficients.

3.2.5. Copper

The bioavailability of Cu is statistically significantly influenced by pH and SOM, which is in accordance with the previous studies [24,25,50,51,52]. Unlike our prediction models cases, Jahiruddin et al. [50], Vries et al. [52], and Rékási and Filep [46], confirmed the statistical significance of clay effect on Cu mobility. Cu bioavailability is significantly influenced by pseudo total forms, as obtained by Vries et al. [52], Ivezić et al. [51], and Mondaca et al. [56]. The activity of free Cu ions in the study of Mondaca et al. [56] showed a significant relationship with the total content of Cu, pH, and SOM. The study of McBride et al. [54] also showed that the bioavailability of Cu is influenced by pH, SOM, and total Cu.
As the soils consist of heterogeneous mixtures of various organic and organic-mineral substances (minerals of clay, oxides, and hydroxides of Fe, Mn, and Al), other solid components, and a variety of soluble substances, there are the manifold binding mechanisms for TE and forms of their occurrence in soils, varying in accordance with the soil composition and physical properties [5]. There are many studies that document the significant influence of the pH value on the TEs adsorption, describing an increase in TEs solubility with a decrease in pH values [57]. By measuring the pH in water suspension, the values that more closely reflect the true pH values that prevail in natural soil conditions are obtained, while the values of pH in 1M KCl suspension provide better information concerning the chemical properties of the system [58]. The measurement of pH in water generally yields higher pH values than in 1 M KCl suspensions also [59]. According to Kim et al. [60], in soils with usual pH values (pH 3.6–8.2), TEs subjected to our study can be classified into two classes: I) class of relatively high mobility (Ni and Zn); II) class of relatively low mobility (Cu and Pb). Unlike the usual soil pH range, in acidic soils, Brümmer [61] determined the decreases in the mobility of TEs in the following order: Cd > Zn > Ni> Mn > Cu > Pb > Hg.
The level and mobility of TEs are significantly affected by the content of organic matter in soils in combination with other factors. Organic matter affects both the accumulation and release of TEs in soil [62]. Carrillo-Gonzalez et al. [3] indicated that high content of solid organic matter in soil increases the sorption of cations onto humus material (increase mobility and/or bioavailability of TEs), whereas high and soluble humus content increases the complexation for most trace cations (decrease/increase mobility and/or bioavailability of TEs). In addition, they determined that the TEs could also be binded and stripped from the solution by high-molecular-weight organic compounds, since they can be insoluble and, therefore, semi-immobile. In highly acidic soil conditions, organic matter immobilizes TEs, and in weakly acidic to alkaline soil conditions (pH 6–8), it mobilizes TEs by forming insoluble and soluble complexes [63].
As for clay, in the majority of soil types, the TEs are mainly associated with the clay fraction [62]. With respect to soil type and organic content, the clay-rich soils generally have a higher retention capacity than soils with little or no clay [64]. Many studies indicate that cation sorption on clay minerals varies depending on cation properties and clay nature [3,65,66,67]. The insignificant concentrations of TEs may be included in clay minerals as structural components, but their capacities of sorption to TEs play the most important role [68].
The pseudo total forms of TEs affect bioavailability for all tested elements and in all models. Belanović et al. [69] found that the greatest influence on available forms of TEs have the total contents of TEs in soil. In the study of Ivezić et al. [51], the total TEs concentrations were very low, and variations among sites were small, and, therefore, total TEs concentrations were not statistically significant for the regression model. Despite this anomaly, they considered that it is necessary to include the total concentrations of TEs for the prediction of mobile TEs forms in soil, which is in accordance with the study of Lončarić et al. [70].
From the derived prediction models, the models for bioavailable Cu and Pb (R2 = 0.60–0.85), which showed a high degree of reliability, are distinguished. For TEs Cu, Ni, and Zn, it can be stated that, in all models, the same soil parameters have an effect, characteristically for each element. The models for Mn, Ni, and Zn bioavailability have a coefficient of determination R2 = 0.25–0.54, indicating that the model refers to a 50% or smaller bulk of samples.

4. Conclusions

In the conducted research, three prediction models for bioavailability of trace elements (TEs) in soil were formed on samples of different physico-chemical properties, taken from different localities and with various ways of using.
Multiple linear regression showed that the prediction models for TE Cu and TE Pb were the most reliable.
The bioavailability of TEs was most influenced by the content of pseudo total forms of TEs, while the influence of other soil parameters tested (pH, soil organic matter—SOM, clay) was less significant and varied depending on models and TEs.
The inconsistency of the starting hypotheses, made in the formation of these models, in the behavior of individual soil parameters (pH, SOM and clay) on the mobility of TEs could be explained by the different mechanisms of action of the parameters tested in this study, especially SOM and clay, as explained in the paper. More reliable prediction models could be obtained by dividing the sample based on pH, SOM type, and soil type, as previously suggested. The results of the present research have shown that it would be useful to derive a prediction for sub-models by separating samples according to pH, SOM type, with soil type inclusion.
The absence of a significant difference in the values of the determination coefficient for models that include different types of suspensions in determining the pH of the soil indicates that, in future studies, it is not important to select a suspension when forming a prediction model.
The reliability of the prediction model for Cu bioavailability should be verified by applying as many soil samples as possible.
In agriculture, the application of prediction models based on the results obtained in the present research provides information for assessing the suitability of plant species cultivation at given locations, ecological risks, and the need for applying the agro-technical measures based on the bioavailability of TEs.

Author Contributions

Conceptualization, Z.D. and J.M.; methodology, R.P.; software, J.M.; validation, R.P., and A.S.-S.; formal analysis, Z.D.; investigation, Z.D. and J.M.; resources, R.P.; data curation, Z.D., J.M., R.P. and A.S.-S.; writing—original draft preparation, Z.D. and J.M.; writing—review and editing, R.P. and A.S.-S.; visualization, Z.D. and J.M.; supervision, R.P.; project administration, R.P.; funding acquisition, R.P.

Funding

Serbian Ministry of Education, Science and Technological Development, Project TR 37006 and TR 31018.

Acknowledgments

This research work was carried out with the financial support of the Ministry of Education, Science and Technological Development of the Republic of Serbia, project TR 37006 and TR 31018.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Kabata-Pendias, A. Soil—plant transfer of trace elements—An environmental issue. Geoderma 2004, 122, 143–149. [Google Scholar] [CrossRef]
  2. Podlesáková, E.; Nemecek, J.; Vácha, R. Mobility and Bioavailability of Trace Elements in Soils. In Trace Elements in Soil: Bioavailability, Flux, and Transfer; Iskandar, I.K., Kirkham, M.B., Eds.; Lewis Publishers: Boca Raton, FL, USA, 2001; pp. 21–42. ISBN 1-56670-507-X. [Google Scholar] [CrossRef]
  3. Carrillo-Gonzalez, R.; Simunek, J.; Sebastien, S.; Adriano, D. Mechanisms and pathways of trace element mobility in soils. Adv. Agron. 2006, 91, 111–179. [Google Scholar] [CrossRef]
  4. Hough, R.L. Copper and Lead. In Trace Elements in Soils; Hooda, P.S., Ed.; Wiley-Blackwell Publishing: West Sussex, UK, 2010; pp. 441–461. ISBN 978-1-405-16037-7. [Google Scholar]
  5. Kabata-Pendias, A.; Mukherjee, A.B. Trace Elements from Soil to Human; Springer-Verlag: Berlin/Heidelberg, Germany, 2007; E-Book; ISBN 978-3-540-32714-1. [Google Scholar] [CrossRef]
  6. Bogdanović, D.; Ubavić, M.; Hadžić, V. Heavy metals in soil. In Heavy Metals in Environment; Kastori, R., Ed.; Institute of Field and Vegetable Crops: Novi Sad, Serbia, 1997; pp. 97–193. ISBN 86-80735-38-8. [Google Scholar]
  7. Adriano, D.C. Trace Elements in Terrestrial Environments: Biogeochemistry, Bioavailability, and Risks of Metals, 2nd ed.; Springer-Verlag: Berlin/Heidelberg, Germany, 2001; ISBN 978-0-387-21510-5. [Google Scholar] [CrossRef]
  8. Dąbkowska-Naskręt, H.; Jaworska, H. Manganese mobility in soils under the impact of alkaline dust emission. J. Elem. 2013, 18, 371–379. [Google Scholar] [CrossRef]
  9. Kabata-Pendias, A.; Szteke, B. Trace Elements in Abiotic and Biotic Environments; CRC Press: Boca Raton, FL, USA, 2015; ISBN 9781482212792. [Google Scholar] [CrossRef]
  10. Wenzel, W.W.; Kidd, P.S.; Puschenreiter, M.; Rosenkranz, T. Nickel Biogeochemistry at the Soil–Plant Interface. In Nickel in Soils and Plants; Tsadilas, C., Rinklebe, J., Selim, M., Eds.; CRC Press: Boca Raton, FL, UAS, 2018; pp. 21–50. ISBN 9781498774604. [Google Scholar]
  11. Dalton, D.A. Essentiality of Nickel for Plants. In Nickel in Soils and Plants; Tsadilas, C., Rinklebe, J., Selim, M., Eds.; CRC Press: Boca Raton, FL, USA, 2018; pp. 1–20. ISBN 9781498774604. [Google Scholar]
  12. Schulin, R.; Johnson, A.; Frossard, E. Trace Element-Deficient Soils. In Trace Elements in Soils; Hooda, P.S., Ed.; Wiley-Blackwell Publishing: West Sussex, UK, 2010; pp. 175–194. ISBN 978-1-405-16037-7. [Google Scholar]
  13. Rutkowska, B.; Szulc, W.; Bomze, K.; Gozdowski, D.; Spychaj-Fabisiak, E. Soil factors affecting solubility and mobility of zinc in contaminated soils. Int. J. Environ. Sci. Technol. 2015, 12, 1687–1694. [Google Scholar] [CrossRef] [Green Version]
  14. Somasundaram, J.; Krishnasamy, R.; Mahimairaja, S.; Savithri, P. Dynamics of Lead (Pb) in Different Soil Conditions. J. Environ. Sci. Eng. 2006, 47, 123–128. [Google Scholar]
  15. Zan, N.R.; Datta, S.P.; Rattan, R.K.; Dwivedi, B.S.; Meena, M.C. Prediction of the solubility of zinc, copper, nickel, cadmium, and lead in metal-contaminated soils. Environ. Monit. Assess. 2013, 185, 10015–10025. [Google Scholar] [CrossRef] [PubMed]
  16. MvV. Circular on Target Values and Intervention Values for Soil Remediation. Dutch Target and Intervention Values, 2000 (the New Dutch List). Annexes. Ministerie van Volkshulsvesting, Ruimtelijke Ordeningen Milieubeheer; Ministry of Housing, Spatial Planning and the Environment: Barendrecht, The Netherlands; Available online: http://esdat.net/Environmental%20Standards/Dutch/annexS_I2000Dutch%20Environmental%20Standards.pdf (accessed on 19 September 2019).
  17. Martens, D.C.; Lindsay, W.L. Testing soils for copper, iron, manganese, and zinc. In Soil Testing and Plant Analysis, 3rd ed.; Westerman, R.L., Ed.; Soil Science Society of America, Inc.: Madison, WI, USA, 1990; pp. 229–264. ISBN 0-89118-793-6. [Google Scholar]
  18. Kashem, A.; Singh, B.R. Solid-Phase Speciation of Cd, Ni, and Zn in Contaminated and Noncontaminated Tropical Soils. In Trace Elements in Soil: Bioavailability, Flux, and Transfer; Iskandar, I.K., Kirkham, M.B., Eds.; Lewis Publishers: Boca Raton, FL, USA, 2001; pp. 213–228. ISBN 9780429137907. [Google Scholar] [CrossRef]
  19. Ivezić, V. Trace metal availability in soils under different land uses of the Danube basin in Croatia. Ph.D. Thesis, Department of Plant and Environmental Sciences Norwegian University of Life Sciences, As, Norway, October 2011. [Google Scholar]
  20. Versluijs, C.W.; Otte, P.F. Accumulation of Metals in Plants. A Contribution to the Technical Evaluation of the Intervention Values and Site-Specific Risk Assessment of Contaminated Sites; RIVM report 711701024; National Institute for Public Health and the Environment: Bilthoven, The Netherlands, 2001. [Google Scholar]
  21. Otte, P.F.; Lijzen, J.P.A.; Otte, J.G.; Swartjes, F.A.; Versluijs, C.W. Evaluation and Revision of the C Soil Parameter Set; RIVM report 711701021; National Institute of Public Health and the Environment: Bilthoven, The Netherlands, 2001. [Google Scholar]
  22. Krauss, M.; Wilcke, W.; Kobza, J.; Zech, W. Predicting heavy metal transfer from soil to plant: Potential use of Freundlich-type function. J. Plant. Nutr. Soil Sci. 2002, 165, 3–8. [Google Scholar] [CrossRef]
  23. Liang, Z.; Ding, Q.; Wei, D.; Li, J.; Chen, S.; Maa, Y. Major controlling factors and predictions for cadmium transfer from the soil into spinach plants. Ecotoxicol. Environ. Saf. 2013, 93, 180–185. [Google Scholar] [CrossRef]
  24. de Souza Braz, A.M.; Fernandes, A.R.; Ferreira, J.R.; Alleoni, L.R.F. Distribution coefficients of potentially toxic elements in soils from the eastern Amazon. Environ. Sci. Pollut. Res. 2013, 20, 7231–7242. [Google Scholar] [CrossRef]
  25. Buchter, B.; Davidoff, B.; Amacher, M.C.; Hinz, C.; Iskandar, I.K.; Selim, H.M. Correlation of Freundlich Kd and n retention parameters with soils and elements. Soil Sci. 1989, 148, 370–379. [Google Scholar] [CrossRef]
  26. Sauvé, S.; Hendershot, W.; Allen, H.E. Solid-Solution Partitioning of Metals in Contaminated Soils: Dependence on pH, Total Metal Burden, and Organic Matter. Environ. Sci. Technol. 2000, 34, 1125–1131. [Google Scholar] [CrossRef]
  27. Sauvé, S.; Manna, S.; Turmel, M.C.; Roy, A.G.; Courchesne, F. Solid−Solution Partitioning of Cd, Cu, Ni, Pb, and Zn in the Organic Horizons of a Forest Soil. Environ. Sci. Technol. 2003, 37, 5191–5196. [Google Scholar] [CrossRef] [PubMed]
  28. Gavriloaiei, T. The Influence of Electrolyte Solutions on Soil pH Measurements. Rev. Chim. 2012, 63, 396–400. [Google Scholar]
  29. Thomas, G.W. Soil pH and Soil Acidity. In Methods of Soil Analysis Part 3: Chemical Methods; SSSA Book Series 5; Sparks, D.L., Ed.; Soil Science Society of America: Madison, WI, USA, 1996; pp. 475–490. ISBN 978-0-89118-866-7. [Google Scholar]
  30. SRPS ISO 11464:2004 Soil Quality—Pretreatment of Samples for Physico-Chemical Analyses; SRPS, Institute for Standardisation of Republic of Serbia: Belgrade, Serbia, 2004.
  31. SRPS ISO 10390:2007 Soil Quality—Determination of pH; SRPS, Institute for Standardisation of Republic of Serbia: Belgrade, Serbia, 2007.
  32. SRPS ISO 10694:2005 Soil Quality—Determination of Organic and Total Carbon after Dry Combustion (Elementary Analysis); SRPS, Institute for Standardisation of Republic of Serbia: Belgrade, Serbia, 2005.
  33. Hadžić, V.; Belić, M.; Nešić, L. Determination of the mechanical composition (texturally gravimetric) of the soil. In Methods of Research and Determination of Physical Properties of the Soil; Bošnjak, Đ., Ed.; JDPZ: Novi Sad, Serbia, 1997; pp. 17–32. (In Serbian) [Google Scholar]
  34. SRPS ISO 14870:2005 Soil Quality—Extraction of Trace Elements by Buffered DTPA Solution; SRPS, Institute for Standardisation of Republic of Serbia: Belgrade, Serbia, 2005.
  35. ISO 22036:2008 Soil quality—Determination of trace elements in extracts of soil by inductively coupled plasma—atomic emission spectrometry (ICP—AES); International Organization for Standardization ISO Central Secretariat: Geneva, Switzerland, 2008.
  36. SRPS ISO 11466:2004 Soil Quality—Extraction of Trace Elements Soluble in Aqua Regia; SRPS: Institute for Standardisation of Republic of Serbia: Belgrade, Serbia, 2004.
  37. Official Gazette of Republic of Serbia. Rule Book of Allowed Concentrations of Dangerous and Hazardous Materials in Soil and in Water for Irrigation and Methods for Analysis, Official Gazette of Republic of Serbia 23; Official Gazette of Republic of Serbia: Belgrade, Serbia, 1994. (In Serbian)
  38. Barrett, W.; Jones, N.; Chu, P.; Pohlman, K.; Coleman, G.; Poole, H.; Goff, D.; Spiva, C.; Griffith, L.; Zwiep, J.; et al. Micronutrients. In Agronomy Handbook, Soil and Plant Analysis Resource Handbook; Ankerman, D., Large, R., Eds.; Midwest laboratories, INC.: Omaha, NE, USA, 2017; pp. 65–76. [Google Scholar]
  39. Hazelton, P.; Murphy, B. Interpreting Soil Test Results: What do all the Numbers Mean? 3rd ed.; CSIRO Publishing: Clayton, VIC, Australia, 2016; ISBN 9781486303960. [Google Scholar]
  40. Mrvić, V.; Zdravković, M.; Sikirić, B.; Čakmak, D.; Kostić-Kravljanac, L. Harmful and dangerous elements in soil. In Fertility and Content of Dangerous and Harmful Substances in Soils of Central Serbia; Mrvić, V., Antonović, G., Martinović, L., Eds.; Institute of soil science: Belgrade, Serbia, 2009; pp. 75–134. ISBN 978-86-911273-1-2. [Google Scholar]
  41. Moreira, C.S.; Alleoni, L.R.F. Adsorption of Cd, Cu, Ni and Zn in tropical soils under competitive and non-competitive systems. Sci. Agric. 2010, 67, 301–307. [Google Scholar] [CrossRef] [Green Version]
  42. Gandois, L.; Probst, A.; Dumat, C. Modelling trace metal extractability and solubility in French forest soils by using soil properties. Eur. J. Soil Sci. 2010, 61, 271–286. [Google Scholar] [CrossRef] [Green Version]
  43. Zhang, X.; Li, J.; Wei, D.; Li, B.; Ma, Y. Predicting Soluble Nickel in Soils Using Soil Properties and Total Nickel. PLoS ONE 2015, 10, e0133920. [Google Scholar] [CrossRef] [Green Version]
  44. Sungur, A.; Soylak, M.; Ozcan, H. Investigation of heavy metal mobility and availability by the BCR sequential extraction procedure: Relationship between soil properties and heavy metals availability. Chem. Speciat. Bioavailab. 2014, 26, 219–230. [Google Scholar] [CrossRef] [Green Version]
  45. Mahmoudabadi, E.; Sarmadian, F.; Nazary Moghaddam, R. Spatial distribution of soil heavy metals in different land uses of an industrial area of Tehran (Iran). Int. J. Environ. Sci. Technol. 2015, 12, 3283–3298. [Google Scholar] [CrossRef] [Green Version]
  46. Rékási, M.; Filep, T. Factors determining Cd, Co, Cr, Cu, Ni, Mn, Pb and Zn mobility in uncontaminated arable and forest surface soils in Hungary. Environ. Earth Sci. 2015, 74, 6805–6817. [Google Scholar] [CrossRef]
  47. Behera, S.K.; Shukla, A.K. Total and Extractable Manganese and Iron in Some Cultivated Acid Soils of India: Status, Distribution and Relationship with Some Soil Properties. Pedosphere 2014, 24, 196–208. [Google Scholar] [CrossRef]
  48. Hseu, Z.Y.; Chen, Z.S. Nickel in Serpentine Soils. In Nickel in Soils and Plants; Tsadilas, C., Rinklebe, J., Selim, M., Eds.; CRC Press, Taylor & Francis Group: Boca Raton, FL, USA, 2018; pp. 181–198. ISBN 978-1498774604. [Google Scholar] [CrossRef]
  49. Nikoli, T.; Matsi, T. Methods of Ni Determination in Soils and Plants. In Nickel in Soils and Plants; Tsadilas, C., Rinklebe, J., Selim, M., Eds.; CRC Press, Taylor & Francis Group: Boca Raton, FL, USA, 2018; pp. 199–216. ISBN 9781498774604. [Google Scholar] [CrossRef]
  50. Jahiruddin, M.; Harada, H.; Hatanaka, T.; Islam, M.R. Status of trace elements in agricultural soils of Bangladesh and relationship with soil properties. Soil Sci. Plant. Nutr. 2000, 46, 963–968. [Google Scholar] [CrossRef]
  51. Ivezić, V.; Almas, A.R.; Singh, B.R. Predicting the solubility of Cd, Cu, Pb and Zn in uncontaminated Croatian soils under different land uses by applying established regression models. Geoderma 2012, 170, 89–95. [Google Scholar] [CrossRef]
  52. de Vries, W.; Curlik, J.; Muranyi, A.; Alloway, B.; Groenenberg, B.J. Assessment of relationships between total and reactive concentrations of cadmium, copper, lead and zinc in Hungarian and Slovakian soils. Ekol. -Bratisl. 2005, 24, 152–169. [Google Scholar]
  53. Jopony, M.; Young, S.D. The solid solution equilibria of lead and cadmium in polluted soils. Eur. J. Soil Sci. 1994, 45, 59–70. [Google Scholar] [CrossRef]
  54. McBride, M.; Sauve, S.; Hendershot, W. Solubility control of Cu, Zn, Cd and Pb in contaminated soils. Eur. J. Soil Sci. 1997, 48, 337–346. [Google Scholar] [CrossRef]
  55. Dragović, S.; Mihailović, N.; Gajić, B. Heavy metals in soils: Distribution, relationship with soil characteristics and radionuclides and multivariate assessment of contamination sources. Chemosphere 2008, 72, 491–495. [Google Scholar] [CrossRef]
  56. Mondaca, P.; Neaman, A.; Sauvé, S.; Salgado, E.; Bravo, M. Solubility, partitioning, and activity of copper-contaminated soils in a semiarid region. J. Plant. Nutr. Soil Sci. 2015, 178, 452–459. [Google Scholar] [CrossRef]
  57. Blume, H.P.; Brümmer, G.W.; Fleige, H.; Horn, R.; Kandeler, E.; Kögel-Knabner, I.; Kretzschmar, R.; Stahr, K.; Wilke, B.M. Threats to the Soil Functions. In Scheffer/Schachtschabel Soil Science; Springer: Berlin/Heidelberg, Germany, 2016; pp. 485–559. ISBN1 978-3-642-30942-7. ISBN2 978-3-642-30941-0. (eBook). [Google Scholar] [CrossRef]
  58. Tan, K.H. Principles of Soil Chemistry, 4th ed.; CRC Press, Taylor & Francis Group: Boca Raton, FL, USA, 2010; ISBN 9781439894606. [Google Scholar] [CrossRef]
  59. Bloom, P.R.; Skyllberg, U. Soil pH and pH Buffering. In Handbook of Soil Sciences: Properties and Processes, 2nd ed.; Huang, P.M., Li, Y., Sumner, M.E., Eds.; CRC Press, Taylor & Francis Group: Boca Raton, FL, USA, 2012; pp. 19:1–19:14. ISBN 9781439803059. [Google Scholar] [CrossRef]
  60. Kim, R.Y.; Yoon, J.K.; Kim, T.S.; Yang, J.E.; Owens, G.; Kim, K.R. Bioavailability of heavy metals in soils: Definitions and practical implementation—A critical review. Environ. Geochem. Health 2015, 37, 1041–1061. [Google Scholar] [CrossRef]
  61. Brümmer, G.W. Heavy Metal Species, Mobility and Availability in Soils. In The Importance of Chemical “Speciation” in Environmental Processes; Bernhard, M., Brinckman, F.E., Sadler, P.J., Eds.; Springer: Berlin/Heidelberg, Germany, 1986; Volume 33, pp. 169–192. ISBN 978-3-642-70443-7. [Google Scholar] [CrossRef]
  62. Mleczek, M.; Gąsecka, M.; Kaniuczak, J.; Goliński, P.; Szostek, M.; Magdziak, Z.; Rutkowski, P.; Budzyńska, S. Dendroremediation: The Role of Trees in Phytoextraction of Trace Elements. In Phytoremediation; Ansari, A.A., Gill, S.S., Gill, R., Lanza, G.R., Newman, L., Eds.; Springer: Cham, Switzerland, 2018; pp. 269–295. ISBN 978-3-319-99650-9. [Google Scholar] [CrossRef]
  63. Brümmer, G.; Herms, U. Influence of Soil Reaction and Organic Matter on the Solubility of Heavy Metals in Soils. In Effects of Accumulation of Air Pollutants in Forest Ecosystems, Proceedings of a Workshop held at Göttingen, West Germany, 16–18 May 1983; Ulrich, B., Pankrath, J., Eds.; Springer: Heidelberg, Germany, 1983; pp. 233–243. [Google Scholar] [CrossRef]
  64. Murray, K.S.; Rogers, D.T.; Kaufman, M.M. Heavy metals in an urban watershed in Southeastern Michigan. J. Environ. Qual. 2004, 33, 163–172. [Google Scholar] [CrossRef]
  65. Abd-Elfattah, A.; Wada, K. Adsorption of lead, copper, zinc, cobalt, and cadmium by soils that differ in cation-exchange materials. J. Soil Sci. 1981, 32, 271–283. [Google Scholar] [CrossRef]
  66. Tiller, K.G.; Gerth, J.; Brümmer, G. The relative affinities of Cd, Ni and Zn for different soil clay fractions and goethite. Geoderma 1984, 34, 17–35. [Google Scholar] [CrossRef]
  67. Helios Rybicka, E.; Calmano, W.; Breeger, A. Heavy metals sorption/desorption on competing clay minerals; an experimental study. Appl. Clay Sci. 1995, 9, 369–381. [Google Scholar] [CrossRef]
  68. Kabata-Pendias, A. Trace Elements in Soils and Plants, 4th ed.; CRC Press, Taylor & Francis Group: Boca Raton, FL, USA, 2010; ISBN 9781420093681. [Google Scholar] [CrossRef]
  69. Belanović, S.; Čakmak, D.; Kadović, R.; Beloica, J.; Perović, V.; Alnaass, N.; Saljnikov, E. Availability of some trace elements (Pb, Cd, Cu and Zn) in relation to the properties of pasture soils in Stara Planina Mountain. Bull. Fac. For. 2012, 106, 41–56. [Google Scholar] [CrossRef]
  70. Lončarić, Z.; Popović, B.; Karalić, K.; Rékási, M.; Kovačević, V. Regression model for prediction availability of essential heavy metals in soils. In Proceedings of the 19th World Congress of Soil Science, Soil Solutions for a Changing World, Brisbane, Australia, 1–6 August 2010; Gilkes, R.J., Prakongkep, N., Eds.; IUSS (International Union of Soil Sciences): Brisbane, Australia, 2010; pp. 92–95, ISBN 978-0-646-53783-2. [Google Scholar]
Agronomy 09 00856 g001 550
Figure 1. Soil sampling locations in the Republic of Serbia.
Figure 1. Soil sampling locations in the Republic of Serbia.
Agronomy 09 00856 g001
Table
Table 1. Descriptive statistics of the physico-chemical properties of the soil samples tested: Minimum, Maximum, Mean, Median, and N (number of soil samples tested) for model 1.
Table 1. Descriptive statistics of the physico-chemical properties of the soil samples tested: Minimum, Maximum, Mean, Median, and N (number of soil samples tested) for model 1.
ParametersMinimumMaximumMeanMedianNMPL 1/ High Values 2
pH in 1M KCl3.607.555.264.90240
SOM %1.118.743.162.88240
Clay %5.6056.427.728.2240
Aqua regia—Extractable TE (mg kg−1)
Mn2011670612587240/
Ni14.019440.431.524050
Pb4.3512823.214.7240100
Zn17.716061.253.5240300
Cu5.5212022.618.4240100
DTPA—Extractable TE (mg kg−1)
Mn1.1221.930.628.024013
Ni0.3917.62.031.82240/
Pb1.4445.52.441.75240/
Zn0.4060.21.710.882403
Cu2.9238.53.192.372401.3
1 MPL for TEAR, according to the Official Gazette of Republic of Serbia [37]; 2 High values for TEDTPA, according to Barrett at al. [38].
Table
Table 2. Descriptive statistics of the physico-chemical properties of the soil samples tested: Minimum, Maximum, Mean, Median, and N (number of soil samples tested) for models 2 and 3.
Table 2. Descriptive statistics of the physico-chemical properties of the soil samples tested: Minimum, Maximum, Mean, Median, and N (number of soil samples tested) for models 2 and 3.
ParametersMinimumMaximumMeanMedianNMPL 1/ High Values 2
pH in H2O4.508.306.155.80204
pH in 1M KCl3.607.555.154.80204
SOM %1.388.343.062.81204
Clay %6.2056.429.029.5204
Aqua regia-Extractable TE (mg kg−1)
Mn2011670604580204/
Ni14.010236.630.920450
Pb4.3592.617.713.6204100
Zn17.713656.850.7204300
Cu5.5210021.216.6204100
DTPA-Extractable TE (mg kg−1)
Mn5.6023229.727.620413
Ni0.116.341.891.81204/
Pb0.3311.02.131.68204/
Zn0.2410.61.110.802043
Cu0.4330.23.142.282041.3
1 MPL for TEAR, according to the Official Gazette of Republic of Serbia [37]; 2 High values for TEDTPA, according to Barrett at al. [38].
Table
Table 3. Prediction model 1.
Table 3. Prediction model 1.
Estimated Regression Model for TEDTPAR2N
log(TEDTPA Mn) = −1.38 − 0.09log(pH) + 0.10log(SOM) − 0.27log(Clay) + 1.30log(TEAR Mn)0.50240
log(TEDTPA Ni) = −0.73 − 0.11log(pH) + 0.09log(SOM) + 0.21log(Clay) + 0.75log(TEAR Ni)0.25240
log(TEDTPA Pb) = −0.66 − 0.03log(pH) + 0.05log(SOM) + 0.09log(Clay) + 0.76log(TEAR Pb)0.60240
log(TEDTPA Zn) = −1.77 + 0.02log(pH) + 0.29log(SOM) − 0.70log(Clay) + 1.42log(TEAR Zn)0.47240
log(TEDTPA Cu) = −0.67 − 0.05log(pH) − 0.24log(SOM) − 0.03log(Clay) + 1.16log(TEAR Cu)0.76240
Table
Table 4. Prediction model 2.
Table 4. Prediction model 2.
Estimated Regression Model for TEDTPAR2N
log(TEDTPA Mn) = −1.42 − 0.09log(pH) − 0.09log(SOM) − 0.06log(Clay) + 1.23log(TEAR Mn)0.54204
log(TEDTPA Ni) = −1.23 − 0.07log(pH) − 0.28log(SOM) + 0.89log(Clay) + 0.41log(TEAR Ni)0.38204
log(TEDTPA Pb) = −0.85 − 0.03log(pH) + 0.17log(SOM) − 0.043og(Clay) + 1.02log(TEAR Pb)0.83204
log(TEDTPA Zn) = −1.80 + 0.01og(pH) + 0.12log(SOM) − 0.88log(Clay) + 1.66log(TEAR Zn)0.41204
log(TEDTPA Cu) = −0.78 − 0.04log(pH) − 0.32log(SOM) − 0.04log(Clay) + 1.26log(TEAR Cu)0.85204
Table
Table 5. Prediction model 3.
Table 5. Prediction model 3.
Estimated Regression Model for TEDTPAR2N
log(TEDTPA Mn) = −1.35 − 0.10log(pH) − 0.13log(SOM) − 0.03log(Clay) + 1.24log(TEAR Mn)0.54204
log(TEDTPA Ni) = −1.16 − 0.08log(pH) − 0.31log(SOM) + 0.91log(Clay) + 0.43log(TEAR Ni)0.38204
log(TEDTPA Pb) = −0.83 − 0.03log(pH) + 0.15log(SOM) − 0.03log(Clay) + 1.03log(TEAR Pb)0.83204
log(TEDTPA Zn) = −1.80 + 0.012og(pH) + 0.13log(SOM) − 0.89log(Clay) + 1.66log(TEAR Zn)0.41204
log(TEDTPA Cu) = −0.74 − 0.05log(pH) − 0.34log(SOM) − 0.02log(Clay) + 1.26log(TEAR Cu)0.85204

Share and Cite

MDPI and ACS Style

Dinić, Z.; Maksimović, J.; Stanojković-Sebić, A.; Pivić, R. Prediction Models for Bioavailability of Mn, Cu, Zn, Ni and Pb in Soils of Republic of Serbia. Agronomy 2019, 9, 856. https://doi.org/10.3390/agronomy9120856

AMA Style

Dinić Z, Maksimović J, Stanojković-Sebić A, Pivić R. Prediction Models for Bioavailability of Mn, Cu, Zn, Ni and Pb in Soils of Republic of Serbia. Agronomy. 2019; 9(12):856. https://doi.org/10.3390/agronomy9120856

Chicago/Turabian Style

Dinić, Zoran, Jelena Maksimović, Aleksandra Stanojković-Sebić, and Radmila Pivić. 2019. "Prediction Models for Bioavailability of Mn, Cu, Zn, Ni and Pb in Soils of Republic of Serbia" Agronomy 9, no. 12: 856. https://doi.org/10.3390/agronomy9120856

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop