JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 11 May 2005
Published in J Environ Qual 34:1026-1035 (2005)
DOI: 10.2134/jeq2003.0238
© 2005 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (9)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ingwersen, J.
Right arrow Articles by Streck, T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ingwersen, J.
Right arrow Articles by Streck, T.
Agricola
Right arrow Articles by Ingwersen, J.
Right arrow Articles by Streck, T.
Related Collections
Right arrow Plant and Soil Interactions
Right arrow Municipal Wastes
Right arrow Heavy Metals
Right arrow Other Models
Right arrow Soil Pollution

TECHNICAL REPORTS

Heavy Metals in the Environment

A Regional-Scale Study on the Crop Uptake of Cadmium from Sandy Soils

Measurement and Modeling

Joachim Ingwersen* and Thilo Streck

University of Hohenheim, Institute of Soil Science and Land Evaluation, Biogeophysics Section, D-70593 Stuttgart, Germany

* Corresponding author (jingwer{at}uni-hohenheim.de)

Received for publication June 18, 2004.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Plant uptake is one of the major pathways by which cadmium (Cd) in soils enters the human food chain. This study was conducted to investigate the uptake of Cd by crops from soils within the wastewater irrigation area (WIA) of Braunschweig (Germany) and to develop a simple process-oriented model that is suited to predict Cd uptake at the regional scale. The sandy soils within the WIA (4300 ha) have received considerable loads of heavy metals by irrigation using municipal wastewater for up to 40 years. In 1998 and 1999, we sampled soil and plant material at 40 potato (Solanum tuberosum L.), 40 sugar beet (Beta vulgaris L.), and 32 winter wheat (Triticum aestivum L.) fields. In both years and for all three crops, we found close linear relationships between the Cd content of plant material and the Cd concentration in soil solution. For all three crops, we observed a trend of relatively increased Cd uptake in the year with the higher saturation deficit of the atmosphere. We interpret this to indicate that transpiration plays an important role in the Cd uptake of crops under the conditions of the WIA. In modeling the uptake of Cd by crops, we assume that uptake is proportional to mass flow, that is, the product of water transpired, Cd concentration in soil solution, and a plant-specific empirical parameter. The simulations agreed well with the observed Cd contents in crops. Our model explained between 66 and 87% of the observed variance.

Abbreviations: WIA, wastewater irrigation area


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
SINCE THE industrial revolution, the Cd input in soils has continuously increased (Jones et al., 1987). Cadmium enters agricultural soils mainly via atmospheric deposition, fertilization, as well as the application of sewage sludge, wastewater, or compost. Cadmium is one of the most mobile and bioavailable heavy metals in soil and may cause human and ecotoxicological impacts even at low concentrations. With regard to human intake, the most important pathway is food. The uptake of soil Cd by crops is mediated by both soil and plant factors. Upon entering the soil, Cd is mainly sorbed on the surface of organic compounds, clays, and iron oxides. Only a minor fraction remains in solution. This fraction, however, is one of the most important key variables in controlling bioavailability of Cd (Grant et al., 1998). Only dissolved Cd can be transferred to plants. The relationship between the sorbed and dissolved phase depends strongly on soil properties such as pH, organic carbon content, and clay content (Bergkvist and Jarvis, 2004; Krishnamurti and Naidu, 2003; Streck and Richter, 1997).

The uptake of Cd and its distribution in crops differs among species and among cultivars within a species. Petterson (1977), for example, observed that the Cd content of plants grown in solution culture increased in the order: oats, wheat < bean, pea, sunflower, cucumber < corn, mustard < radish, kale, rape < tomato, carrot, sorre < lettuce. Variation of Cd uptake patterns among cultivars within a species has been reported in potato (McLaughlin et al., 1994b), wheat (Chaudri et al., 2001; Oliver et al., 1995), durum wheat (Tahvonen and Kumpulainen, 1993), maize (Florijn and van Beusichem, 1993), and oat, carrots, and spinach (He and Singh, 1994).

Cadmium is transported from soil to plant roots by mass flow, diffusion, and interception. Mass flow and diffusion are considered to be the most important supply mechanism for ions in soil (Marschner, 1995). Mass flow means that ions dissolved in soil solution are transported to the roots with the transpiration flux. If the uptake by roots is greater than the supply by mass flow, the ion concentration will decrease at the root surface and diffusion will become increasingly important. For macronutrients such as potassium or phosphate, plants are known to be able to deplete the direct vicinity of their roots (Hendriks et al., 1981). The relative importance of mass flow and diffusion for the supply of plant roots varies for different species, soils, and solutes (Olsen and Kemper, 1968).

To predict the uptake of solutes by plants several deterministic models have been developed during the last decades. These models differ in their conceptual approach and degree of complexity. Following Addiscott and Wagenet (1985), deterministic models can be grouped in mechanistic and functional models. While mechanistic model incorporate the most fundamental mechanisms of processes, functional models treat root solute uptake in a simplified manner. A good example of a mechanistic model is the Barber–Cushman model (Barber and Cushman, 1981). The model simulates the radial flow of nutrients from soil to plant roots taking into account diffusive and convective flow. In combination with active solute uptake described by Michaelis–Menten kinetics, nutrient-depletion zones can be modeled around the root. However, the assumption this model makes about the mechanism of the solute uptake by the root differs fundamentally from other approaches, for example, those given in Dalton et al. (1975) or Nobel (1999). In the Barber–Cushman model, solute uptake by roots is an exclusively active process, whereby in the two citations mentioned above the solute flux across a membrane is assumed to be composed of a diffusive, a convective, and an additional active component. These authors consider root solute uptake to be coupled with the root water uptake. Accordingly, the root solute uptake may depend on the water uptake rate even when active uptake is dominant. In the Barber–Cushman model root solute uptake is only indirectly affected by water uptake.

Functional models do not explicitly consider the underlying mechanisms governing root solute uptake. Processes are considered from a more macroscopic view, and the process model is typically simpler and uses effective parameters. Simplifications are accepted for the benefit of less model data input. This point becomes particularly important in moving from the lab to the field or to even larger spatial scales. In general, the larger the scale the less data are available. A widely used functional model approach for simulating the root solute uptake assumes that solute uptake is linearly proportional to the product of soil solution concentration and water uptake (Christensen and Tjell, 1984; Behrendt et al., 1995; Trapp, 2000; Schoups and Hopmans, 2002). Some of these functional models have been successfully tested against data from laboratory and outdoor lysimeter experiments, but none has yet been thoroughly tested against field data.

The purpose of the present study was to (i) investigate the Cd uptake of potato, sugar beet, and winter wheat from sandy soils that have received considerable loads of heavy metals due to irrigation with municipal wastewater for up to 40 years under field conditions, (ii) identify the factors controlling the plant uptake of Cd, (iii) develop a functional model suitable for predicting the Cd uptake by crops at the regional scale, and (iv) test the model against an extensive field data set.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Investigation Area
The wastewater irrigation area (WIA) of Braunschweig (Germany) is located about 10 km northwest of the city center. It is the largest WIA in Germany and even in Europe. The WIA covers about 43 km2 today, whereby 28.7 km2 of this area are in agricultural use. The main crops are sugar beet, potato, winter cereal, and summer cereal. The remaining 14.3 km2 are mainly forests and villages. The sandy soils of the WIA have received considerable loads of heavy metals due to irrigation with municipal wastewater for up to 40 years. In the early 1980s, the annual Cd load to the WIA reached maximum values of about 200 g ha–1. Today annual Cd loads declined to values below 5 g ha–1. The dominant soil type is a Dystric Cambisol (Food and Agriculture Organization, 1990) or Typic Haplumbrept (Soil Survey Staff, 1994). In areas with a high ground water table Cambic Gleysols (Soil Taxonomy: Typic Haplaquept) are also present. Toward the north, the portion of Humic Cambisols (Soil Taxonomy: Typic Hapludoll) increases. The average pH and organic carbon content of the topsoils are 5.8 (in 0.01 M CaCl2) and 1.0% by weight, respectively. The total Cd content (extractable in aqua regia) averages 0.36 mg kg–1. For more information on the WIA see Ingwersen (2001).

Sampling
From 1998 to 1999, we took soil and plant samples from 40 potato, 40 sugar beet, and 32 winter wheat fields. The fields were randomly selected. From potato and sugar beet fields a single plant was collected. In winter wheat fields a plot of 0.25 m2 (0.5 x 0.5 m) was sampled at each site. At each site we took three soil monoliths with 0.07-m diameter and 1.0-m length using a mechanical soil auger (Cobra; Atlas Copco, Essen, Germany). Each monolith was divided into eight soil layers (0–0.3, 0.3–0.4, ..., 0.9–1.0 m) and from the center of each layer a sample was taken. The large diameter of the probe made a spreading-free sampling possible. Soil samples from identical layers were mixed to a homogeneous sample. The samples were stored in plastic cups, dried at 40°C, and passed through a 2-mm sieve.

The potato plant was divided into straw (stem and leaves) and tuber, the sugar beet was split up into leaf and hypocotyl, and the wheat plant material was divided into straw and grain. The plant material was cleaned with distilled water. Particularly the hollow potato stems were carefully cleaned. Each stem was cut open and checked for the absence of soil material. The potato tubers were peeled and the peel was further prepared separately. Before drying the plant material at 60°C the total fresh mass was determined. The hypocotyls of sugar beet plants were dried by freezing. The dry mass was determined and plant material was first minced with a cutting mill and afterward ground with a mortar grinder. The wheat material was cleaned with distilled water. Fresh and dry mass were determined and the material was milled in two stages to a fine, "floury" consistency. The samples were stored in plastic cups. The cups were kept inside an exsiccator until analysis.

No winter wheat plants were sampled in 1998. Fortunately, the Braunschweig Municipal Waste Water Association (BMWA) gave us the data of their long-term control fields (each field covers between 5 and 10 ha). In 1998, they sampled winter wheat grain at 13 control fields along a plot-specific sampling path. The wheat grain was analyzed for total Cd content by the laboratory of the BMWA. Because the BMWA did not take soil samples we resampled each control field in September 1998. Along each sampling path six soil monoliths were drilled as described above. Additionally, soil samples were taken from the Ap horizon every 25 m (about 12 Ap horizon samples per field) and were mixed to a homogeneous soil sample.

Measurements
All soil samples were analyzed for pH, organic C content, and Cd and Zn concentration in 0.0025 M CaCl2. For details see Streck and Richter (1997). We assume that under the conditions of the WIA, a 24-h equilibration of soil with 0.0025 M CaCl2 is suited to determine the concentration in soil solution to a good approximation (Streck, 1993; Streck and Richter, 1997). The ionic strength of the background electrolyte corresponds to the average ionic strength of the irrigation water and chloride and calcium are the main anion and cation, respectively, of the wastewater.

The 0.0025 M CaCl2 extracts were analyzed for Cd using a graphite furnace atomic absorption spectrometer with deuterium background compensation (AAS 4001; PerkinElmer, Wellesley, MA). Two modifiers were used: (i) a mixture of palladium(II) chloride, 1 M HNO3, and butanol-(1) with a volume ratio of 10:10:1 and (ii) 0.015 M Mg(NO3)2·6H2O. The average detection limit of this method was 0.25 µg Cd L–1. For the analysis of Zn we used the flame atomic absorption system of the AAS 4001.

The total Cd content of plant material was determined using a microwave-assisted extraction. About 0.5 g of dry matter received 7 mL of 65% HNO3 and 2 mL of 35% H2O2 in sealed vessels for microwave digestion (MDS 2000; CEM, Matthews, NC). Afterward samples were cooled, transferred into 25-mL flasks and filled to the mark by adding 65% HNO3. The Cd concentration in the digest was measured using graphite furnace atomic absorption spectrometry as described above. For quality control, some samples of the first campaign were reanalyzed in 1999 and were seen to be in agreement.

Modeling
Cadmium is supplied to the root system by interception, mass flow, and diffusion. If we assume that the dominant process is mass flow, then the uptake rate {phi} (kg m–3 yr–1) of a plant may be approximated as:

[1]
where qR (yr–1), expressed as volume of water extracted per volume of soil and year, is the root water uptake and C (kg m–3) denotes the concentration in soil solution. Because Cd dissolved in the transpiration stream cannot pass the Casparian strip, it must cross a biological membrane before entering the symplast. As a result, the composition of the transpiration stream may change. This is accounted for by introducing a crop-specific and/or cultivar-specific empirical parameter {eta}, termed the transpiration stream concentration factor (Briggs et al., 1982; Trapp, 2000) or plant factor (Christensen and Tjell, 1984). At {eta} = 1, solute uptake is solely determined by the concentration in soil solution and water uptake ("passive"). A value of {eta} > 1 corresponds to active uptake. Values for {eta} may be estimated from an expression derived by Dalton et al. (1975), which accounts for diffusive, convective, and active uptake mechanisms.

Under field conditions the root water uptake is not constant with depth. Provided that the root water uptake is proportional to the root length density and that the latter decreases exponentially with soil depth the root water uptake may be expressed as:

[2]
where {omega} (m–1) is an empirical parameter that describes the distribution of root length density with depth, z (m) is soil depth, and TP denotes the transpiration rate (m yr–1). Equation [2] ensures that:

[3]

Consequently, the uptake rate {phi}(z) (kg m–2 s–1) as a function of depth may be written as:

[4]

If we divide the soil into n compartments of thickness {Delta}z (m), then the Cd uptake rate {phi}i of the ith compartment is defined as:

[5]

Integration of Eq. [5] yields:

[6]
where Ci (kg m–3) denotes the solution-phase concentration in the ith compartment.

The average Cd content of total dry matter, (kg kg–1), is the quotient between total Cd uptake and total dry matter production Y (kg m–2 yr–1). Therefore, it can be calculated by:

[7]

Following Bierhuizen and Slatyer (1965) the relationship between transpiration and dry matter production is given by:

[8]
where {Delta}e (Pa) is the average saturation deficit of the atmosphere during the main vegetation period, kP (Pa) is a crop-specific constant, which is generally denoted as k factor in plant-physiological studies, while {rho}w (kg m–3) is the density of water. Insertion of Eq. [8] in Eq. [7] yields:

[9]

The translocation inside the plant with respect to unprocessed (e.g., straw) and processed (e.g., grain) plant parts is modeled by means of empirical crop-specific parameters:

[10]
and:

[11]
where QHM denotes the Cd content ratio and QY the dry matter yield ratio between unprocessed and processed plant parts. The terms CU and CP stand for the Cd content of the unprocessed and processed plant parts, respectively. The terms YU and YP are the respective dry matter yields. Because the average Cd content of dry matter can be expressed as:

[12]
the Cd content of the processed plant part (e.g., wheat grain) becomes:

[13]

Combining Eq. [9] and Eq. [13] leads to:

[14]

Competitive effects on the uptake of Cd, for example by H+ and Zn2+, may be considered by extending Eq. [14] to:

[15]
whereby X denotes the concentration (mol L–1) of the competing ion and KX (L mol–1) and VX (1) are empirical parameters.

Model Parameterization
Hourly values of the saturation deficit {Delta}eji (Pa) of the atmosphere at the ith hour of the jth day of a year were calculated as follows (Maidment, 1993):

[16]
where Tji (°C) denotes the air temperature and rHji (%) the relative humidity at the ith hour of the jth day of a year. The meteorological data were taken from the weather station Braunschweig-Völkenrode, which is run by the German Meteorological Office. The station is located about 5 km south of the WIA. The station provided air temperature T (°C) and relative humidity rH (%) data in hourly resolution. The mean saturation deficit during the main vegetation period ({Delta}e) was obtained by averaging hourly values for the period between 0630 and 2030 h during the main vegetation period (Ehlers, 1989). For winter wheat, {Delta}e was taken as the average value between 1 April and 31 July, for sugar beet between 15 April and 30 September, and for potato between 15 April and 15 August.

For potato and winter wheat kP factors were taken from a study by Ehlers (1996)(Table 11.6), who gives 4.0 Pa for winter wheat and 6.2 Pa for potato. For sugar beet, no value was found in the literature. Therefore, kP had to be fitted to measured data and was estimated as 5.5 Pa (see below).

Using the root distribution function introduced above (Eq. [2]), the fraction of plant roots within an Ap horizon with a thickness of zAp is:

[17]

Integration of Eq. [17] and rearrangement results in:

[18]

This equation was used to calculate the empirical root distribution parameter {omega} from literature data of R(zAp). In sandy soils R(zAp) was found to be 0.79 for winter wheat (Meuser et al., 1985), 0.93 for sugar beet (Kücke and Löffler, 1989), and 0.97 for potato (Schmidtke et al., 1999). These values lead to {omega} = 5.202 m–1 for winter wheat, {omega} = 8.864 m–1 for sugar beet, and {omega} = 11.689 m–1 for potato. All other model parameters (QHM, QY, and {eta}) and the state variable Ci were obtained or derived from field measurements (see below).

Statistics and the Model Assessment Criterion
Whether the slopes of two linear regressions were significantly different was tested by using the two sample Student's t test. The procedure was performed at the {alpha} = 0.05 significance level. The slope value of each regression is given with its standard error. The parameters of nonlinear equations were fitted by applying the nonlinear regression procedure implemented in SPSS Version 6.2.1 (SPSS, 1999). For grouping data in clusters we used a two-step cluster analysis using the Log-Likelihood distance measure. The model with the "best" number of clusters was selected with the help of the Schwarzsches Bayes information criteria (SPSS Version 12.0.1; SPSS, 2003).

For model evaluation we used the modeling efficiency EF (Loague and Green, 1991). The modeling efficiency EF is defined as the proportion of the total variance of observed data explained by the model:

[19]
where Pi and Oi denote predicted and observed values, respectively. The term n is the number of samples and stands for the mean of the observed data. Applying Eq. [19] is more appropriate than regressing modeled values on observed ones because a good modeling performance requires that observed and predicted data are identical rather than simply linearly related. Note, however, that the meaning of EF is similar to that of the coefficient of determination, R2. The EF can be seen as the R2 for a regression line with a slope of unity and an intercept of zero. Thus, the EF is a measure of the extent to which predicted values approach a corresponding set of measured observations. If measured and predicted data are equal EF is unity. If EF is less than zero, then the predicted values are less accurate than the observed mean.


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Observed Cadmium Uptake
Table 1 summarizes the Cd analysis of the crop material, important soil properties, and weather conditions. For all three crops the QHM value was larger than unity, that is, the Cd contents of unprocessed plant parts were higher than those of the processed parts. However, the translocation pattern of Cd to the different plant organs differed widely between the three species. The QHM value of potato was up to 10 times higher than that of wheat. Highest Cd outputs were calculated for sugar beet and potato (only in 1999). For both crops the higher Cd output compared with wheat reflects the larger dry mass yield and is not caused by a higher Cd accumulation. In 1999, because of higher temperatures, the saturation deficit of the atmosphere was markedly higher than in 1998. For sugar beet, for example, this deficit was 46% higher in 1999 than in 1998. Consequently, in 1999 the crops transpired more water to produce a given quantity of dry mass than in 1998. The average Cd concentration in soil solution varied between investigated fields and years from 1.8 to 4.0 µg L–1.


View this table:
[in this window]
[in a new window]
  		Table 1. Results of the Cd analysis of the crop material and overview of important topsoil properties and atmospheric conditions during the years 1998 and 1999.

 
In both years, we observed a significant linear relationship between the Cd solution-phase concentration and Cd content of wheat grain (R2 = 0.84 in 1998 and R2 = 0.70 in 1999; both significant at the 0.001 probability level) (Fig. 1A) . In the year with the higher saturation deficit (1999) the slope of regression was steeper (0.115 ± 0.018 m3 kg–1 versus 0.082 ± 0.011 m3 kg–1), but the difference was statistically not significant.



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 1. Relationship between topsoil Cd concentration in soil solution and plant Cd contents in 1998 and 1999 for (A) winter wheat, (B) sugar beet, and (C) potato. For winter wheat, ordinate values give the Cd content in grain. For sugar beet and potato, ordinate values represent the Cd content of the whole plant.

 
The situation was similar in sugar beet. Here again, we found significant linear relationships between solution-phase concentration and Cd content of plant (R2 = 0.84 in 1998 and R2 = 0.65 in 1999; both significant at the 0.001 probability level) (Fig. 1B). In 1999, the average saturation deficit during the main vegetation period was 46% higher than in 1998. As in winter wheat, we observed a steeper slope of regression (0.146 ± 0.025 m3 kg–1 versus 0.086 ± 0.009 m3 kg–1) in the year with the higher saturation deficit. Moreover, the two regression slopes were statistically different at {alpha} = 0.05. Note, however, that the one extreme value in 1998 distinctly affects the result of the t test. Although the slope of regression was steeper in 1999, the year with the higher saturation deficit of the air, the average Cd content in crops was not higher than in 1998 (Table 1). This is because in 1999, the Cd concentration in soil solution of sampled soils was, on average, lower than in 1998.

Also in potato, the Cd content was closely related to the Cd solution-phase concentration (R2 = 0.49 in 1998 and R2 = 0.67 in 1999; both significant at the 0.001 probability level) (Fig. 1C). In 1999, the slope of the linear regression model was 118% higher than in 1998 (0.181 ± 0.03 m3 kg–1 versus 0.083 ± 0.02 m3 kg–1). This large difference, which is statistically significant at {alpha} = 0.05, may partly be due to the 33% higher saturation deficit in 1999 (917 Pa in 1999 versus 692 Pa in 1998). This climatic effect, however, cannot fully explain the much steeper slope of regression in 1999. Additional factors probably affected the Cd uptake. A simple explanation would be that we investigated different cultivars of potatoes with different Cd uptake mechanism. The uptake behavior is known to differ between potato cultivars. For example, McLaughlin et al. (1994b) reported that tuber Cd concentration between two cultivars differed by up to a factor of two.

Simulated Cadmium Uptake
Table 2 summarizes all parameters used for modeling the crop uptake. Figure 2A shows the scatterplot of measured versus predicted Cd contents of winter wheat grain. The prediction assumes a passive uptake ({eta} = 1) and is based on parameters measured or taken from the literature. In view of the fact that no parameters were fitted, the agreement between measured and predicted values is remarkably good (EF = 68%). The model overestimated the observed Cd contents by 12% on average. By fitting {eta} to the data, the EF could be slightly increased to 71%, but at {alpha} = 0.05 the fitted value ({eta} = 0.94 ± 0.06) was not different from unity.


View this table:
[in this window]
[in a new window]
  		Table 2. Parameterization of the plant uptake model for winter wheat, sugar beet, and potato. For parameters QHM and QY, see Table 1.

 


View larger version (16K):
[in this window]
[in a new window]
  		Fig. 2. Measured versus predicted Cd contents of (A) winter wheat, (B) sugar beet, and (C) potato. Predictions assume a passive uptake and are based on parameters either measured or taken from the literature; only sugar beet kP was fitted. In each panel the modeling efficiency (EF) is given. The black open circle in (C) points to four potato samples with a particularly strong disagreement between model and measurement.

 
In the literature, no unambiguous value could be found for the kP factor of sugar beet. In the present study kP was therefore estimated from field observations to be 5.5 ± 0.3 Pa. The scatterplot of measured versus predicted Cd contents of sugar beet plants (Fig. 2B) shows acceptable spreading around the bisector (EF = 66%). Note, however, that it is not possible to estimate both kP and {eta} simultaneously. An identical EF could be achieved by fixing kP to any other value and fitting {eta}. Thus, the fitted kP should be interpreted with caution. The high Cd content of about 1.5 mg kg–1 strongly affects the result. When this value was ignored the fitted kP was 4.9 ± 0.3 Pa and the EF decreased to 48%.

Figure 2C shows a scatterplot of measured versus predicted Cd contents of potato plants. For this prediction no parameter was fitted and a passive uptake was assumed. Some predicted values agree fairly well with the measured data. However, the model systematically underestimates measured data and, overall, the agreement between measured and independently predicted Cd contents is unsatisfactory (EF = 0%). The disagreement between model and measurement was particularly strong at the four sites surrounded by the black line. While predicted values were between 0.26 and 0.32 mg kg–1 measured values ranged from 0.82 to 1.12 mg kg–1. Because of these unsatisfying modeling results we investigated whether it is possible to group the data according to their residuum between predicted and measured value. Using a two-step cluster analysis, three clusters (in the following alpha [N = 13], beta [N = 23], and gamma [N = 4]) were identified (Fig. 3) . For each of the clusters a value for {eta} was fitted, which was found to be 0.91 ± 0.07, 1.65 ± 0.07, and 3.22 ± 0.2 for alpha, beta, and gamma, respectively. Using this procedure distinctly increased the model efficiency from zero to 87% (Fig. 4) . Two possible explanations for the observed clusters are that we investigated different potato cultivars with different Cd uptake mechanisms (McLaughlin et al., 1994b) or that sites differed in some soil properties that govern plant uptake but were not taken into account by the model. The chlorid concentration could be such a soil property (McLaughlin et al., 1994a).



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 3. Measured versus predicted Cd contents of potato plants. With the help of a two-step cluster analysis the data pairs were grouped in three clusters. The cluster criterion was the residuum between predicted and measured value.

 


View larger version (20K):
[in this window]
[in a new window]
  		Fig. 4. Measured versus predicted Cd contents of potato plants (modeling efficiency [EF] = 87%). Data were grouped in three clusters. For each cluster a specific plant factor {eta} was fitted.

 
For all three crops the influence of Zn and proton competition on Cd plant uptake was tested by fitting of KX and VX to the data (Eq. [15]). Zinc concentrations and pH values of investigated soils are given in Table 1. None of the fitted KX values were significantly different from zero at {alpha} = 0.05, which indicates that competitive effects of these two cations were negligible under the conditions of our investigation.

Sensitivity Analysis
Figure 5 shows the influence of different model parameters on the Cd content of processed plant part CP (e.g., wheat grain). The reference simulation was done using the following input values: C(z = 0–30 cm) = 1 µg L–1, C(z > 30 cm) = 0 µg L–1, kP = 4 Pa, {Delta}e = 700 Pa, QHM = 10, QY = 0.5, {eta} = 1, and {omega} = 8 m–1. All parameters were changed one by one from 50 to 200% of their initial values. Because CP is a linear function of C, {eta}, and {Delta}e, the sensitivity of CP to changes in C, {eta}, and {Delta}e is the same for all three parameters. For example, if the saturation deficit of the atmosphere increases, more water has to be transpired to produce a given quantity of biomass and therefore CP increases. The influence of kP is the other way round. As kP increases the water efficiency of the plant increases, that is, the plant needs less water to produce a given quantity of biomass and therefore CP decreases. Increased QHM and QY values result in decreased CP because more Cd is translocated to the unprocessed plant parts (e.g., straw). Cadmium content of processed plant part CP has the lowest sensitivity to the root distribution parameter {omega}. Increasing {omega} means that less roots are located in subsoil (>30 cm). Because the analysis assumes that, in the subsoil, the solution-phase concentration C is equal to zero, CP increases with increasing {omega}.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 5. Sensitivity analysis of model input parameters. The reference simulation was done using the following parameters: C(z = 0–30 cm) = 1 µg L–1, C(z > 30 cm) = 0 µg L–1, kP = 4 Pa, {Delta}e = 700 Pa, QHM = 10, QY = 0.5, {eta} = 1, and {omega} = 8 m–1.

 

   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
The present study was not designed to explore the Cd uptake mechanism of plants and therefore provides only a rather coarse picture on how the uptake may function. However, the observed trend of relatively increased Cd uptake in the year with the higher saturation deficit of the atmosphere suggests that transpiration affects the uptake. Whether the relation between metal uptake and mass flow is caused by an enhanced supply of Cd from bulk soil to the rhizosphere or rather by a coupled transport of water and solute across root membranes cannot be deduced from our data. Our finding that transpiration seems to affect metal uptake by crops is in line with observations of several other studies. Kashem and Singh (2002) and Lorenz et al. (1994) report that the uptake of Cd and Zn into plants (radish, oat, and water spinach) was positively correlated with mass flow of metals. Kashem and Singh (2002) point out that the contribution of mass flow to the supply of plants with major cations and heavy metals is highly dependent both on metal type and plant species. Grifferty and Barrington (2000) observed a 3-times-higher Zn uptake by winter wheat plants at a 2.3-times-higher transpiration rate. Hardiman and Jacoby (1984) showed that the total amount of Cd absorbed by bean plants was elevated by inducing higher transpiration rates. Mullins et al. (1986) simulated the uptake of Cd and Zn by maize from soils treated with sewage sludge using the Barber–Cushman model. A sensitivity analysis suggested that mass flow could contribute a significant portion of the Zn and Cd supply from soil to root. On the other hand, Florijn and van Beusichem (1993) investigated the uptake and distribution of Cd in maize but found no relationship between shoot Cd content and the transpiration coefficient. They concluded that internal distribution rather than uptake is mainly responsible for the genotypic variation in the shoot Cd content of maize.

Another effect of transpiration in soil is that it lowers the soil water content, which may affect the solution-phase concentration of Cd and in consequence its plant uptake. Whether transpiration and root water uptake, respectively, increases or decreases the Cd solution-phase concentration depends on the Cd uptake mechanism of the plant under study. Provided that the proposed Eq. [2] holds for describing plant uptake, three cases can be distinguished [for details see Ingwersen (2001)(p. 130–131)]. If the uptake is passive ({eta} = 1), water removal by plant roots has no effect on the Cd solution-phase concentration. If the uptake is active ({eta} > 1), soil water is depleted and Cd concentration decreases. If the plant employs a mechanism to restrict the uptake via massflow ({eta} < 1), the dissolved metal concentration increases. However, because Cd is strongly sorbed in soil, a change in the solution-phase concentration is buffered by sorption and desorption processes. Whether sorption and desorption are effective in buffering a concentration change depends on their kinetics.

In our study a significant fraction of the variance in potato Cd contents could only be explained after we grouped the investigated potatoes into three empirical clusters. This weakness clearly demonstrates the limits of the proposed model. If there are additional important factors beneath the solution-phase concentration that govern the plant uptake of a metal, the model needs to be calibrated or extended. Such additional factors might depend on the cultivars, which may have different mass flow transfer rates, or soil processes such as metal complexation or a competition with other metal ions. Moreover, our study was performed in a relatively homogeneous area with relatively similar soil properties, a uniform climate, and a low level of Cd. The model robustness will probably decrease as it is used under more heterogeneous conditions.

For all three crops, we found a significant relationship between shoot Cd content and solution-phase concentration. This finding agrees with many other studies that indicate that the solution-phase concentration strongly controls Cd uptake (Grant et al., 1998). The solution-phase concentration itself, unless the equilibrium is controlled by precipitation, depends on the sorbed-phase concentration and on the sorptive strength of soil. This, in turn, is influenced by soil physicochemical properties such as soil pH and organic carbon content. Several methods are available to measure the solution-phase concentration (centrifuge method, suction cups, equilibration with a background electrolyte, etc.). Note, however, that different methods may yield different solution-phase concentrations for the same soil. Because ionic strength as well as chloride and calcium concentrations affect Cd sorption (Boekhold et al., 1993), the use of different background electrolytes (e.g., 0.01 M CaCl2) will yield different solution-phase concentrations and thus result in different estimates, for example of the plant parameter {eta}. Therefore, it must be pointed out that the success of any future experiment to apply or test our plant uptake model will strongly depend on the accurate choice of the method to determine the solution-phase concentration. A possible method to derive the composition of an appropriate background electrolyte is given in Matschonat et al. (2003).

Many investigators observed that Cd was retained to a high degree in roots. For example, Jarvis et al. (1976) used solution culture techniques to investigate 23 species and found that between 35 and 91% of the Cd taken up was retained in the roots. Winter wheat, for example, retained 68%. The percentages given are probably much higher than under field conditions because the experiments did not last for a complete vegetation period: the distribution between roots and shoots was only examined four days after a single, three-day exposure to a nutrient solution containing 10 µg L–1 added Cd. In another study, Cataldo et al. (1981) reported that, especially at maturity, high amounts of Cd are remobilized to the shoots or rather the seeds. Reevaluating of wheat data from Lübben's pot experiments (Lübben, 1993, p. 206–207) revealed that on average about 20% of the total Cd taken up was retained in roots. This might explain why Cd contents of wheat grain were overestimated by 12% in our study. The model assumes that all the taken-up Cd is transported to the shoots. Consequently, if some Cd is retained in the roots, the model overestimates the Cd content of shoots. Expanding our model by a root compartment would probably improve the results. Unfortunately, sampling as well as sample preparation of roots is extremely difficult under field conditions. Any application to plants with considerable root-shoot barriers like soybean or bush beans, however, would certainly require the model to be extended by a root compartment.

In the year with the higher saturation deficit of the atmosphere we observed a relatively higher Cd uptake of crops. According to our modeling concept this was because of a higher mass flow, which resulted in enhanced Cd transfer from soil solution to plant. However, Eriksson et al. (1996) report a contrary finding. They reviewed results from 20 years of field studies focused on evaluating the influence of soil Cd content on Cd levels in agricultural crops. They observed that Cd levels in winter wheat and oat were lower in years with a dry summer. The authors did not measure the saturation deficit of the atmosphere but it is very probably that in the dry summers the saturation deficit of the atmosphere was increased. The authors discuss several possibilities and speculate, among others, that in dry years roots penetrate deeper into the soil where Cd concentrations are generally lower. In such a situation, of course, our model would fail, because in the model, climatic factors do not affect the root distribution. However, in our study area even in dry summers crops should not suffer from water shortness because of the high irrigation heights.

Several studies have shown that Zn and other cations inhibit Cd uptake (Cataldo et al., 1983; Jarvis et al., 1976; Shanker et al., 1996). Whether competitive inhibition becomes important depends strongly on the concentration of competing cations. For example, the Cd uptake by ryegrass fell at a Zn concentration of about 6500 µg L–1 (Jarvis et al., 1976). The present study found no evidence for a depression of the Cd uptake by Zn competition. In the soils of the WIA, the maximum Zn concentration in Ap horizons was about 350 µg L–1 and, hence, markedly lower than that used in the above study.


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
In our study we observed that the uptake of Cd by crops was strongly correlated with the solution-phase concentration. This finding supports the concept that the soil solution plays a critical role in controlling the uptake of heavy metals by plants. The solution-phase concentration itself depends strongly on soil factors such as total Cd content, pH, and organic carbon content. Moreover, the Cd uptake of crops was related to the saturation deficit of the atmosphere. This indicates that, beneath several other plant and soil factors, the climate also plays a role in controlling Cd uptake. Therefore a report on the climatic conditions during a study should be helpful to interpret results and to compare different studies. Our simulations, which are based on a mass flow–governed uptake of heavy metals, agreed well with field observations of Cd contents in crops. The chosen modeling approach was suited to model Cd uptake by crops under the conditions of this study. An application of the model to other study areas, crops, or metals would need calibration and/or model extension. More research is needed to systematically investigate which factors control to which extent mass flow, diffusive, and active uptake mechanisms contribute to the uptake of heavy metals by crops.


   ACKNOWLEDGMENTS
 
Special thanks to Prof. Eggers, Mr. Seeßelberg, and Mr. Blickwede of the Braunschweig Municipal Waste Water Association for their cooperation. We thank Mrs. B. Heine, Mrs. C. Marheineke, Mrs. D. Müntner, and Mr. A. Küsters for laboratory assistance. The research work presented in this paper was funded by the Deutsche Forschungsgemeinschaft (German Research Foundation).


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
This work was funded by the German Research Foundation (DFG).


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 J Exp BotHome page
L. Van der Vliet, C. Peterson, and B. Hale
Cd accumulation in roots and shoots of durum wheat: the roles of transpiration rate and apoplastic bypass
J. Exp. Bot., September 4, 2007; (2007) erm119v1.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
J. Ingwersen and T. Streck
Modeling the Environmental Fate of Cadmium in a Large Wastewater Irrigation Area
J. Environ. Qual., August 9, 2006; 35(5): 1702 - 1714.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (9)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ingwersen, J.
Right arrow Articles by Streck, T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ingwersen, J.
Right arrow Articles by Streck, T.
Agricola
Right arrow Articles by Ingwersen, J.
Right arrow Articles by Streck, T.
Related Collections
Right arrow Plant and Soil Interactions
Right arrow Municipal Wastes
Right arrow Heavy Metals
Right arrow Other Models
Right arrow Soil Pollution

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome