Plant Soil (2011) 338:423–434 DOI 10.1007/s11104-010-0556-2 REGULAR ARTICLE Soil solution dynamics and plant uptake of cadmium and zinc by durum wheat following phosphate fertilization Xiaopeng Gao & Donald N. Flaten & Mario Tenuta & Mark G. Grimmett & Eugene J. Gawalko & Cynthia A. Grant Received: 18 June 2010 / Accepted: 23 August 2010 / Published online: 4 September 2010 # Springer Science+Business Media B.V. 2010 Abstract A growth chamber study was conducted to evaluate the effect of application of phosphate fertilizer on soil solution dynamics of cadmium (Cd) and Cd accumulation in durum wheat (Triticum turgidum L. var. durum). Treatments consisted of three phosphate fertilizer sources containing 3.4, 75.2, and 232 mg Cd kg−1 applied at three rates (20, 40 and 80 mg P kg−1) plus a no fertilization control. An unplanted treatment at 40 mg P kg−1 was included to separate the effects on soil solution Cd dynamics of the crop from that of the fertilizer. Soil solution M. G. Grimmett Agriculture and Agri-Food Canada, Crops and Livestock Research Centre, Charlottetown, PE, Canada C1A 4N6 samples were obtained using soil moisture samplers every 10 days after germination. The experimental results indicated that plant biomass significantly increased with P application rates and decreased with increased Cd concentration in the phosphate fertilizers. Total cadmium concentration in soil solution was not consistently affected by phosphate fertilization rate and fertilizer sources, and therefore Cd concentration in the fertilizer. Application of phosphate fertilizer, however, increased the concentration and accumulation of Cd and shoot Cd/Zn ratio, and decreased shoot Zn concentration in durum wheat. Phosphate sources had a marginally significant effect (P=0.05) on shoot Cd concentration and did not affect Cd accumulation in durum wheat. Concentration of Cd in soil solution was unrelated to Cd concentration in durum wheat. These results suggest that the immediate increase in Cd concentration and Cd accumulation in durum wheat with phosphate application is due more to competition between Zn and Cd for absorption into plants, enhanced root to shoot translocation and enhanced root development, than to a direct addition effect from Cd contained in phosphate fertilizer. In the short term, application of phosphate fertilizers can increase Cd concentration in the crops, regardless of the Cd concentration of the fertilizer. An optimal P fertilization, possibly in combination with Zn application, may offer an important strategy for decreasing Cd concentration and accumulation in crops. E. J. Gawalko Canadian Grains Commission, Winnipeg, MB, Canada R3C 3G8 Keywords Cadmium . Cd/Zn ratio . Durum wheat . Phosphate . Soil solution . Zinc Responsible Editor: Fangjie Zhao. X. Gao : D. N. Flaten : M. Tenuta Department of Soil Science, University of Manitoba, Winnipeg, MB, Canada R3T 2N2 X. Gao e-mail: gaox@agr.gc.ca X. Gao : C. A. Grant (*) Agriculture and Agri-Food Canada, Brandon Research Centre, Box 1000A, R.R.#3, Brandon, MB, Canada R7A 5Y3 e-mail: Cynthia.Grant@AGR.GC.CA 424 Introduction Cadmium (Cd) is a potentially toxic trace metal naturally present in soils and it may also be added to soil as a contaminant in fertilizer, manure, sewage sludge and from aerial deposition (Grant et al. 1998). Excessive Cd concentrations represent a threat to soil productivity, and environmental and human health because of the accumulation of Cd in the food chain. There is, therefore, considerable interest in management of Cd in plant-soil systems and in strategies for reducing Cd accumulation in crops (Grant et al. 2008). One of the major inputs of Cd into agricultural soils is application of phosphate fertilizers (Grant and Sheppard 2008). Cadmium can be present in phosphate fertilizer at concentrations varying from near 0 to more than 300 mgkg−1, depending on the provenance of the phosphate rock (Mortvedt and Osborn 1982). In field studies in Sweden (Andersson and Siman 1991) and in the Canadian prairies (Grant and Bailey 1997), Cd concentration in grain and seeds of several crops consistently increased with increasing phosphorus (P) application, which was attributed to the Cd contained in the phosphate fertilizer and the fertilizer effect on soil properties and plant growth. Similar to the longterm accumulation of Cd in the soil over time from repeated application of P fertilizer, P application has also been shown to increase Cd concentration and accumulation in crops over the short term. For example, uptake of Cd by rape (Brassica napus L.) and oats (Avena sativa L.) in pot experiments was greater when the P source was a single superphosphate than when it was a NPK source having a lower Cd concentration (Singh 1990). As well as adding Cd as a contaminant directly to the soil, phosphate fertilizer may also influence Cd availability indirectly through its effects on soil properties and plant growth. In pot trials, application of reagent-grade monoammonium phosphate (MAP) was shown to increase Cd concentration in shoot and grain or seed of durum wheat (Triticum turgidum L. var. durum) and flax (Linum usitatissimum L.) (Choudhary et al. 1994; Jiao et al. 2004). To better understand the effect of phosphate fertilizer on phytoavailability of Cd in soil-plant system, it is therefore necessary to distinguish the fertilizer’s direct effect of adding Cd to the soil from its indirect effects through its influence on other soil and plant factors. Improved understanding of the mechanisms through which phosphate affects Cd would also Plant Soil (2011) 338:423–434 help to determine the optimal strategy for P fertilization management to decrease Cd accumulation in soils and crops. Assessing the impact of phosphate fertilization on the accumulation of Cd in soils and its transfer to plants requires adequate knowledge of how various factors affect Cd phytoavailability. Soil properties that can influence Cd availability include pH, clay type, chloride content and the content of soil organic matter and Fe and Mn oxides (He and Singh 1993; McLaughlin et al. 2000). Of these soil properties, soil pH is often regarded as the most important factor (Gavi et al. 1997; Hooda et al. 1997; Grant et al. 1999). In a laboratory study, Levi-Minzi and Petruzzelli (1984) showed that addition of MAP decreased soil pH and the amount of Cd adsorbed by soil, while application of diammonium phosphate increased soil pH and led to Cd precipitation. Similarly, Lambert et al. (2007) showed that MAP application at commercially relevant rates decreased soil pH and increased Cd concentrations in soil extracts of field and the laboratory experiments. Although changes in soil solution concentrations of Cd and pH after the application of phosphate fertilizers have been investigated, little is known about the temporal variation in soil solution Cd and pH following phosphate fertilization. The relationship between plant Cd concentration and Cd pools or soil properties is inconsistent (Singh and Kristen 1998; McBride 2002; Huang et al. 2004; Oporto et al. 2009). Generally, the concentration and accumulation of Cd in plants is positively associated with the soil soluble Cd fractions (Adams et al. 2004; Kamewada and Nakayama 2009). However, François et al. (2009) found that total Cd added to soil and some key soil properties were better estimators than water extractable Cd when predicting the Cd concentration in durum wheat following application of phosphate fertilizers varying in Cd concentration. Similarly, by examining the relationship between the solid-solution partition coefficient and the concentration of Cd in plant tissues obtained from 70 crop production fields in California, Chen et al. (2009) found that the total soil Cd concentration seems a more appropriate estimator of plant uptake than the concentration in the soil solution. Reasons for the inconsistent results remain unclear but are mostly related to variation among experiments in factors such as soil types, plant species and soil Cd extraction methods. Though a single universal soil test to Plant Soil (2011) 338:423–434 accurately assess the phytoavailability of Cd does not exist (McLaughlin et al. 2000), measuring the total dissolved Cd in the soil solution seems a more direct means of estimating the potential for Cd uptake by a crop (McBride 2002). In a recent pot trial, Cd uptake by garland chrysanthemum (Chrysanthemum coronarium L.) was successfully predicted from the total soluble Cd in the soil solution, independent of soil type (Kamewada and Nakayama 2009). Phosphorus fertilization is frequently reported to reduce Zn concentration in plant tissues. Grant et al. (2002) found phosphate fertilization generally decreased grain Zn concentration but increased Cd concentration of durum wheat grown at various locations in Canada. Verna and Minhas (1987) found that P fertilization increased the translocation of Zn into above ground of wheat and maize (Zea mays L.). In a pot trial, high P supply decreased Zn uptake of three sorghum (Sorghum bicolor L.) genotypes (Chand et al. 1995). There is little information available in the literature about the effect of the Pinduced reduction in Zn concentration on Cd uptake by crops, even though the competition between Zn and Cd for uptake and translocation by the plant reduced concentration and accumulation of Cd in several studies (McLaughlin et al. 1994; Oliver et al. 1994; Grant and Bailey 1997; Gao et al. 2010a). To further understand the influence of phosphate fertilization on Cd concentrations in soil solution and Cd uptake in durum wheat, we conducted a pot experiment to investigate the immediate effect of application of monoammonium phosphate fertilizers containing varying concentrations of Cd on the concentrations of Cd, P and Zn in soil solution and plant tissue, as well as total plant accumulation of Cd, P and Zn. We further explored the relationships between plant biomass, Cd, P and Zn concentrations in soil solution and distribution in plant tissue to determine the factors that affect the concentration of Cd in durum wheat following phosphate application. Materials and methods Soil was collected from the 0 to 15 cm depth of a cropped field at Agriculture and Agri-Food Canada’s Brandon Research Centre farm, located near Brandon, MB, Canada. The soil was a clay loam soil classified as a Haplic Chernozem based on FAO soil classification 425 (FAO 1998). Soil was air-dried and passed through a 1 cm mesh screen before being characterized and used in experiments. The soil had a pH of 7.7 (water:soil ratio 2:1), organic matter content of 45 gkg−1, particlesize distribution of 38% sand, 32% silt and 30% clay, electrical conductivity (water:soil ratio 2:1) of 388 μS cm−1, 0.05 M NaHCO3 extractable-P of 12 mgkg−1, 1 M NH4OAc extractable-K of 265 mgkg−1, DTPA-Cd, Zn, and -Cu of 0.16, 0.97, and 1.35 mg kg−1, respectively. The water holding capacity of the soil was determined by saturating the soil and subsequently placing it in a Büchner funnel to allow free drainage. When all excess water was drained from the soil by gravity, the gravimetric water content was determined and set at 100% (Brookes 2009). The measured water holding capacity of the tested soil was 424 g H2O kg−1 dry soil. The experiment was conducted in a plant growth chamber (model PGW36, CONVIRON, Winnipeg, Canada) at the Department of Soil Science, University of Manitoba. Plastic cylinder pots with a diameter of 14 cm and a height of 20 cm were used. Each pot contained 3.0 kg dry equivalent soil. The experimental setup included three different sources of MAP fertilizer at three rates, plus a control which did not receive phosphate, for a total of 10 treatments. Unplanted pots for all three types of MAP fertilizers at 40 mg P kg−1 and control were included to separate the effects on soil solution Cd dynamics of the crop from that of the fertilizer. All treatments were replicated three times in a randomized compete block design for a total of 42 pots. As determined by acid digestion, the MAP fertilizers contained 3.4, 75.2 and 232 mg Cd kg−1 for fertilizers obtained from Ontario (ON-Canada), North Carolina (NC-USA) and Idaho (ID-USA), respectively. They also contained 117, 869 and 3,500 mg Zn kg−1, respectively. Use of fertilizers with same Zn concentration but varying Cd concentration to avoid the co-variance would be ideal. In this study, however, we preferred using the commercial sources of MAP over reagent grade sources because the availability of the trace elements in the complex mineral forms in the commercial fertilizers may be different from those in physical blends of reagent grade chemicals. Rates of phosphate applications were 20, 40 and 80 mg P kg−1 soil. For the 20 mg P kg−1 rate, the Cd input to soil was 0.3, 6.6 and 20.3 μg Cd kg−1 for each fertilizer, respectively. For 426 Zn, this corresponded to an addition of 10.2, 76.0 and 306.0 mg Zn kg−1, respectively. The 20 mg P kg−1 rate is representative of the high end of the typical range of fertilization rates that are applied on an annual basis for wheat production in Western Canada, whereas 40 and 80 mg P kg−1 are used to represent the effects of excessive input of fertilizer. The MAP fertilizers were mixed thoroughly and uniformly with the soils. All treatments received mineral nutrients applied in solution at rates of 150 mgN (reagent grade NH4NO3 was used to balance the N applied with the MAP) and 100 mgK (reagent-grade K2SO4) kg−1 soil. To extract soil solution, Rhizon In Situ Samplers (RISS) (Rhizon Research Products, Wageningen, The Netherlands) were installed horizontally, 10 cm below the soil surface in each pot. Each sampler was made of a hydrophilic porous polymer tube, with a typical pore diameter of 0.2 μm, extended with a polyvinyl chloride tube. The outer diameter of a Rhizon was 2.4 mm, and the filter section had a length of 10 cm. These simple porous tubes allow for extraction of pore waters with a syringe and are suitable for trace metal work after being rinsed with 5% HNO3 followed by ultra-pure water. Ten seeds of durum wheat (cv. AC Melita) were sown in each pot and thinned to four seedlings 7 days after emergence. The pots were watered daily with reverse osmosis water, maintaining moisture content at 70% water holding capacity. Emergence occurred 3 days following sowing. The temperature in the growth chamber was 25°Cday and 15°C night, 16 h light and 8 h dark, relative humidity was 60% and photosynthetically active radiation light intensity was 762 μmolm−2 s−1. Plants were grown for 8 weeks before the shoot material was harvested. Plants were at boot stage at harvest according to the Feekes scale (Miller 1992). Soil solution was extracted on a total of six occasions (at 10 day intervals for a total of 252 samples) by attaching sterile, 10 mL plastic syringes (BD Syringe, Franklin Lakes, USA) to the RISS and exerting a vacuum pressure equivalent to 0.3 Atm. The volume of each sample was approximately 7 mL and allowed for a range of chemical analyses including Cd, P and Zn, together with pH. Solution pH was determined by using an Accumet basic AB15 pH meter with a glass electrode 13–620–531 (Fisher Scientific, Ottawa, Canada). Elemental concentrations were determined by an ICP-MS (Perkin Elmer Elan Plant Soil (2011) 338:423–434 6000, PerkinElmer, Massachusetts, USA) with prior acidification to 2% using trace-metal grade HNO3. The detection limit was 0.01 μgL−1 for Cd and 0.2 μgL−1 for Zn. At harvest, the plant shoots were cut off at ground level and soil was washed from the roots with tap water followed by deionized water. The plant shoots and roots were then oven dried at 70°C for 48 h, and weighed. Dried and ground plant samples were digested in a boiling acid mixture (HNO3 + HClO4) for element analysis (Westman 1990). Cadmium in plant digests was analyzed using a Varian Spectra AA400 (Mulgrave, Australia) graphite furnace atomic absorption spectrophotometer with deuterium background correction. Phosphorus and Zn concentration in the digests were determined using an ARL 3520 (Chermside, Australia) inductively coupled plasma spectrophotometer. Reliability of the analysis was assessed by including certified plant reference materials in each set of digests. Statistical analysis All data were tested and met the requirements of normality (Kolmogorov–Smirnov) and homogeneity of variance (Levene’s test). Analysis of variance (Proc Mixed) was performed for shoot and root dry weight, element concentrations in plant tissue and element accumulation, as well as for pH and element concentrations in soil solution. Element accumulation was calculated as the total plant element content (biomass yield × concentration). Means were compared using a Tukey’s Honestly Significant Difference test at α=0.05. Linear regression analyses were performed between element concentrations in soil solution and concentration or accumulation in plant tissues. All analyses were performed with SAS Release 9.0 (SAS Inc., NC, USA). Results Soil solution pH and concentrations of Cd, P, and Zn In general, the pH of soil solution in the rooted soil of durum wheat increased during the first 30 d, then declined (Fig. 1). Neither phosphate fertilization rate nor fertilizer source/Cd concentration in fertilizer affected solution pH consistently. In comparison, for the unplanted treatments (Table 1), soil solution pH in the Plant Soil (2011) 338:423–434 427 Soil solution pH and concentrations of Cd, P, Zn in the planted soil of durum wheat following the application of three sources of monoammonium phosphate fertilizers (Ontario-ON, North Carolina-NC and Idaho-ID) at three rates (20, 40, and 80 mg P kg−1 soil) ƒFig. 1 pH in soil solution 8.0 7.5 7.0 6.5 10 20 30 40 50 60 50 60 50 60 Cd concentration in soil solution (ng L-1) Days after germination 100 75 50 25 0 10 20 30 40 4000 3000 2000 1000 Zn concentration in soil solution (µg L-1) P concentration in soil solution (µg L-1) Days after germination 0 10 20 30 40 Days after germination control without phosphate was always 0.2–0.3 pH units greater than that in the pots with 40 mg P kg−1 phosphate application. The planted treatments generally resulted in greater pH values compared to the unplanted treatments and the differences increased substantially in the last half of the incubation period (Table 1). The concentration of Cd in the soil solution in the rooted soil changed with time but neither phosphate fertilization (P=0.12) nor fertilizer source (P=0.15) showed consistent effects (Fig. 1). Phosphate fertilization significantly increased soil solution P concentration compared to the control (P<0.001), but showed no consistent effect on Zn (P=0.71). Increasing the P application rate led to greater P concentration in the soil solution. Phosphate concentration in the soil solution continued to decrease in the first half of the experimental period (0–30 d after germination) and then showed little change afterwards. Solution pH was significantly correlated with P (r=−0.39, P<0.001) and Zn (r=0.46, P<0.001), but not with Cd (r=−0.12, P= 0.11) concentrations in the soil solution. Similar to the effect on Cd, fertilizer sources did not affect the concentrations of P (P=0.11) or Zn (P=0.12) in soil solution. Under unplanted conditions, application of fertilizer increased Cd concentration in soil solution at three out of the six sampling times, resulting in an overall significant positive effect (Table 2). In contrast, Table 1 Comparison of soil solution pH between control and monoammonium phosphate fertilized (MAP at 40 mg P kg−1) treatments under both planted and unplanted conditions. Data for MAP treatment were means of three types of monoammonium phosphate fertilizers 25 20 15 Treatment Days after germination Mean 10 10 20 30 40 50 60 control 7.13 a 7.37 a 7.47 a 7.27 a 7.23 a 7.03 a 7.25 a MAP 6.87 b 7.14 b 7.28 b 7.10 b 7.08 b 6.93 a 7.07 b 5 Unplanted 0 10 20 30 40 50 60 Days after germination Control ID20 ON20 ON40 ID40 ON80 NC20 NC40 NC80 ID80 Planted control 7.30 a MAP 7.07 b 7.30 a 7.33 a 7.50 b 7.53 a 7.73 a 7.57 a 7.49 a 7.70 a 7.71 a 7.48 a 7.49 a 7.69 a Within planted or unplanted treatment, per column means followed by the same letter are not significantly different (Tukey, P<0.05)` 428 Plant Soil (2011) 338:423–434 Table 2 Comparison of Cd concentration (ngL−1) in soil solution between control and monoammonium phosphate fertilized (MAP at 40 mg P kg−1) treatments under both planted and unplanted conditions. Data for MAP treatment were means of three types of monoammonium phosphate fertilizers Treatment Days after germination 10 20 30 Mean 40 50 Table 3 Shoot and root dry weight of durum wheat as affected by the application of three sources of monoammonium phosphate fertilizers (Ontario-ON, North Carolina-NC and Idaho-ID) at three rates (20, 40, and 80 mg P kg−1 soil) Fertilizer Shoot dry weight source (g plant−1) 0 None 15.9 1.4 20 ON 22.5 2.9 60 Unplanted control 38.1 b 26.3 b 36.7 a 44.2 b 42.8 a 28.1 a 38.3 b MAP 57.3 a 52.4 a 56.2 a 70.2 a 43.1 a 37.1 a 52.7 a Planted control MAP 43.1 a 49 .0 a 51.5 a 41.8 a 56.3 a 29.5 a 45.2 a 43.2 a Root dry weight (g plant−1) P application rate (mgkg−1) 40 33.3 b 29.6 b 55.9 a 23.0 b 24.7 a 35.0 b Within planted or unplanted treatment, per column means followed by the same letter are not significantly different (Tukey, P<0.05) 80 NC 19.1 2.2 ID 19.1 2.4 ON 22.1 2.4 NC 22.6 2.7 ID 20.1 2.0 ON 25.8 2.8 NC 23.8 2.3 22.3 2.3 ID Contrast opposite results were found in the planted treatment, where the overall mean concentration of Cd in the fertilized treatment was lower than in the control. There were no relationships between Cd concentration in soil solution and P (r=−0.02, P=0.8) or Zn (r=0.02, P=0.8), but there was a weak but significantly negative relationship between P and Zn concentrations (r=−0.16, P=0.04). DF 1 Control vs others Analysis of variance Pr≥F <0.001 <0.001 Factor DF Rate of P 2 Pr≥F <0.001 0.74 Source of P 2 <0.001 0.02 Rate × Source 4 0.01 0.21 Values are the mean of three replicates DF Degrees of freedom Concentrations and accumulations of Cd, P, and Zn in plant tissues Adding phosphate fertilizer significantly increased both shoot and root dry weight compared to control (Table 3), indicating P was a growth limiting factor in the control plants. Shoot dry weight increased with increasing fertilization. In contrast, root dry weight was not affected by fertilization rate. Fertilizer containing high Cd concentration (Idaho) resulted in lower shoot and root dry weight than fertilizer with low Cd concentration (Ontario). A rate by source interaction indicated that the shoot biomass increased more consistently for MAP from North Carolina and Idaho than for MAP from Ontario. Application of phosphate increased Cd and P concentrations and decreased Zn concentration in both shoot and root of durum wheat, with the effect being greater at the higher rates of fertilization (Table 4). Fertilizer source, therefore, Cd concentration in fertilizer showed only a marginally significant Two-way analysis of variance was done on data at P treatments including 20, 40 and 80 mgkg−1 rates effect (P= 0.05) on shoot Cd concentration, but significantly affected shoot P (P=0.01) and Zn (P= 0.002) concentrations. Shoot P concentrations at 20 and 40 mgkg−1 application rates were greater for Idaho fertilizer than for Ontario and North Carolina fertilizers but the opposite was observed at 80 mg kg−1 application rate, resulting in a significant fertilizer rate by source interaction. Accordingly, the shoot Zn concentration for the Idaho fertilizer treatment was similar to that for the Ontario and North Carolina fertilizer treatments at 20 and 40 mg kg−1, but a greater concentration was observed for Idaho fertilizer at the 80 mgkg−1 rate. Neither Cd concentration nor Zn concentration in the plant shoot was related to the initial concentration of Cd (r= 0.12, P= 0.52) or Zn (r=−0.28, P=0.13) in soil solution. Phosphorus concentration in plant shoot, Plant Soil (2011) 338:423–434 429 Table 4 Cadmium, P and Zn concentrations in shoot and root of durum wheat as affected by the application of three sources of monoammonium phosphate fertilizers (Ontario-ON, North Carolina-NC and Idaho-ID) at three rates (20, 40, and 80 mg P kg−1 soil) P application rate Fertilizer source Element concentration in durum wheat (mgkg−1) Cd (μgkg−1) P (mgkg−1) Shoot Root Shoot Zn (mgkg−1) Cd/Zn×103 P/Zn Root Shoot Root Shoot Root Shoot Root 0 None 151 550 1548 953 27.0 22.4 5.6 24.6 57.3 42.5 20 ON 214 608 2513 1273 24.4 20.1 8.8 30.2 103.0 63.3 40 80 Contrast NC 220 624 2337 1395 23.3 19.9 9.4 31.4 100.3 70.1 ID 267 667 2722 1487 23.9 20.7 11.2 32.2 113.9 71.8 ON 312 597 3017 1598 20.7 19.1 15.1 31.3 145.7 83.7 NC 274 634 2648 1447 16.5 18.2 16.6 34.8 160.5 79.5 ID 324 685 3387 1403 20.3 17.6 16.0 38.9 166.8 79.7 ON 301 584 3838 1540 12.6 15.7 23.9 37.2 304.6 98.1 NC 322 734 3663 1748 12.4 17.1 26.0 42.9 295.4 102.2 ID 324 794 3120 1847 13.5 20.1 24.0 39.5 231.1 91.9 DF Pr≥F Control vs others 1 <0.001 0.02 <0.001 <0.001 0.007 0.009 <0.001 <0.001 <0.001 <0.001 Analysis of variance Factor DF Rate of P 2 Pr≥F <0.001 0.002 <0.001 0.002 <0.001 0.02 <0.001 0.001 Source of P 2 0.05 <0.001 0.01 0.40 0.002 0.34 0.28 0.01 0.003 0.77 Rate × Source 4 0.28 0.02 <0.001 0.13 0.02 0.13 0.53 0.15 <0.001 0.81 <0.001 <0.001 Values are the mean of three replicates DF Degree of freedom Two-way analysis of variance was done on data at P treatments including 20, 40 and 80 mgkg−1 rates however, correlated significantly and positively with the initial P concentration in soil solution (r=0.79, P<0.001). The Cd/Zn ratio in the shoot was 1.5 to 3.4 times less than that in the root. Conversely, the P/ Zn ratio in the shoot was 1.3 to 2.1 times greater than that in the root (Table 4). Adding phosphate increased the Cd/Zn and P/Zn ratios in the shoot and root tissue, with the effect being greater with increasing fertilization rate. Greater concentrations of Cd in fertilizer source resulted in greater Cd/Zn ratios in roots, but did not affect Cd/Zn ratios in shoots. Conversely, fertilizer sources had no effect on P/Zn ratios in roots and a significant, but inconsistent effect on P/Zn ratios in shoots. The concentration of Cd in the shoot correlated positively and strongly with the shoot P concentration (r=0.82, P<0.001) and the shoot P/Zn ratio (r=0.70, P< 0.001), and negatively with the shoot Zn concen- tration (r = −0.68, P < 0.001) (Fig. 2). Shoot Zn concentration was negatively correlated with shoot P concentration (r=−0.75, P<0.001). Phosphate fertilization greatly increased the total accumulation of Cd of durum wheat compared with the control plants (P < 0.001), with the effect being greater with increasing fertilization rate (Fig. 3). Fertilizer source had no significant effect (P= 0.24) and there was no interactive influence (P= 0.30) of fertilizer source and rate on Cd accumulation in the whole plant. Similar to Cd concentration in plant tissue, Cd accumulation was unrelated to initial Cd concentration in soil solution (r = 0.13, P= 0.51). Element translocation from root to shoot refers to the proportion of element present in shoot relative to the total accumulation in the plant. Compared to the control, phosphate application had no overall effect on Cd and P translocation, Shoot Cd concentration (µg kg-1) Shoot Cd concentration (µg kg-1) Shoot Zn concentration (mg kg-1) Cd accumulation (µg plant-1) Plant Soil (2011) 338:423–434 Shoot Cd concentration (µg kg-1) 430 400 300 200 r = 0.82 P < 0.001 100 1000 12 Root 10 1 Control 2 ON20 3 NC20 4 ID20 5 ON40 6 NC40 7 ID40 8 ON80 9 NC80 10 ID80 8 6 4 2 0 1 2000 3000 4000 Shoot P concentration (mg kg-1) Shoot 2 3 4 5 6 7 8 9 10 Treatment Fig. 3 Cadmium accumulation in durum wheat as affected by application of three sources of monoammonium phosphate fertilizer (Ontario-ON, North Carolina-NC and Idaho-ID) fertilizers at three rates (20, 40, and 80 mg P kg−1 soil) 400 300 Discussion 200 r = -0.68 P < 0.001 100 10 15 20 25 Shoot Zn concentration (mg kg-1) 30 400 300 200 r = 0.70 P < 0.001 100 0 100 200 Shoot P/Zn ratio 300 30 25 20 15 10 1000 r = -0.75 P < 0.001 2000 3000 4000 Shoot P concentration (mg kg-1) Fig. 2 Relationships between element concentrations in shoot of durum wheat but decreased translocation of Zn (Table 5). With increasing fertilization rate, both Cd and P translocation show a small but significant increase. However, the increase was not consistent across fertilizer sources. Cadmium concentration in fertilizers had no effect on translocation of Cd, P or Zn. In the present study, Cd concentration in soil solution ranged from 20 to 80 ngL−1. Similar ranges of Cd in soil solution have been frequently reported in previous studies (Bonito 2005; Meers et al. 2007; Beesley and Dickinson 2010) that also used the Rhizon soil moisture sample method to extract the soil. In a recent work, Gao et al. (2010b) used the same extraction method on a Chinese agricultural soil and measured total Zn concentrations ranging from 9 to 15 μgL−1, which was similar to the present study (5–20 μgL−1, Fig. 1). In addition, P concentration in soil solution correlated well with the phosphate fertilization rate (Fig. 1). These facts suggest that the Rhizon extraction method is a reliable method for collecting soil solution in situ, even on an uncontaminated site. The lack of association between shoot Cd concentration of durum wheat and the initial concentration of Cd in soil solution, plus the observation that phosphate sources had only a small effect on the shoot Cd concentration (Table 4) and did not affect total plant Cd accumulation (Fig. 3), suggest that the increase in Cd uptake following phosphate fertilization was not due to a direct addition effect of Cd contained in fertilizer. It should be noted that this observation may be limited to this particular soil or similar soils with relatively higher buffering capacity. Use of other soils with lower buffering capacity might have different effects. For example, Lambert et al. (2007) found that increasing the phosphate application rate or Cd contamination in fertilizer inconsistently enhanced the extractability of Cd in soils on seven experimental field sites. The inconsistency between field sites was related to variability of soil parameters, specifically Plant Soil (2011) 338:423–434 431 Table 5 The translocation of Cd, P and Zn from root to shoot of durum wheat as affected by the application of three sources of monoammonium phosphate fertilizers (Ontario-ON, North Carolina-NC and Idaho-ID) at three rates (20, 40, and 80 mg P kg−1 soil). Translocation was calculated as the percentage of element present in shoot relative to the total accumulation P application rate (mgkg−1) Fertilizer source Element root to shoot translocation (%) Cd P Zn 0 None 75.2 94.7 92.9 20 ON 73.3 93.9 90.3 NC 75.0 93.5 90.8 ID 76.4 93.7 90.4 ON 82.5 94.4 90.7 NC 78.6 94.0 88.6 ID 82.3 96.0 91.9 ON 82.8 95.9 88.3 NC 81.6 95.5 88.0 ID 79.8 94.3 86.8 Contrast DF Pr≥F Control vs others 1 40 80 0.15 0.87 0.01 Analysis of variance Factor DF Pr≥F Rate of P 2 0.001 0.002 0.004 Source of P 2 0.74 0.52 0.68 Rate × Source 4 0.42 0.02 0.21 Values are the mean of three replicates DF Degree of freedom Two-way analysis of variance was done on data at P treatments including 20, 40 and 80 mgkg−1 rates the soil pH buffering capacity. However, our results agree with those of Choudhary et al. (1994) and Jiao et al. (2004), who used growth chamber studies to demonstrate an increase in Cd concentration in durum wheat occurred after application of reagent grade MAP. Similar results have also been confirmed under field conditions (Grant et al. 2002). These reported results, in combination with ours, indicate that the application of low-Cd phosphate fertilizers as compared to high-Cd fertilizers is unlikely to reduce Cd concentration and accumulation in crops from this type of soil in the short term. It is, however, important to differentiate between the immediate, short-term impact of P fertilizer in the year of application as opposed to the potential long-term accumulation of Cd in soils from repeated application of high Cd fertilizers over time, as these are two distinctly different issues. As opposed to the lack of a short-term effect of Cd concentration in fertilizer, some long-term field studies showed that repeated application of high-Cd phosphate fertilizers resulted in high Cd concentration and accumulation in soils and crops (Andersson and Siman 1991; Grant and Bailey 1997) whereas low-Cd fertilizers did not (Mortvedt 1987; Richards et al. 1998). Therefore, the low-Cd fertilizers, while unlikely to reduce Cd concentration in the crops in the short term, should still be recommended considering the long-term environmental issue that may affect crop production and quality. Addition of high rate of phosphate has been shown to reduce metal bioavailability and proposed as a remedial technology in metal-contaminated soils including Cd (Zwonitzer et al. 2003; Basta and McGowen 2004). However, other field and laboratory studies, using agronomic rates of application on relatively uncontaminated agricultural soils, phosphate fertilizer was found to cause soil acidification and increase Cd concentration in soil extracts (Lambert et al. 2007). We observed similar results under the unplanted conditions, where MAP fertilization decreased soil solution pH (Table 1) and increased Cd concentration (Table 2) in soil solution. These fertilizer effects on pH and Cd concentration in soil solution were, however, absent under in the planted pots (Tables 1 and 2; Fig. 1), suggesting the importance of root activity on Cd availability. As documented by others, root-induced chemical changes in the rhizosphere, especially the pH changes and release of root exudates, could cause rhizosphere acidification and increase the solubilization of particulate-bound Cd into soil solution (Cieśliński et al. 1998; Nigam et al. 2001). Therefore, in the present study, it is possible that root activity had a much greater effect on Cd uptake than phosphate fertilization. Various hypotheses have been suggested to explain the P-induced inhibition of Zn uptake by crops, including (i) P-Zn interaction in soil (Agbenin 1998), (ii) less translocation of Zn from the roots to the shoots (Olsen 1972; Verna and Minhas 1987), (iii) metabolic disorder within plant cells (Haldar and Mandal 1981) and (iv) dilution effect (Olsen 1972). The P-induced decrease in Zn concentration in this study was likely due to a combination of these factors because a weak but significantly negative relationship between Zn and P concentrations in soil solution (r=−0.16, P=0.04), enhanced plant growth (Table 3), reduced root to shoot translocation of Zn (Table 5) and a negative relationship 432 between shoot Zn and P concentration (Fig. 2) were all observed in response to phosphate fertilization. Although Cd and Zn have many chemical similarities as they belong to same group of the periodic table, similar P effects on Cd for an uncontaminated soils is not expected because there will always be much less Cd than Zn in the soil solution, as well as in the plant tissues. Although addition of phosphate can directly reduce Cd bioavailability during remediation of metalcontaminated soils (Zwonitzer et al. 2003; Basta and McGowen 2004), we did not find a similar reduction of Cd bioavailability by phosphate in this study. In our experiment, application of phosphate fertilizer increased plant Cd concentrations and accumulations, indicating that there is clearly a difference in response between agronomic applications of phosphate fertilizer on uncontaminated agricultural soils and remediation use of phosphate on the contaminated soils where the concentrations of both P and Cd would be much higher. The increase in Cd concentration in plant tissue that we observed with application of P fertilizers may be due to a three-way interaction among P, Zn and Cd. Previous reports suggested that Zn can interfere with uptake and root to shoot transfer of Cd in plants (Oliver et al. 1994; McLaughlin et al. 1995). In this study, the increase in Cd concentration in durum wheat following phosphate application could be mainly due to a P-induced reduction in Zn concentration and consequently the antagonistic interactions between Cd and Zn in plant. Firstly, Cd and Zn are chemically similar and may compete for uptake sites on the root surface and for transporters inside the plants (Kabata-Pandias 2000). In durum wheat, Cd2+ and Zn2+ were found to share a common transport system in the root cell plasma membrane (Hart et al. 2002). Secondly, P-induced Zn deficiency may increase root exudation of amino acids, sugars and phenolics into the rhizosphere. The increased root exudates may further increase Cd uptake by plants due to their chelating/complexing properties for mobilization of Cd. In durum wheat, low-molecularweight organic acids in rhizosphere soils were found to play an important role in the solubilization of particulate-bound Cd into soil solution and its subsequent uptake by plants (Cieśliński et al. 1998). Thirdly, the increase in Cd concentration by Pinduced Zn deficiency could be related to the role of Zn in stimulating the biosynthesis of antioxidant Plant Soil (2011) 338:423–434 enzymes and maintaining root cell plasma membrane integrity (Cakmak 2000). For example, Zn deficiency resulted in decreased enzyme activity of Cu/Zn superoxide dismutase (SOD) and consequently an impairment of the root cell plasma membrane (Cakmak and Marschner 1993; Cakmak et al. 1997), which may increase the membrane permeability and lead to an increase of Cd into the plant through mass flow (Cakmak and Marschner 1988). In this study, increasing phosphate application rate increased not only Cd concentration and accumulation in plants, but also the translocation of Cd from root to shoot (Table 5), suggesting a positive association between shoot Cd concentration and the translocation capacity. Similarly, root-to-shoot Cd translocation via the xylem is the major and common physiological process determining Cd accumulation in shoots and grains of rice (Orazy sative L.) (Uraguchi et al. 2009) and maize (Florijn and van Beusichem 1993). Therefore, control of Cd movement from roots to shoots is of great importance when decreasing Cd concentration and accumulation in the above-ground portion of crops. In this study, application of phosphate fertilizer decreased the root to shoot translocation of Zn while increasing the translocation of Cd (Table 5), confirming the competition between Cd and Zn in plants. The antagonistic interaction suggested that Zn application might be also an effective practice to decrease Cd in crops. Determination of distribution/speciation of metal among various chemical forms in the soil solution is of importance in determining phytoavailability in soil solution. In the current study, however, Cd distribution/speciation in the soil solution was not measured or modeled because of limited access to measurements on free Cd2+ and major cations and anions. Another limitation of this study is that only one soil and one durum genotype vegetative growth were investigated and Cd uptake was measured only during the vegetative stage of growth. However, the soil used is a typical Chernozemic soil in the Canadian prairies, where 50–60% of the durum wheat traded internationally is produced each year (http://www.cwb.ca/public/ en/library/research/popups/durum_pasta_brochure.jsp). Accumulation of Cd in durum wheat has been identified as a concern on the Chernozemic soils of the Canadian prairies (Clarke et al. 1997; Grant and Bailey 1998). Also, the genotype used in the study is widely grown on the Canadian prairies and past studies Plant Soil (2011) 338:423–434 have shown the pattern of response of different durum cultivars to P fertilizer applications is similar (Grant and Bailey 1998). The uptake of Cd in vegetative growth has been shown to closely reflect accumulation in the grain of several crops, including durum wheat (Jiao et al. 2004). The results of this paper therefore provide information on the factors affecting impact of P fertilization on Cd accumulation on soils that produce a large portion of the durum wheat consumed world-wide. However, direct measurement of grain Cd concentration and accumulation with more soils and more genotypes would be desirable to extrapolate the results of this study more generally. In conclusion, results from the present study suggest that increased concentration and accumulation of Cd in durum wheat shoots immediately following phosphate fertilization was due primarily to reduced competition between Zn and Cd for absorption into plants, an enhanced root to shoot translocation of Cd and enhanced root development, rather than to a direct addition effect of Cd contained in phosphate fertilizer. In the short term, application of phosphate fertilizers can increase Cd concentration in the crops, regardless of the Cd concentration of the fertilizer. An optimal P fertilization strategy, likely in combination with Zn application, is of great importance to decrease Cd concentration and accumulation in crops. Further investigations on the complex P-Zn-Cd interaction are needed to more clearly understand the plant-soil mechanisms involved in their uptake and translocation, as well as the agronomic effectiveness of Zn addition on Cd in crops. Acknowledgements The authors gratefully acknowledge the financial support of the Natural Science and Engineering Research Council (NSERC) of Canada, Metals in the Human Environment (MITHE) Research Network and the Canada Research Chair Program in Applied Soil Ecology. The technical assistance of Josh Price is greatly appreciated. References Adams ML, Zhao FJ, McGrath SP, Nicholson FA, Chambers BJ (2004) Predicting cadmium concentrations in wheat and barley grain using soil properties. J Environ Qual 33:532–541 Agbenin JO (1998) Phosphate-induced zinc retention in a tropical semi-arid soil. Eur J Soil Sci 49:693–700 Andersson A, Siman G (1991) Levels of Cd and some other trace elements in soils and crops as influenced by lime and fertilizer level. Acta Agric Scand 41:3–11 433 Basta NT, McGowen SL (2004) Evaluation of chemical immobilization treatments for reducing heavy metal transport in a smelter-contaminated soil. Environ Pollut 127:73–82 Beesley L, Dickinson N (2010) Carbon and trace element mobility in an urban soil amended with green waste compost. J Soils Sediments 10:215–222 Bonito MD (2005) Trace elements in soil pore water: a comparison of sampling methods. PhD thesis, University of Nottingham, UK Brookes P (2009) Laboratory methods of soil microbial biomass research group in rothamsted research http:// www.rothamsted.ac.uk/aen/smbweb1/methods.php? id=896. Accessed 15 December 2009 Cakmak I (2000) Role of zinc in protecting plant cells from reactive oxygen species. New Phytol 146:185–205 Cakmak I, Marschner H (1988) Increase in membrane permeability and exudation in roots of zinc deficient plants. J Plant Physiol 132:356–361 Cakmak I, Marschner H (1993) Effect of zinc nutritional status on activities of superoxide radical and hydrogen peroxide scavenging enzymes in bean leaves. Plant Soil 155/ 156:127–130 Cakmak I, Oztürk L, Eker S, Torun B, Kalfa HI, Yılmaz A (1997) Concentration of zinc and activity of copper/zincsuperoxide dismutase in leaves of rye and wheat cultivars differing in sensitivity to zinc deficiency. J Plant Physiol 151:91–95 Chand K, Dixit ML, Gupta VK (1995) Influence of phosphorus fertilization on Fe and Zn in forage sorghum genotypes. Ann Arid Zone 34:313–315 Chen W, Li L, Chang AC, Wu L, Chaney RL, Smith R, Ajwa H (2009) Characterizing the solid-solution partitioning coefficient and plant uptake factor of As, Cd, and Pb in California croplands. Agric Ecosyst Environ 129:212–220 Choudhary M, Bailey LD, Grant CA (1994) Effect of zinc on cadmium concentration in the tissue of durum wheat. Can J Plant Sci 74:549–552 Cieśliński G, Van Rees KCJ, Szmigielska AM, Krishnamurti GSR, Huang PM (1998) Low-molecular-weight organic acids in rhizosphere soils of durum wheat and their effect on cadmium bioaccumulation. Plant Soil 203:109–117 Clarke JM, Leisle D, Kopytko GL (1997) Inheritance of cadmium concentration in five durum wheat crosses. Crop Sci 37:1722–1726 FAO (1998) World reference base for soil resources. Food and Agriculture Organization of the United Nations, Rome Florijn PJ, Van Beusichem ML (1993) Uptake and distribution of cadmium in maize inbred lines. Plant Soil 150:25–32 François M, Grant CA, Lambert L, Sauvé S (2009) Prediction of cadmium and zinc concentration in wheat grain from soils affected by the application of phosphate fertilizers varying in Cd concentration. Nutr Cycl Agroecosys 83:125–133 Gao X, Akhter F, Tenuta M, Flaten DN, Gawalko EJ, Grant CA (2010a) Mycorrhizal colonization and grain Cd concentration of field-grown durum wheat in response to tillage, preceding crop and phosphorus fertilization. J Sci Food Agric 90:750–758 Gao X, Schröder TJ, Hoffland E, Zou C, Zhang F, van der Zee SEATM (2010b) Geochemical modeling of zinc bioavailability for rice. Soil Sci Soc Am J 74:301–309 434 Gavi F, Basta NT, Raun WR (1997) Wheat grain cadmium as affected by long-term fertilization and soil acidity. J Environ Qual 26:265–271 Grant CA, Bailey LD (1997) Effect of phosphorus and zinc fertiliser management on cadmium accumulation in flaxseed. J Sci Food Agric 73:307–314 Grant CA, Bailey LD (1998) Nitrogen, phosphorus and zinc management effects on grain yield and cadmium concentration in two cultivars of durum wheat. Can J Plant Sci 78:63–70 Grant CA, Bailey LD, Harapiak JT, Flore NA (2002) Effect of phosphate source, rate and cadmium content and use of Penicillium bilaii on phosphorus, zinc and cadmium concentration in durum wheat grain. J Sci Food Agric 82:301–308 Grant CA, Bailey LD, McLaughlin MJ, Singh BR (1999) Management factors which influence cadmium concentration in crops. In: McLaughlin MJ, Singh BR (eds) Cadmium in soils and plants. Kluwer, Dordrecht, pp 151–198 Grant CA, Buckley WT, Bailey LD, Selles F (1998) Cadmium accumulation in crops. Can J Plant Sci 78:1–17 Grant CA, Sheppard SC (2008) Fertilizer impacts on cadmium availability in agricultural soils and crops. Hum Ecol Risk Assess 14:210–228 Haldar M, Mandal LN (1981) Effect of phosporus and zinc on the growth and phosphorus, zinc, copper, iron and manganese nutrition of rice. Plant Soil 59:415–425 Hart JJ, Welch RM, Norvell WA, Kochian LV (2002) Transport interactions between cadmium and zinc in roots of bread and durum wheat seedlings. Physiol Plant 116:73–78 He QB, Singh BR (1993) Plant availability of cadmium in soils I. Extractable cadmium in newly and long-term cultivated soils. Acta Agric Scand B 43:134–141 Hooda PS, McNulty D, Alloway BJ, Aitken MN (1997) Plant availability of heavy metals in soils previously amended with heavy applications of sewage sludge. J Food Agric 73:446–454 Huang B, Kuo S, Bembenek R (2004) Availability of cadmium in some phosphorus fertilizers to field-grown lettuce. Water Air Soil Pollut 158:37–51 Jiao Y, Grant CA, Bailey LD (2004) Effects of phosphorus and zinc fertilizer on cadmium uptake and distribution in flax and durum wheat. J Sci Food Agric 84:777–785 Kabata-Pandias A (2000) Trace elements in soils and plants, 3rd edn. CRC, Florida Kamewada K, Nakayama M (2009) Cadmium uptake by garland chrysanthemum can be predicted from the cadmium in the soil solution, independent of soil type. Soil Sci Plant Nutr 55:441–451 Lambert R, Grant C, Sauvé S (2007) Cadmium and zinc in soil solution extracts following the application of phosphate fertilizers. Sci Total Environ 378:293–305 Levi-Minzi R, Petruzzelli G (1984) The influence of phosphate fertilizers on Cd solubility in soil. Water Air Soil Pollut 23:423–429 McBride MB (2002) Cadmium uptake by crops estimated from soil total Cd and pH. Soil Sci 167:62–67 McLaughlin MJ, Palmer LT, Tiller KG, Beech TA, Smart MK (1994) Increased soil salinity causes elevated cadmium Plant Soil (2011) 338:423–434 concentrations in field-grown potato tubers. J Environ Qual 23:1013–1018 Mclaughlin MJ, Maler NA, Freeman K, Tiller KG, Williams CMJ, Smart MK (1995) Effect of potassic and phosphatic fertilizer type, fertilizer Cd concentration and zinc rate on cadmium uptake by potatoes. Fert Res 40:63–70 McLaughlin MJ, Zarcinas BA, Stevens DP, Cook N (2000) Soil testing for heavy metals. Commun Soil Sci Plan 31:1661– 1700 Meers E, Laing GD, Unamuno V, Ruttens A, Vangronsveld J, Tack FMG, Verloo MG (2007) Comparison of cadmium extractability from soils by commonly used single extraction protocols. Geoderma 141:247–259 Miller TD (1992) Growth stages of wheat: identification and understanding improve crop management. Better Crops 76:12–17 Mortvedt JJ (1987) Cadmium levels in soils and plants from some long-term soil fertility experiments in the United States of America. J Environ Qual 16:137–142 Mortvedt JJ, Osborn G (1982) Studies on the chemical form of cadmium contaminants in phosphate fertilizers. Soil Sci 134:185–192 Nigam R, Srivastava S, Prakash S, Srivastava MM (2001) Cadmium mobilisation and plant availability—the impact of organic acids commonly exuded from roots. Plant Soil 230:107–113 Oliver DP, Hannam R, Tiller KG, Wilhelm NS, Merry RH, Cozens GD (1994) The effect of zinc fertilization on Cd concentration in wheat grain. J Environ Qual 23:705–711 Olsen SR (1972) Micronutrient interaction. In: Micronutrients in agriculture. Soil Sci Society of America Inc. Madison, Wisconsin, pp 243–264 Oporto C, Smolders E, Degryse F, Verheyen L, Vandecasteele C (2009) DGT-measured fluxes explain the chlorideenhanced cadmium uptake by plants at low but not at high Cd supply. Plant Soil 318:127–135 Richards IR, Clayton CJ, Reeve AJK (1998) Effects of longterm fertilizer phosphorus application on soil and crop phosphorus and cadmium contents. J Agric Sci 131:187– 195 Singh BR (1990) Cadmium and fluoride uptake by oats and rape fromphosphate fertilizers in two different soils: Cadmium and fluoride uptake by plants from phosphorus fertilizers. Norw J Agric Sci 4:239–250 Singh BR, Kristen M (1998) Cadmium uptake by barley as affected by Cd sources and pH levels. Geoderma 84:185– 194 Uraguchi S, Mori S, Kuramata M, Kawasaki A, Arao A, Ishikawa S (2009) Root-to-shoot Cd translocation via the xylem is the major process determining shoot and grain cadmium accumulation in rice. J Exp Bot 60:2677–2688 Verna TS, Minhas RS (1987) Zinc and phosphorus interaction in a wheat-maize cropping system. Fert Res 13:77–86 Westman RL (1990) Soil testing and plant analysis. Soil Science Society of America, Madison Zwonitzer JC, Pierzynsky GM, Hettiarachchi GM (2003) Effects of phosphorus additions on lead, cadmium, and zinc bioavailability in metal-contaminated soil. Water Air Soil Poll 143:193–209