Cadmium concentrations in New Zealand pastures: relationships to soil and climatic variables新西兰牧场中的镉浓度：与土壤及气候变量的关系

Journal of Environmental Quality

HeAvy MetAls in tHe environMent

René Reiser, Michael Simmler, Denise Portmann, Lynne Clucas, Rainer Schulin, and Brett Robinson*

Cadmium (Cd) is a nonessential element that occurs at abovebackground concentrations in many New Zealand (NZ) soils. Most of this Cd is due to the historical application of single superphosphate that was made from Nauru phosphate rock containing between 400 and 600 mg Cd kg⁻¹ P. Pasture Cd uptake exacerbates the entry of Cd into animal products. We sought to determine the critical environmental factors affecting Cd uptake in NZ pastures and to calculate the likely Cd intake of sheep and cattle. We tested 69 pastures throughout NZ for a range of variables, including Cd. Soil Cd and pasture Cd were positively correlated with soil P and soil concentrations of other elements found in phosphate fertilizers. We found that no single environmental variable adequately predicted pasture Cd uptake. Nevertheless, pseudo-total soil Cd and Cd extracted using a 0.05 mol L⁻¹ Ca(NO3)2 solution were positively correlated with pasture Cd. Although soil pH, soil Fe, and soil Cd provided an excellent predictor of the Ca(NO3)2–extractable soil Cd fraction, regression models explained just 38% of the variation of the Cd concentration in pasture grasses. Incorporating the effect of pasture species composition is a crucial next step in improving these models. A calculation of the likely exposure to Cd of sheep and cattle revealed that no pastures tested resulted in sheep and cattle ingesting Cd at a rate that would result in breaching muscle-tissue food standards. For offal products, which the NZ meat industry does not sell for human consumption, food safety standards exceedence was calculated in a few cases.

镉（Cd）是一种非必需元素，在新西兰（NZ）许多土壤中的浓度高于背景值。这些Cd大多源于历史上施用的过磷酸钙，其原料为瑙鲁磷矿石，Cd含量为400～600 mg Cd kg⁻¹ P。牧草对Cd的吸收加剧了Cd进入动物产品。本研究旨在确定影响新西兰牧场Cd吸收的关键环境因素，并估算绵羊和牛可能的Cd摄入量。我们对新西兰各地69个牧场的一系列变量（包括Cd）进行了测定。土壤Cd和牧草Cd均与土壤P以及磷肥中所含其他元素的土壤浓度呈正相关。结果表明，没有任何单一环境变量能够充分预测牧草对Cd的吸收。尽管如此，土壤准全量Cd以及0.05 mol L⁻¹ Ca(NO3)2溶液提取的Cd均与牧草Cd呈正相关。虽然土壤pH、土壤Fe和土壤Cd能够很好地预测Ca(NO3)2可提取态土壤Cd，但回归模型仅能解释牧草中Cd浓度变异的38%。纳入牧草物种组成的影响是改进这些模型的关键下一步。对绵羊和牛Cd暴露量的估算表明，所测牧场均不会导致绵羊和牛的Cd摄入速率超过肌肉组织食品标准。对于新西兰肉类行业不作为人类食用销售的内脏产品，计算结果显示少数情况下超过了食品安全标准。

Cup by plants and transferred through fodder and food

admium (Cd) is a nonessential element (Kabata- Pendias and Mukherjee, 2007) that is readily taken Cup by plants and transferred through fodder and food products into animals and humans (Cadmium Working Group, 2011a). Enrichment of Cd in soils has been reported worldwide through the land application of biosolids, industrial effluents, and phosphate fertilizers with high Cd concentrations (Hooda, 2010; Kabata-Pendias and Pendias, 2001; Kabata-Pendias and Mukherjee, 2007). Atmospheric deposition of Cd is often as significant as input via soil amendments in some highly populated and industrialized countries, and fertilizer Cd dominates in other countries with lower populations and less industrialization, such as New Zealand (NZ) (Hooda, 2010; Taylor, 2007). Several authors have reported enrichment of Cd in NZ’s pasture soils (Gray et al., 1999; Kim, 2005; Roberts et al., 1994; Schipper et al., 2011; Taylor, 1997; Taylor, 2007). Most Cd in these soils originates from mineral phosphate fertilizers (primarily single superphosphate), which may contain up to 280 mg Cd kg⁻¹ P (Loganathan et al., 2003) (Cadmium Working Group, 2011b). In a NZ soil survey, Taylor et al. (2007) reported an average Cd concentration in pasture soils of 0.43 mg kg⁻¹ (n = 825). This is more than double the background concentration of 0.16 mg kg⁻¹ (n = 372) (Taylor, 2007).

镉（Cd）是一种非必需元素（Kabata- Pendias and Mukherjee, 2007），易被植物吸收，并经饲料和食品转移至动物和人体（Cadmium Working Group, 2011a）。世界各地均有报道，由于土地施用生物固体（污泥）、工业废水以及高Cd含量的磷肥，土壤中Cd发生富集（Hooda, 2010; Kabata-Pendias and Pendias, 2001; Kabata-Pendias and Mukherjee, 2007）。在一些人口稠密、工业化程度高的国家，大气Cd沉降往往与土壤改良剂带入的Cd同等重要；而在人口较少、工业化程度较低的国家（如新西兰（NZ）），肥料来源的Cd占主导地位（Hooda, 2010; Taylor, 2007）。多位作者报道了新西兰牧场土壤中的Cd富集（Gray et al., 1999; Kim, 2005; Roberts et al., 1994; Schipper et al., 2011; Taylor, 1997; Taylor, 2007）。这些土壤中的Cd大多来源于矿物磷肥（主要是过磷酸钙），其Cd含量可高达280 mg Cd kg⁻¹ P（Loganathan et al., 2003）（Cadmium Working Group, 2011b）。在一项新西兰土壤调查中，Taylor et al. (2007) 报道牧场土壤的平均Cd浓度为0.43 mg kg⁻¹（n = 825），是背景浓度0.16 mg kg⁻¹（n = 372）的两倍以上（Taylor, 2007）。

Pastureland is the main type of farming in NZ, covering 13.8 million hectares (Loganathan et al., 2003). Grazing animals take up Cd mainly via consumption of herbage; intake through soil ingestion is generally lower (Lee et al., 1996; Loganathan et al., 2003). Food safety standards have been set to limit the concentration of Cd in foodstuffs. Exceeding such standards might result in trade restrictions and serious economic damage for producers (Cadmium Working Group, 2011a). The most important factor affecting food-chain transfer is soil–plant uptake (Hooda, 2010). The bioavailability of Cd in soil depends on its (total) concentration, pH, organic matter content, clay content, soil moisture conditions, and availability of macro- and micronutrients (Grant et al., 1999; Welch and Norvell, 1999). The use of a single soil-to-plant transfer coefficient for a given

牧场是新西兰主要的农业类型，面积达1380万公顷（Loganathan et al., 2003）。放牧动物主要通过采食牧草摄入Cd；通过摄食土壤摄入的量一般较低（Lee et al., 1996; Loganathan et al., 2003）。为限制食品中的Cd浓度，已制定了相应的食品安全标准。超过这些标准可能导致贸易限制，并给生产者造成严重经济损失（Cadmium Working Group, 2011a）。影响食物链转移的最重要因素是土壤–植物吸收（Hooda, 2010）。土壤中Cd的生物有效性取决于其（全量）浓度、pH、有机质含量、黏粒含量、土壤水分状况以及大量和微量养分的有效性（Grant et al., 1999; Welch and Norvell, 1999）。对某一给定

Copyright © American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America. 5585 Guilford Rd., Madison, WI 53711 USA. All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.

版权所有 © 美国农学会、美国作物科学学会和美国土壤科学学会。5585 Guilford Rd., Madison, WI 53711 USA。保留所有权利。未经出版者书面许可，不得以任何形式或任何手段（电子或机械手段，包括影印、录音或任何信息存储与检索系统）复制或传播本期刊的任何部分。

J. Environ. Qual. 43:917–925 (2014) doi:10.2134/jeq2013.09.0367 Supplemental data file is available online for this article. Received 17 Sept. 2013.

*Corresponding author (brett.robinson@lincoln.ac.nz).

R. Reiser and D. Portmann, Agroscope, Reckenholzstrasse 191, CH-8046 Zürich; M. Simmler, Institute of Biogeochemistry and Pollutant Dynamics, ETH Zürich, CHN F 71, Universitätstrasse 16, 8092 Zürich; R. Schulin, ETH Zürich, Institute for Terrestrial Ecosystems, CHN F 31.1, Universitätstrasse 16, 8092 Zürich; L. Clucas and B. Robinson, Dep. of Soil and Physical Sciences, B222, Lincoln 7647, New Zealand. Assigned to Associate Editor Cynthia Grant.

R. Reiser和D. Portmann，Agroscope, Reckenholzstrasse 191, CH-8046 Zürich；M. Simmler，Institute of Biogeochemistry and Pollutant Dynamics, ETH Zürich, CHN F 71, Universitätstrasse 16, 8092 Zürich；R. Schulin，ETH Zürich, Institute for Terrestrial Ecosystems, CHN F 31.1, Universitätstrasse 16, 8092 Zürich；L. Clucas和B. Robinson，Dep. of Soil and Physical Sciences, B222, Lincoln 7647, New Zealand。责任副主编：Cynthia Grant。

Abbreviations: ADICd, acceptable daily intake for cadmium; BAF, bioaccumulation factor; DICd, daily intake of cadmium; dw, dry weight; NZ, New Zealand.

plant species is not appropriate to describe plant Cd uptake in a wide range of soils (Rodrigues et al., 2012; Römkens et al., 2009).

植物物种采用单一的土壤–植物转移系数，并不适合描述植物在多种土壤中对Cd的吸收（Rodrigues et al., 2012; Römkens et al., 2009）。

Hooda (2010) suggested that Zn and Cd should be considered together to understand either of them in detail. The two elements are often cocontaminants (the typical Zn:Cd ratio is 200:1), have many similar properties in soil and plants, and are readily absorbed and translocated to plant shoots (Hooda, 2010). Synergistic and antagonistic effects of Zn on Cd uptake by plants are known, with antagonistic effects usually observed under non–Zn deficiency conditions (Grant et al., 1999). Whereas Zn is often deficient in agricultural systems, Cd is often a food-chain toxicity risk (Hooda, 2010).

Hooda (2010) 建议应将Zn和Cd结合起来考虑，以便深入理解其中任一元素。这两种元素常常是共存污染物（典型Zn:Cd比为200:1），在土壤和植物中具有许多相似性质，且易被吸收并转运至植物地上部（Hooda, 2010）。Zn对植物吸收Cd存在协同和拮抗作用，在不缺Zn的条件下通常观察到拮抗作用（Grant et al., 1999）。农业系统中常常缺Zn，而Cd则常常构成食物链毒性风险（Hooda, 2010）。

Mechanistic models capable of predicting soil–plant transfer have not been developed (Swartjes, 2011). Various empirical models have been proposed to describe Cd accumulation by plants as a function of soil properties (e.g., pH, total Cd, Cd that is extractable using a dilute salt solution and soil organic matter content) (Smolders, 2001; Tudoreanu and Phillips, 2004). Empirical models are more limited in scope than mechanistic models; however, because they are based on observations, they provide a degree of accuracy not necessarily present in mechanistic models (Tudoreanu and Phillips, 2004). Empirical models are only useful in specific environments within the range of their calibration. They must be validated or recalibrated with local soil and environmental conditions (Swartjes, 2011). Little information about the proposed model variables is known other than under experimentspecific conditions. There is a lack of studies investigating the large number of environmental and plant factors suspected to influence Cd uptake under field conditions.

目前尚未建立能够预测土壤–植物转移的机理模型（Swartjes, 2011）。已有多种经验模型被提出，用以将植物Cd累积描述为土壤性质（如pH、全量Cd、稀盐溶液可提取态Cd及土壤有机质含量）的函数（Smolders, 2001; Tudoreanu and Phillips, 2004）。经验模型的适用范围较机理模型更为有限；但由于其基于观测数据，因此具有机理模型未必具备的一定精度（Tudoreanu and Phillips, 2004）。经验模型仅在其校准范围内的特定环境中有效，必须结合当地土壤和环境条件进行验证或重新校准（Swartjes, 2011）。除特定试验条件外，对所提出的模型变量知之甚少。目前缺乏在田间条件下对大量可能影响Cd吸收的环境因素和植物因素开展的研究。

We aimed to elucidate the most important environmental and edaphic factors affecting Cd concentrations in NZ pastures with a view to identifying factors that could be measured along with soil Cd to predict the risk of plant Cd uptake. Furthermore, we sought to use the model of Rodrigues et al. (2012) to assess the daily intake of Cd by sheep and cattle on various pastures.

本研究旨在阐明影响新西兰牧场Cd浓度的最重要环境因素和土壤因素，以期找出可与土壤Cd一同测定、用于预测植物Cd吸收风险的因子。此外，我们还尝试采用Rodrigues et al. (2012) 的模型评估不同牧场上绵羊和牛的Cd日摄入量。

1Materials and Methods材料与方法

Soil and Pasture Sampling土壤与牧草采样

Soil and pasture samples from 69 sites were collected during the NZ summer (Dec. 2009 to Feb. 2010). The sampling locations, given in Supplemental Table S1, were determined using a GPS receiver (accuracy, 8–10 m). Most of the pastures were intensive cattle or dairy farms, which typically receive high rates of P fertilization. Pastures were in the vegetative growth phase, and a visual inspection indicated that they contained minimal senescent material. The median total annual rainfall at the collection sites was 1838 mm (range, 424–5406 mm). The median average annual air temperature ranged from 8.6 to 16.4°C. The range of rainfall and temperature characteristics reflects the geographical spread and difference in elevation. Pasture samples were cut 2 to 3 cm above ground on an area of approximate 0.2 m². Nongraminaceous weeds were removed from samples to minimize the effect of species composition in the study. The composition of graminaceous species was typical for dairy farm pastures in NZ, dominated by perennial ryegrass [Lolium perenne (L.)], with some pastures containing small amounts of brown top [Agrostis capillaris (L.)]. Roberts and

于新西兰夏季（2009年12月至2010年2月）从69个样点采集土壤和牧草样品。采样位置（见补充表S1）采用GPS接收器确定（精度8～10 m）。大多数牧场为集约化肉牛或奶牛农场，通常施用较高量的磷肥。牧草处于营养生长期，目视检查表明其中衰老组织极少。采样点年降水总量中位数为1838 mm（范围424～5406 mm）。年平均气温中位数为8.6～16.4°C。降水和气温特征的变化范围反映了样点的地理分布和海拔差异。在约0.2 m²的面积上于离地2～3 cm处刈割牧草样品。为尽量减少物种组成对研究的影响，去除了样品中的非禾本科杂草。禾本科物种组成为新西兰奶牛牧场的典型组成，以多年生黑麦草［Lolium perenne (L.)］为主，部分牧场含少量细弱剪股颖［Agrostis capillaris (L.)］。Roberts and

Longhurst (2002) reported that a typical NZ pasture contains approximately 70% ryegrass, 20% white clover, and 10% weeds. A composite sample of the underlying topsoil was obtained by means of a soil corer (10–15 cores per sampling location taken from a depth of 0–10 cm).

Longhurst (2002) 报道，典型的新西兰牧场约含70%黑麦草、20%白三叶草和10%杂草。采用土钻采集下层表土混合样（每个采样点取10～15钻，深度0～10 cm）。

Chemical Analyses化学分析

Plant samples were washed in deionized water and dried at 60°C until a constant weight was obtained. Samples were milled using a Cyclotech type 1093 cyclone grinder with an aluminum rotor. Plant material (0.5 g) was digested in 5 mL HNO3. The digests were diluted with Milli Q (EASYpure RF, 18.3 MW-cm, Barnstead) to a volume of 25 mL and filtered with Whatman 52 filter paper (pore size, 7 mm).

植物样品用去离子水洗净，于60°C烘干至恒重。样品用配有铝制转子的Cyclotech 1093型旋风研磨机粉碎。取0.5 g植物样品用5 mL HNO3消解。消解液用Milli Q水（EASYpure RF, 18.3 MW-cm, Barnstead）定容至25 mL，并用Whatman 52滤纸（孔径7 mm）过滤。

Soil samples were dried for 7 d at room temperature. The aggregates were separated and passed through a 2-mm nylon sieve. Soil pH was measured in Milli Q at a solid:water ratio of 1:2.5 using a Metler Toledo pH meter. The grain size distribution of the soils was determined using the pipette method according to the reference methods of the Swiss federal agricultural research stations (FAL et al., 1996).

土壤样品在室温下风干7 d。将团聚体分散后过2 mm尼龙筛。采用Metler Toledo pH计在土水比1:2.5条件下以Milli Q水测定土壤pH。土壤粒径分布采用吸管法，按照瑞士联邦农业研究站的参考方法测定（FAL et al., 1996）。

Pseudo-total element concentrations in soil (henceforth referred to as “total”) were measured in acid digests. Half a gram of soil was digested in 5 mL aqua regia (BDH Aristar nitric acid 69%, BDH Aristar hydrochloric acid 37%). Extractable elemental concentrations in soil were measured by weighing 5 g of soil into 40-mL centrifuge tubes and adding 30 mL 0.05 mol L⁻¹ Ca(NO3)2 (BDH AnalaR [Ca(NO3)2×4H2O]). Tubes were agitated for 2 h on an end-over-end shaker, centrifuged for 10 min at 15,000 rpm, and filtered using Whatman 52 filter paper. Oxalate-extractable concentrations were measured according to the method described elsewhere (Blakemore et al., 1987). All extracts were stored at 4°C before the analysis.

土壤中元素准全量浓度（以下称“全量”）在酸消解液中测定。取0.5 g土壤用5 mL王水（BDH Aristar 69%硝酸，BDH Aristar 37%盐酸）消解。土壤可提取态元素浓度的测定方法为：称取5 g土壤于40 mL离心管中，加入30 mL 0.05 mol L⁻¹ Ca(NO3)2（BDH AnalaR [Ca(NO3)2×4H2O]）。离心管在翻转振荡器上振荡2 h，以15,000 rpm离心10 min，并用Whatman 52滤纸过滤。草酸盐可提取态浓度按文献所述方法测定（Blakemore et al., 1987）。所有提取液分析前均于4°C保存。

Element concentrations were measured using ICP–OES (Varian 720 ES). Wageningen (ISE 921, IPE 100) and NIST (1573a) reference materials were analyzed for quality assurance. The results of these analyses are given in Table 1. The C and N content of soil and plant material was determined using an Elementar Vario MAX CN element analyzer.

元素浓度采用ICP–OES（Varian 720 ES）测定。分析Wageningen（ISE 921, IPE 100）和NIST（1573a）标准参考物质以进行质量保证，其结果见表1。土壤和植物样品的C、N含量采用Elementar Vario MAX CN元素分析仪测定。

Correction for Cadmium Arising from Surface Contamination表面污染所致镉的校正

Because sample washing cannot remove all soil particles from field-collected plant material (Robinson et al., 2009), the plant concentrations were corrected for soil contamination using a method described by Robinson et al. (2008). Iron was used as the indicator element with the minimal measured value in plant of 44.4 mg kg⁻¹ dry weight (dw) as baseline value. The masses of soil calculated using Fe as an indicator were in reasonable agreement with the corresponding masses when Al or Cr was used as indicator element (R = 0.68; R = 0.76 [Pearson]).

由于样品清洗无法去除田间采集植物样品上的全部土壤颗粒（Robinson et al., 2009），采用Robinson et al. (2008) 所述方法对植物浓度进行了土壤污染校正。以Fe为指示元素，以植物中实测最低值44.4 mg kg⁻¹干重（dw）作为基线值。以Fe为指示元素计算得到的土壤质量，与以Al或Cr为指示元素时的相应质量具有较好的一致性（R = 0.68；R = 0.76［Pearson］）。

Geographic Information System Analysis地理信息系统分析

Climate data (raster dataset with 500 m spatial resolution) representing medians of the period 1971 to 2000 were obtained from the National Institute of Water and Atmospheric Research (Tait and Zheng, 2007). The value of the raster cell overlying the sampling point was ascribed to that sample using the ESRI ArcGIS 9.3 software. The average annual solar radiation is henceforth referred to as radiation, the total annual rainfall as rainfall, and the average annual air temperature as temperature.

气候数据（空间分辨率500 m的栅格数据集）代表1971—2000年期间的中位数，来源于新西兰国家水与大气研究所（Tait and Zheng, 2007）。利用ESRI ArcGIS 9.3软件，将覆盖采样点的栅格单元值赋予相应样品。以下将年平均太阳辐射简称为辐射，年降水总量简称为降水，年平均气温简称为气温。

Statistical Analysis统计分析

Correlations between variables were described by Spearman rank correlation coefficients. To identify the best linear regression models, an exhaustive search (i.e., an all-subset regression) across all predictor variables was conducted. The best model for each number of explanatory variables was selected given that all estimated coefficients were significant (a = 0.05, F-test using type III sums of square). Variables showing positively skewed frequency distributions [organic C, total Mn, total S, oxalate-extractable Zn (ox-Zn), oxalate-extractable Cd (ox-Cd), Ca(NO3)2–Zn, Ca(NO3)2–Cd, rainfall, pasture N, pasture Zn, and pasture Cd] were log-transformed before regression analysis. The relative importance metric LMG (Lindemann, Merenda, and Gold) was calculated to describe the importance of explanatory variables (Lindeman et al., 1980). Statistical analyses were undertaken with the software R (R-Core-Team, 2012) and Excel 2007. For correlation analysis, the R package Hmisc (Harrell, 2012) was used. For model selection and assessment of the importance of model variables, the R packages leaps (Lumley, 2009) and relaimpo (Grömping, 2006) were used.

变量间的相关性采用Spearman秩相关系数描述。为确定最佳线性回归模型，对全部预测变量进行了穷举搜索（即全子集回归）。在所有估计系数均显著（a = 0.05，采用III型平方和的F检验）的前提下，为每一解释变量个数选取最佳模型。呈正偏态频率分布的变量［有机C、全Mn、全S、草酸盐可提取态Zn（ox-Zn）、草酸盐可提取态Cd（ox-Cd）、Ca(NO3)2–Zn、Ca(NO3)2–Cd、降水、牧草N、牧草Zn和牧草Cd］在回归分析前进行对数转换。计算相对重要性指标LMG（Lindemann, Merenda, and Gold）以描述各解释变量的重要性（Lindeman et al., 1980）。统计分析采用R软件（R-Core-Team, 2012）和Excel 2007完成。相关分析使用R软件包Hmisc（Harrell, 2012）。模型选择和模型变量重要性评估使用R软件包leaps（Lumley, 2009）和relaimpo（Grömping, 2006）。

2Results and Discussion结果与讨论

Total Cadmium and Zinc Concentrations in Soil土壤中全量镉和锌浓度

Table 2 shows that the Cd concentration in topsoil ranged from 0.25 to 1.58 mg kg⁻¹ (mean, 0.89 mg kg⁻¹). Earlier NZ surveys reported mean Cd concentrations of 0.4 mg kg⁻¹ for pasture soils (Roberts et al., 1994; Taylor, 2007). The discrepancy in our results may be because most of our samples were taken from farmland that was under intensive dairy farming, which receives higher fertilizer inputs than the extensive grazing systems included in the aforementioned studies. Comparing them with the mean background concentration of 0.16 mg kg⁻¹ (Taylor, 2007), our values are consistent with the frequently reported accumulation of Cd in NZ soils due to anthropogenic inputs.表2显示，表土中Cd浓度为0.25～1.58 mg kg⁻¹（平均0.89 mg kg⁻¹）。早期的新西兰调查报道牧场土壤平均Cd浓度为0.4 mg kg⁻¹（Roberts et al., 1994; Taylor, 2007）。本研究结果的差异可能是由于大部分样品采自集约化奶牛养殖农田，其肥料投入高于上述研究中所包含的粗放放牧系统。与平均背景浓度0.16 mg kg⁻¹（Taylor, 2007）相比，本研究数值与频繁报道的新西兰土壤因人为输入而导致Cd累积的结论一致。

The clay contents ranged from 3 to 40% (mean, 23%). These values are typical values for relatively young soils (Scheffer et al., 2010), as are the soils found in the NZ landscape. Total soil Cd was positively correlated with clay content, total Fe, Mn, Al, P,

黏粒含量为3%～40%（平均23%），这是相对年轻土壤的典型数值（Scheffer et al., 2010），新西兰地貌中的土壤即属此类。土壤全量Cd与黏粒含量以及全Fe、Mn、Al、P、

Ca, and Zn (Table 3). Total Cd was negatively correlated with sand content and site temperature. The positive correlations with clay content and total concentrations of Fe, Mn, and Al are consistent with the binding of Cd by metal oxides and clay minerals (Scheffer et al., 2010).

Ca和Zn呈正相关（表3）。全量Cd与砂粒含量及样点气温呈负相关。与黏粒含量以及Fe、Mn和Al全量的正相关关系，与金属氧化物和黏土矿物对Cd的结合作用相一致（Scheffer et al., 2010）。

That Cd was positively correlated with P and Ca is consistent with the numerous reports that these elements coaccumulate due to fertilizer use (Loganathan et al., 2003). Despite the presence of 11% S in single superphosphate, the total Cd concentration did not correlate with the total concentration of S in soil (Table 3). This is not surprising because S as sulfate leaches from NZ soils, only being retained in significant quantities as organic S (Edmeades et al., 2005; Ghani et al., 1991).

Cd与P和Ca呈正相关，这与大量关于这些元素因施肥而共同累积的报道一致（Loganathan et al., 2003）。尽管过磷酸钙中含有11%的S，但土壤全量Cd浓度与全S浓度并无相关性（表3）。这并不意外，因为硫酸盐形态的S易从新西兰土壤中淋溶，仅以有机S形态大量保留（Edmeades et al., 2005; Ghani et al., 1991）。

There was a positive correlation between Zn and P and Ca in addition to that between Cd and Zn, indicating that fertilizer was an important source of Zn. Like Cd, Zn occurs as impurity in phosphate rocks used for the production of superphosphate fertilizers (Syers et al., 1986). McGrath and Tunney (2010) observed a significant increase in soil Zn (on average, 8.19 mg kg⁻¹ in the 0- to 20-cm soil layer) in a grazed pasture after 31 yr of superphosphate addition (930 kg kg⁻¹ over 31 yr) compared with a control pasture receiving no superphosphate (McGrath and Tunney, 2010). For many years, phosphate rock from Nauru Islands, known to contain high concentrations of Cd (100 mg kg⁻¹) and Zn (1100 mg kg⁻¹), was used for the production of superphosphate applied to New Zealand and Australian soils (Cadmium Working Group, 2008). In a survey of Iranian arable soils used for wheat production, Karami et al. (2009) found a correlation between total Zn and available P (Olsen P). They concluded that superphosphate fertilizer with high Zn (100–1100 mg kg⁻¹), which is commonly applied in Iran, is an important source of Zn in these soils. Zinc might also accumulate in NZ pasturelands because it is used as a prophylactic against facial eczema in sheep and cattle (Anderson et al., 2012). Therefore, another explanation for the observed correlations may be that pastures that received higher rates of phosphate fertilizer also received higher rates of zinc as a result of the prophylactic facial eczema treatment.

除Cd与Zn之间存在正相关外，Zn与P和Ca之间也呈正相关，表明肥料是Zn的重要来源。与Cd一样，Zn也作为杂质存在于生产过磷酸钙肥料所用的磷矿石中（Syers et al., 1986）。McGrath and Tunney (2010) 观察到，与未施过磷酸钙的对照牧场相比，连续31年施用过磷酸钙（31年共930 kg kg⁻¹）的放牧牧场土壤Zn显著增加（0～20 cm土层平均增加8.19 mg kg⁻¹）（McGrath and Tunney, 2010）。多年来，已知含高浓度Cd（100 mg kg⁻¹）和Zn（1100 mg kg⁻¹）的瑙鲁群岛磷矿石被用于生产施用于新西兰和澳大利亚土壤的过磷酸钙（Cadmium Working Group, 2008）。在一项对伊朗小麦耕地土壤的调查中，Karami et al. (2009) 发现全量Zn与有效P（Olsen P）之间存在相关性。他们认为，伊朗普遍施用的高Zn（100～1100 mg kg⁻¹）过磷酸钙肥料是这些土壤中Zn的重要来源。由于Zn被用于预防绵羊和牛的面部湿疹（Anderson et al., 2012），Zn也可能在新西兰牧场中累积。因此，对所观察到的相关性的另一种解释可能是：施用较高量磷肥的牧场，由于面部湿疹预防处理，同时也接受了较高量的锌。

Ca(NO3)2–Extractable Cadmium and Zinc ConcentrationsCa(NO3)2可提取态镉和锌浓度

The Ca(NO3)2–extractable fractions of soil Cd and Zn have been reported to be good predictors of plant Cd and Zn uptake in pasture soils (Black et al., 2011; Black et al., 2012). Table 2 shows the Ca(NO3)2–extractable fractions of Cd and Zn in the soils tested. The extractable concentrations were approximately three orders of magnitude lower than the total concentrations, consistent with observations that these elements are immobile in weakly acidic to alkaline soils (Loganathan and Hedley, 1997). Mean K_{D values (total/extractable concentration} coefficients) for Cd and Zn were approximately 3000 and 6000, respectively.

据报道，土壤Cd和Zn的Ca(NO3)2可提取态是牧场土壤中植物吸收Cd和Zn的良好预测指标（Black et al., 2011; Black et al., 2012）。表2列出了所测土壤中Cd和Zn的Ca(NO3)2可提取态含量。可提取态浓度比全量浓度约低三个数量级，这与这些元素在弱酸性至碱性土壤中不易移动的观测结果一致（Loganathan and Hedley, 1997）。Cd和Zn的平均K_{D值（全量/可提取态浓度}系数）分别约为3000和6000。

All sampled soils had pH values below 7.0 (mean, 5.5) (Table 2). Calcium nitrate–cadmium was negatively correlated with pH (Table 3). Variable charged binding sites are increasingly protonated at low pH and unavailable for binding Cd. In addition, there is increased competition with acidic metal cations (e.g., Al³⁺) for binding sites under acidic conditions (Loganathan et al., 2012; Scheffer et al., 2010). Most of the sampled soils probably had a history of liming because acidity limits fertility, particularly N availability, on many NZ farms (Carran, 1992). There was a negative correlation between Ca(NO3)2–Cd and total Ca. This is likely due to an indirect effect related to the pH increase as a result of liming. The positive correlation of total Ca with pH (Supplemental Table S2) is consistent with this rationale. The positive correlation between Ca(NO3)2–Cd and rainfall is also suspected to be related to soil pH because higher rainfall was negatively correlated with pH (Supplemental Table S2). Basic cations tend to leach with the percolating rainwater, and the moist soil conditions allow increased biological activity, which produces acidity (Sparks, 2003). The negative correlation between Ca(NO3)2–Cd and the clay fraction and total Al is consistent with the aforementioned likely importance of Al-bearing minerals of the clay fraction in binding of Cd. The correlation with Ca(NO3)2–Zn reflects the geochemically similar behavior of Cd and Zn (Hooda, 2010). The Ca(NO3)2–Cd concentration was governed by factors influencing the solubility rather than by the total soil Cd, as there was no significant correlation with total Cd.

所有采样土壤的pH值均低于7.0（平均5.5）（表2）。Ca(NO3)2–Cd与pH呈负相关（表3）。在低pH条件下，可变电荷结合位点日益质子化，无法结合Cd。此外，在酸性条件下，酸性金属阳离子（如Al³⁺）对结合位点的竞争加剧（Loganathan et al., 2012; Scheffer et al., 2010）。由于酸度限制了新西兰许多农场的肥力（尤其是N的有效性），大多数采样土壤可能有施用石灰的历史（Carran, 1992）。Ca(NO3)2–Cd与全Ca呈负相关，这可能是施用石灰导致pH升高所产生的间接效应。全Ca与pH之间的正相关（补充表S2）与这一推论一致。Ca(NO3)2–Cd与降水之间的正相关也可能与土壤pH有关，因为降水越多，pH越低（补充表S2）。盐基阳离子易随下渗雨水淋溶，而湿润的土壤条件使生物活性增强，从而产生酸度（Sparks, 2003）。Ca(NO3)2–Cd与黏粒组分及全Al呈负相关，这与前述黏粒组分中含Al矿物在结合Cd方面可能具有的重要作用一致。与Ca(NO3)2–Zn的相关性反映了Cd和Zn相似的地球化学行为（Hooda, 2010）。由于Ca(NO3)2–Cd浓度与全量Cd无显著相关，其浓度受影响溶解度的因素控制，而非受土壤全量Cd控制。

Tables 4 and 5 show regression models that use edaphic and climate variables to calculate Ca(NO3)2–extractable concentrations of soil Cd and Zn. In each model, between one and six environmental variables are used, with successive variables increasing the explanatory power (R²_adj) but decreasing the ease of parameterization of the model. For Cd and Zn, pH was the most important soil parameter in explaining the Ca(NO3)2–extractable concentrations. Incorporating organic C, total Fe, total Cd, ox-Cd, and rainfall, linear regression explained 86% of the variation in Ca(NO3)2–Cd. Total Fe is incorporated with a negative sign, consistent with Fe oxides being important adsorbents of Cd. Because organic carbon is known to bind Cd, the incorporation of organic C with a negative sign is unsurprising. The organic carbon contents where medium to high (range, 2.0–15.9 wt%; median, 5.5 wt%). This is characteristic for NZ soils, which typically have higher organic matter contents than soils from arid countries (Cadmium Working Group, 2008). Compared with total Fe, the relative importance of organic C as explanatory variable was minor. Linear regression models to predict Ca(NO3)2– Zn differed slightly from Cd, with total P, total Mn, and temperature in the set of explanatory variables (Table 5).

表4和表5列出了利用土壤和气候变量计算土壤Cd和Zn的Ca(NO3)2可提取态浓度的回归模型。每个模型采用1～6个环境变量，依次增加的变量提高了解释力（R²_adj），但降低了模型参数化的简便性。对于Cd和Zn，pH均是解释Ca(NO3)2可提取态浓度最重要的土壤参数。纳入有机C、全Fe、全量Cd、ox-Cd和降水后，线性回归可解释Ca(NO3)2–Cd变异的86%。全Fe以负号纳入，与Fe氧化物是Cd的重要吸附剂相一致。由于已知有机碳能结合Cd，有机C以负号纳入并不意外。有机碳含量为中等至较高水平（范围2.0～15.9 wt%；中位数5.5 wt%），这是新西兰土壤的特征，其有机质含量通常高于干旱国家的土壤（Cadmium Working Group, 2008）。与全Fe相比，有机C作为解释变量的相对重要性较小。预测Ca(NO3)2–Zn的线性回归模型与Cd略有不同，其解释变量中包括全P、全Mn和气温（表5）。

† Standard deviation range.

† 标准差范围。

Pasture Cadmium and Zinc Concentrations牧草镉和锌浓度

Most (90%) pasture samples had Cd concentrations <0.2 mg kg⁻¹. The median and mean were 0.08 and 0.13 mg kg⁻¹, respectively (Table 2). Seven samples had concentrations >0.2 mg kg⁻¹, with the highest sample having a Cd concentration of 1.55 mg kg⁻¹. A previous NZ survey reported mean Cd concentrations in pasture grass of 0.1 mg kg⁻¹. However, the pasture soil surveyed by Roberts et al. (1994) and most probably the species composition in this work are not comparable to our study. Although there can be significant differences in Cd uptake among various plant species and cultivars (Crews and Davies, 1985; McLaughlin et al., 1994), there is a lack of information on variations in Cd uptake between various pasture species. The Zn concentrations ranged from 9 to 89 mg kg⁻¹, with a mean of 28 mg kg⁻¹. In terms of soil-to-plant transfer coefficient (plant/soil concentration quotient), the median values for Cd and Zn were 0.09 and 0.42, respectively. Similarly the median plant/Ca(NO3)2–extractable quotients for Cd and Zn were 157 and 1352, respectively. This indicates that pasture takes up Zn, an essential micronutrient, more effectively than Cd, a nonessential heavy metal. This is probably due to the presence of Zn-specific transporters that are adapted to take up this essential element (Reid and Hayes, 2003).

大多数（90%）牧草样品的Cd浓度<0.2 mg kg⁻¹，中位数和平均值分别为0.08和0.13 mg kg⁻¹（表2）。7个样品浓度>0.2 mg kg⁻¹，最高样品的Cd浓度为1.55 mg kg⁻¹。此前一项新西兰调查报道牧草中平均Cd浓度为0.1 mg kg⁻¹。然而，Roberts et al. (1994) 所调查的牧场土壤以及该研究中很可能存在的物种组成与本研究并不具有可比性。尽管不同植物物种和品种之间Cd吸收可能存在显著差异（Crews and Davies, 1985; McLaughlin et al., 1994），但目前缺乏关于不同牧草物种间Cd吸收差异的资料。Zn浓度为9～89 mg kg⁻¹，平均28 mg kg⁻¹。就土壤–植物转移系数（植物/土壤浓度商）而言，Cd和Zn的中位数分别为0.09和0.42。同样，Cd和Zn的植物/Ca(NO3)2可提取态浓度商中位数分别为157和1352。这表明牧草对必需微量养分Zn的吸收比对非必需重金属Cd的吸收更为有效。这可能是由于存在适应于吸收这一必需元素的Zn特异性转运蛋白（Reid and Hayes, 2003）。

The concentration of Cd in pasture was positively correlated with the total soil concentrations of P and Cd, the ox-Cd, the Ca(NO3)2–Cd, rainfall, and pasture concentrations of P, S, Ca, and Zn (Table 3). The correlation with Ca(NO3)2–extractable

牧草中Cd浓度与土壤全P和全量Cd浓度、ox-Cd、Ca(NO3)2–Cd、降水以及牧草中P、S、Ca和Zn浓度呈正相关（表3）。与Ca(NO3)2可提取态

Cd (R = 0.41) was stronger than with ox-Cd (R = 0.36) and total Cd (R = 0.31). Thus, of the three Cd fractions, Ca(NO3)2– extractable Cd explained the bioavailable Cd pool best, in line with the findings of Black et al. (2012).

Cd的相关性（R = 0.41）强于与ox-Cd（R = 0.36）和全量Cd（R = 0.31）的相关性。因此，在三种Cd形态中，Ca(NO3)2可提取态Cd对生物有效性Cd库的解释最好，这与Black et al. (2012) 的研究结果一致。

Total Cd and Ca(NO3)2–extractable Cd in soil correlated significantly with pasture Cd, whereas total Cd and Ca(NO3)2– extractable Cd did not correlate with each other. This may imply that the readily soluble fraction of Cd in soil was not the only source for pasture Cd (i.e., plants had access to stronger bound Cd, possibly through Cd mobilized by root exudates). Analyses of pasture throughout the growing season may reveal a different relationship between total Cd and plant Cd due to physiological changes during the plants’ life cycles.

土壤全量Cd和Ca(NO3)2可提取态Cd均与牧草Cd显著相关，而全量Cd与Ca(NO3)2可提取态Cd彼此之间并不相关。这可能意味着土壤中易溶态Cd并非牧草Cd的唯一来源（即植物能够获取结合较牢固的Cd，可能是通过根系分泌物活化的Cd）。在整个生长季对牧草进行分析，可能由于植物生命周期中的生理变化而揭示全量Cd与植物Cd之间不同的关系。

The uptake and root-to-shoot translocation of micronutrients by plants can be affected by interactions among these elements (Nan et al., 2002a) and vary with the plant’s nutrient status (Robinson et al., 2009). Synergistic and antagonistic effects have been reported between Zn and Cd uptake (Kabata-Pendias and Pendias, 2001). Although our results do not reveal any mechanisms of Cd–Zn interactions, they are not consistent with a strong antagonistic interaction between Cd and Zn. The pasture concentrations of the two metals were positively correlated (R = 0.39) but were not correlated with the soil concentrations (total, Ca(NO3)2– extractable) of each other. Compared with Zn, which is often reported to behave similar to Cd in the soil–plant system (Hooda, 2010), Cd was less dependent on the plant nutrient status but more strongly related to the total soil Cd concentration and soil factors affecting Cd solubility. Pasture Cd concentration was only weakly correlated with pasture P (R = 0.29) and S (R = 0.24), was not correlated with pasture N, but was strongly correlated with total Cd, Ca(NO3)2–Cd and ox-Cd in soil. In contrast, pasture Zn concentration strongly correlated with pasture N (R = 0.63), P (R = 0.50), and S (R = 0.55) but did not correlate with total Zn, Ca(NO3)2– extractable Zn, or ox-Zn. Similar observations were made by other researchers. Nan et al. (2002b) used stepwise regression analysis to produce a meaningful model relating wheat grain Cd concentration with the explanatory variables total Cd in soil, pH, and available P. Starting with the same set of potential explanatory variables (pH, OM, available P, and total soil metal concentrations), they found no meaningful multiple regression model for grain Zn concentration of wheat. They suggested that, in contrast to Cd, under field conditions the uptake of Zn was largely affected by the crop’s physiological characteristics.

植物对微量养分的吸收及其从根部向地上部的转运会受到这些元素间相互作用的影响（Nan et al., 2002a），并随植物养分状况而变化（Robinson et al., 2009）。已有报道指出Zn与Cd吸收之间存在协同和拮抗作用（Kabata-Pendias and Pendias, 2001）。虽然本研究结果未揭示Cd–Zn相互作用的任何机制，但与Cd和Zn之间存在强拮抗作用的观点并不一致。两种金属在牧草中的浓度呈正相关（R = 0.39），但与对方的土壤浓度（全量、Ca(NO3)2可提取态）均不相关。Zn在土壤–植物系统中的行为常被报道与Cd相似（Hooda, 2010），但与Zn相比，Cd较少依赖于植物养分状况，而与土壤全量Cd浓度及影响Cd溶解度的土壤因素关系更为密切。牧草Cd浓度与牧草P（R = 0.29）和S（R = 0.24）仅呈弱相关，与牧草N不相关，但与土壤全量Cd、Ca(NO3)2–Cd和ox-Cd强相关。相反，牧草Zn浓度与牧草N（R = 0.63）、P（R = 0.50）和S（R = 0.55）强相关，但与全量Zn、Ca(NO3)2可提取态Zn或ox-Zn均不相关。其他研究者也有类似发现。Nan et al. (2002b) 采用逐步回归分析，建立了将小麦籽粒Cd浓度与解释变量土壤全量Cd、pH和有效P相关联的有意义模型。以相同的潜在解释变量（pH、OM、有效P和土壤金属全量浓度）为起点，他们未能找到小麦籽粒Zn浓度的有意义多元回归模型。他们认为，与Cd不同，在田间条件下Zn的吸收在很大程度上受作物生理特性的影响。

† Significance code: a < 0.001 < b < 0.01 < c < 0.05.

www.agronomy.org • www.crops.org • www.soils.org

www.agronomy.org • www.crops.org • www.soils.org

† All coefficient significant at a = 0.001 (F-test using type III sums of square). ‡ LMG/R², in the order of explanatory variables in the model. § Model with the complete set of explanatory soil and climate variables (# 19, Table 3) determined R² = 0.90. ¶ Maximum adjusted R² for model size of 12 explanatory variables (R²_adj = 0.87). # ox-Cd, oxalate-extractable Cd.

† 所有系数在a = 0.001水平显著（采用III型平方和的F检验）。‡ LMG/R²，按模型中解释变量的顺序排列。§ 包含全部土壤和气候解释变量的模型（# 19，表3）的R² = 0.90。¶ 12个解释变量模型的最大调整R²（R²_adj = 0.87）。# ox-Cd，草酸盐可提取态Cd。

† Significance code: a < 0.001 < b < 0.01 < c < 0.05 (F-test using type III sums of square). ‡ LMG/R², in the order of explanatory variables in the model. § Model with the complete set of explanatory soil and climate variables (# 19, Table 3) determined R² = 0.80. ¶ Maximum adjusted R² for model size of 12 explanatory variables (R²_adj = 0.74). # ox-Zn, oxalate-extractable Zn.

† 显著性代码：a < 0.001 < b < 0.01 < c < 0.05（采用III型平方和的F检验）。‡ LMG/R²，按模型中解释变量的顺序排列。§ 包含全部土壤和气候解释变量的模型（# 19，表3）的R² = 0.80。¶ 12个解释变量模型的最大调整R²（R²_adj = 0.74）。# ox-Zn，草酸盐可提取态Zn。

Tables 6 and 7 show the best multiple linear regression models we derived to predict pasture concentrations of Cd and Zn, respectively. Generally, edaphic and climate variables only poorly accounted for pasture Cd concentration, with a sixfactor regression model having an R²_adj value of just 0.38. The Ca(NO3)2–extractable, ox-Cd, and total Cd concentrations were heavily weighted in the regression models (Table 6).

表6和表7分别列出了我们建立的预测牧草Cd和Zn浓度的最佳多元线性回归模型。总体而言，土壤和气候变量对牧草Cd浓度的解释能力较差，六因子回归模型的R²_adj值仅为0.38。Ca(NO3)2可提取态Cd、ox-Cd和全量Cd浓度在回归模型中权重较大（表6）。

The multiple regression models for Zn had more explanatory power, with just four terms providing a fit with an R²_adj of 0.51.

Zn的多元回归模型解释力更强，仅用四项即可获得R²_adj为0.51的拟合。

Besides Ca(NO3)2–Zn, these models were heavily weighted in pasture N and include pasture Cd (with a positive sign), strengthening the above-mentioned dominant role of plant factors in Zn uptake.

除Ca(NO3)2–Zn外，这些模型中牧草N的权重较大，并包含牧草Cd（正号），进一步印证了上述植物因素在Zn吸收中的主导作用。

Daily Intake of Cadmium for Animals Grazing on Sampled Pasture采样牧场放牧动物的镉日摄入量

Assuming that animals graze in the field all the time and always with the same Cd concentration in the forage, the daily intake of Cd (DICd) (mg d⁻¹), can be estimated according to Eq. [1] (Rodrigues et al., 2012). Using daily consumption parameters Ifeed (kg d⁻¹ dw) and Isoil (kg d⁻¹ dw) from the literature (Rodrigues et al., 2012) and the measured soil and pasture Cd concentrations, [Cd]soil and [Cd]feed (mg kg⁻¹), we calculated a DICd for all sites for cow and sheep. We used food standards of Australasia (ANZFSC, 2012) and the European Union (Rodrigues et al., 2012), respectively, and bioaccumulation factors (BAFs) from the literature (Rodrigues et al., 2012) to calculate an acceptable daily intake for Cd (ADICd) using Eq. [2] (Rodrigues et al., 2012). The BAF is defined as the Cd concentration in the organ (mg kg⁻¹ fresh weight) divided by the concentration in the feed (mg kg⁻¹ dw).

假设动物全天在田间放牧，且饲草中Cd浓度始终不变，则Cd日摄入量（DICd）（mg d⁻¹）可按式1估算（Rodrigues et al., 2012）。利用文献中的日消耗参数Ifeed（kg d⁻¹ dw）和Isoil（kg d⁻¹ dw）（Rodrigues et al., 2012）以及实测的土壤和牧草Cd浓度[Cd]soil和[Cd]feed（mg kg⁻¹），计算了所有样点牛和绵羊的DICd。分别采用澳新食品标准（ANZFSC, 2012）和欧盟标准（Rodrigues et al., 2012），以及文献中的生物累积因子（BAF）（Rodrigues et al., 2012），按式2计算Cd的可接受日摄入量（ADICd）（Rodrigues et al., 2012）。BAF定义为器官中Cd浓度（mg kg⁻¹鲜重）除以饲料中Cd浓度（mg kg⁻¹ dw）。

Table 8 shows that all calculated DICd values were lower than ADICd when calculated considering muscle tissue food standards. Considering offal food standards (kidney, liver), some calculated DICd values exceeded the ADICd. The few exceptions are consistent with the reported exceedence of food safety standards for offal products of gracing animals in NZ, which led to表8显示，以肌肉组织食品标准计算时，所有DICd计算值均低于ADICd。以内脏（肾脏、肝脏）食品标准计算时，部分DICd计算值超过了ADICd。这少数例外与已报道的新西兰放牧动物内脏产品超过食品安全标准的情况一致，这也导致了

the current policy of the NZ meat industry not to use offal of animals over 2.5 yr of age for human consumption (Roberts et al., 1994). Although continued application of Cd-containing phosphate fertilizers will increase Cd concentrations in animal products, this effect may be lessened by the Tiered Fertilizer Management System (Rys, 2011). Using this system, soils with <0.6 mg kg⁻¹ Cd (19% of our samples) have no restriction placed on the application of P fertilizers. Soils with concentrations >0.6 and <1.0 (51% of our samples) are permitted restricted use of P to ensure that the soil does not exceed 1.8 mg kg⁻¹ Cd after 100 yr. More intensive monitoring and modeling is required for soils with Cd concentrations >1 and <1.8 mg kg⁻¹ (30% of our samples). No further Cd accumulation is permitted in soils with Cd concentrations >1.8 mg kg⁻¹ (0% of our samples) (Rys, 2011).

新西兰肉类行业现行的政策，即2.5岁以上动物的内脏不供人类食用（Roberts et al., 1994）。尽管持续施用含Cd磷肥将提高动物产品中的Cd浓度，但分级肥料管理系统（Tiered Fertilizer Management System）可能减轻这一影响（Rys, 2011）。根据该系统，Cd含量<0.6 mg kg⁻¹的土壤（占本研究样品的19%）施用P肥不受限制。浓度>0.6且<1.0的土壤（占本研究样品的51%）允许限量施用P肥，以确保100年后土壤Cd不超过1.8 mg kg⁻¹。Cd浓度>1且<1.8 mg kg⁻¹的土壤（占本研究样品的30%）需进行更密集的监测和模拟。Cd浓度>1.8 mg kg⁻¹的土壤（占本研究样品的0%）不允许Cd进一步累积（Rys, 2011）。

† Significance code: a < 0.001 < b < 0.01 < c < 0.05 (F-test using type III sums of square).

† 显著性代码：a < 0.001 < b < 0.01 < c < 0.05（采用III型平方和的F检验）。

‡ LMG/R², in the order of explanatory variables in the model.

‡ LMG/R²，按模型中解释变量的顺序排列。

§ Model with the complete set of explanatory variables (# 25, Table 3) determined R² = 0.55.

§ 包含全部解释变量的模型（# 25，表3）的R² = 0.55。

¶ Maximum adjusted R² for model size of 10 explanatory variables (R²_adj = 0.42).

¶ 10个解释变量模型的最大调整R²（R²_adj = 0.42）。

# ox-Cd, oxalate-extractable Cd.

# ox-Cd，草酸盐可提取态Cd。

†† ox-Zn, oxalate-extractable Zn.

†† ox-Zn，草酸盐可提取态Zn。

† Significance code: a < 0.001 < b < 0.01 < c < 0.05 (F-test using type III sums of square).

† 显著性代码：a < 0.001 < b < 0.01 < c < 0.05（采用III型平方和的F检验）。

‡ LMG/R², in the order of explanatory variables in the model.

‡ LMG/R²，按模型中解释变量的顺序排列。

§ Model with the complete set of explanatory variables (# 25, Table 3) determined R² = 0.67.

§ 包含全部解释变量的模型（# 25，表3）的R² = 0.67。

¶ Maximum adjusted R² for model size of 12 explanatory variables (R²_adj = 0.57).

¶ 12个解释变量模型的最大调整R²（R²_adj = 0.57）。

† Calculation according to Eq. [1] and Eq. [2] with parameters Ifeed = 16.9 kg d⁻¹ dw; Isoil = 0.41 kg d⁻¹ dw; BAFfeed-animal organ = 2.99 (kidney), = 0.554 (liver), 0.0033 (muscle) for cow; and parameters Ifeed = 2.5 kg d⁻¹ dw, Isoil = 0.10 kg d⁻¹ dw, and BAFfeed-animal organ = 2.08 (kidney), = 1.85 (liver), = 0.0029 (muscle) for sheep (Rodrigues et al., 2012).

† 按式1和式2计算，牛的参数为：Ifeed = 16.9 kg d⁻¹ dw；Isoil = 0.41 kg d⁻¹ dw；BAFfeed-animal organ = 2.99（肾脏）、0.554（肝脏）、0.0033（肌肉）；绵羊的参数为：Ifeed = 2.5 kg d⁻¹ dw，Isoil = 0.10 kg d⁻¹ dw，BAFfeed-animal organ = 2.08（肾脏）、1.85（肝脏）、0.0029（肌肉）（Rodrigues et al., 2012）。

‡ Daily intake of Cd.

‡ Cd日摄入量。

§ EU, European Union; NZ, New Zealand. ¶ Acceptable daily intake of Cd.

§ EU，欧盟；NZ，新西兰。¶ Cd可接受日摄入量。

Intake from pasture was more important than from soil ingestion (80 and 70% of total intake for cow and sheep, respectively). As discussed above, there are numerable variables that affect concentration of Cd in pasture. Therefore, it is unsurprising that there was no significant correlation between the DICd and the total soil Cd concentration. The soils where DICd exceeded ADICd did not have significantly higher Cd than the remaining samples. However, this does not indicate that total soil Cd is unimportant. Other things being equal, increasing soil Cd will invariably increase pasture Cd uptake, as indicated by the positive correlation between the two (Table 3). Soil samples >0.88 mg kg⁻¹ Cd (50% of the samples) had a significantly (p < 0.001) higher DICd (1.32 and 1.50 mg for cow and sheep, respectively) compared with soils with <0.88 mg kg⁻¹ Cd (0.74 and 0.82 mg for cow and sheep, respectively). Increasing soil Cd concentrations may result in a greater number of breaches of food safety standards.

来自牧草的摄入比来自土壤摄食的更为重要（牛和绵羊分别占总摄入量的80%和70%）。如上所述，影响牧草Cd浓度的变量众多。因此，DICd与土壤全量Cd浓度之间无显著相关并不意外。DICd超过ADICd的土壤，其Cd含量并不显著高于其余样品。然而，这并不表明土壤全量Cd不重要。在其他条件相同的情况下，土壤Cd增加必然会提高牧草对Cd的吸收，二者之间的正相关即说明了这一点（表3）。Cd含量>0.88 mg kg⁻¹的土壤样品（占样品的50%）的DICd（牛和绵羊分别为1.32和1.50 mg）显著（p < 0.001）高于Cd含量<0.88 mg kg⁻¹的土壤（牛和绵羊分别为0.74和0.82 mg）。土壤Cd浓度升高可能导致更多超过食品安全标准的情况。

3Conclusions结论

Soil Cd and pasture Cd were positively correlated with soil P and soil concentrations of other elements found in phosphate fertilizers. Our study indicated that no single environmental variable could be used to adequately predict pasture Cd uptake. Nevertheless, total soil Cd and Ca(NO3)2–extractable soil Cd were positively correlated with pasture Cd. Although soil pH, soil Fe, and soil Cd provided an excellent predictor of the Ca(NO3)2–extractable soil Cd fraction, regression models explained just 38% of the variation of the Cd concentration in pasture grasses. Incorporating the effect of pasture species composition is a crucial next step in improving these models. Our calculation predicted no exceedences of food safety standards for muscle tissue but predicted some exceedences for offal products. Although continued application of phosphate fertilizers containing Cd will increase the concentration of Cd in animal products, the Tiered Fertilizer Management System may lessen this effect. Future work should involve determining the relative accumulation of various pasture species and examining the rhizosphere chemistry in a range of edaphic conditions. This will allow a better understanding of when the acceptable daily intake may be exceeded and permit the development of management tools, such as pasture species selection, that will reduce the exposure of stock to Cd.

土壤Cd和牧草Cd均与土壤P以及磷肥中所含其他元素的土壤浓度呈正相关。本研究表明，没有任何单一环境变量可用于充分预测牧草对Cd的吸收。尽管如此，土壤全量Cd和Ca(NO3)2可提取态Cd均与牧草Cd呈正相关。虽然土壤pH、土壤Fe和土壤Cd能够很好地预测Ca(NO3)2可提取态土壤Cd，但回归模型仅能解释牧草中Cd浓度变异的38%。纳入牧草物种组成的影响是改进这些模型的关键下一步。计算结果预测肌肉组织不会超过食品安全标准，但内脏产品会出现部分超标。尽管持续施用含Cd磷肥将提高动物产品中的Cd浓度，但分级肥料管理系统可能减轻这一影响。今后的工作应包括测定不同牧草物种的相对累积能力，并研究多种土壤条件下的根际化学。这将有助于更好地了解可接受日摄入量何时可能被超过，并有助于开发管理工具（如牧草物种选择），以减少牲畜的Cd暴露。

4References参考文献

Anderson, C.W.N., B.H. Robinson, D.M. West, L. Clucas, and D. Portmann. 2012. Zinc-enriched and zinc-biofortified feed as a possible animal remedy in pastoral agriculture: Animal health and environmental benefits. J. Geochem. Explor. 121:30–35. doi:10.1016/j.gexplo.2012.01.009

ANZFSC. 2012. Australia New Zealand Food Standards Code, Standard 1.4.1: Contaminants and natural toxicants. Australia New Zealand Food Authority, Canberra, Australia.

Black, A., R.G. McLaren, S.M. Reichman, T.W. Speir, and L.M. Condron. 2011. Evaluation of soil metal bioavailability estimates using two plant species (L. perenne and T. aestivum) grown in a range of agricultural soils treated with biosolids and metal salts. Environ. Pollut. 159:1523–1535. doi:10.1016/j. envpol.2011.03.004

Black, A., R.G. McLaren, S.M. Reichman, T.W. Speir, and L.M. Condron. 2012. Examining the integrity of soil metal bioavailability assays in the presence of organic amendments to metal-spiked soils. Soil Use Manage. 28:89– 100. doi:10.1111/j.1475-2743.2011.00384.x

Blakemore, L.C., P.L. Searle, and B.K. Daly. 1987. Methods for chemical analysis of soils. New Zealand Soil Bureau Scientific Report 80. DSIR, Wellington, New Zealand.

Cadmium Working Group. 2008. Report one: Cadmium in New Zealand. Agriculture Ministry of Agriculture and Forestry, Wellington, New Zealand.

Cadmium Working Group. 2011a. Cadmium and New Zealand agriculture and horticulture: A strategy for long term risk management. Ministry of Agriculture and Forestry, Wellington, New Zealand.

Cadmium Working Group. 2011b. A national cadmium management strategy for New Zealand agriculture. 24th Annual FLRC Workshop, Palmerston North, New Zealand.

Carran, R.A. 1992. Managing soil acidity: Old solutions can cause new problems. Proc. N. Z. Grassl. Assoc. 54:77–80.

Crews, H.M., and B.E. Davies. 1985. Heavy-metal uptake from contaminated soils by 6 varieties of lettuce (Lactuca sativa L.). J. Agric. Sci. 105:591–595. doi:10.1017/S0021859600059499 Edmeades, D.C., B.S. Thorrold, and A.H.C. Roberts. 2005. The diagnosis and correction of sulfur deficiency and the management of sulfur requirements in New Zealand pastures: A review. Aust. J. Exp. Agric. 45:1205–1223. doi:10.1071/EA01173

FAL, RAC, and FAW. 1996. Schweizerische Referenzmethoden der Eidgenossischen Landwirtschaftlichen Forschungsanstalten. FAL- Eigenverlag, Zurich, Switzerland. Ghani, A., R.G. McLaren, and R.S. Swift. 1991. Sulfur mineralization in some New Zealand soils. Biol. Fertil. Soils 11:68–74. doi:10.1007/BF00335838 Grant, C.A., L.D. Bailey, M.J. McLaughlin, and B.R. Singh. 1999. Management factors which influence cadmium concentrations in crops. Dev. Plant Soil Sci. 85:151–198. Gray, C.W., R.G. McLaren, A.H.C. Roberts, and L.M. Condron. 1999. The effect of long-term phosphatic fertiliser applications on the amounts and forms of cadmium in soils under pasture in New Zealand. Nutr. Cycl. Agroecosyst. 54:267–277. doi:10.1023/A:1009883010490 Grömping, U. 2006. Relative importance for linear regression in R: The package relaimpo. J. Stat. Softw. 17:1–27.

Harrell, F.E.J. 2012. Hmisc: Harrell miscellaneous. R package version 3.10-1. http://CRAN.R-project.org/package=Hmisc (accessed Oct. 2013). Hooda, P.S. 2010. Trace elements in soils. Wiley Online Library.

Kabata-Pendias, A., and H. Pendias. 2001. Trace elements in soils and plants. CRC Press, Boca Raton, FL.

Kabata-Pendias, A., and A.B. Mukherjee. 2007. Trace elements from soil to human. Springer, Berlin, Germany.

Karami, M., M. Afyuni, A.H. Khoshgoftarmanesh, A. Papritz, and R. Schulin. 2009. Grain zinc, iron, and copper concentrations of wheat grown in central Iran and their relationships with soil and climate variables. J. Agric. Food Chem. 57:10876–10882. doi:10.1021/jf902074f

Kim, N. 2005. Cadmium accumulation in Waikato soils. Environment Waikato, Hamilton, New Zealand.

Lee, J., J.R. Rounce, A.D. Mackay, and N.D. Grace. 1996. Accumulation of cadmium with time in Romney sheep grazing ryegrass white clover pasture: Effect of cadmium from pasture and soil intake. Aust. J. Agric. Res. 47:877–894. doi:10.1071/AR9960877 Lindeman, R.H., P.F. Merenda, and R.Z. Gold. 1980. Introduction to bivariate and multivariate analysis. Scott, Foresman, Glenview, IL.

Loganathan, P., and M.J. Hedley. 1997. Downward movement of cadmium and phosphorus from phosphatic fertilisers in a pasture soil in New Zealand. Environ. Pollut. 95:319–324. doi:10.1016/S0269-7491(96)00142-X Loganathan, P., S. Vigneswaran, J. Kandasamy, and R. Naidu. 2012. Cadmium sorption and desorption in soils: A review. Crit. Rev. Environ. Sci. Technol. 42:489–533. doi:10.1080/10643389.2010.520234

Loganathan, P., M.J. Hedley, N.D. Grace, J. Lee, S.J. Cronin, N.S. Bolan, and J.M. Zanders. 2003. Fertiliser contaminants in New Zealand grazed pasture with special reference to cadmium and fluorine: A review. Aust. J. Soil Res. 41:501–532. doi:10.1071/SR02126

Lumley, T. 2009. Leaps: Regression subset selection. R package version. Using Fortran Code by Alain Miller 2.9. http://CRAN.R-project.org/ package=leaps (accessed Oct. 2013).

McGrath, D., and H. Tunney. 2010. Accumulation of cadmium, fluorine, magnesium, and zinc in soil after application of phosphate fertilizer for 31 years in a grazing trial. J. Plant Nutr. Soil Sci. 173:548–553. doi:10.1002/ jpln.200900140

McLaughlin, M.J., C.M.J. Williams, A. McKay, R. Kirkham, J. Gunton, K.J. Jackson, R. Thompson, B. Dowling, D. Partington, M.K. Smart, and K.G. Tiller. 1994. Effect of cultivar on uptake of cadmium by potato-tubers. Aust. J. Agric. Res. 45:1483–1495. doi:10.1071/AR9941483

Nan, Z.R., J.J. Li, J.M. Zhang, and G.D. Cheng. 2002a. Cadmium and zinc interactions and their transfer in soil-crop system under actual field conditions. Sci. Total Environ. 285:187–195. doi:10.1016/ S0048-9697(01)00919-6

Nan, Z.R., C.Y. Zhao, J.J. Li, F.H. Chen, and W. Sun. 2002b. Relations between soil properties and selected heavy metal concentrations in spring wheat (Triticum aestivum L.) grown in contaminated soils. Water Air Soil Pollut. 133:205–213. doi:10.1023/A:1012962604095

R-Core-Team. 2012. R: A language and environment for statistical computing. R Foundation for Statistical Computing Vienna, Austria. http://www.R-project.org/ (accessed Oct. 2013).

Reid, R., and J. Hayes. 2003. Mechanisms and control of nutrient uptake in plants. In: K.W. Jeon, editor, International review of cytology: A survey of cell biology. Vol. 229. Elsevier, Amsterdam, the Netherlands. p. 73–114. Roberts, A.H.C., and R.D. Longhurst. 2002. Cadmium cycling in sheep-grazed hill-country pastures. N. Z. J. Agric. Res. 45:103–112. doi:10.1080/0028 8233.2002.9513499

Roberts, A.H.C., R.D. Longhurst, and M.W. Brown. 1994. Cadmium status of soils, plants, and grazing animals in new-zealand. N. Z. J. Agric. Res. 37:119–129. doi:10.1080/00288233.1994.9513048

Robinson, B.H., G. Banuelos, H.M. Conesa, M.W.H. Evangelou, and R. Schulin. 2009. The phytomanagement of trace elements in soil. Crit. Rev. Plant Sci. 28:240–266. doi:10.1080/07352680903035424

Robinson, B.H., S. Bischofberger, A. Stoll, D. Schroer, G. Furrer, S. Roulier, A. Gruenwald, W. Attinger, and R. Schulin. 2008. Plant uptake of trace elements on a Swiss military shooting range: Uptake pathways and land management implications. Environ. Pollut. 153:668–676. doi:10.1016/j. envpol.2007.08.034

Rodrigues, S.M., M.E. Pereira, A.C. Duarte, and P.F.A.M. Römkens. 2012. Soil–plant–animal transfer models to improve soil protection guidelines: A case study from Portugal. Environ. Int. 39:27–37. doi:10.1016/j. envint.2011.09.005

Rys, G.J. 2011. A national cadmium management strategy for New Zealand agriculture. In: L.D. Currie and C.L. Christensen, editors, Adding to the knowledge base for the nurient manager: Occasional report no. 24 of the Fertiliser and Lime Research Centre. Massey University, Palmerston North, New Zealand.

Scheffer, F., P. Schachtschabel, H.-P. Blume, and S. Thiele-Bruhn. 2010. Lehrbuch der bodenkunde. Spektrum Akademischer Verlag, Heidelberg, Germany.

Schipper, L.A., G.P. Sparling, L.M. Fisk, M.B. Dodd, I.L. Power, and R.A. Littler. 2011. Rates of accumulation of cadmium and uranium in a New Zealand hill farm soil as a result of long-term use of phosphate fertilizer. Agric. Ecosyst. Environ. 144:95–101. doi:10.1016/j.agee.2011.08.002

Smolders, E. 2001. Cadmium uptake by plants. Int. J. Occup. Med. Environ. Health 14:177–183.

Sparks, D.L. 2003. Environmental soil chemistry. Academic Press, San Diego, CA.

Swartjes, F.A. 2011. Dealing with contaminated sites. Springer, Dordrecht, the Netherlands.

Syers, J.K., A.D. Mackay, M.W. Brown, and L.D. Currie. 1986. Chemical and physical characteristics of phosphate rock materials of varying reactivity. J. Sci. Food Agric. 37:1057–1064. doi:10.1002/jsfa.2740371102

Tait, A., and X. Zheng. 2007. Analysis of the spatial interpolation error associated with maps of median annual climate variables. National Institute of Water and Atmospheric Research, Auckland, New Zealand.

Taylor, M.D. 1997. Accumulation of cadmium derived from fertilisers in New Zealand soils. Sci. Total Environ. 208:123–126. doi:10.1016/ S0048-9697(97)00273-8

Taylor, M.D. 2007. Soil maps of cadmium in New Zealand. Ministry of Agriculture and Forestry, Ltd., Manaaki Whenua-Landcare Research, Lincoln, New Zealand.

Tudoreanu, L., and C.J.C. Phillips. 2004. Modeling cadmium uptake and accumulation in plants. Advances Agron. 84:121–157.

Welch, R.M., and W.A. Norvell. 1999. Mechanisms of cadmium uptake, translocation, and deposition in plants. Dev. Plant Soil Sci. 85:125–150.

Römkens, P.F.A.M., H.Y. Guo, C.L. Chu, T.S. Liu, C.F. Chiang, and G.F. Koopmans. 2009. Prediction of Cadmium uptake by brown rice and derivation of soil–plant transfer models to improve soil protection guidelines. Environ. Pollut. 157:2435–2444. doi:10.1016/j. envpol.2009.03.009

↑ Back to top↑ 返回顶部