cd concent potatoes(1)

7/29/2019 CD Concent Potatoes(1)

1/6

Elevated Cadmium Concentrations in Potato Tubers Due to Irrigation with River WaterContaminated by Mining in Potos, Bolivia

Carla Oporto,* Carlo Vandecasteele, and Erik Smolders

ABSTRACT

Risk of cadmium (Cd) in the human food chain in Cd-contaminatedareas is often limited by phytotoxicity from zinc (Zn) that is associated

with the Cd contamination. A semiarid area, 60 km downstream

of a tin mine in Bolivia, was surveyed where irrigation with Cd-

contaminated river water (65240 mg Cd L21) has increased median

soil Cd to 20 mg kg21 while median soil Zn was only about 260 mg kg21.

Cadmium concentrations in potato tubers increased from background

values (0.05 mg kg21 dry wt.) in soils irrigated with spring water to a

median value of 1.2 mg kg21 dry wt. in the affected area. Median

concentration of Cd in soil solutions was 27 mg L21 and exceeded the

corresponding value of Zn almost twofold. Soil-extractable chloride

ranged from 40 to 1600 mg Cl2 kg21 and was positively correlated

with soil total Cd. Increasing soil solution Cl2 decreased the solid-

liquid distribution coefficient of Cd in soil. Soil total Cd explained 64%

of the variation of tuber Cd concentration while only 3% of the varia-

tion was explained by soil extractable Cl2 (n 5 49). The estimated

dietary Cd intake from potato consumption by the local population is

about 100 mg d21 which exceeds the WHO recommended total dailyintake. It is concluded that the food chain risk of Cd in the irrigation

water of the semiarid area is aggravated by the association with Cl2

and, potentially, by the relatively large Cd/Zn ratio.

CADMIUM naturally occurs in the biosphere and mayaccumulate in soils due to anthropogenic activities.The main sources of this contamination are industry,mining, agriculture, and the application of sewage sludgeand fertilizers (Adriano, 2001). Risk assessments revealthat food chain contamination with Cd is the criticalpathway for human exposure. A well-known case ofhuman poisoning by Cd is the Itai Itai disease that wascaused by the consumption of unpolished rice thatcontained elevated Cd concentrations. The major sourceof that rice contamination was a zinc-lead mine up-stream from the endemic area. The average Cd concen-tration in rice was more than ten times higher in theendemic area (0.51.0 mg kg21 fresh wt.) than in otherareas (0.050.2 mg kg21 fresh wt.; WHO, 1992). This andother epidemiological studies have led to the definitionof the WHO recommended tolerable daily Cd intake(TDI) of 1 mg per kg body weight day (WHO, 1992).

It has been argued that the Cd transfer from soil to thefood chain is unique in the case of rice, because of the

relatively large Cd transfer to rice and the relatively

elevated bioavailability of orally ingested Cd to subsis-tence rice farmers (Chaney et al., 1999). Preliminarysurveys in the Chayanta catchment in Bolivia haveshown elevated concentrations of Cd, both in agricul-tural soils and associated potato tubers. These soils areirrigated with water from the Chayanta river, whichupstream receives acid mine drainage from waste dumpsand tailing ponds from the tin mine district Siglo XXLlallagua. Potatoes are locally grown and food habits inthe Andean zone are based on the consumption of thiscrop, which accounts for more than 60% of the totaldaily calories (Ugarte and Iriarte, 2005). Potatoes gen-erally contain small Cd concentrations, median tuberconcentrations being about 0.03 mg kg21 fresh wt. inuncontaminated areas (McLaughlin et al., 1997). Pota-

toes grown on a soil with elevated Cd due to sludgeapplication (soil Cd: 20 mg Cd kg21) contained between0.15 to 0.41 mg kg21 dry wt (about 0.030.08 mg Cd kg21

fresh wt.; Harris et al., 1981). A kitchen garden survey inthe Belgian Kempen area contaminated by Cd and Znshowed that average tuber Cd concentrations onlyexceed EU food limits (0.1 mg kg21 fresh wt.) if soiltotal Cdexceeds10 mgkg21 (Ruttens, personal communi-cation, 2006). However, elevated tuber Cd concentra-tions, up to about 0.2 mg kg21 fresh wt. (about 1.0 mgkg21 dry wt.), have been identified in southern Australia(McLaughlin et al., 1997). This was not related to soilCd, but rather to soil chloride (Cl2) salinity that in-creases Cd availability.

The Chayanta River is contaminated with Cd and hasmoderate Cl2 concentrations. The semiarid climate canmarkedly increase soil solution Cl2 and thus increase Cdavailability. In addition, Zn concentrations are relativelyless elevated in the river and large Cd/Zn ratios some-times increase soil Cd availability (Grant et al., 1999).The aim of the present study was to survey the Cd con-centrations in agricultural soils and potato tubers in anarea irrigated with water from Chayanta River. The rela-tionship between both was assessed to identify factorsthat control the soil Cd bioavailability.

MATERIALS AND METHODS

Sampling Area

The survey area is located 60 km downstream from Siglo

XXLlallagua mining district (18j

27

S, 66j

27

W; elevation3180 m) in the Department of Potos , Bolivia (Fig. 1). Theclimate is classified as semiarid steppe (BSk) in the Koppensystem, with an annual precipitation of 534 mm and 10jC

C. Oporto, Centro de Aguas y Saneamiento Ambiental, U.M.S.S,Calle Sucre final (Campus), P.O. Box 5783, Cochabamba, Bolivia.C. Vandecasteele, Lab. for Applied Physical Chemistry and Environ-mental Technology, K.U.Leuven, W. de Croylaan 46, 3001 Heverlee,

Belgium. E. Smolders, Lab. for Soil and Water Management,K.U.Leuven, Kasteelpark Arenberg 20, 3001 Heverlee, Belgium. Re-ceived 26 Sept. 2006. *Corresponding author ([email protected]).

Published in J. Environ. Qual. 36:11811186 (2007).Technical Reports: Heavy Metals in the Environmentdoi:10.2134/jeq2006.0401 ASA, CSSA, SSSA677 S. Segoe Rd., Madison, WI 53711 USA

Abbreviations: CEC, cation exchange capacity; dry wt., dry weight;EC, electrical conductivity; fresh wt., fresh weight; TDI, tolerabledaily intake.

Re

producedfrom

JournalofEnvironmentalQuality

.PublishedbyASA

,CSSA

,andSSSA

.Allcopyrightsreserved

.

1181

Published online June 27, 2007


2/6

average temperature. The main food crops include potato(Solanum tuberosum), corn (Zea mays), and beans (Vicia

faba), which are the predominant components of the popu-lation dietary intake, which is characterized by a very lowprotein content. Parcels are located both on slopes and in flatareas near the river; their sizes range between 4 and 400 m 2.The crops are mainly used for food consumption of the farmers

household. The soils are traditionally irrigated by furrowirrigation with Chayanta River water, which has been pollutedby mining activity since approximately 1920. The total surveyedarea is approximately 30 ha and the estimated population is250 inhabitants. The river is under strong regime change, be-tween dry (Q5 32 m3 s21) and rainy season (Q 5 120 m3 s21)and absolute concentrations in the water are highly time-variable. During the period of this survey, the Cd concentra-tion in the water of the irrigation channels from ChayantaRiver varied seasonally between 65 and 240 mg L21, the Cl2

concentration between 2 and 6 mM, and the electrical con-ductivity between 0.5 to 1.3 dS m21. Older data (1999) of Cdconcentrations in the Chayanta river reported values between260 and 2620 mg L21, while Zn concentrations ranged between15 and 26 mg L21, and tin (Sn) concentrations were less than10 mg L21 (Rojas and Vandecasteele, 2006). A survey wasperformed between September 2004 and February 2005.

Based on the monitoring of irrigation in the pilot fields duringthe agricultural campaign, it was estimated that the annualwater use in the field is about 430 mm. This is equivalent to anaverage annual input of Cd by irrigation of about 0.9 kg Cdha21 and an annual increase of soil Cd of about 0.4 mg kg21.Forty nine samples of soil and associated potatoes were takenfrom fields exposed to irrigation with Chayanta River water intwo villages along the river, Quila Quila (Location 1) andAsiruri (Location 2). The distance between the two villages isapproximately 1 km (Fig. 1). Additionally, 11 paired crops and

soils were sampled in fields located about 25 m higher and at adistance of 150 m from Location 2, where soils are irrigatedwith spring water (Location 3 5 reference location). Soil wascollected from the plow layer (20 cm) using a stainless steelshovel. Samples of 1 kg were taken as a composite of foursubsamples and were transported in plastic bags. Potato tuberswere sampled when mature from the same sampling points as

the soils; four subsamples were taken and mixed to make a 1-kgsample that was transported in fabric bags to the laboratory.Tuber size ranged from 25 to 60 mm.The potatoes(S. tuberosunssp. andigena) belonged predominantly to native varieties Alqapali and Alqa imilla.

Soil Analysis

Soil samples were air-dried, disaggregated in a wood mor-tar, and sieved to ,2 mm. Soil pH was determined in a CaCl20.01 M, 1:5 soil/water suspension (Corning 315 pH/ion). Theelectrical conductivity (EC) was measured in a 1:5 soil/watersuspension with a Cole Parmer conductivity meter (19820-20).Texture was analyzed according to the pipette method afterdestruction of organic matter with hydrogen peroxide, re-moval of carbonate and soluble salts with HCl, and dispersionwith sodium hexametaphosphate (ISO11277:1998[E], Interna-

tional Organization of Standardization, Geneva, Switzerland;Gee and Bauder, 1986). Chloride was extracted in a 1:5 soil/water suspension and measured in the filtered solution usingan ORION 1528 ion selective electrode. The unbuffered silver-thiourea method (Chhabra et al., 1975) was used to measurethe effective cation exchange capacity (CEC) and exchange-able cations. Organic matter was determined with the WalkleyBlack method. Soil total metal concentration was measuredafter digesting 1 g soil with 6 mL HNO3 (65%) and 3 mL HClin a microwave oven (Milestone-Ethos 900). Extracts were

Fig. 1. Location of Quila Quila (Loc1), Asiruri (Loc2), and reference (Loc3) sampling sites.

Re

producedfrom


.PublishedbyASA

,CSSA

,andSSSA


.

1182 J. ENVIRON. QUAL., VOL. 36, JULYAUGUST 2007


3/6

diluted to 50 mL with Milli Q water. Metal concentrationswere determined by flame atomic adsorption spectroscopy(PerkinElmer AAnalyst 100) and by inductively coupledplasmaoptical emission spectroscopy (ICPOES, PerkinEl-mer Optima 3300 DV). The SRM NIST 2710, Montana Soil,with elevated trace metals concentrations, and SRM NIST2709 San Joaquin soil, with baseline values, were run simulta-

neously for assessing accuracy. The measured total Cd concen-tration for the SRM 2710 was 21.9 6 0.8 mg kg21 which agreedwith the certified value, 21.8 mg kg21. The average Zn concen-tration measured on SRM 2709 was 104.2 6 0.97 mg kg21 ingood agreement with the certified value of 106 mg kg21.

Soil Solution Isolation and Analysis

Suitable amounts of MilliQ water were added to 70 g ofdried soil to saturate the soil, which was subsequently stored inplastic containers overnight at 20jC. The level of saturationwas set as described in the recommended method for obtain-ing soil saturation extracts (Rhoades, 1996), i.e., water is addeduntil the paste glistens as it reflects light. The soil moisture con-tent at the saturation ranged between 0.23 and 0.33 g water/gsoil for our samples. The soil solutions were isolated by cen-trifugation at 3000 g followed by membrane filtration (0.45mm)

discarding the first droplets. Chloride, NO32

, and SO422

weredetermined in the soil solution by ion chromatography(Dionex ICS-2000). Metal concentrations were measured inacidified soil solutions with ICPOES (Cd detection limit is0.5 mg L21). The concentration of dissolved organic carbon(DOC) was determined by thermal oxidation (ThermaloxTOC analyzer, Analytical Sciences).

Tuber Analysis

Potato tubers were brushed with a nylon brush under tapwater and rinsed with distilled water. Clean unpeeled tuberswere left to drain on blotting paper for 15 min. After this,tubers were cut unpeeled in small chips with a stainless steelknife. A comparison of Cd concentrations between peeled andunpeeled samples showed larger Cd concentrations (about

10%) in unpeeled samples (n5

51, samples from Belgium; DeTemmerman, personal communication, 2006). Potatoes areusually consumed unpeeled in the test area. Chips were driedat 65jC during 72 h. The dry matter content ranged from 16 to26%. The dried samples were crushed in a mortar and ap-proximately 0.5 g of ground sample was digested in 10 mL ofHNO3 (65%) in a microwave oven; the extract was diluted to50 mL with MilliQ water. The concentration of Cd in thetubers was determined by graphite furnace atomic absorptionspectroscopy (PerkinElmer AAS 3110, HGA 600). Sampleswere digested in duplicate, in batches including SRM NIST8436 Durum Wheat flour with a certified Cd concentration

of 0.11 mg kg21 dry wt. The mean of our results was 0.110 60.006 mg kg21 dry wt. These analyses were performed with adetection limit of 0.002 mg kg21. A subset of potato sampleswere analyzed by ICPOES, which allowed determination ofother elements as well. In that case, 100 mg of dried, crushedsample were digested in 2 mL of HNO3 and diluted to 5 mL.The digestions were performed in batches including two

blanks and an internal reference of plant material. The Cddetection limit of this procedure was 0.025 mg kg21. The inter-nal reference sample was an endive leaf sample from the inter-national plant-analytical exchange program (IPE, UniversityWageningen) with a consensus value of 1.75 mg Cd kg21 dry wt.We found 1.75 6 0.05 mg Cd kg21 dry wt.

Statistical Analysis

Effects of location were identified by ANOVA. Tuber Cdconcentrations were analyzed by correlation analysis andby stepwise multiple regressions (SAS 9.1, SAS Institute, Cary,NC). The significance levels for a regressor staying in themodel were set to 0.05.

RESULTS

Selected soil properties are summarized in Table 1.Soil pH was not significantly different among the threelocations, the standard deviations being within 0.2 unitsin the entire data set. In contrast, other soil parameterswere significantly affected by location. Soils from loca-tion 1 (Loc1) were predominantly sandy in texture witha low content of organic matter, while the clay and theorganic matter percentages were higher for location 2(Loc2) soils. The texture in the reference location 3(Loc3) was not measured, but the CEC values and soiltexture, estimated by palpation, suggest similar textureto location 2. The soil Cd concentration increased in theorder Loc3b Loc1 , Loc 2. The soil EC and extract-able Cl2 followed the same trend indicating that theChayanta River water has been a source of Cd and Cl2

salinity while the spring water is low in Cd and dissolvedsalts. The soil-extractable Cl2 was significantly corre-lated with soil total Cd concentration (R2 5 0.55). Themedian total soil Cd concentrations are 6 and 57 mg kg21

at Loc1 and Loc2, respectively, significantly above valuesin reference location 3 (0.4 mg kg21). Soil total Zn wasalso significantly affected by location; however, the in-crease above the range in the background soils wassmaller than for total Cd (Table 1).

The ionic molar composition of the soil solutionswas generally dominated by Cl2 and Na1, while Ca21

Table 1. Selected soil properties. Soils at locations 1 and 2 have been irrigated with water from the contaminated river, while soils atlocations 3 have been irrigated with spring water.

Location 1 Location 2 Location 3

Parameters n Range Median n Range Median n Range MedianpH 23 7.48.2 7.8 25 7.47.8 7.6 11 7.47.8 7.7EC (dS m

21) 23 0.11.1 0.4 25 0.31.3 0.6 3 ,0.1 0.09Clay (%) 23 328 10 19 1639 26CEC (cmolc kg

21) 23 3.413.2 6.8 25 9.116.6 14.0 10 9.316.2 13.6OM (g kg21) 23 3.227.9 10.5 19 10.553.4 26.6Extr.Cl2 (mg kg21) 23 39809 222 19 961593 698 3 1018 15Total Cd (mg kg

21) 23 3.519.7 6.2 25 19.577.8 57.2 11 0.040.5 0.4Total Zn (mg kg21) 23 91.1273 130 22 262857 616 11 69113 76

EC, electrical conductivity in 1:5 water extract of soil; CEC, cation exchange capacity; OM, organic matter; Extr.Cl2, water-extractable chloride; CdT, total

concentration of Cd; ZnT, total concentration of Zn.

Re

producedfrom


.PublishedbyASA

,CSSA

,andSSSA


.

1183OPORTO ET AL.: ELEVATED Cd IN POTATOES DUE TO MINING-CONTAMINATED WATER


4/6

and SO422 dominated in a few samples (Table 2). The

Cd/Zn weight ratio in solution was larger than 1 in 63%of the samples. This ratio is unusual (e.g., median is 0.01in 62 samples of contaminated and noncontaminatedsamples; Degryse et al., 2003), suggesting relativelylarge Cd bioavailability in soil. Soil solution Cd cor-related with soil solution Cl2 (r5 0.81; P, 0.01). This ispartly due to the covariance between soil-extractableCl2 and soil Cd contamination and partly due to the

effect of Cl2

mobilizing Cd in soil by the formation ofCdCln

2-n complexes (Smolders et al., 1998). The Cdsolid-liquid partition coefficient Kd was calculated fromthe ratio of total soil Cd to that in soil solution. The Kdvalues were unrelated to soil pH (not shown) butsignificantly decreased with increasing soil solution Cl2

(Fig. 2, r5 20.50; P, 0.01) in a log-log relationship.The potato tuber Cd concentration ranged from 0.05

to 4.77 mg kg21 dry wt. and was significantly differentamong the three locations. The European food limit forpotato tubers (0.1 mg kg21 fresh wt., about 0.5 mg kg21

dry wt.) was never exceeded in the reference locationbut was exceeded in 88% of the samples of location 1and 2 (Fig. 3). Tuber Cd concentrations were larger thanin any previously reported surveys (Table 3; see also

surveys reviewed in McLaughlin et al., 1997). The tuberCd concentrations were positively correlated with tuberZn, Na, and Ca concentrations (r5 0.69, 0.66, and 0.65,respectively, P, 0.01). Tuber Cd was highly correlatedwith soil solution Cd, Cl2, and SO4

22, but the latterparameters also positively correlate with the total soilor soil solution Cd concentrations (Table 4). The tuber

Cd concentrations exhibited nonlinear relationshipswith soil total Cd or with soil solution Cl2 while thesetrends appear linear in log-log plots (Fig. 3 and 4). Soil-extractable Cl2 explained additional variation of tuberCd concentrations while pH had no effect (Table 5).

DISCUSSION

Geological Cd background concentrations in soils of

this area are unknown; however, the Cd concentrationsin the reference soils of this study (0.040.5 mg kg21, n511; Table 1), compare with background values reportedworldwide where most geometric mean values in min-eral soils are ,1 mg Cd kg21 (Traina, 1999). The totalCd concentration in the affected area (3.578 mg kg21)clearly exceeds the background. The current annual in-crease in soil total Cd was estimated at 0.4 mg kg21 (seeabove) illustrating that irrigation, since the river con-tamination in 1920, is the major source of soil Cd in thatarea. Major Cd inputs from application of sludge orfertilizer application are unlikely, since farmers do notuse mineral fertilizer and only rely on the use of manurefrom local cattle. The annual input from manure andcompost essentially recycles some of the Cd ingested

by animals feeding on crops produced on the fields.Tuber Cd concentrations reported here are probablythe highest ever reported in field surveys (Table 3). Thesoil Cd contamination in our study, however, also ex-ceeds that of most of these other surveys. The few pre-viously reported data of tuber Cd concentrations in

Table 2. Loc1, Loc2, and Loc3 soil solution composition. Soils were wetted to saturation before isolating the pore water by centrifugation.

Location 1 Location 2 Location 3

Parameters n Range Median n Range Median n Range Median

Cd (mg L21) 21 ,0.150 12.0 24 20630 146.0 7 ,0.1 ,0.1

Zn (mg L21) 21 10190 11.0 24 ,0.190 38.0 7 ,0.110 7.0

Na (mM) 21 3.790.9 16.2 24 6.790.9 45.0 7 0.72.5 0.9Ca (mM) 21 3.419.1 13.2 24 4.832.0 17.2 7 1.84.4 3.2

SO422

(mM) 21 2.232.5 15.1 24 6.642.0 22.3 7 1.64.9 2.4NO32 (mM) 21 0.18.2 2.37 24 0.216.9 2.9 7 0.116.3 1.3

Cl2

(mM) 21 2.471.8 15.3 24 3.7107.3 38.8 7 1.210.6 1.7DOC (mg L21) 21 17.6233.2 69.0 24 33.6232.0 89.3 7 70.6220.8 91.8

DOC, dissolved organic carbon.

0

500

1000

1500

2000

2500

3000

0 20 40 60 80 100 120

Soil solution Cl-

concentration (mM)

Kd(L/kg)

Loc 1

Loc 2

Loc 3

Fig. 2. Soil solution Cl2 decreases the partitioning coefficient Kd ofCdbetween soil and soil solution, i.e., Cl2 mobilizes Cd in soil.

0.0

0.1

1.0

10.0

0 1 10 100

Soil total Cd concentration (mg/kg)

TuberCdconcentration(mg/kgDW)

Loc 1

Loc 2

Loc 3

Fig. 3. The potato tuber Cd concentrations relate to soil total Cd con-centration. The European food limit (0.1 mg kg21 fresh wt. 0.5 mg kg21 dry wt.) is shown by the horizontal line.

Re

producedfrom


.PublishedbyASA

,CSSA

,andSSSA


.



5/6

Cd-contaminated soils show values less than about0.5mgCdkg21 drywt. (see introduction). Soils with simi-lar total Cd concentration e.g., close to metal smeltersor from sludge application, are often phytotoxic due tosoil acidity or increased toxic trace metals, such as Zn orCu (Chaney and Hornick, 1978). The pH was unaffectedhere by soil contamination and there were no overt signsof phytotoxicity. Elevated soil Cd bioavailability mightadditionally explain the elevated tuber Cd concentra-tion. Cadmium bioavailability in the field is promoted

by soil acidity, Cl2

, or SO422

soil salinity and, in somecases, large Cd/Zn ratio (Grant et al., 1999). The Cd/Znratio in soils here are well above the usual value of 0.01,likely due to the relatively larger Cd/Zn ratio in theirrigation water, and is even more than 1 (median 1.42)in soil solution. In addition, soil Cl2 and SO4

22 are ele-vated in soils of this survey. Chloride forms complexeswith Cd21 and mobilizes Cd in soil (Bingham et al.,1983), a mechanism observed here again (Fig. 2). In-creased crop Cd due to Cl2 salinity is reported in the fieldfor potatoes (McLaughlin et al., 1997), sunflower seeds(Li et al., 1994), and wheat grain (Norvell et al., 2000).The potato tuber Cd concentrations increased linearlyby about a factor of 5 with increasing soil-extractableCl2 between 10 and 1000 mg kg21 (Fig. 5 in McLaughlin

et al., 1997). This relationship was observed in a surveywhere soil Cd had no significant effect on the tuber Cdconcentrations, illustrating the effect of Cl2 on Cd bio-availability which is also shown in experimental studies(Smolders et al., 1998). The median soil-extractable Cl2 inboth contaminated locations of this survey was 220 and700 mg Cl2 kg21, sufficient to predict increased soil Cdavailability based on a comparison with the McLaughlinet al. (1997) data. The multiple regression model shownin Table 5 predicts a more than twofold increase in tuberCd by increasing soil-extractable Cl2 between 10 and

1000 mg kg21 at constant total Cd and CEC. However,soil-extractable Cl2 only explains less than 5% of thevariation in tuber Cd after excluding the effect of totalCd (Table 5). This lack of a strong statistical relation-

ship is likely due to the strong covariance between soilCd and Cl2 (Table 4), both originating from the riverwater. Variability in water use among fields may explainthis covariance.

The effect of increasing soil solution SO422 on Cd

bioavailability is less clear than that of Cl2. IncreasingSO4

22mobilizes Cd in soil, yet only small or inconsistenteffects on Cd bioavailability are found (Grant et al.,1999). A multivariate model using soil solution proper-ties showed that increasing SO4

22 increased tuber Cdafter excluding the effect of total dissolved Cd; however,SO4

22 only explained 6% of the variation.Data about daily intake of potatoes in the local popu-

lation are not available; however, potatoes are a majorfood crop for these people. The estimated daily potato

consumption intake in the Andean zone is approxi-mately 0.4 kg fresh wt., representing 60% of daily caloricintake. This intake corresponds to a daily Cd intake of98 mg from potatoes at the average Cd concentrationin the crops of location 1 and 2. This value exceeds theWHO recommended TDI of 70 mg d21 (for an adultof 70 kg). Total dietary Cd intake will be larger due toother sources of dietary Cd and will obviously be largerin highly exposed individuals. This observation triggersa more detailed risk characterization with biomonitor-ing in the local population.

Table 3. Potato tuber cadmium concentrations in various countries.

Country n Median Min Max Reference

mg kg21 dry wt.

Aust ra li a 3 59 0. 16 0. 02 1 .1 6 McLaughli n et a l. , 1 99 7S weden 75 0.04 0.004 0.14 Grawe et al., 2001Norway 79 0.06 ,0.01 0.22 Alne and Gjerstad, 1998Poland 6306 0.11 Kabata-Pendias et al., 2001B elg ium 1 08 0. 12 0. 03 0 .3 4 De Te mmer man, per sona l

communication, 2006Bolivia 59 1.05 0.05 4.77 This study

Fresh wt. values transformed to dry wt. assuming 20% dry matter content.

Table 4. Pearson correlation coefficients (r) between tuber Cdconcentrations (Cdtub), total Cd in soil (CdT), and soil solutionconcentrations of Cd, Cl2, and SO4

22.

Parameter Cdtub CdT Cd Cl SO45

mg kg21 dry wt. mg kg21 mg L21 mM

Cdtub(mg kg21

dry wt.)1.00 0.70** 0.58** 0 .52** 0.53**

CdT (mg kg21) 1.00 0.68** 0.42** 0.32*

Cd (mg L21) 1.00 0.81** 0.49**Cl2 (mM) 1.00 0.72**SO4

22 (mM) 1.00

* Significant at the P, 0.05 level.** Significant at the P, 0.01 level.

0.0

0.1

1.0

10.0

1 10 100 1000

Soil solution Cl-

concentration (mM)

TubersCdconcentration(mg/kgDW)

Loc 1

Loc 2

Loc 3

Fig. 4. Tuber Cd concentrations are positively related to soil solutionCl2; note that soil Cd contamination is positively correlated withsoil solution Cl2 (Table 4).

Table 5. Results of stepwise multiple regression (partial R2 andmodel R2) relating tubers Cd concentration (log-transformed,mgkg21 dry wt.) to soil parameters (log-transformed except pH).

Soil parameters entered (CdT, CEC, Extr. Cl2, pH)

Step logCdT logCEC logExtr.Cl2 Model R2 n

mg/kg cmolc/kg mg/kg

1 0.64 0.642 0.06 0.693 0.03 0.72 49Model: logCdplant520.4310.39logCdT10.16logExtr.Cl

220.44logCEC

CdT, total Cd in soil; CEC, cation exchange capacity; Extr.Cl2, water-

extractable chloride.

Re

producedfrom


.PublishedbyASA

,CSSA

,andSSSA


.

1185OPORTO ET AL.: ELEVATED Cd IN POTATOES DUE TO MINING-CONTAMINATED WATER


6/6

ACKNOWLEDGMENTS

C. Oporto thanks VLIR-IUC program between the FlemishUniversities and the Universidad Mayor de San Simon (UMSS),for a doctoral grant and for the funding of the present research.

REFERENCES

Adriano, D.C. 2001.Trace elements in the terrestrial environment.2nd ed. Springer Verlag, New York.Alne, J.I., and K.O. Gjerstad. 1998. Bly og kadmium i poteter og

grnnsaker (in Norwegian). SNT-report 7:98. Norwegian FoodControl Authority, Oslo, Norway.

Bingham, F.T., J.E. Strong, and G. Sposito. 1983. Influence of chloridesalinity on cadmium uptake by Swiss chard. Soil Sci. 135:160165.

Chaney, R.L., and S.B. Hornick. 1978. Accumulation and effects ofcadmium on crops. p. 125140. In Proc. Int. 1st Cadmium Conf.,San Francisco. Metals Bulletin, Ltd., London.

Chaney, R.L., J.A. Ryan, and S.L. Brown. 1999. Soil cadmium as athreat to human health. p. 219256. In M.J. McLaughlin and B.R.Singh (ed.) Cadmium in soils and plants. Kluwer AcademicPubl., Dordrecht, The Netherlands.

Chhabra, R., J. Pleysier, and A. Cremers. 1975. The measurement ofthe cation exchange capacity and exchangeable cations in soils:A new method. p. 439449. In S.W. Bailey (ed.) Proc. of the Int.Clay Conf. in Mexico. Applied Publ. Ltd., Wilmette, IL.

Degryse, F., K. Broos, E. Smolders, and R. Merckx. 2003. Soil solution

concentration of Cd and Zn can be predicted with a CaCl2 soilextract. Eur. J. Soil Sci. 54:149158.

Gee, G.W., and J.W. Bauder. 1986. Particle-size analysis. p. 383411.In A. Klute (ed.) Methods of soil analysis. Part 1. ASA and SSSA,Madison, WI.

Grant, C.A., L.D. Bailey, M.J. McLaughlin, and B.R. Singh. 1999.Management techniques to reduce cadmium transfer from soils:A review. p. 151198. In M.J. McLaughlin and B.R. Singh (ed.)Cadmium in soils and plants. Kluwer Academic Publ., Dordrecht,The Netherlands.

Grawe, K.P., I. Oborn, and K. Gustaffsson. 2001. Cadmium and othertrace metals in Swedish food potatoesDevelopment project forenvironmental monitoring. Final Rep. Swedish EnvironmentalProtection Agency (contract 222 0006, 222 0010) (in Swedish).

Harris, M.J., S.J. Harrison, N.J. Wilson, and N.W. Le pp. 1981. Varietaldifferences in trace metal partitioning in six potato cultivars grownon contaminated soil. p. 399402. In Proc. Int. Conf. Heavy Metalsin the Environment, CEP Consultants, Edinburgh, UK.

Kabata-Pendias, G., H. Terelak, and C. Pietruch. 2001. Impact of soilfactors on Zn and Cd contents in potato tubers. In Proc. 6th Int.Conf. On the Biogeochemistry of Trace Elements, Guelph, Canada.

Li, Y., R.L. Chaney, and A.A. Schneiter. 1994. Effect of soil chloridelevel on cadmium concentration in sunflower kernels. Plant Soil167:275280.

McLaughlin, M.J., N.A. Maier, G.E. Rayment, L.A. Sparrow, G. Berg,A. McKay, P. Milham, R.H. Merry, and M.K. Smart. 1997. Cad-mium in Australian potato tubers and soils. J. Environ. Qual. 26:16441649.

Norvell, W.A., J. Wu, D.G. Hopkins, and R.M. Welch. 2000. Asso-ciation of Cd in durum wheat grain with soil chloride and chelate-extractable soil cadmium. Soil Sci. Soc. Am. J. 64:21622168.

Rhoades, J.D. 1996. Salinity: Electrical conductivity and total dissolvedsolids. p. 417435. In D.L. Sparks et al. (ed.) Methods of soil analy-sis, Part 3. SSSA, Madison, WI.

Rojas, J.C., and C. Vandecasteele. 2006. Influence of mining activitiesin the north of Potosi, Bolivia on the water quality of chayanta riverand its consequences. (in press).

Smolders, E., R.M. Lambregts, M.J. McLaughlin, and K.G. Tiller. 1998.Effect of soil solution on cadmium availability to Swiss chard.J. Environ. Qual. 27:426431.

Traina, S.J. 1999. The environmental chemistry of cadmium. p. 1137.In M.J. McLaughlin and B.R. Singh (ed.) Cadmium in soils andplants. Kluwer Academic Publ., Dordrecht, The Netherlands.

Ugarte, M.L., and V. Iriarte. 2005. Papas bolivianas. Catalogo de 100variedades nativas. PROINPACIP COSUDE, Bolivia.

WHO. 1992. Cadmium. Environmental health criteria 134. World HealthOrganization, Geneva, Switzerland.

Re

producedfrom


.PublishedbyASA

,CSSA

,andSSSA


.


cd concent potatoes(1)

Documents