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


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (12)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kukier, U.
Right arrow Articles by Roseberg, R. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kukier, U.
Right arrow Articles by Roseberg, R. J.
Agricola
Right arrow Articles by Kukier, U.
Right arrow Articles by Roseberg, R. J.
Related Collections
Right arrow Toxic Trace Metals
Right arrow Heavy Metals
Right arrow Soil Pollution
Published in J. Environ. Qual. 33:2090-2102 (2004).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS

Heavy Metals in the Environment

The Effect of pH on Metal Accumulation in Two Alyssum Species

Urszula Kukiera,*, Carinne A. Petersb, Rufus L. Chaneya, J. Scott Angleb and Richard J. Rosebergc

a USDA-ARS, Animal Manure and By-Product Laboratory, BARC-W, Building 007, Beltsville, MD 20705
b Department of Natural Resource Sciences, University of Maryland, College Park, MD 20742
c Oregon State University, Southern Oregon Experimental Station, Central Point, OR 97502

* Corresponding author (kukieru{at}ba.ars.usda.gov)

Received for publication December 4, 2003.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Nickel phytoextraction using hyperaccumulator plants offers a potential for profit while decontaminating soils. Although soil pH is considered a key factor in metal uptake by crops, little is known about soil pH effects on metal uptake by hyperaccumulator plants. Two Ni and Co hyperaccumulators, Alyssum murale and A. corsicum, were grown in Quarry muck (Terric Haplohemist) and Welland (Typic Epiaquoll) soils contaminated by a Ni refinery in Port Colborne, Ontario, Canada, and in the serpentine Brockman soil (Typic Xerochrepts) from Oregon, USA. Soils were acidified and limed to cover pH from strongly acidic to mildly alkaline. Alyssum grown in both industrially contaminated soils exhibited increased Ni concentration in shoots as soil pH increased despite a decrease in water-soluble soil Ni, opposite to that seen with agricultural crop plants. A small decrease in Alyssum shoot Ni concentration as soil pH increased was observed in the serpentine soil. The highest fraction of total soil Ni was phytoextracted from Quarry muck (6.3%), followed by Welland (4.7%), and Brockman (0.84%). Maximum Ni phytoextraction was achieved at pH 7.3, 7.7, and 6.4 in the Quarry, Welland, and Brockman soils, respectively. Cobalt concentrations in shoots increased with soil pH increase in the Quarry muck, but decreased in the Welland soil. Plants extracted 1.71, 0.83, and 0.05% of the total soil Co from Welland, Quarry, and Brockman, respectively. The differences in uptake pattern of Ni and Co by Alyssum from different soils and pH were probably related to the differences in organic matter and iron contents of the soils.

Abbreviations: DTPA, diethylenetriaminepentaacetic acid • HFO, hydrous ferric oxide


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
HYPERACCUMULATOR PLANTS are species able to accumulate very high amounts of trace metals in their tissues, at concentrations 10 to 100 times higher than tolerated by crop plants. For Ni, Co, Cu, and many other metals the threshold for hyperaccumulation was set at 1000 mg kg–1 of shoot dry matter (Brooks et al., 1977). For Zn, this threshold was established at 10000 mg kg–1 (Reeves and Brooks, 1983). Higher concentrations of accumulated metals are found in the shoots than the roots of hyperaccumulators (Chaney et al., 1997). This is opposite to the strategy used by nonaccumulator species, which accumulate metals in their roots when exposed to high soil metal concentrations. For nonaccumulators, exclusion of metals from shoots and/or roots and retention of metals in root cell walls and vacuoles has been found to be a defense mechanism minimizing metal phytotoxicity. In contrast, metal accumulation in shoots of hyperaccumulator plants is believed to provide a unique method of self-defense against microbial infection and larval feeding on leaves, giving these species an advantage on metal-rich soils (Martens and Boyd, 1994). A large number of species have developed the ability to accumulate Zn, Cd, Ni, Co, Se, and Cu, and are usually endemic to various metalliferous soils (Baker et al., 2000). The greatest diversity of hyperaccumulator species evolved in Ni-rich serpentine soils. The largest number of Ni-accumulating species belongs to the Alyssum genus (Baker et al., 2000). The unique flora endemic to serpentine areas had been studied for decades but the idea of using hyperaccumulator species for metal phytoextraction from soil is relatively new. Growing hyperaccumulator plants on Ni-rich soils and ashing harvested biomass is an economically sound alternative way of producing Ni ore (Chaney, 1983; Nicks and Chambers, 1995; Robinson et al., 1997; Brooks et al., 1998; Chaney et al., 2000; Li et al., 2003b). This new way of removing metals is termed "phytomining." There are two categories of soils suitable for production of Ni crops: serpentine soils naturally rich in Ni, which occur all over the world, and industrially contaminated soils. Nickel phytomining from industrially contaminated soils can be also viewed as environmental cleanup. This dual function makes hyperaccumulator plants an especially promising, cost-effective alternative for decontamination of large areas of Ni-rich land.

The unique ability of hyperaccumulator plants to accumulate and tolerate extraordinarily high concentrations of metals stimulated basic research toward a better understanding of hyperaccumulator physiology and biochemistry. Less attention was paid to the effect of soil properties, other than total or exchangeable Ni concentrations, on the Ni uptake by hyperaccumulator plants.

A very important question raised regarding plant–soil–metal interactions is how much soil pH affects uptake of Ni and other metals by hyperaccumulator plants. This question was addressed in the recently published study by Li et al. (2003a), who investigated the effect of changing soil pH on Ni and Co concentrations in the shoots of A. murale and A. corsicum. Plants were grown in soils contaminated with Ni and Co by industrial emissions. The study resulted in the remarkable observation that an increase in soil pH was associated with increase in shoot Ni concentration. This observation is in contrast to an overwhelming number of published scientific data demonstrating that increased soil pH causes a decrease of Ni concentrations in various nonaccumulator plant species grown in Ni-contaminated or serpentine soils rich in Ni of geogenic origin (Crooke, 1956; L'Huillier and Edighoffer, 1996; Kukier and Chaney, 2004). Cobalt in this study followed the usual pattern: decrease of shoot concentration with pH increase. However, in another study there was an indication that Ni and Co hyperaccumulators Berkheya coddii and A. bertolonii grown in serpentine soils may respond with decreased Ni and Co uptake as soil pH is increased (Robinson et al., 1999; Robinson et al., 1997), which is in contrast to results of Li et al. (2003a). Both studies involving Alyssum covered a narrow pH range. Li et al. (2003a) focused on acidic and very acidic pH ranges expecting that soil acidification would maximize Ni and Co uptake. The study of Robinson et al. (1997) involving a naturally alkaline serpentine soil covered the pH range from 7.37 to 7.94. To date, the accumulation pattern of Ni, Co, and other metals by Alyssum in a broad range of soil pH has not been reported.

The knowledge of how soil pH affects Ni and Co accumulation by Alyssum species is of primary importance for maximizing annual Ni phytoextraction. Growing hyperaccumulators on industrially contaminated soils combines phytomining with soil remediation. The pattern of hyperaccumulator response to changes in soil pH is a key factor for designing a soil cleanup strategy. Liming is a widely used technique for in situ remediation of heavy metal phytotoxicity. High rates of limestone are required not only to achieve the highest possible pH but also to provide an excess of liming material for long-term buffering of any acid input. Adding ferric hydroxy oxide (HFO) to Ni-contaminated soil that has already been alkalinized was shown to further improve amelioration of Ni phytotoxicity for some crops (Kukier and Chaney, 2001). The possible interference of liming and HFO amendment to Ni-contaminated soils with Ni phytoextraction could affect soil management at contaminated sites. Provided that heavy liming of contaminated soils would decrease Ni uptake by hyperaccumulator plants, Ni phytoextraction should be the first step of soil remediation followed by raising soil pH to minimize any residual phytotoxic effect of Ni to agricultural crops.

The objective of the study was to assess the effect of soil pH on the Ni and Co solubility and uptake by A. murale and A. corsicum from the industrially contaminated soils, and from a serpentine soil to which Alyssum is native. Soil pH values were adjusted to cover a broad pH range from acid to alkaline (calcareous). The effect of soil pH on the uptake of Cu and Zn, which are not hyperaccumulated by Alyssum, was also investigated.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soil Collection and Characterization
Quarry muck (Terric Haplohemist; Canadian classification, Orthic Humic Gleysol) and Welland silt loam (Typic Epiaquoll; Canadian classification, Terric Mesisol) soils contaminated by particulate emissions from the Port Colborne, Ontario, nickel refinery were collected from the plow layer of farms located on Inco Ltd. property. A Brockman variant serpentine soil (Typic Xerochrepts) high in Ni of geogenic origin was collected in Josephine County, Oregon. Total soil Ni, Co, Mn, and Fe were determined according to USEPA Method 3050 (USEPA, 1995). Organic carbon content was determined with a LECO (St. Joseph, MI) CNS analyzer. Soil pH was measured in a 1:2 soil to water (by volume) slurry after 1 h of equilibration. Particle size distribution of the Welland soil was determined by hydrometer method after removal of organic matter (Gee and Bauder, 1986). Soils for the pot study were sieved through a 5-mm stainless steel sieve and thoroughly mixed. To preserve microbial activity and minimize redox reactions of Fe, Mn, and other elements that occur during drying, the soils were stored moist in closed plastic containers at 4°C until amended with limestone, HFO, and fertilizers. The moisture content of the soils was measured gravimetrically by oven-drying.

Experimental Design
Two hyperaccumulator species, A. corsicum and A. murale, were tested for their growth and metal uptake response to changes in pH of the Welland, Quarry, and Brockman soils. Both species have previously been tested in field trials and found to be suitable for climatic conditions of the Oregon and Port Colborne regions. Soil pH was altered either by addition of Ca and Mg carbonates or by acidification with HNO3. Nickel-contaminated Quarry muck and Welland soils had five treatments: unaltered pH, limed 1, limed 2, calcareous, and calcareous + amended with HFO. Brockman serpentine soil had four treatments: acidified, natural pH, limed, and calcareous. Each soil treatment had four replicates arranged in a randomized complete block design.

The GLM procedure (SAS Institute, 1988) was used to test statistical significance of the treatment effects on plant yield and elemental composition. The procedure was performed separately for each Alyssum species and soil.

Soil Amendments
pH Adjustment
The initial pH values of Welland and Quarry soils were relatively low, 5.24 and 5.66, respectively. Various rates of a mixture of powdered reagent-grade CaCO3 and MgCO3 (hereafter referred to as "limestone") were applied to raise pH of both soils. The combination of Ca and Mg carbonates (4.8:1 weight ratio) was used to avoid Mg deficiency. Rates of limestone were determined by incubation of moist soils amended with increasing rates of Ca and Mg carbonates. The CaCO3 was applied to the Welland soil at rates of 0, 4.4, 10.9, and 101.7 g pot–1. The accompanying rates of MgCO3 were 0, 0.9, 2.3, and 21.4 g pot–1. The limestone rates (Ca and Mg carbonates) expressed in Mg ha–1 were 0, 2.08, 5.19, and 48.4, respectively. Quarry soil received CaCO3 at rates 0, 4.3, 20.8, and 101.7 g pot–1 and MgCO3 at 0, 0.9, 4.4, and 21.4 g pot–1. This corresponds to 0, 2.04, 9.91, and 48.4 Mg ha–1 of Ca and Mg carbonates. These treatments allowed for covering a broad range of pH in each soil, from acidic to calcareous, and are referred to as "unaltered," "limed 1," "limed 2," and "calcareous," respectively.

Brockman serpentine soil (initial pH 6.30) was both acidified and treated with CaCO3 to cover a broad pH range. Reagent-grade CaCO3 (hereafter referred to as "limestone") was used to raise pH of the Oregon serpentine soil because soils developed from ultramafic rocks are naturally rich in Mg and an excess of this nutrient rather than deficiency is typical for these soils. Limestone was added to the Brockman serpentine soil at rates 0, 2.0, and 127 g CaCO3 pot–1, which corresponds to 0, 0.786, and 50.0 Mg ha–1. Please note that, for brevity, the same term "limestone" will be used for the mixture of CaCO3 and MgCO3 applied to Welland and Quarry soils, and CaCO3 applied to Brockman soil.

Acidification of the Brockman serpentine soil was accomplished by addition of HNO3. Four batches of Brockman soil were placed into a plastic-lined cement mixer and 520 mL of 0.8 M HNO3 was sprayed while soil was mixed. Acid-amended soil was placed in plastic containers, sealed, and incubated for 2 d. After that period, soil was placed in 4-L pots; excess salts generated by soil acidification were leached with deionized water (8 L pot–1) before fertilizer amendment.

Hydrous Ferric Oxide
Hydrous ferric oxide was applied only to the calcareous treatment of the Quarry and Welland soils. The HFO was precipitated at room temperature by an addition of NaOH to a solution of Fe(NO3)3 in quantity exceeding the stoichiometric ratio [NaOH to Fe(NO3)3 3:1] by 1.5%. Following precipitation, the HFO was washed with deionized water to remove NaNO3 until electrical conductivity of the entrained solution dropped to 0.05 to 0.06 mS cm–1 and the pH of the suspension dropped to about 7.3. The HFO slurry was applied at equal rate, 27 g Fe pot–1, to the Quarry and Welland soils in addition to making them calcareous. This corresponded to the addition of 2.5% Fe to the Quarry muck and 1.4% to the Welland mineral soil (oven-dry soil basis).

Fertilizers
Basal fertilizers were added to all pots as salt solutions at the following rates (mg pot–1): N, 224; P, 375; K, 223; Cu, 15; Zn, 15; and B, 2.5. As Port Colborne soils are prone to Mn deficiency, especially when limed (Baldwin and Johnston, 1986), increasing rates of Mn fertilizer, accompanying increasing rates of limestone, were added to the Welland and Quarry soils. Manganese was applied as MnSO4 solution at the rate of 51 mg Mn pot–1 to control treatments (unaltered pH), 127 mg Mn pot–1 to the pots amended with a lower rate of lime (limed 1), and at the rate of 254 mg Mn pot–1 to the pots limed to pH 6.5 to 6.8 (limed 2), calcareous, and calcareous + HFO treatment. These rates corresponded to 20, 50, and 100 kg Mn ha–1. Serpentine soil is naturally rich in Mn and was not amended with Mn fertilizer. After mixing with fertilizers, limestone, and HFO, moist soils in the pots were incubated for 2 d before transplanting seedlings.

Plant Growing Conditions
Alyssum seeds were germinated in peat–vermiculite commercial potting medium amended with fertilizers. After 44 d of growing in this medium, seedlings were transplanted, five plants per pot, to the 4-L freely drained plastic pots filled with soils. Plastic mesh covered the drainage holes to retain soil. At transplant, seedlings were about 4 to 5 cm tall. Saucers were used to prevent loss of leachate. The experiment was conducted in a greenhouse with controlled light and temperature conditions. Photoperiod was set at 16 h and the temperature at 27 and 20°C during the day and night, respectively. High-intensity sodium and incandescent lights capable of supplying 400 µmol m–2 s–1 of photosynthetically active radiation supplemented sunlight if necessary. During the growth period, plants were receiving supplemental N–P–K fertilizer at the rate of 187, 277, and 275 mg pot–1, respectively, applied as solution every 14 d, four times during experimental period. Plants were harvested after 68 d growth in the tested soils.

Plant and Soil Analyses
Immediately after harvest, Alyssum shoots were triple-rinsed in deionized water to remove any adhering soil particles. Plants were oven-dried at 65°C to a constant weight, and shoot dry matter was recorded. Plant material was ground in a stainless steel Wiley mill, weighed into Pyrex beakers, and ashed in a muffle furnace at 450°C for 16 h. Blanks were included every 10 samples. Ash was dissolved in 2 mL of concentrated HNO3 on a hot plate and then refluxed for 2 h with 10 mL of 3 M HCl. After digestion, solutions were filtered and diluted to 25 mL with 0.1 M HCl. Samples were analyzed for Ni, Co, Mn, Fe, Zn, Cu, Ca, Mg, K, and P by inductively coupled plasma (ICP)–atomic emission spectrometry using 40 mg L–1 Y as an internal standard. Several laboratory plant standards prepared from field-grown Alyssum species with more than 1% Ni were used in quality control of Ni analysis. Standard Reference Material 1570a (spinach leaves; National Institute of Standards and Technology, Gaithersburg, MD) was digested (one per 20 samples) for quality control of other element analyses.

Following plant harvest, representative soil samples were collected from each pot. Soil samples were air-dried and crushed before analyses. Soil pH was measured in water slurries at the ratio of 1:2 (soil to deionized water by volume) and a 1-h equilibration time. Neutral salt-extractable Ni, Co, Cu, and Zn were determined in 0.01 M Sr(NO3)2 (Madden, 1988) at soil to solution ratio of 1:4 (10 g to 40 mL) and a 2-h shaking time. The 0.01 M Sr(NO3)2 extraction was successfully used previously for the prediction of pH-induced changes in soil Ni availability to agricultural crops grown in the Ni-contaminated Welland and Quarry soils (Kukier and Chaney, 2001, 2004). Diethylenetriaminepentaacetic acid (DTPA)-extractable metals were determined using the method of Lindsay and Norvell (1978) but the soil to solution ratio used was 1 g to 20 mL to avoid saturation of DTPA by the large pool of phytoavailable metals present in contaminated soils (Kukier and Chaney, 2000). Metals bound to soil organic matter were extracted with sodium pyrophosphate (0.1 M Na4P2O7) for 16 h at 1 g to 40 mL soil to solution ratio (Miller et al., 1986; Asami et al., 1995). Concentrations of metals in the soil extracts were determined by atomic absorption spectrometry (AAS).

Soil Saturation Extracts
Soil saturation extracts were obtained from the soil samples from the lowest pH and calcareous treatments collected after plant harvest. Air-dry soil samples representing four replicates of the same treatment were combined, crushed with a rolling pin, and thoroughly mixed before use for the saturation extract preparation. Soil saturation extracts were prepared in triplicate according to the procedure described by Rhoades (1982). Water content of the Brockman serpentine and Welland soils at the saturated paste was 27 and 68% by weight, respectively. In the Quarry muck at low soil pH, water content was 152%, and it was 134% when soil was made calcareous. After 22 h of incubation, solutions were removed from soils by vacuum filtration using Büchner funnels lined with Whatman (Maidstone, UK) no. 42 filters. Solution volumes were recorded for each extraction, and concentrations of Ni, Co, Cu, and Zn were determined by AAS. Average volumes of extracts obtained in a single extraction from the Brockman and Welland soils (both pH levels) were 49 and 124 mL kg–1 air-dry soil, respectively. Average extract volumes for Quarry muck, low pH and calcareous treatment, were 390 and 290 mL kg–1 air-dry soil, respectively. After removal of the solution, moist soils were left for 2 h to enable soil aeration, and subsequently deionized water was added to the soils to bring their water content back to saturated paste conditions, and after 22 h new saturation extracts were obtained. The procedure was repeated four times, to obtain four consecutive saturation extracts from the same soil sample in 24-h intervals to test the ability of soils to resupply Ni after the soluble Ni was removed.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soils
Soils tested in this study differed in their properties and total metal concentrations (Table 1). Brockman soil contained large quantities of iron and cobalt characteristic of serpentine soils (Slingsby and Brown, 1977; Baker and Brooks, 1989; Robinson et al., 1996), an appreciable amount of Mn, and little organic matter. In contrast, Fe and Mn contents of the Quarry and Welland soils were low. Organic carbon content in the Welland soil was much higher than in the serpentine soil. The Quarry soil was a muck soil with organic carbon content of 36.3%. The highest total Ni concentration was found in the Brockman serpentine soil followed by industrially contaminated Welland and Quarry muck. As a result of particulate emission from the Ni refinery, concentrations of Co, Cu, and, to a lesser extent, Zn in both Port Colborne soils were increased above background levels typical for the region. For all soils, the soil pH range covered in this study was about two units, from strongly acid or medium acid to mildly alkaline. The Sr(NO3)2–extractable Ni (Table 2) was inversely related to soil pH in all soils tested. The relationship between soil pH and Sr(NO3)2–extractable Ni had an exponential decline with increase in pH in the Welland and Quarry soils and a linear character in the serpentine soil (data not shown). Decreased extractability with increase in soil pH was also observed for Co. Based on experience with nonaccumulator species, the liming-induced decrease in extractable Ni should be interpreted as a decrease in plant availability of this element (Crooke, 1956; Bisessar, 1989; Sauerbeck and Hein, 1991; L'Huillier and Edighoffer, 1996; Kukier and Chaney, 2001).


View this table:
[in this window]
[in a new window]
  		Table 1. Selected properties of the soils as collected.

 

View this table:
[in this window]
[in a new window]
  		Table 2. The effect of pH treatments on 0.01 M Sr(NO3)2–extractable metals in the Quarry, Welland, and Brockman soils after plant harvest.

 
Plant Yield
Soil pH treatments did not significantly affect yield of either Alyssum species grown in the Quarry muck or Welland soil (Tables 3 and 4). All treatments in both soils resulted in healthy-looking plants. Acidification of the serpentine soil caused a significant (P = 0.05) yield depression of both Alyssum species in comparison with control soil (Table 5), although plants did not show any toxicity or deficiency symptoms. Liming serpentine soil to pH 6.8 to 7.0 did not have any statistically significant effect on yield, as compared with control (unaltered pH), but making soil calcareous caused a yield depression. Plant elemental composition did not provide explanation for this effect.


View this table:
[in this window]
[in a new window]
  		Table 3. Uptake (phytoextraction) of Ni and Co by Alyssum grown for 68 d in Quarry muck soil.

 

View this table:
[in this window]
[in a new window]
  		Table 4. Uptake (phytoextraction) of Ni and Co by Alyssum grown for 68 d in Welland soil.

 

View this table:
[in this window]
[in a new window]
  		Table 5. Uptake (phytoextraction) of Ni and Co by Alyssum grown for 68 d in Brockman serpentine soil.

 
Nickel in Alyssum Shoots
In both Ni-contaminated soils, Quarry and Welland, Ni concentrations in Alyssum shoots increased with increase in soil pH (Fig. 1) . Shoot Ni concentrations of A. corsicum grown in the Quarry muck soil increased from 2240 to 6430 mg kg–1 when soil pH was raised from 5.69 to 7.30. The HFO amendment to the calcareous Quarry muck soil decreased Ni level in A. corsicum shoots despite further increase in pH induced by ferric hydroxide application. Response of A. murale to soil pH treatment was of similar magnitude. The HFO amendment to Quarry muck, which caused an increase in soil pH, resulted in a slight increase in Ni concentration in A. murale shoots, which was consistent with the response of this species observed in the Welland soil. Shoot Ni concentrations of A. corsicum grown in the Welland soil increased from 2380 to 9060 mg kg–1 dry weight when soil pH was raised from 5.10 to 6.50. Raising soil pH further to 7.71 did not cause further increase in shoot Ni concentration. Alyssum murale responded somewhat less dramatically. The change in soil pH from 5.24 to 7.60 resulted in the corresponding change in shoot Ni from 3440 to 6980 mg kg–1. Slight increase in soil pH induced by HFO application, in addition to making soil calcareous, further increased shoot Ni up to 7480 mg kg–1.



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 1. The effect of pH and hydrous ferric oxide (HFO) treatments of the Quarry, Welland, and Brockman soils on Ni concentrations in shoots of Alyssum murale and A. corsicum. The HFO treatments were not included in the data sets for curve fitting. Error bars represent ±standard errors. Solid symbols, limestone-only treatment; open symbols, limestone + HFO. Lines represent linear or quadratic best fit.

 
In contrast to the industrially contaminated soils, the increase in pH of the Brockman serpentine soil caused a decrease in shoot Ni concentrations of both Alyssum species (Fig. 1). Alyssum murale grown in acidified serpentine soil accumulated 12260 mg kg–1 Ni in shoot dry matter (Table 2); shoot Ni concentration of A. corsicum, at the same pH, was 10950 mg kg–1. Raising soil pH to 7.8 reduced Ni level in A. murale and A. corsicum to 4780 and 5420 mg kg–1, respectively. The response of both species to changes of soil pH could be described by one equation.

While interpreting presented data, the reader should keep in mind that changing of soil pH was achieved by addition of Ca and Mg, elements that may affect Alyssum growth and Ni uptake (Vergnano Gambi et al., 1992). Therefore, testing several pH-changing amendments would be required to elucidate the effect of pH on metal uptake by Alyssum.

Cobalt in Alyssum Shoots
Homer et al. (1991) and Malik et al. (2000) demonstrated that many Alyssum species, including A. murale and A. corsicum, hyperaccumulate Co if it is present in soil at significant levels. High concentrations of Co in Alyssum are not found in their natural environment because Ni concentrations in serpentine soils are substantially higher than Co concentrations. This was also the case in our study, because in all soils tested, Co concentrations were low in comparison with Ni. The pattern of Co accumulation in shoots in response to modification of soil pH was different for each soil (Fig. 2) , and reflects probably not only changes in soil Co solubility but also competition between Ni and Co for plant uptake.



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 2. The effect of pH and hydrous ferric oxide (HFO) treatments of the Quarry, Welland, and Brockman soils on Co concentrations in shoots of Alyssum murale and A. corsicum. The HFO treatments were not included in the data sets for curve fitting. Error bars represent ±standard errors. Solid symbols, limestone-only treatment; open symbols, limestone + HFO. Lines represent exponential decay, exponential growth, or quadratic best fit.

 
Making soil calcareous greatly increased Co uptake from the Quarry soil. To our knowledge, this is the first observation of this kind. Worth noting is a large decrease in shoot Co of both Alyssum species in the calcareous Quarry muck amended with HFO. In the Welland soil, liming largely decreased Co concentrations in both Alyssum species. Because Co in both Port Colborne soils originated from the same emission source, this is the most compelling evidence that soil properties have a pronounced effect on plant Co accumulation. The increase in pH from 5.8 to 6.8–6.9 decreased Co levels in shoots of Alyssum grown in the Brockman serpentine soil but making soil calcareous had the opposite effect resulting in U-shaped response curves for Co.

Copper and Zinc in Alyssum Shoots
Copper and Zn represent metals that are not hyperaccumulated by Alyssum (Baker and Brooks, 1989). Similar effect of soil pH on Cu and Zn accumulation by Alyssum was observed in all soil, therefore only Welland soil is presented as an example (Fig. 3) . Copper concentrations in Alyssum were fairly stable within the entire pH range. Despite soil contamination with Cu by the emissions from the refinery, Cu levels in Alyssum shoots grown on Port Colborne soils were within the range typically found in plants grown on nonpolluted soils, or even at the threshold of deficiency, which is for most plant species at ≤5 mg kg–1 of shoot dry matter (Adriano, 1986).



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 3. The effect of pH and hydrous ferric oxide (HFO) treatments of the Welland soil on Cu and Zn concentrations in shoots of Alyssum murale and A. corsicum. The HFO treatments were not included in the data sets for curve fitting. Error bars represent ±standard errors. Solid symbols, limestone-only treatment; open symbols, limestone + HFO. Lines represent linear best fit.

 
Zinc concentrations in plants harvested from Quarry and Welland soils significantly decreased as soil pH was raised. Making Brockman serpentine soil calcareous also significantly decreased Zn levels in shoots of both Alyssum species.

Nickel and Cobalt Uptake
Total Ni in plant shoots (Ni concentration multiplied by shoot dry matter) is labeled "uptake." In phytomining, Ni uptake is considered more important than Ni concentration because harvested shoots are incinerated to produce a Ni-rich ash. If concentration of Ni in the biomass has little or no effect on the costs of processing the ash to recover the metal, the goal of phytomining must be to produce the highest "uptake" from a Ni-rich soil. The highest Ni uptake from the Welland and Quarry soils by both Alyssum species was obtained at the highest soil pH but without HFO added (Tables 3 and 4), which occurred because pH treatments did not affect plant yield, while increase in soil pH increased shoot Ni concentrations. Nickel concentrations in both Alyssum species grown in the Welland soil exhibited a plateau response indicating that further increase in soil pH would probably not substantially increase Ni uptake in this soil. Linearity of the relationship between soil pH and Alyssum shoot Ni suggests that increasing pH of the Quarry muck soil beyond 7.5 would further increase shoot Ni concentration, and its uptake, but raising pH above that maintained by free limestone is not practical. In the Brockman soil, the maximum Ni uptake occurred at the original soil pH (Table 5). Higher shoot Ni concentrations were obtained in the acidified Brockman soil, but simultaneous yield depression resulted in a lower Ni uptake. The character of the response curve indicated that a further acidification of the serpentine soil would probably increase shoot Ni concentrations in both Alyssum species. Provided that yield depression caused by soil acidification could be overcome, lowering the serpentine soil pH to 5.3 to 5.4 would be a strategy of choice to increase Ni phytoextraction from serpentine soils. Adaptation to very strongly acidic soils might by obtained by breeding Alyssum to tolerate the very acidic soil condition. However, because Alyssum species are generally adapted to neutral or alkaline serpentine soils, it may not be easy to improve yields at low pH.

When comparisons between soils were made (Tables 3, 4, and 5), the highest Ni uptake by both Alyssum species, expressed in mg Ni pot–1 (3.3 dm–3 soil), was recorded for the Welland soil followed by the Brockman and Quarry, but differences between soils were not very pronounced. When Ni uptake was expressed in mg Ni kg–1 of air-dry soil, the highest values for uptake were obtained for the Quarry muck followed by Welland and Brockman soil. Consequently, the order of relative Ni uptake expressed as a fraction of total Ni present in the bulk soil was Quarry muck > Welland > Brockman serpentine soil. The maximum Ni uptake by A. corsicum in both industrially contaminated soils was higher than that of A. murale, which was a result of higher shoot Ni concentrations of A. corsicum in both soils and a slightly higher yield in the Quarry soil.

Cobalt uptake was much smaller than that of Ni both in absolute terms and relative to total soil metal values (Tables 3, 4, and 5). While this low recovery of Co is not commercially viable, it presents some value for soil remediation. In the Welland soil, maximum Co uptake occurred at the lowest soil pH, indicating that phytoextraction of Ni and Co could not be maximized simultaneously in this soil.

Sodium Pyrophosphate– and DTPA-Extractable Metals
The highest amounts of Ni, Co, Cu, and Zn were extracted by Na-pyrophosphate (Na4P2O7) from the Quarry muck soil, followed by Welland and Brockman serpentine, from which very low quantities of all metals were extracted (Fig. 4) . The pool of metals extracted decreased with increase in soil pH. Copper association with organic matter was similar in both industrially contaminated soils, 40 to 55% of total, and was little affected by soil pH. From 30% of the total Ni, at acid pH, to 20%, at alkaline pH, was extracted from the Quarry muck. An even higher proportion of Co, about 50% of total, was bound to organic matter in the Quarry muck soil at low pH, but the abundance of this fraction was significantly decreased with increase in pH. Corresponding numbers for Ni in the Welland soil were 15% at the low pH and 7% at the high pH, and for the Brockman serpentine, 1.27 and 0.25% at the low and high pH, respectively.



View larger version (37K):
[in this window]
[in a new window]
  		Fig. 4. Sodium pyrophosphate–extractable (organically bound) metals in Quarry, Welland, and Brockman soils at plant harvest. Error bars represent ±standard errors. Lines represent linear or quadratic best fit.

 
A similar amount of Ni extracted and similar pattern of Ni extractability in relation to soil type and pH were observed when DTPA (Lindsay and Norvell, 1978) was employed as an extracting agent (at 1 g soil to 20 mL solution to avoid saturating the chelator added) (Fig. 5) . The DTPA-extractable Ni was much higher in the Quarry muck and Welland than in the Brockman serpentine. Extractability of Fe was apparently related to the soil organic matter content, and was the highest in the Quarry muck, and very low in the Brockman serpentine soil, which was extremely rich in total Fe, but very low in organic matter.



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 5. The effect of pH treatment on the DTPA-extractable Ni and Fe in Quarry, Welland, and Brockman soils at plant harvest (1 g soil to 20 mL solution). Error bars represent ±standard errors. Lines represent linear best fit.

 
Soil Saturation Extracts
Copper was the only metal whose concentrations in the saturation extracts of the Quarry muck and Welland soils were higher when soils were made calcareous (Fig. 6) . No depletion of Cu was observed in the consecutive saturation extracts. Concentrations of Cu, as well as Co and Zn, in the saturation extracts of the Brockman soil were too low to be precisely determined. Nickel concentrations in the consecutive saturation extracts of all calcareous treatments were fairly stable, and were around 0.9, 1.8, and 0.09 mg L–1 in the Quarry, Welland, and Brockman soils, respectively (Fig. 6). At low pH, each soil exhibited a distinct pattern of Ni release. Initial nickel concentration in the saturation extract of the Welland soil was fairly high, about 20 mg L–1, and it gradually decreased in consecutive extracts. A more abrupt decrease, as well as lower initial concentration of about 0.60 mg L–1, was observed in the Brockman soil. This soil had the most limited ability to replenish removed Ni when made acidic. Quarry muck, at low pH, maintained a relatively constant level of about 1.75 mg Ni L–1 throughout the series of extractions. In contrast to Cu and Ni, Zn in this soil abruptly dropped after the first extraction.



View larger version (37K):
[in this window]
[in a new window]
  		Fig. 6. Concentrations of Ni, Co, Cu, and Zn in four consecutive saturation extracts of Quarry, Welland, and Brockman soils.

 
Although dissolved organic carbon was not measured in soil saturation extracts, the dark brown color of the Quarry muck, brown-yellowish color of Welland, and very light yellowish color of the Brockman soil saturation extracts indicated a high, medium, and very low concentration of dissolved organic matter in the saturation extracts of the Quarry, Welland, and Brockman soils, respectively. In both Port Colborne soils, saturation extracts from the calcareous treatments were darker than the extracts from the low-pH treatments.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
The increase of Ni concentrations in the Alyssum shoots following a pH increase in the Welland and Quarry soils agrees with responses reported by Li et al. (2003a). The same response of Alyssum grown in hydroponic culture was observed by Peters (2000). Alyssum corsicum accumulated increased Ni concentrations in shoots as solution pH increased within the entire pH range tested (pH 5.75–7.5). In contrast, Ni levels in cabbage (Brassica oleracea var. capitata L.) roots and shoots decreased with increasing pH of the nutrient solution. This response of Alyssum conforms with the biotic ligand model that predicts that, if the free metal activity is held constant, increasing solution pH may cause increased binding of metal cations by biotic ligands (biological membrane transporter proteins), due to reduced competition with protons (Di Toro et al., 2001). The trend of increased intake or adsorption of cationic metals on cell walls with pH increase has been reported for aquatic organisms and plants grown in hydroponic culture (Cheng and Allen, 2001; Lopez et al., 2002; Weng et al., 2003). In soil, increasing pH has two opposite effects: increasing metal uptake per unit of free available metal ions on one hand, but decreasing metal solubility and free metal activity in soil solution on the other. Therefore, the usual reaction of plants grown in soil is a decrease of root and shoot metal concentrations as soil pH increases because factors limiting metal solubility prevail. This usual response pattern was followed by crop species grown in Quarry and Welland Port Colborne soils contaminated by a Ni refinery. Nickel concentrations in shoots of oat (Avena sativa L.), wheat (Triticum aestivum L.), corn (Zea mays L.), barley (Hordeum vulgare L.), ryegrass (Lolium perenne L.), radish (Raphanus sativus L.), soybean [Glycine max (L.) Merr.], bean (Phaseolus vulgaris L.), tomato (Lycopersicon esculentum Mill.), Swiss chard (Beta vulgaris var. cicla L.), and redbeet (Beta vulgaris L.) grown in Welland soil decreased markedly with pH increase (Kukier and Chaney, 2004). A similar effect was observed in the case of oat, wheat, and redbeet grown in the limed Quarry muck soil (Kukier and Chaney, 2001). The mild neutral salt extractant, 0.01 M Sr(NO3)2, extracted decreasing amounts of Ni as soil pH increased and was very good in predicting pH-induced changes in Ni availability to 11 crops grown in the Welland soil. Both extractants (DTPA and 0.01 M Sr-nitrate) were effective for the same purpose for oat and redbeet grown in the Quarry muck soil.

Therefore, it can be concluded that the positive correlation observed between soil pH and Ni concentrations in shoots of A. murale and A. corsicum grown in Quarry and Welland soils was a combination of specific soil properties and unique physiological mechanisms of Ni acquisition inherent in these species.

Investigation of the physiological and/or chemical processes occurring on the root surfaces of hyperaccumulator plants was beyond scope of this research. The effect of the differences in soil properties on metal uptake by Alyssum will be a focus of further discussion. Both industrially contaminated soils are characterized by a high organic matter and a low Fe content because they were formed in colder Ontario with high annual inputs of organic matter, while the Brockman soil is high in Fe and contains little organic carbon because the severe infertility of these soils supports only low organic matter inputs, and the warmer and drier climate promotes biodegradation of organic matter. Large pools of Ni, Co, Cu, and Zn in Quarry and Welland soils were extracted by sodium pyrophosphate solution, which suggests that metals are associated with soil organic matter. In the Brockman soil, less than 1.3% of the total Ni and Co was extracted by pyrophosphate. Nickel in serpentine soils low in organic matter is primarily associated with ferric oxides (Kaupenjohann and Wilcke, 1995), which are chemically very inert as demonstrated by the DTPA extraction test for the Brockman serpentine soil in this study (Fig. 5). The charge of ferric oxide surfaces and sorptive properties are modified by pH. At pH below 6.0 sorption of metallic cations on ferric oxides is very low but it rapidly increases as pH exceeds 6.5. A maximum sorption occurs above pH 7.5 to 8.0 (Lo et al., 1994). Quite different effects of pH can be observed in soils high in organic matter. Metals are immobilized at low pH, but raising pH to neutral or alkaline causes gradual increase of metal solubility as metal–humic complexes become soluble (Herms and Brümmer, 1983; Jeffery and Uren, 1983; Spark et al., 1997). In summary, the soil sorptive system dominated by ferric oxide and clay is characterized by increasing retention of metals by the soil solid phase when pH is increased, while in the system dominated by organic matter, metals are strongly sorbed at low pH, but may become more soluble when pH is increased.

Soil saturation extracts are considered a good approximation of soil solution (Krishnamurti and Naidu, 2002). Metal concentrations in saturation extracts of both industrially contaminated Port Colborne soils were much higher than in the serpentine soil. It is generally accepted that metals from recent contamination are more phytoavailable than metals of geogenic origin (Sauerbeck and Hein, 1991). High concentration of organic matter in the industrially contaminated soils was also a factor increasing metal solubility at high pH. Copper, which has the highest affinity to organic matter among the metals studied, exhibited higher concentrations in the saturation extract solutions and 0.01 M Sr(NO3)2 extracts obtained from the calcareous treatments of the Quarry and Welland soils than in the solutions extracted from the low-pH treatments. The opposite was true for the Brockman soil. Organic matter buffered soluble Ni concentrations in the Quarry muck, minimizing soil pH effect. This was evident in the soil saturation- and Sr-nitrate extracts.

Changes in the shoot concentrations of all metals studied suggest the importance of relative abundances of soil organic matter and mineral phases in regulating Ni concentration in soil solution, and consequently, Alyssum response to changes in soil pH. Shoot Ni levels steadily increased with pH increase in the Quarry muck. It is possible that the increased solubility of metal organic chelates following increase of soil pH promoted Ni uptake either by affecting kinetics of Ni binding to biotic ligands on the root surfaces or by increasing Ni mass flow from solid soil phase to soil solution and further to the root binding sites. Krishnamurti and Naidu (2002), studying wheat, found that phytoavailable Cu and Cd were positively correlated with metal fulvic complexes. At comparable levels of free Cu2+ activity, higher concentrations of Cu in the roots of hydroponically grown lettuce were found when a soluble organic matter was included as a component of the nutrient solution (Cheng and Allen, 2001).

In the Welland soil, lower in organic matter, after a rapid increase in shoot Ni with increasing pH, there was a clear plateau of shoot Ni concentrations at high soil pH, where increased adsorption by the mineral phase of the soil counteracted the metal solubilizing effect of organic matter.

The ferric oxide–rich serpentine soil completely counteracted the natural ability of Alyssum to increase Ni concentration in shoots as pH increased. Acidification and reduction are the two processes involved in Fe acquisition by dicot species. Both actions lead to a destabilization of the structure of the ferric oxide crystals and potentially could cause a release of the sorbed or occluded Ni, making this metal more available for root uptake. However, according to Bernal et al. (1994), reducing and acidifying capabilities of the A. murale rhizosphere are smaller than those of radish, a nonaccumulator dicot species. This suggests that Alyssum does not possess the ability to counteract an increased Ni binding by ferric oxide as pH of serpentine soil increases. Peters (2000) reported that the slope of response line "pH versus Alyssum shoot Ni" is lower when Ni concentrations in the hydroponic solution culture are lower. This finding, in conjunction with very low concentrations of Ni in saturation extracts of Brockman serpentine, when soil was made calcareous, helps to explain why plants in this soil responded differently than in Port Colborne soils.

Physiologically mediated patterns of shoot Co concentrations in response to pH changes in the rhizosphere of Alyssum are unknown. However, it may be hypothesized that it could be similar to that of Ni, an increase in shoot Co with pH increase, as these elements are chemically very similar and both are hyperaccumulated by A. corsicum and A. murale to a similar extent if present in equal quantities in the growth medium (Homer et al., 1991). This would explain the increase of Co concentrations in shoots of Alyssum when the Quarry muck was made calcareous. In the Welland soil, Co exhibited a pattern opposite to that of Ni, which was probably a consequence of the fact that a much smaller fraction of total Co was bound to organic matter in this soil. The hypothesis that the ratio of soil organic matter to soil mineral phase or HFO may modify Alyssum response to changes in soil pH is supported by the finding that the addition of HFO to calcareous Quarry and Welland soils decreased Ni concentrations in A. corsicum, and Co concentrations in shoots of both species grown in Quarry muck. In this context, the U-shaped response curves observed in Brockman soil are somewhat surprising, as a decrease of Co concentrations in shoots with increasing soil pH was expected. The increase of Co concentrations in Alyssum shoots in the alkaline region of pH may be an effect of a decreased Ni competition for plant uptake.

Many researchers have noted that element uptake by plants is a complex function of mass transfer, which depends not only on concentration of phytoavailable form of the element in soil solution but also on soil bulk density, water holding capacity, diffusion processes, and root growth pattern (Whiting et al., 2003). Soils in our study differed tremendously in their bulk densities, water holding capacities, and ability to replenish removed Ni, which affected Ni extraction by Alyssum. It can be assumed that the pH treatments resulting in the highest concentration of Ni in Alyssum shoots provided the optimal chemical conditions in soil solution and at the solution–root interface for Ni acquisition by these plants. At the maximum Ni concentration in Alyssum shoots attainable in each soil (acidified Brockman, calcareous Quarry, and Welland), the amount of Ni extracted by Alyssum was proportional to the quantity of metal removed from the soils by four consecutive saturation extracts (Fig. 7) . Results of the single extraction with 0.01 M Sr(NO3)2 (Table 2) did not have any predictive value in this case, probably because soil to solution ratio was constant for all soils, and a single extraction did not reflect differences in soil capabilities to replenish Ni as plants remove it from soil solution.



View larger version (38K):
[in this window]
[in a new window]
  		Fig. 7. The relationship between amount of Ni extracted in four consecutive saturation extracts obtained from the Quarry, Welland, and Brockman soils, and the amount of Ni phytoextracted by shoots of A. corsicum and A. murale. Error bars represent ±standard errors. Soil pH treatments (Brockman acidified, Quarry calcareous, Welland calcareous) resulting in the highest shoot Ni concentrations attainable on each soil were selected for this graph.

 
Further studies including a greater number of industrially contaminated and serpentine soils are necessary to draw more general conclusions on the effect of pH on Ni accumulation by hyperaccumulator Alyssum species. In particular, Ni-contaminated soils selected for further studies should differ in their organic matter and iron content as well as chemical or mineralogical nature of Ni-bearing contaminant. Serpentine soils should be investigated with respect to their relative content of organic matter and ferric oxides. A broad spectrum of Alyssum species should be screened for the response to soil pH changes. All these efforts will not only help to understand the mechanism of Ni uptake by species of Alyssum genus but will have a direct link to the optimization of Ni phytoextraction technology.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Management of soil pH to maximize annual phytoextraction of Ni by the Ni and Co hyperaccumulators Alyssum corsicum and Alyssum murale is important for economic success of phytomining and phytoremediation. A decrease in shoot Ni concentrations with pH increase was observed in the Brockman serpentine soil, and the highest amount of Ni was phytoextracted at the original (unaltered) soil pH of 6.2 to 6.4. Our finding that making Ni-contaminated soils calcareous stimulated Ni phytoextraction has important implications for phytomining and phytoremediation of the land contaminated by the Ni refinery at Port Colborne, Ontario. Liming, which reduced Ni uptake and phytotoxicity to agricultural crops, not only does not interfere with Ni phytoextraction by Alyssum species, but makes it more efficient. Addition of HFO before phytoextraction is not recommended as it may reduce Ni uptake by Alyssum.

Due to much lower concentrations in all soils tested, Co was accumulated in plants to a much lesser extent than Ni. Different pH management is required for phytoextraction of Ni (high pH) and Co (low pH) from the Welland soil. Shoot concentrations of Cu were not affected by pH, and shoot Zn concentrations decreased with soil pH increase. Neither of these metals was hyperaccumulated in the shoots.


   ACKNOWLEDGMENTS
 
This study was funded by Viridian Resources, LLC, Houston, TX, and Inco Limited, Toronto, Ontario. Dr. Kukier was a visiting scientist from the Institute of Soil Science and Plant Cultivation, Pulawy, Poland, during this research. We gratefully acknowledge the assistance of Dr. C.E. Green in maintaining analytical equipment, Dr. Y.-M. Li for recommending Alyssum genotypes for the study, and S. Khader for technical assistance.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2004 33: 1947-1953. [Full Text]  




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (12)
Right arrow Citing Articles via Google Scholar
Google Scholar