Ecotoxicology and Environmental Safety 137 (2017) 225232

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety

journal homepage: www.elsevier.com/locate/ecoenv

Phytoremediation of cadmium and lead-polluted watersheds

Gordon Kersten , Brian Majestic, Martin Quigley

MARK

University of Denver, 2199 S University Blvd., Denver, CO 80210, United States

A R T I C L E I N F O

A B S T R A C T

Keywords:
Salix
Willow
Heavy metals
Metal tolerance
Mine outwash
Bioaccumulation

Abandoned hard rock mines and the resulting acid mine drainage (AMD) are a source of vast, environmental
degradation that are toxic threats to plants, animals, and humans. Cadmium (Cd) and lead (Pb) are metal
contaminants often found in AMD. In our mine outwash water samples, Cd and Pb concentrations were 300 and
40 times greater than EPA Aquatic Life Use water quality standards, respectively. We tested the phytoremediation characteristics, accumulation and tolerance of Cd and Pb contamination, for annual aboveground biomass
harvest of three montane willows native to the Rocky Mountains: Salix drummondiana, S. monticola, and S.
planifolia. We found S. monticola best suited for Pb remediation based on greater growth and tolerance in
response to the low Pb treatment compared to the high Pb treatment. Salix monticola stems also contained higher
Pb concentrations in control treatment compared to S. planifolia. We found S. planifolia and S. drummondiana best
suited for Cd remediation. Salix drummondiana accumulated higher concentrations of Cd in stems than both S.
monticola and S. planifolia. Salix planifolia accumulated nearly 2.5 times greater concentrations of Cd in stems in
control treatment than did S. drummondiana. Salix planifolia also contained more total Cd in stems than did S.
monticola in Cd treatments. Based on our results, S. drummondiana and S. planifolia could aid in reduction of Cd in
watersheds, and S. monticola is better suited than is S. planifolia for aboveground accumulation and tolerance of
Pb pollution.

1. Introduction
Anthropogenic disturbances, including hard rock mining and the
resulting outwash, are signicant sources of metal pollution in the
global environment over the last century (Jung, 2001; Wahsha et al.,
2012; Fig. A.1; US EPA, 2014b; US EPA, 2016). There are over 500,000
abandoned, hard rock mine sites in the U.S., of which 38,500 are on
National Forest System lands, polluting watersheds and ecosystems for
decades or more (Carr, 2005). Acid mine drainage (AMD) from the
abandoned mines contains toxic concentrations of metals that cause
signicant environmental damage because the metals cannot be
chemically degraded (Salt et al., 1995). Remediation of abandoned
mine lands and watersheds is important, especially in Colorado where
mining activities have been active since the mid-1800s. One-third of
EPA's Region 8 (6 states: MT, WY, ND, SD, CO, and UT) superfund sites
are in Colorado as of 2014 (US EPA, 2014a). 2,100 km of streams in
Colorado are signicantly polluted with AMD and many metals. 89% of
total maximum daily load of Colorado water impairments are due to
many of the 23,000 abandoned mine features (CDPHE, 2012).
Metal contaminants not only accelerate environmental degradation
but also are detrimental to humans as well (Feleafel and Mirdad, 2013;
Jarup, 2003; Leao and Pang, 2010; Thompson and Bannigan, 2008;

Mndez-Armenta and Ros, 2007; Rosas et al., 1984). Dust from dry and
loose tailings causes and exacerbates respiratory diseases in human
populations near abandoned mine disposal sites (Mendez and Maier,
2008). More specically, Cd and Pb are common, biologically nonessential metal pollutants in mine tailings and outwash (Das et al.,
1997, Flora et al., 2012; Fowler, 2009; Nagajyoti et al., 2010).
According to the most recent USEPA watershed assessment for
Colorado (US EPA, 2010), nearly 1,300 and 300 km of streams are
impaired by Cd and lead, respectively. As of 2011, 938 million gallons
of water per year are treated near abandoned mining sites in the
Colorado, which is eective but expensive (CDPHE, 2012).
The US Government Accountability Oce found that the US EPA
spent $2.2 billion on abandoned hard rock mine land remediation
between 1997 and 2008 (Nazzaro, 2008). The most common approach
to mine tailings and waste is piling and containing the waste tailings
without chemical or metal removal treatments (Mendez and Maier,
2008). Isolating tailings with embankments does prevent the escape
and spread of pollutants but does not decrease or remove them
(Cunningham and Berti, 1993).
Phytoremediation is a cheap alternative to complement common,
conventional methods. Phytoremediation is a broad term for using
plants for cleanup of environmental metal pollution through phytoex-

Corresponding author.
E-mail address: gordonkersten@gmail.com (G. Kersten).

http://dx.doi.org/10.1016/j.ecoenv.2016.12.001
Received 12 April 2016; Received in revised form 21 November 2016; Accepted 1 December 2016
0147-6513/ 2016 Elsevier Inc. All rights reserved.

Ecotoxicology and Environmental Safety 137 (2017) 225232

G. Kersten et al.

genets of S. drummondiana, and 18 genets of S. planifolia. Individual

genets (shrub masses) of same species were at least 20 m apart, to
ensure genetic diversity. Genets were tagged in the eld with sequential
numbers on aluminum tags. Furthermore, previous studies of willow
phytoremediation focused on plants from a limited geographic or
genetic range. We selected plants from 10 sites from both eastern and
western slopes of the Rocky Mountains with at least 18 unique genets of
each species, with species often intermixed or at the same site. These
sites oered various redox states and metal concentrations.

traction, rhizoltration, or phytostabilization (Salt et al., 1995). Here

we focused on phytoextraction and permanent removal of metal
contaminants using native montane willows (Salix spp.) of the Central
Rockies in Colorado, near abandoned mine lands at elevations above
2,400 m. Phytoextraction is the ability of plants to accumulate metal
contaminants in their aboveground biomass, for harvesting and contaminant removal (Salt et al., 1998). Best practices of phytoremediation, such as phytoextraction, use regionally appropriate plants including Brassicaceae (mustards), Poaceae (grasses), and Salicaceae (poplars
and willows) (Salt et al., 1995). Willows are the dominant riparian
vegetation at these elevations and found along EPA impaired and BLM
high priority watersheds near abandoned mine lands above elevations
of 2,400 m in Colorado. Willows are also easy to propagate and
establish in the eld with very deep and extensive root systems
(Vangronsveld et al., 2009). Willows are known phytoremediation
agents; they accumulate and concentrate metals, grow rapidly with a
relatively high biomass (Pulford and Watson, 2003; Punshon and
Dickinson, 1997), and translocate metals from roots to aboveground
biomass (Wahsha et al., 2012).
Here we investigated the phytoremediation abilities of three
common willows native to Colorado: Salix drummondiana, S. monticola,
and S. planifolia for suitability in reducing Cd and Pb contamination
along the 1,600 km of cadmium and lead-polluted streams across
Colorado watersheds. These phytoremediation abilities would require
annual harvests before leaf fall for optimal reduction of environmental
Cd and lead. Our study is innovative as it is the rst study to directly
compare these three dominant willow species phytoextraction and
tolerance characteristics for Cd and Pb contamination. We tested the
phytoremediation characteristics of these willows via hydroponic
experiments, which are useful for screening willows tolerance and
accumulation characteristics (Huang and Cunningham, 1996; Watson
et al., 2003; Dos Santos et al., 2007; Zhivotovsky et al., 2011a).
Here, we propose four hypotheses. First, three common native
willow species will dier in stem and leaf accumulation concentrations
and total amounts (concentration x nal mass) after growing in Cd and
Pb treatments. Second, three common willow species will dier in Cd
and Pb stem concentrations after growing in low and high metal
concentration levels of Cd and lead. Third, three common native willow
species dier in their relative tolerance (growth) to environmental
representative Cd and Pb treatments. Lastly, three common native
willow species dier in total metal amounts and concentrations of Cd
and Pb without any additional access to metals in the control treatment.

2.2. Hydroponic greenhouse experiment

We conducted a four week accumulation and tolerance greenhouse
experiment with the three willow species of interest: S. drummondiana,
S. planifolia, and S. monticola. We brought cuttings to the greenhouse at
the University of Denver immediately after eld collections, and placed
in bunches of 12 cuttings in each deepot cone (6.4 cm by 36 cm) in
every other slot in 20 slot support trays (Stuewe & Sons, Corvallis, OR).
We used cupcake papers to prevent cuttings from slipping out of the
cones. We submerged each cone in each slot tray in 17 cm of deionized
(DI) water supplemented with 132 mL FloraGro (2-1-6 NPK ratio)
nutrient solution per 100 L of water. University of Denver's Olin Hall
greenhouse has natural, south facing light. We completely replaced the
water every other week and supplemented with FloraGro fertilizer once
a week.
We grew cuttings for 6 weeks for root development. We discarded
cuttings that did not show signs of root or leaf development. For living
cuttings, we assigned a root score (04) at the beginning and end of
experiment (Fig. A.2). For each genet, we chose ve living cuttings
giving them a letter (a, b, c, d, and e) and then assigned them randomly
to each treatment. We successfully rooted 405 cuttings of 828 collected
from the 69 willow genets.
We conducted the experiment with ve hydroponic treatments:
control (DI water; FloraGro), low Cd (11 ppb or g/L; 0.10 M), high
Cd (300 ppb; 2.56 M), low Pb (15 ppb; 0.07 M), and high Pb (145
ppb; 0.70 M). Hydroponic experiments that use unrealistically high
metal concentrations can lead to forced metal accumulation and their
results have no biological relevance (van der Ent et al., 2013). Here, we
used relatively lower (~1 M) concentrations of metals, representing
more environmentally and biologically relevant levels (Table A.1; US
EPA, 2014a, 2014b; US EPA, 2016). We added metals via a stock
solution of 7.6 mM of cadmium chloride (CdCl2) and 2.2 mM of lead
chloride (PbCl2) weekly to obtain the treatment concentrations. We
placed one cutting from three unique genets of each of the three species
in a 26.5 L Sterilite storage bin (block), for a total of 9 cuttings in each
block. We established a randomized block factorial design, randomizing
the location of each cutting. Out of the 69 willow genets, there were
eight genets that had 10 or more successfully growing cuttings that
were included as replicates for genets in treatments and were nested for
the genet. We lled remaining blocks with these replicate cuttings. We
replicated each treatment block nine times for a total of 45 blocks and a
total of 405 cuttings.
We conducted the experiment for four weeks from December 2014
to January of 2015, with diurnal temperatures ranging from 16 C to
23 C. Growing lights, in addition to natural light, were on for 12 h per
day, from 6 am to 6 pm. We randomly organized blocks on the
greenhouse benches, and moved all blocks each week to randomize
lighting conditions in the greenhouse and minor irregularities of depth
proles in the blocks.
We counted leaves weekly, survival at week 3 and 4, and the
biomass of each cutting before and after the experiment for the percent
change. Survival was the presence of healthy roots or leaves (Figs. A.2,
A.3). We only counted fully developed, healthy leaves. Leaves that were
shriveling or showing chlorosis were considered dead and were not
counted. At the beginning and end of the experiment, we recorded
biomasses of the entire cutting to the nearest 0.1 g.

2. Materials and methods

2.1. Collections
We sampled 160 individual willow genets (unique individual willow
shrubs, not connected by roots) in November 2013, and from April to
June 2014, at elevations above 2,400 m throughout ve counties in
Colorado: San Juan, Ouray, Lake, Clear Creek, and Summit. These areas
of interest were near abandoned mine lands and designated as high
priority (BLM, 2013) or impaired (US EPA, 2010) watersheds (Fig. 1).
We collected all unique willow species in an area within 5 m of
waterway edge. If there were dierent willow species, we collected
from no more than three unique genets (not connected by roots) from
each of the dierent species. If there was only one species in an area, we
collected no more than three unique genets of that species.
We collected for diversity in the areas of interest in April June
2014. Of the 160 individual willow collections, we identied eight
unique willow species. We found the three most common willow
species to be Salix monticola, S. planifolia, and S. drummondiana. Once
we identied the three most common willow species, we collected 12
cuttings (2040 cm in length) from 69 genets of the three most common
willow species in August and October of 2014 for the greenhouse
experiment. We collected cuttings from 32 genets of S. monticola, 19
226

Ecotoxicology and Environmental Safety 137 (2017) 225232

G. Kersten et al.

Fig. 1. (A) Map of Colorado showing BLM high priority watersheds (Upper Arkansas River and Upper Animas River) and US EPA impaired watersheds (Uncompahgre River and Peru
Creek). Counties are highlighted in gray where water and willow collections were made. (B) Zoomed area of willow collection sites and Peru Creek and the Arkansas River headwaters in
Lake, Summit, and Clear Creek counties. (C) Zoomed area of willow collection sites and Animas and Uncompahgre Rivers in San Juan and Ouray counties. Maps were created in ArcMap
version 10.2.2.

dierences.
Our methodology compares each cutting and suciently takes into
account initial soil redox and metal states. Metal uptake in willows can
be aected by their immediate habitat, such as soil redox states
(Vandecasteele et al., 2004). To further suciently eliminate initial
concentrations, we compared each cutting in a metal treatment directly
to cuttings from the same genet with the same initial concentrations in
the control treatment (Ling et al., 2011; Tlusto et al. 2007; Dos Santos
Utmazian et al., 2007; Pulford et al., 2001; Vyslouzilova et al., 2003;
Zhivotovsky et al., 2011b). These were cuttings of the same size from
the same genet compared after the experiment in growth and metal
accumulation. For example, Cd in cuttings increases, relative to initial
concentrations, in contaminated media (metal treatments) but not in
uncontaminated media (control) (Vandecasteele et al., 2015). It is also
important to note that during the experiment, metals can be translocated from stems to leaves in cuttings collected in contaminated
locations (Vandecasteele et al., 2015). With our methodology of
comparing cuttings from the same genet (same location, soil redox
status, and contamination), we can conclude the dierences in accumulation are due to the treatment in the experiment, not the initial
concentrations.
Biomass was measured on a dried weight (DW) basis and compared
to the initial mass of each cutting. Biomass percent change was
calculated: 100((Final MassInitial Mass)/Initial Mass). Biomass
percent change was also calculated as dierences between metal
treatment cuttings biomass percent change and their respective control
treatment cutting's biomass percent change (cutting from same genet in
control treatment). A negative value means the cutting of a genet grown
in a metal treatment had less biomass growth than a cutting from the
same genet in the control treatment. It is important to note that we
excluded four outlier data points from our results for the impossible
accumulation and biomass changes they represented. For weekly leaf
counts, we used a MANOVA with repeated measures of each week using
the Roy's Max Root test statistic.

2.3. Preparation of stem and leaf samples for ICP-MS analysis

For the purpose of permanent removal of metal contamination, we
focused on above-ground biomass accumulation in two compartments:
stems and leaves. Stems included bark, wood, and shoots. Leaves
included leaves attached to stems at the end of the experiment. At the
conclusion of the greenhouse experiment, we dried the cuttings in
paper bags in an oven at 70 C for 72 h. We nely ground the leaves and
stems of the cuttings separately using a bead beater (BioSpec MiniBeadbeater-16) courtesy of Denver Botanic Gardens. In the atmospheric
particulate matter lab at the University of Denver, we prepared the
samples for inductively coupled plasma mass spectrometry (ICP-MS) by
digesting samples with acids. For each leaf and stem sample, we massed
out samples between 5 and 40 mg and then added 750 L nitric acid
(HNO3), 250 L hydrochloric acid (HCl), 100 L hydrouoric acid
(HF), and added 100 L hydrogen peroxide (H2O2) dropwise, due to
reaction volatility. We then combined 1 mL H2O2 with 10 mL of
ultrapure H2O (18.0 M ohm-cm) in each digestion chamber (total of
10 for each round) before heating to 210 C in the acid digestion for 1 h
and 25 min. After acid digestion, we diluted samples to 15 mL by
adding the digested sample to 13.8 mL with ultrapure ( > 18.0 M cm)
H2O. We measured the validity of the prepared samples via ICP-MS
with soil and spinach standard reference materials (SRMs) as well as
reference metal standards for Cd and Pb at low, medium, and high
concentrations: 10, 50, and 100 parts per billion (ppb or g/L)
respectively. We found excellent recovery of our references.
2.4. Analysis
We analyzed the data using a mixed model with JMP (Version 11.0).
We analyzed the biomass percentage change and metal accumulation in
stems and leaves using the mixed model. Fixed eects comprised metal
type, metal level, and willow species with a full factorial of all three
interactions and the willow genet was a random eect. We initially
tested multiple random eects such as site location, county, and block
as well, but they did not contribute to the model (0%). We then omitted
these random eects from the mixed model to provide more analysis
power. We used the Tukey HSD for post hoc analysis of signicant

3. Results
We found four signicant dierences testing the accumulation and
227

Ecotoxicology and Environmental Safety 137 (2017) 225232

G. Kersten et al.

Table 1
Mixed Model analysis results of fixed main and interaction effect sources for a) stem
accumulation concentration (Fig. 2a), b) leaf accumulation concentration (Fig. 2b), c)
total mass of metal per cutting (concentrationnal mass; Fig. 4), and d) biomass percent
change (Fig. 5). Random eect was willow genet.
Source

DF

DFDen

F Ratio

Prob > F

Willow Species
Metal Type
Willow Species*Metal Type
Metal Level
Willow Species*Metal Level
Metal Type*Metal Level
Willow Species*Metal Type*Metal
Level

2
1
2
1
2
1
2

66.4
197.0
197.0
197.0
197.0
197.0
197.0

2.63
0.59
6.43
0.14
1.55
0.34
0.33

0.08
0.44
0.002
0.71
0.22
0.56
0.72

Willow Species
Metal Type
Willow Species*Metal Type
Metal Level
Willow Species*Metal Level
Metal Type*Metal Level
Willow Species*Metal Type*Metal
Level

2
1
2
1
2
1
2

21.5
58.2
58.1
54.8
54.8
54.2
54.1

4.07
0.32
0.75
1.21
0.86
0.32
0.23

0.03
0.58
0.48
0.28
0.43
0.57
0.80

Willow Species
Metal Type
Willow Species*Metal Type

2
1
2

77.6
264.8
264.8

1.22
65.89
7.18

0.30
< 0.0001
0.0009

Willow Species
Metal Type
Willow Species*Metal Type
Metal Level
Willow Species*Metal Level
Metal Type*Metal Level
Willow Species*Metal Type*Metal
Level

2
1
2
1
2
1
2

66.2
196.5
196.5
196.5
196.5
196.5
196.5

0.49
4.97
0.30
0.10
2.35
3.80
4.31

0.61
0.16
0.74
0.75
0.10
0.05
0.01

Fig. 2. (a) Stem concentration of cadmium and lead for three willow species: S.
drummondiana, S. monticola, and S. planifolia. The stem concentrations are the dierences
between cuttings in cadmium or lead treatments with cuttings of the same genet in
control (no metal) treatment. Mixed Model: Willow Species*Metal Type (F=6.43, df=2,
p=0.002; Post hoc Tukey HSD - levels not connected by same letter (e.g. A, B) are
signicantly dierent). One error bar equals one standard error from the mean. (b) Leaf
concentration of cadmium and lead (ppb) for three willow species: S. drummondiana, S.
monticola, and S. planifolia. The leaf concentrations are the dierences between cuttings in
cadmium or lead treatments with cuttings of the same genet in control (no metal)
treatment. Mixed Model: Willow Species (F=4.07, df=2, p=0.03; Post hoc Tukey HSD
species not connected by same letter (e.g. A, B) are signicantly dierent). One error bar
equals one standard error from the mean.

tolerance of Cd and Pb for the three willow species (Table 1). We found
a signicant interaction with willow species and metal type for stem
accumulation concentration and total Cd per cutting. There were
signicant dierences among species for leaf accumulation concentration. We also found a signicant dierence for biomass change for the
interaction between Willow Species, Metal Type, and Metal Level.
Salix drummondiana accumulated more Cd in stems than did both S.
monticola and S. planifolia (Willow Species*Metal Type:F=6.43, df=2,
p=0.002; Table 1; Fig. 2a). We found no dierences in Pb accumulation in stems between the treatments (Fig. 2a).
Similar to stem accumulation results, S. drummondiana accumulated
higher Cd concentrations in leaves than did S. planifolia (Willow
Species: F=4.07, df=2, p=0.03; Table 1; Fig. 2b). Salix drummondiana
was the only species to have a greater concentration of Cd in leaves
than control in the pooled metal treatments. We found no dierences in
Pb accumulation concentration in leaves.
We found that neither the environmental Cd nor Pb concentration
levels in the water aected the accumulation concentration of these
metals in stems for any of the species (Fig. 3a; b; Table 1).
Additionally, S. planifolia contained more total Cd per cutting
(concentration x nal mass) than did S. monticola for Cd treatments
(Willow Species*Metal: F=7.18, df=2, p=0.0009; Fig. 4).
We also found a signicant interaction for biomass percentage
growth (Willow Species * Metal Type * Metal Level: F=4.31, df=2,
p=0.01; Table 1; Fig. 5). Salix monticola demonstrated greater tolerance with a 2% increase in biomass over the four week greenhouse
experiment for cuttings in the Low Pb treatment, compared to the over
5% biomass loss of cuttings in the High Pb treatment. All three species
reacted similarly to the pooled metal treatments of Cd and Pb with an
overall decrease in biomass growth relative to controls (Fig. A.4).
There were also dierences in stem accumulation concentrations for
stems in metal treatments (Willow Species*Metal Type: F=29.22,
df=2, p < 0.0001; Fig. 6a). Salix monticola contained higher concen-

trations of Pb than did S. planifolia (Fig. 6a). We found that all three
species accumulated more and contained higher concentrations of Cd in
stems than each did for lead. Similarly, there were dierences in control
stem concentrations of Cd and Pb as well, represented by cuttings in
control treatments (Willow Species*Metal Type: F=5.93, df=2,
p=0.004; Fig. 6b). Salix planifolia contained the highest concentration
of Cd in stems, twice as much as cuttings of S. drummondiana. Salix
monticola contained the highest Pb stem concentration, nearly two
times that of cuttings of S. drummondiana and three times that of
cuttings of S. planifolia.

4. Discussion
Based on our eld sampling and experimental results, S. drummondiana and S. planifolia should be equally investigated for phytoremediation in reducing metal contamination from 1,600 km of cadmium and lead-polluted streams in Colorado in addition to S. exigua, S.
monticola, and S. geyeriana in Colorado. Using diverse and native plant
material is essential for optimal phytoremediation, and these species
provide diverse options for phytoremediation through their metal
tolerance and aboveground biomass metal accumulation for permanent
removal of Cd and Pb. Overall, all three species contained higher
concentrations of Cd than Pb, ranging from 2.5 to 19 times higher
concentrations of Cd than Pb in stems based on concentrations in
cuttings in control treatment (no metal treatment). Also, all willow
species contain similar amounts of total Pb amount per cutting in metal
228

Ecotoxicology and Environmental Safety 137 (2017) 225232

G. Kersten et al.

Fig. 5. Biomass percent changes over four week hydroponic greenhouse experiment of
cuttings representing three willow species: S. drummondiana, S. monticola, and S.
planifolia. The calculated dierence of percent changes of biomass is between cuttings
in each of the four treatments (Cadmium High, Cadmium Low, Lead High, Lead Low)
subtracting biomass percent change of cuttings from the same genet in the control
treatment (no metals). Biomass percent change was calculated: 100*((Final Mass - Initial
Mass)/Initial Mass). Mixed Model: Willow Species*Metal Type*Metal Level (F=4.31,
df=2, p=0.01; Post hoc Tukey HSD, Metal Treatment Levels * Willow Species not
connected by the same letter (e.g. A, B) are signicantly dierent. One error bar equals
one standard error from the mean.

Fig. 3. (a) Stem concentrations of cadmium in two levels of cadmium treatments for three
willow species: S. drummondiana, S. monticola, and S. planifolia. The stem concentrations
are the dierences between cuttings in low and high cadmium treatments with cuttings of
the same genet in control (no metal) treatment. Mixed Model: Willow Species (F=4.03,
df=2, p=0.02). Post-hoc Tukey HSD, willow species not connected by same letter (e.g. A,
B) are signicantly dierent. One error bar equals one standard error from the mean. (b)
Stem concentrations of lead in two levels of lead treatments for three willow species: S.
drummondiana, S. monticola, and S. planifolia. The stem concentrations are the dierences
between cuttings in low and high lead treatments with cuttings of the same genet in
control (no metal) treatment. One error bar equals one standard error from the mean.

Fig. 4. Total metal (g) at completion of experiment per cutting (metal concentration x
nal cutting mass) for cadmium and lead treatments for three species: S. drummondiana,
S. monticola, and S. planifolia. Mixed Model: Willow Species*Metal Type (F=7.18, df=2,
p=0.0009). Post-hoc Tukey HSD, Willow Species*Metal Type not connected by same
letter (e.g. A, B) are signicantly dierent. One error bar equals one standard error from
the mean.

Fig. 6. (a) Stem concentrations of cadmium and lead after four week greenhouse
experiment of cuttings in cadmium and lead treatments for three willow species: S.
drummondiana, S. monticola, and S. planifolia. Mixed Model using log transformed metal
concentration values: Willow Species*Metal Type (F=29.22, df=2, p < 0.0001). Posthoc Tukey HSD, levels not connected by same letter (e.g. A, B) are signicantly dierent.
One error bar equals one standard error from the mean. (b) Stem concentrations of
cadmium and lead cuttings in the control treatment for three willow species: S.
drummondiana, S. monticola, and S. planifolia. Metal concentrations in cuttings in control
treatment represent eld concentrations. Mixed Model using log transformed metal
concentration values: Willow Species (F=7.51, df=2, p=0.001), Metal (F=131.12,
df=1, p < 0.0001), Willow Species*Metal Type (F=5.93, df=2, p=0.004). Post-hoc
Tukey HSD, levels not connected by same letter (e.g. A, B) are signicantly dierent. One
error bar equals one standard error from the mean.

treatments, but S. planifolia contained more total Cd per cutting than

did S. monticola. Our four week accumulation and plant tolerance
experiment demonstrated S. drummondiana to be a better accumulator
of higher Cd concentrations than both S. monticola and S. planifolia
shown by stem and leaf concentrations relative to control treatments.
Salix planifolia is also highly recommended for Cd remediation based on
the fact it contained 2.5 times higher control stem concentrations of Cd
than S. drummondiana and nearly 2 times more total Cd than S.
monticola in Cd treatments. All three willow species accumulated low
Pb concentrations in leaves during the experiment.
There were no statistical dierences in total metal amounts between
metal and control treatments. Even though stems are increasing in

metal concentrations relative to control, they are losing biomass

slightly relative to control. Therefore, the total mass of metals between
treatment and control stems are not dierent, as shown in Fig. A.5.
However, in Fig. 4, we see dierences between total metal amounts in
metal treatment stems. Salix planifolia contained more total Cd (con229

Ecotoxicology and Environmental Safety 137 (2017) 225232

G. Kersten et al.

centration x nal mass) than did S. monticola in Cd treatments. It is also

important to note that there were no dierences in biomass percent
change between species and treatments (Fig. A.6).
Metal accumulation concentrations during the experiment were the
focus here normalizing for varied cutting sizes (20 40 cm in length)
and leaf numbers, though total amount of metals per stem cutting per
species were addressed as well (Fig. 4). Control stems contained metal
concentrations not aected by metal treatments, as they were taken
from the eld and did not have access to metal contamination thereafter (Fig. 6b). Salix monticola is the best candidate of the three species
for Pb remediation. Salix monticola contained and maintained nearly 6
times higher stem concentrations of Pb in control treatment than did S.
planifolia (Fig. 6b). This suggests S. monticola accumulates and maintains more Pb over longer period of time since the high variety of site
locations and soil redox states had no eect on accumulation. Salix
monticola also had similar accumulation concentrations as both of the
other species with a positive accumulation concentration of Pb relative
to control (Fig. 3b) and mixed biomass growth outcomes between low
and high Pb treatments compared to control (Fig. 5). These results for S.
monticola suggest S. monticola loses biomass in high environmental Pb
contamination but still can maintain (not necessarily accumulate in
short period of time) higher concentrations and total amount of Pb in
the plant compared to S. planifolia (Fig. 6), which is valuable for
removing contaminants in aboveground biomass from environment
when harvested each year before leaf fall. Conversely, in Low Pb
environmental concentrations, S. monticola displayed increased growth,
paired with its ability to maintain higher Pb concentrations compared
to S. planifolia is ideal for permanent Pb removal through harvesting of
plant materials.
Negative relative concentrations, as shown in Figs. 2 and 3, indicate
that stems and leaves of S. monticola and S. planifolia in Cd treatments
contain lower Cd concentrations than their cuttings from the same
genet in control treatment. This suggests the cuttings in Cd treatments
for S. monticola and S. planifolia are translocating more Cd out of their
stems to the leaves than they are accumulating Cd to the stems from the
Cd water treatment. This could be due to plants not translocating metals
from stems (bark or shoots) in the control treatment, while more Cd was
translocated to leaves than taken up from contaminated media
(Vandecasteele et al., 2015), or simply translocating Cd to dying or
senescing plant parts.
Cuttings of S. drummondiana in Cd treatments, however, are
accumulating more Cd from the contaminated water treatment than
they are translocating to leaves (Figs. 2a; 3a). At the same time though,
we see in Fig. 2b that leaves of S. drummondiana contain higher
concentrations compared to their control leaves, showing the translocation to leaves occurs at a higher amount in Cd treatments for S.
drummondiana.
Willows are capable of metal tolerance by increasing biomass and
growth over time (Punshon and Dickinson, 1997; Shanahan et al., 2007;
Meiman et al., 2012; Yang et al., 2015). However, in this experiment,
most cuttings from all three species lost biomass and leaves over the
period of the experiment in all treatments (Figs. A.6, A.7). One possible
source of biomass loss across the board could be the cupcake papers
that were wrapped around roots of the cuttings in the rst ve days of
the experiment. The papers may have caused cavitations in the cuttings
disrupting growth. The papers were removed immediately once the
wrapping at the base of cutting around roots was noticed. Biomass loss
is not typical of willows in metal accumulation and tolerance experiments. For example, S. drummondiana increased in biomass over a two
month experiment on amended tailings (Meiman et al., 2012) and S.
monticola had an 87% survival rate after 4 years of growth on amended
mine tailings consisting of both Cd and lead, at concentrations higher
than our experiment. However, it has recently been shown that a
decrease in biomass may also be considered an appropriate metal
tolerance, by adjusting metabolism to avoid cellular damage suggesting
a more long-term tolerance (Evlard et al., 2014).

Other metals did not have an eect on biomass growth in cuttings.

In a subset of plant stem samples over the total range of biomass
percent change, stem concentrations of metals, including manganese,
magnesium, iron, copper, zinc, strontium, and barium (non-treatment
metals). These metals were not statistically signicant contributors to
biomass percent changes: Fit Least Squares (JMP Version 12.2) for
Willow Species (F=0.8072, df=2, p > 0.05), Metal Type (including
control) (F=0.0947, df=2, p > 0.05), Willow Species*Metal Type
(F=1.0173, df=4, p > 0.05), manganese (F=0.4330, df=1, p >
0.05), magnesium (F=3.9071, df=1, p > 0.05), iron (F=0.1775,
df=1, p > 0.05), copper (F=0.9067, df=1, p > 0.05), zinc
(F=0.1488, df=1, p > 0.05), strontium (F=3.3490, df=1, p >
0.05), and barium (F=1.1461, df=1, p > 0.05). Average concentrations of these metals are shown in Table A.2. These non-treatment
metals were compared with plant toxicity ranges. Manganese concentration levels were in the lower range of critical toxic limits for S.
planifolia and S. drummondiana, whereas copper concentrations for all
species were below toxic limits (Macnicol and Beckett, 1985). There is
much variability within willow species. For example, Shanahan et al.
(2007) showed that toxicity levels causing 50% reduction in biomass
(PT50shoot) could not be determined for Zn or Mn, due to erratic data
that could not be t to regression model.
We tested the broad dierences in metal accumulation and tolerance between these native willow species with high genetic and
geographic diversity. We addressed these three common montane
willow species at a signicant scale by using 1832 unique genets of
each species and by collecting from ten sites throughout ve counties in
Colorado. The three willow species diered in observed Cd concentration ranges from the stems in the control treatment: S. drummondiana
(2699,274 g Cd/kg DW), S. planifolia (20521,161 g Cd/kg DW),
and S. monticola (72111,628 g Cd/kg DW). The three willow species
diered in observed Pb concentration ranges from the stems in the
control treatment: S. drummondiana (335,557 g Pb/kg DW), S.
planifolia (2051,159 g Pb/kg DW), and S. monticola (847,779 g
Pb/kg DW). This provides overarching trends of each species, rather
than results of specic genets or ramets from one location representing
the entire species.
Longer-duration experiments demonstrated higher accumulation
abilities of these willow species for Pb and Cd. For example, Bourret
et al. (2009) demonstrated S. monticola leaves of staked cuttings
concentrated nearly 13 times higher concentrations of Pb after two
years of growth on amended mine tailing deposits than our concentrations of Pb in leaves after our experiment. This supports that the Pb
accumulation continues with duration of metal exposure. Meiman et al.
(2012) reported S. drummondiana concentrated twice the Cd concentration in its aboveground biomass in 2 months than we found in stems of
S. drummondiana. As for S. monticola, Bourret et al. (2009) found nearly
9 times higher concentration of Cd in leaves pre-rooted cuttings after
two years, and Boyter et al. (2008) found 11.5 times and 10 times
higher concentrations of Cd after grown in amended tailings for 4
months in live and senesced leaves, respectively, compared to our leaf
concentrations. In the same experiment, Boyter et al. (2008) also found
17 times higher concentrations of Cd in bark after four month greenhouse study compared to Cd concentrations in stems grown in Cd
treatments in our experiment.
There are many willow species that are relevant for ecosystems
aected by AMD, depending on soil type, metal contamination level
and type, and location. For other willow investigations into metal
remediation, see Pulford et al., 2002, Zalesny and Bauer, 2007, Mleczek
et al., 2009, Kuzovkina et al., 2010, and Meiman et al., 2012. For
example, Meiman et al. (2012) found both S. bebbiana (21.5 mg Cd/kg
DW) and S. geyeriana (24.3 mg Cd/kg DW) contained more Cd than
both S. drummondiana (7.12 mg Cd/kg DW) and S. planifolia (7.36 mg
Cd/kg DW) in aboveground biomass after growing in amended mine
tailings, but didnt nd a dierence between species for Cu, Zn, or Pb
concentrations in aboveground or belowground biomass. Pulford et al.
230

Ecotoxicology and Environmental Safety 137 (2017) 225232

G. Kersten et al.

intellectual content.

(2002) found multiple Salix hybrids that accumulated higher concentrations of Cd after 12 years grown in sewage sludge soil with heavy
metal contamination, with one hybrid Calodendron accumulating
32 mg Cd/kg DW in bark, which is 7, 8, and 10.5 times more than
did S. planifolia, S. monticola, and S. drummondiana in aboveground
biomass in our experiment, respectively.
Increased metal accumulation in leaves can cause harm to herbivores. But plants accumulating metals as an herbivore deterrent is
already a leading reason for the accumulation known as the Elemental
Defense Hypothesis, elemental allelopathy, and simply as a plant
chemical defense (Martens and Boyd, 1994; Boyd and Martens,
1998). Other natural deterrents exist for willows such as salicin
(Markham, 1971) and phenol glycoside compounds (Tahvanainen
et al., 1985). Herbivore risks can also be minimized further with proper
land management is used, such as exclosures or predator urine.
Furthermore, it is possible to breed willow clones that have some of
these additional herbivore repellent characteristics (Greger and
Landberg, 1999). Genetically manipulating willows with anti-herbivorous characteristics, such as salicylate-rich leaves or higher phenol
glycosides in stems, with willows that have metal accumulation properties is a worthwhile phytoremediation venture.

Acknowledgements
We acknowledge Dr. Anna Sher for her help with data analysis and
experimental design. We also acknowledge Dr. Shannon Murphy for
experimental design and writing guidance. We acknowledge the Denver
Botanic Gardens, especially Dr. Melissa Islam and Dr. Jennifer Neale,
for their help and use of their vast herbarium and bead beater
equipment. We thank the SIGMA-XI Grant-in-Aid of Research program
and the Ann Smeltzer Charitable Trust for their contributions for
research. Dr. Gwen Kittel was instrumental in helping with willow
taxonomy and identication. We also thank Dr. Benton Cartledge for
his invaluable guidance and help with ICP-MS preparation and analysis.
We thank Tessa Kersten for providing help in numerous eld work trips.
In addition, we thank Ross Minter for his assistance in eld work
collections and photographs.
Appendix A. Supporting information
Supplementary data associated with this article can be found in the
online version at doi:10.1016/j.ecoenv.2016.12.001.

4.1. Future directions

References
Applying a calculation methodology for metal removal eciency of
willow stands in an actual waterway would greatly benet phytoremediation quality analysis moving forward. For example, Dean et al.
(2013) quantied metal removal remediation of a polluted river using
multiple samples sites near a natural wetland over 14 years. Yao et al.
(2009) calculated eciencies of heavy metal removal using an equation
taking into account metal concentrations and volume of contaminated
mine water. A study building on these using a combination of willows
would be a worthwhile investigation as a template for metal removal
eciency and phytoremediation analysis.
A genet database containing information to native ranges, metal
hyperaccumulation abilities and herbivore deterrent characteristics
would be an essential next step for optimizing phytoremediation.
Further research into manipulating and hybridizing hyperaccumulating
willow genets with herbivore-deterrent willow genets would be a
worthwhile research endeavor. Collaborative phytoremediation techniques should also be included in the genet database, such as the
interaction with microorganisms and organic additions. Also senesced
leaves are of interest with nearly 16 and 46 times more Pb accumulated
and concentrated in senesced leaves than living leaves demonstrated by
Boyter et al. (2008). Lastly, willows are fast-growing, metal hyperaccumulating woody plants already commonly used in short rotation
coppice biomass production making them ideal candidates for the
economic benet of biomining. With all future research endeavors, the
investment in researching mitigation restoration for reducing heavy
metal contamination in watersheds to pre-mining levels remains an
important environmental priority.

Bureau of Land Management (BLM), 2013. Abandoned Mine Lands Program: Feasibility
Study for AML Inventory Validation and Physical Safety Closures.
Bourret, M.M., Brummer, J.E., Leininger, W.C., 2009. Establishment and growth of two
willow species in a riparian zone impacted by mine tailings. J. Environ. Qual. 38,
693701. http://dx.doi.org/10.2134/jeq2008.0153.
Boyd, R.S., Martens, S.N., 1998. The signicance of metal hyperaccumulation for biotic
interactions. Chemoecology 8, 17. http://dx.doi.org/10.1007/s000490050002.
Boyter, M.J., Brummer, J.E., Leininger, W.C., 2008. Growth and metal accumulation of
geyer and mountain willow grown in topsoil versus amended mine tailings. Water Air
Soil Pollut. 198, 1729. http://dx.doi.org/10.1007/s11270-008-9822-9.
Carr, M., (USDA Forest Service), 2005. A multiheaded beast: abandoned mine lands and
the challenge of water protection, Wildland Waters Winter, 4, pp. 120
CDPHE (Colorado Department of Public Health and Environment), 2012. Commission:
Water Quality Control, and Division. Colorado Nonpoint Source Program 2012
Management Plan, pp. 180
Cunningham, S., Berti, W., 1993. Remediation of contaminated soils with green plants: an
overview. Vitr. Cell. Dev. Biol. 29, 207212.
Das, P., Samantaray, S., Rout, G.R., 1997. Studies on cadmium toxicity in plants: a review.
Environ. Pollut. 98, 2936.
Dean, A.P., Lynch, S., Rowland, P., Toft, B.D., Pittman, J.K., White, K.N., 2013. Natural
wetlands are ecient at providing long-term metal remediation of freshwater
systems polluted by acid mine drainage. Environ. Sci. Technol. 47, 1202912036.
http://dx.doi.org/10.1021/es4025904.
Dos Santos Utmazian, M.N., Wieshammer, G., Vega, R., Wenzel, W.W., 2007. Hydroponic
screening for metal resistance and accumulation of cadmium and zinc in twenty
clones of willows and poplars. Environ. Pollut. 148, 155165. http://dx.doi.org/
10.1016/j.envpol.2006.10.045.
van der Ent, A., Baker, A.J.M., Reeves, R.D., Pollard, a.J., Schat, H., 2013.
Hyperaccumulators of metal and metalloid trace elements: facts and ction. Plant
Soil 362, 319334. http://dx.doi.org/10.1007/s11104-012-1287-3.
Evlard, A., Sergeant, K., Printz, B., Guignard, C., Renaut, J., Campanella, B., Paul, R.,
Hausman, J.F., 2014. A multiple-level study of metal tolerance in Salix fragilis and
Salix aurita clones. J. Proteom. 101, 113129. http://dx.doi.org/10.1016/
j.jprot.2014.02.007.
Feleafel, M.N., Mirdad, Z.M., 2013. Hazard and eects of pollution by lead on vegetable
crops. J. Agric. Environ. Ethics 26, 547567. http://dx.doi.org/10.1007/s10806-0129403-1.
Flora, G., Gupta, D., Tiwari, A., 2012. Toxicity of lead: a review with recent updates.
Interdiscip. Toxicol. 5, 4758. http://dx.doi.org/10.2478/v10102-012-0009-2.
Greger, M., Landberg, T., 1999. Use of willow in phytoextraction. Int. J. Phytoremediat. 1,
115123. http://dx.doi.org/10.1080/15226519908500010.
Huang, J.W., Cunningham, S.D., 1996. Lead phytoextraction: species variation in lead
uptake and translocation. New Phytol. 134, 7584. http://dx.doi.org/10.2307/
2558516.
Jarup, L., 2003. Hazards of heavy metal contamination. Br. Med. Bull. 68, 167182.
http://dx.doi.org/10.1093/bmb/ldg032.
Jung, M.C., 2001. Heavy metal contamination of soils and waters in and around the
Imcheon AuAg mine. Korea Appl. Geochem. 16, 13691375. http://dx.doi.org/
10.1016/S0883-2927(01)00040-3.
Kuzovkina, Y.A., Zhivotovsky, O.P., Schulthess, C.P., Pettinelli, D., 2010. Plant selection
for a pilot phytoremediation study at a former skeet range. Remediation 20, 93105.
http://dx.doi.org/10.1002/rem.20242.
Leao, E.M., Pang, K.-L., 2010. Eect of copper(II), lead(II), and zinc(II) on growth and

Declaration of author contributions

Gordon Kersten, gordonkersten@gmail.com responsible for integrity of work as a whole from inception to nished article: conception
and design of study, acquisition of data, analysis and interpretation of
data, and drafting and revising article for important intellectual
content.
Dr. Martin Quigley, maquigle@ucsc.edu responsible for integrity
of work as a whole from inception to nished article: conception and
design of study, acquisition of data, analysis and interpretation of data,
and revising article for important intellectual content.
Dr. Brian Majestic, brian.majestic@du.edu - design of study, preparation of inductively coupled plasma mass spectrometry data,
analysis and interpretation of data, and revising article for important
231

Ecotoxicology and Environmental Safety 137 (2017) 225232

G. Kersten et al.

Tlusto, P., Szkov, J., Vyslouilov, M., Pavlkov, D., Weger, J., Javorsk, H., 2007.
Variation in the uptake of arsenic, cadmium, lead, and zinc by dierent species of
willows Salix spp. grown in contaminated soils. Cent. Eur. J. Biol. 2, 254275. http://
dx.doi.org/10.2478/s11535-007-0012-3.
US EPA, 2010. Colorado Water Quality Assessment Report. Colorado Water Quality
Assessment Report. US Environmental Protection Agency.http://ofmpub.epa.gov/
waters10/attains_state.control?P_state=CO.
US EPA, 2014a. uperfund Sites in Region8. http://www2.epa.gov/region8/superfundsites-region-8.
US EPA 2014b. National Recommended Water Quality Criteria - Aquatic Life Criteria
Table https://www.epa.gov/wqc/national-recommended-water-quality-criteriaaquatic-life-criteria-table
US EPA, 2016. Table of Regulated Drinking Water Contaminants https://www.epa.gov/
ground-water-and-drinking-water/table-regulated-drinking-water-comtaminants
Vandecasteele, B., Quataert, P., De Vos, B., Tack, F.M.G., Muys, B., 2004. Foliar
concentrations of volunteer willows growing on polluted sediment-derived sites
versus sites with baseline contamination levels. J. Environ. Monit. 6, 313321.
http://dx.doi.org/10.1039/b314917j.
Vandecasteele, B., Quataert, P., Piesschaert, F., Lettens, S., De Vos, B., Du Laing, G., 2015.
Translocation of Cd and Mn from bark to leaves in willows on contaminated
sediments: delayed budburst is related to high Mn concentrations. Land 4, 255280.
http://dx.doi.org/10.3390/land4020255.
Vangronsveld, J., Herzig, R., Weyens, N., Boulet, J., Adriaensen, K., Ruttens, A., Thewys,
T., Vassilev, A., Meers, E., Nehnevajova, E., van der Lelie, D., Mench, M., 2009.
Phytoremediation of contaminated soils and groundwater: lessons from the eld.
Environ. Sci. Pollut. Res. Int. 16, 765794. http://dx.doi.org/10.1007/s11356-0090213-6.
Vyslouilov, M., Tlusto, P., Szkov, J., Pavlkov, D., 2003. As, Cd, Pb and Zn uptake
by Salix spp. clones grown in soils enriched by high loads of these elements. Plant,
Soil Environ. 49, 191196.
Wahsha, M., Bini, C., Argese, E., Minello, F., Fontana, S., Wahsheh, H., 2012. Heavy
metals accumulation in willows growing on Spolic Technosols from the abandoned
Imperina Valley mine in Italy. J. Geochem. Explor 123, 1924. http://dx.doi.org/
10.1016/j.gexplo.2012.07.004.
Watson, C., Pulford, I.D., Riddell-Black, D., 2003. Screening of willow species for
resistance to heavy metals: comparison of performance in a hydroponics system and
eld trials. Int. J. Phytoremediat. 5, 351365. http://dx.doi.org/10.1080/
16226510390268748.
Yang, W., Ding, Z., Zhao, F., Wang, Y., Zhang, X., Zhu, Z., Yang, X., 2015. Comparison of
manganese tolerance and accumulation among 24 Salix clones in a hydroponic
experiment: application for phytoremediation. J. Geochem. Explor. 149, 17. http://
dx.doi.org/10.1016/j.gexplo.2014.09.007.
Yao, Z., Zhou, L., Bao, Z., Gao, P., Sun, X., 2009. Erratum to: High eciency of heavy
metal removal in mine water by limestone. Chinese J. Geochemistry 28, 293298.
http://dx.doi.org/10.1007/s11631-010-0120-z.
Zalesny, R.S., Bauer, E.O., 2007. Evaluation of Populus and salix continuously irrigated
with landll Leachate I. Genotype-specic elemental phytoremediation. Int. J.
Phytoremediat. 9, 281306. http://dx.doi.org/10.1080/15226510701476461.
Zhivotovsky, O.P., Kuzovkina, J. a., Schulthess, C.P., Morris, T., Pettinelli, D., Ge, M.,
2011a. Hydroponic screening of willows (Salix L.) for lead tolerance and
accumulation. Int. J. Phytoremediat. 13, 7594. http://dx.doi.org/10.1080/
15226511003671361.
Zhivotovsky, O.P., Kuzovkina, Y. a., Schulthess, C.P., Morris, T., Pettinelli, D., 2011b.
Lead uptake and translocation by willows in pot and eld experiments. Int. J.
Phytoremediat. 13, 731749. http://dx.doi.org/10.1080/15226514.2010.525555.

sporulation of halophytophthora from Taiwan mangroves. Water Air Soil Pollut. 213,
8593. http://dx.doi.org/10.1007/s11270-010-0369-1.
Ling, T., Jun, R., Fangke, Y., 2011. Eect of cadmium supply levels to cadmium
accumulation by Salix. Int. J. Environ. Sci. 8, 493500.
Macnicol, R., Beckett, P., 1985. Critical tissue concentrations of potentially toxic
elements. Plant Soil 85, 107129.
Markham, K.R., 1971. A chemotaxonomic approach to the selection of opossum resistant
willows and polars for use in soil conservation. N. Z. J. Sci. 14, 179186.
Martens, S.N., Boyd, R.S., 1994. The ecological signicance of nickel hyperaccumulation:
a plant chemical defense. Oecologia 98, 379384. http://dx.doi.org/10.1007/
S00442-004-V.
Meiman, P.J., Davis, N.R., Brummer, J.E., Ippolito, J. a., 2012. Riparian shrub metal
concentrations and growth in amended uvial mine tailings. Water Air Soil Pollut.
223, 18151828. http://dx.doi.org/10.1007/s11270-011-0986-3.
Mendez, M.O., Maier, R.M., 2008. Phytostabilization of mine tailings in arid and semiarid
environments an emerging remediation technology. Environ. Health Perspect. 116,
278283. http://dx.doi.org/10.1289/ehp.l0608.
Mndez-Armenta, M., Ros, C., 2007. Cadmium neurotoxicity. Environ. Toxicol.
Pharmacol. 23, 350358. http://dx.doi.org/10.1016/j.etap.2006.11.009.
Mleczek, M., Rissmann, I., Rutkowski, P., Kaczmarek, Z., Golinski, P., 2009.
Accumulation of selected heavy metals by dierent genotypes of Salix. Environ. Exp.
Bot. 66, 289296. http://dx.doi.org/10.1016/j.envexpbot.2009.02.010.
Nagajyoti, P.C., Lee, K.D., Sreekanth, T.V.M., 2010. Heavy metals, occurrence and
toxicity for plants: a review. Environ. Chem. Lett. 8, 199216. http://dx.doi.org/
10.1007/s10311-010-0297.
Nazzaro, R.M., 2008. Hardrock mining: information on abandoned mines and value and
coverage of nancial assurances on BLM land
Pulford, I.D., Watson, C., McGregor, S.D., 2001. Uptake of chromium by trees: prospects
for phytoremediation. Environ. Geochem. Health 23, 307311. http://dx.doi.org/
10.1023/A:1012243129773.
Pulford, I., Riddell-Black, D., Stewart, C., 2002. Heavy metal uptake by willow clones
from sewage sludge-treated soil: the potential for phytoremediation. Int. J.
Phytoremediat. 4, 5972.
Pulford, I.D., Watson, C., 2003. Phytoremediation of heavy metal-contaminated land by
trees a review. Environ. Int. 29, 529540. http://dx.doi.org/10.1016/S01604120(02)00152-6.
Punshon, T., Dickinson, N., 1997. Acclimation of Salix to metal stress. New Phytol. 137,
303314.
Rosas, I., Carbajal, M.E., Gmez-Arroyo, S., Belmont, R., Villalobos-Pietrini, R., 1984.
Cytogenetic eects of cadmium accumulation on water hyacinth (Eichhornia
crassipes). Environ. Res. 33, 386395.
Salt, D.E., Blaylock, M., Kumar, N.P., Dushenkov, V., Ensley, B.D., Chet, I., Raskin, I.,
1995. Phytoremediation: a novel strategy for the removal of toxic metals from the
environment using plants. Biotechnology 13, 468474. http://dx.doi.org/10.1038/
nbt0595-468.
Salt, D.E., Smith, R.D., Raskin, I., 1998. Phytoremediation. Annu. Rev. Plant Biol.,
643668.
Shanahan, J.O., Brummer, J.E., Leininger, W.C., Paschke, M.W., 2007. Manganese and
Zinc Toxicity Thresholds for Mountain and Geyer Willow. Int. J. Phytoremediation 9,
437452. http://dx.doi.org/10.1080/15226510701606323.
Tahvanainen, J., Helle, E., Julkunen-Tiitto, R., Lavola, A., 1985. Oecologia as deterrents
against feeding by mountain hare. Oecologia 65, 319323.
Thompson, J., Bannigan, J., 2008. Cadmium: toxic eects on the reproductive system and
the embryo. Reprod. Toxicol. 25, 304315. http://dx.doi.org/10.1016/
j.reprotox.2008.02.001.

232