Transcript Poster Congreso de Recife

Analysis of Heavy Metal Concentrations in Soil Profiles:
Necessity of a Typology
Luis Rodríguez-Lado1 , Florence Carré & Luca Montanarella
Land Management and Natural Hazards Unit. Joint Research Centre. Ispra, 21020 (VA) Italy ([email protected]) Tel. +39 0332 789977; Fax. +39 0332 786394
RESULTS
FACTOR(3)
FACTOR(1)
NI
CR
NI
CU
ZN
PB
HG
NI
CR
CU
HG
CD PB
CU
CU
ZN
PB
NI
HG
HG
ZN
CU
ZN
CR
ZN
NI
ZN
HG CU
CR
NI
HG
HG
HG
ZN
FACTOR(1)
PB
CU
CUNI
CR
NI
CRPB
HG
HG
CD
NI
CU
CU
FACTOR(2)
PB
CR
CD
FACTOR(4)
ZN
CU
CD
PB
CR
NI
HG
CR
ZN
CD
CR
NI
PB
CD
ZN
CD
PB
HG
CD
PB
CD
NI
CR
CR
CU
ZN
CD
PB
CD
PB
CD
FACTOR(4)
FACTOR(3)
This paper presents a general method to link pollutants to soil types that can be helpful to perform a
quick analysis of the distribution of pollutants over soils. It can be used as a tool for decision makers
to make a faster delineation of problematic areas and to analyze the probable sources of pollution in
such areas.
PCA analyses reveals four groupings of heavy metals.
The four-component model accounts for 83% of the
data variation. The first factor well discriminates Ni,
and Cr (Figure 2). It can be considered that the origin
of these elements in soils is geogenic. The second
factor separates Pb and Cu. These elements are
usually related to human activities, so their
concentration in soils is mainly anthopogenic. In the
third axis is represented Zn, also controlled by
lithology. The fourth axis represents Hg. In this case,
the origin of this element in soils is also anthropic.
For Cd we found an ambiguous situation, it is
represented equally in both the 2nd and 3rd axes.
Seems that its presence in soils can be due to both
human and natural inputs.
FACTOR(2)
FACTOR(2)
Decisions on the remediation of polluted soils are one of the most difficult management issues for
environmental state agencies. The cost of the assessment of soil contamination status at regional or
national level is high and, in most of the cases, this assessment is uncertain. The economic
implications of ensuring soil quality are multiple, thus understanding the spatial distribution of
contaminants is a crucial point for policy making at the EU level.
FACTOR(1)
FACTOR(1)
This problem is also recognized in the recent EU “Thematic Strategy for Soil Protection” (COM, 2002),
where contamination is identified as one of the main threats for soils in the Europe. This Strategy
constitutes the basis for maintaining and improving soil resources quality along Europe. The working
group on “Contamination and Land Management” (Van Camp et al, 2004) states the needs for
measuring heavy metal concentrations in soils, determining the sources of pollution, establishing
background values and critical loads of pollutants for each soil type and determining the risk of
pollution as basis for the development of soil quality standards.
Heavy metal contents in this soil is heterogeneous. Since the soils include in this study are mainly derived
from lime rocks, we observed that most of the samples have HM concentrations lower than the thresholds
fixed in the European legislation for soil pH > 7. There are also many samples with high contents of heavy
metals. This occurs in some samples for Cd (Lazio, Molise), Hg (Lazio) Cr (Basilicata, Toscana, Lazio and
mainly Sardinia), Ni (Basilicata, Sardinia), Zn (Basilicata, Toscana, Lazio, Calabria and Sardinia). We must
note that these thresholds were defined for agricultural soils, so they are not really applicable to natural
soils as those presented in this study and they are merely presented just
as a reference
values.
Factor
Loadings
Plot
FACTOR(2)
Among the negative impacts related to human activities, the mobilization of heavy metals from their
naturals reservoirs to the aquatic and terrestrial ecosystems has become a generalized problem in
almost worldwide (Han et al., 2002; Koptsik et al., 2003; Salemaa et al., 2001). At the present, it is
considered that a great proportion of soils in developed countries present concentrations of some
elements and compounds higher than their expected natural concentration (Jones, 1991).
Nevertheless, in some areas, natural factors as parent material, climate, vegetation, volcanoes, etc.
are highly influencing heavy metal contents in soils (Nriagu, 1989; Nriagu and Pacyna, 1988).
FACTOR(3)
INTRODUCTION
FACTOR(4)
1
ZN
FACTOR(3)
FACTOR(4)
Figure 2.- Factor Loadings Plot
STUDY AREA
Basilicata
Permuted Data Matrix
The study was carried out in soils from Natura 2000 protected areas in the Italian Peninsula. We used
a database containing 218 soil profiles, with a total amount of 664 soil horizons described. Their
spatial distribution is shown in Figure 1.
N
#
#
#
#
#
#
##
##
###
###
##
##
##
##
##
##
#
##
##
##
#
#
#
Samples
Regions
##
##
## #
##
#
# ##
#
##
##
#
#
# ##
#
##
#
##
#
##
## ##
#
#
#
#
#
#
#
## #
#
#
Soil profile descriptions include geographic information
(location, geology, vegetation type, aspect, slope, altitude), and
pedological information as soil type, number and description of
the horizons, soil texture.
Total contents of heavy metals (Hg, Cd, Cr, Cu, Ni, Pb, Zn) were
determined by atomic absorption spectroscopy. Threshold
values for HM in agricultural soils coming from European
legislation and descriptive statistics for these samples are
reported in Table 1.
Cr Ni
CR NI
Hg Cd
Calabria
Zn Pb Cu
HG CD ZN
PB
Emilia Romagna
Permuted Data Matrix
Permuted Data Matrix
Cu U Pb B CdD Ni I Cr R Hg G ZnN
C
P
C
N
C
H
Z
CU
CALCARIC
Calcaric
FluvisolFLU
CHROMIC
PHAE
Chromic
Phaeozem
CuU NiI
C
Humic
Umbrisol
HUMIC
UMBRIS
Gleyic
Phaeozem
GLEYIC
PHAEO
Eutric
Cambisol
EUTRIC
CAMBI
DYSTRIC
LUVI
Dystric
Luvisol
Calcaric
Phaeozem
CALCARIC
PHA
Calcaric
Regosol
CALCARIC
REG
CALCARIC
PHA
Calcaric
Phaeozem
CrR HgG Zn CdD PbB
C
H
ZN C
P
HISTIC
PODZO
Hystic Podzol
Dystric
Fluvisol
DYSTRIC
FLUV
CHROMIC
LUVI
Chromic
Luvisol
DYSTRIC
LUVI
Dystric
Luvisol
N
SAPRIC
HISTO
Sapric
Histosol
DYSTRIC
FLUV
Dystric
Fluvisol
Calcaric
Gleysol
CALCARIC
GLE
Luvic
Phaeozem
LUVIC
PHAEOZ
Haplic
Phaeozem
HAPLIC
PHAEO
Calcaric
Cambisol
CALCARIC
CAM
MOLLIC
REGOS
Mollic Regosol
GYPSIC
CAMBI
Gypsic
Cambisol
DYSTRIC
CAMB
Dystric
Cambisol
GYPSIC
LEPTO
Gypsic
Leptosol
HAPLIC
UMBRI
Haplic
Umbrisol
CALCARIC
CAM
Calcaric
Cambisol
Humic
Umbrisol
HUMIC
UMBRIS
Vitric Andosol
VITRIC
ANDOS
2
1
0
-1
LazioPermuted Data Matrix
Dystric
Regosol
DYSTRIC
REGO
CALCISOL
Calcisol
EUTRIC
REGOS
Eutric Regosol
Marche
Permuted Data Matrix
Hg GPb B CuU NiI Cr R Cd D ZnN
H
P
C
N
C
C
Z
Molise
RENDZIC
LEPT
Rendzic
Leptosol
MOLLIC
ANDOS
Mollic Andosol
CALCARIC
LEP
Calcaric
Leptosol
MOLLIC
LEPTO
Mollic Leptosol
LEPTIC
ANDOS
Leptic Andosol
VITRIC
ANDOS
Vitric Andosol
SILTIC
PHAEO
Siltic
Phaeozem
CUTANIC
LUVI
Cutanic
Luvisol
LEPTIC
PHAEO
Leptic
Phaeozem
MOLLIC
LEPTO
Mollic Leptosol
Chromic
Luvisol
CHROMIC
LUVI
CHROMIC
CAMB
Chromic
Cambisol
SKELETIC
CAL
Skeletic
Calcisol
CALCARIC
CAM
Calcaric
Cambisol
Calcaric
Andosol
CALCARIC
AND
HUMIC
LEPTOS
Humic Leptosol
VITRIC
ANDOS
Vitric Andosol
CALCARIC
REG
Calcaric
Regosol
Rendzic
Leptosol
RENDZIC
LEPT
Luvic
Phaeozem
LUVIC
PHAEOZ
LEPTIC
LUVIS
Leptic Luvisol
Eutric Leptosol
EUTRIC
LEPTO
Eutricc
Phaeozem
EUTRIC
PHAEO
HAPLIC
PHAEO
Haplic
Phaeozem
Chromic
Phaeozem
CHROMIC
PHAE
Calcaric
Cambisol
CALCARIC
CAM
VERTIC
PHAEO
Vertic
Phaeozem
Calcisol
CALCISOL
DYSTRIC
CAMB
Dystric
Cambisol
Fluvic
Cambisol
FLUVIC
CAMBI
Haplic
Calcisol
HAPLIC
CALCI
EUTRIC
CAMBI
Eutric
Cambisol
15
10
5
0
-5
Dystric Regosol
DYSTRIC
REGO
Eutric
Cambisol
EUTRIC
CAMBI
Permuted Data Matrix
CdCD PbPB CuCU Zn
ZN Hg
HG NiNI Cr CR
PbPBZnZN Hg
HG Cr
CR NiNI CdCD CuCU
ACRISOL
Acrisol
4
3
2
1
0
-1
-2
CALCARIC
REG
Calcaric
Regosol
2
1
0
-1
-2
LEPTIC
UMBRI
Leptic
Umbrisol
3
2
1
0
-1
CALCARIC
PHA
Calcaric
Phaeozem
LEPTIC
CAMBI
Leptic
Cambisol
Petric Calcisol
PETRIC
CALCI
10
5
0
-5
Eutric Leptosol
EUTRIC
LEPTO
#
##
Puglia
##
#
##
#
##
#
##
###
#
#
##
#
#
#
## #
#
#
##
#
#
#
#
##
##
#
##
#
#
##
# ##
##
#
Element
#
#
Threshold Value
Soil pH < 7
Threshold Value
Soil pH > 7
Minimum
Maximum
Average
SD
200
Petrocalcic
PETROCALCIC
Cadmium
1
3
0.01
160.6
5.15
19.02
Copper
50
210
1.5
156.2
33.1
21.2
Nickel
30
112
3
774.6
46
46.1
40 0 Kilom e te r s
Typic
Argixeralf
T ARGIXERALF
Typic
Argixeroll
T ARGIXEROLL
Lithic
Argixe
LITHIC
ARGIX
Humic Haplox
HUMIC Lithic
HAPLOX
Lead
50
300
0.3
284.2
37.5
35
Zinc
150
450
5.4
3039.7
132.7
222
1
1,5
0.21
20150
178.09
1120.47
100
150
3.3
866.5
76.9
60.3
Mercury
Figure 1.- Location of soil profiles.
C
LithicLITHIC
Xerorthent
XEROR
#
#
0
CuU Pb
Cu
CU HgHG Zn
ZN Ni
NI CrCR PbPB CdCD
## ##
20 0
Sardinia
Permuted Data Matrix
Permuted Data Matrix
Chromium
Table 1.- Threshold values and descriptive statistics for HM.
Haplox
LITHIC HAPLO
3
2
1
0
-1
Mollic
Cambisol
MOLLIC
CAMBI
Leptic
Phaeozem
LEPTIC
PHAEO
Mollic
Leptosol
MOLLIC
LEPTO
Pachic
Umbrisol
PACHIC
UMBRI
LEPTIC
UMBRI
Leptic
Umbrisol
ALBEL
FragicFRAGIC
Albeluvisol
PACHIC
PHAEO
Pachic
Phaeozem
MOLLIC
REGOS
Mollic
Regosol
CHROMIC
LUVI
Chromic
Luvisol
LUVIC
CALCIS
Luvic
Calcisol
LUVIC
UMBRIS
Luvic
Umbrisol
UMBRIC
LEPTO
Umbric
Leptosol
SKELETIC
Skeletic
LeptosolLEP
EUTRIC
CAMBI
Eutric
Cambisol
HUMIC
LUVISO
Humic
Luvisol
CALCIC
LUVIS
Calcic
Luvisol
CALCARIC
CAM
Calcaric
Cambisol
CALCARIC
Calcaric
ArenosolARE
LEPTIC
CALCI
Leptic
Calcisol
EUTRIC
ARENO
Eutric
Arenosol
PB
Toscana
Permuted Data Matrix
ZnN CuU HgG CdD Pb B Cr R Ni I
Zn HgG Cd D Cr R Ni I
ZN H
C
C
N
Z
C
H
C
P
C
N
Haplic Cambisol
HAPLIC CAMBI
Dystric Regosol
DYSTRIC
REGO
Chromic
Luvisol
CHROMIC
LUVI
CHROMI
SKELE
Skeletic
Cambisol
2
1
0
-1
-2
HAPLIC LUVIS
Haplic Luvisol
CALCARIC CAM
Calcaric Cambisol
HAPLIC CALCI
Skeletic Calcisol
SKELETI-CALC
Dystric Cambisol
DYSTRIC CAMB
Leptic
Cambisol
LEPTIC
CAMBI
10
5
0
-5
Figure 3.- Cluster analyses for HM contents
Hierarchical cluster analyses were performed for
both heavy metals and soil types. These analyses
were performed by administrative regions.
Permuted data matrices on standardized data
(Figure 3) show both the cluster trees for elements
and soils and a colored matrix indicating the
standard deviations of HM content for each soil
type. Rows and columns are ordered according to
the overall similarity to help interpretation.
In general we observe the same pattern of
associations between HM as those obtained in the
PCA using the whole dataset. We observe that soils
in regions like Basilicata and Marche have a higher
content in Cu, probably due to vine cultivation. In
Lazio the most evident is the higher contents in Pb
derived from the emissions of the road transport.
In Molise, leptic soils trend to exhibit higher
contents of Cd, Pb, Cu and Zn. These analyses also
permit to identify special situations. In Sardinia, Cr
and Ni contents in Phaeozems and in Mollic
Cambisols/Leptosols can be related to the
presence of vitric materials. Minning activities were
reported in these areas. Vitric Andosols
in Profile Plots
Cluster
Basilicata present very high contents in Cu.0
1
METHODOLOGY
However the presence of noncristaline materials and the high contents of organic matter provides a high
capacity to retain HM so their bio availability is probably low. The higher Hg contents are located in soil
samples from Tuscany and North of Lazio. Industrial activities in these areas as well as pollution coming
from geothermal plant can be the origin of this pollution.
ZN
HG
CR
NI
CD
PB
We adopted a Three-Step strategy in order to make a risk assessment of pollution with HM in these
areas.
- Firstly we compared the HM concentration in these samples against threshold values for coming
from the European legislation. This allows to identify soils at risk according to such values.
- Although the geographical distribution of heavy metals in soils may be dependent on environmental
factors like geology, topography, etc, and thus may be linked to soil types, it may be also highly
related to climatic variables (precipitation, dry deposition rates, wind, etc), and land use. For this
reason it is necessary to determine the sources of heavy metals (geogenic and anthropogenic) on
soils and their partial contribution to the overall heavy metal concentrations.
In this sense Principal Component Analyses (PCA) were carried out to understand the association
between different heavy metals, to try to explain their distribution into the soil profile and to identify
the possible sources of contamination. PCA with Varimax Rotation were performed on standardized
data, and the analyses were done on the correlation matrix. The four main principal components were
retained based on their Eigenvalues. In bibliography, these analyses are the most used to distinguish
geogenic (concentrations that are inherent to soil types due to their pedogenetic origin) and
anthropogenic (mainly derived from atmospheric deposition or land management practices) sources
of HM.
- Finally we used both a Matrix Cluster Classification and a automated K-means algorithm in order to
introduce two new dimensions in the analysis: taxonomy and location of soil observations. All
observations are then classified according to each other on the basis of heavy metal concentrations
in each horizon within each soil type.
Bibliography:
Han, F.X., A. Banin, Y. Su, D.L. Monts, M.J. Plodinec, W.L. Kingery, and G.E. Triplett, 2002. Industrial age anthropogenic inputs of heavy metals into the pedosphere. Naturwissenschaften 89, 497-504.
Jones, K.C., 1991. Contaminant trends in soils and crops. Environ Pollut 69, 311-25.
Koptsik, S., G. Koptsik, S. Livantsova, L. Eruslankina, T. Zhmelkova, and Z. Vologdina, 2003. Heavy metals in soils near the nickel smelter: chemistry, spatial variation, and impacts on plant diversity. J
Environ Monit 5, 441-50.
Nriagu, J.O. 1989. A Global Assessment of Natural Sources of Atmospheric Trace Metals. Nature,v.338, p.47-49.
Nriagu, J.O. and Pacyna, J.M. 1988. Quantitative Assessment of Worldwide Contamination of Air, Water and Soils by Trace Metals. Nature, v. 333, p.134-139.
Salemaa, M., I. Vanha-Majamaa, and J. Derome, 2001. Understorey vegetation along a heavy-metal pollution gradient in SW Finland. Environ Pollut 112, 339-50.
Van-Camp. L., Bujarrabal, B., Gentile, A-R., Jones, R.J.A., Montanarella, L., Olazabal, C. and Selvaradjou, S-K. (2004). Reports of the Technical Working Groups Established under the Thematic Strategy
for Soil Protection. EUR 21319 EN/4, 872 pp. Office for Official Publications of the European Communities, Luxembourg.
Finally, cluster K-means classification reveals the three
main groups of soils according to their HM contents
(Figure 4). The first group includes 94% of the soils. It is
a highly homogeneous group, with all metal contents
distributed around the mean values and low dispersion
of the data that probably represents the characteristics
of the main natural soils in Italy. The second group
includes five soil profiles. It is characterized by higher
contents of Cd and Zn, probably due to specific natural
conditions, to anthropogenic inputs or to a mix of both .
The third group includes three cases with very high
contents of Hg, Cr and Ni. In these cases specific
studies on HM pollution are convenient to better
understand the real problem of contamination in these
areas.
CU
Cluster Profile Plots
1
3
2
ZN
ZN
ZN
HG
HG
HG
CR
CR
CR
NI
NI
NI
CD
CD
CD
PB
PB
PB
CU
CU
CU
3
Figure 4.- Cluster K-means groups
ZN
HG
CR
NI
CD
PB
CU
CONCLUSIONS
Soil vulnerability to heavy metals are influenced by the diversity, distribution and specific vulnerability of
soils across Europe. In this study we presented a method to perform a simple multi-evaluation on the
status of pollution with heavy metal in soils. In this case we used natural soils coming from Natura 2000
sites in Italy. This approach allows to identify areas at risk, determine the possible sources of pollution and
to find links between heavy metal contents. On the other hand, soil types were ranked and clustered
according to their heavy metal content. To find a typology of polluted soils would help decision makers to
protect specific areas minimizing costs of evaluation in order to protect natural ecosystems and human
health. Accurate results can be obtained by means an adequate soil sampling design covering the most
significant soil types and also taking into account their spatial distribution in the study area. These results
can be improved by adding information on land management practices, location of point sources of
pollution, evaluation of deposition rates, etc. We must note that toxicity risk for heavy metals is not
dependent only on the total metal content in soils but also in the speciation forms they are present and in
their mobility. For more detailed studies deepen surveys are needed.