International Journal of Geosciences, 2012, 3, 50-61
http://dx.doi.org/10.4236/ijg.2012.31007 Published Online February 2012 (http://www.SciRP.org/journal/ijg)
Evaluation of Quality of Some Rehabilitated Mined Soils
within the AngloGold-Ashanti Concession in Ghana
Witmann H. K. Dorgbetor1, Gabriel N. N. Dowuona1, Seth K. A. Danso1, Julius K. Amatekpor1,
Ayoade O. Ogunkunle2, Enoch Boateng3
1Department of Soil Scien ce, School of Agriculture, University of Ghana, Legon, Ghana
2Department of Agronomy, University of Ibadan, Ibadan, Nigeria
3Soil Research Institute, Accra Centre, Accra, Ghana
Received December 23, 2011; revised January 15, 2012; accepted January 31, 2012
ABSTRACT
Land degradation caused by surface mining of gold has been extensive in Ghana. In recent years rehabilitation of some
degraded lands b y re-vegetatio n ha s been und ertaken. Th is stu dy p rovid es q uantitativ e data o n the qu ality o f so me reha-
bilitated and un-rehab ilitated mined soils within the AngloGold-Ashanti gold concession in parts of the semi-deciduous
forest zone of Ghana. Soil properties determined included texture, bulk density and aggregate stability, pH, organic
carbon, available phosphorus, total nitrogen, cation exchange capacity, exchangeable bases, exchange acidity, Fe, Mn,
Ni, Cu, Zn, Cd, and Pb. Aggregate stability as a physical quality indicator revealed that aggregates of the rehabilitated
mined soil had become more stable and similar to the control unmined soil due to litter and carbon additions from
planted trees. The nutrient levels were very low because of the presence of low activity clays inherent in the native soil.
Organic carbon content in the rehabilitated soil had increased above that of the unrehabilitated soil. Variability in soil
properties, especially org anic carbon and aggr egate stability, was minimal in the unmined and rehabilitated so ils imply-
ing that soils at the two sites were most robust and resistant to crushing and rup ture. Quality index of the unmined con-
trol soil was 36.5% indicatin g that the quality of the so il was 63.5% relative to the optimum qu ality because of inh erent
poor soil properties. The mined rehabilitated and unrehabilitated soil had index values of 32.5% and 24.4%, respec-
tively. The marginal difference of 4 % in soil quality between the control and reh abilitated soil shows that it is possible
to maintain the health of soils with inherent physical and biochemical deficiencies if reclamation regulations are ad-
hered to. In this way, the socio-economic dilemma of exploiting natural resources for the benefit of societies is amelio-
rated while maintaining an eco system balance.
Keywords: Aggregate Stability; Mean Weight Diameter; Mined Soils; Soil Rehabilitation; Soil Variability;
Soil Quality Index
1. Introduction
Some countries in West Africa including Ghana are en-
dowed with enormous deposits of mineral resources. In
Ghana gold ranks as the most extensively mined chiefly
because it is highly priced and every effort is made to
increase output. The system of mining in the past, which
was dominated by either underground shafts or by small-
scale gold mining, had little impact on the environment.
However, gold mining today is mostly from the surface;
this involves stripp ing the vegetatio n and topso il, digging
up of huge open pits, blasting of gold-bearing ores and
dripping of cyanide through massive piles of the gold ore
to extract the precious mineral. The operation destroys
farmlands and endangers water resources of the mining
communities.
Complaints about mining-related land degradation and
agitations from local people are based on experiences of
health hazards from chemical spillages into local water
bodies and contamination or destruction of quality of the
soil environment. Consequently, rural farming communi-
ties in major mining areas have had confrontations with
mining companies over right of access to lands or for
destruction of the ecosystem. This concern is informed
by empirical evidence elsewhere [1-4].
Bioavailable concentrations of heavy metals may be-
come higher than the permitted critical levels [5-7] in
mine-degraded soils. With time, the concentrations ad-
versely affect soil-plant relations, water quality, buffer-
ing capacities, availability of nutrients and water to
plants and soil microbes, mobility of contaminants and
certain physical factors including crusting. Chemicals
such as cyanide that directly kill or impair soil microbes
also reduce soil quality and a decline in soil quality
C
opyright © 2012 SciRes. IJG

W. H. K. DORGBETOR ET AL. 51
means a decline in the functions of soil and making life
unsustaina bl e [ 8- 10].
Mining operations generally take up farmlands, dis-
lodge communities and cause socio-economic problems
including poverty and migration. Important positive cha-
nges in soil quality, however, occur if rehabilitation is
undertaken which enables farmers to re-use the land for
improvement in their well-being. It has been noted that
mining by its nature presents both positive and negative
impact on an area and the role of policy makers should
be to mitigate or prevent the negatives while promoting
the positives [11]. To achieve this goal, necessary pre-
requisites for post-closure sustainable development must
be put in place while policy makers, industry and other
stakeholders such as researchers monitor environmental
quality. Although rehabilitation of mined lands could
meet the major goal of sustainability, evaluation of the
extent of reclamation or rehabilitation remains the great-
est challenge. Furthermore, a method for measuring suc-
cessful reclamation has been difficult, elusive and sub-
jective [12].
To address this problem, various approaches have been
developed to evaluate soil quality because consensus has
not been reached on which is the best. Be that as it may,
soil quality can b e measured in many ways depending on
the criteria selected [13]. Ecological approach to sus-
tainability incorporates resilience and requires diversity
so that ecological restoration of mined land could repre-
sent the best approach to ensuring sustainabilit y and mai-
ntenance of biodiversity. A reclaimed land could meet
the major goal of sustainability, which is the land use
options, for future generations [14]. Therefore, for rec-
lamation to be ecologically sustainable, it should be as-
sessed according to ecological p rinciples such as stability
of soils and nutrient cycling, vegetation establishment
and animal recolonization [15].
Various soil quality index models [13,16,17] attem-
pted to integrate information from multiple indicators to
arrive at single values that indicate the level of soil qu al-
ity. However, soil indicators that are predominantly site
specific in predicting soil quality as a function of meas-
ured soil properties are of little use as a routine assess-
ment tool [18]. To quantitatively assess the potential im-
pacts of changes in soil p roperties on the health of forest
soils, it is better to develop a soil quality index that inte-
grates the measured physical and chemical parameters of
the soil into a single parameter that could be used as an
indicator of overall forest soil quality [18]. Nevertheless,
application of this model to assess the quality of degr-
aded soil after their rehabilitation is very rare, especially
in mine-degraded tropical forest soils.
The Environmental Impact Assessment Laws of Ghana
require all mining-affected areas to be returned to the
community in physically and bio-chemically safe and
stable condition. In accordance with these regulations,
the mining companies h a ve embarked on rehabilitation of
mined lands to restore biodiversity after exploiting the
mineral wealth. In spite of these regulations, quantitativ e
data on the success or impact of the restoration are very
limited. It is therefore necessary to provide relevant data
through research to ensure that remediated lands attain
high quality status. This study therefore evaluates quan-
titatively the quality of some rehabilitated mined soils
within the AngloGold-Ashanti concession in the semi-
deciduous zone of Ghana using a quality index [18].
2. Materials and Methods
2.1. Site Characteristics, Soils and Sampling
The study area is located within the land concession of
Anglogold Ashanti mines, a major gold mining conces-
sion at Obuasi in the semi-deciduous forest zone of
Ghana (Figure 1). The climate with a bi-modally distrib-
uted annual rainfall of 1600 mm and a mean annual tem-
perature of 28˚C is characterized by distinct seasons and
is controlled primarily by the tropical continental, as well
as the tropical maritime air mass. The study area is un-
derlain predominantly by phyllites th at belong to ro cks of
the Paleoproterozoic Birimian of Ghana. The rocks pro-
duce similar soils at summits to upper slopes with eleva-
tion of about 250 m (above sea level). The natural vege-
tation of the region is the semi-deciduous forest ecology
of Ghana characterized by the Celtic-Triplochiton Asso-
ciation of plant species [19].
The dominant soil (Nzima series) at the study area oc-
curs at upper slope position of the landscape and is clas-
sified as Plinthic Acrisol or Typic Plinthustult. Quartz
gravel and other rock fragments, as well as iron and
manganese nodules, occur in most part of the soil profile.
The drainage is moderately good and the groundwater
level is considered very deep (lower than 150 cm).
Five land use systems namely, one unmined site and
four mined sites, were selected for the study. Preliminary
investigations indicated that all the sites were degraded
forest with similar landuse practice of shifting cultivation.
The original forest was therefore replaced with the pre-
sent vegetation which comprised a mosaic of fallow
farmlands, thickets, secondary forests and forb regrowth.
The umnined site was used as a control whereas the
mined sites consisted of a 7-year old rehabilitated site
with replanted vegetation made up of a mixture of origi-
nal plants species and exotic leguminous trees such as
Acacia and Leucaena (MR); a 4-year old rehabilitated
site under cultivation (WRF); a site covered with subsoil
material and topsoil but not replanted (MunReh); and a
site covered with a subsoil material with no topsoil
(Wdump).
Th e s tu dy site on each landuse system was loc at ed a ft er

W. H. K. DORGBETOR ET AL. 52
Figure 1. Sketch map of parts of the AngloGold-Ashanti c onc ession showing the location of the study site s.

W. H. K. DORGBETOR ET AL. 53
series of test augering. Grid sampling was undertaken at
each site. Each grid system covered an area of 10 m × 10
m (100 m2). Samples were collected from 20 cm depth at
2 m intervals along the midpoin t section of each grid line
to assess variability in soil properties at each site. After
the grid sampling, a modal profile was prepared at the
centre of the control (unmined) site, characterized and
soil samples collected. For the four mined sites, one pro-
file pit (120 cm deep) was sited at the center of each grid
system and samples collected to assess possible pe-
dological variations after the rehabilitation, especially
with respect to structural stability and particle size dis-
tribution. The disturbed samples were used for analyses
of selected physical and chemical properties whereas the
undisturbed samples were used to determine bulk density
and aggregate stability.
2.2. Laboratory Investigations
Soil colour (moist) was determined using the Munsel
Colour Chart. Particle size distribution of the soil was
determined using the modified Bouyoucos Hydrometer
Method [20]. The core method [21] was used to deter-
mine bulk density. Aggreg ate stability of the undisturbed
soils was measured by the dry sieving method [22]. A
100 g soil aggregate was placed on a nest of sieves with
apertures of 2.0 mm - 1.0 mm, 1.0 mm - 0.5 mm, 0.5 mm
- 0.25 mm, 0.25 mm - 0.106 mm and <0.106 mm in di-
ameters. The nests of sieves were subjected to continuous
shaking along 30 cm amplitude for 5 min using the elec-
tronic shaker; the shaking was repeated also for 40 cm
and 50 cm amplitudes for 10 min each. The dry-stable
aggregate in each nest of sieves was determined as:
 
Ma Mi*10
1
n *Xi*Wi
Total SQIS individual soil property index v a lue
Dry-stable aggregate %  (1)
where Ma is mass of the resistant aggregate and Mi is the
initial dry weight of the aggregate before shaking. The
mean weight diameter (MWD) of the structure aggregate
was calculated using the following equation [23].
MWD  (2)
where Xi is the mean diameter of aggregates separated
by sieving in the individual nest of sieves and Wi is the
weight of the aggregate in a particular size range as a
fraction of the initial dry weight of the aggregate ana-
lyzed.
Chemical properties analysed included soil pH, total
nitrogen, available phosphorus [20], exchangeable bases
and exchange acidity [25]. Effective CEC (ECEC) was
determined from the sum of cations. Exchangeable Na
percent (ESP) was calculated as a proportion of the ele-
ment of ECEC. For the measurement of soil organic car-
bon, the dry combustion meth od involving the use of the
Carbon Analyzer was employed. The concentrations of
Fe, Mn, Ni, Cu, Zn, Cd, and Pb in the soils were deter-
mined on the AAS following extraction with 1 M NH4Cl.
2.3. Calculation of Soil Quality Index
The soil quality model of Amacher et al. [18] was used
to calculate the soil quality inde x of each landuse system.
Analytical data generated from this study were compared
to defined threshold levels; index values were assigned to
each property to express soil adverse effects that are pos-
sible or unlikely. The individual index values for all the
properties measured for each landuse system were sum-
med up to a total soil quality index (SQI) as:
(3)

The maximum value of the total SQI is 26 if all 19 so il
properties are measured. The total SQI is then calculated
as:




SQI %Total SQImax. possible total
SQI for properties measured100

 (4)
2.4. Variability in Soil Properties
Coefficient of variation (CV) was used to estimate the
extent of variability in soil properties within each land
use system. This was calculated as:


(5) CV%s z*100
where s is the standard deviation and z is the mean of the
population sample (36 samples for each study site).
3. Results and Discussion
3.1. Pedological Characteristics
Examination of the modal profile of the undisturbed na-
tive soil (Control) showed that it is deep and moderately
well drained with moist soil colours that vary from dark
brown (7.5YR 3/3) and brown (7.5YR 4/6) in the solum
to yellowish red (5YR 4/6) in the parent material with a
corresponding texture that changes from sandy clay loam
at the surface to clay in the subsoil. The soil also has a
structure which is crumbly in the topsoil but grades into
subangular blocky throug hout the subsoil. Its consistenc e
is moderately or slightly hard. Abundant fine and me-
dium roots are found from the topsoil to about 50 cm
deep. Large amounts of quartz gravel and rock fragments
with iron and manganese concretions occur throughout
the profile. All the sites are underlain by weathered phyl-
lite (parent material) with similar characteristics prior to
removal of the overburden layer for mining.
The texture of the surface soils (0 cm - 20 cm) varies
from sandy clay loam for the Control site, clay loam for
the mined rehabilitated soil (MR), waste dump rehabili-
tated soil (WRF) and mined unrehabilitated soil (Mun-
Reh) to sandy clay loam for the waste dump unrehabili-

W. H. K. DORGBETOR ET AL. 54
tated soil (Wdump). The proportions of the rock frag-
ments and gravels in the mined soil pits are less than
those found in the Control soil profile. The four mined
soils at each respective site showed uniformity in colour
variation namely, reddish brown to reddish yellow (MR),
yellow red (WRF), yellow (Wdump) and reddish yellow
to pink (MunReh). It is apparent that mixing of materials
during the various rehabilitation processes influenced
soil texture, amount of coarse fragments and soil colour,
which were distinctly different from the unmined control
soil.
The colour and textural variations are characteristics
worthy of pedological note. Soil colour of the unmined
site (control) can serve as the standard for which all oth-
ers were measured for their closeness to, or deviations
from it. As indicated, the Control soil showed two prin-
cipal colours, brown and shades of it in the upper section
of the pit and a reddish colour at the lower portions. The
rehabilitated soil (MR site) shows two broad colours
namely, shades of brown at the upper portion and shades
of yellow occupying the lower portion of the profile.
Clearly, the impact of organic matter accumulation on
the soil surface must account for the similarity in colours
of the two soils at the soil surface. Subsoil colour of the
at the Control site showed many years of development
whereas the mined rehabilitated soil (MR) exhibited
seven years of organic matter modification of a degraded
layer of surface soil. Distribution of clay in the soils
shows the presence of an argillic horizon in the control
soil (Figure 2). However, this pedological feature was
not displayed by soil from the four previously mined sites
due to disturbance caused by removal of the soil material.
Kaolinite is the dominant clay mineral in this soil [26].
Figure 2. Distribution of clay content in the soils.
3.2. Variations in Soil Physical Properties
Physical properties such as topsoil depth, bulk density,
porosity, aggregate stability, water content, soil strength,
crushing and compaction of soil and water infiltration
rate, which may change due to management practices,
serve as indicators of soil quality. For this study, bulk
density and aggregate stability which influence the other
physi ca l pr operties was used to assess quality o f t h e s o i ls.
The mean bulk density values are presented in Table 1.
Bulk density values for the different sites were 1.73
Mg/m3 (Control), 1.63 Mg/m3 (MR), 1.69 Mg/m3 (WRF),
1.71 Mg/m3 (Wdump) and 1.61 Mg/m3 for MunReh. The
relatively high bulk density values are consistent with the
compact nature of the soil due the abundant rock frag-
ments and concretions, especially in the control soil and
indicate that the soils may pose problems for root estab-
lishment. Generally, roots grow well in soils with bulk
densities of up to 1.4 Mg/m3; root penetration begins to
decline significantly at bulk densities above 1.7 Mg/m3
[27,28].
Aggregate stability values at selected amplitudes of
vibration are presented in Figure 3. At amplitude of 30
cm there was almost no disruption of soil structural sta-
bility which reflected in all the aggregates staying on the
top sieve and almost none collecting in the pan at the
base. At amplitude 40 cm, differences began to show and
some of the aggregates showed crushing weakness but
these differences were statistically insignificant for all
the soils. All aggregates reduced in size because they
became less stable. While mean size of aggregates from
the rehabilitated site MR fell from 9 mm to 8 mm, those
of the Control site and the WRF site fell from 8.6 mm
and 8.7 mm to 6.8 mm and 6.3 mm, respectively. Ag-
gregates from the Wdump and the MunReh sites reduced
in size from 8.5 mm and 7.8 mm to 4.5 mm and 3.2 mm,
respectively. At amplitude 50 cm marked differences
were observed which suggested different capabilities of
the soils to withstand disruptive forces. The difference
between the mined rehabilitated soil (MR) and the unre-
habilitated soils was statistically significant.
The percentage changes in aggregate sizes (at ampli-
tudes 30 cm, 40 cm and 50 cm) of soils from the four
other sites compared to the Control site showed unequal
amounts of aggregates retained in the larger sieves. At
amplitude 30 cm, percentage changes in aggregate sizes
relative to the Control were not significantly differen t but
aggregates of the rehab ilitated sites (MR and WRF) were
3.0% and 2.0% greater, respectively. Aggregates from
the Wdump and the MunReh sites were 2.1% and 10.2%
less, respectively, relative to the Control site. At ampli-
tude 40, differences in aggregates relative to the Control
were greater. The MR site had an increase in stability
17.0% greater than the Control site while stability at the
WRF site reduced to 92.3% relative to the Control. The

W. H. K. DORGBETOR ET AL.
IJG
55
Table 1. Data on selected physical and chemical properties of the soils.
Study sites
Soil property† Control (C) MR (A) WRF (B) Wdump (D) MunReh (E)
Bulk density (Mg·m–3) 1.73 ± 0.05 1.63 ± 0.06 1.69 ± 0.04 1.71± 0.07 1.61 ± 0.08
Coarse fragments (%) 75.2 ± 1.2 40.8 ± 1.6 45.2 ± 1.4 60.6 ± 1.8 65.4 ± 1.5
Soil pH 5.4 ± 0.2 5.1 ± 0.2 4.8 ± 0.2 5.1± 0.2 4.4 ± 0.1
Organic carbon (g/kg) 13.3 ± 1.2 11.1 ± 0.1 7.2 ± 0.9 2.3 ± 0.7 4.8 ± 0.5
Total N (g/kg) 1.10 ±0.09 0.84 ± 0.05 0.56 ± 0.02 0.18± 0. 03 0.37 ± 0.02
Exch. Ca (cmol/kg) 1.99 ± 0.08 0.81 ± 0.05 0.67 ± 0.03 1.33 ± 0.07 1.29 ± 0.07
Exch. Mg (cmol/kg) 1.65 ± 0.07 0.66 ± 0.05 0.57 ± 0.03 1.68 ± 0.17 0.73 ± 0.04
Exch. K (cmol/kg) 0.16 ± 0.02 0.20 ± 0.01 0.07 ± 0.01 0.08 ± 0.01 0.10 ± 0.01
Exch. Na (cmol/kg) 0.88 ± 0.04 0.14 ± 0.01 0.17 ± 0.01 0.24 ±0.01 0.16 ± 0.01
ESP (%) 13.8 ± 0.04 3.5 ± 0.01 4.3 ± 0.01 4.6 ± 0.01 3.7 ± 0.01
Exch. acidity (cmol/kg) 1.68 ± 0.08 2.16 ± 0.08 2.51 ± 0.08 1.92 ± 0.10 2.04 ± 0.12
Available P (mg/kg) 8.53 ± 0.38 8.08 ± 0.83 6.60 ± 0.31 4.33 ± 0.13 7.89 ± 0.82
Fe (mg/kg) 0.027 ± 0.001 0.011 ± 0.001 0.009 ± 0.001 0.030 ± 0.005 0.033 ± 0.004
Mn (mg/kg) 9.168 ± 1.351 3.017 ± 0.641 2.017 ± 0.445 0.269 ± 0.085 0.272 ± 0.090
Ni (mg/kg) 0.353 ± 0.081 0.250 ± 0.061 0.212 ± 0.041 0.016 ± 0.006 0.018 ± 0.002
Cu (mg/kg) 0. 008 ± 0.003 0.018 ± 0.001 0.016 ± 0.002 0.00 0.00
Zn (mg/kg) 0.029 ± 0.010 0.019 ± 0.001 0.015 ± 0.003 0.029 ± 0.006 0.032 ± 0.004
Cd (mg/kg) 0.015 ± 0.002 0.011 ± 0.003 0.010 ± 0.004 0.020 ± 0.002 0.019 ± 0.003
Pb (mg/kg) 0.079 ± 0. 020 0.052 ± 0.004 0.042 ± 0.008 0.048 ± 0.003 0.048 ± 0.003
†Data represent means of 36 samples.
(c)
Figure 3. Stability of soil aggregates (mean values) at dif-
ferent amplitudes of vibration.
(a)
change in percentage aggregate sizes in comparison with
the Control was greater for the Wdump site (65.3%) and
the MunReh site (47.5%). There was a further increase in
percentage change of aggregate stability of the soils at
amplitude 50 cm. The MR site had a change of 41% rela-
tive to the control site. Aggregates from the WRF,
Wdump and MunReh sites, were 6.3%, 40.1%, and 27.5%
less, respectively. Clearly, aggregates from the waste
dump and mined unrehabilitated sites (MunReh) showed
the least strength at withstanding disruptive forces.
The implication of the observed trends in aggregate
stability and bulk density was that the vegetation of the
rehabilitated site had a positive impact on stabilizing
aggregates within a short term of reclamation, especially
(b)

W. H. K. DORGBETOR ET AL. 56
in soils which are inherently less fertile and unprodu ctive.
It is apparent that aggregate stability can increase or de-
crease depending on other prevailing factors. A high
value can occur through increased forest canopy cover-
age, addition of manure and the presence of other bio-
logical binding agents. Stability of soil aggregates can
decrease through removal of soil cover, burning, crush-
ing, wind action, erosion by water and deforestation as
well as top soil loss as in mining. The binding agents of
aggregates are rooting systems, decaying litter falls, soil
organisms and their decomposit i on products [29-31].
3.3. Variations in Soil Chemical Properties
The pH of the soils which ranged from 5.3 at the control
site to less than 5.0 at the mined sites (Table 1) can be
described as strongly to very strongly acid and reflect the
characteristic of a weathered soil. Total nitrogen and
available phosphor us contents are very low. The low ph-
osphorus content is partly due to the very small amounts
of apatite originally present in the phyllites and the long
period of leaching, and also partly due to the element
being locked in insoluble or less available compounds of
aluminium or iron as noted in similar soils elsewhere
[26,31]. Exchangeable bases concentration is very low
with Ca and Mg as the dominant cations at exchange site.
A fairly high exchange acid ity v alu e was recorded for th e
soils; this accounts for more than 30% of effective cation
exchange capacity. This is consistent with the low pH
environment and suggests an aluminium toxicity problem
is inherent in this soil. It is ap parent that the original soil
(Control site) has been leached of much of its plant nu-
trients and is naturally infertile. The concentrations of
trace elements and heavy metals were very low suggest-
ing that mining operations and subsequent rehabilitation
did not introduce any heavy metals into the soils.
Variations in carbon content of the soils are presented
in Figure 4. Organic carbon contents were 13.3 g/kg for
the unmined control site, 11.1 g/kg for the MR site, 7.2
g/kg for the WRF site, 2.3 g/kg for the Wdump site and
4.8 g/kg at the MunReh site. Clearly, the Control soil
which was not affected by mining recorded the highest
carbon values whereas the soil at the MR site, which has
been re-vegetated for a relatively longer period, ranked
second. The carbon build-up can be attributed to accu-
mulation of leaf litter from the fast growing trees and
other plant materials as well as other associated biologi-
cal activity which contributed to a restoration of lost soil
carbon. The carbon contents of soils at the Wdump and
MunReh sites reflect the impact of mining on degrada-
tion with attendan t loss of soil carbon. At the time of the
field study, re-vegetation work had not commenced at
these two sites.
The carbon content at the MR site (11.1 g/kg) relative
to the WRF site (7.2 g/kg) is worthy of special note. It
Figure 4. Distribution of organic carbon content at the study
sites.
should be noted that desired results of accumulating car-
bon will be achieved only when rehabilitation trees are
kept continuously growing. Although rehabilitation works
were implemented on both soils at the same time, the
trees on soil MR were maintained while those on soil
WRF were cut after about 3 - 4 years and the soil culti-
vated to food crops using traditional methods with no
external nutrient input. In this respect, the gap in carbon
content between MR and WRF could be attributed to
nutrient mining by cultivated crops. It is, therefore, nec-
essary that for the desired effect of rehabilitation to be
achieved, clear guidelines must be established on the
minimum lifespan of rehabilitation trees that must pass
before some trees are cut to allow for farming.
3.4. Coefficient of Variation and Soil Quality
Coefficient of variation values, calculated using data on
bulk density, aggregate stability and organic matter, are
presented in Table 2. The spread of sample values aro-
und the mean value of a set of data is an impo rtant meas-
ure of variability in sample populations. Coefficient of
variation (CV) is a normalized measure of spreading
about the mean. The CV increases as the population
variability also increases. Soil properties with larger co-
efficient of variation are more variable than those with
smaller values.
A classification scheme identifies the extent of vari-
ability for soil properties based on their coefficient of
variation [32]. Consequently, CV values of <0% - 15%,
16% - 35% and >36% indicate little, moderate, and high
variability, or Classes I, II and III soils, respectively. On
the basis of this classification scheme [32], the variability
in bulk density values for the soils can be described as

W. H. K. DORGBETOR ET AL. 57
Table 2. Coefficient of variation in some soil properties at the study sites.
Coefficient of variation and groups at the sites†
Soil property Control MR WRF Wdump MunReh
Bulk density:
CV (%) 7.91 12.07 7.18 13.22 16.01
Group (I) (I) (I) (I) II)
Aggregate stability:
A = 30 cm
CV (%) 4.49 0.58 1.06 4.39 15.68
Group (I) (I) (I) (I) (II)
A = 40 cm
CV (%) 10.05 3.11 6.63 24.66 17.14
Group (I) (I) (I) (II) (II)
A = 50 cm
CV (%) 13.05 12.03 22.09 21.62 39.06
Group (I) (I) (II) (II) (III)
Organic carbon
CV (%) 27.15 29.61 42.76 95.13 29.07
Group (II) (II) (III) (III) (II)
†(Number of sa mples (n) = 36; at 0 - 20 cm depth.
very marginal except for the MunReh site which showed
moderate variability. Variability in organic carbon was
moderate for the Control, MR and MunReh sites but high
for at the WRF and Wdump site. Considering aggregate
stability, variability was minimal for all the soils except
MunReh at vibrating amplitude 30 cm. However, when
the amplitude was raised to 40 cm, variability also in-
creased to moderate for aggregates in the Wdump and
MunReh soils. Remarkably, aggregates at the Control
and MR sites were very stable with little variability re-
gardless of the amplitude of the vibration force. The un-
rehabilitated mined soil (MunReh) showed the greatest
variability in aggregate stability due to absence of vege-
tation. Large variability in soil properties was noted for
the degraded forest soils elsewhere in the humid tropics
[33].
The role of organic carbon in stabilizing soil aggre-
gates is worthy of note. The differences in organic car-
bon contents at the Control and MR sites on one hand,
and those of the Wdump and MunReh sites are signifi-
cant. The implication is that aggregates of the mined re-
habilitated (MR) and Control sites are most robust and
resistant to crushing and rupture. Consequently, the in-
fluence of organic carbon on the stability of soil aggre-
gates may provide evidence on the physico-chemical
quality of soils. Correlation coefficient values showed a
positive dependence of aggregate stability on organic
carbon contents likely from addition of plant residues to
the rehabilitated soil. As noted from the earlier part of the
discussion, at amplitude 30 cm, all the soils withstood the
simulated crushing effect and there was no difference in
strength or stability of aggregates. The correlation coef-
ficient value (r30 = 0.57) shows that at amplitude 30 cm,
the influence of organic carbon was not too important. At
amplitudes 40 cm and 50 cm, however, the soils de-
pended more on their carbon content to resist rupture and
the crushing effect and this was reflected in the correla-
tion coefficient values of, r40 = 0.81 and r50 = 0.79, re-
spectively.
Although organic carbon influenced stabilization of
the aggregates, it is apparent that the greater MWD value
and the lower carbon content at the MR site may suggest
that other factors introduced by rehabilitation also had
effect on aggreg ate stability. The rehabilitat ed but farmed
soil (WRF) had aggregates less stable than the MR and
Control soils because organic carbon addition to the soil
was reduced when the replanted trees were cut to make
way for farming at the WRF site. Aggregates at the
Wdump and MunReh sites ranked the least stable be-
cause they had the least carbon content in them and were
also not rehabilitated or had no trees planted on them.
The implication of this observation is that vegetation
decreases variability in soil prop erties. For this study, the
issues raised by environmentalist and other stakeholders
in respect of effects of surface mining on destruction of
the ecosystem have to be reappraised. One can argue that
it is possible to maintain the health of a soil, especially
soils with inherent physical and biochemical deficiencies,
if reclamation regulations are adhered to. In this regard,
the socio-economic dilemma of exploiting natural re-
sources for the benefit of societies while maintaining
ecosystem balance is addressed.
3.5. Soil Quality Index
Soil quality index values, associated threshold va lues and
calculated percent total soil quality index from all the
study sites are presented in Table 3. The calculated index
value for bulk density is zero for all the soils. Bulk den-
sity values above 1.50 Mg/m3 indicate possible adverse
effects of soil impedance whereas values below 1.50
Mg/m3 suggest a minimal adverse effect to impedance

W. H. K. DORGBETOR ET AL. 58
Table 3. Soil quality index values of the soils.
Index values
No. Soil property Control (C) MR (A) WRF (B) Wdump (D) MunReh (E)
1 Bulk density 0 0 0 0 0
2 Coarse fragments 0 +1 +1 0 0
3 Soil pH +1 +1 +1 +1 +1
4 Organic carbon +1 +1 0 0 0
5 Total N +1 0 0 0 0
6 Exch. Ca 0 –1 –1 0 0
7 Exch. Mg 0 0 0 0 0
8 Exch. K 0 0 0 0 0
9 Exch. Acidity +1 +1 +1 +1 +1
10 ESP +1 +1 +1 +1 +1
11 Available P 0 0 0 0 0
12 Fe 0 0 0 0 0
13 Mn +1 +1 +1 0 0
14 Ni +1 +1 +1 +1 +1
15 Cu 0 0 0 0 0
16 Zn 0 0 0 0 0
17 Cd +1 +1 +1 +1 +1
18 Pb +1 +1 +1 +1 +1
Total 9 8 7 6 6
Soil Quality Index (%)† 36.5 32.5 28.4 24.4 24.4
† = calculated from Equation (4).
[18]. On this basis, the quality index value for all the
soils is consistent with the range in bulk density values
(Table 1). It can be said that generally all the soils can
pose problems for root penetration irrespective of the
state of disturbance from mining effects. Coarse frag-
ment proportion at the MR and MRF sites had a value of
+1 indicating that adverse effects to root penetration are
less likely in these soils.
Soil pH values (5.4, 5.1, 4.8, 5.1, and 4.5 for the Con-
trol, MR, WRF, Wdump and MunReh, respectively) fall
within the high acid range and showed no significant
differences and accordingly were an assigned an index
value of +1. The implication is that all the soils could
pose detrimental effect to the growth of a wide range of
crops except for acid tolerant species. Considering car-
bon, the Control and MR soils were awarded an index
value of +1 because their respective carbon contents of
13.3 g/kg and 11.1g/kg were moderate or adequate for
soil productivity. On the other hand, the recorded carbon
levels of 7.2 g/kg, 2.3 g/kg and 4.8 g/kg for the WRF,
Wdump and MunReh sites, respectively, were low and
were awarded an index of 0 indicating possible loss of
organic carbon, which can attributed to disturbance from
mining. This observation is consistent with studies else-
where in the tropics [34,35] where removal of topsoil as
a result of mining and other land management practices
caused decreases in soil organic matter.
When one compares the Control and MR soils, it is
possible to appreciate the value of keeping soil vegeta-
tive cover as well as promoting rehabilitation of de-
graded soils through afforestation. Of all the soil proper-
ties, the most important change or determinant was soil
organic carbon. Soil organic carbon is a key indicator for
soil quality and biological activity which impacts on the
chemical and physical behaviour of soils [36].
The Control soil was assigned an index of +1 because
it contained moderate levels of total nitrogen whereas
soils from the other four sites with relatively low levels
of nitrogen were assigned an index of zero. Generally,
exchangeable Ca in the control, Wdump and MunReh
soils were low for which a zero index value was assigned.
An index value of –1 was given to the MR and WRF
soils because of the very low Ca in them; this indicates
severe depletion of the element. Exchangeable Mg and K
levels were very low and thus attracted an index of zero
which suggests possible deficiencies of the elements in
all the soils. For all the soils, calculated exchangeable
sodium percentage was below 15% showing that adverse
effect due to sodicity is unlikely, which justified an ind ex
value of +1. The general commonality in index values
support the trend observed for other soil properties and
the generally low exchange capacity of the soils. It is
obvious that removal of the soil material and subsequent
refilling of the excavated pits and rehabilitation have not
had any significant effect on exchange capacity of the
soils.
The range in values of exchange acidity (1.68 - 2.51
cmol/kg) for the soils can be described as moderate using
the classification range of Amacher et al. [14]. A quality
index value of +1 was therefore assigned to all the soils
which suggest that only plants sensitive to Al are likely
to be affected when cultivated in these soils. The level of
available phosphorus in all the soils was low with the
likelihood of deficiencies hence all were awarded an in-

W. H. K. DORGBETOR ET AL. 59
dex value of zero.
For the trace elements, the associated index is zero for
the measured threshold levels of Fe, Cu and Zn in the
soils. An index value of +1 was assigned to all the soils
for Cd and Pb because of their relative concentration
levels (Table 1). The levels of Mn in the Control, MR
and WRF soils were low and fitted into an index value of
+1 whereas the very low levels which indicated defi-
ciency of the element in the Wdump and MunReh soils
accounted for an index value of zero. Although the levels
of Ni were moderate in the Control, MR and WRF soils
and low in the Wdump and MunReh soils they had an
index value +1. An index value of +1 was assigned to the
Control and the two rehabilitated soils b ecaus e of the low
Mn levels and zero for the two unrehabilitated soils as a
result of the very low levels of the element.
Summation of the assigned index values of all the
properties at each site showed that the Control, MR,
WRF, Wdump and MunReh soils total values were 9, 8,
7, 6 and 6, respectively. The corresponding calculated
soil quality index (Equation (4)) expressed in percentages
(% SQI) were 36.5, 32.5, 28.4, 24.4 , and 24.4 (Figure 5)
on the assumption of a proportionate maximum possible
total soil quality index of 24.6 instead of 26 (i.e. 18 out
of 19 measured soil properties). The scale of calculated
SQI indicates that the higher the value the better the
quality of the soil.
In this study, the SQI value of the Control (unmined)
soil suggests a poor quality soils because of the inherent
low fertility status and morphological features generally
associated with upland soils. The poor quality of the un-
mined soil is worthy of special note considering the fact
that it is under a secondary forest with a history of pre-
vious cultivation. Although we did not encounter the
original virgin forest vegetation with its uncu ltiv ated so il,
it is doubtful if the 36.5% quality index could be any dif-
ferent. Field and analytical data on a sim il ar soi l in a virgin
forest elsewhere in Ghana showed similar trend in meas-
ured soil properties except for a relatively higher organic
carbon and nitrogen contents [37]. Notwithstanding
Figure 5. Soil quality index at the study sites.
the shifts in carbon and nitrogen contents, however, the
associated index values based on the range of defined
thresholds [18] would not be different. It can therefore be
concluded that the poor quality of the Control soil is an
inherent characteristic. The difference in quality between
the maximum obtainable and the soil’s inherent quality
was further widened when it was subjected to the impact
of surface mining and this was reflected in lower index
values for all the mine-affected soils.
The mined rehabilitated soil MR h ad an index v alue of
32.5% which is less than the value of the Control by 4%
while the index of the rehabilitated soil cultivated soil
(WRF) was less by 8.1%. Index values for the waste
dump (Wdump) and mined unrehabilitated (MunReh)
soils were 12.1% less than the value for the control soil.
It is apparent that the most important contributing factor
for the relatively lower index values in the mined soils
was the reduction in total organic carbon and total nitro-
gen. Because the deeply weathered deposits are devoid of
weathered mineral residues for fertility, their agronomic
value depends on topsoil organic matter as nutrient
source for plant roots and also on the parent material for
fertility but this fertility is quickly lost when the forest
cover, which produced and protected it, is removed as
happens during surface mining of gold.
The SQI values reported for temperate forest soils [18]
were all greater than 40% because of higher nutrient lev-
els. Be that as it may, this study has shown that the
model can be applied to estimate the quality of tropical
forest soils. For our study, a major implication of the
quality index values of the soils, especially at the MR
and WRF sites, is that after exploiting the underlying
mineral wealth and rehabilitating the land (according the
EPA standard) it is possible to return the land to an equi-
librium state of the natural soil within the short and me-
dium terms, especially when the initial inherent quality is
low.
4. Conclusions
In all the soils studied, variability in organic carbon was
moderate for the Control and MR soils. There was little
variability in aggregate stability except for MunReh at
vibrating amplitude 30 cm. Significantly, aggregates at
the control and MR sites were very stable with little
variability regardless of the amplitude of the vibration
force. At higher amplitude of vibration, only soils at the
Control and MR sites were stable. Variability in soil
properties was very minimal in the Control soil fo llowed
by the MR sites. The Control unmined soil had 36.5%
soil quality index indicating a soil with a poor inherent
physical and biochemical properties. For the mined sites
the quality was less due to reductions in total organic
carbon and breakdown in aggregate stability. In order to
sustain soil productivity and prevent retrogression to the

W. H. K. DORGBETOR ET AL. 60
state of degradation, rehabilitated soils should be pro-
tected from farming activities for 12 - 15 years to allow
organic matter build-up to return to the conditions of the
Control soil. During crop cultivation, leguminous trees
should be incorporated and bush burning that tradition-
ally accompanies land preparation should be discouraged
to allow organic matter build-u p to mature.
A major outcome of this study relates to the issues
raised by environmentalists and other stakeholders on
impact of surface mining on destruction of the ecosystem.
It may be argued that the health of a soil can be main-
tained, especially soils with inherent physical and bio-
chemical deficiencies, if reclamation regulations are ad-
hered to. The period for the rehabilitated soil to reach the
stage of ecological balance may be in the short to me-
dium terms. In this regards, it is possib le to exploit natu-
ral resources for the benefit of societies while conscious
efforts are made to restore the health of the environment.
5. Acknowledgements
The authors wish to acknowledge the kind permission of
the Environmental Division of AngloGold Ashanti for
access to their mined sites and the Department of Soil
Science and the Ecological Laboratory, University of
Ghana, Legon for use of their laboratories for the analy-
ses. Special thanks go to Prince Gyekye of the Soil Re-
search Institute, CSIR, Accra for assistance in producing
the site map.
REFERENCES
[1] R. J. Van Aarde, A. M. Smit and A. S. Claassens, “Soil
Characteristics of Rehabilitating Mined and Unmined
Coastal Dunes at Richards Bay, KwaZulu-Natal, South
Africa,” Restoration Ecology, Vol. 6, No. 1, 1998, pp.
102-110. doi:10.1046/j.1526-100x.1998.06113.x
[2] E. Callender, “Heavy Metals in the Environment—His-
torical Trends,” In: B. S. Lollar, Ed., Treatise on Geo-
chemistry, Environmental Geochemistry, Elsevier, Am-
sterdam, 2004, pp. 67-105.
[3] A. Hüttermann, I. Arduini and D. L. Godbold, “Metal
Pollution and Forest Decline,” In: M. N. V. Prasad, Ed.,
Heavy Metal Stress in Plants from Biomolecules to Eco-
systems, 2nd Edition, Springer-Verlag, Berlin, 2004, pp.
295-312.
[4] K. Telmera, M. Costa, R. Angélica, E. S. Araujof and Y.
Mauriceg, “The Source and Fate of Sediment and Mer-
cury in the Tapajós River Para, Brazilian Amazon: Ground-
and Space-based Evidence,” Journal of Environmental
Management, Vol. 81, No. 2, 2006, pp. 101-113.
doi:10.1016/j.jenvman.2005.09.027
[5] Australian Environment Protection Agency, “Rehabilita-
tion and Revegetation, Best Practice Environmental Man-
agement in Mining,” ACT7 Commonwealth of Australia,
Barton, 1995.
[6] W. H. O. Ernst, “Bioavailability of Heavy Metals and
Decontamination of Soils by Plants,” Applied Geochem-
istry, Vol. 11, No. 1-2, 1996, pp. 163-167.
doi:10.1016/0883-2927(95)00040-2
[7] D. C. Adriano, “Trace Elements in Terrestrial Environ-
ments: Biogeochemistry, Bioavailability and Risk of Met-
als,” Springer-Verlag, New York, 2001.
[8] J. F. Parr, R. I. Papendic k, S. B. Hornick and R. E. Meyer,
“Soil Quality: Attributes and Relationship to Alternative
and Sustainable Agriculture,” American Journal of Al-
ternative Agriculture, Vol. 7, No. 1-2, 1992, pp. 5-11.
doi:10.1017/S0889189300004367
[9] US National Research Council, “Safety Evaluation Re-
port to Amendment No. 127 to Facility Operating Licence
No. DRP-48,” Zion Nuclear Power Station, Unit 2, Wash-
ington DC, 1992.
[10] C. A. Seybold, M. J. Mausbach, D. L. Karlen and H. H.
Rogers, “Quantification of Soil Quality,” In: R. Lal, J. M.
Kimble, R. F. Follett and B. A. Stewart, Eds., Soil Proc-
esses and the Carbon Cycle, CRC Press, Boca Raton,
1998, pp. 387-404.
[11] G. Miller, “The Whitehorse Initiative: A Case Study in
Partnerships, in Management of Commodity Resources in
the Context of Sustainable Development: Social Impact
of Mining,” Proceedings of United Nations Conference
on Trade and Development (UNTAD/ITCD/COMS), Ge-
neva, 1996, pp. 72-81.
[12] A. Clewell and J. P. Rieger, “What Practitioners Need
from Restoration Ecologist,” Restoration Ecology, Vol. 5,
No. 4, 1997, pp. 350-354.
doi:10.1046/j.1526-100X.1997.00548.x
[13] L. D. Karlen and D. E. Stott, “A Framework for Evaluat-
ing Physical and Chemical Indicators of Soil Quality,” In:
J. W. Doran, D. C. Coleman, D. F. Bezdicek and B. A.
Stewart, Eds., Defining Soil Quality for a Sustainable En-
vironment, Special Publication No. 35, Soil Science Soci-
ety of America, Madison, 1994, pp. 53-72.
[14] M. J. Haigh, “Degradation of Reclaimed Lands Previ-
ously Disturbed by Coal Mining in Wales: Causes and
Remedies,” Land Degradation & Development, Vol. 3,
No. 3, 1992, pp. 169-180. doi:10.1002/ldr.3400030306
[15] J. C. Chambers and G. L. Wade, “Evaluating Reclamation
Success: The Ecological Considerations,” Proceedings of
a Symposium, USDA Forest Service, Northern Forest
Experiment Station, Radnor, 1992, p.107.
[16] J. W. Doran and T. B. Parkin, “Quantitative Indicators of
Soil Quality: A Minimum Da ta Set,” In: J. W. Doran and
A. J. Jones, Eds., Methods for Assessing Soil Quality,
Special Publication No. 49, Soil Science Society of
America, Madison, 1996, pp. 25-37.
[17] J. Glover, J. Reganold and P. Andrews, “Systematic
Methods for Rating Soil Quality of Conventional, Or-
ganic and Integrated Apple Orchards in Wa shington State,”
Agriculture, Ecosystems & Environment, Vol. 88, No. 1-2,
2000, pp. 29-45. doi:10.1016/S0167-8809(00)00131-6
[18] M. C. Amacher, K. P. O’Neil and C. H. Perry, “Soil Vital
Signs: A New Soil Quality Index (SQI) for Assessing
Forest Soil Health,” Research Paper RMRS-RP-65WWW,
Forest Service, Rocky Mountain Research Station, Fort
Collins, 2007.

W. H. K. DORGBETOR ET AL. 61
[19] C. J. Taylor, “The Vegetation Zones of the Gold Coast,”
Department Bulletin No. 4, Government Printer, Accra,
1952.
[20] P. R. Day, “Particle and Particle Size Analysis,” In: C. A.
Black, Ed., Methods of Soil Analysis, Part I, Agronomy,
American Society of Agronomy, Madison, 1965, pp. 545-
567.
[21] G. R. Blake and K. H. Hartge, “Bulk Density,” In: A.
Klute, Ed., Methods of Soil Analysis, Part 1—Physical
and Mineralogical Methods, 2nd Edition, ASA-SSA,
Madison, 1986, pp. 363-382.
[22] W. D. Kemper and R. C. Rosenau, “Aggregate Stability
and Size Distribution,” In: A. Klute, Ed., Methods of Soil
Analysis, Part 1—Physical and Mineralogical Methods,
2nd Edition, ASA-SSA, Madison, 1986, pp. 425-442.
[23] D. Hillel, “Introduction to Environmental Physics,” El-
sevier Academic Press, Amsterdam, 2004.
[24] R. H. Bray and L. T. Kurtz, “Determination of Total Or-
ganic and Available Forms of Phosphorus in Soils,” Soil
Science, Vol. 59, No. 1, 1945, pp. 39-45.
doi:10.1097/00010694-194501000-00006
[25] G. W. Thomas, “Extractable Cations,” In: A. L. Page, R.
H. Miller and D. Keeney, Eds., Methods of Soil Analysis,
Part 2—Chemical and Biochemical Methods, 2nd Edition,
ASA-SSA, Madison, 1982, pp. 159-165.
[26] G. N. N. Dowuona, I. A. Taina and R. J. Heck, “Porosity
Analysis of Two Acrisols by X-Ray Computed Microto-
mography, Soil Science, Vol. 174, No. 11, 2009, pp. 583-
893. doi:10.1097/SS.0b013e3181c2a90b
[27] R. F. Sutton, “Soil Properties and Root Development in
Forest Trees: A Review,” Forestry Canada Information
Report O-X-413, Ottawa, 1991.
[28] R. F. Fisher and D. Binkley, “Ecology and Management
of Forest Soils,” 3rd Edition, John Wiley & Sons, New
York, 2000.
[29] R. F. Harris, G. Chester and O. N. Allen, “Dynamics of
Soil Aggregation,” Advances in Agronomy, Vol. 18, 1965,
pp. 107-160. doi:10.1016/S0065-2113(08)60649-5
[30] P. Lavelle, “Earthworm Activities and the Soil System,”
Biology and Fertility of Soils, Vol. 6, No. 3, 1988, pp.
237-251. doi:10.1007/BF00260820
[31] G. N. N. Dowuona, A. R. Mermut, S. G. K. Adiku, E.
Nartey and I. Tete-Mensah, “Improvement in the Quality
of Soils under Agroforestry Practice in Ghana,” In: G.
Renard, A. Neef, K. Becker and M. von Oppen, Eds., Soil
Fertility Management in West Africa Land Use Systems,
Margraf Verlag, Weikersheim, 1998, pp. 251-258.
[32] L. P. Wilding and L. R. Drees, “Spatial Variability and
Pedology,” In: L. P. Wilding, N. E. Smeck and G. F. Hall,
Eds., Pedogenesis and Soil Taxonomy I: Concepts and
Interactions, Elsevier, Amsterdam, 1985, pp. 83-116.
[33] A. O. Ogunkunle and O. O. Eghaghara, “Influence of
Land Use on Soil Properties in a Forest Region of South-
ern Nigeria,” Soil Use Management, Vol. 8, No. 3, 1992,
pp. 121-125. doi:10.1111/j.1475-2743.1992.tb00906.x
[34] G. S. Henderson, “Soil Organic Matter: A Link between
Forest Management and Productivity,” In: W. H. McFee
and J. M. Kelly, Eds., Carbon Forms and Functions in
Forest Soils, Soil Science Society of America, Madison,
1995, pp. 419-435.
[35] W. H. Schlesinger, “Biogeochemistry: An Analysis of
Global Change,” 2nd Edition, Academic Press, San Diego,
1997.
[36] M. Roberts, The Ground: Interface in the Environment,
Resource for Development,” Dunod/Masson, Paris, 1996.
[37] P. M. Ahn, “Soils of the Lower Tano Basin, Southwestern
Ghana,” Memoir No. 1, Department of Soil and Landuse
Survey, Kumasi, 1961.