Content uploaded by Elisa Grieco
Author content
All content in this area was uploaded by Elisa Grieco on Jul 30, 2024
Content may be subject to copyright.
Journal of Environmental Management 367 (2024) 121993
0301-4797/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
Research article
Impacts of deforestation and land use/land cover change on carbon stock
dynamics in Jomoro District, Ghana
Elisa Grieco
a
,
*
, Elia Vangi
a
, Tommaso Chiti
b
,
c
, Alessio Collalti
a
,
c
a
Forest Modeling Lab., Institute for Agriculture and Forestry Systems in the Mediterranean, National Research Council of Italy (CNR-ISAFOM), 06128 Perugia, Italy
b
Dipartimento per la Innovazione nei Sistemi Biologici, Agroalimentari e Forestali (DIBAF), Universit`
a della Tuscia, 01100 Viterbo, Italy
c
National Biodiversity Future Center (NBFC), 90133 Palermo, Italy
ARTICLE INFO
Keywords:
Deforestation
Land use change
Land cover change
Carbon stocks
Sustainable development
ABSTRACT
Tropical deforestation in the African continent plays a key role in the global carbon cycle and bears signicant
implications in terms of climate change and sustainable development. Especially in Sub-Saharan Africa, where
more than two-thirds of the population rely on forest and woodland resources for their livelihoods, deforestation
and land use changes for crop production lead to a substantial loss of ecosystem-level carbon stock. Unfortu-
nately, the impacts of deforestation and land use change can be more critical than in any other region, but these
are poorly quantied. We analyse changes in the main carbon pools (above- and below-ground, soil and litter,
respectively) after deforestation and land use/land cover change, for the Jomoro District (Ghana), by assessing
the initial reference level of carbon stock for primary forest and the subsequent stock changes and dynamics as a
consequence of conversion to the secondary forest and to ve different tree plantations (rubber, coconut, cocoa,
oil palm, and mixed plantations) on a total of 72 plots. Results indicate overall a statistically signicant carbon
loss across all the land uses/covers and for all the carbon pools compared to the primary forest with the total
carbon stock loss ranging between 35% and 85% but with no statistically signicant differences observed in the
comparison between primary forest and mixed plantations and secondary forest. Results also suggest that above-
ground carbon and soil organic carbon are the primary pools contributing to the total carbon stocks but with
opposite trends of carbon loss and accumulation. Strategies for sustainable development, policies to reduce
emissions from deforestation and forest degradation, carbon stock enhancement (REDD+), and planning for
sustainable land use management should carefully consider the type of conversion and carbon stock dynamics
behind land use change for a win-win strategy while preserving carbon stocks potential in tropical ecosystems.
1. Introduction
The increasing decline of African forests is the result of land use
change related mostly to anthropogenic activities, affecting not only
biodiversity, carbon (C) and water cycle but also undermining the
mitigation potential of tropical ecosystems. A relevant aspect of climate
change mitigation strategies is the understanding of the dynamics of
land use changes following deforestation (Masolele et al., 2024). Despite
the widely recognised importance of the Sub-Saharan Africa region to
the global C-budget, the knowledge gap at C-pool level still needs to be
lled (Houghton and Hackler, 2006;Olorunfemi et al., 2022). Land use
change (LUC) activities related to deforestation play a major role in
determining sources and sinks of carbon (Le Qu´
er´
e et al., 2009) being
responsible for emissions of about 1.9 PgC yr
−1
over the period
2013–2022 (Friedlingstein et al., 2023). Africa experienced the largest
annual rate with 3.9 million hectares of forests lost between 2010 and
2020 with an increasing trend since 1990 and long-term management
plans existing for less than 25% of forests, one of the lowest levels after
South America (20%)(FAO, 2020). Agricultural expansion has a key role
in the African deforestation process, leading to 64% of total forest loss,
in favour of small-scale cropland, from 2001 to 2020. Moreover, 7% of
forest conversion in Africa has been attributed to commodity crops such
as rubber, oil palm and cocoa, common in West Africa, notably in Ghana
(Masolele et al., 2024). An IUCN analysis (IUCN, 2008) highlighted the
critical situation of Ghana’s Western Region where most of the patches
of forestland outside forest reserves, that existed in 1990, had been
converted to other land uses by 2007. Findings from Ampim et al. (2021)
also conrm a loss in forest cover in the Western Region (~704.7 km
2
)
* Corresponding author.
E-mail address: elisa.grieco@cnr.it (E. Grieco).
Contents lists available at ScienceDirect
Journal of Environmental Management
journal homepage: www.elsevier.com/locate/jenvman
https://doi.org/10.1016/j.jenvman.2024.121993
Received 10 May 2024; Received in revised form 23 July 2024; Accepted 24 July 2024
Journal of Environmental Management 367 (2024) 121993
2
between 1995 and 2019 despite a general increase at the national level
(+23.3%). National policies, aiming at modernising and expanding tree
crop plantation areas while supporting smallholder farmers, are also
inuenced by global demand in the supply chain and the need to boost
employment and local food security (Ampim ibid.). Tree plantations
could apparently seem a perfect win-win solution for enhancing local
community livelihood while promoting climate change mitigation and
economic development, probably even more than pure afforestation and
reforestation practices (Kongsager et al., 2013). Nevertheless, it is
questionable whether this choice can be sustainable in terms of
long-term C-stock balance at the ecosystem level. As also underlined by
Pendrill et al. (2022), land-use change research in Africa and ‘commo-
dity-specic land use dynamics data’are poorly known and efforts
should be focused on better characterising deforestation in smallholder
shifting agriculture, being one of the main drivers of deforestation. At
the same time, the expansion of commodity crops like cocoa, rubber and
oil palm in large-scale cropland in humid and dry forests in western and
south-eastern Africa, indicates a likely vulnerability to future land use
change and a real obstacle to zero deforestation supply chain (Masolele
et al., 2024). Information on C-sequestration potentialities and C-stock
dynamics of tree crop monoculture systems in developing countries –
and particularly in Africa –are scarce and incomplete (Kongsager ibid.).
It is undeniable that ecosystems’resilience, forest resource protection,
and mitigation potentialities are interlinked with socioeconomic im-
provements but more research is needed to properly assess the actual
impact of forest reduction for tree crop expansion on tropical ecosys-
tems. This study aims to untangle some unclear assumptions on the
supposed benecial effects of tree crop plantations in terms of C-stock
potentials. Despite the extensive literature on land use change in tropical
areas, this study applies stringent eligibility criteria in site selection to
assess carbon pools across various land use and cover types at different
stages post-deforestation. This approach provides a better understand-
ing of how different land uses, covers, and plantation ages inuence
carbon sequestration and storage over time. More specically, the pre-
sent work aims to address the following questions by analysing data and
describing and discussing results from the Jomoro District in Ghana
(Africa): 1) What are the effects of land use/land cover change in trop-
ical forested land, on different carbon pools compared to the primary
forest? 2) What are the changes in relation to the establishment of
different tree plantations? 3) What is the range, rate and magnitude of
these variations and the dynamics in terms of C-stocks across different
C-pools and does it depend on the considered pool, the age or type of
conversion?
2. Material and methods
2.1. Study area
The Jomoro District occupies the Southwestern corner of the Western
Region of Ghana and covers an area of 1.344 km
2
, about 5.6% of the
total area of the Western Region (JDA, 2009,2014). The district has
extensive rainforest and the south-central part includes the Ankasa
Forest Reserve. The potential vegetation is represented by high-forests,
not uniform throughout but divided into several belts which differ in
their oristic composition, general characters and distribution mainly
related to rainfall and soil acidity. The high forest is characterised by
species referred to as indicator trees, such as Cynometra ananta Hutch. &
Dalziel, Lophira alata Banks ex Gaertn. and Tarrietia utilis Sprague also
known as the Cynometra-Lophira-Tarrietia association (Ahn, 1961). The
district lies on four main geological formations: Granites, Lower Birri-
mian, the Tertiary Sands and the Coastal Sands (Hall and Swaine, 1976;
Schlüter, 2005;Grieco, 2011;Chiti et al., 2014). The soils developed
from the weathering of granite (Atsivor et al., 2001) can be divided into
two main groups in the Jomoro District: Ochrosols and Oxysols (Chiti
et al., 2014; USDA, 2010). The two groups differ mainly in the topsoil
with the Ochrosols showing a pH of 5.5 and the Oxisols of less than 5
(Ahn, 1961). The landscape is characterised by slopes varying from 1 to
30% (Wauters et al., 2008). The climate is classied as Equatorial
Monsoon (Kottek et al., 2006). The rainfall regime in the region is
bimodal with four seasons; two rainy seasons: a major in April–July and
a minor in October–November, and two dry seasons: a major in
December–March and a minor in August–September. Monthly rainfall
varies between 0 and 500 mm, while annual rainfall ranges from 1200 to
1800 mm. The average relative air humidity is between 95 and 100%.
The average annual temperature ranges between 24 and 27 ◦C. Absolute
extreme temperatures are 15 and 40 ◦C. The main drivers of deforesta-
tion and forest degradation in the high forest zone of Ghana are strictly
braided into multiple factors which encompass agricultural expansion,
mining, wood/timber extraction and infrastructure extension (Hansen
et al., 2009;Ampim et al., 2021). Among the dominant drivers of forest
loss for agricultural expansion, is the increased expansion of cocoa, oil
palm and rubber with a forest conversion of 25.4, 1.9 and 1.5%,
respectively, with 25.9% of forest converted into small-scale cropland
between 2001 and 2020 (Masolele et al., 2024).
2.2. Site’s selection, eld measurement and sample collection
2.2.1. Sites’selection
The sites’selection aimed at locating the main land uses after the
original rainforest clearance throughout the Jomoro District and to this
end, selection was based both on main land uses driving deforestation in
the area and on results from interviews with local farmers supported by
on eld surveys. It was considered ‘eligible’, the site which has been
deforested to be used for only a single plantation type ever since. Only
areas with a single, consistent land use/cover post-deforestation were
included in the study to ensures that observed changes in carbon pools
can be condently attributed to specic land use/cover types. The
eligibility criteria have been applied to assess the changes in Total
Carbon Stocks (TCS) as a direct effect of the forest clearing on the main
C-pools: above-ground biomass, below-ground biomass, soil, litter, and
standing dead wood. The sampling eldwork was carried out on each
site, in three selected plots per site for soil, litter and above-ground then
followed by laboratory and data analysis. Extensive surveys, performed
in 2009 and 2010 within the Jomoro District, have allowed the identi-
cation, in addition to forest and secondary forest sites, of ve main
commodities plantations for a total of 23 sites (i.e. 3 for forest, 2 for
secondary forest, 2 for oil palm, 2 for cocoa, 4 for rubber, 2 for mixed
and 8 for coconut plantation) consisting in a total number of 72 different
plots (Fig. 1,Table S1 in Supplementary Material).
2.2.1.1. Primary forests. Two primary forest sites were identied. The
former was a patch of land close to Cocotown village, preserved as a
high-forest due to its sacred value for local people. The utilization of
wood and collections of other woody products was strictly forbidden.
The total area of this forested land was 3 ha. The second site was a
portion of the Ankasa conservation area, which covers approximately a
total surface of 50,900 ha. The sites were identied with the codes
‘ANKF’and ‘CTF’, respectively. A third patch of primary forest was
selected near Bawia village. This parcel of land was designated as
community property, reserved for future farming activities. It covered 4
ha.
2.2.1.2. Secondary forests. The secondary forest sites were two, ‘SF
10
’
and ‘SF
20
’. The rst was a land of 2.4 ha, deforested in 2000 and then
abandoned. Similarly, on the second site covering 1.6 ha, the pristine
forest was cleared in late December 1988, so we considered deforested
from 21 years (SF
21
). After deforestation, the land was cultivated for one
year with cassava and coconut but farming activities did not succeed
because of pests, so the land was then abandoned. The age of secondary
forest on this land was thus 20 years.
E. Grieco et al.
Journal of Environmental Management 367 (2024) 121993
3
2.2.1.3. Oil palm plantations. Two oil palm plantation sites were iden-
tied, having been established 8 and 25 years before the survey. These
plantations were established immediately after clearing the primary
forest originally covering those areas. The plantation ‘OP
8
’was a land of
1.22 ha established in 2002 while ‘OP
25
’covered 2.4 ha and was
established in 1985, one year after forest clearance. At the time of the
survey, the plantation OP
25
was in its second generation. The replace-
ment was done in 2005, so that, at the time of the survey, the new
plantation was only 4-year-old.
2.2.1.4. Cocoa plantations. Two cocoa plantations were identied, the
rst at Cocotown village, along the river Tano. This area has been his-
torically famous for cocoa production since the 18th century. The cho-
sen site, ‘CC
120
’, was established in 1890 and continuously cultivated
with cocoa. The assessed plantation of 6 ha was reported never to have
been replaced because the trees were left for natural regeneration
through coppicing. The second site, ‘CC
34
’, was selected in a village
located south of Cocotown. The plantation, which covered 1.4 ha, was
established in 1975 following the clearance of the primary forest and has
been consistently cultivated cocoa ever since.
2.2.1.5. Coconut plantations. Height sites for coconut plantations were
identied. The rst, ‘CN
21
’, was a land of 3.6 ha located near the western
entrance of the Ankasa conservation area and was established in 1988.
The second site, ‘CN
28
’, was situated at Nyamenle Keya Beven near
Sowodzadzem village. It was established in 1982 and covered 30 ha. The
third site ‘CN
44
’was a plantation of 12.14 ha located between Navrongo
crossroad and Abudu village and was established in 1966. The fourth,
‘CN
15
’, covering 2.4 ha, was located close to Betekomoa village and was
established in 1995. The fth, ‘CN
50
’, was a land of 8.1 ha near Aboyele
village and was established in 1960. The sixth one, ‘CN
100
’, was close
Nuba village and was established in 1910 and, at the time of the survey,
the 2 ha plantation was 100-year-old. The seventh site, ‘CN
53
’, was
deforested in 1956 and was a plantation of 4 ha. The eighth, ‘CN
95
’, was
a land covering 12.1 ha deforested in 1915.
2.2.1.6. Rubber plantations. Two sites deforested at different times were
found also for rubber. The rst site of 1.6 ha,‘RP
5
’, was located near
Johnatan village and was established in 2005. The second rubber
plantation,‘RP
10
’, covered 5.3 ha, it was located near New Ankasa
village and established in 2000. At both sites were present the rst
generation of trees. The rubber plantations found on this substrate were
two, the rst, ‘RP
14
’, located near Bawia village was established in 1996
and covered 2.4 ha, while the second one, ‘RP
50
’, located in Mpataba
village was established in 1960 and covered 3 ha.
2.2.1.7. Mixed plantations. The investigated sites cultivated as mixed
plantations in the Jomoro district were two. The tree component has
been considered for measurements. The rst plantation, ‘MP
36
’, was
established in 1974, near Agaege village and covered 5.3 ha. Trees
proportion within the plantation was oil palm (40%) and coconut (60%).
The latter site, ‘MP
50
’, was a land of 10.2 ha, established in 1960, in the
proximity of one of the Ankasa Conservation area. This mixed plantation
was characterised by a composition of oil palm (38%), coconut (19%)
and other species (13%).
2.2.2. Field measurements and sample collection
2.2.2.1. Above- and below-ground biomass. To assess the above-ground
biomass (AGB), we conducted morphometric measurements of the
standing vegetation starting by measuring tree height (H, in meters) and
diameter at breast height (DBH, in cm), according to FAO protocol by
Ponce-Hernandez et al., 2004. The DBH measurement threshold was 5
cm according to VCS (VCS, 2010). The sample plots’dimensions were
10 ×10 m, coinciding with the soil sample plots. For species with a
wider tree planting spacing, the plot dimensions were doubled (e.g.
coconut) in order to ensure a sufcient number of trees within the plot,
providing a more reliable and accurate estimates of tree density per
hectare. Wider spacing can result in fewer trees per unit area, leading to
higher variability and potential sampling errors if the plot size is too
small. By increasing the plot area, we include more trees within the
sample, which helps to reduce the inuence of random spatial variations
and it improves the statistical reliability of the density estimates. This
approach aligns with ecological sampling principles that advocate for
adjusting plot size to accommodate the spatial characteristics of the
species, ensuring that the sample is representative and the estimates are
Fig. 1. Map of the survey area and the selected sites (red dots) in the Jomoro District, Western Region, Ghana.
E. Grieco et al.
Journal of Environmental Management 367 (2024) 121993
4
robust. The tree height was measured with a Suunto clinometer, the DBH
with a dendrometer calliper, and the sampling area bordered with a
measuring tape. Each plot within the selected site was georeferenced
through a GPS device (Garmin/GPS60). Dead-standing trees were also
considered in measurements for tree height and DBH and included in the
Deadwood Biomass pool (DB). Non-destructive method was adopted for
AGB and DB estimation, through the use of allometric equations
(Table 1). We prioritized and adopted biomass models yielded results
within the expected and scientically acceptable range. The
below-ground biomass (BGB) was not collected on the eld directly but
estimated as a percentage of the above-ground (AGB) biomass using the
root-to-shoot ratio as in Cairns et al. (1997),Houghton et al. (2001),
Achard et al. (2002),Ponce-Hernandez et al. (2004),Mokany et al.
(2006) and Ramankutty et al. (2007). BGB was considered 20% of the
AGB based on a predictive relationship established from extensive
literature reviews (Cairns et al., 1997;Houghton et al., 2001;Achard
et al., 2002;Ponce-Hernandez et al., 2004;Mokany et al., 2006;Ram-
ankutty et al., 2007). The xed ratio was adopted to ensure consistency
across plots and species. However, the limitations and uncertainties of
this method are acknowledged. Therefore, future studies should incor-
porate root-to-shoot ratios based on detailed species-specic and
size-specic data as available.
2.2.2.2. Soil and litter. For assessing the changes in soil organic carbon
stock, it was adopted the soil sampling protocol proposed by Stolbovoy
et al. (2005) and Chiti et al. (2014). Three random sampling plots for
each land-use system were considered. In each plot, 20 ×20 m, soil
samples were collected along a grid of 25 sampling points at three
depths: 0–10, 10–20, 20–30 cm. Then the 25 samples for each depth
were pooled together to have three composite samples per depth and per
land use area. In the middle of each plot a minipit was opened to collect
bulk density samples at the same depths, to have one sample per depth
per plot and three samples per depth and per land use area, following the
core method (Blake and Hartge, 1986).
Where present, the litter layer was collected within a frame 40 ×40
cm (e.g., in the case of primary and secondary forests, rubber and cocoa
plantations) and 3 ×3 m in case of big leaves such as those of the oil
palm plantations. In the rest of the cases, the number of collected sam-
ples was 5 in each plot for a total amount of 15 in each site. For the
species with big leaves, a single leaf was collected in each plot and it has
been counted as the number of the leaves in the 3 ×3 m frame.
2.2.2.3. Deadwood biomass. Dead wood biomass was estimated in terms
of dead-standing trees. Per each dead standing tree, the biomass was
estimated using allometric equations (see Table 1). Results of DB were
expressed as a percentage of the AGB.
2.2.2.4. From biomass to carbon. According to IPCC (2003),Losi et al.
(2003),Sarmiento et al. (2005),Chave et al. (2005),Pearson et al.
(2005), and Fonseca et al. (2012), a coefcient of 0.5 was used to
convert tree biomass to carbon. Carbon per tree, expressed in kg, was
then multiplied by the number of trees per hectare to determine C in
above- and below-ground biomass on an area basis (MgC ha
−1
). The
exception was made for Musa acuminata, for which a conversion factor
of 0.46 was used (Hairiah et al., 2010;Danarto and Hapsari, 2015).
Overall the analyses, including the statistical ones, were performed,
described and discussed at the carbon level.
2.3. Data analysis
Annual carbon loss and gain was estimated comparing carbon pool
values of different land uses/covers to the reference values of forest
carbon pools (being primary forest the original land cover for all sites).
Considering the age of the plantations at the time of the survey, we
calculated the total change in carbon (loss or gain) and divided it by the
years since deforestation. This approach enabled the assessment of
annual changes across all study sites.
2.3.1. Laboratory analysis
2.3.1.1. Soil and litter. Soil samples were oven-dried at 60 ◦C, and
sieved at 2 mm to separate the rock fragments from the ne earth. Both
of the fractions were weighed. Fine earth was analyzed for total carbon
and nitrogen (N) using CN analyzer by dry combustion (Thermo Fin-
nigan Flash EA112 CHN). Bulk density samples were oven-dried at
105 ◦C until constant mass, then the oven-dry weights of the soil samples
were divided by the cylinder volume and calculated as Mg m
−3
. The soil
organic carbon stock was calculated as follows:
SOC =∑
Horizon=n
Horizon=1
SOCHorizon (Eq. 1)
Where
SOCHorizonx= [SOCx]⋅Bulk Densityx⋅Depthx⋅(1−frag)⋅10 (Eq. 2)
where SOC is soil C-content per unit area (MgC ha
−1
), [SOC] is carbon
concentration in soil sample (kgC kg
−1
soil), the Bulk Density is the soil
density of the ne earth expressed as (Mg m
−3
), Depth is the thickness of
the horizon within the considered section (cm) and frag is the percentage
of rock fragments, and xis for the different soil horizons (Poeplau et al.,
2017).
The litter samples were oven-dried at 60 ◦C, weighed, grounded and
analyzed for total C and N by dry combustion (Thermo Finnigan Flash
EA112 CHN). The C-stock was estimated by multiplying the weight of
dry-matter of the sample by the C-concentration and reporting the value
on a surface basis, as follows:
C=[(C%
100)⋅W]⋅(109
G)(Eq. 3)
Where C is the nal value of C in the litter layer (MgC ha
−1
), C% is the
percentage of carbon concentration in the considered sample, W is the
weight of the litter sample and G is the surface of the grid (cm
3
).
2.3.2. Statistical analysis
Statistical tests were performed to check for signicant differences in
C-pools across different land uses/covers. We t a one-factor ANOVA to
test if land uses have an overall signicant effect, and if so, we tested all
pairwise comparisons between the seven land uses. For this purpose, the
Table 1
Allometric equations used to assess Above-Ground Biomass (AGB) and Below-
Ground Biomass (BGB) in this study. AGB and BGB are expressed in kg of Dry
Matter per tree, Diameter at Breast Height (DBH) in cm, Tree Height (H) in m,
wood density (
ρ
) in g⋅cm
−3
and girth (g) in m.
Land use/species Equation References
Forest and secondary Forest AGB =0.17 ⋅DBH
1.97
⋅
H
0.55
Henry et al. (2010)
Cocoa plantation
Theobroma cacao L.
AGB =0.7217 ⋅
ρ
⋅
(DBH
2
)
0.921
Donkor et al. (2023)
Rubber plantation
Hevea brasiliensis (Willd. ex A.
Juss.) Müll.Arg.
AGB =0.0673 ⋅0.47 ⋅
DBH
2
⋅H
0.976
BGB =0.207 ⋅DBH
1.668
Chave et al. (2014)
Yang et al. (2017)
Coconut plantation
Cocos nucifera L.
AGB =H⋅g
2
⋅41.14142
BGB =13.5961 ⋅H
0.6635
Kumar et al. (2008)
Zahabu et al. (2018)
Oil palm plantation (young
trees)
Elaeis Guineensis Jacq.
Based on leaf carbon
content
Mixed plantation:
- Plantain Musa acuminata
Colla
- Coconut Cocos nucifera L.
- Oil palm (adult trees) Elaeis
Guineensis Jacq.
AGB =0.0303 ⋅
DBH
2.1345
AGB =H⋅g
2
⋅41.14142
AGB =(71.797 H) –
7.0872
Danarto and
Hapsari (2015)
Kumar et al. (2008)
Asari et al., (2013)
E. Grieco et al.
Journal of Environmental Management 367 (2024) 121993
5
post hoc general linear hypothesis test (glht)(p =0.05) was used to test
multiple hypotheses concerning a linear function of interest obtained
from the matrix of parametric contrasts calculated on the land use factor
by Tukey’s method (Bretz et al., 2010). We also performed a Dunnett
post hoc test (p =0.05) using the forest land use as a control group. We
adjusted the p-value based on the joint normal distribution of the linear
function (Dunnett and Tamhane, 1991). We used a sandwich estimator
that provides a heteroscedasticity-consistent covariance matrix estimate
for both post hoc tests when needed. Before proceeding, normality and
homogeneity of variance were checked via the Shapiro-Wilk and Lev-
ene’s tests, respectively (for both p =0.05). All analyses were performed
using the statistical software R, with the package ‘multcomp’and ‘sand-
wich’(Zeileis, 2004).
3. Results
Results for all carbon pools (as singularly taken and overall
accounted in TCS) are described in comparison to forest and other land
uses/covers and in absolute terms as also considering differences, within
the same land use/cover, between ages.
3.1. Carbon stocks and dynamics
3.1.1. Above-ground carbon (AGC)
The overall differences in AGC between forest, secondary forest and
all the plantation types result to be statistically signicant (p <0.05)
except for mixed plantation (p =0.19) (Fig. S1 in Supplementary Ma-
terial). Compared to the pristine forest, all the plantations show a
diminished AGC which varies both with the age and the type of plan-
tation (Table 2). It is noteworthy that, considering the age of the plan-
tations, the AGC of secondary forest (SF
10
), rubber (RP
14
and RP
50
),
mixed plantation (MP
36
, MP
50
) and coconut (CN
95
), is not statistically
different from the forest (Fig. S2 in Supplementary Material). The AGC
from seven subplots of primary forest, results to be, on average, 263.9 ±
8.5 MgC ha
−1
(here and elsewhere, ‘±’denotes one standard deviation).
This was considered the reference value to be confronted with. Sec-
ondary forest results in an AGC of 78.7 ±5.2 (SF
10
) and 33.1 ±2.5 MgC
ha
−1
(SF
20
), showing a loss of 70.2 and 87.5%, respectively (Table 3).
After deforestation, the observed mean AGC annual increment is 7.9 in
SF
10
and 1.65 MgC ha
−1
yr
−1
in SF
20
. In oil palm plantations, the AGC is
3.8 ±0.03 (4-year-old) and 3.4 ±0.1 MgC ha
−1
(8-year-old) with a
statistically signicant loss from the forest reference level of 98.6 and
98.7%, respectively. The estimated mean AGC annual increment is 1 ±
0.5 and 0.4 ±0.1 MgC ha
−1
yr
−1
in 4 and 8-year-old stands respectively.
For the 120-year-old cocoa plantation, AGC results in 19.5 ±0.1 while
the 34-year-old stand has 21.2 ±0.3 MgC ha
−1
. The loss in this case
accounts respectively for 92.6 and 92% with a mean AGC annual
increment since deforestation, accounting for 0.6 ±0.2 and 0.2 ±0.05
MgC ha
−1
yr
−1
in CC
34
and CC
120
. Four rubber plantation AGC stock
ranges between 10.7 ±0.3 (RP
5
) and 162.9 ±5 MgC ha
−1
(RP
50
). The C-
loss in this case varies from 95.9 to 38.3%, respectively, while the mean
AGC annual increment since forest clearing changes from 2.1 ±0.4 to
3.3 ±0.7 MgC ha
−1
yr
−1
, in RP
5
and RP
50
respectively. The AGC in MP
36
is 109.02 ±3.73 and 164.22 ±8.20 MgC ha
−1
in MP
50
, with a loss
accounting for 58.7 and 37.8%, respectively. In terms of mean annual
AGC increment, MP
36
is 3 ±0.5 while MP
50
was 3.3 ±2.1 MgC ha
−1
yr
−1
. Coconut AGC ranges from 44.4 ±0.5 (CN
15
), to 46.7 ±1.7 MgC
ha
−1
(CN
100
). The AGC loss on the eight coconut sites varies between
93.3 (CN
21
) to 56.7% (CN
95
) while the mean annual AGC increment
changes from 3 ±0.08 in CN
15
to 0.5 ±0.2 MgC ha
−1
yr
−1
in CN
100
(Tables 2 and 3, and Table S2 in Supplementary Material).
3.1.2. Below-ground Carbon (BGC)
As for the AGC pool, also BGC results are statistically different for all
the land uses compared to the forest with the only exception of mixed
plantation (p =0.07)(Fig. S3 in Supplementary Material). In this pool,
considering the age of the plantations, only the 50-year-old mixed
plantation (MP
50
) does not signicantly differ from the forest BGC
(Fig. S2 in Supplementary Material). Overall, forest shows the highest
BGC stock (68.24 ±2.20 MgC ha
−1
) among the land uses. Secondary
forest, after 10 years from deforestation, accounts for 20.5 ±1.6, while
after 21 years, the BGC was 8.69 ±1.4 MgC ha
−1
with a loss of 70.2 and
87.5%, respectively (Table 3). The estimated BGC annual increment
since deforestation is 2 ±0.7 and 0.4 ±0.2 MgC ha
−1
yr
−1
in SF
10
and
SF
20
. In the oil palm plantations, BGC is 0.8 ±0.01 (OP
25
, second
rotation, 4-year-old plantation) and 0.7 ±0.01 MgC ha
−1
(OP
8
). BGC
loss in the oil palm is respectively 99 and 98.9% while the mean annual
increment is 0.2 ±0.1 (OP
25(4)
) and 0.1 ±0.02 (OP
8
) MgC ha
−1
yr
−1
.
Cocoa BGC is 5.5 ±0.1 (CC
35
) and 5.1 ±0.03 (CC
120
) MgC ha
−1
with a
loss of 92 and 92.6%. The BGC mean annual increment is 0.2 ±0.05 in
Table 2
Carbon stocks of different pools (AGC, BGC, SOC, Litter) and TCS (AGC +BGC +Litter +SOC) in the forest, secondary forest and tree plantations (in MgC ha
−1
). The
subscript number on the site code indicates the years since deforestation, in parentheses the age of the plantation, ‘±’denotes one standard deviation. NA =Not
Available.
Land use/cover Site code Time from LUC
Years
Tree density
Trees ha
−1
AGC
MgC ha
−1
BGC
MgC ha
−1
SOC 0-30
MgC ha
−1
Litter
MgC ha
−1
DB
% of AGB
Total C stock
MgC ha
−1
Forest ANKF, CTF,
NANAF
–1086 263.9 ±8.45 68.24 ±2.20 87.10 ±30.07 7.47 ±5.08 NA 427.09 ±30.61
Secondary Forest SF
10
10 667 78.71 ±5.21 20.46 ±1.35 74.29 ±13.92 3.39 ±1.41 NA 176.85 ±14
SF
20
21 (20) 258 33.05 ±2.53 8.59 ±1.35 62.37 ±4.56 2.90 ±0.73 NA 106.92 ±5.62
Oil palm OP
8
8 400 3.38 ±0.07 0.68 ±0.01 65.85 ±13.07 NA NA 69.91 ±13.07
OP
25
25 (4) 500 3.80 ±0.03 0.76 ±0.01 57.27 ±7.85 5.78 ±3.16 NA 67.61 ±7.85
Cocoa CC
34
34 1500 21.23 ±0.33 5.08 ±0.03 57.03 ±5.11 2.95 ±0.84 7% 86.73 ±5.11
CC
120
120 2600 19.54 ±0.13 5.52 ±0.08 34.92 ±9.83 3.94 ±0.60 3% 63.48 ±9.83
Rubber RP
5
5 633 10.72 ±0.27 3.54 ±0.07 64.38 ±8.76 1.86 ±0.29 NA 80.50 ±8.77
RP
10
10 700 42.66 ±1.08 7.67 ±0.12 56.76 ±9.35 4.51 ±1.26 NA 111.60 ±9.34
RP
14
14 600 91.64 ±2.11 12.57 ±0.18 55.03 ±7.83 2.90 ±0.73 NA 162.14 ±7.79
RP
50
50 400 162.89 ±5 11.42 ±0.19 48.18 ±7.70 2.94 ±0.80 NA 225.43 ±7.29
Mixed MF
36
36 596 109.02 ±3.73 21.14 ±0.69 66.66 ±10.66 NA 5.4% 196.82 ±10.66
MF
50
50 567 164.22 ±8.70 32.92 ±1.73 68.51 ±13.17 NA 6.5% 265.65 ±13.00
Coconut CN
15
15 200 44.36 ±0.54 9.44 ±0.05 47.31 ±3.64 NA NA 101.11 ±3.66
CN
21
21 133 17.58 ±0.19 4.57 ±0.02 68.85 ±7.94 NA NA 91.00 ±7.91
CN
28
28 133 38.57 ±1.0 6.49 ±0.07 51.80 ±5.58 NA NA 96.86 ±5.58
CN
44
44 125 41.33 ±1.5 6.52 ±0.09 39.58 ±4.64 NA NA 87.43 ±4.6
CN
50
50 175 45.11 ±0.74 8.85 ±0.05 43.76 ±6.75 NA NA 97.72 ±6.7
CN
53
53 325 56.92 ±1.02 12.17 ±0.11 40.90 ±1.99 NA 6% 109.99 ±1.96
CN
95
95 225 114.14 ±2.51 14.11 ±0.16 28.17 ±21.78 NA NA 156.43 ±1.81
CN
100
100 183 46.65 ±1.72 9.32 ±0.19 41.58 ±1.99 NA 3% 97.55 ±2.27
E. Grieco et al.
Journal of Environmental Management 367 (2024) 121993
6
the 34-year-old plantation and 0.08 ±0.01 MgC ha
−1
yr
−1
in the 120-
year-old plantation. BGC in rubber plantation ranges from 3.5 ±0.1
(RP
5
) to 12.6 ±0.2 MgC ha
−1
(RP
14
). The BGB loss varies between 94.8
(RP
5
), to 81.7 (RP
14
), while the estimated mean annual increment is 0.7
±0.09 (RP
5
), 0.8 ±0.01 (RP
10
), 0.9 ±0.07 (RP
14
) and 0.2 ±0.05 MgC
ha
−1
yr
−1
(RP
50
), respectively. Mixed plantations have 21.1 ±0.7
(MP
36
) and 32.9 ±1.7 MgC ha
−1
(MP
50
), with 69.2 and 52% of total loss
compared to the forest level with BGC mean annual increment ac-
counting for 0.6 ±0.01 and 0.7 ±0.4 MgC ha
−1
yr
−1
on 36-year-old and
50-year-old plantations, respectively. Below-ground carbon stocks in
coconut plantations range from 4.6 ±0.02 (CN
22
) to 14.11 ±0.2 (CN
95
)
MgC ha
−1
, respectively. The loss for this land use is for all the ages,
above 70%, specically ranging from 79.4 (CN
95
) to 93.3% (CN
21
),
respectively, while mean annual BGC increment varies from 0.6 ±0.01
(CN
15
) to 0.1 ±0.05 MgC ha
−1
yr
−1
(CN
100
)(Tables 2 and 3, and
Table S2 in Supplementary Material).
3.1.3. Litter carbon
The litter layer was absent in several sites due to common practices
of removal, burning and utilization as fodder. Forest litter C-stock is 7.5
±5.1 MgC ha
−1
, higher than any of the other investigated plantations.
No statistically signicant differences were found for oil palm plantation
(p =0.82)(Fig. S4 in Supplementary Material). The litter layer was
collected in both the secondary forest sites, accounting for 3.4 ±1.4 in
SF
10
and 2.9 ±0.7 MgC ha
−1
in SF
20
, with differences compared to the
forest of 54.7 and 61.1%, respectively. On oil palm plantations, the litter
layer was collected in OP
25
(4-year-old plantation) since in OP
8
site was
regularly removed for other uses (mainly as fuelwood and fodder). The
litter carbon estimated on a leaf number basis results in 5.7 ±3.2 MgC
ha
−1
, differing from forest stock by 22.7%. Indeed litter in cocoa plan-
tations at 35 and 120 years is 2.95 ±0.8 and 3.9 ±0.6 MgC
ha
−1
respectively, having a percentage difference of 60.5 and 47.3%.
Rubber plantations has, in all the sites (5-10-14-50-year-old plantations,
respectively), a lower value of carbon litter compared to the forest’s.
Litter pools varies from 1.9 ±0.3 in RP
5
to 4.5 ±1.3 MgC ha
−1
in RP
10
,
while almost identical values are estimated in RP
14
(2.9 ±0.7 MgC
ha
−1
) and RP
50
(2.9 ±0.8 MgC ha
−1
). This plantation type has a litter C-
stock difference, compared to the forest, of about 75.1, 39.6, 61.2 and
60.7% respectively. The litter layer in any of the mixed plantations and
coconut sites investigated was regularly removed for other uses
(Tables 2 and 3, and Table S2 in Supplementary Material).
3.1.4. Soil carbon
Changes in SOC stocks at 0–30 cm depth, after deforestation, result in
a statistically signicant decrease (p <0.05) in all conversion into
plantations except for mixed plantation (p =0.095) and secondary forest
(p =0.115)(Fig. S5 in Supplementary Material). Forest SOC in the 0–30
cm layer, is 87.1 MgC ha
−1
. Comparing SOC in different land uses and
plantation ages, no signicant differences (p >0.05) are observed with
secondary forest (SF
10
and SF
21
), oil palm (OP
8
and OP
25(4)
), cocoa
(CC
34
), rubber (RP
5
), mixed (MP
36
, MP
50
) and coconut (CN
21
). SOC in
secondary forest ranges between 74.3 ±13.9 (SF
10
) and 62.4 ±4.6 MgC
ha
−1
(SF
21
) with a total loss of 14.7 and 28.4%, respectively (Table 2,
Table 3,Fig. 3). Mean annual SOC loss is 1.5% (1.3 ±1.4 MgC ha
−1
yr
−1
) and 1.4% (1.2 ±0.2 MgC ha
−1
yr
−1
) after 10 and 21 years from
deforestation, not differing signicantly from primary forest values. In
oil palm plantations, SOC is 57.3 (OP
25(4)
) and 65.9 MgC ha
−1
(OP
8
)
with decreases of 21.25 (34.3%) and 29.84 MgC ha
−1
(24.4%) after 8
and 25 years from deforestation. The mean SOC annual loss is 3% and
1.4% (2.7 ±2 and 1.2 ±1.6 MgC ha
−1
yr
−1
) in OP
25(4)
and OP
8
,
respectively. SOC in cocoa plantations accounts for 57 ±5.1 and 34.9 ±
9.8 MgC ha
−1
in CC
34
and CC
120
respectively. Cocoa plantation, 34 years
after deforestation, lead to a total SOC decrease of 30.1 MgC ha
−1
(34.5%) consisting of a mean annual loss of 0.9 ±0.1 MgC ha
−1
yr
−1
(1%) while in 120-year-old plantation the total loss accounts for 52.2
MgC ha
−1
(59.9%) with a mean annual soil loss of 0.43 ±0.08 MgC ha
−1
yr
−1
(0.5%). SOC in rubber plantations is 64.4 (RP
5
), 56.8 (RP
10
), 55
(RP
14
) and 48.2 (RP
50
) MgC ha
−1
but, compared to forest reference
level, was observed a SOC loss of 22.7 MgC ha
−1
(26.1%) in RP
5
, 30.3
(34.8%) in RP
10
, 32.1 (36.8%) in RP
14
and 38.9 (44.7%) in RP
50
. The
Table 3
Changes in carbon stock (in %) for different C-pools (AGC, BGC, Litter, SOC) and in TCS considering the primary forest as benchmark reference vs. secondary forest and
other tree plantations (positive values represent carbon accumulation and negative values carbon losses compared to the primary forest). The subscript number on the
site code indicates the years since deforestation, in parentheses the age of the plantation. AVG are mean losses in % between the primary forest and other land use/
cover. ***p-value <0.001, **p-value <0.01, *p-value <0.05. NA =Not Available Data.
Land use/cover Site code Time from LUC
Years
AGC BGC Litter SOC TCS
Secondary Forest AVG −78.82%* −78.82%* −57.91%*** −21.55% −66.79%*
SF
10
10 −70.17%* −70.17%* −54.68%** −14.71% −58.59%
SF
20
21 (20) −87.48%* −87.48%** −61.15%*** −28.39% −74.96%**
Oil palm AVG −98.64%** −98.95%** −61.33% −29.32%* −83.90%**
OP
8
8−98.72%** −99.01%** NA −24.40% −83.63%**
OP
25
25 (4) −98.56%** −98.89%** −22.65% −34.52% −84.17%**
Cocoa AVG −92.28%** −92.28%** −53.90%*** −47.22%*** −82.41%**
CC
34
34 −91.96%** −91.96%** −60.52%*** −34.52% −79.69%**
CC
120
120 −92.60%** −92.60%** −47.28%** −59.91%*** −85.14%**
Rubber AVG −70.83%* −87.18%** −59.14%*** −35.61%*** −66.07%*
RP
5
5−95.94%** −94.84%** −75.11%*** −26.09% −81.15%**
RP
10
10 −83.84%* −88.83%** −39.60%* −34.83%* −73.87%*
RP
14
14 −65.27% −81.68%* −61.19%*** −36.82%* −62.04%*
RP
50
50 −38.28% −83.36%* −60.66%*** −44.69%** −47.22%
Mixed AVG −48.23% −60.61% −22.41% −45.86%
MF
36
36 −58.69% −69.19%* NA −23.47% −53.92%
MF
50
50 −37.77% −52.02% NA −21.35% −37.80%
Coconut AVG −80.83% * −86.98%* −48.06%*** −75.47%**
CN
15
15 −83.19%* −86.24%** NA −45.68%* −76.33%**
CN
21
21 −93.34%** −93.34%** NA −20.95% −78.69%**
CN
28
28 −85.39%* −90.54%** NA −40.53%* −77.32%**
CN
44
44 −84.34%* −90.49%** NA −54.56%*** −79.53%**
CN
50
50 −82.91%* −87.10%** NA −49.76%*** −77.12%**
CN
53
53 −78.43%* −82.27%* NA −53.04%*** −74.25%*
CN
95
95 −56.75% −79.43%* NA −67.66%*** −63.37%*
CN
100
100 −82.32%* −86.42%* NA −52.27%*** −77.16%**
E. Grieco et al.
Journal of Environmental Management 367 (2024) 121993
7
mean annual SOC loss is 4.5 ±1.7 (5.2%) in RP
5
, 3 ±0.9 (3.5%) in RP
10
,
2.3 ±0.07 (2.6%) in RP
14
and 0.78 ±0.05 MgC ha
−1
yr
−1
(0.9%) in
RP
50
. Mixed plantation SOC is 66.7 and 68.5 MgC ha
−1
after 36 and 50
years from plantation establishment, respectively. SOC decreased of
20.4 MgC ha
−1
(23.5%) after 36 years with a mean annual loss of 0.57 ±
0.3 MgC ha
−1
yr
−1
(0.7%). After 50 years the total loss accounts for 18.6
MgC ha
−1
(21.3%) with a mean annual loss of 0.4 ±0.2 MgC ha
−1
yr
−1
(0.4%). Differences in SOC between forest and mixed are not statistically
signicant. SOC in coconut ranges from 28.17 ±21.78 in CN
95
to 68.85
±7.94 in CN
21
. Compared to forest level, SOC loss varies from 18.2
(21%) in CN
22
to 58.9 (67.7%) MgC ha
−1
in CN
95
. The SOC mean annual
loss changes from 2.7 ±0.2 (3%)(CN
15
) to 0.5 ±0.02 (0.5%) (CN
100
)
(Tables 2 and 3, and Table S2 in Supplementary Material).
3.1.5. Dead standing carbon
The dead wood accounts for 3% of the total biomass in the 120-year-
old cocoa plantation and 7% in the 34-year-old cocoa plantation. In
coconut plantations, these percentages are 3% (CN
100
) and 6% (CN
53
),
while is 6% for 50-year-old mixed plantations (Table 2).
3.1.6. Total carbon
Total carbon differences between forest, secondary forest and tree
plantations are all statistically signicant (p <0.05), except for mixed
plantation (p =0.12)(Fig. 2).
Considering total carbon (TCS) as the sum of all C-stocks for each
pool (but excluding the dead standing wood), forest accounts for 427.1
±30.6 MgC ha
−1
. This value results higher than any plantation type
observed in this study (Fig. 3). The estimated TSC in the secondary forest
is 176.8 ±14 and 106.9 ±5.6 MgC ha
−1
with a total C-loss of 250.2
(58.6%) and 320.2 MgC ha
−1
(75%) in SF
10
and SF
21
respectively. TCS
loss is 25 ±2 (5.9%) and 15.2 ±1.3 MgC ha
−1
yr
−1
(3.6%) after 10 and
21 years since deforestation. Oil palm plantation total carbon is 69.9 ±
13.1 (OP
8
) and 67.6 ±7.8 (OP
25(4)
) MgC ha
−1
. These losses, compared
to the forest reference level, account for 357.2 (83.6%) and 359.5
(84.2%) MgC ha
−1
, respectively, with an annual mean TSC loss of 44.6
±1.7 (10.5%) and 15 ±3.3 MgC ha
−1
yr
−1
(3.5%) after 8 and 25 years
from deforestation. Cocoa TSC is 86.7 ±5.1 and 63.5 ±9.8 MgC ha
−1
,
which compared to the forest reference values, represent a loss of 340.4
and 363.6 MgC ha
−1
in CC
34
and CC
120
respectively. This means 10 ±
0.4 (2.3%) and 3 ±0.1 MgC ha
−1
yr
−1
(0.7%) of total mean annual loss,
respectively, after 34 and 120 years since forest clearing. Rubber
plantation TSC ranges from 80.5 (RP
5
) to 225.4 MgC ha
−1
(RP
50
) with
mean C-losses varying from 346.6 (81.2%) in RP
5
and 201.7 (47.2%) in
RP
50
. Mean annual total C-loss changes from 69.3 ±1.6 (16.2%) in RP
5
to 4 ±0.9 MgC ha
−1
yr
−1
(0.9%) in RP
50
. Mixed plantations present a
TCS of 196.8 ±10.7 and 265.6 ±13 MgC ha
−1
in MP
36
and MP
50
with
mean total C-loss of 230.3 (53.9%) and 161.4 MgC ha
−1
(37.8%) after 36
and 50 years from forest clearing, corresponding to 6.4 ±0.5 (1.5%) and
3.2 ±2.5 MgC ha
−1
yr
−1
(0.8%) of mean annual TSC loss. In coconut
plantations, TCS ranges from 87.4 ±4.6 (CN
44
) to 156.4 ±21.8 MgC
ha
−1
(CN
95
). The total loss varies from 339.7 (79.5%) in CN
44
to 270.7
(63.4%) MgC ha
−1
in CN
95
. The mean annual total carbon loss changes
from 21.7 ±0.3 (5.1%) in CN
15
to 2.8 ±0.6 (0.7%) MgC ha
−1
yr
−1
in
CN
95
(Tables 2 and 3, and Table S2 in Supplementary Material).
4. Discussions
This work aimed to evaluate the effects of land use change in tropical
forest land converted to tree plantations or abandoned to grow as sec-
ondary forest. To this aspect, we questioned whether these tropical Af-
rican ecosystems experience C-stock changes in C-pools at the ecosystem
level after deforestation and land use/land cover change and if the range
of this variation depended on the pool, the stand age and the type of
conversion considered.
4.1. Primary forest and mixed plantation lead in C-stock potential
Primary forests generally do not serve as direct sources of economic
income for local communities, especially when compared to revenue
generated from tree plantations. In our study, these areas, as well as
secondary forests, are considered sacred by community members and
only a few activities are allowed (e.g. very sporadic hunting and non-
timber product harvesting in secondary forests)(Grieco, 2011). How-
ever, their role at the ecosystem level remains undeniable considering
their C-storage potential. Indeed, comparing forest C-stock with sec-
ondary forests and tree plantations, we observed that both the total
C-stock, and for each single pool, carbon is consistently higher in the
primary forest (427.1 ±30.6 vs. 122.6 ±56 MgC ha
−1
as the mean
value of the other TSC’s land uses). Few studies in the literature have
been found to consider and describe C-stocks and dynamics in all the
C-pools at the same time (i.e. AGC, BGC, Litter and SOC) for identifying
the carbon ecosystem potential as we considered here. Our results are,
Fig. 2. Bar plots represent different Total Carbon Stock (TCS in MgC ha
−1
) in different land uses. Within the same bar, different colours represent different stocks in
different C pool (in MgC ha
−1
)(left panel). Statistical signicance is represented with Dunnet plot, where condence interval crossing the reference dotted line (0)
indicates no statistical difference (p >0.05)(right panel).
E. Grieco et al.
Journal of Environmental Management 367 (2024) 121993
8
however, within the range observed by Sierra et al. (2007) for a tropical
forest in Colombia (299.4–467.9 MgC ha
−1
). We found that after
deforestation and conversion, only mixed plantations showed no sta-
tistically signicant differences with forest in terms of total C-stock, and
at any of the considered ages of the stand. This appears to align with
ndings discussed by Hulvey et al. (2013) about the benets of
tree-mixed plantations for their higher C-storage potential compared to
monocultures. He et al. (2013) also observed that ecosystem C-storage in
mixed plantations is higher than in monoculture in subtropical regions
in China, particularly evident considering the soil C-pool. In our study,
tree monoculture plantations had similar behaviour to mixed ones with
no signicative differences compared to forest if not at some specic
stand age and for some of the C-pool as in SF
10
(AGC, SOC, TCS),
SF
20
(SOC), OP
8
(SOC), RP
5
(SOC), RP
14
(AGC), RP
50
(AGC, TCS),
CN
21
(SOC) and CN
95
(AGC, SOC, TCS).
4.2. AGC and SOC are key contributors to total C-stock
In the current study, the estimated reference value of AGC stocks for
primary forest is higher than the values observed by Sierra et al. (2007)
(247.8 ±40.5 MgC ha
–1
) in Colombia, by Gineste et al. (2008)(from
154.2 to 171.1 MgC ha
−1
) in Ghana, by Lewis et al. (2009) (202 MgC
ha
−1
) from 79 plots in tropical African forest and by Adu-Bredu et al.
(2010) in Ghana (202.07 MgC ha
−1
) but perfectly in line with values
referred to the specic forest-vegetation zone in Ghana by Nyarko et al.
(2024)(254–278 MgC ha
−1
) and Houssoukp`
evi et al. (2022) in Benin
(279 ±74 MgC ha
−1
). We estimate that the AGC in primary forest is
61.7% of the total C-stock, followed by SOC (20.4%), BGC (16.1%) and
litter carbon (1.7%). Although with different proportions, SOC and AGC
are the pools contributing in a larger part to the total ecosystem C-stock
in forestland. At the global scale, this is also conrmed by FAO (2020),
with 44% of TCS in living biomass, 4% in deadwood, 6% in litter and
45% in soil, respectively. Considering values reported by Sierra et al.
(2007), proportions of SOC and AGC were recalculated for a proper
comparison, taking into account only SOC in the rst 30 cm, resulting in
AGC being 44% of the TCS while SOC 38% and BGC 15%. Tree
plantations in this study also follow this pattern. It is particularly pro-
nounced the contribution of AGC in older coconut and rubber planta-
tions (73% in CN
95
and 72.3% in RP
50
) as well as in forest and MP
50
(both 61.8%). Over 50% of the total C-stock is found in CN
54
(51.8%),
MP
36
(55.4%) and RP
14
(56.5%). Rubber proportion of AGC in 14-year--
old plantation, perfectly coincides with 56% found by Wauters et al.
(2008), for the same stand age in Ghana. Along with AGC, SOC is a
prevalent contributor to the total C-stock at the ecosystem level, and this
is particularly in younger plantations representing 94.2 and 84.7% of
the total C-stock in oil palm plantations (OP
8
and OP
25(4)
), 80% in RP
5
,
75.7% in CN
22
and the 65.8% in CC
34
, respectively. These results are
similar to those of Houssoukp`
evi et al. (2022) who reported SOC
contribution to TCS of 63% in tree plantations and young palm SOC
contributing for 54% in southern Benin. The higher contribution of SOC
to TCS could be attributed to lower AGC due to young stand age and to
the original forest SOC still kept after deforestation.
The importance of forest soils at the ecosystem level is widely
acknowledged, along with their vulnerability to substantial depletion of
SOC upon conversion in agricultural land and tree plantation (Hassink
and Whitmore, 1997;Van Noordwijk et al., 2000;Tan et al., 2009;Chiti
et al., 2014,2016;Quezada et al., 2022). In our study, forest SOC values
in the rst 30 cm lie well within 89.6 ±4.6 MgC ha
−1
value reported by
Chiti et al. (2014) in Ghana and the 84.5 ±40.8 MgC ha
−1
reported by
Ledo et al. (2020) from a global dataset. Higher levels of SOC depletion
are found in the secondary forest compared to Don et al. (2011) in the
0–38 cm layer for a 28-year-old secondary forest (12.6 MgC ha
−1
, with a
loss of 13% from the previous forest C), perhaps depending on the data
derived from 39 different tropical countries. The SOC values we found in
the cocoa plantation after 120 years from deforestation were similar to
the value of 35.9 ±3.1 MgC ha
−1
observed in 30-year-old cocoa in
Ghana, by Adiyah et al. (2023), at a soil depth of 20–60 cm. Adiyah et al.
(2023) presented trends with peculiar directions regarding gain in SOC
after an initial decline during the rst three years of plantation. We do
not have comparable data from young stands but we observed a deple-
tion rate in SOC rate varying from 1 to 0.5% in 34 and 120-year-old
plantations respectively, that could not be considered as an increasing
Fig. 3. Total Carbon Stocks (TCS =AGC +BGC +SOC +Litter, in MgC ha
−1
) for different land use systems at different stand ages (Age)(AGC =Above Ground
Carbon, BGC =Below Ground Carbon, SOC =Soil Organic Carbon).
E. Grieco et al.
Journal of Environmental Management 367 (2024) 121993
9
trend but probably as a slowing of the rate of loss with plantation
because of ageing. The total loss of SOC (0–30 cm) results we found are
higher, in absolute values, in older stands but the mean annual loss rate
declines with ageing. Indeed, the average annual loss is greater for all
plantations during the rst years after deforestation with percentages
declining from 5.2% (RP
5
) to 0.9% after 50 years (RP
50
). The same trend
was estimated in coconut plantations with an annual decrease of 3% in
the rst 15 years, reaching 0.5% after 100 years. This might suggest that
in the long term, the depletion of SOC slows down even though in ab-
solute values of loss, the difference remains greater for older plantations.
Comparing even-aged plantations (CN
50
, RP
50
and MP
50
), SOC depletion
is highest in coconut with almost 50% followed by rubber with little less
than 45%. Mixed plantations had a no-signicant SOC depletion ac-
counting for 21.3%. The absence of statistically signicant loss in SOC
compared to forest emphasizes the importance of mixed plantations,
especially in tropical ecosystems where they are widely adopted by local
communities but also because they may represent a win-win solution in
terms of livelihoods, food security and, at the same time, for controlling
SOC erosion and promoting SOC recovery, overall guaranteeing a higher
systemic sustainability.
4.3. AGC and BGC accumulation rate is higher in younger stands
Several studies show the prominent effect of age in reducing carbon
sequestration capacity and efciency with ageing (Fonseca et al., 2011;
Carbone et al., 2013;Collalti et al., 2020;Luo et al., 2024). Our results
show that among all considered land uses/land covers, AGC annual
accumulation rate is higher than any other age class in young secondary
forest (SF
10
) and young rubber plantation (RP
14
). Secondary forests in
humid tropical forest zones are reported to have a greater C-accumu-
lation (due to faster-increasing biomass) during the rst 10 years
(Fonseca et al., 2011) and then declining with ageing. This is evident in
rubber AGC accumulation which we found to be similar to the 5.1 MgC
ha
−1
yr
−1
reported by Kongsager et al. (2013) for a 12-year-old plan-
tation and to 5.4 MgC ha
−1
yr
−1
derived by Wauters et al. (2008) for a
14-year-old rubber plantation (AGC accounting 76.3 MgC ha
−1
) both in
western Ghana. With stand ageing, we observed a decline in the accu-
mulation rate for the 50-year-old rubber, nearing the reported values of
3.6 MgC ha
−1
yr
−1
by Brahma et al. (2018) in a 40-year-old plantation in
India and, similarly, close to the 4.9 MgC ha
−1
yr
−1
value reported by
Kongsager et al. (2013) in a 40-year-old in Ghana. In our oil palm AGC
estimates, values for the 8-year-old are about six times lower than those
from Kongsager et al. (2013) for the 7-year-old plantation (21.7 vs. 3.4
MgC ha
−1
). Such a difference may be attributable to a different AGB
assessment method highlighting the importance of the method used to
estimate the C-pool and the relative uncertainty related to the allometric
equations adopted (Vorster et al., 2020). In our case, we did not estimate
the AGB through allometric equations but we directly estimated the AGC
based on leaves C-content due to the young age of the stands given the
absence of stems in such species in the juvenile phase. Another aspect to
be considered could be the sites’peculiarity. Indeed, despite the sites
having suitable characteristics for oil palm cultivation, management is
undertaken by smallholders, likely with limited resources for ensuring
optimal and enhanced plant growth compared to large-scale plantations
or, as in the case of Kongsager et al. (2013) study, within an agricultural
research station. Considering the annual C-accumulation rate in cocoa
plantation we found a reduction in the annual rate from 0.6 for the
34-year-old to 0.2 MgC ha
−1
yr
−1
for the 120-year-old plantation. Our
values are lower than those reported by Somarriba et al. (2013)(1.3–2.6
MgC ha
−1
yr
−1
) for cocoa plantations in Central America (but under the
agroforestry system) and lower than the rate of 3.1 observed by
Kongsager et al. (2013) in Ghana and referred to a younger stand (i.e.
21-year-old). Different values could be the result, in the rst case, of a
more complex and rich stand (agroforestry vs. monoculture) and in the
second case could be derived by the younger age in the Kongsager et al.
(2013) study. Coconut plantation AGC accumulation rate shows the
same declining pattern, as in the other plantations, as the stands grow
and reach a maturity stage. Studies from Bhagya et al. (2017) reported
51.14 MgC ha
−1
for the AGC in 50-year-old stand in India, which
resulted in 1.02 MgC ha
−1
yr
−1
, consistently similar to 0.9 MgC ha
−1
yr
−1
from our rate of the same stand age.
A contributing pool to the total ecosystem C-stock is the BGC, which
also regulates nutrient cycling and participates in carbon sequestration
and climate change mitigation. BGC in cocoa monoculture was esti-
mated at 5.4 MgC ha
−1
by Borden et al. (2019) in Ghana, which is
similar to values ranging from 5.08 to 5.52 MgC ha
−1
from our study
(CC
120
, CC
34
). In the 14-year-old rubber plantation, our estimates of
BGC are higher than 7.8 MgC ha
−1
found by Wauters et al. (2008) in
Ghana for the same stand age. We argue that the clone type, which may
be different from ours, could affect the results as well as the allometric
equations adopted. Overall among the sites, the higher annual accu-
mulation rate of BGC is found in secondary forest. 10 years after forest
clearing, the BGC has a comparable value to what was assessed by
Brown et al. (2020) for a 44-year-old secondary forest (20.5 vs. 20.7
MgC ha
−1
) in Ghana. All the rest of the sites had shown annual accu-
mulation rates lower than 1 MgC ha
−1
, implying a less predominant role
in the TCS balance. No signicant differences were found in MP
50
compared to the reference level of BGC in the forest, following the same
outcomes as for AGC.
5. Conclusions
Deforestation and land use changes and signicantly impact carbon
stock dynamics in various ecosystems, inuencing development policies
and natural resources conservation strategies. Our study compares the
potentiality of C-sequestration of some of the main and more common
land uses in tropical ecosystems for a deeper understanding of how
different land uses/covers and plantation ages impact carbon seques-
tration and storage over time. Although primary forest ecosystem has
the highest amount of carbon in any of the assessed pools, we found that
mixed plantation and secondary forest show comparable carbon
sequestration potential, particularly in AGC and SOC pools. This un-
derlines the crucial role of local smallholders in preserving and restoring
C-stock at ecosystem level, alongside the key role of local government in
protecting and enhancing forest resources. Further studies are needed
for secondary forests as carbon reservoirs, particularly in the tropics. We
found that SOC is higher in younger stands, likely representing residue
of the previous forest cover asset which decreases over time once a
plantation is established. Our work underscores the growing need for a
comprehensive carbon estimates that account for and consider all C-
pools within the ecosystem.
Our study suggests that tailored land management, such as the
establishment of mixed tree plantations on degraded or agricultural
lands, can improve smallholder livelihood, food security as well as
carbon sequestration, contributing signicantly to climate change
mitigation. Understanding ecosystem carbon sequestration and stock
dynamics post-deforestation is a key point in developing sustainable
mitigation strategies and forest protection planning. This requires
assessing the complex environmental and socio-economic interactions
within deforestation-risk areas for effective and long-term sustainable
rural development.
CRediT authorship contribution statement
Elisa Grieco: Writing –review &editing, Writing –original draft,
Investigation, Formal analysis, Data curation, Conceptualization. Elia
Vangi: Writing –original draft, Investigation, Formal analysis. Tom-
maso Chiti: Writing –original draft, Investigation, Formal analysis,
Data curation. Alessio Collalti: Writing –review &editing, Writing –
original draft, Investigation, Formal analysis, Conceptualization.
E. Grieco et al.
Journal of Environmental Management 367 (2024) 121993
10
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgements
Fieldwork was funded by grants from Africa-GHG: 247349 under the
European Union’s Seventh Framework Programme (FP7/2007–2013).
T.C. and A.C. also acknowledge the project funded under the Na-
tional Recovery and Resilience Plan (NRRP), Mission 4 Component 2
Investment 1.4 - Call for tender No. 3138 of December 16, 2021, recti-
ed by Decree n.3175 of December 18, 2021 of Italian Ministry of
University and Research funded by the European Union –NextGener-
ationEU under award Number: Project code CN_00000033, Concession
Decree No. 1034 of June 17, 2022 adopted by the Italian Ministry of
University and Research, CUP B83C22002930006, Project title “Na-
tional Biodiversity Future Centre - NBFC”.
We would like to gratefully acknowledge the Wildlife Division of the
Forestry Commission in Ghana, Ankasa Park managers Kareem A. F. and
Balangtaa C. and all the park rangers for their unwavering helpfulness.
Special thanks go to the eld team Mensa J.J., Cudjoe E. and Cobienna J.
who helped with data collection on the eld. Gratitude and great
appreciation are expressed to the local communities and traditional
authorities, for their warm welcome and trust, which made our work
possible.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.jenvman.2024.121993.
References
Achard, F., Eva, H.D., Stibig, H., Mayaux, P., Gallego, J., Richards, T., Malingreau, J.,
2002. Determination of deforestation rates of the world’s humid tropical forests.
Science 297, 999–1002. https://doi.org/10.1126/science.1070656, 10.1126/
science.1070656.
Adiyah, F., Csorba, ´
A., Dawoe, E., Ocansey, C.M., Asamoah, E., Szegi, T., et al., 2023. Soil
organic carbon changes under selected agroforestry cocoa systems in Ghana.
Geoderma Regional 35, e00737. https://doi.org/10.1016/j.geodrs.2023.e00737.
Adu-Bredu, S., Abekoe, M.K., Tachie-Obeng, E., Tschakert, P., 2010. Carbon stock under
four land use systems in three varied ecological zones in Ghana. In: Presented at the
Open Science Conference on “Africa and Carbon Cycle: the Carbo Africa Project”.
Accra.
Ahn, P., 1961. Soils of the lower tano basin, south-western Ghana. Ministry of Food and
Agricolture, Scientic Service Division, Soil and Land use Survey Branch Kumasi.
Book: Memo. 2 Soil Ld-Use Surv. Br. Ghana Minist. Fd Agric., p. 266
Ampim, P.A., Ogbe, M., Obeng, E., Akley, E.K., MacCarthy, D.S., 2021. Land cover
changes in Ghana over the past 24 years. Sustainability 13 (9), 4951. https://doi.
org/10.3390/su13094951.
Asari, N., Suratman, M.N., Jaafar, J., Khalid, M.M., 2013. Estimation of aboveground
biomass for oil palm plantations using allometric equations. November. In: 4th Int.
Conf. on Biology, Environment and Chemistry. https://www.cabidigitallibrary.org/
doi/full/10.5555/20143181981.
Atsivor, L., Dowuona, G.N., Adiku, S., 2001. Farming system-induced variability of some
soil properties in a sub-humid zone of Ghana. Plant Soil 236, 83–90. https://doi.org/
10.1023/A:1011907420292.
Bhagya, H.P., Maheswarappa, H.P., Bhat, R., 2017. Carbon sequestration potential in
coconut-based cropping systems. https://krishi.icar.gov.in/jspui/bitstream
/123456789/13435/1/Carbon.pdf.
Blake, G.R., Hartge, K.H., 1986. Methods of Soil Analysis. Bulk Density, Part 1, second
ed. ASA and SSSA, Madison, pp. 363–375. Agronomy 9.
Borden, K.A., Anglaaere, L.C., Adu-Bredu, S., Isaac, M.E., 2019. Root biomass variation of
cocoa and implications for carbon stocks in agroforestry systems. Agrofor. Syst. 93,
369–381. https://doi.org/10.1007/s10457-017-0122-5.
Brahma, B., Nath, A.J., Sileshi, G.W., Das, A.K., 2018. Estimating biomass stocks and
potential loss of biomass carbon through clear-felling of rubber plantations. Biomass
Bioenergy 115, 88–96. https://doi.org/10.1016/j.biombioe.2018.04.019.
Bretz, F., Hothorn, T., Westfall, P., 2010. Multiple Comparisons Using R. CRC Press.
Brown, H.C., Berninger, F.A., Larjavaara, M., Appiah, M., 2020. Above-ground carbon
stocks and timber value of old timber plantations, secondary and primary forests in
southern Ghana. For. Ecol. Manag. 472, 118236 https://doi.org/10.1016/j.
foreco.2020.118236.
Cairns, M.A., Brown, S., Helmer, E.H., Baumgardner, G.A., 1997. Root Biomass
Allocation in the World‘s Upland Forests –Oecologia- 111, pp. 1–11. https://doi.
org/10.1007/s004420050201.
Carbone, M.S., Czimczik, C.I., Keenan, T.F., Murakami, P.F., Pederson, N., Schaberg, P.
G., et al., 2013. Age, allocation and availability of nonstructural carbon in mature
red maple trees. New Phytol. 200 (4), 1145–1155. https://doi.org/10.1111/
nph.12448.
Chave, J., Andalo, C., Brown, S., Cairns, M.A., Chambers, J.Q., Eamus, D., F¨
olster, H.,
Fromard, F., Higuchi, N., Kira, T., Lescure, J.-P., Nelson, B.W., Ogawa, H., Puig, H.,
Ri´
era, B., Yamakura, T., 2005. Tree allometry and improved estimation of carbon
stocks and balance in tropical forests. Oecologia 145, 87–99. https://doi.org/
10.1007/s00442-005-0100-x.
Chave, J., R´
ejou-M´
echain, M., Búrquez, A., Chidumayo, E., Colgan, M.S., Delitti, W.B.,
et al., 2014. Improved allometric models to estimate the aboveground biomass of
tropical trees. Global Change Biol. 20 (10), 3177–3190. https://doi.org/10.1111/
gcb.12629.
Chiti, T., Grieco, E., Perugini, L., Rey, A., Valentini, R., 2014. Effect of the replacement of
tropical forests with tree plantations on soil organic carbon levels in the Jomoro
district, Ghana. Plant Soil 375, 47–59. https://doi.org/10.1007/s11104-013-1928-1.
Chiti, T., Perugini, L., Vespertino, D., Valentini, R., 2016. Effect of selective logging on
soil organic carbon dynamics in tropical forests in central and western Africa. Plant
Soil 399, 283–294. https://doi.org/10.1007/s11104-015-2697-9.
Collalti, A., Ibrom, A., Stockmarr, A., Cescatti, A., Alkama, R., Fern´
andez-Martínez, M.,
Matteucci, G., Sitch, S., Friedlingstein, P., Ciais, P., Goll, D.S., Nabel, J.E.M.S.,
Pongratz, J., Arneth, A., Haverd, V., Prentice, I.C., 2020. Forest production efciency
increases with growth temperature. Nat. Commun. 11 (1), 5322. https://doi.org/
10.1038/s41467-020-19187-w.
Danarto, S.A., Hapsari, L.I.A., 2015. Biomass and carbon stock estimation inventory of
Indonesian bananas (Musa spp.) and its potential role for land rehabilitation.
BIOTROPIA 22 (2), 102–108. https://doi.org/10.11598/btb.2015.22.2.376.
Don, A., Schumacher, J., Freibauer, A., 2011. Impact of tropical land-use change on soil
organic carbon stocks–a meta-analysis. Global Change Biol. 17 (4), 1658–1670.
https://doi.org/10.1111/j.1365-2486.2010.02336.x.
Donkor, E., Adu-Bredu, S., Jnr, E.M.O., Andam-Akorful, S.A., Mohammed, Y., 2023.
Biomass estimation models for cocoa (theobroma cacao) plantations in Ghana, West
Africa. Open J. Appl. Sci. 13, 1588–1618. https://doi.org/10.4236/
ojapps.2023.139126.
Dunnett, C.W., Tamhane, A.C., 1991. Step-down multiple tests for comparing treatments
with a control in unbalanced one-way layouts. Stat. Med. 10 (6), 939–947.
FAO, 2020. Global Forest Resources Assessment 2020: Main Report. https://doi.org/
10.4060/ca9825en. Rome.
Fonseca, W., Benayas, J.M.R., Alice, F.E., 2011. Carbon accumulation in the biomass and
soil of different aged secondary forests in the humid tropics of Costa Rica. For. Ecol.
Manag. 262 (8), 1400–1408. https://doi.org/10.1016/j.foreco.2011.06.036.
Fonseca, W., Alice, F.E., Rey-Benayas, J.M., 2012. Carbon accumulation in aboveground
and belowground biomass and soil of different age native forest plantations in the
humid tropical lowlands of Costa Rica. N. For. 43, 197–211. https://doi.org/
10.1007/s11056-011-9273-9.
Friedlingstein, P., O’Sullivan, M., Jones, M.W., Andrew, R.M., Bakker, D.C.E., Hauck, J.,
Landschützer, P., Le Qu´
er´
e, C., Luijkx, I.T., Peters, G.P., Peters, W., Pongratz, J.,
Schwingshackl, C., Sitch, S., Canadell, J.G., Ciais, P., Jackson, R.B., Alin, S.R.,
Anthoni, P., Barbero, L., Bates, N.R., Becker, M., Bellouin, N., Decharme, B., Bopp, L.,
Brasika, I.B.M., Cadule, P., Chamberlain, M.A., Chandra, N., Chau, T.-T.-T.,
Chevallier, F., Chini, L.P., Cronin, M., Dou, X., Enyo, K., Evans, W., Falk, S., Feely, R.
A., Feng, L., Ford, D.J., Gasser, T., Ghattas, J., Gkritzalis, T., Grassi, G., Gregor, L.,
Gruber, N., Gürses, ¨
O., Harris, I., Hefner, M., Heinke, J., Houghton, R.A., Hurtt, G.C.,
Iida, Y., Ilyina, T., Jacobson, A.R., Jain, A., Jarníkov´
a, T., Jersild, A., Jiang, F.,
Jin, Z., Joos, F., Kato, E., Keeling, R.F., Kennedy, D., Klein Goldewijk, K., Knauer, J.,
Korsbakken, J.I., K¨
ortzinger, A., Lan, X., Lef`
evre, N., Li, H., Liu, J., Liu, Z., Ma, L.,
Marland, G., Mayot, N., McGuire, P.C., McKinley, G.A., Meyer, G., Morgan, E.J.,
Munro, D.R., Nakaoka, S.-I., Niwa, Y., O’Brien, K.M., Olsen, A., Omar, A.M., Ono, T.,
Paulsen, M., Pierrot, D., Pocock, K., Poulter, B., Powis, C.M., Rehder, G.,
Resplandy, L., Robertson, E., R¨
odenbeck, C., Rosan, T.M., Schwinger, J., S´
ef´
erian, R.,
Smallman, T.L., Smith, S.M., Sospedra-Alfonso, R., Sun, Q., Sutton, A.J., Sweeney, C.,
Takao, S., Tans, P.P., Tian, H., Tilbrook, B., Tsujino, H., Tubiello, F., van der Werf, G.
R., van Ooijen, E., Wanninkhof, R., Watanabe, M., Wimart-Rousseau, C., Yang, D.,
Yang, X., Yuan, W., Yue, X., Zaehle, S., Zeng, J., Zheng, B., 2023. Global carbon
budget 2023. Earth Syst. Sci. Data 15, 5301–5369. https://doi.org/10.5194/essd-15-
5301-2023.
Gineste, M., Martel, S., Henry, M., Adu-Bredu, S., Saint-Andre, L., 2008. Estimating the
impact of selective logging on aboveground carbon stocks in Boi Tano Forest
Reserve. In: Proceedings of the Carboafrica Annual Meeting, pp. 25–27. Citeseer
Accra, Ghana.
Grieco, E., 2011. Land Use Change and Carbon Stock Dynamics in Sub-saharan Africa-
Case Study of Western Africa–Ghana. University of Tuscia. Ph.D. Thesis. https://ds
pace.unitus.it/bitstream/2067/2435/1/egrieco_tesid.pdf.
Hairiah, K., Dewi, S., Agus, F., Velarde, S., Ekadinata, A., Rahayu, S., Van Noordwijk, M.,
2010. A Manual: Measuring Carbon Stocks across Land Use Systems. Bogor (ID):
WorldAgroforestryCentre (ICRAF).
Hall, J.B., Swaine, M.D., 1976. Classication and ecology and closed canopy forest in
Ghana. Ecol. 64 (3), 913–951.
E. Grieco et al.
Journal of Environmental Management 367 (2024) 121993
11
Hansen, C.P., Lund, J.F., Treue, T., 2009. Neither fast, nor easy. The prospect of reducing
emissions from deforestation and degradation (REDD) in Ghana. Int. For. Rev. 11
(4), 439–455.
Hassink, J., Whitmore, A.P., 1997. A model of physical protection of organic matter in
soils. SSSA (Soil Sci. Soc. Am.) J. 61, 131–139.
He, Y., Qin, L., Li, Z., Liang, X., Shao, M., Tan, L., 2013. Carbon storage capacity of
monoculture and mixed-species plantations in subtropical China. For. Ecol. Manag.
295, 193–198. https://doi.org/10.1016/j.foreco.2013.01.020.
Henry, M., Besnard, A., Asante, W.A., Eshun, J., Adu-Bredu, S., Valentini, R.,
Bernoux, M., Saint-Andr´
e, L., 2010. Wood density, phytomass variations within and
among trees, and allometric equations in a tropical rainforest of Africa. For. Ecol.
Manag. 260 (8), 1375–1388. https://doi.org/10.1016/j.foreco.2010.07.040.
Houghton, J.T., Ding, Y., Griggs, D.J., Noguer, M., van der Linden, P.J., Dai, X.,
Maskell, K., Johnson, C.A., 2001. Summary for policymakers. In: Climate Change
2001: the Scientic Basis. Cambridge University Press, Cambridge, UK, pp. 2–20.
Houghton, R.A., Hackler, J.L., 2006. Emissions of carbon from land use change in sub-
Saharan Africa. J. Geophys. Res.: Biogeosciences 111 (G2). https://doi.org/10.1029/
2005JG000076.
Houssoukp`
evi, I.A., Aholoukp`
e, H.N.S., Fassinou, D.J.M., Rakotondrazafy, M.N.,
Amadji, G.L., Chapuis-Lardy, L., Chevallier, T., 2022. Biomass and soil carbon stocks
of the main land use of the Allada Plateau (Southern Benin). Carbon Manag. 13 (1),
249–265. https://doi.org/10.1080/17583004.2022.2074314.
Hulvey, K.B., Hobbs, R.J., Standish, R.J., Lindenmayer, D.B., Lach, L., Perring, M.P.,
2013. Benets of tree mixes in carbon plantings. Nat. Clim. Change 3 (10), 869–874.
https://doi.org/10.1038/nclimate1862.
IPCC, 2003. Good Practice and Guidance for Land Use, Land Use Change and Forestry.
Intergovernmental Panel on Climate Change - IPCC/OECD/IEA/IGES, Hayama,
Japan.
IUCN, 2008. Poverty in the Landscape: Capturing Variation and Learning from it. Map
Originally created by Gill Shepherd, produced by Johannes F¨
orster.
JDA, 2009. Medium Term Development Plan 2006-2009 - Jomoro District Assembly.
JDA, 2014. Draft District Medium-Term Development Plan (2014-2017) - Jomoro
District Assembly.
Kongsager, R., Napier, J., Mertz, O., 2013. The carbon sequestration potential of tree
crop plantations. Mitig. Adapt. Strategies Glob. Change 18, 1197–1213. https://doi.
org/10.1007/s11027-012-9417-z.
Kottek, M., Grieser, J., Beck, C., Rudolf, B., Rubel, F., 2006. World Map of the K¨
oppen-
Geiger climate classication updated Meteorol. Zeitschrift 15 (No3), 259–263.
https://doi.org/10.1127/0941-2948/2006/0130. June 2006.
Kumar, S.N., Bai, K.V.K., George, J., 2008. A method for non-destructive estimation of
dry weight of coconut stem. https://www.cabidigitallibrary.org/doi/full/10.5555/
20103048770.
Le Qu´
er´
e, C., Raupach, M.R., Canadell, J.G., Marland, G., Bopp, L., Ciais, P., et al., 2009.
Trends in the sources and sinks of carbon dioxide. Nat. Geosci. 2 (12), 831–836.
https://doi.org/10.1038/ngeo689.
Ledo, A., Smith, P., Zerihun, A., Whitaker, J., Vicente-Vicente, J.L., Qin, Z., et al., 2020.
Changes in soil organic carbon under perennial crops. Global Change Biol. 26 (7),
4158–4168. https://doi.org/10.1111/gcb.15120.
Lewis, S., Lopez-Gonzalez, G., Sonk´
e, B., W¨
oll, H., 2009. Increasing carbon storage in
intact African tropical forests. Nature 457, 1003–1006. https://doi.org/10.1038/
nature07771.
Losi, C.J., Siccama, T.G., Condit, R., Morales, J.E., 2003. Analysis of alternative methods
for estimating carbon stock in young tropical plantations. For. Ecol. Manag. 184
(1–3), 355–368. https://doi.org/10.1016/S0378-1127(03)00160-9.
Luo, X., Zhao, R., Chu, H., Collalti, A., Fatichi, S., Keenan, T., et al., 2024. Deciduous
forests use carbon more efciently than evergreen forests. https://doi.org/10.21203
/rs.3.rs-3989566/v1.
Masolele, R.N., Marcos, D., De Sy, V., Abu, I.O., Verbesselt, J., Reiche, J., Herold, M.,
2024. Mapping the diversity of land uses following deforestation across Africa. Sci.
Rep. 14, 1681. https://doi.org/10.1038/s41598-024-52138-9.
Mokany, K., Raison, R.J., Prokushkin, A.S., 2006. Critical analysis of root: shoot ratios in
terrestrial biomes. Global Change Biol. 12 (1), 84–96. https://doi.org/10.1111/
j.1365-2486.2005.001043.x.
Nyarko, I., Nwaogu, C., Diagi, B.E., H´
ajek, M., 2024. The dynamics and potential of
carbon stocks as an indicator of sustainable development for forest bioeconomy in
Ghana. Forests 15, 256. https://doi.org/10.3390/f15020256.
Olorunfemi, I.E., Olufayo, A.A., Fasinmirin, J.T., Komolafe, A.A., 2022. Dynamics of land
use land cover and its impact on carbon stocks in Sub-Saharan Africa: an overview.
Environ. Dev. Sustain. 24 (1), 40–76. https://doi.org/10.1007/s10668-021-01484-z.
Pendrill, F., Gardner, T.A., Meyfroidt, P., Persson, U.M., Adams, J., Azevedo, T., et al.,
2022. Disentangling the numbers behind agriculture-driven tropical deforestation.
Science 377 (6611), eabm9267. https://doi.org/10.1126/science.abm9267.
Pearson, T.R.H., Brown, S., Ravindranath, N.H., 2005. Integrating Carbon Benets
Estimates into GEF Projects, 56p.
Poeplau, C., Vos, C., Don, A., 2017. Soil organic carbon stocks are systematically
overestimated by misuse of the parameters bulk density and rock fragment content.
Soils 3 (1), 61–66. https://doi.org/10.5194/soil-3-61-2017.
Ponce-Hernandez, R., Koohafkan, P., Antoine, J., 2004. Assessing Carbon Stocks and
Modelling Win-Win Scenarios of Carbon Sequestration through Land-Use Changes,
vol. 1. Food &Agriculture Org.
Quezada, J.C., Guillaume, T., Poeplau, C., Ghazoul, J., Buttler, A., 2022. Deforestation-
free land-use change and organic matter-centered management improve the C
footprint of oil palm expansion. Global Change Biol. 28 (7), 2476–2490. https://doi.
org/10.1111/gcb.16069.
Ramankutty, N., Gibbs, H.K.W., Achard, F., Defries, R., Foley, J.A., Houghton, R.A.,
2007. Challenges to estimating carbon emissions from tropical deforestation. Global
Change Biol. 13 (1), 51–66. https://doi.org/10.1111/j.1365-2486.2006.01272.x.
Sarmiento, G., Pinillos, M., Garay, I., 2005. Biomass variability in tropical American
lowland rainforests. Ecotropicos 18 (1), 1–20.
Sierra, C.A., del Valle, J.I., Orrego, S.A., Moreno, F.H., Harmon, M.E., Zapata, M., et al.,
2007. Total carbon stocks in a tropical forest landscape of the Porce region,
Colombia. For. Ecol. Manag. 243 (2–3), 299–309. https://doi.org/10.1016/j.
foreco.2007.03.026.
Schlüter, T., 2005. Geological Atlas of Africa. Springer-Verlag, Berlin, pp. pp110–113.
Somarriba, E., Cerda, R., Orozco, L., Cifuentes, M., D´
avila, H., Espin, T., et al., 2013.
Carbon stocks and cocoa yields in agroforestry systems of Central America. Agric.
Ecosyst. Environ. 173, 46–57. https://doi.org/10.1016/j.agee.2013.04.013.
Stolbovoy, V., Montanarella, L., Filippi, N., Selvaradjou, S.K., Panagos, P., Gallego, J.,
2005. Soil Sampling Protocol to Certify the Changes in Organic Carbon Stock in
Mineral Soils of European Union - EUR 21576 EN. Joint Research Centre.
Tan, Z., Liu, S., Tieszen, L.L., Tachie-Obeng, E., 2009. Simulated dynamics of carbon
stocks driven by changes in land use, management and climate in a tropical moist
ecosystem of Ghana. Agric. Ecosyst. Environ. 130 (3–4), 171–176. https://doi.org/
10.1016/j.agee.2009.01.004.
USDA, 2010. Soil Survey Staff. 2010. Keys to Soil Taxonomy, 11th ed. USDA- Natural
Resources Conservation Service, Washington, DC.
Van Noordwijk, M., Hairiah, K., Sitompul, S.M., 2000. Reducing uncertainties in the
assessment at national scale of C stock impacts of land use change. Contribution to
IGES/NIES Workshop on GHG Inventories for Asia-Pacic Region, 9-10 March 2000.
Shonan Village Center, Japan.
VCS, 2010. Approved VCS methodology VM0006 - version 1.0- methodology for carbon
accounting in project activities that reduce emissions from mosaic deforestation and
degradation- sectoral scope 14. Terra Global Capital LLC.
Vorster, A.G., Evangelista, P.H., Stovall, A.E., Ex, S., 2020. Variability and uncertainty in
forest biomass estimates from the tree to landscape scale: the role of allometric
equations. Carbon Bal. Manag. 15, 1–20.
Wauters, J., Coubert, C., Grallien, E., Jonard, M., Ponette, Q., 2008. Carbon stock in
rubber tree plantation in Western Ghana and Mato Grosso (Brazil). For. Ecol. Manag.
255, 2347–2361. https://doi.org/10.1016/j.foreco.2007.12.038.
Yang, X., Blagodatsky, S., Liu, F., Becksch¨
afer, P., Xu, J., Cadisch, G., 2017. Rubber tree
allometry, biomass partitioning and carbon stocks in mountainous landscapes of sub-
tropical China. For. Ecol. Manag. 404, 84–99. https://doi.org/10.1016/j.
foreco.2017.08.013.
Zahabu, E., Mugasha, A.W., Malimbwi, R.E., Katani, J.Z., 2018. 10 Allometric Biomass
and Volume Models for Coconut Trees. Allometric Tree Biomass and Volume Models
in Tanzania, vol. 93.
Zeileis, A., 2004. Econometric computing with HC and HAC covariance matrix
estimators. (October 2004 ed.). Institut für Statistik und Mathematik, WU Vienna
University of Economics and Business. https://doi.org/10.57938/56b40e45-d5a1-
4507-81cf-934332c61890. Research Report Series/Department of Statistics and
Mathematics No. 10.
E. Grieco et al.
