Carbon storage in Ghanaian cocoa ecosystems
© The Author(s). 2016
Received: 14 October 2015
Accepted: 9 May 2016
Published: 23 May 2016
The recent inclusion of the cocoa sector as an option for carbon storage necessitates the need to quantify the C stocks in cocoa systems of Ghana.
Using farmers’ fields, the carbon (C) stocks in shaded and unshaded cocoa systems selected from the Eastern (ER) and Western (WR) regions of Ghana were measured. Total ecosystem C (biomass C + soil C to 60 cm depth) ranged from 81.8 to 153.9 Mg C/ha. The bulk (~89 %) of the systems’ C stock was stored in the soils. The total C stocks were higher in the WR (137.8 ± 8.6 Mg C/ha) than ER (95.7 ± 8.6 Mg C/ha).
Based on the cocoa cultivation area of 1.45 million hectares, the cocoa sector in Ghana potentially could store 118.6–223.2 Gg C in cocoa systems with cocoa systems aged within 30 years regardless of shade management. Thus, the decision to include the cocoa sector in the national carbon accounting emissions budget of Ghana is warranted.
KeywordsCarbon stocks Cocoa ecosystem Shaded and unshaded cocoa systems
Cocoa is cultivated in the forest regions of Ghana where an estimated area of 1.45 million hectares of forest land has been displaced . A substantial volume of literature is replete with evidence that the reductions in forest cover produced net sources of carbon dioxide (CO2), the main greenhouse gas of the atmosphere [2, 3]. According to the Intergovernmental Panel on Climate Change (IPCC), global C stocks in terrestrial biomass have decreased by 25 % over the past century [3, 4]. This corresponds to an annual decline of 1.1 Gt of the global carbon stocks in forest biomass . Stern  note that deforestation alone is responsible for 18 % of the world’s greenhouse gas emissions.
Cocoa intensification for higher yields has led to a drastic reduction in shade tree density and, on many farms total elimination of the shade trees in cocoa ecosystems . Essentially, cocoa expansion in Ghana has been closely linked to deforestation [7, 8]. One option to redress deforestation and create a carbon sink is to encourage the establishment of tree-crop farming or agroforestry systems [9–11]. Cocoa agroforestry is an age-old practice in the tropics . Various recommendations have been made to farmers with regard to the number of non-cocoa trees to provide shade for cocoa during planting. However, the decision on how much shade is optimal often depends on the ecological system, social factors, biodiversity interests, ecological services and pod yields [7, 11].
With the recent inclusion of the cocoa sector in the national C emission accounting budgets of Ghana , the need to quantify the carbon sequestered in cocoa ecosystems is urgent. In addition to measuring the amounts of carbon stored in cocoa and shade tree biomass in the cocoa systems, the soil organic carbon content needs to be determined. Globally, the amount of C stored in soils is estimated to be 1.5–3 times more than in vegetation . Thus, if Ghana is to include the C sequestered in the cocoa sector in its proposal for developing a national carbon accounting strategy, as outlined in its Readiness Plan Proposal , the C quantities stored both in the vegetation and the soils of the cocoa ecosystems must be included.
This paper evaluates the C storage in cocoa ecosystems from two regions of Ghana under two shade management systems and two cocoa stand age categories. It was hypothesised that; (a) the distribution of the total C stocks in the cocoa ecosystem differs between vegetation and soils, and (b) the C stocks differ between regions and shade management. The objectives were: (1) to quantify the total carbon stocks and distribution in the cocoa ecosystem, and (2) to assess the influence of shade management and the region of cocoa production on the C stocks.
Results and discussion
Selected properties of the soils under the cocoa ecosystems
Grand mean ± standard error of selected properties of the soils in the cocoa ecosystems for region (n = 24), system (n = 24) and depth (n = 16)
Bulk density (Mg/m3)
1.5 ± 0.1
8.3 ± 0.6
51 ± 2
40 ± 1
14.7 ± 0.7
0.7 ± 0.1
1.6 ± 0.1
11.7 ± 0.6
52 ± 2
38 ± 1
14.7 ± 0.7
1.5 ± 0.1
1.6 ± 0.1
10.5 ± 0.6
53 ± 2
36 ± 1
14.8 ± 0.7
1.0 ± 0.1
1.5 ± 0.1
9.6 ± 0.6
49 ± 2
41 ± 1
14.7 ± 0.7
1.3 ± 0.1
1.1 ± 0.1
6.6 ± 0.7
53 ± 2
41 ± 1
17.9 ± 0.8
2.0 ± 0.1
1.6 ± 0.1
9.9 ± 0.7
51 ± 2
39 ± 1
12.6 ± 0.8
0.8 ± 0.1
1.9 ± 0.1
13.6 ± 0.7
50 ± 2
37 ± 1
13.7 ± 0.8
0.6 ± 0.1
Biomass C concentrations
With the exception of litter C (a proportion of which is lost through respiration as it decomposes), the other components had a narrow range of 42.0–45.6 % C, with cocoa trees having the least C and Persea americana (dominant shade species in the Western region) having the highest (Fig. 1). Although few studies on agroecosystem C stocks present direct measurements of carbon with the aid of a C-analyser [16, 17], several studies have used constant values ranging from 45 to 50 % as the proportion of C for all parts of tree biomass [18, 19]. The organic carbon levels in the shade trees in the current study are not markedly different from the constant 45 % C for forest species being used by other studies [20, 21].
Soil organic carbon concentration
The soil total organic carbon concentrations differed significantly (P < 0.05) between regions, systems and soil depths (Table 1). Soil C concentration decreased with soil depth from the surface. Similar trends with depth have been noted by Cifuentes-Jara  and Dawoe . The topsoil, 0–20 cm, contained approximately 58.8 % of the soil organic C in the 0–60 cm soil profile. This undoubtedly reflects the great mass of litter fall in cocoa ecosystems. In addition, the high C concentration in the topsoil is in accordance with the presence of 80–85 % mat of lateral roots of cocoa trees being predominantly found within the top 0–30 cm [23–25], although visible roots were excluded in sampling for the current study. The soil C concentration range of 0.6–2.0 % lies within the soil C concentration range of 0.4–2.6 %, reported by Dawoe  for 15 and 30 year old cocoa ecosystems in the Ashanti region, Ghana.
Above-ground carbon stocks in cocoa ecosystems
Mean C stocks ± standard error (Mg/ha) in cocoa trees, shade trees, stumps and litter components as influenced by region [Eastern (E), Western (W)] and system [shaded (S), unshaded (U)] and their interactions, (n = 12)
15.2 ± 1.0
10.2 ± 6.4
0.16 ± 0.02
2.3 ± 0.2
13.5 ± 1.0
16.4 ± 6.4
0.12 ± 0.02
2.4 ± 0.2
12.7 ± 1.1
13.3 ± 4.1
0.07 ± 0.02
2.6 ± 0.2
16.1 ± 1.1
0.21 ± 0.02
2.2 ± 0.2
Region * System
E * S
13.6 ± 1.5
10.2 ± 6.4
0.12 ± 0.03
2.3 ± 0.2
E * U
16.9 ± 1.5
0.19 ± 0.03
2.4 ± 0.2
W * S
11.8 ± 1.6
16.4 ± 6.4
0.01 ± 0.03
2.9 ± 0.2
W * U
15.2 ± 1.5
0.24 ± 0.03
1.9 ± 0.2
Pearson correlation coefficients (r) for linear relationships among biomass C components in cocoa ecosystems
Overall, the mean carbon storage of cocoa trees was similar to that estimated for cocoa trees in a 30 year old cocoa system in Cameroon (14.4 Mg C/ha) reported by Norgrove and Hauser . The present study estimated C stock in cocoa trees similar to those reported by Isaac et al.  as 10.3 Mg C/ha in an 8 year-old cocoa system in Ghana . Isaac et al.  estimated the C storage of a 15 year-old cocoa system in Ghana as 16.8 and 15.9 Mg C/ha for a 25 year-old system, both of which agreed with the present finding that the average carbon storage of cocoa trees ranged between 11.8 and 16.9 Mg C/ha.
Soil organic carbon stocks
Understanding the effects of land use/land cover changes on ecosystem functions is often inferred from changes in soil organic carbon. However, measurements of SOC have often been excluded in many studies on land-use change because of methodological uncertainties. Jones et al.  reported a measurement standard error of 1000 kg/ha for SOC, due largely to wide variation in the soil C estimation at deeper soil profiles. In the current study, uncertainty was reduced in the characterization of the soil C pools from the surface to 60 cm depth by measuring C stocks in different soil layers.
Mean soil organic C stocks ± standard error (Mg/ha) at 0–20, 20–40 and 40–60 cm layers as influenced by region [Eastern (E), Western (W)], and system [shaded (S), unshaded (U)], (n = 12)
40.2 ± 3.4
16.6 ± 3.6
14.3 ± 1.6
58.4 ± 3.4
33.3 ± 3.6
25.7 ± 1.6
45.4 ± 3.4
19.0 ± 3.6
18.5 ± 1.6
53.2 ± 3.4
30.9 ± 3.6
21.4 ± 1.6
Region * System
E * S
44.7 ± 4.8
18.2 ± 5.0
17.1 ± 2.2
E * U
35.7 ± 4.8
15.0 ± 5.0
11.5 ± 2.2
W * S
46.1 ± 5.1
19.9 ± 5.4
20.0 ± 2.4
W * U
70.7 ± 4.8
46.7 ± 5.1
31.3 ± 2.2
Total cocoa ecosystem carbon stocks and accumulation
Mean cocoa ecosystem carbon stocks ± standard error, distributed between the biomass and soil (0–60 cm depth) components according to region [Eastern (E), Western (W)], and system [shaded (S), unshaded (U)], (n = 12)
Biomass C (Mg/ha)
Soil C (Mg/ha)
Total C (Mg/ha)
25.2 ± 1.6
70.5 ± 5.4
95.7 ± 8.6
18.4 ± 1.6
113.0 ± 5.4
137.7 ± 8.6
25.8 ± 1.6
83.7 ± 5.5
115.5 ± 8.6
17.8 ± 1.6
99.8 ± 5.5
117.9 ± 8.6
Region * System
E * S
31.3 ± 2.2
79.3 ± 7.7
109.5 ± 12.0
E * U
19.0 ± 2.2
61.7 ± 7.7
81.8 ± 12.0
W * S
20.2 ± 2.3
88.1 ± 8.2
121.5 ± 12.0
W * U
16.7 ± 2.2
137.8 ± 7.7
153.9 ± 12.1
Total SOC pools from the topsoil to 60 cm depth varied considerably from a minimum of 61.7 ± 7.7 Mg C/ha in unshaded cocoa system in the Eastern region to a maximum C stock of 137.8 Mg/ha in unshaded system in the Western region (Table 5). Results from this study estimated higher SOC stocks than the mean SOC value of 60.4 Mg/ha in Dawoe  for 0–60 cm depth of cocoa soils in the Ashanti region, Ghana. Cumulative (0–60 cm depth) SOC indicated significant (P < 0.05) variations between regions and also between management systems (Table 5).
The total ecosystem C stock of cocoa systems was estimated as the sum of soil C within 0–60 cm depth and above-ground biomass C (trees, stump and litter C). Total ecosystem C was higher in the Western region (137.7 ± 8.6 Mg C/ha) than in the Eastern region (95.7 ± 8.6 Mg C/ha). These C estimates are very high when compared with data from Dawoe . This is attributed to the low soil C stocks (35.5–80.4 Mg C/ha) from 0–60 cm depth reported by Dawoe , that were equivalent to the estimated C stocks in the current study’s topsoil, 0–20 cm (35.7–70.7 Mg C/ha) (see Table 4). Notably, in the current study, the soils contributed between 3 and 5 times more C than the above-ground pools of the cocoa ecosystems. Given the age range (7–28 years) of farms used in the current studies, as well as the extensive cultivation of 1.45 million hectares of cocoa in Ghana , it appears that approximately 118.6–223.2 Gg C could be stored in cocoa systems with stands aged within 30 years, irrespective of the shade-management system.
With respect to the effects of cocoa systems on soil C, there appears to be a general decline of the soil C stocks as time progressed. Whereas the shaded systems indicate a slight increase, the unshaded systems showed a slight decrease in soil C (Fig. 2). The two systems have similar soil C stocks at stand age of 25 years onwards.
The primary source of soil C is from litter and so the quantity and quality of the litter inputs affect the soil C dynamics . Of the systems’ contribution to the total C in the ecosystems, the trend follows that of the soil C since the bulk of C (>80 %) is stored in the soil (Table 5). The trends indicate that total carbon in shaded and unshaded systems are the same at age 17 years, but the shaded system thereafter, increased in the total C higher than that of the unshaded system (Fig. 2).
The need to quantify the carbon stocks in cocoa systems in Ghana is necessitated by the recent inclusion of the sector as an option that could result in a net increase in terrestrial carbon stocks. Hence, this paper estimated the carbon stocks in shaded and unshaded cocoa systems at different age categories; the fields were selected from the Eastern region (E) and Western region (W) of Ghana. Total ecosystem carbon was higher in the W than E. While the biomass C stock from shaded systems was twice that in unshaded systems, the two systems did not differ significantly with respect to total ecosystem C stocks. The bulk of the C stock was in the soil. The estimated high C stocks suggest that the cocoa sector holds a large amount of carbon and should be included in the national carbon accounting emission budget of Ghana.
Physiology of the study area
The Eastern region covers a land area of 19,323 km2 representing 8.1 % of the total land area of Ghana. It is located between latitude 6° and 7°N and longitude 1°30′W and 0°30′E. The region lies within the wet semi-equatorial zone which is characterized by double-maxima rainfall in June and October. The natural vegetation of the region is humid deciduous forest. Temperatures in the region are high and range between 26 °C in August and 30 °C in March. The relative humidity which is high throughout the year varies between 70 and 80 %.
The Western region occupies a land area of 23,921 km2 which is approximately 10 % of the total land area of Ghana. The region lies in the equatorial climatic zone that is characterized by a double maxima rainfall occurring in May–July and September/October. Its vegetation is that of humid deciduous forest. The region is the wettest part of Ghana with an average rainfall of 1600 mm per annum and harbours about 24 forest reserves that account for about 40 % of the forest reserves in Ghana. The climate creates much moisture culminating in high relative humidity, ranging from 70 to 90 % in most part of the region. Temperatures range between 22 °C at nightfall and 34 °C during the day.
Thus, the two regions experience similar climate and vegetation. The major soils found in both regions are mostly well drained Ochrosols or Oxisols suitable for the production of industrial crops such as cocoa, pineapple, pawpaw cola nut and oil palm. However, the Eastern region has been producing cocoa long before cultivations started in the Western region.
Selection of farms
Eight farms, comprising four from the Duodukrom community in the Suhum district of the Eastern region, and four from the Anyinabrim community in the Sefwi-Wiawso district of Western region were selected for sampling cocoa stands on the basis of shade management (shaded, unshaded). Selected farms had cocoa stand ages of 10, 14, 25 and 28 years in the Eastern region (E) and 7, 13, 17 and 27 years in the Western region (W).
At each farm, plot sizes of 30 × 90 m were demarcated for sampling. Two 30-m transects dividing the plot into three of 30 × 30 m (~0.23 acre or 0.09 ha) sub-plots were demarcated to give three pseudo-replications of each farm. The common shade tree species identified on the cocoa farms included Terminalia ivorensis, Terminalia superba, Entandrophragma cylindricum, Entandrophragma angolense, Newbouldia laevis, Persea americana, Celtis mildbraedii, Cola nitida, Carica papaya, Palmae sp., Spondia smombin, Ficus exasperate, Citrus sinensis (L.) Osbeck, Acacia mangium, and other forest tree species. Avocado (Persea americana) was the dominant shade tree in cocoa farms found in the Western region whilst Newbouldia laevis was the dominant shade tree in the Eastern region’s cocoa farms.
Soil moisture and bulk density
Soil samples at 0–20, 20–40 and 40–60 cm depths were taken from a total of 16 plots comprising 2 micro-plots of (50 × 50 cm) that were established at random within the eight cocoa farms. Two core soil samples per depth were taken randomly at each micro plot using an auger after removing visible litter from the soil surface. Soil bulk density gives an indication of the level of soil compaction . Soil bulk density and moisture contents at each sampling depth were determined on the undisturbed core samples, as outlined in Blake and Hartge .
Another set of soil samples from the same micro-plots was air-dried for 72 h, and ground to pass through a 2-mm mesh sieve to yield the fine earth fraction for chemical analysis. The soil particle size distribution was determined by laser granulometry, using a Coulter LS230 particle size analyser connected to a Windows-based computer [38–40].
Weights between 0.9–1.1, and 8.0–12.0 mg were taken respectively, from plant and soil samples the determination of C concentrations. The organic C concentration in the samples was determined using the Europa Roboprep connected to a VG 622 Mass Spectrophotometer.
There are various C pools, or compartments, within cocoa ecosystems. These include the soil C pool, the litter C pool and the woody biomass C pool in trees. The quantity of C stored in each pool is reported as the C stock, and the sum of the C stocks from the different pools constitutes the total ecosystems C stocks. On each farm, the total biomass-C stock was estimated as the sum of the C stocks in cocoa tree components (root, stem, branch, and leaf litter), floor litter and shade trees (if any) as expressed in Eq. (2). The cocoa tree component-C stocks were calculated as the product of the mean C concentration and the biomass per hectare . The mean C concentration of leaves of shade trees was used as the average C concentration of the whole shade tree in estimating the C stock of the shade trees.
where SOC denotes soil organic carbon stock (Mg/ha); ρ = soil bulk density (g/cm3); i = 0–20, 20–40, and 40–60 cm sampling depth; d = depth over which the sample was taken (cm); and %C = soil carbon concentration (%). The total cocoa ecosystem carbon stock for each farm/system was then estimated as the sum of Eqs. (2) and (3).
The data were tested for normality using q–q plot with Anderson–Darling P values in MINITAB v16. Where the tested component C was found to be non-normal, the appropriate transformation was determined with the help of Box-Cox transformation and optimal or rounded lambda that suggested one of the following transformational method as appropriate: square root, reciprocal square root, natural logarithm or inverse transformation method, according to the skewness of the data . Specifically, litter, stumps and total ecosystem C data were normal (P > 0.05) without transformation; biomass and soil C were inversely transformed; cocoa tree C was transformed with square root and shade tree C was normalized using natural logarithms. The transformed data were analysed by the Linear MIXED Model of IBM SPSS statistics 20th edition to determine significant differences between Eastern and Western regions and between shaded and unshaded systems as well as the interactions on carbon stocks controlling for the ages (covariate) of cocoa farms. The means were then estimated by restricted maximum likelihood (REML) and back-transformed to maintain the original form of the measurement. Correlation analyses by Pearson’s rank matrix were also carried out to determine any relationships among some of the ecosystem variables.
All authors (AMM, JSR, DM and AV) took part in the development of the main idea and the writing of the manuscript. The field data and laboratory analysis were done by AMM who also performed the majority of the statistical analysis. The data and analytical results were discussed among all four authors. All authors read and approved the final manuscript.
This paper is an output from a Ph.D. study in the University of Reading, UK, with funds from the Ghana Education Trust Fund, and the authors are grateful for the permission granted to publish this work. The support by cocoa farmers in Ghana to use their fields for data collection is also gratefully acknowledged.
The authors declare that they have no competing interests.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
- Anim-Kwapong G, Frimpong E. Vulnerability of agriculture to climate change-impact of climate change on cocoa production. Final Report submitted to the Netherlands climate change studies assistance programme: NCAP; 2005. 2.Google Scholar
- Stern NH. Stern review: the economics of climate change, vol. 30. London: HM treasury London; 2006.Google Scholar
- IPCC. Climate change 2007 the physical science basis. Agenda. 2007;6:07.Google Scholar
- IPCC. Climate change: the scientific basis. Cambridge: Cambridge University Press; 2001.Google Scholar
- Marland G, Boden TA, Andres RJ, Brenkert A, Johnston C. Global, regional, and national fossil fuel CO2 emissions. Trends: a compendium of data on global change; 2003. p. 34–43.Google Scholar
- Seeberg-Elverfeldt C, Schwarze S, Zeller M. Payments for environmental services: incentives through carbon sequestration compensation for cocoa-based agroforestry systems in Central Sulawesi. Grauer; 2008.Google Scholar
- Wade ASI, Asase A, Hadley P, Mason J, Ofori-Frimpong K, Preece D, Spring N, Norris K. Management strategies for maximizing carbon storage and tree species diversity in cocoa-growing landscapes. Agric Ecosyst Environ. 2010;138(3–4):324–34.View ArticleGoogle Scholar
- Gockowski J, Sonwa D. Cocoa intensification scenarios and their predicted impact on CO2 emissions, biodiversity conservation, and rural livelihoods in the Guinea Rain Forest of West Africa. Environ Manage. 2011;48(2):307–21.View ArticleGoogle Scholar
- Dixon R. Agroforestry systems: sources of sinks of greenhouse gases? Agrofor Syst. 1995;31(2):99–116.View ArticleGoogle Scholar
- UNFCCC. Calculation of the number of sample plots for measurements with A/R CDM project activities version 02; 2009.Google Scholar
- Oke D, Olatiilu A. Carbon storage in agroecosystems: a case study of the cocoa based agroforestry in Ogbese forest reserve, Ekiti State,Nigeria. J Environ Prot. 2011;2(8):1069–75.View ArticleGoogle Scholar
- Oke D, Odebiyi K. Traditional cocoa-based agroforestry and forest species conservation in Ondo State, Nigeria. Agric Ecosyst Environ. 2007;122(3):305–11.View ArticleGoogle Scholar
- Chagas T, O’Sullivan R, Bracer C, Streck C. Consolidating national REDD + accounting and subnational activities in Ghana. UNDP, Global Environment Facility, NORAD, and Gordon and Betty Moore Foundation; 2010. p. 36.Google Scholar
- Smith JE, Heath LS. Identifying influences on model uncertainty: an application using a forest carbon budget model. Environ Manage. 2001;27(2):253–67.View ArticleGoogle Scholar
- Anglaaere LCN. Improving the sustainability of cocoa farms in Ghana through utilization of native forest trees in agroforestry systems, Ph.D. Thesis, University of Wales, Bangor; 2005. p. 340.Google Scholar
- Kraenzel M, Castillo A, Moore T, Potvin C. Carbon storage of harvest-age teak (Tectona grandis) plantations, Panama. For Ecol Manage. 2003;173(1–3):213–25.View ArticleGoogle Scholar
- Losi CJ, Siccama TG, Condit R, Morales JE. Analysis of alternative methods for estimating carbon stock in young tropical plantations. For Ecol Manage. 2003;184(1–3):355–68.View ArticleGoogle Scholar
- Snowdon P, Raison J, Keith H, Ritson P, Grierson P, Adams M, Montagu K, Bi H-Q, Burrows W, Eamus D. Protocol for sampling tree and stand biomass. Canberra: Australian Greenhouse Office; 2002. p. 76.Google Scholar
- IPCC. Good practice guidelines for National Greenhouse gas inventories. Hayama: Institute for Global Environmental Strategies (IGES); 2006.Google Scholar
- Noordwijk MV, Rahayu S, Hairiah K, Wulan Y, Farida A, Verbist B. Carbon stock assessment for a forest-to-coffee conversion landscape in Sumber-Jaya (Lampung, Indonesia): from allometric equations to land use change analysis; 2002. p. 1–12.Google Scholar
- Smiley G, Kroschel J. Temporal change in carbon stocks of cocoa–gliricidia agroforests in Central Sulawesi, Indonesia. Agrofor Syst. 2008;73(3):219–31.View ArticleGoogle Scholar
- Cifuentes-Jara M. Aboveground biomass and ecosystem carbon pools in tropical secondary forests growing in six life zones of Costa Rica, Ph.D. Thesis, Oregon State University; 2008. p. 195.Google Scholar
- Dawoe E. Conversion of natural forest to cocoa agroforest in lowland humid Ghana: impact on plant biomass production, organic carbon and nutrient dynamics, Ph.D. Thesis, Kwame Nkrumah University of Science and Technology; 2009. p. 279.Google Scholar
- Wood GAR, Lass R. Cocoa. Hoboken: Wiley-Blackwell; 2008.Google Scholar
- de Oliveira Leite J, Valle RR. Nutrient cycling in the cacao ecosystem: rain and throughfall as nutrient sources for the soil and the cacao tree. Agric Ecosyst Environ. 1990;32(1):143–54.View ArticleGoogle Scholar
- Isaac M, Gordon A, Thevathasan N, Oppong S, Quashie-Sam J. Temporal changes in soil carbon and nitrogen in west African multistrata agroforestry systems: a chronosequence of pools and fluxes. Agrofor Syst. 2005;65(1):23–31.View ArticleGoogle Scholar
- Norgrove L, Hauser S. Carbon stocks in shaded Theobroma cacao farms and adjacent secondary forests of similar age in Cameroon. Trop Ecol. 2013;54(1):15–22.Google Scholar
- Kürsten E, Burschel P. CO2-mitigation by agroforestry. Water Air Soil Pollut. 1993;70(1):533–44.View ArticleGoogle Scholar
- Polzot CL. Carbon storage in coffee agroecosystems of southern Costa Rica: Potential applications for the clean development mechanism, M.Sc. Thesis, York University, Toronto; 2004. p. 162.Google Scholar
- Isaac ME, Timmer VR, Quashie-Sam SJ. Shade tree effects in an 8-year-old cocoa agroforestry system: biomass and nutrient diagnosis of Theobroma cacao by vector analysis. Nutr Cycl Agroecosyst. 2007;78(2):155–65.View ArticleGoogle Scholar
- Jones J, Graham W, Wallach D, Bostick W, Koo J. Estimating soil carbon levels using an ensemble Kalman filter. Trans Am Soc Agric Eng. 2004;47(1):331–42.View ArticleGoogle Scholar
- Post WM, Kwon KC. Soil carbon sequestration and land-use change: processes and potential. Glob Change Biol. 2000;6(3):317–27.View ArticleGoogle Scholar
- Taylor CJ. The vegetation zones of the Gold Coast. Forestry Department Bulletin 4, Accra. 1952; pp. 48–51.Google Scholar
- FAO. Estimating biomass and biomass change of tropical forests: a primer, in FAO Forestry Paper 134. Rome: FAO; 1997. p. 55.Google Scholar
- UNFCCC. Monitoring methodologies for selected small-scale afforestation and reforestation project activities under the Clean Development Mechanism. Bonn; 2006.Google Scholar
- Hillel D. Environmental soil physics: fundamentals, applications, and environmental considerations. New York: Academic Press; 1998. p. 19–201.Google Scholar
- Blake G, Hartage K. Bulk density. Methods of soil analysis: part 1—physical and mineralogical methods, 1986(methodsofsoilan1). p. 363–75.Google Scholar
- Buurman P, Pape T, Muggler C. Laser grain-size determination in soil genetic studies 1. Practical problems. Soil Sci. 1997;162(3):211–8.View ArticleGoogle Scholar
- Muggler C, Pape T, Buurman P. Laser grain-size determination in soil genetic studies 2. Clay content, clay formation, and aggregation in some brazilian oxisols. Soil Sci. 1997;162(3):219–28.View ArticleGoogle Scholar
- Arriaga FJ, Lowery B, Mays MD. A fast method for determining soil particle size distribution using a laser instrument. Soil Sci. 2006;171(9):663–74.View ArticleGoogle Scholar
- Subedi BP, Pandey SS, Pandey A, Rana EB, Bhattarai S, Banskota TR, Charmakar S, Tamrakar R. Forest carbon stock measurement: guidelines for measuring carbon stocks in community-managed forests; 2010. p. 69.Google Scholar
- Hamilton LC. Modern data analysis: a first course in applied statistics. California: Brooks/Cole Pacific Grove; 1990.Google Scholar