International Journal of Applied Earth Observations and Geoinformation 112 (2022) 102938 Contents lists available at ScienceDirect International Journal of Applied Earth Observations and Geoinformation journal homepage: www.elsevier.com/locate/jag Livelihood, carbon and spatiotemporal land-use land-cover change in the Yenku forest reserve of Ghana, 2000–2020 Joseph Ofori Acheampong a, Emmanuel Morgan Attua b, Michael Mensah c, Benedicta Y. Fosu-Mensah d, Roland Akuka Apambilla d, Eric Kofi Doe b,* a Department of Public Administration, University of Ghana Business School (UGBS), Legon, Accra, Ghana b Department of Geography and Resource Development (DGRD), University of Ghana, Legon, Accra, Ghana c Department of Business Administration, University of Professional Studies (UPSA), Accra, Ghana d Institute for Environment and Sanitation Studies (IESS), University of Ghana, Legon, Accra, Ghana A R T I C L E I N F O A B S T R A C T Keywords: Tropical forests are important sources of securing basic human needs (livelihoods) for both the deprived and Deforestation well-endowed but are also critical for reducing metric tonnes of carbon (tC) emitted from deforestation and land Forest fringe community degradation. However, inequalities of human population and land-use land-cover change (LULCC) are existential Google earth engine remote sensing threats to sustainable tropical forest reserve management and their aboveground biomass carbon stock (AGBCS) Perception Carbon neutrality in Africa. This study examines the extent of LULCC, AGBCS and perception of livelihood effects on the Yenku Land degradation neutrality Forest Reserve (YFR) in the Central Region of Ghana. Google Earth Engine remotely sensed Landsat data analysis using supervised classification, change detection, mixed with qualitative data from individual in-depth in- terviews and focus group discussions with inhabitants were used. The overall classification accuracy was 89.1%, 90.8% and 89.8% for the LULC in 2000, 2010 and 2020 respectively. Farming, charcoal production, hunting and harvesting non-timber forest products were the main livelihood activities impacting LULCC and AGBCS in the reserve. Open degraded forest was estimated at 1627ha, 1764ha, 1784ha out of 2293ha, corresponding to 36,349.6tC, 39,395.70tC, 39,840.0tC respectively in 2000, 2010 and 2020. Dense degraded forest cover yielded the least carbon stock of 938.6tC compared to 39,840.0tC from less dense degraded forest cover. These findings would aid policy decisions toward achieving United Nations land degradation neutrality and sustainable development goals (SDGs) one, ten and fifteen while ensuring YFR sustainability. Deprived forest-fringe com- munities, traditional authorities and other relevant stakeholders need to actively adopt gendered livelihood objectives to achieve SDGs, carbon and land degradation neutralities within YFR. 1. Introduction lands (Hou Jones and Franks 2015). The anthropogenic land use land cover change (LULCC) constitutes 12.5% of the forest CO2e (Friedling- Carbon neutrality and deforestation are major challenges of climate stein et al. 2010; Houghton et al. 2012). Exploiting forest resources to change mitigation and the United Nations (UN) Sustainable Develop- secure livelihood objectives is well-known for causing deforestation and ment Goals (SDGs). About 1.2% of the world’s tropical forest lands land degradation leading to land cover change (Baccini et al. 2012; representing 21 million hectares, are destroyed annually through human Sobeng et al. 2018). Legal/illegal logging and agricultural activities livelihood development activities (FAO 2016). These anthropogenic caused 15.6 million hectares (ha) of Africa’s forest land cover loss be- activities release sequestered carbon (C) from the forest, which increases tween 2010 and 2015 (Hou Jones and Franks 2015; FAO 2019). atmospheric carbon (CO2), thereby exacerbating carbon emissions Forest lands support more than 1.6 billion people’s livelihoods (CO2e), global warming and climate change (CC) (Baccini et al. 2012; (World Bank 2004; Sobeng et al. 2018). Some protected forest reserves Rittenhouse and Rissman 2012). Rittenhouse and Rissman (2012) appear crucial to the livelihoods of forest-fringe community dwellers, contend clearing forests correlates negatively with forest carbon stocks (Langat et al., 2016; Sobeng et al., 2018). While close to 90% of these and positively with CO2e. Africa accounts for 17% of the global forest forests support terrestrial biodiversity, about 90% of 1.2 billion * Corresponding author. E-mail address: ekdoe@st.ug.edu.gh (E. Kofi Doe). https://doi.org/10.1016/j.jag.2022.102938 Received 14 March 2022; Received in revised form 16 July 2022; Accepted 24 July 2022 Available online 3 August 2022 1569-8432/© 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). J. Ofori Acheampong et al. I n t e r n a t i o n a l J o u r n a l o f A p p l i e d E a r t h O b s e r v a t i o n a n d G e o i n f o r m a t i o n 112 (2022) 102938 residents in forest-fringe communities, live in extreme poverty less than the forest resources, residents continue to exploit the resources due to one US$ per day (World Bank 2004). Therefore, sustainable manage- limited alternative livelihoods (Amoah and Wiafe 2012; Cobbinah et al. ment of forest reserves and livelihood activities in forest-fringe com- 2015a). Although fringe community dwellers appreciate the need for munities is essential in averting indiscriminate perturbation and conserving forest reserves, they cannot protect the forest when socio- deforestation CO2e. Notably, for achieving the UN SDGs 1 (reduce economically constrained and have to rely on Non-timber Forest Prod- poverty), 10 (gender equality) and 15 (life on land) as well as the carbon ucts (NTFPs) for survival (Amoah and Wiafe 2012; Cobbinah et al. neutrality and land degradation neutrality (LDN) declarations toward 2015a; Shackleton et al. 2015). Indigenes may be aware of illegal forest climate change mitigation (Lal 2019). reserve activities but less concerned about the danger of such activities In Ghana, one of the forest reserves in the Southern savanna dry on YFR CO2e and sustainability. forest zone under anthropogenic threat is the Yenku Forest Reserve Reducing Emissions from Deforestation and Degradation (REDD+) (YFR). Dottey (2005), suggests human activities in the reserve have necessitates local communities’ participation (Forestry Commission depleted essential forest resources and its sustainability is at risk with 2015; Appiah et al. 2016). REDD+ is a financing mechanism of the increasing population and urbanization. Some of the livelihood activ- United Nations Framework Convention on Climate Change UNFCCC) ities undertaken in the YFR are fetching fuelwood, illegal hunting, which is also linked to SDGs 13 and 15. The concept illustrates how unstainable agricultural practices and charcoal production. forested African countries and donors traditionally engaged in forestry Managing the effects of livelihood activities on the YFR has become a and agroforestry development finance by requiring proponents and Sisyphean task. Despite several restrictions governing the utilization of stakeholders to demonstrate emissions reduction impacts for payments. Fig. 1. Study area map showing the YFR in the Central Region (insert) of Ghana. Source: Authors’ construct based on extracts from forest reserves and towns in Ghana. 2 J. Ofori Acheampong et al. I n t e r n a t i o n a l J o u r n a l o f A p p l i e d E a r t h O b s e r v a t i o n a n d G e o i n f o r m a t i o n 112 (2022) 102938 Thus, it aims at reducing greenhouse gas (GHG) emissions from defor- Perception is “how something is regarded, understood and inter- estation and forest degradation (Forestry Commission 2015; Appiah preted by people”. The perception of fringe community dwellers is used et al. 2016). Unlike REDD, REDD+ includes conservation, sustainable to examine the effects of their livelihood activities on the YFR LULCC forest management (SFM) and forest carbon stocks enrichment. Ghana and the sustainability of the reserve. Sustainability is the use or man- started to engage in REDD+ in 2008, as a multi-sectoral program pro- agement of resources to meet current needs without compromising the moting SFM for improved livelihoods and opportunities to increase ability of future generations. forest cover (Forestry Commission 2015). The sustainable livelihood framework (SLF) developed by the Insti- However, there is a little knowledge of indigenes’ perception of the tute of Sustainable Development (IDS) was adopted for this study impact of their livelihood activities on the sustainability of the YFR. because it captures the diverse components of sustainable rural liveli- Furthermore, little is known about LULCC from 2000 to 2020 in the YFR. hoods (Scoones 1998, Natarajan et al., 2022). The SLF shows how in The present research, therefore, sought to examine the perception of the different contexts, household livelihoods are achieved through access to inhabitants of YFR fringe communities concerning threats posed by their a range of resources such as natural, economic, human and social capital activities on the reserve and carbon stock. Specifically, the study in pursuit of different livelihood strategies and activities (Scoones assessed the (1) magnitude, trends and rate of LULCC in the YFR; (2) 1998). The use of these resources is directly influenced by a structural effect of LULCC on carbon stock; (3) socio-economic characterization of process such as the laws and policies governing the management and use the endowment of the inhabitants; (4) reasons for their continued live- of land resources. SLF is defined by Scoones (1998) as human activities lihood activities in the reserve and (5) roles in YFR management in based on their assets, capabilities and entitlement required to achieve a Ghana. basic standard of living; including their ability to cope with and recover from external stresses (policies) and shocks (climate, disaster) to their 2. Methods livelihood activities, without undermining the natural resource base and structures (Chambers and Conway 1992). The theory allows for a 2.1. Study area diverse aspect of livelihood activities to be inculcated into the exami- nation and understanding of rural life and development out of poverty. Fig. 1 shows the study area. The YFR is located within longi- tudes’0045′W and’0041′W and latitudes’5019′N and’5028′N. The area of 2.3. Study design the reserve is about 29.31km2. It falls within the Effutu Municipality, Gomoa East and Gomoa West District Assemblies. It is managed by The study used exploratory mixed methods involving quantitative Winneba Forestry Service Division (FSD) of the Forestry Commission of and qualitative data. Primary quantitative data, enabled socioeconomic Ghana as part of Forest Management Unit 29 under the SFM systems of description of the participants (N = 90, Female = 45, Male = 45). These Ghana. It was Gazetted as a forest reserve on 6th April 1940 under participants came from nine (9) YFR fringe communities, namely Forests Ordinance Cap 122 (WFD 2004; Dottey 2005). Onyadze, Ostew-Jukwa, Mankoadze, Bewadze, Aseibu, Oguakrom, The YFR has two primary vegetation types, the coastal savannah and Gomoa Dahom, Amenfi and Gomoa Lome. As summarized in Fig. 2, pre- semi-deciduous forest. The former consists mainly of grassland and short processed secondary quantitative data from remotely sensed Landsat trees, patches of shrubs, while the latter is characterized by tall trees data was applied for the spatiotemporal classification of land use-land interspersed with grasses, shrubs, and soft woody species. There are two cover (LULC) change and magnitude of carbon emission. These sec- rainfall seasons, the major (April-July) and minor (September- ondary data include Enhanced Thematic Mapper Plus (ETM+) Landsat 7 November) rainy seasons (WFD 2004; Dottey 2005). The mean annual of 2000 and 2010 and Operational Land Imager (OLI) Landsat 8 of 2020 rainfall ranges between 70cm and 90cm in the southern coastal belt and at 30 m × 30 m (pixel) resolution from Google Earth Engine (GEE). The 90cm to110cm in the northern and north-western semi-deciduous forest imageries were acquired in February of each year, screened for less than areas. The annual maximum and minimum temperatures of 29◦C and 10–12 % cloud cover and haze. 26◦C occur in February-March and August, respectively. The forest The LULC classification was achieved using an automated supervised livelihood activities are agroforestry (taungya), firewood cutting, char- random forest classification algorithm, using GEE JavaScripts (Breiman coal production, hunting and farming (WFD 2004; Dottey 2005). 2001; Liaw and Wiener 2002; Gorelick et al. 2017). The GEE JavaScripts are available in supplementary material 1 in the online version of this 2.2. Theoretical framework: Land use and livelihood sustainability paper. GEE offers widely accepted reproducible and effective means of large-scale spatial data analysis in the 21st century (Banerjee et al. This study is underpinned by land use and livelihood sustainability 2021). Surface reflectance (level 2) of the LandSat image pixels (Chen theories. In most rural settings, land is the primary means of livelihood et al. 2021) depicting five classes of the land cover types, aided quan- (Scoones 1998). These theories elicit biophysical and perceived un- tification of each LULC type in hectares, using the random forest clas- derstandings of people’s livelihood dynamics and effects on LULCC. The sifier. These LULC types are close (dense) degraded forest (close-df) and sustainable livelihood theory explains how people use their assets, ca- open (less dense) degraded forest (open-df). The rest are close (dense) pabilities (entitlements) to develop strategies for their survival amidst grasses/herbaceous vegetation (close-ghv), open (less dense) grasses/ state policies, institutional framework and environmental occurrences herbaceous vegetation (open-ghv) and bare land surface. (Chambers and Conway 1992). Three Landsat datasets representing February of 2020, 2010 and 200 Land use describes the management or activity to which people put a were used. Unlike temperature data which varies rapidly over short land cover (Doe et al. 2018) or the management within each land cover periods, rigidity of land cover data to short-term change allows LULCC type (class) (Houghton et al. 2012). LULCC is the conversion from one trend analysis to use a few data points over long intervals. In China for land cover type to another. It is a biophysical process that alters forest example, Wu et al. (2021) classified urban LULCC in Wuhan using three land resources and ecosystem services, such as the magnitude of vege- Landsat data sets. Gao and Liu (2010) used only two Landsat images to tative area and its carbon emission or sequestration potentials. Forested detect trends of LULCC at 10-year intervals caused by soil salinization land size is directly proportional to the total metric tonnes of carbon and waterlogging. emission (tCe) or otherwise sequestration. Nature and people define the As demonstrated by Olofsson et al. (2014), Crowson et al. (2019) and rate of LULCC and its consequential effect on carbon sequestration or Wu et al. (2021), training datasets (351 sites) for the supervised classi- emissions, hydrological cycles, waste and pollution abatement, biodi- fications were validated using key informants and visual assessment of versity, soil conservation and climate change (Verburg et al. 2019; Lal unchanged pixel reflectance of the respective LULC types. In addition, 2020). the key informants confirmed the training data as semi-permanent like 3 J. Ofori Acheampong et al. I n t e r n a t i o n a l J o u r n a l o f A p p l i e d E a r t h O b s e r v a t i o n a n d G e o i n f o r m a t i o n 112 (2022) 102938 Fig. 2. Workflow of YFR LULC spatial and temporal changes and carbon emission. rocky hills, sacred groves and bare lands that hardly changed during the study period, making them suitable for training the 2000, 2010 and LULCC *EFRate of change in TCS tC class class( /year) = i i *100 (5) 2020 LULC classifications. The use of these semi-permanent training time interval data ensured parity in the classification process to permit comparisons of The primary qualitative data was necessary for exploring the expe- the classified images. For cross-validation purposes, ground-truthing of riences and perceptions of participants concerning their livelihood ac- the sites was done where possible. tivities and implications in the management of the YFR. This was The classification accuracy of the 351 training data from different achieved using in-depth interviews (IDIs) and Focus Group Discussions LULC types was examined using a confusion matrix following Olofsson (FGDs), following Ardayfio-Schandorf et al., (2007) and Kamwi et al. et al., (2014). Please, refer to the confusion matrix results and training (2015). These authors used sustainable livelihood framework (SLF) and data available in the supplementary file online. For each LULC type, interview guides to assess socioeconomic drivers of LULCC in forest 40% of the training data were used as test samples and 60% as training regions. The focus was on commons such as farmers, firewood cutters samples to validate the accuracy of the classification. The accuracy of (FC), hunters and charcoal producers (CP) from nine YFR fringe com- the classified LULC was tested using the confusion matrix to compare the munities. Purposive sampling was employed to select these participants predicted LULC types and the actual samples. The accuracy was also based on their in-depth knowledge and lived experiences. Selection of examined using expert knowledge and visual assessment. these participants is essential for narrowing the inequality (SDG-10) Post-classification change detection was also conducted (Hassan between less-endowed and well-endowed (SDG-1) commons in the area et al. 2021). The magnitude and rate of change of each LULC type while enhancing the involvement in sustainable management of the YFR (LULCCclass i) were estimated following Doe et al., (2018) as expressed in lands (SDG-15) through carbon and degradation neutralities. The sam- equation (Eq.) one (1): ples consisted of eight farmers and two hunters, making ten people drawn from each of the nine communities. Four of the farmers were LULC LULCC ha class icurrentyear − LULCclass ipastyear class i( ) = (1) firewood cutters and charcoal producers. LULCclass ipastyear The interview guides cover questions like the type of livelihood ac- LULC LULC tivities undertaken at the reserve, the implication of these human ac- LULCC % class icurrentyear − class ipastyear class i( ) = *100 (2) tivities on the reserve and the level of knowledge about climate change. LULCclass ipastyear They also include questions on community perception of sustainable The computation was repeated for years 2000 to 2010, 2010 to 2020 management of YFR. Informed consent was sought before the in- and 2000 to 2020. Dividing each estimate by the observed time interval terviews. We ensured participant anonymity and confidentiality of in- yields the annual rate of LULCC per land cover type (Doe et al. 2018). formation provided using “farmer”, “hunter”, “firewood cutters In this study, a local carbon emission factor (EF) for the different (fetching)” and “charcoal producers” for purposes of anonymity. Field categories of savanna forest vegetation (Forestry Commission 2017) notes of observations were also taken. The IDIs and FDGs were recorded enabled estimates of the vegetative carbon stock for each land cover type using a digital audio-type recorder. The recordings were later tran- (Houghton et al. 2012; Goslee et al. 2014). The EF was obtained from the scribed for thematic content analysis. Thematic analysis based on the Ghana Forestry Commission provided in the refined 2006 Inter- steps proposed by Braun and Clarke (2006) was followed to enable governmental Panel on Climate Change (IPCC) guidelines of 2019. systematic coding and generation of the themes. The EF covers carbon stored in live and dead biomass such as trees, wood and litter, excluding soil carbon (Forestry Commission 2017). The 3. Results and discussion default EF for LULCC of class “i” (LULCCclass i) such as the close savanna forest LULC type is 18.79tC ha− 1 (EFclass i) per year (Forestry Commis- 3.1. The magnitude and rate of land use land cover change within the YFR sion 2017). It is 22.34tC ha− 1 for open savanna forest LULC type, 15.82tC ha− 1 for grassland LULC type and 0.0tC ha− 1 for bare land LULC Fig. 3 presents the spatiotemporal classification of the YFR LULC type. Based on these EF values, the total carbon stock (TCS) for each types as of 2000, 2010 and 2020. The confusion matrix estimated an LULC type was estimated (in tC) using equations (3)–(5). overall classification accuracy of 89.1%, 90.8% and 89.8% respectively. The estimated producer accuracy (PA) and user accuracy (UA) are in the TCS of LULCclassi = Hectarage of LULCclassi*EFclassi (3) supplementary file. Fig. 3 shows evidence that the common LULC types are the close-df (49.95ha), open-df (1783.35ha), close-ghv (287.37ha), Change in TCS of LULCclassi = LULCCclassi*EFclassi (4) open-ghv (120.60ha) and Bare surface (51.66ha). As of the year 2020, the open-df dominates (77.8% of 2293.00 ha). This is followed by the 4 J. Ofori Acheampong et al. I n t e r n a t i o n a l J o u r n a l o f A p p l i e d E a r t h O b s e r v a t i o n a n d G e o i n f o r m a t i o n 112 (2022) 102938 Fig. 3. Spatiotemporal land use land cover (ha) within YFR in 2000 (A), 2010 (B) and 2020 (C). Source: Authors’ construct based on Landsat data. close-ghv (12.5%) and open-ghv types (5.3%). The bare land surface 3.2. Change in aboveground biomass carbon stocks of the YFR (2.3%) and the close-df were the least in area coverage as of 2020. Concerning the trends (please, see in supplementary Fig. 1 or The YFR carbon stock (tC) appears to have increased at the expense Table A2.2) of the LULCC, the open-df which depicts open savanna of dense forest biomass aboveground. The total amount of carbon woodland vegetation increased in size (land area) steadily from 71.0% decreased from 46,196.8tC in 2000 to 46,052.6tC in 2010 and then (1627.11ha) in 2000 to 77.8% (1783.35ha) in 2020. According to the increased to 47,233.7tC in 2020 (Fig. 4). A majority of the increase came interviewees, new plantations were established between 2002 and 2009. from expansion in the open-df instead of the close-df (Fig. 4). The close- Concurrently, the size of close-df depicting close savanna woodland df decreased in carbon stock alongside decreases in the close-ghv and vegetation decreased from 2.5% (56.79ha) to 2.2% (49.95ha). open-ghv types. This implies the close-df carbon stock of 1,067.1tC in The two grass/herbaceous vegetation (ghv) types dwindled in size. 2000 and 938.6tC in 2020 was traded for open-df carbon which The close-ghv reduced from 18.5% (423.90ha) in 2000 to 12.5% increased from 36,349.6tC to 39,840.0tC. Suggesting that the YFR is (287.37ha) in 2020 while the open-ghv dropped from 5.7% (131.04 ha) losing aboveground biomass carbon stock (128.5tC) from its close-df at to 5.3 % (120.60ha), respectively. Although, the bare land surface an annual rate of 6.5%. The close-ghv also reduced in carbon from showed an increase from 2.4% (54.09ha) to 5.2% (120.15ha) between 6,706.1tC to 4,546.2tC at the rate 1.6% per year. The decreasing pattern 2000 and 2010 and it decreased to 2.3% (51.66ha) by 2020. of endemic vegetation was reported by Massetti and Gil (2020) at 2.0% These estimates portray that the YFR is losing relatively high vege- elsewhere. tation quality (dense cover) for low-quality vegetation cover which as- Although the rate of forest cover loss is lower than the national rate sociates with biodiversity loss and reduced carbon sequestration of 2 % (Forestry Commission 2015; FAO 2016), the forest vegetation capacity. These forest and land quality degradation patterns are cover loss in the YFR corresponds to overexploitation of resources in the consistent with Rittenhouse and Rissman (2012) and Dymond et al. reserve. The concept of sustainability recommends that the rate of (2012). The observations are buttressed by the magnitude of loss in the resource exploitation should match the regenerative capacity of the close-df by 12.0% (6.84ha) and an increase in the less degraded forest by resource to avoid overexploitation and degradation of the resource 9.6% (156.24ha) over the 30 years observed. The close-ghv type expe- quality. Due to legal or illegal harvesting, encroachment, farming and rienced the largest decline of 32.2% (136.53ha) at an annual rate of wildfires, the degradation rate outweighs the rate of restoration of the 1.1%. YFR. This finding can be likened to the report of Doe et al., (2018), who concurred that human livelihood activities and population growth 5 J. Ofori Acheampong et al. I n t e r n a t i o n a l J o u r n a l o f A p p l i e d E a r t h O b s e r v a t i o n a n d G e o i n f o r m a t i o n 112 (2022) 102938 Fig. 4. Estimates of carbon stock of vegetation cover types in the YFR, tC = EF * hectarage of LULC type. resulted in the loss of the natural forest vegetation cover within the Kakum Conservation Area in the Central Region of Ghana. Despite, the Table 1 Modified Tuangya System (MTS), and its management challenges Reasons for continued livelihood activities in the YFR. (Acheampong et al. 2016), the failure of the plantations could be Community Reasons for livelihood activities in the YFR attributed to livelihood activities that cause wildfire outbreaks and other 2 FC for “mushrooms, medicines, chewing sticks, tree trunk for preparing forms of forest land degradation like charcoal production (Global Fire pestles, bushmeat, snails, firewood, and building posts, among others” Monitoring Center (GFMC) 2007). It could also be due to climate (IDI) change, drought, inadequate rainfall, poor supervision and maintenance 3 FC “land scarcity is a major threat to our livelihoods. We are food crop farmers doing subsistence farming for our household income to survive” contrary to the recommendations proposed by Ayivor et al. (2011). (IDI) The dependence on NTFPs is because “what they want cannot be obtained outside the forest reserve” 3.3. Socioeconomic characteristics of respondents around the YFR To enhance YFR sustainability, “tree planting on farmlands (agroforestry) was initiated by the FSD to recover lost vegetation, improve An equal number of women (50.0 %) and men (50 %) participated in soil fertility and NTFPs production” (FGD) the study. Most (77.0 %) of whom were above 40 years, married (72.2 5 CP We carry tree stems, branches and stumps from the reserve to burn %) with no formal education (48.9 %) and at least primary school ed- charcoal outside the forest reserve” (IDI) F “NTFPs are harvested for personal use and sold by those who do not farm ucation (51.1 %). The respondents (participants) were mainly indigenes in the forest” (IDI) (77.8 %) and a few migrants (22.2 %) making less than US$34.5 7 FC We collect medicinal plant parts and use them to treat ailments and other (Ghc200) of income a month from their livelihood activities, covering diseases like stomach aches, broken limbs and malaria” (IDI) charcoal producers (26.7 %), farmers (26.7 %), firewood cutters (26.7 Such traditional healing methods are fading away making young ones ignorant about the most effective medicinal plants in the forest” (FGD) %) and hunters (20.0 %). This socioeconomic characterisation is 8 H “I go there for farming, hunting and to collect NTFPs consistent with Ayivor et al. (2011) but poorer than the description of 9 F “I usually visit the forest to collect medicinal plant parts, but others enter Sobeng et al. (2018). Sobeng et al., (2018) indicate an income of US$1.9 the forest to produce charcoal, hunt grass cutters and antelopes. These per day for fringe community dwellers near the Tano forest reserve in animals feed on our food crops anyway” (IDI) the Atwima Mponua District of the Ashanti Region in Ghana. This F = Farmer; FC = Fuelwood cutter; CP = Charcoal producer; H = Hunter; IDI = finding confirms that poor households depend on protected forest In-depth interview; FGD = Focus group discussion. reserve resources (Langat et al. 2016). to survive” (IDI with a FC, Community 3). Others said, “We carry tree 3.4. Reasons for continued human livelihood activities in the YFR stems, branches and stumps from the reserve to burn charcoal outside the forest reserve” (IDI with a CP, Community 5). Another explained that Table 1 reveals the fringe community dwellers’ reasons for continued “We collect medicinal plant parts and use them to treat ailments and other dependence on the YFR for their livelihoods. diseases like stomach aches, broken limbs and malaria” (IDI with FC, Contrary to perceptions that YFR is a sacred grove where all forms of Community 7). “Such traditional healing methods are fading away making human activities are strictly forbidden except for the local “Aboakyir” young ones ignorant about the most effective medicinal plants in the forest” festival, the current study corroborates Dottey, (2006) on the occurrence (FGD, community 7). of human activities in the reserve. The participants revealed various Most participants specified that NTFPs were harvested in different motives for visiting the YFR. “I go there for farming, hunting and to collect proportions and for several purposes. The “NTFPs are harvested for per- NTFPs” (IDI with Hunter, Community 8). A farmer said “I usually visit sonal use and sold by those who do not farm in the forest” (IDI with the forest to collect medicinal plant parts, but others enter the forest to pro- Farmer, Community 5). These NTFPs include “mushrooms, medicines, duce charcoal, hunt grass cutters and antelopes. These animals feed on our chewing sticks, tree trunk for preparing pestles, bushmeat (game), snails, food crops anyway” (IDI with Farmer, Community 9). Another firewood, and building posts, among others” (IDI with a FC, Community participant indicated that “land scarcity is a major threat to our livelihoods. 2). However, the daily collection in recent times has been much lower We are food crop farmers doing subsistence farming for our household income compared to the past 30 years. The dependence on NTFPs is because 6 J. Ofori Acheampong et al. I n t e r n a t i o n a l J o u r n a l o f A p p l i e d E a r t h O b s e r v a t i o n a n d G e o i n f o r m a t i o n 112 (2022) 102938 “what they want cannot be obtained outside the forest reserve”. To enhance 3.6. Perceived effects of human activities in the YFR YFR sustainability, “tree planting on farmlands (agroforestry) was initiated by the FSD to recover lost vegetation, improve soil fertility and NTFPs pro- As shown in Table 3, the participants believe that YFR has dwindled duction” (FGDs in Community 3). These findings are consistent with in its status in terms of hectarage of trees, the greenness of the close-df Daur et al., (2016) and Cobbinah et al., (2015a) that conservation ap- and soil fertility compared to the past 30 years. This finding corroborates proaches must be localized within the socio-cultural, economic and Kyere-Boateng and Marek (2021) and Doe et al., (2018) that access to political context of the people without compromising on biodiversity. forests, illegal timber harvesting and human pressure account for The findings also confirm Trædal and Vedeld, (2018) and Kyere-Boateng reserve vegetation loss in the long term. Participants in the current study and Marek, (2021) that there are complexities in using forest resources. believe that the YFR LULCC is caused by “human pressure, over- These complexities call for including livelihood objectives of forest exploitation of NTFPs and unsustainable farming activities. These have fringe community dwellers in the YFR management more effectively. resulted in deforestation and degradation of the forest resulting in inadequate rainfall, erratic rainfall and adverse temperature variability” (IDI with 3.5. Local perception about the role of YFR management Hunter, Community 6). “There is a little rain, which comes with heavy storms and wildfires due to prolonged drought” (IDI with CP, Community The results in Table 2 showed that the participants knew some of the 8). These views confirm the observations and estimates of the LULCC in mandates of the FSD concerning the YFR management. Fig. 3 as demonstrated in Appiah et al. (2016). They acknowledged that the “FSD release degraded portions of the reserve for agroforestry purposes” (FGD, Community 9). This is known as 3.7. Local knowledge about climate change and climate variability effects the Modified Taungya System (MTS). The “MTS module permits forest of LULCC fringe community farmers to plant food crops within tree plantations until the tree canopy closes, usually within three (3) years” (IDI with a Farmer, Concerning climate change (CC) in the study area, the participants Community 2) or more, which corroborates Acheampong et al. (2016). affirm a Community Education was conducted by the FSD Office at The “farmers provide labour for land clearing, tending, and protection of the Winneba. According to the participants (Table 4), “Climate Change is plantations to its rotation period” (IDI with a FC, Community 4). The about the excessive heat from sunlight, drought, variability in rainfall pat- practice is consistent with the reports of Cobbinah et al. (2015a) and terns, sea-level rise and the effect of sea waves along the Winneba coast, Mbanze et al. (2020), who demonstrate the potential of agroforestry which has resulted in corrosion of metallic roofs of buildings in the area” (IDI systems in advancing forest frontiers. with Farmer, Community 7). “Salt from the sea waves can cause crop However, YFR “MTS project expired in 2009 and the farmers now have disease and death of forest plants… a 16.0 ha teak plantation was destroyed difficulties in forest farming’ (IDI with a CP, Community 1). When the by bushfires in 2011 at the Bewadze portion of the reserve” (IDI with participants were asked about the Benefit-Sharing Agreement (BSA) of Hunter, Community 1). Other participants attributed CC to God; “The the MTS, it was confirmed that the MTS gave farmers legitimate right to “making of God” who has promised that drastic changes can occur toward enter the reserve to farm as reported by Acheampong et al. (2016). The the end of man’s existence on the earth” (IDI with FC, Community 3). “farmers, stool landowners and communities take 40 %, 15 % and 5 % of the These findings corroborate previous literature on local CC knowledge share respectively from the proceeds, while the FSD takes 40 %” (IDI with a (Derbile et al. 2016). support the global call for CC mitigation and Hunter, Community 3). The BSA enhanced forest protection and adaptation (IPCC 2019). planting of Cedrela odorata (Cedrela), Cassia spp, (Cassia), Tectona grandis (teak), Ceiba pentandra (ceiba) and Khaya spp (mahogany) in the 3.8. Perceived challenges of sustainable management of the YFR reserve. As a result, “illegal farming stopped entirely between 2002 and 2009… because it became difficult for people who were not involved in the Table 5 presents some of the perceived challenges of the YFR. The plantation to sneak into the forest illegally” (IDI with a Farmer, Com- study participants admitted they are partly responsible for the sustain- munity 9). These findings imply that in the absence of resources and able management of the YFR. favourable forest management systems, people are likely to take the law However, as earlier reported by Shackleton et al., (2015), unem- into their hands, as shown in previous literature (Cobbinah et al. 2015a; ployment and poverty that lead to desperation are forcing the people to Mbanze et al. 2020; Kyere-Boateng and Marek 2021). overexploit or degrade the forest resources reluctantly. Hence the hes- itance of the people to collaborate effectively with the FSD. For instance, Table 2 an IDI revealed that “lack of employment leads to inadequate household Local perceptions about the role of YFR management. incomes, hunger, inability to pay for education and health facilities. This situation is worsened by low crop yields from off-reserve farmlands” (IDI Community Perceptions about YFR management with Farmer/CP, Community 6). Another major hindrance is the 1 CP The “MTS project expired in 2009 and the farmers now have difficulties in “failure of farmers to maintain their MTS plantations due to high mainte- forest farming’ (IDI) nance cost, even though some operational costs such as weeding, climber 2 F The “MTS module permits forest fringe community farmers to plant food crops within tree plantations until the tree canopy closes, usually within cutting, protection and patrols were borne by the FSD” (IDI with a Farmer three (3) years” (IDI) in Community 4). Some participants reiterated that “there used to be a 3 H The “farmers, stool landowners and communities take 40 %, 15 % and 5 Community Forest Committees between 2003 and 2007. It collapsed because % of the share respectively from the proceeds, while the FSD takes 40 %” of lack of local incentives to protect the forest” (IDI with FC, Community (IDI) 4 FC “farmers provide labour for land clearing, tending, and protection of the plantations to its rotation period” (IDI) Table 3 9 F The BSA enhanced forest protection and planting of Cedrela odorata Perceived effects of human activities in the YFR. (Cedrela), Cassia spp, (Cassia), Tectona grandis (teak), Ceiba pentandra (ceiba) and Khaya spp (mahogany) in the reserve. As a result, “illegal Community Perceived effects of human activities on YFR farming stopped entirely between 2002 and 2009… because it became 6 H The YFR LULCC is caused by “human pressure, overexploitation of difficult for people who were not involved in the plantation to sneak into NTFPs and unsustainable farming activities. These have resulted in the forest illegally” (IDI) deforestation and degradation of the forest resulting in inadequate “FSD release degraded portions of the reserve for agroforestry purposes” rainfall, erratic rainfall and adverse temperature variability” (IDI). (FGD) 8 CP “There is a little rain, which comes with heavy storms and wildfires due to F = Farmer; FC = Fuelwood cutter; CP = Charcoal producers; H prolonged drought” (IDI). = Hunter;; IDI = In-depth interview; FGD = Focus group discussion. CP = Charcoal producers; H = Hunter; IDI = In-depth interview. 7 J. Ofori Acheampong et al. I n t e r n a t i o n a l J o u r n a l o f A p p l i e d E a r t h O b s e r v a t i o n a n d G e o i n f o r m a t i o n 112 (2022) 102938 Table 4 the forestry laws and regulations. YFR decision-making should engender Local knowledge about climate change and climate variability effects. active involvement of deprived fringe community dwellers’ livelihood Community Knowledge about climate change and variability effects objectives, integrating SDGs 1, 10, 15, carbon/land degradation neu- tralities and climate actions alongside REDD+. The authors believe 1 H “Salt from the sea waves can cause crop disease and death of forest plants… a 16.0 ha teak plantation was destroyed by bushfires in 2011 at active and informed participation of the less endowed (deprived) fringe the Bewadze portion of the reserve” (IDI) community dwellers and stakeholders is critical to achieving carbon 3 FC “The “making of God” who has promised that drastic changes may occur neutrality and land degradation neutrality. Vigorous public awareness, towards the end of man’s existence on the earth” (IDI). communication and education are required for effective participation in 7 F “Climate Change is about the excessive heat from sunlight, drought, variability in rainfall patterns, sea-level rise and the effect of sea waves decision-making and sustainable integration of livelihood objectives. along the Winneba coast, which has resulted in corrosion of metallic roofs There is a need for rejuvenation of persuasive enforcement, monitoring, of buildings in the area” (IDI). supervision and participatory action to mitigate forest cover and carbon F = Farmer; FC = Fuelwood cutter; H = Hunter;; IDI In-depth interview. loss. Further research of the YFR LULCC should estimate the carbon = stock using remote sensing and integrate qualitative insight from all categories of stakeholders. Table 5 The study contributes to the general body of knowledge on earth Perceived challenges of sustainable management of the YFR. observation and geospatial science, social and forest ecology toward Community Challenges of managing YFR sustainable forest management and human livelihood development. This is particularly true in the context of YFR LULCC and carbon stock. It 4 F “failure of farmers to maintain their MTS plantations due to high maintenance cost, even though some operational costs such as weeding, advances the theory and practice of sustainable forest fringe community climber cutting, protection and patrols were borne by the FSD” (IDI). development and beefs up techniques for rapid forest carbon estimation 5 “Poor farming practices and illegal harvesting of timber and NTFPs lead to methodologies. It is, therefore, applicable to multiple academic disci- land degradation, including soil erosion leading to loss of vegetation, plines and multi-sectoral sustainable development practice. biodiversity and drying-up of water bodies (FGD). Description of Author’s Responsibilities 6 F, “lack of employment leads to inadequate household incomes, hunger, CP inability to pay for education and health facilities. This situation is All authors worked on, read and approved the final manuscript. worsened by low crop yields from off-reserve farmlands” (IDI). JOA-initial draft of the manuscript and qualitative data analysis. EMA, FC “there used to be a Community Forest Committees between 2003 and MM, BYF-M and RAA - development of methodology, review of the 2007. It collapsed because of lack of local incentives to protect the forest” manuscript, results and discussion. EKD- geospatial and sustainable (IDI). development conceptualization, geospatial methodology, quantitative F = Farmer; CP = Charcoal producers;; IDI = In-depth interview; FGD = Focus analysis, LULC classification and carbon stock estimation, results/dis- group discussion. cussion and final draft of the manuscript. 6). These findings also confirm Akamani et al., (2015) concerning bar- Declaration of Competing Interest riers to collaborative forest management initiatives. When the participants were asked about the implications of Illegal The authors declare that they have no known competing financial logging, collection and collection of forest fragments (NTFPs), a group of interests or personal relationships that could have appeared to influence the participants shared their acknowledgement. “Poor farming practices the work reported in this paper. and illegal harvesting of timber and NTFPs lead to land degradation, including soil erosion leading to loss of vegetation, biodiversity and drying-up Data availability of water bodies (FGD at Community 5). This finding is not different from the general opinion of most forest fringe community dwellers (Ayivor Data is available in the supplementary file attached et al. 2011; Cobbinah et al. 2015b, a). It affirms the risk forewarned by Derbile et al., (2016) that sustainability aims can be undermined when Acknowledgement local land-use decisions fail to contextualize a balance between liveli- hood, ecosystems and forest reserve sustainability objectives. Thanks for the technical assistance provided by the Ghana Forestry Commission, the Winneba Forestry Service Division, Mr. George Owusu 4. Conclusion and Mrs. Mary Amponsah of CERSGIS and the RS/GIS Lab at the Uni- versity of Ghana. Appreciation goes to the Queen Elizabeth Scholarship - We examined the magnitudes, trends and rates of YFR LULCC and Advanced Scholars Program (QES) of Carleton University, Canada and their effects on carbon stock of the reserve. We also examined the socio- the Centre for Climate Change and Sustainability Studies of the Uni- economic characteristics of the inhabitants and the reasons for their versity of Ghana for their support. livelihood activities in the reserve as well as their roles in the reserve management. Based on Google Earth Engine supervised LULC and car- Appendix A. Supplementary data bon emission coefficients, this study provides new insights into the extent of human livelihood activity impacts on vegetation and above- Supplementary data to this article can be found online at https://doi. ground carbon stock (AGBCS) in the YFR. Generally, the YFR LULCC is org/10.1016/j.jag.2022.102938. related to socio-economic and ecological factors. The carbon stock increased at the expense of close-df forest biomass aboveground. Most of References this increase came from expansion of the open-df while the close-df decreased in size and carbon stock, alongside the close-ghv and open- Acheampong, E., Insaidoo, T.F.G., Ros-Tonen, M.A.F., 2016. Management of Ghana’s ghv declines. The YFR is losing AGBCS mostly from its close-df at an modified taungya system: challenges and strategies for improvement. Agrofor Syst annual rate of 6.5%. The close-ghv is also losing its carbon stock at 1.6% 90, 659–674. https://doi.org/10.1007/s10457-016-9946-7. Akamani, K., Wilson, P.I., Hall, T.E., 2015. Barriers to collaborative forest management per year. Inadequate synergies between indigenes’ livelihood expecta- and implications for building the resilience of forest-dependent communities in the tions and YFR management appear to be yielding marginal results due to Ashanti region of Ghana. J Environ Manage 151, 11–21. https://doi.org/10.1016/j. conflicting interests of livelihood actors and reserve management goals. jenvman.2014.12.006. Amoah, M., Wiafe, E.D., 2012. Livelihoods of fringe communities and the impacts on the Legal/Illegal harvesting of NTFPs, charcoal production, agriculture and management of conservation area: the case of Kakum National Park in Ghana. Int For other activities occur in the reserve despite the people’s knowledge of Rev 14, 131–144. https://doi.org/10.1505/146554812800923381. 8 J. Ofori Acheampong et al. I n t e r n a t i o n a l J o u r n a l o f A p p l i e d E a r t h O b s e r v a t i o n a n d G e o i n f o r m a t i o n 112 (2022) 102938 Appiah, D.O., Bugri, J.T., Forkuo, E.K., Yamba, S., 2016. Agricultural and Forest Land Gorelick, N., Hancher, M., Dixon, M., Ilyushchenko, S., Thau, D., Moore, R., 2017. Use Potential for REDD+ among Smallholder Land Users in Rural Ghana. Int J For Google Earth Engine: Planetary-scale geospatial analysis for everyone. Remote Sens Res 2016, 1–7. https://doi.org/10.1155/2016/7218305. Environ 202, 18–27. Ardayfio-Schandorf, Y.PW., Asiedu, A., et al., 2007. Socio-economic perspectives of off- Goslee K, Brown S, Casarim F (2014) Forest Carbon Monitoring System : Emission reserve forest management in the Goaso Forest District of Ghana Fringe Factors and their Uncertainties , Version 2. Communities of the Goaso Forest District. Woeli Publishing Services, Accra, Ghana. Hassan, A., Abbasi, H., Rakip, I., 2021. Remote Sensing Applications : Society and Ayivor, J.S., Gordon, C., Adomako, J.K., Ntiamoa-Baidu, Y., 2011. Challenges of Environment A review on change detection method and accuracy assessment for managing forest reserves: case study of Atewa range forest reserve, Ghana. Special land use land cover. Remote Sens Appl Soc Environ 22, 100482. https://doi.org/ Issue: Economic and Social Significance of forests for Africa’s Sustainable 10.1016/j.rsase.2021.100482. Development. Nature and Faune. Hou Jones X, Franks P (2015) Food vs forests in sub-Saharan Africa: a challenge for the Baccini, A., Goetz, S.J., Walker, W.S., Laporte, N.T., Sun, M., Sulla-Menashe, D., SDGs. Hackler, J., Beck, P.S.A., Dubayah, R., Friedl, M.A., Samanta, S., Houghton, R.A., Houghton, R.A., House, J.I., Pongratz, J., van der Werf, G.R., DeFries, R.S., Hansen, M.C., 2012. Estimated carbon dioxide emissions from tropical deforestation improved by Le Quéré, C., Ramankutty, N., 2012. Carbon emissions from land use and land-cover carbon-density maps. Nat Clim Chang 2 (3), 182–185. change. Biogeosciences 9 (12), 5125–5142. Banerjee, A., Chen, R., Meadows, M.E., Sengupta, D., Pathak, S., Xia, Z., Mal, S., 2021. IPCC (2019) Summary report on the SBSTA – IPCC special event : Unpacking the new Tracking 21st century climate dynamics of the Third Pole: An analysis of topo- scientific knowledge and key findings in the IPCC Special Report on Climate Change climate impacts on snow cover in the central Himalaya using Google Earth Engine. and Land. Madrid , Spain , 4 December 2019 Note by the Chairs of the SBSTA and the Int J Appl Earth Obs Geoinf 103, 102490. https://doi.org/10.1016/j. IPCC Contents. Madrid, Spain. jag.2021.102490. Kamwi, J.M., Chirwa, P.W.C., Manda, S.O.M., Graz, P.F., Kätsch, C., 2015. Livelihoods, Braun, V., Clarke, V., 2006. Using thematic analysis in psychology. Qual Res Psychol 3 land use and land cover change in the Zambezi Region, Namibia. Popul Environ 37 (2), 77–101. (2), 207–230. Breiman, L., 2001. Random Forests. Mach Learn 45. https://doi.org/10.1023/A: Kyere-Boateng, R., Marek, M.V., 2021. Analysis of the Social-Ecological Causes of 1010933404324. Deforestation and Forest Degradation in Ghana : Application of the DPSIR Chambers R, Conway G (1992) Sustainable Rural Livelihoods: Practical Concepts for the Framework. Forests 1–29. https://doi.org/10.3390/f12040409. 21st Century, IDS Discussion Paper 296. Bright Inst Dev Stud. Lal, R., 2020. Managing soils for resolving the conflict between agriculture and nature: Chen, S., Woodcock, C.E., Bullock, E.L., Arévalo, P., Torchinava, P., Peng, S., The hard talk. Eur J Soil Sci 71, 1–9. https://doi.org/10.1111/ejss.12857. Olofsson, P., 2021. Monitoring temperate forest degradation on Google Earth Engine Lal R (2019) Carbon Cycling in Global Drylands. Curr. Clim. Chang. Reports. using Landsat time series analysis. Remote Sens Environ 265, 112648. https://doi. Langat, D.K., Maranga, E.K., Aboud, A.A., Cheboiwo, J.K., 2016. Role of Forest Resources org/10.1016/j.rse.2021.112648. to Local Livelihoods: The Case of East Mau Forest Ecosystem. Kenya. Int J For Res Cobbinah, P.B., Black, R., Thwaites, R., 2015a. Biodiversity conservation and livelihoods 2016, 1–10. in rural Ghana: Impacts and coping strategies. Environ Dev 15, 79–93. https://doi. Liaw, A., Wiener, M., 2002. Classification and regression by random forest. R News 2. org/10.1016/j.envdev.2015.04.006. Mbanze, A.A., Viera da Silva, C., Ribeiro, N.S., Silva, J.F., Santos, J.L., 2020. A Livelihood Cobbinah, P.B., Black, R., Thwaites, R., 2015b. Ecotourism implementation in the Kakum and Farming System approach for effective conservation policies in Protected Areas Conservation Area, Ghana: Administrative framework and local community of Developing Countries: The case study of the Niassa National Reserve in experiences. J Ecotourism 14, 223–242. https://doi.org/10.1080/ Mozambique. Land use policy 99, 105056. https://doi.org/10.1016/j. 14724049.2015.1051536. landusepol.2020.105056. Crowson, M., Hagensieker, R., Waske, B., 2019. Mapping land cover change in northern Natarajan, N., Newsham, A., Rigg, J., Suhardiman, A., 2022. A sustainable livelihoods Brazil with limited training data. Int J Appl Earth Obs Geoinf 78, 202–214. https:// framework for the 21st century. World Development 155, 105898. https://doi.org/ doi.org/10.1016/j.jag.2018.10.004. 10.1016/j.worlddev.2022.105898. Daur, N., Adam, Y.O., Pretzsch, J., 2016. A historical political ecology of forest access Olofsson, P., Foody, G.M., Herold, M., Stehman, S.V., Woodcock, C.E., Wulder, M.A., and use in Sudan: Implications for sustainable rural livelihoods. Land use policy 58, 2014. Good practices for estimating area and assessing accuracy of land change. 95–101. https://doi.org/10.1016/j.landusepol.2016.06.016. Remote Sens Environ 148, 42–57. Derbile, E.K., Jarawura, F.X., Dombo, M.Y., 2016. Climate change, local knowledge and Rittenhouse, C.D., Rissman, A.R., 2012. Forest cover, carbon sequestration, and wildlife climate change adaptation in Ghana. Adaptation to Climate Change and Variability habitat: Policy review and modeling of tradeoffs among land-use change scenarios. in Rural West Africa. Environ Sci Policy 21, 94–105. https://doi.org/10.1016/j.envsci.2012.04.006. Doe EK, Aikins BE, Njomaba E, Owusu AB (2018) Land use land cover change within Scoones, I., 1998. Sustainable Rural Livelihoods a Framework for Analysis. Analysis 72, kakum conservation area in the Assin South District of Ghana, 1991-2015. West 1–22. https://doi.org/10.1057/palgrave.development.1110037. African J Appl Ecol 26. Shackleton CM, Pandey AK, Ticktin T (2015) Ecological sustainability for non-timber Dottey, R.T., 2005. Effects of Deforestation on Livelihoods of Forest Fringed forest products: Dynamics and case studies of harvesting. Communities in the Awutu-Effutu-Senya District. University of Ghana. Sobeng, K.A., Agyemang-Duah, W., Thomas, A., Appiah, O.J., 2018. An assessment of the Dymond, J.R., Shepherd, J.D., Newsome, P.F., Gapare, N., Burgess, D.W., Watt, P., 2012. effects of forest reserve management on the livelihoods of forest fringe communities Remote sensing of land-use change for Kyoto Protocol reporting: the New Zealand in the Atwima Mponua District of Ghana. For Trees Livelihoods 27, 158–174. case. Environ Sci Policy 16, 1–8. https://doi.org/10.1080/14728028.2018.1490668. FAO (2016) State of the World’s Forests and Agriculture: land-use challenges and Trædal, L.T., Vedeld, P., 2018. Cultivating forests: The role of forest land in household opportunities. Rome. livelihood adaptive strategies in the Bac Kan Province of northern Vietnam. Land use FAO (2019) Agriculture and climate change – Challenges and opportunities at the global policy 73, 249–258. https://doi.org/10.1016/j.landusepol.2018.02.004. and local Level. Verburg, P.H., Alexander, P., Evans, T., Magliocca, N.R., Malek, Z., Rounsevell, M.DA., Forestry Commission (2015) Ghana National REDD+ Strategy. van Vliet, J., 2019. Beyond land cover change: towards a new generation of land use Forestry Commission (2017) Ghana’s National Forest Reference Level. 171. models. Current Opinion in Environmental Sustainability 38, 77–85. Friedlingstein P, Houghton RA, Marland G, et al (2010) Update on CO2 emissions. Nat. WFD (2004) Winneba Forest District Management Plan: Yenku Forest Reserve. Geosci. 3. World Bank (2004) Sustaining forests : a development strategy. Agriculture and rural Gao, J., Liu, Y., 2010. Determination of land degradation causes in Tongyu County, development. Northeast China via land cover change detection. Int J Appl Earth Obs Geoinf 12, Wu, H., Lin, A., Xing, X., Song, D., Li, Y., 2021. Identifying core driving factors of urban 9–16. https://doi.org/10.1016/j.jag.2009.08.003. land use change from global land cover products and POI data using the random Global Fire Monitoring Center (GFMC) (2007) Forestry Commission on wildfire forest method. Int J Appl Earth Obs Geoinf 103, 102475. https://doi.org/10.1016/j. education. jag.2021.102475. 9