The research used qualitative and geospatial techniques (mixed methods) for site selection, data collection, and data analysis. These include stakeholder workshops, focused group discussions, key informant interviews, documentary research, content analysis, geospatial analysis, and water quality analysis.
Conceptual framework
At the center of a sustainable SEL is the improvement of the management of land and the natural resource base in such a way that land use concurrently meets three goals: (i) provision of products (e.g., food) and services on a sustainable basis, (ii) support for sustainable livelihoods for all social groups and (iii) conservation of the full complement of biodiversity and ecosystem services. Globally, sustainable SEL approaches such as inclusive landscape management, agroecology, eco-agriculture and integrated landscape management are already being applied, with promising results in places where food production, poverty alleviation and conservation of biodiversity, water, and ecosystem services are all high priorities [14], [15]. These approaches are applied on productive landscapes with different forms of land use (e.g., forestry, agriculture, extraction of minerals, conservation/protected areas, and settlements) that are symbiotic. Therefore, SEL state and performance assessment frameworks that focus exclusively on, for example, the conservation of natural resources on the one hand or agriculture and other land uses, on the other hand, can at best give an inadequate overview of the SEL. Considering the varied SEL goals, there is need for a comprehensive and iterative assessment framework that considers the drivers of land use and the complex interactions among different land uses and interventions across the landscape.
The purpose here is not to present a new conceptual framework for analyzing SEL sustainability. The intended aim is to draw insights from the works of [13]–[18] and the case studies of the Satoyama Initiative [19] to comprehensively understand the driving forces and pressures that underpin changes in the state of SELs as well as their implications for human wellbeing, ecosystems services and sustainable landscape management in Ghana and Nigeria. The DPSIR-SEL assessment framework (Figure 1) is a coupled social-ecological framework informed by systems thinking. It is a tool that can inform the assessment of landscape-level phenomena.
Essential aspects of the DPSIR-SEL include the five key components which interact at the landscape level and have a significant influence on the benefits derived from the landscape:
Driving Forces motivate human activities and fulfill basic human needs, consistently identified as the necessary conditions and materials for a good life, good health, good social relations, security, and freedom. Driving forces describe “the social, demographic, and economic developments in societies”. Social determinants also have a strong influence on SEL dynamics. Therefore, for this framework, driving forces have been broadened to include socio-cultural and political factors. Accordingly, during the stakeholder workshops, focus group discussions (FGDs) and key informant interviews (KIIs), respondents were asked about what they thought motivated the type of land uses in and around their communities.
Dynamic Pressures, in this paper's context, are human activities derived from the functioning of natural, social and economic driving forces that induce changes in the environment or human systems. Pressures are not stressors. Stressors are the components of the state that are changed by pressures (e.g., land development [the pressure] - increases sediment [the stressor] in urban watersheds, which then may stress the ecological components of the water system to define its state). For this study, participants were asked about:
- Land use changes resulting from alterations in the natural landscape;
- Discharges of pollutants that may result from the operation of industries or vehicles or the diffused distribution of contaminants from agricultural lands, mine sites, or roads through groundwater or storm-water run-off, etc.
- Contact uses activities that lead to a direct alteration or manipulation of the open/closed vegetation, water resources, or land, including:
- Physical damage – direct degradation through mining, dredging and filling, deforestation;
- Biological addition – ballast discharge, the release of non-natives, feeding, creation of artificial habitat;
- Biological harvest – harvesting, fishing, accidental by-catch, clear-cutting
The State of the landscape in this work refers to the state of the natural and built environment. From respondents, information was sought on the quantity and quality of the following components of the landscape:
- physical,
- chemical,
- biological,
- And human systems
With regards impacts on ecosystems and human well-being, the notion is that changes in the structure, functioning and composition of an ecosystem will impact the production of ecosystem goods and services and, ultimately, human well-being through, for example, health and food insecurity. For the impacts on ecosystem goods and services, the study sort data/information on:
- Provisioning services
- Regulating services
- Cultural services
- Supporting processes
On the impacts on human wellbeing, as abstract as the concept is, the study tried to get qualitative information on a mixture of people’s life circumstances and the degree of fulfilment of basic human needs for food, water, health, security, culture, and shelter. Human well-being reflects a positive physical, mental and social state. For this paper, human well-being includes Economic prosperity (e.g., productivity, ability to work, income), Health and safety (e.g., life span, medical or insurance costs, sick days, pain and suffering), Cultural and social well-being (e.g., “happiness”, sense of belonging, community vibrancy, spiritual fulfilment)
Responses: A key benefit of using the DPSIR-SEL framework is that it explicitly includes an Action or Response component that can be taken at any level of the causal network. In the DPSIR-SEL assessment framework, responses are considered actions taken by groups or individuals in society and (non) governmental institutions to:
- prevent
- compensate
- ameliorate
- adapt to changes in the state of the environment
- modify human behaviors that contribute to health risks
- directly modify health through medical treatments or to compensate for social or economic impacts of the human condition on human well-being.
Responses may be directed at driving forces, pressures, landscape state, or impacts. Responses were solicited from participants and review of official reports.
Case study profile
It is important to mention that this work is part of the ONE CGIAR initiative called West and Central Africa AgriFood Transformation (TAFS-WCA). Consequently, the criteria for selecting these target landscapes were informed by two expert workshops held in Ghana by International Water Management Institute (IWMI-Ghana team) and in Ivory Coast by the TAFS-WCA partner institutions. During these engagements, the existence of the following outlined themes guided the selection of case landscapes:
- Location of case landscape: For Ghana, the site must be in the forest transition zone.
- Significant competing land uses and related degradation (e.g., Agriculture, Forestry, Mining, settlement expansion, Chain-saw operations, etc.).
- Types of crops: vegetables, sweet potato, rice, cassava, plantain, cowpea, cocoa, Yam, maize, cocoyam.
- Fishing and aquaculture
- Watersheds and related issues: quantity and quality of water and water productivity.
- Existing landscape management initiatives/Low-hanging-fruits (preferably ONE CGIAR institutions and related projects- e.g., IITA, CIAT, AfricaRice, TAAT II, AICCRA, CCAFS, WAAPP, IAR4D, etc.).
- Existing multi-stakeholder platforms/forums
In the context of this paper, the selected target landscapes in Ghana and Nigeria are used as cases. In Ghana, the target landscape (i.e., Mankran landscape, case study-1) is located in the Mankran watershed in the Offin sub-basin. Administratively, the site is located in the Ahafo Ano South West District (AASWD) in the Forest Transition Belt of Ghana (Figure 2). Ahafo Ano South West District is located at 6°42' north, 1°45' east, and 2°20' west. The district, where case study-1 is located, is approximately 645.54sq/km, about 2.6% of the total land area of the Ashanti region. It is bounded to the northeast by Tano North Municipal in the Ahafo Region, to the northwest by Ahafo Ano North Municipal, to the south by Atwima Nwabiagya Municipal, and to the east by the Offin North District in the Ashanti Region. According to the 2021 Population and Housing Census, the district's population is 65,770, with 33 641 men and 32,129 females [20].
The AASWD comprises 119 settlements, with Mankranso as the capital located approximately 35km from Kumasi. A significant social challenge within the district is the notably high poverty level. This is particularly evident in the human settlements, where rural poverty is starkly manifested. Many communities experience a lack of basic social amenities such as health, education, water, and sanitation, contributing to the overall state of poverty. This hardship is further reflected in the deteriorating infrastructure and the overall decline of the built environment. The district's physiography features numerous rivers/streams, a moist semi-deciduous rainforest, double maxima rainfall, and fertile soils suitable for agriculture, alongside mineral-rich rock formations. However, recent illegal mining and logging activities have escalated, degrading rivers, and transforming tributaries into seasonal water systems. This alteration contributes to water insecurity, particularly in dry seasons. Despite sufficient rainfall for agriculture, its erratic and unpredictable nature poses challenges for rain-fed farming. Climate variability also leads to frequent flooding, exacerbated by river siltation. Natural vegetation is diminishing rapidly, with secondary forests replacing the original cover. Deforestation stems from excessive tree felling, primarily by illegal chain saw operators, and poor farming practices.
In Nigeria, the Doma-Rutu Landscape (case study-2) in the Nasarawa State in Nigeria is the selected case. It is located at latitudes 80 17’ 32” to 80 26’ 48” N and longitude 80 12’ 34” to 80 23’ 16” E, with altitude ranging from 73 m above sea level around the Mada River to 217 m above sea level southwest of the Doma Dam (Figure 3). The landscape is within the Doma Local Government Area of Nasarawa State, Nigeria. The landscape covers an area of approximately 192.26 km2 (19,226 ha). The Mada River borders the West, and the Doma Dam borders the southeast. The main river in the Doma-Rutu Landscape is the Ohina River, which originates from the Shandam Plateau hills, enters the landscape from the southeastern flange, flows through the full length of the landscape, and drains into the Mada River to the northwest. The Ohina River is dammed within the landscape (the Doma Dam) and then flows approximately 15.4 km (from the Doma Dam) to the Mada River. The Mada River is a larger river flowing from the plateau hills and passing by the border of the landscape in the northeastern part of Rutu village. The communities within the landscape include parts of Doma and Mukaiya towns in the northeast, Dogon Kurmi/Iwashi village in the north, Rutu village in the northwest, and Alagye village in the southwest [21], [22]. Records of the human population of the communities in the landscape are not within reach; however, the population of the Doma Local Government Area as of 2022 is estimated at 214,600 people at a growth rate of 2.8% per annum [23].
Qualitative data collection and analysis
This study used primary and secondary data from both countries. The primary data were largely qualitative. In AASWD (case study-1), nine parallel FGDs and Participatory Rural Appraisal (PRA) sessions were carried out. These sessions engaged not only community members but also community leaders. Simultaneously, a comprehensive set of 12 institutional interviews was conducted, encompassing vital entities such as the Agricultural Department, Forestry Department, Education Directorate, Health Directorate, Ambulance Services, Social Welfare and Community Development, Environmental Health and Sanitation, Judiciary (Registrar), National Commission for Civic Education (NCCE), Commission for Human Rights and Administrative Justice (CHRAJ), and the Police. Consent forms were methodically employed to ensure a foundation for informed consent before the participants' involvement. Privacy safeguards have been established by applying pseudo-identification methodologies. Participants were educated on the study's objectives to mitigate potential response biases rooted in social desirability [24]. Comprehensive records were maintained for all interviews and focus group discussions, each with the participants' explicit consent.
At the Doma-Rutu Landscape (DRL) (Case study-2), a series of nine Focus Group Discussions (FGDs) and Participatory Rural Appraisal (PRA) sessions were conducted, engaging diverse community members, including men, women, youth, herders, institutions, and organizational leaders. These interactive sessions were held across three communities and at the Doma Irrigation Site. Additionally, 19 Key Informant Interviews (KIIs) were conducted involving community leaders, ministry representatives, and various organizations. Consent for participation was obtained through consent forms and vocal agreements, accompanied by pseudo-identification measures to ensure participant anonymity. The overarching research objectives were communicated transparently to mitigate potential social desirability bias. The technological prowess of the Online Data Kit (ODK) facilitated seamless data collection during both KIIs and FGDs.
The FGDs were categorized into four cohorts: men, women, youth, and herders, hailing from diverse communities. Inhabitants with a minimum of a decade's community residency constituted the FGD participants, and their extensive local experience contributed valuable and dependable insights into the landscape. The ambit of discussions encompassed an array of themes spanning food production, land use dynamics, community socioeconomic structures, food security paradigms, livelihood alternatives, conservation initiatives, and the preservation and revival of wild biodiversity and ecosystem services. This comprehensive approach reverberated consistently across the focused group discussions and participatory rural appraisal (PRA) sessions. Content analysis was used to analyze the reports and transcripts from key informant interviews and FGDs [25], [26].
Geospatial data acquisition and analysis
Secondary data were mainly geospatial, sociodemographic, and economic (Table 1). Data were collected to evaluate changes in land use and land cover (LULC) in Ghana and Nigeria, utilizing data from 2008, 2015, 2018, and 2021 for Ghana and 2000, 2010, and 2021 for Nigeria (Table 2). The Google Earth Engine was used to extract and process historical images of a given year. The GEE platform was used for data processing for both sites, as it provides better solutions for assessing and processing large amounts of freely available online data [33]. GEE also offers an opportunity for more rapid analysis of LULC images using a set of pixel-based classifiers with different classification techniques for land mapping [34]–[36]. One of the challenges of satellite imagery in developing countries is cloud cover [37]–[39]; however, all images captured during the dry season were used to avoid this. Subsequently, the images were georeferenced within ArcGIS Pro version 3.0 software. To delineate the boundaries of the case study landscapes, the pour point technique was used. In Mankran case, pour points located at the southernmost confluence of the Mankranso river were identified and used to delineate the watershed after a hydrologically conditioned DEM was created. Extraction of flow characteristics (flow direction, flow accumulation, stream order, flow length, stream link and stream feature) was then carried out to delineate the boundary of case study-1. Same was replicated for the Nigeria case.
Table 1: Secondary data used in the study
Available Data
|
Source
|
Purpose
|
Year
|
|
Ghana (case study-1)
|
|
Socio-demographic and economic data
|
Ghana Statistical Service [27]
|
Profile of study districts/communities
|
2021 Population Census
|
Satellite images
|
European Space Agency [28], National Aeronautics and Space Administration (NASA)[29]
|
Land cover/ Land use mapping
|
1986 - 2022
|
Digital elevation models
|
NASA
|
Watershed and Topographic mapping
|
2013-2017
|
Forest/Game Reserves
|
Forestry Commission
|
Protected area mapping
|
|
Roads
|
Roads & Highways
|
|
2021
|
Streams and Rivers
|
Hydrological services Department [30]
|
Drainage Density, watershed delineation, water quantity and quality assessment
|
2021
|
|
Nigeria (case study-2)
|
|
Socio-demographic and economic data
|
Nigeria Bureau of Statistics [31]
|
Profile of study districts/communities
|
2021 Population Census
|
Satellite images
|
European Space Agency, NASA
|
Land cover/ Land use mapping
|
2000, 2010, 2022
|
Digital elevation models
|
NASA
|
Watershed and Topographic mapping
|
2021
|
Forest/Game Reserves
|
Forestry Commission
|
Protected area mapping
|
|
Roads
|
Roads & Highways
|
|
2021
|
Streams and Rivers
|
Nigeria Hydrological Services Department [32]
|
Drainage Density, watershed delineation, water quantity and quality assessment
|
2022
|
Table 2: Characteristics of the satellite data used for both Ghana and Nigeria
Satellite
|
Sensor
|
Period covered
|
Spatial Resolution (m)
|
Ghana (case study 1-)
|
Landsat -7
|
TM
|
2008
|
30
|
Landsat -7
|
TM
|
2015
|
30
|
Landsat-8 OLI
|
OLI
|
2018
|
30
|
Landsat-8 OLI
|
OLI
|
2021
|
30
|
Nigeria (case study 2-)
|
Landsat -7
|
TM
|
2000
|
30
|
Landsat -7
|
TM
|
2010
|
30
|
Landsat-8 OLI
|
OLI
|
2022
|
30
|
For image classification, a modified version of the Food and Agricultural Organizations’ land cover classification system was adopted to identify land cover and land use types in the landscapes (Table 3). Approximately 178 training points covering all land use land cover classes were purposefully collected from the watershed to aid interpretation. Segmented images were created from the object-based classification process. The derived image and 110 stratified and randomly selected training samples were used to identify and classify land use/land cover types. Specifically, object-based image classification with the Support Vector Machine algorithm was used. Unlike pixel-based classification, which considers only spectral information at the individual pixel level, object-based classification considers spatial and spectral characteristics of features of interest. It can produce results comparable to visually interpreted images. Object-based classification suits high-resolution images with spectrally heterogeneous features [40]. The quality of the classified image was checked using the image using the remaining 68 field samples. In measuring the accuracy of classification, the proportion correctly classified (PCC) index and the Kappa statistic, derived from an error matrix, were used. Post-classification change detection techniques were employed to account for land use/land cover transfers between the period in question (i.e., January 2008 – December 2021in the case of Mankran landscape). This involved an overlay of independently classified images. It is the most used qualitative method of change detection [41]. It operates on two or more independently classified images as inputs, resulting in a change map and a change matrix. The classified thematic map of 2008, 2015, 2018 and 2021 was loaded and analyzed using tools in ArcGIS Pro 3.0 to indicate changes between the images in the form of a change map and change matrix, which was then used for the analysis.
Table 3: Adapted Land use / Land cover classification systems of the FAO
FAO LAND COVER CLASSIFICATION SYSTEM
|
ADAPTED LAND USE/ LAND COVER CLASSES
|
Cultivated and Managed Terrestrial Vegetation
|
Cash Crop
|
Subsistence Farming
|
Natural - Semi-Natural Terrestrial Vegetation
|
Dense Natural Vegetation
|
|
Sparse Natural Vegetation
|
Artificial Surface
|
Road
|
Built-Up
|
Bare Surface
|
Degraded Areas
|
Recovering Mining Sites
|
|
Clear Natural Water Body
|
Natural - Semi-Natural Aquatic Vegetation
|
Natural Aquatic Vegetation
|
Artificial Aquatic Vegetation
|
Rice
|
Artificial Waterbody
|
Artificial Water Body (Irrigation)
|
Source: [42]
Water quality testing
For the ecosystem degradation assessment of the cases, we considered only the water quality of water bodies in the landscape as an indicator. For the case study in Ghana, there was no primary data collection by the authors, but a literature review from recent studies in the landscape was conducted. At case study site 1, 17 water samples in the offshore sub-basin were collected randomly from the river systems where the questionnaire was administered between May and June 2019 similar to [43]. The precise coordinates of the sampling points were documented using a handheld Garmin Etrex GPS device. The sampling procedure adhered to the guidelines set by the American Public Health Association (APHA) [44]. Plastic bottles of 500 mL capacity were thoroughly rinsed within the river following the flow direction to collect the samples. After collection, the samples were promptly placed in a cooler box containing ice blocks to ensure preservation during transportation to the laboratory, maintaining their natural state.
The analysis of heavy metals, including arsenic (As), cadmium (Cd), copper (Cu), iron (Fe), mercury (Hg), and lead (Pb), was executed at SGS Laboratory Services Ltd in Tema, Ghana, employing an inductively coupled plasma–optical emission spectrometry instrument (Nexion 300x ICP). Other analytical procedures were conducted at the Environmental Quality Engineering Laboratory of Kwame Nkrumah University of Science and Technology in Kumasi, Ghana. Parameters such as total dissolved solids (TDS), electrical conductivity (EC), and pH were determined using a Palin test multimeter. Total suspended solids (TSS) were evaluated using the gravimetric method. The analysis of nitrate-nitrogen (N-NO3) and ammonia-nitrogen (N-NH4) was performed using a DR 3900 spectrophotometer. Acid-persulfate digestion, cadmium reduction, and Nessler methods were applied to determine these parameters. For total nitrogen (TN), the Kjeldahl method was employed using Velp 139 distillation equipment.
In case study 2, 15 water samples were collected from the river systems by the co-authors where the questionnaires were administered. Subsequently, in situ, rapid water quality testing was carried out for the different streams using a C-600 7 digital water quality tester. Each test was repeated three times, and the means were recorded. The parameters measured included pH (pH scale), temperature (°C), DO (mg L−1), EC (μS/cm), TDS (mg L−1), DO (mg L−1), oxidation-reduction potential (ORP), Specific Gravity (SG), and water temperature using the ‘Standard Methods for the Examination of Water and Wastewater’ [44]. These water quality parameters were compared with the WHO standards for drinking water [45].