Takoutsing Bertin | ICRAF - Academia.edu (original) (raw)

Papers by Takoutsing Bertin

Geoderma

Geostatistics and machine learning have been extensively applied for modelling and predicting the... more Geostatistics and machine learning have been extensively applied for modelling and predicting the spatial distribution of continuous soil variables. In addition to providing predictions, both techniques quantify the uncertainty associated with the predictions, although geostatistics is more developed in this respect. Despite the increased use of these techniques, most algorithms ignore that the soil measurements are not error-free. Recently, concern has also arisen about the extrapolation risk of these techniques, be it in geographic space, feature space, or both. In this paper, regression kriging (RK) and random forest (RF) were compared with respect to their ability to deliver accurate predictions and quantify prediction uncertainties, while accounting for measurement errors in the soil data. The sensitivity of results of both models to soil measurement errors was also evaluated, as well as their spatial extrapolation potential. This was done for a case study in Cameroon where soil pH, clay and organic carbon were mapped from measurements obtained using both conventional and proximal soil sensing methods. The results showed that both models produced comparable ranges and maps of predicted values for the soil properties of interest. Compared to RF, RK outperformed RF by presenting generally a higher Model Efficiency Coefficient (MEC), lower Root Mean Squared Error (RMSE) values and better extrapolation performance. The improvement in RMSE was about 10, 12 and 2 % while the improvement in MEC was on average 5, 22 and 1 % for pH, clay and SOC, respectively Overestimation of the local uncertainty observed for RK was larger than that of RF as shown by accuracy plots, indicating that prediction uncertainties were better quantified by the RF model. Better extrapolation performance was obtained with RK that derived better predictions than RF at unsampled locations as shown by cross-validation metrics and scatter plots, particularly when RK and RF were used for spatial extrapolation. The effects of incorporating measurement errors were not significant both for the predictions and for the prediction uncertainties due to the fact that most calibration data had the same measurement error variance. Model comparison should go beyond common validation metrics that only evaluate prediction accuracy but must also account for the ability to quantify prediction uncertainty at unsampled locations.

Geostatistics and machine learning have been extensively applied for modelling and predicting the... more Geostatistics and machine learning have been extensively applied for modelling and predicting the spatial distribution of continuous soil variables. In addition to providing predictions, both techniques quantify the uncertainty associated with the predictions, although geostatistics is more developed in this respect. Despite the increased use of these techniques, most algorithms ignore that the soil measurements are not error-free. Recently, concern has also arisen about the extrapolation risk of these techniques, be it in geographic space, feature space, or both. In this paper, regression kriging (RK) and random forest (RF) were compared with respect to their ability to deliver accurate predictions and quantify prediction uncertainties, while accounting for measurement errors in the soil data. The sensitivity of results of both models to soil measurement errors was also evaluated, as well as their spatial extrapolation potential. This was done for a case study in Cameroon where soil pH, clay and organic carbon were mapped from measurements obtained using both conventional and proximal soil sensing methods. The results showed that both models produced comparable ranges and maps of predicted values for the soil properties of interest. Compared to RF, RK outperformed RF by presenting generally a higher Model Efficiency Coefficient (MEC), lower Root Mean Squared Error (RMSE) values and better extrapolation performance. The improvement in RMSE was about 10, 12 and 2 % while the improvement in MEC was on average 5, 22 and 1 % for pH, clay and SOC, respectively Overestimation of the local uncertainty observed for RK was larger than that of RF as shown by accuracy plots, indicating that prediction uncertainties were better quantified by the RF model. Better extrapolation performance was obtained with RK that derived better predictions than RF at unsampled locations as shown by cross-validation metrics and scatter plots, particularly when RK and RF were used for spatial extrapolation. The effects of incorporating measurement errors were not significant both for the predictions and for the prediction uncertainties due to the fact that most calibration data had the same measurement error variance. Model comparison should go beyond common validation metrics that only evaluate prediction accuracy but must also account for the ability to quantify prediction uncertainty at unsampled locations.

European Journal of Soil Science, 2022

Digital soil mapping (DSM) approaches provide soil information by utilising the relationship betw... more Digital soil mapping (DSM) approaches provide soil information by utilising the relationship between soil properties and environmental variables. Calibration of DSM models requires measurements that may often have substantial measurement errors which propagate to the DSM outputs and need to be

Agroforestry Systems, 2022

The restoration of degraded lands has received increased attention in recent years and many commi... more The restoration of degraded lands has received increased attention in recent years and many commitments have been made as part of global and regional restoration initiatives. Well-informed policy decisions that support land restoration, require spatially explicit information on restoration potentials to guide the design and implementation of restoration interventions in the context of limited resources. This study assessed ecosystems indicators of land degradation using a systematic approach that combines field surveys and remote sensing data into a set of multi-criteria analyses to map restoration potentials in the semi-arid areas. The indicators considered were soil organic carbon, erosion prevalence, enhanced vegetation index, Normalized differences water index and the Net Primary productivity. Three classes of restoration potential were established: (1) areas not in need of immediate restoration due low degradation status, (2) areas with high potential for restoration with moderate efforts required and (3) areas in critical need of restoration and require high level of efforts. Of the total area of the study site estimated at 88,344 km 2 , 59,146.12 km 2 , or 66.94% of the theoretically recoverable area, was considered suitable for restoration, of which 38% required moderate efforts while 28% require less efforts. The recoverable areas suitable for restoration could be restored through tree planting, soil and water conservation practices, farmers managed natural regeneration, and integrated soil fertility management. These results can help to spatially identify suitable multifunctional restoration and regeneration hotspots as an efficient way to prioritize restoration interventions in the context of limited resources.

Afrika Focus, 2015

Pentaclethra macrophylla Benth (Fabaceae) has emerged as one of the top underutilized agroforestr... more Pentaclethra macrophylla Benth (Fabaceae) has emerged as one of the top underutilized agroforestry species with considerable economic and nutritional value in West and Central Africa where the species is endemic and exploited by farmers. However, the species suffers from a level of exploitation that is unsustainable. This dramatically decreases the natural resource base. Key information such as tree variation is needed to develop an ef cient strategy for its domestication. This study therefore assesses the species phenotypic variation in two populations in the forest zone of Cameroon. Overall, 49 trees (19 in Lomie and 30 in Djoum) were sampled. Tree characteristics (diameter at breast height (DBH), height, crown diameter), pod (number of seeds per pod), seed (width, length, thickness, weight) and kernel traits (weight) were recorded. All these parameters were subjected to a combination of multivariate, PCA and cluster analyses. Results revealed four clusters of P. macro- phylla tre...

Land Degradation & Development, 2018

Soil properties are useful for assessing the potential of landscapes to provide terrestrial ecosy... more Soil properties are useful for assessing the potential of landscapes to provide terrestrial ecosystem services, but they are affected by anthropogenic activities and environmental factors including landscape attributes. This study assessed how soil properties are influenced by landscape attributes and their interactions in the highlands of Cameroon using the Land Degradation Surveillance Framework as a data collection tool. Soil properties [soil organic carbon (SOC), clay content, exchangeable bases (ExBas), electric conductivity (EC), boron (B), manganese (Mn), phosphorus (P), pH] were quantified within classes of landscape attributes. Soil samples were collected on 160 (1000 m 2) plots randomly located in a sentinel site of 100 km 2 and were analyzed using a combination of conventional laboratory methods and midinfrared spectroscopy. Soil properties were highly affected by soil depths, land use types, slope gradients and topographic positions, but less by their interactions. Significant interactions existed between land use types and topographic positions for SOC, EC, ExBas and pH, and between slope gradients and topographic positions for pH, while Mn was influenced by the interaction between land use types and slope gradients. Most soil properties were higher in low altitude plots and those with higher vegetation cover but decreased in the upslope direction. The pH and clay content were less affected by slope gradient confirming the inherent nature of the properties. These results are useful for site-specific implementation of ecological intensification in areas with complex topography such as the highlands of Cameroon, offering a reference for future ecological policies and landscape restoration.

Journal of Geochemical Exploration, 2017

Understanding spatial variability of soil properties is essential to support land management deci... more Understanding spatial variability of soil properties is essential to support land management decisions. However, despite the growing worldwide emphasis on integrated landscape management, soil variations resulting from land use changes have rarely been documented. The study used the land health surveillance concept in combination with simple geostatistical approaches to describe key soil properties among land use types and characterize their spatial variability in the highlands of Cameroon. A total of 320 soil samples were collected in two sites with contrasting land uses (agricultural and pasture) and were analyzed for granulometric fraction, soil organic carbon (SOC), nitrogen (N), soil reaction (pH), phosphorous (P), calcium (Ca), potassium (K), magnesium (Mg), aluminum (Al) and zinc (Zn). The spatial correlations between the soil properties revealed the factors responsible for the observed differences and showed that wide ranges were obtained in agricultural site as opposed to pasture. SOC and N decreased in the order of forest > grassland > fallow > croplands > pasture due to inherent soil properties, anthropogenic activities, land cover/land use and topographic factors. Kriged maps provided detailed visualization of soil properties at landscape scale, and helped to identify critical areas for targeted land management interventions to improve land quality. The spatial distribution of selected soil properties showed a well-defined pattern of higher concentrations in the lowlands and valleys and areas with permanent vegetation cover in both sites.. These results are useful for improving the efficient use of inputs such fertilizers. Context-specific land management based on spatial variability of soil properties is highly recommended and more research is required to generalize our knowledge about spatial variability of soil health indicators and the casual factors in the highlands of Cameroon. ACCEPTED MANUSCRIPT Highlights  Geostatistics was used to analyze spatial structures of soil quality parameters  Clay content has a negative effect on soil properties  Land use types have significant effects on spatial patterns of soil properties  Soil variability was due to inherent conditions, land use types and management  Nutrient spatial patterns are useful for site specific land management practices

ABSTRACT A field experiment was conducted in the Western Highlands of Cameroon to study the effec... more ABSTRACT A field experiment was conducted in the Western Highlands of Cameroon to study the effects of Calliandra calothyrsus, sterilized compost, non-sterilized compost and mineral fertilizers (NPK 11:11:22) on physical characteristics, yield components and late blight disease severity of potato (Solanum tuberosum). A two factorial treatment combination made up of fertilization schemes and sanitary measures were laid out in a randomized complete block design (RCBD) with four replicates and 10 treatments. Data collected were subjected to a multivariate ANOVA and means separated with the Dunnett t-test with Calliandra calothyrsus considered as the main treatment. Mineral fertilizers and Calliandra calothyrsus significantly augmented stem diameter, plant height and plant vigor, total and marketable yields, despite the fact that late blight severity was high in the two treatments. The correlation matrix showed that total yield had significant and positive correlation with stem diameter (r = 0.74), plant height (r = 0.61), plant vigor (r = 0.61) and marketable yield (r = 0.99) and negative correlation (r = -37) with late blight severity. The increase observed is attributed to the nutrient contents of the treatments applied. Leafy prunings of C. calothyrsus increase soil organic matter which probably improves moisture retention and biological activity. The highest late blight severity recorded between the 45 and 60 coincides with active development of plant canopy which influences disease severity and create a canopy microclimate conducive to disease development. This study suggests that Calliandra calothyrsus has the potential of improving potato production provided appropriate fungicide is used.

Journal of Agricultural Science, 2012

Collective action has been promoted as a strategy to improve the incomes of small scale producers... more Collective action has been promoted as a strategy to improve the incomes of small scale producers in many developing countries, primarily by reducing transaction costs and improving farmers' bargaining power. This paper evaluates producers' perception of collective action initiatives which have been used in Cameroon for the production and marketing of kola nuts. Using a perceptual evaluation and cluster analysis based on interviews with 203 kola producers in western highlands of Cameroon, findings reveal that producers evaluate effectiveness of collective action based on 5 main groups of criteria: reduced transaction costs, new learning and skills acquisition, market and financial status, social networks and status, and unfavorable dimensions. Furthermore, two main types of producers who evaluate the collective initiatives differently are identified: the positive group and the unimpressed group. Higher proportions of women and youth are in the unimpressed group than in the positive group, suggesting that more needs to be done to understand their perspectives and better target collective action initiatives to their needs and circumstances. The paper concludes that promoters of collective action initiatives need to adopt differentiated strategies to enhance its adoption in the study area.

Small-scale Forestry, 2013

Your article is protected by copyright and all rights are held exclusively by Steve Harrison, Joh... more Your article is protected by copyright and all rights are held exclusively by Steve Harrison, John Herbohn. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com".

Small-scale Forestry, 2012

Availability of high quality tree planting material within proximity of farmers and at affordable... more Availability of high quality tree planting material within proximity of farmers and at affordable prices is one of the prerequisites for larger uptake of tree cultivation. This study examines whether rural small-scale nurseries can produce a diversity of tree planting material and whether resource-poor farmers have access to it. Twelve nurseries supported by the tree domestication program in the West and Northwest regions of Cameroon were compared to 12 nurseries in similar conditions, but not in contact with the program. Nurseries using the domestication approach were found to provide tree planting material that responds better to farmers' needs in terms of quantities, species and propagation methods used. Their clientele is more diverse including farmers from the communities where the nurseries are located, but also from far beyond. However, prices of vegetatively propagated material were considered the most prohibitive factor. It is concluded that tree planting initiatives should refocus efforts towards technical training and business support to small-scale nurseries to increase efficiency. Research efforts should look for ways of reducing production costs and improving nursery productivity.

The Journal of Agricultural Education and Extension, 2014

ABSTRACT Purpose: Formal agricultural research has generated vast amount of knowledge and fundame... more ABSTRACT Purpose: Formal agricultural research has generated vast amount of knowledge and fundamental insights on land management, but their low adoption has been attributed to the use of public extension approach. This research aims to address whether and how full participation of farmers through the concept of Rural Resource Centre (RRC) provides new insights for the development of alternative and farmers-based extension methods. Design/Methodology/Approach: Using the Concept of RRC, this research assesses the role of farmers in on-farm demonstrations and scaling-up of land management practices, and investigates effective ways to enhance beneficial interactions between researchers, extension workers and farmers in view of improving adoption. Findings: The findings suggest that farmers can effectively participate in demonstrations and scaling-up of agricultural practices. This participation is enhanced by judicious incentives such as higher crop yields that motivate farmers and influence adoption. The current success of the approach stems from the fact that farmers, extension workers and researchers jointly implement the activities and their different aims were achieved simultaneously: scientific results for researchers, better agricultural practices for extension workers, and economic success and free choice for farmers. Practical implications: This research concludes that farmers have the capacities to play an innermost role in demonstrations and scaling-up of agricultural practices. However, there is a need to build and strengthen their capacities to facilitate their participation and contribution. Originality/Value: The article demonstrates the value of the preponderant role of farmers in onfarm demonstrations and scaling-up practices by exhibiting the beneficial interactions between researchers, extension workers and farmers.

International Journal of Agricultural Sustainability, 2011

The restoration of degraded lands has received increased attention in recent years and many commi... more The restoration of degraded lands has received increased attention in recent years and many commitments have been made as part of global and regional restoration initiatives. Well-informed policy decisions that support land restoration, require spatially explicit information on restoration potentials to guide the design and implementation of restoration interventions in the context of limited resources. This study assessed ecosystems indicators of land degradation using a systematic approach that combines field surveys and remote sensing data into a set of multicriteria analyses to map restoration potentials in the semi-arid areas. The indicators considered were soil organic carbon, erosion prevalence, enhanced vegetation index, Normalized differences water index and the Net Primary productivity. Three classes of restoration potential were established: (1) areas not in need of immediate restoration due low degradation status, (2) areas with high potential for restoration with moderate efforts required and (3) areas in critical need of restoration and require high level of efforts. Of the total area of the study site estimated at 88,344 km 2 , 59,146.12 km 2 , or 66.94% of the theoretically recoverable area, was considered suitable for restoration, of which 38% required moderate efforts while 28% require less efforts. The recoverable areas suitable for restoration could be restored through tree planting, soil and water conservation practices, farmers managed natural regeneration, and integrated soil fertility management. These results can help to spatially identify suitable multifunctional restoration and regeneration hotspots as an efficient way to prioritize restoration interventions in the context of limited resources.

Digital soil mapping (DSM) approaches provide soil information by utilising the relationship betw... more Digital soil mapping (DSM) approaches provide soil information by utilising the relationship between soil properties and environmental variables. Calibration of DSM models requires measurements that may often have substantial measurement errors which propagate to the DSM outputs and need to be

Geoderma

Geostatistics and machine learning have been extensively applied for modelling and predicting the... more Geostatistics and machine learning have been extensively applied for modelling and predicting the spatial distribution of continuous soil variables. In addition to providing predictions, both techniques quantify the uncertainty associated with the predictions, although geostatistics is more developed in this respect. Despite the increased use of these techniques, most algorithms ignore that the soil measurements are not error-free. Recently, concern has also arisen about the extrapolation risk of these techniques, be it in geographic space, feature space, or both. In this paper, regression kriging (RK) and random forest (RF) were compared with respect to their ability to deliver accurate predictions and quantify prediction uncertainties, while accounting for measurement errors in the soil data. The sensitivity of results of both models to soil measurement errors was also evaluated, as well as their spatial extrapolation potential. This was done for a case study in Cameroon where soil pH, clay and organic carbon were mapped from measurements obtained using both conventional and proximal soil sensing methods. The results showed that both models produced comparable ranges and maps of predicted values for the soil properties of interest. Compared to RF, RK outperformed RF by presenting generally a higher Model Efficiency Coefficient (MEC), lower Root Mean Squared Error (RMSE) values and better extrapolation performance. The improvement in RMSE was about 10, 12 and 2 % while the improvement in MEC was on average 5, 22 and 1 % for pH, clay and SOC, respectively Overestimation of the local uncertainty observed for RK was larger than that of RF as shown by accuracy plots, indicating that prediction uncertainties were better quantified by the RF model. Better extrapolation performance was obtained with RK that derived better predictions than RF at unsampled locations as shown by cross-validation metrics and scatter plots, particularly when RK and RF were used for spatial extrapolation. The effects of incorporating measurement errors were not significant both for the predictions and for the prediction uncertainties due to the fact that most calibration data had the same measurement error variance. Model comparison should go beyond common validation metrics that only evaluate prediction accuracy but must also account for the ability to quantify prediction uncertainty at unsampled locations.

Geostatistics and machine learning have been extensively applied for modelling and predicting the... more Geostatistics and machine learning have been extensively applied for modelling and predicting the spatial distribution of continuous soil variables. In addition to providing predictions, both techniques quantify the uncertainty associated with the predictions, although geostatistics is more developed in this respect. Despite the increased use of these techniques, most algorithms ignore that the soil measurements are not error-free. Recently, concern has also arisen about the extrapolation risk of these techniques, be it in geographic space, feature space, or both. In this paper, regression kriging (RK) and random forest (RF) were compared with respect to their ability to deliver accurate predictions and quantify prediction uncertainties, while accounting for measurement errors in the soil data. The sensitivity of results of both models to soil measurement errors was also evaluated, as well as their spatial extrapolation potential. This was done for a case study in Cameroon where soil pH, clay and organic carbon were mapped from measurements obtained using both conventional and proximal soil sensing methods. The results showed that both models produced comparable ranges and maps of predicted values for the soil properties of interest. Compared to RF, RK outperformed RF by presenting generally a higher Model Efficiency Coefficient (MEC), lower Root Mean Squared Error (RMSE) values and better extrapolation performance. The improvement in RMSE was about 10, 12 and 2 % while the improvement in MEC was on average 5, 22 and 1 % for pH, clay and SOC, respectively Overestimation of the local uncertainty observed for RK was larger than that of RF as shown by accuracy plots, indicating that prediction uncertainties were better quantified by the RF model. Better extrapolation performance was obtained with RK that derived better predictions than RF at unsampled locations as shown by cross-validation metrics and scatter plots, particularly when RK and RF were used for spatial extrapolation. The effects of incorporating measurement errors were not significant both for the predictions and for the prediction uncertainties due to the fact that most calibration data had the same measurement error variance. Model comparison should go beyond common validation metrics that only evaluate prediction accuracy but must also account for the ability to quantify prediction uncertainty at unsampled locations.

European Journal of Soil Science, 2022

Digital soil mapping (DSM) approaches provide soil information by utilising the relationship betw... more Digital soil mapping (DSM) approaches provide soil information by utilising the relationship between soil properties and environmental variables. Calibration of DSM models requires measurements that may often have substantial measurement errors which propagate to the DSM outputs and need to be

Agroforestry Systems, 2022

The restoration of degraded lands has received increased attention in recent years and many commi... more The restoration of degraded lands has received increased attention in recent years and many commitments have been made as part of global and regional restoration initiatives. Well-informed policy decisions that support land restoration, require spatially explicit information on restoration potentials to guide the design and implementation of restoration interventions in the context of limited resources. This study assessed ecosystems indicators of land degradation using a systematic approach that combines field surveys and remote sensing data into a set of multi-criteria analyses to map restoration potentials in the semi-arid areas. The indicators considered were soil organic carbon, erosion prevalence, enhanced vegetation index, Normalized differences water index and the Net Primary productivity. Three classes of restoration potential were established: (1) areas not in need of immediate restoration due low degradation status, (2) areas with high potential for restoration with moderate efforts required and (3) areas in critical need of restoration and require high level of efforts. Of the total area of the study site estimated at 88,344 km 2 , 59,146.12 km 2 , or 66.94% of the theoretically recoverable area, was considered suitable for restoration, of which 38% required moderate efforts while 28% require less efforts. The recoverable areas suitable for restoration could be restored through tree planting, soil and water conservation practices, farmers managed natural regeneration, and integrated soil fertility management. These results can help to spatially identify suitable multifunctional restoration and regeneration hotspots as an efficient way to prioritize restoration interventions in the context of limited resources.

Afrika Focus, 2015

Pentaclethra macrophylla Benth (Fabaceae) has emerged as one of the top underutilized agroforestr... more Pentaclethra macrophylla Benth (Fabaceae) has emerged as one of the top underutilized agroforestry species with considerable economic and nutritional value in West and Central Africa where the species is endemic and exploited by farmers. However, the species suffers from a level of exploitation that is unsustainable. This dramatically decreases the natural resource base. Key information such as tree variation is needed to develop an ef cient strategy for its domestication. This study therefore assesses the species phenotypic variation in two populations in the forest zone of Cameroon. Overall, 49 trees (19 in Lomie and 30 in Djoum) were sampled. Tree characteristics (diameter at breast height (DBH), height, crown diameter), pod (number of seeds per pod), seed (width, length, thickness, weight) and kernel traits (weight) were recorded. All these parameters were subjected to a combination of multivariate, PCA and cluster analyses. Results revealed four clusters of P. macro- phylla tre...

Land Degradation & Development, 2018

Soil properties are useful for assessing the potential of landscapes to provide terrestrial ecosy... more Soil properties are useful for assessing the potential of landscapes to provide terrestrial ecosystem services, but they are affected by anthropogenic activities and environmental factors including landscape attributes. This study assessed how soil properties are influenced by landscape attributes and their interactions in the highlands of Cameroon using the Land Degradation Surveillance Framework as a data collection tool. Soil properties [soil organic carbon (SOC), clay content, exchangeable bases (ExBas), electric conductivity (EC), boron (B), manganese (Mn), phosphorus (P), pH] were quantified within classes of landscape attributes. Soil samples were collected on 160 (1000 m 2) plots randomly located in a sentinel site of 100 km 2 and were analyzed using a combination of conventional laboratory methods and midinfrared spectroscopy. Soil properties were highly affected by soil depths, land use types, slope gradients and topographic positions, but less by their interactions. Significant interactions existed between land use types and topographic positions for SOC, EC, ExBas and pH, and between slope gradients and topographic positions for pH, while Mn was influenced by the interaction between land use types and slope gradients. Most soil properties were higher in low altitude plots and those with higher vegetation cover but decreased in the upslope direction. The pH and clay content were less affected by slope gradient confirming the inherent nature of the properties. These results are useful for site-specific implementation of ecological intensification in areas with complex topography such as the highlands of Cameroon, offering a reference for future ecological policies and landscape restoration.

Journal of Geochemical Exploration, 2017

Understanding spatial variability of soil properties is essential to support land management deci... more Understanding spatial variability of soil properties is essential to support land management decisions. However, despite the growing worldwide emphasis on integrated landscape management, soil variations resulting from land use changes have rarely been documented. The study used the land health surveillance concept in combination with simple geostatistical approaches to describe key soil properties among land use types and characterize their spatial variability in the highlands of Cameroon. A total of 320 soil samples were collected in two sites with contrasting land uses (agricultural and pasture) and were analyzed for granulometric fraction, soil organic carbon (SOC), nitrogen (N), soil reaction (pH), phosphorous (P), calcium (Ca), potassium (K), magnesium (Mg), aluminum (Al) and zinc (Zn). The spatial correlations between the soil properties revealed the factors responsible for the observed differences and showed that wide ranges were obtained in agricultural site as opposed to pasture. SOC and N decreased in the order of forest > grassland > fallow > croplands > pasture due to inherent soil properties, anthropogenic activities, land cover/land use and topographic factors. Kriged maps provided detailed visualization of soil properties at landscape scale, and helped to identify critical areas for targeted land management interventions to improve land quality. The spatial distribution of selected soil properties showed a well-defined pattern of higher concentrations in the lowlands and valleys and areas with permanent vegetation cover in both sites.. These results are useful for improving the efficient use of inputs such fertilizers. Context-specific land management based on spatial variability of soil properties is highly recommended and more research is required to generalize our knowledge about spatial variability of soil health indicators and the casual factors in the highlands of Cameroon. ACCEPTED MANUSCRIPT Highlights  Geostatistics was used to analyze spatial structures of soil quality parameters  Clay content has a negative effect on soil properties  Land use types have significant effects on spatial patterns of soil properties  Soil variability was due to inherent conditions, land use types and management  Nutrient spatial patterns are useful for site specific land management practices

ABSTRACT A field experiment was conducted in the Western Highlands of Cameroon to study the effec... more ABSTRACT A field experiment was conducted in the Western Highlands of Cameroon to study the effects of Calliandra calothyrsus, sterilized compost, non-sterilized compost and mineral fertilizers (NPK 11:11:22) on physical characteristics, yield components and late blight disease severity of potato (Solanum tuberosum). A two factorial treatment combination made up of fertilization schemes and sanitary measures were laid out in a randomized complete block design (RCBD) with four replicates and 10 treatments. Data collected were subjected to a multivariate ANOVA and means separated with the Dunnett t-test with Calliandra calothyrsus considered as the main treatment. Mineral fertilizers and Calliandra calothyrsus significantly augmented stem diameter, plant height and plant vigor, total and marketable yields, despite the fact that late blight severity was high in the two treatments. The correlation matrix showed that total yield had significant and positive correlation with stem diameter (r = 0.74), plant height (r = 0.61), plant vigor (r = 0.61) and marketable yield (r = 0.99) and negative correlation (r = -37) with late blight severity. The increase observed is attributed to the nutrient contents of the treatments applied. Leafy prunings of C. calothyrsus increase soil organic matter which probably improves moisture retention and biological activity. The highest late blight severity recorded between the 45 and 60 coincides with active development of plant canopy which influences disease severity and create a canopy microclimate conducive to disease development. This study suggests that Calliandra calothyrsus has the potential of improving potato production provided appropriate fungicide is used.

Journal of Agricultural Science, 2012

Collective action has been promoted as a strategy to improve the incomes of small scale producers... more Collective action has been promoted as a strategy to improve the incomes of small scale producers in many developing countries, primarily by reducing transaction costs and improving farmers' bargaining power. This paper evaluates producers' perception of collective action initiatives which have been used in Cameroon for the production and marketing of kola nuts. Using a perceptual evaluation and cluster analysis based on interviews with 203 kola producers in western highlands of Cameroon, findings reveal that producers evaluate effectiveness of collective action based on 5 main groups of criteria: reduced transaction costs, new learning and skills acquisition, market and financial status, social networks and status, and unfavorable dimensions. Furthermore, two main types of producers who evaluate the collective initiatives differently are identified: the positive group and the unimpressed group. Higher proportions of women and youth are in the unimpressed group than in the positive group, suggesting that more needs to be done to understand their perspectives and better target collective action initiatives to their needs and circumstances. The paper concludes that promoters of collective action initiatives need to adopt differentiated strategies to enhance its adoption in the study area.

Small-scale Forestry, 2013

Your article is protected by copyright and all rights are held exclusively by Steve Harrison, Joh... more Your article is protected by copyright and all rights are held exclusively by Steve Harrison, John Herbohn. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com".

Small-scale Forestry, 2012

Availability of high quality tree planting material within proximity of farmers and at affordable... more Availability of high quality tree planting material within proximity of farmers and at affordable prices is one of the prerequisites for larger uptake of tree cultivation. This study examines whether rural small-scale nurseries can produce a diversity of tree planting material and whether resource-poor farmers have access to it. Twelve nurseries supported by the tree domestication program in the West and Northwest regions of Cameroon were compared to 12 nurseries in similar conditions, but not in contact with the program. Nurseries using the domestication approach were found to provide tree planting material that responds better to farmers' needs in terms of quantities, species and propagation methods used. Their clientele is more diverse including farmers from the communities where the nurseries are located, but also from far beyond. However, prices of vegetatively propagated material were considered the most prohibitive factor. It is concluded that tree planting initiatives should refocus efforts towards technical training and business support to small-scale nurseries to increase efficiency. Research efforts should look for ways of reducing production costs and improving nursery productivity.

The Journal of Agricultural Education and Extension, 2014

ABSTRACT Purpose: Formal agricultural research has generated vast amount of knowledge and fundame... more ABSTRACT Purpose: Formal agricultural research has generated vast amount of knowledge and fundamental insights on land management, but their low adoption has been attributed to the use of public extension approach. This research aims to address whether and how full participation of farmers through the concept of Rural Resource Centre (RRC) provides new insights for the development of alternative and farmers-based extension methods. Design/Methodology/Approach: Using the Concept of RRC, this research assesses the role of farmers in on-farm demonstrations and scaling-up of land management practices, and investigates effective ways to enhance beneficial interactions between researchers, extension workers and farmers in view of improving adoption. Findings: The findings suggest that farmers can effectively participate in demonstrations and scaling-up of agricultural practices. This participation is enhanced by judicious incentives such as higher crop yields that motivate farmers and influence adoption. The current success of the approach stems from the fact that farmers, extension workers and researchers jointly implement the activities and their different aims were achieved simultaneously: scientific results for researchers, better agricultural practices for extension workers, and economic success and free choice for farmers. Practical implications: This research concludes that farmers have the capacities to play an innermost role in demonstrations and scaling-up of agricultural practices. However, there is a need to build and strengthen their capacities to facilitate their participation and contribution. Originality/Value: The article demonstrates the value of the preponderant role of farmers in onfarm demonstrations and scaling-up practices by exhibiting the beneficial interactions between researchers, extension workers and farmers.

International Journal of Agricultural Sustainability, 2011

The restoration of degraded lands has received increased attention in recent years and many commi... more The restoration of degraded lands has received increased attention in recent years and many commitments have been made as part of global and regional restoration initiatives. Well-informed policy decisions that support land restoration, require spatially explicit information on restoration potentials to guide the design and implementation of restoration interventions in the context of limited resources. This study assessed ecosystems indicators of land degradation using a systematic approach that combines field surveys and remote sensing data into a set of multicriteria analyses to map restoration potentials in the semi-arid areas. The indicators considered were soil organic carbon, erosion prevalence, enhanced vegetation index, Normalized differences water index and the Net Primary productivity. Three classes of restoration potential were established: (1) areas not in need of immediate restoration due low degradation status, (2) areas with high potential for restoration with moderate efforts required and (3) areas in critical need of restoration and require high level of efforts. Of the total area of the study site estimated at 88,344 km 2 , 59,146.12 km 2 , or 66.94% of the theoretically recoverable area, was considered suitable for restoration, of which 38% required moderate efforts while 28% require less efforts. The recoverable areas suitable for restoration could be restored through tree planting, soil and water conservation practices, farmers managed natural regeneration, and integrated soil fertility management. These results can help to spatially identify suitable multifunctional restoration and regeneration hotspots as an efficient way to prioritize restoration interventions in the context of limited resources.

Digital soil mapping (DSM) approaches provide soil information by utilising the relationship betw... more Digital soil mapping (DSM) approaches provide soil information by utilising the relationship between soil properties and environmental variables. Calibration of DSM models requires measurements that may often have substantial measurement errors which propagate to the DSM outputs and need to be