Do livelihood typologies influence local perceptions of forest cover change? Evidence from a tropical forested and non-forested rural landscape in western Uganda. (February 2017)

• Record Type: Journal Article
• Authors: Twongyirwe, Ronald; Bithell, Mike; Richards, Keith S.; Rees, W.G.
• Is Part Of: Journal of rural studies. Volume 50(2017)
• Page Start: 12
• Page End: 29
• Publication Date: 2017-02
• Subjects: Household demographics -- Time- and Labour-Budgets -- Energy use -- Cropping and livestock husbandry -- Agro-ecological zones -- Perceptions -- Albertine rift landscape; Sociology, Rural -- Periodicals; Country life -- Periodicals; Rural development -- Periodicals; Land use, Rural -- Planning -- Periodicals; Rural conditions -- Periodicals; Sociologie rurale -- Périodiques; Vie rurale -- Périodiques; Développement rural -- Périodiques; Sol, Utilisation agricole du -- Planification -- Périodiques; Conditions rurales -- Périodiques; Country life; Land use, Rural -- Planning; Rural conditions; Rural development; Sociology, Rural; Periodicals; 307.72
• DOI: 10.1016/j.jrurstud.2016.12.009
• Languages: English
• ISSNs: 0743-0167

Abstract

Abstract: Validation of scientific findings from satellite remote sensing against local ecological knowledge could make the interpretation of forest cover patterns more robust. In this paper, we examine local perceptions of forest cover change in parishes around Budongo and Bugoma for a 30-year period (1985–2014), compare the results with those obtained from remote sensing (Twongyirwe et al., 2015), and interrogate whether the perceptions could be related to livelihood typologies. First, we characterise household strategies for the entire landscape to place livelihood strategies of communities in deforestated areas in a broader local context. An in-depth questionnaire was administered to 706 households in 13 parishes situated in 4 Agro-Ecological Zones (AEZs). The data included household demographic characteristics, energy use, cropping and livestock husbandry, and seasonal time- and labour-budgets. Principal Components Analysis (PCA) and Cluster Analysis (CA) were employed to help identify dominant structures in the data. Secondly, the 375 households in 7 parishes around Budongo and Bugoma forests (part of the 706) responded to additional questions that sought their perceptions on the forest cover trend. The PCA results for the entire landscape show that significant variation amongst households is mainly related to the cultivation time input, on-farm income particularly from cropping activities, livestock husbandry, demographic characteristics, agricultural extension activities, and cultivation labour input. Hierarchical CA shows that households at the landscape level fall into about nine different types, with variation in spatial distribution. The analysis suggests that poor households do live near forested regions, and that the rural poor are more reliant on forest products than peri-urban populations. Regarding perceptions of forest cover change, the majority (70.1%: n = 375) of the respondents in the parishes think that there has been a decline in forest cover, and this percentage is larger than the percentage of non-respondents (18.9%), those that thought it had increased (5.6%), not changed (3.7%), and those that did not know (1.6%). In addition, perceptions on forest change were significantly related to the household livelihood typologies (X 2 = 623.4, df = 4, p = 0.000): respondents who perceived forest cover as having declined and those that provided no response belonged to cluster 2 ("low income mixed farming households"), which is also the dominant livelihood typology around these forests. While the data largely suggest that there is a remarkable agreement between remote sensing results and local knowledge on forest change, and that local people may play a big role in filling data gaps where a dearth of information is prevalent (or where remote sensing evidence is fuzzy), there is a clear signal that people in different social classes and age groups can have very different views on what the change in forest cover might be despite what the remote sensing data show. This might have policy implications if decision makers tend to come from the groups that are not likely to have perceived forest cover change, or base their judgement on views from certain social classes. This implies that it is important to have the remote sensing data available as a counter balance to local perception (and vice versa) and therefore these data should be considered concurrently. Highlights: Rural poor are mostly situated in regions where deforestation was widespread. Local ecological knowledge of forest cover change largely corroborates remote sensing data. People's perceptions of forest change are however strongly related to social class and age group.