1 Introduction

Land degradation (LD)¹ is a complex phenomenon triggered by the interplay of environmental, economic and social factors (Warren, 2002; Geist and Lambin, 2004*; Reynolds et al., 2007). It is reaching a significant dimension especially in rural areas of developing countries where its effects are more severe (Eswaran et al., 2001*; Safriel, 2007*; Bai et al., 2008*). Research in developing regions has been substantial during recent decades (e.g., Thapa and Weber, 1991; Oldeman, 1998; Mertz et al., 2009*; Kondolf and Podolak, 2014; Vu et al., 2014*). LD patterns have been also studied at the local scale (Douglas, 2006; Waswa et al., 2013; Dahal et al., 2014), and causative (proximate and underlying) forces are receiving increased attention (Geist and Lambin, 2001*; Schreiber et al., 2012*; Agyemang, 2012*; Bisaro et al., 2014*). However, the contribution of the social perception of these forces at the local level has not been sufficiently emphasized (Kiage, 2013*).

Underlying driving forces are decisive; in some instances, proximate causes may be merely symptoms of underlying causes. For instance, poverty does not have a linear relationship with degradation, but in marginal lands, and where natural resources are scarce, it may be an underlying force (Gisladottir and Stocking, 2005*). Both the causes and the effects of degradation are time- and site-dependent (Santibañez and Santibañez, 2007; Mertz et al., 2009). Thus, it is desirable to understand the driving forces that are decisive at the local level, particularly as perceived by local agricultural producers (Gisladottir and Stocking, 2005; Verstraete et al., 2008; Maitima et al., 2009*; Kiage, 2013). Assessment of LD at the local level may help to mitigate its development in a given rural area. Blaikie and Brookfield (1987) suggested that public policy and socio-economic and cultural contexts decisively influenced degradation; only during the past decade has this been widely accepted. The Millennium Ecosystem Assessment (MEA, 2005*, p. 9) concluded that policies leading to unsustainable resource use, and lack of supportive infrastructure, are major contributors to degradation, yet both the academic and the technical communities have been reluctant to accept this; the literature reveals an emphasis on rather technical aspects (e.g., Safriel, 2007), biophysical indicators (e.g., Sankhayan et al., 2003*) and measurements of degradation patterns through remotely sensed data (e.g., Bai et al., 2008*). Approaches have been more descriptive of processes than explanatory of causality; top-down projects dominated, and a gap developed between science and successful decision-making. Programs addressing land conservation are not succeeding where they are most needed. Understanding, preventing and mitigating land degradation at the local scale seem to require more than technical knowledge and perception by external agents such as agricultural advisors and government officials, an issue also raised by (Hammad and Børresen, 2006*).

Hence, two questions may require research attention. Should local community perception feature strongly in technical assessments of land degradation? What factors influence farmers’ attitudes to coping with it? In this paper we explore the relevance of peasant perceptions of causes and implications of land degradation processes in rural areas of developing countries. The main purpose is to identify the factors determining farmers’ decisions to adopt land conservation practices in the local context. We argue that peasant decision-making procedures are strongly based on their perceptions of the forces that drive degradation. Moreover, if perception were taken in conjunction with technical evaluation to construct a hybrid vision, particularly in the field, simple solutions to these complex problems would be feasible. We believe that the absence of this conjunction contributes to the failure of top-down approaches to land conservation. First, we summarize and rank prominent driving forces in LD at the local level. Next, we discuss how local perception and traditional knowledge (TK),² including local indicators, have been addressed in published studies. Finally, we inspect the attitudes and strategies to cope with degradation from the perspective of local communities as reported in the scientific literature.