Environmental Cognitions, Land Change and Social-Ecological Feedbacks: Local Case Studies of Forest Transition in Vietnam

Download Environmental Cognitions, Land Change and Social-Ecological Feedbacks: Local Case Studies of Forest Transition in Vietnam

Post on 23-Dec-2016




0 download


<ul><li><p>Environmental Cognitions, Land Change and Social-EcologicalFeedbacks: Local Case Studies of Forest Transition in Vietnam</p><p>Patrick Meyfroidt</p><p>Published online: 27 January 2013# Springer Science+Business Media New York 2013</p><p>Abstract Forest transition i.e., the shift from decreasingto expanding forest cover in the northern mountains ofVietnam was analyzed at the local scale in four villages fromthe 1970s to 20072008 to understand feedbacks from localenvironmental degradation on land uses, the conditions un-der which such feedbacks occur, and their possible roles inthe transition. Remote sensing data were combined with fieldsurveys including interviews, group discussions, mental andparticipatory mapping, observations and secondary sources.The feedbacks from environmental degradation and changesin the provision of ecosystem services to land practices viaenvironmental cognitions were analyzed. The case studiesshowed that forest scarcity was perceived, interpreted andevaluated before possibly affecting land use practices.</p><p>Keywords Deforestation . Reforestation . Ecosystemservices . Vietnam . Land use change . Shifting cultivation</p><p>Introduction</p><p>Explaining land use/cover change requires understanding thedynamic coupling between human societies and their ecosys-tems embedded in social-ecological systems (SES) (Lambin etal. 2003; Ostrom 2009). Human-induced environmentalchange can affect the production of ecosystem services andgoods, and in turn these changes can impact human behaviors.Land use systems undergo structural transformations throughthe interplay between local social-ecological feedbacks and</p><p>exogenous socioeconomic and policy changes (Lambin andMeyfroidt 2010). Tropical deforestation has long been viewedas a continuous linear process, but nonlinear dynamics inforest cover changes are increasingly studied. Vietnam hasrecently experienced a forest transition shift from shrinkingto expanding forests (Meyfroidt and Lambin 2008a, b). Hu-man responses to the decline of forests caused by humanactivities can drive forest transitions (Meyfroidt and Lambin2011), but this depends on perceptions of forest degradationand scarcity. Scarcity is neither exclusively a social constructnor a physical reality, but the product of human cognitiveprocesses triggered by environmental change. Understandingforest and land use transitions therefore requires analyzinghuman cognition, i.e., mental processes. This paper describesan empirical study of land changes, environmental cognitions,perceptions of forest degradation and feedbacks in four vil-lages of the northern mountains of Vietnam, to show howenvironmental and land changes were perceived, understoodand evaluated by local agents and affected their behavior andland use practices, and thus subsequent environmental andforest cover changes.</p><p>In land change science and human ecology, local percep-tions of environmental change have rarely been directlylinked with actual practices and behaviors. Yet local actorsoften perceive environmental services differently than landplanning authorities and react in unanticipated ways to newpolicies. Current models of decision-making and behavior inland change science, usually based on variants of rationalchoice theory or of heuristics, often assume mechanisticresponses to environmental signals and a stable set ofpreferences, beliefs and decision-rules (Meyfroidt 2011;Villamor et al. 2011). Although useful in stable or linearlychanging contexts, these perspectives are insufficient tounderstand how agents may modify their beliefs, attitudes,preferences or utility function, and/or heuristics to respondto qualitatively changing environments. Many studies ex-plore how local agents adapt to environmental change that</p><p>P. Meyfroidt (*)F.R.S.-FNRS, Louvain-La-Neuve, Belgiume-mail: patrick.meyfroidt@uclouvain.be</p><p>P. MeyfroidtGeorges Lematre Center for Earth and Climate Research,Earth and Life Institute, Universit catholique de Louvain,Louvain-La-Neuve, Belgium</p><p>Hum Ecol (2013) 41:367392DOI 10.1007/s10745-012-9560-x</p></li><li><p>they cannot mitigate typically climate change (e.g.,Mertz et al. 2009), but few focus on how local land managersrely on their own environmental perceptions, knowledge,evaluation and decision-making tools to directly act on thecauses of environmental change (Lauer and Aswani 2010).</p><p>Feedbacks from environmental change on cognitions can bedirect when they modify only the perceived state of theenvironment and parameters of the decision-making, or indi-rect or transformative when they modify beliefs, values,and/or decision-making strategies and possibly subsequentlysocial norms (Fig. 1; Meyfroidt 2011; Reed et al. 2010).Social-ecological feedbacks are mediated by the cognitiveprocesses of perception, interpretation and evaluation of envi-ronmental change (Meyfroidt 2011). The rate and saliency ofchange (Clitheroe et al. 1998, Dayton et al. 1998, Rogan et al.2005), traditional ecological knowledge (Berkes 2008) anddifferent forms of practical engagement with the environment(Ingold 2000) influences whether environmental change isperceived early enough to allow for effective reactions. Causalattribution of the change in particular attribution to naturalvariability versus anthropogenic causes depends on thebeliefs, worldviews and values of the agents, and influencesthe perceived severity of the change, responses seen as appro-priate and expected consequences (Brown et al. 2005;Meyfroidt 2011; Schad et al. 2011). Cognitions about theenvironment can also be acquired through social learning i.e., a change in understanding that is situated within wider</p><p>social units or communities of practice and operates throughsocial interactions (Reed et al. 2010). Evaluation of the impor-tance and valence of the resource and its change, influenced bybeliefs, values and emotions, will affect the commitment tomitigate or reverse these changes. Pro-environmental attitudesbased on utilitarian beliefs and values are associated withdistant and indirect behaviors such as policy support (Rauwaldand Moore 2002), while emotional cognitions such as a senseof connectedness to nature are associated with local and prac-tical actions (Hinds and Sparks 2008; Kaltenborn 1998;Stedman 2002). The perceived possibility to predict, controland improve the future state of resources affects decisions andcommitment (Armitage and Conner 2001; Bandura 1997).Without a belief in their ability to control environmentalevents, local agents will tend to assess changes as beingirreversibly engaged. This study explores four hypothesesabout the role of local-scale cognitions on responses to envi-ronmental degradation: (i) early perception of environmentalchange is necessary to allow for effective reaction; (ii) thecausal attribution for environmental degradation will influencethe type of reactions selected; (iii) more highly valued resour-ces will elicit more committed responses to degradation; and(iv) a lower perception of self or group efficacy will limitreactions to mitigate environmental change.</p><p>A national-scale statistical analysis of forest transition inVietnam (Meyfroidt and Lambin 2008b) showed that dis-tricts with low forest cover at the turning point had higher</p><p>Fig. 1 Conceptual model of environmental feedbacks via cognitions</p><p>368 Hum Ecol (2013) 41:367392</p></li><li><p>rates of reforestation, suggesting that local forest scarcitycould motivate investments by local landowners in forestconservation, tree planting and more intensive forest man-agement by driving up timber prices (Hyde et al. 1996), and/or by increasing awareness of adverse effects of deforestationsuch as floods and land degradation. To date, the latter hasonly been shown to influence national level policies (Durst etal. 2001; Lambin and Meyfroidt 2010; Mather et al. 1999;Mather and Fairbairn 2000; Wenhua 2004). Possible mitiga-tion responses by local actors remain unclear. Profoundsocioeconomic and political changes affected the uplandsof Vietnam during the last decades (Meyfroidt and Lambin2008b). Cooperative agriculture of paddy lands, establishedduring the 1950s1960s, was progressively dismantled dur-ing the 1980s with the privatization of agricultural produc-tion and the liberalization of markets. Introduced in 1991,land zoning defined three types of forestry land defined asland covered by forest or planned for forestry uses: forestsfor protection of watersheds and from land degradation;production forests; and special-use forests of high biologicalor cultural value. Through the 1993 Land Law, forestry landswere allocated to households along with rights and dutiesdepending on the category of forest and actual vegetation(MARD 2003). The government also initiated reforestationprojects by providing seedlings and labor fees to participat-ing households. In the study areas these policies constitutethe main exogenous changes that interacted with local per-ceptions of environmental change to cause locally specificchanges in forest cover.</p><p>In the four village case studies we quantitatively measuredand linked the three components of the feedback loop: (i)environmental changes, (ii) cognitive features that are part ofthe perception, interpretation, evaluation and decision-makingsteps, and (iii) the actual behaviors and land use practices(Fig. 1; Meyfroidt 2011). The two objectives were to explorethe four hypotheses proposed above and the further hypothesisthat forest scarcity may contribute to a forest transition byinfluencing local land uses. We discuss the conditions underwhich this takes place, and, although not the primary focus ofthe paper, the main outcomes of the forest transition for locallivelihoods in the four study villages.</p><p>Materials and Methods</p><p>Overview</p><p>Different methods were combined to analyze the three com-ponents of the feedback loop (Fig. 1). Forest-cover changeswere measured by remote sensing, validated by fieldwork.Environmental cognitions of the local land managers wereassessed through several types of interviews and participa-tory approaches and mappings. Land use practices were</p><p>reconstructed using participatory mapping, remote sensing,interviews, and land use statistics. Additional data includingland tenure and land zoning maps and histories of land usepolicies were collected. Then, process-tracing analysis wasused to describe the detailed steps of the causal chains of thefeedback loop from environmental cognitions to outcomesin three variables transition in forest area, in forestdensity, and in the satisfaction about forests by villagers(George and Bennett 2004). Analysing the spatio-temporalpatterns of the land use/cover changes and land zoningallowed validating the reforestation processes described bylocal informants. Alternative hypotheses were discussed.Cognitions and decisions vary among agents, but the objec-tive was to reconstruct the main social-ecological dynamicsin each village as a unit of analysis. Disagreements ordivergences possibly affecting these conclusions wereassessed. Sensitive issues, including land use policies, tim-ber harvesting, and possible illegal activities, wereaddressed by cross-checking and confronting data fromdifferent sources. For full methodological details and addi-tional results, see Meyfroidt (2009).</p><p>Selection of Study Sites</p><p>The four villages were selected to allow for within-casestudy and controlled comparison (George and Bennett2004). To study the role of environmental cognitions in landuse transitions, the communes were chosen to control themain exogenous forces affecting both forest cover and en-vironmental cognitions of local actors. Dynamics of dea-grarianization (Rigg 2006) and outmigration (Phan andCoxhead 2010) occurred over the recent decades in manyagricultural regions of Southeast Asia, and in some contextssimilar dynamics have been shown to lead to land abandon-ment and reforestation (Meyfroidt and Lambin 2011). Thus,we focused on places that were poor, remote and with anagriculture-based economy, excluding the areas with largedeagrarianization dynamics and/or agricultural intensifica-tion previously shown to allow for smooth setting aside offorestry land (Sikor 2001, Castella and Quang 2002), as wellas areas with large-scale reforestation programs. The objec-tive was not to find places that were entirely preserved fromexogenous influences, but controlling these allowed easierassessment and separation of the respective influences oflocal social-ecological feedbacks.</p><p>In each of two communes selected in two differentregions of Vietnam we chose two villages,1 one in a valley,populated by an ethnic group traditionally combining paddyrice with shifting cultivation (Tran 2003), and the other inuplands and populated by an ethnic group traditionally</p><p>1 The four villages were Khang and Na Da in Xuan Lac commune, andHoc and Pa Dong in Ta Hoc commune.</p><p>Hum Ecol (2013) 41:367392 369</p></li><li><p>Fig. 2 Location of study sites, overlayed on Landsat infrared color composites</p><p>370 Hum Ecol (2013) 41:367392</p></li><li><p>relying mostly on shifting cultivation. This allowed under-standing the respective roles of the administrative, biophys-ical and economic environments. Finally, the selectedcommunes were thought to have experienced a forest tran-sition, so that effects of forest decline on environmentalperceptions and responses of local agents could both beanalyzed. The study sites were therefore not representativeof the average Vietnamese upland commune and we do notclaim that it is possible to extrapolate our observations forthe whole of Vietnam. As described above, strong politicaland socioeconomic forces were important causes of theforest transition (Meyfroidt and Lambin 2008b), but in thisanalysis they are treated only as control variables. Consid-ering these criteria, the study sites were selected by prelim-inary surveys, with collaboration from the provincial anddistrict administrations. Xuan Lac is an isolated commune inthe north of Cho Don district in Bac Kan province innortheastern Vietnam (Fig. 2; Tables 1 and 2). Khang is aTay village in a small valley bottom occupied by paddyland(Fig. 3a). Tay farmers traditionally practice a compositeagricultural system with paddy rice as the cornerstone, com-plemented by shifting cultivation of upland rice and maizewith short fallows (Castella and Erout 2002). Na Da is a Daovillage upstream of the main valley of Xuan Lac commune(Fig. 3b), where farmers traditionally practice upland ricecultivation in short cycles of 23 years followed by longfallows, typically 20 years (Mellac 2000). Ta Hoc is locatedin Mai Son district in Son La province, in the mountainsbordering the Da River. The commune was affected byseveral land zoning and reforestation projects specificallyaimed at protecting the watershed of the Hoa Binh dam, themost important hydroelectric facility in Vietnam, locateddownstream on the Da river. Situated along a deep valleyat the confluence with the Da river (Fig. 3c), Hoc is occu-pied by Black Thai farmers who traditionally combine pad-dy rice farming in the valley bottoms with...</p></li></ul>


View more >