In transition: post-abandonment landscapes in Trás-os-Montes, Portugal.

Financiador

Organização

Publicações

Trajectories and drivers signalling the end of agricultural abandonment in Trás-os-Montes, Portugal

Publication . Imbrechts, Lien; Azevedo, João; Verburg, Peter

Agricultural abandonment has given rise to novel landscape dynamics worldwide. This paper investigates abandonment and post-abandonment dynamics in continental Portugal as a hotspot of landscape change. We mapped the spatial patterns and drivers of recent (1995–2018) land use changes in a remote mountainous region as post-abandonment trajectories, based on detailed land use/land cover data made available by the Portuguese government. We showed that ‘Revegetation’ trajectories, indicative of agricultural abandonment, were still widespread between 1995 and 2007. However, between 2007 and 2018, the landscape was much more stable with ‘Return to agriculture’ as the dominant change trajectory. To understand what drives landscape changes after abandonment, we explored the influence of a wide range of potential biogeophysical and socio-economic drivers on the observed trajectories. We contrasted different landscape outcomes in binary logistic regression models with the potential underlying drivers as independent variables. The regressions revealed that the most significant determinants of these alternating dynamics are existing land use, climate, slope, protection regime and accessibility. The results of the regressions are at times counterintuitive and give important indications of the changing spatio-temporal scales at which these variables exert influence on the landscape outcomes. However, the regression models’ limited accuracies highlight the need for deeper investigation of the socio-economic and historic context of the observed changes. Improved understanding of the (drivers of) alternative dynamics following agricultural abandonment can help inform policy decisions regarding agriculture and cultural landscape preservation.

2024Artigo científico

Acesso aberto

Ver mais

Publication . Imbrechts, Lien; Azevedo, João C.; Dossche, Rebekka; Bürgi, Matthias; Verburg, Peter H.

Large-scale land use changes can lead to a fundamental reorganisation of the land and corresponding socioecological regime. These regime shifts are notably hard to detect, predict or model, often arising from unprecedented changes in technology, markets or policies. In a remote rural region in Portugal, we tracked the land use history in three parishes between 1899 and 2018, capturing multiple disruptive socioeconomic and political circumstances, to assess whether a regime shift was triggered. We used a causal historical approach that included the spatio-temporal mapping of LULC changes and a socioecological event timeline to track policy changes and other important events or circumstances. We used these, and other available (historic) literature, to contextualise local information provided in oral history interviews (OHI) that revealed land managers’ decision-making during the last 70 years. We found that during the Estado Novo dictatorship, productivism-based policies had a strong influence on the observed land use intensification, yet OHI revealed that the high level of self-sufficiency agriculture made alternative trajectories unlikely if not impossible. After the 1974 revolution, recalibration took place in the form of a rural exodus and associated land reorganisation, including tenure and production systems. Overall, semi-natural areas, which usually depend on grazing or pastoralism, strongly diminished, while natural and plantation forests expanded. Arable land areas remained relatively stable across the entire study period. Our study shows that land systems that have undergone disruptive changes in the past may continue to evolve discontinuously afterwards, without causing a socioecological regime shift or breaking path dependence.

2026Artigo científico

Acesso aberto

Ver mais

Entidade financiadora

Fundação para a Ciência e a Tecnologia

Número da atribuição

2022.12731.BD