International Mountain Conference

Abstract ID 117 | Date: 2022-09-13 13:45 – 14:00 | Type: Oral Presentation | Place: SOWI – Seminar room SR1 |

Pachoud, Carine (1,2)

Studies on sustainable transitions have multiplied in recent years, driven by the need for radical change in our practices and values to escape crises. Such radical transformations are situated in space. A spatial approach towards transformative processes can thus inform about the conditions and contexts of such dynamics. Recent publications on the geography of transition are tackling this issue, but still give little importance to the constructed and relational nature of place and the role of multi-scalar relationships. This contribution presents a conceptual framework for the socio-spatial analysis of transformation dynamics by adopting the French-speaking literature on territoire. This framework is applied to a case study in mountains, located in Gresse-en-Vercors, in the French Alps. We show that the concept of territoire appears highly relevant in the study of sustainability transitions, especially in mountains, because it allows to grasp three fundamental dimensions of transformation, i.e the material, institutional and ideal dimensions. This concept enables thus to study transformation dynamics in a systemic way. It also reveals the political dimension among actors involved in such dynamics, enabling researchers to grasp multiple and intertwined power relationships. Lastly, it integrates the influence of administrative levels in shaping transformation dynamics.

Abstract ID 868 | Date: 2022-09-13 14:00 – 14:15 | Type: Oral Presentation | Place: SOWI – Seminar room SR1 |

Black, Benjamin Samuel; Gao, Tian; Wicki, Sergio; Grêt-Regamey, Adrienne

Investigating pathways for the transformative adaptation of socio-ecological systems (SESs) frequently requires simulating the development of aspects of the system in both space and time, often into the future. In many cases this is achieved through the creation of predictive models which are calibrated and validated using historical data before being applied to generate future projections under the assumption of stationarity (i.e. that the relationships between the phenomena being modelled and its predictors are constant). Of course, this assumption is inherently flawed given that the calibration of models often highlights the presence of clear non-stationarity, for example between different historical periods, indicating systemic change within the SES.

Instead of disregarding non-stationarity this research seeks to demonstrate how characterising it can be used to improve understanding of how a given SES has, and is changing, and thus inform planning for deliberative interventions to encourage transformative adaptation.

Our research will demonstrate this in the context of modelling future Land Use and Land Cover (LULC) change scenarios for several alpine regional nature parks in Switzerland. Specifically, we use Random Forests supervised classification to statistically model the relationships between class-class LULC transitions and a wide set of environmental, socio-economic and neighbourhood predictor variables. Performing this modelling at a regional scale across multiple time periods allows for the identification of both spatial and temporal non-stationarity within the predictors of LULC transitions.

This non-stationarity is characterised through two approaches. Firstly, through changes in the inclusion of predictors in LULC transition models as a result of the feature selection process (filtering to produce the most parsimonious set of predictors by minimising redundancy). Secondly, through changes in the partial dependence plots which visualise the relationships between predictors and the probability of LULC transitions.

Such analysis provides insights into how the relationships between, socio-environmental factors and the likelihood of LULC transitions change over time and space, which should inform the development of scenarios of future LULC change for alpine parks in Switzerland. For example, comparing the degree of temporal non-stationarity exhibited across different LULC transitions, in combination with historical rates of LULC change (unit area per time period), could highlight which transitions are more ‘unstable’ and thus subject to greater uncertainty. This knowledge should be taken into consideration when devising scenarios based around interventions intended to elicit transformative change specifically related to these LULC transitions.

Abstract ID 655 | Date: 2022-09-13 14:15 – 14:30 | Type: Oral Presentation | Place: SOWI – Seminar room SR1 |

Stritih, Ana (1); Senf, Cornelius (1); Kuemmerle, Tobias (2,3); Bleyhl, Benjamin (2); Seidl, Rupert (1,4)

Mountain forests help mitigate climate change, provide a source of renewable energy and support the adaptation of mountain communities by providing ecosystem-based disaster risk reduction. While the demand for these ecosystem services is growing, forests’ capacity to provide them may be jeopardized by climate change and increasing disturbance rates. At the same time, forest landscape development is partly determined by legacies of past land-use, so the potential to steer development on the short term might be limited. In this study, we make use of recent developments in spaceborne lidar and forest disturbance mapping to investigate mountain forest structure and dynamics comparatively, at broad scale across the European Alps and the Caucasus Mountains. Both regions share similar natural vegetation types, a long history of human land use, and landscapes shaped by mountain agriculture. However, the development of both regions has diverged during the last centuries. The Alps experienced a rapid development of tourism, agricultural intensification and abandonment, and protection and expansion of forests. Today, these forests are experiencing an increasing frequency of natural disturbances. While the Caucasus has also experienced land abandonment and an increase of protection since regime changes in the 1990s, many communities still rely on traditional mountain agriculture and use forests as a source of firewood and for livestock grazing, sometimes leading to conflicts with conservationists in this biodiversity hotspot. We investigate how these land- and forest-use legacies influence today’s forest structure, and, in turn, forests’ capacity to provide ecosystem services. Our results show that across both mountain forest regions, forest structures converge to a similar distribution with two main basins of attraction – open and closed forests. The pathways between both states are shaped by natural disturbances and forest management, which are also influenced by past land use. Recognizing these path dependencies may help us understand the potential future trajectories of mountain landscapes and identify strategies to address the changing needs for forest ecosystem services under global change.

Abstract ID 508 | Date: 2022-09-13 14:30 – 14:45 | Type: Oral Presentation | Place: SOWI – Seminar room SR1 |

Dollinger, Christina Elisabeth; Seidl, Rupert; Rammer, Werner

The forests of the European Alps are expected to face major changes in the next decades due to the interplay of accelerated climate change and intensifying disturbance regimes. Particularly vulnerable are regions where past timber-oriented forest management has led to even-aged Norway Spruce (Picea abies) dominated forests. Areas where biodiversity conservation is of high priority, such as Berchtesgaden National Park (BGNP) in the German Alps, thus promote natural succession towards more natural forest conditions. In pursuit of this goal the park is running a restoration management program in the park’s management zone (~5000 ha, 25 % of the total park area).

We here analyse the long-term effects of different restoration strategies on forest composition and structure. Specifically, we contrasted two strategies which both focus on planting the currently underrepresented species Silver Fir (Abies alba) and European Beech (Fagus sylvatica). The more intensive “proactive” strategy introduces fir and beech in patches affected by natural disturbance and in canopy gaps artificially created by management, while the “reactive” strategy only replanted in gaps that occur naturally. To test the overall effectiveness of active management, a third “No Management”-strategy was considered. The strategies were implemented within the forest landscape model iLand and simulated over a period of 80 years (2020 – 2100) under past and future climate conditions (historical climate, RCP 2.6, RCP 4.5 and RCP 8.5). The model was evaluated with regard to its ability to reproduce tree regeneration densities observed in the field.

The forests of the management zone of BGNP developed from even-aged, strongly Spruce-dominated (73.2 % of all trees) forests to more mixed and structurally diverse forests under all evaluated management strategies, with the prevalence of Spruce decreasing by between 34 and 43 %. The “proactive” strategy was most successful in restoring species composition, while the “reactive” and “No Management”-strategies were best at increasing structural diversity. It remains unclear whether future forests will be resilient to the emerging environmental conditions, but younger and more diverse forests might have a better chance of adapting to change. In this study these indicators of resilience increased over time, regardless of human intervention.