Combining spatiotemporal corridor design for reindeer migration with harvest scheduling in Northern Sweden

Abstract

Reindeer husbandry and commercial forestry seek to co-exist in the forests of Northern Sweden. As interwoven as the two industries are, conflicts have arisen. Forest practices have reduced the distribution of lichen, the main winter diet for reindeer. Forest practices have also increased forest density, compromising the animals' ability to pass through forested areas on their migration routes. In an attempt to reduce impacts on reindeer husbandry, we present a spatially explicit harvest scheduling model that includes reindeer corridors with user-defined spatial characteristics. We illustrate the model in a case study and explore the relationship between timber revenues and the selection and maintenance of reindeer corridors. The corridors are not only to include sufficient lichen habitat, but they are also supposed to ensure access for reindeer by connecting lichen areas with linkages that allow unobstructed travel. Since harvest scheduling occurs over a planning horizon, the spatial configuration of corridors can change from one time period to the next in order to accommodate harvesting activities. Our results suggest that maintaining reindeer corridors in harvest scheduling can be done at minimal cost. Also, we conclude that including corridor constraints in the harvest scheduling model is critical to guarantee connectivity of reindeer pastures.

Optimizing the Geometry of Wildlife Corridors in Conservation Reserve Design

“…St John et al (2016) embedded the OCCA into a forest harvest scheduling model to select reindeer corridors in Sweden. The case study consisted of 3,823 management units (1,996 forested and 1,827 non-forested), and yielded 4,461 polygons and 120,572 gate pairs.…”

“…In order to test our approach, we created hypothetical landscapes by varying the values of these three spatial attributes using a Vorronoi Tesselation-based landscape generator called rlandscape by Passolt et al (2013). Table 3 provides three levels for each characteristic to be used in generating landscapes, based on seven realistic landscapes in FMOS 2014 (St John et al 2016). Each value has a small margin around it for ease of landscape generation.…”

“…The MIP start allowed CPLEX to find a reasonable solution (with a 3.42% optimality gap) within 30 hours. For more details, see St John et al (2016). The case study showed that (1) the OCCA can select corridors of maximum width or minimum length, and (2) it can be incorporated as a new functionality in large harvest scheduling models.…”

Protecting wildlife habitat in managed forest landscapes—How can network connectivity models help?

“…An alternative approach combines harvest planning and habitat connectivity models in a single optimization problem (St. John, Tóth, & Zabinsky, 2018; St. John et al, 2016; Yemshanov et al, 2020). St. John et al (2016) proposed a multitemporal MIP model for forest harvesting and protection of reindeer habitat and migration corridors. For each harvest planning period t , the model allocated a connected corridor of reindeer habitat while meeting the harvest objective.…”

“…This makes the joint habitat connectivity/harvest problem more difficult combinatorially than the replanning approach of Martin et al (2016). In fact, both St. John et al (2016) and Yemshanov et al (2020) required multistage warm start techniques to solve the problem.…”