by André Boily | 30 January 2019
Published in PeerJ 4: e2218. https://doi.org/10.7717/peerj.2218
Projecting suitable conditions for a species as a function of future climate provides a reasonable, although admittedly imperfect, spatially explicit estimate of species vulnerability associated with climate change. Projections emphasizing range shifts at continental scale, however, can mask contrasting patterns at local or regional scale where management and policy decisions are made. Moreover, models usually show potential for areas to become climatically unsuitable, remain suitable, or become suitable for a particular species with climate change, but each of these outcomes raises markedly different ecological and management issues. Managing forest decline at sites where climatic stress is projected to increase is likely to be the most immediate challenge resulting from climate change. Here we assess habitat suitability with climate change for five dominant tree species of eastern North American forests, focusing on areas of greatest vulnerability (loss of suitability in the baseline range) in Quebec (Canada) rather than opportunities (increase in suitability). Results show that these species are at risk of maladaptation over a remarkably large proportion of their baseline range. Depending on species, 5-21% of currently climatically suitable habitats are projected to be at risk of becoming unsuitable. This suggests that species that have traditionally defined whole regional vegetation assemblages could become less adapted to these regions, with significant impact on ecosystems and forest economy. In spite of their well-recognised limitations and the uncertainty that remains, regionally-explicit risk assessment approaches remain one of the best options to convey that message and the need for climate policies and forest management adaptation strategies.
by Marie-Claude Boileau | 30 January 2019
Published in Forestry 90(1): 4-17. https://doi.org/10.1093/forestry/cpw024
Gap-based silvicultural systems were developed under the assumption that richness, and diversity of tree species and other biota positively respond to variation in size of harvest-created canopy gaps. However, varying gap size alone often does not meet diversity objectives and broader goals to address contemporary forest conditions. Recent research highlights the need to consider site factors and history, natural disturbance models, within-gap structure and recruitment requirements in addition to light resources for desired tree diversity. This synthesis brings together silvicultural developments and ecological literature on gap-based management, highlighting interactions with other factors such as microsite conditions, non-tree vegetation and more. We pose a revised concept for managers and researchers to use in prescriptions and studies focused on integrated overstory and understory manipulations that increase structural complexity within and around canopy openings.
by Claire Morin | 30 January 2019
Published in Forest Ecology and Management 430: 157-165. https://doi.org/10.1016/j.foreco.2018.08.007
Ecosystem based management in Québec is framed by reference conditions defining percentage of old-growth forest (>100-years-old) and forest composition characterizing pre-industrial forest landscapes. In the western spruce-moss bioclimatic subdomain (154 184 km2) a fire cycle estimated at 150 years was used to target that 49% of the landscape has to be composed of old-growth forest. Yet, this target was developed using past (19th–20th C.) climate and vegetation data and assume that environment and ecosystem processes are homogeneous for the entire western spruce-moss bioclimatic subdomain. The wide spatial and narrow temporal windows limit the application of reference conditions under ongoing climate change. Our aim was to classify current vegetation heterogeneity of the western spruce-moss subdomain into homogeneous zones and to study the long-term history of fire and vegetation within these zones. This approach will help to refine forest management targets that are based upon short-term records by providing a long-term perspective that is needed for the forests to be managed within their natural range of variability. Modern forest inventories data were used along with climate, physical variables, and natural and human disturbances to study the current vegetation-environment interactions among the western spruce-moss subdomain. We also used 18 published sedimentary pollen and charcoal series to reconstruct Holocene vegetation and Fire Return Intervals (FRI). Contemporary data revealed 4 zones with homogeneous interactions between vegetation and environment. Pollen analysis revealed three long-term vegetation paths: early successional species dominance, late to early species transition and late successional species dominance. These suggest that modern forest composition results from Holocene trajectories occurring within each zone. Holocene mean FRI (mFRI) ranged from 222 to 258 years across the subdomain, resulting in old-growth forests ranging between 64% and 68%, depending upon the zone. Paleoecological and contemporary results support that to make forest management more sustainable, current landscape heterogeneity that arises from millennial forest composition trajectories and fire cycle dynamics should be taken into account by down-scaling the previously established reference conditions.
by Marie-Claude Boileau | 30 January 2019
Published in New Forests 50(4): 677-698 https://doi.org/10.1007/s11056-018-9690-0
Vast areas of hardwood and mixedwood forests of eastern North America have been high-graded in the past and need silvicultural treatments to increase their value and productivity. To rehabilitate a high-graded temperate mixedwood stand, in Quebec, Canada, we used a split–split-split plot design with three replicates to assess different seed-tree and strip cutting systems, in combination with scarification and planting. The experiment consisted of three regeneration cuts in main plots: 10 seed-trees/ha, 40 seed-trees/ha and a 40-m wide strip clearcut (0 seed-tree/ha) with 60 seed-trees/ha in leave strips, thereby resulting in four levels of tree retention, and all included understory brushing. We applied two types of scarification (patch scarification or disk-trenching) to subplots, two regeneration modes (natural regeneration or planting with white spruce [Picea glauca]) to sub-subplots and two mechanical release treatments (softwood or mixedwood production) to planted sub–sub-subplots. Density of seed-trees did not affect the natural regeneration dynamics after 5 years, but disk-trenching was more efficient for the establishment of yellow birch (Betula alleghaniensis) and sugar maple (Acer saccharum). Few seed-trees of desirable white spruce were present and most died standing, confirming the importance of supplemental planting. Height growth of planted seedlings was 15% higher in the 0 and 10 (26–27 cm/year) than in the 40 and 60 (23 cm/year) seed-trees/ha treatments, and release doubled mean height growth (33.1 vs. 16.6 cm/year). Despite intensive site preparation, pre-established beaked hazel (Corylus cornuta) and mountain maple (Acer spicatum) were present at high densities in the regeneration stratum. Controlling this recalcitrant layer might be the greatest challenge for rehabilitating degraded stands of the mixedwood forest, especially since the use of herbicides is prohibited on Quebec’s public lands.
by Claire Morin | 30 January 2019
Published in Biological Conservation 195: 187-195. https://doi.org/10.1016/j.biocon.2015.12.029
For a given species, reducing the extent, connectivity and health status of source populations will diminish the probability of colonizing suitable habitats. In this study, we evaluated if metapopulation size helped predict species presence in suitable habitats, by using boreal understory plants and epiphytic lichens as case studies. We sampled a network of 792 vegetation plots located across a 240,000-km2 section of pristine boreal forests. Landscape-level species abundance (LLSA), a proxy for metapopulation size, was quantified in a 50 km radius surrounding each plot. The effects of local environmental factors and LLSA on species presence were modelled with general linear models (GLMs). Forty-eight vascular plant species and 8 pendulous epiphytic lichens were abundant enough to be considered in the statistical analyses. Habitat variables explained the presence of 30 vascular plants and 4 lichen species to a relatively strong degree (Tjur’s R2 > 0.15). Among these species, 16 vascular plants and 4 lichens were significantly influenced by LLSA, independently of local habitat characteristics. Vascular plant species whose presence was significantly affected by LLSA did not exhibit particular morphological traits or habitat preferences compared with those that were not affected. By contrast, the 4 epiphytic lichen species that were affected by LLSA were all associated with older forests. No species exhibited obvious metapopulation size thresholds below which the probability of species presence in suitable habitats dropped drastically. In terms of conservation, these results suggest that coarse-filter approaches are probably more robust and easy to implement than species-specific threshold approaches in this kind of ecosystem.
