To what extent does surrounding landscape explain stand-level occurrence of conservation-relevant species in fragmented boreal and hemi-boreal forest? - a systematic review.

Background: Forestry and land-use change are leading causes of habitat loss, degradation, and fragmentation worldwide. The boreal forest biome is no exception, and only a small proportion of this forest type remains intact. Since forestry will remain a major land-use in this region, measures must be taken to ensure forest dependent biodiversity. Stand level features and structures promoting conservation relevant species have received much attention, but the landscape level perspective is often missing. Hence, we review the literature that has related fragmentation in the surrounding landscape to occurrence of threatened, declining, red-listed, rare, or deadwood dependent species as well as those considered to be indicator, flagship, umbrella, and/or keystone species in a given boreal forest stand.

Methods: A comprehensive search string was developed, benchmarked, and adapted for four bibliographic databases, two search engines, and 37 specialist websites. The online evidence synthesis tool Cadima was used for screening of both abstracts and full texts. All articles meeting the inclusion criteria were subject to study validity assessment and included in a narrative table. Studies reporting means and variance were included in quantitative meta-analysis when more than 3 comparable studies were available.

Results: The searches resulted in 20 890 unique articles that were reduced to 172 studies from 153 articles. These studies related stand level presence, abundance, species richness, and/or composition of conservation relevant species to landscape factors such as: categorical fragmentation intensity (higher vs. lower), amount of habitat or non-habitat, distance to habitat, and/or habitat configuration, on scales ranging from tens to tens of thousands of ha. Forty-three studies were suitable for meta-analysis. These showed a significant negative effect of fragmentation on both presence and abundance of conservation relevant species, as well as a near significant trend for species richness. This was particularly clear when fragmentation was measured as distance to surrounding habitat for presence, and as habitat amount for abundance. The organism groups with the strongest support for a negative effect of fragmentation were wood fungi and birds.

Conclusion: As hypothesised, there is strong support for negative effects of fragmentation in boreal forest. These results emphasize the negative consequences of the intensive forestry and associated landscape transformation that has been the norm for the last century. We argue that this should have direct implications for policy makers to shift towards including a landscape perspective in all planning of harvesting, preserving, and restoring forest. In addition, we found that research effort has been very uneven between organism groups, that studies on landscape change over time were rare, and that many studies have not quantified the difference in fragmentation intensity among landscapes making it difficult to quantify the extent of the negative effect. One way forward would be to revisit the studies included here in to incorporate change over time, as well as a true quantification of landscape fragmentation. By doing so, the scale of the negative effects would be much better analysed, which would greatly assist conservation practitioners all throughout the boreal forest biome.