Location relative to a retention patch affects the ECM fungal community more than patch size in the first season after timber harvesting on Vancouver Island, British Columbia

Silviculture systems that include retention of green trees are becoming more common in North America. The goals of green tree retention are to maintain forest structural diversity, preserve species associated with mature forests and to support faster post-harvest recovery of biodiversity. We studied the proportion of living roots and ectomycorrhizal (ECM) fungal communities in, and adjacent to, aggregated retention patches of coastal western hemlock forest on Vancouver Island, 4-6 months after harvest. Our objectives were to determine, for the window of time during which replanting typically occurs, (i) whether aggregated patches of green trees had retained ECM fungal communities similar to uncut forest and whether this depended on patch size; (ii) how far the influence of the patch extended into the harvested area, and whether this depended on patch size. These factors will influence the effectiveness of the aggregated patches as inoculum sources for seedlings planted in adjacent harvested areas. Soil samples were collected at the center and edge of 16 patches: four replicates each of 5, 10, 20, and 40m diameter patches, as well as at 10 and 20m into the harvested area around each patch. A control-forested area was also sampled. The state of the stele was used to designate 25 lateral roots from each sample as live or dead. One hundred active mycorrhizas per sample were then examined and described morphologically. The internal transcribed spacer region of the fungal rDNA was amplified and sequenced from representative tips of each morphotype. ECM communities were indistinguishable between uncut forest and the aggregated retention patches. This was true for patches as small as 5m in diameter, with no significant overall effect of patch size on ECM fungal species richness, Shannon Diversity Index, or the proportion of live root segments. Sampling location, however, significantly affected all these variables, with the influence of the patch disappearing by 10m into the harvested area. The only indication of a patch size effect was that ECM species richness at the edges of the 5m plots was slightly lower (P <0.1) than the edges of larger patch sizes. Based on these results, we recommend that patch sizes be at least 10m in diameter for coastal western hemlock forests. Since the edge:area ratio of smaller patches is higher, more small patches of at least 10m diameter would be more effective than a few large patches in supplying ECM inoculum to adjacent harvested areas during the first year after harvest.