Edge habitats resulting from the construction and maintenance of forest roads favour pioneer, shade-intolerant and disturbance-adapted plant species. The effect of roads on the spread of non-native species has been frequently studied, but few studies have focused upon their effects on native tree species. We studied the effect of forest roads on the expansion dynamics of trembling aspen (Populus tremuloides Michx.) in a boreal forest landscape of eastern Canada. We determined whether roads act as a habitat and dispersal corridor for trembling aspen, and whether populations that established along roads act as a starting point for aspen expansion into adjacent stands. We evaluated the effect of forest roads on the distribution of trembling aspen by surveying the vegetation along 694 km of roads. In 19 stands, we compared the density and age of individuals in 100 m transects established parallel and perpendicular to roads, to determine the role of roads. Trembling aspen is abundant along the forest road network. Forest roads act sometimes as habitat corridors for trembling aspen, but their effects on its density extend only over a short distance (10 m) on each side of the roads. The forest roads did not act as a starting point for the expansion of trembling aspen into adjacent stands. Forest roads are particularly favourable habitats for trembling aspen. Although roads did not act as a starting point for aspen dispersal away from roads, these habitats would be vulnerable to invasion following a disturbance that would reduce the thickness of the organic layer.

Postglacial vegetation colonization that followed ice retreat and proglacial lakes drainage in north-eastern America occurred rapidly, more rapidly than expected based on the modern dispersal capacities of boreal mixedwood trees. Paleo-islands from proglacial Lake Ojibway in Québec (Canada) were afforested early, before the final drainage of the lake. We hypothesized that these paleo-islands could have acted as outposts of migration and thus, could explain the rapid afforestation of lowlands observed. To determine if postglacial colonization occurred as a south-north front from the southern margin of Lake Ojibway or if islands acted as migration outposts, we estimated the date of first arrival of the main taxa of the current boreal mixedwood forest. We studied southern sites never covered by proglacial Lake Ojibway, sites that were islands within Lake Ojibway, and northern lowland sites that were liberated after the final drainage of proglacial Lake Ojibway. Taxa arrival was estimated as a sharp rise of the pollen percentage or as the occurrence of macro-remains within the sediments of small lakes dated with radiocarbon. Then we compared migration scenarios where colonization occurred gradually from south to north from the southern margin of proglacial Lake Ojibway and where paleo-islands of Lake Ojibway were first colonized through long-distance dispersal, thus becoming sources of seeds readily available to colonize lowland sites after the final drainage of Lake Ojibway. Finally, we compared the migration rates from the scenarios with the current mean dispersal capacities of the studied taxa. The migration rates estimated without taking the paleo-islands into account are too slow to explain the rapid afforestation observed following the final drainage of proglacial Lake Ojibway. Only the migration rates estimated from the scenarios with paleo-islands were comparable to the current mean dispersal capacity of the boreal mixedwood taxa. Thus, paleo-islands acted as stepping stones during postglacial migration, which explains why the lowlands were rapidly colonized. Larger paleo-islands and those located closer to the southern margin of the proglacial Lake were colonized first, in line with the theory of island biogeography.