Priority effects will impede range shifts of temperate tree species into the boreal forest

Temperate tree species are expected to expand their distribution into the boreal forest in response to climate change. However, it is becoming increasingly clear that many species will experience significant setbacks in capacity to migrate due to a series of unfavourable conditions impacting their recruitment success, and thus their ability to colonize new locations. We quantify the relative influence of a series of factors important for tree seedling recruitment at range margins: propagule dispersal, substrate favourability, and the influence of the local hetero-specific species canopy composition. We hypothesized that boreal trees are responsible for priority effects that influence the establishment of temperate tree species seedlings. To do so, we analyzed two recruitment stages (first-year seedlings and older seedlings) for seven tree species; Abies balsamea (ABBA), Acer rubrum (ACRU), Acer saccharum (ACSA), Betula papyrifera (BEPA), Betula alleghaniensis (BEAL), Populus tremuloides (POTR), and Fagus grandifolia (FAGR) commonly found within the temperate-boreal ecotone forests of northeastern North America. Overall, we found that boreal canopy trees influence the distribution of substrates, more specifically the occurrence of needle cover and decayed wood in recruitment plots. Interestingly, decayed wood was a poor substrate in almost all cases. This association between canopy and substrate led to highly unfavourable substrates that affected the seedling densities of all temperate tree species. In addition, we found that seedling dispersion was highly localized, where the mean dispersal distance of all trees occurred in close proximity of parent trees. Unfavourable substrates and limited MDD (mean dispersal distance) of trees due to resident boreal trees generate (strong) priority effects within the temperate-boreal ecotone. Together these conditions promise to cause significant lags in temperate tree species migration into the boreal forest in the future.

Methods
A grid of 20 m x 20 m quadrats were established at each site to map overstorey canopy trees. Each site varied in plot size, where the total area of mapped grids ranged from 200-320 m in width and 200-1000 m in plot length. All trees within the study grid had their coordinates mapped using Leica Flexline© TS06 and TCR805 Total Stations. Each tree with a diameter at breast height (DBH) ≥10.0 cm was mapped, identified by species, given a health status (living, dead, declining, fallen) and its DBH (1.3 m) measured. In the late spring of 2015, four recruitment plots (1 m2) were established within each of the 20 x 20 m grids at a 5 x 5 m spacing (3,424 total recruitment plots across all three sites). Recruitment was measured as stem tallies by species over two years (2015 and 2016) for two recruitment stages: (i) first-year seedlings (referred to as stage 1 seedlings); identified by the presences of cotyledons, absence of terminal bud scars, suppleness of the stem, and number of leaves, (ii) older seedlings (referred to as stage 2 seedlings); which were not first year and were <30.0 cm in height. We visually estimated the percent coverage (to the nearest 5%, total summing to 100%) of the recruitment plots by the following substrates: grass, leaf, needles, moss, decayed logs, solid logs, and bole cover from overstorey/sapling trees. In addition to the substrates, we also assessed the cover of deciduous plant cover (non-canopy tree species) within the recruitment plot.