Results
A final cluster analysis was completed based on similar response coefficients from the response function analysis. The new response groups were then plotted to see where groups with similar limiting climatic variables exist on the landscape (Figure 10).
Figure 10. A map displaying a cluster analysis where response coefficients and external climate variables were grouped based on similarities. The response coefficients were calculated using a bootstrapped response function analysis in the Treeclim package in R Programming environment. The corresponding barplots show the positive or negative growth responses of each new response group according to average precipitation and temperature over a 16 months period: The first 5 months are the previous growing season, the next 6 months are the winter dormancy period, and the last 5 months are current growing season.
Response Groups 1 and 6, which generally occupy areas of interior Alaska and the Canadian boreal forest respectively, appeared to be the most precipitation limited groups. The results of this study show that these groups are particularly susceptible to drought, as they prefer colder temperatures and increased precipitation for most of the year. Similarly, Response Group 5 prefers cooler temperatures with higher precipitation in early spring with a consistent preference for cool temperatures for most of the summer. This group may also be negatively impacted by hotter, drier conditions related to climate change.
Response Group 2 in the coastal Alaska region responded positively to precipitation in May, but abruptly switched to prefer hotter, drier conditions in late spring to early summer. This followed a similar trend as Response Group 4, which also occupies northern coastal regions. Response Group 3 preferred drier, colder winters and did not seem to require as must precipitation in the growing season when compared to other response groups. Overall, these regions appear to be primarily limited by temperature as opposed to precipitation during the latter part of the growing season.
Response Groups 1 and 6, which generally occupy areas of interior Alaska and the Canadian boreal forest respectively, appeared to be the most precipitation limited groups. The results of this study show that these groups are particularly susceptible to drought, as they prefer colder temperatures and increased precipitation for most of the year. Similarly, Response Group 5 prefers cooler temperatures with higher precipitation in early spring with a consistent preference for cool temperatures for most of the summer. This group may also be negatively impacted by hotter, drier conditions related to climate change.
Response Group 2 in the coastal Alaska region responded positively to precipitation in May, but abruptly switched to prefer hotter, drier conditions in late spring to early summer. This followed a similar trend as Response Group 4, which also occupies northern coastal regions. Response Group 3 preferred drier, colder winters and did not seem to require as must precipitation in the growing season when compared to other response groups. Overall, these regions appear to be primarily limited by temperature as opposed to precipitation during the latter part of the growing season.
Discussion
The results of this study produced similar results to D’Orangeville et al. and Chen et al., where tree populations in western interior regions of North America are particularly vulnerable to increasing temperatures. In coastal areas that receive higher quantities of annual precipitation, we see a consistently positive growth response to increased temperatures for most of the growing season.
It is important to highlight that all of the response groups displayed a negative response to hotter temperatures in early spring. Spring snowmelt helps regulate soil water retention and soil temperature, but hotter conditions too early in the spring could reduce the soil moisture required for positive growth trends throughout the growing season. This remained a consistently vulnerable time of year for all response groups, which has implications for impeded growth and tree mortality if drought conditions become common in April to May. Areas within the western Canadian interior, which includes the boreal forest, are reported to already be experiencing a critical water-balance in warming conditions (Hogg and Bernier, 2005). These forests are particularly vulnerable to drought when compared to their eastern and western coastal counterparts as dry, prairie-like conditions expand northward (Hogg and Bernier, 2005). Previous tree-ring studies examining the impacts of climate variation on white spruce consistently found that populations in Western Canadian interior showed a positive growth response to increased precipitation (Hogg and Wein, 2005; Chhin et al., 2004). This was also exemplified by the results of this study, where negative growth responses were associated with increasing temperatures as opposed to increasing precipitation.
Further research could include a closer analysis of drought-tolerance and resistance according to stand age. Hember et al., (2016) discovered that water-stress tree mortality in North America has increased over the last six decades as a result of interannual variations and increasingly significant drought events, but the susceptibility to drought increased with the age of the tree. Perhaps this could help shed some insight on how to incorporate stand age into forest management models as a means of mitigating the effects of future climate change. A more robust analysis that incorporates other common tree species could also allow us to see the interplay occurring within forest communities to compare the susceptibility between mixed forests and conifer-dominant forests.
It is important to highlight that all of the response groups displayed a negative response to hotter temperatures in early spring. Spring snowmelt helps regulate soil water retention and soil temperature, but hotter conditions too early in the spring could reduce the soil moisture required for positive growth trends throughout the growing season. This remained a consistently vulnerable time of year for all response groups, which has implications for impeded growth and tree mortality if drought conditions become common in April to May. Areas within the western Canadian interior, which includes the boreal forest, are reported to already be experiencing a critical water-balance in warming conditions (Hogg and Bernier, 2005). These forests are particularly vulnerable to drought when compared to their eastern and western coastal counterparts as dry, prairie-like conditions expand northward (Hogg and Bernier, 2005). Previous tree-ring studies examining the impacts of climate variation on white spruce consistently found that populations in Western Canadian interior showed a positive growth response to increased precipitation (Hogg and Wein, 2005; Chhin et al., 2004). This was also exemplified by the results of this study, where negative growth responses were associated with increasing temperatures as opposed to increasing precipitation.
Further research could include a closer analysis of drought-tolerance and resistance according to stand age. Hember et al., (2016) discovered that water-stress tree mortality in North America has increased over the last six decades as a result of interannual variations and increasingly significant drought events, but the susceptibility to drought increased with the age of the tree. Perhaps this could help shed some insight on how to incorporate stand age into forest management models as a means of mitigating the effects of future climate change. A more robust analysis that incorporates other common tree species could also allow us to see the interplay occurring within forest communities to compare the susceptibility between mixed forests and conifer-dominant forests.