Discussion

This study has provided a visually powerful and empirically derived analysis of the spatiotemporal variation in the suitable habitat range for white spruce under different climate conditions. Observed trends reveal the complex interplay of bioclimatic variables in shaping the future composition of boreal forest in the Ontario-Quebec bottleneck region. The results from this analysis are further elaborated and explained in the context of climate change in the following section.

Habitat Fragmentation

Increased number of patches, small class size, and lower cohesion indicate that white spruce stands had become more isolated and fragmented over the time period of 2001 to 2011. Although the ecological impacts of habitat fragmentation may be considerably less than direct habitat loss, it can still contribute to the lowering the overall quality of the remaining patches and undermine its ability to provide ecosystem services such as regulating climate and offering habitats to an array of fauna and flora. The decreasing trend is expected to continue, as Murray et al. suggest that white spruce will experience a great extent of range loss by 2080 (2017). Climate-induced fragmentation will likely couple with anthropogenic habitat fragmentation, such as the change in land use and deforestation, in altering the meta-population dynamics, gene flow, range continuity, and ecosystem integrity of boreal forests (Murray et al., 2017).

A slight increase in suitable habitats

under a small temperature rise

The reduction in suitable habitats in the long run and the northward advancement of suitable climate conditions

Heat and water stress are the major culprits of shrinking white spruce habitats under climate change (Martin et al., 2010; Martin et al., 2016; Stinziano and Way, 2014; D’Orangeville et al., 2016). The highest temperature associated with the optimal climate white spruce breeding zones for Ontario is around 19°C (Thomson et al., 2010). Comparing the favourable climate conditions for white spruce and the future climate projection data, the impacts of heat stress can be clearly demonstrated. Under the current climate conditions, only 11 grids exceeded the mean temperature for the hottest month, whereas in the RCP 8.5 scenario for 2050 and 2070, 26 and 189 grids surpassed the highest monthly temperature. Looking at a histogram of maximum temperature values at the know occurrences for white spruce, the most occurrences have a maximum temperature between 9.3-9.6°C (Chart. 2). In the 2050 RCP 8.5 scenario, however, the most occurrences are found around the 12.2-12.6°C range. Therefore, it is predicted that such most locations will not be suitable for white spruce in the future, so white spruce might not survive in most of its current range. This is in accordance with Martin et al finding that all needleleaf species within the genera Picea respond negatively to summers in the year before tree growth (2016).

What makes the condition for white spruce more precarious is that white spruce is extremely intolerant to fire. The changing fire regime and the heightened chance of forest fire as a result of a warmer and drier climate pose a challenging threat to the survival of white spruce.

Moreover, precipitation indexes drop noticeably in different future climate conditions (an example is RCP 8.5 in 2050 in Figure). This is mainly due to the alteration of precipitation regimes, higher evapotranspiration rates, lowering of soil moisture, and trees' increased demand for water under higher temperature (Lloyd et al., 2013; Stinziano and Way, 2014). The decrease in precipitation is expected to be the prime reason for habitat decline (D’Orangeville et al., 2016). Thus, the results of this study raise a red flag immediately. The number of grids classified as suitable habitats for white spruce is summarized in Chart 3.

To adjust to the changing climate, tree species may migrate northward in search of favourable and habitable locations. According to D’Orangeville, Northeast North American would be the climate refugium for boreal forest (2016). They discover a divide in Ontario boreal forests’ responses to climate change along the latitude of 49°N. Forests north of 49°N are found to benefit from a warmer climate in terms of growth rates and growing season as the precipitation in these areas is anticipated to increase. MaxEnt results for the two 8.5 scenarios especially illustrate this divide as the narrow medium-low probability of suitable conditions is identified along latitude 49°N.

The maximum migration rates of tree species in Canada were on average 10 km per century since post-glacial (Price et al., 2013). Actual migration rate may be slower due to physical and climatic constraints. However, the flatter terrain characteristic of the Canadian boreal zone means that trees have to travel a longer distance in a shorter time frame compare to mountainous regions. Long distances pose a challenge on the seed dispersal ability of white spruce, whose almost all of its regeneration is from seed (Gartner et al., 2011). This may impede the natural colonization process of white spruce, and leading to the shrinkage of presence (Price et al., 2013), predicted in the MaxEnt models. It is even estimated that climate zones suitable for boreal conifers will disappear completely from the southern extents of their present distributions (and in the study area) by 2100 (Price et al., 2013).

Unchanging suitability around Lake Superior

It is interesting to note that there is expected to be a retainment of suitable habitat in close proximity to Lake Superior under all future projections. In particular, in the two RCP 8.5 scenarios, the small patch of the suitable locations are the only sites remaining (Figure 4). That area also has a higher score for suitability relative to other highlighted sites throughout difference scenarios under 2050. One explanation for that is the moderating ability of water on climate. Water has a high heat capacity at 4.18 J/g*C, which means it takes more heat to warm a water body than landmass. As a result, the temperature increase in the boreal forest around Great Lakes tends to be milder compared to inland boreal zones (Scott and huff, 1996).

The influence of Lake Superior, in particular, is enormous in terms of precipitation (Scott and huff, 1996) and this is evident in the bioclimatic variables. The large body of water releases water vapour into the atmosphere and forms precipitation that falls within the Great Lakes Basin. This allows its surrounding areas to withstand droughts and other water stressors.

In the example of 2050 RCP 8.5 scenario, the variations of all temperature (BIO 2,4,6) and precipitation (BIO 11, 13, 14, 15, 16) variables around Lake Superior are noticeably smaller than in other regions (highlighted in Figure 6). Despite the potential temperature rise in water temperature (Dietz & Bidwell, 2012), the temperature and precipitation stressors imposed on areas adjacent to Lake Superior will be smaller; thus, they remain to provide potential habitats for what white spruce.

Contrary to expectations, the extent of habitable areas for white spruce is estimated to expand to some extent under RCP scenarios 4.5 in 2050 and 2.6 in 2070 (Figure 5). The number of high suitability score (>0.5) in 2050 RCP 2.6 is almost comparable to that for the current baseline condition. A possible explanation for this phenomenon is the “fertilizer effect” of carbon dioxide CO2. Research has found that the increase in CO2 concentration in the atmosphere has a positive impact on the net primary productivity (NPP) in the Canadian boreal forest (). CO2 triggers RuBisCO carboxylation capacity (fixing of carbon in photosynthetic organisms by an enzyme) and can enhance photosynthesis rate; thus, CO2 increase can moderately promote plant growth in the short term (Martin et al., 2010; Stinziano and Way, 2014). Furthermore, warmer temperatures, especially in winter months, reduces total snowpack and lengthen growing season consequently (Goldblum and Rigg, 2005). However, the benefit of global warming has a threshold of 5 °C increase in growth temperatures (Stinziano and Way, 2014). Climate conditions exceeding this threshold is nevertheless detrimental to plant growth. This explains why there is an increase in range and then shark decrease after RCP 4.5.

Figure 5: Changing suitable climate conditions and habitat range under different climate scenarios.

Chart 2. BIO5 (Max Temperature of Warmest Month) values in for current climate condition and (left) and RCP 8.5 in 2050 (right).

Chart 3. Number of suitable habitats for white spruce under current and future climate conditions.

Figure 6. Maps of the 8 environmental variables under current climate conditions (left) and RCP 8.5 in 2050 (right)

Limitations

In this investigation, there are several sources of error.

Different types of white spruce might respond to climate change differently due to their distinct glacial origin and post-glacial migration routes. Research has found that white spruce in the northernmost margin of east Canada continue to experience northward migration, whereas their genetically distinct counterparts located in central Quebec have “reached their Holocene maxima” and are slowly replaced by black spruce (Lafontain et al., 2010: 741). The phylogeography of white spruce is a complex process is out of the scope of this research; hence, the different demographic trajectories were not taken into consideration.
Other biotic and abiotic factors are equally, if not more, important to the changes in suitability of boreal forest for white spruce in the age of climate change. These factors include, but not limited to, the alterations in competitive interactions, changes to pest distributions, dispersal barriers, and soil characteristics (Goldblum and Rigg, 2005; Martin et al., 2010).
MaxEnt model does not give an account as to why certain climate conditions or locations are not suitable given the bioclimate variables. For example, a higher temperature may make a location unsuitable due to the increased potential for forest fires, pest infestation, replacement of other species evapotranspiration, or the combination of them. Hence, further examination is needed to explain the interrelationships between bioclimatic variables and suitable conditions for white spruce. Also worthy to note is that MaxEnt only models and predict the degree of suitability of the study area. The actual occurrences of white spruce in 2050 and 2070 may differ from these predictions.