Environmental conditions
In this experiment, we examined the initial physiological response of eight tree species and three provenances (climate analogues) in two contrasting environments with the aim to optimize seedling acclimation to current conditions in a context of forest assisted migration. As expected, cutting treatment had the most significant impact on the microclimatic environment surrounding seedlings, particularly for light availability and temperatures. Light availability in the shelterwood cut represented 51% of the patch clearcut light availability during the growing seasons of 2020 to 2022, based on mean maximum PAR. Under a similar 40% basal area removal in yellow birch – conifer stands in Quebec, Prévost (2008) found that light availability was 22% of the availability in the patch clearcut during the first post-cut year. In our experiment, less vigorous trees with an open crown and with lesser structural attributes could explain the higher shelterwood post-cut light transmission.
In the context of climate change and forest assisted migration, a higher minimum air temperature during the bud break period appears interesting to limit late frost damage on seedlings (e.g., Groot and Carlson 1996; Langvall and Örlander 2001; Silvestro et al. 2019; Mura et al. 2022). The presence of an overstory generally decreases air and soil temperatures during the day and increases night temperatures in comparison with open area like a clearcut (e.g., Childs and Flint 1987; Carlson and Groot 1997; Langvall and Örlander 2001; Likens et al. 2004). In our experiment, cutting affected mainly extreme monthly minimum, daily maximum, and extreme monthly maximum air temperatures, generally elevating the minimum and decreasing the maximum temperature in the shelterwood cut compared to the patch clearcut (Fig. 1). This difference looks effective for alleviating frost damage risk: northern red oaks were damaged by late frost in year 2 (June 2020) and year 3 (May 2021), but the damage mainly occurred in the patch clearcut where 86% of the seedlings were damaged vs. only 5% in the shelterwood cut (Parent 2022, unpublished data). Similarly, Groot and Carlson (1996) reported that nearly 75% of the white spruce seedlings in a clearcut had medium to heavy spring frost damage, while only 2% of the understory seedlings were frost damaged.
Concerning soil temperature, it was around 2°C lower in the shelterwood cut (14.7°C) than in the patch clearcut (16.6°C), between June and August. Similarly, Prévost (2008) found that soil temperature under uniform partial cuts removing 40–60% of the basal area was lower by 4–5°C (year 1 post-cut) to 1–2°C (year 5) than in a patch clearcut. Difference between the two post-cut contrasting environments with the years resulted from decrease of solar radiation reaching the soil surface associated with post-cut vegetation regrowth. Carlson and Groot (1997) also reported that the difference in soil temperature among overstory openings were greatest early in the season, and diminished as regeneration of shrubs, herbs, and grasses developed. Similarly, in our experiment conducted in stands growing on rich sites, cutting effect on soil temperature was detected only in the first three post-cut years. In the patch clearcut, rapid regrowth of the shrub layer and other competing species sheltering soil surface from solar radiation likely wipe out significant soil temperature difference with shelterwood cut from the year 4 (2022). Thus, all effects of cutting on microclimate reported here likely affected physiology and growth of planted seedlings.
Effects on seedling physiology
Contrary to our first hypothesis, cutting had little impact on short-term photosynthetic capacity (Amax) of seedlings, suggesting that shade created by the residual overstory was insufficient to affect this physiological trait for most of the studied species (5 out of 8). However, consistent higher specific leaf area in the shelterwood cut in comparison with the patch clearcut (Fig. 3) suggests adjustment in leaf morphology because the residual overstory did not provide optimal light conditions to maximize carbon gain over the whole growing season. Such partial overstory seems adequate to acclimate and establish seedlings by reducing microclimate extremes (e.g., late frost) or the occurrence of other damaging agents (e.g., pine weevil attack on white pine; Ostry et al. 2010). However, overstory closure (e.g., Prévost 2008) and the rapid development of understory competing vegetation (e.g., Paquette et al. 2006) in the next years could impair physiology, growth, and survival of least shade-tolerant species like pines and red oak (Hannah 1988; Shepperd et al. 2006; Prévosto et al. 2011). Removal of the protective shelterwood cover is usually performed 5–15 years after cutting (Prévost and DeBlois 2014), either totally, partially, or progressively (Day et al.2011; Cogliastro and Paquette 2012). Trends in Amax and specific leaf area (SLA) with the years can inform us about the evolution of light conditions and the relevance to open the overstory to maintain or increase physiological and growth performance, and ultimately to avoid seedling suppression and mortality. For example, to satisfy light requirements of underplanted red oak and black cherry and produce a significant number of competitively trees, Cogliastro and Paquette (2012) showed that double overstory thinning at year 3 and 7 was necessary. This highlights the importance to continue ecophysiological monitoring, particularly to determine the best timing to manage the shelterwood cover, especially for translocated species and provenances that might be more susceptible to extreme conditions.
For some species (3 out of 8), photosynthetic capacity was influenced by cutting. This was the case for the shade-intolerant black cherry where Amax was higher in the patch clearcut than in the shelterwood cut at year 4 (Fig. 2a), corroborating our first hypothesis. However, low initial Amax or delayed Amax difference between the two cutting treatments and increasing water use efficiency (WUE) with the years suggest that this displaced species experienced significant planting stress (Grossnickle 2005). Black cherry is a newcomer to the balsam fir – yellow birch bioclimatic domain, representing a riskier type of assisted migration, i.e., assisted species migration, and thus perhaps more susceptible to planting stress. Such a stress could arise from low initial root growth (Burdett 1990; Margolis and Brand 1990) and/or mycorrhizal development (Refsland et al. 2023), particularly in a stressful post-clearcut environment. Poor root system size and distribution generally affect seedlings water uptake and nutrition which are susceptible to decrease Amax (Burdett 1990; Grossnickle 2005). Moreover, the abnormally warm and dry spring of 2020 (year 2) coupled with simultaneous competing vegetation release treatment could have a direct negative impact on Amax and WUE. It is reported that a partial overstory is required for the optimal establishment of black cherry, but that more sunlight is necessary afterwards to ensure continued growth (Marquis 1990; Verheyen et al. 2007; Cogliastro and Paquette 2012). Once the critical establishment period done, the drought tolerance of black cherry (Gilman and Watson 1994) makes it a good candidate species in the context of adaptation to climate change. Black cherry was the only species in our experiment whose SLA consistently decreased with the years (Fig. 3a), likely because its rapid height growth allows it to rapidly escape from light interference created by competing vegetation. By contrast, Amax of white spruce was higher in the patch clearcut than in the shelterwood cut from the moment it was planted, decreasing and equalling the Amax of seedlings in the shelterwood cut by year 4 (Fig. 2b). High regrowth of competing vegetation, forcing shade foliage development could explain this result. Acclimation to shade in spruces is known to occur at the expense of the Amax (Grassi and Bagnaresi 2001; Dumais and Prévost 2008). In accordance with increasing shade, SLA of white spruce increased in year 4 (Fig. 3d), and Amax of red spruce also decline with the years (Fig. 2c). However, adjusting leaf area can allow spruce to limit the impact of decreasing light availability on its global carbon gain (Evans and Poorter 2001). Nevertheless, competing vegetation influence on physiological performance was reported in other enrichment planting studies including spruce species (e.g., Hébert et al. 2013; Dumais et al. 2020; Bourque et al. 2022). Finally, white cedar was the only species for which the Amax was globally higher in the patch clearcut than in the shelterwood cut (Fig. 2f). Conversely, we previously found that Amax of planted white cedar decreased with increasing basal area removal intensity because competing vegetation cover increased simultaneously (Bourque et al. 2022). The contrasting result of the present experiment could be explained by the release treatment applied at year 2, which was probably sufficient for cedar to avoid Amax decline during the observation period, contrary to spruce species. It is also possible that faster acclimation and higher height growth of cedar allowed it to free itself from competing vegetation regrowth.
In accordance with our second hypothesis, a major finding emerging from our experiment is the lower seedling water stress (higher daytime xylem water potential, Ψd) for the two southern climate analogues of deciduous species in the patch clearcut (Fig. 4a–c), where environment was generally warmer and dryer than in the shelterwood cut. That result suggests that southern provenances of black cherry, red oak and sugar maple could be more tolerant to drought than seedlings from local seed source and/or that current climate change started to affect local populations. This is contradictory to a greenhouse experiment using the same, unplanted, seedlings, where climate analogues of these species presented similar levels of tolerance to a reduction in water availability (Champagne et al. 2021). Long-term seedling physiological and growth monitoring in field conditions will be important to determine the real benefits of forest-assisted migration in changing climate, although our results suggest that assisted range expansion and assisted species migration can be undertaken without affecting the initial establishment of seedlings.
For mid-century and end-century analogues of the red oak, lower Ψd values were observed in the shelterwood cut than in the patch clearcut (Fig. 4b). Yet, no effect of cutting on air relative humidity and soil water content was detected (Table 2). So, lower Ψd values in the shelterwood cut suggest a limited water uptake in partial shade that could be associated to low root growth. Indeed, initial rapid height growth of sheltered deciduous tree seedlings can occur at the expense of the root system (Burger et al. 1996). In an underplanting red oak experiment, Bardon et al. (1999) reported that reduction in growth and survival of sheltered seedlings could be the results of unbalanced shoot-root ratio, competition for nutrients and other resources, net loss of daily carbon gain, or a combination of these factors. According to Parker and Dey (2008), regulation of the overstory density is determining for photosynthesis and both tree water status and water use efficiency of underplanted oak. It is also possible than southern red oak provenances used in our experiment (southern Quebec, Canada and Pennsylvania, USA) were less adapted to regenerate in the understory, unlike local oak. Finally, for white and red spruces, cutting effect on Ψd rapidly disappeared in year 4, indicating some acclimation to clearcut conditions. This could indicate that root system of these spruce species and provenances was sufficiently extended after four years. Despite a negative impact on Amax, it is also possible that the shade from competing vegetation regrowth was beneficial to reduce water stress of the spruces during acclimation phase (e.g., Man and Greenway 2011; Dumais and Prévost 2016).