Effects of Temperature and Light Conditions on Growth of Current-Year Seedlings of Warm-Temperate Evergreen Tree Species and Cool-Temperate Deciduous Tree Species

It is suggested that global warming affects plant distribution along latitudinal and altitudinal gradients because vegetation changes with thermal conditions. Simulation studies predicted that global warming largely affects plant distribution (e.g., Morin et al., 2008). Actually, vegetation change during several decades has been observed (Penuelas et al., 2007; Lenoir et al., 2008). By contrast, some other studies did not observe vegetation changes (Holtmeier & Broll, 2007; Harsch et al., 2009). Interpretation of results of simulation models also needs caution because simulation results are different according to modeling methods even for same species (Thuiller, 2003). Therefore, there is still uncertainty of effects of global warming on plant distribution. Plant distribution is determined by integrated demographic processes such as seed dispersal, seed germination, growth and survival of individual plants. Since early demographic phase such as seedling establishment is more susceptible to environmental conditions than the adult phase (Kullman, 2002), it is important to clarify effects of temperature on seedling growth to predict effects of global warming on plant distribution. There are many experimental studies that examined effects of temperature on growth of tree seedlings (Danby & Hik, 2007; Hoch & Korner, 2009; Munier et al., 2010). For example, Yin et al. (2008) reported that seedling growth of Betula albo-sinensis increased in the warm condition with 0.51oC higher than the ambient air condition. Many experimental studies that examined effects of temperature were conducted at bright conditions (e.g. Danby & Hik, 2007; Way & Sage, 2008). However, most seedlings distribute in dark closed-canopy conditions in forests. Therefore, it is necessary to examine effects of temperature on seedling growth not only in bright conditions but also in dark conditions. Plants plastically change morphology according to light conditions. For example, relative biomass allocation to leaves is greater in dark conditions than in bright conditions,


Introduction
It is suggested that global warming affects plant distribution along latitudinal and altitudinal gradients because vegetation changes with thermal conditions. Simulation studies predicted that global warming largely affects plant distribution (e.g., Morin et al., 2008). Actually, vegetation change during several decades has been observed (Penuelas et al., 2007;Lenoir et al., 2008). By contrast, some other studies did not observe vegetation changes (Holtmeier & Broll, 2007;Harsch et al., 2009). Interpretation of results of simulation models also needs caution because simulation results are different according to modeling methods even for same species (Thuiller, 2003). Therefore, there is still uncertainty of effects of global warming on plant distribution. Plant distribution is determined by integrated demographic processes such as seed dispersal, seed germination, growth and survival of individual plants. Since early demographic phase such as seedling establishment is more susceptible to environmental conditions than the adult phase (Kullman, 2002), it is important to clarify effects of temperature on seedling growth to predict effects of global warming on plant distribution. There are many experimental studies that examined effects of temperature on growth of tree seedlings (Danby & Hik, 2007;Hoch & Körner, 2009;Munier et al., 2010). For example, Yin et al. (2008) reported that seedling growth of Betula albo-sinensis increased in the warm condition with 0.51ºC higher than the ambient air condition. Many experimental studies that examined effects of temperature were conducted at bright conditions (e.g. Danby & Hik, 2007;Way & Sage, 2008). However, most seedlings distribute in dark closed-canopy conditions in forests. Therefore, it is necessary to examine effects of temperature on seedling growth not only in bright conditions but also in dark conditions. Plants plastically change morphology according to light conditions. For example, relative biomass allocation to leaves is greater in dark conditions than in bright conditions, accompanied with reduction of leaf mass per area (LMA) (Ellsworth & Reich, 1992;Gould, 1993;Niinemets et al., 1999;Takahashi et al., 2005). These morphological changes increase light capture per plant, which is an adaptive strategy in dark conditions. Thus, plants adapt to changing environments through morphological plasticity. This study compared growth responses to temperature and light conditions among three species with different climatically distribution ranges. This comparison is important to clarify effects of global warming on distribution shift of vegetation. Central Japan is a latitudinal vegetation ecotone between warm-temperate evergreen broadleaved forests and cool-temperate deciduous broad-leaved forests. Ishigami et al. (2003) simulated net primary production (NPP) of forests by the modified model of BIOME3 (Haxeltine & Prentice, 1996). They predicted that NPP of warm-temperate evergreen broadleaved forests is greater than that of cool-temperate deciduous broad-leaved forests above the current northern distribution limit of warm-temperate evergreen broad-leaved forests if air temperature is increased by global warming. According to the simulation result of NPP, Ishigami et al. (2003) suggested that the northern distribution limit of warm-temperate evergreen broad-leaved forests will move to the north. However, the northern distribution limit would not move to the north easily because of competition with existing northern vegetation (Kohyama & Shigesada, 1995). Therefore, it is important to compare growth responses of seedlings to temperature and light conditions between warm-temperate evergreen broad-leaved species and cool-temperate deciduous broad-leaved species to clarify vegetation changes due to global warming. It is considered that warm-temperate evergreen broad-leaved species cannot distribute in the cool-temperate zone due to low winter temperature, not due to low summer temperature for the growth (Kira, 1949). Cold winter temperature decreases photochemical efficiency of leaves in warm-temperate evergreen broad-leaved species (Aranda et al., 2005;Taneda & Tateno, 2005). Reduction of photochemical efficiency due to cold temperature is more conspicuous in bright conditions than dark conditions (Matsuki et al., 2003). Therefore, it is necessary to examine growth responses to temperature and light conditions not only during growth period but also during dormant period (winter) to clarify how temperature affects growth of warm-temperate evergreen broad-leaved species. This study examined effects of temperature and light conditions on seedling growth of a warm-temperate evergreen broad-leaved species and two cool-temperate deciduous broadleaved species to answer the following two questions.
(1) Do the warm-temperate evergreen and cool-temperate deciduous broad-leaved species grow more in higher temperature and light conditions? (2) Do high temperature conditions mitigate reduction of photochemical efficiency in winter for the warm-temperate evergreen broad-leaved species?  Horikawa (1972) and Okuyama (1982), respectively. Blue and green dots indicate locations of two study sites, Matsumoto and Nobeyama, respectively. temperatures in January and in August were -0.3 and 24.8˚C, respectively. The mean annual precipitation of Nobeyama was 1435.3 mm during 2000-2009. The mean annual temperature was 7.2˚C. The mean monthly temperatures in January and in August were -5.4 and 19.2˚C, respectively.

Study species
This study examined a warm-temperate evergreen broad-leaved species (Quercus myrsinaefolia Blume) and two cool-temperate deciduous broad-leaved species (Betula platyphylla var. japonica Hara and Quercus crispula Blume) (Fig. 1). Q. myrsinaefolia is a shadetolerant tall tree species that distributes in the northern part of warm-temperate zone (Horikawa, 1972). B. platyphylla var. japonica and Q. crispula are shade-intolerant and midshade-tolerant tall tree species, respectively (Samejima, 1979;Koike, 1988;Masaki et al., 1992). B. platyphylla var. japonica and Q. crispula often form pure stands after disturbances such as strong wind, forest fire and clear cutting (Samejima, 1979;Kamitani, 1993;Namikawa et al., 1997). Q. myrsinaefolia distributes below 700 m a.s.l. in the warm-temperate zone, and the northern distribution limit is about N38˚ (Horikawa, 1972). Distribution of Q. myrsinaefolia is limited to below 500 m a.s.l. in the north of N36˚. Although the southern part of Nagano Prefecture was the northern distribution limit of Q. myrsinaefolia (Horikawa, 1972), the distribution of Q. myrsinaefolia is recently observed in central part (Matsumoto area) of Nagano Prefecture probably because of escapes of seeds from planted Q. myrsinaefolia trees (Otsuka & Ozeki, 2008). B. platyphylla var. japonica distributes in central and northern Japan (Okuyama, 1981). B. platyphylla var. japonica distributes mainly at about 1000 m and 1300 m a.s.l. in Matsumoto and Nobeyama, respectively (Editorial Board of Flora of Nagano Prefecture, 1997). Q. crispula is one of the dominant species in cool-temperate deciduous broad-leaved forests, and widely distributes in the cool-temperate zone of Japan (Horikawa, 1972). Q. crispula distributes mainly at about 1000 m and 1300 m a.s.l. in Matsumoto and Nobeyama, respectively, like B. platyphylla var. japonica. Therefore, Matsumoto is the northern distribution limit for the warm-temperate evergreen species Q. myrsinaefolia, and Nobeyama is optimal thermal conditions for the two cool-temperate deciduous species B. platyphylla var. japonica and Q. crispula.

Seedling growth experiments
Growth experiments were conducted at Matsumoto and Nobeyama to examine effects of temperature and light conditions on the current-year seedling growth of Q. myrsinaefolia, B. platyphylla var. japonica and Q. crispula. A greenhouse experiment was also conducted at Matsumoto to make optimal temperature conditions for the warm-temperate evergreen Q. myrsinaefolia because Matsumoto is the northern distribution limit of this species. Windows of the greenhouse were opened in summer to avoid extreme rising of air temperature inside the greenhouse. Increase of air temperature in the greenhouse was 3˚C on the average. Therefore, air temperature increased in the order of Nobeyama, Matsumoto and the greenhouse at Matsumoto. Nobeyama, Matsumoto and the greenhouse at Matsumoto are referred to T1, T2 and T3, respectively, in this study. Two light conditions (20% and 54% light) were set at each temperature condition by using shade cloth. 20% and 54% light conditions are referred to L1 and L2, respectively. were sown again at T1 in early June because seeds did not germinate. Pots were watered once a week or according to the need during the experiment period. All pots were randomly moved within each treatment once a week to minimize effects of pot position on seedling growth. Seedlings of the three species were excavated and washed carefully in early October in 2009 after the cease of the current-year growth. In terms of Q. myrsinaefolia, only the half number of seedlings was excavated at each light and temperature condition, and the other seedlings were remained for the measurement of photochemical efficiency (see next paragraph). Stem diameter at the base and stem height were measured at the harvest for the three species. Seedlings were divided into stem, root and leaf, and then all leaves were scanned by using the free graphic software ImageJ (http://rsbweb.nih.gov/ij/) to measure total leaf area per seedling. Each organ was oven-dried at 80˚C for two days, and was weighed. The ratio of light-induced variable to maximum fluorescence of chlorophyll (Fv/Fm) of Q. myrsinaefolia was measured as a surrogate for photochemical efficiency of PSII (Demmig- Adams et al., 1989). Photochemical efficiency was measured around the noon once a month from summer of 2009 to the next summer by using a chlorophyll fluorometer OS-30p (Opti-Science, NH, USA). The measurement was conducted after 30 min of dark acclimation. Two-way ANOVA was done to examine effects of temperature and light conditions on stem height, stem diameter, LMA and seedling dry mass (including root) for the three species (Q. myrsinaefolia, B. platyphylla var. japonica and Q. crispula).

Growth of the three species
Effect of temperature on stem height was significant for the three species (Table 1), i.e., stem height was greater in higher temperature conditions, irrespective of light conditions (Fig. 2). However, stem height of Q. crispula was not different between T1 and T2 because of the shoot re-growth from the base after the shoot die-back in late May at T2. The difference in stem height of Q. myrsinaefolia between T2 and T3 was also not large (Fig. 2). Light conditions did not affect stem height for the three species (Table 1). By contrast, stem diameter of the three species was different not only among the three temperature conditions but also between the two light conditions (Table 1). Stem diameter of the three species was greater at L2 than L1 for each temperature condition (Fig. 3). Stem diameter was greater at T2 and T3 than T1 for Q. crispula. Stem diameter of B. platyphylla var. japonica increased in the order of T1, T2 and T3. However, stem diameter of Q. myrsinaefolia was slightly smaller at T3 than T2 for the two light conditions. Effects of temperature and light conditions on seedling dry mass were significant for the three species (Table 1), i.e., seedling dry mass was greater at higher temperature and brighter light conditions (Fig. 4). However, seedling dry masses of Q. myrsinaefolia at T3 were similar to and smaller than those at T2 for L1 and L2, respectively.
LMA was greater at L2 than L1 for the three species (Fig. 5, Table 1). LMA of Q. myrsinaefolia and Q. crispula was significantly different among the three temperature conditions (Table 1).
Although LMA of the two species was greater at T2 than T1 and T3, the difference was not large (Fig. 5).

Photochemical efficiency of Q. myrsinaefolia
Values of photochemical efficiency (Fv/Fm) differed, according to temperature and light conditions and season (Fig. 6). In August of 2009, Fv/Fm values were similar between the two light conditions for the three temperature conditions. Although Fv/Fm value was about 0.8 at T3 in August, Fv/Fm values slightly decreased as temperature decreased from T3 to T1. Fv/Fm value was about 0.7 at T1 in August of 2009. Fv/Fm value decreased from autumn to winter in each temperature condition. This tendency was more conspicuous at L2 than L1 in each temperature condition. In addition, Fv/Fm value decreased to almost zero in February at T1, and all seedlings died by March (Fig. 7). By contrast, although Fv/Fm values decreased in winter at T2 and T3, these values increased again from spring to summer (Fig. 6) and the seedlings grew well by autumn (Fig. 8).

Discussion
This study showed (1) that high temperature and light conditions increased the current-year seedling growth of the warm-temperate evergreen and the two cool-temperate deciduous species, except for the effect of light on stem height, and (2) that high temperature mitigated reduction of photochemical efficiency (Fv/Fm) of Q. myrsinaefolia in winter. Furthermore, winter temperature was so cold at T1 for Q. myrsinaefolia that all seedlings of this species died in winter. Therefore, this study showed that high temperature and light conditions increased the seedling growth of the three species and survival of Q. myrsinaefolia in winter. Fig. 7. A current-year seedling of Quercus myrsinaefolia at T1 with L1 light condition on February 16 and March 11, 2010. These photos were taken for a same seedling. Die-back was observed on March 11.
Light conditions did not affect stem height of the three species. Seiwa & Kikuzawa (1989) also reported that stem height of current-year seedlings was not different between bright and dark conditions for deciduous broad-leaved species with small seeds, while stem height was smaller in bright conditions than dark conditions for those of large seeds. Thus, stem height is not always taller in brighter conditions for current-year seedlings. This may reflect a survival strategy in forests. Litter accumulation on forest floor is a factor reducing seed germination and seedling survival (Goldberg & Werner, 1983). Therefore, the current-year seedlings germinated below litter layer have to grow over the litter layer. Furthermore, high growth of stem height is also advantageous for survival because of competition with other plants. Therefore, seedlings of the three species would preferentially grow stem height more than stem diameter in dark conditions as compared with bright conditions. LMA was smaller at darker conditions for the three species at the three temperature conditions, which is advantageous for light capture efficiency per unit leaf mass. LMA of Q. myrsinaefolia and Q. crispula was also affected by temperature. Although LMA of the two species tended to be greater at T2 than T1 and T3, these differences were not large. By contrast, Woodward (1979) observed that low temperature increased LMA of Phleum bertolonii and P. alpinum because of increase of cell diameter. Unfortunately, this study did not examine cell diameter of leaves at the three temperature conditions. However, it is possible that light dominates the control of LMA of the current-year seedlings of the three species, compared with temperature. Many previous studies about effects of temperatures on plant growth have been conducted so far in cold environments such as alpine region and tundra (Chapin & Shaver, 1985;Molau, 1997;Arft et al., 1999;Hollister & Webber, 2000). It is often reported that high temperature increases plant growth in such cold environments (Chapin & Shaver, 1996;Takahashi, 2005), meaning that cold temperature limits plant growth. Increase of temperature enhances photosynthetic rates in cold environments because of temperature dependency of photosynthetic rates (DeLucia & Smith, 1987). Furthermore, high temperature throughout a year expands the growth period of plants (Chmielewski & Rotzer, 2001). Nevertheless, seedling dry mass, stem height and diameter of Q. myrsinaefolia at T3 were similar to or smaller than those at T2 (the temperature conditions of the northern distribution limit). This suggests that temperature during the growth period is not a major factor limiting the growth of Q. myrsinaefolia. Northern distribution limits of warm-temperate evergreen broad-leaved forests tend to be determined by coldness in winter, not by warmth during the growth period (Kira, 1991). Thus, winter temperature may be important for the distribution of Q. myrsinaefolia along latitudinal gradients. Increase of temperature clearly mitigated reduction of photochemical efficiency (Fv/Fm) of Q. myrsinaefolia during winter. This tendency was more conspicuous at brighter conditions. In forests, seedlings hardly grow in understory dark conditions, and grow vigorously in canopy gaps (Takahashi et al., 2001;Takahashi & Rustandi, 2006). Therefore, it is suggested that increase of photochemical efficiency of Q. myrsinaefolia by increase of temperature would enhance the growth in bright conditions. This effect may be large in the northern distribution limit of Q. myrsinaefolia because Q. myrsinaefolia c o -d o m i n a t e s w i t h m a n y deciduous broad-leaved species there (Otsuka et al., 2004). Much light penetrates into forest floor in winter because of leaf fall of deciduous broad-leaved species. Understory seedlings and saplings of Q. myrsinaefolia can assimilate during winter after leaf fall of deciduous broad-leaved species, and the photosynthetic production during winter increases the growth and survival of Q. myrsinaefolia in the understory (Takenaka, 1986). Therefore, increase of winter temperature would enhance the growth and photochemical efficiency of Q. myrsinaefolia in the northern distribution limit. The growth of the two cool-temperate deciduous broad-leaved species (Q. crispula and B. platyphylla var. japonica) was greater at T2 and T3 conditions than T1 (the optimal temperature conditions). Many studies showed that tree growth is greater nearer the southern and lower distribution limits along latitudinal and altitudinal gradients, respectively (Persson, 1998;Mäkinen et al., 2000;Li et al., 2003;Takahashi & Yoshida, 2009). Although the results of the previous studies were conducted in natural distribution ranges, the result of this study showed that tree growth increases in the temperature conditions warmer than the natural distribution range. Drought stress increases near southern and lower distribution limits (Buckley et al., 1997;Takahashi et al., 2003a;Adams & Kolb, 2005;Hart et al., 2010;Lebourgeois et al., 2010). On the contrary, increase of temperature prolongs the growth period (Sparks et al., 2000;Chmielewski & Rotzer, 2001;Fujimoto, 2008). Although increase of temperature induces drought stress to some extent, it would increase annual growth of plants by increasing annual carbon gain through increase of growth period. The high growth of Q. crispula and B. platyphylla var. japonica at the two temperature conditions (T2 and T3) warmer than the current distribution range suggests that the natural distribution ranges of the two species are not determined by optimal temperature conditions alone (i.e., the concept of ecological niche). Although Q. crispula is a common species in oldgrowth deciduous broad-leaved forests (Masaki et al., 1992;Takahashi et al., 2003b), Q. crispula is recognized as a gap-dependent species (Yamamoto, 1989). It is reported that saplings of deciduous broad-leaved species distributed mainly in canopy gaps in an evergreen broad-leaved forest (Miura et al., 2001). In deciduous broad-leaved forests, understory saplings increases carbon gain by leafing before leaf expansion of canopy trees (Seiwa & Kikuzawa, 1996). However, understory of evergreen broad-leaved forests is dark conditions throughout a year, which have restricted saplings of deciduous broad-leaved species into canopy gaps. Therefore, actual distribution ranges are largely affected not only by temperature conditions but also by competition with other species (Takahashi, 2003).

Conclusion
This study predicted that high winter temperature mitigates the reduction of photochemical efficiency and reduces winter mortality of Q. myrsinaefolia. Furthermore, this study showed that increase of temperature enhances the growth of not only Q. myrsinaefolia but also Q. crispula and B. platyphylla var. japonica. It is possible that global warming intensifies competition between warm-temperate and cool-temperate broad-leaved species at their latitudinal ecotones. Therefore, it is unclear whether Q. myrsinaefolia will gain parts of the habitat now covered by B. platyphylla var. japonica and Q. crispula under global warming. This study examined only one warm-temperate broad-leaved species and two cooltemperate broad-leaved species, and the results of this study are not enough to show general responses of the two life forms to temperature and light conditions because of large variations in responses to temperature and light conditions within a life form (Parsons et al., 1994;Takahashi et al., 2003a;van der Werf et al., 2007). Therefore, it is necessary to analyze many species for each life form to clarify general growth responses to temperature and light conditions.