Leaf Adaptation of Eurya japonica Thunb. (Pentaphylacaceae) in Coastal Area

To clarify the process of plant adaptation to coastal areas, we conducted morphological and anatomical analyses of Eurya japonica Thunb. (Pentaphylacaceae). There was no significant difference in leaf shape between the inland and coastal populations, although the leaves in coastal populations tended to be thicker. However, our anatomical analysis revealed significant differences in stomatal size and adaxial and abaxial epidermal cell sizes. The smaller stomata of the coastal population of this species were effective in reducing transpiration during gas exchange. Furthermore, the expansion of epidermal cells could be an adaptive strategy to retain water in leaves in coastal environments.


Introduction
Various adaptation processes frequently result from spatial variations in selective environmental forces acting on phenotypic diversity (Hereford, 2009). As selective forces operate, they may reduce heritable variation within a population, leading to the specialization of individuals (Kawecki & Ebert, 2004). Conversely, in highly stochastic environments, selection can increase the potential of a species for phenotypic plasticity (Albert et al., 2011). Therefore, research on adaptation is essential for understanding the response of a species to environmental changes (Jump & Peñuelas, 2005;Aitken et al., 2008). The environment provided by the land-sea interface supports many plant species that are not found further inland (van der Maarel, 2003). Harsh conditions in many coastal areas limit the range of plant species that can be successfully grown (Rozema et al., 1985). Most coastal plant species have gained characteristics which make them ideally suited to withstand the effects of the coastal environment such as strong persistent on-shore winds carrying sand and salt spray which dry out the leaves and soil (Hesp, 1991;Maun, 1998;Ciccarelli et al., 2009). Therefore, coastal adaptation is often a product of environmental variations in geographical space (Ciccarelli et al., 2010). Coastal plants are known for their rich biodiversity and high level of endemism, including a concentration of rare and threatened taxa and high diversity of endemic plant species (Burgess et al., 1998;Lovett, 1998;Myers et al., 2000;Azeria et al., 2007), and they have implications for biodiversity conservation (van der Meulen & Udo de Haes, 1996;Cori, 1999;Schlacher et al., 2008). Particular attention has been focused on understanding ecophysiological mechanisms involved at the cellular and molecular levels (Elhaak et al., 1997;Migahid & Elhaak, 2001;Tunala et al., 2012;Kumekawa et al., 2013;Ohga et al., 2013;Sunami et al., 2013;Takizawa et al., 2022). It is uncertain whether plant species widely distributed from inland to coastal areas have adapted with or without morphological changes. For example, Tunala et al. (2012) indicated that the epidermal cells of the coastal variety Aster hispidus Thunb. var. insularis (Makino) Okuyama (Asteraceae) were larger but fewer than those of As. hispidus Thunb. var. hispidus and had succulent leaves to store water. Additionally, Sunami et al. (2013) reported that leaf hairs on the abaxial side of leaves were correlated with the stomatal density of this variety; the fewer the hairs on the leaf, the lower the stomatal density to avoid transpirational water loss. Ohga et al. (2013) suggested that the coastal population of Adenophora triphylla var. japonica (Regel) H.Hara (Campanulaceae) has evolved relatively thick leaves via a heterochronic process to store water. Generally, stomata are the major gates for gas exchange in leaves (Raschke, 1975;Schoch et al., 1980;Brent & Ram, 2000). Guard cells surrounding the stomata contain chloroplasts, and they increase stomata sugar concentration, which in turn causes water absorption and swelling of cells (Jarvis, 1976;Ewers et al., 2001). Additionally, stomatal conductance depends on leaf characteristics, such as the size, number, and frequency of stomata and leaves (Cole & Dobrenz, 1970;Teare et al., 1971;Ciha & Brown, 1975). Therefore, plants in wide habitats, including coastal areas, adapt to coastal conditions through morphological and anatomical changes.
Grasses differ from trees due to the spatial separation of rooting niches and the differential utilization of below-ground resources such as water. Trees exclusively access water in deeper soil horizons, whereas grasses use water primarily in the topsoil (Hesla et al., 1985;Knoop & Walker, 1985;Sala et al., 1989). Therefore, water stress on herbs in coastal areas is considered to be much greater than that on trees because of shore winds carrying sand and salt spray. However, Cruz et al. (2019) reported that Avicennia schaueriana Stapf & Leechm. ex Moldenke (Acanthaceae), a dominant coastal tree, had small leaves with a low stomatal conductance and transpiration. However, Takizawa et al. (2022) reported that the stomatal size of Ligustrum japonicum Thunb. (Oleaceae) was significantly smaller in coastal areas than that in inland areas. Thus, analysis of the adaptations of tree species in coastal areas suggested that tree species also undergo some morphological changes to adapt to coastal areas. Furthermore, these stomatal studies revealed that the adaptation pattern in coastal areas reduced the density of stomata in grass species, while tree species decreased stomata size, suggesting that stomatal adaptation in coastal areas had different modes between grass and tree species. However, only a few studies have reported on the coastal adaptation of tree species.
Eurya japonica is a broad-leaved evergreen woody perennial approximately 1-3.5 m tall that is widely distributed in warm temperate forests in Japan (from Honshu to Ryukyu) and Korea (Tsuyama, 1989). Nagamasu (2006) and Nakajima and Yoshizaki (2018) reported that E. japonica comprises a coastal forest with Eurya emarginata (Thunb.) Makino, Machilus thunbergii Siebold et Zucc. (Lauraceae), and Ilex integra Thunb. (Aquifoliaceae). Shiba et al. (2021) indicated that E. japonica in riverside populations develops thicker leaves and a higher stomatal density than those in inland populations for adaptation to strong solar radiation, and waterlogged environments. Moreover, serpentine populations of this species have small and thick leaves and reduced stomatal size to minimize water loss via gas exchange . Findings on E. japonica suggested that this species has adapted to unique environments by changing its morphological and anatomical characteristics. This species is expected to undergo some morphological variations when colonizing coastal areas. Therefore, investigating the coastal adaptation of E. japonica is important when considering coastal forests that play a major role as ecosystems with high conservation value. Therefore, this study aimed to clarify the E. japonica adaptation leaf traits by comparing individuals between coastal and inland areas using morphological and anatomical approaches ( Figure 1).

Method
The Eurya japonica samples examined in this study were collected from a field in Shikoku, Japan. The collection locations are shown in Figure 2 and Table 1. The inland populations near each coastal area were used as controls. Two hundred and ten individuals (30 individuals per population) representing 7 populations involved in 4 coastal areas and 3 populations were sampled. Among these, Ishima in Tokushima Pref. is a small island in the Kii Channel in the Pacific Ocean between Honshu and Shikoku at the eastern-most tip of Shikoku ( Figure 2). Although it is difficult to define coastal forests on this island, individuals sampled from these islands were used as the coastal area samples in this study, because E. japonica had been grown together with the following coastal endemic species: Pittosporum tobira (Thunb.) W.T.Aiton (Pittosporaceae) and E. emarginata in the coastal forests. According to the Japan Meteorological Agency, our coastal sampling areas had stronger winds than inland areas (Figure 3).

Figure 2. Sampling localities in this study
Number in corresponds to that given in Table 1. Blue and brown circles indicate coastal and inland areas, respectively. Blue and brown triangles indicate anemometer locations near coastal and inland areas, respectively. 33°30'N 133°20'E *: locality no. corresponds to that given in Figure 2.
For morphological analysis, individuals were measured for leaf blade length and width and leaf thickness using a digimatic caliper (CD-15CXR; Mitutoyo Corporation, Kanagawa, Japan) and digimatic outside micrometer (MDC-SB; Mitutoyo Corporation). Leaf measurements were taken from a fully expanded leaf at the midpoint of plant height. Three leaves were randomly selected from one individual, and the average value was calculated.
Fully expanded leaves were collected from each individual for anatomical analysis. To count the number of stomata on the leaf blade, the abaxial surface of leaves was peeled off using Suzuki's universal micro-printing (SUMP) method (Kijima, 1962). Briefly, the middle part of the blade along the midrib was examined to determine the number and size of stomata. Replicas of each leaf (1 cm 2 ) were prepared to determine the stomatal density (Number/µm 2 ) and size of five leaves per leaf. The stomatal size was calculated using the following formula: stomatal length × stomatal width/2, based on a study by Kumekawa et al. (2013). The copied SUMP images for each individual were examined once using a light microscope (CX41; OLYMPUS, Tokyo, Japan).  183.85 ± 6.34 µm, respectively. There were no significant morphological differences in E. japonica between the coastal and inland areas. In this study, the leaf index values were calculated as the leaf length to width ratio based on Tsukaya (2002). A significant difference in leaf index values indicates that the population with a large value has narrow leaves. The leaf indexes of the coastal populations Tei, Inomisaki, Kashiwajima, and Ishima were 2.45 ± 0.06, 2.39 ± 0.04, 2.58 ± 0.06, and 2.46 ± 0.05, and those of inland controls Shigeto, Noichi, and Okina were 2.66 ± 0.05, 2.37 ± 0.04, and 2.47 ± 0.05, respectively. However, there was no significant difference in leaf index between the coastal and inland populations.        70, 17.18 ± 0.58 and 17.65 ± 0.63, 63.20 ± 3.90, 59.88 ± 3.36 and 61.58 ± 3.45, 105.38 ± 6.33, 82.30 ± 5.26 and 90.43 ± 3.82, and 9.58 ± 0.49, 13.65 ± 0.34 and 14.20 ± 0.63 µm, respectively. The results showed that the heights of epidermal cells on both sides were significantly different between coastal and inland populations. However, those of palisade and spongy tissues showed no significant differences. Furthermore, we measured the mean stomatal size (guard cell area based on guard cell pair length and width) and density (Table 5 & Figure 6). There were significant differences between the stomatal sizes of the four coastal populations (87.03 ± 1.60, 88.69 ± 1.36, 90.25 ± 1.19, and 89.27 ± 1.49 µm 2 in Tei, Inomisaki, Kashiwajima, and Ishima, respectively) and those of the inland populations (104.75 ± 1.34, 107.02 ± 1.93, and 100.17 ± 1.64 µm 2 ). However, there were no significant differences between the stomatal density (233.95 ± 6.49, 261.78 ± 3.89, 288.18 ± 10.44, and 238.64 ± 4.28 N/µm 2 in Tei, Inomisaki, Kashiwajima, and Ishima, respectively) of the coastal populations and those of the inland populations (243.33 ± 6.49, 233.26 ± 5.67, and 259.19 ± 4.90 N/µm 2 in Shigeto, Noichi, and Okina, respectively).

Discussion
Some coastal adaptive species, such as Pittosporum tobira, Camellia japonica L. (Theaceae), Cirsium maritimum Makino (Asteraceae), and Dianthus japonicus Thunb. (Caryophyllaceae) were present in our coastal sampling areas, and the above-mentioned coastal adaptive species Adenophora triphylla var. japonica was grown in neighboring areas where we collected samples, suggesting that strong coastal selection pressures affect Eurya japonica and these species. Changes in the leaf morphology of coastal species are commonly observed in plants subjected to increased drought and salinity stresses. These were caused by increased leaf water content and subsequent accumulation of large amounts of solute without increasing the osmotic pressure of cells (Suá rez & Sobrado, 2000). In the present study, we found that coastal E. japonica has significantly smaller stomata and larger epidermal cells than inland area species, indicating that the former gained adaptive characteristics in coastal areas. Many studies have reported that coastal adaptation included a decrease in stomatal density (Tunala et al., 2012;Ohga et al., 2013;Kumekawa et al., 2013). Contrastingly, Takizawa et al. (2022) reported that woody Ligustrum japonicum had significantly smaller stomatal sizes in coastal areas than those in inland area species and hypothesized that woody plants could not invade coastal areas from inland areas without stomatal changes, although woody plants could adapt to coastal areas by reducing stomatal size. In fact, herbaceous plants have limited capacity for water retention, but woody plants can retain water in the trunk under dry conditions (Takizawa et al., 2022); however, since the results were for only one woody plant (L. japonicum), studies on different woody taxa are needed to test this hypothesis. Our findings are consistent with the hypothesis that L. japonicum (Oleaceae) has a large phylogenetic distance from E. japonica (Pentaphylacaceae) based on angiosperm phylogeny (Chase et al., 1993;Soltis et al., 2000;Savolainen & Chase, 2003;Soltis & Soltis, 2004). Therefore, smaller stomatal modifications to introduce inland area species to coastal areas suggest that this process is a common trend in the coastal adaptation of woody plants. To gain a deeper understanding of coastal adaptation in woody plants, it would be interesting to clarify whether changes in stomatal size are genetic or plastic. Frankes et al. (2009) indicated that such stomatal changes were likely due to plasticity; therefore, the modification in stomatal size associated with coastal adaptation of these woody plants may be plastic. In the future, it is necessary to examine plasticity through the cultivation experiment of E. japonica.
Our anatomical analyses showed that the epidermal cells of coastal E. japonica were larger in size but fewer in number than those of inland. Liu et al. (2005) reported larger cell size in drought-tolerant genotypes in Boehmeria nivea (L.) Gaud. (Urticaceae). Moreover, Liu et al. (2011) indicated that artificial tetraploid plants of Dendranthema nankingense (Nakai) Tzvel. (Asteraceae), which have larger cells than those in normal diploid plants, showed higher levels of abiotic stress tolerance, including drought and salinity stresses. In addition, Kumekawa et al. (2013) showed that the expansion of cells of coastal endemic taxa, Aster hispidus var. insularis, could be considered an adaptive strategy to coastal environments that allowed the saving of water in plant bodies. These studies suggested that large cell size of leaf could be resistant of dry condition such as coastal areas. However, the size of the leaves is increased when the cell size are enlarged, and therefore the influence of physical stress such as wind and flying sand is increased in coastal areas. From these reasons, we considered that the number of epidermal cell in coastal E. japonica was tended to reduce than that in inland area. It will be necessary for further studies to investigate the correlation between cell size and water content in leaves of these species.
Some genetic studies have provided evidence for the coordination of cell proliferation and expansion of the leaf. For example, the reduction in the final leaf area is compensated for by an increase in the size of individual leaf cells when cell proliferation in a leaf is reduced because of certain mutations, which could aid in understanding the regulation of cell proliferation and expansion at the organ level (Tsukaya, 2002;Horiguchi et al., 2005;Tsukaya, 2008). Kim and Kende (2004) reported that the loss-of-function mutation of model plants, which positively regulate cell proliferation in leaf primordia, causes the typical compensation syndrome (Horiguchi et al., 2005). Additionally, several other mutations that affect leaf cell proliferation cause the compensation syndrome (Mizukami & Fischer, 2000;Ullah et al., 2001;Autran et al., 2002;Nelissen et al., 2003;Clay & Nelson, 2005;Ali et al., 2007;Fujikura et al., 2007). These genetic backgrounds of cell proliferation and leaf expansion are useful considering evolutionary developmental studies of leaf differentiation between coastal and inland E. japonica. Therefore, it would be interesting to compare the leaf morphologies of transgenic model plants by introducing these candidate homologs isolated from them.
We analyzed the coastal adaptation of E. japonica using morphological and anatomical data. These results provide an unbiased interpretation of the coastal adaptation. Our analyses were highly effective in revealing the coastal adaptation process of E. japonica. However, these data were least effective in providing definitive answers to the question of their patterns of adaptation along the seaside. It has been shown that E. japonica adapts to the characteristic environment, such as along riversides (Shiba et al, 2021) and serpentine areas , and to coastal environment where selective pressure is strong by changing its morphology. For example, in the genus Eurya, E. emarginata is a coastal dioecious (Tsuyama, 1989), broad-leaved, and woody perennial species distributed in the Southern China and Japan (western region to Chiba Pref. of Honshu, Shikoku, Kyushu, and Ryukyu Islands) and a few locations in the southern coastal area of the Korean Peninsula (Tsuyama, 1989). In Japan, E. emarginata is cultivated along roads and hedges far from coastal areas (data not shown). These inland E. emarginata individuals may be subject to different selective pressures in coastal areas. These may have gained adaptive morphology from inland to coastal areas, contrary to the findings of our study. Coastal species are scattered across various taxa in angiosperms and have a history of coastal adaptation. Research on coastal adaptation may lead to environmental conservation of coastal areas, where a high rate of environmental change has been observed.