Photosynthetic Properties of Miscanthus condensatus at Volcanically Devastated Sites on Miyake-jima Island

How photosynthetic-related leaf traits of non-nitrogen (N)-fixing pioneer species respond to extreme habitat conditions of primary succession is still not well-elucidated, especially in volcanically N-deplete habitats. The effect of N-deplete soil on photosynthetic-related leaf traits can provide a basis for predicting how plants adjust their strategies to adapt to such habitats. To examine the responses of leaf traits to extreme conditions, we investigated Miscanthus condensatus (a non-N-fixing C4 pioneer grass) which grows on a volcanically devastated area on Miyake-jima Island, Japan, in which the volcanic ash has been deposited for 17–18 years since the 2000-year eruption. Leaf N content (Narea), light-saturated photosynthetic rate (Amax), and photosynthetic N use efficiency (PNUE) in three contrasting study sites: bare land (BL), shrub land gap (SLG), and shrub land under canopy (SLUC) were determined. Results indicated that compared to previous studies and internal comparison of Miyake Island, M. condensatus in BL was able to maintain a relatively high Amax, Narea and PNUE. The higher Amax was in part a result of the higher PNUE. This is a characteristic necessary for its successful growth in N-deplete soils. These results suggest that M. condensatus has photosynthetic-related advantages for adaptation to volcanically N-deplete habitats.


Introduction
Volcanic ecosystems are usually characterized by high levels of stress and disturbance [1]. In early volcanic successional systems, plant growth is severely limited by nitrogen (N) [2,3]. Plants inhabiting volcanic ecosystems have to withstand the lack of N, which limits their growth and ability to invade and establish new sites. Furthermore, surviving in volcanic devastated sites requires coordinated physiological responses [4][5][6]. Pioneer species seem to be highly adaptable to adverse environments. Many dominant pioneer species can adapt to environmental changes by adjusting their growth strategies [7,8]. For example, altered root morphology, improved resource use efficiency, resorption efficiency (e.g., N, P), and N-fixing ability are all important adaptation strategies that allow plants to cope with poor nutrient habitats. Many studies have focused on N-fixing pioneer species at new volcanically devastated sites [9][10][11][12], because N is scarce in new volcanic materials, such as lava, scoria, and volcanic ash [2]. However, there are also many pioneer species without the N-fixing ability in volcanic succession. This raises the question of how non-N-fixing pioneer species adjust Table 1. Location and morphometric characteristics of bare land (BL), shrub land gap (SLG) and shrub land under canopy (SLUC). Different letters indicate significant differences between the study sites revealed by Steel Dwass post hoc test at a significance level p < 0.05 after Kruskal-Wallis test. Abbreviations are meters above sea level (m a.s.l); photosynthetic photon flux density (PPFD); soil total N (STN) (n = 30); soil total carbon (STC) (n = 30); Canopy openness (n = 48).

Monthly and Site Changes in Leaf Traits of M. condensatus
The main effect of site was statistically significant for most of the leaf traits of M. condensatus (Table 2, p < 0.05), whereas water use efficiency (WUE) and PNUE were not affected by site. In BL, M. condensatus had significantly lower N area , A max , maximum quantum yield of PS II (Fv/Fm), and transpiration rate (E), but significantly higher LMA and WUE than those in the SLG or SLUC (Table 3, p < 0.05). Light compensation point (LCP) value in the SLUC was significantly lower than that in the BL and SLG (Table 3, p < 0.05), but the difference in mean between BL and SLG was not significant. In terms of PNUE, there was no significant difference in values from the three sites (Table 3, p > 0.05).  Table 2, p < 0.05), and relative values among the three sites varied in each month ( Figure 1). The site by month interaction was significant for all leaf traits of M. condensatus (Table 2, p < 0.05). In the BL, the A max , E, and PNUE values in October 2017 and June 2018 were higher than those recorded in the other months (Figure 1b,d,h); also, the PNUE values in these two months were higher in the three sites ( Figure 1h). Figure 2 shows the estimated actual photosynthetic rates (A actual ) in the three sites. The A actual in SLUC only reached 23-30% of the A actual recorded in SLG.

Effect of Environment Factors on the Leaf Traits of M. condensatus
Based on the generalized linear mixed model (GLMM) results (Table 4), canopy openness had significant positive effects on the Fv/Fm, LMA and LCP. No association between canopy openness

Effect of Environment Factors on the Leaf Traits of M. condensatus
Based on the generalized linear mixed model (GLMM) results (Table 4), canopy openness had significant positive effects on the Fv/Fm, LMA and LCP. No association between canopy openness and the remaining leaf traits was found. STN had a significant positive effect on N area and A max . The precipitation had significant positive effect on A max and PNUE. As for PNUE, both canopy openness and STN significantly affected it negatively (Table 4). Table 4. Effect of environment factors on each leaf traits (n = 144) of M. condensatus (generalized linear mixed model-GLMM). Response variables are maximum quantum yield of PSII (Fv/Fm); transpiration rate (E); leaf N content (Narea); light compensation point (LCP); leaf mass per area (LMA); light-saturated photosynthetic rate (Amax); water use efficiency (WUE); photosynthetic N use efficiency (PNUE), respectively. MR 2 Significance of the coefficients: ns p > 0.05, * p < 0.05, ** p < 0.01, *** p < 0.001.

Advantages of M. condensatus in Leaf Traits at Miyake Volcanically Devastated Site
Our aim in this study was firstly to examine photosynthetic-related advantages of M. condensatus in volcanically devastated sites. Normally, N area for C 4 plants is considered to range from 1.68 to 2.52 g·m −2 , while that of C 3 plants ranges from 2.80 to 3.64 g·m −2 [45]. Ghannoum et al. reported a N area range of 0.48 to 1.23 g·m −2 for C 4 NADP-ME grasses [46]. An analysis by Kattge et al. reported an average of 0.93 ± 1.45 g·m −2 for 232 C 4 grass species [47]. Compared to the above research reports, the present study reported leaf N values within the normal range for M. condensatus despite the low soil N content. In addition, it should be noted that the soil N in the BL was about 10% of that in SLG; however, the N area reached 70% of what was recorded in SLG (Table 3). This seems to indicate that M. condensatus has a unique nitrogen acquisition conservation strategy in extreme N-limited habitats.
Values of A max for M. condensatus (18.81 ± 3.72 µmol·m −2 ·s −1 ) in BL were lower than in SLG and SLUC (Table 1). However, it should be noted that M. condensatus achieved 75% A max compared to that in SLG despite low soil N conditions. Generally, C 4 grasses have a high capacity for photosynthesis. Kattge et al. reported an A max value of 19.78 ± 1.58 µmol·m −2 ·s −1 based on 97 species of C 4 grass in various conditions [47]. However, the Amax of M. condensatus (18.81 ± 3.72 µmol·m −2 ·s −1 ) in BL was not drastically lower than this value. All these above indicated that the photosynthetic advantages of M. condensatus in Miyake.
To further evaluate the photosynthetic adaptive advantage of M. condensatus in the volcanically devastated site, its photosynthetic activity was compared to those of plants growing in other nutrient-poor habitats (e.g., sand dunes, glacier retreated sites, volcanic deserts). A recent study in Australia's sand dunes compared the photosynthetic activity of several species across three different successional stages, including the early stage of primary succession [22]. Seven pioneer species (not including C 4 grass species) growing in the early stage of primary succession (plant growth is limited by N) had A max values ranging from 11.3 to 17.8 µmol·m −2 ·s −1 [22]. The highest A max value was reported for the native shrub species, Acacia rostellifera. Compared to the species monitored by Guilherme Pereira et al. [22], M. condensatus appears to have a higher photosynthetic capacity in the early stage of primary succession. However, there are no comparable studies which investigated photosynthesis of C 4 plants in volcanic deserts. In terms of C 3 plants, including trees, there are a few studies comparable to our data. Choi et al. reported that the A max value for A. sieboldiana (an N-fixing C 3 tree species) at Miyake-jima Island [12], in a site with low exposure to the disturbances from the 2000 eruption, was 12 µmol·m −2 ·s −1 . This value is lower than that recorded for M. condensatus, even though the study site was highly disturbed and had low soil N. In the volcanically devastated site of Mt. Fuji, Japan, Sakata et al. reported the photosynthetic rate of two C 3 herb species, Reynoutria japonica and Aconogonum weyrichii [43]. Both species inhabited an old volcanic desert (last eruption at 1707) exposed to heavy wind, and they had A max values of 18.60 and 22.70 µmol·m −2 ·s −1 for R. japonica and A. weyrichii, respectively. On the other hand, the A max value of M. condensatus appears to be higher than that of Metrosideros polymorpha, a C 3 tree species in Hawaii [44]. The A max values of M. polymorpha species growing on a 26-year-old lava flow (from an eruption in 1959) and on a 195-year-old lava flow (from an eruption in 1790) in the Kilauea Iki crater were 8.68 µmol·m −2 ·s −1 and 8.54 µmol·m −2 ·s −1 , respectively [3,44]. These comparisons support the finding that M. condensatus has a higher photosynthetic capacity in nutrient-poor habitats.
PNUE describes leaf photosynthetic economics, and plants with higher PNUE values are regarded as being more adaptive to N-deplete habitats. In general, C 4 plants have higher PNUE than C 3 plants [48][49][50]. Kattge et al. [47] reported the average PNUE value of 80 C 4 grasses as 263.34 ± 23.23 µmol·CO 2 mol·N −1 ·s −1 , which is similar to the value recorded for M. condensatus in the present study. We also found that there was no significant difference in PNUE among three study sites. This indicated that the advantages in N use strategy of M. condensatus. However, there are no reported studies on the PNUE of species growing on similar nutrient deplete habitats; hence, there is no basis for comparison with the PNUE of M. condensatus. This higher PNUE of M. condensatus appears to be primarily a result of its C 4 pathway [46], which can eliminate photorespiration by increasing CO 2 levels in the vicinity of the enzyme Rubisco. However, to fully understand the relative adaptability to N poor habitat among C 4 grass, further comparative studies are necessary.

Physiological Cause for the Variation in PNUE
The GLMM results showed that there were significant negative effects of STN on PNUE, positive effect on N area and A max , and no effect on LMA (Table 4). This is consistent with the conclusions of several preceding studies that a higher PNUE facilitates the establishment of plants [46,50,51]. However, in the present study, the effect of the different sites on PNUE values was not significant (Tables 2 and 3). This is because PNUE is the ratio of A max to N area and the common trend observed in them as a result of soil N levels may be responsible for the non-significant difference in PNUE values between the three study sites. As for the physiological causes of PNUE variation, numerous studies focused on inter-species differences in PNUE suggest that an increase in PNUE with increasing LMA is caused by decreasing photosynthetic rates and lower N partitioning into Rubisco compared to higher N partitioning into cell walls [52][53][54]. However, in this study (about a single species), it was discovered that LMA did not significantly reduce PNUE (Table 3). This is consistent with a previous study, which reported similar PNUE but different LMA values, 40-50 g·m −2 and 15-20 g·m −2 in the leaves of Chenopodium album grown in low nutrient and high average light intensity, and those grown in high nutrient and low average light intensity, respectively. In addition, it was observed that the main factor affecting LMA in the present study was the light condition (Table 4). Hikosaka (2004 for review) [31] also suggested that within a single species, differences in PNUE do not occur when LMA changes with growth light or nutrient availability. Therefore, in evaluating the physiological factors responsible for the variation in PNUE in M. condensatus, it is necessary to investigate the relationship between PNUE and the allocation of N among the major foliar N fractions. In addition, GLMM shows the strong positive relationship between precipitation and PNUE. During the growing season, Amax, PNUE showed strong monthly changes, and relative values among the three habitats in each month also varied ( Figure 1, Table 2). In BL, Amax increased in June and October (Figure 1d), and the PNUE value in these two months was also markedly higher among the three habitats (Figure 1c). All the changes observed in June and October occurred when measurements were carried out after rainfall. The rainfall in June and October were higher than in other months ( Figure A1, Appendix A). These are consistent with the previous finding that temporal rainfall may help plant growth [55,56], and indicate that M. condensatus in BL are more sensitive to water deficits. Water deficit is another factor that limits growth, and probably had a large effect on M. condensatus grown in newborn soil. In this study, under water stress, M. condensatus could not perform all its photosynthetic abilities. Therefore, in the Miyakejima volcanically devastated site, the photosynthetic capacity of M. condensatus is achieved when there is a sufficient water supply. The site-specific pattern of M. condensatus responses to rainfall suggests that rain may be an important contributor to the growth and survival of M. condensatus on the volcanically devastated site.

Physiological Responses to Light Availability
Light availability gradients affect leaf traits, which determine leaf carbon acquisition [57,58], and previous studies have demonstrated that leaves that inhabit low-light environments exhibit shade-acclimatized traits in the long run [59,60]. Results from this study indicated that under low light environments (SLUC), M. condensatus had significantly lower LMA and LCP ( Table 2), but higher Fv/Fm than those exposed to higher light intensities. The GLMM shows a strong positive relationship between canopy openness and Fv/Fm. These results demonstrate that M. condensatus acclimatized to the shade by decreasing LCP and by increasing its ability to capture light (higher Fv/Fm). The higher Fv/Fm observed in SLUC (Table 2, Figure 1) suggests that M. condensatus leaves are more efficient at trapping light in the pigment range of PSII under low light conditions [61], enabling them to use light during the constant low-light periods in the SLUC. In addition, Fv/Fm can provide insights on the ability of plants to tolerate environmental stresses and the extent to which those stresses can damage their photosynthetic apparatus [61]. Results from this study indicated that extremely low soil N did not significantly affect Fv/Fm (Table 3). This also proves the high resistance of the species to environmental stress. Furthermore, reduction in LMA is a typical shade acclimatization response, which can improve the light harvested per unit of resource invested in the construction of photosynthetic tissue [59,62].
Although the capacity to photosynthesize and grow under shaded conditions has been proven in this study, due to the limitation of the actual light condition, the actual photosynthesis rate in SLUC was not high (Figure 2). The low A actual in SLUC indicates that the productivity of M. condensatus is low under low-light conditions. The lower productivity of M. condensatus under the pioneer trees of A. sieboldiana can explain the successional change from grassland consisting of M. condensatus to shrub land comprising A. sieboldiana and other tree species [42].

Materials and Methods
This study was conducted between 2017 and 2018 on a volcanically devastated site (17 and 18 years after the eruption). At the time of this study, the landscape of the area was composed of bare land, M. condensatus grasslands, and A. sieboldiana shrubs.

Research Area
This study was conducted at the Miyake-jima active volcanic island, Japan (34 • 05 N, 139 • 55 E), which covers an area of 55.44 km 2 and has an altitude of 775.1 m a.s.l. (Mt. Oyama). The island has a humid, temperate climate, with an average annual temperature of 17.7 • C, with the average temperatures of the hottest and the coolest month being 26.2 • C (August) and 9.6 • C (February), respectively. The annual precipitation is 2953.6 mm, and the average precipitation for the months with the minimum and maximum rainfall are 140.3 mm (December) and 383.3 mm (October), respectively [63]. At the time of this study, there were no current effects of volcanic gas as the emissions had already ceased.

Experimental Design and Measurements
To examine the effects of environment conditions on the leaf traits of M. condensatus, three study sites in different successional stages were selected: bare land (BL) with M. condensatus patches ( Figure A2a), a site with a relatively open tree canopy (SLG: shrub land gap), and a site with a closed canopy (SLUC: shrub land under canopy) ( Figure A2b,c). In BL and SLG M. condensatus were the absolutely dominant species. The other species such as A. sieboldiana (N-fixing species) and F. japonica var. hachidyoensis (a perennial herb) were also existed sparsely. In SLG, the upper leaves of M. condensatus were almost fully exposed to ambient sunlight, while in SLUC, M. condensatus was sparsely distributed under the closed canopies, and some dead M. condensatus were found. The recovery of vegetation in the study area proceeded in a direction from lower to higher elevation, and from further away from the Oyama crater to closer to the crater [54]. This recovery process may be caused by many factors, such as heavy wind and seed dispersal, the initial damage of volcanic ash and toxic gas [54]. Thus, the three study sites fully reflected the recovery of the vegetation. In addition, differences in actual air temperature and wind caused by changes in altitude may also have affected plant growth in the study site.
In each study site, eight M. condensatus plants were selected for the measurement of leaf traits. The M. condensatus selected in BL were between 0.5 and 1.5 m in height. The distance between the selected plants was at least 5 m. In SLG and SLUC, the plants were between 1.8 and 2.3 m in height. In total, 144 leaf samples were measured. In each study site, leaf traits were measured at an interval of 1 month from July to October 2017 and from May to June 2018 (6 months in total) which contained the entire growing season of M. condensatus. The specific measurement time for each study site is shown in Figure A3. We measured a total of 48 leaves (eight leaves every month) at each study site. All study sites were measured within 3 days of each other.

Gas Exchange Parameters
To measure the activities of healthy leaves located in the upper part of each sample, the second or third (counted from the top of the shoot) fully expanded leaves were selected every month. Based on continuous field observations (every month from May to October 2018), it was discovered that the leaves of M. condensatus in Miyake-jima Islands had a relatively short leaf life span (4 months), and that the production of new leaves was continuous throughout the growing seasons. The relative position of leaves became sequentially lower with the progression of seasons; therefore, leaf samples were collected from the same position in the plant for measurement to make sure they have similar leaf age. In addition, the sampled leaves were evaluated for the determination of LMA and leaf N every month. Intelligent portable photosynthesis system LCPro + (ADC BioScientific, UK) was used to calculate the measurement of gas exchange. At each study site, eightleaves (one leaf per plant) were measured per month. Measurements were taken from the leaves in the middle, and all gas exchange parameters were measured between 9:00 a.m. and 11:00 a.m. During all of the measurements, the chamber CO 2 concentration, temperature, and vapor pressure deficit (VPD) were maintained at 420 µmol·mol −1 , 25 • C, and 1.0 kPa, respectively. Light-response curves were determined at irradiances between 2000 and 0 µmol·m −2 ·s −1 using a built-in LED light source in seven PPFD steps and were fitted using a non-rectangular hyperbola model [64]. A max , and LCP were calculated using light-response curves. WUE was calculated as the ratio of A max to the E at photosynthetic photon flux density (PPFD) saturation [65].

Structural and Biochemical Characteristics of Leaves
When the measurements of photosynthetic parameters were completed, the sampled leaves were collected and wiped; the measured areas were cut out and scanned, and the images were processed using ImageJ software (National Institutes of Health, Bethesda, MD, USA, imagej.nih.gov/ij) for leaf area. For leaf dry mass/leaf area determination, cut-out leaf samples were dried at 80 • C for 48 h to constant weight and measured for their final weight. The remaining leaf samples were also dried for the determination of leaf N using an NC analyzer (SUMIGRAPH NC-220F). The PNUE was calculated as A max divided by N area (PNUE, µmol·CO 2 mol·N −1 ·s −1 = A max (µmol·m −2 ·s −1 )/(1/14N area ) [52].

Chlorophyll Fluorescence
While measuring gas exchange parameters, three light-adapted leaves (the middle part) of each plant were selected for the measurement of chlorophyll fluorescence using MINI-PAM-II-Walz. The minimum fluorescence (F0) and maximum fluorescence (Fm) were determined following a 30 min dark adaptation using leaf clips, in order to calculate the maximum photochemical efficiency of PSII (Fv/Fm).

Rainfall Distribution before Measurements
To assess the effects of precipitation on the leaf traits of M. condensatus, the rainfall distribution before the measurement day of each month was analyzed. We used the 3-day integrated rainfall from before the photosynthetic measurements, from the daily rainfall data from the Japan Meteorological Agency (2019) [63]. The highest levels of rainfall (before the measurement day) were recorded in June and October ( Figure A1).

Soil Measurements
In October 2017, soil samples from the root zones of the sampled M. condensatus plants were collected using 100 mL core samplers of 5 cm depth for the determination of the soil properties of the study sites. Surface soils (5 cm in depth) were collected because M. condensatus develops its root system on the ground surface and its fine roots are concentrated in shallow soils. Three cores were sampled from the soil around the root crown (approximately 10 cm away from the root crown) and were subsequently combined and mixed. After air-drying, the samples were passed through 2 mm and 0.5 mm sieves, respectively, and the roots were removed. Soil samples between 0.5 mm and 2 mm were ground to sizes lesser than 0.5 mm. The soil total C (STC) and STN were analyzed with SUMIGRAPH NC-220F, using samples smaller than 0.5 mm.

Light Measurements
Every time we measure photosynthetic parameters, Hemispherical photographs were taken using a Nikon Coolpix 990 and a Nikon FC-E8 Fisheye Converter (Nikon, Tokyo, Japan). Photographs were taken above each plant (totally, 48 were taken in each study site) and were analyzed to calculate canopy openness using Gap Light Analyzer (GLA Version 2.0) [66]. Canopy openness was defined as the fraction of open sky in the hemisphere that was visible from a point beneath the canopy and was used as an index of the light available to the M. condensatus plants.
A representative location was selected in the three sites for the measurement of photosynthetic photon flux density (PPFD) at sunny or cloudy days, with the aid of IKS-27 sunshine sensors (KOITO MANUFACTURING CO., LTD, Yokohama, Japan) which were mounted on poles at a height of 1 m. The measurements were done simultaneously in the three sites between 8:00 a.m. and 4:00 p.m. in the month of August 2016 and reading of data was done every 30 s. We used maximum daily PPFD for each site and light-response curve for each leaf sample (48 samples in each study site) to estimate the actual photosynthetic rates (A actual ). This parameter can be used to explain the photosynthetic rate in natural light condition.

Data Analysis
All analyses were carried out using the statistical package R version 3.6.0 [67]. For each leaf traits variable, data were analyzed by two-way ANOVA, with study sites and month as main fixed factors plus a site × month interaction term ( Table 2). The site differences of annual mean value of each leaf traits were compared (Table 3). Prior to analyses, the assumptions of normality and homogeneity of variances were tested using Shapiro-Wilk and Bartlett tests. For leaf traits that meets homogeneity of variance and normal distribution (E, N area , LMA, A max , PNUE), one-way ANOVA was applied to assess the differences of them followed by Tukey's post hoc test used for multiple comparisons between the sites. A non-parametric Kruskal-Wallis test was performed to check for differences in environmental factors and some leaf traits (Fv/Fm, LCP, and WUE) among the three study sites. Steel Dwass post hoc test was performed to detect significant differences in the mean value of environmental factors, Fv/Fm, LCP, and WUE. The generalized linear mixed model (GLMM) with the package lme4 [68] was used to assess the effect of environment factors on leaf traits. The predictors (fixed effects) including the soil N, canopy openness, temperature and precipitation (3-day integrated rainfall), while the response variables were leaf traits (Fv/Fm, E, N area , LCP LMA, and A max , WUE, PNUE), respectively. The study site and month were included as random effect factors. Marginal and conditional R 2 were calculated.

Conclusions
We investigated the leaf traits of M. condensatus, a non-N-fixing pioneer species that grows in Miyake-jima, a volcanically devastated site. The results showed that compared to previous studies and internal comparison of Miyake Island, M. condensatus could maintain a relatively higher N area , A max despite under extreme N-deficient conditions. This is partly as a result of its higher PNUE. PNUE in M. condensatus appears to be high even for a C 4 grass; this is a characteristic necessary for its successful growth in N-deficient soils. The higher PNUE of M. condensatus may be an intrinsic function of its C 4 pathway. We also found that PNUE values in BL were higher than in the other two habitats in June and October. Considering the level of rainfall before our measurements, the higher PNUE in BL was related to rainfall.