**1. Introduction**

In forests, individual plants from a single species often experience various light environments, from well-lit clearings or large gaps to shaded understories [1–4]. For plants, as sessile organisms, phenotypic plasticity is essential for survival in such heterogeneous environments [3,5–8]. This phenotypic plasticity and the consequent intraspecific variation also greatly influence community-level plant traits and productivity [9–15], highlighting the importance of the quantification of phenotypic plasticity of plant traits under different light environments.

In shaded understories, maximizing net carbon gain [3,6,16–18] and maximizing stress tolerance [19–22] are two major determinants of plant survival [18,23]. For maximizing net photosynthetic carbon gain, acclimation of leaf physiological traits [24,25] and biomass allocation patterns [18,24,25] are both important strategies. Within a species, plants grown in shaded places have leaves with a lower light-saturated photosynthetic rate [1–7,15,18–20,26–31] and a lower dark respiration rate [1–7,18], have thinner leaves with a lower leaf mass per unit area associated with their lower biomass investment per unit area [4,15,18–20,27–29,32–34], and have a higher leaf mass ratio (i.e., leaf mass relative to whole-plant mass) [5,18,25,28] than plants grown in well-lit places. Analogous leaf acclimation to different light environments has also been reported for sunlit and shaded leaves within a single canopy or within a single plant [26,29,34–47]. A low dark respiration rate of a shade leaf leads to a lower photosynthetic light compensation point (LCP) [3,5,6,17,18,24,30,48]. It has been frequently suggested that the net daily carbon gain would increase by lowering the LCP in the shade [3,17,24]. Such a simple consideration, however, has limitations because it only evaluates static photosynthetic parameters. In the forest understory, light intensity changes diurnally due to the diurnal elevation of the sun and fluctuates dynamically due to sunflecks [2,49–60]. A comparison of only static photosynthetic parameters, such as light-saturated photosynthetic rates and dark respiration rates, may therefore poorly reflect actual daily photosynthesis in field environments [42,51,56,61,62]. Given this, it has been questioned whether simple sun vs. shade acclimation can be understood based on the steady-state photosynthetic rate [23,42]. Additionally, the results of laboratory experiments under controlled low light environments [6,20] or those of field shading experiments using shade cloths [19,30,33] may not provide an accurate estimate of carbon gain in the understory, because they do not take into consideration sunflecks. To understand the effect of shade acclimation on daily net carbon gain, therefore, the effects of sunflecks also should be considered.

Here, we investigated the shade acclimation of *Petasites japonicus* subsp. *giganteus* that naturally grew in either well-lit or shaded places in a temperate forest. In a previous study on the same species [32], the phenotypic plasticity of some leaf traits under different light environments was reported. However, because the authors did not measure photosynthetic parameters and local light intensity, they did not clarify whether such plasticity contributed to maximizing carbon gain under each light environment. The objectives of this study, therefore, were (1) to quantify the photosynthetic and morphological acclimation to different light environments for this species, and (2) to test whether leaf physiological acclimation contributed to maximizing leaf-level carbon gain under diurnally changing light environment due to sunflecks.

#### **2. Materials and Methods**

#### *2.1. Study Species*

Butterbur (*Petasites japonicus* (Siebold et Zucc.) Maxim. subsp. *giganteus* (G.Nicholson) Kitam.) (Asteraceae) is a perennial herb distributed in Northeast Asia [63]. This species is found naturally in environments of varying amounts of light, such as roadsides, well-lit forest gaps, and in shaded forest understories. This species also is grown as a vegetable in eastern Asia, including Japan, Korea [64], and Taiwan [65]. Large radical leaves (often reaching 1–2 m in height) elongate from an underground shoot in this species (Figure 1a–c). Therefore, investigating the leaves is equivalent to investigating the entire above-ground part (ramet) for this clonal species. These leaves are usually horizontally arranged on the ground so as to prevent overtopping others, but small immature leaves that are not fully expanded often exist below fully expanded leaves.

#### *2.2. Study Site and Sampling*

We performed the study at two sites in the same forest (clearing [C] and understory [U]), which were approximately 100 m apart, in the Forest of Obihiro (Obihironomori) (42◦53 N, 143◦09 E, altitude: 86 m a.s.l.). This secondary forest comprises a mixture of planted and regenerated trees and is located in Obihiro City in eastern Hokkaido in a cool-temperate region in Japan. The mean annual temperature and precipitation at the Japan Meteorological Agency Obihiro Weather Station (6 km from the study site) between 1998 and 2017 were 7.2 ◦C and 937 mm, respectively [66]. In the clearing site (approximately 30 × 30 m), few trees were taller than the investigated leaves (Figure 2a). The understory site (Figure 2b) (approximately 15 × 10 m) was located under a young birch forest (*Betula platyphylla* Sukaczev var. *japonica* (Miq.) H.Hara; DBH: 17.5–25.1 cm), in which some walnut (*Juglans mandshurica* Maxim. var. *sachalinensis* (Komatsu) Kitam.) grew as subcanopy trees. Within each plot, the investigated leaves were selected along a transect. Although we found multiple separate patches of leaves at each site, the number of genets was unknown. Therefore, the investigated leaves were selected as evenly as possible along the entire length of each transect.

**Figure 1.** The measured morphological parameters of the leaves of the butterburs (*Petasites japonicus* (Siebold et Zucc.) Maxim. subsp. *giganteus* (G.Nicholson) Kitam.) investigated in this study. Leaves in (**a**) the clearing and (**b**) the understory, in addition to (**c**) the measured leaf parameters, are shown. H: The highest point on the leaf lamina. O: The point located at the opposite side of H on the lamina edge. P: The point of attachment of the lamina to the petiole. G: The point of attachment of the petiole to the ground. *d*high: The distance between H and P. *h*high: The vertical distance between H and P. *d*opp: the distance between O and P. *h*opp: the vertical distance between O and P. *pl*: Above-ground petiole length (the distance between P and G). α: Lamina openness angle. Photographs were taken in June 2020 by Kohei Koyama.

**Figure 2.** The study sites, (**a**) the clearing and (**b**) the understory, in the Forest of Obihiro. Photographs were taken on (**a**) June 28 and (**b**) 3 July 2020 by Kohei Koyama.

*Forests* **2020**, *11*, 1365

In June 2020, we marked 62 leaves (32 from plants in the clearing and 30 from plants in the understory). Leaf three-dimensional structure was measured using measuring tapes on 24–25 June 2020, and the lamina openness angle [67,68] was calculated (Figure 1c). Photosynthetic light response curves were measured for a total of 12 leaves (6 at each site) on 21, 22 and 24 June 2020 with a portable photosynthesis system (LI-6400; LI-COR, Lincoln, NE, USA) equipped with an LED light source (LI-6400-02B) (Figure 3a). Due to the amount of rainfall prior to the measurement days (June 18 (3 mm), June 19 (7.5 mm), June 20 (4 mm), and June 23 (1 mm), data from [66]), the soil in the fields was wet during the measurements. Measurements were taken in the morning (7:30–12:00) each day under cloudy and humid conditions, and the environment inside the chamber showed favorable conditions for photosynthesis: leaf temperature (measured by a thermocouple inside the chamber) ranged from 17.93 to 24.48 ◦C, and the vapor pressure deficit (VPD) based on leaf temperature was always less than 0.9 kPa. In the understory, we first induced the leaves by keeping incident photosynthetic photon flux density (PPFD) on the leaf surface at 1000–1500 μmol m−<sup>2</sup> s−<sup>1</sup> until equilibration. This process was omitted for most of the leaves in the clearing, which showed a very quick response under PPFD 2000 μmol m−<sup>2</sup> s<sup>−</sup>1. Then, we progressively lowered the incident PPFD on the leaf surface (2000, 1500, 1000, 750, 500, 250, 125, 63, 32, and 0 μmol m−<sup>2</sup> s<sup>−</sup>1). On each occasion of changing light intensity, we kept the PPFD constant until the equilibration of the leaves. The CO2 concentration of the air entering the leaf chamber was controlled at 400 ppm. All the data recorded by the LI-6400 (e.g., photosynthesis, stomatal conductance, transpiration, humidity, temperature at each PPFD, etc.) are available in the Supplementary Materials.

**Figure 3.** Measurements of (**a**) photosynthesis and (**b**) the diurnal course of incident light in the clearing site. The two panels show the same leaf. The position on the leaf lamina, at which photosynthetic traits and incident light were measured, was marked with a light-resistant ink pen (red box). In the cases when that part of the lamina was inclined, the light sensor was inclined such that the lamina and the top of the sensor were parallel to one another. Leaf mass per area (LMA) was subsequently measured by sampling the lamina part within the same red box. Photographs were taken on (**a**) June 21 and (**b**) 1 July 2020 by Kohei Koyama.

(**a**) (**b**)

Net photosynthetic rate per unit area of each leaf (*P*net μmol m−<sup>2</sup> s<sup>−</sup>1) was assumed to be expressed by the non-rectangular hyperbola (NRH) [69]:

$$P\_{\rm net} = \frac{\Phi l + P\_{\rm g\\_max\\_area} - \sqrt{\left(\Phi l + P\_{\rm g\\_max\\_area}\right)^2 - 4\theta \,\Phi l P\_{\rm g\\_max\\_area}}}{2\theta} - R\_{\rm area} \tag{1}$$

where *I* (μmol m−<sup>2</sup> s−1) indicates the incident PPFD for each leaf at each moment, and *P*g\_max\_area (μmol m−<sup>2</sup> s<sup>−</sup>1) indicates the maximum gross photosynthetic rate when *I* approaches infinity. Φ (mol CO2 mol−<sup>1</sup> quanta) and θ (dimensionless) indicate the initial slope and the convexity, respectively. *R*area (μmol m−<sup>2</sup> s−1) indicates the dark respiration rate. These parameters were fitted for each leaf

with the R function *nls*. Light compensation point (LCP) was calculated by solving the quadratic form of NRH [68] for *I* on the condition that *P*net = 0 [70] using the software Maxima (Maxima project, USA) [71]:

$$\begin{aligned} \left(\Phi(P\_{\text{net}} + R\_{\text{area}})\right)^2 - \left(\phi I + P\_{\text{g\\_max\\_area}}\right) \left(P\_{\text{net}} + R\_{\text{area}}\right) + \phi I P\_{\text{g\\_max\\_area}} &= 0\\ P\_{\text{net}} = 0 \Rightarrow I \equiv \text{LCP} &= \frac{R\_{\text{area}} \left(R\_{\text{area}} 0 - P\_{\text{g\\_max\\_area}}\right)}{\left(R\_{\text{area}} - P\_{\text{g\\_max\\_area}}\right) \Phi} \end{aligned} \tag{2}$$

#### *2.3. Measurement of PPFD*

#### 2.3.1. Diurnal Course of Incident PPFD on the Leaves

We measured a time-series of incident PPFD on the selected leaves on two days: an overcast day (June 24; clearing, *n* = 4; understory, *n* = 4) and a sunny day (July 3; clearing, *n* = 3; understory: *n* = 4) in 2020. The sample size difference between these two days was due to a measurement failure caused by an operational error. On the days between June 24 and July 3, PPFD data were not obtained due to intermittent disruptions by rain. Those leaves were selected from the samples for which photosynthetic light response curves were measured. PPFDs were measured for the same parts of the leaves as for the photosynthetic parameters (the red box, Figure 3). For each target leaf, we set one quantum sensor (MIJ-14PAR Type2/K2; Environmental Measurement Japan, Fukuoka, Japan) on the pole. If the leaf part was inclined, the sensor was inclined to measure the incident PPFD on the leaf surface (Figure 3b). Each sensor was connected to a voltage logger (LR5041; HIOKI, Ueda, Japan). Voltage was recorded every 10 min for 24 h. These voltage values were transformed into PPFD using sensor-specific coefficients.
