Diurnal and seasonal variations in photosynthetic rates of dwarf pine Pinus pumila at the treeline in central Japan

ABSTRACT

It is important to develop a convenient method for predicting photosynthetic rates in cold regions such as high elevations. Dwarf pine Pinus pumila is a representative treeline species in Japan. In this study, we investigated how environmental factors affect diurnal and seasonal variations of the in situ photosynthetic rate of P. pumila. Five environmental factors (photosynthetic photon flux density [PPFD], air temperature [T_a], soil temperature [T_s], vapor pressure deficit [VPD], and soil water content [SWC]) were examined. Photosynthetic measurements were conducted for four needle ages at hourly intervals on fifteen days from July to October 2018. The net photosynthetic rate decreased with needle age and changed seasonally and diurnally for each needle age. The net photosynthetic rate showed unimodal patterns with T_a, T_s, and SWC. The seasonal and diurnal changes in the net photosynthetic rate were related to seasonal changes in the four environmental factors (PPFD, T_a, T_s, SWC); the net photosynthetic rate decreased in mid-summer when T_a, T_s, and PPFD were high due to a decrease in SWC. Seasonal and diurnal changes in net photosynthetic rates could be reproduced by the multiplicative model of the four environmental factors and needle age. Therefore, the photosynthetic rate of P. pumila can be explained and predicted using these environmental factors and needle age. This information will be useful for predicting the effects of climate change on the photosynthetic production of plants in cold regions.

KEYWORDS:

• Climate change
• dwarf pine
• growth
• photosynthesis
• treeline

Introduction

The Intergovernmental Panel on Climate Change (2013) reported that the mean global air temperature increased by 0.12°C per decade from 1951 to 2012 and predicted that the mean annual air temperature between 2081 and 2100 will be up to 4.8°C higher than that during 1986 to 2005. Air temperature affects plant growth and distribution by altering physiological traits, such as photosynthesis and respiration (Sage and Kubien 2007; Ganjurjav et al. 2015; Hu et al. 2016). In general, climate change is expected to increase both photosynthesis and respiration (Wu, Su, and Jiang 2011); however, the degree of change varies between ecosystems (Oberbauer et al. 2007; Craine et al. 2011). In colder regions, the correlation between plant growth and summer temperatures is stronger (King et al. 2013; Y. Zhang et al. 2014; Jørgensen et al. 2015). Therefore, climate change is expected to increase plant growth and photosynthetic production in cold regions (Takahashi and Okuhara 2013; Reyer et al. 2014; Sánchez-Salguero et al. 2017; Kellomäki et al. 2018).

The photosynthetic rates of plants change seasonally and diurnally in response to climate conditions. Photosynthesis is a temperature-dependent process, and the optimal air temperature for photosynthesis (T_a•opt) varies among species and individuals (Farquhar, Von Caemmerer, and Berry 1980). Photosynthesis temperature dependence is the result of physiological acclimation because T_a•opt is lower for individual plants growing in colder environments (Wieser et al. 2010; Zaka et al. 2016). Soil temperature affects stomatal conductance and transpiration (Foster and Smith 1991; Cochard et al. 2000; Mellander, Bishop, and Lundmark 2004), which further influences photosynthetic rates through stomatal control (DeLucia, Day, and Öquist 1991). Vapor pressure deficit (VPD) and soil water content (SWC) also affect photosynthesis and stomatal closure (Francey and Farquhar 1982; Granier and Bréda 1996; Reich et al. 2018). For example, the photosynthetic rate of black spruce decreases with low SWC due to stomatal closure (Major and Johnsen 1996). Midday depressions in photosynthesis are often observed due to reduced stomatal conductance and relative humidity and increased temperature and VPD (S. B. Zhang et al. 2005). In addition, the assimilative capacity or maximum photosynthetic rate of plants generally decreases with leaf age because of decreases in leaf nitrogen concentration (Reich, Walters, and Ellsworth 1992; Suzuki and Takahashi 2020). Therefore, it is important to examine how climatic conditions affect seasonal and diurnal variations in the photosynthetic rates of leaves of different ages to accurately determine the effects of climate change on the photosynthetic production of plants in cold regions.

Many photosynthesis studies employ biochemical models (Farquhar, Von Caemmerer, and Berry 1980; Kattge and Knorr 2007; Sharkey et al. 2007; Bellasio and Farquhar 2019); however, biochemical models are complex and difficult to apply in field studies, especially in cold regions and at high elevations where measuring in situ photosynthetic rates can be challenging. Alternatively, annual or long-term photosynthetic production can be estimated using empirical photosynthesis models as functions of environmental factors (e.g., light, temperature, VPD) and leaf traits (e.g., leaf mass per area, leaf age; Peri et al. 2002; Liu, Xie, and Li 2019). Therefore, developing empirical models will facilitate evaluations of the effects of climate change on the photosynthetic production of plants in cold regions.

Pinus pumila Regel is a dwarf pine distributed in the uppermost subalpine zone in central Japan, forming the treeline (Okitsu and Ito 1984; Miyajima, Sato, and Takahashi 2007; Ohdo and Takahashi 2020). It is well known that the annual shoot elongation rate of P. pumila is positively correlated with the summer temperatures of the previous year (Takahashi 2006). Therefore, increases in summer temperatures are expected to increase the photosynthetic production of P. pumila. Although Ishida et al. (2001) investigated diurnal changes in the in situ photosynthetic rates of P. pumila on three days, diurnal and seasonal changes over an entire growing season have not been determined. The present study investigated diurnal and seasonal variations in the in situ photosynthetic rate of P. pumila to reveal the effects of environmental factors (light, air and soil temperatures, SWC, and VPD) on photosynthetic production. The purpose of the study was to determine whether diurnal and seasonal variations in the in situ photosynthetic rate of P. pumila can be explained by environmental factors and needle age. This information will be useful for predicting the effects of climate change on photosynthetic production in cold regions.

Methods

Study site

This study was conducted at a P. pumila krummholz (2,770 m.a.s.l., 36°06′ N, 137°33′ E) on Mount Norikura (summit elevation 3,026 m.a.s.l.) in central Japan in 2018 (Figure 1). The krummholz height was about 1 m (Takahashi and Yoshida 2009). The P. pumila krummholz was distributed in a vast area, so the study site is a representative area of this species. Empetrum nigrum var. japonicum was also distributed in the area. The plant height of E. nigrum var. japonicum was only several centimeters. In 2018, mean monthly temperatures of the coldest month (February) and the hottest month (July) were −15.0°C and 13.1°C, respectively, and the mean annual temperature was −0.48°C at the study site, according to Agro-Meteorological Grid Square Data (AMGSD), National Agricultural Research Organization (Ohno et al. 2016; Kominami, Sasaki, and Ohno 2019; 1 × 1 km resolution; https://amu.rd.naro.go.jp). An air temperature of 5.0°C is often used as a threshold for the growing season of plants (Flanagan and Syed 2011; Weijers et al. 2013); the mean monthly temperature at the study site exceeded 5.0°C between June and September in 2018. However, snow covered the forest floor until mid- to late June.

Figure 1. A photo of the study site (2,770 m.a.s.l.). Dwarf pine Pinus pumila is distributed in the vast area near the top of Mount Norikura (summit elevation 3,026 m.a.s.l.) in central Japan.

Photosynthetic measurements

Photosynthetic measurements were performed at weekly intervals from early July to late October for a total of fifteen days. Photosynthetic rates were measured hourly from early morning to late afternoon on each measurement day. However, from summer to autumn, the start of the measurement was delayed and the end time was earlier. The measurement time was shortened due to bad weather and logistic problems on eight of fifteen days.

The maximum longevity of P. pumila needles is five years. However, most four- and five-year-old needles in the study area had fallen, so we measured the net photosynthetic rate of current, one-, two-, and three-year-old needles (Appendix 1; Figure S1). Measurements began in July for one- to three-year-old needles and in August for current year needles because current year needles were immature in July. Net photosynthetic rates (P_n) and stomatal conductance (g_s) were measured using a portable photosynthesis system (model LI-6400; LI-COR Biosciences, Lincoln, NE) equipped with a standard 2 × 3 cm chamber under natural conditions of light intensity, air temperature, relative humidity, and carbon dioxide concentration in ambient air. The soda lime and desiccant columns were bypassed. Air temperature (T_a), carbon dioxide concentration, relative humidity, and VPD in the chamber and photosynthetic photon flux density (PPFD) beside the chamber were measured simultaneously with each photosynthetic measurement. Carbon dioxide concentration ranged between 370 and 430 ppm for each measurement day.

P. pumila has five needles per fascicle, and the needle length ranges between 3 and 7 cm (Kajimoto 1993). Five needles of a fascicle were used to measure the photosynthetic rate at each hour for each needle age. The photosynthetic rate of the same five needles was recorded at 1-minute intervals for 10 to 30 minutes in each hour for each needle age. Different needles were measured each hour for each needle age. Approximately fifty P. pumila trees in the krummholz were used for the measurements. The environmental conditions were the same among these trees because these trees were grown at the same krummholz (Figure 1). The five needles used for the photosynthetic measurement were sampled immediately after the measurement of each hour and transported to the laboratory to measure the projected leaf area of five needles. The projected leaf area of five needles was estimated by the product of the mean needle diameter, needle length in the chamber (3 cm), and the number of needles (five).

Data on outliers and measurement errors were removed, and the total number of data used for the statistical analysis was 8,205 for four needle ages between July and October. Photosynthetic rates and stomatal conductance of each needle age were averaged for each hour. The hourly average of photosynthetic rates and stomatal conductance were 91, 120, 117, and 116 for the current and one-, two-, and three-year-old needles, respectively.

Environmental factors

Daily mean air temperature, daily total solar radiation, daily mean relative humidity, and daily precipitation at the study site in 2018 were obtained from the AMGSD (1 × 1 km resolution). Hourly air temperature and relative humidity were estimated from the AMGSD (Appendix 2). VPD was calculated from the estimated temperature and relative humidity for each hour (Howell and Dusek 1995).

On each photosynthesis measurement day, air temperature and relative humidity were measured at hourly intervals at a height of 1 m in an open site using a thermohygrometer (HOBO H8 Pro Series; Onset Computer Corporation, Bourne, MA) in a radiation shield (CO-RS1; Onset Computer Corporation). PPFD was also measured at 10-minute intervals between 0:00 and 23:50 using a light quantum sensor (IKS-27; Koito Electric Corporation, Shizuoka, Japan) connected to a data logger (MES-101; Koito Electric Corporation). However, measurements of temperature, relative humidity, and PPFD were limited to the period when the photosynthetic measurements were conducted in October.

Soil temperature (T_s) and volumetric SWC (unitless) at a soil depth of 5 cm were measured in the P. pumila krummholz at hourly intervals throughout 2018 using a digital thermometer (TR-52i, T & D Corporation, Matsumoto, Japan) and five soil moisture sensors (ECH₂O EC-5; METER Environment, Pullman, WA) connected to a data logger (Em50; METER Environment), respectively.

Abbreviations and units of measurement for variables measured in this study are listed in Table 1.

Table 1. Definitions and units of abbreviations

Abbreviation	Definition	Unit
age	Needle age	Year
DOY	Day of year	Unitless
gs	Stomatal conductance	mol H2O m^−2 s^−1
Pn	Net photosynthetic rate	μmol CO2 m^−2 s^−1
Pmax	Maximum gross photosynthetic rate	μmol m^−2 s^−1
Rd	Dark respiration rate	μmol m^−2 s^−1
PPFD	Photosynthetic photon flux density	μmol m^−2 s^−1
SWC	Volumetric soil water content	Unitless
SWCopt	Optimal SWC for photosynthesis	Unitless
Ta	Air temperature	°C
Ta·opt	Optimal air temperature for photosynthesis	°C
Ts	Soil temperature	°C
Ts·opt	Optimal soil temperature for photosynthesis	°C
VPD	Vapor pressure deficit	kPa

Statistical analysis

This study constructed an empirical photosynthetic model that was the product of the functions of environmental factors to examine the effects of environmental factors (PPFD, VPD, SWC, T_a, and T_s) and needle age on the net photosynthetic rate (P_n) of P. pumila. Data for three environmental factors (PPFD, VPD, T_a) recorded by the LI-6400 were used for the analysis. In the model, the maximum gross photosynthetic rate (P_max) was multiplied by the function of the relative effect of each factor, and the dark respiration rate (R_d) was subtracted (Equation (1)(1) $P_{\mathrm{n}}=P_{\mathrm{m}\mathrm{a}\mathrm{x}}\times f\left(\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}\right)\times f\left(T_{\mathrm{a}}\right)\times f\left(\mathrm{V}\mathrm{P}\mathrm{D}\right)\times f\left(\mathrm{S}\mathrm{W}\mathrm{C}\right)\times f\left(T_{\mathrm{s}}\right)\times f\left(\mathrm{a}\mathrm{g}\mathrm{e}\right)-R_{\mathrm{d}}.$(1) ):

(1) $P_{\mathrm{n}}=P_{\mathrm{m}\mathrm{a}\mathrm{x}}\times f\left(\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}\right)\times f\left(T_{\mathrm{a}}\right)\times f\left(\mathrm{V}\mathrm{P}\mathrm{D}\right)\times f\left(\mathrm{S}\mathrm{W}\mathrm{C}\right)\times f\left(T_{\mathrm{s}}\right)\times f\left(\mathrm{a}\mathrm{g}\mathrm{e}\right)-R_{\mathrm{d}}.$(1)

The functions of the environmental factors are shown in Equations (2)(2) $f\left(\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}\right)=\hspace{0.17em}\frac{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}}{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}+a}$(2) to (7) and range between 0 and 1:

(2) $f\left(\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}\right)=\hspace{0.17em}\frac{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}}{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}+a}$(2)

(3) $f\left(T_{\mathrm{a}}\right)=1-b\times {\left(T_{\mathrm{a}}-T_{\mathrm{a}\bullet \mathrm{o}\mathrm{p}\mathrm{t}}\right)}^2$(3)

(4) $f\left(\mathrm{V}\mathrm{P}\mathrm{D}\right)=1-c\times \mathrm{V}\mathrm{P}\mathrm{D}$(4)

(5) $f\left(\mathrm{S}\mathrm{W}\mathrm{C}\right)=1-d\times {\left(\mathrm{S}\mathrm{W}\mathrm{C}-\mathrm{S}\mathrm{W}{\mathrm{C}}_{\mathrm{o}\mathrm{p}\mathrm{t}}\right)}^2$(5)

(6) $f\left(T_{\mathrm{s}}\right)=1-e\times {\left(T_{\mathrm{s}}-T_{\mathrm{s}\bullet \mathrm{o}\mathrm{p}\mathrm{t}}\right)}^2$(6)

(7) $f\left(\mathrm{a}\mathrm{g}\mathrm{e}\right)=1-f\times \mathrm{a}\mathrm{g}\mathrm{e}$(7)

where a to f are constants. T_a•opt (Equation (3)(3) $f\left(T_{\mathrm{a}}\right)=1-b\times {\left(T_{\mathrm{a}}-T_{\mathrm{a}\bullet \mathrm{o}\mathrm{p}\mathrm{t}}\right)}^2$(3) ), SWC_opt (Equation (5)(5) $f\left(\mathrm{S}\mathrm{W}\mathrm{C}\right)=1-d\times {\left(\mathrm{S}\mathrm{W}\mathrm{C}-\mathrm{S}\mathrm{W}{\mathrm{C}}_{\mathrm{o}\mathrm{p}\mathrm{t}}\right)}^2$(5) ), and T_s•opt (Equation (6)(6) $f\left(T_{\mathrm{s}}\right)=1-e\times {\left(T_{\mathrm{s}}-T_{\mathrm{s}\bullet \mathrm{o}\mathrm{p}\mathrm{t}}\right)}^2$(6) ) are the optimal air temperature, SWC, and soil temperature for photosynthesis, respectively, and age (Equation (7)(7) $f\left(\mathrm{a}\mathrm{g}\mathrm{e}\right)=1-f\times \mathrm{a}\mathrm{g}\mathrm{e}$(7) ) is needle age. The age of current year needles is zero. All constants were estimated by nonlinear regression analysis using mean photosynthetic rates of four needle ages at each hour on the fifteen measurement days. The Akaike information criterion (AIC) was used to assess the relative effect of each function (Equations (2(2) $f\left(\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}\right)=\hspace{0.17em}\frac{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}}{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}+a}$(2) )‒(7)) and R_d on the whole model (Equation (1)(1) $P_{\mathrm{n}}=P_{\mathrm{m}\mathrm{a}\mathrm{x}}\times f\left(\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}\right)\times f\left(T_{\mathrm{a}}\right)\times f\left(\mathrm{V}\mathrm{P}\mathrm{D}\right)\times f\left(\mathrm{S}\mathrm{W}\mathrm{C}\right)\times f\left(T_{\mathrm{s}}\right)\times f\left(\mathrm{a}\mathrm{g}\mathrm{e}\right)-R_{\mathrm{d}}.$(1) ). The model with the lowest AIC is essentially the best model; however, another model may be considered the same as the best model if the difference in AIC values (ΔAIC) is smaller than two (Burnham and Anderson 2002). If a certain function is important on the whole model, it is expected that the AIC value largely increases when the function is removed from the model. We compared AIC values among fifteen patterns of variable combinations to select a best model.

All statistical analyses were conducted using R v3.5.1 (R Core Team 2018). The nls function in R was used for the nonlinear regression analysis.

Results

Diurnal and seasonal changes in net photosynthetic rate

The daily maximum net photosynthetic rate and stomatal conductance were highest in late August (day of year [DOY] 230) and decreased in October (after DOY 275) for each needle age (Figures 2a, 2b). The seasonal variation in the daily mean soil temperature was smaller than that in the daily mean air temperature (Figures 2c, 2d). Although the daily mean air and soil temperatures and daily solar radiation were highest from mid-July to early August (DOY 190‒220; Figures 2c–2e), the daily maximum net photosynthetic rate and stomatal conductance decreased (Figures 2a, 2b). VPD increased and SWC decreased during this period (DOY 190‒220), with less precipitation (Figures 2f–2h). Stomatal conductance temporarily increased at the end of July (DOY 212) because of the increase of SWC due to precipitation (Figures 2b, 2g, 2h).

Figure 2. Seasonal changes in the daily maximum net photosynthetic rate and daily maximum stomatal conductance of Pinus pumila and six environmental factors at the study site in 2018. (a), (b) The black, blue, green, and red circles indicate the current and one-, two-, and three-year-old needles, respectively. The photosynthetic rate and stomatal conductance of the three-year-old needles on 10 July (DOY 191) were not plotted because of the limited measurement time in late evening between 16:00 and 18:00. The six environmental factors are (c) daily mean air temperature, (d) daily mean soil temperature at a soil depth of 5 cm, (e) daily total solar radiation, (f) daily maximum VPD, (g) daily mean volumetric SWC at a soil depth of 5 cm, and (h) daily total precipitation. Air temperature, solar radiation, precipitation, and relative humidity were obtained from the Agro-Meteorological Grid Square Data, National Agricultural Research Organization (1 × 1 km resolution). VPD was calculated from the temperature and relative humidity.

Decreases in photosynthetic rate and stomatal conductance were observed during daytime in July and August, even when PPFD and temperature were high (Figure 3). VPD increased during daytime with high PPFD (Figure 3). Stomatal conductance decreased with the increase of VPD for each needle age (Appendix 1; Figure S2d, Table S1).

Figure 3. Examples of diurnal changes in the net photosynthetic rate (P_n) and stomatal conductance (g_s) of Pinus pumila and four environmental factors on four days. The black, blue, green, and red lines indicate the current and one-, two-, and three-year-old needles, respectively. The four environmental factors are air temperature (T_a), PPFD, VPD, and volumetric SWC at a soil depth of 5 cm.

Regression model of photosynthetic rates

We estimated the parameters of Equation (1)(1) $P_{\mathrm{n}}=P_{\mathrm{m}\mathrm{a}\mathrm{x}}\times f\left(\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}\right)\times f\left(T_{\mathrm{a}}\right)\times f\left(\mathrm{V}\mathrm{P}\mathrm{D}\right)\times f\left(\mathrm{S}\mathrm{W}\mathrm{C}\right)\times f\left(T_{\mathrm{s}}\right)\times f\left(\mathrm{a}\mathrm{g}\mathrm{e}\right)-R_{\mathrm{d}}.$(1) for the three best models (ΔAIC < 2), and many parameters were statistically significant, except for the dark respiration rate (R_d) of the first and third models, the parameter c of the third model, and the parameter d of all three models (Table 2). The respiration rate (R_d) in the first and third models was a negative value (Table 2), although the respiration rate should be a positive value. Therefore, the second model without R_d was selected:

(8) $P_{\mathrm{n}}=10.22\times \frac{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}}{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}+315.5}\times \left(1-0.003\times {\left(T_a-13.53\right)}^2\right)\times \left(1-12.35\times {\left(\mathrm{S}\mathrm{W}\mathrm{C}-0.22\right)}^2\right)\times \left(1-0.0098\times {\left(T_{\mathrm{s}}-11.14\right)}^2\right)\times \left(1-0.066\times \mathrm{a}\mathrm{g}\mathrm{e}\right).$(8)

Table 2. Estimated parameters of Equation (1)(1) $P_{\mathrm{n}}=P_{\mathrm{m}\mathrm{a}\mathrm{x}}\times f\left(\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}\right)\times f\left(T_{\mathrm{a}}\right)\times f\left(\mathrm{V}\mathrm{P}\mathrm{D}\right)\times f\left(\mathrm{S}\mathrm{W}\mathrm{C}\right)\times f\left(T_{\mathrm{s}}\right)\times f\left(\mathrm{a}\mathrm{g}\mathrm{e}\right)-R_{\mathrm{d}}.$(1) in the main text. Parameter values of three best models are shown. AIC values of the other combinations of parameters are shown in the Appendix 1 (Table S2)

	Model
Parameter	#1	#2	#3
Pmax	10.14**	10.22**	10.15**
a	380.2**	315.5**	382.0**
b	0.0035**	0.0030**	0.0035**
c			−0.011
d	13.79	12.35	13.16
e	0.0105**	0.0098**	0.0100**
f	0.070**	0.066**	0.070**
Rd	−0.42		−0.42
Ta•opt	13.78**	13.53**	13.67**
Ts•opt	11.17**	11.14**	11.31**
SWCopt	0.22*	0.22*	0.23*
AIC	1,561.8	1,562.5	1,563.7
n	444	444	444

*p < .01. **p < .001.

VPD was not included in this model, indicating that the effect of VPD on the photosynthetic rate was smaller than that of the four other factors (PPFD, T_a, T_s, SWC).

The photosynthetic rates of the four needle ages across diurnal and seasonal measurements were plotted with the environmental factors PPFD, air temperature, soil temperature, and SWC (Figure 4). Variations in net photosynthetic rates were large at certain values of each environmental factor due to the effects of the three other factors (e.g., the photosynthetic rate decreased in dark conditions even if the three other factors were suitable for photosynthesis).

Figure 4. Relationships of net photosynthetic rates (P_n) of Pinus pumila with four environmental factors. The black, blue, green, and red circles and lines indicate the current and one-, two-, and three-year-old needles, respectively. The four environmental factors are (a) PPFD, (b) air temperature (T_a), (c) soil temperature (T_s), and (d) volumetric SWC. PPFD beside the chamber and air temperature in the chamber at the photosynthetic measurement were used for the analysis. A regression line in each panel was drawn for each needle age using Equation (8)(8) $P_{\mathrm{n}}=10.22\times \frac{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}}{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}+315.5}\times \left(1-0.003\times {\left(T_a-13.53\right)}^2\right)\times \left(1-12.35\times {\left(\mathrm{S}\mathrm{W}\mathrm{C}-0.22\right)}^2\right)\times \left(1-0.0098\times {\left(T_{\mathrm{s}}-11.14\right)}^2\right)\times \left(1-0.066\times \mathrm{a}\mathrm{g}\mathrm{e}\right).$(8) on the assumption that the three other environmental factors do not limit the photosynthetic rate; that is, the equation in each panel is (a) $P_{\mathrm{n}}=P_{\mathrm{m}\mathrm{a}\mathrm{x}}\times f\left(\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}\right)\times f\left(\mathrm{a}\mathrm{g}\mathrm{e}\right),\hspace{0.17em}$ (b) $P_{\mathrm{n}}=P_{\mathrm{m}\mathrm{a}\mathrm{x}}\times f\left(T_{\mathrm{a}}\right)\times f\left(\mathrm{a}\mathrm{g}\mathrm{e}\right)$, (c) $P_{\mathrm{n}}=P_{\mathrm{m}\mathrm{a}\mathrm{x}}\times f\left(T_{\mathrm{s}}\right)\times f\left(\mathrm{a}\mathrm{g}\mathrm{e}\right)$, and (d) $P_{\mathrm{n}}=P_{\mathrm{m}\mathrm{a}\mathrm{x}}\times f\left(\mathrm{S}\mathrm{W}\mathrm{C}\right)\times f\left(\mathrm{a}\mathrm{g}\mathrm{e}\right)$. Each regression line indicates the change in the maximum photosynthetic rate for a certain environmental factor.

Regression lines (Equation (8)(8) $P_{\mathrm{n}}=10.22\times \frac{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}}{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}+315.5}\times \left(1-0.003\times {\left(T_a-13.53\right)}^2\right)\times \left(1-12.35\times {\left(\mathrm{S}\mathrm{W}\mathrm{C}-0.22\right)}^2\right)\times \left(1-0.0098\times {\left(T_{\mathrm{s}}-11.14\right)}^2\right)\times \left(1-0.066\times \mathrm{a}\mathrm{g}\mathrm{e}\right).$(8) ) were drawn for each environmental factor for four needle ages under the assumption that the three other environmental factors do not limit the photosynthetic rate; that is, each regression line indicates the change in the maximum photosynthetic rate for a certain environmental factor (Figure 4). The photosynthetic rate decreased by 6.6 percent with increasing leaf age for each environmental factor (Figure 4, Table 2). The net photosynthetic rate had unimodal relationships with air and soil temperatures and SWC, reflecting the existence of optimal values for the three factors (Figures 4b–4d). The T_a•opt, T_s•opt, and optimal SWC were 13.53°C, 11.14°C, and 0.22, respectively (the second model in Table 2; Figure 4). Air temperature (T_a) positively correlated with soil temperature (T_s): T_s = 0.563 × T_a + 3.756 (R² = 0.865, p < .001, n = 119) during DOY 182–300 (1 July–27 October). T_s (11.37°C) estimated by the regression at T_a•opt (13.53°C) was almost equal to T_s•opt (11.14°C).

Equation (8)(8) $P_{\mathrm{n}}=10.22\times \frac{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}}{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}+315.5}\times \left(1-0.003\times {\left(T_a-13.53\right)}^2\right)\times \left(1-12.35\times {\left(\mathrm{S}\mathrm{W}\mathrm{C}-0.22\right)}^2\right)\times \left(1-0.0098\times {\left(T_{\mathrm{s}}-11.14\right)}^2\right)\times \left(1-0.066\times \mathrm{a}\mathrm{g}\mathrm{e}\right).$(8) tended to underestimate photosynthetic rates for higher values because the slope of the regression of expected photosynthetic rates to observed photosynthetic rates was smaller than one (Appendix 1; Figure S3). However, seasonal and diurnal changes in net photosynthetic rates were well reproduced by the model for each needle age (Figures 5, 6). Therefore, the photosynthetic rate of P. pumila could be explained by the four environmental factors (PPFD, T_a, T_s, SWC) and needle age.

Figure 5. Comparison between observed values (dots) and expected values (lines) for the seasonal change in daily mean photosynthetic rates of Pinus pumila. The daily mean photosynthetic rate was the mean value of photosynthetic rates at hourly intervals. The panels show (a) current year needles, (b) one-year-old needles, (c) two-year-old needles, and (d) three-year-old needles.

Figure 6. Examples of model fitting (lines) to observed values (dots) of net photosynthetic rates of Pinus pumila on four days in 2018: (a)–(c) 24 July, (d)–(g) 19 August, (h)–(k) 28 September, and (l)–(o) 16 October. The panels show the current and one-, two-, and three-year-old needles, respectively, from left to right.

Discussion

In this study, the in situ photosynthetic rate largely decreased from summer to autumn for each needle age, and the photosynthetic rate was lower for older needles. The seasonal and diurnal changes in P. pumila photosynthesis were related to changes in the air and soil temperatures, SWC, and solar radiation. Solar radiation (or PPFD) decreased from summer to autumn, and it is no doubt that solar radiation greatly influences the diurnal change in the photosynthetic rate of P. pumila (Ishida et al. 2001).

Other studies have also reported similar results in terms of air and soil temperatures; that is, the photosynthetic rate changed according to seasonal changes in air and soil temperatures and the photosynthetic rate increased with warming (Babalola, Boersma, and Youngberg 1968; Kajimoto 1990; Niu et al. 2008; Gunderson et al. 2010; Wieser et al. 2010). Plants usually acclimate to short- and long-term changes in thermal conditions; that is, the maximum photosynthetic rate and optimal temperature for photosynthesis increase in proportion to growth temperature (Gunderson et al. 2010; Wieser et al. 2010; Sendall et al. 2015; Slot and Winter 2017). The optimal air temperature for photosynthesis (13.5°C) of P. pumila was similar to the monthly mean air temperature in July (13.1°C) and August (12.6°C) in 2018, indicating the adaptation to the current thermal conditions. The highest photosynthetic rate was observed at an air temperature of 16°C to 20°C in the chamber, which was higher than the optimal air temperature (13.5°C). This may be due to the difference in temperature between the inside and outside of the chamber (i.e., air temperature in the chamber was 4.3°C higher than outside on average).

The optimal soil temperature (11.1°C) was also similar to the monthly mean soil temperature in July (10.4°C) and August (11.3°C). Soil temperature affected the photosynthetic rate of P. pumila probably because of the stomatal control (DeLucia, Day, and Öquist 1991). Soil temperature affects stomatal conductance through fine root growth, membrane fluidity, and permeability (Ameglio et al. 1990; Carvajal, Cooke, and Clarkson 1996; Pregitzer et al. 2000). In particular, soil temperature often restricts transpiration below a certain threshold temperature (Mellander, Bishop, and Lundmark 2004). In this study, stomatal conductance was always low at a soil temperature below 9°C, and there was a large variation in stomatal conductance at soil temperature above 9°C (Appendix 1; Figure S2c). The soil temperature of 9°C was seasonally about DOY 260 (mid-September; i.e., half a month before the end of the growing season of P. pumila). Although it is not unclear whether 9°C is the threshold soil temperature for the P. pumila, it is possible that soil temperature relates to the end of the growing season of P. pumila by regulating water transportation.

The photosynthetic rate decreased slightly due to a decrease in SWC (i.e., drought stress) in late July and early August when the air temperature and solar radiation were highest and precipitation was lower. Although VPD also increased during this period, VPD was not important for the seasonal and diurnal variations in photosynthetic rates. In this study site, the daily maximum VPD scarcely exceeds 1.5 kPa (Appendix 1; Figure S4); a VPD of 1.5 kPa is not extremely high (cf. Grossiord et al. 2020). Takahashi and Miyajima (2008) measured the predawn water potential in the leaves of four dominant species along an elevational gradient from 1,600 to 2,400 m.a.s.l. on Mount Norikura and found that drought stress hardly occurred, irrespective of elevation. Precipitation in this region is greater at higher elevations, which also endure lower air temperatures and shorter insolation durations (Takahashi, Tokumitsu, and Yasue 2005). In addition, fog often occurs on summer afternoons at high elevations in central Japan. Such local climatic conditions are unlikely to cause drought stress for P. pumila at the treeline (Takahashi 2003). Therefore, the seasonal and diurnal changes in the photosynthetic rate of P. pumila can be explained by the four environmental factors (T_a, T_s, PPFD, SWC) and needle age.

This study showed that diurnal and seasonal variations in the photosynthetic rate of P. pumila could be explained by four environmental factors (T_a, T_s, PPFD, SWC) and needle age. A biochemical process–based model of photosynthesis is often used for the modeling of photosynthesis (Farquhar, Von Caemmerer, and Berry 1980; Kattge and Knorr 2007; Sharkey et al. 2007; Bellasio and Farquhar 2019). However, these models are complex, and parameter estimation can be difficult in field studies, especially at high elevations where the measurement of in situ photosynthetic rates can be challenging. We developed an empirical model of the photosynthetic rate of dwarf pine P. pumila that will facilitate evaluations of the effects of climate change on the photosynthetic production of P. pumila at treelines. In fact, shoot elongation in P. pumila is positively correlated with the mean summer temperatures of the previous year (Takahashi 2006). It is expected that the increase of air and soil temperatures due to climate change will increase the growth of P. pumila by increasing photosynthetic production. However, the model developed in this study is not sufficient to predict the photosynthetic production of P. pumila under climate change because of the acclimation to long-term change in temperatures. Giving the similarity of T_a•opt and T_s•opt with the monthly mean air and soil temperatures of July and August, it is expected that the increase in air and soil temperatures in summer will reach suboptimum levels for the photosynthesis if T_a•opt and T_s•opt of the model (Equation (8)(8) $P_{\mathrm{n}}=10.22\times \frac{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}}{\mathrm{P}\mathrm{P}\mathrm{F}\mathrm{D}+315.5}\times \left(1-0.003\times {\left(T_a-13.53\right)}^2\right)\times \left(1-12.35\times {\left(\mathrm{S}\mathrm{W}\mathrm{C}-0.22\right)}^2\right)\times \left(1-0.0098\times {\left(T_{\mathrm{s}}-11.14\right)}^2\right)\times \left(1-0.066\times \mathrm{a}\mathrm{g}\mathrm{e}\right).$(8) ) do not increase in proportion to the increase in summer temperature, which decreases photosynthetic production in summer (Wentz, Neff, and Suding 2018). Therefore, the empirical model of the photosynthetic rate in this study cannot be used for the prediction of effects of climate change on the photosynthetic production of P. pumila. Further studies are necessary to examine the temperature dependency of photosynthetic characteristics for the precise prediction of climate change effects on the photosynthetic production of P. pumila.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental material for this article can be accessed on the publisher’s website.

Additional information

Funding

This work was partially supported by the joint research program of the Institute for Cosmic Ray Research, University of Tokyo. This work was partially supported by grants from the Japan Society for the Promotion of Science (21K06331) and the Sumitomo Foundation for Environmental Research Projects (193109).