Introduction

Gelidiella acerosa (Forsskål) Feldmann et Hamel (Gelidiales, Rhodophyta) is an alga that can be found commonly on the surf-exposed and moderately wave-sheltered rocks and reefs in the intertidal and upper sublittoral zones in the tropical and subtropical regions of the world (Santelices and Stewart 1985; Abbott 1999; Lin and Freshwater 2008). This species is also known as one of the more important sources of raw material for agar production (Armisen and Galatas 1987; Ganzon-Fortes 1994), contributing to the regional fisheries in the Indo-Pacific region (e.g., Philippines, China, and India; McHugh 1991; Roleda et al. 1997a, b; Ohno and Largo 1998; Zemke-White and Ohno 1999).

In Thailand this species can be commonly found along the coastlines facing the Andaman Sea and the Gulf of Thailand (Lewmanomont and Ogawa 1995; Coppejans et al. 2011; Tsutsui et al. 2012). Despite the importance as an agarophyte, not to mention its high quality and potential as a fisheries resource, G. acerosa has not been harvested in Thailand, unlike species of Gracilaria and Hydropuntia (Edwards and Tam 1984; Praiboon et al. 2006; Bunsom and Prathep 2012).

The stable production of G. acerosa is largely dependent upon various environmental factors driven by the local climate, and especially by the environmental conditions of its habitat (i.e., temperature and irradiance). Indeed, changes in regional seawater temperature are also of some concern, especially along the coast of the Andaman Sea (Brown et al. 1996). For example, seawater temperatures in Phuket, Andaman Sea, have increased at a rate of 0.161 °C per decade (Brown et al. 1996; Tanzil et al. 2009). Presently, the monthly average seawater temperatures in Phuket range from 28 to 30 °C (Brown et al. 1996; Tanzil et al. 2009), suggesting that any further increases in temperature may be detrimental to many of the corals and macroalgae presently found in the region and leading to change in biodiversity.

Despite the commercial importance of G. acerosa and the presence of previous studies on the ecological and physiological aspects from the Philippines and India (Rao and Mehta 1973; Thomas et al. 1975; Chennubhotla et al. 1986; Ganzon-Fortes 1997a, b, 1999; Roleda et al. 1997a, b; Villanueva et al. 1999), there is a little information regarding its ecophysiology, especially with respect to detailed investigations dealing with the response of photosynthesis to temperature and irradiance.

Our recent studies have reported the temperature and irradiance characteristics of photosynthesis of a number of algae species using pulse amplitude modulated (PAM) fluorometer and dissolved oxygen sensors (Lideman et al. 2013; Vo et al. 2014; Watanabe et al. 2014a, b; Fujimoto et al. 2014a, b). We have great interest regarding the temperature response of the effective quantum yield (ΦPSII) in photosystem II and comparing this response to the temperature response of gross photosynthesis. The information derived from these experiments provides detailed insight on how algae respond to temperature, and provides some knowledge of physiologically important ranges of temperature. Indeed, this yield information on the physiological status of the regulatory processes of photosynthetic energy allocation has been evaluated to be a useful tool for the assessment of photosynthetic activity under various environmental conditions (Schreiber et al. 1986; Edwards and Kim 2010; Schagerl and Möstl 2011).

This study was undertaken to provide insights regarding the effects of water temperature and irradiance on the photosynthesis of G. acerosa collected from the coastal areas of Krabi, Thailand, facing the Andaman Sea. More specifically, we focused on elucidating the temperature and irradiance optima of photosynthetic from measurements by oxygen sensors and PAM fluorometry and applied Bayesian techniques to provide estimates of these optima. We expect that this knowledge will lead to sustainable utilization and conservation of the G. acerosa as an important agar resource.

Materials and methods

The algae (ca. 60 individuals) were collected from Railay Beach, Krabi, Thailand (8°0′14.57″ N, 98°50′39.43″E) on January 24, 2013, and used for the measurement of oxygen evolution and PAM fluorometry. Collected algae were stored in 1 L plastic bottles with seawater and transported to Japan via an overnight airline trip in a cooler at about 28 °C, which is approximately the same temperature as the sampling date. The samples were maintained for 1 to 3 days before examination at the Faculty of Fisheries, Kagoshima University, in 1 L flasks containing sterile seawater at a salinity of 33 psu. The flasks were placed in an incubator at a water temperature of 24 °C, and under photosynthetic active radiation (PAR) of ca. 100 μmol photons m−2 s−1 (12:12-h light/dark cycle). The PAR was adjusted by changing the distance of cool fluorescent lamps measuring irradiance using a quantum (2π) sensor (LI-190, LI-250A, LI-COR, USA). Voucher specimens were deposited in the Herbarium of Kagoshima University Museum, Kagoshima.

Effect of temperature on photosynthesis

The measurement methods are described in detail in Terada et al. (2013) and Vo et al. (2014). Briefly, the materials were divided into seven temperature treatment groups (16, 20, 24, 28, 32, 36, 40 °C, N = 5 replicates/treatment) and held under 200 μmol photons m−2 s−1, which is higher than saturation irradiance (E k) as revealed by the photosynthetic–irradiance (P–E) curve (as detailed below). Light was provided by a metal-halide lamp (e.g., Nishihara et al. 2004) and temperature was controlled using a glass water bath. PAR in the photosynthetic experiments was measured in the glass water bath by using a submersible spherical (4π) quantum sensor (LI-193, LI-250A, LI-COR, USA).

Dark respiration and net photosynthetic rates were determined by measuring the dissolved oxygen concentration (in milligram per liter) every 5 min for 30 min after a 30-min pre-incubation to acclimate the samples to each experimental condition, and the slope of the linear regression was determined from the data of 30-min measurements to estimate rates. Dissolved oxygen (DO) was measured using a polarographic sensor and a DO meter (Model 58, and 5100, YSI Incorporated, USA).

Explants used in this experiment were approximately 300 mg wet weight (mgww) and were acclimated overnight with sterilized seawater in the incubator at the same temperature and PAR condition as during pre-incubation (Muraoka et al. 1998; Serisawa et al. 2001). To start the experiment we randomly selected at least five explants and placed them in 100-mL BOD bottles containing sterilized natural seawater. The DO sensors were placed in sterilized natural seawater so that no bubbles were trapped and seawater was continuously stirred during the measurement. The exact volume of the BOD bottles was determined after the experiments and was used in the estimates of photosynthesis and dark respiration rates. Seawater medium was renewed after every measurement to avoid any affects due to the depletion of nutrients and dissolved carbon dioxide.

Irradiance effect on the photosynthesis

Photosynthetic rates were determined at 0, 30, 60, 100, 150, 200, 250, and 500 μmol photons m−2 s−1 (N = 5 replicates/level) at 28 °C and the procedure follows those of the temperature experiment.

Temperature and irradiance effect on photosynthetic parameters

Maxi Imaging-PAM (Heinz Walz, Germany) measurements were based on procedures detailed in Lideman et al. (2013) and Vo et al. (2014). Ten replicates of pre-incubated whole algae were randomly selected and placed in a stainless-steel tray (12 × 10 × 3 cm) containing sterilized seawater. Tray temperature was controlled with a block incubator (BI-535A, Astec, Fukuoka, Japan) by placing the tray on the aluminum block of the incubator. Water temperature in the stainless-steel tray was measured with a thermocouple to confirm that the appropriate temperature condition was achieved.

The maximum effective quantum yields (ΦPSII at 0 μmol photons m−2 s−1) were measured from 14 to 40 °C in 2 °C increments. Each increment in temperature occurred over a 30-min period with an additional 30 min allowed for dark and temperature acclimation. One set of experiments typically took more than 6 h to complete.

Modeling the photosynthetic response to irradiance and temperature

The temperature response of gross respiration and effective quantum yield (ΦPSII) was assumed to follow a non-linear exponential function (Alexandrov and Yamagata 2007; Eq. 1), where y is the response variable, which in this case is the gross photosynthetic rate and K is the temperature in Kelvin. There are four parameters in this model, where y max is the maximal rate of y occurring when the temperature is K opt, H a is the activation energy in kilojoule per mole and H d is the deactivation energy in kilojoule per mole. R in this model is the ideal gas constant, and has a value of 8.314 J kJ mol−1. The gross respiration rates were calculated by adding the dark respiration rates to the net photosynthetic rates, after assuming that the dark respiration rates approximate photorespiration.

$$ y=\frac{y_{\max}\times {H}_{\mathrm{d}}\times \exp \left(\frac{H_{\mathrm{a}}\times \left(K-{K}_{\mathrm{opt}}\right)}{K\times R\times {K}_{\mathrm{opt}}}\right)}{\left({H}_{\mathrm{d}}-{H}_{\mathrm{a}}\times \left(1- \exp \left({H}_{\mathrm{d}}\times \frac{\left(K-{K}_{\mathrm{opt}}\right)}{\left(K\times R\times {K}_{\mathrm{opt}}\right)}\right)\right)\right)} $$
(1)

The relationship between the dark respiration rate and temperature was initially assumed to follow a simple linear model; however, a second order polynomial was also fitted and statistically examined.

The response of photosynthesis to irradiance was examined by modeling the data using an exponential equation (Jassby and Platt 1976; Webb et al. 1974; Platt et al. 1980; Henley 1993) that included a respiration term, which had the form:

$$ {P}_{\mathrm{net}}={P}_{\max}\left(1- \exp \left(-\frac{\alpha }{P_{\max }}E\right)\right)-{R}_{\mathrm{d}} $$
(2)

where, P net was the net O2 production rate, P max was the maximum O2 production rate, α was the initial slope of the photosynthesis versus irradiance curve, E was the incident irradiance, and R d was the dark respiration rate. From this model, the saturation irradiance (E k) was calculated as P max/α and the compensation irradiance (E c) was \( {P}_{\max } \ln \left(\frac{P_{\max }}{\left({R}_{\mathrm{d}}-{P}_{\max}\right)}\right)/\alpha \).

Statistical analysis

Statistical analyses of all the models were done using R version 3.0.1 (R Development Core Team 2013) and model fitting was done using rstan version 2.10 (Stan Development Team 2013). The parameters were examined by fitting the relevant models (i.e., Eq. 1 or Eq. 2) using Bayesian inference, since these models are difficult to fit using least squares or maximum likelihood methods. rstan primarily uses a variant of a Hamiltonian Monte Carlo sampler to construct the posterior distributions of the parameters, and four chains of at least 500,000 samples/chain were generated and assessed for convergence. Weakly informative normal priors were placed on all of the parameters of the model, and a half-cauchy prior was placed on the scale parameter of the models (Gelman 2004, 2006). A generalized linear model was used to analyze the dark respiration–temperature relationship, assuming a normal error distribution. A linear and second-order polynomial was fitted to the data and compared using an F test.

Results

Effect of irradiance on the net photosynthesis

The net photosynthetic rates at 28 °C steadily increased from −10.5 mg O2 gww −1 min−1 (95 % confidence interval (CI) of −13.5 to −7.5 mg O2 gww −1 min−1) at 0 μmol photons m−2 s−1 to a high of 21.2 mg O2 gww −1 min−1 (CI, 18.3–24.2 mg O2 gww −1 min−1) at 500 μmol photons m−2 s−1, respectively (Fig. 1a).

Fig. 1
figure 1

The response of the oxygenic photosynthesis and dark respiration of G. acerosa to temperature and irradiance. a The net photosynthetic rates along an irradiance gradient determined at 28 °C. b The gross photosynthetic rates along a temperature gradient determined at 200 μmol photons m−2 s−1. c The dark respiration rates along a temperature gradient. The dots and vertical lines indicate the mean and 95 % confidence interval of the data (n = 5) and the model lines indicate the expected value. The shaded region in a and b indicates the 95 % Bayesian credible interval of the model, and for c, the shaded region indicates the 95 % confidence interval of the model

Full size image

Given the model (Eq. 2) and the data, the posterior distribution of the parameters to describe the model was determined to be 27.7 mg O2 gww −1 min−1 (95 % Bayesian credible interval (BCI) of 24.3–31.0 mg O2 gww −1 min−1) for the maximum net photosynthetic rate (P max), 7.8 mg O2 gww −1 min−1 (BCI, 5.0–10.5 mg O2 gww −1 min−1) for the dark respiration rate (R d), and 0.25 mg O2 gww −1 min−1 (μmol photons m−2 s−1)−1 (BCI, 0.17–0.34 mg O2 gww −1 min−1 (μmol photons m−2 s−1)−1) for the initial slope (α) of the model, respectively.

From these parameters, we estimated that the 95 % BCI of the compensation irradiance (E c) was from 29 to 45 μmol photons m−2 s−1, with a mean value of 37 μmol photons m−2 s−1, and that the 95 % BCI of the saturation irradiance (E k) was from 52 to 128 μmol photons m−2 s−1, with a mean value of 83 μmol photons m−2 s−1, respectively.

Effect of temperature on the gross photosynthesis and dark respiration

The measured gross photosynthetic rates was highest at 28 °C, with a maximum value of 17.4 mg O2 gww −1 min−1 (CI, 12.4–22.4) and lowest at 40 °C with a minimum value of 3.3 mg O2 gww −1 min−1 (CI, 1.5–5.1 mg O2 gww −1 min−1) (Fig. 1b).

The model (Eq. 1) fit to the data indicated that the optimal temperature (T GPopt ) where the maximal gross photosynthetic rates (GP max = 18.4 mg O2mgww chl-a −1 min−1 (BCI, 15.9–20.8 mg O2mgww chl-a −1 min−1)) would occur was 29.2 °C (BCI, 26.8–32.1 °C). The activation energy (H GPa ) was determined to be 62.1 kJ mol−1 (BCI, 28.3–111 kJ mol−1) and the deactivation energy (H GPd ) was 296 kJ mol−1 (BCI, 197–533 kJ mol−1).

The measured dark respiration rates increased in magnitude from 0.9 ± 0.9 mg O2 gww −1 min−1 °C−1 (mean ± SE) at 16 °C to 6.7 ± 1.4 mg O2 gww −1 min−1 °C−1 at 32 °C (Fig. 1c). Two generalized linear models of the dark respiration rate were fitted to the data assuming that linear model was the null model and the second order model was the alternative. An F test revealed that the second order model best fit the data, with F (1,38) = 9.27 (P < 0.01). Therefore, we assumed that the dark respiration rates and temperature relationship could be modeled in the form, y = b 0 + b 1 x + b 2 x 2, were y is the respiration rate and x is the temperature. The coefficients b0, b1, and b2 were all significant (P < 0.01), and were estimated to be 13.8, −1.24, and 0.019, respectively. Hence, the expected values of the respiration rates were −0.49 mg O2 gww −1 min−1 at 16 °C and −0.93 mg O2 gww −1 min−1 at 40 °C.

Effect of temperature on and maximum quantum yield (ΦPSII)

The temperature response of the effective quantum yield (ΦPSII) at 0 μmol photons m−2 s−1 was similar to that of the photosynthesis–temperature relationship (Fig. 2). Indeed, ΦPSII was low at low temperatures, and the measured value (mean ± SE) was 0.22 ± 0.01 at 14 °C. However, the ΦPSII climbed to a peak of 0.59 ± 0.01 at 26 °C, and then decreased to another low of 0.20 ± 0.04 by 40 °C.

Fig. 2
figure 2

The temperature response of the maximum effective quantum yield (ΦPSII at 0 μmol photons m−2 s−1) in G. acerosa. The dots and vertical bars indicate the mean value of the data (n = 10) and the 95 % confidence interval of the mean. The model lines indicate the expected value. The shaded regions indicate the 95 % Bayesian confidence interval of the model

Full size image

Given the model (Eq. 1) and the data, the maximum ΦPSII (Φ PS I I max) occurring at the optimal temperature (T ΦPSIIopt  = 27.7 (BCI, 26.5–28.9) °C) was 0.57 (BCI, 0.55–0.58). The activation and deactivation energy was determined to be 10.4 (6.4–15.6) kJ mol−1 and 247 (194–308) kJ mol−1, respectively.

Discussion

In Thailand, G. acerosa can be found on the substrata throughout the year and is reported to be particularly abundant during the rainy season (e.g., October; Prathep 2005). The laboratory study of G. acerosa, clearly revealed that temperature influences their photosynthetic activity in addition to the rates of dark respiration. Indeed, the temperature response of oxygen evolution (gross photosynthetic rate) showed a characteristic dome-like shape, and the highest gross photosynthetic rate occurred at 29.2 °C (Fig. 1b). Likewise, an empirical second-order linear model suggested that respiration rates increased nonlinearly with temperature and are restricted at higher temperatures (Fig. 1c). Although this model is difficult to interpret, perhaps the restriction of the respiration rate above 36 °C might be related to denatured enzymes and damage of protein structures that was caused by the thermal stress. More experiments will be needed to confirm this hypothesis.

Nevertheless, the apparent optimal temperatures determined for photosynthetic activity was within the ranges reported for their natural habitat in the Philippines (27 to 31 °C; Roleda et al. 1997b), Taiwan (18 to 25 °C; Lin and Freshwater 2008), and the Ryukyu Islands, Japan (18 to 28 °C; Chihara and Kamura 1963; Vo et al. 2014). It is relevant to note that in a prior study of the macroalgal community found in the lower intertidal zone and near the collection site, the seawater temperatures were recorded to range from 27.8 to 34.5 °C (Prathep et al. 2007).

It is also notable that the maximum effective quantum yield (ΦPSII at 0 μmol photons m−2 s−1) was temperature dependent, given that it gradually increased with increasing temperature until 27 °C, and then quickly decreased above 30 °C (Fig. 2). More importantly, it closely corresponded with the temperature response of oxygen evolution (gross photosynthesis) and the optimum temperature for maximum effective quantum yield and oxygen evolution were almost identical.

Previous studies also indicated that the maximum effective quantum yield was temperature dependent and its optimum temperature was almost related to the temperature in the natural habitat during the period when each species or sexual phase occurred, as reported for Pyropia tenera (Kjellman) Kikuchi et al. (Watanabe et al. 2014a) and Agardhiella subulata (C. Agardh) Kraft et Wynne (Vo et al. 2014). For this study, the maximum effective quantum yield also corresponded with those of the natural habitat in the Andaman Sea coast of Thailand (Prathep et al. 2007).

Despite the temperature response in the ΦPSII in the present study, the maximum quantum yield (Fv/Fm) is typically known as being temperature independent. Nevertheless, the effective quantum yield (ΦPSII) is temperature dependent in physiologically appropriate temperature ranges and will vary with species (Dongsansuk et al. 2013; Salvucci and Crafts-Brandner 2004). As we mentioned in Watanabe et al. (2014a), the reduction in gross photosynthesis at high temperatures are partly attributable to increased respiration rates, as demonstrated by the dark respiration experiments and in part to reduced CO2 uptake as rubisco activity declines (Salvucci and Crafts-Brandner 2004). However, the mechanisms associated with the decline in ΦPSII are uncertain. Typically in higher plants, the decrease of the ΦPSII at high temperatures can be associated PSII reaction centers inactivation (Roháček 2002).

However, unlike higher plants and other orders of macroalgae, red algae are known to have light harvesting antennae that include phycobilisomes, which capture a wide band (580 to 680 nm) of light energy (Yokono et al. 2011; Larkum 2003; Larkum and Vesk 2003). Approximately half of the energy captured by phycobilisomes is then transferred to photosystem II (PSII), and when all the photochemical traps of PSII are closed, up to 100 % of this energy can be transferred to PSI (Kowalczyk et al. 2013). Therefore, for high-temperature PAM experiments, the decreases in ΦPSII might be due to changes in the state of PSII or to the proportion of energy transferred to PSI and PSII.

Despite the concerns of tolerable temperature in the tropical waters, G. acerosa was considered to be adapted to the current range of seawater temperatures in this region. This insight might be related to the habitat characteristics of this species that is growing on the lower intertidal zone, which is exposed by the lowest spring tide. In the tropical region, seawater temperature in the shallow intertidal zone is somewhat unstable (Prathep et al. 2007; Tanzil et al. 2009). Direct sunlight in shallow waters undergoing tidal fluctuation have different thermal properties associated with decreased water mass during shallow periods and increased water mass during relatively deep periods. Indeed, characteristics of thermal and desiccation tolerance have been reported for some intertidal species (Dring and Brown 1982; Lüning 1984; Smith and Berry 1986; Bell 1993; Kübler and Davison 1993; Davison and Perason 1996) including Gelidium pusillum (Stackhouse) Le Jolis from Thailand (Prathep et al. 2009). It is also likely that G. acerosa can tolerate short-term desiccation, because it is sometimes exposed from the seawater by the lowest spring tide. However, the eventual bleaching of this species after long periods of emersion were also reported by Ganzon-Fortes (1997a, b), suggesting that further studies of the desiccation tolerance must be required in future.

The experiments revealed that photosynthetic rates of this species were optimal within the range of 26 to 32 °C, but are close to their marginal temperatures in the natural environment. If summer seawater temperatures continue to rise in these regions, as has been shown in previous studies (Tanzil et al. 2009), the continuous and compounded effects of high temperature, direct sunlight, and emersion may lead to the decreased abundance of this species (Roleda et al. 1997b).

Regarding the PAR response of G. acerosa, the experiments revealed that the net photosynthetic rate at 28 °C was strongly dependent on irradiance when PAR < 83 μmol photons m−2 s−1, which is the E k determined from the model (Fig. 1a). Therefore, we can hypothesize that the vertical distributional limit of these species is connected to the magnitude of E k. In contrast, we did not detect any evidence of photoinhibition, which was similar to other intertidal species (e.g., Gracilariopsis chorda (Holmes) Ohmi and P. tenera) determined from our previous studies (Terada et al. 2013; Watanabe et al. 2014a). Consider that G. acerosa can be found in the lower intertidal and upper sublittoral zones (i.e., up to 2 m in deep), and therefore, it is likely that this species is periodically exposed to direct sunlight during the daytime low tide. Therefore, the evidence for the lack of photoinhibition remains inconclusive, given that the experimental PAR did not exceed 500 μmol photons m−2 s−1.

We stress that the results of this study are based on short-term laboratory measurements of photosynthesis and respiration using both dissolved oxygen and fluorescence measurements, and that the temperature characteristics of these species remains to be fully elucidated. For example, temperature acclimation has not been examined for photosynthesis and respiration, which was observed in higher plants (Atkin and Tjoelker 2003). The results are extrapolated with caution, and we acknowledge that studies of longer timescale be conducted to verify our hypotheses, especially regarding climate change induced increases in temperature.

In conclusion, we recommend the continued monitoring of seawater temperature to better forecast the future state of G. acerosa growing in the tolerable environments and to include dissolved oxygen derived estimates of photosynthetic activity, given that fundamentally different processes of light energy harvesting and transfer of the photosynthetic apparatus in red algae.