Can the exceptional chilling tolerance of C₄ photosynthesis found in Miscanthus × giganteus be exceeded? Screening of a novel Miscanthus Japanese germplasm collection

Abstract

Background and Aims A clone of the hybrid perennial C4 grass Miscanthus × giganteus (Mxg) is known for achieving exceptionally high rates of leaf CO₂ uptake during chilling. This is a requisite of success in the early spring, as is the ability of the leaves to survive occasional frosts. The aim of this study was to search for genotypes with greater potential than Mxg for photosynthesis and frost survival under these conditions.

Methods A total of 864 accessions representing 164 local populations of M. sacchariflorus (Msa), M. sinensis (Msi) and M. tinctorius (Mti) collected across Japan were studied. Accessions whose leaves survived a natural late frost in the field were screened for high maximum photosystem II efficiency (F_v/F_m) following chilling weather, as an indicator of their capacity for light-limited photosynthesis. Those showing the highest F_v/F_m were transferred to a high-light-controlled environment and maintained at chilling temperatures, where they were further screened for their capacities for high-light-limited and light-saturated leaf uptake of CO₂ (Φ_CO2,max and A_sat, respectively).

Key Results For the first time, relatives of Mxg with significantly superior capacities for photosynthesis at chilling temperatures were identified. Msa accession ‘73/2’ developed leaves in the spring that survived night-time frost, and during growth under chilling maintained a statistically significant 79 % higher Φ_CO2,max, as a measure of light-limited photosynthesis, and a 70 % higher A_sat, as a measure of light-saturated photosynthesis. A second Msa accession, ‘73/3’ also showed significantly higher rates of leaf uptake of CO₂.

Conclusions As remarkable as Mxg has proved in its chilling tolerance of C₄ photosynthesis, this study shows that there is still value and potential in searching for yet more superior tolerance. Msa accession ‘73/2’ shows rates of light-limited and light-saturated photosynthesis at chilling temperatures that are comparable with those of the most cold-tolerant C₃ species. This adds further proof to the thesis that C₄ photosynthesis is not inherently limited to warm climates.

INTRODUCTION

C₄ photosynthesis is recognized as providing potentially greater light, nitrogen and water use efficiency than C₃, and is therefore a desired property for sustainable bioenergy crops (Heaton et al., 2004, 2008). However, in practice, the C₄ photosynthetic apparatus is usually impaired by chilling (≤15 °C), limiting realization of these advantages to tropical and sub-tropical climates or to the warm months of temperate climates. Most C₄ grasses emerge significantly later than C₃ perennial grasses in the same environment and, as a result, are unable to use a significant portion of the potential growing season. To overcome this limitation, germplasm is needed that shows early emergence in forming leaves that are photosynthetically competent at chilling temperatures. Since night frosts may occur during spring, leaves must also be able to survive moderate freezing. Under field conditions, total carbon assimilation requires maintenance of a high efficiency under both light-limited and light-saturated conditions. Very different factors control these two efficiencies, although both are impaired in most C₄ species during chilling in the light (Long and Spence, 2013). Impairment of light-limited photosynthesis typically results from loss of functional D1 protein at photosystem II (PSII), impairment in the synthesis of key proteins of the photosynthetic membrane and/or accumulation of the de-epoxidated xanthophyll, zeaxanthin. The latter can cause dissipation of most absorbed light energy as heat, so decreasing the efficiency of PSII, but protecting it from longer lasting damage (Baker, 2008). In contrast, impairment of light-saturated C₄ photosynthesis has been attributed to the cold lability of pyruvate-Pi-dikinase (PPDK) and/or insufficient ribulose-1:5-bisphosphate carboxylase/oxygenase (Rubisco) (Naidu et al., 2003; Naidu and Long, 2004; Farage et al., 2006; Sage and Kubien, 2007; Baker, 2008; Wang et al., 2008; Sage et al., 2011; Long and Spence, 2013). In summary, in searching for more chilling-tolerant germplasm, three properties are needed; the ability of the leaves to: (1) survive night frosts; (2) maintain a high efficiency of light-limited photosynthesis expressed as the maximum quantum yield of PSII (Φ_PSII); and (3) achieve a high light-saturated rate of leaf CO₂ uptake (A_sat) in chilling conditions.

Among C₄ plants, one clone of a sterile triploid form of Miscanthus × giganteus has proved to be exceptional in being able to maintain high photosynthetic efficiency under chilling conditions (Long and Spence, 2013). In England, this clone has achieved harvested yields of >20 t(dry matter) ha^–1 (Beale and Long, 1995), and has shown little loss of maximum quantum yield under chilling conditions (Beale et al., 1996) while achieving the high nitrogen and water use efficiencies associated with C₄ photosynthesis (Beale and Long, 1997; Beale et al., 1999; Christian et al., 2008). The first replicated trials of this genotype in North America gave even higher yields, averaging 23 t ha^–1 over 8–10 years across seven sites in Illinois, with yields exceeding 40 t ha^–1 at individual sites in the best years (Heaton et al., 2008; Arundale et al., 2014). As a result of these, and similar trials, it has become a major emerging bioenergy crop in both Europe and North America (Jones and Walsh, 2001; Clifton-Brown et al., 2004; Heaton et al., 2010; Somerville et al., 2010; Jones, 2011). In a side by side replicated comparison with a modern maize cultivar in the US Corn Belt, M. ×giganteus produced 59 % more biomass. This was shown to result from its ability to form and maintain photosynthetically functional leaves under chilling conditions, allowing it to capture and utilize light earlier in the spring and into the autumn, by comparison with its close relative maize (Dohleman and Long, 2009). Controlled environment analysis has shown that on transfer to chilling conditions, or during growth under chilling conditions (10–15 °C), photosynthetic capacity in the closely related species of the same grass tribe – maize, sorghum and sugarcane – declines until leaves are virtually unable to assimilate CO₂ (Long and Spence, 2013; Głowacka et al., 2014a). This loss is characterized by a decline in light-limited photosynthesis accompanied by loss of PSII efficiency and in light-saturated photosynthesis accompanied by loss of both PPDK and Rubisco (Long and Spence, 2013). In contrast, photosynthesis in M. ×giganteus acclimates to maintain a high photosynthetic rate, corresponding to maintenance of a high PSII efficiency, an increase in the amount of the protein PPDK and its transcript, (Wang et al., 2008) and an increase in transcripts for several key chloroplast membrane proteins (Spence et al., 2014).

Most analyses, trials and commercial deployment of M. ×giganteus appear to be with a single clone that has sometimes been termed the ‘Illinois’ clone in the USA (USDA-NRCS, 2011). It is assumed to have originated from a single plant most probably collected in southern Honshu, Japan, in the 1930s, which was transferred first to Denmark, and then distributed onward to various botanical gardens in Europe and the USA (Greef and Deuter, 1993; Linde-Laursen, 1993; Hodkinson and Renvoize, 2001; Hodkinson et al., 2002a; Głowacka et al., 2014b). Since the clone was grown for many years at Hornum in Jutland, Denmark (Jørgensen, 1997), a more appropriate term is the ‘Hornum’ clone or heritage clone, and it is referred to here as Mxg. Hodkinson and Renvoize (2001) first defined Miscanthus × giganteus J.M.Greef & Deuter ex Hodk. & Renvoize, as a nothospecies hybrid between M. sinensis (Msi) and M. sacchariflorus (Msa). Although Msa can be both diploid and tetraploid, the ‘Hornum’ clone (Mxg) is a sterile triploid implying that the Msa parent was tetraploid, as confirmed by cytological analysis (Hodkinson et al., 2002b). Agronomic field trials of this clone indicate that it is a sustainable bioenergy crop that can realize the advantages of C₄ photosynthesis in cool temperate climates, such as that of Western Europe, or during the cooler weather of spring and autumn in continental/microthermal climates, such as that of the North American corn belt (Beale and Long, 1995, 1997; Beale et al., 1999; Lewandowski et al., 2000; Clifton-Brown et al., 2004; Lewandowski and Schmidt, 2006). Despite these qualities, the widespread use of a single clone opens up obvious risks of epidemics of pests and diseases (Prasifka et al., 2009; Ahonsi et al., 2010; Bradshaw et al., 2010). A single genotype is clearly not going to be the optimal form of a crop species for all climates and soils. Further, given the geographic ranges of the two parent species, extending through Hokkaido and into southern Siberia, it is unlikely that a clone collected from the warmer climate of lowland Honshu will realize the full chilling tolerance of this clade. It would seem likely that the northward spread of the two parental species into colder climates has involved increased genetic tolerance to chilling. However, screening of chilling tolerance in collections of the two parent species has failed to show germplasm with a clear advantage in chilling tolerance of photosynthesis over the ‘Hornum’ clone (Purdy et al., 2013; Głowacka et al., 2014a). This has raised the possibility that its tolerance results from hybrid vigour. However, a recent analysis of newly synthesized hybrids between Msi and Msa found these to be significantly inferior in chilling tolerance to the ‘Hornum’ clone, showing that heterosis is not necessarily the basis of this tolerance (Głowacka et al., 2014a). An alternative interpretation is that the germplasm that has been screened is limited to material imported to Europe and North America for horticultural purposes and that it is too limited to provide further chilling tolerance. This material was most probably selected for aesthetic qualities for landscaping rather than for productivity. Indeed a recent analysis using a high density of nuclear and plastid DNA markers has shown that most of the horticultural material is closely associated with native genotypes found in southern Japan, and represents a very narrow portion of the available genetic diversity (Clark et al., 2014). This suggests that with a wider range of germplasm it might still be worth pursuing the following question. Are there genotypes with greater photosynthetic capacity under chilling conditions and with greater capacity to form leaves that survive spring frosts than the ‘Hornum’ clone of M. ×giganteus?

As noted above, in 1995 a collection was made in Japan of material appearing to be productive and suitable for cultivation to provide thatching material in Denmark (Kjeldsen et al., 1999). This collection of 864 accessions of Msa, Msi and a close relative, M. tinctorius (Mti) was planted into field plots of the Experimental Farm of Aarhus University in central Jutland. The original clone of M. ×giganteus was also represented.

In early May 2012 a late frost visibly scorched much of the material, but some accessions remained largely green. From this fortuitous screen, we selected the 29 accessions which showed little or no frost damage to their leaves, providing a unique opportunity to assess this reduced and tractable number of genotypes from the original 865 for both light-limited and light-saturated photosynthetic capacity during chilling. In the study presented here, these remaining plants were screened in the field for maintenance of a high maximum efficiency of PSII (F_v/F_m) under chilling conditions, as a proxy for light-limited photosynthetic capacity. Those accessions showing the highest F_v/F_m values were then transferred to controlled environments, to maintain chilling conditions, and to conduct a final screen of A_sat as a measure of light-saturated photosynthetic capacity. The initial slope of the response of leaf CO₂ uptake to photon flux (Φ_CO2,max) was also determined on a sub-set of accessions as a more exact, but lower throughput screen, of photosynthetic capacity under light-limited conditions This study tests the hypothesis that there is germplasm with greater chilling tolerance of light-limited and light-saturated photosynthesis than the ‘Hornum’ clone of M. ×giganteus, hereafter termed Mxg.

MATERIALS AND METHODS

Plant material in the field

In 1995, 164 local populations of Miscanthus sacchariflorus (Msa), M. sinensis (Msi) and M. tinctorius (Mti) were collected as seed, or in a few cases rhizomes, from across Japan (Kjeldsen et al., 1999). The populations were raised in a greenhouse for quarantine purposes. They were then transferred to the field and planted 75 cm apart in spring 1996, except for accessions of M. sacchariflorus, which were planted 150 cm apart due to their spreading habit. Since these species are self-incompatible, each seed collection yielded several different genotypes, giving a total of 864 distinct plants. Miscanthus ×giganteus J.M.Greef & Deuter ex Hodk. & Renvoize ‘Hornum’ (Mxg) was also included among these for comparison and was planted from rhizomes (Larsen et al., 2014). These trials were conducted on a sandy loam soil (typic Fragiudalf; USDA soil taxonomy) at Aarhus University Research Centre at Foulum, in central Jutland, Denmark (56 °30'N, 9 °35'E) (Kjeldsen et al., 1999; Clifton-Brown et al., 2001; Jørgensen et al., 2003). Soil temperature at a depth of 10 cm and air temperature at a height of 20 and 150 cm were recorded every 10 min by the meteorological station which was located within the trial site (Fig. 1).

Fig. 1.

Soil and air temperatures before and during the period of sampling. Temperatures were measured at 10 min intervals at 10 cm soil depth, and at 20 cm and 150 cm above the soil surface in the centre of the Miscanthus genotype trials at Foulum. The arrows indicate days when F_v/F_m measurements were made. The plants with a high F_v/F_m were transferred to a controlled environment on 25 May.

Field measurements of F_v/F_m

After the late frosts in May 2012, the plants showing the least frost damage, based on visual necrosis of the leaves, were chosen for assessment of maximum efficiency of PSII (F_v/F_m) in situ by chlorophyll fluorescence. At this point, the plants were still experiencing chilling (≤15 °C) on most days, and measurements were taken at the end of the photoperiod (Fig. 1). The most recent fully expanded leaves, as judged by ligule emergence, were dark adapted by enclosing the lamina with ‘dark adaptation’ clips (9964-091; LI-COR, Lincoln, NE, USA). To minimize variation in the light exposure received prior to measurement, portions of lamina were selected that were naturally horizontal and not shaded by any other leaves. On completion of 2 h dark adaptation, the clip enclosing the lamina was attached to the fluorometer chamber via a light-tight seal, the clip shutter opened and a saturating pulse applied, to measure the ratio of variable to maximal leaf chlorophyll fluorescence (F_v/F_m) with an attached pulse amplitude-modulated fluorometer (LI-6400-40; LI-COR). If the accession in the field trial was represented by a single plant, four replicate leaves each on different shoots were measured. Otherwise, four replicate plants of the clone were measured. Minimum fluorescence (F_o) and maximum fluorescence (F_m) were recorded, to compute F_v (F_v = F_m – F_o) and F_v/F_m, i.e. Φ_PSII,max, the maximum dark-adapted quantum efficiency of PSII (Maxwell and Johnson, 2000).

Leaf photosynthetic gas exchange

A sub-set of the 11 accessions that showed the highest F_v/F_m during mid-May was chosen to analyse capacity for leaf photosynthetic CO₂ assimilation under chilling conditions. A single division of each of these accessions was transferred into a 6 L pot in their field soil, taking care to minimize any damage to the emerging root system. They were then transferred to a walk-in controlled-environment room (MB-Teknik, Brøndbytoften, Denmark) under a 15 °C/15 °C 14 h day/10 h night cycle at 850 µmol photon flux m^–2 s^–1 (Powerstar HQI-BT 400 W/D PRO Daylight lamps, OSRAM) and a relative humidity of 65 %. To avoid confounding variation in the environment within the chamber with accession, the position of each plant within the chamber was changed following a randomized design, each day. Leaf photosynthetic gas exchange was measured on the uppermost leaf that was fully expanded on the shoot, as judged by ligule emergence, using an open gas exchange system incorporating differential infrared CO₂ and water vapour analysers (LI-6400, LI-COR). In this system, the leaf was enclosed in a controlled-environment cuvette which reproduced the temperature and humidity conditions of the controlled-environment room. Measurements were conducted under ambient air (21 % O₂) at 390 µmol mol^–1 [CO₂], a leaf temperature of 15 °C, 1500 µmol m^–2 s^–1 photon flux and 65 % relative humidity. Actinic light in this leaf cuvette was supplied by light-emitting diodes (90 % red light, 630 nm; 10 % blue light, 470 nm).

All measurements were recorded during the daylight hours on light-adapted leaves when steady-state rates of leaf CO₂ uptake were obtained after 20–25 min. For each accession, four replicate leaves, each on independent shoots, were measured. Net leaf CO₂ uptake per unit leaf area (A), stomatal conductance to water vapour (g_s) and intercellular CO₂ concentration (c_i) were calculated, as described previously (von Caemmerer and Farquhar, 1981; Long and Bernacchi, 2003).

A/Q and A/c_i responses

For the five accessions which showed the highest A, as measured above, further measurements were made to determine the responses of A to Q and to intercellular CO₂ concentration (c_i) to determine the mechanistic basis of variation in A_sat and Φ_CO2,max between accessions. The A/Q and A/c_i responses were measured on the same population of leaves within the walk-in controlled-environment plant growth chamber.

Responses of A to Q were obtained starting at a photon flux of 1500 µmol m^–2 s^–1, which was then decreased through nine steps to complete darkness. At each light level, sufficient time was allowed for a new steady-state A to be attained. In order to describe the response of A to varied Q, a non-rectangular hyperbola was fitted to the nine data points for each leaf. The parameters were: the initial slope of the response of A to Q (Φ_CO2,max), the light-saturated rate of A (A_sat; asymptote), the convexity of the transition from light-limited to light-saturated A (θ; acuteness of the transition from light-limited to light-saturated photosynthesis with increasing photon flux) and the mitochondrial respiration rate in the light (R_d; y-intercept). These were obtained by fitting the quadratic equation describing this four-parameter non-rectangular hyperbola (Long and Hällgren, 1993) by non-linear regression analysis using the NLIN procedure with the MARQUARDT fitting method (Marqurdt, 1963; SAS v. 9.3, SAS Institute, Cary, NC, USA). Linear regression of the response of A to Q (Q = 0–100 µmol m^–2 s¹) was also used to obtain the initial slope (Φ_CO2,max) fitted with the REG procedure (SAS v. 9.3). This second method avoided any influence of variation in A_sat and θ on the estimation of Φ_CO2,max; data are presented only for this method, but agreed well with the first method. The mean and standard error of Φ_CO2,max, A_sat, θ and R_d for each accession were determined from the four replicate shoots.

Measurements of A vs. c_i were made at 15 °C leaf temperature, 1500 µmol m^–2 s^–1 photon flux and 65 % relative humidity starting at a cuvette [CO₂] of 390 µmol mol^–1. Once a steady-state A was obtained, [CO₂] was decreased stepwise to 35 µmol mol^–1. The cuvette [CO₂] was then returned to 390 µmol mol^–1 and increased stepwise to 1500 µmol mol^–1. Each curve consisted of nine separate measurements to which a non-rectangular hyperbola was fit. The initial slope of the A/c_i response reflects the in vivo capacity for phosphoenol-pyruvate (PEP) carboxylation (V_pmax), while the horizontal asymptote of a non-rectangular hyperbolic function for each A/c_i curve was used for estimation of the [CO₂]-saturated rate of A (V_max) determined by capacity for PEP regeneration and leakage of CO₂ from the bundle sheath (von Caemmerer, 2000). Both of these parameters when determined from leaf CO₂ uptake rates will be reduced by leakage of CO₂ from the bundle sheath (von Caemmerer, 2000). Values given are the means of 3–4 replicate shoots.

Statistical analysis

All statistical analyses were performed in SAS (SAS v. 9.3). The statistical significance of variation in F_v/F_m, A, g_s, c_i/c_a, A_sat, Φ_CO2max, θ, R_d, V_max and V_pmax across all accessions were determined by one-way analysis of variance in GLM. Where a significant effect of accession was detected, the difference between the mean for each accession and that of the Mxg clone was assessed in LSMEANS using Dunnett’s test at P ≤ 0·1; P ≤ 0·05; P ≤ 0·01.

RESULTS

Temperature at the field site

Shoots of the Miscanthus accessions began to emerge in late April. This was followed by three successive frosts on the nights of 6–8 May (Fig. 1). The temperature at 20 cm above ground level, the height of most of the emerged shoots, dropped to −2·0, −2·8 and −3·6 °C, respectively, and remained below freezing for 6–8 h on these nights (Fig. 1). After this period, air temperatures remained above freezing, but largely within the chilling range (≤15 °C), until late May. For most of the 18 and 19 May, daytime air temperatures were ≤15 °C (Fig. 1). Dark-adapted F_v/F_m was measured in the late evenings of these two days (Table 1).

Table 1.

List of 29 Miscanthus accessions that had emerged with green leaves and showed only little visible damage following the heavy frosts of 6–8 May (Fig. 1), out of a total of 865

Accession	Collected material	Origin/location of collection in Japan	Fv/Fm
Msi ‘56/1’	S	5 km south-west from Kanazawa, Honshu (280 m)	0·655 (0·010)
Msi ‘56/2’	S	5 km south-west from Kanazawa, Honshu (280 m)	0·641 (0·006)
Msa ‘44/1’	S	Biratori (south-west from Sapporo), Hokkaido	0·641 (0·016)
Msi ‘56/3’	S	5 km south-west from Kanazawa, Honshu (280 m)	0·637 (0·013)
Msa ‘44/2’	S	Biratori (south-west from Sapporo), Hokkaido	0·620 (0·018)
Msa ‘73/3’	S	Hakusan National Park, Honshu (900 m)	0·618 (0·016)
Msi ‘22/1’	S	Yoichi, Hokkaido; OUAV collection	0·617 (0·009)
Msa ‘73/2’	S	Hakusan National Park, Honshu (900 m)	0·611 (0·015)
Msi ‘11/1’	S	Iwanai, Hokkaido; OUAV collection	0·604 (0·022)
Mti ‘132/1’	S	Ainukura, Honshu (350 m)	0·598 (0·021)
Mxg	S	Hornum, Denmark (S. Honshu, site unknown)	0·598 (0·038)
Msi ‘54/1’	S	5 km south-west from Kanazawa, Honshu (280 m)	0·577 (0·017)
Mti ‘119/1’	S	South of Ogimachi, Shirakawa-Go, Honshu (500 m)	0·573 (0·013)
Mti ‘105/1’	R	South from Shirakawa, Honshu (600 m)	0·562 (0·015)
Mti ‘122/1’	S	South of Ogimachi, Shirakawa-Go, Honshu (500 m)	0·562 (0·014)
Msi ‘24/1’	S	North from Obihiro, Hokkaido	0·561 (0·028)
Mti ‘119/2’	S	South of Ogimachi, Shirakawa-Gun, Honshu (500 m)	0·560 (0·027)
Msi ‘32/1’	S	North from Obihiro, Hokkaido	0·550 (0·015)
Mti ‘121/1’	S	South of Ogimachi, Shirakawa-Go, Honshu (500 m)	0·545 (0·011)
Mti ‘132/2’	S	Ainukura, Honshu (350 m)	0·543 (0·020)
Msi ‘66/1’	S	Hakusan National Park, Honshu (850 m)	0·543 (0·021)
Mti ‘115/1’	S	South of Ogimachi, Shirakawa-Go, Honshu (500 m)	0·541 (0·014)
Mti ‘121/2’	S	South of Ogimachi, Shirakawa-Go, Honshu (500 m)	0·541 (0·024)
Msi ‘32/2’	UN	North from Obihiro, Hokkaido	0·536 (0·025)
Mti ‘133/1’	S	Ainukura, Honshu (350 m)	0·536 (0·027)
Mti ‘118/1’	S	South of Ogimachi, Shirakawa-Go, Honshu (500 m)	0·520† (0·010)
Mti ‘133/2’	S	Ainukura, Honshu (350 m)	0·501* (0·015)
Mti ‘117/1’	S	South of Ogimachi, Shirakawa-Go, Honshu (500 m)	0·494** (0·026)
Mti ‘116/1’	S	South of Ogimachi, Shirakawa-Go, Honshu (500 m)	0·454** (0·012)
Mean			0·570

The species and accession number or name are given on the left. Species codes are: M. sacchariflorus (Msa), M. sinensis (Msi) and M. tinctorius (Mti). Accessions sharing the same first number (e.g. 56/1 and 56/2) indicate seed or rhizome collected from the same plant.

The second column shows whether seed (S) or rhizome (R) was collected or if the type of collected material is unknown (UN).

The location of the original collection, including altitude measured with an altimeter is given in the third column, and mapped (Supplementary Data Fig. S1).

The right-hand column shows the mean (n = 4) maximum dark-adapted quantum efficiency of photosystem II (F_v/F_m) with the standard error in parentheses. The measurements were taken in the field. Mean F_v/F_m for each accession is compared with the value for M. ×giganteus (Mxg). Statistical separation based on Dunnett’s test is indicated (†P ≤ 0·1; *P ≤ 0·05; ** P ≤ 0·01).

OUAV, Obihiro University of Agriculture and Veterinary Medicine.

Maximum efficiency of photosystem II (F_v/F_m)

The night frosts of early May caused necrotic damage to the leaves of many of the accessions, providing a fortuitous early screen of frost tolerance of the leaf material. Twenty-nine accessions were identified which had emerged and showed little or no visible damage of the leaves (Table 1). These were then surveyed for evidence of chilling-dependent reduction in F_v/F_m in the evenings following on from chilling throughout the days of 18 and 19 May. Values ranged from 0·454 to 0·655; of these, nine showed values that were apparently higher than that of Mxg.

Leaf gas exchange in a chilling controlled environment

Maintained and measured at 15 °C in a controlled environment, four of the selected accessions showed a higher A than Mxg. Both Msa ‘73/2’ and ‘73/3’ showed significant increases of approx. 58 and 43 %, respectively, relative to Mxg. All accessions of Msi and Mti showed a lower A, although only one line was significantly lower (Fig. 2A). This pattern was paralleled by stomatal conductance to water vapour, where Msa ‘73/3’ showed the highest g_s (Fig. 2B). The ratio of intercellular to ambient CO₂ concentration (c_i/c_a) showed little variation between accessions, except that the Mti accession showed a significantly higher c_i than Mxg (Fig. 2C). Transfer of plants with their soil ball to pots and the controlled environment would inevitably result in some root damage, which could affect water supply to the leaves. However, there was no evidence of water deficit effects in the gas exchange measurements. All leaves showed high conductances (g_s >70 mmol m^–2 s^–1) and high ratios of intercellular to external CO₂ concentrations (c_i/c_a >0·5) (Fig. 2B, C). Four Msa accessions found to have a higher A than Mxg at 15 °C (Fig. 2A) had higher rates of leaf CO₂ uptake (A) at all light (Q) and c_i levels (Fig. 3A, B). Inspection of the A/c_i response (Fig. 3) shows that the higher A was the result of a significantly higher V_max, even though V_pmax was also higher in these accessions (Table 2).

Fig. 2.

(A) The net leaf CO₂ uptake rate (A) at a photon flux (Q) of 1500 μmol m^–2 s^–1, (B) stomatal conductance of water vapour (g_s) and (C) ratio of intercellular to atmospheric CO₂ concentration (c_i/c_a) for 11 Miscanthus accessions selected from the 29 given in Table 1. Measurements were made at a leaf temperature of 15 °C and 390 μmol mol^–1 of CO₂ in the controlled-environment growth chamber (15 °C/15 °C 14 h day/10 h night cycle at Q = 850 μmol m^–2 s^–1). Differences between Mxg (black bars) and the other accessions were examined by Dunnett’s test (†P ≤ 0·1; *P ≤ 0·05; **P ≤ 0·01). Values are mean ± (1 s.e.); n = 4. Msa, M. sacchariflorus; Msi, M. sinensis; Mti, M. tinctorius; Mxg, M. ×giganteus.

Fig. 3.

(A, C, E) The response of net leaf CO₂ uptake rate (A) to incident photon flux (Q) at an ambient CO₂ concentration of 390 μmol mol^–1, and (B, D, F) A vs. intercellular CO₂ concentration (c_i) at Q = 1500 μmol m^–2 s^–1 for the four Miscanthus accessions that showed a higher A than Mxg in Fig. 2. Each curve represents a four-parameter non-rectangular hyperbola fit to the measurements of four replicates. (C–F) The fitted curves and individual data points to which the curves were fit for the two accessions with the highest rates (C and D for Msa ‘73/2’ and E and F for Mxg). Measurement conditions and species abbreviations are as in Fig. 2.

Table 2.

Parameters of the non-rectangular hyperbola fit to the responses of net leaf CO₂ uptake rate (A) to incident photon flux (Q) and to internal CO₂ concentration (c_i) (Fig. 3)

Parameter	Accession
	Msa ‘44/2’	Msa ‘44/1’	Msa ‘73/2’	Msa ‘73/3’	Mxg
Asat (μmol m–2 s–1)	16·00 (0·99)	17·49 (0·78)	24·76** (1·52)	21·32* (1·29)	14·58 (1·45)
ΦCO2,max	0·040* (0·003)	0·032 (0·002)	0·050** (0·003)	0·042* (0·003)	0·028 (0·001)
θ	0·581 (0·086)	0·356 (0·308)	0·350 (0·146)	0·452 (0·059)	0·455 (0·125)
Rd (μmol m–2 s–1)	1·03 (0·08)	1·14 (0·07)	1·54† (0·10)	1·41 (0·27)	1·00 (0·18)
Vmax (μmol m–2 s–1)	15·37 (0·91)	16·30 (1·85)	21·69** (1·99)	20·63** (2·36)	11·66 (0·82)
Vpmax (μmol m–2 s–1)	28·13 (2·60)	26·66 (1·39).	38·91** (3·04)	36·68** (1·92)	19·33 (2·60)

Measurements were taken in a controlled environment growth room (15 °C/15 °C 14 h day/10 h night cycle at 850 μmol photon flux m^–2 s^–1). Significant differences between Mxg and the Msa accessions were calculated by Dunnett’s test (†P ≤ 0·1; *P ≤ 0·05; **P ≤ 0·01). Numbers are the mean ± (s.e.) n = 4.

Species codes are: M. sacchariflorus (Msa), M. sinensis (Msi) and M. tinctorius (Mti).

A_sat, light-saturated rate of net CO₂ uptake; the asymptote of the A/Q response; Φ_CO2,max, maximum apparent quantum yield of CO₂ uptake (dimensionless); initial slope of the A/Q response; θ, convexity of the A/Q response; R_d, mitochondrial respiration rate in the light; intercept of the A/Q response; V_max, [CO₂]-saturated rate of A; asymptote of the A/c_i response; V_pmax, maximal rate of PEP carboxylation; initial slope of the A/c_i response.

DISCUSSION

This study tested the hypothesis that greater capacity for photosynthesis during early-season chilling exists in populations of the parent species of M. ×giganteus. Our three-stage screen narrowed the many lines of this broader collection of the parent species down to two clones that were clearly superior (Table 2). The successive screens showed that two accessions of M. sacchariflorus developed leaves in the spring that not only survive night-time frost, but also show superior light-limited and light-saturated photosynthesis under chilling conditions, by comparison with Mxg. During growth under chilling, Φ_CO2,max was equivalent to that of unstressed C₄ leaves as a measure of light-limited photosynthesis. Similarly, A_sat as a measure of light-saturated photosynthesis attained values that would be adjudged high for most plants under non-stress conditions. These last two measures were both substantially and significantly higher than in Mxg (Table 2). Accession Msa ‘73/2’ showed the highest Φ_CO2,max and A_sat of all when grown and measured at 15 °C. In comparison with Mxg, Φ_CO2,max was 79 % higher and A_sat 70 % higher (Table 2). Although the projected A_sat at the hypothetical asymptote of the fitted hyperbola was 25 µmol m^–2 s^–1 for the best of these clones (Table 2), the actual rate achieved at a photon flux of 1500 µmol m^–2 s^–1 was lower at 19 µmol m^–2 s^–1 (Fig. 3C). This results because the hyperbola projects to infinite Q, and C₄ photosynthesis is typically not saturated, even in full sunlight: Q = approx. 2000 µmol m^–2 s^–1. Nevertheless the rates are remarkable for any plant, C₃ or C₄, at this temperature (Long and Spence, 2013). For example, it equals the rate obtained at 15 °C by a well-fertilized modern winter wheat grown under a similar temperature regime (Nagai and Makino, 2009). It also equals the rate obtained at this temperature for the fast growing cool temperate C₃ weed, Chenopodium album, when measured at the current atmospheric [CO₂] (Sage, 2002). In accession Msa ‘73/3’, stomatal conductance (g_s) was also significantly greater than in Mxg; however, this does not explain the higher A, since c_i/c_a was almost identical to that of Mxg (Fig. 2B, C). A higher g_s would cause a higher c_i, which in turn would increase A_sat, if photosynthesis is not [CO₂] saturated (von Caemmerer, 2000). However, inspection of the A/c_i response shows that c_i was saturating, nor did it increase (Fig. 3D, Table 2). This rules out the possibility that the higher A results from higher g_s, rather the two changing in concert to maintain a constant c_i and water use efficiency at the leaf level. This suggests that the higher g_s is simply an adjustment to the higher photosynthetic capacity of this accession at 15 °C and not causal. This contrasts with chilling-sensitive C₄ species which can show a loss of stomatal control and decline in A/g_s as a measure of instantaneous water use efficiency. In some cases, this loss of stomatal control in C₄ leaves is sufficient to induce wilting, even in well-watered plants (Rodriguez and Davies, 1982). Three further Msa accessions also appeared to have a higher A and A_sat than Mxg at 15 °C (Figs 2A and 3). Further analysis of the responses of A_sat to c_i for these four accessions showed a V_pmax, the in vivo maximum rate of PEP carboxylation, that was almost double that of Mxg for the best two accessions (Table 2). By reference to Fig. 3D, this higher V_pmax would have little influence on A at the operating c_i (195 µmol mol^–1). The operating c_i is the value obtained in Fig. 2C at the growth ambient c_a of 390 µmol mol^–1. However, if factors that could cause a lower g_s occur in combination with chilling, e.g. drought, these lines might then have an additional advantage through their higher V_pmax (von Caemmerer, 2000). This is because decreased g_s would lower water loss, but would not decrease A, until g_s has declined to lower c_i below the transition point of the A/c_i response from PEP regeneration limitation to PEP carboxylation limitation.

Not only did Msa ‘73/2’ and ‘73/3’ show a high A_sat, but, perhaps more importantly a superior rate was maintained at all light levels, with a maximum quantum yield of CO₂ uptake (Φ_CO2,max) of 0·042–0·050 (Table 2). Maximum quantum yield determines the rate of CO₂ uptake by a leaf under light-limited conditions. Since canopy carbon uptake over a diurnal course is as dependent on light-limited photosynthesis as it is on light-saturated photosyntheis, this has important implications for potential crop productivity (Long, 1993). For leaves of NADP-ME species, in the absence of stress, the highest Φ_CO2,max is considered to be approx. 0·065 (Ehleringer and Pearcy, 1983), but this is on an absorbed light basis. If it is assumed that approx. 20 % of the incident light is transmitted or reflected, then the highest value would be reduced to 0·052, very close to that for the two Msa accessions in Table 2. Light-dependent reduction in Φ_CO2,max is a pervasive early symptom of chilling intolerance in C₃ and C₄ plants, and is commonly observed in crops of tropical and sub-tropical origin (Long and Spence, 2013). In the least tolerant plants it typically represents damage to PSII, and in more tolerant species a protective increase in the de-epoxidated xanthophyll, zeaxanthin, which is associated with heat dissipation of excess energy at PSII so avoiding damage to the reaction centre (Demmig-Adams and Adams, 1992; Long et al., 1994; Havaux and Niyogi, 1999; Farage et al., 2006). While accumulation of zeaxanthin protects PSII, it also lowers the efficiency of PSII and in turn light-limited CO₂ assimilation expressed as Φ_CO2,max. This is because of diversion of absorbed light energy into heat dissipation under conditions where light is limiting A, as observed in Mxg (Farage et al., 2006). The high Φ_CO2,max in these two Msa accessions under chilling conditions is close to the maximum achieved in C₄ NADP-ME grasses under optimal conditions (Ehleringer and Pearcy, 1983). This suggests that there is very little photodamage or photoprotection at 15 °C in these accessions (Table 2). The ability of these Msa accessions to achieve high light-saturated and light-limited rates at 15 °C will in itself provide protection by dissipating excitation energy from PSII (Long et al., 1994).

As explained in the Introduction, the bases of chilling impairment of light-limited photosynthetic capacity expressed here as Φ_CO2,max and light-saturated photosynthesis expressed here as A_sat are very different. The former results from either damage to PSII or increased thermal deactivation of excitation energy at PSII. The latter results from loss of activity of key enzymes of carbon metabolism. Dark-adapted F_v/F_m provides a direct measure of the maximum efficiency of PSII, and so should be functionally correlated with Φ_CO2,max, but not A (Bolhár-Nordenkampf et al., 1989). The independence of these processes probably explains why when ranked on A the order of genotypes is different from ranking on F_v/F_m (Table 1; Fig. 2). However, the screen indicated an approx. 5 % higher F_v/F_m for Msa ‘73/2’ and ‘73/3’ by comparison with Mxg, while Φ_CO2,max in these accessions was >50 % higher. This might in part be explained by the different measurement conditions, field vs. controlled environment, and much greater variability in the field measures. In particular, small variations in leaf angle can result in significant variation in light exposure (Nobel et al., 1993), which would in turn affect photoinhibition measured as a decline in F_v/F_m. Nevertheless, F_v/F_m is a rapid screen, easily used in the field and, while its correlation with Φ_CO2,max determined in the controlled environment is weak, it did identify the two accessions that proved to be superior to Mxg in terms of chilling tolerance. Although M. tinctorius ‘132/1’ was initially expected to have cold tolerance, based on observed early growth in the field and a similar F_v/F_m to Mxg following chilling in the field, it proved the least tolerant of the accessions with respect to carbon assimilation (Table 1; Fig. 2). Not only did it show the lowest A when maintained at 15 °C, but it was the only accession to show a significant increase in c_i/c_a. As noted earlier, this suggests that under chilling conditions, stomata failed to adjust to decreased photosynthetic capacity, at the expense of efficiency of water use at the leaf level (Rodriguez and Davies, 1982).

This work has identified relatives of Mxg with significantly superior capacity for photosynthesis at chilling temperatures, apparently for the first time. When grown and measured at 15 °C, these accessions of Msa achieved photosynthetic rates at low and high light equivalent to those of productive cold-tolerant C₃ plants. They showed very little evidence of the chilling-dependent photoinhibition that normally characterizes C₄ photosynthesis at these temperatures (Long et al., 1994; Long and Spence, 2013). Although no Msi accession was found that was superior to Mxg, accessions that were equivalent in tolerance were identified, notably Msi ‘11/1’. Since Msa provides approx. 67 % of the genome of Mxg, when produced from hybridization of tetraploid Msa with diploid Msi, this is arguably the more important parent.

Conclusions

The findings from this wider collection, have identified important material for breeding new synthetic M. ×giganteus with a greater capacity for photosynthesis under chilling conditions than in the widely grown ‘Hornum’ clone (Mxg). It has shown the value of searching more widely across the natural geographic range of Msa and Msi to obtain breeding material, particularly material observed to have high productivity in the field at cold locations. Most importantly it shows that evolution within this C₄ clade has overcome barriers, previously assumed to be inherent (Sage, 2002) in achieving efficient C₄ photosynthesis at chilling temperatures.

SUPPLEMENTARY DATA

Supplementary data are available online at www.aob.oxfordjournaljournal.org and consist of Figure S1: known collection sites of 29 Japanese accessions used in the current study.