Temperature responses are a window to the physiology of dark respiration: differences between CO2 release and O2 reduction shed light on energy conservation

Authors


M. A. Adams. Fax: +61 2 9385 2120; e-mail: Mark.Adams@unsw.edu.au

ABSTRACT

We showed that temperature responses of dark respiration for foliage of Pinus radiata could be approximated by Arrhenius kinetics, whereby E0 determines shape of the exponential response and denotes overall activation energy of respiratory metabolism. Reproducible and predictable deviation from strict Arrhenius kinetics depended on foliage age, and differed between RCO2 and RO2. Inhibition of oxygen reduction (RO2) by cyanide (inhibiting COX) or SHAM (inhibiting AOX) resulted in reproducible changes of the temperature sensitivity for RO2, but did not affect RCO2. Enthalpic growth – preservation of electrons in anabolic products – could be approximated with knowledge of four variables: activation energies (E0) for both RCO2 and RO2, and basal rates of respiration at a low reference temperature (RREF). Rates of enthalpic growth by P. radiata needles were large in spring due to differences between RREF of oxidative decarboxylation and that of oxygen reduction, while overall activation energies for the two processes were similar. Later during needle development, enthalpic growth was dependent on differences between E0 for RCO2 as compared with RO2, and increased E0(RO2) indicated greater contributions of cytochrome oxidase to accompany the switch from carbohydrate sink to source. Temperature-dependent increments in stored energy can be calculated as the difference between RCO2HCO2 and RO2HO2.

INTRODUCTION

Respiration provides energy (ATP), reducing power (NADH) and carbon skeleton intermediates for the biosynthesis and maintenance of cell structures (Fernie, Carrari & Sweetlove 2004). On a global scale, autotrophic dark respiration is associated with CO2 fluxes into the atmosphere that exceed those originating from fossil fuel combustion by one order of magnitude (Atkin et al. 2005). Nevertheless, respiratory carbon release is counterbalanced by photosynthetic carbon gain, and most terrestrial ecosystems constitute a sink for atmospheric CO2 of variable strength (Schulze, Beck & Müller-Hohenstein 2002, pp. 506–510). At present, it is uncertain how the balance between photosynthesis and respiration will be affected by predicted changes in climate. Most carbon balance models that address this question predominantly rely on our well-established understanding of photosynthetic processes. In some of these models, dark respiration is treated rather simplistically by assuming that respiratory CO2 release is a constant fraction of photosynthetic carbon gain, and thereby, directly linked to photosynthesis (Dewar, Medlyn & McMurtrie 1998; Waring, Landsberg & Williams 1998; Gifford 2003). This notion has some experimental support, insofar as the substrate supply for respiration ultimately depends on photosynthetic performance (Turnbull et al. 2003; Whitehead, Griffin & Turnbull 2004; Wright et al. 2005; Atkin, Scheurwater & Pons 2006, 2007; Hartley et al. 2006). However, a recent meta-analysis of the literature suggests that, although the slope of the relationship between net primary productivity and gross primary productivity (defined as carbon use efficiency) was 0.53, values varied from 0.23 to 0.83 for different forest types (DeLucia et al. 2007). Furthermore, there is also experimental evidence that respiration does not directly depend on previously synthesized carbohydrates (Nogues et al. 2006), and that rates of dark respiration can be buffered, at least for some time. This is also reflected in recent findings that acclimation of photosynthesis and respiration is asynchronous in response to changes in temperature (Campbell et al. 2007).

Given the importance of respiratory processes for the carbon balance between atmosphere and biosphere, it seems crucial to better understand the mechanistic control of respiration (more or less independently from that of photosynthesis), in order to elucidate how that balance might be altered in the future. Given that respiratory processes are embedded within many other cellular biochemical mechanisms, elucidation of mechanistic control is a non-trivial task, much more complex than for photosynthesis.

It is perhaps telling that considerable advances in developing current theories as to how photosynthetic processes are regulated were based on the interpretation of gas exchange (mostly CO2 exchange at leaf level) in response to variable light or CO2 concentrations (measuring light response and A/Ci curves, respectively; Farquhar, Caemmerer & Berry 1980; Lambers, Chapin & Pons 1998, pp. 10–61). Is it possible, by analogy, to gain insight into the physiology of dark respiration by measuring its temperature sensitivity?

Respiration is one of the most temperature dependent of all plant processes (Atkin et al. 2005). In general, respiration increases exponentially over the range of temperatures most commonly experienced by plants, albeit with a sharp reduction at higher temperatures (i.e. >30 °C; Criddle et al. 1994; Anekonda et al. 2000). As a result, a temperature coefficient, Q10, that denotes the proportional change of respiration per 10 °C increase in temperature, has long been used to describe the temperature response of respiration. In practise, the approaches used in calculation of Q10 result in a coefficient that is itself temperature dependent, and not clearly related to underlying physiological mechanisms (also compare our companion study Kruse & Adams 2008a).

An increasing number of studies rely on a different approach, based on Arrhenius kinetics (Turnbull et al. 2001; Griffin, Turnbull & Murthy 2002; Shapiro et al. 2004; Davidson & Janssens 2006). The Arrhenius equation originates in physical chemistry (Arrhenius 1889) and defines, in biological systems, the activation energy of a particular reaction (Thornley & Johnson 1990),

image(1)

where k is the rate constant and Ea is the activation energy (kJ mol−1) of this particular reaction, T is the absolute temperature (K), R is the gas constant (8.314 J mol−1 K−1) and A is a constant. The employment of Arrhenius functions to describe the temperature dependency of respiratory metabolism was pioneered by Criddle et al. (1994) and Taylor et al. (1998). They defined a temperature coefficient µ (K), that defines the slope of the exponential function as µ = Ea/R.

More recently, Shapiro et al. (2004) modified the Arrhenius equation and used the parameter Eo (instead of Ea) as a temperature coefficient, related to the overall activation energy for the metabolic process of respiration (RCO2 in their study). The term ‘overall activation energy’, measured in kJ mol−1, and describing the temperature response of the entire respiratory metabolism, is perhaps more widely used than other equivalent measures (Davidson & Janssens 2006). It conveys the idea that the temperature response of gas exchange has a physiological underpinning and depends on metabolic pathways and, thus, the activity of enzymes involved in respiratory processes.

For example, mitochondrial oxygen reduction (RO2) is catalyzed by two enzymes [cytochrome oxidase (COX) and alternative oxidase (AOX)] that differ in catalytic properties (Buchanan, Gruissem & Jones 2000). Shifts in their relative activity are theoretically reflected in the temperature response of RO2. By contrast, the interpretation of the temperature sensitivity of CO2 production (RCO2) seems less straightforward, given the multitude of decarboxylation reactions during dark respiration. However, carbon flux through the main catabolic pathways [glycolysis and tricarboxylic acid (TCA) cycle] is generally under the control of ‘pacemaker’ enzymes that are themselves highly regulated and catalyse irreversible reactions. Consequently, the activation energy for RCO2 may well be dominated by the activation energy of one or more limiting reactions within the pathway (Mathews, van Holde & Ahern 2000, pp. 360–375). Notwithstanding such theoretical considerations, it is possible to calculate the overall activation energy of respiration, bearing in mind that this coefficient reflects the activation energy of more than one reaction contributing to either RCO2 or RO2.

Clearly, the accuracy of Arrhenius models in describing temperature responses of RCO2 and RO2 remains subject to some conjecture. Arrhenius models may aptly describe respiratory temperature responses if substrate supply remains non-limiting (i.e. enzymes are operating close to Vmax). At higher temperatures, substrates are seldom likely to reach concentrations sufficient to saturate the enzyme(s), and the temperature sensitivity of Km becomes more important, which could result in significant departures from Arrhenius kinetics (Davidson, Janssens & Luo 2006).

In the present study, we sought first to elucidate if the overall activation energy (Eo) is a useful parameter to describe the temperature response of respiration over a broad range of measurement temperatures. Secondly, we sought to establish a link between the temperature sensitivity of respiration and the physiological mechanisms involved. For this purpose, we tested the effects of specific inhibitors of COX (cyanide) and AOX [salicylhydroxamic acid (SHAM)] on the overall activation energy for oxygen reduction (RO2), and if respiratory CO2 release was affected by these inhibitors. Thirdly, we characterized changes in the temperature sensitivity of respiration associated with developmental changes in Pinus radiata needles. These changes are accompanied by significant changes in metabolism and foliage chemistry. Importantly, when we recognize the energy content or oxidation state of anabolic products as a component of ‘growth’ in its broader context (Hansen et al. 2002), it becomes evident that relations between rates of oxidative decarboxylation (that are coupled to the reduction of NAD) and rates of oxygen reduction (that depend on NADH oxidation) are important determinants of rates of growth (when defined as an increase of enthalpy). We employed an enthalpy balance model to characterize ontogenetic influences on energy conservation and its temperature dependency. We believe such a characterization assists the interpretation of the temperature sensitivities of RCO2 and RO2 in physiological terms.

MATERIALS AND METHODS

Respiration measurements and determination of respiration parameters

Respiration was determined using two, multi-cell, differential-scanning calorimeters (CSC 4100, MC-DSC, Calorimetry Sciences Corporation, Provo, UT, USA). Routine measurements with calorimeters usually include two steps: first, the heat rate of the sample (q) is measured and then secondly, it is re-measured in the presence of NaOH (qNaOH). This second heat rate is generally greater than the first, owing to the exothermic formation of carbonate from CO2 released from the plant material (Criddle, Breidenbach & Hansen 1991; Hansen et al. 1994; Criddle & Hansen 1999). The calorimetric method is sensitive and precise, owing to the exact control (±0.01 °C) of measurement temperature and the low background noise associated with the instruments (typically less than 1–2% of the signal strength of a respiring plant material). Rates of CO2 production (RCO2) are calculated from the difference between the second and first heat rate, taking into account the enthalpy change for carbonate formation (−108.5 kJ mol−1) (Criddle & Hansen 1999). More than 90% of the metabolic heat (q) from a cell is generated from the reduction of oxygen in the mitochondria (Hopkin 1991). The heat rate (q) is thus directly proportional to RO2. Indeed q = −RO2ΔHO2. The term ΔHO2 varies only slightly with the class of compound used as substrate for respiration, and is usually substituted by Thornton's ‘constant’ or the oxycaloric equivalent (−455 ± 15 J mmol−1, Hansen et al. 2004). In the current study, we assumed that carbohydrates were the sole substrate of respiration so that ΔHO2 and ΔHCO2 (the enthalpy change for combustion of one mole of substrate carbon to carbon dioxide, or −469 J mmol−1) are of the same magnitude. Hence, RCO2ΔHCO2 denotes the energy made available by carbohydrate oxidation and RO2ΔHO2 the amount of energy lost during oxygen reduction. The enthalpic growth rate can thus be calculated according to Eqn 2,

image(2)

where RSG is the specific rate of conversion of substrate carbon to biomass carbon (nmol g−1 s−1) and ΔHB is the total enthalpy change for incorporation of one mole of substrate carbon (carbohydrates) into structural carbon (‘biomass’, Hansen, Hopkin & Criddle 1997), including the enthalpy effects from all elements (kJ nmol−1). Therefore, RSGΔHB is a measure of the instantaneous rate of growth. It is the energy stored in the chemical bonds of anabolic products, or, the conservation of electrons originating from oxidative decarboxylation.

The enthalpy conversion efficiency (ηH), that is the ratio of energy output to energy input, was calculated as

image(3)

Combining Eqns 2 and 3 gives

image(4)

It follows from Eqn 3 that ηH can, theoretically, assume values between zero and unity. If only carbohydrates are respired, the enthalpic efficiency is proportional to the respiratory quotient CO2/O2, and a respiratory quotient of 1 is equivalent to ηH = 0.

Temperature response of respiration

In the present study, we calculated the overall activation energy (Eo) of oxygen reduction (RO2) and CO2 release (RCO2) form the linearized Arrhenius equation:

image(5)

The original Arrhenius equation covers temperature regimes that are irrelevant to biological systems. It can be modified as shown in Eqn 6 to suit temperature regimes relevant to plant metabolism and can be tailored to any reference temperature,

image(6)

where Rsp is the respiration rate (either RCO2 or RO2), RREF is the respiration rate at a low reference temperature (K) and Eo is the overall activation energy (kJ mol−1) of either RCO2[henceforth, denoted Eo(RCO2)] or RO2[henceforth, denoted Eo(RO2)].

All experiments reported here were conducted using P. radiata growing close to the School of Forest and Ecosystem Science at Creswick, about 15 km north of Ballarat, Victoria, Australia. Experiments 1 and 2 employed 1-year-old needles collected from an individual tree, while experiment 3 was based on sampling in 10-year-old pine plantations (planted 1994, see Experiment 3 description for more details).

Experiment 1a: methods

The restricted volume (1 cm3) of measurement ampoules of calorimeters used here generally requires that plant material be cut before calorimetric measurements. This may affect the availability of substrates within tissues or cause a wound response (Criddle & Hansen 1999). Furthermore, measurements of respiratory responses to changing temperature typically last for 4–6 h, which could lead to significant substrate depletion. Such effects could confound observed temperature responses.

A simple test to assess substrate availability and a putative wound response, as suggested by Criddle & Hansen (1999), was employed here: cut needles from eight different, independent branches were left for 30 min prior to measurement. Different quantities of cut needles – ranging from 5 to 35 pieces – were measured at 15 °C. The remaining cut needles were left for 2, 4 and 6 h at room temperature (15–20 °C) in the dark, before measurement.

Experiment 1b: methods

In Experiment 1b, we tested the extent to which RCO2 and RO2 can be described by strictly linear Arrhenius kinetics, and if measurement temperature caused significant departures from Arrhenius kinetics. In this experiment, we used fully expanded current year and 2-year-old source foliage from one branch. Needles were cut into 1 cm long pieces and left for 30 min at room temperature. The needle cuttings were mixed thoroughly and aliquots of ∼150 mg needle fresh weight were placed into each of the six available ampoules. These were arranged pairwise, and we added capsules containing NaOH to one ampoule of every pair. Ampoules were placed in the calorimeters and left for 15 min at 10 °C. Subsequently, the temperature response of respiration was measured in 2.5 °C steps between 12.5 and 40.0 °C. After the 25.0 and 32.5 °C measurements, ampoules were opened, flushed with air and returned to the calorimeters. RCO2 was determined by subtracting the heat rate of the ampoule without NaOH from the heat rate of the ampoule containing NaOH (after accounting for the exact fresh weight of needles in every ampoule).

Experiment 2: methods

In a second experiment, we explored the effects of application of specific inhibitors of COX and AOX on temperature response of RO2, and whether these inhibitors also affect Eo(RCO2).

Inhibitor studies were again performed with 1-year-old needles. Needles were cut into ca. 0.3 cm long pieces, mixed and left for 1 h (air control) or were incubated in a solution containing either 10 mm KCl (control), 10 mm KCN or 10 mm SHAM. Initial experiments showed that 1 h incubation in solution had little effect on respiration as compared with needles left in air. CN- concentrations >5 mm significantly inhibited respiration. We chose concentrations of 10 mm for both KCN and SHAM studies based on this and a review of the literature. The samples were shaken for 1 h in solution and, subsequently, dried for 10 min. Aliquots of 150–200 mg were placed in ampoules and left for 15 min at 7 °C in the calorimeter cells. Subsequently, the temperature response of respiration (both RO2 and RCO2) was measured at four different temperatures (10, 15, 20 and 25 °C). At each temperature, heat rates (q) were recorded after a stabilization period of 40 min. Subsequently, capsules containing NaOH were added to the ampoules, and qNaOH was measured after another 40 min. RCO2 and RO2 were determined as described above. Inhibitor experiments were performed with six independent replicates per treatment. Data were subjected to one-way analysis of variance (anova), and differences between treatments were analysed by Tukey's Honest Significant Difference (HSD) post hoc test and deemed significant if P < 0.05.

Experiment 3: methods

This principal set of experiments was designed to characterize the changes in respiration rates at a low reference temperature, and especially Eo(RO2) and Eo(RCO2) during needle ontogeny. The results were also interpreted in terms of enthalpic growth rates and the enthalpy conversion efficiency.

Experiments were conducted in 10-year-old pine plantations (planted 1994), 15 km north of Ballarat, Victoria, Australia. The climate of the region is Mediterranean, with cool wet winters and warm dry summers. Research plots were established in October 2003 (0.05 ha in size) and received different amounts of nitrogen fertilizer [nil, 100, 200 and 400 kg N ha−1, applied as (NH4)2SO4]. The addition of N-fertilizer was designed to alter the growth rates and thereby, the ontogeny of needles (see also Kruse & Adams 2008b). In early spring of the following year, six branches (from different trees) per plot were detached from the lower canopy level (i.e. 7 m above ground level). The branches were re-cut under water and taken to the laboratory in close proximity to the research plots. Branches were also collected in mid-December 2004 and at the end of February 2005 for respiration measurements. Respiration measurements were conducted at 15, 20, 25, 30 and 35 °C as described in Methods for experiment 2.

In December and February, heat rates (q) were constant (after a 40 min instrument stabilization) at every measurement temperature, but heat rates (qNaOH) declined at 35 °C. In October 2004, at a time of strong sink activity and rapid growth, both heat rates (q and qNaOH) declined predictably during the stabilization period at every measurement temperature – and the extent depended on temperature. In this case, we extrapolated q and qNaOH to the mid-point between the two measurements to obtain the best possible estimate of RCO2.

For the determination of Eo(RCO2) in February and December, we omitted the measurement at 35 °C that showed the clear departure from linear Arrhenius kinetics. This finding will be discussed later. All other activation energies (for RO2 and RCO2) were calculated from the linear fits over the whole temperature range (15–35 °C).

Determination of nitrogen

After weighing, dried samples were ground to a fine powder with a ball mill. The nitrogen concentration was determined by Dumas combustion. Aliquots of 50 mg of the finely ground foliage samples were combusted to N2 and CO2 and quantified by means of thermoconductivity (LECO CHN2000, St Joseph, MI, USA).

Statistical analysis

Statistical analysis was performed with the results of six independent replicates per season and treatment (i.e. three seasons and four treatments). The effects of season, treatment and their interaction were assessed by anova (Statistica, version 6.0, Tulsa, OK, USA) for the parameters needle-N and needle length, and for the activation energies of RCO2 and RO2. The respiration parameters RCO2ΔHCO2, RO2ΔHO2, RSGΔHB and ηH were assessed by three-factorial anova (including season, treatment and temperature, i.e. 15 °C versus 25 °C, as factors). Statistical significance of main effects and their interaction is indicated by *, if P < 0.05, **, if P < 0.01 and ***, if P < 0.001. Differences between individual means were analysed by Tukey's HSD post hoc test and deemed significant if P < 0.05.

RESULTS

Experiment 1a

A linear correlation between the number of cut needles and their heat rate, measured at different times after cutting, explained 95% of the observed variation (Fig. 1). The remaining variance could be ascribed to differences among the eight independent replicates. Hence, neither wound responses (30 min after cutting the needles) nor substrate depletion after 6 h at 15–20 °C significantly affected the heat rate of needles measured at 15 °C.

Figure 1.

Correlation between the number of cut needle pieces and heat rate. Needles from eight independent branches were cut into 1 cm long pieces and left for 0.5, 2.0, 4.0 and 6.0 h at 15–20 °C in the dark. After this period of time, heat rates of different numbers of pieces were measured at 15 °C. The correlation between the number of cut pieces and their heat rate was linear and explained 95% of the observed variation.

Experiment 1b

Figure 2a depicts the linearization of the exponential function given by Eqn 5. In the current year needles, RCO2 deviated slightly from the Arrhenius kinetics within the broad temperature interval under consideration (12.5–40.0 °C). Still, the Arrhenius model provided a good approximation of the observed temperature response, in particular, over narrower temperature intervals. Likewise, the temperature dependency of oxygen reduction (RO2) of the current year needles was almost perfectly described by Arrhenius kinetics over the entire temperature range (Fig. 2b). Note that the slight increase in RO2 at 27.5 and 35 °C was due to the opening and flushing of the ampoules with air as indicated by the arrows in Fig. 2. In older needles, respiration (both RCO2 and RO2) significantly deviated from strictly linear Arrhenius kinetics, in particular, for RCO2 (Fig. 2c,d). Subsequent experiments were performed with the current year needles, for which respiration could be well approximated by linear Arrhenius kinetics. This response was reproducible in three consecutive temperature runs for a temperature range of 10–25 °C (results not shown).

Figure 2.

Linearization of respiration rates measured across a wide temperature (T) range (12.5–40 °C). Linearization of the exponential temperature response of respiration according to the Arrhenius model is used to calculate the overall activation energy (Eo) of respiration (the slope of the linear function equals Eo/R, compare Eqn 5) and helps assess deviation from Arrhenius kinetics. We assessed such deviation for CO2 release (RCO2) of the current year (a) and 2-year-old needles (b), and for the oxygen reduction (RO2) of the current year (c) and 2-year-old needles (d). Oxygen reduction of the current year needles could be described almost perfectly according to Arrhenius kinetics (b), but deviation was observed for the older tissue (d). By contrast, RCO2 could only be approximated by the Arrhenius model, and the deviation was stronger for 2-year-old (c) than for younger needles (a).

Experiment 2

Inhibition of COX with cyanide significantly reduced the activation energy of RO2 and respiration rates (by ∼50%) at a low reference temperature (measured at 10 °C, Table 1). On the other hand, inhibition of AOX with SHAM decreased RO2 by ca. 13% at 10 °C and slightly increased Eo(RO2). Note that the activation energy for RO2 was not constant after inhibitor application – as compared to the control – because of the partial readjustment of O2 reduction during these rather lengthy experiments (compare Fig. 3a). Eo(RO2) was thus not constant, but changed with increasing measurement temperature, with time, or both (as is evident from the poorer R2 values – when compared to the control). While inhibition of COX with CN- reduced RO2 by ca. 50% at 10 °C, RCO2 was reduced by only 30% (compare Table 1). Crucially, the addition of KCN did not affect the activation energy of RCO2 when compared with the control needles, despite its large effect on Eo(RO2). As a consequence, the CO2/O2 quotient that declined exponentially with increasing temperatures in the control (and SHAM treated) plants, increased slightly for needles treated with KCN (compare Fig. 3b).

Table 1.  Effects of KCN and salicylhydroxamic acid (SHAM) on basal respiration rates (rates at a low measurement temperature like 10 °C, R10) and the activation energies (Eo) of oxygen reduction(RO2) and CO2 release (RCO2)
 AirKCl (control)KCNSHAM
  1. Data were subjected to one-way analysis of variance, and differences between treatments were analysed by Tukey's Honst Significant Difference post hoc test. Data shown are means ± SD of six independent replicates, and different indices indicate significant differences between means (P < 0.05), as affected by treatment.

R10 (O2) (nmol g−1 s−1)1.37 ± 0.13c1.34 ± 0.06c0.71 ± 0.09a1.17 ± 0.08b
Eo (RO2) (kJ mol−1)67.1 ± 4.2c,d61.5 ± 3.2b,c26.5 ± 9.8a76.3 ± 5.3d
R10 (CO2) (nmol g−1 s−1)3.2 ± 0.8b3.2 ± 0.4b2.3 ± 0.3a3.2 ± 0.5b
Eo(RCO2) (kJ mol−1)46.5 ± 6.5a38.2 ± 6.9a38.5 ± 6.3a39.1 ± 6.4a
Figure 3.

Effects of the inhibitors KCN and salicylhydroxamic acid (SHAM) on the activation energy (Eo) of oxygen reduction (RO2) (a) and the temperature (T) response of the respiratory quotient CO2/O2 (b). Cyanide is an inhibitor of cytochrome oxidase (COX), and SHAM inhibits alternative oxidase (AOX). The effects on the activation energy (Eo) can be deduced from the slope of the respective line, showing that KCN application significantly decreased the overall activation energy of RO2[Eo(RO2)], as a result of COX inhibition (also compare Table 1). Note that, in contrast to the control treatment, Eo was not constant after inhibitor application. The application of KCN and SHAM had contrasting effects on the temperature response of the respiratory quotient (b), owing to the different effects on Eo(RO2) versus Eo(RCO2) (also compare Table 1).

Experiment 3

Nitrogen concentration and length of needles

Concentrations of N in needles were significantly dependent on season and fertilizer treatment (Table 2). Concentrations averaged across all treatments were 1.57% in October, 1.20% in December, 1.25% in February and 1.18% in August. Fertilizer application increased needle-N from 1.13% in control trees to 1.29% in trees treated with 100 kg N ha−1, to 1.38% when 200 kg N ha−1 had been applied and to 1.44% when 400 kg N ha−1 had been applied. The fertilizer effect was greatest for the N400 treatment, particularly in October 2004, but had almost disappeared by August 2005 (compare Table 2).

Table 2.  Nitrogen concentration and length of needles of Pinus radiata
SamplingTreatmentN-concentration (%)Needle length (cm)
  • Research plots were treated with (NH4)2SO4 at a rate of nil, 100, 200 and 400 kg ha−1 in October 2003. Data shown are the means ± SD of six replicates, each. Different indices indicate significant differences between means (P < 0.05) within each season (i.e. the treatment effect within the respective sampling season).

  • *

    Calculated as % of the respective needle length in August.

October 2004Control1.34 ± 0.15a<0.6
N1001.57 ± 0.14a,b<0.6
N2001.59 ± 0.13a,b<0.6
N4001.77 ± 0.17b<0.6
December 2004Control1.08 ± 0.07a5.0 ± 1.4a; 47%*
N1001.16 ± 0.23a,b5.1 ± 1.5a; 48%
N2001.22 ± 0.14a,b5.1 ± 1.5a; 47%
N4001.36 ± 0.15b6.4 ± 0.7b; 62%
February 2005Control1.05 ± 0.12a9.7 ± 1.6a; 92%
N1001.22 ± 0.12a,b9.7 ± 2.1a; 91%
N2001.34 ± 0.08b9.8 ± 2.1a; 90%
N4001.39 ± 0.14b10.2 ± 1.7a; 98%
August 2005Control1.06 ± 0.09a10.5 ± 1.3a
N1001.11 ± 0.10a10.7 ± 1.5a
N2001.21 ± 0.15a10.9 ± 1.9a
N4001.23 ± 0.11a10.4 ± 0.3a

Needle length increased from 0.6 cm in October 2004 to 10.6 cm in August 2005, and the overall treatment effect was not significant (Table 2). However, in December 2004, needles from the N400 treatment were significantly longer compared with those from other treatments, and had reached 62% of their final length (as compared with 47–48% of the other treatments, Table 2).

The temperature response of RCO2ΔHCO2, RO2ΔHO2 and enthalpic growth (RSGΔHB)

In this study, the temperature response of RCO2ΔHCO2, RO2ΔHO2 and RSGΔHB are described by functions that provide the best fit to the data (Fig. 4).

Figure 4.

CO2 production, O2 consumption, and enthalpic growth rates as a function of temperature of Pinus radiata needles. RCO2ΔHCO2 was calculated assuming that carbohydrates are the sole substrate for respiration (i.e. that ΔHCO2 is −469 J mol−1) and RSGΔHB was calculated as the difference between RCO2ΔHCO2 and RO2ΔHO2. Research plots were treated with fertilizer (NH4)2SO4 at a rate of nil, 100, 200 and 400 kg ha−1 in October 2003. The first measurements were conducted in the spring of the following year. Data shown are the means ± SD of six replicates, each. RCO2ΔHCO2, energy made available by carbohydrate oxidation; RSGΔHB, measure of the instantaneous rate of growth; RO2ΔHO2, amount of energy lost during oxygen reduction.

In October, a logistic function provided the best fit for the temperature response of RCO2ΔHCO2 and RO2ΔHO2. Reductions in heat rates during measurements, in particular at higher temperatures, presumably contributed to the shape of the curves in October (see Materials and Methods). Still, temperature responses of enthalpic growth rates (RSGΔHB) were linear (Fig. 4g), as both q and qNaOH decreased during the measurements. The slope of the response (i.e. dRSGΔHB/dT) depended significantly on fertilizer application and needle-N (not shown, but see Fig. 5).

Figure 5.

Correlation between needle-N and enthalpic growth rates of needles (at 25 °C) in October (circles), December (squares) and February (triangles). Rates of enthalpic growth were well correlated with needle-N during cell division in October, but not during other stages of ontogenetic development. RSGΔHB, measure of the instantaneous rate of growth.

In December and February, temperature responses of RO2ΔHO2 could be described by exponential functions, while temperature responses of RCO2ΔHCO2 (and consequently also RSGΔHB) were better approximated by third-order polynomials.

In order to test the effects of fertilizer treatment and season on RCO2ΔHCO2 and RO2ΔHO2 (and in particular, enthalpic growth rates and the enthalpy conversion efficiency), and their interactions with the measurement temperature, we chose respiration rates at 15 and 25 °C. Respiration measurements at 15 and 25 °C were closest to the mean minimum and maximum temperature in December (9.9 and 22.3 °C, respectively) and February (9.7 and 22.5 °C, Australian Bureau of Meteorology), and respiration rates at 25 °C were close to the optimum temperature for growth in December and February, prior to the significant decline in RCO2 at higher temperatures (compare Fig. 4h,i).

Respiration and enthalpic growth and efficiency at 15 and 25 °C

Unsurprisingly, measurement temperature exerted the greatest influence on rates of respiration (Tables 3 & 4). RCO2ΔHCO2 averaged 2.47 µW mg−1 at 15 °C and 4.78 µW mg−1 at 25 °C. RO2ΔHO2 averaged 1.91 µW mg−1 at 15 °C and 3.81 µW mg−1 at 25 °C. The effect of sampling date (season) on RCO2ΔHCO2 and RO2ΔHO2 was also highly significant, albeit smaller than the temperature effect (Table 4). By contrast, sampling date had a larger impact on enthalpic growth rates (RSGΔHB), averaging 1.46 µW mg−1 in October, 0.36 µW mg−1 in December and 0.42 µW mg−1 in February, than the measurement temperature (0.55 µW mg−1 at 15 °C and 0.95 µW mg−1 at 25 °C). Note that, although RCO2ΔHCO2 decreased from 4.19 µW mg−1 in December to 2.47 µW mg−1 in February, enthalpic growth rates remained almost constant, as RO2ΔHO2 changed in a similar manner (from 3.83 µW mg−1 in December to 2.05 µW mg−1 in February). Enthalpic growth was four to five times greater in spring than later in the growth season –RCO2ΔHCO2 was significantly greater than RO2ΔHO2 (compare Table 3).

Table 3.  Respiration variables at 15 and 25 °C
SamplingTreatmentRCO2ΔHCO2 (µW mg−1, at 15 °C)RO2ΔHO2 (µW mg−1, at 15 °C)RSGΔHB (µW mg−1, at 15 °C)ηH (efficiency, at 15 °C)
October 2004Control2.95 ± 0.231.83 ± 0.171.12 ± 0.240.38 ± 0.06
N1003.18 ± 0.522.05 ± 0.161.13 ± 0.520.34 ± 0.12
N2003.17 ± 0.211.94 ± 0.351.23 ± 0.380.39 ± 0.11
N4003.57 ± 0.452.29 ± 0.191.28 ± 0.270.36 ± 0.03
December 2004Control2.64 ± 0.472.65 ± 0.570.05 ± 0.370.02 ± 0.15
N1002.34 ± 0.332.09 ± 0.320.25 ± 0.330.10 ± 0.14
N2002.92 ± 0.342.99 ± 0.420.09 ± 0.350.03 ± 0.12
N4002.69 ± 0.332.46 ± 0.330.23 ± 0.190.08 ± 0.07
February 2005Control1.48 ± 0.091.08 ± 0.170.40 ± 0.150.27 ± 0.11
N1001.74 ± 0.341.42 ± 0.180.31 ± 0.230.17 ± 0.10
N2001.49 ± 0.181.20 ± 0.140.29 ± 0.210.18 ± 0.12
N4001.37 ± 0.180.96 ± 0.150.41 ± 0.050.30 ± 0.03
  RCO2ΔHCO2 (µW mg−1, at 25 °C)RO2ΔHO2 (µW mg−1, at 25 °C)RSGΔHB (µW mg−1, at 25 °C)ηH (efficiency, at 25 °C)
  1. RCO2ΔHCO2 was calculated assuming that carbohydrates are the sole substrate for respiration (i.e. that ΔHCO2 is −469 kJ mol−1) and RSGΔHB was calculated as the difference between RCO2ΔHCO2 and RO2ΔHO2. The enthalpy conversion efficiency (ηH) was calculated as the ratio RSGΔHB/RCO2ΔHCO2 and is proportional to the respiratory ratio CO2:O2. Data shown are the means ± SD of six replicates, each. For statistical results, compare Table 4.

  2. RSGΔHB, measure of the instantaneous rate of growth; RCO2ΔHCO2, energy made available by carbohydrate oxidation; RO2ΔHO2, amount of energy lost during oxygen reduction.

October 2004Control4.69 ± 0.283.09 ± 0.201.60 ± 0.320.35 ± 0.05
N1004.86 ± 0.613.36 ± 0.251.50 ± 0.510.31 ± 0.07
N2004.92 ± 0.483.23 ± 0.421.69 ± 0.150.36 ± 0.03
N4006.00 ± 0.813.85 ± 0.342.15 ± 0.480.37 ± 0.03
December 2004Control5.54 ± 0.915.16 ± 0.970.37 ± 0.260.07 ± 0.05
N1004.96 ± 0.734.28 ± 0.610.67 ± 0.220.14 ± 0.03
N2006.58 ± 1.085.95 ± 0.850.63 ± 0.480.09 ± 0.06
N4005.80 ± 0.715.06 ± 0.540.74 ± 0.250.13 ± 0.03
February 2005Control3.43 ± 0.342.77 ± 0.310.65 ± 0.240.19 ± 0.07
N1003.53 ± 0.223.37 ± 0.400.16 ± 0.420.04 ± 0.11
N2003.47 ± 0.312.95 ± 0.300.51 ± 0.200.11 ± 0.06
N4003.22 ± 0.302.54 ± 0.270.67 ± 0.170.21 ± 0.04
Table 4.  Results of analysis of variance (anova), showing the significance of the main effectors season (sampling date), treatment and their interaction for needle-N and length, and for the activation energies of RCO2 and RO2
 NitrogenLength
FPFP
Sampling (S)49.2***514***
Treatment (T)29.5***0.6n.s.
S × T1.2n.s.3.1*
 RCO2ΔHCO2RO2ΔHO2RSGΔHBηH
FPFPFPFP
S185***225***183***128***
T5.0*3.6*4.5*3.2*
Temperature (Tp)745***733***56.4***1.2n.s.
S × T7.2***12.6***3.1*4.1**
S × Tp21.8***26.1***5.7*9.8.***
T × Tp1.6n.s.0.4n.s.1.8n.s.0.6n.s.
 Eo(RCO2)Eo(RO2)
FPFP
  • The variables RCO2ΔHCO2, RO2ΔHO2, RSGΔHB and ηH were assessed by three-factorial anova (sampling, treatment and temperature, i.e. 15 °C versus 25 °C).

  • *

    P < 0.05;

  • **

    P < 0.05;

  • ***

    P < 0.05.

  • RCO2ΔHCO2, energy made available by carbohydrate oxidation; RO2ΔHO2, amount of energy lost during oxygen reduction; RSGΔHB, measure of the instantaneous rate of growth; ηH, enthalpy conversion efficiency; RCO2, CO2 release; RO2, oxygen reduction; Eo(RCO2), overall activation energy of RCO2; Eo(RO2), overall activation energy of RO2; n.s., not significant.

S91***425***
T2.1n.s.5.5***
S × T0.8n.s.3.8**

Rather surprising was that overall, temperature had no effect on enthalpic efficiency, but there was a significant temperature × sampling date interaction (Table 4). In October, enthalpic efficiency varied little with temperature (0.36 at 15 °C and 0.34 at 25 °C across all treatments). In December, ηH increased from 0.04 at 15 °C to 0.12 at 25 °C, whereas in February, ηH decreased from 0.23 at 15 °C to 0.14 at 25 °C (see also Table 3).

Nitrogen treatment had a small, but statistically significant impact on all respiration variables (Table 4), in particular, the N400 treatment (Table 3). This impact was most pronounced in October, during rapid cell division and growth (significant interaction, Table 4). In October, rates of enthalpic growth were correlated positively with the needle-N concentration. This correlation was absent during later stages of needle ontogeny (compare Fig. 5).

Activation energies for RO2 and RCO2

The concept of activation energy simplifies the assessment of temperature effects on respiration, enthalpic growth and efficiency. Only two variables [Eo(RCO2) and Eo(RO2)] are necessary to comprehensively describe temperature effects for the range over which the temperature response can be approximated by Arrhenius kinetics.

In December and February, Eo(RO2) could be determined with high precision, and R2 values of the linear fits were between 0.992 and 0.996 (Table 5). Precision was poorer in October as the temperature response was not strictly exponential owing to a decline in heat rates during measurements (see Materials and Methods). In December and February, we omitted the last temperature step (35 °C) for the determination of Eo(RCO2). Even so, the R2 of the linear fit for the calculation of Eo(RCO2) remained less than for Eo(RO2), particularly in December (Table 5). These results confirm that a strict Arrhenius approach to describing temperature sensitivity works better for RO2 than for RCO2 (compare experiment 1b).

Table 5.  Activation energies of decarboxylation (RCO2) and oxidation reactions (RO2) during respiration
SamplingTreatmentEo(RCO2) (kJ mol−1)R2Eo(RO2) (kJ mol−1)R2
  1. In every case, Eo(RO2) was determined from the linearization of the Arrhenius equation with n = 5. Eo(RCO2) was determined with n = 5 in October, but with n = 4 in December and February (i.e. omitting the 35 °C measurement). Data shown are the means ± SD of six replicates, each. Average R2 values of the linear fits are given. Different indices indicate significant differences between means (P < 0.05) within each season (i.e. the treatment effect within the respective season).

  2. RCO2, CO2 release; RO2, oxygen reduction; Eo(RCO2), overall activation energy of RCO2; Eo(RO2), overall activation energy of RO2.

October 2004Control33.2 ± 4.1a0.95137.6 ± 3.4a0.937
N10030.7 ± 6.8a0.92435.5 ± 2.8a0.957
N20031.2 ± 7.8a0.95836.9 ± 3.8a0.959
N40036.9 ± 3.3a0.92237.1 ± 2.3a0.944
December 2004Control54.2 ± 5.5a0.97248.1 ± 2.9a0.992
N10053.5 ± 8.8a0.95951.1 ± 2.0b0.992
N20057.5 ± 7.2a0.95649.0 ± 2.6a,b0.993
N40054.7 ± 3.1a0.98551.6 ± 2.2b0.993
February 2005Control59.7 ± 3.4a0.98966.5 ± 5.6a,b0.996
N10051.6 ± 9.2a0.98462.7 ± 3.3a0.995
N20058.4 ± 5.3a0.98664.5 ± 3.9a0.995
N40060.9 ± 6.6a0.98572.9 ± 4.4b0.995

In October, activation energies for RO2 and RCO2 were similar, averaging 36.7 and 35.6 kJ mol−1, respectively. Eo(RO2) increased to 50.0 kJ mol−1 in December, and increased further to 66.7 kJ mol−1 in February. Eo(RCO2) changed little, averaging 60.5 kJ mol−1 in December and 59.1 kJ mol−1 in February. Consequently, activation energies for RO2 and RCO2 were similar in October, and Eo(RCO2) was greater than Eo(RO2) in December, but less in February (Fig. 6b). On the other hand RCO2 and RO2 at a low reference temperature were similar in December and February, as indicated by the strong correlation (Fig. 6a). In October, however, RCO2 at 15 °C significantly exceeded those of RO2 (compare Fig. 6a).

Figure 6.

Respiration (RCO2 versus RO2) at a low measurement temperature (15 °C, b), and activation energies for RCO2 plotted against the activation energies for RO2 (a). Circles, October; squares, December; triangles, February; RCO2, CO2 release; RO2, oxygen reduction; Eo(RO2), overall activation energy of RO2; Eo(RCO2), overall activation energy of RCO2.

Nitrogen treatment had a slight, though still significant, impact on the activation energy of RO2 (Table 3). Averaged over all seasons, Eo(RO2) was 50.1 kJ mol−1 in control trees, 49.4 kJ mol−1 in the N100 treatment, 50.3 kJ mol−1 in the N200 treatment and 53.9 kJ mol−1 in the N400 treatment. The greater activation energy of the N400 treatment was particularly pronounced in February (significant sampling × treatment interaction, Table 4).

DISCUSSION

Arrhenius kinetics as a means to describe the temperature response of respiration

In the present study, we demonstrated that Arrhenius functions describe well the temperature responses of respiration in some cases, but only serve as an approximation in other cases. Best linear fits, describing the exponential relationship between respiration rates and the measurement temperature, were obtained for mitochondrial oxygen reduction by young needles. By contrast, the same young needles exhibited slight deviation from strictly linear Arrhenius kinetics for respiratory CO2 release. Such departures from strict Arrhenius kinetics became more prevalent in older foliage and, here again, RCO2 was affected more strongly than RO2.

We also showed that respiratory temperature responses have a clear physiological underpinning. Inhibition of mitochondrial pathways to oxygen reduction in young source foliage, using specific inhibitors of COX (cyanide) and AOX (SHAM), resulted in significantly lower (in CN- treated) or higher (in SHAM-treated plants) activation energies for overall oxygen reduction, as compared with the control treatments. Complete inactivation of specific metabolic pathways is hard to achieve with chemical inhibitors. Nonetheless, the results of the inhibitor study show (1) that partial inhibition of one pathway causes the direction of Eo(RO2) to change in a predictable manner and (2) that the temperature sensitivity of respiratory CO2 release is not affected by CN- or SHAM treatments. These results illustrate that the temperature sensitivity of respiration is influenced by distinct metabolic pathways, catalysed by specific enzymes (with defined activation energies). Temperature responses of respiration, when precisely measured using calorimetry, can thus be used to elucidate the underlying physiology of gas exchange processes in the dark.

It is also clear that Arrhenius kinetics cannot alone comprehensively describe the temperature response of respiration. It is a common observation that respiration – most often measured as CO2 release – increases exponentially with higher temperatures within a low temperature range. However, respiration deviates from exponential functions at higher temperatures and reaches a peak at some ‘optimum temperature’, before declining at still greater temperatures (Thornley & Johnson 1990; Atkin & Tjoelker 2003). It has been argued that this slowed increase at suboptimal temperatures primarily reflects the different temperature sensitivities of Vmax and Km of contributing enzymes (Davidson & Janssens 2006; Davidson et al. 2006). Provided substrate concentrations are significantly greater than Km of the ‘pacemaker’ enzyme(s), then the temperature response of respiration depends mainly on the temperature sensitivity of Vmax (Davidson et al. 2006). At higher temperatures, substrate concentrations are much less likely to be sufficient to saturate the enzyme(s), and the temperature sensitivity of Km becomes more important. This shift can explain, at least in part, why respiration rates do not continue to increase exponentially, even within the suboptimal temperature range.

A notable exception in the present study was the behavior of CN--treated needles, for which we observed increasing rates of oxygen reduction as measurement temperature increased. The rather long duration of the experiment may have allowed some relief from inhibition. However, this dynamic, positive re-adjustment is apparently more strongly dependent on measurement temperature, and can also be related to differences in the catalytic properties of COX versus AOX (compare with our companion study, Kruse & Adams 2008a).

Physiological interpretation of the respiratory temperature response

As cyanide inhibition of COX resulted in reduced overall activation energy for oxygen reduction [Eo(RO2)], it can be inferred that a large activation energy for RO2 indicates a strong involvement of COX in oxygen reduction (also compare Armstrong et al. 2006). Involvement of COX activity increased, more or less, continuously with needle age, while there was a decrease in total electron transfer at low temperatures (see also Kruse & Adams 2008a). Such a continuous increase of Eo(RO2) in ageing foliage was also observed for eucalypts (Anekonda et al. 2000).

The sink to source transition of foliage has profound effects on cell metabolism. During this transition, there is a reduced requirement for carbon skeleton intermediates and reducing power (NADH), and increased requirement for ATP for maintenance processes (Amthor 2000). Furthermore, the demand for ATP to propel the formation and export of sucrose, increases in ageing foliage. It seems likely that, in relative terms, a larger fraction of total electron transfer was catalyzed by COX in older foliage (for further discussion, compare Kruse & Adams 2008a). At greater rates of N-fertilizer (N400), P. radiata exhibited greater activation energies for RO2 in February, indicating an earlier switch from sink to source. For example, we recently found that needles from the upper canopy of pine plantations exhibited greater N-concentrations and showed advanced development as compared with needles from the lower canopy (Raison, Myers & Benson 1992; Kruse & Adams 2008b). These developmental changes were also reflected by alterations of the overall activation energy for oxygen reduction (compare with Kruse & Adams 2008a). Clearly, temperature responses of respiration yield valuable insight into cell metabolism. The temperature sensitivity of oxygen reduction changes quite predictably according to the developmental stage of foliage, as affected by nitrogen status and environmental conditions.

The results of the inhibitor study also made it clear that Eo(RO2) and Eo(RCO2) are not coupled, as might be expected, given the substantial differences in biochemical pathways. Likewise, Taylor et al. (1998) observed that the temperature responses of RCO2 and RO2 are flexible and mostly independent. By contrast, basal RCO2 appears to depend, at least partially, on basal RO2, as KCN decreased the R10 of both RCO2 and RO2. It seems logical that the decarboxylation of organic acids within the mitochondrial matrix, electron transfer in the inner mitochondrial membrane and subsequent oxygen reduction are dependent on each other to some degree (compare Fig. 6a). However, other decarboxylating pathways operating in the dark, in particular, the oxidative pentose phosphate pathway, are fully active during the early stage of development when needles are not completely photoautotrophic and require the Calvin cycle intermediates or ribose-5-phoshate for the synthesis of RNA and DNA (ap Rees 1980). This pathway is a major source of reducing power for anabolic reactions (Kruger & von Schaewen 2003) and may explain why during cell division in October (when requirements for highly reduced compounds like membrane lipids and proteins were large), basal RCO2 was significantly greater than basal RO2. At present, this interpretation in terms of the developmental control of decarboxylating pathways remains speculative and needs to be tested in future studies.

An important point is that we require the knowledge of both CO2 production (RCO2) and oxygen consumption (RO2) to comprehensively characterize energy conservation (i.e. ‘growth’ defined as accumulation of energy) during ontogenic development (Hansen et al. 1997).

Adding another level of complexity: temperature dependency of enthalpic growth

The respiratory quotient (CO2/O2) is closely related to the oxidation state of substrate carbon respired during ‘steady state’ or ‘maintenance’ respiration (Davey et al. 2004; Hansen et al. 2004). By contrast, for many actively growing plants (including cell division), the respiratory quotient is often correlated with growth rate when expressed on a mass basis (Lambers et al. 1998). In the current study, we defined ‘growth’ as the rate of energy conservation during respiration. This enthalpic growth rate depends on the balance between rates of oxidative decarboxylation and those of oxygen reduction (i.e. RSGΔHB = −RCO2ΔHCO2 + RO2ΔHO2, i.e. it is the conservation of electrons or, more figuratively, the energy stored in chemical bonds of anabolic products (Macfarlane, Adams & Hansen 2002; Kruse & Adams 2008b).

Linking respiration to growth requires that temperature effects be taken into account. The temperature dependency of growth is complex and is not fully understood (Loveys et al. 2002). Fortunately, temperature-dependent enthalpic growth can be estimated as the difference between two sets of functions – describing the temperature response for RCO2ΔHCO2 and for RO2ΔHO2 (Criddle et al. 1994; Taylor et al. 1998).

Bearing in mind that the Arrhenius model is a first approximation of the temperature response of respiration, it is possible to calculate the growth increment from this difference, integrated over a distinct temperature regime and time interval (Eqn 7),

image(7)

where ISGΔHB is the increase in enthalpy, i.e. the growth increment in the dark. It is evident from Eqn 7 that the calculation of growth increment requires the knowledge of four parameters: (1) basal rates of respiration at a low reference temperature (for both O2 consumption and CO2 production) and (2) activation energies of both CO2 release and O2 uptake. The basal respiration rate (i.e. the pre-exponential factor RREF in Eqns 4 & 5) gives the respiration rate at a low reference temperature.

As at every temperature, the basal rate of respiration is limited by the activity of enzymes (Vmax) involved in either RCO2 or RO2 (Atkin & Tjoelker 2003). However, basal respiration is also coupled, to some extent, to the activation energy of the process (either RCO2 or RO2). High activation energy of respiration means that the respiration rate is restricted at low temperatures. This phenomenon is discussed more thoroughly in our accompanying paper (Kruse & Adams 2008a).

In the present study, activation energies for RCO2 and RO2 were similar in October (Fig. 6b), but rates of respiration, measured (at 15 °C) as the production of CO2 (RCO2), were significantly greater than those measured as O2 reduction (RO2). We conclude that the activity of enzymes involved in oxidative decarboxylation was significantly greater than that of enzymes catalysing oxygen reduction (compare Fig. 6a). Note that in October, the enthalpy conversion efficiency (ηH) did not change with the temperature (Table 3). The temperature dependency of ηH, or that of the respiratory quotient CO2/O2 can be predicted with knowledge of the four parameters introduced earlier,

image(8)

It follows from Eqn 8 that CO2/O2 can be defined in terms of the ‘initial’ respiratory quotient at low temperatures (RREFCO2/RREFO2), if Eo(RCO2) and Eo(RO2) are equal – as observed in October. If however, Eo(RCO2) > Eo(RO2), the respiratory quotient will increase with increasing temperature. This situation was observed in December, or after the application of KCN to pine needles in the inhibitor experiment. In February (or in the control treatment of the inhibitor study), enthalpic efficiency decreased with temperature, as Eo(RCO2) < Eo(RO2).

Equation 7 is a means to describe temperature-dependent increments of energy in phenomenological terms. In order to be of use in predictive growth models, future research needs to address the following issues:

  • 1How to reliably describe, and explain, deviation from strictly linear Arrhenius kinetics. That is, combining Arrhenius with the Michaelis–Menten kinetics could help to fine tune the model described by Eqn 7. Such an approach has been outlined in our companion study for the process of oxygen reduction (Kruse & Adams 2008a).
  • 2We need to establish the links between the activities of decarboxylating pathways and the temperature sensitivity of RCO2. For a start, we may characterize changes in Eo(RCO2) as affected by the main catabolic pathways. For example, the temperature sensitivity of RCO2 is likely to be different in the light as compared with the dark because of photorespiration (Atkin et al. 2005).
  • 3Valid application of Eqn 7 requires that the temperature range and time interval used for integration be carefully considered. In the current study, we demonstrated that respiratory variables change with plant ontogeny, and it will be a major challenge to better identify the time frame (i.e. the outer integral in Eqn 7) over which the four variables characterizing respiration remain valid. Secondly, we need to assess if short-term temperature changes (i.e. the inner integral in Eqn 7) causes longer term alterations of these four variables (i.e. compare Bruhn et al. 2007). Put differently, we may have to re-consider what is meant by ‘respiratory acclimation’. Is it a term applicable to irreversible changes, at least within a well-defined time frame? And does the term ‘acclimation’ encompass responses to the environment only or to developmental changes as well? Some of these questions have been addressed in our companion study (Kruse & Adams 2008a), but they leave much scope for future research.

ACKNOWLEDGMENTS

We wish to thank Hancock Victorian Plantations Pty Limited and the Australian Research Council for financial support. We thank Ross Bickford for assistance in the field.

Ancillary