Author for correspondence: Owen K. Atkin Tel: +44 1904 32 8560 Fax: +44 1904 32 8505 Email: OKA1@york.ac.uk
• The impact of nitrogen (N) supply on the temperature response of root respiratory O2 uptake (R) was assessed in several herbaceous species grown in solution culture. Warm-grown (25 : 20°C, day:night) plants differing in root N concentration were shifted to 13 : 8°C for 7 d to cold-acclimate.
• Log–log plots of root R vs root N concentration both showed that R increased with increasing tissue N concentration, irrespective of the growth temperature. Although the regression slopes of the log–log plots did not differ between the warm-grown and cold-acclimated plants, cold-acclimated plants did exhibit a higher y-axis intercept than their warm-grown counterparts. This suggests that cold acclimation of root R is not entirely dependent on cold-induced increases in tissue N concentration and that scaling relationships (i.e. regression equations fitted to the log–log plots) between root R and N concentration are not fixed.
• No systematic differences were found in the short-term Q10 (proportional change in R per 10°C change in temperature), or degree of cold acclimation (as measured by the proportional difference between warm- and cold-acclimated roots) among roots differing in root N concentration. The temperature response of root R is therefore insensitive to tissue N concentration.
• The insensitivity of Q10 values and acclimation to tissue N concentration raises the possibility that root R and its temperature sensitivity can be predicted for a range of N supply scenarios.
In contrast to the abundance of studies investigating the relationship between N and metabolism in leaves, relatively few studies have focused on coupling between root N concentration and root metabolism. Rates of R are positively correlated with the concentration of N in fine roots (Pregizter et al., 1998; Reich et al., 1998; Burton et al., 2002; Tjoelker et al., 2005; Bahn et al., 2006), and as roots age, root R declines in parallel with a sharp decline in root N concentration (Volder et al., 2005). The availability of N in soil can also affect root respiration, with rates of root R being lower in plants growing on N-deficient soils compared with plants well supplied with N (Van der Werf et al., 1992). The linkage between root R and N probably reflects the respiratory costs associated with the uptake and assimilation of N, protein turnover and maintenance of solute gradients (Scheurwater et al., 1998). Bahn et al. (2006) showed that the relationship between root R and N concentration for temperate mountain grasslands varied between sites, suggesting that root R–N relationships are not fixed. Whether differences in the relationship between root R and N concentration vary systematically with environmental factors such as growth temperature is not known.
In addition to playing an important role in determining rates of photosynthesis, N supply can also affect the extent of cold acclimation of the photosynthetic apparatus (Martindale & Leegood, 1997), a phenomenon that requires additional N investment in chloroplastic and cytosolic proteins (Stitt & Hurry, 2002). Cold acclimation can be defined as adjustments in metabolic processes in such a way that plant performance is improved in the cold (Körner & Larcher, 1988); full acclimation can result in complete metabolic homeostasis, i.e. identical rates of metabolism in plants growing at contrasting temperatures (Stitt & Hurry, 2002; Atkin & Tjoelker, 2003). Although the impact of N supply on respiratory acclimation is not known (either for leaves or for roots), there is strong a priori evidence that N availability could influence respiratory acclimation. In leaves, growth in the cold increases both R and N concentrations (e.g. Ryan, 1995). Similarly, species adapted to the cold often exhibit higher tissue N concentrations and R (Körner & Larcher, 1988). Tjoelker et al. (1999) found greater cold acclimation of leaf R in conifers than in broad-leaved trees, with foliar N concentrations increasing with decreasing growth temperature in the conifers but not in the broad-leaved trees. Cold acclimation of leaf R is dependent on increases in the capacity for mitochondrial respiration (Armstrong et al., 2006), which in turn would probably require an increase in N investment in respiratory proteins (and thus organic N); limitations in N supply could therefore restrict the extent to which R acclimates to low growth temperatures. Increases in R per unit organic N could occur if cold treatment increases respiratory flux via increases in substrate supply and/or reductions in adenylate restriction of electron transport (Atkin et al., 2000). Alternatively, higher rates of R in cold-acclimated roots might reflect an increase in demand for respiratory energy associated with a higher organic N concentration.
N supply might also impact on the short-term temperature dependence of R [i.e. the proportional change in R per 10°C change in temperature (Q10)]. Atkin & Tjoelker (2003) highlighted how variations in respiratory capacity, adenylates and substrate supply could influence Q10 values. If respiratory capacity is altered by changes in N supply and/or changes in N supply alter the fixation of CO2 and the production of sugars, then one might expect that the Q10 of R would vary with N availability and N concentration in the tissues. Alternatively, the Q10 of root R may vary in response to the turnover of adenylates associated with N uptake, transport and assimilation. Support for the suggestion that nutrient availability and the Q10 of R are linked comes from recent work by Turnbull et al. (2005), who found that the Q10 of leaf R in a temperate rainforest varied with nutrient availability (determined via analyses of extractable soil nitrate, ammonium and phosphorus concentrations; Richardson et al., 2004) along a soil chronosequence in New Zealand; leaf R Q10 values were highest at sites with the greatest soil nutrient concentration. The importance of N supply per se (as opposed to variations in other nutrients) in determining the Q10 of R in leaves or roots was not, however, established. Understanding how N supply impacts on the temperature response of root R is necessary for predicting future rates of R in environments where both N availability and temperature regimes are altered; in addition to an increase in mean annual temperatures, increases in variation about the mean temperature are expected, with potential for an increase in seasonal differences and an increase in the frequency of extreme events (IPCC, 2001).
In this study, we investigated the role external N supply plays in determining the response of root R of several herbaceous plant species after 7 d at a lower growth temperature. The specific questions addressed were as follows.
• To what extent do variations in N supply and N tissue concentration affect rates of root R of several herbaceous plant species?
• Does the short-term temperature sensitivity of root respiration (Q10) vary with N concentration?
• Is the degree of acclimation of roots to a decrease in temperature greater at high N supply than at low N supply?
• Is cold acclimation of root R associated with an increase in root N concentration (with higher R being required to meet the increased energy demands associated with higher protein concentrations, etc.) or are higher rates of R in cold-acclimated plants achieved via increased rates of R per unit organic N?
Materials and Methods
Acclimation of respiration in roots was assessed in warm-grown plants (25 : 20°C day:night regime) which had been grown at a range of different concentrations of N (experiment 1) or subjected to N depletion (experiment 2) before shifting to a lower growth temperature (13 : 8°C day:night regime) for several days. A 13 : 8°C regime has previously been shown to induce cold acclimation of root R in herbaceous species grown on high N supply (Covey-Crump et al., 2002).
The plant species used in the two experiments included a range of fast- and slow-growing herbaceous species: Achillea millefolium L., Achillea ptarmica L., Plantago lanceolata L., Plantago euryphylla BG Briggs, Silene dioica L. and Silene uniflora Roth. Details of the origin, natural distribution and maximum relative growth rates of these species can be found in Loveys et al. (2002). Seeds were sown on trays of Levington F2s compost (Scott's Professional, Ipswich, UK) and placed in growth cabinets at 22 ± 2°C (day) and 14 ± 2°C (night) with a 16-h day. Once the roots had reached at least 3 cm in length, seedlings were carefully removed from the compost and thoroughly washed with water. They were then transferred to 16-l hydroponics tanks filled with fully aerated modified Hoagland's nutrient solution (containing 2000 µm N; Poorter & Remkes, 1990). Nitrogen was supplied in the form of Ca(NO3)2 and KNO3. In each case where the N concentration was reduced (see later), the ions were balanced using CaSO4 and KCl at appropriate concentrations. The solution was changed weekly and maintained at pH 5.8. The tanks were placed in Conviron E15 growth cabinets (Conviron, Winnipeg, Canada) with a 16-h day and 300 µmol m−2 s−1 photosynthetic photon flux density provided by a combination of 400-W metal halide and 400-W high-pressure sodium bulbs. The plants were maintained at a 25 : 20°C day:night regime for 14 d before initial N treatments (experiments 1 and 2) commenced.
In the first part of experiment 1, 25 : 20°C-grown plants [herein called ‘warm-grown’ (WG) plants] of one species (A. ptarmica) were exposed to a wide range of different concentrations of N (2000, 400, 100, 50, 25 and 3 µm N) for extended periods [7 d (2000 and 400 µm N), 14 d (100 and 50 µm N) and 21 d (25 and 3 µm N)]; these plants were then shifted to 13 : 8°C (day:night) for 7 d [herein called ‘cold-acclimated’ (CA) plants] to assess whether the degree of cold acclimation of root R varies with root N concentration. The duration of exposure of warm-grown plants to each new N concentration was varied so that plants grown at different N concentrations were approximately the same size at the time of the temperature shift. The solutions were changed weekly and maintained at pH 5.8.
Root respiration [nmol O2 (g dry mass)−1 s−1] of detached whole root systems (four replicates per treatment) was measured at two temperatures (25 and 13°C) following the 7 d of temperature treatment (for both 25 : 20°C-grown and 13 : 8°C-acclimated plants). Separate individual whole root systems were used for measurements at the two temperatures (to avoid problems associated with extended measurements on single roots). Root R was measured using Clark type O2 electrodes (Rank Brothers, Cambridge, UK) coupled to a data acquisition system (NI-DAQ for Windows, 2000; National Instruments, Newbury, UK). The roots were put into cuvettes containing 50 ml of fully aerated modified Hoagland's nutrient solution (pH 5.8) buffered with 20 mm Morpholine ethane sulphonic acid. In all cases, the N concentration of the solution was identical to that in which the plants had been grown. O2 depletion was recorded for 5 min following stabilization: to avoid O2 limitations to respiration, we ensured that all measurements were terminated before the oxygen concentration in the cuvette fell below 40% of O2 saturation. Following measurement of respiration, roots were frozen in liquid N2 and freeze-dried in an Edwards EF4 Modulyo freeze-drier (Northern Scientific, York, UK) and the dry mass was recorded.
In the second part of experiment 1, we assessed whether the responses exhibited by A. ptarmica were also exhibited by four other herbaceous species (P. lanceolata, P. euryphylla, S. dioica and S. uniflora). Because of space and workload limitations, we restricted our analysis to plants that had experienced N concentrations of 2000 and 25 µm N; we chose 25 µm N as the earlier experiments with A. ptarmica showed that the resulting root N concentrations were reduced by ≈ 90%. 25 : 20°C-grown plants were again shifted to 13 : 8°C as already described for A. ptarmica, with root R and dry mass being measured as described for the first part of the experiment.
To assess whether the relationships between root R and N concentration obtained in roots experiencing differences in N supply were maintained in roots that experienced progressive depletion, we undertook a second experiment using two species: A. millefolium and A. ptarmica. Sixty-four plants of A. millefolium and 32 plants of A. ptarmica were initially grown at 2000 µm N as already described for experiment 1. Thirty-two plants of A. millefolium and 16 of A. ptarmica were then transferred to solutions containing 0 µm N. Four plants of both N treatments (2000 and 0 µm N) were then shifted to 13 : 8°C each week (with some plants being kept at 25 : 20°C as controls) to assess the impact of increasing duration of N depletion on temperature responses of root R. Root R was measured after 7 d as in experiment 1, with four replicates per treatment at each sampling date. Measurements were made at weekly intervals (for 4 wk in A. millefolium, but only for 2 wk for A. ptarmica as the plants began to flower after this point).
The freeze-dried roots were ground to a fine powder using a 31-700 Hammer Mill (Glen Creston Ltd, Stanmore, UK). Individual roots were often too small to allow analysis of N concentrations in each replicate used for respiration measurements; as a result, the four replicates for each treatment were pooled before grinding. Total N concentration in the pooled root samples was analysed by mass spectroscopy (CE Instruments NA2100 Brew Analyser; ThermoQuest Italia Sp A, Milan, Italy) using dried barley leaves as a reference material (N concentration 1.94%; Leco Instruments, Stockport, UK). The nitrate concentration of the pooled samples was analysed as described by Cataldo et al. (1975) with absorbance measured at 410 nm in a spectrophotometer (ELx 800 universal microplate reader; Bio-Tek Instruments, Inc., Winooski, VT, USA) and used to estimate the organic N concentration (i.e. total N minus NO3−). Total and organic N were also determined for a second sample of four pooled root systems for each treatment. N values are therefore the mean of the two pooled samples, except for a few samples where one of the two samples did not undergo a complete burn in the mass spectrometer.
Calculations and statistical analysis
The short-term temperature sensitivity of R (Q10) of both 25 : 20°C-grown (i.e. warm-grown) and 13 : 8°C-treated (i.e. cold-acclimated) plants was calculated over the 13–25°C measurement temperature range, using:
Q10 = (R25/R13)[10/(25−13)]
(R25 and R13, the mean respiration rates of roots measured at 25 and 13°C, respectively.)
Acclimation of respiration was assessed using the ‘homeostasis’ and ‘set temperature’ methods, as described in Loveys et al. (2003). Briefly, the homeostasis method requires a comparison of R rates at the respective growth temperatures of two contrasting growth temperature treatments; in our case, we compared rates of R of warm-grown and cold-acclimated plants at 25 and 13°C, respectively (i.e. at the respective daytime temperatures of each treatment). For the set temperature method, we divided rates of R exhibited by cold-acclimated plants by rates exhibited by warm-grown plants; the rates compared were measured at 25°C in each case.
Statistical analyses were carried out using SPSS v10, Sigmaplot v8.02 (SPSS Science, Birmingham, UK), and Microsoft Excel 2000 (Microsoft Inc.). In order to normalize the data, log10 transformations were carried out and linear regressions then fitted. The F-ratio method for homogeneity of linear regression slopes was used to compare the WG and CA treatments. The slopes were then tested for parallelism and the intercepts compared using a t-test.
Is root respiration affected by N supply and N tissue concentration?
In the first part of experiment 1, we assessed what impact differences in supply of N in hydroponic solutions had on the root N concentrations and root R rates of the slow-growing A. ptarmica. Both total and organic N concentrations of roots of A. ptarmica increased with increasing nitrate concentration supplied in hydroponic solution (Fig. 1a). Comparison of R rates and organic N concentrations in roots of A. ptarmica in experiment 1 showed that root R increased with increasing concentration of N supply (Fig. 1b) and with increasing root N concentration (see data for experiment 1 in Fig. 2).
Past studies investigating the linkage between plant R and N concentration have used log–log plots (e.g. Tjoelker et al., 2005). To determine whether relationships between root R and N are affected by growth temperature, we constructed log–log plots for A. ptarmica root R vs organic N concentration (Fig. 2); we chose to plot rates of root R against organic N concentrations, as differences in the proportion of total N present as organic N (Table 1) suggested that in some species there was ‘luxury accumulation’ at the highest rates of nitrate supply (log–log plots of root R vs total N concentration are provided in Fig. S1 in the supplementary material available online). There was a similar scaling in the relationship between root R and organic N concentration in plants of A. ptarmica either grown at different nitrate supply or depleted of N (see data for experiment 2 in Fig. 2), with no significant differences between the two experiments for either the WG or CA treatments (P > 0.05 in each case). This indicates that the same relationship between tissue root R and N concentration was obtained whether variations in N were achieved through differences in steady-state supply or via N depletion. Data from the two experiments are therefore combined in the following analyses.
Table 1. Total and organic root nitrogen (N) concentrations, the ratio of organic to total N and root respiration rates measured at 25°C for warm-grown (WG) and cold-acclimated (CA) plants grown on solutions containing 2000 or 25 µm N for Achillea ptarmica and four additional species grown in experiment 1
Solution N concentration (µm)
Total N concentration (mmol N g−1 DM)
Organic root N concentration (mmol N g−1 DM)
Ratio of organic:total N
Root respiration (nmol O2 g−1 DM s−1)
N concentration values were obtained from pooled samples of four replicate roots.
Root respiration rate values are means (± standard error, n = 4). Failure of our data acquisition system led to no respiration rate data being available for S. dioica at 25 µm N.
DM, dry matter.
10.8 ± 0.5
16.7 ± 1.6
31.7 ± 2.4
42.6 ± 9.2
23.5 ± 0.9
36.5 ± 1.8
32.6 ± 7.8
43.6 ± 5.1
12.6 ± 1.8
15.9 ± 2.0
46.6 ± 8.4
55.5 ± 7.5
33.0 ± 3.3
46.0 ± 5.0
23.8 ± 4.2
32.1 ± 8.1
36.7 ± 5.5
56.6 ± 10.9
For A. ptarmica, rates of root R at a given root N concentration were higher in the CA plants than in their WG counterparts (i.e. a single equation did not describe the log–log relationship between root R and N concentration in both CA and WG plants; Fig. 2; P < 0.05). To assess whether such differences were maintained in comparisons of multiple species, we exposed P. euryphylla, P. lanceolata, S. dioica and S. uniflora to low (25 µm) and high (2000 µm) N supply in hydroponics solution in the second part of experiment 1; Table 1 shows the effect of N supply on rates of root R and N concentrations in those four species, as well as data for A. ptarmica exposed to 25 and 2000 µm N. Low N supply substantially reduced the concentration of total and organic N in all species (Table 1). We then plotted the root R vs N concentration using all available data from experiments 1 and 2; root R was plotted against organic N on a log–log basis (Fig. 3). The relationship was significantly different between WG and CA plants (P < 0.05). The slope of the regression lines shown in Fig. 3 did not differ between the WG and CA plants (Table 2; P > 0.05); importantly, however, the intercept was higher in CA plants than in their WG counterparts [i.e. for any given N concentration, the CA plants exhibited a higher rate of root R than the WG plants; P < 0.01 for total N concentration (Table 2) and P < 0.05 for organic N concentration (Table 2, Fig. 3)]. Taken together, these results and those in Fig. 2 demonstrate that, while rates of root R were positively correlated with root N concentration (below a maximum total N concentration threshold), differences in scaling between root R and N were exhibited by WG and CA plants, with rates of R for a given N concentration being consistently higher in cold-acclimated roots than in their warm-grown counterparts.
Table 2. Regression analysis of log10-transformed dataa for warm-grown (WG) or cold-acclimated (CA) plants
y-axis intercept (log10)
P (y-axis intercept)
x = log10 (N concentration in mmol N g−1 DM), y = log10 (root respiration in nmol O2 g−1 DM s−1), where DM is dry matter.
P < 0.01
P > 0.05
P < 0.05
P > 0.05
Effects of N concentration on Q10 values and acclimation ratios
Q10 values did not vary with the organic N concentration of the roots (Fig. 4), nor was there any significant difference between the short-term Q10 of root R between WG [mean ± standard error (SE) = 1.61 ± 0.08] and CA (mean ± SE = 1.63 ± 0.07) plants. Acclimation ratios showed no systematic variation with organic N concentration, as measured using the ‘homeostasis’ (Fig. 5) and ‘set temperature’ (data not shown) methods (mean values ± SE = 0.77 ± 0.04 and 1.31 ± 0.07, respectively), regardless of whether the acclimation ratios were plotted against the N concentration of the WG plants (as shown in Fig. 5) or that of the CA plants (data not shown).
Do variations in N supply and N tissue concentration affect rates of root respiration?
We modulated root tissue N concentration through changes in nitrate supply and by N depletion; both factors are potentially important in determining variations in the N concentration of roots in natural environments. We found that, for a given growth temperature, scaling between root R and N concentration was maintained irrespective of how N concentration was manipulated. The rate of root R in whole roots increased with increasing total and organic N concentration, as has been reported for leaf R (e.g. Reich et al., 1998; Tjoelker et al., 1999, 2005; Wright et al., 2001), tree fine root R (Burton et al., 2002; Tjoelker et al., 2005; Volder et al., 2005) and fine root R from temperate mountain grasslands (Bahn et al., 2006).
In their recent study comparing 39 species grown under common garden conditions, Tjoelker et al. (2005) concluded that the fine root R–N relationship was fixed across species (and tissue type); as a result, they suggested that N concentrations could be used to model root (and leaf) R rates. Given that root respiration is a major component of net ecosystem CO2 exchange, with up to two-thirds of CO2 released by soils coming from root respiration (Ekblad & Hogberg, 2001; Bhupinderpal et al., 2003), the ability to model rates of R based on other traits such as tissue N concentration would greatly enhance our ability to understand and predict net ecosystem CO2 exchange. Do our results support the suggestion that R–N relationships are fixed? We compared our across-species log–log R–N plots [organic N basis (Fig. 3) and total N basis (Fig. S1, supplementary material)] with those of Tjoelker et al. (2005); although R increased with N in both our data set (organic and total N basis) and that of Tjoelker et al. (2005), the two data sets could not be superimposed on each other. Rather, for any given root N concentration, rates of root R were much higher in our data set than in that reported by Tjoelker et al. (2005). Moreover, we found that the slope of the relationships differed from that shown in Tjoelker et al. (2005). Several factors might explain such differences, including our use of hydroponically grown roots under controlled environment conditions, whereas Tjoelker et al. (2005) used field-grown roots. Secondly, unlike Tjoelker et al. (2005), we did not restrict our measurements to fine roots; rather, we measured root R on whole root systems, which would have consisted of a population of roots at various stages of development rather than just fine roots. In addition, unlike Tjoelker et al. (2005), many of the whole root systems sampled in our study may have exhibited responses indicative of ‘luxury uptake’ of N at high rates of N supply (Table 1). Whatever the cause of the differences, our results suggest that it may not be possible to assume a fixed relationship between root R and N in all environmental conditions and/or developmental stages.
Does the Q10 of root respiration vary with N concentration?
Theory tells us that variations in respiratory capacity, substrate supply and/or adenylates all potentially influence Q10 values of R (Atkin & Tjoelker, 2003). Given that respiratory capacity could be altered by changes in N supply (Noguchi & Terashima, 2006), we hypothesized that the Q10 of R would vary with N availability and N concentration in the tissues. Support for this hypothesis came from the work of Turnbull et al. (2005), who found that the Q10 of leaf R in temperate rainforest trees increased with increasing nutrient availability along a soil chronosequence in New Zealand. Although we found wide variation in the Q10 of roots, there were no systematic responses to N concentration (Fig. 4); if the same applies to the Turnbull et al. (2005) study on leaves, then variations in Q10 along the New Zealand soil chronosequence might not have reflected variations in soil N availability. Rather, variations in soil phosphate availability along such soil chronosequences (Fitter & Parsons, 1987; Richardson et al., 2004) may have been important. Whilst phosphate availability will have implications for photosynthesis, and hence on R through substrate availability, it may also affect R directly. Changes in soil phosphate availability have been shown to affect the extent to which mitochondrial electron transport is adenylate restricted (Gonzalez-Meler et al., 2001); moreover, changes in adenylate restriction can impact on the Q10 of R (Atkin et al., 2002; Covey-Crump et al., 2002). Whether variations in soil phosphate (or soil N) supply are responsible for the variations in Q10 reported for leaves by Turnbull et al. (2005) has yet to be determined.
The short-term Q10 values of root R in our study (measured using whole root systems) were quite low in comparison to those obtained for forest trees by Burton et al. (2002; Q10 values of 2.4–3.1), but were in a similar range to those obtained by Loveys et al. (2003) for R of hydroponically grown whole roots of herbaceous species. Bahn et al. (2006) found a negative correlation between the Q10 of root R and root diameter in alpine grasslands, implying that Q10 may decline with root age. In Loveys et al. (2003), Q10 values were lower for whole roots than for fully expanded leaves. Large differences in Q10 values can occur, therefore, between leaves and roots, and between fine roots and whole root systems. Further work is necessary if we are to understand the mechanistic basis for such differences in Q10 values.
The dependence of thermal acclimation of root R on N availability
One of the main objectives of our study was to quantify the dependence of thermal acclimation of root respiration on nitrogen supply. As stated in the Introduction, there is strong a priori evidence that N availability could influence respiratory acclimation, with cold acclimation being associated with increases in respiratory capacity (Atkin & Tjoelker, 2003; Armstrong et al., 2006) and increased concentrations of tissue N (Körner & Larcher, 1988; Ryan, 1995; Tjoelker et al., 1999). Changes in respiratory capacity would necessitate an overall increase in protein (and thus N) investment in the respiratory apparatus. Because of this, we hypothesized that the degree of acclimation of roots to a decrease in temperature would be greater at high N supply than at low N, and that cold acclimation of root R would be associated with an increase in root N concentration (with no change in the slope or intercept of the scaling relationship between R and N concentration). However, our results indicate that there was no consistent difference in the degree of acclimation with organic N concentration (Fig. 5) or total N concentration (data not shown). Moreover, although the slope of the scaling relationship between R and N was unaffected by growth temperature (Fig. 3), the results shown in Figs 2 and 3 demonstrate that CA plants had higher rates of R for a given N concentration than their WG counterparts; this was the case regardless of whether N concentration was expressed on a total and organic basis. Such results suggest that the higher rates of root R exhibited by CA roots were not simply a result of an increase in tissue N and associated changes in respiratory capacity. While changes in capacity (via increased mitochondrial numbers and/or capacity of individual mitochondria; Armstrong et al., 2006) might have contributed, in part, to the changes in respiratory flux, other factors, such as increased substrate availability and/or demand for adenylates, are likely to have been largely responsible for the higher rates of root R at a given N concentration in CA roots compared with their WG counterparts.
Although we found no systematic relationship between acclimation and N concentration (Fig. 4), considerable variability in the degree of acclimation was exhibited among the species. Variability in the degree of thermal acclimation of root R is common (Atkin et al., 2005). For example, both Burton et al. (2002) and Weger & Guy (1991) found no evidence of acclimation in roots of woody species, whereas acclimation of root R has been observed in saplings or young trees grown at constant elevated temperatures (Bryla et al., 1997; Tjoelker et al., 1999) or at elevated field temperatures (Bryla et al., 2001). Similarly, the degree of acclimation of root R was highly variable amongst 16 species in the study by Loveys et al. (2003). One factor that may contribute to such interspecific differences in the degree of acclimation is whether root tissues developed at the growth temperature in question or developed at an earlier temperature, after which they then experienced a change in growth temperature. Previous work with leaves has shown that tissue developed at a new temperature acclimates more fully to changes in temperature than previously developed tissue (photosynthesis: Strand et al., 1999; respiration: Talts et al., 2004; Armstrong et al., 2006). Work by Loveys et al. (2003) indicated that this may also be true for root systems, as the degree of acclimation of root respiration was greater in plants grown at a particular temperature than in those transferred to a new temperature regime where the tissues had developed at the previous temperature. Species with a high relative growth rate (RGR) might be expected to produce more new root tissue in a certain time period than plants with a low RGR (and thus exhibit a higher degree of acclimation), depending on their ability to maintain growth at low temperature or N supply.
Whilst we found that there was variation among species in their ability to acclimate to a new growth temperature at different levels of N supply, over a range of species there was a decrease in the absolute rates of R in response to reduced N supply and an increase in the absolute rates of R in roots that cold acclimate. Our results suggest that, while the slope of scaling relationships between root R and N concentration is not affected by growth temperature, cold-acclimated plants do exhibit a higher y-axis intercept than their warm-grown counterparts. From this we conclude that cold acclimation of root R is not dependent on cold-induced increases in N concentration, and that scaling relationships between root R and N concentration are not fixed.
In addition to finding that short-term Q10 values of root R did not vary systematically with root N concentration and were unaffected by long-term changes in growth temperature, we found that the proportional change in R in plants that are cold-acclimated (as measured by acclimation ratios) remained relatively constant at the different N concentrations. Such constancy of Q10 values and acclimation ratios, irrespective of N availability, may enable root R rates to be predicted for a range of growth temperature/N supply scenarios. Nevertheless, because Q10 values were similar across N concentrations, and at higher N there were higher rates of R, the absolute change in R at high N will be greater. With increased N availability through N deposition and changes in composition as a result of global climate change, these larger absolute changes at high N may have important implications for global carbon cycling.
This work has been supported by a Daphne Jackson Fellowship funded by Natural Environment Research Council (LJA) and a NERC research grant to OKA (NER/B/S/2001/00875). We would like to thank David J. Sherlock for expert technical assistance and Greg S. Thurland for assistance with the practical work.