1. Photosynthetic characteristics of an annual herb, Chenopodium album, and an evergreen tree, Quercus myrsinaefolia, were compared to clarify causes of the difference in photosynthetic nitrogen-use efficiency (photosynthetic capacity per unit nitrogen) between leaves of herbaceous and evergreen species.
2. When leaves with the same nitrogen content on an area basis were compared, photosynthetic capacity of C. album was twice as high as that of Q. myrsinaefolia. Gas-exchange measurements showed higher intercellular CO2 concentration in C. album. Biochemical analyses indicated larger allocation of nitrogen into ribulose-1,5-bisphosphate carboxylase, a key enzyme of photosynthesis, and higher specific activity of ribulose-1,5-bisphosphate carboxylase in C. album. However, these differences were not large.
3. Compositional deviation of 13C in leaves of the two species suggested that the drop of CO2 level between the intercellular space and the chloroplast was slightly larger in Q. myrsinaefolia when compared between the leaves with the same photosynthetic capacity.
4. It is concluded that the difference in photosynthetic nitrogen-use efficiency between C. album and Q. myrsinaefolia is not caused by a sole factor that is markedly different between the two species but by several factors each of which is slightly disadvantageous to Q. myrsinaefolia compared with C. album.
It is well known that a correlation between photosynthetic capacity and nitrogen content of the leaves for any species is fairly strong. This may reflect the fact that more than half of leaf nitrogen is allocated to the photosynthetic apparatus (Evans 1989). However, the dependence of photosynthetic capacities on nitrogen contents strikingly varies among species when both are expressed on a leaf area basis (Field & Mooney 1986; Evans 1989). In particular, leaves of evergreen trees and shrubs show very low photosynthetic capacities relative to those of crop and herbaceous plants. For example, in leaves with 0·1 mol N m–2, photosynthetic capacities are lower than 10 μmol CO2 m–2 s–1 in trees or shrubs, while those of some crop species achieve 25 μmol CO2 m–2 s–1 (Evans 1989).
Because the first step of CO2 assimilation is an enzymic reaction, difference in its capacity may be attributed to the amount of the substrate, the amount of the enzyme or the specific activity of the enzyme. Therefore, hypotheses as to the cause of differences in photosynthetic nitrogen-use efficiency (PNUE, photosynthetic capacity per unit nitrogen) are: (1) photosynthetic capacities in evergreen leaves of woody plants may be limited by the large resistances to diffusion of CO2 (the substrate) because their mesophyll cell walls are thick or impermeable; (2) lower photosynthetic capacities may result from smaller allocation of nitrogen to photosynthetic enzymes; (3) inefficient allocation of nitrogen among photosynthetic enzymes may result in lower photosynthetic capacities; (4) the specific activity of photosynthetic enzymes may be lower in evergreen plants (Field & Mooney 1986).
Recently, measurements of 13C discrimination with an open gas-exchange system and of chlorophyll fluorescence enabled the estimation of a conductance for CO2 diffusion from the intercellular spaces to chloroplasts (Evans et al. 1986; von Caemmerer & Evans 1991; Harley et al. 1992; Loreto et al. 1992). Using 13C-discrimination measurement, Lloyd et al. (1992) showed that the CO2 concentration at chloroplasts was lower in species with lower PNUE. Such tendencies were also obtained by Epron et al. (1995). Lloyd et al. (1992) also estimated the amount of RuBPCase (ribulose-1,5-bisphosphate carboxylase, a key enzyme of photosynthesis) per unit leaf area from gas-exchange characteristics. They suggested that the amount of RuBPCase per unit of nitrogen is higher in species with higher PNUE. However, they did not directly measure the amount or activity of RuBPCase. Although some investigators measured the amount or activity of RuBPCase in woody leaves (Vu & Yelenosky 1988; Vapaavuori & Vuorinen 1989; Kursar & Coley 1992), dependency of photosynthetic capacities on the RuBPCase content has not been addressed. For kinetic parameters of RuBPCase, the specific activity, the Michaelis constant and the activation state of RuBPCase of Citrus sinensis, an evergreen tree, were shown to be comparable to those of herbaceous species (Vu & Yelenosky 1988).
Because these previous studies focused on each of the above-mentioned factors separately, relative importance of these four factors for the low PNUE in evergreen species is still unclear. Makino, Mae & Ohira (1988) analysed PNUE of rice and wheat. They showed that, relative to wheat leaves, rice leaves had higher stomatal conductance and higher allocation of leaf nitrogen to RuBPCase but had the lower specific activity of RuBPCase. Consequently, these two species had similar PNUEs. This fact suggests that each of the above-mentioned factors is potentially variable even among species exhibiting similar PNUE. Therefore, these factors should be compared for the same data set.
In the present study, we quantified causes of different PNUE between an annual, Chenopodium album L., and an evergreen tree, Quercus myrsinaefolia Blume. For each leaf, gas-exchange characteristics, nitrogen content, RuBPCase content and 13C composition were determined. We evaluated the quantitative importance of each factor in determining the PNUE of the two species. Field & Mooney (1986) indicated that NUE is lower in leaves with lower photosynthetic capacity within a species. This is because the regression of photosynthetic capacity on nitrogen content generally has the positive x-intercept. In the present study, plants were grown under three light intensities to obtain leaves with various nitrogen contents.
Materials and methods
Two species were used in the present study: C. album (Chenopodiaceae), an annual herb, which colonizes disturbed habitats and Q. myrsinaefolia (Fagaceae), an evergreen broad-leaved tree, which is one of the dominant species in the climax broad-leaved evergreen forests in Japan. Seeds of these species were collected at Tokyo, Japan.
Plants were grown in a greenhouse of Tohoku University, Sendai, Japan (38° N, 141° E). The greenhouse was made of steel frames covered with transparent sheet. During the growth season, the side sheet of the house was removed to make the air conditions inside the greenhouse similar to those on the outside. The plants were raised in pots (15 cm diameter, 12 cm height) filled with soil. For Q. myrsinaefolia, current-year leaves of 3 year-old plants were used. Every pot was automatically watered every morning and evening. 100 ml of Hoagland solution (12 mM nitrate, see Hikosaka, Terashima & Katoh 1994 for detailed composition of the nutrition) was supplied every week. Plants were grown under three light conditions, 10%, 30% and 100% of full sunlight. For shade treatments, plants were kept in two shade frames (1·5 m × 1 m × 1 m high each) covered with shade cloth.
MEASUREMENT OF PHOTOSYNTHETIC PROPERTIES
Rates of CO2 exchange and transpiration of leaves were determined with an open gas-exchange system. CO2 concentration was detected with an infrared gas analyser (Model ZRC, Fuji, Japan). Water vapour pressure was detected with a dew-point sensor (Model D-2, General Eastern, USA). Leaf temperature was measured with a copper-constantan thermocouple and was adjusted to 25·7 ± 0·3 °C. Vapour pressure deficit was maintained at less than 1 kPa.
CO2 dependence of the light-saturated rate of photosynthesis was determined. First, the photosynthetic rate under high CO2 concentration was measured and CO2 concentration was decreased stepwise. Then, respiration rates were determined in the dark.
After the photosynthetic measurements, four leaf discs of 1 cm diameter were punched out from one leaf. With one of these discs, the chlorophyll (chl) content was determined spectrophotometrically after extraction with dimethylformamide (Porra, Thompson & Kriedemann 1989). The other three discs were dried in an oven for more than 3 days and their nitrogen content was determined with an NC analyser (NC-80, Shimadzu, Kyoto, see Hikosaka et al. 1994). The residual part of leaves was frozen in liquid nitrogen and stored at – 70°C. The residual part was further divided for determination of RuBPCase content and for 13C composition (see below). For determination of RuBPCase content, the frozen leaf was homogenized in a 100 mM Na-phosphate buffer (pH 7·5) containing 0·4 M sorbitol, 10 mM NaCl, 2 mM MgCl2, 5 mM iodine acetate, 1 mM phenylmethyl sulfonyl fluoride, 5 mM dithiothreitol and 2% (w/v) polyvinylpyrrolidon. After filtration through 20 μm mesh, concentration of chl in the filtrate was determined with 80% acetone (Porra et al. 1989). Then, the filtrate was applied to SDS-PAGE. The gel was stained with Coomassie Brilliant Blue R-250 (CBB). The band of the large subunit of RuBPCase was extracted with formamide for spectrophotometric determination of the RuBPCase content (Makino, Mae & Ohira 1986). Calibration curves were made with the RuBPCase purified from Spinacia oleracea L. and the protein concentration was determined using the method of Lowry et al. (1951). The RuBPCase content per unit leaf area was calculated as a product of the RuBPCase/chl ratio and chl content per unit leaf area.
The RuBPCase activity of another set of leaves was determined spectrophotometrically (UV-160, Shimadzu, Kyoto, Japan) by coupling phosphoglycerate kinase and glyceraldehyde-3-phosphate dehydrogenase (Usuda 1985). Leaves were incubated under saturating light for several 10 min periods to activate RuBPCase. Then, leaf discs of 1 cm diameter were immediately punched out and frozen in liquid nitrogen. These discs were stored at – 70°C until measurement of the RuBPCase activity. A leaf disc was ground with a chilled mortar and pestle in 600 μl of an extraction buffer containing 50 mM Mops-NaOH (pH 7·5), 1 mM EDTA, 15 mM MgCl2, 5 mM dithiothreitol, 1% Triton X-100, 1 mM phenylmethyl sulfonyl fluoride and 1% polyvinylpolypyrrolidon. The homogenate was centrifuged at 10 000 g for several 10 s periods. The supernatant was incubated for 5 min at 25°C in a reaction buffer containing 0·1 M Bicine-KOH (pH 8·2), 1 mM EDTA, 30 mM MgCl2, 25 mM NaHCO3, 2·5 mM dithiothreitol, 0·2 mM NADH, 5 mM ATP, 2 mM phosphocreatine, three units ml–1 each of phosphocreatine kinase and carbonic anhydrase, and 10 units ml–1 each of phosphoglycerate kinase and glyceraldehyde-3-phosphate dehydrogenase. Changes in the absorbance at 340 nm were measured immediately after addition of 40 μl of 15 mM ribulose-1,5-bisphosphate. Enzymes used were purchased as sulfate-free powder from Sigma. A part of the supernatant was applied to SDS-PAGE to determine the specific activity of RuBPCase as described above.
STABLE CARBON ISOTOPE MEASUREMENTS
Part of the frozen leaves was dried at 70°C for 3 days and ground to a fine powder. The carbon isotope ratios were measured for subsamples of 0·1–0·2 mg with a system combining an elemental analyser (EA1108, Carlo Erba, Italy) for combustion of the samples and a stable isotope ratio mass spectrometer (MAT 252, Finnigan MAT, Bremen, Germany). The carbon isotope ratios are expressed as the compositional deviation (δ), calculated according to the following equation:
δ13C is presented according to per mil (‰) deviations from the international standard (Pee Dee belemnite: PDB). Rsample and Rstandard are the 13C/12C ratios of the samples and the standard, respectively. The precision for isotope ratio measurement was ± 0·08‰.
13C discrimination (Δ) in leaves is calculated as follows:
where δair and δLeaf are δ13C in air and leaf, respectively. According to the model of Evans et al. (1986), Δ is given as follows:
where Ca, Cs, Ci and Cc are the partial pressure of CO2 in air, at the leaf surface, in the intercellular spaces and in the chloroplasts, respectively, ab, a and ai are the fractionations owing to diffusion through the boundary layer, through stomata and from the intercellular spaces to chloroplasts, respectively, and b is the net fractionation by RuBPCase and phosphoenolpyrvate carboxylase. The symbols e and f represent the fractionations associated with dark respiration R and photorespiration, k stands for the carboxylation efficiency of RuBPCase and Γ* is the CO2 compensation point in the absence of dark respiration.
It is possible to estimate the carbon isotope discrimination (Δi) that would occur if discrimination during respiration and photorespiration were zero and the partial pressure of CO2 in chloroplasts equals the intercellular partial pressure of CO2 (Cc = Ci) as:
Subtracting eqn 3 from eqn 4 shows that the difference in Δ and Δi is proportional to the drop in CO2 partial pressure from the intercellular space to the chloroplasts (Evans et al. 1986):
PHOTOSYNTHESIS, STOMATAL CONDUCTANCE AND RUBPCASE
Relationships between photosynthetic capacity and nitrogen content of the leaves are shown in Fig. 1. Photosynthetic capacity is defined as the rate of photosynthesis at Ca = 35·40 ± 0·67 Pa under saturating light. In both species, a significant correlation was observed (see Table 1 for the regression coefficients). When compared at the same nitrogen content, the difference in photosynthetic capacity between C. album and Q. myrsinaefolia was larger than twofold.
Table 1. . Regression coefficients of relationships between various photosynthetic characteristics of Chenopodium album and Quercus myrsinaefolia. Results of ANCOVA calculated according to Sokal & Rohlf (1981) are also shown as probability (P). Units of each parameter are μmol m-2 s-1 for photosynthetic capacity and RuBPCase activity, mol m-2 for nitrogen content, μmol m-2 s-1 Pa-1 for the initial slope of the Ci response curve, g m-2 for RuBPCase content and per mil for δ13C and Δi–Δ
In Fig. 2, the ratio of CO2 partial pressure in the intercellular spaces to that in ambient air, Ci/Ca, is compared. The intercept of the regression was significantly lower in Q. myrsinaefolia (Table 1). Mean and SD of Ci/Ca were 0·79 ± 0·06 and 0·69 ± 0·05 for C. album and Q. myrsinaefolia, respectively. The lower CO2 level at the intercellular space is therefore one of the causes of lower PNUE in Q. myrsinaefolia. The initial slope of the Ci response curve of photosynthesis is plotted against the nitrogen content (Fig. 3). This parameter reflects photosynthetic activity at the same Ci (previously called ‘mesophyll conductance’. See von Caemmerer & Farquhar 1981; Evans 1989). Leaves of Q. myrsinaefolia had the lower initial slopes at a given nitrogen content (Table 1).
Figure 4 compares the relationship between contents of RuBPCase and nitrogen. For the leaves of the same nitrogen content, Q. myrsinaefolia had significantly lower contents of RuBPCase (Table 1), indicating that this also contributes to lower PNUE in Q. myrsinaefolia. Assuming that the molecular mass of RuBPCase is 550 000 and that 16% of mass of protein is nitrogen, the calculated fraction of nitrogen allocated to RuBPCase was 29·2 and 23·7% for C. album and Q. myrsinaefolia at 0·1 mol N m–2, respectively. Photosynthetic capacities were still lower in Q. myrsinaefolia when compared at the same RuBPCase content (Fig. 5).
Mean ± SD of the specific activity of RuBPCase (the maximum velocity per unit enzyme amount at saturated CO2 concentration and saturated RuBP) was 20·8 ± 2·4 and 18·2 ± 3·4 mol CO2 mol–1 protein s–1 for C. album and Q. myrsinaefolia, respectively. In Fig. 6, the initial slope of the Ci response curve of photosynthesis is plotted against the amount (Fig. 6a) and activity (Fig. 6b) of RuBPCase both expressed on a leaf area basis. The latter was calculated as the RuBPCase content multiplied by its specific activity. When plotted against the RuBPCase content, the initial slope was significantly higher in C. album (Fig. 6a, Table 1). 6Figure 6b shows the relationship between the initial slope and the RuBPCase activity. Although the relationship between the two species was similar, the difference in the slope of the regression was marginal (Table 1).
The compositional deviation of 13C (δ13C) value is considered to reflect the CO2 concentration in the chloroplast and is known to be correlated with the instantaneous discrimination of CO2 determined under saturating light (Evans et al. 1986). Chenopodium album and Q. myrsinaefolia leaves had δ13C values of – 30·4 ± 1·6 and – 29·0 ± 1·5‰, respectively. 7Figure 7a shows δ13C of the leaf plotted against Ci/Ca. The negative correlation between δ13C and Ci/Ca indicates that the difference in δ13C is partly owing to the difference in Ci/Ca between the species. 7Figure 7b shows Δi–Δ, which is considered to be proportional to Ci–Cc (see Materials and methods), plotted against photosynthetic capacity. For calculation of Δi, a and b are assumed to be 4·4 and 29‰, respectively (Evans et al. 1994) and the boundary resistance is ignored (Ca = Cs). It is assumed that δ13C of air is – 8‰. For the leaves of the same photosynthetic capacity, Q. myrsinaefolia leaves tended to have high values of Δi–Δ, suggesting higher Ci–Ca in the Q. myrsinaefolia.
SUN–SHADE ACCLIMATION: ORGANIZATION OF PHOTOSYNTHETIC COMPONENTS
Figure 8 shows dependence of photosynthetic characteristics on growth irradiance. Dependence of the nitrogen content on PFD was different between the two species: the nitrogen content was steeply decreased with decreased PFD in C. album while it was relatively stable in Q. myrsinaefolia. Both ratios of RuBPCase/chl and of chl a/b decreased with decreased PFD. Each of the ratios was similar between the two species when compared at the same PFD. 8Figure 8d also shows the changes in the leaf mass per area. In both species, leaf mass per area increased with increasing growth irradiance and Q. myrsinaefolia always had larger leaf mass per area.
The difference in the PNUE between C. album and Q. myrsinaefolia was more than twofold (Fig. 1). The difference in PNUE between these species was attributed to the difference in several factors: CO2 concentration at chloroplast owing to differences in Cc/Ci and in Ci/Ca, nitrogen allocation to RuBPCase, and the specific activity of RuBPCase. However, as discussed below, each of the differences was not considerable.
The ratio of Ci to Ca was 14% higher in C. album (Fig. 2). According to the biochemical model of Farquhar, von Caemmerer & Berry (1980), if other factors are the same, this difference in Ci/Ca would cause c. 11% difference in the photosynthetic rate. It has been believed that the Ci/Ca ratio is similar among most of C3 species. Yoshie (1986) determined Ci/Ca of 27 temperate species with different life forms and from different microhabitats and found that values of Ci observed in those species were similar to each other and in the range of values obtained in other studies. He suggested common mechanisms working to maintain the stomatal conductance proportional to photosynthetic capacity over a wide variety of C3 species. The values of Ci/Ca obtained by Yoshie (1986) were 0·67–0·82. Although C. album had higher values, the Ci/Ca ratios of both species are included in this range.
Low Ci/Ca in Q. myrsinaefolia leaves suggests that this species realizes high water-use efficiency (WUE, photosynthesis per unit transpiration) owing to relatively low stomatal conductance (Farquhar & Sharkey 1982). This theoretically suggests an intrinsic trade-off relationship between PNUE and WUE (Field, Merino & Mooney 1983). However, several studies showed that this trade-off is not always true. Sobrado (1991) compared several deciduous and evergreen species in tropical dry forest and showed that both WUE and PNUE were always lower in evergreen species. Poorter & Farquhar (1994) examined water use and photosynthesis of 24 wild species differing in relative growth rate. The relative growth rate was positively correlated with nitrogen availability of habitat of species (Poorter & Remkes 1990) and with PNUE (Poorter, Remkes & Lambers 1990). However, they reported that the relative growth rate was not correlated with either Ci/Ca or water-use efficiency. These results suggest that Ci/Ca is not always the major factor responsible for difference in PNUE among species.
Higher allocation of nitrogen into RuBPCase in C. album also contributed to higher PNUE (Fig. 4). From literature survey, Evans (1989) suggested that the fraction of leaf nitrogen invested in RuBPCase varies from 15·9 to 28·0% among herbaceous species. Makino et al. (1992) also reported that the fraction of nitrogen allocated to RuBPCase ranged from 15 to 36% among five crop species. Therefore, the difference in allocation of nitrogen into RuBPCase between C. album and Q. myrsinaefolia was also not exceptional.
Two possible causes may be suggested for the smaller allocation of nitrogen to RuBPCase in woody species: smaller allocation of nitrogen to the photosynthetic apparatus and difference in the nitrogen partitioning among photosynthetic components (Field & Mooney 1986). Because both of RuBPCase/chl and chl a/b ratios were similar between C. album and Q. myrsinaefolia when compared at the same PFD (Fig. 8), the latter may not be true in this case. Hikosaka & Terashima (1996) also compared nitrogen partitioning among photosynthetic components in C. album and a shade plant, Alocasia odora. Except for C. album grown at very low light intensity (5% of sunlight), these species showed nitrogen partitioning similar to each other. Hikosaka & Terashima (1995) theoretically showed that suboptimal nitrogen partitioning can decrease daily photosynthesis by 10%. It was also suggested that nitrogen partitioning in A. odora and C. album was close to the theoretical optimum (Hikosaka & Terashima 1996). Nitrogen partitioning among photosynthetic components may be similar among C3 species.
On the other hand, it is still unclear where nitrogen allocated to other than RuBPCase is allocated in Q. myrsinaeflia. There are several possibilities. First, the secondary compounds, such as alkaloids, accumulated to avoid herbivory. Second, because evergreen leaves have thick cell walls of mesophyll cells, larger investment of cell-wall proteins may be necessary. Allocation to such nitrogen compounds may be necessary to maintain leaves with longer longevity.
It is also possible that affinity of CBB to RuBPCase differs depending on species. However, homology in amino acids sequence of large subunit of RuBPCase between Chenopodiaceae and Fagaceae is 94–95% (J. Yokoyama, personal communication). Therefore, the affinity of CBB to RuBPCase would not differ much between these species.
Higher specific activity of RuBPCase in C. album may be responsible for higher PNUE. It is known that the specific activity of RuBPCase varies among species (Evans 1989). For example, Makino et al. (1988) obtained 22·9 and 15·5 mol mol–1 s–1 for wheat and rice, respectively, and Seemann et al. (1987) obtained 29·0 and 19·7 mol mol–1 s–1 for common bean and Alocasia macrorrhiza, respectively. The difference in the specific activity of RuBPCase between C. album and Q. myrsinaefolia is not as large as those studies.
When the initial slope of the Ci response curve of photosynthesis is plotted against RuBPCase activity per unit leaf area, the difference between the two species became very small (Fig. 6b), while a clear difference was found when the initial slope was plotted against the RuBPCase content (Fig. 6a). However, because the specific activity was determined at saturated CO2 concentration, it may not represent the carboxylation rate in situ if the Michaelis constants of RuBPCase (Km) vary. For example, Makino et al. (1988) reported that both the specific activity and Km were higher in wheat than in rice. Consequently, the difference in the net activity of RuBPCase at the physiologically realistic CO2 concentration between rice and wheat was 22% although the difference in the specific activity was 49%. If there is such a positive correlation between the specific activity and Km, the difference in the initial slope of the Ci response curve of photosynthesis between the two species may be much larger than that shown in 6Fig. 6b). This suggests that the photosynthetic rate per unit RuBPCase activity may still be different between the two species even when Ci is the same.
As the cause of this difference, difference in (Ci–Cc) may be suggested. Several authors indicated that CO2 levels at chloroplasts are lower in species with lower PNUE (Lloyd et al. 1992; Epron et al. 1995). In the present study, when leaves with the same photosynthetic capacity are compared, Δi–Δ was slightly larger in Q. myrsinaefolia (Fig. 7b). If we assume that b–ai = 27·2 (Evans et al. 1994) and that (eR/k + fΓ*)/Ca = – 3‰, the difference in (Ci–Cc)/Ca between the two species is 5–10%. Although it is still possible that this estimation may be invalid because obtained values of Δ are derived not from gas-exchange measurement but from leaf material, this difference is comparable to that shown by Epron et al. (1995). According to a model incorporating the Cc response of photosynthesis (Hikosaka 1997), difference in Cc by 2 Pa [difference in (Ci–Cc)/Ca by c. 10%] is expected to cause a difference in the photosynthetic rate by 9%.
Medina (1971) speculated that CO2 diffusion in evergreen leaves is limited because they have thick or impermeable cell walls. Recent works have suggested that the cell wall of mesophyll cells is the most responsible factor for limitation of CO2 diffusion in a leaf. Evans et al. (1994) showed that conductance of CO2 diffusion in a leaf was correlated with the area of chloroplast exposed to cell walls, rather than with intercellular characteristics, such as leaf thickness. In a theoretical study, we also suggested that the intercellular characteristics have minor effects on CO2 diffusion (Terashima et al. 1995). We also obtained preliminary data indicating that the ratio of photosynthetic rate to RuBPCase content is negatively correlated with cell-wall thickness (Terashima et al. 1995). These results suggest the validity of the speculation of Medina (1971). Thick or impermeable cell walls may be necessary to have longer leaf life span.
Between C. album and Q. myrsinaefolia, several factors were involved in the more than twofold difference in PNUE. However, quantitative difference in each factor was rather small. If other factors were the same, the difference in the photosynthetic rate caused by Ci/Ca, Cc/Ci, allocation of nitrogen to RuBPCase and the specific activity of RuBPCase would be 11%, 4–9%, 23% and 14%, respectively. For an evergreen tree, Q. myrsinaefolia, we could not find any characteristics being considerably different from herbaceous species. The present results suggest that the difference in PNUE among species results from small differences in several factors, not from a sole factor. Evergreen species have to maintain their leaves with longer life span. Adaptation necessary for the maintenance, probably such as high WUE, thicker cell walls and allocation to secondary compounds, may impose low PNUE. However, this adaptation may enable longer leaf life span, which would contribute to the lifetime integral of nitrogen-use efficiency owing to longer residence time of nitrogen (Aerts 1990).
We thank Dr Y Suzuki for generous gift of purified RuBPCase, K. Satoh for helping to raise plants, Drs A. Makino and H. Usuda for advice regarding the determination of RuBPCase amount and activity, Dr J. Yokoyama for information about amino acid sequences of RuBPCase. This study was supported in part by National Institute of Agro-Environmental Sciences, by grants from Japan Ministry of Education, Science and Culture, and by a grant from Japan Ministry of Agriculture, Forestry and Fishery [Biocosmos project]. Y.T.H. was supported by Research Fellowships of the Japan Society for the Promotion of Science for Young Scientists.
Present address: Department of Biology, Graduate School of Science, Osaka University, Machikaneyama, Toyonaka, Osaka, 560–0043, Japan.