Author for correspondence: R. Ceulemans Tel: +32 03 8202256 Fax: +32 03 8202271 Email:email@example.com
• Morphological and anatomical characteristics of needles from different whorl levels were examined on Scots pines (Pinus sylvestris), grown for 4 yr under either ambient or elevated (ambient + 400 µmol mol−1) CO2 concentrations in open-top chambers.
• Needle characteristics were studied using light microscopy, scanning electron microscopy and laser scanning confocal microscopy.
• Under the elevated CO2 treatment stomatal density was reduced on both adaxial and abaxial needle surfaces although the number of rows of stomata did not change significantly. Needle cross-sectional area increased by 10%; this was largely the result of an increase in needle thickness and, to a lesser extent, needle width. The increase in needle thickness was due to a large increase in mesophyll tissue. The relative area (i.e. proportion of the total area) of epidermis plus hypodermis, of resin canal, of xylem and of central cylinder decreased, whereas the relative area of needle phloem significantly increased.
• The results suggest that a prolonged exposure to elevated CO2 has an effect on needle structure, anatomical and stomatal characteristics of Scots pine needles.
Elevated CO2 often increases total leaf area, leaf weight and leaf weight-to-area ratio (Ceulemans, 1997; Norby et al., 1999). Leaf area development is an important determinant of total plant productivity and varies with environmental conditions (Taylor et al., 1994). However, much uncertainty remains as to how elevated CO2 affects leaf growth and anatomy. Short-term responses to elevated CO2 have indicated changes in leaf anatomy (Wullschleger et al., 1997), but it remains questionable whether individual leaf size, total leaf number, or both are influenced. Moreover, highly variable responses have been observed between tree species with determinate growth as compared with those with indeterminate growth (Gardner et al., 1995; Jach & Ceulemans, 1999). Anatomical features are plastic because leaf structure, leaf shape and cell distribution can change, within limits, together with leaf functions, which enable an adaptive response to rising CO2 (Long, 1998; Pritchard et al., 1998).
The leaf is the key organ for photosynthesis and transpiration. Therefore, leaf morphology and cell distribution may be important in influencing physiological processes (Parkhurst, 1986). The leaf contains both the assimilating and conducting tissues, and either or both tissues could be affected by CO2 (Assmann, 1999). Although leaf morphological features have been considered for understanding plant responses to rising CO2, the present knowledge based on morphology is still insufficient. It seemed necessary, therefore, to discriminate between responses of mesophyll and vascular tissues to elevated CO2, and also to take into account other microscopic features related to source-sink relations.
There have been studies concerning the responses of leaf anatomy to rising CO2, but most have focused on stomatal acclimation (Beerling, 1997; Jarvis et al., 1999). Few studies have examined changes in internal leaf structure and in leaf surface wax under elevated CO2, which may be important for bridging data collected at the physiological level to whole plant and canopy level processes (Prior et al., 1997; Pritchard et al., 1999). The aim of this study was to investigate the responses of Scots pine (Pinus sylvestris L.) needles after 4 yr of exposure to elevated CO2 to elucidate the response mechanism of different tissue types and to provide data for further interpretation of whole-plant source-sink relationships.
Materials and Methods
Plant material and growth conditions
Three-yr-old, pot-grown, dormant Scots pine seedlings obtained from the Institute for Forestry and Game Management, Groenendaal, Belgium, were planted in the ground in four open-top fumigation chambers (OTCs) on the campus of the University of Antwerp (UIA), Belgium, on March 21 1996. Mean annual temperature and rainfall at the site are 11.98°C and 769 mm, respectively. All seedlings were from the same Belgian provenance (south of the Samber and Maas rivers) and were about 0.4 m tall at the time of planting. Eleven trees per chamber were planted in a circular pattern, 70 cm apart from each other and from the walls. To reduce boundary effects, seedlings from the same batch of seeds were planted around each OTC. Before planting, the original heavy loam soil was excavated to a depth of 0.5 m and replaced with forest soil (about 0.12% N on a dry mass basis). No nutrients or water were applied during the experiment.
Each decagonal OTC (diameter 3 m, height 4 m) was made of 1 m wide perspex (Plexiglas, Belgium) sheets, and had a usable ground area of 7.1 m2. Incoming air was supplied at a rate of about 5000 m3 h−1, or nearly two air volume changes min−1. Air distribution within each OTC was effected through a flexible duct supplying a perforated, fiber-reinforced polythene annulus positioned 1 m above the ground. Two OTCs were assigned to each CO2 treatment. The ambient CO2 treatment provided about 350 µmol mol−1 of CO2 and the elevated CO2 treatment provided a CO2 concentration of ambient +400 µmol mol−1. The CO2 enrichment started on April 11996, and the treatment has been applied since then on a 24-h basis continuously throughout the year. For a more detailed description of the experimental conditions see Jach & Ceulemans (1999, 2000).
Samples for anatomical study
For anatomical observations, two trees from each OTC (two OTCs per treatment) and four whorl levels (WL3, WL4, WL5 and WL6) from each tree were selected. Most or nearly all needles from WL1 and WL2 had dropped off. Five current-year needles from each of the four whorl levels were randomly sampled from the trees after 4 yr of experimental treatment. Samples were collected on October 21 1999 and immediately placed in a solution containing FAA (5% Formalin, 5% Acetic acid, 90% alcohol). Needles were cut into pieces of 0.5–1.0 mm in length and kept in fixative until use. Afterwards, samples were washed in water and softened in a mixture of glycerin and ethanol for 1 d. Samples were then dehydrated through an ethanol series, embedded with paraffin and sectioned at 10 µm thickness with a rotary microtome. All sections were dewaxed by a xylin series, stained with safranin and counter-stained with fast green.
All sections were examined with a light microscope and measured using electronic image analysis equipment (Leica Q5000MC and QWin v.1.00 software, Wetzlar, Germany). On each slide, three to four replicates of needle width and needle thickness, thickness of epidermis plus hypodermis, resin canal diameter and frequency, width and thickness of the central cylinder were measured. Needle cross-sectional area, the relative area occupied by the epidermis plus hypodermis, central cylinder, mesophyll tissue, vascular bundles, and xylem and phloem within the vascular bundle were determined (Fig. 1).
Determination of stomatal density
Two trees from each OTC (two OTCs per treatment) were selected and five current-year needles per whorl level from each tree were sampled for both CO2 treatments. The needles were washed in 20% sodium hypochlorite, and rinsed in distilled water. They were then stained with a 1% diphenylboric acid 2-aminoethylester (Sigma chemicals, St. Louis, MO, USA) in methanol following the method described by Schnitzler et al. (1996) for about 5 min, rinsed in methanol, mounted under cover-slips in distilled water and examined with an epifluorescence light microscope. The number of rows of stomata on both needle surfaces was counted from samples excised from the mid-portion of needles along the longitudinal axis. Stomatal density was counted on both needle surfaces; measurements were again made in the mid-portion of each needle to minimize variability due to spatial heterogeneity (Yoshie & Sakai, 1985). Twenty counts were made on both adaxial and abaxial needle surfaces, respectively.
Scanning electron microscope (SEM) observations
Needle sections of about 5 mm long from two needles per whorl level on each of two trees per OTC (two OTCs per treatment) were sampled. To determine the differences in epicuticular waxes, two 2-yr-old needles from each whorl were additionally sampled. They were soaked in a 20% solution of sodium hypochlorite until the surface lost color, and then washed in running water. After dehydration through a graded alcohol series, the needle samples were critical point-dried, mounted on stubs, and coated with gold in a high-vacuum evaporation unit. Samples were examined at 15 kV under a JSM 6300 scanning electron microscope.
Needle samples were sectioned in the transverse plane at a thickness of 20 µm using a cryomicrotome, stained with 0.1% safranin for 20 min, and washed in water. Sections were then mounted in water under a cover-slip and examined on a Bio-Rad MRC600 (Cambridge, MA, USA) laser scanning confocal microscope attached to a Nikon microscope (Tokyo, Japan). The Bio-Rad confocal microscope was set to detector channel 1 (photomultiplier 1), which detects emissions larger than 560 nm. Images were collected at 4- and 2-µm intervals with 4× and 10× lenses, respectively.
Because there were only two OTCs (replicates) per treatment, subreplicates (trees) were used rather than real replicates (chambers), in the statistical analysis. All trees from two chambers belonging to the same treatment were pooled, giving four subreplicates. For the quantitative characteristics, five needles from the same whorl level (four whorl levels for each tree) were averaged. Initial soil characteristics, air temperature and photon flux density (PPFD) at tree level were similar in all four OTCs (Jach & Ceulemans, 1999, 2000). Therefore, differences in tree characteristics between OTCs were most probably attributable to differences in CO2 concentration of the air. To test the main effect of CO2 concentration (treatment) and position within the canopy, data were subjected to a two-way ANOVA. When the CO2 effects were significant, a pairwised comparison was further used to determine in which level the CO2 effects occurred. All statistical tests were performed with SPSS vs10.0 software.
Stomatal density and epicuticular wax
The abaxial needle surface frequently contained one or two additional rows of stomata than the adaxial surface in both treatments. The distance between the stomata along the rows, however, did not differ between adaxial and abaxial surfaces (data not shown). At whorl levels 3 and 4, the number of rows of stomata on the adaxial surface was significantly higher in the elevated CO2 treatment than in the ambient treatment. On the abaxial surface at whorl levels 3, 4 and 5, there was a trend of more rows of stomata in the elevated CO2 treatment, but the effect was not significant. On the adaxial needle surface of whorl level 5 and on both adaxial and abaxial surfaces of whorl level 6, not much difference was found in the number of rows of stomata. For all whorl levels combined, the elevated CO2 treatment did not affect the number of rows of stomata (Tables 1 and 2).
Table 1. Needle characteristics of needles of Pinus sylvestris trees grown under ambient or elevated CO2 conditions. Five current-year needles were collected from each of the four whorl levels on two trees per open top chamber (two OTCs per treatment). WL3, WL4, WL5 and WL6 indicate the whorl numbers where WL3 is the lowest and WL6 the highest whorl layer on the tree. Means and standard error (within brackets) of four trees per treatment are shown. Significance values are indicated as: ** P < 0.01, * P < 0.05 and ns = P ≥ 0. 0.05. Detailed results of the ANOVA are presented in Table 2
Number of rows of stomata on adaxial surface (No.)
Table 2. Statistical significance (results of two-way ANOVA) of the effects of whorl level (location within the crown of the tree), CO2 treatment (CO2 effect) and treatment × whorl level interactions on different characteristics of needles of Pinus sylvestris trees grown under ambient or elevated CO2 conditions. Values are presented in Table 1. Levels of significance are indicated as: ** P < 0.01, * P < 0.05 and ns = P ≥ 0.05
Whorl level effect
CO2 × whorl level interaction
Number of rows of stomata on adaxial surface (No.)
The responses of stomatal density to elevated CO2 showed no consistent pattern among needle surfaces and whorl levels on the tree. At whorl levels 3 and 4, there was no difference between CO2 treatments on the stomatal density of the adaxial surface, while stomatal density on the abaxial needle surface of whorl levels 3 and 4 was slightly reduced following exposure to elevated CO2. At whorl levels 5 and 6, stomatal density was reduced by 20.1% and 10.6% on the adaxial needle surface and by 11.1% and 23.2% on the abaxial needle surface in trees grown at high CO2. When all whorl levels were combined, a significant reduction in stomatal density was observed in the elevated CO2 treatment as compared with the ambient CO2 treatment (Tables 1 and 2).
The morphology of Florin rings surrounding the stomata and epicuticular waxes showed little difference between the CO2 treatments based on the observation of the samples excised from the mid-portion of the needle. This was not true for the samples close to and inside the needle sheath where Florin rings and waxes were obscure or absent. Young or developing needles generally had a higher density of wax tubes than older needles, but little difference was observed in morphology and in the density of wax tubes among different whorl levels in 2-yr-old needles (Fig. 2).
Anatomical needle characteristics
There was a significant interaction between CO2 treatment and whorl level on anatomical characteristics of needles. At all whorl levels, needle width did not differ between treatments. At whorl levels 3 and 4, needle thickness significantly increased under the elevated CO2 treatment. At whorl levels 5 and 6, needle thickness increased by 3.6% to 6.2% but the effect of elevated CO2 treatment was not significant. The average thickness of needles from all whorl levels increased by 6.4% under the elevated CO2 treatment, while only a 2.6% increase for the average needle width was observed under the elevated CO2 treatment (Tables 1 and 2).
The central cylinder width did not show changes at the upper whorl levels except for whorl level 3 where a slight increase was observed under the elevated CO2 treatment compared with the ambient CO2 conditions. There was a tendency for a slightly thicker central cylinder under elevated CO2 at all whorl levels. Nevertheless, the elevated CO2 treatment had no significant impact on the thickness of epidermis plus hypodermis, nor on the number of resin canals per needle at different whorl levels (Tables 1 and 2).
Needle area and the relative area occupied by different types of tissue
Much variation was observed in the needle cross-sectional area and the relative area occupied by different tissue types. Within the crown, the largest needles were found in the uppermost whorl level and the smallest needles in the lowest whorl levels. When needles from all whorl levels were combined, their overall average cross-sectional area increased very significantly by 10.4% under the elevated CO2 conditions. The relative area of the epidermis plus hypodermis, and of the resin canal slightly decreased under the elevated CO2 although epidermis plus hypodermis thickness as well as the number of resin canals per needle remained unaltered in the two treatments. The proportion of the central cylinder was significantly lower although its thickness was slightly higher in the elevated CO2 treatment compared with the ambient treatment (Tables 1 and 2). In contrast, the relative area of mesophyll tissue in the elevated CO2 treatment significantly increased by 5.7%, resulting from more and larger mesophyll cells in needles as shown in Fig. 3.
Despite a significant decrease in the relative area of the central cylinder, the proportion of the vascular bundles did not significantly change by the CO2 treatment. It was of interest, however, to notice that the average relative area of the phloem increased by 4.4% in contrast to a slight decrease for the relative area of the xylem in the elevated CO2 treatment compared with the ambient treatment (Tables 1 and 2).
Stomatal density affects gas exchange, stomatal conductance and instantaneous water-use efficiency (Woodward & Bazzaz, 1988). Thus, stomatal acclimation to the prevailing CO2 environment has been the subject of various studies (Peñuelas & Matamala, 1990; Woodward & Kelly, 1995; Poole et al., 1996; Bettarini et al., 1998). These studies generally reported a reduction in stomatal density when plants, including coniferous trees, were exposed to elevated CO2. For example, Beerling (1997) separated out the different stomatal density responses of the needle surfaces in P. sylvestris and reported a reduction in stomatal density in response to elevated CO2 in yr 1 of treatment. Contradicting observations, however, have been reported for P. banksiana (Stewart & Hoddinott, 1993) and P. pinaster (Guehl et al., 1994). The results presented here indicated that Scots pines, after 4 yr exposure to elevated CO2, reduced their stomatal density. There were no significant differences in the average number of rows of stomata between the two CO2 treatments. This might indicate that the lower stomatal density under elevated CO2 could largely be explained by an increase in epidermal cell expansion that gave rise to the larger needle size in the elevated CO2 treatment (Ceulemans et al., 1995). It was further observed that both adaxial and abaxial surfaces responded differently to the elevated CO2 treatment in terms of their number of rows of stomata and stomatal density, as also noticed in other experiments (Pearson et al., 1995). Such differences in CO2 sensitivity of the adaxial vs the abaxial surface could be attributed to the different light environments at each surface (Ceulemans et al., 1995) or different intrinsic cellular development (Sachs et al., 1993). Although we did not measure stomatal conductance in this experiment, Beerling (1997) and Medlyn et al. (2001) provided a direct evidence of reduction in stomatal conductance for P. sylvestris under elevated CO2. Considering the role of stomata in various physiological processes, the parallel changes in the reduction of stomatal density and stomatal conductance may have important consequences for the response of pine trees to water stress under elevated CO2 (Wang & Kellomaki, 1997). Recently a gene involved in the signal transduction pathway responsible for controlling stomatal numbers at elevated CO2 has been identified in Arabidopsis (Gray et al., 2000).
Many investigators observed that elevated CO2 stimulated leaf development (Radoglou & Jarvis, 1990; Taylor et al., 1994). The experiments described here indicated that needle cross-sectional area increased by 10.4% in the elevated CO2 treatment, confirming an earlier observation (Jach & Ceulemans, 1999). Nevertheless, the CO2 treatment had different effects on needle width and needle thickness. When all whorl levels were combined, needle thickness significantly increased, by on average 6.4%, in contrast to a non-significant increase in needle width.
Cell size and the number of cells dictate needle size. Changes in the differentiation of tissues within the needle may result in different physiological functions. Therefore, the relative area of mesophyll cells may be more closely related to the rate of photosynthesis than are the epidermis and the vascular bundles (Parkhurst, 1986; Evans, 1999; Roderick et al., 1999). In the present experiment, we found that the relative area of epidermis plus hypodermis, resin canal and central cylinder remained similar or decreased in the elevated CO2 treatment; only the relative area of the mesophyll tissue showed a significant increase. In some cross-sections, the development of a third mesophyll cell layer could be observed, giving rise to a large increase in area. This observation agreed with earlier findings made in P. taeda (Thomas & Harvey, 1983) and in Populus (Radoglou & Jarvis, 1990). In both studies it was concluded that the enlargement of cross-sectional area of needles resulted primarily from an increase in mesophyll tissue. Given the link between the proportion of mesophyll tissue and total chloroplast number per needle, this is an important finding, suggesting that the higher photosynthetic rate was likely to result from a larger area of mesophyll tissue. A higher photosynthetic rate under elevated CO2 was indeed observed (Gunderson & Wullschleger, 1994; Beerling, 1997; Jach & Ceulemans, 1997;2000).
Responses of vascular tissues to elevated CO2 have already been reported for several pine species (Thomas & Harvey, 1983; Conroy et al., 1986; Pritchard et al., 1997). Thomas & Harvey (1983) demonstrated that the area of the vascular tissue increased in P. taeda following exposure to elevated CO2. A similar result was also reported for P. radiata (Conroy et al., 1986) and for P. ponderosa (Pushnik et al., 1995). Pritchard et al. (1997), however, found that the phloem area decreased under elevated CO2 as a result of fewer sieve cells in the needles of P. palustris. The present study revealed a significant increase in the relative area of the phloem although the CO2 treatment did not have a significant effect on the relative area of the vascular bundle. The enhanced formation of phloem under elevated CO2 was in agreement with the findings of Ewers (1982). This author reported that the unifacial cambium in the vascular bundle produced secondary phloem rather than secondary xylem in several coniferous species including P. sylvestris grown in natural conditions. Such changes of the phloem area may imply that the capacity of the vascular bundle to transport fluids is enhanced since there is a link between vein structure and photoassimilate translocation (Körner et al., 1995; Jokela et al., 1997). After exposure to elevated CO2, photosynthetic rate is enhanced with the production of extra photoassimilates in P. eldarica (Garcia et al., 1994), P. taeda (Teskey, 1997) and P. sylvestris (Beerling, 1997; Jach & Ceulemans, 1999). Therefore, the increase in the area of the phloem cells may be interpreted as a positive response of the anatomical structure to the extra photoassimilates produced under elevated CO2 conditions.
The anatomical structure of the conifer needle has been considered an important diagnostic tool in the field and has been used as a bioindicator of environmental pollution (Sutinen & Koivisto, 1995). Likewise, structural characteristics of conifer needles can also be applied to determine the effects of elevated CO2 (Pritchard et al., 1997). In the light of the variability of certain anatomical characteristics, it was important to use strictly comparable material since the variation encountered among needles from different developmental stages or from different crown levels can be larger than the effects of different treatments (Kinnunen et al., 1999). The limitation of small samples together with the use of plant materials from different sources have been overlooked in some publications, and this may have accounted for much of the disagreement in the conclusions. It is strongly suggested that the mid-portion of mature needles should be selected preferably from one single crown level to further confine the variation to the minimum for comparative analysis.
In conclusion, changes in morphological and anatomical characteristics were found in Scots pine needles following a 4-yr exposure to elevated CO2. Leaf development was altered after exposure to elevated CO2. Nevertheless, stomatal density was reduced on both adaxial and abaxial needle surfaces. The significant increase in needle thickness together with a small increase in needle width accounted for a 10% increase in the needle cross-sectional area. The relative area of the epidermis plus hypodermis, resin canal and central cylinder remained similar or decreased in the elevated CO2 treatment, whereas the relative area of mesophyll significantly increased. Furthermore, the formation of sieve cells was enhanced after exposure to elevated CO2. The change of vascular bundles to CO2 enrichment may be linked to the transport of extra photoassimilates produced following the exposure to elevated CO2. In contrast with less competing species, Scots pine needles might structurally benefit from a prolonged exposure to elevated CO2.
This study was supported by the EC Fourth Framework Programme through its Environment & Climate RTD Programme (research contract ENV4-CT95–0077). JXL acknowledges support from the research Fund of the University of Antwerpen (UIA), Belgium. The authors thank Jie Le and W. Dorrine for their technical assistance during the course of this study, P. Van Espen (Department of Chemistry, UIA) for the use of electron microscope facilities, as well as D. A. Sampson, B. Gielen and two anonymous reviewers for their useful comments and suggestions on earlier drafts of this manuscript. Drawings for Fig. 1 were made by A. Muys (UIA).