- Top of page
- Materials and Methods
Forests play a significant role in the global carbon cycle and in the control of carbon dioxide concentration in the atmosphere. They not only passively undergo global climatic changes but are also driving factors that may influence the course of climatic change. The scientific community has therefore sought to assess and quantify both the response and the contribution of forests (Dixon et al., 1994). Our current knowledge of tree responses to elevated atmospheric CO2 concentrations under different experimental conditions has been summarized in recent reviews (Eamus & Jarvis, 1989; Ceulemans & Mousseau, 1994; Jarvis, 1998; Saxe et al., 1998; Norby et al., 1999). Almost all experiments demonstrated that tree growth – quantified in terms of stem height, total biomass or leaf area development – increased under enhanced levels of CO2 concentration. Elevated CO2 also stimulated biomass production of plants with indeterminate growth more than for plants with determinate growth, presumably because of differences in sink strength (Oechel & Strain, 1985). As stem growth and standing biomass are the most obvious indications of the performance of a forest ecosystem, above-ground growth has primarily been used to quantify the response of forests to global change (Norby et al., 1999).
The Free-Air CO2 Enrichment (FACE) technology has been developed to treat and examine entire ecosystems while minimizing the environmental disturbance between the CO2-treated and the surrounding ‘control’ plant communities (Hendrey et al., 1993; Norby et al., 1999).
Poplar was chosen for this experiment in order to quantify the carbon sequestration capacity of an intensively managed tree plantation with potential as a partial sink for surplus, man-made CO2 emissions. Poplars are the most promising trees for short rotation cultures (SRC), a kind of agro-forestry system in which trees are intensively managed with the aim of maximizing biomass production for pulp, timber or energy (Beaton et al., 1991; Galinski et al., 1991; Ceulemans et al., 1992). A number of studies have been carried out on the effects of elevated atmospheric CO2 on Populus (Radoglou & Jarvis, 1990; Bosac et al., 1995; Ceulemans et al., 1995; Curtis et al., 1995; Gardner et al., 1995; Pregitzer et al., 1995; Kalina & Ceulemans, 1997; Tognetti et al., 1999), but all of them had a limited duration of treatment and were performed on individual or isolated plants except for the Aspen FACE experiment on Populus tremuloides (Isebrands et al., 2001; King et al., 2001). Before now there have been very few manipulative experiments where tree growth has been studied until the closed-canopy stage (Norby et al., 1999; Norby et al., 2001). In the present experiment growth dynamics were followed during three growing seasons, from planting until canopy closure, i.e. with LAI values from 0 until about 8 (Gielen et al., 2003).
Because growth was initially stimulated by FACE (Calfapietra et al., 2001), competition may differ between control and FACE treatments, and some assumptions have been made about competition in our study (Gielen et al., 2001, 2002). Therefore, to understand the growth of a stand exposed to elevated CO2, particularly of a high-density plantation, information about the competition in the stand is strongly needed. Generally, two models of competition are supported in the literature, one-sided (asymmetric) competition and two-sided competition (symmetric or asymmetric) (Perry, 1985; Weiner, 1985, 1986; Toméet al., 1994). When competition is one-sided, larger trees obtain a disproportionate part of resources and suppress the growth of smaller individuals, whereas the term two-sided competition is used when resources are shared equally or in proportion to size (Soares & Tomé, 1996). An important hypothesis relates one-sided competition to competition for light, and two-sided competition to the use of water and/or nutrients (Weiner, 1986; Soares & Tomé, 1996).
At the POPFACE experiment a complete harvest was carried out after 3 yr of enrichment, i.e. at the end of the first rotation cycle, evidencing a significant stimulation of total biomass in FACE treatment (Calfapietra et al., 2003). Given that, the objective of the present research was to study the above-ground growth dynamics throughout the first 3 yr of CO2 enrichment at the POPFACE experiment to understand whether the stimulation under elevated CO2 changed with time and in particular if it was also sustained after canopy closure. The second objective was to study whether this response in poplar was species-specific.
- Top of page
- Materials and Methods
Although the first evident result is a significant FACE effect on growth parameters after 3 yr of exposure – also confirmed by biomass data (Calfapietra et al., 2003) – this paper aims to analyse the dynamics of growth in each of the 3 yr characterized by extremely different conditions of the stand. The response to elevated CO2 differed between the first and the second year, whereas fewer differences were evidenced between the second and the third year. The maximum LAI reached values between 4 and 5 in the second year (Gielen et al., 2001) and between 5 and 8 in the third year, depending on the species and treatment (Gielen et al., 2003), providing evidence of canopy closure and the onset of strong competition amongst trees. Growth stimulation by the FACE treatment started immediately after planting for P. nigra and P. × euramericana as demonstrated by the dynamics of BA and height (Fig. 1a,b) and the CO2 effect on D2H (Fig. 2). This initial stimulation during the first phase of CO2 treatment is in line with that demonstrated for various herbaceous and tree species (Bazzaz et al., 1993; Poorter et al., 1996) as well as for clones of cherry and Sitka spruce (Centritto et al., 1999).
The POPFACE experiment clearly demonstrated species-specific differences in the extent of CO2 stimulation. The response was weak for P. alba and only became evident 2 months after planting, whereas it was particularly rapid and strong for P. nigra, resulting in a doubling of the D2H in FACE when compared with the control treatment at the end of the first year. The positive response was caused by an increase in BA rather than in height, which was stimulated by about 10% in the FACE treatment for all species (Table 1). Nevertheless the treatment effect on height was significant at the end of the first and the third years (Table 2). Similar observations were made by Tognetti et al. (1999) on poplar and by Percy et al. (2002) on aspen.
During the second year, stimulation under elevated CO2 was lost and RGR of D2H became higher in ambient CO2 (Fig. 3). However, at the end of the second year the size of FACE trees, namely BA and D2H, remained significantly bigger than control trees.
It should be emphasized that comparing growth in the control and FACE treatments was complicated at the onset of the second growing season because of treatment related size differences. Many studies have shown that elevated CO2 causes an enhancement of plant development, especially during the early stages (Bazzaz et al., 1993; Norby et al., 1995), but later growth can be very similar when analysing plants of similar size rather than at the same time (Centritto et al., 1999), as was observed in our experiment. When comparing the AGR in control and FACE treatments at similar tree sizes the stimulating effect of FACE was very prominent in the first year but not in the others, which was supported by our analysis of covariance (Table 3). In the second year, slope was higher in control but intercept was higher in FACE, which suggests that overall growth was not different between treatments. We hypothesize that higher values of second-year RGR of D2H in the control treatment are simply due to the larger size of trees in the FACE treatment, which had ontogenetically reduced the response of plants to elevated CO2 (Jarvis & Jarvis, 1964; Evans, 1972). This is confirmed by negative correlations between RGR of BA and BA in the second year. The reduction in growth rate has also been explained as a reduction of assimilation referred to as acclimation (Brown, 1991; Ceulemans & Mousseau, 1994). However, this explanation does not fit our observations since the ratio between AGR and corresponding values of leaf area during the second year was similar between both treatments, or even higher in FACE (data not shown). Since leaf area is one of the main factors determining tree productivity, it is possible to conclude that there was no reduction of growth in FACE compared to the control treatment in the second year. Moreover, no physiological acclimation emerged from leaf gas exchange measurements (Bernacchi et al., 2003).
With regard to this issue, we investigated whether the FACE-induced first-year's stimulation could have affected competition in these high-density stands. The coefficient of variation in this study decreased from the first to the second year, and increased again in the third year, which is usually explained by two-sided competition during the second year, and the onset of asymmetric competition in the third year (Weiner, 1985; Weiner & Thomas, 1986). This was confirmed by our observations of RGR of BA vs BA; small trees exhibit larger RGR as compared to large trees in the absence of asymmetric competition, and RGR is similar among tree size classes in an early stage of competition (Ford, 1976; Perry, 1985; Soares & Tomé, 1996). Our results are similar to those of eucalypt plantations that evolved from two-sided to asymmetric competition near the time of canopy closure (Toméet al., 1994; Soares & Tomé, 1996). Therefore, competition was probably mainly for nutrients and/or water in the second year, and was more asymmetric in the third year when light became limiting (Weiner, 1986; Soares & Tomé, 1996). Coefficients of size variation were observed to be lower in FACE; however, these observations may be caused by initial differences in variation.
In the third year we did not observe a significant variation of individual CI values between treatments, although the relationship between individual CI values and corresponding RGR of BA was significantly affected by the FACE treatment. Trees grown under elevated CO2 showed an increased RGR at the lowest competition levels and a decreased RGR at the highest competition levels compared to the control (Fig. 5). This cannot be explained by the mean bigger size of the FACE trees (Table 4) because this was evident for all the competition classes, and moreover RGR of BA had a positive correlation with size in the third year. Our finding agrees with McDonald et al. (2002) who observed the greater positive response to elevated CO2 for competitively advantaged trees.
The biomass present on branches should not be neglected since it made up a relevant part of the total production in particular for P. alba and P. nigra. The stimulation of sylleptic branch production during the first year was one of the most evident effects of the FACE treatment, except for P. alba. Increased branch production in elevated CO2 was also observed by Idso et al. (1991), Norby et al. (1996) and on poplars by Ceulemans et al. (1995) and Tognetti et al. (1999). This seems to represent a way for the plant to invest surplus carbohydrates derived from the increased photosynthetic rate and/or leaf area, which usually occurs under elevated CO2. The first year-stimulation of sylleptic branch production was confirmed by increased LAI values (Gielen et al., 2001), particularly for P. nigra. Increased LAI was due to an increased number of branches but also to increased leaf size in the FACE treatment (Ferris et al., 2001; Gielen et al., 2001). The number of sylleptic branches per tree could have been influenced by the larger size of trees in elevated CO2, although the length of the stem differed only slightly between treatments. Values of RSBPR confirmed the differences between treatments, even taking into account the differences in the length of the growing season (Tables 5 and 6). Stimulation of sylleptic branch production by FACE was lost in the second year in line with observations of other growth parameters.
The number of proleptic branches produced in the second year per unit of stem length was higher in control than in FACE, possibly because the larger space not occupied by sylleptic branches increased bud availability for generating proleptic branches. The number of proleptic branches produced during the second year was very similar for all species, unlike the number of proleptic branches produced in the third year. Most of them developed very quickly at the beginning of each growing season and died after a few months, even at canopy levels with high light levels. At the end of the second year, the component of BA represented by proleptic branches was particularly large in P. × euramericana, whereas it was relatively small in P. nigra and P. alba, which invested mainly in sylleptic branches (Fig. 6).
After 2 yr the BA of stem and of total branches were shown to be highly correlated with a constant ratio typical for each species (Fig. 6).
Dates of bud burst did not differ significantly among species or between treatments. Bud set, however, differed by almost 50 d between P. × euramericana and P. alba, although the longer growing season was not sufficient for P. alba to catch-up with the stem growth of the two other species. FACE treatment did not significantly influence date of bud set in any year; there were small differences for P. nigra in the first year and for P. × euramericana in the second. This finding disagreed with previous reports of a clear effect of elevated CO2 on bud phenology. For example Ceulemans et al. (1995) observed delayed bud burst in elevated CO2 for one genotype of Populus grown in open top chambers, whereas another genotype exhibited advanced bud set. The same behaviour was observed by Sigurdsson (2001) on P. trichocarpa. Picea sitchensis and Castanea sativa growing in pots in field chambers exhibited both delayed bud burst and advanced bud set (El Kohen et al., 1993; Murray et al., 1994). However, in different experiments on trees no effect of elevated CO2 on bud phenology was found (Johnsen & Seiler, 1996; Rey & Jarvis, 1997; Murray & Ceulemans, 1998).
In conclusion, size of three poplar species was significantly enhanced under CO2 enrichment in POPFACE after 3 yr of exposure. The stimulation was strong in the first year but did not occur in the second and third year. This result was confirmed by comparing tree growth in different treatments at the same size. Expected differences among species were observed in both treatments. Moreover during the third year, characterized by the presence of strong competition (asymmetric competition), growth resulted stimulated under FACE treatments for competitively advantaged trees but not for disadvantaged ones.