Eucalyptus globulus Labill., a globally significant plantation species, is grown commercially in a multiple rotation framework. Second and subsequent crops of E. globulus may be established either by allowing the cut stumps to resprout (commonly referred to as coppice) or by replanting a new crop of seedlings. Currently, long-term growth data comparing coppice and seedling productivity in second or later rotations in southern Australia is limited. The capacity to predict productivity using these tools is dependent on an understanding of the physiology of seedlings and coppice in response to light, water and nutrient supply. In this study, we compared the intrinsic (independent of the immediate environment) and native (dependent on the immediate environment) physiology of E. globulus coppice and second-generation seedlings during their early development in the field. Coppice not only grew more rapidly, but also used more water and drew on stored soil water to a depth of at least 4.5 m during the first 2 years of growth, whereas the seedlings only accessed the top 0.9 m of the soil profile. During the same period, there was no significant difference between coppice and seedlings in either their stomatal response to leaf-to-air vapour pressure difference (D) or intrinsic water-use efficiency; CO2- and light-saturated rates of photosynthesis were greater in seedlings than that in coppice as were the quantum yield of photosynthesis and total leaf chlorophyll content. Thus, at a leaf scale, seedlings are potentially more productive per unit leaf area than coppice during early development, but this is not realised under ambient conditions. The underlying cause of this inherent difference is discussed in the context of the allocation of resources to above- and below-ground organs during early development.
Eucalyptus globulus Labill. is a globally significant plantation species and more than 450,000 ha are now planted in Mediterranean environments across southern Australia (Parsons et al. 2006). Most of these plantations were established after 1997 and over the past 3 years the harvested area has increased exponentially. After harvest, plantation managers can choose to re-establish the plantation either with new seedlings or by allowing the cut stumps to resprout (coppice). Process-based models of plantation growth (e.g., CABALA – Battaglia et al. 2004) are one of the options for simulating potential yield. Unfortunately, such models have not been parameterized to account for genetic differences in planting stock within species, genetic gain or for second rotation coppice and seedlings. Realising the potential of these tools will therefore require a quantitative understanding of key processes that affect growth in each such system. For example, information on leaf-scale carbon gain and water use in coppice and seedlings would extend our understanding of the long-term trade-off between stand productivity and drought susceptibility in multiple rotation systems employing these growth forms. Growth form in this context describes the structural contrast between coppice and seedlings.
After an E. globulus stem is harvested, the root system is potentially able to supply large quantities of nutrients and water relative to the size of the developing shoots, because it has access to more soil resources and a large intrinsic store of carbon in the form of starch reserves (Fleck et al. 1996, Poorter and Nagel 2000). This imbalance between above- and below-ground biomass may also mean that starch reserves and current photosynthate can be allocated to above-ground biomass components instead of roots. Regardless of this mechanism, vigorous above-ground growth has been observed during early coppice development in E. globulus (Antonio et al. 2007). Seedlings, in contrast, allocate assimilated carbon to biomass components to bring the supply of nutrients and water into balance with demand so that under resource (water or nutrient) limiting conditions, assimilated carbon will be directed to the growth of roots and away from the above-ground components (Santantonio 1989, Battaglia and Sands 1997). In addition, the disproportional amount of live below-ground biomass in coppice could mean that coppice and seedlings experience different soil moistures because of contrast in the spatial distribution of roots. It is also possible that microclimatic differences, including atmospheric vapour pressure deficit (VPD), will occur between coppice and seedling canopies.
The physiological behaviour of E. globulus coppice is less well understood than seedlings, despite acknowledgement that seedlings and coppice will experience different conditions at the same site. Several studies conducted on northern hemisphere species have observed lower drought tolerance in seedlings compared to that in coppice (Oechel and Hastings 1983, Fleck et al. 1996, 1998, Williams et al. 1997, Clemente et al. 2005), but no studies have compared the intrinsic (independent of the immediate environment) physiological function of these growth forms. There have been many studies of the water relations, gas exchange and productivity of first rotation E. globulus seedlings or saplings (Pereira et al. 1987, White et al. 1996, 1999, Pita and Pardos 2001, Cernusak et al. 2003, Costa E Silva et al. 2004, Macfarlane et al. 2004a, O’Grady et al. 2008), but there is no exploration of these properties in coppiced E. globulus. Seedling-grown E. globulus has a well-defined stomatal response to VPD (White et al. 1999) and reduces stomatal aperture with exposure to cyclical soil water stress events (White et al. 2000). These types of physiological responses to environment are used in process-based growth models including CABALA (Battaglia et al. 2004), but are poorly understood for coppice despite the opportunities for coppice regrowth to be used in second and later rotation E. globulus plantations. Similarly, there has been no assessment of how the early imbalance in biomass allocation in the E. globulus coppice growth form affects intrinsic photosynthetic function. We hypothesize that the disproportionate amount of below-ground biomass in coppice will: (a) modify the manner by which below-ground resources are allocated to the newly developed above-ground coppice canopy and (b) create a different environment for coppice compared to seedlings. Either alone or collectively, these processes could affect both native (dependent on the immediate environment) and intrinsic (independent of the immediate environment) physiological function. The objectives of this study were to: (1) evaluate seedling and coppice native physiology under conditions of contrasting soil and atmospheric water deficit and (2) investigate differences in photosynthetic responses to CO2 and light, as indicators of intrinsic photosynthetic capacity, under ideal conditions.
Materials and methods
Plant material and field site
All measurements were taken on field-grown E. globulus at a second rotation plantation located in south-western Australia (34°15′ S and 115°31′ E). The climate of the region is Mediterranean, characterized by cool wet winters and warm dry summers (Gentilli 1972). The average maximum and minimum air temperatures are 21 and 11 °C, respectively. Air temperature and VPD can exceed 35 °C and 3.5 kPa, respectively, in summer (December to February), and the average annual rainfall for the site is 1040 mm. The soil has a sandy textured A horizon (0.5–1 m deep), over a clay B horizon, separated by a lateritic hardpan.
The first rotation was planted in 1996 and an experiment was established at the site in 1998 to explore the effects of nitrogen and thinning on leaf area development and response to drought. Plot dimensions were 40 × 40 m, with an internal measure plot of 20 × 22 m. The plantation was originally planted at 1250 stems ha−1 and was supplied with a basal fertilizer application at age 2 comprising phosphorus (100 kg ha−1 P), potassium (125 kg ha−1 K), magnesium (10 kg ha−1 MgSO4·7H2O), manganese (10 kg ha−1 MnSO4·H2O), zinc (10 kg ha−1 ZnSO4·7H2O) and copper (5 kg ha−1 CuSO4·5H2O). Well-fertilized plots, which were used for the current study (see the following paragraphs), were supplied with nitrogen (as urea) at 250 kg ha−1 N year−1. When harvested in January 2006 (summer), the trees had accessed the full depth of the B horizon, down to 6–10 m below the surface, and had depleted the plant available soil stored water down to this depth (White, unpublished data).
Following the harvest of the first rotation, six second rotation experimental plots (three coppice plots and three seedling plots) were established by either (a) replanting in July 2006 (winter) with 7-month-old nursery-raised seedlings at a density of 1250 stems ha−1 or (b) allowing first rotation stumps to coppice. Survival of the coppice was 82% giving a final density of 1030 stems ha−1 for coppice plots (note that a stem in this context represents a coppice stump comprising many, up to 50, individual emerging stems). To suppress coppicing in the seedling plots, the stumps were painted with glyphosate soon after harvesting. Nitrogen was applied at non-limiting rates in both treatments (250 kg N ha−1 year−1 from year 2). A foliar analysis during the first rotation indicated that the remaining, potentially limiting, macro- and micronutrients were within the normal range for eucalypts (data omitted).
At the time of planting, seedlings were ~ 0.25 m tall and coppice, which resprouted soon after harvest, was ~ 0.5 m tall. In December 2006, coppice and seedling heights were ~ 2.0 and 0.8 m, respectively, despite a similar age in both growth forms. This early vigour in resprout E. globulus is typical of that commonly observed (Blake 1983), hence comparisons made in this study are likely to be general for the early development of second or later rotation E. globulus plantations.
Specific leaf area, leaf area index and stand biomass
In December 2006, fully hydrated (Garnier et al. 2001) and expanded leaves were collected from each coppice and seedling plot (n = 5 leaves per tree from 20 coppice and 20 seedlings spread across the six measure plots). Leaf areas (m2) were measured with a leaf area meter (model Li-3100C Li-Cor Inc., Lincoln, NE) and leaves were then oven dried at 70 °C for 48 h and dry weight (kg) was determined. Specific leaf area (SLA) was calculated for each leaf as m2 of leaf surface area per kg of leaf dry weight.
Also in December 2006, the seedling stem conical volume (Mendham et al. 2003) was calculated from the stem diameter (m, at a distance of 0.05 m from the ground surface) and height (m) of each seedling was measured from each measure plot. Because coppice stumps comprised many individual stems, the total stem conical volume for each stump was calculated from a relationship derived between three-dimensional crown dimensions (width, length and height, m) and individual stem dimensions (height and diameter at 0.05 m) of a subset of 60 stumps (r2 = 0.81, P < 0.001). Stem conical volume has been shown to be a good correlate for total above-ground biomass components (Mendham et al. 2003).
Also in December 2006, total above-ground biomass (t dry weight ha−1) and leaf area index (LAI) in seedling and coppice plots were estimated using allometric functions derived from a harvest sample of coppice and seedlings comprising stems of a range of diameter and height (n = 20 seedling and n = 20 coppice stems originating from the six coppice and seedling plots). To develop allometric functions, we measured stem height, stem diameter (0.05 m from the soil surface for seedlings and 0.05 m from the point of emergence for coppice), total plant dry weight (kg) and leaf area (m2) from each harvested sample. For seedlings, the allometric functions linking stem conical volume to total plant dry weight and leaf area allowed us directly to calculate the above-ground biomass and LAI in the seedling plots. For coppice, we developed allometric functions linking crown volume to total plant dry weight and leaf area (derived from the 60 subsamples), allowing estimates of above-ground biomass and LAI in coppice plots.
Neutron moisture access tubes were installed within each plot (n = 2 for each) to a depth of 8 m. The tubes were made from polyvinyl chloride with an internal diameter of 0.04 m installed in 0.05 m holes back-filled with kaolin and cement (Prebble et al. 1981). Soil moisture was monitored at 0.2 m increments to 1.5 m from the soil surface, then at 0.5 m increments for the remaining soil profile using a neutron moisture meter (Model CPN 503, Campbell Pacific Nuclear, Concord, CA). Soil moisture was monitored through the inter-rotation period and during coppice and seedling establishment at intervals that were approximately the same as for native leaf water potential and gas exchange measurements (see below). Neutron moisture meter counts were converted to soil volumetric water content (θ*, %) using a calibration-based approach where soil samples, taken over the entire profile, were collected simultaneously with neutron moisture meter readings over a range of values of θ* (Hingston et al. 1998). Volumetric soil water for each vertical interval was multiplied by the respective increment (mm) and, by summing all intervals, an estimate of total soil water content (θ*tot, mm) was obtained for each sampling period. Soil water deficit (W, mm) was then calculated for the entire soil profile as the deviation from capacitance, which was determined at the end of winter when soil profiles are typically recharged by winter rainfall. In this study, September 2000 was used as the capacitance reference, this being a period early during the development of the first rotation and when seasonal rainfall recharge was high.
Native leaf water potential and gas exchange
All measures of native and inherent properties were taken on juvenile leaves, these being the dominant leaf type in both growth forms throughout the experiment. The diurnal course of leaf water potential (Ψleaf, MPa) was observed on six occasions between December 2006 and April 2008, encompassing the wet–dry seasonal oscillations typical of a Mediterranean-type habitat. On each measurement day, Ψleaf was measured at intervals from before dawn to just before dusk at ~ 19:00 h local standard time. One leaf was collected from each of the five plants selected at random from within each plot. The leaves were wrapped in plastic film and kept on ice until all leaves were collected. The leaf water potential was then measured using a Scholander-type pressure chamber (model PMS 1000, PMS Instrument Co., Corvallis, OR). Sources of error associated with measurement of water potential via the pressure chamber technique, as outlined in Ritchie and Hinkley (1975), were minimized. The maximum root-to-leaf hydrodynamic water potential gradient (ΔΨ, MPa) was calculated for each measurement day as the difference between the maximum and the minimum values observed.
On five of the six diurnal Ψleaf measurements, in situ photosynthetic assimilation rate (A, μmol CO2 m−2 leaf s−1), leaf transpiration rate (E, mmol H2O m−2 leaf s−1), stomatal conductance to water vapour (gw, mmol H2O m−2 leaf s−1), leaf temperature (T, °C) and the partial pressure of ambient and intercellular CO2 (Pa and Pi, respectively, μmol CO2 mol−1 air) were determined with a portable photosynthesis system (model Ciras-1, PP Systems, Hitchin, Herts, UK). These measurements were initiated after sunrise and coincided with sampling of leaves for leaf water potential. Briefly, in situ leaf gas exchange properties were taken from five randomly selected plants (n = 1 fully expanded, well-lit leaf per plant) from each of the measurement plots at a Pa of 350 μmol mol−1, ambient temperature, ambient light intensity and ambient relative humidity.
Steady-state leaf gas exchange
Between September and December 2006, when the soil was the wettest and therefore the likelihood of abscisic acid induced stomatal closure was low, steady-state leaf gas exchange measurements were made with a portable photosynthesis system (model Li-6400, Li-Cor Inc., Lincoln, NE). All measurements were initiated early in the morning (~ 08:00 local standard time) and were concluded within the natural daylight photoperiod. Measurements were made on one leaf from each of the four plants from one coppice and one seedling plot. The leaves were sealed in the instrument leaf chamber and allowed to reach a steady state under the following conditions: Pa = 350 μmol mol−1, T = 25 °C, photosynthetically active radiation (PAR) = 1500 μmol m−2 s−1 and leaf-to-air vapour pressure difference (D) = 1 kPa. The relationship between A and Pi was then obtained by manipulating Pa over the range 50–2000 μmol mol−1 and described empirically by fitting a rectangular hyperbola of the form (Olsson and Leverenz 1994):1 and an iterative least square fit approach was used to refine the model over the measured values of A and Pi.
The leaf was then allowed to return to a steady state at a Pa of 350 μmol mol−1 (no hysteresis was observed between the first and the second steady states) and the response of A to PAR was established by controlling PAR over the range 0–2000 μmol m−2 s−1. Photosynthetic light response curves were described by fitting a non-rectangular hyperbola of the form (Marshal and Biscoe, 1980):2 and the model was refined over the data range through an iterative least square fit approach.
Leaf chlorophyll and nitrogen content
In November 2006, one newly expanded leaf was collected from the upper canopy of five individuals from each plot. Taking care to avoid the midvein, 16.0 cm2 of the leaf material was subsampled from each leaf. Each subsample was ground (in the dark) in 80% acetone and acid-washed sand. The mixture was then centrifuged to remove cellular debris and the supernatant was made up to 2.5 ml. Absorbance was then immediately measured in an ultraviolet–visible spectrophotometer (model UV-1601, Shimadzu, Kyoto, Japan) at 652 nm. Total chlorophyll was determined according to Sestak et al. (1971) as
The residual leaf tissue from chlorophyll determination was analysed for nitrogen content. This tissue was oven dried at 70 °C then finely ground with a ball mill. Exactly 0.050 g of the ground material was then weighed and the nitrogen elemental composition (%) was determined using a mass spectrometer (model 20-20 IRMS, Europa, Crewe, UK).
Wherever appropriate, mean values were compared via t tests or one-way analyses of variance (ANOVAs) at the 0.05 level of significance (SPSS for Windows Version 13.0, SPSS Inc., Chicago, IL). Analyses of covariance were used to detect any differences between coppice and seedlings in correlations among various parameters (SPSS for Windows, Version 13.0) and the strength of the regression analyses was quantified as correlation coefficient.
Diurnal and seasonal patterns of water status for coppice and seedlings
During the measurement period, the pattern of rainfall was typical of a Mediterranean climate. For example, more than 77% of the annual total rainfall occurred between late autumn (May) and mid-spring (October) in 2007 (Figure 1A). Soil water deficits increased during summer for both coppice and seedlings, and this was associated with a decrease in predawn leaf water potential (Figure 1B and C).
Although the total soil water deficit (W) was higher under coppice than seedlings throughout the measurement period (Figure 1B), predawn leaf water potential was generally lower for seedlings than coppice and this difference was significant in March 2007, June 2007 and December 2007 (Figure 1C). The maximum root-to-leaf hydrodynamic pressure gradient (ΔΨ) also fluctuated seasonally, but there was no significant difference between coppice and seedlings (Figure 1D).
From December 2006, when the soil was close to being fully recharged, to March 2007 there was only 112.9 mm of rainfall (Figure 1A) and the coppice extracted water from the top 450 cm of soil while, in contrast, the seedlings only used water from the top 90 cm of the soil profile (Figure 2). Despite this disparity in effective rooting depth between coppice and seedlings, diurnal fluctuations in Ψleaf for each growth form followed a similar pattern throughout the study. In December 2006 (Figure 3A), there was no distinction in Ψleaf between coppice and seedlings at any time of the day. Leaf water potential, with the exception of the predawn phase, for coppice and seedlings was also similar in March 2007 (Figure 3B) except that Ψleaf for seedlings increased at around midday. At all other times the measurement of Ψleaf between December 2006 and March 2007 was similar throughout the day for coppice and seedlings.
Biomass, LAI and leaf characteristics of coppice and seedlings
In December 2006, when coppice and seedlings were ~ 11 months old, each coppice stump had an average of 20 stems compared to an individual seedling stem and this translated to around 200-fold more above-ground biomass and LAI for coppice compared to seedlings (Table 1). The SLA of juvenile leaves was, however, 44% higher in seedlings compared to coppice (Table 1).
|SLA (m2 kg−1)||12.33 ± 0.60||17.73 ± 0.24||ANOVA||< 0.01||15|
|LAI||0.7696 ± 0.2639||0.0045 ± 0.0009||t test||0.04||3|
|t dry weight (kg ha−1)||1.6008 ± 0.5582||0.0077 ± 0.0016||t test||0.05||3|
|Chlorophyll (mg m−2)||331.11 ± 10.66||453.21 ± 20.19||ANOVA||< 0.01||15|
|SLA (m2 kg−1)||12.33 ± 0.60||17.73 ± 0.24||ANOVA||< 0.01||15|
|LAI||0.7696 ± 0.2639||0.0045 ± 0.0009||t test||0.04||3|
|t dry weight (kg ha−1)||1.6008 ± 0.5582||0.0077 ± 0.0016||t test||0.05||3|
|Chlorophyll (mg m−2)||331.11 ± 10.66||453.21 ± 20.19||ANOVA||< 0.01||15|
Leaf-scale gas exchange characteristics of coppice and seedlings
On each diurnal sampling date, net carbon assimilation rate (A) was linearly related to the rate of transpiration (E) and there was no significant effect of growth form on either the slope or the intercept of these relationships at any time of the year (Figure 4). The slope of the relationship between A and E, which is analogous to leaf-scale water-use efficiency (WUE) integrated over a day, was found to vary significantly with sampling date. Leaf-scale WUE (expressed as A/E and derived as a slope) was not significantly correlated to Ψpd (P > 0.05), but an exponential decay function with daily average leaf-to-air vapour pressure difference (D) (Figure 5) was significant and it explained 86% of the variation in A/E. Also shown in Figure 5 is the short-term steady-state response of A/E to a step-wise change in D from 1.0 to 2.0 kPa under well-watered conditions and the trajectory of this response was similar to that observed under native conditions.
Leaf-scale photosynthetic rate was positively correlated with leaf conductance to water vapour (gw) on all measurement dates (Figure 6). Neither the y-intercept nor the slope differed significantly between coppice and seedlings (P > 0.05), hence data from both growth forms were pooled for analysis. The data in Figure 6 also compare the results in this study with the relationship published by Macfarlane et al. (2004b), and it indicates a similar relationship for droughted E. globulus trees. The slope or the intrinsic water-use efficiency (WUEi) was twofold greater in the study of Macfarlane et al. (2004b) (0.12 μmol mmol−1) than that for both growth forms in this study (0.06 μmol mmol−1).
Seedlings had a significantly higher CE (P = 0.03), CO2-saturated photosynthetic rate (Amax, P = 0.02) and foliar respiration rate (incorporating both light and dark respiratory processes) (r, P = 0.02) than coppice (Figure 7). Moreover, seedlings tended to have a higher light-saturated photosynthetic rate (Amax*) and apparent quantum yield (Φ) in response to light (Figure 8), but these differences were not significant. These differences in the inherent gas exchange characteristics between seedlings and coppice were associated with a significantly higher total leaf chlorophyll content (expressed on a per unit leaf area basis) in seedlings compared to coppice (t test, P < 0.05, Table 1). Seedlings, on average, had 122.11 mg m−2 (37%) more total leaf chlorophyll than coppice. Higher total leaf chlorophyll contents in seedlings were associated with higher leaf nitrogen and the correlation between leaf chlorophyll and nitrogen content across growth forms could be described by a linear model (Figure 9, P < 0.05).
Different intrinsic gas exchange properties and chlorophyll and nitrogen content between coppice and seedlings suggested a larger resource investment by seedlings at the leaf scale in structures and enzyme systems associated with photosynthetic capacity. However, this did not translate to differences between coppice and seedlings in either leaf water potential or observed rates of gas exchange in the field and was more than offset by the higher early biomass and LAI of coppice compared to that of seedlings. However, the more rapid growth in coppice was also associated with a faster depletion of soil water stores and this may lead to earlier competition for resources and convergence of growth rates later in the rotation. Our hypotheses were that the disproportionate amount of below-ground biomass in coppice would: (a) modify the manner by which below-ground resources were allocated to the newly developed above-ground coppice canopy and (b) create a different environment for coppice compared to seedlings. The most likely theoretical explanation for the different intrinsic properties of coppice and seedlings is the manipulation in root-to-shoot ratio and the reallocation of resources brought about by harvesting of the previous rotation. Thus, our results lead us to accept hypothesis (a), but reject hypothesis (b).
Leaf and soil properties under ambient conditions
Each coppice stump supported a higher leaf area and total above-ground biomass than the seedling stems, and this was reflected in a significantly higher LAI of coppice compared to seedlings. Coppice also extracted water deeper from the soil profile and generated a much larger soil water deficit than that which was observed for seedlings. Put together, these observations confirm that coppice retains a substantial amount of live below-ground biomass from the original tree and that this temporary imbalance in biomass allocation during early development provides for a greater supply of soil resources to above-ground biomass components compared to seedlings.
Contrary to our original hypothesis, leaf water potential, gas exchange and WUEi did not differ markedly between growth forms during the first 2 years of growth. Under natural conditions, plants typically experience oscillation in irradiance, temperature, VPD and soil water potential at various time scales. Stomatal guard cells interpret this oscillation and manipulate stomatal aperture to optimize gas exchange such that carbon uptake is maximized for a given water use (Cowan 1977). Under water limitation, stomata typically close. Hence, it might be anticipated that the greater soil water deficit under coppice would be associated with a lower Ψpd and depressed rates of leaf gas exchange. The similarity in these traits suggests that the supply of plant available soil water (PASW) was balanced with demand for both growth forms and that, in this case, total soil water deficit was not a good measure of underlying water stress. Some process-based models of plantation growth, including CABALA (Battaglia et al. 2004), use relative soil water deficit to predict water potential and in turn leaf conductance and net carbon assimilation. These results highlight the importance of expressing PASW as a function of effective rooting depth or root density in second rotation E. globulus plantations.
The persistence of roots from the previous rotation in coppice trees has been less studied. The depletion of deep soil moisture layers underlying coppice in this study, to our knowledge, is the first evidence that at least some residual deep roots are retained and are actively involved in water uptake in second rotation E. globulus systems. Wildy et al. (2004) made a similar observation for coppice Eucalyptus kochii Maiden and Blakely subsp. Plenissima Gardener (Brooker) growing in tree belts. Coppice are often characterized by vigorous growth, at least some of which must be attributable to the presence of the existing root system. However, the respiratory cost of such a system must prohibit retention of root biomass over and above that required for resource capture. For example, fine roots could be lost soon after harvest due to diminished resource demand from the canopy. This was shown by Wildy and Pate (2002) in the study on E. kochii. Further investigation on root biomass turnover in multiple rotation forestry systems would be valuable to complement our current understanding.
In our study, mean daily integrated WUE varied from 1.81 to 7.27 μmol CO2 mmol−1 H2O depending on season. This difference was driven by a mean 36% increase in E and a mean 34% decrease in A during the first growing season. Instantaneous leaf-scale WUE (A/E) in glasshouse grown E. globulus seedlings has been reported from 4.71 μmol CO2 mmol−1 H2O in low N status plants to 17.60 μmol CO2 mmol−1 H2O in well-watered and fertilized plants (Sheriff 1992). White et al. (2000) showed that much of the variation in E in well-watered E. globulus trees can be explained by the positive relationship between E and VPD. Our observation that A/E is related to D provides additional insight into the drivers for carbon and water exchange in this species. It remains to be seen, however, whether this strong coupling is consistent in both growth forms under a severely depleted soil moisture profile where the root-to-shoot water stress signal may override atmospheric drivers for gas exchange.
The WUEi, calculated as the slope of the relationship between A and gw, did not differ between growth forms and was about half of that reported for 4-year-old trees of the same species in the study of Macfarlane et al. (2004b). This is probably due to the much greater level of water stress, induced by lower mean annual rainfall, reported by Macfarlane et al. (2004b); the average value for Ψpd in Macfarlane et al. (2004b) was less than −1.8 MPa, whereas in this study the minimum value for Ψpd was −0.32 ± 0.02 MPa for coppice and −0.38 ± 0.03 MPa for seedlings. This suggests that gw is more sensitive to soil moisture than A and therefore the impact of drought on photosynthesis is likely to be mediated by stomatal limitation. Despite this, E. globulus is anisohydric, i.e., the combination of soil and atmospheric water deficit results in an increase in the maximum root-to-leaf hydrodynamic pressure gradient (Tardieu and Simonneau 1998, Franks et al. 2007). This functional attribute implies that, despite modification to stomatal aperture with the onset of soil or atmospheric water deficit, this species cannot necessarily regulate water potential within a range to sustain physiological processes. This current study did not capture native conditions that would test this possibility. Hence, we were unable to assess the relative advantage of coppice versus seedlings under such conditions.
Inherent photosynthetic capacity
The capacity to assimilate CO2 was higher in seedling leaves than in coppice leaves, as indicated by a higher maximum rate of CO2-saturated photosynthesis. However, under ambient CO2 concentrations, the observed difference between the growth forms was not significant. The greater photosynthetic capacity of seedlings at high Pi suggests that the regeneration of RuBP in the C3 cycle is faster (von Caemmerer and Farquhar 1981) in seedlings compared to that in coppice. The photochemical partial processes contributing to the regeneration of RuBP are light interception, energy funnelling, phosphorylation of ADP and reduction of NADP. The aggregate of these processes was characterized by comparing photosynthesis as a function of irradiance. The estimated quantum yield of photosynthesis was lower in coppice compared to seedlings, at 0.039 μmol CO2 μmol−1 versus 0.045 μmol CO2 μmol−1 PAR, respectively. The quantum yield of seedlings in this study was similar to that found in 4-year-old E. globulus (0.05; Battaglia and Sands 1997), and is similar to the value reported more generally for C3 species by Ehleringer and Björkman (1977). The lower quantum yield in coppice resprouts compared to seedlings was associated with a lower light-saturated rate of photosynthesis in coppice (17.40 μmol m−2 s−1) compared to seedlings (20.32 μmol m−2 s−1). Battaglia et al. (1996), also using the model of Marshal and Biscoe (1980), recorded light-saturated rates of photosynthesis in E. globulus in the order of 14.35–15.70 μmol m−2 s−1.
In this study, coppice had less total leaf chlorophyll than seedlings. The range of 330 mg m−2 (coppice) to 452 mg m−2 (seedlings) was within the range of total chlorophyll values of 220–600 mg m−2 found by Pinkard et al. (2006) in a range of glasshouse- and field-grown E. globulus seedlings. Chlorophyll pigments intercept sunlight and represent the start of the photosynthetic process. In this way, photosynthesis at light saturation and the quantum yield of photosynthesis are linked to the quantity of chlorophyll in leaf chloroplasts. Combined with the observed response of photosynthesis to Pi, it would appear that the protein available to leaves for the construction of chlorophyll pigments and electron transport capacity is lower in coppice compared to seedlings. The linkage between leaf nitrogen, a fundamental component for protein synthesis, and total leaf chlorophyll implied in this study provides insight into the contrasting mechanisms of resource allocation in coppice and seedling forms of E. globulus during early development.
Resh et al. (2003) estimated that E. globulus allocates ~ 1 kg dry weight m−2 soil surface year−1 into coarse roots (roots > 2 mm in diameter). Although the conditions of our study differed from that of the study of Resh et al. (2003), it is conceivable that coppice accumulated around 10 kg roots m−2 soil surface during the first rotation. This imbalance in biomass components during early development of the second rotation increased the amount of resources available to the young coppice canopy. With potentially more reserves for early canopy development, it could be expected that the leaf level concentration of resources would be elevated in coppice, but this was not the case in this study. Both leaf chlorophyll content and leaf nitrogen content were lower in the coppice growth form, compared to seedlings, resulting in a lower CE and CO2-saturated rate of photosynthesis. In coppice-grown E. globulus, the existing root system, retained after the original tree was harvested and during the development of regenerative buds, represents a large pool of carbohydrate. Theoretically, this carbohydrate pool can be utilized to maintain the living root system, and contribute to rapid early growth of support structures. Coppice biomass rapidly increases under favourable conditions (Schlesinger and Gill 1980, Kruger and Reich 1993, Fleck et al. 1998), resulting in the production of multiple stems, which are the support structures for many leaves (evidenced as a greater LAI than seedlings). This production of above-ground support structures to deploy leaves for photosynthesis is a rapid-start feed-forward system that contributed to a greater above-ground biomass (quantified as dry weight) than seedlings. Coppice also have the advantage of not requiring substantial development of a below-ground support structure to deploy fine roots. In contrast, seedlings are initially dependent on the carbohydrate reserves of the seed. With this they grow a few leaves, and all functional and supporting structures above and below ground must be grown thereafter using the current photosynthate. In this way, the accumulation of leaf area, in balance with root area, is necessarily slower in seedlings than that in coppice. Therefore, seedling development can, in the case of this study, be considered primarily limited by a capacity to physically deploy leaves and fine roots on support structures.
As a result of contrasts in the construction of leaf support structures, the resource investment in leaves of coppice and seedlings is likely to be different. Seedlings can only support a few leaves, and are likely, therefore, to invest all of their nutritional (mostly nitrogen) resources into them to optimize carbon assimilation per unit leaf area. Our observation of a high SLA in seedlings is consistent with the general hypothesis that a greater concentration of nitrogen in leaves translates to leaf lamina expansion. Coppice with the advantage of rapid construction of leaf support structures can deploy many leaves and accordingly distribute nutritional resources more diffusely across the leafy canopy. We hypothesize that this is the most likely underlying reason for the different inherent photosynthetic properties and leaf-area-scaled total chlorophyll content across the growth forms of this study.
The similarity in native response to environment across growth forms, exemplified by analogous WUEi and leaf water potential, is suggestive that the contrast in below-ground root distributions did not alter the growth conditions of E. globulus in this study. Instead, the substantial contrast in the inherent properties of coppice and seedlings supports the idea that a large root-to-shoot ratio and storage reserves are the drivers for early vigorous growth in E. globulus coppice. It remains to be seen whether the initial advantage of coppice, i.e., the early expression of above- and below-ground support structures and access to a significant carbohydrate store, will persist over the second rotation period.
The authors thank the Co-operative Research Centre for Forestry.
We thank Scott Walker, Tammi Short and Jessie Rutter for technical assistance and Chris Beadle, Jenny Carter and Bernie Dell for helpful comments. We are grateful to Hansol P.I. and Great Southern Plantations for site access.