Introduction

Invasive plants can threaten biodiversity (Dukes and Mooney 1999) and potentially affect both the structure and function of ecosystems (Vitousek and Walker 1989; Mack et al. 2000). Biological invasion is estimated to cause approximately $137 billion in global losses every year (Pimentel et al. 2000). Therefore, identifying the physiological and environmental factors that influence invasions by exotic plants is very important for improving prediction and control of potentially invasive species.

One aspect of the environment which is of obvious interest is global climate change. Since the late 1950s, atmospheric CO2 has risen from 280 to 379 μmol mol−1 in 2005 and currently ongoing increases at 1.9 μmol mol−1 per year on average (IPCC 2007). The increasing CO2 concentration has been suggested to play a substantial role in some invasions, such as cheatgrass (Bromus tectorum; Smith et al. 2000; Salo 2005), kudzu (Pueraria lobata; Weltzin et al. 2003; Forseth and Innis 2004) and Japanese honeysuckle (Lonicera japonica; Sasek and Strain 1991). In comparison with native species, many invasive species have shown more responsive to elevated CO2 and benefit from it (Ziska and Bunce 1997; Smith et al. 2000; Ziska 2003). Some studies suggested the differences in plant architecture, construction cost and reproductive allocation between invasive and native species were the possible mechanisms to explain why these invasive plants increase at the expense of existing native plants in elevated CO2 (Vilà et al. 2006). However, it was still not well understood to date. Since plant productivity involves both the assimilation and expenditure of energy (Griffin 1994), examination of energetic properties could provide mechanistic insight into growth responses to increased atmospheric CO2 concentration (Nagel et al. 2005).

Photosynthesis performs an important function of energy supply by converting solar energy into carbohydrate molecules. Higher photosynthetic rate has been recognized as a significant characteristic for invasive species (Pattison et al. 1998; Baruch and Goldstein 1999; Durand and Goldstein 2001; Deng et al. 2004). Conversely, biomass construction cost (CC) is a quantifiable index of the energy demand for biomass production. It has been defined as the amount of glucose required to provide carbon skeletons, reductant and energy for the synthesis of organic compounds (Williams et al. 1987). Construction cost is always related to resource-use efficiency (Williams et al. 1987; Griffin 1994) and relative growth rate (Lambers and Poorter 1992; Poorter and Villar 1997). Therefore, low construction cost is expected to provide invaders a universal growth advantage. Some earlier studies found the lower leaf CC for some invasive vs. native species (Baruch and Gómez 1996; Baruch and Goldstein 1999; Nagel and Griffin 2001; McDowell 2002; Song et al. 2007), suggesting that invasive species might require less energy for biomass construction and allow saved energy to invest in other competitive strategies. In addition, the photosynthetic energy-use efficiency (PEUE) has been used to evaluate the ratio of energetic gains to costs (Nagel et al. 2005), reflecting the specific energy-use strategy of plant species. A strong correlation between PEUE and biomass accumulation was reported in the recent study of Nagel et al. (2005). These results suggest energy properties are influential to competitive success of invasive species.

The increasing CO2 concentration may stimulate photosynthetic activity (Bowes 1993; Saralabai et al. 1997) and alter energy assimilation, investment and allocation (Nagel et al. 2005). The different responses of energetic properties to CO2 enrichment are likely to affect plant growth and interspecific competitions. A recent study found that CO2 enrichment reduced the energetic cost of biomass construction in an invasive desert grass (Bromus madritensis ssp. rubens) compared with its co-occurring native species, and thus increased invasive species dominance (Nagel et al. 2004). Although some researches examined the effect of increasing CO2 on some separate aspects of energy-use properties such as photosynthetic activity or construction cost, few studies were involved the whole energy-use process (including energy assimilation, investment and allocation) and the differences between invasive and native species. Further studies would be useful in exploring the physiological mechanism of invasive plant success at elevated CO2 concentration.

In this study, we compared the effect of elevated CO2 on energetic properties of alien invasive species (Wedelia trilobata (L.) Hitchc.) with its native congener (Wedelia chinensis (Osbeck.) Merr.). Wedelia trilobata is a creeping herb native to the tropics of Central America, which has invaded many areas of tropics and subtropics (Thaman 1999). It has been listed as one of the 100 world’s worst invasive alien species (IUCN 2001). In the 1970s, W. trilobata was introduced to China as an ornamental groundcover, and rapidly escaped from gardens to roadsides and plantations (Li and Xie 2002). Once W. trilobata establishes in plantations, it will overgrow plants and develop into a thick ground cover, crowding out or preventing regeneration of other species. It has become recognized as a serious weed in southern China (Wu et al. 2005b). Wedelia chinensis is the native congener of W. trilobata in China. They appear to be similar species, as they share similar morphologies and life histories. In contrast to W. trilobata, W. chinensis grows slowly and has not been found harmful to native plants or habitats in China. Studies on W. trilobata were focused on its reproductive biology (Wu et al. 2005a), gas exchange characteristics (Wu and Hu 2004; Liu and Li 2005) and allelopathic effects (Vieira et al. 2001; Nie et al. 2004; Zhang et al. 2004), and there is little information about the effect of elevated CO2 concentration on this invasive species.

The main objectives of this study were to determine (1) how these two species responded to CO2 enrichment in growth and energy-use process including energy assimilation, investment and allocation; (2) whether the specific energy-use strategy of invasive species under elevated CO2 facilitated its growth. These results could provide valuable insight into the prediction of how these plants could change in the future, and the proximal mechanisms underlying the success of invasive species.

Materials and methods

Experimental design

Plant materials were grown from stem cuttings in about 5 cm length collected from Guangzhou, China. After 2–3 weeks of propagation, uniform seedlings (approximately 10 cm tall) were selected and transplanted three into each pot (23 cm depth, 25 cm diameter). The initial biomass was measured on seedlings of the same size. The pots were filled with equal proportions of river sand and pool mud. The plants were watered each morning to saturation with deionized water. Additionally, all pots were watered with Hoagland’s solution once every other week. All plants were grown in four controlled environment growth chambers (S10H, Conviron, Canada). The chamber was 1.23 m length × 0.76 m breadth × 1.5 m height. The CO2 concentration was maintained at 400 ± 50 μmol mol−1 (ambient CO2 treatment) in two chambers and 700 ± 50 μmol mol−1 (elevated CO2 treatment) in two others. For each species, six replicate pots were randomly assigned to one of the two CO2 treatments. In all chambers, air temperature was maintained at 28/20°C (day/night) with a photoperiod of 16 h; relative humidity was ca. 50% and photosynthetic photon flux density (PPFD) was ca. 400 μmol m−2 s−1 during the day. All pots were moved biweekly across the chambers, and weekly within the chambers, to minimize any effect of individual chambers or locations. The CO2 treatment lasted from 8th May to 18th August (103 days).

Growth characteristics

All plant materials were harvested in twice, respectively, on 13th June and 18th August. Plants in one pot were considered as one mixed sample and sorted into leaves, stems and roots. The leaf area was determined with a leaf area meter (Li-3100A, Li-Cor, USA) as soon as possible. The total stem length was also measured. After drying at 70°C for 72 h, leaf, stem and root materials were weighed, respectively, determining the total biomass, the percent of total biomass allocation to various plant components (i.e. leaf, stem and root) and leaf mass per unit area (LMA). The relative growth rate (RGR) was calculated according to: RGR = (ln W 2−ln W 1)/t, where W 1 was the average individual biomass at the first harvest and W 2 was the final individual biomass of each species, t was the time between two harvests in days.

Energetic properties

The energy gain throughout the growth period was evaluated as the total carbon fixation per unit leaf area via measured photosynthetic rate (P n). During the growth period, P n was measured at three different times (respectively, on 13th June, 14th July and 5th August) with a programmable, open-flow gas exchange system (Li-6400, Li-Cor, USA). On each measurement date, the diurnal pattern of P n was measured on five fully expanded youngest leaves per species from 08:00 to 18:00 at 2-h intervals. Leaf chamber environmental conditions were set to mimic the chamber environmental conditions (28°C and 400 μmol m−2 s−1 PPFD). The photosynthetic rate was little variable throughout the diurnal period because of the constant chamber environment in the present experiment. Therefore, the daily carbon fixation was calculated by scaling the daily average photosynthesis (P day) for 16 h photoperiod. The total carbon gain per unit leaf area (A total, mol CO2 m−2) was calculated by summing daily carbon gain across the different measurement dates throughout the growth period.

Carbon (C) and nitrogen (N) concentrations of various plant components were determined with an elemental analyzer (Vario, Elmentar, Germany). The total biomass [C] and [N] were calculated by a weighted average of various plant components [C] and [N]. The photosynthetic nitrogen-use efficiency (PNUE) was calculated as the ratio of A total to the total biomass [N] based on unit area. Ash content (Ash) was measured by burning pre-weighed samples in a 500°C muffle furnace (Vulcan A-550, Vulcan, UK) for 6 h and weighing the remaining mass. To obtain ash-free heat of combustion (∆Hc), three 0.5 g pellets from each sample were pressed and combusted using a calorimeter (HWR-15E, Shanghai, China) with correction for nitric acid formation and ignition wire. The ∆Hc values obtained for the triplicate pellets of each sample were then averaged. The construction cost of leaf, stem and root biomass (CC, equivalent to grams glucose per gram dry mass) was calculated according to the methods described by Williams et al. (1987) as: CC = [(0.06968∆Hc − 0.065) (1 − Ash) + 7.5(k N/14.0067)]/E G, where k was the oxidation state of the N substrate (+5 for nitrate or −3 for ammonium). The relative proportion of these forms of soils N in each pot was measured with flow injection analysis (QC8000, Lachat, USA). We estimated CC as a weighted average calculated with each value of k. EG was the growth efficiency and estimated across species to be 0.87 (Penning de Vries et al. 1974). The total biomass CC was calculated by a weighted average of various plant components CC. To calculate leaf CCarea (equivalent to grams glucose per square meter), leaf CC was multiplied by LMA.

Calculations of the total construction cost (CCtotal) of leaf, stem, and root biomass were made by multiplying CC of each plant component with its biomass. The percent of total energy allocated to various plant components was ascertained by dividing CCtotal of a given plant component by the sum of CCtotal of all components and then multiplying this value by 100 for each species. Photosynthetic energy-use efficiency (PEUE) was calculated as the ratio of A total to leaf CCarea according to Nagel et al. (2005).

Statistical analysis

All statistical tests were performed using SPSS 11.5 software (SPSS Inc., USA). A two-way analysis of variance (ANOVA) was performed to evaluate the main effects and interactive effect of CO2 treatment and species on measured variables. Treatment means were compared to determine if means of the dependent variable were significant at the 0.05 probability level with the least significant difference (LSD) posthoc analysis. Linear regression analysis was used to determine the degree of association between energetic properties and growth characteristics (i.e. biomass and RGR).

Results

RGR was affected significantly by species (P = 0.001). W. trilobata grown in elevated [CO2] exhibited a significant RGR increase from 12.89 to 19.49 mg d−1 (P = 0.014). However, the RGR of W. chinensis did not change with CO2 enrichment (9.03 vs. 8.82 mg d−1, P = 0.923; Fig. 1a). The [CO2]-induced changes in RGR were associated with a consistent tendency in the total biomass. The total biomass was affected significantly by CO2 treatment (P = 0.001), species (P < 0.001) and interactive effect (P = 0.023). The total biomass of W. trilobata exhibited a significant increase at elevated [CO2] (3.77 vs. 5.99 g, P = 0.001), whereas that of W. chinensis did not differ significantly between CO2 treatments (1.33 vs. 1.97 g, P = 0.147; Fig. 1b).

Fig. 1
figure 1

Mean relative growth rate (RGR, mg d−1) and total biomass (g individual−1) of invasive Wedelia trilobata and native Wedelia chinensis at ambient (400 μmol mol−1, open bars) and elevated [CO2] (700 μmol mol−1, filled bars). Error bars represent 1SE. Means with a common letter do not differ from each other based on LSD posthoc analysis at P = 0.05 level

Full size image

The increasing CO2 concentration had a significant effect on energy assimilation (P = 0.001). W. trilobata and W. chinensis showed significant increases in A total by 71.8% (P = 0.004) and 55.1% (P = 0.016), respectively, (Fig. 2a). Total biomass CC was affected significantly by species (P < 0.001) and interaction of species and CO2 treatment (P = 0.031). W. trilobata grown in elevated [CO2] exhibited a slight decrease in total biomass CC (1.16 vs. 1.15 g glucose g−1, P = 0.414). However, a significant increase was observed in W. chinensis (1.31 vs. 1.35 g glucose g−1, P = 0.022; Fig. 2b). At both ambient and elevated [CO2], W. trilobata showed significantly lower CC than its native congener (P < 0.001 for both treatments). Mean PEUE was affected significantly by CO2 treatment (P < 0.001), species (P = 0.001) and interactive effect (P = 0.004). W. trilobata and W. chinensis showed significant increases in PEUE at elevated [CO2] by 85.3 and 43.8% (Fig. 2c). Although mean PEUE was not significantly different between these two species at ambient [CO2] (0.34 vs. 0.32 mol CO2 g−1 glucose, P = 0.415), it was significantly higher for W. trilobata than for W. chinensis at elevated [CO2] (0.63 vs. 0.46 mol CO2 g−1 glucose, P < 0.001).

Fig. 2
figure 2

Mean total carbon fixation (A total, mol CO2 m−2), construction cost (CC, g glucose g−1) and photosynthetic energy-use efficiency (PEUE, mol CO2 g−1 glucose) of invasive Wedelia trilobata and native Wedelia chinensis at ambient (400 μmol mol−1, open bars) and elevated [CO2] (700 μmol mol−1, filled bars). Error bars represent 1SE. Means with a common letter do not differ from each other based on LSD posthoc analysis at P = 0.05 level

Full size image

Mean [C] was affected significantly by CO2 treatment (P = 0.023) and species (P < 0.001). Both W. trilobata and W. chinensis exhibited a slight increase in [C] at elevated [CO2] (P = 0.069 and P = 0.087; Table 1). In addition, W. chinensis kept significantly greater [C] than that of W. trilobata at either ambient or elevated [CO2] (P = 0.004 and P = 0.002). Mean [N] was affected significantly by CO2 treatment (P = 0.006) and species (P < 0.001). W. trilobata showed a significant [N] decrease at elevated CO2 (P = 0.008), while W. chinensis did not differ significantly with CO2 enrichment (P = 0.106; Table 1). This resulted in a significant difference in C:N between them at elevated [CO2] (P = 0.002), although there was no significant difference at ambient [CO2] (P = 0.095; Table 1). Meanwhile, CO2 enrichment significantly increased the PNUE of W. trilobata and W. chinensis by 108.3% (P < 0.001) and 60% (P < 0.001; Table 1). Ash content was affected significantly by CO2 treatment (P = 0.007) and species (P < 0.001). Elevated CO2 significantly decreased the ash content in both W. trilobata and W. chinensis (P = 0.024 and P = 0.045), and W. trilobata kept significantly higher ash content than that of W. chinensis at either ambient or elevated [CO2] (P = 0.001 and P < 0.001; Table 1). The mean value of ∆Hc was affected significantly by species (P < 0.001). W. trilobata showed significantly lower ∆Hc than that of W. chinensis at both CO2 treatments (P = 0.004 and P < 0.001; Table 1).

Table 1 Carbon (C) and nitrogen (N) concentrations, ratio of carbon to nitrogen (C:N), photosynthetic nitrogen-use efficiency (PNUE), ash content (Ash) and ash-free heat of combustion (∆Hc) of invasive Wedelia trilobata and native Wedelia chinensis at ambient (400 μmol mol−1) and elevated [CO2] (700 μmol mol−1)
Full size table

The increasing CO2 concentration had different influences on the energy allocation pattern of these two species (Table 2). In W. trilobata, elevated CO2 significantly decreased the percent of CCtotal allocation to leaves (P = 0.009), but increased energy allocation to stems (P = 0.017), and the percent of CCtotal allocation to roots did not differ with CO2 enrichment (P = 0.431). Consequently, W. trilobata grown in elevated CO2 significantly extended its stem length by 69.0% (231.11 vs. 390.67 cm, P = 0.017). In contrast, the energy allocation to each plant component of W. chinensis did not differ significantly with CO2 enrichment (P = 0.996 for leaves, P = 0.543 for stems and P = 0.263 for roots). Comparison of these two species within ambient CO2 treatment, there was no significant difference in the percent of CCtotal allocation to leaves (P = 0.840) or stems (P = 0.083), while the energy allocation to roots of W. trilobata was significantly lower than that of W. chinensis (P = 0.019). Within elevated CO2 treatment, W. trilobata exhibited significantly lower energy allocation to leaves (P = 0.044) and roots (P = 0.024), but greater allocation to stems than those for W. chinensis (P = 0.024).

Table 2 The percentage of total construction cost allocation to various plant components (%) of invasive Wedelia trilobata and native Wedelia chinensis at ambient (400 μmol mol−1) and elevated [CO2] (700 μmol mol−1)
Full size table

When all cases from both CO2 treatments were considered across species, there were no significant correlations between A total and either total biomass (r 2 = 0.301, P = 0.065) or RGR (r 2 = 0.326, P = 0.053; Fig. 3a, b). However, total biomass CC was negatively correlated with total biomass (r 2 = 0.766, P < 0.001) and RGR (r 2 = 0.662, P = 0.001), respectively, (Fig. 3c, d). Additionally, there were positive associations between PEUE and either total biomass (r 2 = 0.534, P = 0.007) or RGR (r 2 = 0.458, P = 0.016; Fig. 3e, f).

Fig. 3
figure 3

Relationships of total biomass or relative growth rate (RGR) of invasive Wedelia trilobata (triangles) and native Wedelia chinensis (squares) at ambient (400 μmol mol−1, open) and elevated [CO2] (700 μmol mol−1, filled) with total carbon fixation (A total), construction cost (CC) and photosynthetic energy-use efficiency (PEUE). The solid regression lines represent significant correlations between theses factors and the dashed lines represent insignificant correlations between theses factors

Full size image

Discussion

Increasing CO2 generally stimulates plant growth rates and biomass accumulation (Curtis and Wang 1998; Poorter and Navas 2003). However, the relative stimulation varies greatly between plant species (Poorter 1993). In the current study, the invasive W. trilobata showed greater growth stimulation (58.9%) by elevated CO2 than its indigenous congener (48.1%; Fig. 1b). The stronger growth response of invasive species to elevated CO2 was expected to increase invasive success in the future (Ziska 2003). The invasive Bromus tectorum exhibited significantly greater biomass and seed rain at elevated CO2 than several species of native annuals, which enhanced the success of this invasive species in western North America (Smith et al. 2000). Similar results were reported for Japanese honeysuckle (Lonicera japonica; Sasek and Strain 1991) and 6 C4 weeds (Ziska and Bunce 1997). The present results suggest that the invasive W. trilobata is likely to become more troublesome in a future high CO2 world. Although the present experiment was conducted in the environment-controlled chambers which were lack of competitions and resource availability limitation in natural communities, the role of increasing CO2 concentration as a possible factor favoring invasiveness of alien species should deserve additional consideration. Further field experiments are necessary to develop this prediction.

Since plant growth is always related to energy-use process (Griffin 1994; Nagel et al. 2004), the greater growth stimulation in invasive W. trilobata was expected to be related to its specific energy-use strategy at elevated CO2 concentration. Actually, the invasive W. trilobata indeed showed a different response pattern to CO2 enrichment in energetic properties relative to native W. chinensis.

Our earlier studies found that CO2 enrichment stimulated the average photosynthesis of W. trilobata and W. chinensis by 75.0 and 36.5%, respectively, (Song et al. 2009), which resulted in a greater A total increase in invasive W. trilobata than that of W. chinensis (Fig. 2a). This provided the invader an advantage in energy gain over its native congener with increasing CO2. Nagel et al. (2005) had found A total was positively associated with biomass production. In the present study, RGR and biomass were shown enhanced with the increasing A total (Fig. 3a, b), although the positive correlations were not significant partly due to the paucity of cases.

Studies examining the effect of elevated CO2 on construction cost reported a small mean change of 3–4% (Poorter et al. 1997; Wullschleger et al. 1997), and the high-CO2 effect on construction cost differed depending on species. The present results indicated opposite effects of CO2 on these two species (Fig. 2b). The change of construction cost was realized to be related to the altered chemical composition induced by CO2 enrichment (Poorter et al. 1997). Nitrogen typically is contained in protein and amino acid, which are relatively expensive to synthesize (Penning de Vries et al. 1974). It exhibits a positive correlation with CC (Griffin et al. 1996; Nagel and Griffin 2001). For W. trilobata, CO2 enrichment caused a significant decrease in [N] (Table 1), and thus reduced its energetic cost of biomass construction. In contrast, CO2 enrichment did not significantly influence [N] of W. chinensis (Table 1), and thus did not decrease its CC. Additionally, the root CC of W. chinensis increased with CO2 enrichment (1.25 vs. 1.60 g glucose g−1, unshown data), which led to the increase of its total biomass CC. Previous studies had reported that the CC of maize shoots and loblolly pine fine-root increased with CO2 enrichment (Amthor et al. 1994; George et al. 2003). The increase in root construction cost by elevated CO2 was suggested to be related to increases in the lignin concentration of fine roots (George et al. 2003). In terms of glucose equivalents, lignin is one of the most expensive compounds to produce (Penning de Vries et al. 1974) and elevated CO2 has been found to increase the lignin concentration of roots (Booker et al. 2000).

Construction cost was also reported positively correlated with ∆Hc and negatively correlated with ash content (Villar and Merino 2001). For invasive W. trilobata, it always exhibited lower construction cost at both ambient and elevated [CO2], mainly due to its lower [N] and ∆Hc, and higher ash content relative to W. chinensis (Table 1). Several similar cases have already been mentioned (Baruch and Gómez 1996; Baruch and Goldstein 1999; Nagel and Griffin 2001; Song et al. 2007). In the present study, CC was shown negatively correlated to RGR and total biomass, respectively, (Fig. 3c, d), and thus the lower CC in W. trilobata might play an important role in accelerating its growth. Furthermore, elevated CO2 magnified the discrepancy in construction cost between these two species (Fig. 2b), implicating that the competitive advantage in invasive W. trilobata would become more outstanding in the future.

In comparison with W. chinensis, W. trilobata exhibited more sensitive to CO2 enrichment and significantly enhanced energy allocation to stems (Table 2). Similar observations had been found in yellow-poplar (Wullschleger et al. 1997) and Xanthium strumarium (Nagel et al. 2005). Under competitive conditions, stem growth is thought to enhance plant fitness by increasing light interception (Sasek and Strain 1991; Weining 2000). In contrast, W. chinensis allocated more energy to construct root at both CO2 treatments (Table 2). Plants tend to allocate resources to the organ necessary for collecting the resource most limiting to growth (Rogers et al. 1996). Elevated CO2 concentration increases plant nitrogen requirement for growth (Cotrufo et al. 1998). W. trilobata exhibited greater PNUE than that of W. chinensis, which could compensate for N limitation at elevated CO2. For W. chinensis, the more energy allocated to roots could facilitate its nutrient absorption from soil and reduce N limitation.

For invasive W. trilobata, elevated CO2 significantly stimulated energy assimilation via photosynthetic activity, but decreased energetic cost of biomass construction, and thus greatly increased its photosynthetic energy-use efficiency. In contrast, the indigenous W. chinensis exhibited a slight increase in PEUE. These results indicated that the invasive species could assimilate more carbon per unit of energy invested in biomass construction under elevated CO2. Considering the significantly positive correlations between PEUE and growth characteristics, it was inferred that the greater PEUE of invasive W. trilobata in elevated CO2 could enhance its growth and competitive ability over W. chinensis. The present results indicated that elevated CO2 concentration increased energy-use efficiency of invasive species over its indigenous congener, which might be an alternative physiological mechanism promoting success of invasion with ongoing increasing CO2.