Elevated CO2 and ammonium nitrogen promoted the plasticity of two maple in great lakes region by adjusting photosynthetic adaptation

Introduction Climate change-related CO2 increases and different forms of nitrogen deposition are thought to affect the performance of plants, but their interactions have been poorly studied. Methods This study investigated the responses of photosynthesis and growth in two invasive maple species, amur maple (Acer ginnala Maxim.) and boxelder maple (Acer negundo L.), to elevated CO2 (400 µmol mol-1 (aCO2) vs. 800 µmol mol-1 (eCO2) and different forms of nitrogen fertilization (100% nitrate, 100% ammonium, and an equal mix of the two) with pot experiment under controlled conditions. Results and discussion The results showed that eCO2 significantly promoted photosynthesis, biomass, and stomatal conductance in both species. The biochemical limitation of photosynthesis was switched to RuBP regeneration (related to Jmax ) under eCO2 from the Rubisco carboxylation limitation (related to Vcmax ) under aCO2. Both species maximized carbon gain by lower specific leaf area and higher N concentration than control treatment, indicating robust morphological plasticity. Ammonium was not conducive to growth under aCO2, but it significantly promoted biomass and photosynthesis under eCO2. When nitrate was the sole nitrogen source, eCO2 significantly reduced N assimilation and growth. The total leaf N per tree was significantly higher in boxelder maple than in amur maple, while the carbon and nitrogen ratio was significantly lower in boxelder maple than in amur maple, suggesting that boxelder maple leaf litter may be more favorable for faster nutrient cycling. The results suggest that increases in ammonium under future elevated CO2 will enhance the plasticity and adaptation of the two maple species.


Introduction
Global change can affect the structure and productivity of ecosystems via its effects on the physiological processes of individual plants (Campoy et al., 2021).Elevated atmospheric CO 2 , nitrogen (N) deposition and plant invasion are drivers of ecosystem changes (Bäurle et al., 2023;de Souza, 2023).A good understanding of plant phenotypic plasticity in response to elevated CO 2 and N deposition is essential for predicting the survival and growth of plant species under the predicted future climate conditions (Du et al., 2019;Liu et al., 2022).
Studies have shown that elevated CO 2 levels tend to have a fertilization effect on plants (Zhu et al., 2021)However, the increases in photosynthetic rate induced by elevated CO 2 do not necessarily always lead to increased biomass accumulation and growth (Skinner et al., 2018;Cabon et al., 2022;Green and Keenan, 2022).Elevated CO 2 can result in a downregulation of leaf photosynthetic capacity (maximum rate of ribulose-1,5bisphosphate carboxylation, V cmax ; maximum of photosynthetic electron transport rate, J max ) and stomatal conductance (g s ) (Sperry et al., 2019;Tcherkez et al., 2020).Photosynthetic downregulation under elevated CO 2 is generally resulted from increases in photosynthetic carbohydrate production and subsequent dilution in leaf nitrogen concentration (Yin et al., 2019).Under elevated CO 2 , the limitation of photosynthesis In C3 plants can shift to RuBP regeneration (indicated by J max ) from Rubisco carboxylation (indicated by V cmax ) (Dusenge et al., 2019;Smith and Keenan, 2020) although co-limitation by the two processes is generally the norm in most C3 species (Smith and Keenan, 2020).Photosynthesis can also be limited by CO 2 diffusions, such as g s and mesophyll conductance (g m ) (Sakoda et al., 2021).The adaptability of invasive tree species under elevated CO 2 should be fully explored.
Elevated CO 2 can inhibit the absorption and assimilation of nitrate but magnify the effects of ammonium addition (Domiciano et al., 2020).Elevated CO 2 may affect nitrate assimilation by inhibiting photorespiration because the assimilation process depends on reductants produced by photorespiration (Ainsworth and Long, 2021).Some studies suggest that elevated CO 2 directly inhibits the activity of nitrate reductase (Wujeska-Klause et al., 2019).The relative abundance of ammonium and nitrate depends vary with soil type, microbial community, forest successional stage and atmospheric N deposition (Poucet et al., 2021;Wang et al., 2023).Ammonium is generally the main nitrogen source in late successional stages (Bloom, 2015).With the increasing proportion of ammonium in atmospheric N deposition, N source forms become more important in the study of plant response to climate change (Luo et al., 2022).
Plant invasion is also a driver of changes in plant communities associated with global change (Petruzzellis et al., 2021).Invasive plants can benefit more from elevated CO 2 than local species and hence climate change can further promote plant invasion and changes in species composition (Rathee et al., 2021).Plants usually have higher phenotypic plasticity in photosynthesis and biomass allocation, especially in key leaf traits (Specific leaf area (SLA), N area , and net photosynthesis rate (A n )) (Onoda et al., 2017;Liu et al., 2022).Invasive plants also have higher capability for resource capture and utilization efficiency by reducing the cost of leaf construction (Barros et al., 2020).Plant invasion threatens biodiversity and ecosystem services (Petruzzellis et al., 2021).However, invasive plants can also lead to soil eutrophication and improve vegetation productivity (Lee et al., 2017).These paradoxes make it difficult to predict the response of plants to future global changes, especially elevated CO 2 and different forms of N deposition.It has been found that increases in ammonium in the soil are conducive to plant invasion, but its interaction with elevated CO 2 is poorly understood (Chen et al., 2021).
Amur maple (Acer ginnala Maxim.) is native tree species in northeast Asia and is popular as its leaves produce antioxidants similar to that of green tea (Bi et al., 2016).Amur maple has high ornamental value and is widely introduced in North America in the 1960s.Due to its tenacious adaptability, amur maple is listed as an invasive plant species by the Natural Resources Conservation Service and the Forestry Service of the United States Department of Agriculture (USDA, 2005).Boxelder maple (A.negundo L.) is a native pioneer tree species in North America with strong adaptability and has successfully invaded Europe, Asia, South America and Australia (Saccone et al., 2010;McEvoy et al., 2022).Boxelder maple has strong phenotypic plasticity and resource allocation ability but can promote the invasion by other species because of the fast nitrogen cycling of its litter (Portéet al., 2011).A recent genomic study suggests that boxelder maple has a smaller genome with recent gene family evolution which might be related to tendencies (McEvoy et al., 2022).However, there are few studies on the responses of these two invasive maple species to elevated CO 2 and N forms.T The purpose of this study was to verify the following scientific questions: 1. Elevated CO 2 promoted the photosynthetic rate of two maple trees and changed the biochemical limits of photosynthesis; 2. The morphological plasticity of leaves was sensitive to climate change; 3. Compared with nitrate, ammonium nitrogen was conducive to the growth of two maple under elevated CO 2 condition.

Plant materials and treatments
The seeds of Amur maple and boxelder maple were collected from 6 mature trees at least 50 m apart from each other in the natural forests (amur maple at 48.416N, 89.267W; boxelder maple at 48.429N, 89.261W) near Lakehead University Thunder Bay campus (Ontario, Canada).The seeds were stratified in wet sand at 4°C for about 60 days before being sown in germination trays.Seedlings with 10 cm high were transplanted into 4 L plastic pots filled with a mixture of vermiculite and peat moss (1:1, v:v) (Sun Gro ® , 770 Silver Street, Agawam, MA, USA) and were treated with different CO 2 and nitrogen form.
The experiment was conducted in four research greenhouses (G1, G2, G3, G4 see Table 1) on the Thunder Bay campus of Lakehead University.The experiment followed a split plot design with two [CO 2 ] levels [ambient 400 µmol mol -1 (aCO 2 ) and elevated 800 µmol mol -1 (eCO 2 )] as the main plot, and three nitrogen treatments as the split plot (10 mM ammonium, 10 mM nitrate, 10 mM N with equal proportion of ammonium and nitrate) (Table 1).The two tree species were nested within the N-CO 2 combination.Each CO 2 level had two replicates (two separate greenhouses).Each replicate of each treatment combination had 8 seedlings of each species (a total of 96 seedlings per species: 2 CO 2 * 2 replicates * 3 N forms * 8 seedlings).
The nitrogen sources used ammonium-sulfonate (NH 4 ) 2 SO 4 for ammonium and sodium-nitrate NaNO 3 for nitrate (BioBasic Inc. 20 Konrad Crescent,Markham,ON,Canada).All N treatments received identical amounts of other nutrient elements: 5 mM Potassium phosphate and 4 g L -1 Micromax Micronutrients Granular (Calcium 6%, Magnesium 3%, Sulphur 12%, Boron 0.1%, Copper 1%, Iron 17%, Manganese 2.5%, Molybdenum 0.05%, Zinc 1%) (Everris NA Inc. P.O.Box 3310, Dublin, OH, USA).The extra S provided with the ammonium formulation and the extra Na provided with the nitrate formulation were compensated in the nutrient solution to avoid differences in other elements between the different N treatments.The seedlings were fertilized twice a week and irrigated with 500 ml water within fertilizer according to the corresponding treatment every other day.The seedlings were randomly changed the position within the same greenhouse in each treatment every two weeks.
The CO 2 concentration in each greenhouse was maintained using by a CO 2 generator (GEN-2E, Custom Automated Products Inc., Riverside, California, USA).Other environmental conditions in all the greenhouses were 25/16°C (day/night) temperature, 50% RH and 16-h photoperiod.High-pressure sodium lamps (P.L. Systems, Grimsby, ON, Canada) were used to supplement then natural light when the ambient light intensity was below 500 µmol m -2 s -1 or the natural daylength was shorter than 16-h.All the environmental variables in each greenhouse were monitored and controlled by an Argus Titan Environment System (Argus Control Systems Ltd.Vancouver, BC, Canada).

Gas exchange and pigment measurement
After three months treatments, six seedlings were randomly selected from each treatment combination and foliar gas exchange was measured between 9:00 am and 16:00 pm on a unshaded mature leaf on the near the top of the canopy using a LI-6800 Portable Photosynthesis System (LI-COR Biosciences, 4647 Superior Street, Lincoln, Nebraska, USA).The gas exchange under treatment CO 2 was measured at the corresponding growth [CO 2 ] (aCO 2 at 400 µmol mol -1 , eCO 2 at 800 µmol mol -1 ), 25°C temperature, 1.1 -1.3 kPa water vapor pressure deficit, 1000 µmol m -2 s -1 photosynthetically active radiation flux density.At least 30 minutes was allowed before a steady state reading was taken.The net photosynthetic rate (A n-g ), stomatal conductance (g s ), intercellular [CO 2 ] to ambient [CO 2 ] ratio (C i /C a ), and photosynthetic nitrogen uses efficiency (PNUE = A n-g / N area ) were subjected to statistical analyses.
Photosynthetic response to intercellular [CO 2 ] (A/C i ) curves were measured at 400,300,200,100,50,400,500,600,800,1000,1200, 1500 µmol mol -1 [CO 2 ], 25°C temperature, 1.1 -1.3 kPa VPD, and 1000 µmol m -2 s -1 PAR.The maximum rate of Rubisco carboxylation V cmax and maximum rate of photosynthetic electron transport J max were estimated using the fitaci function, and the transition point (C i-t , A n-t ) of biochemical limitation from Rubisco to RuBP regeneration was estimated using findCiTranstion function in the Plantecophys R package from A/Ci data (Duursma, 2015).The initial slope of A/C i was estimated as apparent carboxylation efficiency (ACE), and the X-axis intercept was evaluated as the CO 2 compensation point (G ACi ).The photosynthetic light response curve (lrc) was measured at 400 µmol mol -1 [CO 2 ] and 1000, 1500, 1200, 900, 600, 300, 150, 50, 0 µmol m -2 s -1 PAR (Figure 1A).Other measurement environment conditions were the same as A/C i curve measurement.Curvature (q) and photosynthetic rate of saturation light (A n-max ) were fitted from the lrc data by non-rectangular hyperbola model using nonlinear least squares in R (Salter et al., 2019).The initial slope of the lrc was estimated as the apparent quantum yield (AQY) and the Xaxis intercept was estimated as the light compensation point (LCP).The fitting line of A n vs. C i from lrc dataset (Figure 1B) was used to estimate A n-total by the Y-axis intercept and A n-total /C a as the initial slope (Wang and Dang, 2023).
The initial line portions of three A/C i were measured with 200, 150, 100, and 50 µmol mol -1 of [CO 2 ], and 300, 150, and 75 µmol m -2 s -1 PAR, from which the daytime respiratory rate (R d ) and intercellular CO 2 compensation point (C i *) were calculated using Walker's slope intercept method (Walker et al., 2016).The variable J method was employed to calculate mesophyll conductance (g m ) (Harley et al., 1992), using the equation Walker et al., 2016), where G* is the CO 2 compensation point in absence of day respiration.
After the gas exchange measurement, the tested leaves were immediately removed and stored in a refrigerator (-70°C).Leaf pigment was extracted using 80% acetone and the absorbance at 645 nm and 663 nm measured.Chlorophyll concentration (Chl) of the leaves was calculated with Chl = 20.2* A645 + 8.02 * A663 (Wellburn, 1994).

Relative photosynthetic limitation and nitrogen partitioning
The relative limitations of photosynthesis by biochemical (l b ), mesophyll conductance (l m ) and stomatal conductance (l s ) were calculated according to Grassi and Magnani (Grassi and Magnani, 2005): l b = g t /(g t + әA n /әC c ); l s = (g t /g s * әA n /әC c )/(g t + әA n /әC c ); l m = (g t /g m * әA n /әC c )/(g t + әA n /әC c ), where g t is the total conductivity of CO 2 diffusion (g t = 1/(1/g m + 1/g s )), әA n /әC c is the initial slope of the A n to C c response curve and was calculated using the equation әA n / әC c = V cmax /(G* + K m ) (Farquhar et al., 1980), where G* was as noted previously and K m was calculated by K m = K c (1 + O/K o ) (Bernacchi et al., 2001).
We investigated the partitioning of leaf N into carboxylation (N cb ), electron transfer (N et ), light capture (N lc ) and respiration (N resp ).The calculation is based on Trouwborst's method (Trouwborst et al., 2011).

Biomass and nitrogen measurement
After three months CO 2 and nitrogen form treatments, the height of all the seedlings was measured.Leaf area was measured using a Regent WinFolia system (Regent Instruments Inc., Quebec City, QC, Canada).The seedlings were then harvested and separated into roots, stems, and leaves, and oven-dried at 75°C for 48 hours to constant weight to determine the biomass of different organs and specific leaf area (SLA).Leaf mass-based N concentration (N mass ) and carbon concentration (C) were measured using the dry combustion method on a CNS-2000 (LECO Corp., St. Joseph, MI, USA) at the Lakehead University Centre for Analytical Services.Area-based leaf N concentration (N area ) and total leaf N content per plant (N leaf ) were determined from leaf biomass and SLA.

Statistical analysis
The effects of CO 2 and N forms on photosynthetic and growth traits of individual tree species were investigated using two-way analysis of variance using the agricolae package in R. A three-way (CO 2 , N forms and species) ANOVA was performed for A n-g and A sample of a photosynthetic light response curve (lrc) with relevant parameters (A) the photosynthetic rate at saturation light and 400 µmol mol −1 CO 2 (A n- max ), the curvature of the lrc (Theta, q), apparent quantum yield (the initial slope of lrc, AQY), and light compensation point (LCP).A n vs. C i derived from the lrc database (B), where A n-total and A n-total/ C a are the y-intercept and slope of the Wang and Dang 10.3389/fpls.2024.1367535 Frontiers in Plant Science frontiersin.orgPNUE.Logarithmic or power transformation was used to transform variables that did not meet the normality and homogeneity assumptions before ANOVA was carried out.When a interaction was statistically significant, Tukey post hoc pairwise comparisons were conducted for the means.The principal component analysis (PCA) function of the FactoMineR package was used for PCA to investigate the relationship and clustering of physiological and growth-related parameters.The Bowen method (Bowen et al., 2017) was used to investigate the direct/indirect effects of CO 2 and N sources as categorical variables on biomass and A n-g using the structural equation model (SEM) and the psem function in the pieceweSEM package (Lefcheck, 2016).Based on the leaf economic spectroscopy (LES) and a prior structural model (Onoda et al., 2017;Liu et al., 2022), biomass was explained by A n-g , N area and SLA, and A n-g was explained by photosynthetic capacity, Chl and g t .The photosynthetic capacity was obtained by the regression equations from V cmax and J max .All the statistical analyses were performed using R.

Growth in response CO 2 and N form
The eCO 2 significantly increased the biomass of both maple species but the effect was much bigger seedlings fertilized with NH4 or a combination of NH4 and NO3 than those fertilized with NO3 only (Figure 2).While the general response patterns were similar in the two species, there were differences in the responses between the two species: under the aCO 2 , no significant effect of N form on the biomass was observed in amur maple (Figure 2A) but it was significantly lower in the boxelder maple seedlings that were supplied with ammonium only than those of seedlings that were fertilized with the other two N treatments (Figure 2B); Under eCO 2 treatment, the nitrate nitrogen only treatment significantly decreased the biomass of both maple species (Figure 2).CO 2 treatment significantly affected seedling height and leaf area (Tables 1, 3).However, these values of the boxelder maple were influenced by the interaction of the two treatments and showed a trend similar to that of biomass (Table 3).In general, the eCO 2 treatment significantly reduced SLA, except for amur maple treated with nitrate and boxelder maple treated with ammonium (Tables 2, 3).

Leaf nitrogen and PNUE
Treatment of eCO 2 significantly increased N area in amur maple (Table 2) but not in boxelder maple (Table 3).In general, eCO 2 treatment promoted the N leaf of the two tree species with the except of boxelder maple fertilized with nitrate (Table 3).The eCO 2 treatment significantly promoted the leaf C/N value of boxelder maple (Table 3) but not of amur maple (Table 2).
Amur maple distributed relatively larger proportions of leaf N to N cb , Net and N lc than did of boxelder maple while N allocation to N resp was larger in boxelder maple than amur maple (Figure 3).Nitrate and eCO 2 treatment slightly limited leaf N allocation to carboxylation in both species (Figure 2).In general, the PNUE of boxelder maple was higher than that of amur maple, except for NN treatment under aCO 2 and nitrate treatment under eCO 2 (Figure 4A).Ammonium significantly reduced PNUE in amur maple under aCO 2 treatment (Figure 4A).

Photosynthesis traits
A n-g was promoted by eCO 2 and its response in both tree species was similar to that of biomass under different treatments Biomass responses to CO 2 and N form treatments in amur maple (A) and boxelder maple (B)."a-" means ambient CO 2 (400 µmol mol −1 ) and "e-" means elevated CO 2 (800 µmol mol −1 ) treatments.NH4: fertilized 10 mM (NH 4 ) 2 SO 4 ; NN: fertilized with a combination of 5 mM (NH 4 ) 2 SO 4 and 5 mM NaNO 3 ; NO3: fertilized with 10 mM NaNO 3 .Means (± SE, n=6) with different letters indicated significant differences between treatments (Tukey host hoc test, P<0.05).(Figure 4B).The promoting effect of eCO 2 on A n-g was smaller or not statistically significant under nitrate treatment (Figure 4B).Nitrate decreased both V cmax and J max in both species under eCO 2 although the effect was not always statistically significant but did not significantly affect either variable in either species aCO 2 treatment (Tables 4, 5).Ammonium significantly increased leaf Chl concentration of amur maple grown under eCO 2 (Table 4) while nitrate significantly increased leaf Chl of boxelder maple grown under aCO 2 (Table 5).
Nitrate significantly decreased g m of amur maple under aCO 2 condition but increased it under eCO 2 (Table 4).In contrast, the g m of boxelder maple was intermediate in seedlings fertilized with nitrate under eCO 2 while there was not significant difference among N treatments under aCO 2 (Table 5).C i /C a was generally higher in eCO 2 than under cCO 2, especially in boxelder maple (Tables 4, 5).

A/C i traits and photosynthetic limitations
The transition point (C i-t , A n-t ) of photosynthetic limitation from Rubisco carboxylation to RuBP regeneration ion shifted to higher C i-g and higher A n-g under eCO 2 in both species (Figure 5).The transition point was much lower in amur maple seedlings fertilized with ammonium than other N treatments under aCO 2 , but G ACi was greater (Figures 5A-C).Ammonium significantly promoted (C i-g , A n-g ) and (C i-t , A n-t ) in amur maple (Figures 5D-F), nitrate lowered the transition point in boxelder maple under eCO 2 (Figures 5J-L).
Relative photosynthetic limitation analysis showed that photosynthesis was primarily limited by g s in amur maple seedlings treated with aCO 2 and ammonium but primarily  .Two-way ANOVA was performed to analyze CO 2 and nitrogen (N) as well as their interactive effects (CO 2 : N).Significant effects (P ≤ 0.05) are shown in bold and ":" indicated interaction.Different letters within the same column indicated statistically significant differences between treatments (Tukey post hoc test, P<0.05).aCO 2 : ambient CO 2 (400 µmol mol −1 ); eCO 2 : elevated CO 2 (800 µmol mol −1 ); NH4: fertilized 10 mM (NH 4 ) 2 SO 4 ; NN: fertilized 10 mM N with 5 mM (NH 4 ) 2 SO 4 and 5 mM NaNO 3 ; NO3: fertilized 10 mM NaNO 3 .SLA: specific leaf area; N area : leaf N content based on leaf-area; N leaf : whole plant leaf N content; C/N: leaf carbon-nitrogen ratio.limited by mesophyll conductance in all other treatment with the exception of seedlings treated with nitrate under eCO 2 where it was limited almost equally by mesophyll conductance and biochemistry (Figure 6A).In contrast, the photosynthesis of boxelder maple was mainly limited by biochemical and g s (Figure 6B), particularly in the combination of nitrate and eCO 2 (Figure 6B).

Adaptation to CO 2 and N form
PCA showed an interesting result that the photosynthetic capacity parameters (V cmax , J max ) did not orient to the ellipse of aCO 2 in the two species (Figures 7A, C).Seedlings grown under the eCO 2 were grouped into the ellipses with growth parameters (biomass and leaf area) and A n-g (Figures 7A, C).It is also interesting to note that N area and N leaf clustered in the eCO 2 ellipse in amur maple (Figure 7A) while C/N clustered in the eCO 2 ellipse in boxelder maple (Figure 7B).Different N forms had no obvious effect on the response to CO 2 in amur maple since the three ellipses largely overlapped (Figure 7B).In boxelder maple, in contrast, nitrate clustered with N mass and A n-total /C a , while the mixed N treatment was clustered with growth parameters (Figure 7D).It is worth noting that A n-max /A n-total from lrc was negatively correlated with the growth parameters in both tree species at about 180-degree angles (Figures 7A-D), and a similar effect was found in SLA in boxelder maple (Figures 7C, D).
We constructed SEM (Structural Equation Model) to evaluate the direct/indirect effects on biomass and A n-g by the two treatments (CO 2 and N) as categorical variables using piecewiseSEM package in R (Figures 8A, B).The results showed that CO 2 treatment had direct effects on the biomass of both tree species (Figures 8C, E), indirect effects on the biomass through N area in amur maple (Figure 8C), and indirect effects on the biomass through A n-g and SLA in boxelder maple (Figure 8E).N forms had a significant effect on the SLA in amur maple, but its impact on biomass did not reach a significant level (Figure 8C).However, N treatment had a significant indirect effect on biomass through A n-g and SLA in boxelder maple (Figure 8E).
CO 2 had a significant direct effect on the A n-g of both maple (Figures 8D, F), and a significant indirect effect on the biomass through photosynthesis in amur maple (Figure 8D).No significant effect of N form was observed on A n-g and related parameters in amur maple (Figure 8D, not shown).In boxelder maple, N treatment affected total CO 2 conductance, but had no significant effect on A n-g (Figure 8F).

Adaptation to elevated CO 2
Elevated CO 2 and associated climate change affect plant growth and distribution through via influencing physiological processes (Lauriks et al., 2022;Ma et al., 2023).Elevated CO 2 usually causes photosynthetic adaptation, which is typically manifested by decreases in photosynthetic capacity (V cmax , J max ), N area and g s (Hao et al., 2023), but promotes biomass production and photosynthetic rate (Dusenge et al., 2020;Tcherkez et al., 2020).However, there was no reduction in photosynthetic capacity or N area observed in either species in this study.To the contrary, eCO 2 increased V cmax and J max in seedlings fertilized with ammonium.The photosynthetic downregulation induced by eCO 2 is generally related to the dilution of leaf N and reduction in Rubisco as a result of increased carbohydrate production and growth (Dusenge et al., 2019).
Our PCA analysis shows that under aCO 2 treatment, both amur maple and boxelder maple converged on the ellipse in the direction of A n-max /A n-total , (the ratio of A n-max and A n-total from lrc, see Figure 1B and Supplementary Table S1), and that A n-max /A n-total was always inversely proportional to C i /C a .From Figure 1B, A Leaf N partitioning into carboxylation (N cb ), electron transfer (N et ), light capture systems (N lc ), and respiratory (N resp ) in response to CO 2 and different N forms in amur maple (A) and boxelder maple (B)."a-" means ambient CO 2 (400 µmol mol −1 ) and "e-" means elevated CO 2 (800 µmol mol −1 ) treatments.NH4: fertilized with 10 mM (NH 4 ) 2 SO 4 ; NN: fertilized with a combination of 5 mM (NH 4 ) 2 SO 4 and 5 mM NaNO 3 ; NO3: fertilized with 10 mM NaNO 3 .
which means that A n-max /A n-total + C i /C a = 1.This seems to demonstrate an intrinsic inverse relationship between A n-max /A n-total and C i /C a .Elevated CO 2 reduced A n-max /A n-total , but increased C i /C a in both species (Supplementary Tables S3, S4).These results indicated that the photosynthetic adaptation processes in two maple species appeared diverse and highly plastic (Moejes et al., 2017).
The SEM revealed that eCO 2 had positive effects on the biomass and A n-g of both species, but only the effect on A n-g in boxelder maple was statistically significant.CO 2 is considered as a signal to control stomatal movement and growth metabolisms (Hao et al., 2023).Our results show that the biomass of both maple species was closely related to A n-g and CO 2 in the absence of other stresses.

Biochemical and CO 2 diffusion limitations to photosynthesis
This study shows that an important effect of eCO 2 is that it shifted the biochemical limitation of photosynthesis from Rubisco carboxylation to RuBP regeneration, i.e., from A n-g < A n-t under aCO 2 to A n-g > A n-t under eCO 2 (Wang et al., 2022).This result suggests that under the condition of the projected future climate, photosynthesis may be mainly limited by J max rather than V cmax (Smith and Keenan, 2020).It should be noted that when photosynthesis is limited by J max , the sensitivity of A n to C i variation is reduced (Dusenge et al., 2019).The result that C i /C a ratio was higher under eCO 2 than under aCO 2 suggests that Effects of CO 2 and N forms on photosynthetic nitrogen use efficiency [PNUE, (A) and photosynthetic rate at growth CO 2 (A n-g , (B)] in amur maple (am) and boxelder maple (bm).aCO 2 : ambient CO 2 (400 µmol mol −1 ); eCO 2 : elevated CO 2 (800 µmol mol −1 ); NH4: fertilized with 10 mM (NH 4 ) 2 SO 4 ; NN: fertilized with a combination of 5 mM (NH 4 ) 2 SO 4 and 5 mM NaNO 3 ; NO3: fertilized with 10 mM NaNO 3 .Means (± SE, n=6) with different letters indicated significant differences between treatments (Tukey host hoc test, P<0.05).
photosynthesis was less limited by CO 2 diffusion in the two species grown under eCO 2 (Lamba et al., 2018).
Photosynthetic capacity represents the CO 2 fixation potential of leaves and does not represent the photosynthetic rate under actual growth conditions (Wang et al., 2022), since the net photosynthetic rate of C3 plants is also affected by substrate CO 2 supply (Stefanski et al., 2020).For example, the photosynthetic capacity of amur maple treated with ammonium under aCO 2 was maintained at a high level, while the actual A n was the lowest, which was obviously limited by low CO 2 supply and low stomatal conductance (Table 2, Figure 3A).Our results support the theory that higher photosynthetic capacity does not always translate into higher A n (Xu et al., 2020).
The SEM showed that the resistance to CO 2 diffusion was not a primary driver for differences in A n between treatments (Figures 8D, F) and the relationship appeared to be opposite between the two maple species.The resistance to CO 2 diffusion mainly includes g s and g m and has a significant effect on photosynthesis (Grassi and Magnani, 2005).However, using g t [g t = g s * g m /(g s + g m )] to express the overall conductivity of CO 2 seems to blur the individual effect of g s or g m (Grassi and Magnani, 2005).Although g s and g m jointly describe the diffusion of CO 2 from leaf surface to carboxylation sites, they are independent processes with different regulatory mechanisms and physiological significance.The g s may be closely related to water status (Stefanski et al., 2020), while g m integrates biochemical and TABLE 4 Photosynthetic capacity, chlorophyll concentration, conductance to CO 2 diffusion, and intercellular to external CO 2 concentration ratio in amur maple seedlings grown at ambient or elevated CO 2 in response to different forms of N source.(n=6).Two-way ANOVA was performed to analyze CO 2 and nitrogen form (N) as well as their interactive effects (CO 2 : N).Significant effects (P ≤ 0.05) are marked in bold and ":" indicated interaction.Different letters within the same column indicated statistically significant differences between treatments (Tukey post hoc test, P<0.05).aCO 2 : ambient CO 2 (400 µmol mol −1 ); eCO2: elevated CO 2 (800 µmol mol −1 ); NH4: fertilized 10 mM (NH 4 ) 2 SO 4 ; NN: fertilized 10 mM N with 5 mM (NH 4 ) 2 SO 4 and 5 mM NaNO 3 ; NO3: fertilized 10 mM NaNO 3 .V cmax : maximum rate of ribulose-1,5-bisphosphate carboxylation; J max : maximum photosynthetic electron transport rate; Chl: leaf chlorophyll concentration.g m : mesophyll conductance; g s : stomatal conductance; C i /C a : intercellular (C i ) to ambient (C a ) CO 2 concentration ratio.   .Two-way ANOVA was performed to analyze CO 2 and nitrogen (N) as well as their interactive effects (CO 2 : N).Significant effects (P ≤ 0.05) are shown in bold and ":" indicated interaction.Different letters within the same column indicated statistically significant differences between treatments (Tukey post hoc test, P<0.05).aCO 2 : ambient CO 2 (400 µmol mol −1 ); eCO 2 : elevated CO 2 (800 µmol mol −1 ); NH4: fertilized 10 mM (NH 4 ) 2 SO 4 ; NN: fertilized 10 mM N with 5 mM (NH 4 ) 2 SO 4 and 5 mM NaNO 3 ; NO3: fertilized 10 mM NaNO 3 .V cmax : maximum rate of ribulose-1,5-bisphosphate carboxylation; J max : maximum photosynthetic electron transport rate; Chl: leaf chlorophyll concentration.g m : mesophyll conductance; g s : stomatal conductance; C i /C a : intercellular (C i ) to ambient (C a ) CO 2 concentration ratio.

N forms and partitioning effects
It is believed that eCO 2 is beneficial to plants that prefer ammonium over nitrate (Andrews et al., 2019;Dusenge et al., 2019).Our results show that ammonium fertilization led to the highest while nitrate resulted in the lowest photosynthesis and growth among the three fertilization treatments under eCO 2 .This result may be related to the lower energy consumption and higher metabolic efficiency of ammonium compared with nitrate (Bloom, 2015;Rubio-Asensio and Bloom, 2017).However, the accumulation of more ammonium in cells may cause ammonium toxicity, resulting in the decrease of g s and A n , chlorosis and growth inhibition (Vega-Mas et al., 2017).Our results that Chl and g s were lowest in seedlings fertilized with ammonium under aCO 2 may indicate a toxic effect of ammonium to those trees.Ammonium toxicity is attributed to excessive ammonium ions that lead to cell ion disturbances, the depletion of organic acid, and acid stress caused by the proton's mass formation (Hachiya et al., 2021).The eCO 2 can promote the synthesis of organic acids by providing the carbon framework and alleviating ammonium toxicity (Vega-Mas et al., 2017;La Peña et al., 2022).Our results showed that ammonium might be beneficial to the growth of the two maple species in the future when CO 2 is elevated.However, the mechanism of this synergistic effect of eCO 2 and ammonium needs further study.
It is common in C3 plants that eCO 2 inhibits the assimilation of nitrate (Rachmilevitch et al., 2004).The reason may be that eCO 2 inhibits photorespiration, and nitrate assimilation depends on  S2).

B A FIGURE 6
The relative photosynthesis limitations of biochemistry (l b ), stomatal resistance (l s ), and mesophyll resistance (l m ) in response to CO 2 and N form treatment in amur maple (A) and boxelder maple (B).Abbreviations are provided in Supplementary Table S1.
Principal Component Analysis (PCA) on growth and photosynthetic parameters in amur maple exposed to CO 2 (A) and N resource (B) treatments and in boxelder maple (C, D).The arrows point to near overlap, vertical, and reverse, which represent positive, no, and negative correlations between these parameters respectively.aCO 2 : ambient CO 2 (400 µmol mol −1 ); eCO 2 : elevated CO 2 (800 µmol mol −1 ); NH4: fertilized 10 mM (NH 4 ) 2 SO 4 ; NN: fertilized 10 mM N from 5 mM (NH 4 ) 2 SO 4 and 5 mM NaNO 3 ; NO3: fertilized 10 mM NaNO 3 .J max : maximum of photosynthetic electron transport rate; PNUE: photosynthesis nitrogen use efficiency; A n-t : net photosynthesis rate at transition point (C i-t , A n-t ) between Rubisco limitation and RuBP regeneration limitation based on A/C i curve; A n-g : net photosynthesis rate at a growth [CO 2 ] which eCO 2 at 800 µmol mol −1 and aCO 2 at 400 µmol mol −1 ; g s : stomatal conductance; N leaf : total N of the whole-plant leaf; Chl: leaf chlorophyll concentration; C i /C a : the ratio of C i and C a ; N area : leaf N per unit area; SLA: specific leaf area; A n-max /A n-total : the ratio of photosynthetic rate of saturation light at 400 µmol mol −1 (A n-max ) and the y-intercept of A n vs. C i fitting line (A n-total ) from light response curve database; A n-total/ C a : the slope of A n vs. C i fitting line from light response curve database; C/ N: leaf carbon and nitrogen ratio; V cmax : maximum rate of ribulose-1,5-bisphosphate carboxylation; See Supplementary Table S1 for other explanations.
photorespiration-related reductants (Ainsworth and Long, 2021).Furthermore, eCO 2 may also inhibit the activity of nitrate reductase (Wujeska-Klause et al., 2019).The mixed nitrate and ammonium treatment appeared more beneficial to the growth of boxelder maple than the application of only nitrate, this suggests that ammonium and nitrate may have some interactive effects on boxelder maple.Notably, N allocation may also play an important role in the regulation of photosynthesis (Wang et al., 2021).Amur maple treated by aNH4 and eNH4 had equivalent N allocation to photosynthesis (Figure 4A), but the photosynthetic rate and biomass in the latter treatment were more than twice that of the former.This was probably because the N in the eNH4 treatment was distributed more evenly to different components such as N et , N lc and N resp , than in the aNH4 treatment where there was a slight greater N allocation to N cb .

Plasticity and leaf morphological traits
Invasive plants generally have higher plasticity because they exhibit favorable phenotypes and robust adaptability in response to changes in environmental conditions (Godoy et al., 2012;Liu et al., 2017).Plasticity is reflected in the trade-offs between leaf structure and physiological progress described in the leaf economic spectrum The priori (showing all tested paths) piecewise structural equation model (pSEM) relating to direct and indirect effects of CO 2 and N form on biomass (A) and A n-g (B).Two CO 2 treatments (ambient CO 2 : 400 µmol mol −1 and elevated CO 2 : 800 µmol mol −1 ) and N source (10 mM N source by three forms: NH 4 + , NH 4 + +NO 3 -in 1:1, NO 3 -) are treated as categorical variables.A n-g : net photosynthesis rate at a growth [CO 2 ] which eCO 2 at 800 µmol mol −1 and aCO 2 at 400 µmol mol −1 ; N area : leaf N per unit area; SLA: specific leaf area; Chl: leaf chlorophyll concentration; g t : total conductance to CO 2 between the leaf surface and carboxylation sites (1/ gt = 1/g s + 1/g m ).Multiple regression from V cmax and J max to A n-g was used to construct photosynthetic capacity as a component variable (hexagonal frame).The SEM (using psem function in piecewiseSEM package R) of biomass and A n-g in amur maple are in (C, D), while that of boxelder maple are in (E, F).Arrows mean the directional influence between the variables and the solid lines represent significant relationships, and the dashed lines refer to nonsignificant relationships (P > 0.05).The numbers on top of the arrows represent the standardized path coefficients (for continuous variables), and the numbers under the box with R 2 refer to the degree of variation of the variable interpreted by all paths.Curved double arrows represent a significant correlation between variables (with correlation coefficient).The values on the line from the categorical variable indicated the continuous variable estimated marginal means by the treatment levels.(Liu et al., 2022), and in the optimal combination of key leaf traits SLA, N area and A n (Onoda et al., 2017).Successful invasive tree species often show higher SLA (lower leaf structure cost) and higher N area (higher nutrient resources), which is often associated with higher efficiency of resource acquisition, higher photosynthetic rates and faster investment returns (Xiong and Flexas, 2018).Our results show that both amur maple and boxelder maple exposed to eCO 2 increased leaf construction cost (SLA lowered) and had higher N area .The high phenotypic plasticity may further facilitate the spread of specific invasive species in the future when the atmospheric CO 2 elevation continues (Liu et al., 2017).

Adaptation strategy of amur maple and boxelder maple
Despite displaying comparable reactions to elevated CO 2 and different nitrogen sources in terms of biomass and leaf nitrogen allocation, Amur maple and boxelder maple seem to have adopted contrasting adaptation strategies for future climate change.Amur maple appears to prioritize maximizing photosynthetic capacity per unit area of leaves, whereas boxelder maple seems to excel in increasing biomass through adaptations in leaf morphology and nitrogen utilization efficiency (Supplementary Figures 3S, C).This trend is also evident in relative photosynthetic limitation (Figure 6) and structural equation models (Figures 8D, F).
Our results provide further support for the conclusions of previous studies that the great plasticity has facilitated the successful invasion of boxelder maple into Europe, Asia and South America (Portéet al., 2011;McEvoy et al., 2022).A study in Lithuania finds that boxelder maple facilitates invasion by increasing the rate of foliar decomposition and nutrient cycling (Manusadzǐanas et al., 2014).Our results showed that the C/N and N leaf of boxelder maple were significantly higher than those of amur maple, suggesting that boxelder maple may speed soil nutrient cycling and the invasion of other species (Lee et al., 2017).A wellknown paradox is that invasive plants reduce biodiversity but increase plant productivity (Rout and Callaway, 2009).A recent genomic comparison has found that boxelder maple has a smaller genome with recent gene family evolution which might be related to its robust adaptability (McEvoy et al., 2022).Even though the two maple trees have chosen different adaptation strategies for the future climate, they have both demonstrated a favorable reaction to the elevated CO2.It is important to note that the boxelder maple's strategy of optimizing leaf morphology and nitrogen utilization efficiency not only showcases its impressive phenotypic plasticity, but also enhances its invasive potential by facilitating nutrient cycling between leaves and soil.

Conclusions
Invasive tree species generally have greater plasticity and more efficient utilization of carbon and nitrogen sources in the environment.We found that amur maple and boxelder maple both showed strong plasticity in response to eCO 2 and variation in N sources.Furthermore, they seem to exhibit a coordinated response to the two treatment factors.Ammonium was not conducive to the growth and physiology of both species under current CO 2 but significantly improved their performance under the elevated CO 2 .However, the effects of nitrate were the opposite.Boxelder maple seems to have stronger adaptability to future climate change than Amur maple.Because boxelder maple invests more N in its leaves and has a larger specific leaf area, indicating that boxelder maple could promote soil nutrient cycling and ecosystem function.We conclude that increases in soil ammonium will be beneficial to the plasticity and adaptation of amur maple and boxelder maple in the future as atmospheric CO 2 continues to rise.
FIGURE 5The A/Ci curves in response to different N forms in amur maple (am, A-F) and boxelder (bm, G-L) seedlings grown under ambient CO2 (aCO2, A-C in am and G-I in bm) and elevated CO2 (eCO2, D-F in am and J-L in bm).NH4: fertilized with 10 mM (NH4)2SO4 (left column); NN: fertilized with a combination of 5 mM (NH4)2SO4 and 5 mM NaNO3 (middle column); NO3: fertilized with 10 mM NaNO3 (right column).Each point denotes the means of six (C i , A n ) values in A/C i curve.The circles represent the transition point (C i-t , A n-t ) from Rubisco carboxylation to RuBP regeneration of photosynthesis limitation.The triangles indicated the photosynthetic rate (C i-g , A n-g ) under growth C a (400 µmol mol −1 versus 800 µmol mol −1 ).ACE: apparent carboxylation efficiency estimated from the initial slope of A/C i curve; G ACi : CO 2 compensation point estimated from A/C i curve intersects point on X-axis.Different letters of the same parameter in the same species are significantly different between treatments (Tukey host hoc test, P<0.05, see Supplementary TableS2).

TABLE 3
Hight, leaf area, specific leaf aera (SLA), and nitrogen-related traits of boxelder maple grown at ambient and elevated CO 2 response different forms of N supply.
each value represents mean ± SE (n=6
each value represents mean ± SE (n=6).Two-way ANOVA was performed to analyze CO 2 and nitrogen (N) as well as their interactive effects (CO 2 : N).Significant effects (P ≤ 0.05) are marked in bold and ":" indicated interaction.SLA: specific leaf area; N area : leaf N content based on leaf area; N leaf : whole plant leaf N content; C/N: leaf carbon-nitrogen ratio.

TABLE 5
Photosynthesis capacity, chlorophyll concentration, and CO 2 diffusion conductance in boxelder maple seedlings grown at ambient and elevated CO 2 in response to different forms of N source.