Local ecotypic and species range-related adaptation influence photosynthetic temperature optima in deciduous broadleaved trees
- First Online:
- Cite this article as:
- Robakowski, P., Li, Y. & Reich, P.B. Plant Ecol (2012) 213: 113. doi:10.1007/s11258-011-0011-3
- 525 Views
Given prior evidence for local ecotypic and species-specific adaptation in trees, we hypothesized that: (1) Acer rubrum and Quercus rubra provenances with different climate origins should differ in photosynthetic temperature optimum (Topt) even after long-term growth in a common environment; (2) congeneric species Populus tremuloides and Populus deltoides with differing but overlapping ranges should not differ in Topt when co-occurring, due to the likelihood of both ecotypic thermal adaptation and phenotypic thermal acclimation. To address these questions, we investigated the temperature responses of pairs of A. rubrum and Q. rubra provenances planted in a common garden and the temperature responses of P. tremuloides and P. deltoides at four sites where the species ranges overlap in Minnesota, USA. Both studies showed significant signals of temperature adaptation. The provenances of both A. rubrum and Q.rubra that originated from northern sites with lower ambient temperature had lower Topt. This supported the hypothesis about the dominance of local ecotypic adaptation of photosynthesis to temperature despite opportunity for both long-term (12-year) acclimation to the common-garden temperature regime and short-term temperature acclimation. However, acclimation capacity to the temperatures experienced in the days and weeks before the gas exchange measurements differed among the contrasting provenances suggesting that the observed differences in Topt could be due to either fixed genotypic differences (e.g., adaptive Topt), acclimation of Topt, or both. In contrast, the Populus species with the colder home range, P. tremuloides, showed significantly (P < 0.05) lower Topt on average than co-occurring P. deltoides. Thus, despite the opportunity for both ecotypic adaptation and local acclimation, phylogenetic inertia still constrained the species with the colder overall range to a different temperature optimum than the one with a warmer overall range. Our results also imply that rapid but modest climate change may create mismatches between photosynthetic physiology and local climate because of lags in population or species-level adaptation.
KeywordsAdaptation to temperatureBroadleaved tree speciesPhotosynthesisPhotosynthetic temperature optimumProvenancePhotosynthetic temperature response curve
There is evidence that species or ecotypes growing in colder environment often but not always have lower photosynthetic temperature optima (Topt) than those growing at higher temperatures (Björkman et al. 1972; Fryer and Ledig 1972; Slatyer 1977; Berry and Björkman 1980; Cavieres et al. 2000; Cunningham and Read 2002; Gunderson et al. 2010). Taxa can be adapted to their different thermal origins and thus differ in Topt when they grow in a common environment (Berry and Björkman 1980; Cunningham and Read 2002; Gunderson et al. 2010). However, it is not clear whether they would still show differences in Topt even when co-occurring, since strong ecotypic adaptation to the common environment might erase any species-scaled differences. Additionally, acclimatization to the field conditions may erase any adaptive differences. Such variation in Topt can be due to genetic differences in temperature optima (that would arise from adaptation to local thermal environments), physiological acclimation (i.e., phenotypic) to growth temperature, or both. Intra-specific differences in Topt of chamber-grown tree seedlings have been long noted among populations originating from different thermal environments (Björkman et al. 1972; Fryer and Ledig 1972; Slatyer 1977). In some forest tree species inter- and intraspecific variation in Topt have been studied, but mostly in controlled conditions. Only a few studies included temperate or boreal tree species (e.g., Fryer and Ledig 1972; Gunderson et al. 2000; Ledig and Korbobo 1983; Wieser et al. 2010).
For temperate deciduous tree species, differences in Topt of species adapted to contrasting temperature environments were significant but small due to acclimation-temperature-driven shifts of Topt (Gunderson et al. 2010). However, another study (Dillaway and Kruger 2010) found no evidence of either thermal acclimation of Topt or of adaptive differences among species. However, there is to our knowledge no evidence of whether field grown trees of closely related taxa differing in their range differ in their Topt after growing for many years in common conditions (which gives extended opportunity for convergent acclimation).
In the present study, the photosynthetic temperature responses of deciduous broadleaved trees: Acer rubrum, Quercus rubra, Populus deltoides, and Populus tremuloides were investigated. In a 10-year-old common garden, pairs of southern and northern provenances of A. rubrum and Q. rubra presumably adapted to different temperatures of their origins were compared for their Topt derived from temperature–photosynthesis response curves. The photosynthetic temperature responses of congeneric P. deltoides and P. tremuloides were determined in situ, along a latitudinal gradient within the region where the geographic ranges of both species overlap.
Based on evidence of strong ecotypic thermal adaptation of photosynthesis (Fryer and Ledig 1972; Slatyer 1977; Berry and Björkman 1980; Ishikawa et al. 2007), we hypothesized that the A. rubrum and Q. rubra southern ecotypes adapted to warmer climates would show higher photosynthetic temperature optima than northern, cold adapted provenances. This would indicate that adaptation to the thermal conditions of origin would dominate over 10-year acclimation to common temperature in a common environment. An alternate hypothesis was that strong acclimation to local conditions would erase genotypic tendencies after a decade of common conditions.
In the trial along a geographical gradient, we tested the hypothesis that if P. tremuloides and P. deltoides ecotypes are primarily adapted to their local environment, individuals from a species with a colder overall range (P. tremuloides) should still have the same temperature optima for photosynthesis as co-occurring individuals from a species with a warmer overall range (P. deltoides). Alternatively, however, if genetic predisposition with respect to photosynthetic temperature optima is closely related to the overall range of a species, individuals of a species with a colder overall range might have lower temperature optima than nearby individuals of a species with a warmer overall range even when growing in comparable thermal conditions.
Materials and methods
Mean annual temperature (°C)
1 to 24
3 to 18 (4–16)
12 to 16
−3 to 11
Mean maximum temperature of warmest month (°C)
20 to 34
23 to 33
22 to 30
16 to 23
Mean minimum temperature of coldest month (°C)
−24 to 13
−18 to 5
−10 to 12 (8)
−30 to −3
Absolute minimum temperature (°C)
Mean annual rainfall (mm)
620 to 1,670
600 to 1,300 (760 to 2,030)
380 to 3,000 (380 to 1,400)
180 to 1,300
Dry season duration (month)
0 to 6
0 to 3
0 to 1
0 to 3
Acer rubrum (L.) is one of the most abundant and widespread trees in eastern North America. Its range extends between the latitudes 50°N and 26°N, and between the altitudes 0 and 1,800 m a.s.l. A. rubrum is classified as a sub-climax or mid-seral species, generally tolerant of shade in young age.
Quercus rubra (L.) is one of the most important trees in eastern North America. Its range extends between the latitudes 48°N and 31°N and between the altitudes 0 and 1,800 m a.s.l. Red oak can be considered a fire sub-climax species, mid-tolerant of shade that is eventually replaced by more shade-tolerant species in the absence of periodic fires.
Populus deltoides (Bartr. ex Marsh.) has a very wide natural distribution range in eastern, central, and southern North America (between the latitudes 50°N and 28°N). It mostly grows along streams and on bottom lands. Cottonwood is very intolerant of shade showing the traits of a pioneer species.
Populus tremuloides (Michx.) is the most widely distributed tree in North America. It is quick to pioneer many different sites between the latitudes 69°N and 24°N. Trembling aspen is a shade-intolerant, short-lived, wind-firm tree with a rapid initial growth rate.
Common-garden experiment with A. rubrum and Q. rubra
The description of geographical origin and climate characteristics of Acer rubrum and Quercus rubra used in the provenance experiment
Name of Provenance
altitude a.s.l. (m)
Average annual precipitation (mm)
Greatlakes Nursery Co., Wausau
Coweeta High & Med. Seed Source (North Carolina)
Saratoga Tree Nursery, Saratoga Springs
Vallonia State Nursery
Latitudinal transect of P. deltoides and P. tremuloides
The geographical coordinates of the locations where the photosynthetic temperature optima of Populus deltoides and Populustremuloides were determined, their climate characteristics, and the coordinates of the nearest meteorological stations
Lat/long of the locations
Lat/long of the station
Average annual precipitation (mm)
Ely (Hubachek Wilderness Research Center)
Ely Forestry Center
Cloquet (Clquet Forestry Center)
Cloquet Forestry Center
Saint Paul (U of Minnesota campus)
University of Minnesota
Kellogg (Weaver Dunes Scientific and Natural Area)
Alma DAM 4, Wisconsin
Measurements of gas exchange
Thermal responses of photosynthesis were measured with a Li-Cor 6400 portable photosynthesis system (Li-Cor, Lincoln, NE, USA). In the common garden, gas exchange was measured in A. rubrum and Q. rubra between 23 July and 29 September 2008. Both provenances of Acer and both provenances of Quercus were alternately measured within similar temperature conditions over a short period. The gas exchange of P. deltoides and P. tremuloides was measured along a geographical gradient from 30 August to 19 September. Prior to each measurement, a healthy leaf from the southern lower crown part situated ≈1.8–2.2 m above ground was chosen and entered into the infra-red gas analyser leaf chamber at the ambient temperature for 20 min. There was one temperature response curve per tree. In the leaf chamber PPF was maintained at the saturating level of 1,200 μmol m−2 s−1. Air flow was set at 400 μmol s−1, and CO2 concentration at 360 μmol mol−1. The leaf temperature was decreased to ≈3–7°C below the ambient temperature using the thermoelectric bloc and Li-Cor 6400-88 Expanded Temperature Control Kit within the cuvette, and then progressively increased at intervals until a distinct maximum and lowering of net CO2 assimilation rates were shown. The photosynthetic temperature response curves started from the mean initial leaf temperature between 14.7–20.8 (September—“cool period”) and 17.3–23.9°C (July–August—“hot period”). The lowest initial leaf temperature reached by the system was 13.1°C at the end of September (online resource 2). At each target temperature value the data were logged after around 6–8 min needed for equilibration of gas exchange, and three readings, 1 min apart, were taken at each set temperature. Cuvette humidity was not controlled, except for avoiding condensation at low temperatures. The amount of incoming air was routed through desiccant allowing incoming relative humidity (RH) not to exceed 65%. In the experiment with Q. rubra and A. rubrum the increase in temperature in leaf chamber was accompanied by a decrease in RH and greater values of leaf-to-air vapor pressure deficit (VPD) ranging from 0.9 to 3.6 kPa. When the gas exchange was measured in P. deltoides and P. tremuloides VPD in leaf chamber was more stable and ranged from 0.9 to 2.5 kPa.
Fitting of short-term temperature response curves of photosynthesis
The hierarchical analysis of variance in general linear model (GLM) with the independent variables: species and provenance nested in species, and the dependent variables: Aopt, Topt was applied to compare the photosynthetic temperature responses between A. rubrum and Q. rubra and between their provenances. The provenances of A. rubrum and Q. rubra within each species were compared a posteriori using the analysis of contrasts. All the statistical analyses were conducted at the significance level α = 0.05.
Populusdeltoides and P.tremuloides photosynthetic temperature responses were compared for the species and location effects using the two-way analysis of variance with interaction species × location. Both species of poplar in the different locations were compared a posteriori using the analysis of contrasts. The mean values of the parameter b (Eq. 1) describing the spread of the parabola were compared between the populations within each species using a non-parametric Mann–Whitney U test.
To determine potential short-term temperature acclimation of photosynthesis the relationships between Topt of the Acer and Quercus provenances, the Populus species and different temperatures (mean annual temperature, temperature of growth season, moving averages from 1, 3, 5, 7, or 10 days before the gas exchange measurements) were tested using the Pearson correlation coefficients and linear regression. The linear regression was also applied to analyze the relationships between the leaf temperature and ratio of internal CO2 partial pressure in leaf (Ci), leaf-to-air VPD deficit, and between Ci and net CO2 assimilation rate. The mean values of slope and constant were compared between the species and/or provenances with ANOVA or t test at P < 0.05. All the statistical analyses were conducted using Statistica 8.0 (Statsoft, Inc., Tulsa, OK, USA).
Photosynthetic temperature responses of the common-garden-grown provenances
The results of the hierarchical ANOVA for Acer rubrum and Quercus rubra in which the species or provenance nested in species were the source of variance
Source of variance
Effect of lack humidity control on photosynthetic temperature optimum
Linear regression between the leaf temperature and internal CO2 partial pressure (Ci)
Confidence interval −95% to 95%
−4.69 to 5.75
−7.05 to 7.86
−3.96 to 7.51
−7.24 to 15.13
−7.19 to −3.16
−12.26 to −0.74
−6.08 to −2.50
−9.03 to 1.63
Confidence interval −95% to 95%
42.2 to 289.3
−38.4 to 340.5
−15.1 to 267.9
−177.3 to 351.2
208.6 to 308.2
146.5 to 413.9
218.1 to 318.8
70.8 to 386.9
Correlations between growth temperatures and optimum temperature for photosynthesis
The relationship between the mean, mean minimum, and maximum temperatures observed 1, 3, 5, 7, or 10 days before the gas exchange measurements and photosynthetic temperature optima of the Acer rubrum and Quercus rubra provenances
Photosynthetic temperature responses of P. deltoides and P. tremuloides
The results of the two-way ANOVA with interaction in which species (Populusdeltoides and P. tremuloides), location, and interaction between the species and location (S × L) were the sources of variance (n = 37)
Source of variance
S × L
S × L
Topt did not depend on local location, however, indicating no evidence for long-term acclimation within the modest climate gradient among the four sites. For both species, Topt at the two sites cooler during the measurement period (mean daily maxima 21.8 and 24.8°C, July–September 2008) did not differ from at the two warmer sites (mean daily maxima 26.9 and 26.1°C, July–September 2008). For P. deltoides, Topt was generally higher than mean daily maximum, but for P. tremuloidesTopt was generally lower than mean daily maximum except for Ely. Topt of both poplars were higher than Tmean(3) and Tmean(10), but in the range or lower than Tmax(3) and Tmax(10) (online resource Table 2). Additionally, the more northern species had higher Aopt than the southern species at the two northern sites (Cloquet and Ely), but not at the southern ones which was confirmed by the analysis of contrasts between-species within each location (P < 0.05) (Fig. 3).
Ci was ≈220 μmol mol−1. The mean slopes and intercepts of the regression between Tleaf and Ci did not significantly differ between the species and between the species within each location in ANOVA (slope: F = 1.3, P = 0.3; F = 0.3, P = 0.8; intercept: F = 1.9, P = 0.2, F = 0.25, P = 0.9). There were not significant differences between the species in the parameters of regression between Tleaf and VPD. Topt of P. deltoids and P. tremuloides were not correlated with moving average temperatures 1, 3, 5, 7, and 10 days prior to the gas exchange measurements. In comparison with some Acer and Quercus photosynthetic response curves having a shorter left side at Tleaf < Topt, most of the photosynthetic response curves of the poplars had a regular parabolic shape.
Mean values (±SE) of the parameter “b” (Eq. 1) determining a spread of photosynthetic temperature response curves of Populus deltoides and Populus tremuloides in each locations separately and pooled
0.065 ± 0.019
0.058 ± 0.015
0.047 ± 0.017
0.026 ± 0.005
0.095 ± 0.033
0.035 ± 0.014
0.044 ± 0.011
0.009 ± 0.002
Mean “b” for the species
0.064 ± 0.012
0.034 ± 0.007
The provenances of A. rubrum and Q. rubra originating from modestly colder growing season thermal conditions showed lower photosynthetic temperature optima (Topt) than those from warmer conditions, even after growing for many years together in a common garden (Fig. 1). These results confirmed the evidence of significant ecotypic thermal adaptation of photosynthesis in trees (shown previously for seedlings or saplings: Fryer and Ledig 1972; Berry and Björkman 1980; Cavieres et al. 2000; Cunningham and Read 2002), in this case being stronger than any tendency for long-term convergent acclimation. However, in natural conditions, it was impossible to separate genetic differences from differences in acclimation to prevailing temperatures in common garden. The differences observed among Topt of the contrasting taxa could be due to either fixed genotypic differences (e.g., adaptive Topt), or genotypic differences in their decadal, seasonal and short-term acclimation of Topt, or all of these (e.g., Battaglia et al. 1996; Gunderson et al. 2010).
Intraspecific variation of Topt has rarely been investigated among deciduous broadleaved trees populations and those experiments were conducted in greenhouse, in open top chambers, or climate chambers after a short period of acclimation to growth temperature compared with our study (Ledig and Korbobo 1983; Gunderson et al. 2000; Weston and Bauerle 2007). It should be emphasized that in contrast to other studies where Topt for tree species and/or ecotypes was determined (Battaglia et al. 1996; Cavieres et al. 2000; Gunderson et al. 2000; Cunningham and Read 2002; Dillaway and Kruger 2010; Gunderson et al. 2010; Wieser et al. 2010), our results are the first providing evidence of intraspecific differences in Topt for field grown deciduous broadleaved trees growing over the long-term in a common environment. In contrast to our findings, though, Gunderson et al. (2000) did not find significant ecotypic differences in temperature optima between the seedlings populations of Acer saccharum acclimated for 3 years to growth temperature in open top chambers. However, our results are in agreement with those of Ferrar et al. (1989) who found differences in Topt between populations from contrasting thermal environments in Eucalyptus pauciflora Sieber ex A. Spreng. In earlier studies, ecotypic variation in Topt was found in evergreen trees originated from different altitudes (Fryer and Ledig 1972; Ferrar et al. 1989; Cavieres et al. 2000; Wieser et al. 2010), but not in deciduous trees from different altitudes or latitudes (Ledig and Korbobo 1983; Gunderson et al. 2000). However, Weston and Bauerle (2007) showed differences in photosynthesis between the thermally contrasting genotypes of A. rubrum. The Florida A. rubrum genotype had the higher Topt than the Minnesota genotype.
Topt for the study provenances ranged between 17.1 and 21.8°C and only the lowest value obtained for the northern A. rubrum provenance diverged from the range 20–25°C compiled for temperate deciduous tree species in Larcher (2003). Gunderson et al (2010) observed Topt ranging from 23 to 30°C for a Q. rubra provenance with a warmer thermal origin and environment than our provenances. Moreover, Gunderson et al. (2010) found in Liquidamber styraciflua that in autumn Topt averaged 17.3°C, and in the warmest period (June) Topt was 30.9°C indicating strong plasticity in the optimum temperature. Hence it is not surprising to see differences in Topt values for a single species across sites and seasons.
Topt of all four taxa were low in comparison with mean daily maximal temperatures July–September at the study site in the year of measurements or at their sites of origin. However, Topt was higher than the average growth temperature (May–Sept) at the sites of origin. In our study, generally, Topt was lower than maximal daytime summer temperatures. This is not necessarily sub-optimal because if Topt were scaled to the daily temperature maximum or higher, trees might photosynthesize at the highest level only for a very short period within a day when temperature was close to a daily maximum. Nonetheless, we cannot exclude that the low Topt of the provenances growing in common garden might be due to a lack of humidity control in the leaf chamber. A rising leaf temperature accompanied by an increase in VPD, stomatal closure, thus a decrease in Ci and net CO2 assimilation rate at higher temperatures might lead to a lower Topt. Even though rising VPD might be a contributing reason for the low Topt of oaks and maples, it did not influence significantly the differences in Topt between the study provenances because they did not differ in the mean slopes and constants of linear regression between the Tleaf and Ci (Table 5), and VPD. All the temperature response curves of the compared provenances were generated at the similar VPD range. Gunderson et al. (2010) observed on few occasions, Ci decreased at the highest cuvette temperatures. Even then, dividing A by Ci revealed nonstomatal decreases in A, i.e., the rate of A per unit CO2 decreased at high temperatures, and the decline was far steeper than that produced from an A/Ci relationship. Decreasing Ci was thus not the cause of the decline in A above Topt, nor did it account for treatment differences, as VPD correlated with cuvette temperature, without regard to treatment. This is in agreement with our observations.
When P. deltoides and P. tremuloides sympatric populations were compared along a geographical gradient within the region of their overlapping ranges, genetic differences between the species with respect to their photosynthetic temperature responses outweighed any possible effects of adaptation and acclimation to local thermal environments. The lower Topt was shown by P. tremuloides—which has a large natural range area extended to northern Canada—compared with the more southern ranging temperate P. deltoides (Table 1).
Gunderson et al. (2010) also found significant differences in Topt among deciduous tree species differing in ranges. The differences in Topt to some extent corresponded to their thermal origins: two species from cooler, more northerly habitats had slightly lower Topt values than the others. It is noteworthy that in our study the difference in Topt of around 3°C was found in the area where the ranges of P. tremuloides and P. deltoides overlapped. However, in common gardens established along a 900 km latitudinal transect from northern Wisconsin to southern Illinois, these poplars did not differ in Topt (Dillaway and Kruger 2010). Although it is impossible to completely reconcile these differences, the inconsistency with our results could be caused by differences in “Materials and methods” section. We used individuals growing in natural conditions and measured gas exchange in situ in four different locations, whereas poplars used by Dillaway and Kruger (2010) originated from one seed source, they had initially been grown in a greenhouse, and then they were planted in watered and fertilized common gardens.
Our results are one of the first providing evidence of interspecific differences in Topt for field grown deciduous broadleaved trees within the region of the study species overlapping ranges. Both P. tremuloides and P. deltoides have large geographic ranges, therefore we hypothesized that their ecotypes at the study sites would be both adapted to and acclimated to the local thermal niches. However, we did not find differences in Topt between trees from the different localities. This may have been because the study sites were not far apart. A lack of ecotypic acclimation to local growth temperature with respect to Topt might also be caused by a high variation of uncontrolled factors which influenced photosynthetic temperature responses, such as thermal history, radiant energy received by the plant before gas exchange measurements, and other interacting factors, such as drought, water vapor deficit, nutrient availability, and leaf age (Fryer and Ledig 1972; Reich et al. 1998; Zhou et al. 2007). Additionally, the more northern species had higher Aopt than the southern species at the northern but not at the southern sites which might suggest better adaptation of P. tremuloides to colder climate when compared with the co-occurring species.
In our study, significant correlations were found between mean, mean maximal, and mean minimal temperatures observed for the 1, 3, 5, 7, and 10 days before the gas exchange measurements and Topt of the Acer CWH provenance and between Tmax(1) and Topt of the Quercus Indiana provenance. These relationships found between the moving average temperatures and Topt suggest that short-term thermal acclimation influenced Topt of the study provenances. Similarly, Gunderson et al. (2010) found an even stronger relationship of Topt to temperature history for four deciduous tree species including Q. rubra. Compared with our experiment, these authors gathered Topt from May to November encompassing a longer period with higher temperature differences. In contrast, there was no correlation between the moving average temperatures and Topt of both Populus. The gas exchange measurements in poplars were accomplished within a short period in September which did not encompass seasonal temperature changes. Thus, all the curves of photosynthetic temperature responses for P. deltoids and tremuloides were generated in “cold” period. Short-term or seasonal temperature acclimation was also not found in Eucryphia lucida (Cunningham and Read 2002) and in P. tremuloides, Betula papyrifera, and P. deltoides (Dillaway and Kruger 2010).
The results of our study confirmed that the differences between two closely related species and pairs of provenances with differing climate origins could all arise from genotypic differences, and/or genotypic differences in acclimation of Topt to prevailing temperatures. Both adaptation and acclimation (e.g., Battaglia et al. 1996; Gunderson et al. 2010) of Topt of tree species and ecotypes to thermal growth conditions could have important consequences for their potential responses to global warming. Temperate tree species with high intrinsic Topt and/or acclimation potential may respond positively to increased temperatures. The ability to sustain net CO2 assimilation rates close to maximum across a wider interval of leaf temperature was greater in P. tremuloides than P. deltoides. Hence it is possible that our study species had different plasticity in instantaneous response to temperature. If generalizable, the between-species differences in Topt and in ability to sustain A close to a maximum at a great interval of leaf temperature might influence the responses of P. tremuloides and P. deltoides to global warming at sites where they co-occur. The lower Topt of A could in theory help to constrain the southern limits of its range if no acclimation occurred as temperatures grow warmer. In contrast, temperate P. deltoides with high Topt may positively respond to increased temperatures. However, differences in physiological mechanisms of adaptation and acclimation plasticity of photosynthetic parameters, and enormous uncertainty about these, make it difficult to model plant responses to increased temperatures. Better understanding of patterns of Topt—which would reflect photosynthetic adaptation and ability to adapt to the changing thermal environment—could be used to improve bioclimatic models.
In conclusion, our results provided evidence that provenances of A. rubrum and Q. rubra originated from different thermal environment exhibited persistent differences in photosynthetic temperature responses likely due to their ecotypic adaptation to the thermal conditions of their origin, and that such differences persisted despite years of potential acclimation to a common growth condition and short-term temperature acclimation. A genetic predisposition, also was shown by differences in Topt between sympatric populations of P. deltoides and P. tremuloides, consistent with the generally warmer-ranging species being better matched to warmer temperatures than the northerly distributed one. All three pairs of comparable taxa suggest that climate origins matter to temperature-based photosynthesis, even for plants long acclimated to local sites.
We thank Prof. Jacek Oleksyn for thoughtful comments on the manuscript, Ms Cindy Buschena for excellent technical assistance, and the Wilderness Research Foundation for funding support. Piotr Robakowski was supported by the research scholarship from the Kosciuszko Foundation and Yan Li by the Chinese Government PhD Scholarship.
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.