Decomposition of Calcium Oxalate Crystals in Colobanthus quitensis under CO2 Limiting Conditions

Calcium oxalate (CaOx) crystals are widespread among plant species. Their functions are not yet completely understood; however, they can provide tolerance against multiple environmental stress factors. Recent evidence suggested that CaOx crystals function as carbon reservoirs since its decomposition provides CO2 that may be used as carbon source for photosynthesis. This might be advantageous in plants with reduced mesophyll conductance, such as the Antarctic plant Colobanthus quitensis, which have shown CO2 diffusion limitations. In this study, we evaluate the effect of two CO2 concentrations in the CaOx crystals decomposition and chlorophyll fluorescence of C. quitensis. Plants were exposed to airflows with 400 ppm and 11.5 ppm CO2 and the number and relative size of crystals, electron transport rate (ETR), and oxalate oxidase (OxO) activity were monitored along time (10 h). Here we showed that leaf crystal area decreases over time in plants with 11.5 ppm CO2, which was accompanied by increased OxO activity and only a slight decrease in the ETR. These results suggested a relation between CO2 limiting conditions and the CaOx crystals decomposition in C. quitensis. Hence, crystal decomposition could be a complementary endogenous mechanism for CO2 supply in plants facing the Antarctic stressful habitat.


Introduction
Calcium oxalate (CaOX) is a salt of oxalic acid (C 2 H 2 O 4 ) and calcium (Ca 2+ ) that forms insoluble crystals of diverse morphology [1]. Detected in at least 215 families, CaOx crystals are widespread among plant kingdom [2,3]. Crystals occur in roots, stems, leaves, flowers, fruits and seeds, and within epidermal, ground, and vascular tissues [4]. They are formed in the vacuoles of specialized cells called crystal idioblasts, which possess distinct structure and content from the surrounding cells. In leaves, crystals idioblasts are commonly located within mesophyll and/or bundle sheath extensions [5]. Some plants (including mainly succulents) like Cactus senilis, accumulate CaOx by as much as 85% by dry weight [6,7]. The huge variation in distribution among organs, tissues, and cells among plant species suggests that crystals may have independent origins of formation and multiple functions [8,9].
Recent evidence showed that indeed CaOx crystals represent multifunctional tools which are essential especially under stress conditions [8]. They are dynamic storage systems supplying calcium Plants 2020, 9,1307 3 of 11

Calcium Oxalate Crystal Decomposition
A CO 2 restriction experiment was performed with C. quitensis plants to test whether a CO 2 limiting condition might trigger the decomposition of CaOx crystals. The experimental setup allowed to compare the CaOx crystal dynamics under a limiting CO 2 condition on C. quitensis plants, with CO 2 concentration close to 11.5 ppm, which is below the CO 2 compensation point for this species (25-30 ppm) [22] and ambient 400 ppm CO 2 (control). There were significant main effects for both CO 2 concentration and time. There was a statistically significant interaction between CO 2 concentration and time, where the relative area of crystals decreased as time passed under low [CO 2 ]. Tukey's post hoc test showed that, after ten hours of treatment, the mean of total crystal area per leaf was significantly lower in the 11.5 ppm-CO 2 -concentration group compared to the control group; a significant crystal decomposition was evident (Figure 1a, Figure 2 and Figure S1).

Calcium Oxalate Crystal Decomposition
A CO2 restriction experiment was performed with C. quitensis plants to test whether a CO2 limiting condition might trigger the decomposition of CaOx crystals. The experimental setup allowed to compare the CaOx crystal dynamics under a limiting CO2 condition on C. quitensis plants, with CO2 concentration close to 11.5 ppm, which is below the CO2 compensation point for this species (25-30 ppm) [22] and ambient 400 ppm CO2 (control). There were significant main effects for both CO2 concentration and time. There was a statistically significant interaction between CO2 concentration and time, where the relative area of crystals decreased as time passed under low [CO2]. Tukey's post hoc test showed that, after ten hours of treatment, the mean of total crystal area per leaf was significantly lower in the 11.5 ppm-CO2-concentration group compared to the control group; a significant crystal decomposition was evident (Figures 1a, 2, and S1). . The horizontal line indicates the mean and length of each whisker indicates the interquartile range (IQR); n = 14, different letters represent statistically significant differences (Twoway ANOVA; p < 0.05). In (a), plants were kept over time under constant light in airtight chambers injected either with ambient air (Control, 400 ppm CO2, white boxes) or filtered air with soda lime (CO2 limitation, 11.5 ppm CO2, gray boxes). (b), Diurnal fluctuations in CaOx crystal area in C. quitensis leaves. Plants were grown under optimal growth conditions. Yellow background denotes light hours, while gray background dark hours. Error bars denote SE of mean; n = 15. Different letters represent statistically significant differences between time of the day (one-way ANOVA; p < 0.05).
The imposed carbon limitation, generated by the low [CO2], triggered a significant reduction of CaOx crystals areas in the C. quitensis leaves, which is in accordance with our proposed hypothesis. The results are also in agreement with those reported by   [10], where a condition that boosted the stomatal closure and limited the availability of CO2 (e.g., exogenous application of abscisic acid or drought stress), increased the CaOX crystal decomposition in the leaves of Amaranthus hybridus. Furthermore, the number of crystals per leaf area decreased significantly in the low [CO2] treatment after 10 h ( Figure S2). Therefore, the observed reduction of total crystal area per leaf (%) was the sum of both the reduction in size and number of crystals per leaf area. Consequently, given that complete crystal formation (maximum size) and maximum number of crystals per leaf are observed early in leaf development [25]; the observed differences in the crystals per leaf area should be a result of full crystal decomposition and not of differences in the number of idioblasts.
The CaOx crystal decomposition obtained here was also similar to that observed by Tooulakou et al. (2019) [26], where they showed that A. hybridus plants that were grown under CO2 restrictive . The horizontal line indicates the mean and length of each whisker indicates the interquartile range (IQR); n = 14, different letters represent statistically significant differences (Two-way ANOVA; p < 0.05). In (a), plants were kept over time under constant light in airtight chambers injected either with ambient air (Control, 400 ppm CO 2 , white boxes) or filtered air with soda lime (CO 2 limitation, 11.5 ppm CO 2 , gray boxes). (b), Diurnal fluctuations in CaOx crystal area in C. quitensis leaves. Plants were grown under optimal growth conditions. Yellow background denotes light hours, while gray background dark hours. Error bars denote SE of mean; n = 15. Different letters represent statistically significant differences between time of the day (one-way ANOVA; p < 0.05).
The imposed carbon limitation, generated by the low [CO 2 ], triggered a significant reduction of CaOx crystals areas in the C. quitensis leaves, which is in accordance with our proposed hypothesis. The results are also in agreement with those reported by   [10], where a condition that boosted the stomatal closure and limited the availability of CO 2 (e.g., exogenous application of abscisic acid or drought stress), increased the CaOX crystal decomposition in the leaves of Amaranthus hybridus. Furthermore, the number of crystals per leaf area decreased significantly in the low [CO 2 ] treatment after 10 h ( Figure S2). Therefore, the observed reduction of total crystal area per leaf (%) was the sum of both the reduction in size and number of crystals per leaf area. Consequently, given that complete crystal formation (maximum size) and maximum number of crystals per leaf are observed early in leaf development [25]; the observed differences in the crystals per leaf area should be a result of full crystal decomposition and not of differences in the number of idioblasts.
The CaOx crystal decomposition obtained here was also similar to that observed by Tooulakou et al. (2019) [26], where they showed that A. hybridus plants that were grown under CO 2 Plants 2020, 9, 1307 4 of 11 restrictive conditions exhibited a considerable reduction in the leaf CaOx crystal volume over time, compared to the control group. Both Tooulakou et al. reports (2016;2019) consider that leaf CaOx crystals act as dynamic carbon reservoirs, capable of providing CO 2 for photosynthesis when the entry of atmospheric CO 2 into the mesophyll is limited by an environmental factor. Therefore, the CaOx crystal decomposition that is observed during the day might be associated to carbon requirements for photosynthesis [10,12,26].
Plants 2020, 9, x FOR PEER REVIEW 4 of 11 conditions exhibited a considerable reduction in the leaf CaOx crystal volume over time, compared to the control group. Both Tooulakou et al. reports (2016;2019) consider that leaf CaOx crystals act as dynamic carbon reservoirs, capable of providing CO2 for photosynthesis when the entry of atmospheric CO2 into the mesophyll is limited by an environmental factor. Therefore, the CaOx crystal decomposition that is observed during the day might be associated to carbon requirements for photosynthesis [10,12,26]. Moreover, the observation of the leaves crystals for 24 h, under optimal growth conditions, showed that these structures display diurnal fluctuations, similar to those reported in A. hybridus and Oxalis corniculate [12,27]. According to Figure 1b, during the first hours of light the mean crystal area per leaf kept stable. However, by 20:00 to 00:00 h the mean crystal area decreased significantly, while during the dark hours (00:00 to 08:00), these structures undergo a full recovery. During light time, and therefore the period where the plant is photosynthetically active, the crystals undergo a decomposition process. This process could be associated with supplementing-through the supply of subsidiary CO2 released from crystals-the CO2 requirement of this plant, since this species particularly suffers from a strong limitation in the acquisition of environmental CO2 [22].
Throughout dark hours, as there are no electrons (from light-dependent reactions) to fix carbon, the plants could restore the crystals. The recovery of the oxalate can occur through several metabolic pathways; however, its origin is exclusively biological [8]. Diurnal fluctuations evidence that crystals decomposition is not always associated to an environmental stress per se in Colobanthus quitensis, and more a complementary process that could be suppling subsidiary CO2 to its daily cycle. The obtained results suggest that the CaOX crystals in C. quitensis are a dynamic system, which respond to environmental stimuli, such as limitation of CO2 (Figure 1a), and fluctuates in a daily course ( Figure  1b).

Chlorophyll Fluorescence and Oxalate Oxidase Activity Measurements
Tooulakou et al. (2016) also showed that crystal decomposition was accompanied by boosted oxalate oxidase enzymatic activity (OXO; transforms oxalate into CO2) [10]. The enzymatic analysis for oxalate oxidase activity on C. quitensis leaves showed that there were significant main effects for Moreover, the observation of the leaves crystals for 24 h, under optimal growth conditions, showed that these structures display diurnal fluctuations, similar to those reported in A. hybridus and Oxalis corniculate [12,27]. According to Figure 1b, during the first hours of light the mean crystal area per leaf kept stable. However, by 20:00 to 00:00 h the mean crystal area decreased significantly, while during the dark hours (00:00 to 08:00), these structures undergo a full recovery. During light time, and therefore the period where the plant is photosynthetically active, the crystals undergo a decomposition process. This process could be associated with supplementing-through the supply of subsidiary CO 2 released from crystals-the CO 2 requirement of this plant, since this species particularly suffers from a strong limitation in the acquisition of environmental CO 2 [22].
Throughout dark hours, as there are no electrons (from light-dependent reactions) to fix carbon, the plants could restore the crystals. The recovery of the oxalate can occur through several metabolic pathways; however, its origin is exclusively biological [8]. Diurnal fluctuations evidence that crystals decomposition is not always associated to an environmental stress per se in Colobanthus quitensis, and more a complementary process that could be suppling subsidiary CO 2 to its daily cycle. The obtained results suggest that the CaOX crystals in C. quitensis are a dynamic system, which respond to environmental stimuli, such as limitation of CO 2 (Figure 1a), and fluctuates in a daily course (Figure 1b).

Chlorophyll Fluorescence and Oxalate Oxidase Activity Measurements
Tooulakou et al. (2016) also showed that crystal decomposition was accompanied by boosted oxalate oxidase enzymatic activity (OXO; transforms oxalate into CO 2 ) [10]. The enzymatic analysis for oxalate oxidase activity on C. quitensis leaves showed that there were significant main effects for both CO 2 concentration and Time, and also there was a statistically significant interaction between both effects on the oxalate oxidase activity. Tukey's post hoc test showed that there were no significant differences between treatments at 0 h or 5 h of the test. However, statistically significant differences between the two studied groups were observed after 10 h (Figure 3a). The observed OXO activity was similar to that reported for Podophyllum peltatum after water stress [10]. The obtained results allow us to observe an association between the increase in oxalate oxidase activity and the decomposition of the CaOx crystals.
both effects on the oxalate oxidase activity. Tukey's post hoc test showed that there were no significant differences between treatments at 0 h or 5 h of the test. However, statistically significant differences between the two studied groups were observed after 10 h (Figure 3a). The observed OXO activity was similar to that reported for Podophyllum peltatum after water stress [10]. The obtained results allow us to observe an association between the increase in oxalate oxidase activity and the decomposition of the CaOx crystals.
Chlorophyll fluorescence (ChlF) measurements showed that the electron transport rate (ETR) of the CO2-limited plants decreased significantly compared to the control group (Figure 3b). There were significant main effects for CO2 concentration, but not for Time. In addition, there was a statistically significant interaction between both effects on the ETR. However, despite the statistical differences, the plants exposed to low [CO2] for 10 h still maintain high ETR values (~50 μmol e m −2 s −1 ). This ETR values are similar to those reported before for this species under different temperatures and light intensity [22,28]. Following the AP hypothesis, a plant that does not have the AP mechanism and is exposed to a low [CO2], should experience an inhibition of photosynthesis also evident by a reduced ETR. It is known that in C3 plants exposed to low [CO2] the rate of carboxylation of Rubisco is reduced and consequently, the net photosynthetic rates are also affected due to substrate limitations; a situation that boosts photorespiration rates [29]. However, the obtained results showed just a 25% reduction of ETR in C. quitensis plants exposed to low [CO2] ( Figure S3). The linearity between ETR and net CO2 assimilation is commonly absent in C3 species, especially due the existence of alternative electron sinks [30]. Furthermore, under excess light, reducing equivalents from photosynthetic electron transport (NADPH) are exported from the chloroplasts to the cytosol, via malate/oxaloacetate shuttle, and the mitochondrial non-phosphorylating pathways may facilitate the dissipation of these excess reductants in the cell [31]. Therefore, the occurrence of sufficient ETR values alone is not a satisfactory indication that AP is responsible for the use of electrons because there are other alternative electron sinks, such as photorespiration and mitochondrial respiratory chain. In C. quitensis, the photosynthetic electron transport is insensitive to variations in oxygen concentration under non-photorespiratory conditions, indicating that electron transport to oxygen (Mehler reaction) is negligible [32]. However, it has also been shown that under low CO2 availability, the relationship between ETR and Gross photosynthesis (AG) in C. quitensis leaves present high values, indicative of enhanced photorespiration rates [22]. Chlorophyll fluorescence (ChlF) measurements showed that the electron transport rate (ETR) of the CO 2 -limited plants decreased significantly compared to the control group (Figure 3b). There were significant main effects for CO 2 concentration, but not for Time. In addition, there was a statistically significant interaction between both effects on the ETR. However, despite the statistical differences, the plants exposed to low [CO 2 ] for 10 h still maintain high ETR values (~50 µmol e m −2 s −1 ). This ETR values are similar to those reported before for this species under different temperatures and light intensity [22,28]. Following the AP hypothesis, a plant that does not have the AP mechanism and is exposed to a low [CO 2 ], should experience an inhibition of photosynthesis also evident by a reduced ETR. It is known that in C3 plants exposed to low [CO 2 ] the rate of carboxylation of Rubisco is reduced and consequently, the net photosynthetic rates are also affected due to substrate limitations; a situation that boosts photorespiration rates [29]. However, the obtained results showed just a 25% reduction of ETR in C. quitensis plants exposed to low [CO 2 ] ( Figure S3).
The linearity between ETR and net CO 2 assimilation is commonly absent in C3 species, especially due the existence of alternative electron sinks [30]. Furthermore, under excess light, reducing equivalents from photosynthetic electron transport (NADPH) are exported from the chloroplasts to the cytosol, via malate/oxaloacetate shuttle, and the mitochondrial non-phosphorylating pathways may facilitate the dissipation of these excess reductants in the cell [31]. Therefore, the occurrence of sufficient ETR values alone is not a satisfactory indication that AP is responsible for the use of electrons because there are other alternative electron sinks, such as photorespiration and mitochondrial respiratory chain. In C. quitensis, the photosynthetic electron transport is insensitive to variations in oxygen concentration under non-photorespiratory conditions, indicating that electron transport to oxygen (Mehler reaction) is negligible [32]. However, it has also been shown that under low CO 2 availability, the relationship between ETR and Gross photosynthesis (A G ) in C. quitensis leaves present high values, indicative of enhanced photorespiration rates [22].

Calcium Oxalate Crystal Decomposition under Non-Photorespiratory Conditions
In order to eliminate the effect of photorespiration as an alternative electron sink on the observed level of ETR under the 11.5 ppm [CO 2 ] treatment (50 µmol e m −2 s −1 ) (Figure 3b), as well to reduce the mitochondrial respiratory CO 2 efflux, and the putative contribution of mitochondria electron chain consuming chloroplast redox power, a second individual experiment under non-photorespiratory conditions (100% N 2 ) was performed. The C. quitensis plants showed a significant reduction in the crystals area per leaf between the beginning and the end of the treatment after 10 h under 100%-N 2 atmosphere in the glass container (Figure 4a), as well as a high percentage of crystal decomposition ( Figure S1). However, under non-photorespiratory conditions, the ETR values decreased intensely after 2 h of treatment but no further reduction was observed, and until the end of the treatment, the ETR values were kept constant close to~20 µmol e m −2 s −1 (Figure 4b). The difference of ETR between low CO 2 and 100% N 2 was about 30 µmol e m −2 s −1 , about 40% of total ETR; this is the putative contribution of photorespiration, mitochondrial respiration and Mehler reaction as alternative electron sinks to ETR in C. quitensis leaves. This is consistent with a high contribution of oxygen as an alternative electron sinks observed in other plant species [33,34]. Therefore, as previously reported by Saez et al. (2017) [22], it seems that photorespiration is enhanced in this species, which may help to counteract the harmful consequences that are generated as a result of a limited carbon assimilation. Photorespiration could also play a beneficial role in the dynamic and fast response of photosynthetic metabolism under CO 2 limitations, as has been observed in Arabidopsis thaliana [35].

Calcium Oxalate Crystal Decomposition under Non-Photorespiratory Conditions
In order to eliminate the effect of photorespiration as an alternative electron sink on the observed level of ETR under the 11.5 ppm [CO2] treatment (50 μmol e m −2 s −1 ) (Figure 3b), as well to reduce the mitochondrial respiratory CO2 efflux, and the putative contribution of mitochondria electron chain consuming chloroplast redox power, a second individual experiment under non-photorespiratory conditions (100% N2) was performed. The C. quitensis plants showed a significant reduction in the crystals area per leaf between the beginning and the end of the treatment after 10 h under 100%-N2 atmosphere in the glass container (Figure 4a), as well as a high percentage of crystal decomposition ( Figure S1). However, under non-photorespiratory conditions, the ETR values decreased intensely after 2 h of treatment but no further reduction was observed, and until the end of the treatment, the ETR values were kept constant close to ~20 μmol e m −2 s −1 (Figure 4b). The difference of ETR between low CO2 and 100% N2 was about 30 μmol e m −2 s −1 , about 40% of total ETR; this is the putative contribution of photorespiration, mitochondrial respiration and Mehler reaction as alternative electron sinks to ETR in C. quitensis leaves. This is consistent with a high contribution of oxygen as an alternative electron sinks observed in other plant species [33,34]. Therefore, as previously reported by Saez et al. (2017) [22], it seems that photorespiration is enhanced in this species, which may help to counteract the harmful consequences that are generated as a result of a limited carbon assimilation. Photorespiration could also play a beneficial role in the dynamic and fast response of photosynthetic metabolism under CO2 limitations, as has been observed in Arabidopsis thaliana [35]. Despite having neither atmospheric CO2 nor O2 as final electron acceptors, C. quitensis plants were able to maintain a stable ETR value for about 10 h of stress. Theoretically, four electrons are required to reduce 2NADP + to NADPH to fix one CO2 [36]. In practice, because not all the electron really flows linearly and alternative electron sinks are active, about 8-12 electrons are required per CO2 fixed. Therefore, with around ~20 μmol e m −2 s −1 of ETR, the C. quitensis plants could hypothetically fix about 2 μmol CO2 m −2 s −1 , or even higher (about 5 μmol CO2 m −2 s −1 ) if we consider that under low oxygen the main alternative electron sinks, photorespiration, and Mehler reaction are constraint. If these CO2 molecules are effectively being supplied by the CaOx crystals decomposition Despite having neither atmospheric CO 2 nor O 2 as final electron acceptors, C. quitensis plants were able to maintain a stable ETR value for about 10 h of stress. Theoretically, four electrons are required to reduce 2NADP + to NADPH to fix one CO 2 [36]. In practice, because not all the electron really flows linearly and alternative electron sinks are active, about 8-12 electrons are required per CO 2 fixed. Therefore, with around~20 µmol e m −2 s −1 of ETR, the C. quitensis plants could hypothetically fix about 2 µmol CO 2 m −2 s −1 , or even higher (about 5 µmol CO 2 m −2 s −1 ) if we consider that under low oxygen the main alternative electron sinks, photorespiration, and Mehler reaction are constraint. If these CO 2 molecules are effectively being supplied by the CaOx crystals decomposition needs to be probed.
Alarm photosynthesis could be a process that may allow plants maintain a baseline level of photosynthesis when face stress situations. This mechanism is advantageous as a quenching regulator for the energy excess accumulated from the electron transport chain, when the light-dependent reactions are not in pace with photosynthetic CO 2 assimilation from the atmosphere [10]. To this respect, it is possible that some unknown molecules could be involved as final electron acceptors in C. quitensis plants under a CO 2 limiting condition, among them, the CO 2 released from the CaOx crystals decomposition.
The multiple functional role(s) of CaOx crystals in plants is not yet well understood and even some researchers have doubts about the functionality of the bioavailable calcium stocks in plants [13]. However, the obtained results generate a contribution on the understanding of the functions that have been attributed to CaOx crystals, and their relation with the AP mechanism, even though more evidence is still required to ensure that CO 2 release by CaOx decomposition is being used for photosynthesis in C. quitensis.
In this study, we found that Colobanthus quitensis plants exposed to a CO 2 limitation significantly increased the CaOx crystal decomposition, as well as the oxalate oxidase activity in its leaves. This means that under stress conditions, the crystal decomposition could provide CO 2 molecules to the mesophyll tissue [8]. In parallel, ETR decreased but remained stable when compared to the control group. Moreover, under non-photorespiratory conditions a significant CaOx crystal decomposition was also observed, whereas ETR decreased around 40% but was still adequate for the maintenance of a baseline level of photosynthesis if required.
For the Antarctic plant C. quitensis, alarm photosynthesis could play an important role as a complementary endogenous mechanism that could facilitate a CO 2 supply given the limitations in the CO 2 diffusion that have been widely studied. Alarm photosynthesis is considered as a process that may enable the prevention of water losses when plants are under stressful conditions such as drought or strong winds coupled with low relative humidity. Further investigations on other extreme environment plant species is envisioned besides the genetic mechanisms of alarm photosynthesis, especially with focus on the CaOx genetic process of decomposition and regeneration (biosynthetic-degradation pathways), as well as the role of calcium ion during crystal recycling cycles, will allow a more detailed understanding of plant responses to intense drought scenarios. Future research should also focus on further understanding the diurnal fluctuations that have been observed in this plant, its relationship with daily hours and exposure to light. Likewise, attention should be paid to the description and characterization of the oxalate oxidase protein/gene and its regulation, not only because of their participation in the crystal decomposition processes, but also because of their association with stress tolerance processes. (photosynthetically active radiation) of 200 µmol m −2 s −2 and ambient air conditions (approximately 400 ppm CO 2 )).

Experimental Design and Sample Collection
The individual plant cushions were placed inside a transparent airtight borosilicate container and were supplied with air (7.0 L/min) either at ambient (400 ppm, control) or low [CO 2 ] (11.5 ppm, treatment) ( Figure S4). A soda lime scrubber was used to achieve the low [CO 2 ], which was continuously monitored by an infrared gas analyzer (IRGA-LI−6400XT, LI-COR Inc., Lincoln, NE, USA). During the experiment (10 h, from 08:00 to 18:00) plants were kept under 200 µmol m −2 s −2 PAR and 16 • C temperature. Leaf samples (24) were collected for crystal decomposition measurements. Additionally, 20 mg of leaves from 5 plants were collected at time 0, 5, and 10 h for oxalate oxidase activity determination.
A second independent group of plants was kept at optimal growth conditions (temperature 16 • C, PAR of 200 µmol m −2 s −2 and ambient air conditions (approximately 400 ppm CO 2 )) for monitoring the leaves crystals for 24 h. Plants were exposed to 16 h of light (from 08:00 to 23:59) and 8 h of dark (00:00 to 07:59). Leaf samples (15) were collected for crystal decomposition measurements every four hours.

Chlorophyll Fluorescence
The electron transport rate (ETR) of C. quitensis plants exposed to both [CO 2 ] was measured in vivo during the CO 2 -restrictive experiment. Chlorophyll fluorescence measurements were performed using a Maxi-Imaging-PAM Chlorophyll Fluorimeter (Walz, Effeltrich, Germany). Ten areas of interest (AOI) were selected from each cushion and the ETR was calculated according: ETR = Φ PSII ·PAR·αL·(PSII/PSI); where Φ PSII is the quantum efficiency of the photosystem II (PSII), PAR is the photosynthetically active radiation, αL the leaf absorptance (0.73 for control leaves, 0.68 for CO 2 Limitation), and PSII/PSI the distribution of absorbed energy between the two photosystems (assumed to be 0.5). Φ PSII was measured with a Saturation Pulse (6000 µmol photons m −2 s −1 , 800 ms) applied after 3 min illumination at 750 µmol quanta m −2 s −1 of actinic light (AL) ( Figure S5). The leaf absorptance was directly measured in the plants using the Maxi-Imaging PAM as described by Saéz et al., 2018 [37]. Briefly, successive illumination of the samples with red (R) and near infrared (NIR) light and the capture of each remission image were used by the equipment software to calculate pixel by pixel as follows: Abs = 1 − R/NIR. The minimum number of cushions used for the analysis (n) was at least 5.

Measurements of Crystal Degradation
The collected C. quitensis leaves were bleached in sodium hypochlorite solution (5% p/p) according to   [10]. Briefly, whole mature leaves were put in an aqueous solution of commercial bleach for 48 h until full depigmentation. Depigmented leaves were rinsed with abundant distillated water and then put between two microscope slides. Samples were observed under an optical microscope adapted with a polarizing filter at 10× magnification Leica DM750-Camera Leica ICC50W (Leica Microsystems, Wetzlar, Hesse, Germany). Several images were taken per leaf, covering the total leaf area. The area of each crystal was calculated by digital image analysis (ImageJ-Fiji v 2.0.0-rc69/1.52i) [38]. For each individual leaf, several images were taken comprising the total leaf area. Each individual image was analyzed as follows: (1) image was converted to 8 bits, (2) 8 bits image was converted to Mask, and (3) the tool "Analyzing Particles" was used to count and measure all crystals area in the picture using the following parameters: Size 400-5000 pixel 2 , Circularity 0.35-1.00. The total counts (crystals area) of all images from an individual leaf were sum together to obtain the total area of crystals per leaf. Then, this value is divided by the total leaf area to obtain a ratio area crystals/area leaf. The minimum number of leaves used for the analysis at each time (n) was at least 14.

Oxalate Oxidase Activity Determination
The activity of the oxalate oxidase enzyme was determined at time 0, 5, and 10 h of the CO 2 restrictive experiment using the Oxalate Oxidase Activity Assay Kit (BioVision, Inc., Milpitas, CA, USA). Assays were performed following kit protocol using 20 mg of C. quitensis fresh leaves from 5 plants (n = 5). In this test, the decomposition of oxalate by an oxalate oxidase release hydrogen peroxide, which generates a fluorescent signal directly proportional to the amount of active oxalate oxidase present in samples.

Statistical Analysis
Two-way analysis of variance (ANOVA) at a 95% level of significance (p < 0.05) were applied using JASP software (Version 0.13.1) [39] to assess the effects of both time and CO 2 concentration. Tukey Post Hoc Test was carried out in those cases where ANOVA was significant. One-way ANOVA was used to analyze one factor multiple comparation and Independent samples t-test were applied for simple comparison. The assumption of data normality, and homoscedasticity were tested with the Shapiro-Wilk and Levenes's test (data regarding all the statistical tests can be found in Tables S1-S8) Supplementary Materials: The following are available online at http://www.mdpi.com/2223-7747/9/10/1307/s1, Figure S1: Decomposition percentage of CaOx crystals in C. quitensis leaves under adequate (Control, 400 ppm CO 2 ), low CO 2 (CO 2 limitation, 11 ppm CO 2 ) and non-photorespiratory conditions (100% N 2 , 2 ppm CO 2 ) at the end of the treatment. The horizontal line indicates the mean and length of each whisker indicates the interquartile range (IQR); n = 5. Different letters denote significant differences between groups (one-way ANOVA; p < 0.05). Figure S2: CaOx crystals number counted in whole C. quitensis leaves under adequate (Control, 400 ppm CO 2 ) or low (Treatment, 11 ppm CO 2 ) CO 2 concentration. The horizontal line indicates the mean and length of each whisker indicates the interquartile range (IQR); n = 14. Different letters represent statistically significant differences (two-way ANOVA; p < 0.05). Figure S3: Light responses curves (ETR/ETR max %) of Colobanthus quitensis under adequate (Control, 400 ppm CO 2 ) and low CO 2 (CO 2 limitation, 11 ppm CO 2 ). Measurements were performed at 16 • C. Error bars denote SD of mean; n = 6. Figure S4: Experimental setup: System used to run the CO 2 restriction experiment on C. quitensis plants. Figure S5: Quantum yield of PSII (ϕPS II) in C. quitensis leaves under adequate (Control, 400 ppm CO 2 ) or low (Treatment, 11 ppm CO 2 ) CO 2 concentration. Error bars denote SD of mean; n = 15. Table S1: Two-way independent ANOVA results for Figure 1a data. Table S2: One-way independent ANOVA results for Figure 1b data. Table S3: Two-way independent ANOVA results for Figure 3a data. Table S4: Two-way independent ANOVA results for Figure 3b data. Table S5: Independent samples t-test results for Figure 4a data. Table S6: One-way independent ANOVA results for Figure 4b data. Table S7: One-way independent ANOVA results for Figure S1 data. Table S8: Two-way independent ANOVA results for Figure S2