Anthocyanins Carbon Gain And Different Light Environments Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Abstract The possible functions of color change from green to red during autumn months due to anthocyanin synthesis in the acclimation to changes of light environment was evaluated in Acer platanoides and A. saccharum. Seedlings were grown in a greenhouse for 8 month in either 21% (medium light, M) or 4.9% (low light, L) of ambient irradiance, and, in the following month, subsets of the medium light-grown seedlings were transferred to full light (high light, T). Gas exchange, chlorophyll fluorescence imaging, pigments, antioxidant activity, nitrogen (N) resorption were examined in seedlings of the three treatments (M, L, T) at three leaf color (green, yellow and red) stages. Leaf ontogeny during the three leaf color stages was characterized by a progressive anthocyanin pigments accumulation. Across the three leaf color stages, the reddest leaves of an individual occurred in the sunniest treatment. In comparison with plants grown in medium light, plants transferred from medium light to full light had lower chlorophyll, carotenoid and chlorophyll fluorescence parameters (Fv/Fm, qP, ΦPSII, NPQ and ETR) at yellow and red leaf stages. This demonstrates there is evident damage to the leaf photosystem, and light climate induced reductions in fluorescence indicate the higher levels of photoinhibition in the seedlings exposed to high light. Even so, we suggest that anthocyanins reduces the risk of photo-oxidative damage to leaf cells as they senesce, otherwise, plants in the high light would become even more photoinhibited. We also noted that there was a positive correlation (r =0.63, p=0.07 for Norway maple; r =0.69, p=0.04 for sugar maple) between N resorption and anthocyanin levels in leaves. As a consequence we proposed that leaf redness is a way to squeeze a bit more photosynthesis out of the leaves before winter, allowing for the resorption of critical foliar nutrients to occur. The diminished photosynthetic capacity and higher N resorption in leaves under high light conditions may indicate the adaptive function of photoprotection would not only be the protection of leaves per se, but also that functional leaves enable a better resorption of nutrients, especially nitrogen. Our results suggest that it could be important to look at the significant functional role of leaf anthocyanins at both the leaf and the plant scale.

Key words: Nitrogen absorption, anthocyanins, chlorophyll, photoprotection.


While the autumn coloration of tree foliage remains a fascinating spectacle every year, the mechanisms and reasons why temperate and boreal deciduous trees get colored are still subject to discussion. At present, there seem to be two main hypotheses on the topic, focusing on the role of antocyanides and on coloration as a signal to herbivores (Archetti et al., 2009). The functional significance of anthocyanin pigments in leaves has received substantial attention in the recent literature (Manetas, 2006; Archetti, 2009). Anthocyanins are thought to minimize photo-oxidative damage by either absorbing green light, thereby reducing the amount of light absorbed by photo-pigments (Lee and Gould, 2002), and/or through neutralizing reactive oxygen species (ROS) directly as antioxidants (Kytridis and Manetas, 2006). Although there is strong experimental support for a photoprotective role of anthocyanins in many plants (Feild et al., 2001; Hughes et al., 2005), it does not seem to be universal as others found no evidence for a photoprotective role (Kyparissis et al., 2007; Esteban et al., 2008). Moreover, in the case of winter leaf reddening in the Mediterranean shrub Cistus creticus, it was shown that the red-leaf phenotype was physiologically weak and less tolerant than the green phenotype to the combination of low temperature and high light in the field (Zeliou et al., 2009). The reasons for these often opposing results remain unclear.

On the other hand, the primary value of leaf senescence to plant fitness is the resorption and reuse of breakdown products (Aerts, 1996). Hoch et al. (2003) proposed a resorption protection hypothesis in which anthocyanins of senescing foliage shade the photosynthetic system and prevent photoinhibition, thereby allowing for enhanced resorption of nutrients, particularly nitrogen (N). Similarly, Feild et al. (2001) emphasized the dual role of anthocyanins in senescing foliage of red osier dogwood (Cornus stolonifera Michx.) as scavengers of ROS and facilitators of nutrient recovery. However, Lee et al. (2003) did not find a better resorption of N in species with red leaves, although they did find an intraspeci¬c correlation (individuals with a higher anthocyanin concentration had more N resorbed). These results therefore remain open to further investigation (Lee et al., 2003).

Lev-Yadun and Holopainen (2009) discussed that while yellow leaf autumn colors prevail in Europe, red colors seem to be more important in Eastern-Asia and North-America. They argued that different glacial extinction rates and evolutionary histories could be responsible for these differences. Higher extinction rates of both trees and their insect herbivores in Europe as opposed to North America and East Asia seem to indicate that red autumn leaves are probably a relict Tertiary adaptation of temperate floras to past climates and herbivore faunas. Norway maple (Acer platanoides L.), is a Eurasian introduced tree species which has invaded the North American range of its native congener, sugar maple (A. saccharum Marsh.).

In this study, we characterized anthocyanins in the A. platanoides and A. saccharum for plants grown in medium light (23% of full light), plants grown in low light (3% of full light) and plants from medium light switched to a full light environment (high light). Some specific questions investigating leaf redness responses to different light environments were addressed (1) plants transferred from medium light to full light have more anthocyanins than plants in medium light and low light since anthocyanin synthesis is known to be inducible under high light, (2) anthocyanins protect red leaves against photoinhibition, (3) anthocyanins elevate radical-scavenging capacity, (4) higher anthocyanins concentrations is associated with higher leaf N resorption.

Materials and methods

Experimental design and treatments

The experiment was conducted at the Montreal Botanical Garden, Quebec, Canada (+45°33.7' - 073°34.3') as a complement to an ongoing experiment on the comparative ecophysiology of Norway and sugar maple in response to light (Paquette et al., submitted) identifying the former species' characteristics that would confer invasiveness. For that larger experiment, Acer platanoides and A. saccharum seedlings were raised from seeds collected from mature trees in Montreal for the former, and from Québec forest nursery for the latter. Seeds were stratified and then sown in humid sand boxes composed of layers of sand, minced leaf litter, 2.5cm extruded polystyrene foam, and a white plastic sheet, for insulation and protection from rodents. Germinants were transferred to 320 mL multi-cell containers and placed at random in their respective light regime for two months. The seedlings were then transferred to larger 6.7 L pots.  Germinated seedlings were raised in dynamic greenhouses under two light levels 21% (medium light, M treatment) and 4.9% (low light, L treatment) of full ambient photosynthetic photon flux (PPF) (measured on September 9, 2009), mimicking conditions found under forest gaps and closed forest understories, respectively. These light levels were obtained by varying the size of roof openings and calibrated using whole-day PPF measurements.In a previous experiment, Paquette et al. (2010) demonstrated the inadequacy of homogenous shade-cloth greenhouses for mimicking forest understories (see that study for more on the calibration of dynamic greenhouses). A sunny day is around 1600 µmol m-2s-1 in Montreal. All seedlings were arranged into four replication blocks, each comprising the two light treatments assigned at random, and the two species.

All seedlings were well watered throughout the experiment and fertilized using Nutricote 20-7-10 type 180 at a rate of 15g per pot. There were 8 seedlings per species and per treatment, which were spatially arranged into four replication blocks (two seedlings per treatment in each block). On august 25, 2009, four plants per species, one from each block, were moved from the 21% of full ambient PPF treatment to full sunlight (8 months after leaf emergence; high light, T treatments).

I think the above section is unclear and may be misinterpreted, especially that we have changed the "design" as mentioned above regarding substrates for the parallel study. Also, here and elsewhere in the text it is impossible to know exactly what your sample size was. So here's a modified version of the above paragraph:

All seedlings were well watered throughout the experiment and fertilized using Nutricote 20-7-10 type 180 at a rate of 15g per pot. On august 25, 2009 (8 months after leaf emergence), we took a total of 24 seedlings from the larger experiment and assigned them to the present study on leaf redness. One seedling per species and per original light treatment (M and L) was chosen from each of the blocks to be part of the present study. These 16 seedlings remained in their original location for the present experiment. To make our high light treatment (T), an additional 8 trees, four per species (one from each block), were taken and moved from the M treatment (21% of full ambient PPF) outdoors and placed about a meter along the eastern walls of their respective greenhouse block, providing high-light conditions (~86% of full sunlight) and protection against dominant winds, similar to a large forest opening.

Measurements of gas exchange, chlorophyll fluorescence imaging, pigments, antioxidant activity, N resorption were performed on green leaf stage (August 25- September 5), yellow leaf stage (September 10-October 3) and red leaf stage (October 12-31). The yellow phase was considered a transition phase.

Pigment determination

Leaf discs were collected for the determination of pigment concentrations. Leaf discs were sampled on all trees at each leaf color stage, and were immediately frozen on dry ice in the field, and then stored at -80 °C until analysis. Frozen discs were ground in 100% acetone with a small amount of quartz sand in a chilled mortar. Chlorophyll (Chl) and total carotenoid (Caro) concentrations were determined using a multiwavelength analysis at 470, 645, 662 and 710 nm (Lichtenthaler and Buschmann, 2001) with a CARY 300 UV-Visible spectrophotometer. For anthocyanin determination, four leaf discs were disrupted in liquid nitrogen and extracted in 1.25 mL of 3M HCl: H2O: MeOH (1:3:16 by vol.) using a tissue homogenizer. Absorbance of anthocyanins at 530 nm was estimated according to Murray and Hackett (1991).

Gas exchange measurements

For each species, photosynthetic rates were determined in one leaf per plant. Photosynthetic rates (A) were determined on September 3 and 4, September 29, October 15, 2009 with a portable photosynthesis system (GFS-3000, Walz, Effeltrich, Germany) at a saturating photon flux (1400 µmol m-2 s-1) and ambient CO2 concentration (~400 ppm). Leaf temperatures (mean ± SD) were 25.0 ± 2.9 on September 3 and 4, 15.2± 0.1 on September 29, and 10.4 ± 0.8 °C on October 15, respectively.

Chlorophyll fluorescence measurements

For each treatment, one fully mature leaf from each tree was randomly selected for the fluorescence measurements. Fluorescence measurements were carried out with an IMAGING-PAM chlorophyll fluorometer (Heinz Walz GmbH, Effeltrich, Germany). Leaves were dark-adapted for at least 30 min prior to the measurements in order to completely reoxidize PSII electron transporters. Fluorescence was measured with relatively weak light pulses (<1 Î¼mol m-2 s-1) at a low frequency (1 Hz) for measurement of minimal fluorescence (Fo). Maximal fluorescence yield of a dark-adapted leaf (Fm) was measured during an 800-ms exposure to a photon flux of approximately 2600 Î¼mol m-2 s-1. Actinic light of a PPF of 400 µmol m-2 s-1 was applied to induce electron transport, and saturating pulses were applied to determine maximal fluorescence yield of a light-adapted leaf (Fm′) levels. All lighting (modulated measuring light, actinic light and saturation pulses for measurement of Fm and for the Fm′ was provided from blue light-emitting diodes (450 nm). When performing a measurement, an area of interest (AOI) with a diameter of 1 cm was selected in the center of the leaf. Maximal PSII quantum yield (Fv/Fm, equivalent to (Fm − Fo)/Fm), effective PSII quantum yield (ΦPSII), non-photochemical quenching (NPQ) and coefficient of photochemical quenching (qP) were the average of the AOI. In addition, their images were simultaneously derived from the IMAGING-PAM software.

Rapid light curve measurements were carried out using 30-s exposures to stepwise increased PPF (1, 24, 54, 103, 265, 532, 599, 831, 1029, 1322, 1617, 2001, and 2603 μmol m-2 s-1). Simultaneously, apparent electron transport rate (ETR) values were estimated as ETR =ΦPSII Ã- 0.5 Ã- 0.84 Ã-

PPF according to White and Critchley (1999), where PPF is the photosynthetic photon flux incident on the leaf, 0.5 is a factor that assumes equal distribution of energy between the two photosystems, and 0.84 is the assumed assumed leaf absorption coefficient (Genty et

al., 1989).

Antioxidant activity assay

Antioxidant activity of leaf extracts was assessed by determining their ability to scavenge 1,1-diphenyl-2-picryl hydrazyl (DPPH), a stable free radical. Leaf discs were sampled at each leaf color stage as above. Reaction mixtures containing 0-100 µL leaf extract and 1.5 mL of 18 µM DPPH in MeOH were diluted with MeOH to a ¬nal volume of 1.6 mL, vortexed, and then held at room temperature for 30 min, after which the absorbance of the mixtures at 517 nm was measured. Antioxidant activity of the leaf extracts was expressed as an effective concentration for radical scavenging (IC50): the concentration of fresh leaf material (mg mL-1) required to produce a 50% reduction in A517 relative to the control mixture to which only methanol was added.

Leaf nitrogen analysis

To examine foliar nitrogen (N) resorption patterns, we collected fully expanded young to medium-aged green leaves. Red leaves that were still attached to branches just before defoliation were sampled on October 25, 2009, when some leaves had fully senesced, falling readily at the? touch. Also, shed leaves were counted at intervals of 6 days. Leaf samples were ground and passed through a 20 mesh screen after being first dried at 70 °C for 36 h. The total concentrations of N were determined by the semi-micro Kjeldahl method. For each treatment, N resorption efficiency was calculated as (Ng - Ns)/ NsÃ-100% in which Ng is the green leaf N concentration and Ns is the senescent leaf N concentration (Sanz-Pérez et al., 2009).

Leaf Nitrogen in cell walls

Leaf proteins can be divided into water-soluble, detergent-soluble, and detergent-insoluble fractions. Water-soluble proteins include soluble enzymes such as Rubisco; the detergent-soluble fraction includes membrane associated proteins such as enzymes and electron transport; and the detergent-insoluble fraction is the proteins in cell walls, which contribute to leaf toughness (Takashima et al., 2004). Total leaf protein content was calculated as the sum of the contents of the 3 protein fractions; the ratio of cell wall proteins to total leaf proteins was also calculated. N content in cell walls was calculated from cell wall proteins with a conversion coefficient (0.16g N g-1 wall proteins). The proportion of leaf N allocated to cell walls was calculated as N content in cell walls / total leaf N.

Statistical analysis

The measurements showed that the parameters tested in our study (gas exchange, chlorophyll fluorescence, pigments, antioxidant activity, N resorption) were unaffected by soil type, hence, we did not account for soil in the statistical analysis. Data were first checked for normality and the homogeneity of variances, and log-transformed to correct deviations from these assumptions when where? needed. For each species, two-way ANOVA for light and leaf stage were conducted for each parameter assessed. A one-way ANOVA was used to examine the significance of the parameter between leaf color stages in the different treatments. For significant light effects, multiple comparisons were also performed to permit separation of effect means using Tukey's post hoc test at a global significance level of P < 0.05. Simple linear regression was used to determine the relationships between N resorption and anthocyanin levels in leaves. All statistical analyses were conducted with SPSS software (SPSS 11.0 for windows, SPSS Inc., Chicago, IL, USA).


Effects of light on pigment contents

Leaf changes occurred during the experimental period as the fluorescence imaging colors of ΦPSII changed to red from the original green (Fig. 1). In Norway maple, high light treatment (T) significantly (p < 0.05) decreased Chl (38-91%) at yellow and red leaf stages (Fig. 2a), and increased anthocyanins (53-95%) across the three leaf stages in comparison with the medium light treatment (M) (Fig. 2c). Compared to the medium light, the high light-induced reductions were 15-62% for Chl across the three leaf color stages in sugar maple, although not significant. By contrast, there was significant (p < 0.05) enhancement (57-100%) for anthocyanins at green and red leaf stages in sugar maple and marginally significant (p = 0.07) at yellow leaf stage. Caro was unaffected by light treatments until red leaf stage for both species (Fig. 2b) when the reductions were 67% for Norway maple and 58% for sugar maple. For both species, IC50 was not significantly affected by light treatments at any leaf color stage (Fig. 2d)

Effects of light on photosynthetic rates and chlorophyll fluorescence parameters

Compared to the medium light treatment, the high light treatment significantly (p < 0.05) reduced A of 62-71% in Norway maple and 36-75% in sugar maple at the green and yellow leaf stages, whereas no significant decreases were observed at the red leaf stage (Fig. 3a).

For both species, at yellow and red leaf stages, Fv/Fm was lower in seedlings grown in high light compared with seedlings in medium light (Fig. 3b). The changing tendency of ETR was compatible with that of ΦPSII (Fig. 3c). At red leaf stage, ΦPSII sharply decreased at high light (Fig. 3e). NPQ increased to the highest point at yellow leaf stage under high light conditions for both species, and at red leaf stage under low light conditions (L), which indicated that these seedlings had the strongest capability to regulate excessive energy dissipation at this time. Compared to medium light treatment, the dynamic changes in ΦPSII and NPQ under high light treatment with photooxidation seem to involve three stages of alteration: at green leaf stage, both ΦPSII and NPQ decreased or showed no significant change; at yellow leaf stage, ΦPSII slightly decreased and NPQ increased; and at red leaf stage, both parameters decreased. The former stage demonstrated the regulative response to moderate oxidative stress, and the later stage indicates irreversible damage of PSII. ETR decreased in all species as light intensity increased, especially at the red leaf stage (Fig. 4). For both species, above 500 μmol m-2 s-1 PPF, light-dependent ETR in plants under high light conditions at red leaf stage was lowest. Across all light treatments, ETR was higher (p < 0.05) in medium light and low light than in high light stress.

Effects of light on leaf N and N resorption efficiency

Norway maple seedlings exposed to high light showed 29% lower leaf N relative to medium light at red leaf stage, although that difference was not statistically significant (p = 0.11). For sugar maple, high light treatment decreased N by 33% (p < 0.05) at yellow leaf stage and by 43% (p < 0.05) at red leaf stage relative to medium light treatment (Fig. 5a). The N concentration in cell wall was not affected by light treatments in either species. No significant differences in fraction of leaf N in cell wall were detected among treatments at the green leaf stage (Fig. 5c). During the yellow leaf stage, the fraction of leaf N in cell wall of seedlings exposed to the high light increased so that by the red stage, these seedlings had a significantly (p < 0.05) higher fraction of leaf N in cell wall relative to seedlings under medium light conditions (Fig. 5c). N resorption of seedlings under high light conditions increased by 42% (p < 0.05) and 36% (p < 0.05) in comparison with seedlings under medium light conditions in Norway maple and sugar maple, respectively. N concentration in cell wall, fraction of leaf N in cell wall and N resorption were similar in seedlings under medium light and low light conditions (Fig. 5).

Effects of light on number of leaves shedding

In both species, the accumulated proportions of leaves shed in late September and October differed among treatments. For both species in all treatments, leaf abscised sharply at yellow leaf stages (September 10-October 3). In contrast, at red leaf stages (October 12-31), both species shed fewer leaves (Fig. 6). In addition, low light treatment decreased leaf fall at yellow leaf stage, compared with high light treatment.


In our study, sugar maples had greater anthocyanin expression. Nevertheless we found no indication that sugar maple seedlings have a photosynthetic advantage compared with Norway maple seedlings. It is possible that the weaker production of anthocyanins by Norway maple during autumn senescence is compensated by alternative mechanisms (Feild et al., 2001). On the other hand, light-induced reductions in fluorescence indicated higher levels of photoinhibition in the high light exposed seedlings. We find that seedlings grown in high light are clearly irradiance stressed and that high light clearly speeds up the senescence process. This can be shown by the decreased Fv/Fm and the increased accumulation of anthocyanins. This demonstrates a direct association between anthocyanin production and the period of increased vulnerability to photo-inhibition during autumn, and points to a higher need for a buildup of a photo-protection system. Furthermore, redness was strongly coupled with light environment (Fig. 2d, f) with the reddest leaves (more anthocyanins) of an individual occurring in the sunniest treatments (high light). Our results indicate that leaf redness is inducible and related to photo-oxidative stress. The stimuli for, and the timing of autumnal anthocyanin production are all consistent with a photoprotective role for these pigments in autumn leaves. However, light stressed and shaded seedlings had a similar proportion of nitrogen that was relocated to other tissues. This is contrary to our initial hypothesis that photoinhibiton would reduce leaf N re-translocation.

Leaf ontogeny during the three leaf color stages was characterized by a progressive anthocyanin pigments accumulation (Fig. 2d). Chlorophyll and carotenoid levels in plants under high light conditions were lower than under the medium light and low light conditions. Previous results indicated shade adaptation in red leaves (Kyparissis et al., 2007). The need to increase light capture in shaded leaves results in an increased ratio of light harvesting antennae per reaction center, and, in turn, to lower chlorophyll a:b ratios. However, in our study, red leaves (high-light stressed leaves) had a higher chlorophyll a:b ratio than green (shaded) leaves, which indicates biochemical adaptation to high irradiance as plants shift resources from light harvesting to photochemical processing.

As shown in Fig. 3, a decreasing trend in photosynthetic rates and Fv/Fm is evident as light increases. A decrease in dark-adapted Fv/Fm is generally used as a measure of photoinhibition (Perron and Juneau, 2011). At green leaf stage, Fv/Fm values were between 0.75 and 0.80 across all treatments of both species, indicating the absence of photoinhibited regions (Björkman and Demming, 1987). However, during the yellow and red leaf stages, the rapid decrease in Fv/Fm of plants under high light conditions compared to plants under medium light conditions indicates the onset of photodamage to PSII. Leaf-level avoidance of high light stress can be attributed to a variety of physiological processes (Lovelock et al., 1998). Energy consumption via photosynthetic electron transport connected to PSII is an important factor in photoinhibition avoidance. In addition to higher rates of electron transport, heat dissipation acts as a photoprotective mechanism (Kitao et al., 2006). Heat dissipation reduces ΦPSII and qP (Fig. 3). The lower chlorophyll fluorescence parameters (qP, ΦPSII, NPQ and ETR) on the upper surface of the plants grown in high light are further indicators of the occurrence of photoinhibition in the high light-exposed seedlings. Even so, we suggest that anthocyanins reduce risk of photo-oxidative damage to leaf cells as they senesce. Otherwise the plant under the high light conditions would become even more photoinhibited, which is analogue to the reduced photoprotective capacities of the anthocyanin-deficient mutants of woody species, Cornus sericea, Vaccinium elliottii (Chapmn.), and Viburnum sargentii (Koehne) (Hoch et al., 2003). Their ETR and photosynthesis would drop even more because their photosystems would be damaged. In September, the light in Montreal is 74% of the maximum (July values) and in October it is still 40%. Therefore the physiological gain of continued photosynthetic activity can be substantial. By increasing photoprotection and maintaining a photosystem during cold nights, trees could increase the length of their photosynthetically active period by a few weeks. As a consequence, we hypothesize that leaf redness is a mechanism able to squeeze a bit more photosynthesis out of the leaves before winter. This would concord with the relatively high rates of photosynthesis we still observed in the leaves at the yellow stage (Fig. 3a).

For both species foliar N concentrations during green and yellow periods were well within the 16.0 to 23.2 mg g-1 range of Kolb and McCormick (1993) for sugar maple trees. In contrast, N concentrations during red leaf periods concentrations averaged only 6.0 g kg-1, which is far below the concentrations reported for sugar maple seedling with N-limited growth (Walters and Reich, 1997). Leaves of high-light stressed plants were also characterized by lower N concentrations at all sampling dates, encompassing both the green and the red leaf period. Similarly, low levels of foliar N have been associated with early and high levels of leaf pigments, such as anthocyanins (Baltzer and Thomas, 2005; Sinkkonen, 2008). During the highly coordinated processes of leaf senescence, nutrients are re-absorbed back into woody tissues. Approximately 69 -75% of the N in green leaves was resorbed during senescence in both species. This represents a higher resorption efficiency than the mean (50-52%) reported for many plants (Chapin and Kedrowski, 1983; Aerts, 1996), but is consistent with previous measurements of resorption in Acer rubrum (Grizzard et al., 1976). In our study, the resorption of N is more efficient under light stress. This is in disagreement with Hoch et al. (2003). However, we also observed a positive correlation between N resorption and anthocyanin levels (r = 0.63, p = 0.07 for Norway maple; r = 0.69, p = 0.04 for sugar maple) in leaves. One hypothesis is that anthocyanins facilitate the resorption of nutrients by reducing the oxidative activity of the breakdown products of photosynthesis sequestered in vacuoles during the orderly breakdown of chlorophyll (Matile et al., 1999), as free radicals could disrupt the movement of nitrogen and/or phosphorus from leaves back into branches (Lee et al., 2003). However, we observed that the antioxidant activity of leaves in the different light treatments was equal from green to red leaf periods. The similarity in antioxidant activity suggests anthocyanins could be a strategy employed to maintain overall antioxidant status during senescence when other antioxidant compounds and systems may be less abundant or functional (van den Berg and Perkins, 2007). Therefore, although we did not find that red leaves had increased free-radical scavenging in the two species, we still suggest anthocyanins could increase the efficiency of nutrient resorption and reduce the residual nitrogen in fallen leaves (Aerts, 1996), through photoprotection and free-radical scavenging (Hoch et al. 2003) which otherwise may lower the efficiency of nutrient retrieval from senescing autumn leaves. More focused physiological studies, particularly of species with phenotypes differing in anthocyanin production, will add to our understanding of the functional role of anthocyanins during senescence.

A recent study by Schaberg et al. (2008) demonstrates a relationship between foliar coloration and abscission zone formation and suggests that red coloration in sugar maple may allow for an extended period of nutrient and sugar translocation compared with yellow leaves. In the present study, the number of abscised leaves at the yellow leaf stage was higher than at the red leaf stage (Fig.6), further strengthening the hypothesis that anthocyanins may benefit species like sugar maple by allowing for prolonged retention. Additionally, N concentration in the cell-wall was nearly constant across different leaf color periods at different light treatments, which may contribute to maintaining leaves functional for a long period (Takashima et al., 2004). We considered two possible explanations for red leaves being retained in high-light stressed plants. First, leaves would be retained as long as they have a positive daily carbon gain, and shed when the carbon gain becomes negative (Ackerly, 1999). This would concord with the suggestion that anthocyanins accumulation could be a maximization of the carbon gain late in the season. Second, we suggest that red leaves are involved in a conservative function, increasing N resorption and mean residence time during red leaf stage, while other green or yellow-leaves in the maple trees are mainly involved in a photosynthetic function. Especially for seedlings, extending the useful life of leaves is well compensated by potential carbon gains. However it may not be so for adults, as this comes at the risk of them freezing and dying before the tree could remove the nutrients therein, and indeed this is exactly what we find in forests: adult shedding leaves earlier than understory seedlings.

In conclusion, it appears that in spite of a large amount of de novo anthocyanin synthesis in senescing leaves, photoinhibition occurs in red and yellow maple leaves, when they are exposed to high light levels. We also note that there was a positive correlation between N resorption and anthocyanin levels in leaves. We proposed that autumnal anthocyanins protect senescing foliage from photoinhibitory irradiances (although inadequate photoprotection), and leaf redness is a way to squeeze a bit more photosynthesis out of the leaves before winter, allowing for the resorption of critical foliar nutrients to occur. The diminished photosynthetic capacity but higher N resorption in high light leaves may indicate the adaptive function of photoprotection would not only be the protection of leaves per se, but also that functional leaves enable a greater resorption of nutrients, especially nitrogen. Our results suggest it may be rewarding to view the significant functional role of leaf anthocyanin from the leaf to the whole plant level.

Acknowledgements We gratefully acknowledge funding by the Fonds québécois de la recherche sur la nature et les technologies (FQRNT). We also thank the Montreal Botanical Garden for providing the space necessary for this experiment.