Rapid changes in xanthophyll cycle-dependent energy dissipation and photosystem II efficiency in two vines, Stephania japonica and Smilax australis, growing in the understory of an open Eucalyptus forest
W. W. ADAMS III,
Department of Environmental, Population, and Organismic Biology, University of Colorado, Boulder, CO 80309–0334, USA and,
Leaves of Stephania japonica and Smilax australis were characterized in situ on the coast of north-eastern New South Wales, Australia, where they were growing naturally in three different light environments: deep shade, in the understory of an open Eucalyptus forest where they received frequent sunflecks of high intensity, and in an exposed site receiving full sunlight. In deep shade the xanthophyll cycle remained epoxidized during the day and the vast majority of absorbed light was utilized for photosynthesis. In the exposed site both deepoxidation and epoxidation of the xanthophyll cycle and changes in the level of xanthophyll-dependent thermal energy dissipation largely tracked the diurnal changes in photon flux density (PFD). In the understory the xanthophyll cycle became largely deepoxidized to zeaxanthin and antheraxanthin upon exposure of the leaves to the first high intensity sunfleck and this high level of deepoxidation was maintained throughout the day both during and between subsequent sunflecks. In contrast, thermal energy dissipation activity, and the efficiency of photosystem II, fluctuated rapidly in response to the changes in incident PFD. These findings suggest a fine level of control over the engagement of zeaxanthin and antheraxanthin in energy dissipation activity, presumably through rapid changes in thylakoid acidification, such that they became rapidly engaged for photoprotection during the sunflecks and rapidly disengaged upon return to low light when continued engagement might limit carbon gain.
The important role that sunflecks play in the overall carbon balance of plants growing in the understory has been well documented (Pearcy et al. 1994). Not only is there an enhancement of CO2 fixation during a sunfleck, but CO2 continues to be assimilated at elevated rates for a short period of time subsequent to a sunfleck. As important as sunfleck-stimulated carbon gain may be, positive CO2 uptake during exposure to the low light prevailing in the absence of sunflecks (filtered and diffuse sky radiation) also plays an important role in the success of plants found in the understory (Pfitsch & Pearcy 1992; Koizumi & Oshima 1993; Pearcy et al. 1994).
Leaves of plants growing in low light or understory environments typically have a relatively low capacity for photosynthetic electron transport and CO2 fixation (Björkman 1981; Pearcy & Sims 1994). Thus, when receiving sunflecks of moderate to high photon flux density (PFD) it appears that only a fraction of the absorbed light can be used in photosynthetic electron transport (Logan et al. 1997; Watling et al. 1997). The remaining excitation energy absorbed by the light-harvesting complexes must either be dissipated harmlessly or it has the potential to damage the photosynthetic apparatus. The majority of thermal energy dissipation in photosystem II (PSII) is obligatorily dependent on the presence of zeaxanthin (Z) and antheraxanthin (A), the deepoxidized components of the xanthophyll cycle (for reviews, see Demmig-Adams & Adams 1996a; Demmig-Adams, Gilmore & Adams 1996b; Gilmore 1997). The engagement of Z + A in thermal dissipation is induced by thylakoid acidification, resulting in the protonation, and conformation change in, Chl- and xanthophyll-binding light-harvesting complexes.
Königer et al. (1995) found that the xanthophyll cycle can be deepoxidized to a high degree in leaves of understory plants receiving sunflecks of high intensity. Depressions in PSII efficiency observed during exposure to full sunlight in canopy gaps (Krause & Winter 1996; Thiele, Krause & Winter 1998) or during brief exposures to sunflecks (Watling et al. 1997) were largely attributed to increases in photoprotective energy dissipation. In the latter study it was suggested that decreases in PSII efficiency that persisted subsequent to a sunfleck might represent a cost to the plant in terms of a reduced capacity to fix CO2 in the low light following the passage of a sunfleck. Logan et al. (1997) examined xanthophyll cycle-dependent energy dissipation in Alocasia brisbanensis that experienced sunflecks of low intensity and short duration and reported that depressions in PSII efficiency were largely rapidly reversible.
In this study the leaves of two vines, Stephania japonica and Smilax australis, were intensively characterized where the plants were growing in sites exposed to full sunlight as well as in the understory of an open Eucalyptus forest where they received frequent sunflecks of high intensity and varying duration. Leaves were characterized in terms of xanthophyll cycle-dependent energy dissipation as well as a number of parameters derived from chlorophyll fluorescence determined in situ from predawn to dusk. In addition, leaves of the same species growing in deep shade were examined for comparison with those growing in the understory and in full sunlight. This study thus represents the first comprehensive characterization of the dynamics of energy dissipation activity and changes in carotenoid composition over the entire spectrum of possible light environments from full sun to sunflecks to deep shade.
MATERIALS AND METHODS
Site description and species
Two species of vines, Stephania japonica (Thunb.) Miers var. discolor (Blume) Forman (family Menispermaceae) and Smilax australis R. Br. (family Smilacaceae), were characterized in three different light environments just to the south of Middle Head on the north-eastern coast of New South Wales, Australia (30°50 'S, 153°00 'E, approximately 2–15 m). The deep-shade site in which both species were characterized was under the canopy of coastal scrub species and immediately south of Middle Head (a rock outcrop of volcanic origin) that shielded the plants from direct sunlight throughout the day (midday PFD reached 29–30 μmol photons m–2 s–1). Both species were also characterized in an exposed area on the first dune above the beach that was stabilized by vegetation, approximately 75 m south of Middle Head, where they received full sunlight from sunrise until midday at which time direct sunlight was obstructed by Middle Head. The third site was located approximately 300 m inland under a canopy of predominantly Eucalyptus trees. In this site both species received sunflecks of varying duration and intensity throughout the day. The plants growing in deep shade as well as those in the fully exposed site adjacent to the beach were characterized on 17 June 1994. Both species growing under the canopy of Eucalyptus trees were characterized on 18 June 1994, and Stephania japonica was characterized more intensively in the understory a second time on 20 June 1994.
Characterization in the field
Chlorophyll fluorescence emission was measured using a PAM-2000 chlorophyll fluorometer (Walz, Effeltrich, Germany) and portable chart recorder as described in Demmig-Adams & Adams (1996b) and Demmig-Adams et al. (1996a). The leaf temperatures and PFDs incident upon the leaves were ascertained using the sensors integrated into the leaf clip holder of the PAM-2000 unit, and more continuous monitoring of incident PFDs on 18 June was achieved with a Li-Cor quantum photometer (model LI-189; Lincoln, NE, USA). Small samples of leaf tissue (0·25 cm2) were removed from the leaves concomitant with the characterization of chlorophyll fluorescence. These tissue samples were immediately frozen in liquid nitrogen in which they were maintained until analysis of carotenoid and chlorophyll content in the laboratory. Additional tissue samples of 0·77 cm2 were frozen in liquid nitrogen for determination of ascorbate content. For characterization of leaves growing in full sunlight and in the understory of the forest, adjacent leaves comparable in appearance and orientation (exposure to light) were chosen for the diurnal characterizations on 17 and 18 June, with one leaf used for the determination of chlorophyll fluorescence and the second for pigment analysis and ascorbate content. The third Stephania japonica leaf characterized on 20 June was an adjacent leaf comparable in appearance and orientation to the two leaves characterized on 18 June.
Calculations from Chl fluorescence
The fraction of absorbed light utilized through photochemistry was estimated from (Fm' –F)/Fm' (see Genty, Briantais & Baker 1989). An estimate of the flux of photons going into photochemistry, which can be used as an estimate of the rate of photochemistry at a given PFD, was derived from (Fm' –F)/Fm' · PFD. Fv/Fm is a measure of the efficiency of open PSII units in darkness (predawn), and Fv'/Fm' of the efficiency of open PSII units during illumination. 1 –Fv/Fm and 1 –Fv'/Fm' can be utilized to estimate the fraction of absorbed light that is dissipated thermally, and in analogy to the estimation of the rate of photochemistry (1 –Fv'/Fm') · PFD can be used to estimate the flux of photons going into thermal dissipation, i.e. the rate of thermal energy dissipation (Demmig-Adams et al. 1996a). In this study the incident PFDs were used since an estimate of leaf absorptance was not obtained. Therefore the reported rates of photochemistry and thermal dissipation are somewhat overestimated, but only to the same degree as the reported incident PFDs are greater than that which was actually absorbed by chlorophyll. The fractional estimates of photochemistry and thermal energy dissipation are unaffected by this fact, as they represent the fate of the excitation energy that had already been absorbed by the chlorophyll.
Non-photochemical quenching of Fm, from Fm/Fm' – 1 (NPQ or Stern-Vollmer quenching; see Bilger & Björkman 1990), is an additional estimate of the level of energy dissipation activity that was quantified. These were derived using the highest value of Fm ascertained prior to sunrise, as well as from estimates of fully relaxed values of Fm (designated as Fm*) based upon an Fv/Fm value of 0·82 using the equation of Kitajima & Butler (1975; see Demmig-Adams 1990). The Fv/Fm value of 0·82 was chosen as a conservative estimate of the fully relaxed level of PSII efficiency (values greater than 0·86 were measured from Smilax australis growing in the shade in 1986; see Adams et al. 1988). Such calculations of Fm* are valid if the predawn depression in PSII efficiency was due to a sustained level of energy dissipation activity that remained engaged during the night, resulting in a persistent decrease in the efficiency of energy delivery from the light-harvesting Chls to the reaction centres, i.e. Fv/Fm (see, e.g. Adams & Demmig-Adams 1995; Adams et al. 1995).
In response to the diurnal increase in incident PFD (Figs 1a & 2a) there were strong decreases (or ‘quenching’) of maximal fluorescence Fm and, to a lesser extent, minimal fluorescence Fo (Figs 1b & 2b) in both plant species. The efficiency of open PSII units (Fv/Fm predawn and Fv'/Fm' during the day) as well as, to an even greater extent, the efficiency of PSII at the actual degree of closed or reduced centres [(Fm' – F)/Fm'] declined during exposure to sunlight and subsequently increased in the afternoon when the leaves received only diffuse sky radiation (Figs 1c & 2c). Neither species exhibited fully relaxed levels of PSII efficiency (maximal value 0·8 or higher; Adams et al. 1990) prior to sunrise, and this nocturnally sustained depression of Fv/Fm was greater in Smilax australis than in Stephania japonica (Figs 1c & 2c).
In both plant species the flux of excitation energy processed through photochemistry increased to a maximal rate with increasing PFD and declined once the leaves were no longer receiving direct sunlight (Figs 1d & 2d). The estimated mean maximal rate of photochemistry in the two sun leaves of Stephania japonica (305 μmol photons processed m–2 s–1) was approximately twice as high as that in the two sun leaves of Smilax australis (144 μmol photons processed m–2 s–1). Whereas the flux of excitation energy through photochemistry was apparently saturated above an incident PFD of approximately 500 μmol photons m–2 s–1, the flux of excitation energy dissipated thermally varied more or less in proportion to the incident PFDs up to full sunlight (Figs 1d & 2d). The flux of photons dissipated thermally reached maximal levels of approximately 4× and 8× that utilized through photochemistry in Stephania japonica (Fig. 1d) and in Smilax australis (Fig. 2d), respectively.
Analysis of the fractional allocation of absorbed excitation energy to thermal dissipation and photochemistry (Figs 1e & 2e) revealed that greater than 50% of the absorbed light was utilized for photochemistry when incident PFD was low during the afternoon, but that leaves of Stephania japonica used approximately 20% and Smilax australis only 10–15% of the absorbed light for photochemistry between 0730 and 1300 or 1200 h, respectively, when incident PFDs were high. During this period of exposure to direct sunlight the fraction of absorbed excitation energy dissipated thermally was 60–70% in Stephania japonica and 70–80% in Smilax australis (Figs 1e & 2e).
In sun leaves of both species 25–40% of the xanthophyll cycle was retained nocturnally as Z + A (Figs 1f & 2f). During exposure to full sunlight most of the remaining violaxanthin (V) was converted to Z + A, with maximal conversion states (A + Z)/(V + A + Z) for Stephania japonica leaves 1 and 2 reaching 0·86 and 0·89 and for Smilax australis leaves 1 and 2 reaching 0·92 and 0·93, respectively. Whereas the midday level of Z + A as a fraction of the total xanthophyll cycle pool (V + A + Z) was only slightly higher in the leaves of Smilax australis compared to Stephania japonica, the midday level of Z + A per total chlorophyll was approximately 30% greater in Smilax australis (Figs 1f & 2f; chlorophyll contents of the sun leaves of both species were similar, see Table 1). Epoxidation of Z and A proceeded during the afternoon once the leaves were no longer receiving direct sunlight, but was somewhat delayed in Stephania japonica that had received high PFD for longer than Smilax australis (cf. Figures 1a & 2a). The diurnal increase and decrease in the conversion state of the xanthophyll cycle (Figs 1f & 2f) generally mirrored the diurnal decreases and increases in maximal and minimal fluorescence emission (Figs 1b & 2b) and PSII efficiency (Figs 1c & 2c), the diurnal increases and decreases in the rate of dissipation (Figs 1d & 2d), and the fraction of absorbed excitation energy dissipated thermally (Figs 1e & 2e).
Table 1. . Total daily integrated PFD (estimated from Figs 1a, 2a, 3a & 5a), chlorophyll and carotenoid composition, and ascorbate content of Stephania japonica and Smilax australis leaves growing in deep shade, in the understory of an open Eucalyptus forest, and in full sunlight at Middle Head on the coast of north-eastern New South Wales in June of 1994. Values represent the mean for each leaf ±SD, with n = 2 for shade, n = 16 for the leaves receiving sunflecks and n = 8–11 for the leaves receiving full sunlight, except for the ascorbate values which each represent single measurements
Predawn and midday characterization of shade leaves
Whereas exposed leaves of Stephania japonica and Smilax australis reached predawn minima between 4 and 6 °C (Figs 1a & 2a), the shade leaves of both species growing in the shelter of the coastal scrub species experienced warmer temperatures during the night (pre-dawn minima of 10–11 °C). In contrast to the sun-exposed leaves of both species, which reached a midday temperature of approximately 30 °C (Figs 1a & 2a), the shade leaves of both only reached approximately 17 °C. During the day, when these leaves received approximately 30 μmol photons m–2 s–1 incident radiation, 71% of the absorbed excitation energy was utilized in photochemistry and only 27% was dissipated thermally (which largely represents constitutive, and not regulated, thermal loss). The conversion state of the xanthophyll cycle [(A + Z)/(V + A + Z)] in these leaves was low prior to sunrise (0·02 and 0·04 in Stephania japonica and Smilax australis, respectively), and remained low throughout the day (0·06 and 0·04 at midday in Stephania japonica and Smilax australis, respectively). The predawn PSII efficiency, Fv/Fm, was 0·78 in the shade leaves of both species, and Fv'/Fm' remained high at midday (at 0·73).
Diurnal characterization of leaves in the forest understory
Whereas the absolute magnitude of the most intense sunflecks received by the leaves of both species in the understory of the Eucalyptus forest was similar (between 1400 and 1600 μmol photons m–2 s–1), the frequency and duration of the sunflecks was greater for the Smilax australis leaves compared with the Stephania japonica leaves characterized in this study (Figs 3a & 4a). A fairly continuous characterization of sunflecks experienced by both sets of leaves was possible, but the appropriate care in the measurement of chlorophyll fluorescence and sampling of leaf tissue for pigment analysis necessitated a more selective characterization of these latter parameters (the two species were located approximately 10 m from one another). The leaves were thus characterized just before receiving sunflecks, during exposure to selected sunflecks, and immediately following the transition from high light to low light after a sunfleck had passed. At 1050 h the PFD varied so rapidly that it was not possible to obtain an accurate measure of Fo' (and thus Fv'/Fm', the rate of thermal dissipation, nor of the fraction of absorbed light dissipated thermally) in Stephania japonica, and during the early afternoon (between 1200 and 1330 h) characterization of Smilax australis interfered with the characterization of Stephania japonica (see Fig. 4b–4f).
As was the case for the sheltered shade leaves, the nocturnal temperatures experienced by the leaves in the understory of the Eucalyptus trees were higher (approximately 13 °C predawn; Figs 3b & 4b) than those experienced by the fully exposed leaves (Figs 1a & 2a). Mid-sunfleck temperature maxima for the leaves in the understory reached 23·6 °C (Stephania japonica; 4Fig. 4b: 26·6 °C on 20 June during the sunflecks that were not characterized on 18 June, see Fig. 5a) to 25·8 °C (Smilax australis; Fig. 3b). Also depicted in these figures (Figs 3b & 4b) are the PFDs experienced by the leaves at the times that they were characterized for PSII efficiency, photochemistry, thermal dissipation, and the conversion state of the xanthophyll cycle.
During exposure to sunflecks both maximal fluorescence Fm' and minimal fluorescence Fo' decreased, whereas both levels of fluorescence increased between sunflecks when the leaves received only diffuse and filtered radiation (Figs 3c & 4c). Both the efficiency of open PSII units (Fv/Fm predawn and Fv'/Fm' during the day) and to a greater extent the efficiency of PSII at the actual degree of closure [(Fm' –F)/Fm'] declined during exposure to sunflecks and increased again rapidly between sunflecks (Figs 3d & 4d). Neither species exhibited fully relaxed levels of PSII efficiency prior to sunrise, and the nocturnally sustained depression of Fv/Fm was greater in Smilax australis (0·65) than in Stephania japonica (0·75; Figs 3d & 4d). Whereas rapid changes in PSII efficiency tracked the sunflecks, the highest efficiencies of open PSII units during the periods of low PFD were 0·66 and 0·58 in Stephania japonica and Smilax australis, respectively, and thus lower than those observed prior to sunrise (Figs 3d & 4d).
The estimated rates of both photochemistry and thermal dissipation increased markedly during the sunflecks and decreased to minimal levels between sunflecks (Figs 3e & 4e). As was the case for the leaves growing in full sunlight (Figs 1d & 2d), the estimated maximal rate of photochemistry in the leaf of Stephania japonica (245 μmol photons processed m–2 s–1; Fig. 4e, see also Fig. 5b) growing in the understory was greater than that in the leaf of Smilax australis (168 μmol photons processed m–2 s–1; Fig. 3e).
The fractional allocation of absorbed excitation energy through photochemistry versus thermal dissipation varied dramatically throughout the day depending upon the PFD the leaves were experiencing (Figs 3f & 4f). During sunflecks a majority of absorbed excitation energy was dissipated thermally, whereas during the periods when incident PFD was low (between sunflecks and early in the morning and late in the afternoon) greater than 50% of the absorbed excitation energy that was not dissipated constituitively (the fraction above the dashed lines in Figs 3f & 4f) was utilized through photochemistry.
In Smilax australis almost 50% of the total xanthophyll cycle pool had been retained as Z + A during the night (Fig. 3g), whereas only 16% of the pool was retained nocturnally as Z + A in Stephania japonica (Fig. 4g). Upon exposure to the first sunflecks the conversion state reached almost 80% in Smilax australis and more than 70% in Stephania japonica (Figs 3g & 4g). There was a small degree of epoxidation of the xanthophyll cycle between sunflecks but for the most part the cycle remained highly deepoxidized throughout the day (the maximal conversion state reached 0·93 at 1245 h in Smilax australis and 0·77 at 1330 h in Stephania japonica) until 1500 h after which the PFD remained low and epoxidation occurred (Figs 3g & 4g).
Whereas the maximal level of A + Z on a chlorophyll basis was 30% greater in sun leaves of Smilax australis compared with the sun leaves of Stephania japonica (Figs 1f & 2f), in the understory the level of A + Z was almost 50% higher on a chlorophyll basis in Smilax australis relative to Stephania japonica (Figs 3g & 4g), although this was due in part to a lower leaf chlorophyll content in Smilax australis (Table 1). In contrast to the fully exposed leaves (Figs 1 & 2), the maximal and minimal levels of fluorescence (Figs 3c & 4c), the intrinsic efficiency of PSII Fv'/Fm' (Figs 4d & 5d), and the level of energy dissipation activity (Figs 3e, 3f, 4e & 4f) fluctuated considerably during the day in response to changes in PFD (Figs 3a, 3b, 4a & 4b) with little fluctuation in the level of A + Z between 0900 and 1500 h (Figs 3g & 4g).
Given the spotty nature of the measurements on 18 June (Figs 3 & 4), the fact that the leaves had to be darkened briefly each time fluorescence was ascertained in order to obtain Fo', and the fact that for Stephania japonica a portion of the afternoon was not characterized (Fig. 4), another Stephania japonica leaf was characterized continuously on 20 June in an effort to elucidate the rapidity with which PSII responded to rapid changes in the incident PFD in the understory (Fig. 5). Comparison of the incident PFDs shows that the light environment experienced by this adjacent leaf (Fig. 5a) was almost identical to that experienced by the leaf examined on 18 June (Fig. 4a). On 20 June the leaf was never darkened so as not to interfere with the rapid changes characterized, and therefore the minimal level of fluorescence Fo' (and all parameters derived therefrom) was not determined during the day.
The estimated rate of photochemistry increased rapidly in response to sunflecks and declined just as rapidly upon return to low light between sunflecks (Fig. 5b), reaching maxima comparable to those observed in the leaf on 18 June (Fig. 4e). Stern-Vollmer quenching of Fm' or NPQ, a measure of energy dissipation activity that does not require the level of Fo' for its calculation, responded equally rapidly to changes in PFD (Fig. 5c). NPQ was also calculated correcting for the nocturnally sustained depression in PSII efficiency (Fv/Fm = 0·766 at 0612 h, not shown) under the assumption that the sustained depression was due to persistent energy dissipation activity (Fig. 5d; see Materials and Methods). Both estimates of energy dissipation activity increased rapidly during exposure to sunflecks and decreased rapidly to relatively low levels, albeit not to zero, between sunflecks. During sunflecks the maximal rates of photochemistry were slightly lower in the morning compared to the afternoon (Fig. 5b), whereas the maximal levels of energy dissipation were slightly higher in the morning compared with the afternoon (Figs 5c & d), possibly due to the differences in leaf temperature between morning and afternoon (Fig. 5a).
The efficiency of PSII at the actual degree of closure was equally responsive to the rapid changes in the light environment (Fig. 5e). The PSII efficiency dropped to very low levels during exposure to high PFD sunflecks (to a greater extent in the morning than during the afternoon), and increased rapidly to high levels during the low PFD periods between sunflecks. Nevertheless, during periods of very low PFD PSII efficiency remained below 0·7 and did not reach the predawn value of 0·766.
Leaf ascorbate and pigment composition
The leaves of Stephania japonica characterized in the understory received approximately one-third of the daily integrated PFD received by the leaves characterized in full sunlight (Table 1). The ascorbate contents of the Stephania japonica leaves from deep shade and the understory were 6 and 17%, respectively, of that found in the leaf growing in full sunlight (Table 1). In contrast to Stephania japonica, the leaves of Smilax australis characterized in the understory received 80% of the daily integrated PFD received by the leaves characterized in full sunlight (Table 1). The ascorbate content of the understory leaf of Smilax australis was likewise greater than that of the understory leaf of Stephania japonica, at 39% of the ascorbate content found in the Smilax australis leaf growing in full sunlight.
On a leaf area basis the chlorophyll and carotenoid contents of Stephania japonica were higher in the understory than in full sunlight or in deep shade, whereas in Smilax australis both were highest in the deeply shaded leaves (Table 1). The carotenoids of the xanthophyll cycle were present at higher levels in full sunlight compared to deep shade, whereas the levels of α-carotene were highest in deep shade in both species on a leaf area basis (Table 1). The other carotenoids (β-carotene, lutein, and neoxanthin) were present at the highest levels in either the understory (Stephania japonica) or in deep shade (Smilax australis; Table 1).
When expressed on a chlorophyll basis, however, the total carotenoids were lowest in deep shade, higher in the understory, and highest in full sunlight in both species (Table 1). This increase with increasing growth PFD was due predominantly to the carotenoids of the xanthophyll cycle (Table 1 and Fig. 6), although per chlorophyll both β-carotene and lutein were also somewhat higher as the growth PFD increased (Table 1). On the other hand, the per chlorophyll neoxanthin was the same in the leaves of both species from all light environments, and α-carotene was highest in deep shade, intermediate in the understory, and lowest in full sunlight (Table 1). Consequently, when expressed as a fraction of the total carotenoids present in the leaves, the carotenoids of the xanthophyll cycle showed progressive increases, β-carotene and lutein did not differ, and neoxanthin and especially α-carotene showed progressive decreases from shade to full sunlight (Fig. 6).
The Chl a/b ratio was lowest in the deep-shade leaves of both species and similar in the leaves growing in the understory and in full sunlight (Table 1). The ratio of total carotenes to total xanthophylls, as well as that of β-carotene to the carotenoids of the xanthophyll cycle, was greatest in deep-shade leaves, intermediate in the understory leaves, and lowest in the leaves growing in full sunlight (Table 1). Over the course of the day the level of variation in both of these carotenoid ratios was greater in the understory leaves than in the leaves growing in the exposed site (Table 1), and this was due primarily to fluctuations in the ratio of β-carotene to the carotenoids of the xanthophyll cycle that roughly corresponded to exposure to high PFD during sunflecks (when the ratio declined) and to low PFD between sunflecks (when the ratio increased; Fig. 7). During each sunfleck this ratio appeared to approach the ratio found in the leaves growing in full sunlight (Figs 7a & b). Changes in this ratio were due to both decreases in β-carotene and increases in the carotenoids of the xanthophyll cycle during exposure to high-PFD sunflecks, and to increases in β-carotene and decreases in the xanthophyll cycle pool when the leaves experienced low PFDs between sunflecks (Figs 7c & d).
Acclimation of foliar carotenoid and ascorbate levels to growth PFD for both Stephania japonica and Smilax australis (Table 1 and Fig. 6) were generally consistent with previous studies (Königer et al. 1995; Logan et al. 1996; Demmig-Adams 1998). The differences in carotenoid composition between the fully exposed leaves and those growing in the Eucalyptus understory for the first time permit some evaluation as to whether acclimation occurs in response to the maximal PFD that the leaves experience or to the total integrated PFD they receive over the entire day. Given that the leaves in the understory received sunflecks reaching PFDs that were equivalent to the maximal PFDs received by the fully exposed leaves (compare Figs 1a & 2a with Figs 3a–5a), acclimation of foliar carotenoid composition (e.g. the level of xanthophyll cycle carotenoids) would thus appear to respond to some measure of the daily integrated PFD (Table 1).
Stephania japonica, the more mesophytic of the two species, utilized a greater fraction of the absorbed light through photochemistry than Smilax australis (cf. Figs 1 & 2). The higher levels of xanthophyll cycle-dependent energy dissipation activity observed in Smilax australis (Figs 1e & f) relative to Stephania japonica (Figs 2e & f) are consistent with these differences in the utilization of the absorbed light through photosynthesis, as are the differences in foliar ascorbate level (Table 1). Ascorbate, which in the chloroplast serves a dual role in photoprotection as a reductant in the enzymatic conversion of V to A and Z as well as in the detoxification of H2O2 to water (see discussion in Logan et al. 1996), was found in the highest levels in full sunlight, and was also higher in the species (Smilax australis) that utilized a smaller fraction of the absorbed light through photochemistry.
The diurnal changes in Fm' and Fo' (Figs 1b, 2b, 3c & 4c) are consistent with changes in the level of energy dissipation activity underlying the diurnal changes in the efficiency of open PSII units (Figs 1c, 2c, 3d & 4d; see Kitajima & Butler 1975). The diurnal changes in the conversion state of the xanthophyll cycle in the leaves growing in the fully exposed site (Figs 1f & 2f) largely mirror the pattern of allocation of absorbed light that is dissipated thermally (Figs 1e & 2e), consistent with the involvement of Z + A in energy dissipation in these leaves (for reviews, see Demmig-Adams & Adams 1996a; Demmig-Adams et al. 1996b). In contrast, in the understory energy dissipation activity (Figs 3e, 3f, 4e, 4f, 5c & 5d) and PSII efficiency (Figs 3d, 4d & 5e) responded rapidly to changes in incident PFD (Figs 3a, 3b, 4a, 4b & 5a) whereas the xanthophyll cycle experienced deepoxidization following the first major sunfleck in the morning and remained largely deepoxidized throughout the day (Figs 3g & 4g), even during periods when the leaves received only diffuse radiation of low PFD (Figs 3a, 3b, 4a & 4b). These data suggest that Z + A were retained throughout the day but that their actual engagement in energy dissipation was finely controlled. This control is probably exerted through changes in thylakoid acidification (Gilmore & Yamamoto 1993; Goss, Richter & Wild 1995; Horton, Ruban & Walters 1996; Ruban, Young & Horton 1996; see Gilmore 1997) resulting from rapid changes in the rate of electron transport (Figs 3e, 4e & 5b; Schönknecht et al. 1995). Thus, the leaves in the understory maintained a pool of Z + A that could be rapidly engaged in energy dissipation during exposure to sunflecks (excess light) and rapidly disengaged once the sunfleck had passed and the leaves received only diffuse sky radiation (limiting light). Rapid engagement of (Z + A)-dependent energy dissipation upon exposure to direct sunlight may provide the photoprotection necessary to prevent damage to the photosynthetic apparatus during sunflecks, and its rapid disengagement upon the transition to low PFD between sunflecks permits a rapid increase in the efficiency of photosynthetic energy conversion (Fig. 5e). Were Z + A to remain engaged in energy dissipation upon transition to low light it would have the potential to limit carbon gain during the exposure to diffuse PFD between sunflecks.
Given that changes in thylakoid acidification appear to be able to modulate the level of energy dissipation activity at a more or less constant level of A + Z, one may ask whether the xanthophyll cycle itself is even necessary. However, whereas strong and rapid fluctuations in PSII efficiency mirrored the rapid fluctuations in incident PFD in the understory environment (Figs 3, 4 & 5), the highest levels of PSII efficiency both prior to sunrise as well as during the periods of low PFD during the day remained below the maximal possible level of 0·8–0·83 (Adams et al. 1990). This suggests that retention of high levels of A + Z does not allow maximally high levels of PSII efficiency. Thus low levels of both A + Z and acidification may be necessary for the very maximal possible levels of PSII efficiency. For the present study we cannot exclude an involvement of drought stress and low temperatures not only in the nocturnal retention of A + Z (1Figs 1f, 22f & 33g; see, e.g. Demmig et al. 1988; Adams et al. 1995; Adams & Barker 1998) but also in the diurnal retention of A + Z between sunflecks.
The increasing Chl a/b ratio from deep shade to understory to fully exposed leaves of both species (Table 1) would tend to suggest an increased ratio of Chl a-binding core proteins to outer, light-harvesting Chl a- and b-binding proteins from shade to more exposed sites. On the other hand, the decreasing ratio of total carotenes to total xanthophylls suggest a decreasing ratio of core proteins to light-harvesting proteins from shade to the more exposed sites (Table 1, Fig. 6; see Green & Durnford 1996; Yamamoto & Bassi 1996). This apparent inconsistency may indicate that either the levels of Chl b (Tanaka & Melis 1997) or V + A + Z (Demmig-Adams 1998) bound to given pigment-protein complexes, or both, may vary with growth PFD.
Interestingly, there was some indication for rapid changes in the levels of β-carotene and the total pool of xanthophyll cycle carotenoids in the understory leaves that appeared to correspond to the periods of exposure to sunflecks (when decreases in β-carotene and increases in V + A + Z were observed) and to low light (when increases in β-carotene and decreases in V + A + Z were observed), whereas no consistent changes were observed over the course of the day in the fully exposed leaves (Fig. 7). Decreases in β-carotene content may reflect a release of this carotenoid from core antenna proteins, followed by hydroxylation to zeaxanthin (Depka, Jahns & Trebst 1998; see also Demmig-Adams et al. 1998).
Taken together the results of this study, as well as previous characterizations of the xanthophyll cyle (Königer et al. 1995), chlorophyll fluorescence and the xanthophyll cycle in canopy gaps (Krause & Winter 1996; Thiele et al. 1998), and of the xanthophyll cycle and energy dissipation in the low PFD sunflecks of a subtropical rainforest (Logan et al. 1997) and sunflecks and light flecks in rainforest species (Watling et al. 1997), suggest a central role for xanthophyll cycle-dependent energy dissipation in the regulation of the delivery of excitation energy to photosynthesis in response to the whole range of growth PFD, including the dynamic light environment encountered by leaves in the forest understory. Rapid engagement of A + Z during sunflecks is likely to provide the photoprotection necessary upon sudden exposure to excess PFD, and its rapid disengagement upon return to low PFD is likely to permit high levels of carbon fixation when light is limiting.
We wish to express our gratitude to Drs W. S. Chow and J. M. Anderson, previously at the Division of Plant Industry, CSIRO, in Canberra, for the use of their spectrophotometer. We are indebted to Terry Parkhouse and the Yarrahapinni Ecology Study Centre for their continued open door policy and hospitality at the study site. This work was supported by a National Science Foundation grant, award number IBN-9207653, a visiting fellowship from the ANU, and a fellowship from the David and Lucile Packard Foundation to B.D.-A.
Permanent address: Department of Biology, Bowdoin College, Brunswick, Maine 04011, USA