I. The discovery of crassulacean acid metabolism (CAM) in the trees of Clusia: arrival in the limelight of international research 8
II. Phylogeny 8
III. Photosynthetic physiotypes 10
IV. Metabolic flexibility: organic acid variations 12
V. The environmental control of photosynthetic flexibility 13
VI. Phenotypic plasticity: physiotypes and morphotypes 16
VII. Ecological amplitude and habitat impact 16
VIII. Conclusions and outlook 21
It is the aim of this review to present a monographic survey of the neotropical genus Clusia on scaling levels from molecular phylogeny, metabolism, photosynthesis and autecological environmental responses to ecological amplitude and synecological habitat impact. Clusia is the only dicotyledonous genus with real trees performing crassulacean acid metabolism (CAM). By way of introduction, a brief historical reminiscence describes the discovery of CAM in Clusia and the consequent increase in interest in studying this particular genus of tropical shrubs and trees. The molecular phylogeny of CAM in the genus is compared with that in Kalanchoë and the Bromeliaceae. At the level of metabolism and photosynthesis, the great plasticity of expression of photosynthetic physiotypes, i.e. (i) C3 photosynthesis, (ii) CAM including CAM idling, (iii) CAM cycling and (iv) C3/CAM-intermediate behaviour, as well as metabolic flexibility in Clusia is illustrated. At the level of autecology, the factors water, irradiance and temperature, which control photosynthetic flexibility, are assessed. The phenotypic plasticity of physiotypes and morphotypes is described. At the level of synecology, the ecological amplitude of Clusia in the tropics and the relations to habitat are surveyed.
I. The discovery of crassulacean acid metabolism (CAM) in the trees of Clusia: arrival in the limelight of international research
As we shall see, species of Clusia are handsome woody plants. Some have small flowers only, but others develop rather large beautiful flowers; for example, the flowers of Clusia grandiflora Engl. have a diameter of 150 mm. Nevertheless, Clusia most likely would have remained just one genus among a vast number of others in the large biodiversity of trees and shrubs in the tropics if it had not turned out that species of Clusia are performing crassulacean acid metabolism (CAM). This mode of photosynthesis is basically characterized by nocturnal uptake of CO2 and dark fixation via phosphoenolpyruvate carboxylase (PEPC), where the resulting fixation product, mainly malic acid, is stored in the central cell sap vacuole. Behind closed stomata and in the absence of any overt gas exchange, the organic acid is remobilized during the light period and decarboxylated, and the CO2 regained is refixed via ribulose-bis-phosphate carboxylase/oxygenase (Rubisco) and assimilated in the Calvin cycle.
Clusia is the only genus of trees with this mode of photosynthesis. True, there are a number of sizeable plants with CAM in the families of Cactaceae, Euphorbiaceae and Didieraceae as well as the monocotyledonous Yucca, all of which have been considered ‘fantastic trees’ (Menninger, 1967; Ellenberg, 1981). Moreover, giant cactus communities in Venezuela have been called ‘cactus forests’ (Vareschi, 1980). However, Clusia is the only genus of trees sensu stricto with typical dicotyledonous secondary growth that performs CAM.
Clusia early raised the curiosity of researchers because of the peculiarities of its photosynthetic physiology. During field work in February 1800 in Venezuela, Alexander von Humboldt discovered that Clusia rosea Jacq. had no obvious overt gas exchange in the light period but built up a high internal gas pressure with an oxygen concentration near 40% (Krätz, 2001; Lüttge, 2002). In 1937, Willy Hartenburg reported observations from glasshouse experiments showing that on bright days Clusia mexicana Vesque did not take up CO2 in the light period but even released some CO2, which was not explicable by respiration. However, on overcast days the CO2 exchange pattern was as expected of photosynthesizing plants (Hartenburg, 1937; Lüttge, 1995b). Neither Alexander von Humboldt nor Willy Hartenburg was able to really explain their observations. Now, as we know that Clusia species are performing CAM, we realize that they saw particular features of CAM, where stomata are closed in the light period when CO2 remobilized from nocturnally stored organic acids is photosynthetically assimilated, which is associated with the build up of high internal partial pressures of oxygen (Spalding et al., 1979). Decarboxylation of nocturnally stored organic acids behind closed stomata also leads to high internal CO2 concentrations, i.e. 2–60 times atmospheric (Lüttge, 2002), which may even cause some diffusive loss of CO2 to the atmosphere in spite of stomatal closure. It remained for the Mexicans Tinoco Ojanguren & Vazquez-Yanez (1983) to demonstrate and clearly explain the performance of CAM in Clusia for the first time, and with an additional seminal study by Ting et al. (1985) this became widely known and created wide interest, moving Clusia into the limelight of international research activities. This now allows me to present Clusia in this review, and we are also preparing a monographic book on Clusia covering all aspects of its biology, i.e. its taxonomy and phylogeny, diversity and phytogeography, ecology and physiology, and circadian clock (Lüttge, 2006c).
The question is why there are not more trees with CAM. Similarly, although there are woody plants, shrubs and small trees with C4 photosynthesis, there are no real large trees with this mode of photosynthesis. The two modes, C4 and CAM, share many biochemical features, particularly the primary fixation of CO2 by PEPC and synthesis of organic acids (mainly malate and citrate in CAM and malate and aspartate in C4 photosynthesis) and the subsequent decarboxylation of the organic acids and refixation and assimilation of the CO2 by Rubisco in the Calvin cycle. There are, of course, important differences between C4 photosynthesis and CAM which are well known and need not be reiterated here. However, the observation that trees do not use phosphoenolpyruvate (PEP) carboxylation as widely as CAM and C4 photosynthesis are otherwise distributed among angiosperms shows an intriguing parallel between both modes of photosynthesis. Phylogenetic studies address the question of why this is so, but we must note here, at the outset, that in both cases the question has not been answered to date.
CAM is widespread among the taxa of vascular plants including ferns and fern allies (Isoëtes). There is no doubt that CAM arose polyphyletically many times during the evolution of vascular plants (Lüttge, 1987). This must have been facilitated by the fact that CAM does not require any particular metabolic trait as compared with non-CAM plants. A reorganization and specific management of metabolic modules common in the housekeeping functions of all photosynthesizing plants are sufficient for CAM performance (Lüttge, 2005). The evolution of CAM has been specifically studied in the genus Kalanchoë and the family Bromeliaceae and more recently also in the genus Clusia. There are conspicuous differences among the three taxa. A comparison (Lüttge, 2004, 2005) is useful as it reveals interesting peculiarities of Clusia.
The radiation centre of Kalanchoë is in the moist forests of eastern Madagascar (Kluge et al., 1991; Gehrig et al., 2001; Kluge, 2005). The genus comprises species with obligate C3 photosynthesis and obligate CAM and there are a small number of species that are C3/CAM intermediate and can switch between the two modes of photosynthesis (Brulfert et al., 1973, 1975). The majority of the species in the genus are CAM species. Morphological (Boiteau & Allorge-Boiteau, 1995) as well as molecular (Gehrig et al., 1997, 2000, 2001) evidence suggests that CAM must have evolved monophyletically in the genus from a C3-photosynthesis ancestor during radiation from the moist east to the drier areas up to the arid areas in the west and south-west of Madagascar, where CAM functions as an important adaptation to the limited availability of water.
Conversely, in the neotropical family of the Bromeliaceae with its three subfamilies Pitcairnioideae, Bromelioideae and Tillandsioideae and very many epiphytic species, again as supported by morphological/anatomical, biochemical and molecular genetic evidence (Smith, 1989; Crayn et al., 2004), CAM and the epiphytic habit evolved a minimum of three times. Epiphytism and CAM have evolved independent of each other, although evidently CAM is an adaptation to the water shortage stress of the epiphytic habitat (Zotz & Hietz, 2001). The last common ancestor of the Bromeliaceae was a terrestrial C3 mesophyte. It is important to note that there has also been a reversion from CAM back to C3 photosynthesis in some taxa of the bromelioid line following subsequent radiation into less xeric habitats (Crayn et al., 2004). Diversity is very large in the family in many respects but photosynthetic plasticity is low. There are obligate C3 and CAM species but in the entire large family there appears to be only one clearly C3/CAM-intermediate species, namely Guzmania monostachia (L.) Rusby ex Mez (Maxwell et al., 1994, 1995, 1999; Maxwell, 2002).
The subfamily Clusioideae with the genus Clusia in the family Clusiaceae has a minimum phylogenetic age of 90 × 106 years (Gustafsson et al., 2002). Clusia is a large genus of shrubs and trees with about 300–400 species (Pipoly et al., 1998). The genus is considered monophyletic. However, internal transcribed spacer (ITS) sequencing of nuclear ribosomal DNA has shown that the evolution of CAM in the genus was polyphyletic (Vaasen et al., 2002; Gehrig et al., 2003; Holtum et al., 2004). ITS analyses of some 80 species of Clusia and a derived phylogenetic tree suggest that CAM evolved at least twice in the genus. A reversion from CAM to C3 was noted in eight individual species and in one group of three species (Gustafsson et al., 2006). Thus the situation in Clusia resembles that in the Bromeliaceae and not that in Kalanchoë. However, in contrast to the Bromeliaceae, Clusia is characterized by an enormous plasticity. There are very many C3/CAM-intermediate species. It appears that this even may be the rule in the genus and obligate C3 and especially obligate CAM species are much fewer (Holtum et al., 2004).
This expectation is, however, based on a much smaller sample size. To some extent, the occurrence of CAM can be screened by analysing carbon isotope ratios (δ13C) because, as a result of the much lower 13C discrimination of PEPC (−7 relative to CO2) than of Rubisco (+27 relative to CO2) (Ziegler, 1994), in CAM-performing plants δ13C values are significantly less negative than in C3 plants. However, it has been noted that C3/CAM-intermediate species often make little use of their CAM option in the field, so that δ13C values are C3-like. Although Winter & Holtum (2002) and Winter et al. (2005) have presented a calibration for deducing from δ13C values the extent of primary dark fixation (PEPC) and light fixation (Rubisco) of CO2, respectively, for a definitive identification of a CAM potential in a species additional physiological evidence is required, for example from gas exchange measurements and analyses of diurnal organic acid fluctuations. It is much more tedious to obtain such information for many species than to just analyse δ13C values, but it appears to be very important to do this for as many species of Clusia as possible. This would not only provide insights into the evolution of CAM within the genus. In addition, because of the extraordinary inherent evolutionary plasticity in the genus and its high speciation rate (Gustafsson et al., 2006), the investigation of Clusia may also advance understanding of CAM evolution in general.
In summary, the comparison of the three taxa Kalanchoë, Bromeliaceae and Clusia, in which CAM phylogeny has been studied in the greatest depth, shows that the evolution of CAM not only occurred independently in the three taxa but also went rather different ways, i.e. monophyletically within the genus Kalanchoë, polyphyletically but without generation of much plasticity in the Bromeliaceae and polyphyletically with production of enormous plasticity in the genus Clusia.
The key enzyme of CAM as well as C4 photosynthesis is PEPC. Recently, the molecular characterization of PEPC isoforms has been investigated in detail in C4 photosynthesis (Engelmann et al., 2003; Svensson et al., 2003; Westhoff & Gowik, 2004) and also in CAM plants, with a strong phylogenetic bias. It has been found that Kalanchoë pinnata (Lam.) Pers. has seven PEPC isoforms, only one of which is CAM relevant (Gehrig et al., 2005). Taybi et al. (2004) studied the expression of PEPC and PEPC-kinase (PEPCK) genes in four different species of Clusia. PEPCKs regulate the activity of PEPC which is more active and less sensitive to feedback inhibition by malate in the phosphorylated state. The species compared were C. rosea, an obligate CAM species, Cluisa minor L., a C3/CAM-intermediate species, Clusia aripoensis Britt., a weakly CAM-inducible species, and Clusia multiflora H.B.K., an obligate C3 species. They found that transcriptional control of PEPC abundance is a key factor in determining the genotypic capacity of CAM, while the PEPCK gene is under the control of metabolic feedback, and there are no C3- or CAM-specific isoforms of the PEPCK genes. The obligate CAM species Clusia hilariana Schlecht. produces three isoforms of PEPC, one of which is root-specific and is probably a housekeeping isoform in this organ, while the C3 species C. multiflora and the C3/CAM-intermediate species C. minor only have one isoform (Vaasen, 2005; Vaasen et al., 2006). For C. minor the observation that there is only one isoform each of PEPC and PEPCK is most noteworthy because this implies that just one isoform can support both the housekeeping functions of the C3 state and the particular requirements of the CAM state. This may facilitate C3–CAM switches. PEPC transcript abundance is increased in the CAM state (Taybi et al., 2004; Vaasen, 2005; A. Vaasen, D. Begerov, R. Hampp, Botanisches Institut, Universität Tübingen, Tübingen, Baden-Württemberg, Germany). The observation that a single isoform each of PEPC and PEPCK can support performance of both C3 photosynthesis and CAM also strengthens the speculation that this has facilitated CAM evolution and that C3/CAM flexibility, which is so widespread in the genus Clusia, has played a role in this. However, whether the absence of more isoforms in C. minor is a result of a lack of evolution of more isoforms or the loss of redundant isoforms during evolution remains unresolved at present. More comparative work clearly is highly desirable. However, in the earlier literature it was not C3/CAM-intermediate behaviour but ‘CAM cycling’ that was suggested to have been a starting point for CAM evolution (Guralnick et al., 1986; Guralnick & Jackson, 2001), but this now takes us to the next section discussing photosynthetic plasticity in Clusia.
III. Photosynthetic physiotypes
In Fig. 1, I have reduced metabolic complexity to the extent of only using two modules, namely PEPC and organic acids and Rubisco and carbohydrate ([CH2O]n), respectively, and external and internal CO2 for a schematic illustration of four different photosynthetic physiotypes, all of which are found among Clusia species.
1C3 photosynthesis, where stomata are open during the light period and CO2 taken up is fixed directly via Rubisco in the light.
2CAM, where stomata are open in the dark period and CO2 taken up is fixed via PEPC leading to the production of organic acid (both malic and citric in Clusia species), organic acid is decarboxylated during the light period and the CO2 regained is refixed via Rubisco. Some direct C3-type fixation is also possible, i.e. in the later light period, when ecological conditions, particularly water relations, allow stomatal opening after nocturnally accumulated organic acid is consumed. CAM idling, as also shown in Fig. 1, is not considered a separate physiotype but a variant of normal full CAM, where stomata remain closed night and day when water relations are adverse and respiratory CO2 is refixed in the dark period and recycled to carbohydrate during the light period.
3C3/CAM-intermediate behaviour with reversible switches between the two modes of photosynthesis in perennial Clusia, where leaves are used for at least two growth periods (Olivares, 1997).
4CAM cycling, where CO2 uptake and fixation occur mainly during the light period as in C3 photosynthesis, but respiratory CO2 is recuperated via PEPC behind closed stomata during the dark period and organic acid is produced which is decarboxylated during the light period, supplementing the CO2 taken up from the atmosphere via the open stomata as substrate for Rubisco. This use of PEPC to prevent loss of respiratory CO2 in the dark period was thought to have been an early step towards the evolution of CAM, as alluded to in the previous section (Guralnick et al., 1986; Guralnick & Jackson, 2001).
The scheme illustrated in Fig. 1 symbolizes the modular nature of photosynthetic physiotypes and can be used to elucidate the links that might be important in long-term and medium-term evolutionary and ontogenetic development, respectively, as well as in short-term environmental responses in ecophysiology. For example, (i) CAM can also involve a strong C3 element; (ii) without light-period CO2 uptake CAM cycling would become CAM idling.
Some representative curves of day:night CO2 exchange in Clusia species are shown in Fig. 2. A bona fide obligate C3 species is C. multiflora and obligate CAM species are Clusia alata Pl. et Tr., Clusia major L., C. rosea and C. hilariana. As noted in the previous section, there are very many C3/CAM-intermediate species in the genus Clusia. The most widely studied species in this respect is C. minor. Under different conditions with respect to a combination of different environmental factors, such as irradiance during growth and a variety of day:night temperature regimes, it produced a vast number of different responses including full CAM with CAM idling (Haag-Kerwer et al., 1992). In another experiment, irradiance during growth and nitrogen nutrition were varied. Plants grown at high irradiance without supplemental nitrogen showed no CO2 uptake and a minimal loss of CO2 but accumulated appreciable amounts of citrate plus a small amount of malate in the dark period, while light-period CO2 uptake was substantial, i.e. there were clear indications of CAM cycling (Franco et al., 1991). Most of the experiments cited here were performed with one clone of vegetatively propagated C. minor. The enormous phenotypic plasticity with respect to expression of photosynthetic physiotypes of the genome is evident.
IV. Metabolic flexibility: organic acid variations
Clusia also shows flexibility at the level of metabolites in the CAM cycle. Normally the precursor for glycolytic formation of phosphoenolpyruvate (PEP) as a CO2 acceptor for dark fixation is glucan (starch) and the acid produced is malic acid (malate). In Clusia, however, free sugars, i.e. glucose, fructose and sucrose, are also used in addition to starch as PEP precursors (Popp et al., 1987; Ball et al., 1991; Berg et al., 2004). Although noteworthy, this situation is not unique as it also occurs occasionally in other taxa, for example in the bromeliad pineapple Ananas comosus (L.) merr. (Black et al., 1996). More remarkable is the fact that Clusia, in addition or alternatively to malate, can also synthesize citrate. There are other CAM plants that nocturnally form citrate in their CAM cycle but the amounts are far lower than those found in Clusia; for example, day:night oscillations of citrate concentrations in some species of Kalanchoë may be up to 26 mm, and this is considered relatively high (Lüttge, 1988), but C. minor and C. rosea can show values of up to 200 mm (Franco et al., 1992). This really amounts to a qualitative rather than only a quantitative difference. The relative amounts of malate and citrate produced can vary within the same species, clone or plant depending on environmental conditions, where again C. minor is an excellent example (Haag-Kerwer et al., 1992).
I can list three cases where differential responses of nocturnal malate and citrate accumulation in the CAM of Clusia species to environmental factors have been observed.
1Under drought conditions in experiments by Franco et al. (1992) and de Mattos et al. (1999) with three Clusia species (Clusia lanceolata Camb., C. rosea and C. minor) the ratio of malate to citrate accumulation declined by a factor of 1.8–3.4 after 10 to 16 d without watering. This was, however, not observed in a study by Borland et al. (1998) with C. minor, C. rosea and C. aripoensis, but in this case the ratios of malate to citrate were already very low in the nondroughted controls.
2In C. minor the ratio of malate to citrate accumulated also depended on nitrogen nutrition and irradiance during growth (Franco et al., 1991). At low irradiance there was no malate and only citrate accumulation with and without supplemental nitrogen nutrition. At high irradiance the ratio of malate to citrate was 0.22 without and 3.07 with nitrogen.
3In C. minor the ratio of malate to citrate was also affected by day:night (D:N) temperature regimes and irradiance during growth. For example, low irradiance and D:N temperatures of 30 : 15°C produced only malate accumulation; high irradiance and D:N temperatures of 25 : 15°C or 30 : 15°C produced high malate and low citrate accumulation; high irradiance and D:N temperatures of 25 : 20°C produced equal malate and citrate accumulation; low irradiance and D:N temperatures of 30 : 30°C produced only citrate accumulation; high irradiance and D:N temperatures of 20 : 20°C produced high citrate and low malate accumulation (Haag-Kerwer et al., 1992).
Naturally one would expect that a switch in metabolism as important as a change from malate to citrate accumulation should have adaptive benefits in response to the various environmental cues. However, the picture created by the above three examples of environmental responses remains very fuzzy, and we can only try to evaluate comparatively a number of roles of malate and citrate, respectively, which I summarize below in eight points (Table 1; Franco et al., 1992; Lüttge, 1996, 2006a).
Table 1. Potential advantages and disadvantages, respectively, of night:day change in the concentrations of malate (Δmal) and citrate (Δcitr) during the crassulacean acid metabolism (CAM) cycle (Franco et al., 1992; Lüttge, 1996, 2006a)
1. CO2 acquisition
2. H2O saving during CO2 acquisition
3. Carbon recycling in the dark period
4. Carbon recycling in the light period and increase in internal CO2 concentration
CO2 Yes, more than for malate
5. Energy budget in the dark period
ATP consumption and production of redox power
6. Energy budget in the light period
Larger ATP consumption and consumption of redox power
7. Osmotic changes
Yes, less than for malate
8. Buffer capacity
1With respect to day:night carbon balance, the CAM cycle with citrate is futile. Nocturnal citrate accumulation does not contribute to carbon gain because, starting with a C6-hexose unit as precursor, the result is C6-citrate. Conversely, two molecules of C4-malate are formed by nocturnal fixation of 2 CO2 per C6-hexose unit used (Lüttge, 1988).
2As it does not contribute to carbon gain, citrate, of course, also does not contribute to water saving during CO2 acquisition, unlike malate synthesis by nocturnal CO2 uptake when evaporative demand is much lower than during the day.
3Like malate, citrate can contribute to carbon recycling during the dark period, for example in CAM idling. With malate accumulation respiratory CO2 is recycled; with citrate accumulation entire carbon skeletons are recycled.
4Like malate, citrate can contribute to CO2 recycling in the light period. Potentially more CO2 is generated from citrate breakdown in the light period so that citrate can make a larger contribution to the build-up of high internal CO2 concentrations behind closed stomata than malate.
5Energetically, citrate accumulation is marginally cheaper in the dark period. ATP is consumed in similar amounts in both malate and citrate accumulation but reduction equivalents are generated in citrate synthesis which can be used for ATP production in the respiration chain.
6Energetically, the return of carbon from citrate to carbohydrate in the light period appears to be more costly than in the case of malate. The decarboxylation of two molecules of malate gives rise to two molecules of pyruvate, which can be used to regenerate two molecules of triose and then one molecule of hexose via gluconeogenesis. The decarboxylation of one molecule of citrate may generate three molecules of CO2 and only one molecule of pyruvate. The refixation of CO2 to generate triose is much more costly energetically than gluconeogenesis starting from pyruvate.
7The nocturnal accumulation of malate in the vacuoles of CAM plants has osmotic consequences and can support nocturnal water storage (Eller & Ruess, 1986; Lüttge, 1986; Ruess et al., 1988; Eller et al., 1992; Murphy & Smith, 1998). The CAM cycle with citrate is less effective in this respect; for example, for one hexose unit from glucan, two osmotically active malate molecules and only one citrate molecule are formed. If the precursor is free hexose, citrate formation is osmotically neutral.
8Citrate is a more effective pH-buffer compound than malate. The buffer capacity of the vacuole is important for nocturnal organic acid accumulation because it determines the trans-tonoplast proton gradient against which the proton pumps energizing acid accumulation must work (Lüttge et al., 1981). A higher buffer capacity may therefore allow an increased capacity for organic acid accumulation (Franco et al., 1992). Thus, citrate accumulation in Clusia species may be an explanation of the fact that in Clusia species we observe the highest nocturnal accumulation of acids ever found among CAM plants, namely 1410 mm titratable protons in one example of C. minor in the field in Trindiad (Borland et al., 1992).
Hence, there is no benefit from citrate accumulation for carbon gain and improvement of water relations (points 1, 2 and 7 above). Conversely, citrate is useful as it could facilitate a larger total accumulation of acids in the dark period (point 8) which contributes to the build-up of high internal CO2 pressures in the light period, in which citrate is more effective than malate (point 4). This is useful because high internal CO2 concentrations may bring Rubisco closer to substrate saturation, or actually achieve substrate saturation, minimizing photorespiration, photoinhibition and photodestruction at high irradiance during the light period (Lüttge, 2002). The higher energy demand of citrate cycling in the light period (point 6) in this respect may also be considered beneficial as it contributes to energy dissipation, reducing the danger of photoinhibition and photodestruction at high irradiance. In this respect, observations in the field in Trinidad are interesting, where citrate breakdown in C. minor during the day showed a much closer correlation with irradiance as compared with malate (Borland et al., 1996). Thus, one might expect that citrate accumulation might be more favourable than malate accumulation under conditions of drought, when stomata partially or fully close during the dark period and internal carbon recycling and CAM idling become increasingly important, and also because drought is often associated with high-irradiance stress. However, apart from the observation that malate:citrate ratios during nocturnal acid accumulation under drought conditions may decrease, the other results on differential effects of environmental factors on malate and citrate accumulation, respectively, presented above do not show any obvious relations to the theoretically expected benefits from citrate accumulation.
It is astonishing that such a fascinating problem as the putative functional advantage of the metabolic plasticity of performing CAM with malate or citrate is not understood. Even information about the metabolic pathway involved in nocturnal citrate synthesis is scarce (Olivares et al., 1993), and we have no information on light-period citrate metabolism in Clusia, so that we are left with known general biochemistry of intermediate metabolism (see also Holtum et al., 2005), for example in the assessment of energy budgets, as mentioned above. Some classical biochemical studies would be very interesting, and rigorously systematic comparative studies varying environmental parameters may provide important new insights. However, current fashions in plant biology have a very different focus from motivating funding of the study of such an exciting problem in biochemical ecology.
V. The environmental control of photosynthetic flexibility
Environmental input is always received by the phenotypes, and in the case of the different modes of photosynthesis these are the photosynthetic physiotypes distinguished above (Fig. 1). There is then a possible feedback to the genotype which may respond with changes of the phenotype expressed (Lüttge, 2005). More straightforwardly, we can ask the questions of which environmental factors determine the reversible switches between C3 photosynthesis and CAM in Clusia species and how this works. As an example, we have already seen above that quite specifically the expression of PEPC may change during a C3 to CAM transition. The major external factors are water, irradiance and temperature (Lüttge, 2004). In the field they never act individually but always interactively. Even in experiments they are often hard to separate. However, individual factors may become limiting and we may try to consider them separately.
Although the earliest environmental pressure for the evolution of CAM was probably the problem of CO2 acquisition with the advantage of the 60 times higher CO2 affinity of PEPC than Rubisco and the related function of CAM as a CO2-concentrating mechanism (Lüttge, 2002), it is evident that the driving force for the polyphyletic evolution of CAM in so many taxa of the vascular plants must have been limited availability of water.
In the C3/CAM-intermediate C. minor, CAM can be induced by drought. In an experiment where this was done it was shown subsequently that even the two opposite leaves on one node simultaneously could perform C3 photosynthesis and CAM, respectively. The two leaves were kept in gas exchange cuvettes at a low (6.2 mbar bar−1) and high (13.1 mbar bar−1) water vapour pressure deficit (VPD) of the air, respectively, i.e. subjecting the two leaves to different evaporative demand. When the plant was re-watered after 4 d of drought, the leaf in the dry air continued to perform CAM with dominating CO2 uptake and fixation in the dark, while the leaf in the moist air within a few hours switched to CO2 uptake in the light (Schmitt et al., 1988, where C. minor is wrongly called C. rosea).
While this illustrates the plasticity of a single plant, K. Winter and colleagues have tried to carry out a more general survey of Clusia species, assessing water use efficiency (WUE) by carbon isotope analysis (δ13C). WUE is defined as the ratio of moles of CO2 fixed and assimilated to moles of water lost by transpiration. WUE is high in nocturnal CO2 fixation during CAM because CO2 acquisition in the dark occurs at low evaporative demand. WUE is much lower in C3 photosynthesis in the light and also in CAM plants when they make use of their potential for direct CO2 fixation via Rubisco in the later light period (Fig. 1). As the 13C discrimination of PEPC is so much lower than that of Rubisco, analysis of δ13C values allows screening (see section II) of the extent to which plants use primary CO2 fixation via PEPC and Rubisco, respectively (Borland et al., 1993), and allows this to be related to WUE. Winter & Holtum (2002) and Winter et al. (2005) have shown in general that δ13C values in CAM species strongly correlate with WUE. These authors also screened 38 species of Clusia from Panamá and found that under natural conditions in the field Clusia species only sparingly make use of their PEPC option of CO2 fixation (Holtum et al., 2004) and, thus, the related high WUE. This underlines the importance of other factors in addition to water in the habitats of Clusia species.
High irradiance increases VPD and may cause over-energization of the photosynthetic light-harvesting apparatus. Thus, one would expect that it would support the use of the CAM option in plastic Clusia species. However, this is not always so. In well-watered C. minor an increase of the irradiance from 360 to 1200 mol m−2 s−1 suppressed CO2 dark fixation (Schmitt et al., 1988). In studies of the return from CAM to C3 photosynthesis in C. minor, de Mattos & Lüttge (2001) found that, provided water is not limiting and the plant can afford increased transpiration, daily photon use increases when there is unrestricted C3-like CO2 uptake, where the plant may overcome the limitations of the storage capacity of the vacuole for nocturnal organic acid accumulation, improving its daily carbon balance.
A comparative phytotron study of the obligate C3 species C. multiflora and the C3/CAM-intermediate species C. minor (Herzog et al., 1999) is interesting because it also sheds light on the behaviour of the species and their relative use of the CAM option in the field. Plants of the two species were grown at a low daily dose of irradiance (4 mol m−2 d−1) and then transferred to two higher daily doses of 24.5 and 33.5 mol m−2 d−1, respectively. Daily irradiance was applied in a bell-shaped pattern as in nature, i.e. increasing from dawn towards noon and decreasing again towards dusk. Figure 3 shows some results obtained 5 d after the transfer. We can see that the C3 species which had maximum rates of CO2 uptake () between 2 and 4 mol m−2 s−1 at 4 mol m−2 d−1 could make use of the higher irradiance of 24.5 mol m−2 d−1 for increasing up to 6 mol m−2 s−1 but was strongly inhibited by the still higher irradiance of 33.5 mol m−2 d−1 showing only a of < 1 mol m−2 s−1. Conversely, the C3/CAM-intermediate species which had performed C3 photosynthesis with maximum rates of of c. 2–3 mol m−2 s−1 at 4 mol m−2 d−1 had switched to CAM under the higher daily doses of irradiance with high potential rates of even in the transition phase between nocturnal CO2 uptake and daytime stomatal closure in the morning and during C3-like CO2 fixation in the later part of the day, with no difference between the two high daily doses. At 4 mol m−2 d−1 neither species was photoinhibited. However, under the higher irradiances they were similarly photoinhibited and increasingly so in the second part of the light period after irradiance was increased from 24.5 to 33.5 mol m−2 d−1 as given by the potential quantum yield of photosystem II (Fv/Fm, where Fv is the variable and Fm is the maximum fluorescence of dark-adapted leaves and a ratio below 0.83–0.80 indicates photoinhibition; Björkman & Demmig, 1987). Nonphotochemical quenching (NPQ) of chlorophyll (Chl) a fluorescence of photosystem II was also similar for the two species, as were the levels of the xanthophyll zeaxanthin, which is involved in harmless thermal energy dissipation under irradiance stress (Hager 1980; Demmig-Adams, 1990; Demmig-Adams & Adams, 1992; Horton et al., 1994; Pfündel & Bilger, 1994; Schindler & Lichtenthaler, 1996). With the putative protective function of the high internal CO2 concentrations built up during CAM in the light period (see section IV), one would have expected that Fv/Fm would have been higher and NPQ and zeaxanthin levels lower in the C3/CAM-intermediate species after its switch to CAM than in the C3 species. That this is not so shows that CAM species, like C3 species, are not exempt from photoinhibition and have protective mechanisms, such as thermal energy dissipation, to deal with irradiance and oxidative stress, as was also observed in the CAM species C. hilariana (Franco et al., 1999). Measurements by Winter et al. (1990) showed that in the early light period in CAM-performing C. rosea the level of zeaxanthin greatly increased from much lower levels maintained in the dark period, decreased in the middle of the day when stomata closed and organic acid was remobilized, and subsequently increased again as stomata opened in the later light period. The levels of violaxanthin, the double epoxide of zeaxanthin, showed the opposite pattern during the day. With respect to the pair of Clusia species, C. multiflora and C. minor, compared by Herzog et al. (1999), the long-term comportment after transfer to higher irradiance is important. In the C3 species, under higher daily irradiance leaves became necrotic and died. However, the plant was not principally suffering from the high irradiance. It only suffered from the sudden change from low irradiance during growth to higher irradiances, probably because acclimation of the light-harvesting apparatus takes a long time or is totally insufficient. However, C. multiflora then produced new leaves from dormant buds which were adapted to the new situation. Conversely, the C3/CAM-intermediate species C. minor was not damaged after the transfer to higher irradiance. It was probably the much more rapid switch of the biochemical machinery to CAM combined with protective mechanisms which saved it from destruction of its leaves. In fact, separate experiments showed that the capacity to produce zeaxanthin was even higher in C. minor (c. 130 mol m−2 leaf area) than in C. multiflora (c. 60 mol m−2) but it did not make use of this higher capacity, at least up to 5 d after transfer (Fig. 3).
These phytotron experiments suggest that the C3/CAM-intermediate species may have a larger niche width, performing well and with high short-term flexibility both under full sun exposure and in more shaded situations, while the C3 species performs well under high irradiance only when it has long-term development under full exposure, for example starting with seedling growth. Indeed, in the field it has been observed that C. multiflora occupies exposed sites in a secondary savanna while C. minor prefers semi-shaded sites in a deciduous dry forest but can also invade the exposed sites of C. multiflora (Franco et al., 1994; Grams et al., 1997).
Studies on species of Clusia show that temperature effects are often closely related to effects of irradiance (Haag-Kerwer et al., 1992). We have already seen in section IV that, depending on irradiance during growth, day:night temperature regimes may strongly determine the extent to which either malate or citrate or both are used in the CAM cycle. In C. minor it was found that the amplitude of the day:night (D:N) temperature difference is important in eliciting a C3 to CAM switch and this is modulated by irradiance during growth. At low irradiance, plants performed C3 photosynthesis when D:N temperatures were 30 : 30, 30 : 25 and 30 : 20°C and switched to CAM at 30 : 15°C. A smaller D:N temperature difference was required for the switch in high-irradiance-grown plants which performed C3 photosynthesis at 30 : 30°C but already switched to CAM at 30 : 25°C. At 15 : 15°C gas exchange was reduced under both irradiances, but high-irradiance-grown plants performed CAM and low-irradiance-grown plants C3 photosynthesis. The changes in the mode of photosynthesis were always readily reversible when temperature regimes were changed again.
VI. Phenotypic plasticity: physiotypes and morphotypes
Clusia species are woody shrubs and trees with a dichasial cyme. In an earlier review (Lüttge, 1999) I proposed considering all species of Clusia to belong to a single morphotype. Saying this, I referred to the vegetative photosynthetic organs, the leaves, which, although they may have greatly differing sizes in different species, are all somewhat leathery and succulent. Diversity is revealed when we consider the physiotypes with high plasticity of modes of photosynthesis and metabolic pathways during CAM performance, as described in sections III and IV. At the morphotypic level diversity is very high in the generative organs (Engler, 1925; Bittrich & Amaral, 1996; Gustafsson & Bittrich, 2003; Gustafsson et al., 2006). Flower size may range in diameter from 5 to 150 mm. There is substantial diversity in floral morphology with respect to size, degree of fusion and number of floral parts, position of staminodes and stamen and anther morphology. Pollination rewards are nectar, pollen and resin. Floral resin is a rare award. It has evolved polyphyletically in the genus of Clusia (Gustafsson et al., 2006). Diversity at the morphological level is also revealed by the different life forms of Clusia species. They may germinate terrestrially and directly develop into independent free-standing trees. They may also germinate epiphytically, predominantly in humus niches between branches of host trees or of other epiphytes such as epiphyte nests or even the tanks of bromeliads. Then they can establish soil contact via aerial roots, i.e. they are hemi-epiphytes. Other adventitious roots may strangle the bark of the host tree and thus murder the phorophyte, the stem of which may then rot away under a network of strangler roots so that the previous hemi-epiphyte eventually becomes a free-standing tree (Lüttge, 1991). This diversity does not only occur among species, because even individual species can express many of these life forms, so that we have here another example of plasticity in the genus (Fig. 4).
In contrast to other hemi-epiphytes, mainly in the genus Ficus, several morphological/anatomical traits, such as leaf thickness, and stomatal density and size, did not differ much between the epiphytic and rooted tree life form of C. minor (Holbrook & Putz, 1996). Often the hydraulic architecture must be adapted in hemi-epiphytes as compared with rooted trees for sufficient water supply. In a comparative study of the hemi-epiphytic C3/CAM-intermediate Clusia uvitana Pittier with other hemi-epiphytes, mainly of the genus Ficus (Zotz & Winter, 1994a,b; Patiño et al., 1995; Zotz et al., 1997), it turned out that C. uvitana had less effective unit leaf area, a rather large leaf area per unit of stem cross-sectional area and a lower specific stem conductance. This suggests that physiotypic flexibility with the water-saving CAM option also relates to a morphotypic trait such as hydraulic architecture.
VII. Ecological amplitude and habitat impact
Physiological ecology has always largely been autecology. We have noticed, however, that it is increasingly developing the potential to make contributions to synecology (Lüttge & Scarano, 2004). This is in the cast of mind of Sir Arthur Tansley, whose name is heading the series of reviews to which the current contribution belongs, because quite early ‘He made it clear that synecology and autecology are subsumed in the study of the community by methods firstly descriptive and appreciative, and secondly, analytic and experimental’ (Godwin, 1977). Thus, in this vein, in this section I shall describe the community performance of Clusia. However, one may bear in mind here that laboratory studies of autecology mostly cover the dynamics of short-term acclimations, while field observations represent momentary pictures of long-term adaptations. Long-term laboratory observations are as rare as recordings of gradual responses under gradually varying, for example seasonally, natural conditions. The continuous long-term studies in the field which I allude to below are quite rare.
With the great plasticity found in the genus Clusia we may expect that it will span a large ecological amplitude. Indeed, it is observed that the genus, with its single simple vegetative morphotype, occupies a wide range of habitats in the neotropics, as listed in Table 2 and illustrated in Fig. 5. Not only the genus as a whole but also some individual species, for example the C3 species C. multiflora, the CAM species C. rosea and the C3/CAM-intermediate species C. minor, cover almost the entire range of different habitats listed in Table 2, and some species occur in rather contrasting habitats, such as the coastal restingas of Brazil and inselbergs (the C3 species Clusia parviflora Saldanha et Engl.) and the restingas and dry lowland forest (the CAM species Clusia fluminensis Pl. et Tr.). In the following I shall list a number of sites occupied by Clusia species and try to evaluate the environmental impact of the genus. I shall do this by comparing the ecophysiological performance of Clusia species with those of other co-occurring plants with similar life forms, i.e. shrubs and trees. Much field work is in fact reported in the literature and an ecophysiological comparison can be based on traits such as carbon isotope ratios, rates of CO2 exchange, Chl a fluorescence parameters of photosytem II and day:night changes of organic acid levels in CAM-performing species. Space does not allow me to go into as much detail as elsewhere (Lüttge, 1999, 2006b), but the advantage of a briefer treatment is that it provides a more concise overview.
Table 2. Ecological amplitude of Clusia with the range of habitats where Clusia is found, and occupation of a large number of these habitats by three selected species with obligate C3 photosynthesis, obligate crassulacean acid metabolism (CAM) and C3/CAM-intermediate behaviour
C. muliflora (C3)
C. rosea (CAM)
C. minor (C3/CAM)
Gallery forest – cerrado ecotone
Dry lowland forest
Secondary shrub forest
Dry montane karstic limestone forest
Upper montane rainforest
Cloud forest/fog forest/elfin forest
1. The Brazilian restinga complex
The restinga complex consists of a mosaic of plant communities on the sandy coastal plains ranging from open formations to forest ecosystems (Lacerda et al., 1993). The open formation is characterized by shrub islands surrounded by white sand. There are sand dune ridges with dry forest in between and on fixed dunes (Cirne & Scarano, 2001). Species of Clusia are important floristic elements of the restingas, which also have been called ‘Clusia scrub’ (Ule, 1901). According to annual rainfall, height of the ground water table and salinity of the ground water, we may distinguish wet, intermediate and dry restingas. The major Clusia species of the restingas are the C3 species C. parviflora, and the CAM species C. hilariana and C. fluminensis, the former of which is much more dominant than the latter.
In the wet restinga of Jacarepiá (22°47′ to 22°57′ S, 42°20′ to 42°43′ W), C. fluminensis has been compared with the C3 Fabaceae species Andira legalis Vell. Toledo (Geßler et al., 2005a; Scarano et al., 2005). C. fluminensis developed somewhat larger maximum apparent electron transport rates of photosynthesis (ETRmax) and with respect to photoinhibition the two species were very similarly affected at midday and slightly before dawn. In the intermediate restinga of Jurubatiba (22°00′ to 22°23′ S, 41°15′ to 41°45′ W), C. parviflora and C. hilariana were compared with the C3 species A. legalis, Rheedia brasiliensis (Mart.) Pl. et Tr. (another Clusiaceae) and Myrsine parvifolia A. DC. (Myrsinaceae) (Franco et al., 1996; Duarte et al., 2005; Scarano et al., 2005). The dominant C. hilariana (CAM) showed superior photosynthetic capacity (ETRmax) as compared with the other four species, and the C3 species C. parviflora was similar to the other C3 species. Chronic photoinhibition not reversible overnight was absent or low in all species, but an advantage of CAM in avoiding acute photoinhibition at midday was not seen. In the intermediate restinga of Barra de Maricá (22°53′ S, 42°52′ W) C. lanceolata and C. fluminensis were studied (Roberts et al., 1996; Reinert et al., 1997). Both are potential CAM species but δ13C values indicated that C. lanceolata was mostly fixing CO2 by the C3 mode, while C. fluminensis was shifted more towards CAM. Chronic photoinhibition was absent and ETRmax and acute photoinhibition were similar in the two species so that the relative performance of the two modes of photosynthesis did not make much difference. In the dry restinga of Massambaba (22°56′ S, 42°13′ W), C. fluminensis mainly performing CAM and for comparison A. legalis, R. brasiliensis and M. parviflora were investigated (Duarte et al., 2005; Geßler et al., 2005a; Scarano et al., 2005). All species showed similar photosynthetic capacity and acute photoinhibition and there was no obvious advantage of CAM for C. fluminensis. In the dry forest of Buziós (22°49′ S, 41°59′ W), C. fluminensis was compared with three Fabaceae trees, Caesalpinia echinata Lam., Caesalpinia ferrea Mart. ex Tul. and Machaerium obovatum Kuhlm. et Hoehne, and the Euphorbiaceae Croton compressus Lam. (Geßler et al., 2005b; Scarano et al., 2005). C. fluminensis achieved somewhat higher ETRmax than C. compressus and higher rates than the other species. In terms of acute and chronic photoinhibition it was not better off than the two Caesalpinia species or C. compressus but was superior to M. obovatum. C. fluminensis in the very dry forest must be limited by the moisture regime (Scarano et al., 2005) and again there appeared to be no clear advantage of CAM at the dry site. ETRmax measurements (mol m−2 s−1) indicated that C. fluminensis showed a strong dependence on a moisture gradient from wet restinga (290) to dry restinga (160) and dry forest (100).
Overall, the potential of Clusia species for CAM performance may have some subtle benefits but does not appear to provide a very obvious conspicuous advantage as compared with C3 species of similar life form. I have alluded to the fact that Clusia may dominate the restingas in places. Moreover, several studies have shown that Clusia species, especially C. hilariana, can have important pioneer functions in the restingas, starting vegetation islands on the bare sand plain and functioning as nurse trees under the canopy of which other vegetation may become established (Liebig et al., 2001; Dias et al., 2005). In ecology as well as in evolution (Gould, 2002), subtle differences can become decisive. However, functions other than photosynthetic capacity discussed here could be essential, for example reproductive biology. However, in the restingas, sexual reproduction of Clusia species is absent or very rare and we must assume that all growth is clonal.
2. Northern coastal mountain ranges in Venezuela
In the northern coastal range of Venezuela near Caracas (10°24′ N, 66°58′ W, 1500–1740 m asl) the same two Clusia species, C. multiflora and C. minor, investigated in a phytotron study in relation to the irradiance factor (section V.2) and physiotypic plasticity described in section III were measured in the field (Franco et al., 1994; Grams et al., 1997). In comparison to the behaviour discussed in section V.2 the field performance was similar (Fig. 6). The C3 species C. multiflora potentially could achieve higher maximum rates of CO2 uptake, -max, in the light. Notwithstanding the higher -max in C. multiflora at higher irradiance, there was a pronounced midday depression. Chronic photoinhibition at dawn and acute photoinhibition at midday (Fv/Fm values < 0.8) were similar in the two species. Nocturnal CO2 uptake in C. minor when performing CAM increased from low to medium irradiance and then decreased again at the highest irradiance, and this was reflected in the total integrated CO2 uptake for the dark periods and night:day oscillations of organic acids (not shown). Integrated CO2 uptake over 24 h was much higher under high than under low irradiance in C. multiflora, and in C. minor at the highest irradiance day-time CO2 uptake was also somewhat higher and night-time CO2 uptake was reduced as compared with the shaded plants at the lowest irradiance (Franco et al., 1994; Grams et al., 1997). C. minor when performing CAM always had much higher WUE than C. multiflora.
In the Sierra de San Luis in the state Falcón (11°18′ N, 69°45′ W, 1200 m asl) on a karstic limestone ridge the performance of the C3 species C. multiflora was compared with that of three CAM species, C. rosea, C. alata and an unidentified species (Popp et al., 1987; Franco et al., 1994; Haag-Kerwer et al., 1996). The CAM species made considerable use of their CAM option at this site. Photosynthetic capacity was similar in all four species. However, ETRmax values were higher in the CAM species as a result of the high internal CO2 concentrations built up during acid remobilization in the light period. All four species showed slight chronic and pronounced acute photoinhibition at dawn and midday, respectively.
3. Semi-evergreen moist tropical forest, Barro Colorado Island, Panama
In this forest (9°10′ N, 79°51′ W) the hemi-epiphytic C. uvitana was compared with other hemi-epiphytes by Zotz & Winter (1994a,b). These authors made an important contribution to the understanding of a possible advantage of the CAM option in C3/CAM-intermediate species. The advantage of the CAM option in structuring hydraulic architecture in hemi-epiphytes has been mentioned in section VI. However, the two authors also covered an entire 12-month period with numerous measurements, one of the rare studies meeting the need for continuous long-term studies in the field, and thus revealing an important seasonal effect. In the wet season, with respect to both CO2 uptake integrated over 24 h and maximum rates of , the performance of C. uvitana was close to that of a C3 fern Polypodium crassifolium L. and inferior to the that of the C3 orchid Catasetum viridiflavum Hook. However, in the dry season, when C. uvitana switched to CAM it was superior to the other two species.
4. Tropical forest on the island of Trinidad
A similar advantage of the CAM option in response to fluctuations of habitat conditions in wet and dry seasons was also shown for Clusia species in Trinidad: C. minor, C. tocuchensis, C. aripoensis and C. intertexta. This is another example of rather continuous long-term studies, where especially for C. minor the advantage of C3/CAM switches for performance over the seasons was demonstrated (Borland et al., 1992, 1993, 1994, 1996; Roberts et al., 1998).
5. Inselberg, Pão de Açúcar, Rio de Janeiro
On this inselberg (22°57′ S, 43°59′ W) the C3 species C. parviflora and the two Euphorbiaceae C. compressus and Styllingia dichotoma Muell. Arg. form bushes and shrubbery. The performances of all three species are quite similar (Duarte et al., 2005; Scarano et al., 2005).
VIII. Conclusions and outlook
Clusia, the only dicotyledonous tree genus with CAM, presents all facets of CAM and with its extraordinary plasticity it allows the study of CAM in all its ramifications. It suggests that not so much CAM per se but the high flexibility in carbon acquisition under variable environmental conditions in the tropics inherent in CAM (Fig. 1) is its real ecophysiological advantage. Plasticity with segregation and separation of ecotypes based on the high ecological amplitude might relate to the high speciation rate in the genus (Lüttge, 1995a,b, 2005). The radiative power of the genus is also great. This can be seen in the Atlantic rain forest complex of Brazil, where the Atlantic forest itself on tertiary ground is surrounded by a variety of plant communities (Fig. 7), some of which are also older relicts, such as the dry forest, but some of which are quite young, such as the restingas and swamp forests on tertiary ground. The dioecious species C. hilariana in these communities is a migrant from the Atlantic forest with missing or extremely limited generative propagation in the restingas because its specific pollinators from the Euglossini tribe, for example the bee Trigona spinipes, did not follow it to the new habitat (Scarano et al., 2004).
My monographic review on the genus Clusia may have shown that there is a great potential for future study of this single diverse genus spanning all scaling levels from molecular genetics to ecosystems.
1While the molecular phylogeny of Clusia has advanced, its scope and data base need to be extended and the molecular approach must also be used in population genetics.
2Studies of metabolism should evaluate the organization and management of metabolic reaction modules in the flexibility of photosynthetic physiotypes. Metabolomics will be a useful approach, but some classical biochemistry is also required, for instance regarding the plasticity in carbohydrate metabolism and the citric acid enigma.
3More rigorously comparative quantitative laboratory and field studies, such as the comparison of C. multiflora and C. minor in the phytotron and in the field discussed in this review, should be performed to evaluate ecophysiological capacities and potential environmental impact.
4In addition to assessment of ecophysiology, reproductive biology should be advanced to foster understanding of the performance of Clusia at the community level.
I thank Dr Annie Borland, Newcastle-upon-Tyne, UK, for valuable comments.