Ecophysiology of constitutive and facultative CAM photosynthesis

Abstract

In plants exhibiting crassulacean acid metabolism (CAM), CAM photosynthesis almost always occurs together with C₃ photosynthesis, and occasionally with C₄ photosynthesis. Depending on species, ontogeny, and environment, CAM input to total carbon gain can vary from values of <1% to 100%. The wide range of CAM phenotypes between and within species is a fascinating example of functional diversity and plasticity, but poses a significant challenge when attempting to define CAM. CO₂ gas exchange experiments designed for this review illustrate key patterns of CAM expression and highlight distinguishing features of constitutive and facultative CAM. Furthermore, they help to address frequently recurring questions on CAM terminology. The functional and evolutionary significance of contrasting CAM phenotypes and of intermediate states between extremes is discussed. Results from a study on nocturnal malate accumulation in 50 species of Aizoaceae exposed to drought and salinity stress suggest that facultative CAM is more widespread amongst vascular plants than previously thought.

Acidity, carbon assimilation, evolution, facultative CAM, Hatiora, Kalanchoe, ontogeny, photosynthesis, photosynthetic intermediate, Portulaca

Introduction

Research on the functional genomics of crassulacean acid metabolism (CAM) plants is rapidly advancing. Elucidating the relationship between genome and phenotype is key to the understanding of CAM evolution and for introducing CAM into C₃ crop plants to enhance their water use efficiency. Concomitant with improving our understanding of the molecular underpinnings of CAM, ecophysiological research is increasing our appreciation for the large phenotypic variation amongst CAM-exhibiting species, in terms of (i) CAM usage for carbon gain relative to C₃ (or C₄); and (ii) how CAM expression is controlled ontogenetically and environmentally. There is hardly a better example of metabolic flexibility amongst vascular plants than facultative CAM (i.e. the reversible induction of CAM in response to drought stress).

This review discusses the current status of research on phenotypic diversity and plasticity of CO₂ assimilation in plants with CAM photosynthesis. The review contrasts the ontogenetic controls of constitutive CAM and the environmental controls of facultative CAM, and features the wide range of CAM expression relative to C₃ or C₄ photosynthesis within and between species. Frequently recurring questions on CAM biology and CAM terminology are addressed, and topics for future ecophysiological CAM research are identified.

Constitutive CAM

In most species with CAM photosynthesis, CAM expression is constitutive (or obligate); that is, the CAM pathway always manifests itself in mature photosynthetic tissues as part of a developmental routine that typically starts with the C₃ pathway when tissues are young (Box 1). Although generally recognized by CAM researchers and sometimes utilized in comparative physiological and molecular studies (Dever et al., 2015; Hartwell et al., 2016), this ontogenetic C₃–CAM shift has rarely been documented continuously for any species (Winter and Holtum, 2007). Figure 1 shows the ontogenetic progression from C₃ to CAM during the early development of a Kalanchoe pinnata leaf. CO₂ exchange measurements started 12 d after leaf emergence, when the area of the leaf under investigation was <5% of the final area. CO₂ fluxes increased with increased leaf size. Up until day 20, the 24 h CO₂ exchange pattern of the leaf was essentially identical to that of a C₃ plant. Nearly constant rates of CO₂ uptake were displayed in the light and were followed predominantly by constant rates of net CO₂ loss in the dark. On day 21, the first signs of CAM appeared: there was a temporary reduction in diurnal CO₂ uptake paralleled by decreased rates of nocturnal CO₂ loss. At the end of day 23, net nocturnal CO₂ fixation was detected for the first time. From that point onward, net dark CO₂ fixation gradually increased while the temporary depression of CO₂ fixation during the first half of the light period gradually became more pronounced. On day 41, nocturnal carbon gain surpassed diurnal carbon gain. On the last day shown in Fig. 1 (day 46), dark CO₂ fixation contributed 60% to the total daily carbon gain. Nine weeks later, this had increased to 80% of the total daily carbon gain (data not shown), which equates to a further shift from C₃ to CAM. Leaf thickness increased during the course of the experiment, enhancing the storage capacity for nocturnal acid accumulation as leaves matured.

Fig. 1.

Net CO₂ exchange during the early development of a Kalanchoe pinnata leaf. The youngest leaf attached to a 31 cm tall, well-watered plant was enclosed in a GWK-3M gas exchange cuvette (Walz, Effeltrich, Germany) which was connected to a through-flow gas exchange system. The cuvette was supplied with ambient air. The gas exchange cuvette was placed inside a controlled-environment chamber. Twelve hour light periods (650 µmol photons m⁻² s⁻¹; 28 °C) alternated with 12 h dark periods (22 °C). Dark periods are indicated by the gray areas. Measurements began 12 d after leaf emergence. Leaf area on day 12 was 9 cm², and increased to 190 cm² on day 46. Green: CO₂ exchange during light periods. Red: CO₂ exchange during dark periods. Positive values correspond to net CO₂ uptake, and negative values to net CO₂ loss.

Box 1.

Key definitions of CAM expression and CAM species terminology

Constitutive CAM

CAM is always expressed in mature photosynthetic tissues as part of a pre-programmed, irreversible ontogenetic process. Environmental conditions influence rates of dark CO₂ fixation, but particular environmental conditions are not necessary to elicit CAM.

Obligate CAM

The terms obligate and constitutive CAM are often used interchangeably. While constitutive CAM refers to the continual operation of CAM in mature tissues, obligate CAM highlights the need for CAM for growth and reproduction (i.e. survival).

Facultative CAM

Facultative CAM is environmentally triggered, optional CAM. CAM is induced or up-regulated in a reversible manner in response to water-deficit stress in plants that, under well-watered conditions, gain carbon exclusively or predominantly via C₃ photosynthesis, or in some instances via C₄ photosynthesis (e.g. Portulaca). Even though the CAM phenotype is displayed only under conditions of environmental stress, the ability to do so is hardwired (encoded).

CAM cycle

The core CAM metabolic cycle consists of two phases separated in time. (i) At night: glycolytic breakdown of storage carbohydrate to form the CO₂ acceptor phosphoenolpyruvate (PEP), nocturnal assimilation of atmospheric CO₂ via cytosolic PEP carboxylase (PEPC), synthesis of malic acid, and vacuolar storage of malic acid. (ii) During the day: release of malic acid into the cytosol, malate decarboxylation, assimilation of liberated CO₂ via Rubisco into the photosynthetic reduction cycle, and gluconeogenic regeneration of storage carbohydrate from the remaining 3-carbon compound (pyruvate or PEP). Note that the four phases of CAM plant gas exchange of Osmond (1978) cover segments of the 24 h cycle that are not necessarily part of the CAM cycle. Late-afternoon phase-4 fixation of atmospheric CO₂ via Rubisco participates in the CAM cycle only insofar as it contributes to carbon reserves available for PEP production.

CAM plant

A species which, throughout its life, gains most of its carbon by dark CO₂ fixation involving CAM. CAM is typically expressed constitutively. δ¹³C values are less negative than –20‰. Classic examples are most cacti and agaves.

C₃–CAM plant

A species in which CAM is present but in which C₃ photosynthesis contributes to long-term carbon gain more than CAM does. CAM can be expressed constitutively or facultatively.

C₃–CAM intermediate

Similar meaning to C₃–CAM plant. Sometimes the term is also used for plants with facultative CAM. Given the bimodal distribution of δ¹³C values in families with C₃ and CAM species (e.g. Bromeliaceae, Euphorbiaceae) (Winter and Holtum, 2002; Horn et al., 2014; Crayn et al., 2015), C₃–CAM intermediacy with long-term carbon gain derived equally from C₃ photosynthesis and CAM does not seem to be favored ecologically. Possible exceptions from bimodality of δ¹³C values are lineages predominantly composed of facultative CAM species which, during their life cycle, can exhibit the full range of isotopic signatures from C₃-type δ¹³C values to those reflecting pronounced CAM. The net result is intermediate lifetime means of around –20‰. Lifetime means based on monthly averages for Mesembryanthemum crystallinum in Israel and California were –21.1 and –21.6‰, respectively (Winter et al., 1978; Bloom and Troughton, 1979).

Strong CAM

CAM is strongly expressed; that is, the diel CO₂ exchange pattern is characterized by high rates of dark CO₂ fixation in the order of 5 µmol m⁻² s⁻¹, with much higher values having been reported. Strong CAM is typically associated with constitutive CAM.

Full CAM

The diel CO₂ exchange pattern is characterized entirely by net CO₂ uptake in the dark with essentially no net CO₂ uptake in the light. Rates of dark CO₂ fixation may be high (strong CAM) or low. Plants may exhibit full CAM but not strong CAM in response to drought stress when net CO₂ fixation is restricted to the dark but rates of dark CO₂ fixation are low.

Weak CAM

CAM is weakly expressed, either facultatively or constitutively, in plants in which C₃ photosynthesis is typically the principal pathway of carbon acquisition. Dark CO₂ fixation contributes less than ~5% to total carbon gain when CAM is expressed constitutively; when expressed facultatively, dark CO₂ fixation is less than ~5% compared with daytime CO₂ fixation in well-watered plants in the C₃ state.

CAM cycling

An extreme form of weakly expressed CAM in species that exhibit net CO₂ uptake solely in the light. Slightly elevated rates of nocturnal CO₂ fixation result in nocturnal acidification, but nocturnal CO₂ fixation is not sufficient to outweigh nocturnal respiratory CO₂ loss.

CAM-idling

The term CAM-idling was coined for severely droughted CAM plants that maintain an active CAM cycle by utilizing respiratory CO₂ as a carbon source while gas exchange with the atmosphere has largely ceased. CAM-idling is interpreted as a mechanism that extends survival. The ecological significance of CAM idling is unclear.

This ontogenetic C₃ to CAM shift is also commonly observed at the organismal level in young plants of constitutive CAM species, in part because young plants carry mostly immature leaves. The shoot of a young plant of K. pinnata, propagated from a plantlet that originated from the margins of a mature leaf, exclusively showed C₃-type CO₂ fixation in the light at the onset of measurements (Fig. 2). As the plant grew, CAM-type CO₂ fixation developed (Fig. 2, from day 10 onwards) and eventually exceeded C₃ photosynthetic CO₂ fixation in the light (data not shown). Similar responses have been demonstrated for seedlings of stem-succulent cacti (Winter et al., 2011). The initial C₃ photosynthetic phase may be strongly reduced or absent in young cladodes of platyopuntias that develop on mother cladodes, depending upon the extent to which these young cladodes use carbon supplied from the mother cladode for early growth (Winter and Holtum, 2002; Winter et al., 2008).

Fig. 2.

Net CO₂ exchange during the early development of a well-watered Kalanchoe pinnata plant. The shoot of a young plant with recently produced, still immature leaf pairs 2 and 3 was enclosed in a 20×20×15 cm plexiglas cuvette. The cuvette was supplied with ambient air. Roots and the pot were outside the cuvette. Twelve hour light periods (500 µmol photons m⁻² s⁻¹; 28 °C) alternated with 12 h dark periods (22 °C). Dark periods are indicated by the gray areas. Green: CO₂ exchange during light periods. Red: CO₂ exchange during dark periods. Leaf pairs 4, 5, and 6, emerged on days 5, 12, and 21, respectively.

In view of the results shown in Figs 1 and 2, one is automatically reminded of Ernst Haeckel’s 19th century ‘ontogeny recapitulates phylogeny’ concept. Although this concept is now considered an oversimplification and is treated by many authors as biological mythology (Gould, 1977), the ontogenetic C₃–CAM shift is nonetheless a fascinating system for modern evolutionary-developmental-biology (evo-devo) research.

Facultative CAM

In contrast to the ontogenetically controlled irreversible constitutive CAM, facultative CAM is environmentally triggered, typically in response to reduced soil water availability caused by drought or high salinity stress. Facultative CAM was discovered by Winter and von Willert (1972), who demonstrated that the halophytic annual Mesembryanthemum crystallinum (Aizoaceae) exhibited the day–night CO₂ exchange pattern of a C₃ plant when grown under well-watered, non-saline conditions. In contrast, M. crystallinum exhibited the CO₂ exchange pattern of a CAM plant when irrigated with highly saline water containing 400 mM NaCl. The concept of stress-induced facultative CAM was initially met with skepticism as some researchers suggested that the stress treatment merely accelerated a normal irreversible ontogenetic C₃–CAM shift (of the type shown in Figs 1 and 2) (von Willert and Kramer, 1972; Osmond 1978; Adams et al., 1998; Dodd et al., 2002). However, experiments showing that M. crystallinum could complete its life cycle and produce seeds by solely operating in the C₃ mode under non-stress conditions disproved the notion that accelerated aging was the principal driver of CAM in M. crystallinum (Winter and Holtum, 2007). In order to ensure that CAM is truly facultative in a given species (i.e. optional and environmentally triggered), it has become standard procedure not only to demonstrate the induction or up-regulation of CAM in response to stress, but also to demonstrate the reversion to C₃ upon removal of stress. Such information is now available for a wide range of species (Table 1). Long-lived leaves of Clusia ssp. can exhibit multiple C₃–CAM–C₃ swings during repeated wet–dry–wet cycles (K. Winter, unpublished data), emphasizing the crucial role of environmental rather than ontogenetic factors in promoting nocturnal CO₂ fixation in facultative CAM plants.

Table 1.

Species with facultative CAM

Family	Species	Reversibility	Reference
Aizoaceae	Carpobrotus edulis (L.) N.E.Br.	Yes	Winter (1973); K. Winter (unpublished); Treichel and Bauer (1974)
	Delosperma tradescantioides (P.J.Bergius) L.Bolus	Yes	Herppich et al. (1996)
	Disphyma clavellatum (Haw.) Chinnock	ND	Winter et al. (1981)
	Mesembryanthemum crystallinum L.	Yes	Winter and von Willert (1972); Winter and Holtum (2014)
	Mesembryanthemum cordifolium L.f.	ND	Treichel (1975)
	Mesembryanthemum nodiflorum L.	ND	Treichel and Bauer (1974)
Anacampserotaceae	Anacampseros australiana J.M.Black	Yes	Winter and Holtum (2017)
	Anacampseros coahuilensis (S.Watson) Eggli & Nyffeler	ND	Guralnick et al. (2008)
	Grahamia bracteata Gillies ex Hook. & Arn.	ND	Guralnick et al. (2008)
Araceae	Zamioculcas zamiifolia (Lodd.) Engl.	Yes	Holtum et al. (2007)
Basellaceae	Anredera baselloides (Kunth) Baill.	Yes	Holtum et al. (2018)
	Basella alba L.	Yes	K. Winter et al. (unpublished)
Bromeliaceae	Guzmania monostachia (L.) Rusby ex Mez	Yes	Medina et al. (1977)
	Vrisea sanguinolenta Cogn. & Marchal.	ND	Beltrán et al. (2013)
Cactaceae	Pereskia guamacho F.A.C.Weber	ND	Diaz and Medina (1984)
Clusiaceae	Clusia aripoensis Britton	ND	Borland et al. (1998)
	Clusia cylindrica Hammel	Yes	Winter et al. (2009)
	Clusia minor L.	Yes	Borland et al. (1998); Winter et al. (2008)
	Clusia pratensis Seem.	Yes	Winter and Holtum (2014)
	Clusia uvitana Pittier	Yes	Winter et al. (1992)
Commelinaceae	Callisia fragrans (Lindl.) Woodson	ND	Martin et al. (1994)
	Tradescantia brevifolia (Torr.) Rose	ND	Martin et al. (1994)
	Tripogandra multiflora (Sw.) Raf.	ND	Martin et al. (1994)
Crassulaceae	Crassula sieberiana (Schult. & Schult.f.) Druce	Yes	Winter and Holtum (2017)
	Kalanchoe gracilipes (Baker) Baill.	Yes	J. Hartwell et al. (unpublished)
	Kalanchoe miniata Hilsenb. & Bojer ex Tul.	Yes	Brulfert et al. (1996)
	Kalanchoe porphyrocalyx (Baker) Baill.	Yes	Brulfert et al. (1996)
	Sedum acre L.	Yes	Kluge (1977)
	Sedum album L.	Yes	Castillo (1996)
	Sedum pulchellum Michx.	ND	Smith and Eickmeier (1983)
	Sedum sexangulare L.	?	Schuber and Kluge (1981)
	Sedum telephium L.	Yes	Lee and Griffiths (1987)
Didiereaceae	Ceraria fruticulosa H.Pearson & Stephens	ND	Veste et al. (2001)
	Portulacaria afra Jacq.	Yes	Ting and Hanscom (1977)
Lamiaceae	Plectranthus amboinicus (Lour.) Spreng.	Yes	K. Winter (unpublished)
Montiaceae	Calandrinia creethae Tratman ex Morrison	Yes	Holtum et al. (2017a)
	Calandrinia holtumii Obbens & L.P.Hancock	Yes	K. Winter (unpublished)
	Calandrinia pentavalvis Obbens	Yes	Holtum et al. (2017a)
	Calandrinia polyandra Benth.	Yes	Winter and Holtum (2011)
	Calandrinia quadrivalvis F.Muell.	Yes	Holtum et al. (2017a)
	Calandrinia reticulata Syeda	Yes	Holtum et al. (2017a)
	Cistanthe cf. grandiflora (Lindl.) Schltdl.	Yes	K. Winter et al. (unpublished)
Portulacaceae	Portulaca australis Endl.	Yes	Winter and Holtum (2017)
	Portulaca cyclophylla F.Muell.	Yes	Holtum et al. (2017b)
	Portulaca cryptopetala Speg.	Yes	K. Winter et al. (unpublished)
	Portulaca digyna F.Muell.	Yes	Holtum et al. (2017b)
	Portulaca grandiflora Hook.	ND	Guralnick et al. (2002)
	Portulaca molokiniensis Hobdy	Yes	K. Winter et al. (unpublished)
	Portulaca oleracea L.	Yes	Koch and Kennedy (1980); Winter and Holtum (2014)
	Portulaca pilosa L.	Yes	Winter and Holtum (2017)
	Portulaca umbraticola Kunth	Yes	This study
Talinaceae	Talinum fruticosum (L.) Juss.	Yes	Winter and Holtum (2014)
	Talinum paniculatum (Jacq.) Gaertn.	?	Güerere et al. (1996)
Vitaceae	Cissus trifoliata (L.) L.	Yes	Olivares et al. (1984)

Currently accepted species names are used. Some species names employed in the original publications such as Aptenia cordifolia (L.f.) Schwantes, Werauhia sanguinolenta (Cogn. & Marchal) J.R.Grant, Sedum mite Gilib., and Talinum triangulare (Jacq.) Willd. are now considered synonyms of Mesembryanthemum cordifolium L.f., Vriesea sanguinolenta Cogn. & Marchal., Sedum sexangulare L., and Talinum fruticosum (L.) Juss., respectively. ND, not determined. ?, not clear whether or not reversibility was examined.

Figure 3 shows the drought-induced, reversible shift to CAM in a C₄ species, Portulaca umbraticola, for which facultative CAM has not been demonstrated previously. CO₂ exchange of an entire shoot (leaves and stems) was continuously monitored for 20 d. As in the experiment shown in Fig. 2, overall CO₂ fluxes increased as the plant inside the gas exchange system increased in size. In the well-watered state, CAM was absent: net CO₂ uptake occurred during the light period only, and during the dark period CO₂ was released at a relatively constant rate. As drought stress developed, CO₂ gain in the light declined over the course of several days, and nocturnal CO₂ exchange gradually changed from net CO₂ loss to net CO₂ gain. Early in the transition to CAM, nocturnal gas exchange passed through a stage typical of CAM cycling (Fig. 3, day 12 versus day 11; see also Fig. 1, day 22 versus day 18; Box 1). On days 17 and 18, CO₂ exchange was close to zero for most of the light period, and net dark CO₂ fixation had plateaued. Upon re-watering, the plant rapidly returned to solely net CO₂ fixation in the light. Based on the CO₂ exchange responses in Fig. 3, P. umbraticula can be considered a species with facultative, weakly expressed CAM, because nocturnal CO₂ fixation in the stressed state is very small compared with diurnal CO₂ fixation in the non-stressed state (Box 1).

Fig. 3.

Net CO₂ exchange of Portulaca umbraticola during a wet–dry–wet cycle. Measurements were conducted on a shoot that grew inside a 11×11×10 cm plexiglas cuvette. Roots and the pot were outside the cuvette. [CO₂] of the air entering the cuvette was ~400 µl l⁻¹ provided by a Walz CO₂/CO₂-free air mixing system. Twelve hour light periods (600 µmol photons m⁻² s⁻¹; 28 °C) alternated with 12 h dark periods (22 °C). Green: CO₂ exchange during light periods. Red: CO₂ exchange during dark periods. Irrigation stopped on day 9 (open arrow) and resumed on day 19 (filled arrow). At the onset of the experiment, the plant had cotyledons and the leaf length of the first leaf pair was 7 mm. At the conclusion of the experiment (day 23, not shown), total leaf area was 132 cm² and shoot dry mass was 0.602 g.

Thus far, facultative CAM has been reported for at least 54 species from 27 genera and 15 families, many in the order Caryophyllales (Table 1). The number of species will increase significantly when previously unpublished results from a CAM survey of 50 species of the large family Aizoaceae (2271 species in total; www.theplantlist.org) are taken into consideration (Table 2). This survey was a direct follow-up to the first demonstration of high-salinity-induced CAM in M. crystallinum. Most species in Table 2 belong to the large subfamily Ruschioideae (1939 species in total), with the exception of two species in the subfamily Mesembryanthemoideae [Mesembryanthemum lancifolium and Phyllobolus prasinus (formerly Aridaria prasina)]. CAM-type nocturnal malate accumulation was observed in 45 of the 50 species studied. Consistent with facultative CAM, 28 species showed nocturnal malate accumulation under conditions of drought and/or drought plus salinity stress, but not when well watered. In the 17 species in which significant nocturnal malate accumulation was already present in well-watered plants, the dawn:dusk malate level ratio (i.e. the fold nocturnal malate increase) was enhanced upon drought and salinity stress, consistent with a facultative CAM component in addition to constitutive CAM. It seems that CAM, especially facultative CAM, is a common feature amongst the Aizoaceae, and that in the Aizoaceae alone, facultative CAM could be present in well over 1000 species.

Table 2.

Leaf malate content end of day (L) and end of night (D) of 50 species of Aizoaceae cultivated under well-watered conditions, or exposed to drought stress, or drought plus salinity stress

Species	Malate (µmol g-1 DM)
	Control (well-watered)		Drought stress		Drought + salt stress
	L	D	L	D	L	D
Bergeranthus vespertinus ({Berger) Schwantes	123 ± 37	280 ± 98	68 ± 29	257 ± 43	36 ± 25	191 ± 22
Calamophyllum cylindricum (Haw.) Schwantes	52 ± 21	67 ± 18	11 ± 1	44 ± 12	0 ± 0	15 ± 14
Carpobrotus acinaciformis (L.) L.Bolus	221 ± 67	187 ± 59	46 ± 25	89 ± 25	25 ± 4	83 ± 8
Carpobrotus aequilaterus (Haw.) N.E.Br.	183 ± 71	194 ± 62	26 ± 11	73 ± 7	ND	ND
Carpobrotus quadrifidus L.Bolus	99 ± 2	104 ± 13	148 ± 42	249 ± 27	21 ± 2	104 ± 18
Carruanthus ringens (L.) Boom	205 ± 93	236 ± 87	ND	ND	54 ± 29	358 ± 27
Cephalophyllum cupreum L.Bolus	248 ± 167	299 ± 85	106 ± 53	338 ± 93	49 ± 14	217 ± 58
Cephalophyllum purpureoalbum (Haw.) Schwantes	82 ± 24	73 ± 32	29 ± 6	52 ± 5	18 ± 2	36 ± 10
Cephalophyllum subulatoides (Haw.) N.E.Br.	92 ± 49	74 ± 24	29 ± 8	80 ± 21	13 ± 4	61 ± 6
Cephalophyllum tricolorum (Haw.) N.E.Br.	77 ± 18	104 ± 24	25 ± 13	119 ± 11	12 ± 1	67 ± 10
Chasmatophyllum masculinum (Haw.) Dinter & Schwantes	13 ± 7	32 ± 9	7 ± 5	37 ± 2	5 ± 3	23 ± 1
Conicosia pugioniformis (L.) N.E.Br.	4 ± 1	5 ± 5	ND	ND	ND	ND
Delosperma brunnthaleri (A.Berger) Schwantes ex H.Jacobsen	36 ± 18	51 ± 21	122 ± 15	205 ± 39	9 ± 3	35 ± 13
Delosperma hirtum (N.E.Br.) Schwantes	97 ± 3	219 ± 47	60 ± 25	220 ± 22	51 ± 10	239 ± 13
Delosperma luteum L.Bolus	193 ± 59	325 ± 67	24 ± 6	178 ± 26	25 ± 22	149 ± 80
Delosperma macellum (N.E.Br.) N.E.Br.	120 ± 17	143 ± 39	189 ± 18	295 ± 19	61 ± 18	194 ± 31
Delosperma mahonii (N.E.Br.) N.E.Br.	114 ± 20	296 ± 46	46 ± 16	184 ± 51	37 ± 8	202 ± 19
Delosperma steytlerae L.Bolus	126 ± 17	196 ± 47	147 ± 45	156 ± 32	36 ±18	60 ± 22
Delosperma wethamae L.Bolus	4 ± 4	3 ± 5	80 ± 30	152 ± 47	18 ± 20	50 ± 6
Disphyma australe (Sol. ex Aiton) J.M.Black	96 ± 36	93 ± 35	76 ± 24	134 ± 56	102 ± 10	198 ± 11
Disphyma crassifolium (L.) L.Bolus	63 ± 18	100 ± 10	73 ± 35	177 ± 81	23 ± 9	59 ± 9
Erepsia heteropetala (Haw.) Schwantes	48 ± 9	61 ± 10	3 ± 4	30 ± 4	ND	ND
Faucaria felina (L.) Schwantes	129 ± 12	363 ± 74	91 ± 10	216 ± 33	16 ± 14	84 ± 21
Faucaria subintegra L.Bolus	337 ± 166	376 ± 153	55 ± 35	95 ± 33	42 ± 16	92 ± 13
Faucaria tigrina (Haw.) Schwantes	85 ± 2	257 ± 2	112 ± 40	265 ± 33	27 ± 9	151 ± 35
Glottiphyllum depressum (Haw.) N.E.Br.	186 ± 104	261 ± 95	89 ± 14	161 ± 23	36 ± 4	84 ± 7
Glottiphyllum difforme (L.) N.E.Br.	228 ± 108	265 ± 65	57 ± 17	117 ± 17	45 ± 11	99 ± 10
Glottiphyllum longum (Haw.) N.E.Br.	318 ± 134	360 ± 111	156 ± 23	170 ± 81	49 ± 3	84 ± 3
Hereroa calycina L. Bolus	105 ± 25	105 ± 9	44 ± 11	98 ± 13	9 ± 10	64 ± 22
Hereroa gracilis L.Bolus	84 ± 8	212 ± 36	40 ± 4	271 ± 15	32 ± 8	198 ± 31
Hereroa granulata Dinter & Schwantes	130 ± 78	223 ± 82	34 ± 4	198 ± 18	0 ± 0	58 ± 3
Hereroa stanleyi L. Bolus	63 ± 22	267 ± 57	47 ± 21	183 ± 75	16 ± 6	155 ± 29
Lampranthus curviflorus N.E. Br.	19 ± 6	35 ± 9	ND	ND	3 ± 5	20 ± 8
Lampranthus deltoides (L.) Glen ex Wijnands	117 ± 68	166 ± 69	26 ± 7	97 ± 12	19 ± 7	62 ± 8
Lampranthus falciformis (Haw.) N.E.Br.	20 ± 3	38 ± 2	6 ± 4	30 ± 9	7 ± 3	31 ± 9
Lampranthus lunatus N.E. Br.	38 ± 8	63 ± 26	18 ± 6	71 ± 4	27 ± 21	62 ± 35
Lampranthus multiseriatus N.E. Br.	14 ± 4	19 ± 2	55 ± 38	38 ± 17	7 ± 7	34 ± 12
Lampranthus variabilis N.E. Br.	0 ± 0	12 ± 11	4 ± 3	32 ± 6	4 ± 7	34 ± 16
Mesembryanthemum lancifolium (L. Bolus) Klak	72 ± 52	329 ± 111	145 ± 29	945 ± 66	103 ± 19	848 ± 95
Nananthus orpenii (N.E.Br.) L.Bolus	196 ± 36	477 ± 103	ND	ND	ND	ND
Phyllobolus prasinus (L.Bolus) Gerbaulet	144 ± 28	861 ± 83	98 ± 47	782 ± 50	109 ± 6	862 ± 34
Pleiospilos compactus Schwantes	898 ± 431	1430 ±263	487 ± 71	1026 ± 37	227 ± 52	636 ± 40
Pleiospilos magnipunctatus Schwantes	646 ± 238	913 ± 312	472 ± 117	785 ± 8	176 ± 0	439 ± 73
Rabiea cibdela (N.E.Br.) N.E.Br.	77 ± 23	35 ± 13	ND	ND	ND	ND
Rhinephyllum broomii L. Bolus	17 ± 2	41 ± 33	50 ± 7	111 ± 18	ND	ND
Ruschia hexamera L.Bolus	48 ± 36	56 ± 7	28 ± 6	84 ± 2	0 ± 0	53 ± 20
Ruschia vaginata Schwantes	42 ± 4	62 ± 6	15 ± 11	57 ± 1	8 ± 2	49 ± 0
Titanopsis calcarea (Marloth) Schwantes	219 ± 82	405 ± 56	84 ± 23	106 ± 31	31 ± 5	77 ± 26
Trichodiadema barbatum Schwantes	766 ± 78	834 ± 86	151 ± 62	356 ± 24	ND	ND
Trichodiadema stelligerum Schwantes	633 ± 131	728 ± 36	116 ± 40	238 ± 62	153 ± 14	302 ± 29

Plants were grown in terra cotta pots in cold frames at the Botanic Garden of the Technical University of Darmstadt during late spring and summer of 1976. Plants were irrigated daily (well-watered controls), weekly (drought stress), or weekly plus 200 mM NaCl each 4 weeks (drought + salt stress). Values are means ± SD (n=3 samples from different plants, or, in few cases, 2 samples). End-of-night values in bold are significantly higher than end of day values (one-tailed t-test, P <0.05). ND, not determined (K. Winter, previously unpublished results).

Do C₃–CAM shifts always reflect facultative CAM?

The answer is no. What distinguishes facultative CAM from other types of C₃–CAM shifts is the fact that upon stress, the magnitude of nocturnal CO₂ uptake increases. Changes in the relative proportions of diurnal to nocturnal carbon gain in favor of nocturnal CO₂ uptake, but without enhanced nocturnal CO₂ assimilation, do not represent facultative CAM sensu stricto. Figure 4 depicts the response of a mature leaf of K. pinnata to water-deficit stress. The leaf exhibits significant diurnal C₃ photosynthetic CO₂ uptake when well hydrated, in addition to nocturnal CO₂ fixation. Upon the imposition of instantaneous drought stress (petiole cut in the experiment of Fig. 4), night-time CO₂ fixation remains, whereas daytime CO₂ fixation stops almost immediately. In the example of Fig. 4, the shift from a C₃–CAM pattern to an exclusively CAM pattern is driven by a more rapid decline in C₃ photosynthetic CO₂ uptake in the light than in CO₂ uptake in the dark. In this case, the criteria for facultative CAM are not fulfilled.

Fig. 4.

Effect of abruptly induced water-deficit stress on the net CO₂ exchange of a fully expanded Kalanchoe pinnata leaf. The leaf was enclosed in a GWK-3M gas exchange cuvette (Walz, Effeltrich). The cuvette was supplied with ambient air. Twelve hour light periods (650 µmol photons m⁻² s⁻¹; 28 °C) alternated with 12 h dark periods (22 °C). Dark periods are indicated by the gray areas. Green: CO₂ exchange during light periods. Red: CO₂ exchange during dark periods. From day 1 to day 3, the leaf was attached via its petiole to the plant outside the cuvette. At the beginning of day 4, 15 min into the light period, the petiole was cut (blue arrow). Leaf lamina dry mass at the conclusion of the experiment was 1.53 g and the area was 120 cm².

Sealing entire succulent plants or parts of them into gas exchange cuvettes can be challenging. Therefore, CO₂ measurements are sometimes performed on detached CAM organs, assuming that their gas exchange remains largely unchanged, at least for a day or so upon separation from the mother plant, given the high water-holding capacity of succulent tissues (Boxall et al., 2017). Figure 4 demonstrates that this assumption certainly does not apply to leaves of K. pinnata.

During long-term drought stress, net dark CO₂ fixation eventually ceases, but nocturnal acidification may continue at a low magnitude using respiratory CO₂ as substrate for CAM. The recycling of respiratory CO₂ in the absence of gas exchange with the atmosphere has been called CAM-idling and it has been suggested that CAM-idling is an important survival mechanism (Hanscom and Ting, 1978). Supporting evidence for the latter is scarce. Drought experiments were conducted with a range of species where gas exchange was monitored until the exact point in time when net CO₂ uptake in the dark no longer occurred. In the example of Fig. 5, which features an intact attached branch of the epiphytic cactus Hatiora salicornioides, nocturnal net CO₂ uptake stopped at exactly 47 d after watering was withheld. Indeed, a small H⁺ increase from 0.4±0.6 µmol g⁻¹ to 7.3±1.4 µmol g⁻¹ fresh mass was observed during the course of the following night, consistent with CO₂ recycling via CAM. In other species, such as the CAM bromeliad Tillandsia flexuosa, a highly drought-resistant epiphytic air plant, nocturnal acid accumulation was no longer detectable once nocturnal net CO₂ uptake had ceased (K. Winter, unpublished data). Notably, dehydrated CAM tissues do not operate as completely closed, airtight systems, and tissues may rapidly shift to continuous diel net CO₂ loss once net CO₂ dark fixation no longer takes place. Furthermore, non-uniform dehydration is a major complication of long-term drought experiments with CAM plants, and partial death of photosynthetic organs occurs in order to assist the remaining tissues in staying active longer. Survival through partial death (Evenari et al., 1971) may well be of greater adaptive significance to drought stress than CAM-idling per se.

Fig. 5.

Effect of prolonged drought stress on the net CO₂ exchange of an intact, attached branch of the epiphytic cactus Hatiora salicornioides. The branch was enclosed in a GWK-3M gas exchange cuvette (Walz, Effeltrich). The rest of the plant was outside the cuvette. The cuvette was supplied with ambient air. Twelve hour light periods (650 µmol photons m⁻² s⁻¹; 28 °C) alternated with 12 h dark periods (22 °C). Dark periods are indicated by the gray areas. Green: CO₂ exchange during light periods. Red: CO₂ exchange during dark periods. Irrigation was stopped at the onset of the experiment and never resumed.

Can facultative and constitutive CAM co-occur?

The answer is yes. Under well-watered conditions prior to CAM induction, not all species with facultative CAM exhibit pure C₃ physiology with zero nocturnal acidification. Even if they do, full reversibility to C₃ does not always occur. In many species, facultative CAM is observed as stress-induced up-regulation of CAM, superimposed on a pre-existing background of weakly expressed constitutive CAM. Upon the removal of stress, plants either fully or partly return to the pre-stress CAM level. Even plants with pronounced constitutive CAM have been shown to exhibit small transient increases in nocturnal CO₂ uptake in response to water-deficit stress (Winter et al., 2008, 2014). In all these cases, the facultative component of CAM refers to the stress-induced reversible increase of nocturnal CO₂ uptake and nocturnal acidification, irrespective of the magnitude of a possible constitutive CAM background.

Is facultative CAM a transitional state between C₃ and full CAM?

The adaptive significance of facultative CAM in annuals such as M. crystallinum or Calandrinia ssp. appears obvious (Winter et al., 1978; Winter and Ziegler, 1992; Herrera, 2009; Winter and Holtum, 2014). C₃ photosynthesis promotes initial rapid vegetative growth when it rains, while water-use-efficient CAM prolongs the life cycle and aids reproduction during the subsequent drought. This is photosynthetic plasticity par excellence and seems like a perfect strategy for these annuals in their particular habitats. It is hard to imagine that these plants are on their way to becoming perennials with full CAM. Nonetheless, such reasoning does not exclude the possibility that extant species showing pronounced constitutive CAM derived from ancestors with facultative CAM. Recent research demonstrated facultative CAM in leaves and stems of the annual C₃–C₄ intermediate Portulaca cryptopetala (K. Winter et al., unpublished data). Similarly, combinations of weakly expressed facultative and constitutive CAM were detected in leaves and stems of the tropical vine Basella alba (K. Winter et al., unpublished data) and the pan-tropical coastal Sesuvium portulacastrum (Winter et al., 2019). Life forms with these attributes could give rise to long-lived stem succulents with full CAM through the thickening of stems and their green cortex, accompanied by the shift of stem photosynthesis from facultative to strong constitutive CAM, in addition to the permanent loss of leaves. In the Cactaceae, leafy forms with C₃ photosynthesis or facultative CAM are ancestral to non-leafy stem succulents with full CAM (Diaz and Medina, 1984; Edwards and Diaz, 2006). In the genus Euphorbia (Euphorbiaceae) containing leafy and cactiform growth forms, CAM has evolved multiple times in a complex phylogenetic pattern (Horn et al., 2012, 2014). Weakly expressed CAM occurs in Euphorbia milii (Herrera, 2013), but facultative CAM has not yet been demonstrated conclusively in this species-rich genus (Mies et al., 1996; Hastilestari et al., 2013). Within the large genus Kalanchoe (Crassulaceae), species in the basal Kitchingia group are capable of facultative CAM, whereas the most derived taxa exhibit strong constitutive CAM (Hartwell et al., 2016).

Should all plants with CAM be considered CAM plants?

The answer to this recurring question is no. The CAM cycle is believed to be present in well over 5% of vascular plant species (Winter and Smith, 1996). As noted above, in most cases CAM co-occurs with C₃ photosynthesis. Depending on the species, developmental stage, and environmental conditions, the contribution of CAM to daily carbon gain may range from <1% to 100%. The term CAM plant should be reserved for species which, throughout their lives, obtain the majority of their carbon through the CAM pathway (Box 1; Winter et al., 2015). Many cacti and agaves seem to operate at or close to the full-CAM end of the phenotypic C₃–CAM spectrum (Nobel, 1988) and are traditionally and rightly considered CAM plants. Kalanchoe pinnata would also qualify for the CAM plant category: despite significant C₃-type CO₂ exchange during early development, CAM eventually becomes the major contributor to leaf life cycle carbon gain (Fig. 1). On the other hand, the term CAM plant would be inappropriate for species with weakly expressed CAM in which C₃ rather than CAM is the principal mode of carbon assimilation. This is particularly true for species such as Welwitschia mirabilis with very minor roles of CAM relative to C₃. In such circumstances, the term C₃–CAM plant is an option (Winter et al., 2015).

In general, it is much easier to define distinct functional properties (e.g. facultative CAM) of a plant than to categorize organisms on the basis of a specific trait (e.g. facultative CAM plant), because this trait may co-occur in multiple combinations with other traits during different phases of the life cycle. Therefore, when characterizing a species in the context of CAM, detailed case-by-case descriptions of CAM expression relative to C₃ (or C₄) may be more informative than trying to tally a species to one of a multitude of categories created to accommodate all possible C₃/C₄–CAM trait combinations. Even categories that researchers have agreed upon such as constitutive (or obligate) CAM versus facultative CAM are still compromises. For example, besides indicating that a process is always active, the term constitutive also has the connotation of something that is hardwired. Thus, strictly speaking, the ability to engage in facultative CAM is also constitutive. Furthermore, the term obligate, which has the meaning of being biologically essential for survival, also applies to facultative CAM annuals, where the dry season switch to CAM aids reproduction.

All these terminological complications can be largely disregarded in phylogenetic studies on CAM evolution, where the presence of CAM (irrespective of the degree of CAM usage relative to C₃ usage) and the absence of CAM are the key binary traits of the most basic analysis.

How to identify weakly expressed CAM

In recent years, there has been growing awareness of the significant number of species in which the CAM cycle (although present) contributes little to overall CO₂ assimilation when compared with C₃ or C₄ photosynthesis. Carbon isotopic signatures (δ¹³C values) of dried plant material are not suitable indicators of weakly expressed CAM because the amount of CAM-derived carbon is usually too small to be detected against the large background of C₃ carbon. Similarly, isotopic measurements do not reveal weakly expressed CAM in C₄ plants such as Portulaca spp., since the C₄ and CAM isotopic signals are largely identical.

Weakly expressed CAM can only be diagnosed in living material. Highly replicated measurements of titratable acidity of samples from mature tissues, when repeatedly collected at dusk and dawn over the course of several days, can conclusively demonstrate low levels of nocturnal acidification. Acid titrations are highly sensitive and can resolve day–night differences of as low as 1–2 µmol H⁺ g⁻¹ fresh mass. It is advisable to demonstrate nocturnal acidification not only on a fresh mass, but also on a dry mass and area basis. In some species with highly elastic fleshy leaves, leaf fresh mass decreases in the course of the day even in the absence of soil-water deficit stress; this leads to an overestimation of acidity levels at dusk and thus an underestimation of nocturnal acidification. In extreme cases, decreases in fresh mass during the day can completely mask nocturnal increases in H⁺.

Measurements of malate levels at dawn and dusk (e.g. enzymatically or through HPLC) can also be used to detect low-level CAM. However, malate assays do not distinguish between malic acid involved in CAM and malate anions that are electrochemically balanced by cations such as K⁺ and that do not participate in CAM. It is noteworthy that the leaves of some non-CAM species in the Brassicaceae accumulate substantial amounts of malate in the course of the day (i.e. not at night!) (Winter et al., 1976, 1982), when NO₃⁻ taken up as KNO₃ is metabolized in the light and malate²⁻ is synthesized to balance the positive charges of the remaining K⁺. Such diurnal increases in malate are completely unrelated to CAM. Potassium malate, unlike malic acid, is not detectable by titration. Suggestions that nocturnal net acid accumulation is a feature of C₃ species (fig. 2A of Bräutigam et al., 2017) are unfounded.

In order to confirm the presence of weakly expressed CAM in a species, acidity measurements should be accompanied by continuous CO₂ gas exchange measurements during 12 h light/12 h dark cycles. In the past, such measurements generally required elaborate, laboratory-installed custom-built gas exchange systems. Nowadays, miniaturized modern portable photosynthesis systems, when programmed for automatic long-term CO₂ logging, are suitable tools for CAM studies as well. Weakly expressed CAM may or may not be associated with net CO₂ dark fixation, and in many cases the nocturnal carbon balance remains negative. If nocturnal net CO₂ fixation does occur, it is often restricted to a brief phase of the dark period. In the case when dark CO₂ fixation capacity—although elevated when compared with regular C₃ plants—is not sufficient to support net CO₂ fixation, weak CAM nonetheless results in characteristic, curved CO₂ exchange patterns of net CO₂ loss during the course of the night, with lowest rates of net CO₂ loss typically in the middle of the night. Weak, presumably constitutive, CAM has been reported for a large number of species in the Orchidaceae (Silvera et al., 2005). Furthermore, in many species with facultative CAM, the degree of drought-induced CAM is very low compared with C₃ photosynthetic CO₂ uptake in the light in unstressed plants (Box 1).

Demonstrating the absence of the CAM cycle in a given species, especially when the species under investigation belongs to a lineage that contains confirmed CAM species, can be as challenging as demonstrating the presence of weakly expressed CAM in a species. Excluding CAM is of major concern in phylogenetic studies on CAM evolution that may require 100% pure C₃ ancestral material for comparative purposes. It may also be useful to keep in mind the dictum that ‘absence of evidence is not evidence of absence’ when exploring CAM to C₃ reversals in lineages.

Is there a C₃–CAM continuum?

The answer is: phenotypically yes, genotypically no. The entire range of diel CO₂ exchange patterns is possible from 0% CAM (i.e. 100% C₃ or C₄) to 100% CAM, as demonstrated by comparative gas exchange studies of C₃-, C₄-, and CAM-exhibiting species, particularly of CAM species transitioning from C₃ to CAM ontogenetically or facultatively (Figs 1–5). Hence, phenotypically, there is undoubtedly a C₃–CAM continuum. Genotypically, this is not the case as the evolutionary transitioning to CAM is based on discrete changes in the genetic makeup of species resulting in changes in protein sequence and/or gene expression (Yang et al., 2016, 2017). Enhanced succulence in some lineages may have a potentiating effect, increasing the likelihood for CAM to evolve (Heyduk et al., 2016). Even if the development of CAM were to be entirely based on the up-regulation of pre-existing C₃ genes in some species, the ability to do so is a heritable trait that would be encoded in the genome. Unraveling the molecular basis of CAM is one of the most active areas of current CAM research (e.g. Goolsby et al., 2018; Heyduk et al., 2018), although emphasis is largely on iconic CAM species with pronounced CAM such as Agave ssp., Opuntia ssp., and Ananas comosus (Ming et al., 2015; Abraham et al., 2016). Genome and transcriptome studies of plants with weakly expressed constitutive and facultative CAM are currently less in vogue (Brilhaus et al., 2016), but hold particular promise for capturing the early steps of CAM evolution.

Underscoring the distinctness of CAM is the fact that CAM (including weakly expressed CAM) is currently noted in 35 families of angiosperms, plus five families of lycopods, ferns, and gymnosperms (Smith and Winter, 1996; J.A.C. Smith et al., unpublished data), whereas CAM is not known to occur in over 370 angiosperm families. Unquestionably, CAM will be discovered in additional families in the future, but this is unlikely to alter the predominance of non-CAM taxa. One possible candidate for a ‘new’ CAM (or C₄) family is the Capparaceae. For one of its members, Cadaba aphylla, a leafless, succulent-branched shrub from tropical Africa to South Africa and Namibia, Schulze et al. (1976) reported a δ¹³C value of –16.9‰. Further studies into the photosynthetic pathway operating in this species would be rewarding.

Bioengineering CAM into C₃

Introducing CAM into C₃ crops may make them more resilient to hotter and drier conditions in the face of concurrent man-made climate change (Yang et al., 2015). However, even if CAM-into-C₃ engineering would become technically feasible, there is more to the CAM ‘syndrome’ than merely its CAM cycle biochemistry. Full, strongly expressed CAM is typically associated with highly succulent leaves or photosynthetic stems, and hence relatively low surface to dry mass ratios of individual photosynthetic organs. These morphological features tend to lower the ratio of total photosynthetic surface area to total dry mass of plants, thereby lowering rates of growth, even though CAM plants are known for maximizing available surface areas for CO₂ uptake. Considerations of this kind are useful when planning to bioengineer CAM into C₃ plants, especially C₃ trees (Borland et al., 2015), which in terms of potential growth rates are already disadvantaged due to substantial dry mass allocation to non-photosynthetic stem tissue. Leaves with facultative CAM are likely to have lower construction costs in comparison with leaves with full, strongly expressed CAM. Thus, facultative CAM may be the preferred form of CAM to be introduced into C₃ trees, after having evaluated the benefits of improved water use efficiency versus potential constraints on rates of biomass accumulation.

Topics for future ecophysiological CAM research

A range of topics and questions appear to be of particular interest for future ecophysiological CAM research. These include (i) whole-plant in situ CO₂ gas exchange measurements of archetypal constitutive CAM species, such as columnar cacti and platyopuntias, combined with measurements of plant productivity. Does CO₂ fixation during the early morning contribute to growth? Is there substantial net CO₂ loss during hot daytime hours? How ‘airtight’ are these plants under severe drought stress? (ii) How do the net assimilation rate and leaf or stem area ratio determine relative growth rates of leaf and stem succulent CAM plants? Thus far, CAM-exhibiting species have been largely omitted from global analyses of plant functional traits that describe the performance of species along the fast–slow lifestyle continuum (Osnas et al., 2013; Huber et al., 2018; Males and Griffiths, 2018). (iii) How does eddy-flux-based CO₂ and water vapor exchange of CAM-dominated vegetation and of cultivated CAM crops respond to seasonal change (see, for example, Owen et al., 2016)? (iv) How do daily patterns of C₃ and CAM photosynthesis vary during the life cycle of annual and perennial facultative CAM species in situ? This requires the monitoring of C₃ and CAM activities at close (e.g. weekly) intervals, combined with measurements of plant phenology, microclimate, and edaphic conditions. (v) What is the adaptive significance of weakly expressed CAM? (vi) How abundant is CAM in orchids—one of the two largest families of vascular plants—especially in the relatively understudied Asian species? (vi) How many species of the large family Aizoaceae engage in facultative or constitutive CAM? CAM may have been a key innovation that facilitated the rapid, recent diversification of this clade (see, for example, Klak et al., 2004). (vii) How do CAM species respond to atmospheric and climate change? Is CAM down-regulated in favor of C₃ photosynthesis as atmospheric CO₂ concentrations rise? (viii) Eco-transcriptomics: how does the expression of CAM genes and CAM-related genes change during natural day–night cycles with varying diurnal photon flux density (PFD), especially in genes postulated to be under circadian control?

Acknowledgements

This work was supported by the Smithsonian Tropical Research Institute. Aurelio Virgo spotted Portulaca umbraticola in a local supermarket in Panama and prepared the illustrations.