The CAM photosynthetic pathway

The CAM pathway was discovered in the Crassulacean plant Kalanchoe pinnata (now known as Bryophyllum calycinum) and, in consequence, it was named by this species’ family (^{Kluge & Ting 1978}). However, CAM also occurs commonly in over 400 genera in 38 plant families comprising ~6 % of flowering plant species (^{Niechayev et al. 2019}), and about 10,000 species (^{Lambers & Oliveira 2019}). CAM photosynthesis has arisen independently at least 35 times in flowering plants (^{Silvera et al. 2010}) and there is great plasticity in the expression of this metabolism (^{Borland et al. 2011}).

Where CAM plants occur. In families such as the Cactaceae, CAM is most likely existing in all genera with exception of the leaf-bearing cacti of the genus Pereskia (^{Nobel 2003}), although one species from Venezuela, Pereskia guamacho, exhibited C₃-like patterns of stomatal conductance in the rainy season, but during a prolonged drought, nocturnal stomatal opening and high accumulation of titratable acid in leaves were found (^{Edwards & Díaz 2006}).

In the Asparagaceae, subfam. Agavoideae, Agave spp., Hesperaloe spp., and Manfreda spp. are considered to show CAM, whereas Beschorneria and Polianthes are genera having C₃-CAM photosynthesis, and Yucca has C₃ species and CAM species (^{Heyduk et al. 2016}). In the Orchidaceae, more than 10 % of species have CAM (^{Silvera et al. 2009}), whereas in the Bromeliaceae 57 % of species exhibit CAM (^{Crayn et al. 2015}).

CAM plants occupy unfavorable environments for the growth of plants with C₃ photosynthesis. Most of the CAM species inhabit warm, seasonally dry habitats such as semi-deserts with little, but relatively predictable, seasonal rainfall (^{Winter et al. 2015}). The existent richness of CAM species of large stature and other major terrestrial CAM lineages seem to have emerged during the global expansion of arid environments in the late Miocene (^{Arakaki et al. 2011}, ^{Horn et al. 2014}). Among epiphytic lineages, CAM is considered to have facilitated diversification of the more extreme epiphytic life-forms occupying arid microhabitats in forest canopies, remarkably in the tropical Bromeliaceae and Orchidaceae (^{Crayn et al. 2004}, ²⁰¹⁵, ^{Silvera et al. 2009}). Thus, while the physiognomy of semi-deserts outside the tropics is often determined by succulent CAM plants, tropical forests house many more CAM plants in terms of biomass and species diversity (^{Lüttge 2010}).

Vascular epiphytes are elements occurring in almost all ecosystems of the world, but they are particularly diverse in evergreen rain forests and cloud forests of tropical and subtropical regions (^{Silvera & Lasso 2016}). Vascular epiphytes constitute a heterogeneous group that represents approximately 10 % of the tracheophytes, with more than 31,000 species distributed in 79 families, although the vast majority (80 %) belong to the Orchidaceae, Bromeliaceae and Polypodiaceae (^{Zotz et al. 2021}). In Mexican flora, the proportion of vascular epiphytes is 7.8 %, with approximately 1,813 species distributed in 37 families, mainly in the Orchidaceae (50 %), Bromeliaceae (13 %) and Polypodiaceae (7 %) and are found with greater diversity in the cloud forest or mesophyllous mountain forest (^{Espejo-Serna et al. 2021}).

It has been estimated that most epiphytic species exhibit CAM photosynthesis (^{Zotz & Hietz 2001}, ^{Lüttge 2004}, ^{Andrade et al. 2007}, ^{Silvera & Lasso 2016}, ^{Zotz et al. 2021}). About 50 % of the approximately 30,000 epiphytic species of the Orchidaceae family are CAM (^{Winter & Smith 1996}, ^{Silvera & Lasso 2016}). In Bromeliaceae, CAM is common in the Bromelioidae subfamily (90 % of species; ^{Zotz et al. 2021}), and less common in the Tillandsioidae subfamily (28 % of species; ^{Crayn et al. 2015}). Also, all epiphytic cacti are CAM (^{Zotz et al. 2021}). In Mexico, CAM epiphytic bromeliad species predominate in areas with dry climate or seasonal drought in Mexico (^{Andrade et al. 2007}). Although it has been documented that approximately 19 % of epiphytic species used the CAM mechanism in a dry forest (^{Zotz 2004}), a recent analysis showed that the global proportion of CAM species was similar between epiphytic and terrestrial plants and that most vascular plant families with epiphytic species are not CAM (^{Zotz et al. 2021}). However, the facts that CAM epiphytic species are more abundant than C₃ epiphytic species in exposed canopy sites in tropical moist and dry forests (^{Griffiths & Smith 1983}, ^{Zotz & Ziegler 1997}, ^{Lüttge 2004}, ^{Silvera & Lasso 2016}) and that almost all epiphytic species in dry deciduous forests of Mexico use CAM (^{Reyes-García et al. 2012}, ^{Valdez-Hernández et al. 2015}) reveal that it is a highly favorable photosynthetic mechanism for their existence.

CAM plants in Mexico. Mexican CAM plants are distinctive of arid and semiarid regions of Mexico where, sometimes, they are the more conspicuous evergreen component in the Sonoran, Chihuahuan and Tehuacán-Cuicatlán deserts (^{Valiente-Banuet & Godínez-Álvarez 2002}) and even in some endangered and disturbed tropical dry forests (^{Durán & Olmsted 1999}, ^{Arias-Medellín et al. 2014}). In these natural habitats, CAM plants are essential because during the dry season they provide water and food to animals. Indeed, because of this high-water use efficiency, some native and cultivated CAM terrestrial plants, under irrigated and fertilized conditions, surpass in productivity many C₃ and C₄ crops (^{Nobel 1991}, ¹⁹⁹⁶, ^{Andrade et al. 2009}). Furthermore, the conspicuousness of many Mexican terrestrial CAM plants, such as the opuntias, agaves and saguaros, and of those that are extensively cultivated, has made these plants emblematic of Mexican culture (^{Nobel 2002}, ^{De la Barrera & Andrade 2005}).

First studies of CAM in Mexico. The first Mexican study on CAM plants was published by ^{Tinoco-Ojanguren & Vázquez-Yanes (1983)}. These authors found that the epiphytic plant Epiphyllum crenatum (Lindley) Don. (Cactaceae) and the tree Clusia lundellii Standl. (Clusiaceae) have CAM photosynthesis, and this paper was the first one in showing CAM in a tree. Later, ^{Vazquez-Yanes (1989)} and ^{Larqué-Saavedra (1994)} in the Boletín de la Sociedad Botánica de México (now Botanical Sciences) communicated the increase in scientists and studies on plant ecophysiology and plant physiology in Mexico. However, despite the great diversity of plants and ecosystems and the enormous richness of CAM plants in Mexico, physiological studies of native plant species are still incipient (^{De la Barrera & Andrade 2005}, ^{Andrade et al. 2007}, ^{Valdez-Hernández et al. 2015}, ^{Tinoco-Ojanguren et al. 2018}, ^{Briones et al. 2020}).

In this review, advances on the physiological processes of the Mexican CAM plants (seed germination, photosynthesis, and water relations) are highlighted. Also, ecophysiological responses to plants stress and novel ecophysiological research are considered to increase knowledge of potential responses from Mexican CAM plants to global climate change.

Germination of Mexican CAM plants

The first germination study of a native Mexican CAM species was conducted on the saguaro cactus (Carnegiea gigantea) by ^{Alcorn & Kurtz Jr (1959)} in Arizona; the saguaro is a CAM plant distributed in the southern United States as well as in northern Mexico. These authors evaluated some factors affecting the seed germination of the saguaro and found that germination is stimulated by red light and that the optimum temperature for germination was 25 °C. Later, more seed germination experiments on Mexican CAM plants have been done, mainly evaluating soil water potential, hydration memory, light conditions, temperature, mucilage and microbiota, seed bank, and seed ageing.

Soil water potential. In arid and semiarid environments, soil water potential is one of the key factors affecting seed germination (^{Flores & Briones 2001}, ^{Guillén et al. 2009}, ²⁰¹¹, ²⁰¹⁵, ^{Flores et al. 2017}, ^{Barrios et al. 2020}). Seed germination of Mexican cacti occurs mainly at water potentials of between 0 to -0.6 MPa, and species differ in the minimum water potential at which germination can occur (Table 1). Thus, cactus seeds adapted to germinate at high soil moisture, but not necessarily at the soil field water capacity, have an advantage in arid environments (^{Barrios et al. 2020}). For instance, of the 660 Mexican cactus species (^{Ortega-Baes & Godínez-Álvarez 2006}), the effect of water potential on seed germination has been studied in only 17 species: most of them columnar (12), two globose, two barrel, and one epiphytic (Table 1). Most of these studies have used polyethylene glycol (PEG) 6000 or 8000 to simulate water stress on seeds. These PEG solutions are osmotic media with large molecular weight that cannot penetrate the living cell wall, and they have been used to impose a water potential stress on seeds by decreasing the water potential (Table 1).

Of the 402 Mexican taxa of the family Asparagaceae, subfamily Agavaoideae (^{García-Mendoza & Galván 1995}), the effect of water potential on seed germination has been studied in 17 species, 13 of the stemless rosette-like (all Agave species) and four rosette-like with stem (three Yucca species and Beaucarnea gracilis). Species of the subfamily Agavaoideae appear to be more tolerant to water stress for germination than cacti, because eight of the species (seven Agave species plus Yucca elata) show germination > 80 % at water potential of -1 MPa (Table 1).

For Bromeliaceae, water potential for germination has been studied in only one species (Tillandsia recurvata); seeds from this species increase germination at -0.1 MPa and decrease it at -0.6 MPa (Table 1). Mexican species from this and other families showing CAM still need to be evaluated.

Hydration memory. Desert plants have developed strategies to germinate even in places where available moisture is sporadic (^{Dubrovsky 1996}, ^{Contreras-Quiroz et al. 2016a}). Seeds of some cactus species have desiccation tolerance of imbibed seeds or “hydration memory,” a seed ability to retain some of the physiological changes, like the differential expression of proteins (^{López-Urrutia et al. 2014}), induced by hydration-dehydration cycles (^{Dubrovsky 1996}). Reduced germination times have been found after treating the seeds with hydration-dehydration cycles for three cactus species (Stenocereus thurberi, Pachycereus pecten-aboriginum and Ferocactus peninsulae), from the Sonoran Desert in Mexico (^{Dubrovsky 1996}). In addition, higher germination percentage with hydration-dehydration cycles have been found for Ferocactus pilosus and Echinocactus platyacanthus; however, Yucca filifera seeds germinated more with the control than with the hydration-dehydration treatments. Thus, “hydration memory” is not a trait present in all CAM species (^{Contreras-Quiroz et al. 2016a}) nor exclusive to CAM species, e.g., several non-CAM species of the Mediterranean region show seed hydration memory (^{Copete et al. 2021}).

^{Contreras-Quiroz et al. (2016b)} assessed if seed hydration memory depends on climate. These authors studied seven species from the Argentinian Córdoba Mountains (mesic environment): Gymnocalycium capillense, Parodia mammulosa, Echinopsis candicans, Gymnocalycium bruchii, Gymnocalycium mostii, Gymnocalycium quehlianumand, Gymnocalycium monvillei, and two species (Echinocactus platyacanthus and Ferocactus pilosus) from the Mexican Chihuahuan Desert (dry environment). Four hydration (hours)/dehydration (days) treatments were employed: T1 = 24 h/5 days, T2 = three successive cycles of 24 h/5 days, T3 = 72 h/5 days, and T4 = control (untreated seeds). The treatments were carried out in germination chambers at a constant temperature of 25 °C, with 12 h of light and 12 h of darkness, in containers with distilled water. The authors found that the Mexican species respond to at least one hydration-dehydration treatment by increasing their germination (54 % in untreated and 88 % in T3 for E. platyacanthus, and 38 % in untreated seeds and 93.8 % for F. pilosus). For the Argentinian species, only G. mostii increases its germination in a hydration-dehydration treatment (24 % in untreated and 68 % in T2). This species inhabits rocky and hence drier habitats than the other Argentinian species studied. Thus, it appears that seed hydration memory in cacti depends on climate.

Soil temperature. Besides water, soil temperature may also be important in seed germination in arid and semiarid environments (^{Rojas-Aréchiga & Vázquez-Yanes 2000}, ^{Barrios et al. 2020}). There are CAM species in which optimum germination temperatures coincide with the mean germination season temperatures of their local habitats (^{Flores & Briones 2001}). However, when temperature is close to the seed germination optimum, water potential becomes less restrictive (^{Flores et al. 2017}).

Seed germination of Mexican CAM species (Asparagaceae and Cactaceae) takes place at constant temperatures between 20 and 30 °C, and species differ in the minimum and maximum temperature at which germination can occur (Table 2). The effect of constant temperature on seed germination has been studied in 31 CAM species (22 Cactaceae, 8 Asparagaceae, and 1 Bromeliaceae). Twelve species are barrel, seven are columnar, six are stemless rosette-like, three species are rosette-like with stem, two are globose, and one is epiphytic (Table 2).

Alternating temperature effects on seed germination of CAM species are not very clear, e.g., for cacti, in most studies only constant temperatures have been used, and some experiments that included alternating temperatures did not reveal any significantly different effects on germination compared to constant temperatures, or simply do not favor germination (^{Rojas-Aréchiga & Vázquez-Yanes 2000}). However, for Mamillaria gaumeri the highest germination occurs at diurnal/nocturnal temperatures of 30/20 °C and lowering or raising the diurnal/nocturnal temperature by 5 °C reduces germination by 10 % (^{Cervera et al. 2006}).

Because seeds of CAM plants on the surface of desert soils would be subjected to elevated temperatures, exposure of seeds to high temperatures has often been explored as a pre-treatment to promote germination of desert CAM plants. For instance, for the Cactaceae, seeds of Pachycereus pringlei have higher germination at 40 °C for 7 d and lower at 70 °C for 21 d (^{Cancino et al. 1993}). For Mammillaria magnimamma, 90 °C for 4 and 12 h do not influence germination percentage (^{Ruedas et al. 2000}). For Ferocactus histrix, germination is constant at 40 and 70 °C for 7 d and 21 d; also, seeds of Isolatocereus dumortieri show higher germination after exposure to 40 °C for 14 d and lower at 70 °C for 7 d, and Echinocactus platyacanthus has the highest germination after exposure to 70 °C for 14 d and the lowest with no exposure to heat (^{Pérez-Sánchez et al. 2011}). Among the Agavoideae members, germination for Agave salmiana is consistent at 40 and 70 °C for 7 d and 21 d, and for both Agave lechuguilla and Yucca decipiens, seed exposure to 70 °C results in the lowest germination percentage both at 7 and 14 d (Pérez-Sánchez et al. 2011). Thus, it appears that high temperature pretreatments as a means of increasing germination percentages are not so good for members of Agavoidae, but they are good for some cacti. In addition, hot temperature pretreatments are not exclusive to CAM species, because high temperature also promotes seed germination or breaks seed dormancy in fire-prone habitats with non-CAM species (^{Luna et al. 2019}).

Light environment. Light is another important environmental signal to which the life cycle of a plant may respond to in any environment (^{Gutterman 1993}). The promoting effect of light on seed germination was documented by Caspary in 1860 (^{Shinomura 1997}). Of the various photoreceptor systems in the seeds, phytochrome plays an especially important role in seed germination (^{Shinomura 1997}). Phytochrome is a blue protein pigment with a molecular mass of about 125 kDa (^{Taiz et al. 2015}). ^{Borthwick et al. (1952)} discovered the red/far-red photoreversible effect on seed germination. Absorption of red light converts the inactive P_r form of phytochrome (which inhibits germination) to the active P_fr form (which promotes germination); the absorption of far-red light converts P_fr to P_r (^{Shinomura 1997}). Thus, in many species if seeds are exposed to red light they germinate, but if they are exposed to far-red light they do not germinate. However, in some seeds, far-red light can promote germination (^{Baskin & Baskin 2014}); thus, the active form of phytochrome stimulates germination in darkness in some species (^{Shinomura 1997}).

In terms of their responses to light for germination, plants can be classified as follows: (i) those that require light to germinate (positive photoblastic), (ii) those that require darkness to germinate (negative photoblastic), and (iii) those that have a large percentage of seeds neutral to light (neutral photoblastic, ^{Flores et al. 2016}). Positively photoblastic seeds may detect through the phytochrome a transference from a dark to an illuminated environment when buried seeds are exhumed. Such seeds may also detect the change in the red (655-665 nm): far red (725-735 nm) ratio (R:FR) of the light (^{Vázquez-Yanes & Orozco-Segovia 1996}).

Light requirements and small seed mass have coevolved as an adaptation to ensure germination in desert plants (^{Fenner & Thompson 2005}). ^{Flores et al. (2006)} investigated the effect of light on seed germination of 28 cactus species from the Chihuahuan Desert. Only 11 of the species have non-dormant seeds, germinating ≥ 70 % in the light; these species are positively photoblastic, and all of them have small seeds. Ungerminated seeds in darkness do not germinate to a significantly higher percentage when the same set of seeds is transferred from dark to light, suggesting that darkness triggers a secondary dormancy type named skotodormancy, as an adaptive strategy in xeric environments when conditions are not optimal for germination. Similarly, ^{Rojas-Aréchiga & Mandujano-Sánchez (2017)} found that Mammillaria compressa and Turbinicarpus gielsdorfianus develop skotodormancy after darkness incubation for one month. However, skotodormancy is not exclusive to CAM species, because it has also been found in cultivated plants (^{Bewley 1980}).

For cactus species belonging to all tribes (including many Mexican species), lower light requirement is found for seeds from taller than for shorter taxa, and lower for taxa with heavier seeds than for taxa with lighter seeds (^{Flores et al. 2011}). However, for 54 Mexican cactus species belonging to the tribe Cacteae, no relationship between the requirement of light for germination and seed mass is found (^{Rojas-Aréchiga et al. 2013}). These results could be explained by the generally small seed mass in cacti from tribe Cacteae (seed mass from the studied species belonging to 15 genera ranged from 0.025 mg to 3.95 mg; ^{Rojas-Aréchiga et al. 2013}). Seed mass in other taxa is bigger, e.g., from 0.46 mg to 9.6 mg for Pachycereeae, and from 6.99 mg to 16 mg for Opuntioideae (^{Flores et al. 2011}).

Thirteen members of the Agavoideae are neutral photoblastic: Agave americana, A. angustifolia, A. asperrima, A. gentry, A. filifera, A. lechuguilla, A. salmiana, A. striata, Yucca carnerosana, Y. elata, Y. filifera, Y. potosina, and Y. queretaroensis (^{Jiménez-Aguilar & Flores 2010}, ^{Flores et al. 2016}). These findings are not surprising since these species have big seeds; however, more studies are necessary to better know the photoblastic responses of species of the Agavoideae and of the other CAM families.

There are cactus species having small seeds and neutral photoblastism, e.g., Mammillaria compressa and Coryphantha clavata (^{Flores et al. 2016}). In addition, these authors found that seedlings from these two species under darkness conditions were columnar and those exposed to light were cylindrical, on the basis that the cylindrical growth-form has a stem < 2 times higher in length than in diameter, and the columnar one has a stem 2-5 times higher in length than in diameter (^{Vázquez-Sánchez et al. 2012}). It is possible that seeds from this species can germinate in the dark because they produce columnar seedlings with the ability to emerge from greater soil depths where sunlight cannot penetrate.

Mucilage and microbiota. Adaptive traits have been found in some cactus species to germinate in arid and semiarid environments, e.g., seeds from some species have a secretory layer (mucilage) that absorbs water and distributes it over the seed surface; this mucilage surrounds the ripe seed and improves germination under relatively dry conditions (^{Bregman & Graven 1997}, ^{Mascot-Gómez et al. 2020}). The production of seed mucilage is known as myxospermy (^{Western 2012}), and seeds from five Mexican cactus species (Coryphanta maiz-tablasensis, Echinocactus platyacanthus, Ferocactus latispinus, F. pilosus and Stenocereus queretaroensis) from the Chihuahuan Desert have been found to have mucilage (Mascot-Gómez et al. 2020). This mucilage results in higher germination percentage in three species: E. platyacanthus (88.5 vs. 21.1 % without mucilage), F. latispinus (88.5 vs. 48.2 %) and S. queretaroensis (96.0 vs. 1.0 %), and a lower germination time for E. platyacanthus (10.0 vs. 19.5 d without mucilage), F. pilosus (14.1 vs. 16.4 d) and F. latispinus (7.8 vs. 14.0 d) (^{Mascot-Gómez et al. 2020}). Mucilage in seeds has not been investigated for other Mexican CAM species; however, it is not a trait exclusive to CAM species because it occurs also in a broad range of species, from the Acanthaceae to the Brassicaceae to the Linaceae to the Plantaginaceae (^{Western 2012}).

Besides mucilage, some cactus seeds have a microbiota comprising potential pathogens and beneficial microorganisms that could influence the germination percentage. For example, Echinocactus platyacanthus seeds having episeminal microbial communities have better germination percentage and are less susceptible to colonization by fungi than washed seeds (^{Mascot-Gómez et al. 2021}). For Opuntia spp. seeds, fungi attack the testa, eroding or cracking the hard/stony endocarp, and thus reduce the mechanical resistance to germination in dormant seeds (^{Delgado-Sánchez et al. 2010}, ²⁰¹¹, ^{Sánchez-Coronado et al. 2011}). However, the positive effect of fungi on Opuntia seeds depends on the light condition; ^{Delgado-Sánchez et al. (2013a)} found that the fungi Phoma medicaginis, Trichoderma harzianum, T. koningii, and Penicillium chrysogenum erod the funicular envelope and promote seed germination for O. leucotricha and O. streptacantha but this is more so in light than in darkness. The effects of microorganisms on seed germination for other cacti and other CAM species remain to be investigated.

Seed bank. The seed bank is constituted by the viable seeds available for potential germination and recruitment of new plants (^{Montiel & Montaña 2003}). Species that form seed banks show small seed mass and several ecophysiological features that permit them to persist in the soil for some time, including light requirement for germination (positive photoblastism), some dormancy mechanisms, a period of post maturation for germination, and high ecological longevity (^{Bowers 2000}; ^{Rojas-Aréchiga & Batis 2001}). Also, differences in nurse plant availability and in rodent density can explain the differences between seed bank dynamics in desert ecosystems (^{Montiel & Montaña 2003}).

The information in the reviewed literature about the seed bank of Mexican CAM species is scarce, and most of them is about cacti. For O. rastrera, about 15,000 seeds/ha in nopaleras (Opuntia-dominated stands) and 2,500 seeds/ha in grasslands remain stored in the soil and are able to germinate as soon as environmental conditions are suitable (^{Montiel & Montaña 2003}). The seeds of Ferocactus wislizeni can persist in the soil for at least 18 months or up to 3 years and, if not consumed by predators, can form a between-year seed bank (^{Bowers 2000}, ²⁰⁰⁵). Mammillaria grahamii seeds survive in or on the soil as long as six years, forming a long-term persistent seed bank, and also a between-year seed bank (^{Bowers 2005}). Opuntia tomentosa seeds also can persist in the soil for at least 18 months (^{Olvera-Carrillo et al. 2009}). One more species forming long-term persistent soil seed banks is Polaskia chende, whose seeds survive in the soil over a 5-year period (^{Ordoñez-Salanueva et al. 2017}).

Other group of CAM species has the potential to form a short-term persistent seed bank, such as the columnar cacti Carnegiea gigantea (^{Bowers 2005}) and Stenocereus stellatus (^{Álvarez-Espino et al. 2014}). Recently dispersed S. stellatus seeds do not germinate because they have primary dormancy; however, this dormancy is broken after 6 months of burial in the soil. Seeds buried for 10 months enter secondary dormancy and they are not viable at 24 months, probably because of pathogen attack (^{Álvarez-Espino et al. 2014}). Similarly, germination of buried seeds from two Asparagaceae (Agave striata and Yucca filifera) is high in spring and summer, but no seeds remain viable to test for germination in the other seasons (^{Aragón-Gastélum et al. 2018}). Epiphyte seeds are non-dormant and do not build seed banks (^{Benzing 1990}). However, there is evidence of germination from aged Tillandsia caput-medusae seeds (^{Flores-Palacios et al. 2015}). Thus, more research on aerial seed bank is needed.

Seed ageing. Some CAM species show lower germination in fresh seeds than in old seeds, e.g., Ferocactus spp. (^{Zimmer 1980}, ^{Bowers 2000}), Opuntia spp. (^{Potter et al. 1984}, ^{Mandujano et al. 1997}), Mammillaria heyderi (^{Trejo-Hernández & Garza-Castillo 1993}), Stenocereus stellatus (^{Rojas-Aréchiga et al. 2001}), S. queretaroensis (^{De la Barrera & Nobel 2003}), Turbinicarpus lophophoroides and T. pseudopectinatus (^{Flores et al. 2005}), and Ferocactus peninsulae (^{Rojas-Aréchiga & García-Morales 2022}). Therefore, these species are likely adapted to germinate only after heavy rainfall events, forming soil seed banks. These species could have a post-maturation period (after ripening), which is associated with seed bank formation and permanence (^{Rojas-Aréchiga & Batis 2001}). It has been suggested that seed banks result from evolution of dormancy mechanisms as a response to unpredictable environments such as deserts and semi-deserts (^{Jurado & Flores 2005}).

Physiological dormancy is the type of dormancy found in the Cactaceae (^{Rojas-Aréchiga & Vázquez-Yanes 2000}; ^{Barrios et al. 2020}). Seeds with physiological dormancy can cycle through a gradation of dormancy ‘states’ in response to environmental cues, during which the range of conditions under which the seeds are able to germinate expand and contract (^{Long et al. 2015}). Dormancy cycling, which indicates that the ungerminated seeds are viable seeds in the soil seed bank, has been found for Polaskia chende from the Tehuacán Valley (^{Ordóñez-Salanueva et al. 2017}), as well as for Echinocactus platyacanthus seeds from the Southern Chihuahuan Desert (^{Aragón-Gastélum et al. 2018}). Seeds of E. platyacanthus acquired secondary dormancy in the rainy season (summer and autumn), which was alleviated at the end of the subsequent dry season (winter), possibly because of the high variation recorded in mean and minimum soil temperature at the end of winter.

Other CAM species do not show significant differences between fresh and aged seeds, e.g., Echinocereus spp. and Ferocactus spp. (^{Fearn 1977}, ¹⁹⁸¹), as well as Turbinicarpus knuthianus (^{Flores et al. 2005}). For Tillandsia caput-medusae seeds, fresh seeds have 99.5 % germination success and after more than one year they only decline to 81.7 %, which indicates that seeds of this species can germinate in at least two consecutive rainy periods (^{Flores-Palacios et al. 2015}). These results can be interpreted as a strategy to produce both seeds for immediate propagation and seeds for securing the medium and/or long-term survival of the species; thus, these species could form seedbanks under natural conditions, but this remains to be investigated.

Some CAM species show higher germination in fresh seeds than in old seeds, e.g., Cephalocereus senilis, Echinocereus reichenbachii, Escobaria tuberculosa, Ferocactus acanthodes (Ferocactus cylindraceus) (^{Fearn 1977}), Tillandsia recurvata (^{Fernández et al. 1989}; ^{Flores-Palacios et al. 2015}), Agave americana (^{Pritchard & Miller 1995}), Opuntia tomentosa Salm-Dick (^{Olvera-Carrillo et al. 2003}), Turbinicarpus swobodae and T. valdezianus (^{Flores et al. 2005}), Mammillaria huitzilopochtli (^{Flores-Martínez et al. 2008}), Rhipsalis baccifera (^{de la Rosa-Manzano & Briones 2010}), Tillandsia achyrostachys, T. circinnatioides, T. hubertiana, and T. schiedeana (^{Flores-Palacios et al. 2015}), as well as Lophophora williamsii (^{Mandujano et al. 2020}). All these species are CAM plants, but there is no information about the photosynthetic pathway of T. hubertiana (^{Crayn et al. 2015}). These species might not have the ability to form seedbanks in the soil. A higher germination in fresh seeds would appear to represent a special form of adaptation to their desert habitat, which is characterized by short and irregular rainfall periods. In such a climate, seeds must germinate as soon as rainfall occurs (^{Flores et al. 2004}).

Photosynthesis of terrestrial Mexican CAM plants

Environmental conditions in the natural habitats where terrestrial CAM plants occur affect CAM photosynthesis. This is particularly severe during the dry season because drought is accompanied by elevated temperatures and high photosynthetic photon fluxes (PPF); also, during drought, some CAM phases can be reduced or lost as drought evolves (^{Nobel 2003}, ^{Lüttge 2004}, ^{Andrade et al. 2007}). In Mexico, most studies on CAM photosynthesis have been done with plants found in arid or semiarid places (^{Nobel 2003}). However, tropical dry deciduous forests of Mexico are rich in terrestrial CAM plants (^{Ricalde et al. 2010}, ^{Valdez-Hernández et al. 2015}, ^{Tinoco-Ojanguren et al. 2018}).

Light microenvironments. Light varies seasonally for terrestrial CAM plants because the rainy season is cloudier than the dry season. Furthermore, in tropical forests, the light environment is low but extremely variable for the understory terrestrial CAM plants, where CAM provides a more rapid response to sunflecks than for other nearby C₃ shade adapted plants (^{Skillman et al. 1999}). In contrast, in tropical dry deciduous forests, the tree deciduousness makes light environment to change more dramatically (^{Graham & Andrade 2004}, ^{Andrade et al. 2006}, ^{González-Salvatierra et al. 2013}), favoring CAM terrestrial plants over C₃ plants in some sites (^{Ricalde et al. 2010}, ^{Valdez-Hernández et al. 2015}). However, it is not easy to separate the light changes effect on CAM from the lack of water or temperature changes, since all these variables usually correlate (^{Andrade et al. 2009}).

The maintenance of the electron transport during the day as well as the photorespiration confer photoprotection for CAM plants (^{Niewiadomska & Borland 2008}). Also, the xanthophyll cycle helps protecting the photosynthetic apparatus from photoinhibitory damage, particularly during drought (^{Lu et al. 2003}). Acclimation to high light during the dry season also involves the biosynthesis of antioxidant metabolites to neutralize reactive oxygen species (^{González-Salvatierra et al. 2010}, ^{González-Salvatierra & Flores 2019}).

Excess light reduces photosynthesis in tropical CAM plants. For instance, the hemiepiphytic cactus Hylocereus undatus has a maximal net CO₂ uptake at a PPF of 20 mol m^-2 d^-1, above which photoinhibition reduces CO₂ uptake (^{Nobel & De la Barrera 2004}). For desert CAM plants, although water stress usually increases the negative high radiation effects, light effects are more pronounced for young CAM plants (^{Flores et al. 2004}, ^{Andrade et al. 2006}, ^{Cervera et al. 2006}, ²⁰⁰⁷, ^{Gallardo-Vásquez & De la Barrera 2007}). However, when water is available, seedling of several species of cacti can survive high light levels using CAM (^{Hernández-González & Briones Villareal 2007}). Likewise, one of the adaptations for survival of CAM seedlings, in open areas in well-watered conditions, is the chloroplast movement to avoid high PPF (^{Delgado-Sánchez et al. 2013b}).

Leaves and stems of CAM terrestrial plants are massive and mature plants cannot easily orient their leaves or stems to get more PPF or to avoid it. Because of this, PPF levels influence leaves or stems arrangement and the production of new photosynthetic surfaces. For instance, opuntias living between the 27° latitudes north and south from the equator tend to develop cladodes facing east-west, which maximizes PPF incidence on cladode surfaces (^{Nobel 2003}). Comparably, ribs of the columnar cactus Myrtillocactus geometrizans facing south produce more fruits than ribs facing other directions (^{Ponce-Bautista et al. 2017}).

Water, CO ₂ uptake and chlorophyll fluorescence. When there is enough water in the soil, CO₂ uptake occurs mostly at night for terrestrial CAM plants. For instance, in Agave fourcroydes, a cultivated agave from Yucatan (henequen), 91 % of the CO₂ fixed in a 24-h period is at night (^{Nobel 1985}). For A. lechuguilla, in the Chihuahuan desert, 85 % of the daily CO₂ uptake occurs at night (^{Nobel & Quero 1986}). After a slight drought, total net CO₂ uptake for A. fourcroydes decreases to about 22 % and to less than 1 % after a 30-day drought, but fully recovers after 7 d of watering (^{Nobel 1985}). Agave lechuguilla is more sensitive to drought than A. fourcroydes since after 13 d of drought, its daily CO₂ uptake decreases to 0.8 % of the total daily CO₂ uptake for well-irrigated plants (^{Nobel & Quero 1986}).

When soil water is at field capacity and temperatures are moderate (day/night air temperatures of 25/20 °C), many members of the genera Opuntia and Agave have net CO₂ uptake at the beginning of the light period (^{Cui & Nobel 1994}, ^{Nobel 2003}, ^{Andrade et al. 2009}). This CO₂ uptake during the early morning or late afternoon via C₃ makes CAM crops, such as Opuntia ficus-indica or A. angustifolia, highly productive compared to C₃ or C₄ crops (^{Acevedo et al. 1983}, ^{Nobel 1991}, ¹⁹⁹⁶, ^{José Jacinto & García Moya 1995}).

Another way to evaluate the photosynthetic performance of CAM plants is using chlorophyll fluorescence. For instance, for Agave salmiana, drought causes a direct effect on photosystem II (PSII) photochemistry in light-acclimated leaves, as indicated by a decrease in the photosynthetic electron transport rate. Additionally, down-regulation of photochemical activity occurs mainly through the inactivation of PSII reaction centers and an increased non-photochemical quenching (NPQ, thermal dissipation capacity) of the leaves. These results suggest that the thermal dissipation of excess energy can be an important acclimation mechanism to protect the photosynthetic apparatus from over-excitation in droughted Agave plants (^{Campos et al. 2014}).

Similarly, ^{González-Salvatierra & Flores (2019)} report several mechanisms that enable Yucca filifera saplings to endure water shortage. These authors showed that maximum quantum efficiency and effective quantum yield of photosystem II values, as well as titratable acid and chlorophyll a/b ratio, are strongly affected by water shortage, and their values reverted after rewatering. They also revealed higher proline content, antioxidant activity, and high NPQ during the water-shortage treatment, which represent mechanisms for preventing photodamage. The high ability of Y. filifera saplings to withstand stress caused by water shortage indicates that this species can be a key species for conservation and restoration of degraded arid ecosystems.

Temperature and CAM photosynthesis. Daytime air and leaf or stem temperatures are no critical for the net daily CO₂ uptake (^{Nobel 2003}). Because the initial CO₂ fixation by PEP carboxylase is at night, the nocturnal temperature becomes more important for CAM photosynthesis. As nocturnal temperature increases, stomata conductance decreases, which indicates that stomata close at higher nighttime temperatures for desert CAM plants (^{Nobel 2003}). The optimal average nocturnal temperature for CO₂ uptake of O. ficus-indica is 15 °C but significant net CO₂ uptake occurs at 0 °C (^{Cui & Nobel 1994}, ^{Nobel & Bobich 2002}). Optimal temperatures for CO₂ uptake in tropical CAM terrestrial plants are higher than those for desert ones. For instance, for Agave fourcroydes and Hylocereus undatus, these optimal diurnal/nocturnal temperatures are 30/20 °C (^{Nobel 1985}, ^{Nobel & De la Barrera 2004}). Similarly, desert and cultivated cacti can tolerate temperatures as high as 70 °C (^{Nobel 2002}, ²⁰⁰³) but tropical H. undatus and Mammillaria gaumeri cannot tolerate temperatures above 50 °C (^{Nobel et al. 2002}, ^{Cervera et al. 2006}).

Low temperatures are more critical for the distribution or cultivation of CAM plants than are high ones. For instance, stem temperatures of -7 to -10 °C are lethal for Carnegiea gigantea, Opuntia ficus-indica and Stenocereus queretaroensis (^{Nobel 2002}, ²⁰⁰³). Then, global climate warming would favor northward expansion for Mexican CAM terrestrial plants. Indeed, induce warming experiments on seedlings of two CAM species indicate that they acclimate by increasing photosynthetic pigments and NPQ (^{Aragón-Gastélum et al. 2014}, ²⁰²⁰). More physiological research at the cell level for this high-temperature tolerance, and energy balance at the whole individual level would be necessary for restoration and conservation of arid zones under global climate warming scenarios.

Carbon isotope ratios. Nocturnal acidification is restricted to CAM plants (^{Winter & Smith 2021}) but is not necessarily associated to CO₂ uptake, because during drought and/or higher temperatures stomata close and CO₂ released internally by respiration is also fixed (^{Nobel 2003}). A powerful technique for indicating whether a plant species uses the CAM pathway is the use of carbon isotope ratios. This is because CAM plants with CO₂ uptake at night by PEP carboxylase would have δ¹³C values from -11 to -18 ‰ (like values for C₄ species), whereas if all CO₂ is fixed by Rubisco δ¹³C values would range from -25 to -30 ‰ (^{Nobel 2003}, ^{Santiago et al. 2005}). The use of stable isotopes is particularly convenient in broad surveys in a particular location (^{Vargas-Soto et al. 2009}, ^{Santiago et al. 2017}) or for investigating photosynthetic pathway evolution in some plant families (^{Winter & Smith 2021}).

In the coastal dune of northern Yucatán, tissue δ¹³C values for terrestrial CAM plants are 2 ‰ higher than for CAM plants from the tropical dry deciduous forests (^{Ricalde et al. 2010}), which indicates greater CO₂ fixation through the CAM pathway in sites with less water. Also, using tissue δ¹³C values of specimens of the genus Clusia from herbaria in Mexico, it was possible to find out that Clusia species are CAM in tropical dry forests but in the rest of the country there is a preponderance of species with C₃ and C₃-CAM photosynthesis (^{Vargas-Soto et al. 2009}).

Long-lived succulents contain a record of past variation in growth and photosynthesis coupled to climatic conditions. Using the spines of Carnegiea gigantea, it was revealed that their signatures of stable isotopes of carbon (δ¹³C) and oxygen (δ¹⁸O) are coupled to environmental variation (^{English et al. 2007}, ²⁰¹⁰). In fact, working with several saguaro populations, ^{Hultine et al. (2018)} demonstrated that stem growth and photosynthetic physiology recorded by δ¹³C are coupled to a complex set of environmental conditions from the previous summer and winter.

Photosynthesis of Mexican epiphytic CAM plants

CO ₂ uptake. Under greenhouse conditions, plants of the hemiepiphytic cactus Hylocereus undatus increased the nocturnal CO₂ uptake from 3.9 to 4.7 μmol CO₂ m^-2s^-1 when well-watered but decreased it from 4.3 to 2.9 μmol CO₂ m^-2s^-1 when water is withheld (^{Ortiz Hernández et al. 1999}). Similarly, measurements of the instantaneous flux of CO₂ (with 20, 40, 50 and 65 % of PPF) show that the plants of the epiphytic cactus R. baccifera have CO₂ uptake during the day and the night period, restricting daytime stomatal opening after a month without irrigation (^{de la Rosa & Briones 2008}). Likewise, gas exchange measurements for two Tillandsia species from a tropical dry deciduous forest show that for T. brachycaulos 12 % of the total carbon gain is fixed during daytime, whereas for T. elongata, 25 % of the carbon is fixed in the late afternoon via C₃ photosynthesis; also, for both species, after six days of drought, daytime net CO₂ uptake stops, and nighttime CO₂ uptake is reduced as drought progressed (^{Graham & Andrade 2004}).

The reaction norm of a transplant experiment of T. utricularia plants from populations of a coastal scrub to an evergreen forest and from an evergreen forest to a coastal scrub, shows that CO₂ exchange rates are higher under conditions of high light and irrigation in the plants from both provenances, reaching a maximum mean of 2.29 μmol CO₂ m^-2s^-1 in the individuals of the forest population and the lowest values (0.64 μmol CO₂ m^-2s^-1) when they grew in the scrub under drought; by contrast, individuals from the scrub population have less variation in carbon exchange in both environments (^{Rosado-Calderón et al. 2018}).

Tissue acidity. Measurements of foliar or stem acidity show that, in epiphytic CAM plants, the effect of stress depends on the acclimatization capacity of each species to face the diurnal and seasonal environmental changes. For instance, in four CAM epiphytic bromeliads (Aechmea bracteata, Tillandsia brachycaulos, T. dasyliriifolia, and T. streptophylla) the values of nocturnal acidification are higher in the rainy season, compared to those in the dry season in the tropical dry deciduous forest, coastal dune scrub and mangrove forest in the Yucatan Peninsula (^{Cach-Pérez 2008}, ^{González-Salvatierra et al. 2010}, ^{Ricalde et al. 2010}, ^{Cach-Pérez et al. 2014}). While the diurnal variation in tissue acidity (ΔH⁺) of T. dasyliriifolia in a mangrove decreases to about 30 %, and that of T. brachycaulos in the tropical dry deciduous forest to 78 %, compared to ΔH⁺ of the rainy season (^{Cach-Pérez et al. 2014}).

In general, CAM terrestrial plants show high nocturnal accumulation of organic acids with high incidence of PPF (^{Nobel 2003}). For instance, T. usneoides, a CAM epiphytic plant widely distributed in the Americas, shows high nocturnal accumulation at high PPF under laboratory conditions (^{Martin et al. 1985}). However, this high PPF must be accompanied by water in the substrate. For instance, in a transplant experiment with the obligate CAM epiphyte T. utricularia, the titratable acidity gradually decreases from a maximum of 431.9 mmol H⁺ m⁻² to almost zero as the relative water content decreases over time in a drought treatment for a native population of a Yucatán coastal scrub, while tissue acidity decreases even more as the water content decreases, with a maximum of 925.2 mmol H⁺ m⁻² to almost zero, in samples from an evergreen forest in Chiapas (^{Rosado-Calderón et al. 2018}). Similarly, tissue acidity showed that the epiphytic cactus R. baccifera achieves the highest photosynthesis under the highest PPF (65 %) when grown under irrigation but when growing under drought the highest photosynthesis is under moderate PPF (40 %) (^{de la Rosa & Briones 2008}).

Temperature and tissue acidity. The epiphytic CAM T. usneoides shows an optimum carbon dioxide assimilation between temperatures of 15 and 20 °C, but assimilation decreases at temperatures outside this range (^{Medina 1987}). In response to high air temperature at night, which is above 23 °C in the dry and wet seasons in the tropical dry deciduous forest were T. brachycaulos inhabits (^{Hernández-Robinson et al. 2020}), individuals reduce CAM activity, as measured by change in tissue acidity, during the dry season, but increased it by opening stomata in the II and IV phases of CAM during the rainy season in Yucatán, Mexico (^{González-Salvatierra et al. 2021}).

The epiphytic orchid Encyclia nematocaulon increases the leaf ΔH⁺ to 0.2 mol H⁺ m^-2 during the early dry season and decreases it 90 % in the dry season, but contrary to expectations, it shows intermediate acidity values during the rainy season in a tropical dry deciduous forest with a mean annual rainfall of 770 mm in Yucatán (^{de la Rosa-Manzano et al. 2014}). Similarly, the epiphytic orchids Cohniella yucatanensis and Laelia rubescens, in the same tropical dry deciduous forest, show the highest acidity values during the early dry season and decrease them during the rainy and dry seasons. However, the epiphytic orchids E. nematocaulon, C. ascendens and Lophiaris oerstedii show slight variation in foliar acidity and relative water content in the three seasons in a semi-deciduous tropical forest also located in Yucatán, but with a two-fold mean annual rainfall that of the tropical dry deciduous forest (^{de la Rosa-Manzano et al. 2014}). The lower nocturnal temperature in the early dry season, as well as the low vapor pressure deficit, which increases the probability of dew deposition, can favor CAM in these tropical forests (^{Andrade 2003}, ^{de la Rosa-Manzano et al. 2014}, ^{Chávez-Sahagún et al. 2019}).

Light, tissue acidity an electron transport rate. Under three light regimes of 20, 50 and 70 % of total daily radiation, the epiphytic orchid Stanhopea tigrina shows a strong reduction in nocturnal acidity under drought and high radiation, but the orchid Prosthechea cochleata has similar photosynthetic responses to the different PPF conditions during the dry period in a subtropical cloud forest (^{Guevara-Pérez et al. 2019}).

The quantum yield (Φ_PSII) of plants of the epiphytic cactus R. baccifera from the cloud forest under irrigation and low light (20 % of sunlight, 1.81 mol m^-2d^-1 PPF) was maximum (Φ_PSII pre-dawn = 0.8, Φ_PSII noon = 0.7) and decreased (Φ_PSII pre-dawn = 0.7, Φ_PSII noon = 0.3) at high light (65 % of sunlight, 9.8 mol m^-2d^-1 PPF); the electron transport rate (ETR) increases linearly with the increase of light (maximum values of ETR at low light with irrigation = 160-180 μmol m^-2s^-1, ETR after one month without irrigation = 120-140 μmol m^-2s^-1), and at high light it reaches saturation (ETR with irrigation = 138.9 μmol m^-2s^-1, ETR after one month without irrigation = 126.6 μmol m^-2s^-1) (^{de la Rosa & Briones 2008}). In the bromeliad T. brachycaulos, Φ_PSII values differ between shaded and exposed plants and between the dry and the rainy seasons; the lowest average of Φ_PSII (0.57) is in the leaves of exposed T. brachycaulos plants during the dry season and the highest value of Φ_PSII (0.80) is in leaves of shaded plants during the rainy season (^{González-Salvatierra et al. 2021}). Additionally, the maximum ETR of this epiphytic bromeliad is higher during the rainy season, with values of 23 and 32 μmol m^-2s^-1 in exposed and shaded plants, respectively, and the light saturation point of 800 μmol m^-2s^-1 in both light conditions; in contrast, during the dry season the maximum rate of electron transport is 12 and 18 μmol m^-2 s^-1 in exposed and shaded plants, respectively, and the light saturation point decreases to less than 465 μmol m^-2s^-1 in both light conditions (^{González-Salvatierra et al. 2021}). On the other hand, individuals of the epiphytic T. utriculata under a similar light environment (a coastal scrub) have a higher electron transport rate than those individuals of the same species from a population of a broadleaf forest transplanted to a coastal scrub site; after 20 d without irrigation, plants from the coastal scrub reduce ETR to 30 %, while those from the forest reduce ETR to 50 % (^{Rosado-Calderón et al. 2018}). Under the reduced light conditions of the forest site, ETR values do not differ between individuals from both populations. The effect of the drought is more evident in the coastal scrub site, where the maximum ETR values of the individuals of both populations decrease after 20 days without water, while in the forest site the maximum ETR values remain unchanged in the individuals of both populations during the same period (^{Rosado-Calderón et al. 2018}).

In the epiphytic orchids E. nematocaulon, C. yucatanensis, and L. rubescens, the maximum foliar Φ_PSII is high (0.8 on average) during the rainy and early dry seasons in a tropical dry deciduous forest of Yucatán, but it strongly decreases (Φ_PSII = 0.46) in the three species during drought, indicating possible photoinhibition because it coincides with the increase in the amount of light and the decrease in foliar acidity; in contrast, also in Yucatán, the maximum foliar Φ_PSII of E. nematocaulon, C. ascendens and L. oerstedii remains relatively high (0.7 on average) during the drought and high (0.8 on average) in the rainy and early dry seasons in a semideciduous forest (^{de la Rosa-Manzano et al. 2014}). The epiphytic orchids of the low deciduous forest show the highest maximum foliar ETR values (between 18 and 10 μmol m^-2s^-1 on average) during the rainy and early dry seasons and the lowest during the dry season (between 10 and 4 μmol m^-2s^-1); in contrast, epiphytic orchids from the semideciduous forest show the lowest values of maximum ETR during the rainy season (between 7 and 11 μmol m^-2s^-1 on average) and the highest during the early dry season (between 33 and 13 μmol m^-2s^-1 on average) (^{de la Rosa-Manzano et al. 2014}). The increase in foliage density and cloudiness during the rainy season causes a decrease in the amount of light throughout the day and with it, the energy available to capture CO₂.

The epiphytic orchids Prosthechea cochleata and Stanhopea tigrina from a cloud forest show low foliar Φ_PSII predawn values (< 0.6) under 20, 50, and 70 % of the total daily radiation under experimental conditions but are remarkably lower (< 0.1) in S. tigrina in the highest light treatment (^{Guevara-Pérez et al. 2019}). Also, plants of P. cochleata have slightly higher Y predawn values (0.63) under 50 % than those under 20 and 70 % total daily light (Y predawn = 0.46 and 0.57, respectively). No differences in ETR_max (14 μmol m^-2s^-1) are observed for P. cochleata at the three PPF treatments under drought conditions, while S. tigrina has a significant decrease in ETR_max (< 10 μmol m^-2s^-1) at the 70 % of total daily PPF compared to the other light treatments (^{Guevara-Pérez et al. 2019}).

Non-photochemical energy dissipation (NPQ). Excess light causes photoinhibition due to the production of toxic chemical species and damage to molecules of the light reactions of the photosynthetic apparatus (^{Taiz et al. 2015}). Furthermore, NPQ is a mechanism through which leaves avoid damage by dissipating excess energy as heat. In various environments, photoinhibition of photosynthesis has been observed in various CAM species because the process of energy dissipation with the xanthophyll cycle is a competitor of the photosynthetic process (^{Adams III & Demmig-Adams 1996}). However, in epiphytic plants the production of antioxidant compounds and flavonoids in response to high light and other stressful conditions are photoprotectors that prevent photodamage and cause low NPQ values (^{Saito & Harborne 1983}, ^{González-Salvatierra et al. 2010}).

The CAM epiphytes T. makoyana, T. rothii and T. eistetteri that inhabit the upper canopy stratum in a tropical dry forest with high daily PPF show lower NPQ values (1.5-1.8 on average) compared to the epiphytes T. ionantha and T. intermedia (NPQ = 2.4-3.2 on average) preferring the lower canopy stratum with a more shaded environment under experimental conditions of drought, re-wetting and high dew exposure (^{Reyes-García & Griffiths 2009}). Individuals of the epiphyte T. utriculata from a coastal scrub population maintain relatively high NPQ values (around 1.5 on average) under experimental conditions of irrigation or drought and with high and low PPF, while individuals from a forest population had higher values (2.4 on average) with irrigation and high PPF and showed values between 1 and 1.5 with drought and high or low PPF conditions (^{Rosado-Calderón et al. 2018}). During the rainy and dry seasons, NPQ values in T. brachycaulos are consistently low (< 1) in a shaded environment; however, in an exposed environment during the rainy season, NPQ values increase close to 2 during the morning, decreasing to low levels during the day in a tropical dry deciduous forest (^{González-Salvatierra et al. 2021}). Plants of the epiphytic cactus R. baccifera that grow with constant irrigation or after one month without irrigation increase two-fold the NPQ (about 2.8) when exposed to 65 % compared to 20 % of total PPF (^{de la Rosa & Briones 2008}).

Photoprotective pigments. Flavonoid production has been reported in several species of Bromeliaceae as a response to high PPF and other stressful conditions (^{Saito & Harborne 1983}, ^{Benzing 2000}). For T. brachycaulos, anthocyanins (a type of flavonoid) are in a single layer under the epidermis on both leaf faces, and anthocyanin concentrations are higher in exposed plants during the dry season, compared to shaded plants (^{González-Salvatierra et al. 2010}). Also, there is a strong correlation between incident daily high light and total anthocyanin content, which suggests that these molecules are involved in photoprotection. Epiphytic orchids show mechanisms to avoid excess light and heat, including a reduction in leaf area to increase heat flux by conduction and convection, carotenoid production, and photosystem II heat dissipation of absorbed energy (^{Adams III & Demmig-Adams 1996}, ^{de la Rosa-Manzano et al. 2014}, ²⁰¹⁵). These epiphytic orchids produce more carotenoids during the dry season than during the wet and early dry seasons, and their concentrations are higher in leaves of orchids from a tropical dry deciduous forest than those from a tropical semi-deciduous forest (^{de la Rosa-Manzano et al. 2015}). Moreover, orchid leaves from the former forest have higher zeaxanthin retention and lower values of maximum quantum efficiency of photosystem II during the dry season than orchids from the latter forest (^{de la Rosa-Manzano et al. 2015}). These changes are related to high incident light and the prolonged dry period that occur in the tropical dry deciduous forests when all trees shed their leaves.

Water relations of terrestrial CAM plants

Specialized tissues. Leaves or stems of terrestrial CAM plants have succulent tissues with specialized functions. The outer cortex, or palisade cortex, is characterized by multiple layers of green photosynthetic palisade parenchyma cells and the inner cortex is a water storage parenchyma (^{Barcikowski & Nobel 1984}, ^{Goldstein et al. 1991a}, ^{Terrazas Salgado & Mauseth 2002}, ^{Nobel 2003}). These succulent tissues, especially the water storage parenchyma, undergo successive cycles or filling and emptying and, together with the high-water use efficiency of the CAM cycle, allow continuous CO₂ uptake, since water from storage tissues maintains turgor in the photosynthetic tissues (^{Goldstein et al. 1991a}, ^{Lüttge 2004}, ^{Andrade et al. 2009}).

Cell walls of the water storage parenchyma in Opuntia ficus-indica are thinner (0.16 ± 0.01 mm) than those of the chlorenchyma (0.32 ± 0.02 mm), and unlignified (^{Goldstein et al. 1991a}), which allows a high flexibility for changing in volume without a notable change in water potential (^{Goldstein et al. 1991a}, ^{Terrazas Salgado & Mauseth 2002}). Moreover, in well-watered plants of this species, the bulk modulus of elasticity is about 0.35 MPa for the water storage parenchyma and 0.85 MPa for the chlorenchyma, which changes to 0.97 MPa under drought, indicating rigid cell walls in the latter tissue (^{Goldstein et al. 1991a}).

Water redistribution in tissues. Water storage in the inner cortex allows succulent plants to survive drought and, in CAM terrestrial plants, to continue having photosynthesis through CAM. In O. ficus-indica the lower bulk modulus of elasticity of the water storage parenchyma than that of the chlorenchyma allows the former tissue to maintain turgor and facilitate continuous water movement to the latter. Uprooted CAM terrestrial plants usually survive for years, even losing more than 80 % of the leaf or stem water (^{Barcikowski & Nobel 1984}, ^{Goldstein et al. 1991a}, ^{Andrade et al. 2009}).

The water storage capability of leaves and stems of CAM terrestrial plants can be quantified by the volume-to-surface area ratio (V/A). Leaves of agaves and stems of cacti have a mean V/A of about 1 cm and stems of barrel and columnar cacti can have V/A higher than 5 cm (^{Nobel 1994}, ²⁰⁰³). For seedlings of terrestrial CAM plants, the V/A can indicate survival. For instance, a small cactus of about 0.18 of V/A can tolerate a drought of 3.5 months or more if the seedling is under a nurse plant (^{Nobel 1994}). Similarly, seedlings of Mammillaria gaumeri in a coastal sand dune have a V/A two-fold higher when growing at 20 % of ambient PPF (under a nurse plant) than those receiving 50 % of ambient PPF (^{Cervera et al. 2006}).

Developing organs require water from mature organs. For this to occur, water moves, via xylem, from sites with high water potentials to sites with low water potentials. However, in O. ficus-indica, fruits have a higher water potential than underlying cladodes and water movement to fruits must occur via phloem (^{Nobel et al. 1994}).

Mucilage and relative capacitance. Mucilage, which is a complex polysaccharide with a high water-binding capacity, is common in many terrestrial CAM plants (^{Sáenz-Hernández et al. 2002}). In O. ficus-indica, mucilage content is higher in the water storage parenchyma than in the chlorenchyma (^{Goldstein et al. 1991a}). Additionally, in other cactus species, mucilage content is positively related to tissue relative capacitance (^{Nobel et al. 1992}). Tissue relative capacitance is the capacity of the tissue to maintain its water potentials when the water content decreases, and it is related to succulence, V/A, and other water relation properties; this helps desert CAM plants because they face dry periods of several months (^{Andrade et al. 2009}).

Changes in the water potential components. For O. ficus-indica, under well-watered conditions, acidity increases in the chlorenchyma account for about a 0.28 MPa decrease in osmotic potential in the morning, but osmotic potential in the water storage parenchyma do not change significantly in a 24-h period (^{Goldstein et al. 1991b}). After a 3-month drought, although there is a loss of water in the tissues (27 % in the chlorenchyma and 61 % in the parenchyma), osmotic potential does not change significantly (^{Goldstein et al. 1991b}). Diel changes in turgor pressure in the chlorenchyma were similar for well-watered than for droughted O. ficus-indica plants, indicating that this species maintains high water turgor in the photosynthetic tissue during extended periods of drought (^{Goldstein et al. 1991b}). Also, to compensate for the decrease in turgor pressure in the water storage parenchyma during drought, polymerization of solutes occurs (^{Barcikowski & Nobel 1984}, ^{Goldstein et al. 1991a}).

A drought of more than 30 d for individuals of O. rastrera show no change in their tissue osmotic potential (^{Briones et al. 1998}). However, seedlings of O. streptacantha have a decrease in osmotic potential as drought progresses at high light, because of their low V/A, compared to those in the shade (^{Delgado-Sánchez et al. 2013b}). In other study, seedlings of Agave salmiana, with higher V/A than seedlings of A. striata, do not change water potential after a drought of 6 months (^{Ramírez-Tobías et al. 2021}), not only for the higher succulence of the former species but also for a higher proline production in their tissues than the latter species (^{Ramírez-Tobías et al. 2014}).

Root properties. Roots of CAM plants have evolved characteristics that allow them to take water rapidly when soil is wet but reduce water loss during drought (^{Nobel 1994}). This rectifier-like property begins when roots start shrinking, leading to a root-soil air gap that reduces the hydraulic conductivity in the root-soil system (^{Nobel & Cui 1992}). After rewetting, initiation of new roots and the increase of root radial conductivity allows a rapid water uptake (^{North & Nobel 1996}). For most CAM plants, the low root-to-shoot ratio reduces carbon costs, and the efficient root water uptake, coupled to the high stem or leaves water use efficiency, leads to competitive advantage in the locations where they occur (^{Briones et al. 1998}, ^{Andrade et al. 2009}).

Water relations of epiphytic CAM plants

Humidity and light are the main factors that determine the abundance of epiphytic plants. In relation to light, epiphytes have been grouped as exposed (total or almost total exposure to the sun), sun (partial shade) and shade tolerant (^{Pittendrigh 1948}, ^{Benzing 1990}). With respect to humidity, epiphytic species have been grouped as poikilohydric and homohydric, mesophytic and xerophytic hygrophytes (^{Benzing 1990}). Particularly, epiphytic bromeliads have been classified into tank and atmospheric types based on the strategy of acquiring water and mineral nutrients (^{Pittendrigh 1948}). Tank-type bromeliads are mainly C₃ plants of the Bromeloideae subfamily, with succulent rosette leaves in which they store water and trichomes with the ability to absorb water and nutrients at the base of the leaf blade. Atmospheric type bromeliads are CAM and have a poorly developed or absent tank and absorbent trichomes over the entire leaf (^{Benzing 2000}, ^{Reyes-García & Griffiths 2009}). Atmospheric species from dry forests of Mexico show succulent leaves, are adapted to use water from rainfall pulses, and can maintain photosynthetic activity with low water content (^{Reyes-García et al. 2012}).

Drought tolerance. Plants resist the decrease in water availability by temporarily reducing its impact, with phenological mechanisms, or tolerating it with morphological and physiological mechanisms. CAM epiphytes can tolerate drought by preventing dehydration by maintaining cell turgor and high-water potentials by internally storing water in succulent pseudobulbs and spongy velamen in orchids (^{Benzing 1990}), succulent rhizomes in ferns (^{Hietz & Briones 1998}), succulent stem parenchyma (hydrenchyma) in cacti and succulent leaves of bromeliads, orchids, and ferns (^{Benzing 1990}, ^{Andrade & Nobel 1997}, ^{Nobel & De la Barrera 2004}, ^{de la Rosa & Briones 2008}). Tissue dehydration can also be prevented by reducing the conductance of waxy roots and leaves through diurnal closure of stomata, cuticular thickening, and increased boundary layer of air in contact with the atmosphere. However, CAM epiphytes can also tolerate drought at low water potentials, maintaining turgor by increasing cell elasticity. The accumulation of acids by CAM or solutes compatible with the cytoplasm to reduce the osmotic potential are mechanisms used by terrestrial desert species and could be advantageous for epiphytes with access to aerial soil. For instance, during the dry season, water potential values of T. brachycaulos were not significantly different between dawn and noon, but pre-dawn water potential values were lower than midday water potential ones during the rainy season in a tropical dry deciduous forest of Yucatán, Mexico (^{González-Salvatierra et al. 2021}). However, osmotic potential changes are subtle because the water obtained by epiphytes is water with high water potentials (^{Hernández-Robinson et al. 2020}).

Water absorption. Fluctuations in the availability of rainwater, mist or dew are high in the vertical gradient of the canopy and plants that inhabit it have developed a series of morphological and physiological adaptations to capture atmospheric water rapidly. Epiphytic bromeliads possess hygroscopic peltate trichomes, specialized to condense moisture and absorb liquid water through the leaf surface (^{Smith 1989}, ^{Cach-Pérez et al. 2014}). Epiphytic orchids accomplish the same thing through the velamen, a tissue consisting of a spongy epidermis with multiple layers at the roots (^{Goh & Kluge 1989}). In Bromeliaceae, leaf morphology in some tank species allow a decrease in leaf temperature, which is usually below air temperature in the early morning, promoting dew condensation, contributing to maintain a high-water balance (^{Andrade 2003}, ^{Reyes-García et al. 2012}, ^{Chávez-Sahagún et al. 2019}). Also, nebulophyte morphology, characterized by long and thin leaves with a small boundary layer, allow for the small fog droplets to be intercepted (^{Martorell & Ezcurra 2002}, ²⁰⁰⁷, ^{Reyes-García et al. 2012}).

Air humidity benefits epiphyte growth. In the Yucatan Peninsula in Mexico, individual density of epiphytic species is inversely related to vapor pressure deficit (VPD, ^{Cach-Pérez et al. 2014}). Additionally, in locations where trees tap the permanent water table, tree leaf area and transpiration increase, lowering VPD values, which create low-VPD canopy islands of high epiphyte density (^{Chilpa-Galván et al. 2013}). It has been reported that after drought epiphytes recover physiological processes fast (^{Andrade & Nobel 1996}, ¹⁹⁹⁷, ^{Andrade et al. 2009}). Roots of the epiphytic cacti Epiphyllum phyllanthus and R. baccifera have sheaths composed of soil particles, root hairs and mucilage, which reduce water loss and helps taking advantage of episodic rainfalls (^{North & Nobel 1994}).

Water storage. Many epiphytic CAM plants can store water in parenchymal tissues located in the roots, stems, and leaves (^{Benzing 1990}). For instance, in El Cielo Biosphere Reserve, Tamaulipas, Mexico, the most important organ for storing water in the epiphytic CAM orchids Prosthechea cochleata and Stanophea tigrina is the pseudobulb, compared to roots and leaves; the former species, which grows in dry and humid habitats, shows greater succulence in pseudobulbs and leaves and is less sensitive to drought compared to S. tigrina, which only colonizes humid places (^{Guevara-Pérez et al. 2019}). Another example is the stems of R. baccifera, which can withstand one month without irrigation with a little change in succulence (^{Andrade & Nobel 1997}, ^{de la Rosa & Briones 2008}). This allows this species and another epiphytic cactus, Epipyllum phyllanthus, to occupy more exposed sites in the canopy than other epiphytes in a dry forest in Panama (^{Andrade & Nobel 1997}).

Relative capacitance (change in relative water content per unit change in water potential) of epiphytic cacti is high (0.50 MPa^-1) compared to that of epiphytic ferns (0.16 MPa^-1), which allows to maintain transpiration longer for the cacti than for the ferns (^{Andrade & Nobel 1997}). In general, relative capacitance for epiphytic CAM species is in average about half of that of terrestrial CAM species (^{Andrade et al. 2009}).

Internal transport of water. Elastic cell walls possibly allow T. ionantha, a Mesoamerican atmospheric bromeliad that grows in xeric and mesic tropical environments in Mexico, to transport the water stored in the hydrenchyma (spongy tissue of the parenchyma that stores water) to maintain turgidity and high water potential in the leaf chlorenchyma and thus be able to capture or internally recycle CO₂ and remain metabolically active during experimental drought (^{Nowak & Martin 1997}). Similarly, plants of T. brachycaulos, an atmospheric CAM, have the highest capacitance and bulk modulus of elasticity values in the wet season, compared to the dry season in a tropical dry deciduous forest (^{Hernández-Robinson et al. 2020}). Also, the considerable size of the pseudobulb of the orchid Laelia rubescens and the presence of a specialized hypodermis to store water in the orchids Cohniella yucatanensis and C. ascendens, allow the transport of water to the foliar mesophyll to keep the relative water content constant to carry out photosynthesis during all seasons in a tropical dry deciduous forest of Yucatán (^{de la Rosa-Manzano et al. 2014}).

Potential ecophysiological responses of CAM plants to global change

Global environmental change includes both systemic changes that operate globally through the major systems of the geosphere-biosphere, and cumulative changes that represent the global accumulation of localized changes (^{Turner II et al. 1990}). The way CAM plants will cope in the face of global environmental change is crucial for their survivorship. Temperatures, carbon dioxide, and other trace gases are up at an unusual rate (^{IPCC in press}). In addition, habitat disturbance and urbanization, with its expanding and destructive human footprint, continues to affect through and impact all ecosystems (^{Schmidt et al. 2020}; ^{Shen et al. 2020}). Climatic extremes like droughts, frosts, hurricane-associated floods, and warming are now quite common (^{Altwegg et al. 2017}; ^{Montiel-González et al. 2021}). These are World’s ‘new normal’ and plant ecophysiology provides and will provide some of the most critical evidence and fundamental understanding about how the diversity of plant adaptations (including CAM species) will enable plants to operate the conditions of the Anthropocene (^{Lambers & Oliveira 2019}). However, studies on physiological responses of CAM plants to the environment are far from being complete. Most studies have been done in few species from few families. Environmental physiology of growth and reproduction of CAM plants, as well as of seedlings following germination should be given more attention (^{Andrade et al. 2009}).

Desert CAM plants are expected to expand their distribution ranges because of climate change, as these species can quickly adapt to water shortage and high temperatures (^{Aragón-Gastélum et al. 2014}, ²⁰²⁰, ²⁰²¹, ^{González-Salvatierra & Flores 2019}). However, some studies have indicated the negative effects of induced warming on the survival rate of desert CAM plants during early life cycles (^{Aragón-Gastélum et al. 2017}; ^{Pérez-Noyola et al. 2020}, which may prevent the persistence of their populations in future climate scenarios. It appears that seedling tolerance of warming may increase with age in CAM plants (^{Aragón-Gastélum et al. 2017}), although further studies are needed to evaluate this.

CAM plants are excellent plant biosystems for biophysical studies and many of their physiological processes can be modeled. The environmental productivity index has already been analyzed for many CAM species and can be used to model plant productivity in response to global climate change (^{Nobel 2003}, ^{Andrade et al. 2009}). Also, to feed more robust models, it would be necessary to study multiple stress factors for CAM plants and their plasticity to environmental changes to help predicting plant responses to future climate change (^{de la Rosa-Manzano et al. 2017}, ^{Cach-Pérez et al. 2018}). Such information would be vital for conservation, and restoration programs.

Paradoxically, the first CAM species studied in Mexico was Clusia lundellii (^{Tinoco-Ojanguren & Vázquez-Yanes 1983}) and no more significant studies have been done with this remarkable genus with different growth forms and high physiological plasticity (^{Vargas-Soto et al. 2009}). Under climate change scenarios, Clusia C₃-CAM and CAM species could become more ecologically important, and can even been proposed for restoration programs, but more physiological and ecological studies are needed.

The future of epiphytes is uncertain in this changing world (^{Zotz 2016}, ^{Zotz et al. 2021}), new research that highlights the potential of these species as ornamentals or biological indicators is extremely important. Epiphytes are probably among the first life forms to respond to global change because some bromeliad species are very sensitive to environmental changes (^{Cach-Pérez et al. 2014}). If the favorable wet growing season becomes too short, the epiphyte plant cannot have enough reservoirs to survive a prolonged drought when photosynthesis drops and photoinhibition is induced (^{Reyes-García & Griffiths 2009}). In contrast, some of the epiphytic species that have high drought tolerance can be in disadvantage under wetter conditions. The most dramatic concern for Mexican CAM epiphytes is the change of land use of the tropical dry deciduous forests. For hundreds of years these forests have been deforested worldwide and transformed to agricultural lands and towns (^{Portillo-Quintero et al. 2015}, ^{Siyum 2020}).

Predicted temperature rise due to global change can be the main problem for the physiology of CAM plants. This is because the optimal mean nocturnal temperature for CAM terrestrial plants is relatively low (15 °C, but between 20-25 °C for tropical CAM epiphytes and hemiepiphytes) (^{Nobel 2003}, ^{Nobel & Bobich 2002}). However, high diurnal temperatures have little influence in the growth and survival of CAM plants. It seems that daytime air and leaf or stem temperatures are not critical for CO₂ uptake and terrestrial CAM plants can tolerate temperatures as high as 68 °C (^{Nobel 2003}, ^{Nobel & Bobich, 2002}). However, tropical CAM terrestrial and hemiepiphytes cannot tolerate temperatures above 50 °C (^{Nobel et al. 2002}, ^{Cervera et al. 2006}). Certainly, we will see many CAM species migrating to the northern and southern hemispheres. More field and laboratory studies are required to understand the effect of temperature in the photosynthesis, growth, and survival of CAM species.