Having gained some understanding of the consequences of the CO2-concentrating mechanisms in crassulacean acid metabolism (CAM) that internalize the photosynthetic environment of the Cretaceous on a daily basis, it may be time to consider potential long-term effects of the planetary CO2-concentrating mechanism on growth and ecology of these plants in the Anthropocene. This paper emphasizes our limited understanding of the carbohydrate economy of CAM in relation to growth processes and briefly reviews recent studies of the diel cycles of growth in these plants. An inadvertent long-term, regional-scale experiment from the past is revisited in which an Opuntia monoculture grew to occupy >25 million hectares of farmland in central eastern Australia, producing a total biomass of about 1.5 billion tonnes in about 80 years. Although at the time it does not seem to have been recognized that this invasion involved CAM, a botanist from the University of Melbourne, Jean White-Haney emerges as a heroic pioneer in the control of the invader by poison and pioneered its biological control. The Opuntia population was expanding at 10–100 ha h−1 when it was brought to a halt within a decade by the voracious appetite of Cactoblastis cactorum larvae. It is now known that the female parent moth of this predator detects CAM in O. stricta prior to oviposition by deploying the most sensitive CO2 detector system yet found in the Lepidoptera. The O. stricta invasion is a dramatic demonstration of the capacity of CAM plants to attain and sustain high biomass; to sequester and retain atmospheric CO2. In conclusion, experiments are reviewed that show stimulation of CO2 assimilation, growth, and biomass of CAM plants by elevated atmospheric [CO2], and the proposition that these plants may have a role in atmospheric CO2 sequestration is re-examined. This role may be compromised by predators such as Cactoblastis. However the moth CO2 sensors are adapted to pre-industrial atmospheric [CO2] and FACE (free-air CO2 enrichment) experiments show this exquisite system of biological control is also compromised by rising global [CO2] in the Anthropocene.
As with most things in the human experience, early simplifications tend to accumulate complexities over time. So it is with our understanding of curious things to do with the C4 and crassulacean acid metabolism (CAM) pathways of photosynthesis. It has been known for about 40 years that both these pathways of photosynthetic carbon metabolism use analogous biochemical reactions to drive CO2-concentrating mechanisms, that each day effectively recreate the high [CO2] of the planetary atmosphere of the Cretaceous inside cells and tissues of leaves in air. The leaf-level biological CO2-concentrating mechanisms depend to a large extent on virtually closed cycles of carboxylation and decarboxylation in adjacent cells (C4), or in the same cells following stomatal closure in the light (CAM), and are mediated by intermediary pools of malic and or aspartic acids. As a consequence, the efficiencies of light utilization, carbon assimilation, and water use are increased in C4 and CAM plants, especially at higher temperatures.
The elevated CO2 concentrations achieved internally largely mitigate the inevitable inefficiencies of CO2 fixation by Rubisco in the presence of 21% O2 that lead to futile recycling of CO2 in photorespiration. The CO2-concentrating mechanism of CAM is distinct in that the C3-like Rubisco of Kalanchoë (Badger et al., 1975) functions each day with [CO2] ranging from 108 ppm in phase IV (Maxwell et al., 1997) to >3000 ppm in phase III, with [O2] from 21% to 41%, respectively (Spalding et al., 1979). Estimates of the ratio of carboxylation to oxygenation in phase III of CAM range from 7.7 to 27.4, much higher than those of spinach (2.7), mitigating photorespiration in phase III (Osmond et al., 1999), but not in phase IV. More work is required, but Griffiths (2006) hinted that CAM Rubisco is perhaps an exception to the general rule that ‘all Rubiscos may be nearly perfectly adapted to the differing CO2, O2 and thermal conditions in their sub-cellular environments’ (Tcherkez et al., 2006).
Our appreciation of the complexity of these CO2-concentrating mechanisms in the functional biodiversity of photosynthesis has become more sophisticated over time (cf. Hatch et al., 1971; Sage and Monson, 1999). For example, it is now clear that the broad analogies between C4 and CAM are almost certainly incompatible at cellular and tissue levels, and have evolved separately in rather diverse taxa (Sage, 2002). Moreover, one of the cardinal distinctions between the pathways, the spatial (C4) versus temporal (CAM) separation of carboxylation and decarboxylation events, now has been sacrificed on the altar of evidence for exotic single-cell C4 plants from central Asia (Edwards et al., 2004). By the same token, studies of functional biodiversity of CAM in the tropical Americas and creative experiments in European laboratories (Lüttge, 2004) suggest that another 30-year-old simplification of CAM physiology and biochemistry into four discrete phases (Osmond, 1978) has become much more ‘fuzzy’. What at first had seemed to be a closely coordinated yin–yang relationship between acid metabolism in the dark and recovery of carbohydrate reserves in the light for dark CO2 fixation the next night (Osmond, 1984) is now obviously much more plastic, even fantastic (Dodd et al., 2002).
Importantly, the contexts in which these CO2-concentrating mechanisms function, and the selective advantages they confer on C4 and CAM plants in natural habitats, may also be changing rapidly in the current era of Earth's history, the Anthropocene (Crutzen and Stoermer, 2000). The Anthropocene refers to the 200 or so years since the Industrial Revolution in which humankind has inadvertently created another CO2-concentrating mechanism of planetary proportions. Release of CO2 from the combustion of fossil photosynthates, at rates about a million times faster than rates of deposition over geological time, greatly exceeds the present capacity for CO2 fixation/sequestration on Earth. As a consequence [CO2] in the essentially closed atmosphere of the planet is increasing far more rapidly than ever before in the history of human occupation of Earth, and is likely to approach that achieved internally by the CO2-concentrating mechanisms in leaves of C4 and CAM plants.
Over the last 40 years it has come to be appreciated that the global CO2-concentrating mechanism brings important issues associated with global warming and changed continental patterns of precipitation that are already having large effects on species distribution and function in the plant biosphere (Walther et al., 2002; Root et al., 2003). Photosynthesis is a primary transducer of atmospheric chemistry, and, because of the properties of Rubisco, photosynthetic sequestration is a first responder to elevated [CO2]. On the one hand, we need to consider whether the selective advantages of the leaf-level CO2-concentrating mechanisms of C4 and CAM over C3 photosynthesis are being eroded by elevated atmospheric [CO2]. On the other, we need to evaluate the potential of photosynthetic systems of all kinds for biological CO2 sequestration. In the latter context, Park Nobel has often made the point that CAM ecosystems have considerable potential as low-input CO2 sequestration systems in arid environments (Nobel, 1991; Drennan and Nobel, 2000). At bottom, the potential for CAM systems to accumulate high biomass depends on the capacity of these plants to resolve the conflict of interest presented by the need to partition carbohydrates between nocturnal acid metabolism and growth (Borland and Dodd, 2002).
Although our understanding of the signature acid metabolism of CAM is increasingly (and intriguingly) anchored in molecular regulatory cascades, our knowledge of the regulation of starch and sugar metabolism in CAM, and especially its relationship to growth, largely remains to be elucidated (Holtum et al., 2005b). So among other things, this review highlights a few key issues of carbohydrate metabolism and growth in CAM plants in relation to some aspects of the effects of elevated [CO2] on CAM reported by Nobel (Nobel, 1991; Drennan and Nobel, 2000). It illustrates the potential for terrestrial CAM plants to play a role in CO2 sequestration in seasonally arid environments by reviewing another inadvertent experiment, low-input, long-term (80 years) in the Antipodes in which a CAM plant monoculture expanded to occupy 25×106 ha with sustained high biomass (up to 500 tonnes ha−1). Although plant scientists of the time do not seem to have understood that they were dealing with CAM, the experiment was terminated by purposeful introduction of an insect predator that knew precisely how to detect CAM activity. It is now known that the sustained biological control of this CAM invasion by females of Cactoblastis cactorum depends on an exquisitely sensitive biological CO2 detector that is deployed to detect nocturnal CO2 fixation and target CAM plants. The global CO2-concentrating mechanism of the Anthropocene not only potentially compromises the selective advantages of CAM; it also compromises the sensitivity of the CO2 sensors in moth predators that control the expression of these advantages in ecosystems.
When, in the diel cycle, do CAM plants grow?
Although in most cases the diel carbon economy of CAM plants involves CO2 uptake from the atmosphere in phase I at night and in phase IV in the light (Osmond, 1978), very little is known about when they grow and how they manage the basic ‘conflict of interest’ (Borland and Dodd, 2002) of partitioning reserve carbohydrates between those needed for nocturnal acidification and those needed for growth. The observation in Deléens and Garnier-Dardart (1977) and Deléens et al. (1979) that the δ13C value of malic acid and starch is markedly less negative than that of sugar and cellulose in leaves of the CAM plant Kalanchoe diagremontiana, and in leaves of Kalanchoe blossfeldiana grown in short days to maximize CAM activity, continues to challenge our understanding of carbohydrate metabolism in CAM. The clear implication for chemically isolated pools of carbohydrates committed to nocturnal acidification, and to growth, remains to be evaluated. Borland (1996) and Borland and Dodd (2002) have developed modelling approaches to these problems that may ultimately help resolve the conundrum of Deléens.
Most C3 plants grow at night when it seems likely that growth is mainly driven by mobilization of chloroplast starch reserves (Walter and Schurr, 2005). Given that CAM plants perform many other physiological functions in a temporally perverse manner, one might anticipate that CAM plants would grow during the day. Indeed, Gouws et al. (2005) found that growth of cladodes of Opuntia spp. (and leaves of some other CAM plants) peaked during phase III (Fig. 1) when carbon skeletons are conserved in chloroplast starch for the next night of malic acid synthesis, and when 25% of malate carbon could become available for growth and maintenance. Moreover, because turgor also peaks late in phase I, and because cytoplasmic pH becomes more acid in phase III, superficially this phase of CAM seems the most likely to favour expansion growth of leaves and cladodes.
Against the background of a relatively thin understanding of carbohydrate metabolism in mature, fully expanded, non-growing leaves and cladodes, this view of the carbon economy of CAM plant growth is an oversimplification. For example, the comprehensive analysis of Wang et al. (1998) emphasizes that growth of young Opuntia ficus-indica cladodes is sink dominated for at least 14 d, with an abrupt sink–source transition after 25–28 d, by which time the CAM activity of the daughter cladodes showed a 5-fold increase in nocturnal malate accumulation compared with day 14. The cladodes examined in the authors’ growth experiments may have been sinks; the fact that their δ13C values indicated CAM does not necessarily imply that the carbon was derived from in situ CO2 fixation.
Although it was earlier thought that δ13C values can be used with some precision to assess contributions of the CO2 fixation pathway (Osmond et al., 1976) and source–sink relationships (Deléens and Garnier-Dardart, 1977; Deléens et al., 1979) in K. daigremontiana, subsequent studies have emphasized the role of diffusion limitations, both stomatal and mesophyll, in determining δ13C (Winter and Holtum, 2002; Holtum et al., 2005a; Griffiths et al., 2007). Clearly, δ13C values are very blunt instruments in bulky CAM plants. Controlled environment experiments with well-watered cladodes of O. stricta kept in continuous light at 23 °C, or in day–night cycles (17 °C night; 14 h/23 °C; 10 h day) for up to 877 d showed only very slow changes of δ13C values in new cladodes (Fig. 2). The role of circadian control of gas exchange in these plants is unknown, but it seems unlikely that PEPcase would continue to fix CO2 in continuous light for 2–3 years. As the parent cladode persisted in good health throughout the experiment, it seems equally unlikely that the fifth new cladode formed in continuous light was largely derived from pre-existing carbon. The failure to obtain a signal from the fifth new cladode for CO2 fixation due exclusively to Rubisco was probably due to low stomatal conductance and high internal diffusion resistance (Farquhar et al., 1982). Intercellular CO2 concentrations decline to 108 ppm in K. daigremontiana during Rubisco-dominated CO2 fixation in phase IV of CAM (Maxwell et al., 1997), close to that estimated for C3 juniper leaves exposed to CO2 starvation in the last glacial period (Ward et al., 2005). In an interesting more recent development, English et al. (2007) have found that chronosquences of δ13C and δ18O values in spines from long-lived columnar Carnegiea gigantea (Englemann) Britton and Rose (saguaro) provide a sensitive record of precipitation, climate change, and growth responses in the desert in south-western USA, on scales of decades to centuries in the Anthropocene.
For these and other reasons it was thought that the effects of ontogeny and environment on the expression of CAM in Clusia spp. (Lüttge, 2007) might reveal further clues as to the diel relationships between CO2 assimilation and growth. Although it has been found that diel growth cycles change with water stress-induced CAM, and with development of CAM during ontogeny, in these plants (Walter et al., 2008), it is not yet possible to relate them directly to changes in metabolism. Mature, non-growing leaves of the obligate CAM species C. alata (Popp et al., 1987) and the facultative CAM species C. minor used in the authors’ experiments use sucrose rather than starch for carbon skeletons for acid accumulation (Borland et al., 1994). Thus the carbohydrate economies of young growing leaves of Clusia spp. are likely to be a mix of sink–source relationships for growth and a much more complex carbohydrate–acid metabolism (malate and citrate) relationship. It is well known that many CAM plants, such as pineapple, depend almost solely on soluble sugars stored in the vacuole for nocturnal acid synthesis and these must have very different regulatory mechanisms to manage the ‘conflict of interest’. Studies of the diel growth cycle of pineapple leaves might be more revealing. Clearly, there is a plethora of problems in CAM plant growth processes to be researched, from molecular and kinetic aspects of regulated carbohydrate metabolism, to physiological and ecological analysis of growth itself.
An inadvertent, regional-scale experiment in CAM plant productivity
In the process of European colonization, Australia was assaulted by a huge array of exotic animals and plants. These included many seeds that inadvertently came aboard ships in South Atlantic ports and later became major weeds (particularly from southern Africa), as well as purposeful introductions of exotic plants (especially succulents) and animals (rabbits) that later became serious pests. After the rabbit, perhaps the most aggressive and destructive import was a platyopuntia, the common ‘pest pear’ or prickly pear. Having separated from the African and American land masses prior to the expansion of the Cactaceae, the island continent has only one endemic arid-zone succulent shrub with CAM (Sarcostemma australe R.Br, Asclepiadaceae) but evidently presented a fertile landscape for an aggressive immigrant photosynthetic pathway. Currently known as Opuntia stricta Haw., Australian prickly pear has an uncertain taxonomy (Alexander, 1925). Initially thought to be O. vulgaris Miller (by von Mueller and Bailey), then to be O. inermis P. DC. (by Maiden), the latter is evidently also synonymous with O. bentonii Griffiths and O. stricta (by Britton and Rose). Its introduction is also clouded in uncertainty, but it seems to have been cultivated at Parramatta, NSW before 1840 when it is known to have been transported and successfully established at Scone in the Hunter Valley of NSW and near Warwick in Queensland.
Opuntia stricta was by far the most aggressive of 18 Opuntia spp. established in Australian ecosystems, nine of which attained pest proportions (Alexander, 1925). Having briefly lent ambience to early vineyards in south-eastern Australia, prickly pear moved on to make a monumental impact on the livelihood of thousands and on the landscape as a whole. From a few plants in 1840, rough estimates indicate it progressively expanded to occupy 4×106 ha by 1900 (www.nrw.qld.gov.au), 6.3×106 in 1913 (newspaper article; Fig. 3), 17.6×106 in the early 1920s (Alexander, 1925), and most agree that it had spread to 24–30×106 ha before it fell to biological control in 1930 (Mann, 1970). This essentially logarithmic expansion of territory (R2=0.995), 50% of which was rendered useless for farming, was simply astonishing (Fig. 4). In an era in which colonials in the Antipodes tended to calibrate many things in ‘mother country’ terms, within 80 years O. stricta spread from Scone and Warwick and had occupied an area larger than that of the UK (21.3×106 ha).
Although websites now assert that ‘The perfect climate and lack of natural enemies accounted for its amazing spread – still considered by many experts to be one of the botanical wonders of the world’ (http://www.northwestweeds.nsw.gov.au/common_pear.htm), the prickly pear invasion is poorly documented in the plant literature and there are few accurate records from the 1920s of the astonishing growth of this CAM monoculture. Indeed, we believe that, at the time, there was little appreciation by Australian botanists of the distinctive ecophysiology of Opuntia. Estimates vary, but the overall rate of expansion in central-eastern Australia reached about 50 ha h−1 in 1925 and O. stricta attained biomass of >300 tonnes dry weight ha−1 over large areas (Fig. 4). The plight of landowners denied access to an area greater than the whole area of crop cultivation in Australia at the time was highlighted by Osmond and Monro (1981), and Freeman (1992) has presented a more detailed analysis of the socio-economic impact and the somewhat uncoordinated official responses to the invasion. In a graph illustrating the growth and collapse of prickly pear Freeman (1992) used an average biomass of 620 tonnes ha−1 (possibly fresh weight?) to estimate a total O. stricta biomass of 1.5×109 tonnes in 1929.
Prickly pear occupied open and abandoned farmland, invaded open scrubland and forest, especially in the unpredictably, but frequently, wet habitats of central-eastern Australia that correspond approximately to those having ‘non-effective rainfall for the growth of agricultural crops’ (Leeper, 1960; Osmond et al., 1979). The invader found an open niche in an unpredictable summer rainfall environment similar to that dominated by platyopuntias in the Americas (Nobel, 1988). In retrospect, the plant succeeded in its adopted habitat partly because it is a CAM plant with exceptional water-conserving potential, partly because of its extraordinarily low root to shoot ratio dominated by above-ground cladode biomass that is fully invested in photosynthetic surface, but largely because of its extraordinary vegetative and sexual reproductive activities. Depending on the stimulus provided, on environmental conditions, and on position, meristems in the areolae (axillary buds) of cladodes have the capability to become new cladodes, to become a flower, to become spines that facilitate dispersal of whole cladodes, or to form roots on contact with moist surfaces and thereby quickly rehydrate cladodes after months of drought (Gibson and Nobel, 1986). Its succulent fruits with many seeds are particularly attractive to passerine birds so it is not uncommon to find plants developing to maturity in rocky outcrops, upon wooden fence posts, or along wire fence lines.
A blind spot in the war on prickly pear
To this day, we have been unable to find any reference by Australian plant biologists to the distinctive metabolism of O. stricta until the late 1960s. Perhaps CAM was recognized but not considered to be relevant. In the prevailing climate of concern for eradication of the invader, it may have been unseemly and almost certainly ‘politically incorrect’ for expert plant biologists to display an overt interest in the success of ‘the pest pear’. Certainly newspaper cartoons of the time suggest this may have been so (Fig. 3). However, de Saussaure's discovery (de Saussaure, 1804) of nocturnal O2 uptake that greatly exceeded CO2 release in Opuntia sp. should have been well known. Heyne's acid taste-test of Bryophyllum calycinum was published in 1815 and, although the reciprocal acid synthesis and starch metabolism data from German scientists (e.g. Kraus, 1884) may not have been very accessible, Richards’ (1915) treatise on diel acidification in O. versicolor and other species of cacti from the Carnegie Institution's Desert Botanical Laboratory in Tucson was widely circulated (Fig. 5). As pointed out previously (Black and Osmond, 2003), the importance of nocturnal opening and diurnal closure of stomata in CAM, especially in relation to water conservation, was only recognized much later (Nishida, 1963), even though resistance porometry had been available since Darwin and Pertz (1911).
Tom Neales was the first to investigate CAM in Opuntia in Australia and from the early 1970s he featured prickly pear as an exemplary CAM plant in his classes. He had earlier contrasted the gas exchange of CAM (pineapple) with other plants (Neales et al., 1968) and investigated the diurnal gas exchange patterns in Agave americana after observing the resilience of this plant in the Melbourne drought of 1968 (Neales, 1970, 1973). As a student in Armidale and Adelaide in the early 1960s, Barry Osmond was not widely aware that CAM had conquered a UK's worth of central eastern Australia. Although Danny Avadahni had kindled Barry's interest in CAM on a visit to Adelaide in 1964, the penny did not drop (Osmond, 2006). A small US–Australia workshop on photosynthetic carbon metabolism in Canberra (Hatch et al., 1971) brought many aspects of C3, C4, and CAM into focus for both of us. The famous assertion (no longer strictly valid) that C4 plants were simply ‘CAM mit Kranz’ (Laetsch, 1971) still provokes arguments today, and may have triggered our subsequent interest in CAM ecophysiology.
Neales (1975) reported that O. stricta displayed all the gas exchange phases of a ‘full-CAM’ plant and, at the same meeting, Osmond (1975) showed a map that suggested more C3-like δ13C values for prickly pear in more mesic habitats. Tom's student Hong-Huei Lai examined gas exchange (Fig. 6), acidification, and cladode water potential in O. stricta. He found negligible gas exchange day or night after long-term water stress, but nocturnal CO2 fixation was detected within 2 d of watering, and some CO2 fixation in the light was detectable in well-watered plants (Lai, 1977). The speculation that success of prickly pear as a CAM plant might be associated with frequent additional CO2 fixation by Rubisco in phase IV gained some currency from these studies. A subsequent 3-year field study of O. stricta using rather primitive equipment confirmed that, although CO2 fixation in the light was frequently detectable following rainfall, further δ13C analyses of O. stricta kept in continuous light for several years (Fig. 2) cast doubt on the suggestion that phase IV made a large contribution to the carbon economy of the cactus (Osmond et al., 1979). As discussed above, it is now suspected that the isotope signature of Rubisco in phase IV was masked by low stomatal and internal diffusion resistances. So it seems safe to conclude now that stomatal opening in the light (White-Haney, 1915) within a few days of watering (Lai, 1977) and rapid engagement of ‘full-CAM’ following rainfall (Neales, 1975; Osmond et al., 1979) contributed to the success of O. stricta in central eastern Australia.
With phase IV CO2 fixation (about 30 μmol cm−2 d−1) augmenting nocturnal CO2 fixation (to 130 μmol malate cm−2 d−1), and the recycling of this carbon by ‘CAM-idling’ (at 40 μmol malate cm−2 d−1) between infrequent rainfall events, we have plausible explanations for the accumulation and retention of very high CAM biomass in these communities. It was estimated that CO2 assimilation day and night could sustain cladode relative growth rates of 0.043 g g−1 dry weight d−1 at best, similar to a few measures of growth rates (0.05 g g−1 dry weight d−1) under optimal field conditions. It was also estimated that a clump of about 25 cladodes m−2 land surface, with the potential of doubling each year could soon expand to a monoculture (Fig. 4) with a growth rate of some 30–40 g dry weight m−2 land surface per day under optimal conditions. If these conditions prevailed for about 100 d, the productivity of O. stricta is within sight of the modelled maximum productivity of CAM (Nobel, 1991) and comparable with the measured productivity of 38–47 tonnes ha−1 year−1 for CAM plants under intensive cultivation. Clearly, some careful field studies of biomass and cladode area development in moth-proofed O. stricta populations are needed to validate these speculations.
Chemical and biological CAMpaigns against prickly pear
Interestingly, a botanist was among the first to approach the prickly pear invasion from a scientific perspective, but, understandably, the emphasis was on eradication rather than exploration of the reasons for the success of the invader. Jean White-Haney (Fig. 7) was the second woman to be awarded a Doctor of Science degree from the Botany Department of the University of Melbourne in 1909 (Clifford, 2002; Neales, 2007). In 1912 she set up tents in the remote Queensland town of Dulacca, near the centre of the invasion (Fig. 4), and proceeded to establish a field station that was charged with finding the most efficacious ‘poisonous specific’ for control of prickly pear. She found that spraying the cladodes was more effective than applying poison to the root zone, and concluded from a single spraying at 8 pm that ‘the amount of injury done to the plants by the treatment during the day is far more apparent than in those poisoned at night’. This she attributed to stomata, the supposed ports of entry for the poison, and camera-lucida drawings in her reports show them open in the day (White-Haney, 1915). In her last large-scale poisoning experiment at the Explosive Magazine Reserve, Begool, January 1915 ‘7.2 tons of pear was sprayed using 3625 gallons of spray containing 896 lbs (approx) of crude arsenic acid’. Having exhausted her supplies of arsenic pentoxide in the Begool experiment, and facing manpower problems as her field assistants left for the cliffs of Gallipoli and the fields of Flanders, the station was closed in 1916.
Between the war to end all wars and the next one, the eradication of prickly pear slowly grew to assume high priority in eastern Australia. The Commonwealth Prickly Pear Board (recommended by Jean White-Haney in 1916; established December 1919) was the first scientific project jointly funded by the national (Commonwealth) and state governments (New South Wales and Queensland). Early chemical campaigns were based on White-Haney's experiments and on experience in the trenches. Gas attacks with arsenic fume generators were not wildly successful but hand to hand combat with bayonet devices such as the patented Propert Auto-flow Stabber and spraying assaults on Opuntia cladodes, both using arsenic pentoxide 15–20% in sulphuric acid (s.g. 1.7), were effective on isolated patches of the pest. Evidently unaware of the vegetative reproductive biology of Opuntia, Captain A. Pentland, MC, DFC, RAF, who had ‘seen a great deal of the working of tanks’ (http://www.naa.gov.au/education/ontour/dogs/caterpillar.html), recommended their use to crush the cactus. He was prepared to return to Australia upon demobilization and take up land to demonstrate his proposal. The National Australian Archive preserves records of other militarily inspired recommendations, including the use of that devilish German device, the ‘Flammenwerfer’, which was rejected because of the high cost of gasoline. Even so, the scale of the poison campaign prior to biological control (Fig. 8) was such that the price paid for arsenic pentoxide declined from about $150 to $65 per tonne 1924–29, only to rise again (to about $110) in 1932 (Osmond and Monro, 1981).
Jean White-Haney was also acutely aware of the potential for biological control and carried out the first experiments with cochineal insects from India and Ceylon, finding them very effective against the larger ‘tree pear’, O. monocantha Willd. Mann (1970) believed her early successes with cochineal later ‘gave force to the movement to pursue the biological control effort’. In spite of early government indifference (Freeman, 1992), Royal Commissions set up by State and Federal authorities subsequently led to intensive biological control campaigns, ranging from bounties on the heads and eggs of emus (Australian coat of arms notwithstanding) that were blamed for mid-range dispersal of Opuntia seeds (Fig. 8), to evaluation of >150 cactus predators imported from around the world. By far the most successful of these, the female moth Cactoblastis cactorum Berg, is now recognized as possibly the most industrious CAM researcher on the planet (Stange et al., 1995).
In 1924, Alan Dodd, a young entomologist sent by the Commonwealth Prickly Pear Board to collect predators of Opuntia spp. in the southern USA, took the initiative to collect in South America during the northern hemisphere winter. At Concordia in NE Argentina he collected the female moth of C. cactorum whose larvae fed on many species of Opuntia in Uruguay and Argentina. Dodd's collections arrived safely in 1925 and the rest is history. It was a case of second time lucky because the moth had been introduced to Australia previously in 1914 but perished in the hands of staff in the Brisbane Botanic Gardens.
This time the introduction was successful and when the voracious appetite of the larvae for prickly pear was recognized, tens of millions of eggs of the predator were collected from moths in cages under the supervision of entomologists. These were widely distributed by the authorities and over a billion were provided free to landholders. In the field 80 000 eggs could produce 6 million cocoons and >22 million eggs in the next generation (Dodd, 1940), and there are two generations per year! Cactoblastis larvae carried out a regional-scale mining operation analogous to humankind's insatiable appetite for fossil fuels in the Anthropocene. About a century's worth of CO2 sequestration in prickly pear biomass (estimated at 1.5 billion tonnes; Freeman, 1992) was consumed by trillions of larvae (Fig. 8) and returned to the atmosphere in less than a decade. It was said that one could stand in a patch of prickly pear and hear them eating! Photographs of prickly pear infestations before and after the introduction of Cactoblastis are dramatic (Fig. 9).
Within a decade, abandoned farming and pastoral land equal in area to the UK was returned to production (Dodd, 1940; Osmond and Monro, 1981; Freeman, 1992). Erection of Memorial Halls, remembering the fallen of WW I, was a feature of Australian country towns in the inter-war years. The quintessential Cactoblastis Memorial Hall, erected in 1936 by the grateful citizens of Boonarga, Queensland, is the only public building in Australia erected in memory of an insect, and, so far as is known, the only building on the planet dedicated to the memory of trillions of female CAM researchers. It is about to acquire heritage status (http://www.heritage.gov.au/protect-places/boongara.htm).
Boundary layer detection of CAM by female Cactoblastis cactorum
Although it seems clear that those concerned with the early war on prickly pear were little interested in its distinctive water-efficient nocturnal CO2 assimilation, O. stricta was quickly brought to its knees by a female moth predator that is uniquely endowed to identify CAM in these succulents. In retrospect it seems clear that, although saturation of the ecosystem with eggs of the predator was itself effective (Myers et al., 1981), female Cactoblastis were, and remain, discriminating with respect to oviposition. Even though Cactoblastis species differ in their preferences for different populations of Opuntia (Mann, 1970; Osmond and Monro, 1981) they do not oviposit on other plants in the ecosystem.
Gert Stange discovered that this is because the large labial palps in female Cactoblastis (these are poorly developed in males) are richly endowed with possibly the most sensitive CO2 detectors of all insects examined to date (Fig. 10). The female moths deploy these detectors during a brief dance above the boundary layer (a few millimetres) on any surface upon which they alight during their low flying perambulations at dusk (Stange et al., 1995). The sensors are easily able to detect an inward gradient of a few parts per million CO2 and thereby distinguish the early evening CO2 sink activity of an O. stricta cladode in a landscape that is otherwise dominated by CO2 sources at this time. Oviposition follows on the most uncomfortable, but most secure, site on the cladode (barbed glochidia in the areolae). Recent studies show that Cactoblastis can also sense volatiles extracted from O. stricta (Pophof et al., 2005) and that these may be important also in long-distance targeting of the cactus.
While plant biologists of the day were debating whether a specimen was, or was not O. stricta, female Cactoblastis went about selecting the CAM plants in the landscape on the basis of boundary layer CO2 fluxes each evening, and her offspring larvae brought the problem plant to heel (http://www.northwestweeds.nsw.gov.au/images/Cacto). Some elements of the population biology of the well-sustained predator–prey relationship, including host plant food quality for the larvae, have been described by John Monro (Osmond and Monro, 1981). In this respect, Rubisco in O. stricta hardly provides a rich source of protein (cladodes contain 4–10% crude protein), but moth preference for greener cladodes with more nitrogen is probably mediated by detection of more active CO2 fixation, effectively taking out the most aggressive members of the CAM plant population (Myers et al., 1981). The system of biological control has been stable for over 70 years and is sustained by about a single clump of prickly pear per hectare (Osmond and Monro, 1981). After its success in Australia, Cactoblastis was introduced as a biological control agent to South Africa, to Hawai'i, and to the Bahamas, from whence it came to Florida. Having moved up the Florida peninsula, Cactoblastis is now advancing across the south-eastern US and Mexico with an indiscriminate taste for platyopuntias (Perez-Sandez, 2001), causing deep concern for those hoping to preserve CAM-dominated natural ecosystems and Opuntia horticulture.
Conclusions: CAM (and the control of prickly pear) in response to rising [CO2] and global climate change in the Anthropocene
The initially substantial stimulation of growth in C3 plants of various life-forms (and to a lesser extent in C4 plants) by increasing atmospheric [CO2] has attracted much attention from plant biologists since the 1970s. So has the attenuation of this growth stimulation by resource limitations and stress factors such as nutrition, water, salinity, and temperature. We are also increasingly aware that ecosystem level factors can further attenuate the growth and biomass response of plants to elevated [CO2]. For example, although the coppiced Populus deltoides forest mesocosms in the Biosphere 2 Laboratory grown at elevated [CO2] for four seasons showed stimulation of CO2 assimilation, growth, and biomass above ground, below-ground biomass, soil carbon exudation, soil microbial biomass, soil respiration, and soil nutrient depletion were also stimulated (Barron-Gafford et al., 2005). Ecosystem-level carbon fluxes were accelerated and, in this case, there was no evidence of the hoped-for increased sequestration in soil. Indeed, the metabolism of existing soil carbon was dramatically stimulated (Trueman and Gonzalez-Meler, 2005).
Do CAM plants have particular attributes that might confer a distinctive and perhaps useful role in carbon sequestration in the Anthropocene? Or, having assembled a photosynthetic pathway that re-creates the elevated atmospheric [CO2] of the Cretaceous internally behind closed stomata, and having thereby achieved remarkable water use efficiency on a daily basis, is global change now likely to render the selective advantages of CAM irrelevant? We have seen that in the absence of predators, an invasive population of a bulky CAM plant can build very high biomass over a century with very low input. Like their cultivated CAM cousins (and C3 or C4 crops), they can achieve high, near theoretical maximum annual productivity (Nobel, 1991), but over areas in which precipitation is inadequate, or unreliable, and evaporation so great that rainfall is ineffective for crop growth. Key attributes of bulky CAM plants that contribute to their capacity to attain such high biomass include:
(i) precipitation-responsive superficial root systems and low root to shoot ratios (∼0.1) that minimize below-ground diversion of carbon for structure, maintenance, and secretion of carbon to soils;
(ii) above-ground biomass entirely vested in photosynthetic surfaces;
(iii) water-efficient nocturnal CO2 uptake, augmented by up to 20–40% diurnal CO2 assimilation in the light following rainfall;
(iv) total recycling of above-ground respiratory CO2 (stomatal closure in light; nocturnal CO2 fixation);
(v) water conservation by stomatal closure associated with high water capacitance (Comins and Farquhar 1982);
(vi) thermal tolerance limits from –20 °C to 67 °C (Nobel, 1988);
(vii) efficient reproduction, both vegetatively and by seed dispersal.
It is probably too early to predict with confidence the changes that will be wrought in regional climate systems as rising atmospheric [CO2] drives global warming. In general terms, however, higher temperatures and more stochastic extreme precipitation and drought events may be expected in many presently important agricultural regions. The future would seem to portend an expansion of territory suitable for sustainable occupation by large CAM plants.
Rising atmospheric [CO2] is the most certain outcome of fossil fuel consumption in the Anthropocene, and Drennan and Nobel (2000) cite >20 reports of the effects of elevated [CO2] on most aspects of CO2 assimilation and growth of bulky CAM plants from work done in the UCLA laboratories and the field. Unexpectedly perhaps, in long-term experiments (1–17 months) with well-watered CAM plants, a doubling in atmospheric [CO2] stimulated total CO2 uptake by an average 31% for six large cacti, and stimulated growth and biomass by 33%. The physiological attributes listed above evidently conspire to effect a relatively direct conversion of assimilated carbon to biomass.
These responses were unexpected because elevated [CO2] was not anticipated to stimulate CO2 assimilation in the presence of closed stomata in the light, and because it was anticipated that CO2 assimilation in the dark by PEP carboxylase would be saturated at internal [CO2] likely to prevail in the present atmosphere. Nobel and colleagues found that part of the response could be attributed to increased afternoon CO2 fixation by Rubisco (phase IV). Stimulation of dark CO2 fixation in phase I was also commonly observed, presumably because both carbohydrate substrates for PEP carboxylase increased (due to increased Rubisco activity in phase IV) and because PEP carboxylase access to CO2 in phase I is diffusion limited. These responses pretty much accord with what is now known of the diffusion limitations to CO2 fixation in all phases of CAM (Maxwell et al., 1997; Rascher et al., 2001; Nelson et al., 2005; Griffiths et al., 2007). However, as the atmosphere of the Anthropocene continues to be enriched with fossil fuel-derived CO2, our ability to identify CAM plants on the basis of δ13C values and to evaluate the contributions of CO2 assimilation in the dark and light (Winter and Holtum, 2002), seems likely to be further compromised. In any event, opportunities for exploiting succulents to understand physical constraints on carbon isotope signals better remains attractive.
Returning to our earlier focus on the importance of further work on the carbohydrate economy of CAM plants, Wang and Nobel (1996) found that growth of O. ficus-indica for 3 months in elevated [CO2] showed little evidence for down-regulation of photosynthesis of the sort commonly found in herbaceous plants and usually ascribed to sink limitations and feedback effects of sugar on CO2 assimilation and gene expression. Instead they found higher CO2 assimilation (source capacity), greater sucrose transport in the phloem, and stronger sink strength. Having dealt comprehensively with the diversity of morphological and other responses of CAM plants to elevated [CO2] at the organismal level, Drennan and Nobel (2000) conclude that high biomass CAM communities offer potential as a low-input system for atmospheric CO2 sequestration in arid habitats. Moreover, the closed model ecosystem constructed by Rascher et al. (2006) showed that CAM can assimilate and retain a detectable and significant part of nocturnal respiratory CO2 efflux from arid soils. Although much further research is needed, we should not rule out the possibility that long-lived CAM plants in arid ecosystems may present effective regional carbon sequestration systems on time scales of decades to centuries.
Clearly, the regional-scale prickly pear experiment illustrates how pathogens and predators could compromise carbon sequestration in monoculture succulent ecosystems. On the other hand, the CAM discovery capacity of Cactoblastis will almost certainly be attenuated by rising atmospheric [CO2]. Stange and Wong (1993) pointed out that the CO2-sensing system of the moth Helicoverpa was thermally compensated to function most effectively at pre-industrial [CO2], and was predicted to lose sensitivity with increasing background [CO2]. Gert Stange proceeded to put this to the test with Cactoblastis and Opuntia in a series of mini-FACE (free-air CO2 enrichment) ring experiments (Fig. 11) with astonishing results. Oviposition on O. stricta in a fluctuating elevated [CO2] was reduced 3.2-fold, and under a steady 2-fold higher [CO2], oviposition was reduced 1.8-fold and the egg-sticks deposited were smaller (Stange, 1997). Much the same response was found in the Datura–Manduca system (Abrell et al., 2005).
The loss in sensitivity involves many factors, and provides only a limited answer to the question as to what effect the rise in [CO2] will have on the interaction between O. stricta and C. cactorum. In a future global scenario, where CO2 is elevated permanently, such long-term effects will add extra complexity. However, the combined effects on sensory performance more than cancel the increased signal strength caused by the response of the plant to elevated [CO2], and it would appear that this is sufficient to account for the observed reduction of plant attractiveness in doubled CO2 (Stange, 1997). It is overstating the case that increasing atmospheric [CO2] may release O. stricta from the present strangle-hold of female Cactoblastis over a 25×106 ha of central eastern Australia and promote accelerated growth of prickly pear, or that it might save the platyopuntias of the south-eastern US from their attention (Perez-Sandi, 2001). However, these studies underline some of the system level responses that need to be taken into consideration when we contemplate a role for CAM in carbon sequestration in the Anthropocene.
The authors are grateful to Professor Howard Griffiths for the invitation to present this overview paper and to many colleagues for their stimulus to our research in CAM over the past 40 years.