Reconciling continuous and discrete models of C4 and CAM evolution

Abstract

Background: A current argument in the CAM biology literature has focused on the nature of the CAM evolutionary trajectory: whether there is a smooth continuum of phenotypes between plants with C3 and CAM photosynthesis or whether there are discrete steps of phenotypic evolutionary change such as has been modelled for the evolution of C4 photosynthesis. A further implication is that a smooth continuum would increase the evolvability of CAM, whereas discrete changes would make the evolutionary transition from C3 to CAM more difficult.

Scope: In this essay, I attempt to reconcile these two viewpoints, because I think in many ways this is a false dichotomy that is constraining progress in understanding how both CAM and C4 evolved. In reality, the phenotypic space connecting C3 species and strong CAM/C4 species is both a continuum of variably expressed quantitative traits and yet also contains certain combinations of traits that we are able to identify as discrete, recognizable phenotypes. In this sense, the evolutionary mechanics of CAM origination are no different from those of C4 photosynthesis, nor from the evolution of any other complex trait assemblage.

Conclusions: To make progress, we must embrace the concept of discrete phenotypic phases of CAM evolution, because their delineation will force us to articulate what aspects of phenotypic variation we think are significant. There are some current phenotypic gaps that are limiting our ability to build a complete CAM evolutionary model: the first is how a rudimentary CAM biochemical cycle becomes established, and the second is how the 'accessory' CAM cycle in C3+CAM plants is recruited into a primary metabolism. The connections to the C3 phenotype we are looking for are potentially found in the behaviour of C3 plants when undergoing physiological stress - behaviour that, strangely enough, remains essentially unexplored in this context.

Keywords: Phenotype; character evolution; continuous; discrete; heuristic.

Fig. 1.

Earliest depictions of C₄ and CAM evolutionary models, modified from Peisker (1986) and Teeri (1982). (A) Peisker’s model of C₄ evolution. Peisker proposed nine steps to transition between C₃ and C₄ photosynthesis. Here, the lines represent how these nine steps influenced ¹³C/¹²C isotope ratios (continuous line) and CO₂ compensation concentrations (dashed line), with the values at stage 1 being C₃ values and the values at stage 10 being C₄ values. Peisker modelled variation in the order in which the nine steps occurred, which would lead to different phenotypic trajectories (labelled II, I and III) that eventually evolved towards similar endpoints. Although he described these changes as discrete ‘phases’, all nine proposed changes refer to shifts in continually varying characters, such as the relative proportion of Rubisco in the bundle sheath or mesophyll, the capacity for phosphoenolpyruvate regeneration, and the increase in phosphoenolpyruvate carboxylase activity. (B) Teeri’s model of CAM evolution. Teeri proposed that the first step of CAM evolution is to evolve CAM-cycling, which is re-fixation of respired CO₂ at night by the CAM cycle (labelled as ‘recycling present’ in the diagram). In the centre of the diagram, he depicts the ‘upregulation continuum’, evolving from 100 % C₃ fixation to 100 % CAM fixation. The upregulation continuum has persisted as the primary CAM model in the literature, although in the text Teeri references discrete stages of CAM evolution, explaining that the double arrows in the middle indicate that there might be ‘multiple evolutionary steps’ between the phenotypes. He envisions another phenotype that is 100 % flexible between C₃ and CAM pathways, modelled on Kalanchoe blossfeldiana, that he assumes must be an optimal phenotype in all conditions, because there are no evolutionary pathways away from it, only towards it. Teeri does envision reversals from the other main CAM types back to C₃. The continuous arrows are hypothesized transitions supported with evidence from Crassulaceae; dashed arrows are hypothesized transitions that had not been documented at the time. Peisker and Teeri both explicitly modelled continuous characters but described the transitions as discrete evolutionary stages.

Fig. 2.

Discretizing quantitative characters to delineate stages of an evolutionary trajectory. (A) Hypothetical distribution of trait values for a character that is naturally multi-modal. In this case, it is straightforward to bin the standing variation into ‘low’, ‘medium’ and ‘high’ values. Dots below the hypothetical distribution curve are real data from Flaveria for the CO₂ compensation point, demonstrating a clear phenotypic gap between C₃ and C₃–C₄ species, less so between C₃–C₄ and C₄. (B) Discretization can also be justified when the distribution is not multi-modal, perhaps with other analyses that support threshold values where a particular value of one trait influences the evolvability of another. In this case, the bundle sheath-to-mesophyll (BS:M) ratio of leaves might show a relatively flat distribution, but phylogenetic modelling studies demonstrate that a BS:M ratio of 0.15 is required for C₄ evolution (although the BS:M ratio of many C₄ plants is much higher than that number). Dots below the hypothetical distribution curve are real data from Flaveria for the BS:M ratio, demonstrating no clear phenotypic gaps, but the C₃ species lie right at the value of the modelled threshold, with all other C₃–C₄ and C₄ species on the other ‘C₄ side’ of the threshold. We may use thresholds to discretize traits into ‘low’ and ‘high’ bins that are biologically relevant. Flaveria data are from Lyu et al. (2021).

Fig. 3.

Model of CAM evolution showing two potential trajectories from C₃ to strong CAM. Understanding which characteristics of C₃ plants facilitate the establishment of a nascent CAM cycle (‘CAM-enabled’ box) remains an open research question; likely candidates include species that maintain high tissue water potentials and operate with a conservative water-use strategy. The upper trajectory follows the SIEI model, hypothesizing that stress induction results in variation in malate accumulation, which is selected for and eventually genetically assimilated, with a full CAM cycle becoming regulated by stress-induced gene regulatory networks (GRNs). Eventually, a CAM cycle also becomes constitutively expressed, potentially via co-regulation with circadian GRNs and/or increases in succulence (see lower trajectory). The lower trajectory follows a ‘mechanistic CAM–succulence model’, whereby all plants have a latent CAM cycle such that increased fluxes may be selected for when expressed in a facilitating anatomical context. Both trajectories result in a plant with a constitutive but weakly expressed CAM cycle with moderate succulence. The evolutionary transition to strong CAM is driven by a ‘synergistic anatomical pleiotropy’, whereby further anatomical changes towards increased succulence positively influence both CAM and water-storage functions, driving convergent evolution of strong-CAM succulent life forms across the tree of life (sensuEdwards, 2019). Abbreviations: Ψ, photosynthetic tissue water potential; GRN, gene regulatory network; SAP, synergistic anatomical pleiotropy.