Mechanisms for concentrating carbon around the Rubisco enzyme, which drives the carbon-reducing steps in photosynthesis, are widespread in plants; in vascular plants they are known as crassulacean acid metabolism (CAM) and C₄ photosynthesis. CAM is common in desert succulents, tropical epiphytes, and aquatic plants and is characterized by nighttime fixation of CO2. The proximal selective factor driving the evolution of this CO2-concentrating pathway is low daytime CO2, which results from the unusual reverse stomatal behavior of terrestrial CAM species or from patterns of ambient CO2 availability for aquatic CAM species. In terrestrials the ultimate selective factor is water stress that has selected for increased water use efficiency. In aquatics the ultimate selective factor is diel fluctuations in CO2 availability for palustrine species and extreme oligotrophic conditions for lacustrine species. C4 photosynthesis is based on similar biochemistry but carboxylation steps are spatially separated in the leaf rather than temporally as in CAM. This biochemical pathway is most commonly associated with a specialized leaf anatomy known as Kranz anatomy; however, there are exceptions. The ultimate selective factor driving the evolution of this pathway is excessively high photorespiration that inhibits normal C3 photosynthesis under high light and high temperature in both terrestrial and aquatic habitats. CAM is an ancient pathway that likely has been present since the Paleozoic era in aquatic species from shallow-water palustrine habitats. While atmospheric CO2 levels have undoubtedly affected the evolution of terrestrial plant carbon-concentrating mechanisms, there is reason to believe that past atmospheric changes have not played as important a selective role in the aquatic milieu since palustrine habitats today are not generally carbon sinks, and the selective factors driving aquatic CAM are autogenic. Terrestrial CAM, in contrast, is of increasing selective value under extreme water deficits, and undoubtedly, high Mesozoic CO2 levels reduced the amount of landscape perceived by plants as water limited. Late Tertiary and Quaternary reductions in atmospheric CO2, coupled with increasing seasonality, were probably times of substantial species radiation and ecological expansion for CAM plants. C4 photosynthesis occurs in only about half as many families as CAM, and three-fourths of C4 species are either grasses or sedges. Molecular phylogenies indicate C4 is a more recent innovation than CAM and that it originated in the mid-Tertiary, 20–30 Ma, although some data support an earlier origin. While the timing of the origin of C4 remains controversial, the nearly explosive increase in C4 species is clearly documented in the late Miocene, 4–7 Ma. Increasing seasonality has been widely suggested as an important climatic stimulus for this C4 expansion. Alternatively, based on models of photosynthetic quantum yield at different temperatures and CO2 concentration, it has been hypothesized that the late Miocene C4 expansion resulted from declining atmospheric CO2 levels. This model is most appropriate for explaining the transition from C3 grasslands to C4 grasslands but by itself may not be sufficient to explain the more likely scenario of a late Miocene transition from C3 woodland/ savanna to C4 grasslands. A largely unexplored hypothesis is that climatic changes in late Miocene altered disturbance regimes, in particular the incidence of fires, which today are often associated with maintenance of C4 grasslands. Oceanic charcoal sediments that appear to represent Aeolian deposits from continental wildfires follow a strikingly similar pattern of explosive increase in late Miocene. Climate, CO2, and disturbance are not mutually exclusive explanations and probably all acted in concert to promote the expansion of C4 grasslands. More recently, late Quaternary changes in CO2 may have been responsible for driving major changes in the landscape distribution of C4 species. The theory is sound; however, many of the studies cited in support of this model are open to alternative interpretations, and none has eliminated climatic factors as important selective agents. CAM and C4 evolution required coupling of biochemical pathways with structural changes in photosynthetic tissues, succulence in CAM and Kranz in C4. This was apparently accomplished by piecemeal evolution beginning with mechanisms for recapturing respiratory CO2, although this need not have been so in aquatic CAM species. It has been proposed that the extreme rarity of both pathways in the same plant results from biochemical and structural incompatibilities (Sage 2002). Equally important is the fact that the selective environments are quite different, with CAM evolution thriving on stressful sites inhospitable to C3 species whereas C4 evolution has selected for rapid growth capable of outcompeting associated C3 plants.