Published alkenone p records spanning known glacial pCO 2 cycles show considerably less variability than predicted by the diffusive model for cellular carbon acquisition and isotope fractionation. We suggest this pattern is consistent with a systematic cellular enhancement of the carbon supply to photosynthesis via carbon concentrating mechanisms under the case of carbon limitation during low pCO 2 glacial time periods, an effect also manifest under carbon limitation in experimental cultures of coccolithophores as well as diatoms. While the low-amplitude p signal over glacial pCO 2 cycles has led some to question the reliability of p for reconstructing long-term pCO 2 , the [CO 2 ] aq in the tropical oceans during glacial pCO 2 minima represents the most extreme low CO 2 conditions likely experienced by phytoplankton in the Cenozoic, and the strongest upregulation of carbon concentrating mechanisms. Using a statistical multilinear regression model, we quantitatively parse out the factors (namely light, growth rate, and [CO 2 ] aq ), that contribute to variation in p in alkenone-producing algae, which confirms a much smaller dependence of p on [CO 2 ] aq in the low [CO 2 ] aq range, than inferred from the hyperbolic form of the diffusive model. Application of the new statistical model to two published tropical p records spanning the late Neogene produces much more dynamic pCO 2 estimates than the conventional diffusive model and reveals a significant pCO 2 decline over the last 15 Ma, which is broadly consistent with recent results from boron isotopes of foraminifera. The stable isotopic fractionation between coccolith calcite and seawater dissolved inorganic carbon (here ∆ coccolith-DIC ) also shows systematic variations over glacial-interglacial cycles which may, following future experimental constraints, help estimate the degree of upregulation of parts of the algal carbon concentrating mechanism over glacial cycles.This simplified formulation clarifies the dependence of b on variation in the cellular C content and surface area, which scale with cell size; as well as variation in the growth rate and the effective permeability to CO 2 . When the effects of these factors are considered in aggregate, e.g. by empirical derivations of b from photic zone or culture samples, it must be remembered that the covariation and relative weight of each of these factors spatially in the modern ocean, or in culture experiments, may differ from past temporal significance and covariation of these factors. In practice, however, most previous work has interpreted variation in b to reflect either changes only in the growth rate parameter (Bidigare et al., 1997;Seki et al., 2010), or over long timescales also changes in the cell size and consequently in /S (Henderiks and Pagani, 2008;Seki et al., 2010). Potential variations in P have not been evaluated for glacial samples or the full range of published experiments with p determinations in experimental culture, although some previous studies have acknowledged that the b...