Spatiotemporal Variation in the Environmental Controls of C4-Grass Origin and Ecology: Insights from Grass-Pollen δ13C Data

Abstract:

Understanding the environmental factors controlling the origin and shifting abundance of C₄ grasses in Earth’s history is useful for projecting the response of C₄-grass dominated grasslands to future environmental change. Unfortunately, grass pollen is typically morphologically indistinct, making palynological analysis a blunt tool for studying C₄-grasses in the paleorecord. δ¹³C of individual grass-pollen grains using a spooling wire microcombustion device interfaced with an isotope ratio mass spectrometer (Single Pollen Isotope Ratio AnaLysis, SPIRAL) overcomes this challenge and the potential biases of δ¹³C data from other substrates (e.g. leaf waxes). To assess the presence and relative abundance of C₃- and C₄-grass pollen in samples of unknown composition, we developed a hierarchical Bayesian model, trained with ~1,900 δ¹³C values from pollen grains of 31 grass species. Surface-sediment data from Africa, Australia, and North America demonstrate the reliability of this technique for quantifying C₄-grass abundance on the landscape. To investigate the timing and control of the origin of C₄-grasses we analyzed samples from the Oligocene-Miocene from Europe and from the Eocene from North America. Results indicate that C₄ grasses appeared on the landscape of southwest Europe no later than the early Oligocene, implying that low atmospheric pCO₂ may not have been the main driver and/or precondition for the development of C₄ photosynthesis in the grass family. In contrast, we found no evidence for C₄ grasses in the southeast United States before pCO₂ fell. In application of SPIRAL to the late Quaternary, we found that shifts in pCO₂ and moisture balance exerted key controls on the relative abundance of C₃ and C₄ grasses in Africa and Australia. Overall, our results imply that as in the past, future changes in the C₃/C₄ composition of grass-dominated ecosystems will likely exhibit striking spatiotemporal variability as a result of differing combinations of environmental controls.