Grassland expansion as an instrument of hydrologic change in Neogene western North America

- We present 16 δO18 records which show increases of ∼2-6‰ since the Middle Miocene.
- We developed a water vapor transport model to simulate δO18 under different vegetation scenarios.
- Grassland expansion can account for most of the change in Neogene δO18 records.
- Grasslands differ in ET flux, rooting depth and seasonality from prior vegetation, changing δOp18.
- Grasses may not only respond to climate trends, but drive hydrologic change to aid their expansion.

The evapotranspiration (ET) flux accounts for approximately two thirds of terrestrial precipitation worldwide, and in grassland regions ET is equivalent in magnitude to precipitation. Regional contributions to the terrestrial hydrologic budget, however, have been far from constant in the past as distribution of vegetation changed dramatically. The rise of grass-dominated ecosystems is one of the most profound paleoecological changes in the Cenozoic. Why then, would grassland expansion not feature prominently in the record of Neogene hydrologic change? Despite numerous stable isotope paleoenvironmental studies in Neogene North America, the contributions of land cover change have been largely ignored. We present a compilation of 16 oxygen isotope studies of pedogenic carbonate and smectite from western North America, including 4 new records. Nearly all records from California, the Basin and Range, the Rocky Mountains and the Great Plains show increases in δO18 on the order of 2-6‰. In order to assess the role of ET in the hydrologic cycle, we developed an isotopic water vapor transport model wherein we manipulated ET parameters along a specified air mass trajectory. Grasslands lead to δO18 of precipitation (δOp18) values that are up to 5‰ greater than broadleaf and needleleaf vegetation at inland study sites. These results demonstrate that changes in vegetation played a critical role in establishing the modern hydrologic regime in western North America. We suggest that this isotopic increase is due to three primary reasons: 1) Increased evaporation and transpiration fluxes in grassland regions affect water balance, 2) Shallower rooting depths of grasses lead to the transpiration of soil water enriched in O18 due to evaporation, and 3) Grasslands transpire O18-rich waters during a shorter, more punctuated growing season. We argue that the observed isotope signals are indicative of a feedback mechanism wherein grasslands not only respond to regional and global climatic trends, but also act as a driver of hydrologic change. By enhancing seasonality and aridity, grasslands transmit hydrologic disturbances downstream, engineering climatic conditions favorable for their expansion.