COMPARISON OF PALEOPRECIPITATION ESTIMATES FROM STABLE CARBON ISOTOPE ANALYSIS OF FUSAIN AT TWO CRETACEOUS FOSSIL SITES

Stable carbon isotope analysis and foliar physiognomy are two quantitative tools used to estimate Mean Annual Precipitation (MAP). Leaves from modern floras were collected across a precipitation gradient in Texas to evaluate the effectiveness of δ13C to predict MAP compared to foliar physiognomy. Algorithms published by Kohn (2010) and Diefendorf et al. (2010) were accurate in estimating MAP for west Texas sites, but overestimated MAP for mid to east Texas sites compared to foliar physiognomy.

C-isotope precipitation proxies were applied to fossil charcoal separated from two Cretaceous localities: 1) the Xiagou Formation (XF) of northwest China, and 2) the Woodbine Formation (WF) near Arlington, Texas, U.S.A. C-isotopic compositions from the XF averaged -22.92‰ (VPDB) and ranged from -29.01 to -19.54‰. The WF averaged -24.77‰ and ranged from -25.65 to -23.54‰. Precipitation estimates were calculated under varying conditions of δ13Catm with offsets of ±2‰ to account for variation in plant tissues and external factors (e.g.,salinity). Best estimates of paleoprecipitation from the Kohn algorithm for the XF, with no offsets and an estimated δ13Catm of -3.5‰, was 590 (+1071, -484) mm/yr, and was 632 (+827, -358) mm/yr for the Diefendorf et al. global algorithm. Best estimates of paleoprecipitation using Kohn for the WF, with no offsets and an estimated δ13Catm of -4.5‰, was 1091 (+532, -379) mm/yr, and for Diefendorf et al. was 1091 (+439, -312) mm/yr. Using a -2‰ offset to account for salinity effects increases MAP to 2,864 (+1,010, -765) mm/yr for Kohn and 2,610 (+1055, -750) for Diefendorf et al. The XF is interpreted as a palustrine environment with lithologic indicators of both wet (e.g. organic rich mudstones and turtle remains) and dry conditions (e.g. soil carbonates and gilgai structures). WF estimates were drier than its Mississippi Delta-like interpretation unless an offset for salinity is utilized. Examination of existing WF paleoflora for foliar physiognomy will provide a basis for comparison. Considering the tendency for the δ13C-precipitation proxies to overestimate precipitation in wet conditions, actual precipitation for the WF may be less. Precipitation proxies still produce wide MAP estimates, and work to improve accuracy is needed, such as regional or tissue-specific algorithms.