Basal travertine of the Bouse Formation: Geochemistry, diagenesis and implications for the integration of the Colorado River

The Pliocene Bouse Formation is discontinuously exposed in the lower Colorado River region and is a record of the first arrival of the Colorado River to the Gulf of California 5 million years ago. It consists broadly of a lower carbonate member (travertine, marl, and bioclastic units) and an upper siliciclastic member (claystone, mudstone, and Colorado River sands). This paper focuses on the basal travertine (synonymous with “tufa”) unit of the lower carbonate member. Because of its basal position and its chemical encrustation of pre-Bouse topography, the travertine can offer insight into the earliest depositional settings and may be a proxy for the composition of the waters that deposited the first Bouse carbonates. Hence the travertine unit, if it can be shown to preserve a primary geochemical signal, offers the potential to discriminate between alternative hypotheses for marine versus non-marine deposition of the Bouse carbonates of the Blythe Basin. This paper examines the geochemistry of the travertine unit using stable isotopes of carbon and oxygen, 87Sr/86Sr, petrographic examinations of thin sections, and microprobe traverses. Testing for diagenesis included subsampling techniques and textural studies using thin section examinations and SEM investigations. The travertine unit forms an encrustation that drapes and mantles pre-Bouse topography, including volcanic bedrock and fanglomerates. The travertine unit is generally thin, often less than several meters thick although it can reach thicknesses of tens of meters. It is intermittent but fairly widespread in the Blythe Basin, the southernmost of the Bouse basins, and also is present in scattered locations in the more northern basins. Its facies include: porous tufa, microbialite domes (bioherms), vegetation-casts (charophytes of marsh and probable non-marine origin), and botryoidal travertine, all onlapped by and interfingered with marl and high energy bedforms of v bioclastic sandstone that were deposited before the first arriving Colorado River sands. This Walther’s Law relationship suggests that the travertines are broadly coeval with the other facies in the basal carbonate unit of the Bouse Formation. Stable isotope data for travertine reveal a covariation of δ13C with δ18O and a spread of values between (+4,+2‰) and (-16, -9‰) for the southern Blythe Basin and a regression line with R2 = 0.63. Northern basin travertines have a similar covariation trend (R2 = 0.73). Travertines show multiple carbonate generations in thin section, but stable isotope analyses did not show continuous or regular differences in composition of subsamples. Silica diagenesis was observed in the Buzzard’s Peak area where the 4.834 Ma Lawlor Peak tuff is interbedded with carbonates, but this area showed overlapping carbonate chemistry to other areas, although somewhat more positive along the regression line. Compiled and new 87Sr/86Sr analyses show that the basal Bouse carbonates have non-marine values of ~ 0.711 (as opposed to 0.709 for seawater) in all carbonate facies (marls, bioclastic unit, travertine, and numerous fossil types). Double dissolution tests for 87Sr/86Sr values were performed in travertine and marl to evaluate potential diagenetic changes: these revealed little change in values (from 0.71051 to 0.71081; from 0.71056 to 0.71074; and from 0.71088 to 0.71088). Plots of δ18O versus 87Sr/86Sr and versus latitude show no covariation in 87Sr/86Sr over a wide range of δ18O and facies types. The combined data are interpreted as showing only limited carbonate diagenesis within the basal travertine of the Bouse Formation such that carbonate geochemistry can be used as a proxy for the waters that deposited them. Two possibilities are examined to explain the covariation of δ18O with δ13C: 1) mixing of sea water (0, 0) and meteoric water (-16, -7); or 2) evaporation of basin water. We favor evaporation as the dominant explanation based on the similarity of travertine variation in northern (lacustrine) and southern (debated marine versus lacustrine) basins, the presence of non-marine charophytes in travertines, and absence of covariation between δ18O and 87Sr/86Sr, consistent with evaporation but not mixing. Radiogenic 87Sr/86Sr and the presence of localized zones of large volumes of travertine suggest influences from deeply circulated geothermal groundwaters. We