The invasion of woody vegetation into deserts, grasslands and savannas is generally thought to lead to an increase in the amount of carbon stored in those ecosystems. For this reason, shrub and forest expansion (for example, into grasslands) is also suggested to be a substantial, if uncertain, component of the terrestrial carbon sink. Here we investigate woody plant invasion along a precipitation gradient (200 to 1,100 mm yr(-1)) by comparing carbon and nitrogen budgets and soil delta(13)C profiles between six pairs of adjacent grasslands, in which one of each pair was invaded by woody species 30 to 100 years ago. We found a clear negative relationship between precipitation and changes in soil organic carbon and nitrogen content when grasslands were invaded by woody vegetation, with drier sites gaining, and wetter sites losing, soil organic carbon. Losses of soil organic carbon at the wetter sites were substantial enough to offset increases in plant biomass carbon, suggesting that current land-based assessments may overestimate carbon sinks. Assessments relying on carbon stored from woody plant invasions to balance emissions may therefore be incorrect.

No abstract

No abstract

The application of stable isotopes in speleothem records requires an understanding of the extent to which speleothem calcite isotopic compositions refl ect the compositions of the cave waters from which they precipitate. To test for equilibrium precipitation, modern speleothem calcite was grown on glass plates, so that the carbon and oxygen isotope composition of the calcite and the water from which it precipitated could be directly compared. The plates were placed on the tops of three actively growing stalagmites that occupy a 1 m 2 area in Harrison's Cave, Barbados, West Indies. Only some of the plate δ 13 C values and none of the plate δ 18 O values correspond to equilibrium values, indicating signifi cant kinetic isotope effects during speleothem calcite growth. We investigate herein mechanisms that may account for the kinetic isotope effects.On each plate, speleothem calcite was deposited with distinct δ 18 O and δ 13 C compositions that increase progressively away from the growth axis, with up to 6.6‰ 13 C and 1.7‰ 18 O enrichments. The positive δ 13 C versus δ 18 O trends are likely a result of 18 O and 13 C Rayleigh-distillation enrichment in the HCO 3 reservoir owing to progressive CO 2 degassing and CaCO 3 precipitation. The magnitude of the δ 13 C versus δ 18 O slope is likely controlled by the extent to which CO 2 hydration-hydroxylation reactions buffer the oxygen isotope composition of the HCO 3 reservoir during calcite precipitation. Complete oxygen isotopic buffering of the HCO 3 reservoir by CO 2 hydration-hydroxylation reactions will produce a vertical δ 13 C versus δ 18 O slope in calcite sampled along a growth layer. As oxygen isotope buffering of the HCO 3 reservoir decreases to no buffering, the δ 13 C versus δ 18 O slope in calcite sampled along a growth layer will decrease from vertical to 0.52 at the cave temperature. In this study, modern speleothem calcite sampled along the growth layer produced a δ 13 C versus δ 18 O slope of 3.9, indicating incomplete oxygen isotope buffering of the HCO 3 reservoir during calcite precipitation.Both modern and Holocene speleothem calcite from Barbados, sampled temporally along the growth axis, shows similar positive δ 13 C versus δ 18 O slopes. These results, along with the spatial variations in glass plate calcite carbon and oxygen isotope compositions, suggest that the isotopic composition of the Holocene speleothems is in part controlled by non-equilibrium isotope effects. In addition, there is a correlation between stalactite length and oxygen and carbon isotope ratios of calcite precipitated on the corresponding stalagmite and glass plate, which may be due to 13 C and 18 O enrichment of the HCO 3 reservoir during CO 2 degassing-calcite precipitation along the overhanging stalactite.We compiled 165 published speleothem stable isotope records with a global distribution and found that most of these records show a positive covariation between δ 13 C and δ 18 O values. Speleothem stable isotope records may be infl uenced by kinetic isotope effects such...

Variations in growth rates of speleothem calcite have been hypothesized to reflect changes in a range of paleoenvironmental variables, including atmospheric temperature and precipitation, drip-water composition, and the rate of soil CO 2 delivery to the subsurface. To test these hypotheses, we quantified growth rates of modern speleothem calcite on artificial substrates and monitored concurrent environmental conditions in three caves across the Edwards Plateau in central Texas. Within each of two caves, different drip sites exhibit similar annual cycles in calcite growth rates, even though there are large differences between the mean growth rates at the sites. The growth-rate cycles inversely correlate to seasonal changes in regional air temperature outside the caves, with near-zero growth rates during the warmest summer months, and peak growth rates in fall through spring. Drip sites from caves 130 km apart exhibit similar temporal patterns in calcite growth rate, indicating a controlling mechanism on at least this distance. The seasonal variations in calcite growth rate can be accounted for by a primary control by regional temperature effects on ventilation of cave-air CO 2 concentrations and/or drip-water CO 2 contents. In contrast, site-to-site differences in the magnitude of calcite growth rates within an individual cave appear to be controlled principally by differences in drip rate. A secondary control by drip rate on the growth rate temporal variations is suggested by interannual variations. No calcite growth was observed in the third cave, which has relatively high values of and small seasonal changes in cave-air CO 2. These results indicate that growth-rate variations in ancient speleothems may serve as a paleoenvironmental proxy with seasonal resolution. By applying this approach of monitoring the modern system, speleothem growth rate and geochemical proxies for paleoenvironmental change may be evaluated and calibrated.

No abstract

Carbonate rocks and natural waters exhibit a wide range in the concentration and isotopic composition of strontium. This wide range and the quantifiable covariation of these parameters can provide diagnostic tools for understanding processes of fluid‐rock interaction. Careful consideration of the uncertainties associated with trace element partitioning, sample heterogeneity and fluid‐rock interaction mechanisms is required to advance the application of the trace element and isotope geochemistry of strontium to studies of diagenesis, goundwater evolution, ancient seawater chemistry and isotope stratigraphy. A principal uncertainty involved in the application of Sr concentration variations to carbonate systems is the large range of experimental and empirical results for trace element partitioning of Sr between mineral and solution. This variation may be a function of precipitation rate, mineral stoichiometry, crystal growth mechanism, fluid composition and temperature. Calcite and dolomite in ancient limestones commonly have significantly lower Sr concentrations (20–70 p.p.m.) than would be expected from published trace element distribution coefficient values and Sr/Ca ratios of most modern sedimentary pore waters. This discrepancy probably reflects the uncertainties associated with determining distribution coefficient values. As techniques improve for the analytical measurement and theoretical modelling of Sr concentration and isotopic variations, the petrological analysis of carbonate samples becomes increasingly important. The presence of even small percentages of non‐carbonate phases with high Rb concentrations and high 87 Sr86 Sr values, such as clay minerals, can have significant effects on the measured 87 Sr/86 Sr values of carbonate rocks, due to the decay of 87Rb to 87 Sr. For example, a Permian marine limestone with 50 p.p.m. Sr and 1 p.p.m. Rb will have a present‐day 87 Sr/86 Sr value that is >2 × 10−4 higher than its original value. This difference is an order of magnitude greater than the analytical uncertainty, and illustrates the importance of assessing the need for and accuracy of such corrections. A quantitative evaluation of the effects of water‐rock interaction on Sr concentrations and isotope compositions in carbonates strengthens the application of these geochemical tracers. Geochemical modelling that combines the use of trace elements and isotopes can be used to distinguish between different mechanisms of water‐rock interaction, including diffusive and advective transport of diagenetic constituents in meteoric pore fluids during the recrystallization of carbonate minerals. Quantitative modelling may also be used to construct diagnostic fluid‐rock interaction trends that are independent of distribution coefficient values, and to distinguish between mixing of mineral end‐members and fluid‐rock interaction.

No abstract