The carbon-cycle significance of soil carbonate fluxes is subject to large
uncertainties because it is not clear precisely how much calcium is derived
from atmospheric sources compared with that from the chemical weathering o
f silicate minerals. In the petrocalcic horizon (calcrete) of a Pleistocene
soil from the USDA-SCS Desert Project area near Las Cruces, NM, approximat
ely 1.5 g Ca/cm(3) has been added, with an associated expansion of the prof
ile of similar to 200%. Strontium isotope values for the labile cations and
carbonate from the A, B and K soil horizons have Sr-87/Sr-86 values that r
ange From 0.7087 to 0.7093, similar to the values for easily soluble local
dust and rain. The parent material, non-calcareous Camp Rice alluvial sedim
ent, has a Sr-87/Sr-86 ratio of similar to 0.7165. Mixing calculations indi
cate a minimum atmospheric contribution to soil carbonate calcium of simila
r to 94%; the more likely scenarios indicate at least 98% of the Ca origina
ted from atmospheric input. The variations in 87Sr/86Sr ratios of soil sili
cate (0.7131 to 0.7173) are consistent with weathering of volcanogenic sedi
ments and neoformation of clay minerals in the petrocalcic horizon. Moreove
r, the Sr isotope data suggest that 50-70% of silicate in the uppermost 25
cm of the profile could be atmosphere-derived. The isotopic composition of
labile strontium in the A horizon and the mass distribution of silicon and
calcium indicate that the uppermost portion of the profile is the present z
one for the release of cations due to silicate weathering. Steady-state mod
els of the whole profile yield a Sr weathering flux ranging from similar to
200 to 400 mu g cm(-2) Ma(-1). The results indicate that both the present-
day and long-term contribution of calcium from silicate weathering is less
than 2% of that supplied from the atmosphere, and confirm that desert soil
formation is not a significant sink for atmospheric carbon. (C) 1999 Elsevi
er Science B.V. All rights reserved.