Isotope geochemistry is one of those sexy topics that has captured a lot of attention in the past decade (or more), and some of that attention is even deserved (oh! Burn on Isotopes!). Isotopic records have proven to preserve interesting, if sometimes occult and difficult to interpret, patterns of change from many different depositional environments and basins. Often times, these records are used in the dark arts of paleoclimatology, where it is hoped that temperature driven isotopic fractionation can be parsed out from authogenic minerals.
However, there are alternative mechanisms for getting different isotope values into a basin. Very recent work (it came out today) by Carroll et al. (2008) attempts to link isotopic changes preserved within some of the Green River lacustrine strata with evolution of the drainage system in the hinterland. Basically, these authors attribute changes in O and Sr isotopes to drainage capture of high elevation sources in the Cordilleran foreland system.
Lake Gosuite was a Big Ass Lake (BAL in the technical parlance) in the Eocene of southwestern Wyoming; the resultant stratigraphy of these deposits is actually incredibly complicated and cool, though Carroll et al. (2008) focus on the specific interval that saw the deposition of the LaClede "Bed", which are actually a group of related, carbonate rich bedsets that record alternate phases of lake expansion and contraction.
Anyway, previous work in the basin has identified a distinct stratigraphic surface, the "fill to spill surface", across which a considerable change in isotope values in preserved. The figure below is from the Carroll et al. (2008) paper, and is their Figure 2 (on page 792).
What the figure shows is a rapid decrease of delta 18 O values for this interval, dropping from ~+26 parts per mil to ~+20 parts per mil in the Upper LaClede Bed. A concomitant decrease of Sr isotopes, from 0.712296 +/- 0.0004000 to 0.711638 +/- 0.000274 is also recorded (page 792 of Carroll et al. 2008). What could have caused such a dramatic change (the authors are glad you asked)?
Diagenesis is ruled out, on the basis of a lack of textural evidence for alteration in the mudstones. Furthermore, widely dispersed outcrops record the same shifts in O and Sr isotopes. Additionally, diagenesis would be expected to have produced lower delta 18 O values than reported by Carroll et al. 2008. The authors conclude that this is not a diagenetic signature; so what is it?
One explanation is a direct change in climate in the basin, such as wetter conditions in Lake Gosiute. However, based on Carroll et al. (2008)'s mass balance model, a shift of the magnitude reported above would require a 4x increase in precip OR a 50% decrease in evap (or some combination). The problem with this idea, however, is that there is no sedimentological OR paleontological/paleobotanical evidence for such a dramatic climatic change at this time.
The authors conclude that a more probable explanation is river capture of a more northernly drainage, with differently sourced waters bringing in different isotopic values into the basin. This is supported by the fact that, simultaneous with the change in isotope values, is an increase in volcaniclastics (probably sourced from the Idaho and southwestern Montana volcanic fields) within the basin. By simply capturing a river that might be sourcing a higher elevation, the resultant system could dramatically change its isotope value without having to undergo some sort of huge climate shift.
It is an interesting read, and suggests that the record of climate, as derived from isotopes, may be complicated by more processes than previously recognized. Similarly, it makes me wonder about paleoaltimetry estimates; could those records be affected by the size and source of the drainage in the adjacent sed/water delivery system?
Carroll, A.R., Doebbert, A.C., Booth, A.L., Chamberlain, C.P., Rhoades-Carson, M.K., Smith, M.E., Johnson, C.M., and Beard, B.L., 2008, Capture of high-altitude precipitation by a low-altitude Eocene Lake, western U.S.: Geology, v. 36, p. 791-794.