Annual and 5-year averages of δ13C measured on florets of the C4 grass Setaria macrostachya decreased from 10 to 13‰ between 1849 and 2016. Atmosphere-plant δ13C discrimination (Δ) changed from 3 to 4 early in this interval, to 4–5 since 1970. These century-scale changes result from decrease in atmospheric δ13C and, prior to 1970, from changes in Δ related to increase in the partial pressure of CO2 in the atmosphere, both largely caused by the burning of fossil fuels. Short-term fluctuations (interannual to decadal) in slope of the time series of averaged δ13C do not result from natural causes, but in certain cases appear to coincide with human economic activities. Annual δ13C ranges in years with sufficient measurements are consistently about 2‰. It is unlikely that useful information on past atmospheric δ13C will be obtained from S. macrostachya seeds preserved in packrat middens or other archives.

A new variety of the evergreen shrub Choisya arizonica Standl. (Rutaceae) is described. It differs from the typical form in having longer and broader leaflets and thus larger leaves. The new variety C. arizonica var. amplophylla Too lin occurs in a few localities in southeastern Arizona with igneous rock substrates, rather than on soils derived from limestone substrates normally colonized by the species.

Abstract Setaria macrostachya and S. vulpiseta have been treated as distinct species by most authors since they were first described more than 180 years ago. In three recent publications, however, S. macrostachya was listed as a synonym of S. vulpiseta. The present study, which included the examination of some 350 herbarium specimens, has convinced us that these two species are as distinct as many others in the genus, and should continue to be recognized as separate taxa. Communicating Editor: Jerrold Davis

Setaria inopinata Toolin, sp. nov., is described and illustrated, and a key is provided to separate it from similar species in the genus. Ap- parently, this caespitose perennial, which resembles S. tenax, occurs only in northwestern Argentina. During the study of specimens of South Ameri- can Setaria as part of a project to enumerate those species that occur in both North and South Amer- ica, some specimens appeared to share a suite of distinctive characters that did not seem to fit any previously described species. Comparisons re- vealed that these specimens were easily separated from other Setaria taxa, and I here propose them to be elements of a new species. ally pyriform, the base broad, commonly coriaceous, brownish, narrowed to a hyaline upper part, with margins occasionally slightly thickened; the upper part may be only somewhat narrower than the lower (as in the type specimen), but the two parts appar- ently generally of different texture. Lemma of upper (fertile) floret 1.9-2.2 x 1.6-1.9 mm, and 1.5-1.8 mm thick, strongly gibbous, broadly ovate-elliptic, light to dark brown, the apex often becoming suf- fused with maroon at maturity, finely and evenly ru- gose, the cross-wrinkles slighty less pronounced to- ward the shortly apiculate, downturned apex; inrolled margins of the lemma (grasping the palea) are rather broad and somewhat flattened. Stamens 3, yellow. Stigmas 2, dark red.

Accelerator mass spectrometric (AMS) radiocarbon analyses of live-collected, prebomb samples of shell carbonates of the land snails Rabdotus dealbatus and R. alternatus from Texas were carried out to quantify the characteristic age anomalies of land snails from limestone areas. Age anomalies are similar for the two species; they average +700 yr and vary by ±180 yr (1σ) among samples. Serial analysis of 1 shell reveals a significant ontogenetic trend in 14 C age anomalies, with older apparent ages (up to 1200 yr) in the apical part of the shell and younger and uniform ages in the last whorl. No trend in age anomalies was found across a broad range of rainfall conditions (from 300 to 1000 mm mean annual rainfall).

The operation of the NSF-Arizona Laboratory is summarized. The methods used to determine accuracy and precision of radiocarbon measurements, and to make corrections for background contaminations are presented. An insulating support which has been installed, and a new heavy-ion beam line which is under construction, are described.

Samples of Setaria species from packrat middens, herbarium specimens and modern plants preserve a record of δ13C of atmospheric CO2 from 12,600 BP to the present. No secular trend is detected between 12,600 and 1800 BP, when the mean value of δ13C during that period was −6.5 ± 0.1‰ (the error is the standard deviation of the mean). Our value agrees with δ13C averages of pre-industrial CO2 from polar ice cores, and differs significantly from modern regional (-8.2 ± 0.1‰) and global (-7.7‰) values, which are higher because of fossil fuel burning.

We used a continuous krypton ion laser to rapidly oxidize milligram-sized fragments of coniferous driftwood of known ages, and dated the resulting carbon dioxide by accelerator mass spectrometry (AMS). AMS 14 C ages of non-pretreated young wood from different parts of two logs were within 10% of the ages of conventionally determined alkaline insoluble fractions. The age of the oldest whole wood measured after laser oxidation was within the error ranges of conventional values.

Tropical rock varnishes found on Hualalai and Mauna Kea Volcanoes, Hawai'i, vary systematically with time and environment. Radiocarbon dating of encapsulated organic matter, (K+ + Ca2+)/Ti4+ ratios, and Zn, Cu, and Ni trace element concentrations in rock varnish are consistent with lava flowages established by K-Ar and 14C dating, where samples are collected from arid microsites well away from the soil surface. However, inaccurate ages are obtained from rock varnish in subsurface locations and from sites with abundant lichens, cyanobacteria, and fungi that chemically erode varnish. In contrast with continental deserts, Hawaiian varnishes commonly interfinger with and are less common than rock coatings of amorphous silica. Laboratory experiments on Hawaiian rock varnishes indicate that K and Ca are preferentially leached relative to Ti over time and at higher temperatures. The location of in situ leaching has been identified in Hawaiian varnishes as porous textures without abundant detrital grains. THE STEADY GROWTH AND IMPROVEMENT of stratigraphic dating methods have been fundamental components ofQuaternary research during the past two decades. In contrast, surface exposure dating (SED) methods have been a weak sister; they attempt to measure the time since a land surface formed. Even though erosion of landforms is an ongoing process, SED methods utilize fossil remnants of "original surfaces" to date the landform (e.g., patches of glacial polish). Radiometric dating of lava flows and other stratigraphic units can offer valuable insights into rates of earth-surface processes (e.g., Dohrenwend et al. 1987). However, datable materials that provide surface exposure dates have a limited 1 This study has been supported by the National Science Foundation under Presidential Young Investigator Award EAR 87-57014 and a grant from the National Geographic Society. Manuscript accepted 10 April 1991. 2 Department ofGeography, Arizona State University, Tempe, Arizona 85287-0104. 3 Regional Facility for Radioisotope Analysis, University of Arizona, Tucson, Arizona 85721. 4U.S. Geological Survey, Federal Center, MS-903, Denver, Colorado 80225. 5 U.S. Geological Survey, 971 National Center, Reston, Virginia 22092. 6 Crocker Nuclear Laboratory, University of California, Davis, California 95616. distribution-that is, until the recent revolution in SED methods. SED methods can be classified as biological, cosmogenic, and weathering. Biological SED methods rely on the growth of trees or lichens to constrain the age of the underlying deposits. Cosmogenic SED methods measure the time-dependent buildup of stable and unstable isotopes that occur in situ in minerals because of nuclear reactions caused by cosmic rays. Weathering SED methods, in contrast, measure the time-dependent decomposition and diagenesis of minerals in terrestrial weathering environments. Originally proposed by Lal and Peters (1967), the measurement ofin situ cosmogenic isotopes was not possible before the advent of accelerator mass spectrometry (AMS) (Brown 1984, Elmore and Phillips 1987). Helium-3 is the stable cosmogenic isotope that has received the most attention (Cerling 1990, Kurz et al. 1990). Chlorine-36 (Phillips et al. 1990), lOBe and 26Al (Brown 1987, Nishiizumi et al. 1989), 14C (Jull et al. 1989), and 1291 (Fabryka-Martin et al. 1985) have seen the widest application of a large suite of radioactive cosmogenic isotopes. Other in situ cosmogenic isotopes that are being actively studied include 41Ca and 21Ne (Elmore and Phillips 1987, LaI1988).

The production of radioisotopes at the Earth's surface by cosmic-ray effects has been discussed for many years. Only in the past few years, with the higher sensitivity provided by accelerator mass spectrometry (AMS) in detecting 10Be, 26A1 and 36Cl, have the radioisotopes produced in this way been measured. We report here our measurements of cosmogenic 14C in terrestrial rocks at high altitude, and comparisons to other exposure-dating methods.

THE Black Sea, the world's largest anoxic marine basin, is frequently used as a modern analogue for the formation of organic-rich sediments and carbonaceous rocks 1-3, on the widely held assumption that anoxic conditions promote the preferential preservation of organic matter in sediments. Data for testing this hypothesis have so far been equivocal 4-7, but here we use radiocarbon ages obtained using accelerator mass spectrometry for the organic fraction of recent Black Sea sediments to estimate the organic carbon accumulation rates. These range from 0.69 to 2.09 g C m-2 yr-1 and are significantly lower than earlier estimates based on varve counting 6. Depending on the value taken for the rate of primary production in the Black Sea 4,8, between 0.7 and 2.1% of the organic carbon is preserved in the bottom sediments. When compared with carbon accumulation rates in equivalent oxygenated environments 9, these results indicate that the modern Black Sea is not a site of anomalously high organic carbon accumulation. This suggests that anoxic conditions in the water column may not be a prerequisite for the preservation of organic matter in marine sediments, and that models of the origin of carbonaceous facies in the geological record may therefore need to be modified.

Previous radiocarbon ages of detrital moss fragments in basal organic sediments of Lake Emma indicated that extensive deglaciation of the San Juan Mountains occurred prior to 14,900 yr B.P. (Carrara et al., 1984). Paleoecological analyses of insect and plant macrofossils from these basal sediments cast doubt on the reliability of the radiocarbon ages. Subsequent accelerator radiocarbon dates of insect fossils and wood fragments indicate an early Holocene age, rather than a late Pleistocene age, for the basal sediments of Lake Emma. These new radiocarbon ages suggest that by at least 10,000 yr B.P. deglaciation of the San Juan Mountains was complete. The insect and plant macrofossils from the basal organic sediments indicate a higher-than-present treeline during the early Holocene. The insect assemblages consisted of about 30% bark beetles, which contrasts markedly with the composition of insects from modern lake sediments and modern specimens collected in the Lake Emma cirque, in which bark beetles comprise only about 3% of the assemblages. In addition, in the fossil assemblages there were a number of flightless insect species (not subject to upslope transport by wind) indicative of coniferous forest environments. These insects were likewise absent in the modern assemblage.

The use of a particle accelerator to measure the 14C content of carbon-containing materials is briefly described. The application of the method to the determination of the age of some artistic artifacts is discussed.

Fossil insects from the late-glacial deposits at the Lamb Spring archaeological site, near Denver, Colorado, are relatively abundant and diverse, providing considerable paleoecological data for the site. The late Pleistocene insect fauna from the site comprises 72 identified taxa, principally beetles. However, the fauna presented an interpretive problem because it contained a mixture of prairie and alpine tundra species. This was initially considered to be the result of a mixing of faunal elements during the climatic transition of late-glacial times, a “no-modern-analog” fauna. Accelerator dating of insect fossil specimens from the two ecological groups helped resolve the paleoecological problem. Fossil specimens of the prairie-associated species were dated at 17,850 ± 550 yr B.P., while specimens of the tundra-associated species yielded an age of 14,500 ± 500 yr B.P. These dates reveal that what appeared to be an ecological mixing was probably a taphonomic problem, wherein full-glacial-age fossils were probably reworked into latest Wisconsin sediments. While both faunal assemblages reflect climatic conditions substantially colder than present, initial results suggest that the full-glacial fauna represents a cold, dry grassland or steppe environment, while the younger fauna suggests moister and more tundra-like conditions.

The current status of 14C measurements at the University of Arizona accelerator mass spectrometry (AMS) facility are summarized. New developments in measurements of in situ produced 14C by the spallation of oxygen by cosmic-ray neutrons in terrestrial rocks are also discussed. These results indicate that spallogenic 14C can be measured in high-altitude rock samples, and that the measured production rates agree well with theoretical estimates.

Pleistocene lake sediments in the Great Basin typically contain little organic carbon, and thus are difficult to date reliably by conventional radioccarbon methods. Paleoenvironmental data are abundant in these sediments, but are of limited value without adequate age controls. With the advent of accelerator-mass spectrometer (AMS) radiocarbon dating, it is now possible to date these paleolacustrine sediments. AMS dates were obtained on sediment cores from the Bonneville, Franklin, and Lahontan Basins. In the Bonneville Basin, the AMS-based chronology compares well with other chronologies constructed from dated shore-zone features. In the Bonneville and Franklin basins, AMS dates delimit unconformities not apparent by other means. We found that dispersed organic carbon from sediments deposited during relatively freshwater intervals provided apparently reliable AMS radiocarbon dates. Carbonate microfossils from the Lahontan Basin also produced results that appear reasonable, while bulk carbonate yielded erroneous results.

Accelerator mass spectrometry radiocarbon dating of organic matter in the basal layer of rock varnish provides a minimum surface exposure age for the underlying landform. Varnishes 14C ages indicate that the latest Pleistocene highstand of Lake Lahontan was abandoned before 12.7 ka, Searles Lake before 13.6 ka, Lake Manly before 13.0 ka, and Lake Mojave before 11.6 ka. The maxima of the Tioga glaciation at Pine Creek in the eastern Sierra Nevada, California, and the Middle Creek glaciation in the adjacent White Mountains of California-Nevada was just before 18–19 ka; major recessional moraines were deposited in these ranges before 12–13 ka. Penecontemporaneous glacial retreat and high levels in closed-basin lakes cannot be explained by glacial melt water, but probably reflects higher snowlines and more precipitation. A relatively minor glacial advance, perhaps related to the “Younger Dryas” in Europe, may be penecontemporaneous with a slight transgression in lake levels about 10–11 ka in the western Basin and Range, suggesting a temporary return to cooler or wetter conditions, or both at the termination of the Pleistocene.

Very small samples from the Shroud of Turin have been dated by accelerator mass spectrometry in laboratories at Arizona, Oxford and Zurich. As controls, three samples whose ages had been determined independently were also dated. The results provide conclusive evidence that the linen of the Shroud of Turin is mediaeval.