How fast does varnish grow?

Tanzhuo Liu Lamont-Doherty Earth Observatory of Columbia University, Palisades, New York 10964, USA Wallace S. Broecker

ABSTRACT that we were studying the oldest varnish from a Rates of rock-varnish accumulation have never been systematically documented owing to given site. In other words, we wanted to minimize the difficulty in accurately determining both varnish thickness and age. In this study, we quan- the lag time between surface exposure and the on- titatively assess varnish-accumulation rates through thin sectioning and microscopic examina- set of varnishing. Our laboratory criterion is the tion of rock varnish from radiometrically dated geomorphic features of late Quaternary age in pattern of layers seen in rock varnish (cf. Liu, the western United States drylands. Our data indicate that rock varnish grows at rates ranging 1994; Liu and Dorn, 1996; Cremaschi, 1996; Liu from <1 to 40 µm/k.y. on subaerially exposed rock surfaces and rarely reaches a thickness ex- et al., 1998). In the laboratory, we selected layered ceeding 200 µm regardless of age. Our data also indicate that varnish-accumulation rates vary varnish microbasins (i.e., 1–3-mm-wide varnish- greatly from sample to sample at a given site, suggesting that varnish thickness does not corre- filled dimples) showing the oldest microstrati- late with the age of the associated geomorphic feature, invalidating the potential use of varnish graphic unit. thickness as a relative age indicator in geomorphology and archaeology. Nevertheless, being the slowest known accumulating terrestrial sedimentary deposit, rock varnish constitutes a unique Age Control and Thickness Measurement long-term microscale sedimentary archive of past environmental changes in . Varnish ages were inferred from the surface- exposure ages of the sampled geomorphic fea- Keywords: rock varnish, thickness, growth rate, age control, microstratigraphy, western United States. tures. In this study, we collected varnish samples from 27 geomorphic features that have been ra- INTRODUCTION present the first quantitative database on rates of diometrically dated (Table 1). The youngest var- Rock varnish is a dark patina that accumulates varnish formation on landforms that have been nish was from basaltic debris-flow deposits in on rock surfaces. On the average, it consists of radiometrically dated. Grand Canyon, Arizona, that were 3He dated as about 30% manganese and oxides. The re- 1.5 ka (T. Cerling, 1997, written commun.). The mainder is made up by oxides of Si, Al, Mg, K, METHODOLOGY oldest varnish was from a quartzitic boulder on and Ca, some of which is contained in clay min- Sampling Sites and Sample Collection the oldest alluvial-fan surface in Death Valley, erals (Potter and Rossman, 1977). Varnish can The methodology of this study is very simple; California, that was 10Be/26Al dated as 250 ka form on virtually all lithologies including quartz- it involves collecting varnish-coated rocks from (S. Ivy-Ochs, 1998, written commun.). Other var- ite; thus, it is believed to be of sedimentary origin, radiometrically dated surfaces and measuring var- nish samples were from a variety of dated geo- the constituents being supplied from the atmo- nish thickness. Our sampling sites are located in morphic features such as the Bandera lava flow in sphere (Fleisher et al., 1999). Although varnish the regions of the western United States the Zuni-Bandera volcanic field, New Mexico has been found in numerous terrestrial weathering where the environmental setting has fluctuated (10 ka [14C]; Laughlin et al., 1994), the Dry Falls environments including the Antarctic (Dorn et al., between arid and semiarid during the Quaternary in Owens Valley (15.3 ka; Cerling, 1990), the 1992b), Iceland (Douglas, 1987), and Hawaii and where rock varnish is ubiquitous and well Tabernacle Hill lava flow in Utah (14.4 ka [14C]; (Dorn et al., 1992a), it is most common in arid to preserved. Radiometrically dated geomorphic Cerling and Craig, 1994), the Blackhawk land- semiarid deserts of the . features of late Quaternary age such as abandoned slide in the (17.4 ka [14C]; Stout, Rock varnish has been a scientific wonder for shorelines of pluvial lakes, lava flows, alluvial-fan 1977), and the shorelines of major pluvial lakes almost two centuries, dating back to Alexander and terrace surfaces, debris-flow deposits, fault that formed from 10.5 ka (14C) to 120 ka in the von Humboldt’s travels from 1799 to 1804 in scarps, and paleolandslides exist in the regions, Great Basin of the western United States (Table1). South America (von Humboldt, 1812). However, providing age control on the earliest time at which Because all of these dates are for geomorphic two basic questions raised by von Humboldt re- varnish growth was initiated (Table 1). events, they provide maximum ages for the initi- main unanswered. How long does it take for var- As is the case in virtually all Quaternary re- ation of varnish growth. nish growth to initiate? Once started how fast does search, collecting the right type of sample is criti- We collected 42 rock samples, and prepared varnish grow? Certainly, there are anecdotal ob- cal. In order to avoid circular reasoning, we did 74 thin sections, each hosting 1–4 individual mi- servations useful in answering von Humboldt’s not sample based on the field appearance of var- crobasins. Thickness measurements were taken questions. In humid environments, rock surfaces nish darkness, estimates of thickness in hand sam- on 115 such microbasins. Varnish thickness was may be coated within a few decades (Dorn and ples, or surface coverage of varnish. Rather, we accurately measured on thin sections by using an Meek, 1995; Klute and Krasser, 1940). For the judged varnish samples based only on whether the optical microscope. Varnish thin sections were Mojave Desert, suggestions have been made that varnished surface retains primary surface features first prepared by using a new thin-sectioning varnish formation begins within 25 yr of expo- (cf. Liu and Dorn, 1996). For example, we col- technique1 (Liu and Dorn, 1996) and then pho- sure (Engel and Sharp, 1958). But the most com- lected only subaerially exposed rock varnishes on tographed on color print films under the micro- mon observation is that once initiated, varnish boulders (>15 cm in size) with surfaces abraded scope with transmitted polarized light. The mea- formation is slow on subaerially exposed rock by a known geomorphic process, such as alluvial surements of varnish thickness were made on 5" surfaces (Basedow, 1914; Hayden, 1976; Denny, or fluvial polishing, or on lava flows displaying × 3.5" (12.6 cm × 8.7 cm) standard color prints 1965). Yet, after 200 yr of study, there are only a originally formed ropy or aa surface features. few qualitative comments on rates of varnish Thus, we knew that each varnish started to form 1 A varnished rock chip is placed in epoxy and a pol- growth where the ages of the archaeological (cf. sometime after the geomorphic event that created ished cross section of varnish is made on side 1. Then, the section is placed again in epoxy, and the sample is Pyramids, Lucas, 1905) and geomorphological the host surface. ground down on side 2. The varnish is then polished on (cf. paleolake shoreline, Bard et al., 1978) fea- The next step was to use a laboratory criterion, side 2 until it is thin enough for microstratigraphy to be tures are known. The purpose of this paper is to independent of thickness, to maximize the chance visible with light microscope.

Data Repository item 200020 contains additional material related to this article.

Geology; February 2000; v. 28; no. 2; p. 183–186; 3 figures; 1 table. 183 with 477 × magnification of the varnish thin sec- vironmental context: varnish growing in 1–3-mm- nishing.) On a regional scale, for example, var- tions. The apparent varnish thickness was mea- wide microbasins on rock surfaces. nish samples from the ca. 14 ka (14C) shorelines sured in millimeters with a transparent ruler of Searles Lake and Bonneville Lake have thick- placed perpendicular to the outer varnish surface RESULTS nesses of 244 and 121 µm, respectively (Fig. 2, K on the color prints; then true varnish thickness Table 1 lists only the greatest thickness mea- and L). On a local scale, varnish samples from was obtained in micrometers by multiplying the sured on each sample, and Figure 2 presents the the 10.5 ka (14C) highstand of Silver Lake and apparent thickness with a conversion factor of color images of the corresponding varnish mi- 11.5 ka (14C) highstand of Coyote Lake in the 1/477. Measurements when converted to absolute crobasins; however, the full data set of the thick- Mojave Desert display comparable microstrati- varnish thickness are accurate to ±1 µm. On each ness measurements is available.2 The greatest var- graphies that are, however, extremely different in thin section, thickness was measured at the maxi- nish thickness observed in this study is 244 µm thickness, being 212 µm for the former and mum depth of the individual varnish microbasins (Fig. 2K), from the 14.0 ka (14C) highstand of 68 µm for the latter (Fig. 2, E and F). Even on a that display the most complete and consistent mi- Searles Lake in the Mojave Desert (Benson et al., submillimeter scale, different parts of a varnish crostratigraphy (Fig. 1) and thus most closely rep- 1990). The thinnest varnish is 10 µm thick (see microbasin may have different thicknesses owing resent the entire exposure history of the sampled footnote 2) and was collected from an 11.5 ka to topographic variations within the microbasin geomorphic feature (Fig. 2; also see Liu and (14C) alluvial-fan surface in Valley of (Fig. 2, A, E, G, P, and Q). However, varnish sam- Dorn, 1996). When crack(s) appeared in a varnish southern Nevada (Bell et al., 1998). The calcu- ples with similar thicknesses may not be similar microbasin (cf. Fig. 2D), the thickness was de- lated mean thickness of these varnishes is 100 µm, in age. A ca. 120 ka varnish sample from the ducted from the overall thickness measurement. with a standard deviation of 55 µm. This result is Blackwelder shoreline of Lake Manly in Death The layering patterns provide supplementary in good agreement with our field and laboratory Valley (Hooke and Dorn, 1992) is 150 µm thick information. Measuring thicknesses of the oldest knowledge that, regardless of their ages, rock (Fig. 2W); varnish samples with thicknesses layering, however, does not alter the fact that each varnishes in the drylands of the western United close to this value could be either younger, like data point fails to include the lag time between States rarely reach a thickness >200 µm. the one from the 81 ka Lathrop Wells lava flow surface exposure and the initiation of varnish We found that there is no correlation between (Fig. 2U; Zreda et al., 1993), or much older, like growth; no study has yet quantified this lag time varnish thickness and age of the underlying geo- the one from the 250 ka alluvial-fan surface in in a systematic fashion. Postdepositional modifi- morphic surface. Rather, varnish samples of sim- Death Valley (Fig. 2X). cations such as spalling, chemical leaching, com- ilar apparent age can be quite different in thick- Apparent varnish-accumulation rates were cal- paction, and diagenesis (cf. Krinsley, 1998) may ness. (The adjective “apparent” is used herein to culated from the full data set of varnish thick- reduce the original varnish thickness. The influ- indicate that the geomorphic age is not the age of nesses and ages (see footnote 2). Figure 3 presents ences of those factors imply that the varnish the varnish; there is an unknown lag time be- a plot of apparent varnish-accumulation rate vs. growth rates obtained in this study are only mini- tween the ages in Table 1 and the onset of var- varnish age. We found that varnish-accumulation mum values. Furthermore, although some work- rates vary widely, from as high as 40 µm/k.y. on ers (cf. Reneau et al., 1992) do not think that lay- 2 GSA Data Repository item 200020, Varnish-thick- a 1.5 ka varnish in Grand Canyon (Fig. 2A) to as ering patterns offer insight into varnish age, the ness measurements, ages, and apparent accumulation low as 0.6 µm/k.y. on a 250 ka varnish in Death rates, is available on request from Documents Secre- data presented here represent the results of the tary, GSA, P.O. Box 9140, Boulder, CO 80301-9140, Valley (Fig. 2X), with a mean apparent accumu- first systematic study on rates of varnish growth [email protected], or at www.geosociety.org/ lation rate of 6.4 µm/k.y. These rates vary later- on samples selected from a specific, replicable en- pubs/drpint.htm. ally over a single rock surface, being higher in

184 GEOLOGY, February 2000 Figure 1. Idealized rock-varnish microbasin showing most complete microstratigraphy from 14 14C ka geomorphic surface. Dashed line identifies position where maximum var- nish thickness was measured. deep varnish microbasins and lower in shallow microbasins (Fig. 2G; see footnote 2). Within a given microbasin, varnish-accumulation rates also vary; the highest rate is in the center and the lowest is on the edge (Fig. 2, A and E). In addi- tion, we observed that younger varnish samples (<30 ka) appear to have higher apparent accumu- lation rates (3–15 µm/k.y.), whereas older var- nish samples (>50 ka) have lower apparent accu- mulation rates (<3 µm/k.y.) (Fig. 3).

DISCUSSION AND IMPLICATIONS Observations and speculations on rates (or time) of varnish formation are scattered through- out the rock-varnish literature. The most rapid varnish initiation time so far reported in the liter- ature is 25 yr on rocks disturbed during road con- struction in the Mojave Desert (Engel and Sharp, Figure 2. Optical photographs of rock varnish displaying largest thickness and most complete 1958). In southern California, near Fontana, rock microstratigraphy in 1–3-mm-wide microbasins measured for this study. Numbers at upper left corner of each image provide maximum varnish thickness and ages. varnish has been reported on 35-yr-old slag de- posits (Dorn and Meek, 1995). In this study, we observed that the 1.5 ka basaltic debris flow de- develop as visually discernible patchy varnishes; 45 posits in Grand Canyon, Arizona, are coated with in contrast, only the slowest-forming varnishes 40 60-µm-thick patchy varnish (Fig. 2A) and the have the potential of achieving very great age 35 3.0 ka (14C) McCartys lava flows in the Zuni- (without being eroded). The greater long-term 30 Bandera volcanic field, New Mexico, are coated preservation of slower growing varnishes is not a 25 with 52-µm-thick patchy varnish (Fig. 2B). In the new concept in varnish research (cf. Dorn and 20 western United States drylands, many observa- Oberlander, 1982), and it probably explains why

(µm/k.y.) 15 tions of rock varnish on artifacts (Hunt, 1975; the apparent accumulation rates of latest Pleisto- 10 Hayden, 1976), alluvial-fan surfaces (Denny, cene and Holocene visually discernible varnishes 5 1965), and abandoned lake shorelines (Bard in our study areas are significantly higher than Varnish-Accumulation Rate 0 et al., 1978) support the conclusion that it gener- those of the late varnishes (Fig. 3). 0 100 200 300 ally takes about 3000 to 5000 yr to form a visually For decades, geologists and archaeologists Rock-Varnish Age (ka) discernible patchy varnish and about 10000 yr for have used the slow growth of varnish on rock sur- a heavily coated varnish (Elvidge and Iverson, faces as a means of estimating ages for ancient Figure 3. Variations of apparent varnish-accu- mulation rate over varnish age. 1983). Our quantitative data from this study indi- landforms, petroglyphs, and artifacts (Carter, cate that varnishes accumulate at rates of <1 to 1980; Espizua, 1993; Grote and Krumbein, 1993). 40 µm/k.y. Given an average thickness of 50 and The underlying hypothesis of this dating tech- Given the apparent accumulation rates of a few 100 µm for visually discernible patchy varnish nique is that rock varnish accumulates at a com- micrometers per millennium (or a few nanome- and heavily coated varnish, respectively, our var- parable rate in a given region. The results from ters per year) obtained in this study, rock varnish nish-accumulation rates are comparable to these this study provide the first quantitative evidence is probably the slowest known accumulating ter- earlier anecdotal observations. that varnish-accumulation rates change greatly restrial sedimentary deposit. Such finding has a Once initiated, rock varnishes appear to ac- on both local and regional scales, and that even in significant implication for the potential use of crete semicontinuously on rock surfaces (cf. Liu, a single varnish section there are large variations rock varnish as a climate recorder. Despite wide 1994; Cremaschi, 1996; Krinsley, 1998). During in accumulation rates (Figs. 2 and 3). Relative or differences in growth rate, chemical microstratig- the course of rock varnish development, only absolute varnish thickness cannot be used to pro- raphy in rock varnish has been demonstrated to those having the highest accumulation rates will vide a reliable estimate for the age of varnished be regionally consistent and comparable (Liu and have the greatest potential of being able to first landforms. Dorn, 1996; Liu et al., 1998); it carries valuable

GEOLOGY, February 2000 185 information about past environmental fluctua- Cosmogenic ages for earthquake recurrence inter- electron microscopy: Sedimentology, v. 45, tions (Perry and Adams, 1978; Dorn, 1984; Liu vals and debris flow fan deposition, Owens Valley, p. 711–725. and Dorn, 1996; Liu et al., 1998). As seen in Fig- California: Science, v. 270, p. 447–450. Laughlin,A. W., Poths, J., Healey, H. A., Reneau, R., and Carter, G. F., 1980, Earlier than you think: College Sta- WoldeGabriel, G., 1994, Dating of Quaternary ure 2, varnishes from the drylands of the western tion, Texas A&M University Press, 348 p. basalts using the cosmogenic 3He and 14C methods United States contain alternative black layers that Cerling, T. E., 1990, Dating geomorphic surfaces using with implications for excess 40Ar: Geology, v. 22, are rich in Mn and orange layers that are poor in cosmogenic 3He: Quaternary Research, v. 33, p. 135–138. Mn (Fleisher et al., 1999); the alternations prob- p. 148–156. Liu, T., 1994, Visual microlaminations in rock varnish: Cerling, T. E., and Craig, H., 1994, Cosmogenic 3He A new paleoenvironmental and geomorphic tool ably correspond to the fluctuations of moisture production rates from 39°N to 46°N latitude, in drylands [Ph.D. thesis]: Tempe, Arizona State conditions in deserts (Cremaschi, 1996; Liu et al., western USA and France: Geochimica et Cos- University, 173 p. 1998). Similar types of microlaminations are also mochimica Acta, v. 58, p. 249–255. Liu, T., and Dorn, R. I., 1996, Understanding the spa- observed in rock varnish from the Judean Desert Cremaschi, M., 1996, The rock varnish in the Messak tial variability of environmental change in dry- in Israel, Gurbantünggut Desert in western China, Settafet (Fezzan, Libyan ), age, archaeo- lands with rock varnish microlaminations: Asso- logical context, and paleo-environmental impli- ciation of American Geographers Annals, v. 86, and Patagonian Desert in . The slowly cation: Geoarchaeology, v. 11, p. 393–421. p. 187–212. accumulating rock varnish offers a long-term (at Denny, C. S., 1965, Alluvial fans in the Death Valley re- Liu, T., Broecker, W. S., Bell, J. B., and Mandeville, C., least 0–250 k.y.) climate recorder in deserts where gion of California and Nevada: U.S. Geological 1998, Terminal Pleistocene wet event recorded in other climate proxies are either not available or Survey Professional Paper 466, 62 p. rock varnish from southern Nevada, western Dorn, R. I., 1984, Cause and implications of rock var- United States: Geological Society of America Ab- difficult to read. nish microchemical laminations: Nature, v. 310, stracts with Programs, v. 30, no. 7, p. 361. p. 767–770. Lucas, A., 1905, The blackened rocks of the Nile cata- CONCLUSIONS Dorn, R. I., and Meek, N., 1995, Rapid formation of racts and of the Egyptian deserts: Cairo, Egypt, This study provides an answer to Alexander rock varnish and other rock coatings on slag National Printing Department, 58 p. von Humboldt’s 200-yr-old question on the rate deposits near Fontana: Earth Surface Processes Meek, N., 1990, Late Quaternary geochronology and ge- and Landforms, v. 20, p. 547–560. omorphology of the Manix Basin, San Bernardino of varnish growth, by presenting the first quanti- Dorn, R. I., and Oberlander, T. M., 1982, Rock varnish: County, California [Ph.D. thesis]: Los Angeles, tative database from well-dated sites in semiarid Progress in Physical Geography, v. 6, p. 317–367. University of California, 212 p. to arid environments of the western United Dorn, R. I., Jull, A. J. T., Donahue, D. J., Linick, T. W., Meek, N., 1997, New discoveries about the late Pleis- States. Rock varnish is the slowest accumulating Toolin, L. J., Moore, R. B., Rubin, M., Gill, T. E., tocene history of Lake Harper, Mojave Desert, and Cahill, T. A., 1992a, Rock varnish on Hualalai California: Southern California Academy of Sci- terrestrial sedimentary deposit; it has growth and Mauna Kea Volcanoes, Hawaii: Pacific Sci- ences, Programs and Abstracts, v. 28. rates of <1–40 nm/yr. The tremendous variability ence, v. 46, p. 11–34. Perry, R. S., and Adams, J., 1978, Desert varnish: Evi- in growth rates, observed here for varnishes at Dorn, R. I., Krinsley, D. H., Liu, T., Anderson, S., dence of cyclic deposition of manganese: Nature, spatial scales from micrometers to kilometers, in- Clark, J., Cahill, T. A., and Gill, T. E., 1992b, Man- v. 276, p. 489–491. validates varnish thickness as a reliable age indi- ganese-rich rock varnish does occur in : Phillips, F. M., Ayarbe, J. P., Harrison, B., Sharma, P., Chemical Geology, v. 99, p. 289–298. and Elmore, D., 1998, Calibration of fault-scarp cator in geomorphology and archaeology. Douglas, G. R., 1987, Manganese-rich rock coatings diffusion models using cosmogenic nuclides: from Iceland: Earth Surface Processes and Land- Geological Society of America Abstracts with ACKNOWLEDGMENTS forms, v. 12, p. 301–310. Programs, v. 30, no. 7, p. 298. Funded by the U.S. Department of Energy. We thank Elvidge, C. D., and Iverson, R. M., 1983, Regeneration Pohl, M., 1995, First radiocarbon ages on organics from J. Bell, T. Cerling, S. Ivy-Ochs, N. Meek, and F. Phillips of desert pavement and varnish, in Webb, R. H., piedmont alluvium, Ajo Mountains, Arizona: for sample collection, age dating, and thoughtful discus- and Wilshire, H. G., eds., Environmental effects Physical Geography, v. 16, p. 339–353. sions; David Krinsley and an anonymous reader for crit- of off-road vehicles: New York, Springer-Verlag, Potter, R. M., and Rossman, G. R., 1977, Desert varnish: ical reviews. Lamont Doherty Earth Observatory contri- p. 225–243. The importance of clay minerals: Science, v. 196, bution 5984. Engel, C. G., and Sharp, R. S., 1958, Chemical data on p. 1446–1448. desert varnish: Geological Society of America Reneau, S. L., Raymond, R., and Harrington, C., 1992, REFERENCES CITED Bulletin, v. 69, p. 487–518. Elemental relationships in rock varnish stratigra- Anthony, E., and Poths, J., 1992, 3He surface exposure Espizua, L. E., 1993, Quaternary glaciations in the Rio phy, layers, Cima volcanic field, California: Im- dating and its implications for magma evolution Mendoza Valley, Argentine : Quaternary plications for varnish development and the inter- in the Potrillo volcanic field, Rio Grande rift, Research, v. 40, p. 150–162. pretation of varnish chemistry: American Journal New Mexico, USA: Geochimica et Cosmochim- Fitzpatrick, J. A., and Bischoff, J. L., 1993, Uranium- of Science, v. 292, p. 684–723. ica Acta, v. 56, p. 4105–4108. series dates on sediments of the high shoreline of Stout, M. L., 1977, Radiocarbon dating of landslides in Bard, J. C., Asaro, F., and Heizer, R. F., 1978, Per- Panamint Valley, California: U.S. Geological Sur- southern California: California Geology, v. 30, spectives on the dating of prehistoric Great Ba- vey Open-File Report 93-232, p. 1–15. p. 99–105. sin petroglyphs by neutron activation analysis: Fleisher, M., Liu, T., Broecker, W. S., and Moore, W., Stuiver, M., and Reimer, P. J., 1993, Extended 14C data Archaeometry, v. 20, p. 85–88. 1999, A clue regarding the origin of rock varnish: base and revised calib 3.0 14C age calibration pro- Basedow, H., 1914,Aboriginal rock carvings of great an- Geophysical Research Letters, v. 26, p. 103–106. gram: Radiocarbon, v. 35, p. 215–230. tiquity in South Australia: Royal Anthropological Grote, G., and Krumbein, W. E., 1993, Pilze und von Humboldt, A., 1812, Personal narrative of travels to Institute Journal, v. 44, p. 195–211. chemoorganotrophe Bakterien auf Felsbildern in the equinoctial regions of America during the Bell, J. W., Ramelli, A. R., and Caskey, S. J., 1998, Geo- der Wuste: Restauro, v. 1993, no. 2, p. 111–114. years 1799–1804, Volume II (translated and logic map of the Tule Springs Park quadrangle: Hayden, J., 1976, Pre-altithermal archaeology in the edited by T. Ross in 1907): London, George Bell Nevada Bureau of Mines and Geology Map 113, Sierra Pinacate, Sonora, Mexico: American An- & Sons, 521 p. scale 1:24 000. tiquity, v. 41, p. 274–289. Wells, S. G., McFadden, L. D., and Dohrenwend, J. C., Benson, L. V., Currey, D. R., Dorn, R. I., Lajoie, K. R., Hooke, R. L., and Dorn, R. I., 1992, Segmentation of 1987, Influence of late Quaternary climatic Oviatt, C. G., Robinson, S. W., Smith, G. I., and alluvial fans in Death Valley, California: New in- changes on geomorphic and pedogenic processes Stine, S., 1990, Chronology of expansion and con- sights from surface-exposure dating and laboratory on a desert piedmont, eastern Mojave Desert, Cal- traction of four Great Basin lake systems during modelling: Earth Surface Processes and Land- ifornia: Quaternary Research, v. 27, p. 130–146. the past 35 000 years: Palaeogeography, Palaeo- forms, v. 17, p. 557–574. Zreda, M. G., Phillips, F. M., Kubik, P. W., Sharma, P., climatology, Palaeoecology, v. 78, p. 241–286. Hunt, C. B., 1975, Death Valley: Geology, ecology, and and Elmore, D., 1993, Cosmogenic 36Cl dating of Benson, L., Kashgarian, M., and Rubin, M., 1995, archaeology: Berkeley, University of California a young basaltic eruption complex, Lathrop Wells, Carbonate deposition, Pyramid Lake subbasin, Press, 234 p. Nevada: Geology, v. 21, p. 57–60. Nevada: 2. Lake levels and polar jet stream po- Klute, F., and Krasser, L. M., 1940, Über wüstenlack- sitions reconstructed from radiocarbon ages and bildung im Hochgebirge: Petermanns Geographis- Manuscript received June 28, 1999 elevations of carbonates (tufas) deposited in the che Mitteilungen, v. 86, p. 21–22. Revised manuscript received October 8, 1999 Lahontan basin: Palaeogeography, Palaeoclima- Krinsley, D., 1998, Models of rock varnish formation Manuscript accepted November 11, 1999 tology, Palaeoecology, v. 117, p. 1–30. constrained by high resolution transmission Bierman, P. R., Gillespie,A. R., and Caffee, M. W., 1995,

186 Printed in U.S.A. GEOLOGY, February 2000