Controls of tufa development in Pleistocene Lake Bonneville, Utah

Prominent tufa localities along the Provo level (∼14,000 14C yr B.P.) shoreline in Pleistocene Lake Bonneville have been characterized in detail. Three types of tufa are recognized: capping tufa, beachrock, and capping tufa over beachrock. Capping tufa and beachrock are end members of a continuum based on variable clastic content. All three types typically occur on headland environments that had stable substrate and little sediment input. Tufa development correlates with bedrock exposure and landform orientation, which in turn are correlated (R2=0.89) with the longest fetch directions in the basin. Tufa also tends to be located at major subbasin divides and in the western portion of the basin.

Controls of Tufa Development in Pleistocene Lake Bonneville, Utah Alisa Felton, Paul W. Jewell,1 Marjorie Chan, and Donald Currey2 Department of Geology and Geophysics, University of Utah, Salt Lake City, Utah 84112 (e-mail: [e-mail]) ABSTRACT Prominent tufa localities along the Provo level (~14,000 MC yr B.P.j shoreline in Pleistocene Lake Bonneville have been characterized in detail. Three types of tufa are recognized: capping tufa, beachrock, and capping tufa over beachrock. Capping tufa and beachrock are end members of a continuum based on variable clastic content. All three types typically occur on headland environments that had stable substrate and little sediment input. Tufa development correlates with bedrock exposure and landform orientation, which in turn are correlated \R2 = 0.89) with the longest fetch directions in the basin. Tufa also tends to be located at major subbasin divides and in the western portion of the basin. Online enhancements: appendix tables. Introduction Calcium carbonate tufa deposits coat the shorelines of Pleistocene Lake Bonneville throughout north­ern Utah (fig. 1). While King (1878) and Gilbert (1890) described Lake Bonneville tufa and noted the connection between water aeration and tufa de­velopment, the bulk of subsequent Lake Bonneville research focused on clastic deposits and shoreline features of the lake. For this reason, the tufa de­posits lack detailed characterization and a deposi- tional model. In the Lake Lahontan basin, the western Great Basin "cousin" of Lake Bonneville, tufa deposits have been more thoroughly examined. Benson (1994) and Benson et al. (1995) describe physical characteristics and paleoclimatic significance of Lake Lahontan tufa deposits, noting several layers of tufa growth with unique morphologies formed when lake levels stabilized. Benson et al. (1995) cite the role of Lahontan subbasin thresholds in con­trolling the elevation of tufa deposition in the Pyr­amid subbasin. Benson (1994) identifies several conditions that encourage carbonate deposition, in­cluding elevated water temperature, a hydrologi- Manuscript received February 11, 2005; accepted December 15, 2005. 1 Author for correspondence; e-mail: [e-mail]. 2 Department of Geography, University of Utah, Salt Lake City, Utah 84112; deceased. cally closed basin, proximity to a source of calcium, and a solid substrate. These factors are potentially important to Bonneville tufa deposition as well. The relative influence of biological and physical factors influencing tufa development is an unre­solved issue in the broader field of understanding controls of calcium carbonate formation in terres­trial waters. Kelts and Hsu (1978) present a thor­ough discussion of carbonate sedimentation in freshwater that includes the chemistry of calcite precipitation as well as a review of biogenic and physical considerations applied to carbonate de­position. Both algae and photosynthesizing cyano­bacteria play a role in the biology of many tufas. Pedley (2000) indicates that tufas precipitate in part because of conditions in the physical environment and in part as a result of biological activity. Ford and Pedley (1996) note the connection between tu­fas, water aeration, and biology by characterizing tufa and travertine deposits according to physical form. The biological aspect of cyanobacteria cal­cification in relation to availability of C02 and phosphate in a water body is reviewed by Merz- Preifi (2000) and Riding (2000). Two recent studies have examined Lake Bonne­ville tufa from a geochemical perspective. Hart et al. (2004) used strontium isotopes to understand the sources of water within the lake. In the process of [The Journal of Geology, 2006, volume 114, p, 377-389] © 2006 by The University of Chicago, All rights reserved, 0022-1376/2006/11403-0008$15,00 377 378 A . F E L T O N ET AL. Figure 1. Lake Bonneville at highstand 15,000 11C yr ago and tufa locations examined in this study. 1 = north end of the Hogup Mountains, 2 = Terrace Moun­tain, 3 = Lucin Hill and Lion Mountain, 4 = Wcndovcr Knob, 5 = Volcano Peak and Silver Island tombolo, 6 = north end of Fish Springs Range, 7 = Sand Pass, 8 = Table Mountain, 9 = Black Rock Volcano, 10 = north end of the Oquirrh Mountains, 11 = Point of the Mountain, 12 = Beck Street/Wasatch fault, 13 = Cutler Dam; SLC = Salt Lake City. sample collection and analysis, Hart et al. (2004) were able to demonstrate unequivocally the lacus­trine nature of the tufas as well as the fact that most of these rocks have undergone minimal subaerial alteration. Nelson et al. (2005) addressed tufa de­velopment of Lake Bonneville from a wide-ranging petrographic, field, and geomorphologic perspective and established both the local and basinwide con­trols of tufa formation. Tufa Terminology. In order to clarify terms used in this work, it is necessary to distinguish among four often-confused calcium carbonate deposits: tufa towers, waterfall tufa, travertine, and standard tufa. Tufa towers, such as those at Mono Lake in eastern California and the Needles at Pyramid Lake, Nevada, are formed by carbonate-saturated springs flowing into waters (Cloud and Lajoie 1980; Benson 1994). Fluvial waterfall tufas are created by excessive CO, degassing in calcium-saturated wa­ters as they flow over a knick point or hydraulic drop (Zhang et al. 2001). Fluvial waterfall tufas of­ten have a biological component to their formation and are also known as tufa dams or barrages. Trav­ertine is calcium carbonate deposited in hydro­thermal or warm waters. In contrast, tufa forms in ambient-temperature waters. Both travertines and standard tufas form in freshwaters (Ford and Pedley 1996). Lake Bonneville shore-zone calcium carbon­ate deposits belong in the standard tufa category because they were deposited in an alternating open and closed basin in relatively fresh, ambient- temperature waters. Lake Bonneville History and Geologic Setting. Pleistocene Lake Bonneville was a closed-basin plu­vial lake that left a detailed record of climate change in northern Utah, eastern Nevada, and southern Idaho in the form of nearshore sediments, shoreline benches, and offshore sediments (Gilbert 1890; Currey 1990). Lake Bonneville began filling around 27,000 14C yr ago (Oviatt 1997). The pluvial lake rose to the Bonneville highstand around 15,000 14C yr ago (fig. 2), with a maximum surface area of 50,000 km2 (Wambeam 2001). Approximately 14,500 14C yr ago, Lake Bonneville catastrophically flooded through the Red Rock Pass threshold (Gil­bert 1890; Oviatt et al. 1992) and stabilized at the Provo stillstand level, about 140 m below the Bon­neville highstand elevation. During the Provo still- ii 5300 rt! OJ i/l 5200 <u O 5100 < ■w 5000 <U OJ , 4900 C S 4800 5 V 4700 LU "3 4600 HE tj 4500 CC ^4400 rn | 4300 ^ 4200 "O O KJ 103 14Cyr BP (after Oviatt, 1997} Figure 2. Hydrograph of Lake Bonneville in radiocarbon years B.P. versus lake surface elevation. This study fo­cuses on transgrcssivc jprchighstand) shoreline se­quences and Provo shoreline deposits as identified by the highlighted areas in the dashed boxes at elevations be­tween 1420 and 1495 m.Journal of Geology TUFA DEVELOPMENT 3 79 stand, Lake Bonneville had periods of being both hydrologieally closed and open (Sack 1999; Godsey et al. 2005). The closed-basin system during the Provo stillstand potentially affected tufa formation in the basin. After the Provo stillstand, the lake continued its regression to the Gilbert level about 10,000 14C yr ago. The Gilbert lake level marked the end of the Bonneville high-water phase. Lake Bonneville occupied a large section of the eastern portion of the extensional Basin and Range geologic province. Implications of this setting in­clude numerous lake subbasins, an active tectonic regime with concurrent uplift and erosion, and steep-walled basins creating active sediment sup­plies and accommodation spaces. These factors worked in conjunction with an arid Holocene cli­mate to produce and preserve the well-defined shoreline and shore-zone sediments exposed today. Bedrock at selected field sites ranges from Preeam- brian to Tertiary in age and consists primarily of limestones, cherts, granitics, volcanics, and quartz- ites. The amount and extent of limestone bedrock exposure contributing calcium to the lake water and the isolated subbasins with unique water chemistry were two geologic factors that probably influenced tufa development in the Bonneville ba­sin as discussed in this article. Methods Tufa Analysis. Of the original 13 field sites (fig. 1), 10 were mapped on 7.5-min United States Geo­logical Survey quadrangle maps (table 1). Spatial extent, thicknesses, types of tufa, and percentage of tufa-covered area coating a shoreline surface (ta­ble 2) were recorded at each field site. In order to understand the petrographic characteristics of the tufas, two samples of tufa were taken from each of the 10 sites for analysis. Samples were slabbed for hand sample analysis and thin sectioned for petro- gaphic study. Twenty-two thin sections were point counted (95-158 points per thin section; n = 2882 total points) and categorized as algal fila­ments, matrix groundmass, shell fragments, elas­tics, or pore space (table Al in the appendix, avail­able in the online edition or from the Journal of Geology office). The relatively small size of cyano­bacteria prevented their identification in standard thin-section examination. It is possible that a cer­tain fraction of the matrix groundmass is composed of calcified cyanobacteria. The tufa composition was primarily calcium car­bonate, but in thin section alone, the difference between calcite and aragonite cannot be distin­guished. This is an important distinction because of the different water conditions required for the formation of calcite and aragonite (Kelts and Hsu 1978). X-ray diffraction was performed on nine capping tufa samples to determine calcite and aragonite mineral contents. Specific mineralogieal trends could not be established, so the results are not presented here (Felton 2003). Tufa locations at subbasin thresholds were of par­ticular interest because of the possible chemical differences between subbasins and the main basin. Chemical differences and, hence, mineralogieal dif­ferences are likely to exist if there was restricted flow between basins. In order to determine if flow was restricted between subbasins, lake surface ar­eas of subbasins were measured using ESRI Are- View software, and constriction cross-sectional area was calculated from digital topographic maps. Lake surface area was compared to constriction cross-sectional area for each subbasin threshold. A large ratio of subbasin to cross-sectional area of the constriction suggests possible chemical difference between the subbasin and the main body of water. Landform Analysis. The usefulness of process- based landform analyses in interpretation of paleo- lake dynamics and sediment transport analysis has been known for considerable time (e.g., Gilbert 1890; Adams and Wesnousky 1998). Landform analysis is used to determine local depositional controls on tufa development. For this study, aerial photos and topographic maps were studied from the entire Bonneville basin to locate small (~5 km2) Lake Bonneville islands where shore processes could be studied from a 360° perspective (fig. 3). At the 10 field locations, aerial photo interpretation and geomorphic mapping were used to understand the spatial relationships between erosional and de­positional regimes (table 1). All available paleowave energy indicators (described below) were mapped at the 10 field locations. Erosional regimes are characterized by exposed bedrock and tufa formation. Bedrock is eroded and exposed by strong and consistent wave energy. Waves, driven by wind, remove and transport sed­iments from a section of shoreline. The Bonneville basin contains many examples of bedrock exposure on the shoreline. One of these examples is Lucin Hill, where on the eastern side of what was an is­land in the lake, a majority of shorelines have ex­posed bedrock. On the opposite western side, a ma­jority of surfaces are covered with lacustrine sediments, and very little bedrock is exposed (fig. 4). Depositional regimes are characterized by sedi­ment accumulation and constructional landforms such as spits, beach ridges, and tombolos (table 2).Table 1. Summary of Tufa Localities Locality Latitude and longitude Lake level Tufa type Shoreline character3 Percentage of slope on shoreface [+1-5%) Geomorphic setting Northwestern part of the basin: Lucin Hill (west) 41°19'57"N/ 113°54'37"W Provo Both Both 19 Lee side of island/tombolo Lucin Hill (east) 41°19'58"N/ 113°54'25"W Provo Capping tufa Erosional 39 Windward side of island Lion Mountain (west) 41°16'46"N/ 113°55'30"W Provo Beachrock Depositional 6 Lee side of island/tombolo Lion Mountain (east) 41°16'43"N/ 113°54'23"W Provo Capping tufa Erosional 40 Windward side of island Hogup Mountains (north end) 41°35'43"N, 113°11'08"W Provo Both Depositional 18 Spit Western part of the basin: Wendover Knob 40°44'22"N, 114°05'54"W Provo Capping tufa Erosional 27 Windward side of peninsula Volcano Peak (west) 40°47T9"N, 113°59'31"W Provo Both Erosional 45 Windward exposed headland Volcano Peak (east) 40°47T7"N, 113°58'45"W Provo Both Erosional 24 Windward exposed headland Silver Island tombolo 40°53'26"N, 113°52'58"W 4500 ft Both Deposional 4 Tombolo Southern part of the basin: Tabernacle Hill 38°55'20"N, 112°31'40"W Provo Capping tufa Both 10 Synprovo volcanic cone Table Mountain 39°56'54"N, 112°53'46"W Provo Both Both 20 Windward peninsula Black Rock Volcano 38°48'26"N, 112°29'09"W 4900 ft Both Both 20 Volcanic cone Fish Springs Range (north end) 39°52'29"N, 113°25'45"W Provo Both Both 35 Windward peninsula/headland Sand Pass (south side) 39°37T4"N, 113°24'05"W Provo Both Depositional 10 Pass between subbasins Eastern part of the basin: Oquirrh Mountains (north end) 40°43'25"N, 112°13T0"W Provo Both Both 60 Exposed headland " "Both" means there are erosional and depositional portions of the shoreline.Journal of Geology TUFA DEVELOPMENT 381 Table 2. Summary of Field Criteria for Understanding the Erosional and Depositional Regimes of Lake Bonneville Process Field criteria Indicators Erosional Depositional 50%-100% of surfaces with bedrock exposed, 50%-100% of surfaces with tufa coverage Tombolos, spits, beach ridges Bedrock exposure, tufa development Landform orientation Deposition of sediment occurs when sediments are transported from areas of high energy to areas of low energy (King 1972). Erosional shorelines occur where wave energies are sufficient to remove and transport material and are commonly found on windward slopes. Likewise, where wave energies are reduced or convergent, depositional shorelines develop. Depositional shorelines typically occur in conjunction with the lee sides of slopes. Tombolos, beach ridges, and other constructional shorelines form in areas of decreased wave and wind energy (fig. 3). Wave energies impinge on an island and dissipate while being refracted around the island or headland. This creates an eddylike environment on the lee side of an island, where deposition of sed­iment occurs (Zenkovich 1967; King 1972). The lo­cations, extents, and trends of spits, beach ridges, and tombolos associated with the islands were re­corded as field data (table 2). Results Tufa Classification. Tufa deposits occur on var­ious slope profiles and geomorphic settings (table 1). Tufa deposits are located on the basinward edges of Provo shore-zone benches at ~1460 m elevation. Generally, tufa is observed 1 m above to 10 m below mean shoreline bench surface elevation. Distribu­tion of tufa is most common on the break in slope from bench to basin (fig. 5). Three forms of tufa deposits are present in the Bonneville basin: cap­ping tufa, beachrock, and capping tufa over beach- rock (fig. 6). Capping Tufa. Capping tufa is a grayish white, porous foreshore facies that coats exposed bedrock and solidified beachrock (fig. 6 A). Deposits are com­monly 0.2-0.5 m in thickness, with a maximum thickness of 1 m and a minimum tufa film thick­ness of 2 mm (fig. 7 A). Capping tufa is calcium carbonate containing less than 10% clastic material (generally quartz, volcanic sand, and clay). Capping tufa has an average porosity of 18%. Porosity in­creases from the inner portion of the tufa cap to the outer portion (fig. IB). Tufa caps are massive and concentrically accreted around a central nu­cleus or horizontally laminated on a planar surface. Capping tufa also takes pendulous or draping forms (fig. 1C). Contact surfaces between bedrock and tufa are sharp. Encrusting limestones, cherts, vol- canics, and quartzites, capping tufa developed on any appropriately stable substrate and is common on headlands that were exposed to unrestricted lake wave energy. Thin-section counting of 95-158 points for each of 22 samples reveals on average 45% algal fila­ments, 33% micrite groundmass, 18% porosity, 2.5% clastic fragments, and less than 1 % other ma­terial, including shell fragments (table Al). Algal filaments range in diameter from 0.5 to 3 mm. Thin sections show an opaque dirty-brown micrite groundmass in plain light. Micrite is commonly banded between algal filaments (fig. ID) at a scale that was not visible macroscopically. Algae genera and species were not identified. Figure 3. Schcmatic example of wave energy orienting landforms at Lucin Hill. In this case, a tombolo is formed in the lee of an island, which blocks wave energy and allows a tombolo to form. The windward side of the is­land shows erosional characteristics as it absorbs the brunt of wave energy.382 A. F E L TO N ET AL. W E Figure 4. Wave direction indicator of exposed bedrock on Lucin Hill ridge crest looking north, where there is much exposed bedrock on the eastern surface, compared to little bedrock exposure on the western aspect. The east side has erosional character, whereas the west side has depositional character. Camera elevation is 1430 m. X-ray diffraction analysis of seven samples from erosional regimes shows mineralogy of >70% cal- citc and <20% aragonite. Tufas from two areas with depositional character rather than erosional char­acter, Sand Pass and Silver Island tombolo, have >50% aragonite mineralogy (Felton 2003). Beachiock. Bcachrock is a clast-supported cal­cium carbonate deposit (fig. 6B). Bcachrock con­tains >90% clasts, which arc cemented by tufa (fig. 8A). Bcachrock contains clasts of varying size. Sandy bcachrock (0.06-2 mm), pebble bcachrock (2-64 mm), cobble bcachrock (64-256 mm), and boulder bcachrock (>256 mm) arc distinguishable bcachrock fades and refer to the size of the majority of clasts cemented by tufa. Bcachrock fades arc commonly well sorted, which is a consistent fea­ture of high-energy beach foreshore fades. Sandy bcachrock is present in cmbaymcnts and sheltered sides of islands. Cobble and boulder bcachrock oc­curs in reentrant portions of cmbaymcnts and ex­posed headlands. Cobble and boulder bcachrock consists of locally derived clasts. Outcrops vary in thickness from a 1-cm hard ground to a massive 2­m wall of bcachrock. Capping Tufa over Beachrock. At six of 13 field locations, an encrusting tufa deposit, void of clasts, caps a cemented bcachrock (fig. 6C). In this deposit, bcachrock up to 1-2 m thick can bc overlain by up to 0.5 m of capping tufa (fig. 8B). This sequence of capping tufa overlying bcachrock is well defined at the Volcano Peak field locality near Wcndovcr, Ne­vada, where a palcobcach surface with a tufa cap is present (fig. 1). Capping tufa over bcachrock oc­curs in transitional sediment transport regimes, where sediment supply diminishes, and tufa de­velopment is encouraged by the reduction of clastic input. Capping tufa over bcachrock is present in areas that arc transitional between depositional and erosional. The Role of Fetch across Lake Bonneville. Fetch is the distance over water that wind can travel and build wave energy. Bccausc longer fetch distances allow greater wave energy to build, open water ex­posed to long fetch will produce larger waves. The field areas in this study were generally exposed to waters with fetch distances ranging from 32 to 174 km. Fetch distances correspond to interpreted wave directions for cach field area (table A2 in the ap­pendix, available in the online edition or from the Journal of Geology office). The longest fetch direc­tion at cach field area (table A2) is plotted against the field interpretation of predominant energy di­rection (fig. 9). Wave direction field interpretations correlated with directions of greatest fetch and yielded an R2 of 0.89. This implies that the large fetch distances on Lake Bonneville produced large waves that oriented landforms, exposed bedrock, and aided tufa development. Underlying geologic structures and preexisting topography probably also played a critical role in bedrock exposure. Where strata dip steeply shorc- <---------Landward Basinward---------> Figure 5. Schematic drawing of tufa locations on Provo shoreline benches. Tufa commonly occurs in patches on the outer edges of benches.Journal of Geology TUFA D E V E L O PME N T 383 A) Capping tufa encrusting bedrock C) Capping tufa overlying beachrock Figure 6. Three major forms of tufa development in the Bonneville basin. A, Encrusting tufa, thickness from 2 mm to 1 m, commonly occurs as a crust on exposed bedrock surfaces. B, Coated clasts (beachrock). Clast sizes range from sand to boulder, exposed in thicknesses up to 2 m. C, Beachrock with a tufa cap. ward, bcdrock is more easily exposed to lake en­ergy. This exposed bcdrock is then a good substrate on which tufa can form. As a result, tufa formation is more common on headlands with steeply dipping strata. Discussion Tufas arc unique features of ambicnt-tcmpcraturc waters of western U.S. pluvial lakes such as Pleis­tocene Lake Bonneville and Lake Lahontan. In the Bonneville basin, calcium carbonate deposits arc useful tools for recognizing areas of high hydro­dynamic energy. A depositional model for tufa de­velopment is presented here that incorporates the interpretation of a tufa continuum as well as local and basinwidc controls on tufa formation. The two end members of beachrock and capping tufa bracket a continuum of shorc-zonc calcium carbonate deposits (fig. 10A). Tufa deposits in the Bonneville basin arc present in any ratio along the continuum. The continuum has temporal and spa­tial elements. The spatial clement occurs when fa­des along a shoreline change from capping tufa to beachrock. This change can take place because of spatial changes in slope angle or wave energy. Cap­ping tufa overlying beachrock adds a temporal di­mension to the tufa continuum because it is a tran­sitional facics occurring when a shorc-zonc system stops depositing beachrock and starts depositing capping tufa. This transition may be caused by a change in the amount of sediment supply or wave energy. The capping tufa over beachrock deposit would be illustrated by a system moving left to right on the continuum over time (fig. 10). When the Provo level lake initially occupied the Provo shore zone, it is suggested that there was excess sediment available as a result of wave action and reworking of highstand material and that these fac­tors caused beachrock formation. As the stillstand progressed, it is possible that in some places, avail­able sediment supply diminished and capping tufa was able to form on top of the stable substrate cre­ated by the beachrock. Local Depositional Controls Field mapping and laboratory analysis of tufa from around the Bonneville basin provides the basis for interpretation of factors that control tufa devel­opment. These factors include local controls (in the shore zone of the particular field locality) as well as largcr-scalc, basinwidc controls based on spatial distributions of tufas around the Lake Bonneville basin. Chemical Analysis. X-ray diffraction analysis shows predominantly calcitc mineralogy for cap­ping tufa samples. There arc many factors control­ling which carbonate phase precipitates in a lake, including the pH (Friedman 1978; Kelts and Hsu 1978) and the magnesium/calcium (Mg/Ca) ratio of the water. Assuming calcitc is a primary product of deposition and not a secondary replacement of original aragonite, partial pressure of C02 in shorc- zonc waters was reduced enough to precipitate cal­citc at the Provo level of Lake Bonneville but not low enough to precipitate aragonite. This suggests local shorc-zonc waters with a pH of ~8, sufficient to precipitate calcitc (pH > 7) but not high enough384 A. F E L TO N ET AL. A. B. Figure 7. A, Capping tufa adheres to chert bedrock at the Lucin Hill locality. Notice coating nature of these porous tufas. B, Detail of capping tufa facies at Table Mountain with banded and porous tufas. Tufa shows increased porosity toward the outer edge of the deposit. C, Pendulous form of capping tufa coating quartzite at Table Mountain. D, Photomicrograph of tufa showing prominent algal filaments (A) in the middle and upper middle of the photograph. Light areas are pore space [P). to precipitate aragonite (pH > 9; Friedman 1978). Calcite as the primary precipitate suggests a low (<2) Mg/Ca ratio in the shore-zone waters (Kelts and Hsu 1978) of Lake Bonneville. Tufa Precipitation. Shore-zone lacustrine cal­cium carbonate deposition is a function of water temperature, clastic input, calcium concentration of the water body, and local water pH, all of which are influenced by biological factors and water agi­tation. Both the organic (biological factor) and the inorganic (water agitation factor) presumably af­fected tufa deposition on the Provo shoreline of Pleistocene Lake Bonneville. Biomediation, solar heating, and wave agitation (degassing) reduced the partial pressure of CO, and elevated the local pH of shore-zone waters. As pH of water increases, sol­ubility of calcium carbonate decreases, resulting in calcium carbonate precipitation: Ca2+ + 2HCO3 «- H,0 + CO, + CaC03. (1) Release of CO, in equation (1) corresponds to pre­cipitation of CaCO,. In areas where algae and cyanobacteria can flour-Journal of Geology TUFA D E V E L O PME N T 385 A. Figure 8. A, Beachrock at Tabic Mountain, with 1,5-m deposit of pebble to cobble calcium carbonate-cemented clasts overlying Cambrian shale. B, Capping tufa (71 over beachrock jB). This deposit is a transitional facies where stabilized beachrock is overlain by capping tufa. This location at Table Mountain contains boulder beachrock of locally derived quartzite clasts capped by 0.4 m of tufa. ish, such as sediment-limited regimes, local pH rises as a result of biomediation, and tufa can form. It is known that algae played a significant role in the tufa development of Lake Bonneville because thin sections contain an average of 45% algal fil­aments. In areas where there was water agitation, such as headlands with exposure to large wave en­ergies, local pH rose as a result of degassing, and tufa precipitated largely inorganically. Mapping of tufa localities reveals the most prolific tufa deposits on headlands and areas that were exposed to large fetch distances in the lake. It is herein suggested that algal growth and wave agitation together in­fluenced ambient lake chemistry to create locally massive deposits of capping tufa on steep head­lands, where bedrock was exposed and there was little sediment input. In areas where there was some wave action, a less steep slope, and plentiful sediment supplies, beachrock formed (fig. 10). Basinwide Depositional Controls Subbasin Thresholds. Subbasin thresholds are a potentially important control of tufa development because reductions of water flow when the lake level dropped may have isolated the waters of sub­basins. This may be a major factor in inducing tufa deposition (Kelts and Hsu 1978; Benson 1994). In the Bonneville basin, large deposits of tufa occur at thresholds between subbasins. All four major Provo level subbasin thresholds-Old River Bed, Sand Pass, Point of the Mountain, and Cutler Dam- contain tufa deposits (table 3). One of these divides, Sand Pass, is a predominantly depositional regime at ~4800 ft (1463 m) located on the threshold be­tween the Great Salt Lake main basin and the Tule Valley subbasin in the southwest quadrant of Lake Bonneville. Because the area is predominantly de­positional, dense capping tufa like that found at Sand Pass is not expected. In comparison to Sand Pass, the Table Mountain field site contains some of the most laterally extensive and thickest tufa deposits found in the Bonneville basin. This local­ity is near the Old River Bed, which was the con­necting conduit for water flowing from the Sevier subbasin to the Great Salt Lake subbasin of Lake Figure 9. Correlation between longest fetch direction and wave directions jn = 10) interpreted from field map­ping. This relationship illustrates the important role fetch plays in sculpting landforms in the Bonneville basin.386 A. F E L TO N ET AL. Figure 10. A, Field relationships of tufa continuum. In the Bonneville basin, shorc-zonc calcium carbonatc occurs as capping tufa, containing <10% clastic material. Tufa also occurs as the matrix of bcachrock, which consists of calcium carbonatc-coatcd clasts. The bcachrock end member contains >90% clasts. Calcium carbonatc deposits can occur in any concentration along the continuum. B, Detailed continuum of the Lake Bonneville bcachrock to the capping tufa facics. Bcachrock occurs in more depositional regimes where there is adequate sediment input, lower slope angle, and enough wave energy to orient but not transport the clastic material. Capping tufas form where there is sufficient wave energy and slope angle to transport clastic material basinward and allow tufa growth. Bonneville (Gilbert 1890; Oviatt et al. 1992). The Cutler Dam locality is a narrow canyon that carries the present-day Bear River from Cache Valley into the Great Salt Lake basin. Capping tufa coats the walls of this canyon. At Point of the Mountain, which divides Utah Valley and Great Salt Lake Valley, there is beachrock and some capping tufa along the Provo shoreline, especially on the south side of the Traverse Range. The large size of the subbasins (table 3) and the restricted flow channels imply limited mixing and unique water chemistry for the subbasins as com­pared to the main water body. It is suggested that tufa development at these sites was initiated by the isolation and chemical concentration of subbasin waters flowing through these constrictions and mixing with less concentrated waters of the main basin. Water flowing between subbasins mixes at these constrictions, encouraging calcite precipita­tion. An additional factor at the Cutler Dam lo­cality is the watershed runoff from the relatively large Bear River basin that might have contributed a constant supply of calcium to the constriction, promoting calcium carbonate precipitation. Spatial Distribution of Tufa. The Bonneville basin is flanked on the eastern edge by the Uinta and Wasatch mountain ranges, which both contain large drainage basins. The drainage basins feeding the eastern portion of the Great Salt Lake basin make up 97% of modern-day inflow, with the Bear and Weber Rivers contributing 73 % of the total (Ar- now and Stephens 1990). During the last glacial maximum, substantially increased inflow is as­sumed, but the relative size and elevation of the drainage basins and thus the percentage of water entering from the east should have remained equiv­alent to modern values. Is the spatial distribution of tufa related to the distribution of freshwater in the basin? A majority of large tufa deposits at the Provo level of Lake Bonneville do occur in the west­ern portion of the basin. This skewed distribution could be due to two factors. First, a large amount of water coming out of the Wasatch Mountains could have diluted waters enough to suppress car-Table 3. Summary of Subbasins of Lake Bonneville Constriction locality Locality Subbasins involved (subbasin -> main basin) Tufa present? Subbasin water surface area (km2) Constriction cross-sectional area (km2) Ratio of Subbasin subbasin area evaporation with to constriction evaporation of cross-sectional .5 m/yr (m3/s) area8 Old River Bed Table Mountain Sevier basin -> Great Salt Lake basin Yes 5440 .360 78.8 15,111 Sand Pass Sand Pass Tule Valley -> Great Salt Lake basin Yes 971 .0088 15 110,340 Point of the Mountain Point of the Mountain Utah Valley -> Great Salt Lake basin Yes 1230 .298 19 4127 Cutler Dam Cutler Dam Cache Valley -> Great Salt Lake basin Yes 1036 .056 16 18,500 " Subbasin surface areas compared with constriction cross-sectional areas.388 A. F E L TO N ET AL. Figure 11. Lake Bonneville at highstand 15,000 yr ago with major river inputs. Note majority of freshwater in­put is from the eastern half of the basin. This dilution could hinder tufa development in the eastern half of the basin, where larger drainage basins contributed a major­ity of the water to the lake. BR = Bear River, WR = Weber River, PR = Provo River, SR = Sevier River. bonate formation. Second, large amounts of sedi­ment supply produced by rivers and glaciers in the Wasatch Mountains could have overwhelmed any potential tufa development with clastic material (fig. 11). We suggest that these two factors contrib­uted to the lesser amounts of tufa seen in the east­ern portion of the basin. Conclusions Three major forms of tufa occur in the Bonneville basin and can be categorized along a continuum with respect to clastic material entrained in the calcium carbonate deposit. Locally, tufas are prev­alent on headlands and windward sides of islands that were exposed to high wave energy and con­tained a solid substrate. Algal growth and wave ac­tion degassing played a substantial role in the de­velopment of tufa deposits around the Bonneville basin. Calcitc mineralogy, rather than aragonite, in­dicates possible localized shore-zone elevated wa­ter pH and Mg/Ca ratio <2. In terms of basinwidc controls, tufa commonly occurs at basin thresholds, where water is moving between a restricted subbasin and the main body of the lake. Tufa deposition may also be limited by the freshwater and sediment input on the eastern side of the Bonneville basin, resulting in a majority of tufa occurring in the western basin. Identification of tufa deposits in the Bonneville basin demonstrates the importance of paleowater chemistry, wave action, and basin threshold con­trols. Lacustrine basins are widely recognized as valuable paleoclimate records in Earth history. Al­though tufa deposits have been typically over­looked even in the Lake Bonneville basin, this study shows the potential for integrating tufa dis­tribution, mineralogy, lake morphometry, and hy­drodynamic characteristics in interpreting pluvial lake systems. ACKNOWLEDGMENTS H. Godsey, I. Schofield, and M. Gregory assisted in the field and added insightful reviews of this article. Three anonymous reviewers improved the manu­script considerably. This work was supported by Utah Kennecott Copper Corporation and National Science Foundation grants SBR-9817777 and EAR- 9809241. REFERENCES CITED Adams, K. D., and Wesnousky, S. G. 1998. Shoreline pro­cesses and the age of the Lake Lahontan highstand in the Jessup embayment, Nevada. Geol. Soc. Am. Bull. 110:1318-1332. Arnow, T., and Stephens, D. 1990. Hydrologic charac­teristics of Great Salt Lake, Utah 1847-1986. U.S. Geological Survey Water Supply Paper 2332, 32 p. Benson, L. H. 1994. Carbonate deposition, Pyramid Lake subbasin, Nevada. I. Sequence of formation and ele- vational distribution of carbonate deposits (tufas). Pa- laeogeogr. Palaeoclimatol. Palaeoecol. 109:55-87. Benson, L. H.; Kashgarian, M.; and Rubin, M. 1995. Car­bonate deposition, Pyramid Lake subbasin, Nevada. II. Lake levels and polar jet stream positions recon­structed from radiocarbon ages and elevations of car­bonates (tufas) deposited in the Lahontan basin. Pa- laeogeogr. Palaeoclimatol. Palaeoecol. 117:1-30. Cloud, P., and Lajoie, K. R. 1980. Calcite-impregnatedJournal of Geology TUFA DEVELOPMENT 389 defluidization structures in littoral sands of Mono Lake, California. Science 210:1009-1012. Currey, D. R. 1990. Quaternary paleolakes in the evo­lution of semidesert basins with special emphasis on Lake Bonneville and the Great Basin, U.S.A. Palaeo- geogr. Palaeoelimatol. Palaeoeeol. 76:189-214. Felton, A. 2003. Paleowave indicators and a model for tufa development in the Lake Bonneville Basin, Utah. MS thesis, University of Utah, Salt Lake City. Ford, T. D., and Pedley, H. M. 1996. A review of tufa and travertine deposits of the world. Earth Sci. Rev. 41: 117-175. Friedman, G. M. 1978. Principles of sedimentology. New York, Wiley, 791 p. Gilbert, G. K. 1890. Lake Bonneville. U.S. Geological Sur­vey Monograph 1, 438 p. Godsey, H. S.; Currey, D. R.; and Chan, M. A. 2005. New evidence for a late pluvial event and links to global climate trends from lake Bonneville, Utah. Quat. Res. 63:212-223. Hart, W. S.; Quade, J.; Madsen, D. B.; Kaufman, D. S.; and Oviatt, C. G. 2004. The s?Sr/86Sr ratios of lacus­trine carbonates and lake-level history of the Bonne­ville paleolake system. Geol. Soc. Am. Bull. 116:1107­1119. Kelts, K., and Hsu, K. J. 1978. Freshwater carbonate sed­imentation. In Lerman, A., ed. Lakes: chemistry, phys­ics, geology. New York, Springer, p. 295-323. King, C. 1878. Systematic geology: U.S. geological ex­plorations of the fortieth parallel (Vol. 1). Washington, DC, Government Printing Office, 579 p. King, C. A. 1972. Beaches and coasts (2nd ed.). London, Arnold, 570 p. Merz-Preifi, M. 2000. Calcification in cyanobacteria. In Riding, R. E., and Awramik, S. M., eds. Microbial sed­iments. Berlin, Springer, p. 50-56. Nelson, S. T.; Wood, J. M.; Mayo, A. M.; Tingey, D. G.; and Eggett, D. 2005. Shoreline tufa and tufa conglom­erate from Pleistocene Lake Bonneville, Utah, USA: stable isotope and mineralogical records of lake con­ditions, processes, and climate. J. Quat. Sci. 20:3-19. Oviatt, C. G. 1997. Lake Bonneville fluctuations and global climate change. Geology 25:155-158. Oviatt, C. G.; Currey, D. R.; and Sack, D. 1992. Radio­carbon chronology of Lake Bonneville, eastern Great Basin. Palaeogeogr. Palaeoelimatol. Palaeoeeol. 99: 225-241. Pedley, H. M. 2000. Ambient temperature freshwater mi­crobial tufas. In Riding, R. E., and Awramik, S. M., eds. Microbial sediments. Berlin, Springer, p. 179-186. Sack, D. 1999. The composite nature of the Provo level of Lake Bonneville, Great Basin, western North Amer­ica. Quat. Res. 52:316-327. Riding, R. 2000. Microbial carbonates: the geological rec­ord of calcified bacterial-algal mats and biofilms. Sed­imentology 47:179-214. Wambeam, T. 2001. Modeling Lake Bonneville basin morphometry using digital elevation models. MS the­sis, University of Utah, Salt Lake City, 60 p. Zenkovich, V. P. 1967. Processes of coastal development. New York, Wiley, 738 p. Zhang, D. D.; Zhang, Y.; Zhu, A.; and Cheng, X. 2001. Physical mechanisms of river waterfall tufa (traver­tine) formation. J. Sediment. Res. 71:205-216.