Geological Society of America Bulletin

Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Geological Society of America Bulletin

Particle-size fractionation of eolian sand along the Sinai−Negev erg of Egypt and Israel

Joel Roskin, Itzhak Katra and Dan G. Blumberg

Geological Society of America Bulletin 2014;126, no. 1-2;47-65 doi: 10.1130/B30811.1

Email alerting services click www.gsapubs.org/cgi/alerts to receive free e-mail alerts when new articles cite this article Subscribe click www.gsapubs.org/subscriptions/ to subscribe to Geological Society of America Bulletin Permission request click http://www.geosociety.org/pubs/copyrt.htm#gsa to contact GSA Copyright not claimed on content prepared wholly by U.S. government employees within scope of their employment. Individual scientists are hereby granted permission, without fees or further requests to GSA, to use a single figure, a single table, and/or a brief paragraph of text in subsequent works and to make unlimited copies of items in GSA's journals for noncommercial use in classrooms to further education and science. This file may not be posted to any Web site, but authors may post the abstracts only of their articles on their own or their organization's Web site providing the posting includes a reference to the article's full citation. GSA provides this and other forums for the presentation of diverse opinions and positions by scientists worldwide, regardless of their race, citizenship, gender, religion, or political viewpoint. Opinions presented in this publication do not reflect official positions of the Society.

Notes

Particle-size fractionation of eolian sand along the Sinai–Negev erg of Egypt and Israel

Joel Roskin†, Itzhak Katra, and Dan G. Blumberg Department of Geography and Environmental Development, Ben-Gurion University of the Negev, P.O. Box 653, Beer-Sheva, 84105, Israel

ABSTRACT of the very fi ne sand component within the and soils in the central United States (Olson dunes and probably further downwind. Our et al., 1997) and northern Africa (McTainsh, Eolian sand fractions along the west-east results suggest that particle-size distribution 1984) but has not been fully described for transport system of the northern Sinai Pen- can elucidate much about erg history over dunes. At a smaller temporal and spatial scale, insula–northwestern Negev erg of Egypt and time scales of a glacial-interglacial cycle, es- sand plumes of several hundred meters extend- Israel were analyzed in this study with re- pecially in cases where the sand provenance ing downwind of an abandoned sandy agricul- gard to source, dune geomorphology, eolian is of a single dominant source. tural fi eld in the Mojave Desert were found to transport, and paleoclimate. The studied erg exhibit fractionation, with smaller effective par- is composed of active linear (seif) dunes in INTRODUCTION ticle sizes toward the toe of the plume (Okin and northern Sinai (western part), and stabilized Painter, 2004). vegetated linear dunes in the NW Negev dune Particle-Size Changes in Eolian Deposits Particle-size-distribution changes in sand fi eld (eastern part). Linear seif dunes differ along a transport path of mobilized dunes can from vegetated linear dunes in their vegeta- Particle-size distributions of dune sands are result from eolian abrasion of sand grains due tion cover, linearity, internal structure, and an important factor in understanding the mor- to grain collision and erosion of the substrate dynamics. Sand samples were analyzed for phology and dynamic processes of dunes (e.g., underlying the dunes, as well as winnowing sand-grain morphology and particle-size dis- Bagnold, 1937; Tsoar, 1986; Lancaster, 1995; out of coarser eolian particle sizes. The relative tribution. Although both dune types are con- Pye and Tsoar, 2009). Theoretical and empiri- importance of these different processes is still tinuous landforms with similar orientations cal studies have shown that particle-size distri- unclear for most dune bodies. Theoretically, and sand-grain roundness values, the linear butions of eolian sand deposits change along given a particular amount of wind power, the dunes of Sinai are coarser grained than the their transport paths (McLaren, 1981; McLaren particle-size distribution of loose source sedi- Negev vegetated linear dunes. The vegetated and Bowles, 1985). Some dune fi elds, com- ment, which includes sand, silt, and clay, should linear dunes have a variable but higher proposed mainly of parabolic dunes, are reported be differentiated (fractionated) along the trans- portion of very fi ne sand (50–125 μm) con- to undergo a gradual and slight decrease in port pathway as different particle-size fractions tent and a varying but lower sand fi ning ratio particle-size means with downwind distance, have different threshold values for entrainment (defi ned as the ratio of fi ne sand percentage interpreted to be partially due to winnowing of and transport. Silt and clay particles (<50 μm), to very fi ne sand percentage). From these ob- coarse sands (Muhs et al., 1999; Muhs and Hol- carried primarily in suspension (Tsoar and Pye, servations, we infer that fractionation of sand liday, 2001). Particle-size distribution modes 1987), should attain transport distances well occurred along the studied eolian transport have been mapped at various scales for active beyond the region of sand and dune mobiliza- path during periods of enhanced windiness. linear dune bodies, but signifi cant trends have tion (Pye and Tsoar, 1987; Pye, 1995) in accor- Very fi ne sands are suggested to have been not been identifi ed (Lancaster, 1995, and refer- dance to the duration and velocity of the winds. transported by saltation and low suspension ences within). from source deposits and sand sheets. We Particle-size fractionation acts on sediment in Very Fine Sand suggest that a signifi cant proportion of the transport that undergoes selective deposition due very fi ne sand fraction of Nile Delta sands to decreasing energy of the transporting process. Unimodal particle-size distributions of 125– has been transported downwind through As the transport energy decreases, coarser sedi- 250 μm modes (fi ne sand) are common in active northern Sinai during the late Pleistocene, ment has a greater probability of being depos- dune sands, as this particle size is optimal for especially when linear dunes reached the NW ited than fi ner particle sizes (McLaren, 1981). transport by saltation. However, the transport Negev due to last-glacial period windiness Particle-size fractionation of sand and silt has dynamics of very fi ne sand are different than and probably larger sediment supply. Gener- been shown for different arid environments, fi ne sand. Very fi ne sand, being dynamically ally decreasing wind velocities and increas- mainly with regard to accreting soils and eolian transitional between saltation and low, modi- ing precipitation along the west-east dune mantles that are formed by a mixture of sus- fi ed suspension in strong storms (Bagnold, transport path enhanced vegetative cover in pended and saltating particles (McTainsh, 1984; 1941; Chepil, 1951; Pye and Tsoar, 1987; Xu the northern Negev and enabled deposition Holliday, 1989). The process of fractionation of et al., 2007), makes this size fraction poten- sand-grain fractions along transport paths has tially and more uniquely sensitive to current and †E-mail: [email protected] been reported for fi ner-grained eolian deposits past changes in wind velocities. Very fi ne sand

GSA Bulletin; January/February 2014; v. 126; no. 1/2; p. 47–65; doi:10.1130/B30811.1; 13 ﬁ gures; 2 tables.

For permission to copy, contact [email protected] 47 © 2013 Geological Society of America Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Roskin et al. transport by suspension over longer distances Study Outline and Goals et al., 2001; Mohamed, 2012). At Gebel Maghara is inferred from small amounts of very fi ne in the west, mean annual rainfall is ~50 mm sand, interpreted to be eolian, found in offshore The study area, the northern Sinai Peninsula– (Ahmed, 2010), and, along the Mediterranean cores of the Mediterranean Sea (Hamann et al., northwestern Negev erg (Sinai-Negev erg), strad- coast, annual averages are 67 mm at Port Said 2008; Mulitza et al., 2010), on Mediterranean dles the Egypt-Israel border (Figs. 1 and 2). The (Tsoar, 1995), 104 mm at El-Arish, and 200 mm islands (Sevink and Kummer, 1984), and off studied erg is composed of active linear (seif) at Rafah (Abdel Galil et al., 2000) (Fig. 1B). the North African coast (McGee et al., 2013). dunes in northern Sinai (western part) and sta- The northern Sinai dune fi eld east of the Nile The very fi ne sand fraction, defi ned here as the bilized vegetated linear dunes in the NW Negev Delta is spatially continuous over substantial range of 50–125 μm, is a slight extension of the dune fi eld (eastern part) of Israel. The Sinai areas, with dune heights exceeding 30 m (Gad, Krumbein phi (φ) scale for the very fi ne sand dune sands are the immediate source of the 2004). The Sinai sands are mainly sparsely veg- fraction of 63–125 μm, and includes the tail of Negev dunes. etated to bare and uncrusted seif linear dunes (coarse) silt (50–63 μm). The 50 μm cutoff is The Sinai-Negev erg is an ideal “fi eld labora- (Tsoar, 1989; Abdel Galil et al., 2000; Rabie in accordance with the lower cutoff of the U.S. tory” to study eolian particle-size trends along a et al., 2000), and complex braided linear dunes Department of Agriculture (USDA) fi ne sand sand transport system for several reasons. The erg (Tsoar, 1989, 1995) (Figs. 3A and 3B). Com- fraction defi nition and the fact that typical loess is relatively small, confi ned, and young (Roskin plex braided linear dunes characterized by low usually has particles <50 μm (Smalley and Vita- et al., 2011a). The eastern part of the Nile Delta heights, compared with the active linear dunes, Finzi, 1968; Junge, 1979, and references within; is artifi cially separated from the overlapping are mainly found to the east of Wadi Al-Arish, Pye, 1995). The very fi ne sand fraction cen- northwestern section of the northern Sinai dune where the annual rainfall is somewhat higher ters around 80–100 μm, being the particle-size fi eld by the Suez Canal. While the past source (Fig. 1B). Other than in southwest Sinai, where range that has the lowest critical drag velocity deposits of the Sinai sands are not distinctly dune crests extend to the east-southeast (Tsoar needed to initiate particle mobilization (Bag- identifi ed, the Nile Delta is the only reasonable et al., 2004), the dunes extend in a general west- nold, 1941; Chepil, 1951; Pye and Tsoar, 1987). candidate for the past source since the sands east orientation west of the Egypt-Israel border. Impact threshold velocities for fi ne particles are of the delta and the erg are similar mineralogi- This observation is important in ascertaining the much lower than fl uid threshold shear velocities cally, and both are geochemically mature, sug- past direction and upwind Nile Delta source of (Pye and Tsoar, 2009). Thus, even modest wind gesting a sedimentological link between them the sands. velocities in an active dune fi eld can keep very (Muhs et al., 2013). The erg lies to the north and Current annual seif dune elongation rates are fi ne sand and coarse silt particles in motion, downslope of a series of ridges and highlands in accordance with wind data. Strong south- such that cumulative transport distances could of mainly carbonate strata, which reduce pos- eastern eolian sand transport drift potentials amount to hundreds of kilometers (Muhs et al., sibilities of additional sources of quartz grains (DP; defi ned by Fryberger, 1979) (DP = 1139) 2008). These factors can cause particle-size to the erg as initially suggested by Crouvi et al. were calculated for the years 1987–1993 from fractionation of sand. Nevertheless, the scale in (2008) and studied by Muhs et al. (2013). Along meteorological data from the Port Said airfi eld, time and space of very fi ne sand transport and the central and main sand transport corridor of Egypt, at the northeast edge of the Nile Delta deposition is highly variable. the erg, sand-grain redness intensity attributed to (Fig. 1B). The measured winds had annual Recent studies in mid- and low-latitude dune reddish iron-oxide coatings does not vary signifi - monthly averages of 7.4–10 m/s between the fi elds report substantial but differing amounts of cantly (Roskin et al., 2012), further suggesting a years 1989 and 1999 (Roskin et al., 2011a, and very fi ne sand, silts, and clays, mainly in stable lack of additional and secondary sand sources. references within). These winds may explain vegetated linear dunes, though the signifi cance By analyzing particle-size distributions and the occurrence of nonvegetated linear dunes in of the very fi ne sand fraction is not discussed. grain morphologies of northern Sinai sands the western part of the Sinai dune fi eld south Following the recent success of improved meth- and sands from optically stimulated lumines- of Port Said, which are currently elongating to ods for coring dry dune sands (Munyikwa et al., cence (OSL) dated dune profi les of the NW the SSE by several meters per year (Tsoar et al., 2011) and the proliferation of laser diffraction Negev (from Roskin et al., 2011a), we identify 2004). Maximum dune migration rates of 27.3 techniques to measure particle-size distribu- progressive sedimentological changes in parti- m/yr have recently been reported in the south- tions, particle-size analysis along the full profi le cle-size fractions along sand and dune transport western part of the dune fi eld (Hermas et al., of linear dunes has been reported (e.g., Stone corridors. We further explore spatial sedimen- 2012). Wind data from central and eastern parts and Thomas, 2008; Fitzsimmons et al., 2009). tological trends in an effort to understand the of northern Sinai also indicate that wind power Particle-size distributions of full dune profi les source, transport modes, and formational con- decreases to the east toward the Negev. Annual in the Strzelecki and Tirari Deserts of central trols of dunes and downwind eolian deposits elongation rates, usually in the range of 2–15 m Australia showed abundant fi ne sands and silts, in the northern Negev. Specifi cally, we test the (Misak and Draz, 1997; Tsoar, 1989; Tsoar although silt-sized material is most likely to hypothesis of sand particle-size fractionation et al., 2004), generally decrease from west to have been deposited as secondary long-range– and fi ning down the transport pathway of lin- east. Nevertheless, limited sand is transported transported dust (Fitzsimmons et al., 2009). ear dunes, and the association with past regional into the NW Negev dune fi eld (Allgaier, 2008). Stone and Thomas (2008) showed that Kalahari paleoclimates and paleoclimate change. These data and the continuous availability of dune sand particle sizes have modes mainly sand supply imply that today the dynamics within the 150–500 μm range, while some cores STUDY AREA of Sinai sand movement are controlled chiefl y show “coarse silty fi ne sand layers” at their by wind strength. basal parts. However, the very fi ne sand fraction Northern Sinai Dune Field Minor sources of eolian sediments in the has not been analyzed in light of the different dune fi eld environs are coastal dunes and Wadi dune-building dynamics, namely, those between The northern Sinai dune fi eld runs between Al-Arish (Fig. 1A). A 2–4-km-wide strip of active sinuous-shaped linear dunes and veg- hyperarid conditions in the southwest to arid transverse and parabolic dunes along the Sinai etated linear dunes. conditions in its northeast corner (Abou Rayan coast is partially separated from the inland

48 Geological Society of America Bulletin, January/February 2014 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Particle-size fractionation of eolian sand

Figure 1. The study region and area. (A) The distribution of major eolian deposits in the Nile Delta, northern Sinai Peninsula, and in the central and northern Negev Desert, Israel. Note that the eastern parts of the Nile Delta are also covered by sand and dunes (modifi ed after Muhs et al., 2013). (B) Principal dune forms of the Sinai-Negev erg and coastal sands of Israel and (current) sand wind roses. Gen- eralized annual rainfall isohyets for the Sinai Peninsula are after Abou Rayan et al. (2001) and Milewski et al. (2009). The gray- shaded eastern part of the erg is the northwestern Negev dune fi eld, enlarged in Figure 2. The sampling sites are described in Table 1 (modifi ed after Muhs et al., 2013). LGM—Last Gla- cial Maximum.

Sinai dunes. The coastal dunes exhibit a lighter Northwestern Negev Dune Field along a desert fringe between the climatic zones color and a coarser grain size then the inland of the Mediterranean Levant and the subtropi- dunes of Sinai (Tsoar, 1976). These dunes, as The NW Negev dune fi eld (30°N, 34°E) cov- cal desert. The dune fi eld, situated along the with the coastal dunes of Israel (Levin et al., ers ~1300 km2, is of a triangular shape, and, southern part of the wintertime cyclonic tracks 2006, and references within), are associated at its eastern apex, attains a maximum breadth of the Mediterranean Cyprus Low, receives with a historical incursion ca. 1 ka. Wadi Al- of 55 km. The Negev dune fi eld is classifi ed ~150 mm of annual rainfall in the north but only Arish (Fig. 1B), the only wadi that dissects into three dune encroachment (incursion) cor- 80 mm in the south (Fig. 1B). The current wind the Sinai dune fi eld, has been found to be a ridors that differ in length, dune morphology power of the Negev is lower than that of Sinai, negligible source for quartz dune sand (Muhs and spacing, and sand transport rates (Roskin as indicated in the sand wind roses (Fig. 1B). et al., 2013). et al., 2011a) (Fig. 1B). The dune fi eld runs Winds typically encountered today do not cause

Geological Society of America Bulletin, January/February 2014 49 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Roskin et al.

Figure 2. Northwestern Negev dune fi eld development stages, dune encroachment corridors, and optically stimulated luminescence (OSL)–dated logs analyzed for particle-size distribution. Ages are in ka (modifi ed from Roskin et al., 2011a). VLD—vegetated linear dune; ID—interdune. substantial vegetated linear dune elongation nor views linking northern Negev loess sources Paleoclimate of the Sinai-Negev Erg do they rework the dune crest beneath depths of solely to dust entrained from the distant ~3 m (Roskin et al., 2011a). upwind Sahara Desert (Yaalon and Dan, 1974; The paleoclimate that enabled the Negev In contrast to the Sinai dunes, the NW Negev Yaalon, 1987; Ganor et al., 1991) and more dune fi eld development has traditionally been dune fi eld is dominated by stable vegetated proximal Wadi Al-Arish and Sinai fl oodplain interpreted to be the result of past aridity linear dunes, the vegetation cover of which deposits (Ginzburg and Yaalon, 1963; Yaalon ( Goring-Morris and Goldberg, 1990; Magaritz consists of vascular shrubs, with surface cover and Dan, 1974; Ganor et al., 1991; Hunt, and Enzel, 1990; Hunt, 1991; Harrison and Yair, ranging from 5% to 17% (Tsoar and Möller, 1991) have recently been modified. Crouvi 1998). Elsewhere, active inland desert dunes 1986; Allgaier, 2008; Siegal et al., 2013). This et al. (2008, 2009), Enzel et al. (2010), and have also been interpreted to be indicators of vegetation provides minute amounts of organic Amit et al. (2011) proposed that the coarse arid conditions (Sarnthein, 1978; Hesse and material to the upper parts of the dune section silt quartz fraction in the loess is derived from Simpson, 2006), and this paradigm is in accor- (Blume et al., 1995), which are negligible for eolian abrasion of quartz sand grains during dance with the assumption that dune mobiliza- particle-size distribution analysis. The dune the mobilization of dunes in the Sinai and tion thresholds are defi ned in part by a decrease fl anks are currently stabilized by biogenic soil Negev. Negev loess is reported to be enriched in effective precipitation. In many parts of North crusts (Danin et al., 1989; Kidron et al., 2009), in K-feldspar, plagioclase, and calcite, relative America, dunes are active in hyperarid environ- although some dune crests are active. Similar to to quartz, when compared to these minerals in ments where wind strength is low, while veg- a majority of the Sinai linear dunes, the dunes the Negev dunes. These relations may indicate etated dunes are stable in semiarid environments have long axes oriented in a general west-east that these relatively soft minerals underwent where wind strength is high (Muhs and Holli- direction (Ben-David, 2003). eolian abrasion along the Sinai-Negev erg sys- day, 1995). Recent modeling shows, however, The NW Negev dune fi eld is fl anked by tem and could be at least a partial source for that dune activity is controlled dominantly by late Quaternary loess deposits. The traditional loess in Israel (Muhs et al., 2013). wind power, and dunes can be mobilized even

50 Geological Society of America Bulletin, January/February 2014 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Particle-size fractionation of eolian sand

no reliable means to calculate paleowind speeds. A Nevertheless, with an abundant sand supply from the Nile Delta, high-velocity winds would have produced optimal conditions for initiation and growth of the Sinai-Negev erg during the last glacial period (Roskin et al., 2011a, 2011b).

Particle-Size Distributions of the Sinai and Negev Sands

B The particle-size distributions of the sands of the Sinai-Negev erg have not been analyzed in a detailed, or in a regional context, or system- atically along the transport direction of the erg’s sand. The particle-size distribution of northern Sinai dune sands, usually performed for single sites by dry sieving, are reported to have unimodal peaks in the range of 125–250 μm, typical for mobilized linear dunes (Tsoar, 1976, 1978; Goldberg, 1977; Sneh and Weissbrod, 1983; Misak and Draz, 1997). Elsewhere W E in Sinai, the 125–500 μm fraction has been reported for barchan-dominated dunes and sand ramp deposits by Gebel (Mountain ridge in Ara- bic) Maghara and Lagama (Goldberg, 1977) and linear dunes south of the city of Al-Arish (Misak C and Draz, 1997) (Fig. 1B). No distinct difference in sand particle size was found between the dune crests and fl anks (Sneh and Weissbrod, 1983), and the sands displayed meager amounts, e.g., (~5%) of silts and clays (Gad, 2004). For the NW Negev dune fi eld, diverse particle-size distributions are reported. Negev dune sands sampled from the surface are moderately rounded to well rounded and are reported to exhibit bimodal particle-size distributions, predominantly the 125–250 μm and 63–125 μm fractions. Negligible west-east downwind parti- Figure 3. Linear dune types of the Sinai-Negev erg from west to east. cle-size-change trends are interpreted to be due (A) Active linear seif dune in the central part of the Sinai dune fi eld to either varying wind intensities or provenance (courtesy of Haim Tsoar). (B) Active braided (raked) devegetated (Hunt, 1991). Roskin et al. (2011a) conducted a linear dune in northeastern Sinai (from Roskin et al., 2011b). Note comprehensive sampling program of full dune the west-east dune elongation direction and corresponding domi- sections of the NW Negev dune fi eld and used nant paleowind orientation during the late Pleistocene. Modern laser diffraction particle-size distribution analy- reactivations of superimposed northeast-facing dunelets are in ac- sis. These investigators reported the dune fi eld to cordance with current dominant southwesterly winter storm winds. consist of sand to loamy sands, though detailed (C) A stabilized vegetated linear dune (VLD) at the BM site of the analysis of particle-size fractions was not con- NW Negev dune fi eld (Fig. 2) by the border with Egypt. An active ducted. In the southwestern part of the Negev braided dune can be made out beyond the border. dune fi eld, silts and clays derived from the upper 1 m do not exceed 6%–7% (Tsoar and Möller, 1986; Blume et al., 1995, 2008) and have been in humid climates in at least some regions when driving mechanism behind transport of both del- reported to be near zero by Enzel et al. (2010). stripped from vegetation (Tsoar, 2005, 2013; taic and coastal sand inland into northwest Sinai. Yizhaq et al., 2009). Under proposed scenarios of last-glacial synop- METHODS Although there is evidence for measurable tic climatology, it is likely that there were higher- sand movement in northern Sinai at present, velocity, more frequent, west-to-east winds over Sampling there must have been times in the past when sand the area now occupied by the Sinai-Negev erg, advanced at substantially higher rates through possibly due to more frequent eastern Mediterra- In order to retrieve the longest available sand the action of powerful winds. If these winds nean cyclonic systems (Enzel et al., 2008). How- transport paths for the Negev, dune sections occurred in the past, they could have been the ever, as stated by Kocurek (1991), there is still were sampled at the western and eastern ends

Geological Society of America Bulletin, January/February 2014 51 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Roskin et al. of each dune encroachment corridor. The central the Mie scattering model (optical parameters Three frames of fi ne and very fi ne sand grains corridor, composed of several geomorphic units, RI (refractive index) = 1.52; A (absorption) = per sand sample were analyzed and photo- is the widest, longest, and thickest corridor in 0.1). Particle-size distributions of the Sinai and graphed with a DeltaPaix Invenio3SII camera the Negev dune fi eld and is the main conduit for Nile Delta sand samples, obtained later, were attached to a binocular stereomicroscope (Nikon northern Sinai sand (Fig. 2). Here, dune miner- analyzed with a very similar particle range SMZ800). Sand-grain roundness was assessed alogies were found to be less “contaminated” (0.09–1839 µm) and particle-size bands, and for each frame (~50 and 100 grains per fi ne and by local sand additions from sources (mainly the same measurement protocol using the laser very fi ne sand samples, respectively) using the carbonates) around the periphery of the dune diffractometer of the Eolian Simulation Labo- Powers (1953) grain roundness chart. fi eld (Muhs et al., 2013). The northern corridor ratory of Ben-Gurion University of the Negev was also studied, because it has fewer geomor- (Fritsch ANALYSETTE 22 MicroTec Plus). RESULTS phic units (Roskin et al., 2011a), and, therefore, These two laser diffraction analyzers use the trends may be more easily preserved. Nine sites same laser wavelength (632.8 nm). We assumed Sand-Grain Shape and Color were sampled for this study. Approximately half that using the same protocols for both laser of the sections were exposed, allowing precise analyzers provides highly compatible results. The very fi ne sand fraction and the fi ne sand sampling according to stratigraphy. In other Comparison of soil sediment particle-size dis- fraction of the studied samples generally exhibit cases, hand drilling was performed down to tributions between different laser diffraction similar distributions of grain roundness charac- the dune base with Dormer Engineering hand techniques and the sieve/hydrometer methods teristics (Fig. 4). Approximately 15%–25% of augers (Dormer soil samplers [Australia], http:// indicated discrepancies for silt and clay particle- the grains are angular, ~80% are subangular and dormersoilsamplers .com /Dormer). size distributions while sand particle-size dis- 5%, are subrounded (Fig. 5). Though the Negev Northern Sinai sand samples were acquired tributions were found to be in good agreement fi ne sand samples exhibit a higher percentage from the Geological Survey of Israel (GSI) (Cheetham et al., 2008), and previous studies (25%) of angular grains compared to the Sinai archives, courtesy of Dr. Amihai Sneh, who have jointly used both of these analyzers (Sun samples (15%), the grains are less angular than sampled them for the GSI during the late 1970s, et al., 2002). In order to further validate the coarse loess quartz grains found downwind of since the northern Sinai dune fi eld, being part results between the particle-size distributions the NW Negev dune fi eld that were reported to of Egypt, is currently inaccessible. The samples of the Fritsch and Malvern laser diffraction ana- be angular shaped (Enzel et al., 2010). Chipped were usually obtained from shallow depths of lyzers, 12 representative Negev sand samples grains and freshly exposed faces of grains were 0.2–0.4 m in representative dune-geomorphic were measured in both the Malvern and Fritsch observed for grains in some of the samples. For settings throughout the northern Sinai dune fi eld analyzers. The chosen samples included a wide both Sinai and Negev sands, the grains in the (Table 1). We do not suspect that this sampling range of particle-size distributions in order to fi ne sand fraction usually had more prominent constraint affected the results, since globally, as maximize possible differences between the red coatings than grains in the very fi ne frac- in northern Sinai (Sneh and Weissbrod, 1983), two analyzers. The percent volume of particles tion. The Nile sand samples appeared darker no signifi cant trends in particle-size changes per particle-size band for all of the samples for than the Sinai and Negev sands, due to ~5% with regard to their location upon and within the each analyzer was averaged. An equation of y presence of dark (heavy) mineral grains. The active linear dune section have been established = 0.98x + 0.03 (r2 = 0.89) was calculated for heavy mineral component of the Nile sands (e.g., Folk, 1971; Ahlbrandt, 1979; Livingstone, the average percent volume of particles per each may indicate that during earlier stages of trans- 1987; Pye and Tsoar, 2009). Representative equivalent particle-size band of each analyzer. port, the heavier minerals were winnowed out, late Pleistocene sand samples from cores of the A paired t-test was conducted between the aver- while lighter quartz grains were transported in middle and lower Nile Delta were collected by aged particle-size bands of both analyzers. No a downwind direction, as found by Muhs and Jean-Daniel Stanley (see Stanley et al., 1996). signifi cant difference was found between the Holliday (2001) for the Muleshoe dune fi eld in two averages (p < 1.00; mean = 0.00004). New Mexico and Texas. Sedimentology According to the dynamics of sand-grain fraction characteristics, we analyzed the par- Sand-Grain-Size Features Analysis of particle-size distributions for the ticle-size distribution fractions at cutoffs of Negev sands, composed of 100 particle-size >250 μm (medium- to coarse-sized sand), 250– Nile Delta and Sinai dune sands have lower band measurements within the range of 0.1 µm 125 μm (fi ne sand), 50–125 μm (very fi ne sand), amounts of very fi ne sand (Fig. 6) and therefore to 1924 µm, was carried out by laser diffrac- and <50 μm (silt and clay). The silt-plus-clay generally higher average R values, 5.1 and 4.4, tion (using a Malvern Mastersizer MS-2000) fraction represents particles that are transported respectively, compared to the Negev (R = 2.8) at the GSI, Jerusalem, using the GSI protocol in full suspension and are therefore assumed not (Table 2; Fig. 7). The Negev sands also have a for desert sediments. Samples were split into to be relevant to dune mobilization processes. wide range of modes of medium, fi ne, and very 5 g portions, sieved to <2 mm, and stirred for Following the hypothesis that the particle sizes fi ne sand compared to the Sinai dune sands, dispersion for 10 min in Na-hexametaphosphate gradually decrease downwind, due to eolian which exhibit modes usually in the range of solution, followed by ultrasonifi cation for 30 s. sedimentological mechanisms, we defi ne here a 200–300 μm (Fig. 8). Despite these dissimilari- Three replicate aliquots were initially run for sand fi ning ratio (R): ties, the groups have similar average fi ne silt and each sample. Due to good reproducibility of clay content (<10%) and very small amounts of f the results, attributed to the sand particle-size R = FS average silt (<2.5%), which may represent small f of the samples, later runs used two aliquots VFS inputs of suspended eolian material. for each sample. Each aliquot was subjected to where fFS is the percentage of fi nd sand fraction On average, the Sinai sands have substan- three consecutive 5 s runs at a pump speed of (125–250 µm), and fVFS is the percentage of very tial fractions of medium sand (~33.4%) and 1800 rpm. The raw laser diffraction values were fi nd sand fraction (50–125 µm) in the measured fi ne sand (~45.5%) and contain small quanti- transformed into particle-size distribution using sand samples. ties of very fi ne sand (10.4%) (Table 2; Fig. 6).

52 Geological Society of America Bulletin, January/February 2014 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Particle-size fractionation of eolian sand ) ne 2 28.910.8428. 5 16.28.4610.8 44. 16.730.3423.4 68.75.1328.3 02.20.5919.01 93.14. 76.27.5715.7 88. 10.20.8814.7 26.70.4235.3 08.80.1025.5 02.10.0510.41 82.10.25 94.10.7914.52 09.03.1315.8 12.17.93 8 94.20.5811.7 07.48.7815.4 19.10.3813.7 16.20.2614.6 09.17.1028.6 41.80. 93.40. 97.12.7614.9 86.07.1119.21 88.93.9320.4 39.01.3214.11 2 2 64. 37.19.4411.9 2 1 3 ) 2. 9. 5. 0. 3.39.9815. 0.35.7029. 1. 0. R ( 7 1 1 1 3 1 50. 94. 010. 30.2 continued ( sand ratio Fine/very fi 2.7 7.5027. 0. 7.1719. 0. 0 5624. 3 2 3 3 3 1 1 514.11 4 3 0 0 120.5 028.3 0 mm 2 1 14. 1 2 2 2 2 Mode 3.91 7. 8. 9.5 8 8.5 6.7 .4 3 8 1 5 7 01 2 5 5 1 1 1 1 % 0–20 mm ne silt + clay) (fi 2.1 7. 0.1 0.11 5.4 4.3 5.0 7.0 8.1 8.3 4.1 5.3 3. 9.0 7.2 6.0 7.2 4.2 3.4 8.0 7. 2.3 9.1 6.0 7.0 2.3 6.0 9.0 2.2 8.0 6. 7.1 0.1 2. 8.0 5. 7.2 9.0 ) 0 0 2 2 1 1 R % 20–50 mm (coarse–medium silt) 3.2 8.5 8. 7.6 4.81 8.81 2. 0. 8.12 6.22 4.1 9. 5.94 9.13 9.91 3.13 6. 0.02 6.73 3.6 7 4.1 6.22 3.32 7.3 4.12 8.04 8.6 6.22 7.61 1.72 0.31 8.4 3. 1.21 2.82 3.51 1.7 .9 2 3 1 8 9 0 ne sand) 3 3 3 2 1 1 % 50–125 mm (very fi 3.64 4.54 7.85 4.72 4.14 7. 9.54 9.24 9.25 8. 2. 9.65 4.33 2.83 3.35 0.45 2. 7.33 8.4 7.1 1.34 8. 0.82 9.24 7.94 1.83 5.36 4.55 8.85 6. 2.16 4.74 8.15 4. 1.35 6.05 3.64 0.46 7 6 7 3 8 9 5 3 3 4 3 5 4 3 5 3 ne sand) (fi % 125–250 mm 3.64 1.44 4.42 3.72 8.71 5.04 1 7.21 8.2 9.11 6.81 2.31 2.11 2.5 3.42 6. 9.7 4.2 4 6. 8.32 3.3 2.46 6.72 9. 8.02 0.6 5.6 6.11 9 2.8 5.02 0.24 5.1 0. 0.11 9.92 2 . . . . 9 03 13 42 83 03 42 62 2 4 2 1 (medium sand) % 250–1000 mm 7. 4.0 3 0.1 7 7.0 3.0 1. 2.0 7.0 8.0 6.0 1.0 7.0 6.0 9.0 0.1 0.0 1.0 9. 9.0 9.0 2.0 5.0 8.0 0.0 2.0 3.0 1.0 8.0 4 7. 1.0 5.0 3.0 4.0 5.0 5.0 . . . 0 0 0 0 0 0 0 (m) depth Normalized 58. 58. 5 52.0 5 1.6 2.0 5.4 6.4 7.1 6 8.6 2. 6.2 7.5 5. 8.9 5.4 5.1 5.4 4.0 5.1 2 3.3 5.4 2.1 4.2 2.3 9 7. 7 6.4 8.3 8.7 .0 . .2 .3 2 2 1 6 4 9 7 7 3 2 7 5 (m) 1 1 surface Depth from TABLE 1. SITE DESCRIPTION OF PARTICLE-SIZE DISTRIBUTION FRACTIONS AND SAND FINING RATIO ( AND SAND FINING RATIO DISTRIBUTION FRACTIONS 1. SITE DESCRIPTION OF PARTICLE-SIZE TABLE DL DLV DL V V VLD overlying sand deposit VLD overlying sand deposit 1tizu a 6 565- 36 86 02 9 065-FD 417- 425- 0 12 055- 43 73 y 186FD 386 086 5 28 755 40 586FD 096 486FD 345 29 16 2 4 308 id 6 31-FD 6 7 1 2 6 18 47 96 5 67 08 55 5 1 1 5 5 7 7 7 6 5 5 5 - - - - 8 - - 5 6 - 7 6 a - - - - - l laB MM FD FD F F F FD F FD F FD FD FD FD FD FD FD FD F FD F FD F FD FD FD aH F F F FD FD FD F FD D Tzidkiyahu VLD overlying sand deposit D D D D D KD 73 Retamim D Site and sample no. Physiography (DF-) D D NW Negev D D D

Geological Society of America Bulletin, January/February 2014 53 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Roskin et al. ne 49.211.1121.5 53.418.6125.6 79.08.3414.11 49.0 76.33. 54. 03.093.2525.3 61.47.6819.8 73.310.8122.7 32.210.9910.7 29.36.4228.8 21. 55. 50.614. 20.30.3023.8 44.460.5913.8 36.21.1323.6 07.50.8819.5 91.21.7229.5 ) 1 3.1 4.38.20290.4 2. 7. 8.48.22225.9 6.12.59284.5 1.22.59249.4 0.45.7720.21 5.58.22233.3 5.838.22225.3 6.416.4819 8.38.2 3. 5.55.7724.61 6.45.6030.41 9.119.3734.4 0. 9.98.22235.4 1.40. 6. R ( 48.2 2 2 3 5 1 4 2 70. sand ratio Fine/very fi 1.8 3. 8.7 2.2312.21 6.5126.5 2.59282.6 5. 9.3 27461.5 9 8 9 72 7 2 6 1 8 0 2 0 024. 7 mm 22 137.4 1 27. 2 2 2 2 39. Mode 45. 4. 0. 0. 9. 5. 9.9 7 0 0 7 8 4 8 3 1 1 1 1 2 % 0–20 mm ne silt + clay) (fi ) 00. 82.0 1 71.1 3 42.0 0 86.0 36.0 2 23.0 62.0 3 9.1 5. 4. 1.1 1.2 0.1 4.0 1. 6.2 0.4 3.0 5.2 0.0 6. 2.2 0.4 7.1 2.0 5.4 6.1 5.0 5. 5.1 4.7 3.1 2.1 1. 4.2 5. 0.4 6. 0 0 3 4 2 1 0 2 2 ) ( continued % 20–50 mm R (coarse–medium silt) 81.0 25.51 86.41 50 20.4 9 52 06.7 93.1 18.0 1 04.6 7 7.4 5 3. 1.4 5. 1.6 0.41 2.04 1.61 5.0 3.11 4 8.9 7.51 5.1 9.0 7. 3.9 2.3 6.6 8. 0.11 2. 2.4 4.5 3. 2. 5.81 3 . . . . .8 9 2 1 2 8 4 9 0 3 2 4 ne sand) 1 1 1 1 1 1 1 3 1 1 . % 50–125 mm (very fi 38.21 81.35 94.64 0 67.73 9 07.54 15 50.13 05.35 8 16.74 4 6.06 3. 8. 4. 4.85 8. 9.33 1.85 8 8.84 5.64 1.44 8. 1.93 3.14 8.71 3 6.85 4.73 6.15 5.03 3. 8.26 9. 2.65 9.56 0. 5 6.04 6 .83 . . .53 . 0 7 7 5 6 4 2 1 3 9 1 3 4 4 5 6 4 5 4 5 4 2 ne sand) (fi % 125–250 mm 38.1 39.62 96. 8 54.25 79.5 84 51.82 86.55 09 59. 71.14 93 3.9 0.2 0 0. 7. 4 6.14 3.42 7.2 1.7 4.94 8.13 5.5 8. 4 8.72 2.6 0.2 4.12 6 2.13 5.02 1. 7.9 1.72 6 0.13 . . .03 . . . . . 2 92 81 6 23 31 6 04 81 57 83 3 5 4 4 1 2 4 3 8 2 3 3 1 5 5 (medium sand) % 250–1000 mm 9.0 5.0 9. 0.1 5.0 4.0 5.0 8 9.0 6. 6.0 6.0 0. 2.0 0.1 8 3. 2.0 2.0 4.0 . . 0 0 0 1 0 0 (m) depth Normalized 5 58.4 5 5 57.0 3 1.3 6.6 8.9 5.6 6.4 9. 2. 7.7 6. 1.3 5. 6.7 6.2 5.0 5. 4 21 71 0 91 5 . – – – – – – – – – – – – – 4 2 3 4 5 5 4 7 8 1 5 (m) surface Depth from dn a teehsdnasselppi s en e s d u ab na en rusn d TABLE 1. SITE DESCRIPTION OF PARTICLE-SIZE DISTRIBUTION FRACTIONS AND SAND FINING RATIO ( AND SAND FINING RATIO DISTRIBUTION FRACTIONS 1. SITE DESCRIPTION OF PARTICLE-SIZE TABLE nasu ) s latsa en tayi )DL(alazahGidaW u en dnahcraBa DL u DLrim DLrimaHlebeG a D D d H u aZ Lafa Laf M a D DLhsirAlA l laza o da dekro Lhsir c(a E l d i agf E s y h a iaSlE.zEnI gf ll of wadi rf aR kb r hG Hl h .l a a a kb wer A a c eb Km lA Gr Gr idaW E S n nj a l i i Nm saH eG iw i ni S B B k l nbiLlebeG iw ed Sand infi Archeological site a 3 k r d i ara a r 2 B dra a VLD overlying sand deposit eld B h ) KidaW B continued 9 4 1 2 7 1 8 1 The Sinai samples were usually sampled from 0.2 to 0.4 m depths beneath the surface. VLD—vegetated linear dune; LD—linear dune -W,2 -W,5 -W - - -W,6

97 08 007-F 875-FD 107-FD 84 77 545FD 675 575 W,55-S W,68 , 3-SN 4- 2-SN 1-SN 5 5 5F 5 9 2 02-A 4 61 0 3 6 5 8 2 6 9 - - 5 605 8 015 9 115 Note: 2-A 3 1 2 3 4 2 1 2 3 5 5 2 8 5 F FD S FD F F - - -A -A - -S -S -A - -A -A - - - -A - 0 05 0 D 5 A A D Beqa Sekher-VI Northern Sinai duneﬁ D A A Site and sample no. Physiography BM West Nile Delta A N Retamim ( D D 5 S S S

54 Geological Society of America Bulletin, January/February 2014 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Particle-size fractionation of eolian sand

ABCD

Figure 4. Sinai and Negev fi ne and very fi ne sand grains. (A) Fine sand of sample A-30 from a linear dune, near Al-Arish, northern Sinai (Fig. 1B). (B) Very fi ne sand of sample A-30. (C) Fine sand of sample DF-651 from a vegetated linear dune at the Tzidkiyahu site (Fig. 2), in the northwestern Negev dune fi eld, Israel. (D) Very fi ne sand of sample DF-651.

Fine sand, N. Sinai 60 Very fine sand, N. Sinai Figure 5. Sand-grain angular- Fine sand, NW Negev ity for ﬁ ne and very ﬁ ne sand 40 fraction of Negev and Sinai Very fine sand, NW Negev Frequency (%) Frequency sands. 20

0 Sub-rounded Sub-angular Angular Grain roundness 90

60 Nile Delta 50 Sinai % 40 Figure 6. Grain-size fractions of Negev the Nile Delta, northern Sinai, 30 and NW Negev sands.

0 250–1000 125–250 50–125 20–50 0–20 Grain size fraction (µm)

Geological Society of America Bulletin, January/February 2014 55 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Roskin et al. Comments with mode of 472 mm, Coast dune sample A-24, Coast dune sample was excluded from calculations. RN 4.4 6 5.1 12 2.8 58 (mm) Average Mode

Figure 7. Sand ﬁ ning ratio (R) of the northern Sinai and northwestern Negev sands.

Tsoar (1978) also found a particle-size dis- dunes as described by Tsoar (1974, 1976), with tribution cutoff for particle sizes <125 μm for partial westward contribution from the inland (250–1000 mm) a seif linear dune west of Wadi Al-Arish, and Sinai sand, similar to dune sand farther west % medium–coarse sand Misak and Draz (1997) reported a dominantly on the Suez Gulf coast as described by Harris fi ne sand composition for Sinai dune sand south (1958). The coarse-grained coastal fraction is of Al-Arish. The Sabhat (salt marsh in Arabic) not found in the Sinai linear dunes that are far- Bardwil sample (A24) west of Al Arish (Fig. ther from the coast and exemplifi es the lack of ne sand 1B), with a major mode of 472 μm and a minor substantial input from the coast southwards into % fi (125–250 mm) mode ~200 μm (Table 1), is very different and the Sinai dune fi eld. The high-peaked mode of thus is not included in the calculations. The sand sample A1 (Table 1), which has the least content is mainly part of the coarser-grained coastal of very fi ne sand, is from a (currently) active ne sand (50–125 mm) % very fi TABLE 2. PARTICLE-SIZE FRACTION RANGES AND AVERAGES FOR THE SAND SAMPLES FOR AVERAGES AND FRACTION RANGES 2. PARTICLE-SIZE TABLE % silt (20–50 mm)

Figure 8. Particle-size mode (0–20 mm) ne silt and clay frequency of the northern Sinai % ﬁ and NW Negev sands. Range 4.4–16.4 5.0–18.4 1.5–13.2 17.8–52.3 22–75.6 169–456 Range 3.3–28.5 0.2–0.6 1.4–15.5 31.1–53.5 13.5–55.7 206–355 Range 3.5–19.3 0–7.4 0.5–49.5 46.5–65.9 2.8–49.4 111–252 Average: Std. dev.Average: 7.2 7.44: Std. dev.Average: 4.1 11.4: 0.2 0.4: 2.1: 4.7 4.4 8.9: 7.6: 3.9 9.75 45.5: 12.5 32.6: 13.9 37.4: 46.3: 18.2 255: 41 315: 98 Average: Std. dev.Average: 8.9: 2.3 1.8 2.4: 17.5: 12.5 47.6: 9.7 23.6: 11.9 194: 36 Further site data for the Nile Delta sands can be found in Stanley et al. (1996). Note: Nile Delta Sinai Negev Region

56 Geological Society of America Bulletin, January/February 2014 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Particle-size fractionation of eolian sand barchan dune. The barchans may be perceived A Negev sands as a fi eld laboratory exemplifying how active 6 Sinai sands dunes release the very fi ne sand fraction. The Sinai sands do not exhibit any west-east trend 5 Nile Delta sands in fi ning of particle-size distribution modes or 4 increase in the very fi ne sand fractions. The Negev dune fi eld sands have modes in 3 the range of 100–260 μm. The abundant modes are 200–220 μm (Fig. 8), i.e., slightly fi ner than Kurtosis 2 the northern Sinai sands. Nearly half of the Negev sands have fi ner-grained modes than the 1 Sinai sands. On average, the Negev sands have 0 substantial fractions of fi ne sand (~47%) and –1.0 –0.5 0.0 0.5 1.0 substantially larger amounts of very fi ne sand (~17%) compared to the Sinai sands (10.4%) Skewness (Table 2). Thirty out of the 58 Negev samples contain over 15.5% very fi ne sand compared to only one Sinai sample with 15.5% very fi ne B 180 sand. The Negev’s central dune encroachment corridor has very fi ne sand contents in the wide 160 range of 0.5%–40%. The two end sections of the northern corridor have very fi ne sand content in 140 the range of 12%–50%. A mixed analysis of variance (ANOVA) test 120 using different grain-size fractions (μm) of Negev sands 250–1000, 250–125, and 50–125 and different Sorting 100 Sinai sands locations (Nile Delta, Sinai, Negev) as the inde- Nile Delta sands pendent variables was applied to the data. The 80 analyses showed a main effect of grain-size fraction (F[2,156] = 58.73, p <0.001) and a signifi - 60 cant interaction between fraction and location 100 150 200 250 300 350 400 450 500 p (F[4,156] = 14.51, < 0.001). To explore this µ interaction, t-tests comparing between the frac- Grain-size mean ( m) tions from the different locations were conducted Figure 9. (A) Plot of kurtosis vs. skewness values of the sand samples. (B) Plot of sorting vs. in order to validate the signifi cance of grain-size mean particle-size values of the sand samples. fractions between regions (Fig. 6). Since the low number of analyzed samples for the Nile Delta and the Sinai do not fulfi ll statistical require- ments for a parametric test, the data underwent a nonparametric Mann-Whitney test. Both t-tests Particle-Size Variability in the NW parts of all of the central corridor sections have and Mann-Whitney tests resulted in signifi cant Negev Dune Field a relatively narrow particle-size mode range of differences (p < 0.05) between all three frac- 175–225 μm but a wide R range. The young- tions of the Nile Delta and the Negev sands and The northern dune encroachment corridor of est section, MM, the lowest sample of which is between the Nile Delta and Sinai medium sand the NW Negev dune fi eld has both in the west dated to 9.3 ± 2.0 ka (Fig. 2), has a low R and the fraction. Signifi cant differences (p < 0.05) were and east mainly very fi ne sand with modes in lowest mode values, and relatively high very fi ne found also between the Sinai and Negev sands, the range of 100–140 μm, while in the central sand contents (20%–33%; Table 1). Each dune both for the very fi ne sand and the medium sand dune incursion corridor modes are coarser and section seems to have several size modes ranging fraction. These results support the hypothesis more varied, in the range of 125–260 μm (Table between these values (Table 1). Beyond its upper that changes in grain-size fractions occur along 1; Fig. 2). In the central corridor, no trend in R active section and one basal unit (DF-700), the eolian transport system from the Nile Delta values is apparent along the dune profi les, and which has exceptionally high very fi ne sand con- through northern Sinai to the NW Negev. each profi le’s R curve in relation to the normal- tent (40%), the central part of the Retamim sec- The statistical parameters of the regional ized depth is different (Figs. 10A and 10B). The tion has a relatively low very fi ne sand content. sand groups tend to cluster. The Negev sands, northern corridor end sections have a similar Fining along the central corridor is not pro- predominantly positively skewed, are in good trend in particle-size changes, with the midpro- nounced. For example, the Sekher VI section correlation with similar kurtosis values (Fig. fi le containing a lower very fi ne sand fraction at the eastern end of the dune fi eld, dated to ca. 9A). The Sinai samples have slightly negative (Fig. 10C). The eastern end of the Baladiya sec- 12 and ca. 3 ka (Fig. 2), has the narrowest mode skewness values, with kurtosis values in most tion is coarser grained than in the west. range of 189–205 μm (Fig. 11A; Table 1) and is cases similar to the Negev sands. Both Sinai and In the central corridor, the lower parts of the not fi ner grained than the modes of the KD 73 Negev sands usually exhibit moderately good to dune sections have a noticeably broader R value section, 45 km to the west by the Egypt–Israel poor sorting values (Fig. 9B). and mode range of 125–325 μ m. The upper border, where, with the exception of two units, the

Geological Society of America Bulletin, January/February 2014 57 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Roskin et al.

A 0 Very Fine Sand Units in the NW KD73 Negev Sections

TZ Several distinct units of very ﬁ ne sand with mode peaks in the range 104–131 μm occur BM throughout the dune ﬁ eld (Fig. 11). These units, found mainly at the base of eolian sand sections, have different ages that span the full late Pleistocene age range of the NW Negev dune

Normalized Depth (m) fi eld sands. 1 The R profi le generally refl ects the particle- 0 1 10 100 size distribution mode profi le pattern but further Sand fining ratio (R) highlights inherent sedimentological changes and processes (Fig. 11). These changes include an interchanging (zig-zag) pattern between units of the same section with fi ne sand dominance 0 B and negligible very fi ne sand units with substan- MM Figure 10. Sand fi ning ratio vs. tial content of very fi ne sand (Fig. 10). As docu- normalized depth of dune and Retamim mented for the profi led particle-size distribution eolian sand sections per dune of: modes, no clear trend of very fi ne sand content Sekher VI (A) the western part of the cen- occurs with depth, and therefore there is no con- tral dune encroachment corri- Beqa nection between sand OSL age and changes in dor, (B) central and eastern part particle-size distribution. No sand fi ning occurs of the central dune encroach- parallel to the downwind sand transport direc-

ment corridor, and (C) northern Normalized Depth (m) tions in the Negev dune encroachment corri- dune encroachment corridor. 1 dors (Fig. 1B), which probably implies that the For section location, see Fig. 2. changes in particle-size distribution are more of 0 1 10 100 a dynamic eolian character and also are probably Sand fining ratio (R) not solely related to time-transgressive pedogenic breakdown of quartz grains with time. C 0 DISCUSSION

Sedimentological Properties of the Sinai- Haluzit Negev Erg Sands

Baladiya The dominant subangular shape of the sand grains of the Sinai and Negev is consistent with most other arid inland dune sand grains, which

Normalized depth (m) range in the subrounded and subangular classes (Goudie and Watson, 1981). Similar sand- 1 grain roundness distributions are also found for 0 1 10 100 both fi ne and very fi ne sand fractions of desert Sand fining ratio (R) dunes elsewhere (Khalaf and Gharib, 1985; El- Sayed, 1999, and references within). The limited amount of distinct freshly exposed surfaces and chipped grains observed does not indicate modes range from 162 to 211 μm. At the KD-73 A similar discrepancy between Sekher sand substantial grain-to-grain eolian abrasion. More vegetated linear dune section, sample DF-695, particle-size distribution has been reported by angular grains were observed in the Negev fi ne dated to 15.6 ± 1.5 ka (Roskin et al., 2011a), Bacon et al. (2011). The 8-m-thick Beqa sec- sand samples compared to the Sinai samples has a mode of 213 μm, and the adjacent DF-681 tion infi lls a wadi, and its upper part has been (Fig. 5). On the contrary, the higher red-grain of the KD-73 interdune upper section, dated to reworked several times in the Holocene. The coatings of the fi ne sands suggest less abrasion, 13.3 ± 1.4 ka, has a mode of 195 μm, similar to Sekher section is only 3 m thick, mantling a while the lighter color of the very fi ne sand may the Sekher sands deposited slightly later. chalk ridge. The Beqa section particle-size dis- indicate that some of the grains are more likely There are signifi cant sedimentological varia- tribution modes are in the range of 188–224 μm, to be products of eolian abrasion, a hypothesis tions for adjacent dune sections. The Beqa and other than the basal 2.5-m-thick grayish colored that needs further testing. Sekher VI sections, only 800 m apart at the far unit (sample DF-575) dated to 11.6 ± 1.8 ka. Although the original and direct sand source eastern end of the dune fi eld (Fig. 2), exhibit Sample DF-575, in the same age range as the characteristics of the Sinai-Negev erg are prob- different ranges in mode values and particle- lower Sekher sands, also has a distinct particle- ably absent today, late Pleistocene sand of the size distribution curves (Figs. 11A and 11b). size distribution curve and a mode of 132 μm. Nile Delta that has been logged from beneath

58 Geological Society of America Bulletin, January/February 2014 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Particle-size fractionation of eolian sand

A 12 and Stanley, 1987; Frihy et al., 1998; Stanley, DF-577 DF-576 DF-578 2002). Five different eolian microenvironments 10 along the Suez Canal also showed different DF-579 DF-580 DF-575 8 proportions of very fi ne sand content (Harris, 1958). Based on 87Sr/86Sr isotopic similarity, 6 part of the Negev loess particles has been suggested to be of a Nile Delta origin (Vaks et al., c Frequency (%) 4 2013), particularly the coarse fraction (Haliva- Cohen et al., 2012). 2 While Frihy and Stanley (1987) showed the 0 potential of identifying depositional environments of sands based on their sedimentologi- 110 Particle-size (µm) 100 1000 cal characteristics, specifi c identifi cation of the Sinai-Negev erg source deposits and environ- B12 ments is quite complex. Late Pleistocene sands NS-1 NS-2 cored in the northeastern Nile Delta have been 10 found to possess inherent mechanical abrasion NS-3 NS-4 features (Frihy and Stanley, 1987), while mod- 8 ern desert sands are more characterized by solution etching. Most of the sand grains located at 6 the northern edge of the Nile Delta in a beach c environment are rounded (Abd-Alla, 1991), but 4 Frequency (%) this may suggest that these sands and the beach 2 sands of northern Sinai, which have undergone signifi cant fl uvial action, are not the source of 0 the Sinai dunes, strengthening the assumption that the upper and central Nile Delta deposits 1 10 100 1000 Particle-size (µm) located directly west of the erg are the source of C the erg. Desert sand grains, usually of coarse and medium fractions, near the delta shore, are also 12 DF-18 DF-721 rounded and contain v-shaped and crescent-like pits of 10–20 μm. The pits are interpreted to be a 10 DF-42 DF-81 result of eolian sand abrasion (Abd-Alla, 1991). DF-700 DF-575 8 Based on interpretation of scanning electron microscope (SEM) images, it seems that the pits 6 are usually 10–20 μm in size, corresponding to fi ne silt sizes. This observation is in accordance 4 μ Frequency (%) with recent 125 m sand grain-grain abrasion experiments that simulate 1000 km of transport, 2 approximately four times the length of the Sinai- Negev erg. The experiment produced 18.6% ± 0 0.4% of the particles below 125 μm, with 9.0% 1 10 Particle-size (µm) 100 1000 ± 1.6% below 18 μm, while the sand-size distribution values remained similar (Merrison et al., Figure 11. Exemplifi cation of particle-size distribution variations between the adjacent 2010; Merrison, 2012). Based on the sand-grain Beqa (A) and Sekher VI (B) (exposed) sections, located at the eastern end of the NW Negev chemical and mechanical features reported by dune fi eld. Each group has one sand sample with a particle-size distribution similar to the Abd-Alla (1991), the Nile Delta sands may have other section. The higher peaked curves are of the Sekher samples (other than NS-3) and often undergone multiple cycles of transport include DF-578. Sample DF-575, marking the base of the Beqa section, is the outstanding and deposition. Experiments of quartz silt profi ner-grained curve. (C) Negev sand samples with modes around 100 μm. duction from abrasion of highly angular quartz sand grains found that after the fi rst 16 h of a 224 h experiment, 40% of the total fi ne and Holocene silts and clays (Stanley et al., 1996) processes, especially in the windy late Pleisto- coarse silts (not very fi ne sand) were produced may be a good candidate (Muhs et al., 2013). cene, may have acted upon the deltaic sediment (Wright et al., 1998). This is a fair analogue of It seems that very fi ne sand was initially pres- shortly following deposition, initiating winnow- eolian abrasion for fi rst-cycle erosion. However, ent in the delta, as were more abundant fi ner ing. Several studies have reported that the vari- it seems that fi rst-cycle erosion is not the pre- and larger sand particle sizes within the origi- ous depositional settings of the Nile Delta have dominant case of the study area. As earlier work nal dune sources (Fig. 6; Table 2). However, it sediments with particle-size distributions rang- suggests, interpretations of the origin of sands cannot be concluded that the samples were fully ing between medium-sized sand to fi ne silt, but on the basis of grain surface features alone may preserved in the delta since deposition, as eolian they nearly always include very fi ne sand (Frihy be insuffi cient (Frihy and Stanley, 1987, and

Geological Society of America Bulletin, January/February 2014 59 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Roskin et al. references within), which necessitates focusing invading sand, including this fraction, reached of Gebel Maghara (Fig. 1B) were found to on particle-size trends as an indicator of eolian the region from the west, probably from Nile contain signifi cant amounts of very fi ne sand transport processes. Delta sediments (Amit et al., 2011). Sepa- (Goldberg, 1977). The basin may have been a Therefore, we build upon previous studies rately OSL-dated fi ne sand (125–150 μm) and favorable trap for fi ner grain sizes that over- (Tsoar, 1990; Hunt, 1991; Amit et al., 2011; very fi ne sand (88–125 μm) of correlating sand topped the basin’s ridge as climbing dunes. Roskin et al., 2012; Muhs et al., 2013) and infer (sheet) units at ~1.5 m depths within the Qerem This is supported by empirical fi eld investiga- that the very fi ne, fi ne, and larger sand fractions Shalom paleosol section, along the Egypt-Israel tions of a 20° slope of a climbing dune, where, of the Sinai-Negev erg evolved from widespread border, between the Negev dune fi eld and Gaza under a wind shear velocity of 30 cm/s, grains source deposits in the Nile Delta that were Strip, are of very similar ages (13.4 ± 1.7 ka, of 130 μm diameter were found to saltate to the exposed during the last glacial period (Stanley 14.5 ± 2.3 ka, respectively) (Zilberman et al., crest while coarser grains did not (Tsoar et al., and Warne, 1993). This assumption for sand is 2007) and date to the main late Pleistocene 1996; White and Tsoar, 1998). This setting may similar to the genetic association of quartz silt Negev dune incursion from the west (after be analogous to the depositional environment grains between the Nile Delta and the Negev Roskin et al., 2011a, 2011b). This suggests a of the ~100 μm quartz grain mode deposits on loess initially proposed by Smalley and Krinsley genetic synchronicity of very fi ne and fi ne sand the Qeren ridge gully slopes slightly above the (1978). The variability of the Sinai and Negev deposition that originated from the west along elevation of adjacent dunes in the NW Negev sand-grain coatings within the fi ne and very fi ne the Nile Delta–Sinai–Negev eolian transport (Enzel et al., 2010), regardless of their forma- sand fraction may also be attributed to different system, during and prior to dune encroachment. tion processes. depositional environments in the delta, as sug- The different very fi ne sand contents present The high variability of the R and the very gested by Roskin et al. (2012), and may not nec- in several sandy to loamy paleosol units, dated fi ne sand contents in NW Negev dune sections, essarily be a sole result of eolian abrasion. between ca. 200 ka and ca. 30 ka and underly- not identifi ed between the samples of northern Fractionation of very fi ne sand occurred by ing the younger NW Negev dunes, also imply Sinai, suggests fl uctuating rates of transport and the hypothesized Nile Delta source and con- that small quantities of fi ne sand along with deposition in the northern Negev after 23 ka. tinued in the northern Sinai sands. The coarser very fi ne sand reached the Negev prior to the This variability implies that the fractionation sand fractions were winnowed out, while very dune encroachment. The variability is attrib- process and consequent deposition of very fi ne fi ne sand was often transported by saltation and uted to the relative paucity of sediment supply sand usually occurred in short and highly windy low suspension in a general eastward direction or weaker winds (Roskin et al., 2011a, 2013c). events. The sand component of 62–185 μm has beyond the Sinai dune fi eld. Due to SW and Downwind northern Negev loess deposits of been found to be the component with the big- NW dominant winds, as evidenced by the for- similar age ranges, reported to usually have gest standard deviation value and also the one mation of seif dunes (Tsoar, 1989), the very fi ne particle-size distributions (Crouvi et al., 2008) with the highest degree of sensitivity, mainly sands were transported and dispersed to the east that are fi ner than the NW Negev paleosols, may controlled by source-to-sink distance (Guan in a wide angle, which may account for their also imply a fractionation mechanism along the et al., 2013). Several distinct dune units with an presence in the northern Negev loess depos- Nile Delta–Sinai–Negev transport system prior abundance of very fi ne sand are found through- its (Fig. 1B). The signifi cantly higher coarse to the Negev dune encroachment. A jump in the out the dune fi eld in basal units, such as at the sand fractions in the Sinai samples compared very fi ne sand content from 10%–20% to 20%– Beqa section. These units, characterized by to the Negev (Fig. 6; Table 2) and the presence 50% at Qerem Shalom around 190 ka marks fi ner grain sizes, may mark the leading head of coarse-grained zibars at the base of linear the onset of an increased eolian transport sys- of fractionized eolian encroachment. An upper seif dunes in northern Sinai (Tsoar, 1989), not tem of very fi ne sand that ended with the onset unit of an interdune deposit (DF-18 of the MM found in the Negev, exemplify the fractionation of the Holocene. Amit et al. (2011) reported a section) exhibits 33% of very fi ne sand is dated process of fi ner sand from coarser sand. During signifi cant input of coarse silts and very fi ne to 9.8 ± 1.9 ka (Roskin et al., 2011a), similar times of high sand supply, the fi ne sands form sand with a 70 μm mode minimally OSL dated to the time of upper Negev loess deposition at active linear seif dunes, while very fi ne sand to ca. 180 ka in reg soils of the southern Negev ca. 10 ka (Crouvi et al., 2009). This unit may fractions are released from their source mate- Desert (by the border with Egyptian Sinai). represent the fi nal stages of eolian deposition, rial to be partial constituents of vegetated lin- This input was suggested to be connected to the when wind speeds were less profi cient at mobi- ear dunes. Downwind fi ning (grading) of sands exposure of sandy Nile Delta sediments dur- lizing coarser fi ne-grained sand. Eolian erosion from the upper parts of active dunes in the ing a period of glacially eustatically lowered of the late Pleistocene paleosols that underlie Suez Canal area is reported for short distances sea level. Amit et al. (2011) further suggested the dunes (Roskin et al., 2011a) is also a pos- (Harris , 1958). This trend can be inferred to be a genetic connection between the coarse silt sible closer and partial source for these very seen as a small-scale analogue for the whole formation and transport, and dune mobilization fi ne sand units, though further study is required Sinai-Negev erg. in northern Sinai. Therefore, since transport of to identify the relative contribution of the dune very fi ne sand requires lower wind power than substrate to downwind sediments. Evidence for Spatial and Temporal Distribution of the fi ne sand that characterizes linear dunes, the this is hand sample DF-685 (17.9 ± 2.8 ka) at Very Fine Sand assumption of Amit et al. (2011) does not nec- the base of the KD-73 section, which had ~3% essarily imply that dunes were mobilized at the of sand-sized silt pellets. As the equivalent Sedimentation of very fi ne sand is found else- time, especially at the intensities inferred for the (OSL) dose of the sample was strongly biased where in the Sinai and Negev. Very fi ne sand NW Negev during the late Pleistocene. How- and the age error relatively large, we attributed occurred in NW Negev paleosols already by ever, further study of the NW Negev paleosols the pellets to the underlying paleosols dated to the late middle Pleistocene (ca. 190 ka), early is required to validate this hypothesis. ca. 45 ± 13 ka (Roskin et al., 2011a). The rea- marine isotope stage (MIS) 6, long before the Despite the current paucity of very fi ne sand sons behind temporal and spatial variance of appearance of vegetated linear dunes. This in the active Sinai dunes, late Pleistocene fall- very fi ne sand deposition along the 25–45 km observation indicates that small amounts of ing (Mushabi) dune sands in an erosional basin Negev dune fi eld transect may explain why a

60 Geological Society of America Bulletin, January/February 2014 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Particle-size fractionation of eolian sand regional particle-size fi ning trend is not identi- modes. Field experiments with the Negev dune energy and an increase in surface friction. For fi ed in the Negev. sands found that sands with a mean particle certain episodes during the late Pleistocene, Marine and terrestrial very fi ne sand deposits diameter (MPD) of 130 μm have substantially such as during the Heinrich 1 and Younger in the Levant region provide further support- higher sand emission fl uxes than sands with an Dryas, which caused rapid vegetated linear ive data for suspension of very fi ne sand dur- MPD of 180 μm (Bacon et al., 2011). Pye and dune encroachment into the Negev (Roskin ing windy periods. Sharp increases of very fi ne Tsoar (1987) stated that today in typical wind et al., 2011a, 2011b) and probably no earlier sand and coarse silt contents, similar to earlier storms, quartz particles larger than 50 µm are than the last glacial, the Sinai dunes as found glacial times, are found in offshore Mediter- unlikely to be transported more than 60 km from by Goldberg (1977) were probably highly ranean cores during the Heinrich 1 stadial ca. their source, and most will be deposited within active at the same time and probably for some 17–14 ka, a period of persistent higher wind ~30 km. During the Last Glacial Maximum time before. When the linear dunes of the Sinai velocities (Jullien et al., 2007; Hamann et al., (LGM), the low-latitude regions like the study reached the Negev region, their extending noses 2008; Tjallingii et al., 2008; Costas et al., 2012; area may have had more frequent strong winds included the saltating sand along with very fi ne McGee et al., 2013), when the dunes encroached and higher wind speeds (Harrison et al., 2001; sand that were transported via the dunes and into the NW Negev (Roskin et al., 2011a). Suf- Jullien et al., 2007; Enzel et al., 2008; McGee earlier by low suspension (Fig. 12). Following fi cient amounts (several percent) of very fi ne et al., 2010, 2013). Thus, very fi ne sand could reduction in wind speed and turbulence, veg- sand available for luminescence dating in Qua- have been transported to greater distances by a etated linear dune vegetation cover may have ternary terrestrial deposits found throughout combination of saltation and short-term suspen- increased, if there was a generally more humid Israel (N. Porat, 2012, personal commun.), such sion. Taking the Pye and Tsoar (1987) rates at a late Pleistocene climate in the region (Baruch as in late Pleistocene fi ne-grained deposits infi ll- time scale of thousands of years means that very and Goring-Morris, 1997; Vaks et al., 2006). ing open fractures along the western Dead Sea fi ne sand, if independently mobilized, could Vegetation increases the aerodynamic rough- escarpment (Roskin et al., 2013a), also attest to potentially be blown for hundreds to thousands ness of the surface, thereby extracting energy a more distal suspension of this fraction. of kilometers, beyond the Negev dune fi eld and from the airfl ow and reducing shear stress at the loess belt. While small quantities of very fi ne surface (Pye and Tsoar, 1987), causing an expo- Regional Winnowing Mechanism sand are present in marine cores and throughout nential decrease in sand transport and inducing Israel, it seems that the main bulk of late Pleisto- deposition of all of the sand fractions (Allgaier, Our proposition for downwind fractionation cene very fi ne sand is found in the NW Negev 2008). Part of the very fi ne sand also continued of very fi ne sand fi ts eolian sand transport paleosols, vegetated linear dunes, and northern to be transported farther east to form part of the dynamics, the transport mechanism and dis- Negev loess deposits that are no farther than upper sections of the primary northern Negev tance of which are controlled by particle sizes 400 km from the Nile Delta. This observation loess deposits. The very fi ne sand (74–125 µm) (Pye and Tsoar, 2009). A fi ning with distance supports the hypothesis that extremely windy fraction comprises the coarse end of the parti- trend of eolian sediments between the Negev episodes triggered dune mobilizations in the cle-size distribution of these loess deposits east dunes and sandy loess soils was initially sug- NW Negev (Roskin et al., 2011b). and south of the dune fi eld (Crouvi et al., 2008; gested by Ravikovich (1981), who also noted During times of dune mobilization in the Haliva-Cohen et al., 2012). In these depos- that the coarse silt (74–44 μm) and very fi ne Sinai and the Negev, deposition of very fi ne its, the coarse silt and very fi ne sand content sand fractions (144–74 μm) are transported sand in the Negev vegetated linear dunes is increased ca. 30–35 ka and peaked at 20–14 ka signifi cantly longer distances than coarser sand suggested to be due to a decrease in transport (Crouvi et al., 2008).

Figure 12. A conceptual model of eolian fractionation along an eolian transport system from the Nile Delta, through northern Sinai, and into the northern Negev Desert (Israel). The fi g- ure is modifi ed from Figure 11c of Pye (1995), which proposes one of several alternative situ- ations in which loess deposits may be formed. Along the Sinai sand transport path, the wind power decreases, and annual precipitation increases. Me- dium to coarse sand creeps in Sinai to form zibars, fi ne sand forms linear (seif) dunes, and fi ne and very fi ne sand saltates to the east to form vegetated linear dunes in times of large sand supply. Very fi ne sand is also suggested to be transported in low suspension to the east to be a partial component of loess deposits. Beyond the possibility of full suspension above the Sinai-Negev erg, in the past, deltaic silt may have also been deposited in dunes and later released with dune mobilizations, a hypothesis worth testing.

Geological Society of America Bulletin, January/February 2014 61 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Roskin et al.

Postdepositional Translocation of physically bar the possibility of very fi ne sands 2011a) to 200 v.u. at Bir Lahfan and 21–108 Very Fine Sand and Silt and even coarse silt to percolate down-profi le. v.u. in the NW Negev (Tsoar et al., 2008) (Figs. The upper active crest of the vegetated linear 1A and 13). Assuming generally similar wind Allochthonous (postdepositional) input of dunes that are not constantly being resupplied directions in the late Pleistocene (Ben-David, very fi ne sand (after Goudie et al., 1993) into from Sinai shows similar very fi ne sand contents 2003; Enzel et al., 2008), if this gradient in wind the Negev vegetated linear dunes during the to the rest of the profi le. This does not seem to power was enhanced in the late Pleistocene, it Holocene seems unlikely. At the Qerem Shalom indicate that winnowing is occurring today, nor explains the encroachment of dunes from Sinai section, OSL-dated fi ne (125–150 μm) and very are very fi ne sands being incorporated into the into the Negev and their consequent stabiliza- fi ne sand-grain fractions (88–125 μm) of a sam- section by local dune reactivation. tion. The eastern end of the Negev dune fi eld ple from a calcic paleosol, only 25 cm beneath is not fronted by a topographic obstacle on its the paleosol top, are dated to 55 ± 13 ka and Climate Gradient, Winnowing, and downwind side. Therefore, the eastern extent of 42 ± 5 ka, respectively. Though falling within Changing Linear Dune Types the dunes was probably controlled by a decrease errors, the younger very fi ne fraction has been in windiness along with a thicker vegetation interpreted to be postdepositional due to later The current difference between the active cover due to higher annual rainfall (~150 mm dust fall (Zilberman et al., 2007). As the late linear seif dune types of northern Sinai and the today). This decrease in wind speed may have Pleistocene sequence shows fl uctuations of slow vegetated linear dunes of the NW Negev likely also occurred following the Younger Dryas sand sheet accumulation and pedogenic stability refl ects past conditions that promoted winnow- (Enzel et al., 2010), which is consistent with in a hypothesized moister climate that probably ing of sand in Sinai and deposition of very fi ne global data (Roskin et al., 2011b, and references supported more vegetation, incorporation of sand in the Negev. Vegetated linear dunes, being within; Costas et al., 2012; McGee et al., 2013). very fi ne sand into the upper part of the soil is of an accumulative nature (Roskin et al., 2011b; This sedimentological pattern is also appar- possible. However, we suggest that this scenario Telfer, 2011), were ideal candidates for deposi- ent in the change of dune types along the Sinai- is not likely for dunes. tion and preservation of such sediments. Based Negev erg transport path. In the western and Currently, rainfall and consequent percola- upon a climate gradient of rainfall (Mohamed, central Sinai dune fi eld, west of Wadi Al-Arish, tion values in dunes along with wind intensities 2012) and wind intensities, and consequently there are linear dunes and barchans (Abdel- and corresponding dust fall do not support the dune types, across the Sinai-Negev erg from west Galil et al., 2000; Tsoar et al., 2004). East of idea that dust and very fi ne sand are translocated to northeast (Fig. 1B), it seems, that as today, this Wadi Al-Arish, complex-braided linear dunes down the Negev dune profi le in the presiding gradient existed in the past. This gradient was (Tsoar, 1995), found to be currently less active Holocene environment. The Negev dunes are probably characterized by higher rainfall and than their linear seif dunes west of Wadi Al also encrusted with a cynobactaeria type that wind value ranges. In a similar fashion, Amit Arish (Misak and Draz, 1997; Abdel-Galil et al., collates fi nes deposited upon the dune surface, et al. (2006) proposed that the current climatic 2000), extend into the Negev. Northern Sinai and which severely limits translocation. The bio- and rainfall gradient similarly existed in the late the Negev are suggested to have experienced a logical crust has been found to establish itself Pleistocene between the northern and southern long dry period during most of the seventeenth in 4–7 yr (Kidron et al., 2009, and references Negev. Accordingly, dunes, as today, had less and eighteenth centuries, which resulted in the within), and Negev dunes have been proposed vegetative cover and were active in the west and formation of these braided linear dunes. Prior to to have been stabilized by crusts for much of the more vegetated toward the east and northeast. In that, they may have been vegetated linear dunes Holocene (Roskin et al., 2011a, 2013b). Blume times of stability, dunes in the northern part of (Tsoar, 1995). This change in dune types is in et al. (1995, 2008) demonstrated translocation the Negev dune fi eld were also probably covered agreement with the general eastward increase in of clays through the upper 1–2 m vegetated with thicker stabilizing biogenic crusts, preserv- rainfall, which is relatively similar in northeast- linear dune axis in some Negev dunes that are ing the dune profi le as occurs under the current ern Sinai and the NW Negev (Fig. 1b), enabling currently active, but Negev dunes have limited climate (Almog and Yair, 2007). the dunes to support vegetation and develop into clay contents that do not show trends with depth Because of the pathways of cyclonic storms vegetated linear dunes by joint elongation and (Table 2). Annual and seasonal rainfall infi ltra- over the Mediterranean Sea, it is drier and accretion (Roskin et al., 2011b). tion studied in the SW dune fi eld usually does windier in the west (Sinai) and wetter and less not infi ltrate the dune section to depths greater windy in the east (Negev): The rainfall gradient CONCLUSIONS than 1–2 m (Yair, 2008). Water percolation of the Sinai-Negev region (Fig. 1B) is in general down to 4 m in the Negev uncrusted vegetated agreement with a decrease in sand drift potential This study presents a spatial analysis of parti- linear dune axis profi le only occurs in response from west to east. Therefore, the southwestern cle-size distributions of radiocarbon-dated Nile to above-average rainfall exceeding 100 mm part of the Sinai dune fi eld, where precipita- Delta sands, northern Sinai dune fi eld sands, and in 2 mo (Yair et al., 1997). Therefore, only in tion is currently below 50 mm, has been often NW Negev vegetated linear dune sands, dated wetter conditions is percolation expected down extremely arid and unvegetated throughout by OSL. Based on previous studies and the the profi le. The hypothesized wetter late Pleisto- the late Pleistocene and Holocene, similar to available samples for this work, the particle- cene (Baruch and Goring-Morris, 1997; Vaks the southern Negev at similar latitudes to the size distributions of Sinai sands and their lower et al., 2006; Tsoar, 2013) is in theory a favor- east (Amit et al., 2006). As winds are strong and less deviated very fi ne sand content in rela- able time period for deeper water penetration. in western Sinai, the region has been probably tion to the Negev sands leads to a conceptual However, Blume et al. (2008) stated that for the continuously prone to eolian erosion through- eolian sedimentological model of particle-size Nizzana sands, pore radii are in the range of out glacial and interglacial periods, which can fractionation of sand along the Sinai-Negev erg 50–100 μm, and geometric mean radii in dune explain the 190 k.y. of hypothesized activity. during the late Pleistocene (Fig. 12). Though sands of 250 μm have been recently modeled Drift potential values of northern Sinai gener- they are continuous geographic sand deposits and physically calculated not to exceed 43 μm ally decrease to the east, ranging from >1000 with similar orientations and sand-grain round- (Glover and Walker, 2009). These conclusions vector units (v.u.) at Port Said (Roskin et al., ness values, a modern-day gradual decrease

62 Geological Society of America Bulletin, January/February 2014 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Particle-size fractionation of eolian sand

50 Figure 13. Climatic and very W E fi ne sand percentage, changes, Sinai dune elongation (m/yr) and trends along the Sinai- 40 Sinai very fine sand % Negev erg transport system. Negev very fine sand % The very fi ne sand percentage DP/10 of the Negev sections is the av- 30 erage of the full dune profi le. The y axis represents different values as marked in the legend. 20

DP—drift potential. The loca- border Egypt - Israel tions of the DP values are shown in Figure 1B. Very high DP 10 values (DP = 1139) of Port Said near the northwestern corner of the Sinai-Negev erg are not Northern Sinai, Egypt NW Negev, Israel 0 presented. 0 50 100 150 200 250 Km

in windiness and increase in precipitation in a ACKNOWLEDGMENTS frontal systems: Geology, v. 34, p. 509–512, doi: 10.1130 /G22354.1. general west-east orientation along the Sinai- We thank Amihai Sneh for kindly sharing the Sinai Amit, R., Crouvi, O., Simhai, O., Matmon, A., Porat, N., Negev erg are suggested to refl ect a similar samples. We thank Rivka Amit for helpful comments McDonald, E., and Gillespie, A.R., 2011, The role but enhanced climate gradient in the past. This and both Rivka Amit and Onn Crouvi for providing of the Nile in initiating a massive dust infl ux to the Negev in the late to middle Pleistocene: Geological gradient is postulated to have controlled the guidance and access to the sedimentological labora- Society of America Bulletin, v. 123, p. 873–889, doi: proposed mechanism of sand fractionation and tory at the Geological Survey of Israel in Jerusalem. 10.1130 /B30241.1. Interesting discussions with Daniel Hartmann re- Bacon, S.N., McDonald, E.V., Amit, R., Enzel, Y., and downwind deposition of very fi ne sand. sulted in helpful insight for particle-size distribution Crouvi, O., 2011, Total suspended particulate mat- Intermittent periods of enhanced windiness analysis. Naomi Porat is also thanked for discussion ter emissions at high friction velocities from desert in the late Pleistocene are suggested to have and advice. Thanks go also to Vered Refaely for her landforms: Journal of Geophysical Research, v. 116, F03019, doi: 10.1029 /2011JF001965. triggered fractionation of the fi ne and very fi ne guidance with statistics, to Roni Bluestein-Livnon for her assistance with graphics and statistics, and Bagnold, R.A., 1937, The transport of sand by wind: The Geographical Journal, v. 89, p. 409–438, doi: sand fraction of the NW Negev sand source. The to Alexandra Shtein for her assistance with particle- source of the sand is hypothesized to be solely 10.2307 /1786411. size measurements. We warmly thank reviewers Dan Bagnold, R.A., 1941, The Physics of Blown Sand and Desert confi ned to the Nile Delta and northwestern Muhs and W.C. Johnson for their helpful insight and Dunes: London, Methuen, 265 p. Sinai sand sheets. In periods of high sand supply, comments and two additional anonymous reviewers Baruch, U., and Goring-Morris, A.N., 1997, The arboreal fi ne sand formed active linear dunes in the west- for helping in sharpening the scientifi c merit of the vegetation of the Central Negev Highlands, Israel, at manuscript. the end of the Pleistocene; evidence from archaeo- ern and central parts of northern Sinai. Intermit- logical charred wood remains: Vegetation History tent downwind transport of very fi ne sand via REFERENCES CITED and Arachaeobotany, v. 6, p. 249–259, doi:10.1007 /BF01370446. low suspension and rapid saltation along with Abd-Alla, M.A.A., 1991, Surface textures of quartz grains Ben-David, R., 2003, Changes in Desert Margin Environ- saltating fi ne sand led to vegetated linear dune from Recent sedimentary environments along the ments during the Climate Changes of the Upper Qua- accretion and elongation in the Negev where Mediterranean Coast, Egypt: Journal of African Earth ternary [Ph.D. thesis]: Jerusalem, Hebrew University Sciences, v. 13, no. 3–4, p. 367–375, doi: 10.1016 /0899 of Jerusalem, 170 p (in Hebrew). vegetative cover was higher. Based on varying -5362 (91)90100-D. Blume, H.-P., Yair, A., and Yaalon, D.H., 1995, An initial very fi ne sand components in paleosols beneath Abdel Galil, M., Hereher, M., and El-Etr, H.A., 2000, Study study of pedogenic features along a transect across of movement of sand dunes of Northern Sinai and their longitudinal dunes and interdune areas, Nizzana re- the Negev dunes, dominantly coastal sands potential impact on regional development: 2nd Inter- gion, Negev, Israel: Advances in GeoEcology, v. 28, north of the Negev dune fi eld and in the north- national Conference on Earth Observation and Envi- p. 51–64. ern Negev loess deposits, this conceptual model ronmental Information, Cairo, Egypt (CD-ROM). Blume, H.-P., Beyer, L., Pfi sterer, U., and Felix-Henningsen, Abou Rayan, M., Djebedjian, B., and Khaled, I., 2001, P., 2008, Soil characteristics and pattern of the Niz- is suggested to have occurred in lower and vari- Water supply and demand and a desalination option for zana Research Site, in Breckle, S.-W., Yair, A., and able intensities along the same eolian transport Sinai, Egypt: Desalination, v. 136, no. 1–3, p. 73–81, Veste, M., eds., Arid Dune Ecosystems—The Nizzana path, probably since the late middle Pleistocene doi: 10.1016 /S0011 -9164 (01)00167-9. Sands in the Negev Desert: Ecological Studies: Berlin Ahlbrandt, T.S., 1979, Textural parameters of eolian de- Springer, p. 65–77. (ca. 190 ka) around early MIS 6. posits, in McKee, E.D., ed., A Study of Global Sand Cheetham, M.D., Keene, A.F., Bush, R.T., Sullivan, L.A., We fi nd that particle-size distribution is a Seas: U.S. Geological Survey Professional Paper 1052, and Erskine, W.D., 2008, A comparison of grain-size p. 21–51. analysis methods for sand-dominated fl uvial sedi- robust indicator of depositional modifi cation of Ahmed, M.T., 2010, Ecosystems and Human Well-Being— ments: Sedimentology, v. 55, no. 6, p. 1905–1913, doi: eolian dune sand along a transport system from El Maghara, Northern Sinai, Egypt: United Nations 10.1111 /j.1365 -3091 .2008 .00972.x. a defi ned source and climatic gradient of an erg, Environment Programme. Chepil, W.S., 1951, Properties of soil which infl uence wind Allgaier, A., 2008, Eolian sand transport and vegetation erosion: IV. State of dry aggregate structure: Soil which is further expressed by geomorphological cover, in Breckle, S.W., Yair, A., and Veste, M., eds., Science, v. 72, p. 387–402, doi:10.1097 /00010694 differences of the linear dunes. Hopefully, in the Arid Dune Ecosystems: Berlin, Springer, p. 211–224. -195111000 -00007. future, when sediment sampling possibilities can Almog, R., and Yair, A., 2007, Negative and positive effects Costas, S., Jerez, S., Trigo, R.M., Goble, R., and Rebelo, of topsoil biological crusts on water availability along L., 2012, Sand invasion along the Portuguese coast be completed in the same systematic fashion in a rainfall gradient in a sandy arid area: Catena, v. 70, forced by westerly shifts during cold climate events: Egypt as they have been in Israel, a more com- no. 3, p. 437–442, doi: 10.1016 /j.catena.2006 .11.012. Quaternary Science Reviews, v. 42, p. 15–28, doi: Amit, R., Enzel, Y., and Sharon, D., 2006, Permanent 10.1016 /j.quascirev .2012 .03.008. prehensive understanding of the eolian processes Quaternary hyperaridity in the Negev, Israel, result- Crouvi, O., Amit, R., Enzel, Y., Porat, N., and Sandler, A., along the Sinai-Negev erg can be achieved. ing from regional tectonics blocking Mediterranean 2008, Sand dunes as a major proximal dust source

Geological Society of America Bulletin, January/February 2014 63 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Roskin et al.

for late Pleistocene loess in the Negev Desert, Israel: the linear sand ridges of west Kimberley of North-West Livingstone, I., 1987, Grain-size variation on a ‘complex’ Quaternary Research, v. 70, no. 2, p. 275–282, doi: Australia: The Geographical Journal, v. 159, no. 3, linear dune in the Namib Desert, in Frostick, L., and 10.1016 /j.yqres .2008 .04.011. p. 306–317, doi: 10.2307 /3451280. Reid, I., eds., Desert Sediments: Ancient and Modern: Crouvi, O., Amit, R., Porat, N., Gillespie, A.R., McDonald, Guan, Q., Zhang, J., Wang, L., Pan, B., Gui, H., and Zhang, Oxford, UK, Blackwell, p. 281–291. E.V., and Enzel, Y., 2009, Signifi cance of primary hill- C., 2013, Discussion of the relationship between Magaritz, M., and Enzel, Y., 1990, Standing-water deposits top loess in reconstructing dust chronology, accretion dustfall grain size and the desert border, taking the as indicators of late Quaternary dune migration in the rates, and sources: An example from the Negev Desert, southern border of the Tengger Desert and the south- northwestern Negev, Israel: Climatic Change, v. 16, Israel: Journal of Geophysical Research–Earth Surface, ern dust deposit area as an example: Palaeogeography, p. 307–318, doi: 10.1007 /BF00144506. v. 114, F02017, doi: 10.1029 /2008JF001083. Palaeoclimatology, Palaeoecology, v. 386, p. 1–7, doi: McGee, D., Broecker, W.S., and Winckler, G., 2010, Danin, A., Bar-Or, Y., Dor, I., and Yisraeli, T., 1989, The role of 10.1016 /j.palaeo .2013 .01.017. Gustiness: The driver of glacial gustiness?: Qua- cyanobacteria in stabilization of sand dunes in southern Haliva-Cohen, A., Stein, M., Goldstein, S.L., Sandler, A., ternary Science Reviews, v. 29, p. 2340–2350, doi: Israel: Ecologica Mediterranea, v. 15, no. 1/2, p. 55–64. and Starinsky, A., 2012, Sources and transport routes 10.1016 /j.quascirev .2010 .06.009. El-Sayed, M.I., 1999, Sedimentological characteristics and of fi ne detritus material to the late Quaternary Dead McGee, D., deMenocal, P.B., Winckler, G., Stuut, J.B.W., morphology of the eolian sand dunes in the eastern Sea basin: Quaternary Science Reviews, v. 50, p. 55– and Bradtmiller, L.I., 2013, The magnitude, timing and part of the UAE: A case study from Ar Rub’ Al Khali: 70, doi: 10.1016 /j.quascirev .2012 .06.014. abruptness of changes in North African dust deposition Sedimentary Geology, v. 123, no. 3–4, p. 219–238, doi: Hamann, Y., Ehrmann, W., Schmiedl, G., Kruger, S., Stuut, over the last 20,000 yr: Earth and Planetary Science 10.1016 /S0037 -0738 (98)00116-X. J.-B., and Kuhnt, T., 2008, Sedimentation processes in the Letters, v. 371–372, p. 163–176, doi:10.1016 /j.epsl Enzel, Y., Arnit, R., Dayan, U., Crouvi, O., Kahana, R., Ziv, Eastern Mediterranean Sea during the Late Glacial and .2013 .03.054. B., and Sharon, D., 2008, The climatic and physio- Holocene revealed by end-member modeling of the ter- McLaren, P., 1981, An interpretation of trends in grain size graphic controls of the eastern Mediterranean over the rigenous fraction in marine sediments: Marine Geology, measures: Journal of Sedimentary Petrology, v. 51, late Pleistocene climates in the southern Levant and v. 248, p. 97–114, doi: 10.1016 /j.margeo .2007 .10.009. p. 611–624. its neighboring deserts: Global and Planetary Change, Harris, S.A., 1958, Differentiation of various Egyptian McLaren, P., and Bowles, D., 1985, The effects of sediment v. 60, no. 3–4, p. 165–192, doi: 10.1016 /j.gloplacha eolian microenvironments by mechanical composi- transport on grain-size distribution: Journal of Sedi- .2007 .02.003. tion: Journal of Sedimentary Research, v. 28, no. 2, mentary Petrology, v. 55, p. 457–470. Enzel, Y., Amit, R., Crouvi, O., and Porat, N., 2010, Abra- p. 164–174. McTainsh, G., 1984, The nature and origin of the eolian sion-derived sediments under intesnsifi ed winds at the Harrison, J.B.J., and Yair, A., 1998, Late Pleistocene eolian mantles of central northern Nigeria: Geoderma, v. 33, latest Pleistocene leading edge of the advancing Sinai- and fl uvial interactions in the development of the Niz- p. 13–37, doi: 10.1016 /0016 -7061 (84)90087-9. Negev erg: Quaternary Research, v. 74, no. 1, p. 121– zana dune fi eld, Negev Desert, Israel: Sedimentology, Merrison, J.P., 2012, Sand transport, erosion and granular 131, doi: 10.1016 /j.yqres .2010 .04.002. v. 45, no. 3, p. 507–518, doi: 10.1046 /j.1365 -3091 .1998 electrifi cation: Aeolian Research, v. 4, p. 1–16, doi: Fitzsimmons, K.E., Megge, J.W., and Amos, K.J., 2009, .00151.x. 10.1016 /j.aeolia.2011 .12.003. Characterisation of eolian sediments from the Harrison, S.P., Kohfeld, K.E., Roelandt, C., and Claquin, T., Merrison, J.P., Gunnlaugsson, H.P., Knak Jensen, S., and Strzelecki and Tirari Deserts, Australia: Implications 2001, The role of dust in climate changes today, at the Nørnberg, P., 2010, Mineral alteration induced by sand for reconstructing paleoenvironmental conditions: Last Glacial Maximum and in the future: Earth-Science transport: A source for the reddish color of Martian dust: Sedimentary Geology, v. 218, no. 1–4, p. 61–73, doi: Reviews, v. 54, no. 1–3, p. 43–80, doi:10.1016 /S0012 Icarus, v. 205, no. 2, p. 716–718, doi:10.1016 /j.icarus 10.1016 /j.sedgeo .2009 .04.004. -8252 (01)00041-1. .2009 .09.004. Folk, R.L., 1971, Longitudinal dunes of the northwestern Hermas, E., Leprince, S., and El-Magd, I.A., 2012, Retriev- Milewski, A., Sultan, M., Yan, E., Becker, R., Abdeldayem, edge of the Simpson desert, Northern Territory, Aus- ing sand dune movements using sub-pixel correlation A., Soliman, F., and Abdel Gelil, K., 2009, A remote tralia: l. Geomorphology and grain size relationships: of multi-temporal optical remote sensing imagery, sensing solution for estimating runoff and recharge Sedimentology, v. 16, p. 5–54, doi: 10.1111 /j.1365 -3091 northwest Sinai Peninsula, Egypt: Remote Sensing of in arid environments: Journal of Hydrology, v. 373, .1971 .tb00217.x. Environment, v. 121, p. 51–60, doi:10.1016 /j.rse .2012 p. 1–14, doi: 10.1016 /j.jhydrol .2009 .04.002. Frihy, O.E., and Stanley, D.J., 1987, Texture and coarse frac- .01.002. Misak, R.F., and Draz, M.Y., 1997, Sand drift control of tion composition of Nile Delta deposits: Facies analy- Hesse, P.P., and Simpson, R.L., 2006, Variable vegetation selected coastal and desert dunes in Egypt—Case sis and stratigraphic correlation: Journal of African cover and episodic sand movement on longitudinal studies: Journal of Arid Environments, v. 35, no. 1, Earth Sciences, v. 71, no. 1, p. 237–255. Frihy, O.E., El desert sand dunes: Geomorphology, v. 81, p. 276–291, p. 17–28, doi: 10.1006 /jare .1995 .0134. Askary, M.A., Deghidy, E.M., and Mouf addal, W.M., doi: 10.1016 /j.geomorph .2006 .04.012. Mohamed, E.S., 2012, Spatial assessment of desertifi cation in 1998, Distinguishing fluvio-marine environments in Holliday, V.T., 1989, The Blackwater Draw Formation (Qua- north Sinai using modifi ed MEDLAUS model: Arabian the Nile Delta using heavy minerals: Journal of Coastal ternary): A 1–4-plus-m.y. record of eolian sedimenta- Journal of Geosciences, p. 1–13, doi: 10.1007 /s12517 Research, v. 14, no. 3, p. 970–980. tion and soil formation on the southern High Plains: -012 -0723-2. Fryberger, S.G., 1979, Dune forms and wind regime, in Geological Society of America Bulletin, v. 101, no. 12, Muhs, D.R., and Holliday, V.T., 1995, Evidence of active McKee, E.D., ed., A Study of Global Sand Seas: p. 1598–1607, doi: 10.1130 /0016 -7606 (1989)101 <1598: dune sand on the Great Plains in the 19th century from U.S. Geological Survey Professional Paper 1052, TBDFQA>2.3.CO;2. accounts of early explorers: Quaternary Research, p. 137–169. Hunt, S.-J.C., 1991, Provenance and Paleoclimate of Late v. 43, no. 2, p. 198–208, doi: 10.1006 /qres .1995 .1020. Gad, A., 2004, A Parametric Approach in Assessment of Soil Quaternary Eolian Sediments, Northern Negev Desert, Muhs, D.R., and Holliday, V.T., 2001, Origin of late Quater- Degradation Using Remote Sensing and GIS. A Case Israel [Ph.D. thesis]: Cambridge, UK, University of nary dune fi elds on the southern High Plains of Texas Study in North Sinai: National Authority for Remote Cambridge, 122 p. and New Mexico: Geological Society of America Bul- Sensing and Space Sciences, Egypt, presentation for the Jullien, E., Grousset, F., Bruno, M., Duprat, J., Sanchez- letin, v. 113, no. 1, p. 75–87, doi: 10.1130 /0016 -7606 United Nations Offi ce for Outer Space Affairs, 30 p. Goni, M.F., Eynaud, F., Charlier, K., Schneider, R., (2001)113 <0075: OOLQDF>2.0.CO;2. Ganor, E., Foner, H.A., Brenner, S., Neeman, E., and Lavi, Bory, A., Bout, V., and Flores, J.A., 2007, Low-latitude Muhs, D.R., Aleinikoff, J.N., Stafford, T.W., Kihl, R., Been, N., 1991, The chemical composition of aerosols set- dusty events vs. high-latitude icy Heinrich events: J., Mahan, S.A., and Cowherd, S., 1999, Late Qua- tling in Israel following dust storms: Atmospheric En- Quaternary Research, v. 68, no. 3, p. 379–386, doi: ternary loess in northeastern Colorado: Part I—Age vironment, v. 25A, no. 12, p. 2665–2670. 10.1016 /j.yqres .2007 .07.007. and paleoclimatic signifi cance: Geological Society of Ginzburg, D., and Yaalon, D.H., 1963, Petrography and ori- Junge, C., 1979, The importance of mineral dust as an atmo- America Bulletin, v. 111, no. 12, p. 1861–1875, doi: gin of the loess in the Be’er Sheva Basin: Israel Journal spheric constituent, in Morales, C., ed., Saharan Dust: 10.1130 /0016 -7606 (1999)111 <1861: LQLINC>2.3 of Earth Sciences, v. 12, p. 68–70. Chichester, UK, Wiley, p. 49–60. .CO;2. Glover, P.W.J., and Walker, E., 2009, Grain-size to effective Khalaf, F.I., and Gharib, I.M., 1985, Roundness parameters Muhs, D.R., Bettis, E.A., Aleinikoff, J.N., McGeehin, J.P., pore-size transformation derived from electrokinetic of quartz sand grains of recent eolian sand deposits in Beann, J., Skipp, G., Marshall, B.D., Roberts, H.M., theory: Geophysics, v. 74, no. 1, p. E17–E29, doi: Kuwait: Sedimentary Geology, v. 45, p. 147–158. Johnson, W.C., and Benton, R., 2008, Origin and 10.1190 /1.3033217. Kidron, G.J., Vonshak, A., and Abeliovich, A., 2009, Micro- paleo climatic signifi cance of late Quaternary loess in Goldberg, P., 1977, Late Quaternary stratigraphy of Gebel biotic crusts as biomarkers for surface stability and Nebraska: Evidence from stratigraphy, chronology, Maghara, in Bar-Yosef, O., and Phillips, C.J., eds., wetness duration in the Negev Desert: Earth Surface sedimentology, and geochemistry: Geological Society Prehistoric Investigations in Gebel Maghara, Northern Processes and Landforms, v. 34, no. 12, p. 1594–1604, of America Bulletin, v. 120, no. 11–12, p. 1378–1407, Sinai: Hungary, Elsevier. doi: 10.1002 /esp.1843. doi: 10.1130 /B26221.1. Goring-Morris, A.N., and Goldberg, P., 1990, Late Quater- Kocurek, G., 1991, Interpretation of ancient eolian sand Muhs, D.R., Roskin, J., Tsoar, H., Skipp, G., Budhan, J.R., nary dune incursions in the southern Levant: Archaeol- dunes: Annual Review of Earth and Planetary Sciences, Sneh, A., Porat, N., Stanley, J.-D., Katra, I., and Blum- ogy, chronology and palaeoenvironments: Quaternary v. 19, p. 43–75, doi:10.1146 /annurev .ea .19 .050191 berg, D.G., 2013, Origin of the Sinai-Negev erg, Egypt International, v. 5, p. 115–137, doi: 10.1016 /1040 -6182 .000355. and Israel: Mineralogical and geochemical evidence (90)90031-X. Lancaster, N., 1995, Geomorphology of Desert Dunes: Lon- for the importance of the Nile and sea level history: Goudie, A.S., and Watson, A., 1981, The shape of desert don, Routledge, 290 p. Quaternary Science Reviews, v. 69, p. 28–48, doi: sand dune grains: Journal of Arid Environments, v. 4, Levin, N., Kidron, G.J., and Ben-Dor, E., 2006, The spatial 10.1016 /j.quascirev .2013 .02.022. p. 185–190. and temporal variability of sand erosion across a stabi- Mulitza, S., Heslop, D., Pittauerova, D., Fischer, H.W., Goudie, A.S., Stokes, S., Livingstone, I., Bailiff, I.K., and lizing coastal dune fi eld: Sedimentology, v. 53, no. 4, Meyer, I., Stuut, J.-B., Zabel, M., Mollenhauer, G., Col- Allison, R.J., 1993, Post-depositional modifi cation of p. 697–715, doi: 10.1111 /j.1365 -3091 .2006 .00787.x. lins, J.A., Kuhnert, H., and Schulz, M., 2010, Increase

64 Geological Society of America Bulletin, January/February 2014 Downloaded from gsabulletin.gsapubs.org on January 6, 2014 Particle-size fractionation of eolian sand

in African dust fl ux at the onset of commercial agricul- ing and conceptual modeling: Aeolian Research, v. 9, Tsoar, H., and Möller, J.T., 1986, The role of vegetation in ture in the Sahel region: Nature, v. 466, 7303, p. 226– p. 161–173, doi: 10.1016 /j.aeolia.2013 .02.002. the formation of linear sand dunes, in Nickling, W.G., 228, doi: 10.1038 /nature09213. Smalley, I.J., and Krinsley, D.H., 1978, Loess deposits ed., Eolian Geomorphology: Boston, Allen and Unwin, Munyikwa, K., Telfer, M.W., Baker, I., and Knight, C., 2011, asso ciated with deserts: Catena, v. 5, p. 53–66, doi: p. 75–95. Core drilling of Quaternary sediments for lumines- 10.1016 /S0341 -8162 (78)80006-X. Tsoar, H., and Pye, K., 1987, Dust transport and the ques- cence dating using the Dormer Drillmite: Ancient TL, Smalley, I.J., and Vita-Finzi, C., 1968, The formation of fi ne tion of desert loess formation: Sedimentology, v. 34, v. 29, no. 1, p. 15–23. particles in sandy deserts and the nature of ‘desert’ p. 139–153, doi: 10.1111 /j.1365 -3091 .1987 .tb00566.x. Okin, G.S., and Painter, T.H., 2004, Effect of grain size on loess: Journal of Sedimentary Petrology, v. 38, no. 3, Tsoar, H., White, B., and Berman, E., 1996, The effect of remotely sensed spectral refl ectance of sandy desert p. 766–774. slopes on sand transport—Numerical modeling: Land- surfaces: Remote Sensing of Environment, v. 89, no. 3, Sneh, A., and Weissbrod, T., 1983, Size-frequency distribu- scape and Urban Planning, v. 34, no. 3–4, p. 171–181. p. 272–280, doi: 10.1016 /j.rse .2003 .10.008. tions of longitudinal dune rippled fl ank sands com- Tsoar, H., Blumberg, D.G., and Stoler, Y., 2004, Elonga- Olson, C.G., Nettleton, W.D., Porter, D.A., and Brasher, pared to that of slip face sands of various dune types: tion and migration of sand dunes: Geomorphology, B.R., 1997, Middle Holocene eolian activity on the Sedimentology, v. 30, p. 717–725, doi:10.1111 /j.1365 v. 57, no. 3–4, p. 293–302, doi: 10.1016 /S0169 -555X High Plains of west-central Kansas: The Holocene, v. 7, -3091 .1983 .tb00705.x. (03)00161-2. no. 3, p. 255–261, doi: 10.1177 /095968369700700301. Stanley, D.J., 2002, Confi guration of the Egypt-to-Canaan Tsoar, H., Blumberg, D.G., and Wenkart, R., 2008, Forma- Powers, M.C., 1953, A new roundness scale for sedimen- coastal margin and North Sinai byway in the Bronze tion and geomorphology of the NW Negev sand dunes, tary particles: Journal of Sedimentary Petrology, Age, in van den Brink, E.C.M., and Levy, T.E., eds., in Breckle, S.W., Yair, A., and Veste, M., eds., Arid v. 23, p. 117–119, doi: 10.1306 /D4269567 -2B26 -11D7 Egypt and the Levant: Interrelations from the 4th Dune Ecosystems: Berlin, Springer, p. 25–48. -8648000102C1865D. through the Early 3rd Millennium B.C.E.: London, Vaks, A., Bar-Matthews, M., Ayalon, A., Matthews, A., Pye, K., 1995, The nature, origin and accumulation of loess: Leicester University Press, p. 98–117. Frumkin, A., Dayan, U., Halicz, L., Almogi-Labin, Quaternary Science Reviews, v. 14, p. 653–667, doi: Stanley, D.J., and Warne, A.G., 1993, Nile Delta: Recent A., and Schilman, B., 2006, Paleoclimate and loca- 10.1016 /0277 -3791 (95)00047-X. geological evolution and human impact: Science, tion of border between Mediterranean climate region Pye, K., and Tsoar, H., 1987, The mechanics and geologi- v. 260, 5108, p. 628–634, doi:10.1126 /science .260 and the Saharo-Arabian Desert as revealed by speleo- cal implications of dust transport and deposition in .5108.628. thems from the northern Negev Desert, Israel: Earth deserts with particular reference to loess formation and Stanley, D.J., McRea, J.E., Jr., and Waldon, J.C., 1996, Nile and Planetary Science Letters, v. 249, p. 384–399, doi: dune sand diagenesis in the northern Negev, Israel, in Delta Core and Sample Database for Years 1985–1994, 10.1016 /j.epsl .2006 .07.009. Frostick, L., and Reid, I., eds., Desert Sediments: An- MEDIBA Program, Smithsonian Contribution to the Vaks, A., Woodhead, J., Bar-Matthews, M., Ayalon, A., cient and Modern: Oxford, UK, Blackwell, p. 139–156. Marine Sciences: Washington, D.C., Smithsonian In- Cliff, R.A., Zilberman, T., and Frumkin, A., 2013, Pye, K., and Tsoar, H., 2009, Eolian Sand and Sand Dunes: stitution Press, 431 p. Pliocene–Pleistocene climate of the northern margin Berlin, Springer, 458 p. Stone, A.E.C., and Thomas, D.S.G., 2008, Linear dune ac- of Saharan–Arabian Desert recorded in speleothems Rabie, F.H., Abdel Rahman, S.I., Hanna, F., and Kotb, M.M., cumulation chronologies from the southwest Kalahari, from the Negev Desert, Israel: Earth and Planetary 2000, Study of sand dunes migration and their effect Namibia: Challenges of reconstructing late Quater- Science Letters, v. 368, p. 88–100, doi:10.1016 /j.epsl on the agricultural development in North Sinai, Egypt, nary paleoenvironments from eolian landforms: Qua- .2013 .02.027. using remote sensing: 2nd International Conference ternary Science Reviews, v. 27, p. 1667–1681, doi: White, B., and Tsoar, H., 1998, Slope effect on saltation over on Earth Observation and Environmental Information, 10.1016 /j.quascirev .2008 .06.008. a climbing sand dune: Geomorphology, v. 22, no. 2, Cairo, Egypt (CD-ROM). Sun, D., Bloemendal, J., Rea, D.K., Vandenberghe, J., Jiang , p. 159–180, doi: 10.1016 /S0169 -555X (97)00058-5. Ravikovich, S., 1981, The Soils of Israel, Formation, Nature F., An, Z., and Su, R., 2002, Grain-size distribution Wright, J., Smith, B., and Whalley, B., 1998, Mechanisms and Properties: Tel Aviv, Hakibbutz Hameuchad, 332 function of polymodal sediments in hydraulic and of loess-sized quartz silt production and their relative p. (in Hebrew) aeolian environments, and numerical partitioning of effectiveness: Laboratory simulations: Geomorphol- Roskin, J., Porat, N., Tsoar, H., Blumberg, D.G., and Zander, the sedimentary components: Sedimentary Geology, ogy, v. 23, no. 1, p. 15–34, doi:10.1016 /S0169 -555X A.M., 2011a, Age, origin and climatic controls on veg- v. 152, no. 3–4, p. 263–277, doi:10.1016 /S0037 -0738 (97)00084-6. etated linear dunes in the northwestern Negev desert (02)00082-9. Xu, S.-J., Bao-Tian, P., Hong-Shan, G., Guang-Jie, C., and (Israel): Quaternary Science Reviews, v. 30, no. 13–14, Telfer, M.W., 2011, Growth by extension, and reworking, Huai, S., 2007, Changes in sand fractions of Bing- p. 1649–1674, doi: 10.1016 /j.quascirev .2011 .03.010. of a south-western Kalahari linear dune: Earth Surface gou section and the expansion and contraction of Roskin, J., Tsoar, H., Porat, N., and Blumberg, D.G., 2011b, Processes and Landforms, v. 36, no. 8, p. 1125–1135, the Tengger Desert during 50–30 ka: Earth Surface Paleoclimate interpretations of late Pleistocene vege- doi: 10.1002 /esp.2140. Processes and Landforms, v. 32, p. 475–480, doi: tated linear dune mobilization episodes; evidence from Tjallingii, R., Claussen, M., Stuut, J.-B.W., Fohlmeister, J., 10.1002 /esp.1439. the northwestern Negev Desert, Israel: Quaternary Sci- Jahn, A., Bickert, T., Lamy, F., and Rohl, U., 2008, Co- Yaalon, D.H., 1987, Saharan dust and desert loess: Effect ence Reviews, v. 30, no. 23–24, p. 3364–3380, doi: herent high- and low-latitude control of the northwest on surrounding soils: Journal of African Earth Sci- 10.1016 /j.quascirev .2011 .08 .014. African hydrological balance: Nature Geoscience, v. 1, ences, v. 6, no. 4, p. 569–571, doi: 10.1016 /0899 -5362 Roskin, J., Blumberg, D. G., Porat, N., Tsoar, H., and no. 10, p. 670–675. (87)90096-0. Rozenstein, O., 2012, Do dune sands redden with age? Tsoar, H., 1974, Desert dunes morphology and dynamics, Yaalon, D.H., and Dan, J., 1974, Accumulation and distribu- The case of the northwestern Negev dunefi eld, Israel: El-Arish (Northern Sinai): Zeitschrift für Geomorphol- tion of loess-derived deposits in the semi-desert fringe Eolian Research, v. 5, p. 63–75. ogie N.F., v. 20, supplement, p. 41–61. areas of Israel: Zeitschrift für Geomorphologie N.F., Roskin, J., Porat, N., Bowman, D., Katra, I., and Eyal, Y., Tsoar, H., 1976, Characterization of sand dune environments v. 20, supplement, p. 91–105. 2013a, Opening stages of open fractures upon the west- by their grain-size, mineralogy and surface texture, in Yair, A., 2008, Effects of surface runoff and subsurface fl ow ern margin of the Dead Sea Rift, (Israel) based on OSL Amiran, D.H.K., and Ben-Arieh, Y., eds., Geography on the spatial variability of water resources in longitu- age estimates of their sediment fi ll, in Ergozi, R., and in Israel: Jerusalem, Israel National Committee, Inter- dunal dunes, in Breckle, S.-W., et al., eds., Arid Dune Laronne, J., eds., Israel Geomorphological Research national Geographical Union, p. 327–343. Ecosystems: Berlin, Springer, p. 251–269. Group (IGRG) 2013 Conference Abstracts: Ein Gedi, Tsoar, H., 1978, The Dynamics of Longitudinal Dunes: U.S. Yair, A., Lavee, H., and Greitser, N., 1997, Spatial and tem- p. 72. Army, European Research Offi ce, London, 171 p. poral variability of water percolation and movement Roskin, J., Katra, I., and Blumberg, D.G., 2013b, Late Holo- Tsoar, H., 1986, Two-dimensional analysis of dune profi le and in a system of longitudinal dunes, western Negev, Is- cene dune mobilizations in the northwestern Negev the effect of grain size on sand dune morphology, in El- rael: Hydrological Processes, v. 11, no. 1, p. 43–58, dunefield, Israel: A response to combined anthro- Baz, F., and Hassan, M.H.A., eds., Physics of Desertifi - doi: 10.1002 /(SICI)1099 -1085 (199701)11: 1 <43:: AID pogenic activity and short-term intensified windi- cation: Dordrecht, Martinus Nijhoff, p. 94–108. -HYP402>3.0 .CO;2-D. ness: Quaternary International, v. 303, p. 10–23, doi: Tsoar, H., 1989, Linear dunes—Forms and formation: Progress Yizhaq, H., Ashkenazy, Y., and Tsoar, H., 2009, Sand dune 10.1016 /j.quaint .2012 .10.034. in Physical Geography, v. 13, no. 4, p. 507–528, doi: dynamics and climate change: A modeling approach: Roskin, J., Katra, I., Porat, N., and Zilberman, E., 2013c, 10.1177 /030913338901300402. Journal of Geophysical Research–Earth Surface, Evolution of Middle to Late Pleistocene Sandy Cal- Tsoar, H., 1990, Trends in the development of sand dunes v. 114, F01023, doi: 10.1029 /2008JF001138. careous Palaeosols underlying the northwestern Negev along the southeastern Mediterranean coast: Catena, Zilberman, E., Porat, N., and Roskin, J., 2007, The Middle Desert Dunefi eld (Israel): Palaeogeography, Palaeo- v. 18, supplement, p. 51–60. to Late-Pleistocene Sand Sheet Sequence of Kerem climatology, Palaeoecology, v. 387, p. 134–152. Tsoar, H., 1995, Desertifi cation in northern Sinai in the eigh- Shalom, Western Negev—An Archive of Coastal Sand Sarnthein, M., 1978, Sand deserts during glacial maximum teenth century: Climatic Change, v. 29, no. 4, p. 429– Incursion: Geological Survey of Israel, 23 p. and climatic optimum: Nature, v. 272, no. 5648, p. 43– 438, doi: 10.1007 /BF01092427. 46, doi: 10.1038 /272043a0. Tsoar, H., 2005, Sand dunes mobility and stability in rela- SCIENCE EDITOR: NANCY RIGGS Sevink, J., and Kummer, E.A., 1984, Eolian dust deposition tion to climate: Physica A, v. 357, no. 1, p. 50–56, doi: ASSOCIATE EDITOR: ANNE JEFFERSON on the Giara di Gesturi basalt plateau, Sardinia: Earth 10.1016 /j.physa .2005 .05.067. Surface Processes and Landforms, v. 9, no. 4, p. 357– Tsoar, H., 2013, Critical environments: Sand dunes and cli- MANUSCRIPT RECEIVED 25 OCTOBER 2012 364, doi: 10.1002 /esp .3290090408. mate change, in Shroder, J., Lancaster, N., Sherman, REVISED MANUSCRIPT RECEIVED 14 JULY 2013 Siegal, Z., Tsoar, H., and Karnieli, A., 2013, Effects of pro- D.J., and Baas, A.C.W., eds., Treatise on Geomorphol- MANUSCRIPT ACCEPTED 13 SEPTEMBER 2013 longed drought on the vegetation cover of sand dunes ogy, Volume 11: Aeolian Geomorphology: San Diego, in the NW Negev Desert: Field survey, remote sens- Academic Press, p. 414–427. Printed in the USA

Geological Society of America Bulletin, January/February 2014 65