Geology of the Lothidok Range, Northern Kenya

This study provides a stratigraphic and geochronologic context for fossils collected from the Lothidok Range, northern Kenya, that include three new hominoid genera. Age control on the strata is provided by 2 2 potassium-argon determinations at 11 stratigraphic levels. Three fossil bearing horizons in the Lothidok Range are: (1) pre-17.7 Ma, (2) between 17.7 and 16.6 Ma and (3) between 13.6 and 12.0 Ma. Fauna from the lower two horizons are equivalent to the collections from Rusinga and the highest horizon is equivalent to, or slightly younger than, those from Fort Teman. Exposed Tertiary strata of the Lothidok Range are 1500 m thick, but the base is not exposed. In ascending order the section comprises 785 m of basalts and intercalated sedimentary rocks (Kalakol basalts, > 17.7 Ma), 540 m of interbedded sedimentary, pyroclastic and basaltic rocks (lower and upper Lothidok Formation, 17.7 - 12.0 Ma), 120 m maximum of basalts (Loperi basalts, 12.0 - 10.9 Ma) and 50 m minimum of undifferentiated Tertiary sedimentary strata (< 10.9 Ma). Sedimentary rocks of the lower Lothidok Formation consist primarily of polymictic conglomerates and conglomeratic litharenites with southwest paleotransport directions, and are interpreted as braided stream-alluvial fan deposits. The upper Lothidok Formation contains polymictic conglomerates and conglomeratic and feldspathic litharenites near its base, with finer grained lithic to arkosic sandstones near the top. The lower strata are interpreted as braided stream-alluvial fan deposits, and the higher strata are interpreted as meandering stream deposits. All upper Lothidok Formation strata have easterly paleotransport directions. Pyroclastic rocks of the Lothidok Formation consist of mafic alkaline and trachytic tephra and lahars. A disconformity within the Lothidok Formation represents a time gap from -16.4 to ~13.7 Ma. An angular unconformity exists between the Lothidok Formation and the Loperi basalts, and a third unconformity may exist within the undifferentiated Tertiary strata. The structure is dominated by late Tertiary normal faults. Only one fault predates deposition of the Lothidok Formation. Other faults postdate the undifferentiaited Tertiary strata and the range front fault postdates 4.1 Ma. No source terrains for the sediments and pyroclastic rocks are known. The sedimentary and pyroclastic rock record, however, provides evidence for a proximal, early Miocene volcanic highland and mafic alkaline volcanic center close to and east of the Lothidok Range as well as middle Miocene trachytic center(s) to the south.

GEOLOGY OF THE LOTHIDOK RANGE, NORTHERN KENYA by Harold Bradley Boschetto A thesis submitted to the faculty of the University of Utah partial fulfillment of the requirements for the degree of Master of Science in Geology Department of Geology and Geophysics The University of Utah August 1988 Copyright © Harold Bradley Boschetto 1988 All Rights Reserved THE UNIVERSITY OF UTAH GRADUATE SCHOOL SUPERVISORY COMMITTEE APPROVAL of a thesis submitted by Harold Bradley Boschetto This thesis has been read by each member of the following supervisory committee and by major vote has been found to be satisfactory. / . ^ / 7 / _______________________________ Chair: Francis H. Brown X /% ? Maijorie A. Chan X~ ^ / ^ ' t_feennis M. Bramble THE UNIVERSITY OF UTAH GRADUATE SCHOOL FINAL READING APPROVAL To the Graduate Council of the University of Utah: I have read the thesis o f________________ Harold Bradley Boschetto_____________ in its final form and have found that ( 1 ) its format, citations, and bibliographic style are consistent and acceptable; (2 ) its illustrative materials including figures, tables, and charts are in place; (3) the final manuscript is satisfactory to the Supervisory Committee and is ready for submission to the Graduate School. / • X / ? / Date Approved for the Major Department Francis H. Brown Chair/Dean Approved for the Graduate Council Francis H. Brown Chair, Supervisory Committee B. Gale Dick Dean of The Graduate School ABSTRACT This study provides a stratigraphic and geochronologic context for fossils collected from the Lothidok Range, northern Kenya, that include three new hominoid genera. Age control on the strata is provided by 2 2 potassium-argon determinations at 11 stratigraphic levels. Three fossil bearing horizons in the Lothidok Range are: (1) pre-17.7 Ma, (2) between 17.7 and 16.6 Ma and (3) between 13.6 and 12.0 Ma. Fauna from the lower two horizons are equivalent to the collections from Rusinga and the highest horizon is equivalent to, or slightly younger than, those from Fort Teman. Exposed Tertiary strata of the Lothidok Range are 1500 m thick, but the base is not exposed. In ascending order the section comprises 785 m of basalts and intercalated sedimentary rocks (Kalakol basalts, > 17.7 Ma), 540 m of interbedded sedimentary, pyroclastic and basaltic rocks (lower and upper Lothidok Formation, 17.7 - 12.0 Ma), 120 m maximum of basalts (Loperi basalts, 12.0 - 10.9 Ma) and 50 m minimum of undifferentiated Tertiary sedimentary strata (< 10.9 Ma). Sedimentary rocks of the lower Lothidok Formation consist primarily of polymictic conglomerates and conglomeratic litharenites with southwest paleotransport directions, and are interpreted as braided stream-alluvial fan deposits. The upper Lothidok Formation contains polymictic conglomerates and conglomeratic and feldspathic litharenites near its base, with finer grained lithic to arkosic sandstones near the top. The lower strata are interpreted as braided stream-alluvial fan deposits, and the higher strata are interpreted as meandering stream deposits. All upper Lothidok Formation strata have easterly paleotransport directions. Pyroclastic rocks of the Lothidok Formation consist of mafic alkaline and trachytic tephra and lahars. A disconformity within the Lothidok Formation represents a time gap from -16.4 to ~13.7 Ma. An angular unconformity exists between the Lothidok Formation and the Loperi basalts, and a third unconformity may exist within the undifferentiated Tertiary strata. The structure is dominated by late Tertiary normal faults. Only one fault predates deposition of the Lothidok Formation. Other faults postdate the undifferentiaited Tertiary strata and the range front fault postdates 4.1 Ma. No source terrains for the sediments and pyroclastic rocks are known. The sedimentary and pyroclastic rock record, however, provides evidence for a proximal, early Miocene volcanic highland and mafic alkaline volcanic center close to and east of the Lothidok Range as well as middle Miocene trachytic center(s) to the south. v To my mother, Roberta, for her continual support and encouragement, and to Anne Pasch, for sparking my interest in this fascinating science. TABLE OF CONTENTS ABSTRACT....................................................................................................................... iv LIST OF F IG U R E S ....................................................................................................ix LIST OF TABLES.......................................................................................................... xii ACKNOWLEDGMENTS............................................................................................. xiii INTRODUCTION........................................................................................................... 1 Geographic N a m e s ............................................................................................. 1 Previous Work in th Region..................................................................................... 6 Methods of S t u d y ....................................................................................................8 Potassium-Argon Age Determination...................................................................... 11 REGIONAL GEOLOGY OF THE TURK ANA D E P R E S S IO N ...........................14 Regional Structu re ....................................................................................................14 Regional S tra tig rap h y ............................................................................................. 14 STRATIGRAPHY OF THE LOTHIDOK RANGE.....................................................20 Stratigraphy................................................................................................................20 Kalakol Basalts - T k b ............................................................................................. 30 Lothidok Formation - T 1 .......................................................................................35 Loperi Basalts - T i b ............................................................................................. 80 Undifferentiated Tertiary Deposits - T u ............................................................83 INTERPRETATION OF DEPOSITIONAL ENVIRONMENTS...........................8 8 I n t r o d u c t i o n .......................................................................................................... 8 8 Lithologic A s s o c i a t io n s .......................................................................................90 ORIGINS OF THE LOTHIDOK D E PO S IT S ............................................................100 Sedimentary Provenance........................................................................................... 100 Pyroclastic Source .................................................................................................... 101 UNCONFORMITIES IN THE LOTHIDOK STRATIGRAPHY...........................105 Base of the Lokipenata Conglomerate...................................................................105 Base of the Loperi B a s a lts .......................................................................................109 Base of the Tertiary Arkosic S a n d s to n e s ............................................................ 110 STRUCTURE OF THE LOTHIDOK RA N G E ...........................................................113 General S tr u c tu r e ....................................................................................................113 Normal F a u l t s .......................................................................................................... 113 Grabens and H o r s ts ................................................................................................. 119 Temporal Fault R e l a tio n s ...................................................................................... 122 Fault G o u g e .......................................................................................................... 122 FAUNA OF THE LOTHIDOK R A N G E ...................................................................123 I n t r o d u c t i o n .......................................................................................................... 123 Eragaleit Beds, Kalakol B a s a l t s ......................................................................... 134 Lower Lothidok Formation ......................................................................... 134 Upper Lothidok Formation....................................................................................... 136 NGAKORINGORA RIDGE STRATA......................................................................... 138 Description.................................................................................................................138 In te rp re ta tio n .......................................................................................................... 140 GEOLOGIC HISTORY AND PALEOGEOGRAPHY...............................................142 I n t r o d u c t i o n .......................................................................................................... 142 -25 to 17.7 M a ....................................................................................................142 17.7 to 17.3 M a .........................................................................................................143 17.3 t o -16.4 M a ....................................................................................................149 -16.4 t o -13.7 M a ....................................................................................................152 -13.7 t o -13.0 M a ....................................................................................................152 -13.0 to 10.9 M a ....................................................................................................157 10.9 to -5 Ma..............................................................................................................157 CONCLUSIONS................................................................................................................163 APPENDICES A: LIST OF MEASURED SECTIONS IN THE LOTHIDOK RANGE . . 1 6 6 B: FIELD DESCRIPTIONS OF THE COMPOSITE STRATIGRAPHY OF THE LOTHIDOK RANGE......................................................................... 167 REFERENCES . ...........................................................................................................197 viii LIST OF FIGURES Figure Page 1. Location map for the Lothidok Range, northern Kenya....................................... 2 2. Geographic references for the Lothidok Range......................................................4 3. Location map for sections measured in the Lothidok Range......................................9 4. Relation of the Lothidok Range to regional geographic and structural features of the Turkana Depression.........................................................................15 5. Sequence and terminology used for the composite stratigraphy of the Lothidok Range.......................................................................................................... 21 6 . Component sections for the stratigraphy of the Lothidok Range.........................23 7. Generalized north-south cross-section and lateral distribution of the strata exposed in the Lothidok Range......................................................................26 8 . Generalized east-west cross-section and lateral distribution of the strata exposed in the Lothidok Range................................................................................ 28 9. The sequence and terminology used for the composite-stratotype of the Lothidok Formation................................................................................................... 36 10. Correlation and lateral variation in the Kalodirr Tuffs...........................................42 11. Typical sequence of the Kanukurinya tuff beds from section 33.........................45 12. Photograph of leaf imprints in the Kalodirr T u f f s ...............................................48 13. Typical sequence of the Naserte Tuffs from sections 21 and 39......................... 52 14. Photograph of accretionary lapilli in airfall tuffs of the Naserte Tuffs near at section 39....................................................................................................... 54 15. Photograph of fluid escape structures in airfall tuffs of the Naserte Tuffs near section 8 ...............................................................................................................56 16. Photograph of a palm leaf imprint in an unassigned trachytic tuff at section 27 (photo by F. H. Brown).........................................................................64 17. Photograph of a burrow in an unassigned trachytic tuff taken near section 19 (photo by F. H. Brown).........................................................................65 18. Paleocurrent rose diagrams for the Basal Conglomerate Member of the lower Lothidok Formation for data sets from near section 1 and from section 20. All data were corrected for tectonic tilt...............................................67 19. Paleocurrent rose diagrams for unassigned strata of the lower Lothidok. Formation for data sets from multiple outcrops. All were corrected for tectonic tilt................................................................................................................... 69 20. Paleocurrent rose diagrams for the Lokipenata conglomerate of the upper Lothidok Formation for data sets from sections 3 and 19. All were data corrected for tectonic tilt...........................................................................78 21. Paleocurrent rose diagrams for unassigned strata of the upper Lothidok Formation for data sets from sections 27 and 29. All were corrected for tectonic tilt............................................................................ ...........................81 22. Paleocurrent rose diagrams for undifferentiated Tertiary strata taken along the Kalodirr River. All data were corrected for tectonic tilt. . . . 8 6 23. Typical sequence of the Conglomerate-Sandstone-Mudstone (CSM) lithologic association..................................................................................................92 24. Typical sequence of the Sandstone-Siltstone-Mudstone (SSM) lithologic association................................................................................................................... 95 25. Paleocurrent rose diagrams illustrating the complete reversal of sediment transport direction above the disconformity relative to that below. Each diagram represents the combination of every measurement taken in the respective level and all measurements were corrected for tectonic tilt. . . . 1 0 6 26. Schematic geologic cross-sections of the Lothidok Range (no vertical exaggeration)...............................................................................................................114 27. Simplified structure map of the Lothidok Range illustrating location of major structural features and geologic cross sections............................................117 28. Fossil localities of the Lothidok Range: 1) Lothidok north, 2) Lothidok south, 3) Kalodirr, 4) Kanukurinya, 5) Moruorot south, 6 ) Moruorot north, 7) Esha, and 8 ) Atirr...................................................................................... 124 29. Fossil levels and isotopic ages of the Lothidok Range..........................................127 30. Paleogeographic reconstruction of the Lothidok region at ~ 18 Ma and explanation of symbols used for paleogeographic reconstructions of the Lothidok region (Figures 31-36)..............................................................................144 31. Paleogeographic reconstruction of the Lothidok region at 17.7Ma.................... 147 x 32. Paleogeographic reconstruction of the Lothidok region at 16.8 Ma . . . 150 33. Paleogeographic reconstruction of the Lothidok region at~ 13.7 Ma . . . 153 34. Paleogeographic reconstruction of the Lothidok region at~13.0 Ma . . . 155 35. Paleogeographic reconstruction of the Lothidok region at ~ 12.0 Ma . . . 158 36. Paleogeographic reconstruction of the Lothidok region between 10.9 and ~ 5 Ma..................................................................................................... 160 xi LIST OF TABLES Table Page 1. Analytical data for potassium-argon a g e s ............................................................12 2. Lithologic types of the Lothidok Formation, equivalent lithofacies codes, prominant sedimentary structures and interpretation of depositional settings. . 89 3. Measured sections, maximum and minimum ages for the fossil localities.........................................................................................................................126 4. Miocene fauna of selected East African fossil localities. The locations given are listed oldest to youngest; B - Bukwa, K - Karungu , S - Songhor, Le - Lothidok (Eragaleit beds), R - Rusinga, N - Napak, LI -lower Lothidok Fm.,Bu - Buluk, FT - Fort Teman and Lu - upper Lothidok Formation......................................................................................................................129 ACKNOWLEDGEMENTS I would first like to thank my advisor, Frank Brown, who spent seemingly limitless personal and professional time answering my questions and guiding my studies. His insight and knowledge of the Turkana Basin proved invaluable to this study. My other comittee members, Drs. M. A. Chan and D. M. Bramble, reviewed my manuscript and provided valuable comments. I would especially like to thank Richard E. and Meave G. Leakey for their kind help, gracious hospitality and logistical support in Nairobi and in the field. Funding for the 1986 field work was provided by Amoco. Field assistants, gear and vehicles were supplied by the Kenya National Museums. The 1987 field work was supported wholly by the Kenya National Museums. Logistical support and exceptional hospitality was provided by Mike Renolds while in Kenya and by Denise Stone while in Houston, Texas. Amoco also provided airphotos, LandSat images and a suite of thin sections for which I am grateful. I am deeply indebted to my field assistants, John Musa Kyeva and Paul Ngoleni Muthieni, for their impeccable assistance and for making my work in Turkana a pleasure. My sincerest appreciation is due to Musa for seeing the carpet viper that I didn't. Special thanks are also due to Ian McDougall, Tony Ekdale and William Anyonge for their invaluable help in the field. Age determinations for all 1986 samples were made by F. H. Brown under the careful guidance of Ian McDougall. Ian McDougall made all age determinations for the 1987 samples and is greatly appreciated. Terry Davies, Robin Maier and Andrew Sienko provided technical assistance for all dates. Paul Onstatt is credited for his excellent drafting. Other people who deserve special thanks for reveiwing my manuscript include Kent Wheeler, Wanda Taylor, and Rip Langford. xiv INTRODUCTION The Lothidok Range (Lothidok Hills) has been of interest to paleontology since the discovery of vertebrate fossil bearing strata there in 1932 (Arambourg, 1943). The Lothidok Range is located in the Turkana district, Northern Kenya, approximately halfway between Lake Turkana (formerly Lake Rudolf), and the town of Lodwar (Figure 1). Recently, the Kenya National Museums reported three new species of hominoids (Leakey and Leakey, 1986a, 1986b,1987) and collected numerous other vertebrate specimens in the Lothidok Range. The Miocene fossil sites in the Lothidok Range are comparable to several others in East Africa including Rusinga, Napak and Buluk. The principal objectives of this study are to establish the stratigraphy of the Lothidok Range, to place the collected fossils in a stratigraphic context, and to provide numerical age control for these fossils. The study area covers approximately 200 km2 and lies between the southern end of the Lothidok Range and the Kalakol River (Figure 2). Geographic Names Numerous variations in the names and spelling of specific geographic features in the study area are cited in the literature. The names of features referred to in this text are those of the native (Turkana) people, and were verified in the field whenever possible. Three important geographic terms have been used in many earlier publications with varied spelling and usage, and to prevent misunderstanding these are reviewed below. The first of these is 'Lothidok', which is used here to refer to both the entire range and the highest peak in the range, Lothidok Hill (Figure 2). This spelling is used on most 2 Figure 1. Location map for the Lothidok Range, northern Kenya. 3 Figure 2. Geographic references for the Lothidok Range. 5 published maps, including the 1:500,000 Geologic Sheet No. 10 (Walsh and Dodson, 1969) and the 1:250,000 Lodwar sheet (NA-36-4) issued by the Survey of Kenya. Previous workers use either Losodok' (Arambourg, 1935, 1943; Jeremine, 1935), or 'Losidok' (Fuchs, 1939; Dixey, 1945) to refer to the same area and hill. The Lothidok Range is labeled 'Muruarot' on the 1:250,000 Lodwar sheet. The second problematic term is 'Moruorot', which, following the native usage, is used here to distinguish a second large hill near the eastern edge of the range, approximately 11 km south of Lothidok Hill (Figure 2). Previous spellings in the literature include 'Muruaret' (Fuchs, 1939), 'Muruarot' (Walsh and Dodson, 1969), and both Moruaret and Moruarot (Leakey and Leakey, 1986a). Arambourg (1943) refers to this hill as Losodok'. The third problem exists with the terms used for the Kalatum and Alomonet Rivers in the southern end of the Lothidok Range, both of which refer to parts of the same stream course. In accordance with the Turkana people, this streamcourse is called the Kalatum River above its confluence with the Naserte River, and the Alomonet River below the confluence in this report (Figure 2). This river has been previously called the Lopi River (Arambourg, 1935; 1943; Walsh and Dodson, 1969) and the Lopi River Pass (Fuchs, 1939). Previous Work in the Region The first important scientific expeditions along the western margin of the lake began in the 1930s with the Lake Rudolf Expedition led by V. E. Fuchs of Cambridge University. Fuchs (1939) produced a report on the geologic history of the area based on observations during this expedition. A. M. Champion, the Turkana District Commissioner, made the first detailed geologic observations and produced the first reliable physiographic map of the region in 1937 (Walsh and Dodson, 1969). W. C. Smith (1938) provided petrographic descriptions of samples collected by Champion, most of which are well-located, and of considerable value. C. Arambourg led an expedition through Turkana to the Omo Valley, Ethiopia, in 1932-33. Arambourg (1943) briefly discussed Miocene through Quaternary sediments, fossils, stratigraphy and structure of the Lothidok Range. Although Arambourg (1943) referred to the location of fossiliferous strata as Losodok Hill, the actual site is believed to be at Moruorot Hill. Evidence for this is provided by Arambourg's map (1943) and rediscoveiy of his original quarry at Moruorot Hill by a University of California expedition in 1948 (Madden, 1972). Arambourg (1943) regarded the fauna collected from this locality as earliest Burdigalian in age (ca 22 Ma). Jeremine (1935) provided detailed petrographic descriptions and chemical analyses of some samples collected by Arambourg. The Kenyan Geological Survey has completed numerous regional geological reports on several areas within the Turkana Basin. The work most applicable to the Lothidok Range is the report by Walsh and Dodson (1969), which includes cursory discussions of the sedimentary and volcanic rocks in the Lothidok Range. Madden (1972) analyzed the Miocene fauna of the Lothidok Range based on previously collected specimens. Arguing from published descriptions of the fossiliferous sediments, from the associated gastropod fauna , and the presence of hyracoids, proboscideans, anthracotheres, and primates, he suggested that the paleoenvironment was open and semiarid, with shallow swampy lakes. Zanettin et al. (1983) measured potassium-argon ages for samples of volcanic rocks collected in the Lothidok Range. Their results are integrated into the discussions of the relevant strata. Bellieni et al. (1987) reported chemical analyses for a few of these volcanic rocks. 7 Methods of Study Field work during the summers of 1986 and 1987 consisted of mapping on aerial photos flown by the Royal Air Force in 1972 at a scale of approximately 1:50,000. Maps from individual photographs were compiled to produce the final map (Plate, in pocket). All adjustments for distortion were made visually. In addition, 36 stratigraphic sections (Figure 3; Appendix A ; Boschetto, 1988) were measured with a Jacob's staff or taping methods. Each bed in every measured section was described in detail and sampled for representative lithologies. The sections were numbered in sequence of examination. Occasionally, areas with either more than one section in a small region, or those that had no sections but were sampled, were also given numbers. Sections measured in these areas are labeled with the area number first and the section number, following a decimal, second (i.e., sections 6 . 1 and 6.2 are the first and second sections of area 6 ). Areas were not numbered in the 1987 field season and all sections were numbered consecutively. Samples numbers for the 1986 collections consist of the general location abbreviation (Los = Lothidok), and the section number followed by the sample number. For the 1987 collections, the samples are labeled with a K87 for Kenya , 1987, and are followed by a sample number. All sample numbers preceeded with 'K8 6 ' were collected by F. H. Brown. Paleocurrent measurements on sediment transport direction were taken at several localities allowing sufficient three-dimensional exposure of imbricated clasts, trough crossbed axes, and foresets of cross- and ripple stratification. Measurements of imbricated clasts were taken by measuring the azimuth of the long axis of the clast in the direction opposite to clast inclination. Measurements were taken on the azimuth of the dip of the foresets and trough crossbeds axes. The paleocurrent data were corrected for tectonic tilt by using a stereonet computer program written by Adolf Yonkee (University of Utah, 1988). A computer program (Stereonet, version, 2.6 by R. W. Allmendinger for the 8 9 Figure 3. Location map for sections measured in the Lothidok Range. Sections followed by an asterisk represent combined sections. 1 0 Macintosh) was used to calculate the mean direction and to plot rose diagrams for each data set. The results are included in the discussions addressing the stratigraphy. Due to limited exposures, a few data sets contain less than 25 measurements and may not be statistically valid (Miall, 1984). However, these measurements still serve as useful indicators of sediment dispersal. Potassium-Argon Age Determination Potassium-argon ages, determined from sanidine, plagioclase, amphibole, biotite, and whole rock basalt, provide 2 1 new ages from 11 levels to establish the geochronologic framework for the strata of the Lothidok Range. Table 1 provides all ages and the analytical data relevant to these determinations. Sanidine (Or5 5 ) was separated from pumice clasts collected from tuffs. The glass of these pumices has been totally altered to phillipsite, analcite and/or montmorillonite, and the vesicles within the pumice have been filled with calcite. Weathered exterior surfaces of the pumice and large pieces of adhering detrital material were removed with a rocksaw. The calcite was dissolved in dilute (~5%) HNO3 and the remaining material was separated into heavy, intermediate, and light fractions using heavy liquids (s.g. = 2.85 g/cm3 and 2.5 g/cm3 ). The intermediate fraction was sieved (+35 mesh) and the best sanidine phenocrysts were picked by hand under a binocular microscope from the +35 mesh fraction. These crystals were crushed by hand in an agate mortar to -45 +100 mesh fraction to ensure homogeneity between splits taken for argon and potassium analysis. Amphiboles (kaersutite) were separated by similar methods. For small crystals from primary airfall tephra beds, dilute (-5%) HNO3 was used to disaggregate samples. The disaggregated samples were washed to remove fine material and dried. A heavy fraction was obtained using heavy liquids (s.g. = 2.85), and further concentrated using a Frantz isodynamic separator. The best phenocrysts were picked from this concentrate by 11 1 2 Table 1. Analytical data for potassium-argon ages. K Sample number Wt % 40 Ar* 1 0 - 11 mol/g %4uAr* Calculated Age Ma ± 1 s.d. Sample Location Kalodirr Tuff Member Amphibole Los 6 -8 C 1.345,1.401 4.153 56.7 17.4 + 0.5 Section 6 .1 Los 6 -8C 1.385,1.380 4.242 45.3 17.6 + 0.2 Section 6 .1 Los 6 -8A 1.446,1.486 4.513 70.9 17.7 + 0.4 Section 6 .1 K87-3108 1.449, 1.451 4.365 76.1 17.3 + 0.2 Section 6 .1 Los 3A 1.433,1.436 4.398 53.5 17.6 ± 0.2 Section 1 Biotite Los 6 -8 B 7.125,7.168 21.64 60.0 17.4 + 0.2 Section 6.1 21.58 61.9 17.3 + 0.2 Section 6.1 K87-3107 6.72, 6.81 20.42 49.0 17.3 ± 0 .3 Section 6.1 20.84 30.4 17.7 + 0.3 Section 6.1 K86-2742 7.067, 6.976 2 1 . 2 1 42.8 17.4 + 0.2 Section 6.1 Naserte Tuff Member Sanidine separate from pumice K87-3465 7.76, 7.83 22.78 76.1 16.8 + 0 . 2 Section 31 K87-3468A 7.85, 7.86 23.03 78.3 16.8 + 0 . 2 Section 31 Los 4-23B 7.639, 7.639 22.41 8 6 . 1 16.8 ± 0 . 2 Section 4 Unnamed tuff, lower Lothidok Fm Sanidine separate from pumice Los 8-2B 7.907, 7.893 22.80 87.8 16.6 ± 0 . 2 Section 8 Kamurunyang lahar Sanidine separate from pumice K87-3411 5.89, 5.85 13.37 81.2 13.1 + 0.2 Section 25 K87-3437B 5.78, 5.72 13.45 62.3 13.4 + 0.2 Section 29 K87-3437C+D 5.59, 5.53 13.52 70.4 14.0 ± 0.2 Section 29 Kalatum basalt Plagioclase separate from crushed sample K86-2743 0.431,0.437 1.013 76.2 13.6 ± 0 .2 Section 4 Kalakol basalts Whole rock Los 6.2-0 0.923, 0.917 2.839 55.5 17.7 ± 0.2 Section 6.2 Loperi basalts Whole rock B-20 1.522, 1.520 3.185 75.5 1 2 . 0 + 0 . 1 Section 40 B-31 1.411,1.409 2 . 6 6 8 80.9 10.9 ±0.1 Section 29 hand under a binocular microscope. A single large crystal was also prepared by crushing, sieving and washing. Biotites were separated from crushed material using an elutriation tube. Biotite was further concentrated by scattering grains on paper and vibrating the paper while holding it at an angle at which the biotite flakes would remain and the heavier grains would roll off. The final material used for dating was picked by hand under a binocular microscope from the concentrate. Thin sections of each basalt sample collected for isotopic dating were examined for the presence of alteration products such as zeolites and clay minerals. Samples that contained abundant alteration minerals or secondary minerals were excluded, and only one sample (Los 6.2-0) in which such materials were sparse was dated as a whole rock specimen. Even this sample contained minor amounts of alteration products, and was therefore treated with weak acetic acid (5%) to remove carbonate minerals before dating. Plagioclase phenocrysts were separated from another basalt (K86-2743) by magnetic methods following crushing, sieving and washing. All samples were dated in the potassium-argon laboratory in the Research School of Earth Sciences at the Australian National University , Canberra, under the direction of Dr. Ian McDougall. The procedure for determining the potassium-argon age is outlined in McDougall et al. (1980), and McDougall and Watkins (1986). Briefly, potassium was measured by flame photometry and argon by isotope dilution. Analytical precision is about 1 % at the level of one standard deviation, confirmed by satisfactory replication of potassium analyses and the agreement between duplicate argon analyses. Measurements were made on many separates from the same samples or bed with excellent agreement. To insure complete extraction of radiogenic argon from the phenocrysts, temperatures of 1600° C were maintained for a minimum of 40 minutes (F. H. Brown, pers. comm.). Decay constants used are those recommended by Steiger and Jaeger (1977). REGIONAL GEOLOGY OF THE TURKANA DEPRESSION Regional Structure The Turkana Depression (Figure 4), a triangular lowland in northwest Kenya, is bounded on the west by the Turkwel and Ugandan Escarpments (Baker et al., 1972). The Turkwel Escarpment is the oldest and most dissected rift fault escarpment in Kenya (Baker et al., 1972) and the Ugandan Escarpment is interpreted as an eroded monoclinal flexure (Walsh and Dodson, 1969). To the south, the Turkana Depression merges with the Baringo-Suguta graben, a well-defined graben at the northern end of the Gregory Rift. The Kinu-Sogo Fault Zone, essentially the northern continuation of the Baringo-Suguta Graben, forms the eastern boundary of the Turkana Depression (Baker et al., 1972). To the northeast, the depression merges with the well-defined graben presently occupied by Chew Bahir (Lake Stephanie) (Williamson and Savage, 1986). Chew Bahir lies roughly 40 km west of the southern end of the Ethiopian Rift, which is presently occupied by Lake Chamo. Regional Stratigraphy Four major lithostratigraphic units are defined in the Turkana Depression (Williamson and Savage, 1986). These are (1) the Precambrian and ?lowermost Paleozoic gneisses of the Mozambique fold belt; (2) a thick sequence of coarse, immature clastic sedimentary rocks, named the Turkana Grits or Laburr Series; (3) a sequence of Oligocene through Miocene volcanic and interbedded sedimentary rocks; and (4) a thick, heterogeneous assemblage of Plio-Pleistocene volcanic and sedimentary rocks. 1 5 Figure 4. Relation of the Lothidok Range to regional geographic and structural features of the Turkana Depression. 1 6 / 38‘ E L. Chamo ETHIOPIA L. Chew Bahir ,L Turkana, LOTHIDOK RANGE Toror Moroto K a d am Sudan Cv V Uganda p * - Indian Ocean .L./Baringo Kenya Tanzania R h ofrsiO s L. Victoria An early geologic study in northwest Kenya defined a sequence of sedimentary strata as the Turkana Grits (Murray-Hughes, 1933). Subsequently, nearly all sedimentary sequences in the Turkana depression have been collectively referred to as the Turkana Grits' (Dixey, 1934; Fuchs, 1939; Walsh and Dodson, 1969; Madden, 1972). The presence of angiospermous fossil wood (Dryoxylon) prompted numerous workers to consider the 'Turkana Grits' as Oligocene to Miocene in age. Arambourg (1943) alone doubted that the presence of Dryoxylon proved a mid-Tertiary age, and suggested that the wood-bearing sediments might be of Eocene or Cretaceous age. Arambourg (1943) was thus the first to suggest that two distinct episodes of sedimentation occurred in the Turkana depression, the first in the Mesozoic, and the second in the Miocene. Discovery of a sauropod humerus (Arambourg and Wolff, 1969) in the Laburr Series of the Lapurr Range proved Arambourg correct. The regional survey of sedimentary strata in the Turkana Depression by Williamson and Savage (1986) is in strong agreement with Arambourg's original interpretation and division of the sedimentary sequences into two distinct groups. The lower group includes the clastic sequences of Lodwar/Muruanachok, Lapurr (Cretaceous) and lower strata at Kajong (Figure 4). Strata at Lodwar, Muruanachok, and Kajong lie nonconformably on basement, and are unconformably overlain by Miocene deposits. Oligocene basalts (Walsh and Dodson, 1969) overlie the Cretaceous sedimentary rocks in the Lapurr Range. The coarse fraction of these older sediments lacks volcanic minerals and volcanic rock fragments. The upper group includes the sedimentary rocks of Lothidok, Loperot, and upper strata at Kajong (Figure 4). This sequence includes a significant volcaniclastic component and has been shown paleontologically and isotopically to be Miocene in age (Williamson and Savage, 1986). The division of the Turkana Grits into two distinct sequences, however, falls short of solving the problems that currently plague the stratigraphic nomenclature for the sedimentary deposits in the Turkana Depression. Because sedimentary strata contain no substantial volcaniclastic component, he nonconformably on basement, or are nonconformably overlain by Miocene basalts does not prove they belong to either group of sediments. The lack of a volcaniclastic component may simply indicate there was no volcanic source. This is clear at Loperot, roughly 100 km south of the Lothidok Range (Figure 4). Work by Joubert (1966) and by this writer in 1986 and 1987 show that the Miocene sedimentary sequence consists of arkosic sandstones, lying directly on basement, which are overlain by volcaniclastic strata. The sequence is nonconformably overlain by Miocene basalts. This is also the case for the Pliocene and Pleistocene sediments exposed southwest of Lodwar and east of the Lothidok Range. Those near Lodwar were termed Turkana Grits' (Walsh and Dodson, 1969), and those near Lothidok were termed the 'Laburr Series' (Arambourg, 1943) solely because they lack volcaniclastic material. The present strati graphic morass resulted from the assignment of local stratigraphic sequences to one regional unit (Turkana Grits). In a sense Williamson and Savage (1986) continue this style of stratigraphy by correlating the lower sediments at Kajong (Sera Iltomia Formation) with the Labur series even though these sediments contain no direct evidence of deposition during the Cretaceous Period. However, by defining formations on the basis of local sections they helped resolve some of the problems. Should their correlations be disproved in future, the formations will remain valid. The most complete study of Tertiary strata in the region is that of Watkins (1982), who described a 1900 m thick sequence of volcanic rocks and intercalated sedimentary deposits in the Suregei-Asille region northeast of Lake Turkana (Figure 4). Watkins (1982) proposed a formal lithostratigraphic nomenclature for these units consisting of nine formations that span the interval from the early Miocene to the middle Pliocene. Potassium-argon dating of lavas and high-temperature alkali feldspars separated from rhyolitic units in the sequence provide temporal control for these formations (McDougall and Watkins, 1988). The volcanic and sedimentary deposits of the Surgei-Asille region were deposited during the same interval of time as those at Lothidok, and have yielded an important fauna from the site of Buluk (Leakey and Walker, 1985; Harris and Watkins, 1974). The lack of clearly correlative strata between the two regions is of substantial paleogeographic importance, and is treated below. STRATIGRAPHY OF THE LOTHIDOK RANGE Stratigraphy Following the lead set by Watkins (1982) and by Williamson and Savage (1986), a new formation name for the sedimentary strata of the Lothidok Range is defined here on the basis of local sections. The strata are divided into the Kalakol basalts (new informal name), the Lothidok Formation (new stratotype), the Loperi basalts (new informal name), and undifferentiated Tertiary deposits (Figure 5). No single section exposes the entire sequence of strata in the Lothidok Range, and the stratigraphy is therefore constructed from 10 partial sections (Figure 6 ). The Kalakol basalts are described from Sections, 17 and 7. Section 17 includes strata from the base of the section up to a break that results from faulting. Section 7 includes strata above the section break to the upper contact of the Kalakol basalts. The Lothidok Formation is exposed in a series of correlated sections (6.2, 19, 21, 24, 25, 29, 30 and 31; Figure 6 ). The Loperi basalts that overlie the Lothidok Formation are described from section 29, which continues upward through the undifferentiated Tertiary rocks. The composite stratigraphic sequence of the exposed strata is approximately 1500 m thick. Detailed descriptions of these partial sections are given in Appendix B. All correlations between sections are tightly constrained except for one within the Kalakol basalts. A break in the section occurs approximately 200 m above the Eragaleit beds (Figure 5) where section 7 is correlated with section 17. The basalts between the Eragaleit beds and section break were neither measured nor examined in detail, hence there may be missing or duplicated section. If section is missing, the thickness is a minimum; if Figure 5. Sequence and terminology used for the composite stratigraphy of the Lothidok Range. METERS 22 1400 1200 - Tertiary Undifferentiated (Tu) Loperi Basalts (Tib) Lothidok Formation Kalakol basalts (Tkb) 300 200 100 \ Missing Section Explanation Eragaleit beds Lahars *AAAA/ AAAA k ^* Vr 4 V‘ * A ?* 4- Unconformity Tephra Tuffaceous and clastic sediments Basalts Fault Figure 6 . Component sections for the stratigraphy of the Lothidok Range. 24 Measured Sections Unit 6.2 Tertiary Undifferentiated (Tu) Loperi Basalts (Tib) -Lothidok Formation Kalakol basalts (Tkb) Missing Section «> 1 0 0 ^ Explanation w9) Q> s Eragaleit beds Labors Unconformity Tephra Tuffaceous and clastic sediments Basalts Fault duplicated, the thickness is overestimated by 200 m or less. The generalized correlations and lateral variation are illustrated on Figures 7 and 8 . Lithology The lithologies encountered in the Lothidok Range consist primarily of terrestrial sedimentary and volcanic deposits common to many regions. In addition to these, however, a considerable proportion of the strata comprising the Lothidok Range consists of pyroclastic deposits. Because two types of these deposits, airfall tephra and lahars, may not be familiar to the reader, they are briefly described below. For the purposes of this report, pyroclastic rocks are considered to be those composed of volcanic ejecta directly originating from a volcanic eruption, and consist of stratified and massive airfall tephra. These are distinguished by the presence of pumice (recognizable at present alteration state), euhedral volcanogenic minerals (amphibole, pyroxene, biotite, and sanidine) and angular volcanic rock fragments; and by the lack of fluvial sedimentary features. Fluvially reworked tephra are discussed as clastic deposits. The granulometric classification of pyroclasts and pyroclastic deposits is from Schmid (1981). Lahars are mudflows that consist chiefly of volcanic materials (Bates and Jackson, 1987). The lahars form distinctive deposits of matrix dominated/supported, massive to inversely graded conglomerates. They are distinguished from matrix supported conglomerates of fluvial origin by the lack of sedimentary structures and erosional basal contacts, and by the abundance of fine-grained (clay to sand) matrix and volcanogenic material such as pumice. 25 Figure 7. Generalized north-south cross-section and lateral distribution of the strata exposed in the Lothidok Range. Kl Kamurunyang lahar Ab Akwang'a basalt Kab Kalatum basalt Lc Lokipenata conglomerate NEb Nakwel Esha beds NT Naserte Tuf fs KT Kalodirr Tuf fs Be Basa l conglomerate Eb E ra g a le i t beds ? Section (s) Fault •3ol0‘ N & Area boundary 0 ? *<m • Section location Ma r k e r Hor i zons E x p la n a t io n Tertiary undif ferent iated ----------------Un c o n f o rmi t y ? Loperi B a sa l ts (Tib) - Un c o n formi t y Kl (T u ) upper Lothidok Fm. (T lu ) Lc r = - Di s conform i t y NEb lower Lothidok Fm. (T i l ) Con tact -interbedded sediments Kalakol basal ts (Tkb) 7 ' J A A < L.1 A V > i___A7 r < --- 1- 36°E -3 °3 0 N 35o5 0 lE L o t h i d o k Range N) ~4 28 Figure 8 . Generalized east-west cross-section and lateral distribution of the strata exposed in the Lothidok Range. 29 Marker Horizons Kl K am u ru n y a n g la h a r Kab Ka la tum b a s a l t L c L o k ip en a ta c o n g lome ra te NT Na s e r te T u f f s K T K a lo d i r r T u f f s B e B a s a l co n g lome ra te ? S e c t io n (s) Fau It E x p l a n a t i o n T e r t i a r y u n d i f fe r e n t ia te d (T u ) ----------------U n c o n f o rm i t y ? L o p e r i B a s a l t s ( T i b ) -----------------U n c o n f o rm i t y m I Kl upper Lothidok Fm. ( T l u ) Kob Di s c o n fo r m i t y NT lower L o th id o k Fm. ( T i l ) KT B c C o n t a c t --- interbedded sediments Ka la k o l ba s a l ts (Tkb) Kalakol Basalts - Tkb The lowest strata exposed in the Lothidok Range consist of basalt flows and intercalated sedimentary deposits, here informally named the Kalakol basalts (Tkb - Plate, Tvbl - Walsh and Dodson, 1969). On the basis of similar isotopic ages, the Kalakol basalts apparently correlate with the Turkana Basalts of the Lodwar sequence (Zanettin, et al., 1983) and with the Lodwar Formation (Bellieni, et al., 1987). However, the respective authors do not designate a stratotype or provide descriptions for these units. Therefore a new, informal name is used here to refer specifically to the older basalt flows of the Lothidok Range. Most of the basaltic hills in the Lothidok Range north of the Kalatum and Alomonet Rivers consists of Kalakol basalts. Only four minor exposures of the Kalakol basalts occur south of these rivers (Plate). The lowest flows of the Kalakol basalts crop out only between the Eragaleit and Nathuraa rivers below Lothidok Hill (Plate, Figure 2), but the base of the section is not exposed. The Kalakol basalts are a minimum of 785 m thick, and consist of at least 20 flows ranging from 4 to 60 m thick with interbedded sedimentary and pyroclastic rocks ranging from 2 to 50 m thick. The flows are occasionally vesicular near upper contacts. The highest basalt flow ranges from 7 to 17 m thick and thins to the north where it pinches out between sections 7 and 6 .1 (Figure 7). The Kalakol basalts consist predominantly of olivine-augite lava flows that are aphyric to coarsely phyric with an aphanitic groundmass. Phenocrysts of olivine, augite and plagioclase are normally between 1 and 3 mm in long dimension but in one flow, augite reaches 1 cm. Accessory minerals include an Fe-Ti oxide, probably magnetite, and biotite occurs occasionally as poikilitic plates in the groundmass. Fractures filled with drusy quartz and/or calcite are common. All of the basalts are weathered and altered to various degrees. Olivine phenocrysts have been altered to chlorite, or iddingsite with 30 limonite rims. Chlorite alteration of the groundmass is usually extensive, and calcite amygdules are common. Thin, lenticular sedimentary deposits commonly lie between basalt flows and consist of granule to cobble conglomerates, poorly sorted to conglomeratic sandstones, stratified siltstones (rare), and massive mudstones. These deposits are poorly exposed, and generally only a few meters thick. The thickest interval, designated the Eragaleit beds, contains vertebrate fossils and is consequently discussed in detail. Eragaleit Beds The most important sequence of sedimentary and pyroclastic rocks interbedded with the Kalakol basalts is here informally named the Eragaleit beds. These beds are only exposed in the Eragaleit and Nathuraa Rivers near the base of Lothidok Hill and he roughly 600 m below the top of the Kalakol basalts (Figures 5 and 7). Although incomplete, the thickest exposures are found along the Eragaleit river (section 16) where the beds reach a maximum of 50 m. The section appears thicker because it is faulted against lithologically similar strata of an overlying formation (Plate). The only complete sequence of the Eragaleit beds occurs in the small, north flowing tributary to the Nathuraa river (Figure 7). At this location (section 17) the sequence has thinned to 37 m and continues thinning to the north. The section is about 10 m thick where it is truncated by a northwest striking, east dipping normal fault, north of which no exposures of the Eragaleit beds were found. Approximately 500 m south of the Eragaleit River, the beds are truncated by the fault that divides the Eragaleit and Lataagur grabens (see structure discussion, p. 127). Only small exposures of the beds, the bases of which are also truncated by faulting, exist west of the Lataagur graben. The Eragaleit beds consist primarily of polymictic conglomerates and conglomeratic litharenites (McBride, 1963; Folk, 1968), minor siltstones and mudstones, and a stratified 3 1 tephra sequence. The base of the sedimentary sequence consists of a 4 to 6 m thick, massive to crudely bedded, normally graded, grey to reddish grey conglomerate. Clasts are stained with hematite, are poorly imbricated, and range in size from small boulders to large cobbles. The lower contact is an irregular erosional surface with scours up to 1 m deep cut into the underlying basalt. Broad, internal scour and fill structures are common, and large-scale trough crossbeds are present but rare. Overlying the basal conglomerates are pale red to grey granule to cobble conglomerates that generally grade into pale red, poorly sorted, coarse-grained to conglomeratic sandstones. These contain small- to medium-scale trough crossbeds with both upward-coarsening and upward-fining sequences. Basal, pebble to cobble conglomerate lenses and mudstone rip-up clasts are common in the sandstones. Clasts of the conglomerates and conglomeratic sandstones consist of phonolite and basalt. Sandstones of the Eragaleit beds are made up of very fine to coarse, rounded volcanic and lithic grains. The sandstones occasionally grade to ripple stratified siltstones or interfinger with massive, pale red to brown mudstones. The siltstones are moderately to very sandy and generally too poorly exposed to display bedding structures. These fine grained sediments constitute less than 10% of the total section. The highest bed in the sequence is a dark red, well-consolidated, massive siltstone. Columnar jointing in the upper 20 cm o f this siltstone indicates it was baked by the overlying basalt. A moderately resistant con glomerate-sandstone-silts tone sequence (13 m) that forms small ridges near the top of the section (bed 18, section 16; bed 12, section 17) contains abundant vertebrate fossils (see faunal discussion, p. 129). This sequence is dark red with purple streaks and yellow limonite blotches. The sandstones are very poorly sorted and contain abundant clay. The conglomerates contain small to medium pebbles of volcanic rocks and sedimentary rip-up clasts. Medium- to small-scale trough 32 crossbedding, internal scour and fill sequences, normally graded bedding, and basal erosional contacts occur in coarse grained layers. Ripple laminations occur in the fine­grained sandstones and siltstones. Upward-fining sequences average 1 m thick. A tuffaceous interval within the Eragaleit beds lies about 15 m above the basalt-conglomerate contact in section 17. These tuffs are informally named the Nathuraa tuffs, because they are best exposed near the Nathuraa River. The tuffaceous interval consists of interstratified 1 to 3 cm thick, fine tuffs to coarse lapillistones and 10 to 30 cm thick, reworked tephra beds. The lapillistones contain abundant, extremely altered and flattened pumice lapilli in which the glass has been altered to a zeolite or a clay. The tuffs and lapillistone matrix have also been altered to montmorillonite. Several small basaltic dikes intrude the Eragaleit beds in the exposures flanking the Eragaleit River. These extremely altered dikes are generally dark greenish grey to pale olive and average less than 1 m wide by 2 0 m long. Other Interbedded Sedimentary Rocks Sedimentary deposits interbedded within the Kalakol basalts at various levels, differ from the Eragaleit beds mainly in the amount of mud occurring as matrix of coarser rocks and as discrete bodies of mudstone. Although conglomerates and sandstones typically dominate, mudstones are locally the only rocks exposed between the basalt flows. The mudstones are moderately to very sandy, dark red to brown, and grade into muddy sandstones. The mudstones contain numerous white, chalky, calcareous root casts and 1 to 5 cm thick, tabular calcrete? interbeds. Mud-rich sandstones generally contain interbedded mudstone lenses, lithic grains, volcanic mineral grains, and mudstone rip-up clasts. The upper surfaces of some basalts grade to mudstones suggesting that these may represent ancient weathering horizons. 33 The uppermost sedimentary strata within the Kalakol basalts lie 10 to 17 m below the upper contact (Figure 7). These strata are exposed in sections 1, 1.2, 3, 4, 7, 10.1, 10.3, 30, and 31, and are 10 m thick in all exposures along the west side of the Lothidok Range. The only exposure to the east lies on the north side of Moruorot Hill (section 10.3) where this sedimentaiy interval is 80 m thick (Figure 8 ). The deposits consist of pale purple to purplish grey, polymictic conglomerates, conglomeratic litharenites and dark reddish brown mudstones. At most outcrops pebbles or cobbles dominate, but at Moruorot, large cobbles and boulders up to 2.5 m diameter are present. Potassium-Argon Age Determinations The single whole-rock dated sample (Los 6.2-0), collected from the uppermost Kalakol basalt below the Lothidok Formation at section 6.2, yielded an age of 17.7 + 0.2 Ma (Table 1). Although the determinations are replicable, the age should be regarded as a minimum until confirmed by additional work. Zanettin et al. (1983) give an isotopic age of 24.4 +1.0 Ma for a phonolite in this area, but the location from which the dated material was collected is unclear. Coordinates provided for the sample (K 60; Zanettin, et al.,1983) place it near Kakurtua Hi l l , where the Loperi basalts, now known to be much younger (this report), are exposed. Zanettin et al. (1983) map the sample as a Turkana basalt' and indicates that it was taken near Moruorot Hill, approximately 10 km north of the coordinates given. No phonolites were found in this area during this study. Volcanic rocks from this location could belong to either the Loperi or Kalakol basalts although the measured age indicates the latter. Assuming that the sample was collected from the Kalakol basalts exposed at Moruorot Hill and that the age is correct, deposition of the Kalokol basalts in the Lothidok Range occurred from prior ~25 to -18 Ma ago. 34 Lothidok Formation - T1 The Lothidok Formation is defined here as all strata lying between the upper contact of the highest flow of the Kalakol basalts and the lower contact of the Loperi basalts (Figure 9). The formation is a heterogenous assemblage of early and middle Miocene sedimentary and volcanic rocks. The term 'Lothidok' has previously been used informally to designate a formation on the basis of a single dated basalt (Bellieni, et al., 1987). However, Bellieni, et al. (1987) provide no type section, formation description or outcrop location. The International Guide to Stratigraphic Nomenclature (Hedberg, 1976) strictly prohibits the definition of formations based of time of deposition, and requires that some description of the named units be given. It is my intent to formally define the Lothidok Formation and to designate a stratotype for this unit in a subsequent publication; therefore, the terminology is used here in a formal sense to prevent future confusion. Descriptions of the constituent strata are given both in the text below and on columnar sections (Appendix B; Boschetto, 1988). The term 'Lothidok' is retained as it is the most logical name for the formation because the stratotype is a composite with component sections (Figure 6 ) widely scattered throughout the Lothidok Range (Figure 3). The names of all formal and informal subsidiary units within the formation are new, and are defined with reference to particular type sections. The Lothidok Formation is informally divided into lower (Til) and upper (Tlu) units. The general sequence, names, and isotopic ages of the strata are shown in Figure 9. Three members within the lower Lothidok Formation are formally defined: the Basal Conglomerate Member, the Kalodirr Tuffs, and the Naserte Tuffs; a fourth unit is informally named the Nakwel Esha beds. No units are formally defined within the upper Lothidok Formation, but four units are informally named: the Lokipenata conglomerates, the Kalatum basalt, the Akwang'a basalt, and the Kamurunyang lahar. Units are 35 36 Figure 9. The sequence and terminology used for the composite-stratotype of the Lothidok Formation. 500 Loperi Basalts (10.9-12.0 Ma) angular unconformity Kamurunyang lahar (13.1 - 13.4 Ma) Akwang'a basalt 400 300 - 200 - 100 unassigned strata Kalatum basalt (13.6 Ma) Lokipenata Conglomerate disconformity Nakwel Esha beds Naserte Tuff (l6.8Ma) AAA A A AAA A A") ^ A A A A A A A A A J f cAAAAAAAAAA t lAAAAAAAAA A I a a a a a a a a a a a I s unassigned strata Alomonet tuffs Kanukurinya tuffs . Kalodirr Tuffs (17.3­17.7 Ma) Basal Conglomerate {7 7 7 } Kalakol basalts (?-l7.7Ma) upper (TIu)-------------------- *+•--- lower (Ti l ) informally defined when they form useful local markers but are not widespread or well enough exposed for detailed description. The boundary between the lower and upper Lothidok Formation is placed at the base of the Lokipenata conglomerate. Exposures of the Lothidok Formation vary greatly. The most complete exposures occur along the southwestern edge of the Lothidok Range (sections 4 and 30). Section 30 provides a general thicknesses of 250 and 290 m for the lower and upper Lothidok Formation respectively. All other exposures are truncated by faulting, unconformities, or erosion. The predominant trend of the outcrop is north-south (Figure 7), and east-west lateral variation can only be observed in the vicinity of the Kalatum and Alomonet Rivers (Figure 8 , Plate). The poor exposure of sections 4 and 30 necessitates the use of eight component sections (Figure 6 ) to construct a composite-stratotype for the Lothidok Formation (Figure 9). Strata in the lower Lothidok Formation contain enough distinctive marker horizons that correlations are relatively straightforward. By contrast the heterogeneous, areally limited, and laterally variable nature of the upper Lothidok Formation, coupled with the lack of marker horizons and the use of an angular unconformity as the upper boundary, conspire to make correlation very difficult. Marker horizons comprise very little of the thickness of the Lothidok Formation. Most of the formation is composed of laterally variable sedimentary and pyroclastic deposits. Because these can only be broadly correlated on the basis of their stratigraphic position with respect to the marker horizons the Lothidok Formation cannot be exhaustively divided into subordinate stratigraphic units. The following discussions therefore emphasize marker horizons to set the stratigraphic framework and address the strata not assigned to a specific unit in a general way. Interpretations of depositional environments for the sedimentary and pyroclastic rocks are presented in the following discussion. 38 Lower Lothidok Formation - Til Strata of the lower Lothidok Formation are best exposed in sections along or north of the Naserte and Alomonet Rivers with limited exposures farther south. The lower Lothidok Formation consists of volcaniclastic sedimentary rocks, and altered mafic alkaline and trachytic tuffs. Four marker horizons proved useful for correlating outcrops and subdividing the lower Lothidok Formation. These are the Basal Conglomerate Member, the Kalodirr Tuffs, the Naserte Tuffs, and the Nakwel Esha beds (Figure 9). Basal Conglomerate Member The Basal Conglomerate Member is the lowest widespread unit of the Lothidok Formation. Although the member crops out throughout the Lothidok Range, outcrops are generally discontinuous and incomplete, and exposures with more than 40% of the interval represented by this unit are rare. The best exposures are in sections 1 (type section), 1.2, 10.3, and 20; less complete exposures occur in sections 3, 4, 6.1, 6.2, 10.1, 22, 30 and 31 (Figure 3). The Basal Conglomerate Member consists of 10 to 75 m of volcaniclastic conglomerates and sandstones that conformably overlie the Kalakol basalt and interfinger with the conglomerates below the highest Kalakol basalt flow. The unit is thickest in the central and southern exposures and thins to the north. The member consists primarily of polymictic, dark greyish red to light reddish grey and purplish grey, pebble to cobble conglomerates. The largest clasts are large cobbles to medium boulders. Dark red hematite partially or wholly coats phonolite and basalt clasts that are set in a sandstone matrix with medium to very coarse, subangular to subrounded grains. The beds are poorly to moderately cemented with calcite, which may form 2 mm thick rinds on clasts where matrix is absent. 39 Clast-supported conglomerates are massive, or crudely bedded with imbricated clasts. Matrix-supported conglomerates generally exhibit medium- to large-scale trough crossbedding, internal scour and fill sequences and low-angle crossbedding. Strata sets are predominantly normally graded although inverse and symmetric (normal-reverse-normal) grading occurs. Conglomeratic strata cosets generally fine vertically and laterally. Channel fill sequences reach 1 m thick and 3 m wide. Irregular, subhorizontal erosion surfaces are common. The conglomerates grade into or interfmger with poorly sorted to conglomeratic sandstones that occasionally dominate the sequence. The sandstones consist of medium to very coarse, subrounded to subangular lithic grains and are classified as litharenites (McBride, 1963, Folk, 1968) because they contain less than 2% (modal) quartz/feldspar. Most of the quartz grains are euhedral crystals that probably originated from cavity fillings or geodes common in the Kalakol basalt. Small- to large-scale trough crossbeds, low-angle crossbeds and internal scours dominate the sedimentary structures in the sandstones. Thin, massive, conglomeratic lenses commonly overlie basal and internal scours and cosets of cross strata predominantly fine upwards but occasionally coarsen upward. The sandstones occasionally grade into pale red to red siltstones, but these represent less than 1 0 % of the exposed section. Dark red, and reddish brown to brown massive mudstones occur commonly as interbeds. Thin, lenticular mudstones primarily occur within conglomeratic cosets and tabular mudstones lie between cosets. The poorly exposed mudstones are moderately to very sandy and moderately to poorly consolidated. In a few outcrops these contain widely scattered, extremely altered, fine granule to medium pebbles. White, chalky, calcareous rootcasts and thin, tabular, calcrete horizons are common, and are interpreted as pedogenic features. The actual amount of mudstone occurring in this sequence is not clear, although the variation appears great. In all sections except 6 .1 and 6.2, mudstones represent less than 40 20% of the actual outcrop. At sections 6 .1 and 6.2, the massive, dark reddish brown mudstones account for ~15% of the exposed strata, the remainder of the section being conglomerate. K al odirr Tuff Member The Kalodirr Tuffs (new member) lie immediately above the Basal Conglomerate Member, and are internally divided into the Kanukurinya (lower) and Alomonet (upper) tuff beds. The type locality for the Kanukurinya tuff beds is section 6.1, located 12.75 km north of the Kalakol-Lodwar road (Figure 3). The type locality for the Alomonet tuff beds is section 23, located along the Alomonet river (Figure 3). The thickness of the Kalodirr Tuffs ranges from 15 m in section 10.1 to 25 m at section 6.1 (Figure 10), depending on the number and thickness of beds comprising this member. The lower boundary bed is a very fine, massive, pale red tuff at the base of the Kanukurinya tuff beds. The upper boundary is the top of the Alomonet tuff beds. Kanukurinya Tuff Beds The Kanukurinya tuff beds consist of stratified and reworked tephra, lahar deposits and interbedded clastic sediments (Figure 10). The Kanukurinya tuffs occur at every outcrop of this stratigraphic level and can be traced laterally for the length of every continuous exposure. The stratified tephra are olive to pale green, pale red and reddish tan and range from 5 to 50 cm thick. The scale of stratification decreases upwards from thick to thin planar laminations. The laminae commonly drape small scale, topographic irregularities such as mounds or fossil wood, with no evidence for erosion on the basal contact. Planar laminated strata occasionally grade into ripple marked and cross-stratified beds laterally. Platy and elongate grains are oriented parallel or subparallel to bedding. Accretionary 41 Figure 10. Correlation and lateral variation in the Kalodirr Tuffs. s Sect ion 10.1 S e c t ion 31 B; °'& ° 0 6 ^ ■o'-' J ■ • o o o N -25 Explanation - 2 0 Fluvial clastic deposits Scour surface Primary airfall and reworked tephra deposits Accretionary lapilli Non-erosional surface Syenite clasts Miscellaneous clasts Lahor deposits Plant debris ■15 £ <L> E £ Index Map -5 Section 6.2 35°50'E 3°30'N 36° E 3-IO'N 0_______ 5km ^ Area boundary > Section -&• OJ lapilli, ranging from 1 to 4 mm in diameter, occur in one 2 cm thick massive tuff. The stratified and massive tuffs consist of crystals (65%) and lithic lapilli (35%) in an altered matrix, of fine ash. The most abundant crystals are euhedral pyroxene (diopsidic augite), kaersutite, and biotite. The pyriboles (pyroxenes and amphiboles) are up to 2 mm diameter (1 0 ) where as biotite phenocrysts reach 4 cm diameter. Nepheline in various stages of alteration is present but not common. Accessory minerals include apatite, perovskite, mafic clots, extensively altered plagioclase, and an unidentified opaque mineral. Basaltic or basanitic lithic fragments up to 3 mm diameter (~-2 0 ) contain both plagioclase and olivine. The stratified tuffs are predominantly normally graded, fining upwards through the cosets. Symmetric grading (normal-inverse-normal) occurs in the coarsest beds. A very thin (< 1 cm), very fine grained, pale red, thinly laminated to massive tuff commonly forms the highest layer of the beds, and locally has desiccation cracks and raindrop impressions. Gastropods, preserved as calcite casts, are the most common fossils found in the stratified tuff beds. These include Lanistes carinatus, Pila ovata, Cerastua miocenica, and Burtoa nilotica verdcourti (Van Damme and Gautier, 1972; see fauna discussion, p. 123). A turtle carapace and plastron found in a laminated bed (this study) is completely filled with the tuffaceous material of the surrounding bed. Lahar deposits overlie many of the stratified tuff sequences (Figure 11). These range from 15 cm to 4.4 m thick, with the thickest bed being the highest at sections 1 and 33. These beds generally thin to the south. The lahar deposits are pale green to olive, pale yellowish brown to pale reddish brown, and consist of 45 to 65% clay to coarse grained sand matrix, and 35 to 55% granules to boulders. Thin section analysis shows that 35% of the sand grade matrix is made up of volcanic minerals such as euhedral amphibole, pyroxene and biotite. 44 4 5 Figure 11. Typical sequence of the Kanukurinya tuff beds from section 33. meters 46 Bed 15 Bed 14 Bed 13 Bed II A A A A A" A A A A A' f lA' l AA A A A A A A A A A A A A A A A A A A A A A A A A A A A A A-A7V A A a J Bed 10 / G \ A A A A M ® A ^ > A A A A A A A A A A A / A A A A A A A A A A A / 0 / © \ A A A A A A A A k A A Al A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A ]Bed 6 Bed 4 Bed 3 Bed 2 Bed I ^- Rework ed tuf fs Igneous boulders l /Sof t sediment inject ion dikes - L a h a r deposi ts hM u d s t o n e s P l a n t d eb r is A i r f a l l t u f f s G a s t r o p o d s Ac c r e t io n a r y la p i l l i Co n to r te d bedding The lahar deposits contain crystals up to 3 cm diameter (-50) of the same mineralogy as the laminated beds, but platy and elongate particles are randomly oriented in most of these beds. Reverse grading occurs when the beds contain a wide range of clast grades. Basal contacts of these beds are sharp but nonerosional, which is apparent from intact mud flakes of very fine bedded tuffs lying along the basal contact. The curled mud flakes are occasionally overturned and imbricated but rarely broken. Lithic clasts of basalt, phonolite, and nephelinite reach 20 cm in diameter in the lower beds and up to 2.2 m in diameter in the highest bed. The basalt and phonolite clasts were extensively altered and were not collected. Nephelinites are rare and have not been examined in detail. Near sections 1, 1.2 and 33 the highest bed contains numerous well-rounded nepheline-bearing syenite boulders up to 1.2 m in diameter. At sections 4 and 10.1, the boulders are up to 50 cm in diameter. The roundness has been attributed to prolonged periods of water abrasion (Walsh and Dodson, 1969), but may have resulted from magmatic processes (Holmes, 1965). These syenites consist of 75% potassium feldspar, 5-15% aegirine augite, 5-15% alkali amphibole, < 5% nepheline, an Fe-rich biotite and minor apatite. Petrologically similar boulders are discussed by Jeremine (1935) and Smith (1939), but differ from those of the Kalodirr Tuffs by their higher percentage of nepheline and presence of sphene. The lahars also contain abundant plant debris consisting of wood and leaves generally present in the upper parts of individual beds. No preferred orientation of this material was observed. The wood is preserved as calcite filled casts of logs, branches and stumps with very well-preserved exterior surfaces. The largest examples reach 30 cm in diameter and 2 m in length. Leaf and grass imprints are commonly rolled around matrix material (Figure 12), and are most abundant in sections 6.1 and 31. The majority of the leaf imprints have entire margins and are very well-preserved. Some leaf imprints have been tentatively identified as fossils of a broad leafed bamboo (B. Jacobs, pers. comm.). 47 48 Figure 12. Photograph of a leaf imprint in the Kalodirr Tuffs at section 6.1. The knife above and to the right of the imprint is 9 cm in length. A calcite cast of a fossil fruit collected from these beds is believed to belong to the Dicotlyedonae, possibly Burseraceae, Canarum sp. nov. (C. Kabuye, pers. comm.). Thin, fluvially reworked, tephra beds typically overlie the stratified tephra and lahar deposits (Figure 11). These 20 to 50 cm thick beds contain subrounded to rounded lithic and volcanic mineral grains, volcanic rock fragments and minor bedded-tephra intraclasts. Sedimentary structures include basal erosion surfaces, small-scale trough crossbeds and planar to asymmetric ripple stratification. Soft sediment deformation structures are very common in these beds and include contorted and convolute stratified tephra beds. Other features include soft sediment dikes of stratified tephra material injected into overlying lahar deposits, and deformed rip-up clasts (up to 1 m long) of stratified tephra beds within the overlying lahars. Alomonet Tuff Beds The Alomonet tuff beds range from 1.5 to 2.8 m thick, thin to the north and lie less than 4 m above the Kanukurinya tuffs where both occur at the same outcrop. The reddish to yellowish brown color of the Alomonet tuffs is very distinct from the pale to dark olive green to rare pale red colors of the Kanukurinya tuffs. A second difference between the Kanukurinya and Alomonet tuffs is the significantly smaller size and lower abundance of phenocrysts in the Alomonet tuff. Near sections 1, 1.2, 30 and 33 the Alomonet tuffs contain numerous well-preserved footprints of unidentified birds and artiodactyls, and vertebrate fossils are also associated with these beds (see faunal discussion, p. 129). Stratified tuffs comprise roughly 15-25% of the total thickness. The tuffs are primarily very fine to coarse, very thinly laminated and planar to ripple stratified. The beds consist of less than 2 0 % pyroxene, amphibole and biotite phenocrysts,and less than 1 0 % lithic fragments. Pyriboles are < 1 mm and biotites are ~ 2 mm in diameter. 49 The massive tuffs range from 2 to 15 cm thick and overlie laminated beds with sharp, nonerosional contacts. These are very fine to coarse tuffs with biotite flakes from 1 fragments. The wood, preserved as calcite casts, ranges from 3-5 mm diameter and 2 to 5 cm long. Some fine grained massive beds are extensively bioturbated, and the thinly bedded tuffs exhibit various degrees of bioturbation. Several lenticular, 10 to 15 cm thick, fluvially reworked tuffs are interbedded with the tephra beds discussed above. These tuffs are generally very poorly sorted with grains ranging from very coarse sand to small rounded lithic pebbles. They contain abundant 'pyriboles' and very little or no biotite. The 'pyriboles' are commonly abraded and account for ~25-35% of the grains. Sedimentary structures include sharp basal erosion surfaces, internal scours, small-scale trough crossbeds and normal grading. Interbedded Sedimentary Rocks The Kalodirr Tuffs contain minor interbedded conglomerates, sandstones and siltstones (Figure 10). Maximum clasts range from medium pebbles in the conglomeratic sandstones to small cobbles in the conglomerates. At section 31, the sandstones between the Kanukurinya and Alomonet tuffs contain rip up clasts of the underlying tuffs in addition to lithic grains and minor pyroxene and amphibole. The depositional structures are very similar to those in the Basal Conglomerate Member. Small- to medium-scale trough crossbed sets in sandstones fine upward, whereas overall the cosets tend to coarsen upward. In a few instances, the sandstones grade to very sandy, dark reddish brown, lenticular mudstones that reach 30 cm in thickness and contain abundant calcareous root casts. In section 6 .1 the interbedded clastic layers coarsen upward from sandstones in the lower part to conglomerates in the upper part. Potassium-Argon Age Determinations Potassium-argon age determinations on biotite and ampohibole phenocrysts separated from the Kanukurinya tuff beds yielded ages ranging from 17.3 ± 0.3 to 17.7 ± 0.3 Ma (Table 1). The majority of the dates are from samples collected at section 6.1 including Los 6 - 8 A, B and C, K86-2472 , K87-3107, and K87-3108. An additional date is from sample Los 3A from section 1. Naserte Tuffs The Naserte Tuffs are named for the Naserte River, a small southern tributary of the Kalatum, which drains the area of the type locality. The type sequence consists of 40 cm of stratified tephra overlain by a 12 m thick lahar deposit (Figure 13). The best exposures of a complete sequence of the Naserte Tuffs is section 21 (Figure 3), which serves as the type locality but section 39 offers the best exposures of the tephra beds. The interstratified, thinly to medium bedded, stratified and massive tuffs are pale yellow to pale orange. Red to dark red, orange and pale purple liesegang banding is common. The interstratified massive and laminated tuffs range from 2 to 40 cm thick. Accretionary lapilli ranging from 2 mm to 2 cm in diameter account for 5 to 45% of individual massive beds (Figure 14). The accretionary lapilli are cored by fine ash or fragmented accretionary lapilli debris and rimmed with concentric bands of fine ash. Sand sized debris consists predominantly of fine to coarse, angular lithic grains and minor fine to coarse, euhedral to subhedral crystals of sanidine, biotite and amphibole. Altered, fine pumice lapilli occur rarely in a matrix of fine altered ash. Bed geometry is usually tabular, but changes laterally to irregular pinch and swell. The tuffs are bounded by sharp, nonerosional contacts. Massive tuffs dominate the sequence at section 39 where laminated beds occur only once. Sedimentary structures in the stratified beds are dominated by planar laminae with minor ripple stratification. 5 1 Figure 13. Typical sequence of the Naserte Tuffs from sections 21 and 39. Section 21 E x planation SE Breccia deposits "St rat i f ied tephra - Massive tephra ■- Accret ionary lapilli Grading of lithic component ■ None ▼ Inverse Section 39 54 Figure 14. Photograph of accretionary lapilli in airfall tuffs of the Naserte Tuffs near at section 39. A lahar deposit overlies the tephra beds in most sections, and ranges from 40 cm to over 12 m thick (Figure 13). At section 39 the bed is a minimum of 8.5 m thick but is truncated by a normal fault. The deposit exhibits crude inverse grading with altered pumice clasts, up to small boulder size lying on the exposed upper surfaces of lahar deposits. Smaller pumice clasts, 1 to 2 cm in long dimension, have been altered to a clay or zeolite and frequently weather out to produce a vesicular texture over the entire outcrop. Minor amounts of small, calcite casts of wood, and rounded clasts up to large cobbles occur throughout this bed. A discontinuous, massive, coarse granule to coarse pebble, clast supported breccia lies along the contact at sections 19 and 21. The breccia consists of angular phonolite and basalt clasts with imprints of these clasts in the underlying bed. This bed rarely exceeds 3 cm in thickness and overlies a sharp, basal surface. Pumices in these lahar deposits contain abundant euhedral, sanidine phenocrysts and accessory euhedral amphibole and biotite phenocrysts. The acicular amphiboles have been tentatively identified as arfvedsonite. On the basis of phenocryst mineralogy, these tuffs are considered trachytic. The glass of larger pumice clasts has been altered to analcime, and the vesicles in the pumice are filled with calcite. The original matrix of these deposits has been entirely altered to analcime. Fluid escape structures and compaction deformation features are very common in the stratified tephra beds. The fluid escape structures cut across 10 to 25 cm of interbedded stratified and massive beds (Figure 15). Compaction deformation, resulting from the deposition of the overlying lahar deposits, consists primarily of contorted bedding and an irregular pinch and swell bed geometry. Soft-sediment injections of underlying material into overlying lahar deposits are also common. At section 21, the lahar deposit contains a fragment of the slightly deformed stratified tephra about 1 m long and 15 to 25 cm thick. 55 I 56 Figure 15. Photograph of fluid escape structures in airfall tuffs of the Naserte Tuffs near section 8 . The brunton compass is 7 cm wide. Potassium-Argon Age Determinations Potassium-argon age determinations on sanidine phenocrysts separated from pumices of the Naserte Tuffs yielded an average age of 16.8 + 0.2 Ma (Table 1). The samples, Los 4-23 B and K87-3468, were collected at sections 4 and 31. Nakwel Esha Beds The Nakwel Esha beds are important for correlation purposes, but are not well enough exposed to be formally defined as a member. This sequence is composed of stratified and massive tephra; and volcanic sandstones and conglomerates. The best exposures, albeit poor, occur at section 19, and consist of 13 m of tuffs overlain by a minimum of 8 m of conglomerates. The tuffs of the Nakwel Esha beds are grey to pale yellowish grey and thinly to medium bedded with sharp, nonerosional basal contacts. Strata sets consist of 4 to 60 cm thick basal pumiceous conglomerates with normally graded pumice clasts. Elongate pumice lapilli ranging from 2 to 2 0 cm in long dimension lie parallel to bedding surfaces, and constitute up to 45% of some beds. Bed cosets fine vertically as massive beds grade to thinly laminated beds. A few stratified beds appear to have low-angle trough crossbedding. Coarse pumice lapilli are crudely graded (although widely scattered) with very large pumice lapilli in the higher, finer beds. The pumiceous conglomerates grade into dark brown, massive mudstones that contain abundant fine, altered pumice lapilli. Both the groundmass and phenocrysts of these pumices consist of alkali feldspar. The pumice lapilli are primarily yellow or pink and up to 4 cm long, but one deposit contains slate green pumice up to 20 cm long. Pink pumice bearing beds are most common near the top of this unit and can be correlated from the type section to section 3, and to similar beds in section 6 .2 . 57 Above the lower tephra layers of the Nakwel Esha beds lies a yellow, medium to coarse, thinly bedded, horizontally stratified sandstone with a scoured basal contact interpreted as reworked material from tephra and lahar deposits. The sandstone is overlain by a clast supported, phonolite pebble to cobble lag deposit with an irregularly scoured contact. The bed grades to a very coarse, poorly sorted trough cross bedded sandstone. In section 19, the pink pumice beds are overlain by very coarse conglomerates, conglomeratic sandstones and sandstones. These sediments are moderately well-exposed along the east side of Lokipenata Ridge from the Kalakol-Lodwar road south to approximately section 4, and additional exposures occur at the top of section 19. These correlate with the very poorly exposed lahar deposit in section 21 (bed 53). The base of the sedimentary sequence consists of yellow to pale yellow, matrix supported, medium to coarse pebble, clast supported conglomerates consisting of phonolite, minor amounts of basalt and abundant pumice. Sandstones contain discontinuous small cobble to medium boulder lenses overlying irregular, basal scour surfaces. These upward fining lenses are roughly 5 to 8 m long and up to 1 m thick composed predominantly of phonolite clasts. The yellow to pale yellow sandstones occasionally contain abundant altered pumice. Bedding structures are poorly preserved, but a few sandstones grade to dark brown, massive, very sandy mudstone. Unassigned Strata of the Lower Lothidok Formation Two principal stratigraphic intervals, one between the Kalodirr Tuffs and the Naserte Tuffs and the second between the Naserte Tuffs and the Nakwel Esha beds, consist of heterogeneous clastic sedimentary and pyroclastic rocks. Because individual beds are laterally discontinuous, they cannot be correlated from one local section to another and are not assigned a specific stratigraphic rank. For this reason, only the principal lithologic types of these intervals are given below. Detailed descriptions of these strata are given in the measured sections (Appendix B; Boschetto, 1988). Sedimentary Rocks Conglomerates average 1 m in thickness, and are moderately resistant so that they crop out as low ridges. Their colors include pale red to dark reddish brown, yellow to yellowish brown, and pale orange to brownish orange. In some exposures dark red conglomerates contrast strikingly with interbedded pale yellow tephra. They are similar in clast lithology, contact relationships, and bedding structures to the conglomerates described in the Basal Conglomerate Member and within the Kalodirr Tuffs. Etheria elliptica (fresh water oysters) mounds are found in situ near the base of some conglomerates in these intervals. These typically grade laterally and vertically to sandstones. Sandstones generally form low outcrops, but a few form small ridges. Colors are simlar to those of the conglomerates described above, but a few are reddish gray, grey, or pale olive. The variety of colors results from varying amounts of tephra reworked into the sandstones. They range from litharenites to feldspathic litharenites (McBride, 1963; Folk, 1968), and are cemeted with calcite, or, less commonly, hematite. These rocks grade laterally to either coarser or finer sedimentary rock, and also interfinger with other sedimentary beds. Individual beds range from 20 cm to 2.5 m thick, averaging less than 1 m. Most beds are primarily irregular in form (rarely tabular), and basal contacts are predominantly gradational. Where conglomeratic, the clast lithology is the same as that of associated conglomerates. Sedimentary structures in the sandstones are dominated by medium- to small-scale trough crossbeds and internal scour and fill sequences, the latter generally overlain by thin, discontinuous, pebble-grade conglomerates. The beds are normally or inversely graded with cosets that primarily fine upward. Irregular, subhorizontal erosional surfaces commonly cut across subordinate bed boundaries. Less common sedimentary structures include low-angle, planar and epsilon cross-stratification, horizontal stratification, and ripple stratification. Some finer-grained sandstones contain downward branching traces of irregular diameter filled with a white, chalky calcareous material that are interpreted as rootcasts. Vertical traces of constant diameter with a meniscate backfill of darker and finer grained sedimentary material also occur and are interpreted as burrows. Siltstones are most common in southern exposures of the lower Lothidok Formation (sections 10.1, 10.3 and 21). Outcrops of these pale red, reddish tan, and pale yellowish tan strata are generally low but also occur as interbeds in cliff forming, upward fining sequences. They resemble altered tephra deposits, but are distinguished by associated clastic sedimentary rocks and lack of euhedral volcanic mineral grains. The siltstones are moderately to well-consolidated, average 60 cm thick, and are usually lenticular with gradational lower and upper contacts. These are thinly laminated to thinly bedded, or planar to ripple stratified. Some appear massive, perhaps as the result of extensive bioturbation. Vertically and occasionally laterally, the siltstones grade to silty mudstones. Flattened pumice clasts are abundant in these beds, and resemble rip-up clasts. Lenticular (rarely tabular), dark brown to reddish brown, or tan to yellowish brown mudstones crop out poorly. These commonly interfinger with, or are interbedded with coarse, clastic sedimentary rocks. They are generally massive, but some exhibit ripple stratification or lamination. Basal contacts are either sharp or gradational, whereas upper contacts are generally sharp and erosional. White, chalky, calcareous root casts and thin tabular calcretes are common. Tan to dark brown, waxy mudstones commonly overlie tephra beds. These contain angular to subrounded lithic fragments, volcanic mineral grains (sanidine, biotite) and altered pumice lapilli similar in petrology and abundance to the underlying tephra. Fine- to 60 very coarse-grained sand occurs in these beds as thin lenses or scattered throughout. Lighter colored, flattened pumice lapilli give the appearance of intraclasts. Lahar deposit outcrops range from well-defined ridges to unconsolidated piles of angular boulders. Beds range from 30 cm to 15 m thick, and are thickest and are more numerous in the southern part of the Lothidok Range. The lahar deposits commonly overlie tephra beds, with sharp, nonerosional basal contacts. They are massive to inversely graded and contain 10 to 35 % granules to large boulders (up to 2.5 m diameter) in a yellow clay to sand size matrix. Phonolites, coarsely porphyritic and aphyric basalts, altered pumice, and indurated blocks of tephra are the prevalent clast lithologies. The pumice clasts range from 10 to 30 cm in length and 2 to 10 cm in diameter, and have been altered to zeolites with calcite-filled vesicles. Leaf impressions and calcite casts of wood fragments a few centimeters long and a few millimeters across are common constituents of these deposits. Pyroclastic Rocks Numerically, these are the most frequently occurring beds, although they constitute less than 20% of the thickness of strata between defined members. The pyroclastic rocks encompass two petrologically distinct types, one that is mafic-alkaline, and another that is trachytic. The mafic alkaline tephra are confined to the lower levels of the lower Lothidok Formation and consist primarily of the Kalodirr Tuffs. Higher in the section, the pyroclastic rocks consist of trachytic tuffs that can only be broadly correlated by stratigraphic position. Further study based on detailed stratigraphy, clast petrology and alteration geochemistry may provide stronger correlations Directly above the Kalodirr Tuffs, is a sequence of massive, pale red, fine tuffs that have been altered to a zeolite or montmorillonite. The tuffs contain less than 10% euhedral amphibole and pyroxene phenocrysts and 15-25% medium to coarse, angular to subrounded lithic grains. Basal contacts are sharp, but do not appear to be erosional. All pyroclastic deposits higher in the section are stratified and massive trachytic tuffs commonly interbedded on a centimetric to decimetric scale. These tuffs are pale yellow, tan to yellowish brown, and yellowish grey, which makes them very distinct visually. Cumulative bed thicknesses average 10 to 30 cm and rarely exceed 2 m. The best exposures of these trachytic tephra are at sections 18 and 19 (Figure 3). The stratified tephra layers are thinly to thickly laminated with fine to coarse, planar to ripple stratification. Trough cross beds occur in some of these stratified tuffs. The lower contacts are sharp and nonerosional, and the basal strata commonly drape small topographic irregularities. Massive tephra layers occur as thin to thick, tabular interbeds between well-stratified tephra layers. Accretionary lapilli and pumice lapilli are commonly associated with these massive beds. The accretionary lapilli range from 2 to 4 mm in diameter and may account for 15 to 35% of some beds. Very fine to fine euhedral amphibole, sodic sanidine, biotite and angular lithic grains comprise up to 20% of these tuffs. The matrix is now composed of either analcime or montmorillonite believed to represent altered volcanic ash. Very fine to coarse pumice lapilli comprise up to 50% of the tephra. The original glass of the pumice has been altered to clays and/or zeolites, and the pumice clasts are generally flattened and confined 1 to 5 cm thick layers. Calcite fills the vesicles of nonflattened pumice, which is recognizable only by the ghost outlines of the original glass. Pumice replaced by clays and zeolites is recognised by the presence of volcanogenic minerals and/or distinct color differences wtih the matrix. Crude normal grading of lithic fragments and inverse grading of pumice lapilli is common, but symmetric (normal-inverse-normal or inverse-normal-inverse) grading also occurs. Asymmetric ripple cross-stratification with indices of about 0.6 is seen in some tuffs, but at two locations (sections 1 and 31) the ripples appear to be symmetric. Planar and trough cross-stratification, and soft sediment deformation also occurs, but individual thin beds and thick laminae appear structureless. Raindrop imprints and desiccation cracks occur on the upper surfaces of fine tuffs. Contorted and convolute bedding formed by compaction deformation, and water escape structures are common in these beds. The tuffs also contain well-preserved imprints of grass and palm leaves (Figure 16 ). Several types of burrows are commonly well-preserved in the yellow tuffs that have alternate laminated (1-8 cm) and massive (3-8 cm) layers. Vertical, meniscate filled burrows of uniform diameter up to 8 mm across and 8 cm long (Figure 17) typically cut across two to three layers of sediment. These are usually simple, but some are branched, and are best observed in the laminated strata. Similar burrows oriented horizontally are also present. Other trace fossils consist of irregular networks of variable diameter, and horizontal to subvertical tunnels and shafts. The organism(s) that produced the traces is unknown. Fluvially reworked tephra (volcanic sandstones and conglomerates) overlie bedded tephra deposits in many sections. These consist of an altered ash matrix with fine to coarse, rounded to subrounded lithic grains that are petrologically similar to underlying or correlative beds. These beds are lenticular, and commonly have small- to medium-scale trough crossbeds, sharp basal scours, rip-up clasts, abraded volcanic phenocrysts and fine to very coarse pumice pebbles. Pumice is most abundant at the top of reworked deposits where it may constitute as much as 85% of a bed. These beds commonly fine upward to ripple stratified fine sandstones and silts tones. 63 64 Figure 16. Photograph of a palm leaf imprint in an unassigned trachytic tuff at section 27 (photo by F. H. Brown). 65 Figure 17. Photograph of a burrow in an unassigned trachytic tuff taken near section 19 (photo by F. H. Brown). Paleocurrents - Lower Lothidok Formation Paleocurrent measurements from the Basal Conglomerate Member (Figure 18) indicate sediment transport was to the southwest (means = 230° and 237°). The measurements were taken from the upper 30 m of outcrops near section 1 in the north and from two outcrops within the lower 25 m of section 20 in the south. Additional measurements on strata between the Kalodirr and Naserte Tuffs indicate a south (mean = 172°) to southwest (mean = 245°) sediment transport direction (Figure 19). In the north, groups of measurements were taken from five beds in section 1.2 over a total vertical thickness of approximately 50 m. Data for the southern part of the region were collected at one outcrop in section 1 0 .1 from three continuous sequences of sediments over a stratigraphic interval of roughly 2 0 m. Upper Lothidok Formation - Tlu The upper Lothidok Formation is best exposed south of the Kalatum and Alomonet Rivers. Additional exposures exist in the Nakwel Esha graben (see structure discussion, p. 126), along the Nakwel Esha River, and west of Lokipenata Ridge. There are no complete, well-exposed, continuous sections of the upper Lothidok Formation requiring the use of four sections to construct the sequence. Four marker horizons, the Lokipenata conglomerate, the Kalatum basalt, the Akwang'a basalt, and the Kamurunyang lahar (Figure 9), are discussed in detail and the unassigned strata between these levels are also briefly addressed. Lokipenata Conglomerate fNew Informal Name') The Lokipenata conglomerate, best exposed along the east side of Lokipenata Ridge, forms the base of the upper Lothidok Formation. Exposures of this conglomerate extend from the north side of the Kalakol-Lodwar road south to section 4, and also 66 Figure 18. Paleocurrent rose diagrams for the Basal Conglomerate Member of the lower Lothidok Formation for data sets from near section 1 and from section 20. All data were corrected for tectonic tilt. 69 Figure 19. Paleocurrent rose diagrams for unassigned strata of the lower Lothidok.Formation for data sets from multiple outcrops. All were corrected for tectonic tilt. 70 l 3 5 ° 5 0 ' E - 3 ° 3 0 N Lothidok Ran K 3 ° 2 0 'N 5 km A ' Location of measurements j Mean direction of sediment t ranspor t 3 6 ° 0 ‘E N-Mumber of measurements 1 1 occur at the top of section 19. The unit is poorly exposed in the road cut through ' Lokipenata Ridge and at the tops of sections 21, 30 and 31 (Figure 3). The Lokipenata conglomerate is a minimum of 40 m thick at section 30 but is not well enough exposed at other sections to determine thicknesses for comparison. The presence of the conglomerate at section 19 and along the Lokipenata Ridge suggests reasonable lateral continuity, but the overall bed geometry is indeterminate. The lower contact is a broad, irregular erosional surface with intermittent deep, narrow scours. The conglomerate consists of very coarse pebbles to very coarse cobbles in clast (rarely matrix) support, and is pale red to dark red, dark reddish grey, or reddish orange. Clasts are imbricated and crudely graded, extremely weathered, and typically stained with^ hematite. Rip-up clasts from the underlying bed are common. The matrix consists of very fine to very coarse, subrounded to subangular sand grains of quartz, pyroxene and minor potassium feldspar. This marks the first occurrence of significant amounts of basement material in the section. Soil carbonates are common near the base of the contact. Kalatum Basalt ("New Informal Unit) The Kalatum basalt consists of at least two and possibly three flows with a minimum total thickness of 20 m. This basalt is exposed in a series of small hills west of and parallel to the southern end of Lokipenata Ridge (sections 3 and 4), and also farther south where the basalt is offset by small normal faults. Small exposures occur as hillocks in the stream bed of the Kalatum River. In section 30, the Kalatum basalt lies approximately 15 m above the highest exposure of the Lokipenata conglomerate. The thickest exposures lie just north of Ngaletiti Hill (Figure 3), where the basalt is truncated by a normal fault. The lowest part of the Kalatum basalt is an agglomerate (12 m) with blocks up to 2 m in diameter consisting of plagioclase phyric basalt in a green, aphanitic groundmass. 7 1 The basal agglomerate is overlain by a thin, green, aphyric flow ( 8 m) that is either overlain by, or gradational into, a coarse plagioclase phyric flow with a similar groundmass. The porphyritic flow contains coarse plagioclase and amphibole phenocrysts up to a few millimeters long. Potassium-Argon Age Determinations Potassium-argon age determinations on plagioclase phenocrysts separated from the basalt matrix of the Kalatum basalts yielded an average age of 13.6 ± 0.2 Ma (Table 1). The samples, K86-2743, were collected near section 4. Akwang'a Basalt (New Informal Name) The Akwang'a basalt, interpreted to he above the level of the Kalatum Basalt, crops out only in sections 24 and 25 (Figure 3). It is extremely weathered and poorly exposed. In outcrop the basalt is dark grey to olive grey and is coarsely porphyritic with olivine, plagioclase, and pyroxene phenocrysts. Calcite amygdules and cavities filled with euhedral calcite crystals reach 3 cm in diameter. Other cavities are lined or filled with analcime crystals. Kamurunvang Lahar fNew Informal Name) The Kamurunyang lahar forms a continuous yellowish brown ridge of east dipping strata at the type locality (sections 24 and 25). Blocks over 4 m in diameter have weathered from this ridge and litter the scree slope to the west. This bed is here informally named the Kamurunyang lahar after Kamurunyang Hill which lies between sections 24 and 35. At sections 8 and 29, the lahar forms a small continuous ridge. At section 35 the lahar also forms a ridge but is partly buried by recent alluvium derived from the overlying Loperi 72 basalts. The correlations between the type locality and sections 8 , 29 and 35 are not completely secure (see below) and should be used with caution. The Kamurunyang lahar consists of 15 to 40 % pebble to boulder size clasts, of predominantly phonolite but with minor basalt and indurated lahars, in a matrix presently consisting of analcime. Lensoidal, clast supported conglomerates occur above sharp, nonerosional, basal contacts and as discrete lenses within these beds. These exhibit no obvious vertical or lateral gradation. The lahar deposits also contain abundant pumice clasts up to 20 cm in length and 8 cm in diameter weathering from the upper surface. The pumices are very similar in mineralogy and alteration to the lahars from which they are derived. Smaller pumices have altered to clay and weather to producing a vessicular texture of the outcrop. Correlation of Lahar Deposits Correlation of the lahar deposits of the upper Lothidok Formation is important because they overlie fossil localities (see fauna discussion, p. 142) and pumice clasts within them are the only material available for potassium-argon age determinations. These deposits, however, are virtually indistinguishable on the basis of outcrop morphology, rock texture, pumice petrology, or alteration geochemistry, and the lateral relationships are therefore problematic. The lack of associated airfall tephra and proximal marker horizons requires the correlations be based on the stratigraphic position of the deposit relative to the upper boundary of the Lothidok Formation (the base of the Loperi basalts). Because this contact is an angular unconformity (see below), its level is difficult to determine and these correlations are tenuous. Here it is proposed to correlate the lowest lahar deposit at sections 24 and 25 with the thick lahar deposits between 40 to 60 m below the Loperi basalts in sections 8 , 29 and 35. Because section 25 contains three lahars, one might correlate the upper lahar of section 73 25 with the only lahar in section 35. However, the similarities of outcrop morphology, underlying strata and internal geometry indicate that the lower lahar in section 25 is equivalent to the only lahar in section 35. The small distance between these sections requires that the upper two lahars of section 25 were removed by erosion in section 35 prior to deposition of the Loperi basalts. Taking the correlation further, the lahar of section 35 might be equivalent to eroded remnants of a similar deposit in sections 26 and 27. Problems with correlations between the type locality and sections 8 and 29 result mainly from isotopic age determinations, which are addressed below. Potassium-Argon Age Determinations Potassium-argon ages were determined for sanidine phenocrysts separated from pumice of the Kamurunyang lahar deposits at the type locality (section 25) and the tentatively correlative lahar deposit below the Loperi Hill (sections 8 and 29). Application of these ages to the deposits, however, must be made with caution, as the ages represent the time of eruption of the volcanogenic material and not necessarily the time of lahar deposition. One sample (K87-3411; section 25) yielded an age of 13.1 ± 0.2 Ma, but two other samples (K87-3437 B and K87-3437 C+D; section 29), yielded ages of 13.4 ± 0.2 and 14.0 + 0.2 Ma respectively (Table 1). Although the Kalatum basalt and the lahar deposit (Kamurunyang?) do not crop out in a continuous section, the basalt correlates to a position well below the lahar deposit. The age on the basalt (13.6 ± 0.2 Ma), determined from a plagioclase separate, is believed reliable and thus the older age of 14.0 Ma on the lahar appears discrepant. This can be accounted for by either a small amount of contamination in the sanidine separate or the pumice was from an earlier eruption. TTnassigncd Strata - Upper Lothidok Formation Three stratigraphic intervals, one between the Kalatum and Akwang'a basalts, one between the Akwang'a basalt and the Kamaurunyang lahar, and one between the Kamaurunyang lahar and the upper contact of the Lothidok Formation (Figure 9) consist of heterogeneous clastic sedimentary and pyroclastic rocks that are not assigned specific stratigraphic rank. In most respects these strata are essentially the same as those of the lower Lothidok Formation, and therefore only the most distinctive characteristics of the strata will be discussed. Detailed descriptions of these strata are given in the measured sections (Appendix B; Boschetto, 1988). Clastic Sedimentary Rocks The conglomerates are best represented in the lower part of this interval, whereas sandstones and siltstones dominate the upper part. The sedimentary deposits typically fine upwards as does the upper Lothidok Formation. The red to pale red, reddish grey, or white conglomerates of this interval are 2 to 8 m thick. These form discontinuous basal deposits in upward-fming sequences and discrete beds within finer sedimenatry strata that are bounded by sharp upper and lower contacts. Clasts range from coarse pebbles to large cobbles composed primarily of phonolites and basalts. Quartz and perthite clasts are common only near the top of the section. Some beds contain boulders of indurated tephra and lahar deposits altered to analcime and montmorillonite as are the in situ beds of this type. The matrix of these rocks consists of very fine sand to coarse granules of volcanic and basement detritus. The amount of basement detritus increases from less than 10% to -60% upwards in the section. Etheria elliptica. mounds he near the base of some conglomerates. The 2 to 8 m thick sandstones are tan to grey, pale red and pale reddish tan, and occur in upward-fming sequences grading into very fine sandstones and siltstones. The 75 moderately to well-sorted sandstones consist of subrounded to subangular, fine to medium grains. Some sandstones are conglomeratic typically with fine to medium pebbles, but locally with small boulders. Petrologically the sandstones of the upper Lothidok Formation are quite distinct from those of the lower Lothidok Formation. The quartz component of the sand fraction increases upwards in the section and constitutes 25 to 80% of the higher beds. Perthite occurs infrequently at the base of the interval but increases to about 30% upwards in the section. The general increase in abundance of basement detrital component upwards in the section is reflected by feldspathic litharenites near the base of the section, but lithic subarkoses, subarkoses, and occasional arkoses higher in the section. Amphibole and biotite grains vary from 5 to 15% of the total rock and volcanic rock fragments range from 5 to 75%. In contrast with the lower strata, epsilon crossbedding and climbing ripples are common sedimentary structures in the upper Lothidok Formation. Tan, reddish tan, or reddish grey, 50 cm to 2 m thick siltstones are common. These are commonly bounded by gradational contacts, and constitute 35 to 45 % of the upward-fining sequences in the upper part of the section. Horizontal and ripple stratification is well-preserved, and climbing ripples are common. Thinly interbedded massive and laminated layers are similar to those in the fine-grained sandstones in which the massive layers are interpreted to result from bioturbation. The siltstones interfinger with and are overlain by tan to brown, moderately to very sandy mudstones. Dark brown and tan to greyish brown mudstones commonly occur at the top of upward-fining sequences. Laterally continuous mudstones generally have gradational lower and upper contacts, but lenticular mudstones have sharp basal contacts and interfinger with crossbedded sandstones and laminated siltstones. The massive, lenticular mudstones contain abundant very fine- to fine-grained angular sand. Pedogenic features including calcretes and calcareous root casts are common. 76 The numerous lahar deposits in the upper Lothidok Formation are virtually indistinguishable from the Kamurunyang lahar. These ranging from 50 cm to 15 m thick with at least three over 8 m thick. All three of the thickest deposits crop out only at section 25 where the lowest bed forms a pronounced ridge, and the upper two beds crop out very poorly in the riverbed to the east. Thick lahars deposits are also exposed in sections 8 , 24, 26, 29, and 35 (Figure 3). Sections 8 and 24 contain two such deposits while sections 26, 29 and 35 each have only one. These deposits form small, indistinct ridges at each section. Section 27 contains boulders up to 4 m in length that appear to be remnants of an eroded lahar deposit. In contrast to similar deposits in the lower Lothidok Formation none is associated with tephra beds. Pyroclastic Rocks Tephra layers in the upper Lothidok Formation are virtually indistinguishable from those below. Most are interbedded with tan to brown, massive mudstones, but these crop out poorly and the relationship is unclear. The uppermost exposed beds of the Lothidok Formation (< 5 m) consist of extensively weathered massive and stratified tuffs with interbedded mudstones. Thin soil carbonate horizons are common, and the beds contain abundant, small, altered pumice lapilli. The contact between these tuffs and the Loperi basalts is buried by basalt talus. The extreme weathering at this level may be the result of prolonged exposure with consequent pedogenesis. Paleocurrents - Upper Lothidok Formation Paleocurrent measurements from the Lokipenata conglomerate from imbricated clasts at sections 3 and 19 (Figure 20) imply sediment transport was to the east (means = 87° and 90°). Measurements from imbricated clasts and ripple stratification in unassigned 77 78 Figure 20. Paleocurrent rose diagrams for the Lokipenata conglomerate of the upper Lothidok Formation for data sets from sections 3 and 19. All data were corrected for tectonic tilt. VWW V O W V - * 79 1- 3 5 ° 5 0 ' E 4 - 3 ° 3 0 ' N 3 ° I 0 'N 0 Location of measurements ' * * - Mean direct ion of sediment t ra n s p o r t A/ = Number of measurements - r~ 3 6 ° 0 ‘E 1 1 strata of sections 26 and 29 (Figure 21) also indicate sediment transport to the east (means = 71° and 60°). Loperi Basalts - Tib In its southern exposures the Lothidok Formation is overlain by basalt flows (Tib - Plate; Tvb2 of Walsh and Dodson, 1969), which are herein named the Loperi basalts (informal) after Loperi Hill, which has the thickest exposures. These are well-exposed throughout the southern part of the Lothidok Range, and form most of the lava capped hills south of the Kalatum/Alomonet River. Exposures north of the Kalatum and Alomonet Rivers are limited to Moruorot and Esha Hills, and a progressively thinning ridge west of Lokipenata Ridge that terminates in small piles of basalt boulders. The Loperi basalts thin to the east, north and west (Figures 7 and 8 ), but thickness variation to the south is difficult to establish due to extensive faulting and poor exposures. Farther south, the basalt is buried by Quaternary deposits (Qal). These basalts are best exposed at Loperi Hill where they consist of three flows totalling a maximum of 121 m in thickness. Only one flow occurs at Ngaletiti and Esha Hills where it is about 30 m and 50 m thick, respectively. At Moruorot Hill the flow is about 5 m thick. The pronounced thinning to the northeast may have resulted when the flow encountered topographic relief such as a channel margin. The Loperi basalts contain minor (1 to 4 m) interbedded sedimentary and pyroclastic strata. These are only exposed along the north face of Loperi Hill where it is cut by the Kalatum River and were not examined in detail. The very poor exposures consist of very coarse conglomerates, sandstones, and extremely altered tephra. The basalts are predominantly olivine-augite-plagioclase phyric. Titaniferous augite phenocrysts occasionally reach 3 cm diameter, and are strongly zoned. Plagioclase (A1145. 70) phenocrysts are rarely zoned. Olivine phenocrysts are occasionally rimmed with iddingsite. The groundmass consists of plagioclase, olivine, clinopyroxene, iron-titanium 80 Figure 21. Paleocurrent rose diagrams for unassigned strata of the upper Lothidok Formation for data sets from sections 27 and 29. All data were corrected for tectonic tilt. 82 1 3 5 ° 5 0 ‘ E - 3 ° 3 0 ' N Lothidok Range t-3°20'N 3 ° I 0 ‘N 0 Mean direct ion of sediment t ra n s p o r t oxides (magnetite?), minor biotite and apatite, and possibly analcite. Partial alteration of both the groundmass and phenocryst olivine to chlorite is common. Calcite amygdules are common in some flows, and calcite also fills voids and fractures. Columnar jointing, roughly parallel to dip, occurs at or less than 8 m above the unexposed sediment-basalt contact at Moruorot Hill. Excellent exposures of this jointing exist along the eastern and southern flanks of the hill. Potassium-Argon Age Determinations Samples of the Loperi collected from the top of Moruorot Hill (B-20) and from Loperi Hill (B-31) yielded ages of 12.0 + 0.1 and 10.9 + 0.1 Ma respectively (Table 1). Although these determinations are technically good, they probably represent minimum ages. There is only one published isotopic age for the Loperi basalts. Zanettin, et al. (1983) give an isotopic age of 14.6 ± 1.0 Ma for a basalt in this area but the sample location is unclear. The coordinates given for the sample (K64) place it between Lodwar and the Lothidok Hills where there are no exposures of basalt, but Zanettin et al.(1983) map the basalt sample as the Turkana basalt and indicates that the sample was taken near Moruorot Hill (approximately 10 km north of the coordinates given). Undifferentiated Tertiary Deposits - Tu Tertiary deposits overlie the Loperi basalts in the south and the Lothidok Formation in the northern part of the Lothidok Range but were treated informally for this report. Similar strata are poorly exposed above the Loperi basalts on Loperi, Esha, and Konukuangna Hills, west Lokipenata Ridge and within the Kalodirr graben. The maximum measured thickness of these strata is 60 m in section 29, but the lowest beds are not exposed. Above Loperi and Esha Hills these beds consist of pebble and boulder 83 volcaniclastic conglomerates. At Ngaletiti and Kakurtua Hills the sediments directly overlying the Loperi basalts are quartz/perthite granule to pebble conglomerates and very coarse, poorly sorted arkosic sandstones. The conglomerates overlying Loperi Hill are very poorly exposed on the north face of Loperi Hill, where massive conglomerate beds are at least 15 m thick. A small fault block south of the Alomonet River roughly 3 km south of Moruorot Hill also has very poor exposures of these sedimentary rocks overlying basalts. The lowest decent exposures of undifferentiated Tertiary strata in a continuous section lie about 30 m above the Loperi basalts at section 29. Some of the best exposures are found in the cliff on the Kalodirr River (section 15) although their level within the Tertiary section cannot be determined with confidence. The deposits of the undiferrentiated Tertiary strata are dominated by very fine- to coarse-grained, poorly sorted to conglomeratic sandstones. These deposits are grey to pale reddish grey, orange, pale red, and white to grey lithic subarkoses to subarkoses with minor volcanic clasts with fine granule to very coarse pebble clasts. The sandstones commonly grade inversely to massive conglomerates with abundant tabular calcrete? horizons. Exposures of arkoses and subarkoses east of Moruorot Hill cannot be placed in the section Sedimentary structures in the undifferentiated Tertiary strata include small- to medium-scale trough crossbedding, basal and internal scours, and minor planar stratification. These structures are generally poorly defined and are commonly disrupted, possibly from bioturbation. Laminated and disrupted beds contain uniform, 4 to 8 mm diameter, straight burrows filled with hematite-stained, muddy sandstone. Except for minor interbedded volcaniclastic sediments, the conglomerates consist of granules to large pebbles of quartz and K-spar (perthite) with a subarkosic matrix. These grade into, or are equivalent to, the sediments overlying Ngaletiti and Kakurtua Hills. The 84 poorly consolidated matrix of these sediments weathers to leave unconsolidated sheets of quartz and perthite clasts. These cover much of the area surrounding the Lothidok Range including that east of Kakurtua Hill, and the area west of the small basalt outcrops west of Lokipenata Ridge. Sandstones grade vertically to tabular, sandy, massive mudstones up to 2 m thick. The section as a whole fines upwards and lenticular mudstones become more abundant. These mudstones are commonly interbedded with and interfinger with the sandstones, occasionally dominating the sequence. Siltstones occur in gradational sequences but account for little of the section. Pedogenic features such as calcareous root casts, blocky fabric, and tabular calcrete? in the thick, massive mudstones are common. The undifferentiated Tertiary sediments are interpreted to directly overlie the Lothidok Formation at all locations beyond the extent of the Loperi basalts, but the relationship is not clearly displayed. Only in section 6.2, where the contact is not well-defined, can this relationship be seen. Paleocurrents - Undifferentiated Tertiary Deposits Paleocurrent measurements (Figure 22) from imbricated clasts and trough crossbeds imply a sediment transport direction to the east (mean = 55°). Most of these measurements are from a small outcrop along the Kalodirr River. 85 86 Figure 22. Paleocurrent rose diagrams for undifferentiated Tertiary strata taken along the Kalodirr River. All data were corrected for tectonic tilt. 87 INTERPRETATION OF DEPOSITIONAL ENVIRONMENTS Introduction Descriptions of the strata in the preceding section show that only a few sedimentary rock types are represented in the Lothidok Formation and that these lithologies occur in distinct associations. Consideration of sequential and lateral relations of genetically related lithologic types leads to a better understanding of the conditions under which these rocks were deposited. The lack of marine fossils and laminated fine sedimentary rocks coupled with the presence of terrestrial vertebrate and plant remains indicates the rocks were deposited in alluvial and fluvial systems. Numerous studies have addressed fluvial depositional systems. In this study, lithofacies codes (Miall, 1978; Mathisen and Vondra, 1983) are used to define equivalent lithologies of the Lothidok Formation to help interpret depositional environments. The code system, introduced by Miall (1978), consists of a capital letter designation for the dominant grain size followed by one or two lower case letters that refer to principal sedimentary structures. Miall (1978) assigned codes to lithofacies of braided streams and Mathisen and Vondra (1983) assigned codes for meandering stream and pyroclastic deposits. For comparison, their fades codes are given for equivalent lithologies of the Lothidok Range (Table 2). Interpretations of the depositional environments for the sedimentary rocks of the Lothidok Range follow each discussion addressing lithologic associations. Detailed analysis of depositional environments proved difficult because of poorly exposed and laterally limited outcrops. The lack of any three-dimensional or decent two-dimensional exposures precludes defining the external geometry of the clastic sedimentary bodies. The Table 2. Lithologic types of the Lothidok Formation, equivalent lithofacies codes, prominant sedimentary structures and interpretations of depositional settings. Lithology Lithofacies Code Sedimentaiy structures Interpretation massive, matrix supported gravel. Gmsa none to crude inverse grading debris flow massive or crudely bedded gravel Gma horizontal bedding, imbrication longitudinal bar, lag or sieve deposits stratified gravel Gt3 medium to large scale trough crossbeds minor channel fill stratified gravel Gpa planar crossbeds lingoid bars growths from older bars sand, medium to very coarse/pebbly Sta solitary (theta) to grouped (Pi) trough crossbeds dunes (lower flow regime) sand, medium to very coarse/pebbly sPa solitary (alpha) to grouped (omikron) trough crossoeds transverse bars, sand upper flow regime) sand, very fine to coarse Sr3 Ripple marks,small-scale trough crossbeds ripples (lower flow regime) sand, silt, very fine to very coarse/pebbly Sha parallel laminations planar bedding planar bed flows (lower and upper flow regime) sand, fine to medium grained Sla low angle (<10) trough crossbeds scour fills,antidunes sand, fine to coarse with intraclasts Sea crude crossbedding erosional scours scour fills sand, fine to coarse/pebbly Ssa broad, shallow scours scour fills sand, fine grained lenticular coarse/pebbly Sc13 climbing ripple laminations, may appear massive crevasse splay sand, silt, mud Fia laminated, very small ripples climbing ripples overbank or waning flood mud, silt Fma massive overbank or drape deposits a Lithofacies defined by Miall (1978). b Lithofacies defined by Mathisen and Vondra (1983). depositional environments of the sediments of the Lothidok Range are therefore interpreted from internal geometry, nature of contacts and scale of bedding observable in vertical sequences. Allen (1983) and Miall (1985) have recently addressed problems with deciphering the nature of depositional systems from vertical sequences alone. Additional work in the Lothidok Range may provide a more comprehensive analysis of the depositional systems of the Lothidok Formation. Lithologic Associations Sedimentary strata in the Lothidok Range may be grouped into two distinct lithologic associations. These are named for their characteristic lithologies with the dominant lithology listed first followed by subordinate lithologies. Acronyms are used to refer to each lithologic association. The most common lithologic association in these rocks consists of clast or matrix supported conglomerates, trough cross-bedded sandstones, and minor beds of finer grained rocks. This normally graded assemblage is dominated by conglomerates with subordinate finer grained rocks. The acronym used for this association is CSM (conglomerate-sandstone-mudstone). The second lithologic association in the Lothidok Formation consists of well-sorted, trough cross-bedded sandstones, laminated siltstones and mudstones. This assemblage differs from the first mainly in the lack or paucity of conglomerates. Laminated siltstones and massive mudstones are subordinate lithologies. The acronym used for this association is SSM (sandstone-siltstone-mudstone). The sedimentary rocks within the Kalakol basalts and those of the lower Lothidok Formation belong primarily to the CSM association with only minor occurrences of the SSM association. The base of the upper Lothidok Formation consists of rocks primarily of 90 the CSM association, but rocks of the upper part of the unit and the Tertiaiy undifferentiated strata belong to the CSM association. CSM Association The CSM association (Figure 23) consists of upward fining sequences ranging from conglomerates to mudstones that generally, but not always, contain each of the lithologies. Partial assemblages are too rare as distinct deposits to warrant establishing new associations. Clast-supported, massive to crudely bedded conglomerates occur primarily at the base of the association and also as discrete interbeds bounded by sharp contacts. Massive bedding commonly grades vertically into medium- to large-scale trough crossbeds. Conglomerates also occur as thin basal lenses, and as lenses interbedded within sandstones. These grade laterally and vertically into poorly sorted coarse grained or conglomeratic sandstones that are second in abundance to the conglomerates. The sandstones contain deep internal and broad basal scour surfaces, irregular, subhorizontal erosional surfaces, crude low-angle, medium- to small-scale trough and planar cross­stratification and rare horizontal stratification. Epsilon crossbedding (representing lateral accretion surfaces) is very rare but occurs at local sections. Cosets of cross-strata contain normally and inversely graded strata sets. The coarse-grained sandstones typically grade vertically into finer grained strata. Fine-grained sandstones and siltstones are ripple and planar stratified. Massive mudstones occur as interbeds with either sharp or gradational lower contacts, and with sharp, erosional upper contacts. These fine-grained rocks account for an extremely variable amount of this lithofacies. In northern exposures, they constitute less than 15% of the section, but some sequences in the south consist of up to 40% siltstones and mudstones. 9 1 92 Figure 23. Typical sequence of the Conglomerate-Sandstone-Mudstone (CSM) lithologic association. Facies codes are given in Table 2. 93 Thickness in meters o0> c<D k. w3UO O V e r t ic a l sequence ( v a r i a b l e ) 0-1 0-1 common - - rare Nature o f Contact D omin a n t L i t h o lo g y (fades codes) Sed ime nt S t r u c t u r e s - Interpretation General C h a r a c t e r i s t i c s sharp, erosiona! Mudstone (Fm) tabula r c a l c r e t e s - paleosols c a lc a r e o u s root c a s ts -paleosols sharp, rarely gradational Si l ts ton e (Fm ,F I ) bur row, lam s c r ams r a r e cl imbing r ipples - crevasse splay f gradational Sandstone (St, Sh, S r , Ss, S e ) rare r ipple s t r a t i f i c a t io n - lower flow regime smal l t rough crossbeds - lower flow regime planor beds - upper flow regime 1-6 very common •. . « "OO'o ' • *0D#J *f t "* */. basal co n glome r a te s over in t e rn a l scours - channel lag deposits medium t ro u g h c ros sb ed s - dunes (lower flow regime) inte rb ed ded co n g lome r a t e s - channel lag deposits c o n g lome r a t i c o o gradational and erosiona/ Co n g lome r a t e (6m, G t ) 1 -4 very common 0-1 common medium t rough c ro s sb ed s - channel f i l l deposits inte rna l scours la rg e t ro u g h c r o s s b e d s - channel f i l l deposits massive to crude n o rma l g r a d in g , imb r i c a t e d c la s t s - longitudinal bar and channel lag deposits Particle grade Interpretation The CSM association closely resembles facies interpreted as braided stream deposits on alluvial fan complexes (Rust, 1978; Mathisen and Vondra, 1983). The sedimentary structures and dominance of coarse grain sizes of the CSM association indicate high energy stream deposition. The wide range of particle size implies substantial energy level fluctuations. These characteristics are commonly associated with braided streams. In such streams massive, clast supported, imbricated conglomerates are deposited as channel f