SUB- CONTINENTAL LITHOSPHERIC MANTLE DEFORMATION IN THE YERER-TULLU WELLEL VOLCANOTECTONIC LINEAMENT: A STUDY OF PERIDOTITE XENOLITHS By Kaitlyn R. Trestrail        A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Geological Sciences  Master of Science 2016          ABSTRACT SUB- CONTINENTAL LITHOSPHERIC MANTLE DEFORMATION IN THE YERER-TULLU WELLEL VOLCANOTECTONIC LINEAMENT: A STUDY OF PERIDOTITE XENOLITHS By Kaitlyn R. Trestrail Volumetrically, the lithospheric mantle comprises the bulk of the continental lithosphere, yet the mechanisms by which the lithospheric mantle is deformed during rifting are unresolved. Stretching and thermo-mechanical erosion are often cited mechanisms for facilitating lithospheric deformation during continental rift development; however, the infiltration of melt into the lithosphere during rift development also results in chemical alteration of the lithospheric mantle. The purpose of this study is to test the potential mechanisms by which the continental lithospheric mantle is chemically altered during rift development. Here we present a study of mantle xenoliths derived from the lithospheric mantle in Ethiopia that has been deformed during rifting. We find that the lithospheric mantle beneath the this zone exhibits evidence of focused magma-lithosphere interaction, resulting in four distinct types of peridotite xenoliths: a) deformed xenoliths representing the pre-deformation lithospheric mantle; b) granular xenoliths representing overprinted lithospheric mantle; c) replacement dunite xenoliths, evidence of pervasive melt-lithosphere interaction; and d) cumulate xenoliths representing remnants of a metasomatic agent. The deformed xenoliths exhibit a high Mg# (>89) and exhibit little, if any, interaction with melt. The remaining xenolith groups exhibit lower Mg# (<89) suggestive of magma-lithosphere interaction. The high Ni content in olivine and depleted incompatible elements in orthopyroxene of the granular xenoliths are inconsistent with simple metasomatic enrichment and the existence of dunite with olivine (low Ca and Sc) and spinel (Cr# ~60) compositions is inconsistent with a cumulate origin, instead suggesting a replacement dunite. Our samples are derived from a zone of intensely sheared lithosphere and we suggest that melt channeling is preferred over chromatographic metasomatism due to melt focusing along steep topography on the lithosphere-asthenosphere boundary and shear-induced porosity.               Copyright by  KAITLYN R. TRESTRAIL  2016   iv  ACKNOWLEDGMENTS This work was made possible by the support and encouragement from many people and parties:  Special thanks to National Science Foundation (NSF) Petrology & Geochemistry program: EAR- 1219647 for supporting this research. Your funding was instrumental to the data collection for this project. I would also like to thank the Michigan State Department of Geological Sciences for the fellowships and scholarships that made my education possible.  I would like to thank my committee members, Michael Velbel and Julie Libarkin, for all of the guidance and support during this project. Without your valuable feedback, this thesis would have been an impossible task. Thank you to Guillaume Girard for his help with the LA-ICP-MS on multiple data runs, data processing and precision analysis. His knowledge and improved mythology made these data possible.  Thank you to my family, Andy, Laurie, and Andrew Trestrail, for supporting me throughout my education, encouraging me to do my best, and for not making fun of me (too much) for having an unhealthy obsession with rocks.  Thank you Susan Krans, my big geosis, for all of your advice, time spent figuring out my data, and for being an incredible listener. Your friendship has been one of the greatest gifts I have received  Thank you to Charles Hackel, for continuing to be my best friend throughout without you.  Lastly, a very special thank you goes to Tyrone Rooney. You have been by my side for my entire research career and I can safely say that you have molded me into something that I am extremely proud d then ask another), how to write (better), and to always expect a visit from the Spanish Inquisition!   v  TABLE OF CONTENTS LIST OF TABLES ...................................................................................................................................... vii LIST OF FIGURES ................................................................................................................................... viii KEY TO ABBREVIATIONS ...................................................................................................................... ix 1. Introduction ............................................................................................................................................. 1 2. Background ............................................................................................................................................. 4 2.1 East African Rift System ................................................................................................................. 5 2.2 Magmatism in the EARS ................................................................................................................. 6 2.3 Rift evolution ................................................................................................................................... 7 2.4 Yerer-Tullu Wellel Volcanotectonic Lineament ............................................................................. 8 2.5 How mantle xenoliths record lithospheric mantle processes ......................................................... 9 3. Methods .................................................................................................................................................. 11 3.1 Sample selection ............................................................................................................................ 11 3.1.1 Analytical Methods - Laser ablation (LA)- ICP-MS ............................................................. 11 4. Petrography ........................................................................................................................................... 13 4.1 Deformed xenoliths ......................................................................................................................  14 4.1.1 Mineralogy ............................................................................................................................ 14 4.1.2 Textures ................................................................................................................................. 15 4.2 Granular xenoliths ........................................................................................................................ 15 4.2.1 Mineralogy ............................................................................................................................ 15 4.2.2 Textures ................................................................................................................................. 16 4.3 Cumulate xenoliths ........................................................................................................................ 16 4.3.1 Mineralogy ............................................................................................................................ 16 4.3.2 Textures ................................................................................................................................. 17 4.4 Dunite xenoliths ............................................................................................................................. 17 4.4.1 Mineralogy ............................................................................................................................ 17 4.4.2 Textures ................................................................................................................................. 17 5. Results .................................................................................................................................................... 18 5.1 Major, minor, and compatible element chemistry ........................................................................ 18 5.2 Incompatible trace element chemistry .......................................................................................... 19 6. Discussion............................................................................................................................................... 21 6.1 Location of samples within the SCLM .......................................................................................... 21 6.2 Textural evidence of deformation ................................................................................................. 22 6.3 Evidence of metasomatic agents ................................................................................................... 24 6.3.1 Evidence of hydrous metasomatism ...................................................................................... 24 6.3.2 Evidence of carbonatite metasomatism................................................................................. 25 6.3.3 Evidence of melt-lithosphere interaction .............................................................................. 26 6.4 Evidence of focused melt transport ............................................................................................... 28 6.4.1 Pre-existing subcontinental lithospheric mantle: Deformed xenoliths ................................. 29 6.4.2 Overprinted lithospheric mantle: Granular xenoliths .......................................................... 29 vi  7. Conclusions ............................................................................................................................................ 33 APPENDICES ............................................................................................................................................ 35 APPENDIX A TABLES ...................................................................................................................... 36 APPENDIX B FIGURES ..................................................................................................................... 46 APPENDIX C CALIBRATION STANDARDS .................................................................................. 60 APPENDIX D STANDARD DEVIATION OF SELECT CALIBRATION STANDARDS .............. 62 REFERENCES ........................................................................................................................................... 64   vii  LIST OF TABLES Table 1 Detection limits ............................................................................................................................ 37 Table 2 Olivine mineral chemistry .......................................................................................................... 38 Table 3 Orthopyroxene mineral chemistry ............................................................................................ 41 Table 4 Clinopyroxene mineral chemistry .............................................................................................. 43 Table 5 Spinel mineral chemistry ............................................................................................................ 44 Table 6 Regional thermobarometry ........................................................................................................ 45 Table 7 Calibration Standards................................................................................................................. 61 Table 8 Standard Deviation of Select Calibration Standards ............................................................... 63          viii  LIST OF FIGURES Figure 1 East African Rift System (EARS) Location ............................................................................ 47 Figure 2 Main Ethiopian Rift (MER) Location ...................................................................................... 48 Figure 3 Lithospheric Modification of the YTVL .................................................................................. 49 Figure 4 Xenolith Petrography ................................................................................................................ 50 Figure 5 Comparison of Xenoliths ........................................................................................................... 51 Figure 6 Olivine Major and Minor Element Chemistry........................................................................ 52 Figure 7 Orthopyroxene Major and Minor Element Chemistry .......................................................... 53 Figure 8 Spinel Major Element Chemistry ............................................................................................. 54 Figure 9 Orthopyroxene Trace Element Chemistry .............................................................................. 55 Figure 10 Clinopyroxene Trace Element Chemistry ............................................................................. 56 Figure 11 Xenolith Thermobarometry .................................................................................................... 57 Figure 12 Zr/Hf Ratio Chemistry ............................................................................................................ 58 Figure 13 Focused Melt Channel Model ................................................................................................. 59   ix  KEY TO ABBREVIATIONS ANS: Arabian-Nubian Shield BTSH: Boru-Toru Structural High CMER: Central Main Ethiopian Rift EARS: East African Rift System HFSE: High field strength elements HREE: Heavy rare earth elements KED: Kinetic Electron Dispersion LA-ICP-MS: Laser ablation- inductively coupled plasma mass spectrometer  LILE: Large-ion lithophile element LREE: Light rare earth elements MER: Main Ethiopian Rift NMER: Northern Main Ethiopian Rift PGE: Platinum-group element PPL: Plain-polarized light REE: Rare earth element SCLM: Sub-continental lithospheric mantle SMER: Southern Main Ethiopian Rift XPL: Crossed-polarized light YTVL: Yerer-Tullu Wellel Volcanotectonic Lineament   1  1.  Introduction During the rifting of a continent, the thick continental lithosphere must transition to thinner, chemically distinct, oceanic lithosphere. Our current understanding of this transitional process is dominated by observations of crustal extension, which have shown that thick silicic continental crust is converted into thin mafic oceanic crust through a variety of stretching and intrusion mechanisms (e.g., Courtillot et al., 1999; Buck, 2006; Agostini et al., 2011; Corti, 2012). Comparatively few constraints exist as to how the volumetrically more significant subcontinental lithospheric mantle (SCLM) is deformed and removed during continental rifting (Kay and Kay, 1993; Zeyen et al., 1997; Morency et al., 2002).  There are two dominant endmember processes that facilitate deformation of the lithospheric mantle during rifting: (1) simple shear and ductile stretching (e.g., Wiessel & Karner, 1989; Wilson et al., 2005), and (2) thermo-chemical alteration of the lithospheric mantle (e.g., Bedini et al, 1997; Vauchez et al., 1998).  Mechanical deformation and stretching of the lithospheric mantle is well constrained through the application of various geophysical techniques and associated modeling, which have provided details as to the thickness of the continental lithosphere at various stages of rift development (e.g., Begg et al., 2009; Huismans & Beaumont, 2011). However, the subtle geochemical changes imposed by thermo-chemical deformation of the mantle make it difficult to unambiguously resolve how lithospheric structure may be perturbed during rifting. Although evidence of prior thermo-chemical deformation of the continental lithospheric mantle is preserved in portions of the lithospheric mantle exhumed during late-stage extension (e.g., Lavier & Manatschal, 2006), direct investigation of large sections of the lithospheric mantle is not possible during the rifting process. Xenoliths, small pieces of the SCLM brought to the to the SCLM and can provide insight into the processes that occur in the mantle during rifting (e.g., Bedini et al., 1997; Roger et al., 1999; Ferrando et al, 2007; le Roux et al., 2014). 2  Within the East African Rift, mantle xenolith studies have provided an important record of the establishment, evolution, and recent modification of the SCLM. Re/Os studies have provided constraints on the initial lithospheric age (e.g., Chesley et al., 1999; Reisberg et al., 1999); isotopic and trace element studies have revealed evidence of multi-stage metasomatic enrichment of the mantle by a variety of metasomatic agents (e.g., Rudnick et al., 1993; Bedini et al., 1997; Lee et al., 2000); and petrographic and textural evidence has demonstrated evidence of lithospheric chemical deformation (e.g., Kaeser et al., 2006). Increasingly, there is recognition that lithospheric deformation may promote enhanced metasomatic processes. Simple shear of the lithospheric mantle may create topography along the lithosphere-asthenosphere boundary, promoting focused flow of magmas into the lithosphere and creating pronounced thermo-chemical alteration of the lithospheric mantle (e.g., Havlin et al., 2013). While rift-border faults provide a potential nexus for such lithospheric modification (Corti, 2012), the intersection of a rift with a pre-existing lineament provides a particularly effective avenue for the focusing of magmas into the lithospheric mantle (Rooney et al., 2014a). Here we present petrographic and geochemical data from peridotite xenoliths hosted by ~5 Ma lavas collected along the western portion of the Yerer-Tullu Wellel volcano-tectonic lineament (YTVL), an east-west trending feature adjacent to the Main Ethiopian Rift (MER) [Figure 1], which represents a lithospheric weakness formed during the Pan-African orogeny (Rooney et al., 2014a ). We evaluate the characteristics of the lithospheric mantle in this region and how it has responded to extensional strain and metasomatism by examining xenolith petrography and phase geochemistry. Our work builds upon the earlier studies that explored the thermal and chemical conditions of the upper mantle in this region (e.g., Conticelli et al., 1999; Ferrando et al., 2008). A common conclusion of these studies was that the textural and compositional characteristics of the SCLM was the result of recrystallization due to varying degrees of interaction with the Afar plume (e.g., Bedini et al., 1997; Roger et al., 1999; Ferrando et al, 2008). While the Afar plume is influential, the absence of pervasive alteration of the SCLM has led to the development of speculative models suggesting the eruption of the Ethiopian Flood basalts was facilitated 3  by channelized magma flow in the lithospheric mantle (Roger et al., 1999). We present a model of channelized flow in the SCLM beneath the YTVL developed from observations using peridotite xenoliths. Our results show that focused flow resulted in the formation of depleted lithologies (dunite) by the dissolution of phases such as pyroxene. In xenoliths less affected by melt-rock reaction, the formation of granular texture was accompanied by a depletion of the more incompatible trace elements. The results of this study show that focused flow beneath the YTVL has significantly altered the lithological and geochemical composition of the lithospheric mantle, confirming previous conceptual models of channelized flow of magma through the African lithosphere.    4  2.  Background Many of the geologic processes and phenomena that occur today in East Africa are the result of, or influenced by, the geologic history of the Pan-African orogeny. The Pan-African orogeny consisted of a series of collisions that ultimately resulted in the formation of the supercontinents of Gondwanaland and Pannotia (Cutten, 2002) along the Pan-African Mozambique Belt of East Africa (McWilliams, 1981).  The Arabian-Nubian Shield (ANS) (i.e., the area from Jordan south to Sudan and from western Egypt to eastern Saudi Arabia) was formed by the Pan-African Orogeny from approximately 1,200 to 550 Ma (Stein & Goldsetin, 1996). ANS magmatic history is divided into four phases: Phase I being the erupting of tholeiitic basalts from 900 to 870 Ma, phase II being island-arc calc-alkaline magmatism from 870 to ~650 Ma, phase III being pervasive metamorphism and the formation of granitic batholiths from 640 to 600 Ma, and phase IV being the creation of alkaline granites and dolerites from 600 Ma to ~540 Ma (Bentor, 1985). Pb, Sr, Nd, and O isotopes from the core region of the ANS indicate a lack of pre-established continental crust prior to the above mentioned events (Stein & Goldstein, 1996).  Phase I (900 to 870 Ma), the eruption of tholeiitic basalts and sediments such as volcanogenic clastics and breccias, suggests an initial oceanic setting in which magmas were erupted at a very high rate creating an oceanic plateau. REE patterns of tholeiitic basalts from Phase I are relatively flat and similar to those of known oceanic plateaus (e.g., Ontong Java Pleateau) (Stein & Goldstein, 1996), suggesting a similar origin. Phase II (~870  ~650 Ma), island-arc calc-alkaline magmatism, likely occurred as a result of the arrival of the oceanic plateau at a plate margin generating calc-alkaline andesites and diorite-tonalite-trondhjemites (Bentor, 1985). Phase II calc-alkaline magmas show more enriched LREE and fractionated HREE patterns (Stein & Goldstein, 1996) and it is suggested that the parent endmember for these calc-alkaline magams was a garnet lherzolite with a flat REE pattern (Stern, 1994), not unlike the REE pattern 5  seen in Phase I tholeiitic magmas. Phase II magmas were generated at an anomalously high volume suggesting a thermal anomaly as a result of a rising plume head (Stein & Goldstein, 1996). Phase III (640   600 Ma) pervasive metamorphism and formation of granitic batholiths suggest the initiation of the assembly of the Pan-African continent, where the lithosphere is significantly thickened and metamorphosed and large volumes of calc-alkaline granitoids, gabbros, and diorites were generated as a result of a major collisional event with other ANS terrains and other cratons. REE patterns in Phase III are comparable to other Archean and Proterozoic orogenies (Stein & Goldstein, 1996).  Lastly, Phase IV (600   ~540 Ma), the creation of alkaline granite and dolerites, further suggests the assembly of the Pan-African continent followed by the extension of the lithosphere and crust. Phase IV alkaline granites show a negative Eu anomaly and low Sr content, suggesting plagioclase fractionation of mafic melts, likely due to the presence of normal faulting and volcanoclastic sedimentation (Stein & Goldstein, 1996).  The inherent lithospheric weaknesses resulting from the Pan-African Orogeny have influenced the lithosphere and lithospheric mantle in East Africa, creating optimal zones for shearing (Abebe et al., 1998; Keranen et al., 2008). The YTVL is suggested to be a terrain boundary between a northern and southern section of Precambrian crust, with Oligocene underplating to the north but not in the south. Therefore the YTVL, a structural weakness created though the Pan-African orogeny, may aid in focused melt intrusion into the SCLM. 2.1  East African Rift System The East African Rift System (EARS) is the largest active continental rifting system in the world. It stretches from the Afar triple junction to Mozambique in a north-south fashion [Figure 1]. The EARS is comprised of three main components: the Afar dome, the East African dome, and the Turkana depression. The Afar dome, in the north, is a topographic swell that extends from the eastern border of Sudan to the western edge of Ethiopia. It reaches as far north as Yemen and extends south to the southern border of 6  Ethiopia. The East African dome is the southern topographic swell that extends from the eastern border of the Democratic Republic of the Congo to the eastern border of Kenya and from Kenya to the northern border of Tanzania.  The Turkana Depression separates the two domes. The northernmost branch of the EARS, referred to as the Main Ethiopian Rift (MER) lies between the east-west trending Ethiopian and Somalian plateaus   the two major topographic highs in this region [Figure 2]. The MER is divided into three sub-regions based on stages of continental rifting, magmatism, deformation and structure (Corti, 2009): The Northern Main Ethiopian Rift (NMER), which extends southwestward from the Afar MER boundary, bounds the Central Main Ethiopian Rift (CMER) at the Boru Toru Structural High (BTSH) (Bonini et al., 2005); the CMER continues southwestward and is bounded by the Southern Main Ethiopian Rift (SMER), at approximately 7° N.; the SMER stretches southward to the Turkana depression (~5° N) where it is accommodated by a series of basin and ranges [Figure 2].  2.2  Magmatism in the EARS Since the Eocene, the MER has experienced episodic volcanism and magmatic activity which has influenced the structure and tectonics of the region (Burke, 1996; Wolfenden et al., 2004; Corti et al., 2009). Magmatism in the EARS began at approximately 40-45 Ma in the Turkana depression and the EARS has been volcanically active intermittently since then (Furman et al., 2007). At ~30 Ma, magmatism was dominated by the rapid eruption of the 2000 meter thick flood basalt province centered on Ethiopia. Approximately 350,000 km3 of basalt was extruded covering an area of approximately 600,000 km2 in Ethiopia and Eritrea (Mohr, 1983; Furman et al., 2007) and is associated with the breakup of the Afro-Arabian shield   the event responsible for forming the Red Sea and Gulf of Aden (Wolfenden et al., 2004). From ~29-10 Ma, the EARS experienced intermittent episodic volcanism in Turkana and southern Ethiopia (Furman et. al., 2007). Volcanism also occurred along the Ethiopian Plateau with the eruption of isolated shield volcanoes during this time period (Kieffer et al., 2004). Quaternary volcanic activity along the MER occurs as a result of the oblique faults of the Wonji Fault belt system (Corti et al., 2009). Basalts are extruded along the NNE- SSW trending Wonji Fault belts forming small flows, scoria 7  cones and phreatomagmatic deposits (Abebe et al., 2005). Quaternary volcanic activity occurring on the plateau first developed in the Tana graben due to extension-related tectonics. Extension resulted in the formation of fissure-type lava fields, and small to medium tuff cones, rings and maars which erupted alkali basalts in the plateau region (Ferrando et al, 2008). Throughout its magmatic history, the MER has experienced extension as the continental lithosphere is actively deforming and eventually being converted to oceanic lithosphere.  2.3  Rift evolution Evolution of the MER is an actively debated and unresolved topic due to the complexities of the structural, petrological and geochemical mechanisms involved (e.g. Wofenden et al., 2004; Bonini et al., 2005; Buck et al., 2006; Keranen et al., 2008; Corti et al., 2012). Wolfenden et al. (2004) proposes a northward propagation while alternative hypotheses suggests other mechanisms such as rift initiation via dike emplacement (Buck, 2006; Havlin et al., 2014) and stationary hotspot volcanism underlying a northward moving plate (Rogers, 2006) which suggests a southward propagation. Bonini et al. (2005) suggests a three-part evolution which encompasses a northward propagation in the SMER until 11 Ma followed by a southward propagation in the NMER, and eventually the formation of the CMER post 5-6 Ma. Keranen et al. (2008) argues for a north to south rift propagation starting in Afar, moving southward, stalling at the NMER-CMER boundary for some time, then continuing south through the CMER.  While most of the studies previously mentioned are based on surface structural evidence (Wolfenden et al., 2004; Bonini et al., 2005) and modeling approaches (Buck, 2006; Rogers, 2006), only the study by Keranen et al., (2008) takes lithospheric structure heterogeneities into account when assessing the evolution of the MER. They propose that structural evidence seen at the surface extends into the lithospheric mantle to influence rift propagation. They also indicate that the YTVL [Figure 3] plays an extremely important role in rift evolution by accommodating strain in the form of extensional strain from the NMER. 8  2.4  Yerer-Tullu Wellel Volcanotectonic Lineament  The YTVL, which is located within the Afar dome, stretches 700 km from the western edge of the MER at the NMER-CMER boundary to the Sudan border (Abebe et al., 1998). The YTVL is composed of a Proterozoic-Paleozoic basement with Mesozoic continental and marine sediments and Cenozoic volcanics (Abebe et al., 1998). The northern border of the YTVL is defined partially by the Ambo Fault System, an east-west trending structure that divides the Ethiopian plateau [Figure 2] and the edge of the thick mafic underplate beneath the Ethiopian Volcanic Plateau (Keranen et al., 2008) [Figure 3]. The southern border is less well defined. Abebe et al. (1998) proposed that in the late Miocene (~12Ma) the YTVL accommodated extension of the MER along the Precambrian lithospheric weaknesses associated with the Ambo Fault System. This system is likely a remnant lithospheric weaknesses created during the Pan-African Orogeny (Stein & Goldstein, 1996).  Based on P-wave velocities, Keranen et al. (2008) proposes that the YTVL accommodated extension from the NMER due to the influence of the thick mafic underplate under the Ethiopian Volcanic Plateau in addition to the Ambo Fault System. This east-west trending weakness combined with the lack of the thick mafic underplate seen under the plateau allowed volcanism to occur along the YTVL in several phases: 12-7 Ma further west, 6-2 Ma in the middle, and <1 Ma in the east (Abebe et al., 1998; Keranen et al., 2008).  Keranen et al., (2008) emphasizes the importance of understanding how pre-existing weaknesses or pre-rift structures in the crust and upper mantle influence the propagation of rifting [Figure 3]. While crustal structures can be studied easily along the YTVL, constraints on the behavior of the sub-continental lithospheric mantle during extension are lacking. Using peridotite mantle xenoliths, we can examine the state of the lithospheric mantle during the period of extension accommodation by the YTVL.    9  2.5  How mantle xenoliths record lithospheric mantle processes To test how rift evolution influences the SCLM, mantle xenoliths can be examined to infer geochemical, structural and petrological processes that occur in the SCLM  particularly deformation of the lithospheric mantle and metasomatism (e.g. , Bedini et al., 1997; Roger et al., 1999; Ferrando et al, 2008; le Roux et al., 2014).  The xenoliths studied in Ethiopia exhibit the impact of a mantle plume on the continental lithosphere. On the Ethiopian plateau (Lake Tana Region), xenoliths exhibit evidence of interaction between the continental lithosphere and melts/fluids associated with the Afar plume (Conticelli et al., 1999; Roger et al., 1999; Ferrando et al., 2008; Frezzotti et al., 2010). Roger et al. (1999) conducted a comparative study between Oligocene and Quaternary hosted xenoliths from Lake Tana Province revealing that the lithospheric mantle beneath the Ethiopian plateau has been less affected by plume emplacement than the lithosphere beneath the axis of the rift.  In the Sidamo region of Southern Ethiopia, similar studies investigated the influence of fluid percolation in lithospheric mantle to study thermo-mechanical erosion of the lower lithosphere via plume activity (Bedini et al., 1997; Lorand et al., 2003; Reisburg et al., 2004; Bianchini et al., 2014). These studies examined xenoliths for LILE (large-ion lithophile element) enrichment, platinum group element (PGE) chemistry and Os isotope data to support a model of chromatographic metasomatism associated with an ascending mantle plume (Bedini et al., 1997).  Our study area at Nekempte [Figure 2] on the Ethiopian Plateau has likely been affected by similar plume-lithosphere processes; however, it is also located along a significant Precambrian structural lineament, the YTVL. This study area thus provides an opportunity to examine not only metasomatic enrichment associated with an ascending mantle plume, but the impact of a zone of structural weakness on lithologies within the SCLM. To date, the only study known to provide constraints on xenoliths from this region is the thermobarometric study by Conticelli et al. (1999). That study, while limited in scope, provides a basis on which to build our current study. We expand upon earlier work by firstly, examining a 10  much more significant suite of xenoliths; and secondly, by applying laser ablation ICP-MS techniques to reveal trace element systematics of minerals within the xenoliths.   11  3.  Methods  3.1  Sample selection The mantle xenoliths presented in this study were collected from Nekempte, located about 250 km east of the Western border of the YTVL and 25 km south of the northern border of the YTVL [Figure 2]. These xenoliths erupted with Cenozoic lavas aged at 5.92 ± 0.18 Ma (Abebe et al., 1998). Approximately 150 xenolith samples were recovered from a quarry. Xenoliths were commonly peridotite and pyroxenite (examined in a different study) hosted in phonalite. Peridotite xenoliths were selected for analysis on the basis of freshness. 3.1.1 Analytical Methods - Laser ablation (LA)- ICP-MS Through petrographic examination of the sample suite, four thin sections were selected for detailed geochemical analysis on the basis of how representative the sample was of the freshest samples in the overall population. In-situ laser ablation (LA)- Inductively Coupled Plasma Mass Spectrometer (ICP-MS) analyses was conducted using a Thermoscientific ICAP Q quadrupole ICP-MS at Michigan State University.  The system consists of the ICP-MS and a Photon Machines Analyte G2 193 nm excimer laser ablation system equipped with a 15 x 15 HelEx sample chamber. The laser was set at 4.1 mJ/cm-2 fluence and 10 Hz repetition rate. The sample chamber was placed in a 0.75 L/min high-purity helium carrier gas flux at 1 atm. The laser was paired with the mass spectrometer used in kinetic energy dispersion (KED) mode for molecular interference reduction and ability to measure major elements. It was tuned using NIST612 for best signal intensity on Co, In, Th, and U as well as the lowest oxide production rate on ThO/Th (<0.7%) and lowest double charged cations on Ba2+/ Ba (<2%) production rates. The analysis had a reproducibility of 5 replicated analyses of NIST SRM 612 on 59Co, 115In, 232Th, and 238U (1 rsd <5%).  Calibration was achieved from a set of ~ 20 standards including fused powders and natural and synthetic basalt glass standards [See appendix A]. An internal standard correction was performed using two steps. The first step was to measure MgO for pyroxene and olivine and TiO2 for spinel. Two isotopes 12  were measured: one as an internal standard and one as unknown for concentration correction. An arbitrary concentration was assumed to MgO or TiO2 in the minerals. The second step was to normalize all elements for MgO of TiO2 for the arbitrary value. Next, a sum of major element oxides was calculated as the oxides are likely to not equal 100%. Then, all major and trace elements were normalized to 100% oxides because only anhydrous minerals were analyzed and these will have exactly 100% oxides in them. In addition, the NMNH olivine (San Carlos) + diopside standards were analyzed as unknowns [Appendix A].  All chemical analytical data were collected from minerals grains that were at least one mineral grain distance from the xenolith-host lava contact. Olivine, orthopyroxene and clinopyroxene were typically ablated with a spot size of 110 microns for approximately 30 seconds, of which 20-25 seconds of the best signal was kept for processing. Smaller spot size was used sometimes for smaller crystals to avoid overlap. Olivine, orthopyroxene, clinopyroxene, and spinel were chosen prior to chemical analysis via petrographic analysis. Crystals were chosen carefully ensuring that they were large enough to for multiple ablation spots and did not display evidence of secondary alteration or fluid inclusion trails. Olivine, orthopyroxene, and clinopyroxene were analyzed for all major elements as well as trace elements including: Sc, V, Cr, Co, Ni, Cu, Zn, Ga, Sr, Y, Zr, Nb, Ba, La, Ce, Nd, Sm, Eu, Gd, Dy, Er, Yb, and Hf. Spinels were analyzed for major elements MgO, Al2O3, TiO2, MnO, and FeO as well as trace elements Sc, V, Cr, Co, Ni, Cu, Zn, and Ga. Data processing was performed using Qtegra software for masspectrometry. Replicated analyses of JB1a and BHVO-1 fused powder standards interspersed between samples were run every ~1 hour [Appendix A]. Precision was normally within <5% (1) at concentrations typical of basalt standards. Precision at lower concentrations was estimated from peridotite fused powder standards PCC-1 and DTS- 1 run as unknowns [Appendix B]. Detection limit [Table 1] is defined at 3 times standard deviation (1) on gas blank measurements. Any values less than the detection limit were omitted.  13  4.  Petrography Physical deformation (e.g. kink banding and undulose extinction) and chemical enrichment (e.g. pargasitic amphibole/phlogopite, fluid inclusions) processes in the SCLM may be revealed through petrographic examination of mantle xenoliths. Ferrando et al. (2008) classified the spinel-lherzolite xenoliths from Injibara, Ethiopia [Figure 2] based on their petrography, specifically the amount of recrystallization (which they attributed to thermal activity), and provided a comprehensive analysis of peridotite mantle xenoliths from the Lake Tana region in Ethiopia [Figure 2]. To a first order, we adopt the classification scheme presented by Ferrando et al. (2008), with the presumption that large scale processes that have impacted the Northern Ethiopian Plateau, also affect our study area in the South.  Ferrando et al. (2008) divided their xenoliths into three groups: deformed, granular and transitional; deformed having minimal recrystallization (~ less than 30% of the xenolith), granular having a considerable amount (~more than 30% of the xenolith), and transitional forming a bridge between the two endpoints. The deformed spinel lherzolite xenoliths contain two generations of olivine (Ol I, Ol IID) and orthopyroxene (Opx I, Opx IID), the first being coarse (~4 mm) and the second fine (<1 mm). Clinopyroxene (Cpx I), spinel (Spl I), amphibole crystals, and intragranular fluid inclusion trails also occur.  The granular xenoliths are characterized by having more second generation olivine (Ol IIG), orthopyroxene (Opx IIG), and spinel (Spl IIG). There are fewer first generation crystals such as olivine (Ol I), orthopyroxene (Opx I), and clinopyroxene (Cpx I).  Amphibole is present but sparsely distributed. Transitional xenoliths are intermediate between deformed and granular xenoliths. Ferrando et al., (2008) constrained the P-T conditions of these mantle xenoliths to a pressure range of about 1.3- 2.0 GPa and to have undergone multiple thermal events. In deformed and granular samples, temperatures range from 947  1015° C for the first generation crystals. In the granular xenoliths, the second generation olivine and spinel are re-equilibrated at higher temperature ranges of 1043   1167° C (Ferrando et al., 2008).  14  Samples from Nekempte were analyzed microscopically with a Nikon Eclipse E600 Polarized Light microscope. Images were acquired with an attachable Nikon Digital Sight DS-US camera and NIS Elements F software. Following Ferrando et al. (2008), the samples can be divided into two distinct suites of xenoliths: deformed and granular [Figure 4a-c].  In addition, a cumulate [Figure 4d] and dunite [Figure 4e] xenolith were examined.  4.1  Deformed xenoliths Deformed spinel-bearing xenoliths [Figure 4a-b] are classified based on mineralogy, crystal size, amount of recrystallization, kink banding, and other textures present (e.g., fluid inclusion trails). In general, deformed xenoliths are harzburgite to dunite with large kinked crystals of olivine and orthopyroxene (1mm    8mm). 4.1.1 Mineralogy Deformed xenoliths contain two generations of olivine and orthopyroxene: coarse and fine-grained generations [Figure 5]. Other minerals, such as spinel, occur as smaller anhedral crystals and are not as abundant. Samples are on average 70% olivine, 28% orthopyroxene, 1% clinopyroxene, and <1% spinel. Coarse-grained olivine(Ol I) is characterized by larger (3mm  8mm) sub-rounded, deformed, fractured crystals with undulatory extinction. It very commonly contains iddingsite and talc in its fractures. Fine-grained olivin (Ol IID) is small (<1mm  3mm) rounded, subhedral, undeformed crystals that occur at triple junctions of Ol I and Opx I.  Coarse-grained orthopyroxene(Opx I) is large (3mm  5mm), sub-rounded, deformed crystals that are heavily fractured. Fine exsolution lamellae of Cpx I are widely distributed. Fine-grained (<1mm  3mm) Orthopyroxene (Opx IID) are less common but form in triple junctions with Ol and Opx I. Exsolution lamellae never eccur in Opx IID. Spinel (Spl I) (<1 2mm) is sparsely distributed and are always undeformed, anhedral, s that occur interstitially between Ol I and Opx I.  15  Clinopyroxene is present in very low abundances (~1%) in the deformed xenoliths, and is typically only 1 3mm in size. Amphibole is not present in the deformed harzburgite xenoliths.  4.1.2 Textures Deformed xenoliths exhibit little sign of recrystallization (i.e. small crystals, lack of kink banding). Large olivine and orthopyroxene crystals (3mm  8mm) are surrounded by smaller second generation crystals; however, the second generation crystals are less abundant (only about 30% of the sample). Kink banding occurs  in approximately 80% of the olivine of the deformed xenoliths. Kink banding seldom occurs in second generation olivine crystals. Orthopyroxene rarely display exsolution lamellae in first generation crystals; however, they often show a dissolution texture occurring at the rim of the mineral. Intragranular fluid inclusion trails [Figure 4f] occur in coarse-grained olivine and orthopyroxene. 4.2  Granular xenoliths Granular spinel-bearing harzburgite xenoliths [Figure 4c] resemble deformed harxburgite xenoliths in that they also contain two generations of olivine and orthopyroxene: coarse and fine-grained generations [Figure 5]; However, the fine-grained generation is much more abundant. In general, granular xenoliths are distinguished by having more recrystallization indicated by the large amount of small (<1mm 3mm) olivine and orthopyroxene crystals.  4.2.1 Mineralogy Granular xenoliths are approximately 70 % Olivine, 26% Orthopyroxene, and 3% Spinel. Ol I (>3 mm) is present but less abundant than second generation olivine (Ol IIG) which fill the interstices between the first-generation crystals. Similarly, second-generation orthopyroxene (Opx IIG) is more common than Opx I and fills the interstitial spaces between Ol I and Opx I. There are two generations of spinel in the granular xenoliths: the first 16  (Spl IIG) occurs as small rounded blebs approximately 1 3mm in diameter.  No clinopyroxene or amphibole are present in the granular xenoliths.  4.2.2 Textures The granular xenoliths represent the most recrystallized xenolith from the sample suite; about 80% of the xenoliths contain recrystallized olivine and orthopyroxene. The recrystallized olivine (OlIIG) and orthopyroxene (OpxIIG) fill the interstitial spaces of the first generation crystals of Ol IG and OpxIG and often form triple-junctions with other second generation crystals. The granular xenoliths display little to no kink banding in first generation crystals of olivine and orthopyroxene. Kink banding never occurs in second generation crystals. First generation olivine and orthopyroxene are often fractured and filled with secondary minerals (e.g., Iddingsite, Talc). Second generation crystals have minimal fracturing compared to the first generation crystals. Exsolution lamellae were not observed in the granular xenoliths. Fluid inclusion trails are common in granular xenoliths occurring in both first and second generation crystals.  4.3  Cumulate xenoliths The cumulate xenolith [Figure 4d] is classified based on its classic cumulate texture. It has similar characteristics to the granular xenoliths; however, its abundance of subhedral crystals is what makes it unique from other xenolith types of this suite.  4.3.1 Mineralogy The cumulate xenolith is comprised of approximately 80% olivine, 15% orthopyroxene, 3% clinopyroxene, and 2% spinel. Olivine (OlC) range from <1mm to ~4 mm in size and are euhedral to subhedral crystals. It is often zoned and fractured. Anhedral orthopyroxene (OpxC) are <1mm  ~3mm in size and fill the interstices of the olivine crystals. Clinopyroxene (CpxC) are anhedral, and range from 12mm in size. Brown spinel is anhedral and often rounded and fractured. They range from <1mm  ~3mm in size. Minerals often meet at triple junctions. 17  4.3.2 Textures The sample displays a cumulate texture in which euhedral to subhedral olivine crystals (OlC) have accumulated. Orthopyroxene (OpxC), clinopyroxene (CpxC), and spinel (SplC) fill the interstices of the olivine crystals. The classic cumulate texture indicates a lack of recrystallization or replacive textures. There is no kink-banding in the cumulate xenoliths; however, some olivine crystals are moderately zoned. Clinopyroxene sometimes display weak exsolution lamellae. Mineral dissolution is common along the outer edges of clinopyroxenes throughout the sample. Exsolution lamellae are very sparsely distributed in the cumulate xenoliths. Fluid inclusion trails are very widespread in all mineral assemblages.  4.4  Dunite xenoliths The dunite xenoliths [Figure 4e] display similar characteristics to those of the deformed xenoliths; however, due to their significant difference in mineralogy, they must be classified as their own group.  4.4.1 Mineralogy Dunite xenoliths consist of 98% olivine and 2% spinel. Coarse grained olivine(OlRD)is subrounded to anhedral and range from 1mm to 8mm in size and constitute approximately 80% of the sample. Olivine is often fractured and filled with secondary minerals such as iddingsite. Fine grain olivine is rare, anhedral, and is 1mm or less in size. Spinel(SplRD) are anhed3 mm. No orthopyroxene, clinopyroxene, or amphibole occur  in the dunite xenoliths.   4.4.2 Textures The dunite xenoliths contain very large (up to 8mm) kinked olivine constituting ~80% of the entire xenolith. Limited, if any, recrystallization textures (granularity) are present in the dunite xenoliths. Fluid inclusions are abundant in the dunite xenoliths and often occur in linear trails.    18  5.  Results 5.1  Major, minor, and compatible element chemistry  Chemically, the data from the deformed xenoliths, granular xenoliths, and cumulate xenoliths plot in four distinct groups based on % Fo content in olivine, % En content in orthopyroxene, and Cr# in spinel. Mineral chemistry results are provided in Tables 2-5.  Major and minor element analysis for olivines reveals that the xenoliths plot in four distinct groups based on forsterite ranging from ~ 85% Fo to ~91% Fo (Figure 6). The deformed xenoliths have a forsterite content of approximately 89 91 %. Selected element abundances include CaO  (0.05  0.15 weight percent), TiO2 (0.004 0.021 wt %), Al2O3 (0.016 0.43 wt %), Ni (2500 2900 ppm), Cr (~125~425 ppm), and Co (125 145 ppm). The granular xenoliths have forsterite contents that range from ~88  90 %, CaO (0.03 0.1 wt %), TiO2 (0.025 0.01 wt %), Al2O3 (0.015 0.36 wt %), Ni (~28003100 ppm), Cr (~25 ~125 ppm), and Co (135 155 ppm). The cumulate xenolith has a lower Fo content of 85 86%, higher CaO (0.1 0.3 wt %) and lower Ni (~1900  2600 ppm). TiO2 ranges from 0.025 0.02 wt %, Al2O3 (0.02 0.37 wt %), Cr (~100 ~325ppm) and Co from 135 170 ppm. The dunite xenolith has a forsterite content of approximately 86.5 89 % Fo, CaO (0.049 0.16 wt %), TiO2 (0.00250.026), Al2O3 (0.0140.24 wt %), Ni (2300 2800 ppm), Cr (~60 ~110 ppm) and Co (142 168 ppm).  Orthopyroxene mineral analyses reveal the same clustering based on % En, with ranges from 84% En to 90 % En (Figure 7). Deformed xenoliths have orthopyroxene compositions of approximately 88 90 % enstatite, with ranges of CaO (0.79 1.13 wt %), MnO (0.128 0.139 wt %), Zn (31 43 ppm), Sc (16.8 21.9), Cr (4100 5250 ppm), Co (54 79 ppm), V (79 92 ppm), and Ni (680 820). Granular xenoliths have a pyroxene composition of approximately 8889 % enstatite, with ranges of CaO (0.51 0.74 wt %), MnO (0.133 0.148 wt %), Zn (33 38 ppm), Sc (14.1 17.0 ppm), Cr (1500 2200 ppm), Co (57 68 ppm), V (89 101 ppm), and Ni (720 850 ppm). Cumulate xenoliths have an enstatite content of approximately 84.585.5 % enstatite. CaO ranges from 0.750 0.910 wt %, MnO (0.180 0.189 wt %), 19  Zn (46 54 ppm), Sc (18.5 21.9 ppm), Cr (2100 2450 ppm), Co (59 70 ppm), V (122 133 ppm), and Ni (610 680 ppm). Dunite xenoliths contained no orthopyroxene.  Spinel geochemical data were collected for the granular, cumulate and transitional xenoliths [Figure 8]. The analyzed deformed xenolith did not contain sufficient spinel for analysis. Spinel from the three xenoliths have variable chromium number (Cr/ (Cr+Al). Granular xenolith spinels have a Cr# of 0.271 0.334 with varying FeO (12.1714.70 wt %, MgO (20.00 22.92), TiO2 (0.113 0.147 wt %), MnO (0.088 0.112), Sc (0.197 0.968), V (375489), Co (231263), and Ni (3226 4095). Cumulate xenoliths have a Cr# of 0.387 0.421 and variances in FeO (16.27 19.67 wt %), MgO (17.71 19.50 wt %), TiO2 (0.503 0.824 wt %), MnO (0.113 0.155 wt %), Sc (2.084 3.248 ppm), V (725 849 ppm), Co (189.1 212 ppm), and Ni (2365 2922 ppm). Dunite xenoliths have a Cr# of 0.551 0.588 and variances in FeO (15.73 17.89 wt %), MgO (17.63 23.18 wt %), TiO2 (0.3541 0.7586 wt %), MnO (0.111 0.130 wt %), Sc (1.079 3.230 ppm), V (680 776 ppm), Co (226 250 ppm), and Ni (2175 2744 ppm).  5.2  Incompatible trace element chemistry  Trace elements were analyzed in olivine, orthopyroxene and clinopyroxene. Trace element abundances for olivines were below the detection limits of the LA-ICP-MS for most trace elements; therefore, they are not discussed. Trace elements in orthopyroxene reveal three distinct trace element signatures for the deformed, granular, and cumulate xenolith. Orthopyroxene (normalized to chondrite) (Figure 9a) in the deformed xenoliths is slightly more enriched in the LREE (i.e., La, Ce) and depleted in the HREEs (i.e., Er, Yb) when compared to the other two types of xenoliths. The orthopyroxene in the granular xenolith shows a significant depletion in the LREEs and enrichment in HREEs. The cumulate xenoliths have a similar REE pattern to the granular xenoliths although they are slightly more enriched in all of the REEs. Primitive-mantle normalized (Figure 9b) shows the orthopyroxene in the deformed xenolith as being enriched in Ce and Zr, and depleted in Y and Er compared to the other types of xenoliths. Zr/HfPM ratios (primitive-mantle normalized) of orthopyroxene in the deformed xenoliths range 20  from ~40 to 58, while orthopyroxene in all other xenoliths have Zr/HfPM ratios less than the chondritic value of approximately 34.2 (Weyer et al., 2003). Orthopyroxene in the granular and cumulate xenoliths are enriched in Dy, Y, and Er compared to the deformed xenoliths.  Clinopyroxene trace element data was collected from both the deformed and cumulate xenolith. Clinopyroxene was absent in the granular and transitional xenoliths. Clinopyroxene trace element analyses (normalized to chondrite) (Figure 10a), the cumulate xenoliths have a relatively flat REE pattern ranging from approximately ~1020 ppm for all elements. The clinopyroxene in the deformed xenoliths are enriched in LREEs (i.e. La, Ce), though La and Ce behave differently, and are depleted in the HREEs (i.e. Dy, Er, and Yb).   21  6.  Discussion  6.1  Location of samples within the SCLM The SCLM records the complex interaction between pre-existing depleted peridotite and percolating melts/fluids over the lifetime of the lithosphere (e.g., Bodinier et al., 1990). These interactions result in significant modal and chemical heterogeneity that complicate the interpretation of xenolith suites. There is a growing consensus that chromatographic processes, which may result in the geochemical (Bedini et al., 1997) and modal (Pilet et al., 2011) stratification of the SCLM, play an important role in defining the petrographic and geochemical structure of the SCLM. At depth, metasomatic processes may generate pyroxenites within the SCLM; at mid-lithospheric depths, amphibole-rich lithologies are common; while at shallower lithospheric depths, widely disseminated metasomatic phases are evident (e.g., Pilet et al., 2011). Consequently, the interpretation of the geochemical characteristics of mantle xenoliths derived from the SCLM requires constraints as to their depth of formation/equilibration.   Where an aluminum-rich phase is present, the mantle xenolith suite from Nekempte exhibits spinel and lacks garnet, thereby limiting the depth of equilibration of these xenoliths to the spinel peridotite stability range. To more accurately quantify the depth of equilibration, we have applied the two pyroxene geothermobarometer of Putirka (2008). To ensure equilibrium between co-existing pyroxenes we only selected pyroxene pairs that have a KD (Fe-Mg) value of between 0.95 and 1.23. Results of our thermo-barometric calculation show that the pressures and temperatures of equilibration are within a limited range between 1055 and 1114 °C, and 0.99 to 1.08 GPa [Table 6]. These pressure estimates suggest that the samples are derived from the shallowest portion of the lithospheric mantle, located just below the Moho [Figure 11].   Previous thermo-barometric studies of samples from Nekempte are limited (Conticelli et al., 1999), and are based on data from a single harzburgite sample (1.13 GPa, 1045oC). The value is consistent with the results of the present study, and when combined with other xenolith localities in Ethiopia (Conticelli 22  et al., 1999; Roger et al., 1999; Rooney et al., 2005; Ferrando et al., 2007), requires a geotherm in the region that is significantly greater than the 40mWm-2 value expected for a steady state continental shield (e.g., Anderson, 2007). Previous studies have suggested that this elevated geotherm is the result of the Ethiopian SCLM undergoing recrystallization resulting from the thermal contributions from mantle plume(s) (e.g., Conticelli et al., 1999). The majority of regional xenolith samples support such a model; however, granular xenoliths from Injibara exhibit equilibration temperatures ~100oC higher than deformed xenoliths at nominally similar pressures (Ferrando et al., 2007). These observations of significant thermal heterogeneity within a single region prompted Ferrando et al. (2007) to suggest a model whereby infiltration of melt associated with the Afar plume has perturbed the geothermal gradient of the Ethiopian SCLM. Maximum temperatures in the Nekempte xenoliths are observed in the cumulate xenoliths rather than the deformed xenoliths [Figure 11]; however, no thermobarometry calculations were performed for the granular xenoliths due to their inherent lack of clinopyroxene. A processed-based comparison is difficult due to our limited data; however, it is apparent that both the deformed and cumulate xenoliths measured from Nekempte have equilibrated at a temperature of about 100°C hotter than most of the xenoliths observed in other regional studies, but are equivalent to the temperatures observed in granular xenoliths from Nekempte (Conticelli etal., 1999) suggesting a similar thermal history.  6.2  Textural evidence of deformation Suture zones and major structural lineaments are locales whereby shearing observed within the continental crust is also likely reflected in deformation of the SCLM. Along the Ethiopian rift, the YTVL, a Precambrian suture zone that reactivated and likely accommodated extension in the MER during the Cenozoic (Abebe et al., 1999; Keranen et al., 2008), is one of the most significant zones of lithospheric weakness in the region. Previous studies have shown the YTVL to be zone of significant crustal thinning (Keranen et al., 2008) and a region of focused Cenozoic volcanism (Abebe et al. 1999). Xenoliths from 23  Nekempte, which lies along the YTVL, thus have the potential to record deformation associated with the shearing of the lithosphere along this lineament.    The xenoliths analyzed from Nekempte exhibit evidence of dynamic recrystallization (e.g., Roger et al., 1999; Ferrando et al., 2008; Frezzotti et al., 2010). This process occurs where stress is applied to the  identified petrographically in deformed and granular xenoliths from Nekempte with the granular having undergone more dynamic recrystallization than deformed. These observations are similar to those of xenoliths from Injibara, described and classified by Ferrando et al., 2008. Ferrando and coworkers classified their xenoliths from Injibara into three groups: deformed, granular, and transitional. Deformed xenoliths are petrographically characterized by their larger (>3mm) crystal size and stress banding. Deformed xenoliths contain two generations of olivine and orthopyroxene, and one generation of clinopyroxene and spinel. Their overall larger grain size indicates a lack of significant recrystallization. Granular xenoliths are characterized by their smaller (<13mm) crystals and minimal stress banding. The granular samples contain two generations of olivine, orthopyroxene, clinopyroxene, and spinel. Transitional were described as samples having characteristics of both the deformed and granular types.  Following the classification scheme from Ferrando et al. (2008), the sample suite from Nekempte, contains both deformed and granular xenoliths [Figure 4a-c]. Deformed xenoliths contains two generations of olivine (>3mm), and orthopyroxene (>3mm), and one generation of spinel. The other deformed xenolith appears to be a dunite having 100% olivine and spinel. No clinopyroxene was present in the deformed xenoliths. The granular xenolith has two generations of olivine (<13mm), and orthopyroxene (<13mm), and one generation of sparsely distributed clinopyroxene and spinel. This sample suite presents some similar characteristics to the suite from Injibara; however, some differences must be noted. No transitional xenoliths were present in the Nekempte suite, and the fourth xenolith observed from Nekempte had a cumulate texture, indicated by euhedral to subhedral olivines with olivine 24  and orthopyroxene filling the interstices. The peridotite xenoliths from Injibara did not exhibit cumulate textures and consisted of only lherzolite and hartzburgite xenoliths; no dunite compositions were noted (Ferrando et al., 2008). Previous work has postulated that the level of recrystallization was positively correlated with the amount of interaction with plume derived melts (e.g., Roger et al, 1999; Ferrando et al., 2008; Frezzotti et al, 2010).  6.3  Evidence of metasomatic agents The textural and thermobarometric evidence presented above suggest that the dominant mechanism of deformation within the Ethiopian SCLM is melt/fluid-rock interaction. The composition of these metasomatic agents remains less well constrained and requires a geochemical investigation of the phases within the xenoliths. 6.3.1 Evidence of hydrous metasomatism Prior studies have suggested that a Cl-rich H2O  CO2 fluid was a pervasive metasomatic agent within the regional lithospheric mantle (Frezzotti et al., 2010). This conclusion is based upon data collected from amphibole and fluid inclusions in spinel lherzolites, which indicated that this modal metasomatic event occurred at around 1000° C, and interaction between CO2 and H2O - rich fluids and the deformed xenoliths may have caused the cryptic enrichment of Fe and Al in pyroxenes in addition to the growth of new amphibole (Frezzotti etal., 2010).  The origin of the suggested hydrous metasomatism event was be hypothesized to be associated with the earliest stages of mantle upwelling accompanying a rising plume (Frezzotti et al., 2010). The heterogeneity of the lithospheric mantle in East Africa suggests that the hydrous fluids generating the metasomatism percolated through the region via fracture migration, explaining the variation in both volatiles and LREE observed in the lithospheric mantle. Enrichment in Cl and water in mantle xenoliths is indicative of interaction with a recycled crust source (Frezzotti et al., 2010). This is also in agreement with enrichments in Pb, Ba, Th, U, and Sr within the hydrous metasomatic fluid phases (Frezzotti, et al., 25  2010). These data suggest a source that was either rich in marine sediments contained in a subducting oceanic slab (a remnant of the Pan-African Orogeny) or a result of decarbonation or outgassing of hydro-saline melts at depth (Frezzotti et al., 2010). The origin of this hydrous metasomatism event remains uncertain, as isotopic evidence also points to the formation of amphibole cumulates at mid-lithosphere depths associated with the initial formation of the lithosphere during the Pan-African subduction and orogenic event (Rooney et al., 2014a).  While regional evidence suggests that the lithosphere may have been broadly affected by this hydrous metasomatism process, xenoliths at Nekempte exhibit no evidence of hydrous modal metasomatism, requiring a different metasomatic agent. 6.3.2 Evidence of carbonatite metasomatism The geochemical alteration of the continental lithospheric mantle by small-volume carbonate fluids has been noted by previous investigations of the SCLM (e.g., Rudnick et al., 1993); however, this work was undertaken in the southern East African Rift, where thicker continental lithosphere is predominant and carbonatite eruptions have been noted (e.g., Bell & Tilton, 2001). Within the northern East African Rift, carbonatites are absent and prior investigation of xenoliths has not suggested the influence of carbonatitric metasomatism.  However, the absence of this signature from previous xenolith studies in the northern East African Rift might instead reflect an analytical bias towards in-situ analysis of clinopyroxene over orthopyroxene. Due to lower concentrations in incompatible elements, it is analytically challenging to measure trace element abundances within mantle orthopyroxenes, however, newer instrumentation now allows such analyses. In circumstances where complex modal metasomatism is apparent, different xenolith phases (such as opx) may reveal previously undetected complexity.  Orthopyroxene within the deformed xenoliths from Nekempte exhibit unusual trace element values, which may be characteristic of carbonatitic metasomatism [Figure 9a-b].  Elevated values of Zr/Hf in xenoliths and mantle-derived melts have been frequently linked with the influence of carbonatite metasomatism (e.g., Dupuy et al., 1992; Rudnick et al., 1993; Yaxley et al., 1998).  Zr/ Hf ratios of a 26  carbonatitic metasomatic agent are thought to range from ~ 60 80 (Ionov et al., 1993), elevated significantly over the chondritic value of ~34.2 (Weyer et al., 2003). Notably, orthopyroxenes from deformed xenoliths at Nekempte exhibit higher Zr/Hf ratios (4058) when compared to the granular and other xenolith types in the suite (~2045), and are suggestive of a carbonatite metasomatism event. Elsewhere in East Africa, carbonatite metasomatism is has been suggested in peridotite xenoliths from the Omari Cinder Cone (Tanzania), on the basis of a very high Zr/Hf ratio and significant LREE enrichment (Rudnick et al., 1993). The orthopyroxenes from the deformed Nekempte xenoliths exhibit a significant enrichment in LREE in comparison to the granular varieties [Figure 9a-b], consistent with a carbonatitic metasomatism model.  Similarly, clinopyroxene trace element data normalized to primitive mantle [Figure 10a-b] show elevated Zr/Hf ratios (~38  39) [Figure 12] for the deformed xenoliths compared to the cumulate xenoliths (Zr/Hf= ~29  35) further indicating a possible metasomatic enrichment that is carbonatitic in origin. Chazot et al. (1994) observed similar characteristic in spinel-bearing lherzolites from Yemen and indicated that the strong partitioning of Zr and Hf in amphibole and clinopyroxene in their xenoliths may indicate that their xenoliths recrystallized from a carbonatitic melt that interacted with an anhydrous lherzolite or harzburgite mantle. 6.3.3 Evidence of melt-lithosphere interaction Previous studies have suggested that much of the deformation and rerystallization evident in xenoliths derived from the Ethiopian SCLM result from the interaction of silicate melts with the host peridotite (e.g., Roger et al., 1999; Ferrando et al., 2008). The model of Ferrando et al., (2008) suggests a two stage event whereby an initial hydrous metasomatic agent altered the SCLM and formed hydrous phases. The second metasomatic event in this model is caused by a melt which is 100°C warmer than the prior metasomatic agent. This second event promoted recrystallization of second generation crystals of olivine and pyroxenes and increase in Fe and Al, and decrease in Ni, Cr, and Cl. Similar chemistry in the first and 27  second generations of crystals in the granular xenoliths indicate nearly complete re-equilibration during the second cryptic metasomatic event. It is suggested that the physical and chemical variations in these xenoliths may have due to the localization of melts, the deformed xenoliths having been affected by a more uniform pervasive metasomatic event and the granular xenoliths having been impacted by local infiltration of melt.  An alternate model is presented by Bedini et al. (1997), who suggest that, rather than a dual-stage metasomatic event, a single-stage event that induced varying degrees of melt can explain both the physical and chemical variations in the xenoliths of the East African SCLM (in particular at Megga). This model suggests that the deformed xenoliths have been metasomatized with small melt fractions enriched in LILEs and depleted in HFSEs and the granular xenoliths have been metasomatized with large basaltic melt fractions. The metasomatic protolith, the SCLM prior to metasomatism, would have been heterogeneous to account for the mineralogical variations in the deformed xenoliths (i.e., both lherzolites and harzburgites are observed); however, deformed xenolith clinopyroxene trace element chemistry differs significantly, which Bedini et al. (1997)  attributed to chromatographic effects connected to porous flow of small melt fractions where harzburgite will reach equilibrium much faster than lherzolite. In contrast, granular xenoliths are proposed to have interacted with large melt fractions of basaltic melts in order to explain the observed depletion in the LREEs. Granular xenoliths range from harzburgite to cpx-rich lherzolites, yet they still have a relatively uniform chemical composition. This signature is attributed to complete re-equilibration with a LREE-depleted melt  such as a basalt, with large melt/rock ratios associated with melt percolation at a higher porosity and/or over an extended period of time. Because the granular xenoliths contain relic porphyroclasts, it is suggested that the granular xenoliths may have formed at the expense of the deformed xenoliths. This is further indicated by the small variation in mineral chemistry of the granular xenoliths compared to the more heterogeneous mineral chemistry of the deformed facies.  28  Xenoliths from Nekempte exhibit similar features to those of Injibara and Megga. At all three localities, deformed xenoliths have been enriched by a metasomatic agent, although the composition of this agent differs from locality to locality: a Cl-rich H2O- CO2 fluid at Injibara; a LILE enriched/HFSE depleted small degree melt at Megga; and a LILE and HFSE enriched carbonatitic fluid at Nekempte. In contrast, the granular xenoliths at all three localities appear to be derived through the same process of significant melt-rock reaction. This process results in the strongly recrystallized granular xenoliths being depleted in the highly incompatible elements attributed to extensive re-equilibration with large volumes of basaltic melt. We examine this process in the context of melt focusing.  6.4  Evidence of focused melt transport Reactivated structural lineaments such as the YTVL exhibit characteristics that facilitate the transport of melt from the asthenosphere into the lithosphere. It is suggested that the YTVL has aided in focused magmatic intrusion as evidenced by the extensive magmatism along its length over the past ~15 Ma (Abebe et al., 1998). In particular, shear pathways and topography of the lithosphere-asthenosphere boundary (Havlin et al., 2013; Rooney et al., 2014a) may result in the focusing of magma along structural lineaments. Such focused flow results in surface eruptions (e.g., Rooney et al., 2014a), but its passage through the lithosphere may have profound impacts on the SCLM where it may promote metasomatic alteration and recrystallization.  Havlin et al. (2013) indicates that shearing may influence diking and melt intrusion into the lithosphere by increasing porosity within the SCLM. Shear-induced porosity can result in porous zones or channels, which can serve as pathways for percolating magmas and fluids to interact with the SCLM, modifying its composition and other physical properties (Tomassi et al., 2004; Le Roux et al., 2007; Tomassi et al., 200 are focused in ductile shear zones (deformed lithospheric mantle) (Kelemen et al., 1993; Whitehead & Kelemen, 1994) and focused flow in these shear 29  zones facilitates reactions dissolving clinopyroxene and precipitating olivine, ultimately resulting in a zone of precipitated dunite (Kelemen et al., 1992).  The xenoliths examined from Nekempte have a range in textural and geochemical characteristics that help constrain the model of focused melt flow and how it may operate in the Ethiopian SCLM. Based on both petrographic and chemical evidence, our current model incorporates both the textural characteristics and chemical characteristics of all four types of xenoliths: deformed, granular, cumulate, and replacement dunite, into a system that demonstrates the effects of a percolating fluid in a zone of ductile shearing [Figure 13]. Each type of xenolith plays a crucial role in explaining the dynamic system along the YTVL. 6.4.1 Pre-existing subcontinental lithospheric mantle: Deformed xenoliths In our model, the deformed xenolith represents the initial state of the SCLM beneath the YTVL prior to being subjected to focused melt flow. Deformed xenoliths are characterized by their overall larger grain size (> 3mm) and stress banding. The large grain size along with stress banding suggests that they have undergone little recrystallization and therefore have been influenced less by stress or melt interaction. Additionally, deformed xenoliths have the highest Mg content (Fo= ~8991, En= ~8990) [Figure 6,7] suggesting perhaps a more depleted origin. High Zr/Hf ratios are observed in the deformed xenoliths, which we have suggested reflects carbonatite metasomatism (Rudnick et al., 1993; Weyer et al., 2003) [Figure 12]. This is not seen in the granular xenoliths, suggesting a later overprinting event that may have erased the earlier carbonatitic signature. These physical and chemical attributes combined suggest that the deformed xenoliths have undergone minimal chemical interaction with large volumes of melt and  6.4.2 Overprinted lithospheric mantle: Granular xenoliths The granular xenolith in our model represents overprinted SCLM that was once similar in texture and chemistry to the deformed xenolith. Texturally, granular xenoliths are characterized by having overall smaller (<13mm) crystals and minimal to no stress banding suggesting recrystallization (Bedini et al., 30  1997; Ferrando et al., 2008; Frezzotti et al., 2010; Roger et al., 2010). Recrystallization is further indicated by the decrease in magnesium content (Fo= ~8789; En= ~8889) (Ferrando et al., 2008) and depletion in LREE in orthopyroxene trace element chemistry[Figure 9, 10] which suggests a chemical interaction with large volumes of melt (Bedini et al., 1997). The granular xenolith was formed at the expense of the deformed xenolith lithology after being exposed to melt/ fluids generated from a porous melt channel adjacent to the xenolith.  6.4.3 Evidence of pervasive melt-lithosphere interaction: Replacement dunite xenoliths The dunite xenolith discovered within the Nekempte suite is characterized by large olivine crystals (>3mm), the absence of pyroxene, and abundant spinel. Our initial hypothesis for the origin of this lithology was melt depletion of a peridotite during the initial stabilization of the regional lithosphere. On that basis, the dunite xenoliths would simply have experienced a greater degree of melt extraction in comparison to the deformed harzburgites. However, the geochemical characteristics of crystals within the xenoliths preclude this mode of origin. A melt-depletion model for the origin of this sample would result in residual olivine with elevated Fo content; however, our observations show that rather than being more enriched in Fo in comparison to the deformed harzburgite, the magnesium content was actually lower (Fo ~8689). These data, when combined with low Ni content in olivine (~2300   2800 ppm) and high values of TiO2 in spinel (~0.34  0.75 wt%), and a lack of pyroxene, instead suggest this lithology may have formed by a replacement process (e.g., Suhr, 1999).  With Nekempte lying along a zone of lithospheric weakness, one must explore how deformation of the SCLM can affect the ease with which melt can influence the surrounding SCLM during melt localization and focusing. When a melt perturbs the shallow lithospheric mantle, the ascending melts will dissolve the pyroxene and precipitate olivine, forming zones or channe surrounded by harzburgite mantle (Kelemen & Dick, 1995). Channelized flow of melts in the SCLM can generate extreme variations in the more incompatible elements, even on small scales (1  100m) (Speigelman & 31  Kelemen, 2003). This is consistent with data collected from orthopyroxene and clinopyroxene which both show extreme variation in the more incompatible trace elements [Figure 9, 10]. The presence of replacement dunite indicates significant melt-rock reaction has taken place within the SCLM (Spiegelman & Kelemen, 2003). The formation of replacement dunites, which lack the pyroxene phases, is useful for tracing the origin of other xenoliths at Nekempte, but can be challenging to adequately constrain. However, spinel provides some additional information and, in particular, Cr# in spinel can assist in further solidifying the case for focused melt transport. The dunite replacement xenoliths have an extremely high Cr# of ~0.6, much higher that the granular xenoliths (~0.3) and the cumulate xenoliths (~0.4) [Figure 8]. High Cr# in spinel of peridotites typically indicates a higher degree of partial melting of the peridotite (Hellebrand, 2001); however, because the replacement dunite was created at the expense of the orthopyroxene and clinopyroxene, it is plausible that the spinel was also created at the expense of the other minerals (i.e. olivine, pyroxene). High Cr# is observed in ophiolites at contacts between peridotites and pyroxenites and is suggested to be a consequence of dissolution of pyroxenes into a Cr-saturated magma (Bédard & Hébert, 1998). In other words, an exchange in Al and Cr can occur between pyroxene and spinel that will result in an increase in Cr# in spinel (Wilson, 1982). We suggest the high Cr# in the spinel of the replacement dunite xenoliths occurs as a result of the active dissolution of pyroxene during the chemical exchange between the melt channel and surrounding SCLM.  6.4.4 Remnants of the metasomatic agent: Cumulate xenoliths  The final type of xenolith presented in this study is the cumulates, which we suggest represent the remnant of the melt channel represented in the model. Texturally, these xenoliths are characterized by typical cumulate texture, which consists of subhedral olivine and orthopyroxene (<1  3mm) crystals with interstitial anhedral pyroxenes. These xenoliths have the lowest magnesium content (Fo= ~8486;  En= ~85), more consistent with a melt composition. Additionally, igneous olivines have higher and more variable calcium concentrations and lower nickel concentrations (<2200 ppm) due to fractionation (Foley et al., 2013). Calcium concentrations in the olivine of the cumulate xenolith range from ~0.10.25 wt % 32  for CaO [Figure 6], much higher and more variable than the other xenoliths, and nickel concentrations for olivine in the cumulate xenolith are as low as 1900 ppm indicating its cumulate nature.  As melt percolates or flows through the lithosphere, it will fractionate as it slowly cools. As the melt cools, crystals will form and accumulate at the bottom of a magma body, forming a cumulate. The presence of a cumulate xenolith in the SCLM of Nekempte provides further evidence for a need for a model of channelized focused melt flow because a cumulate indicates a remnant of the melt channel itself where igneous processes are actively occurring. The dynamism and heterogeneity of the East-African SCLM has often been explained in a more simplistic fashion involving models which are primarily driven by wide-spread metasomatism (e.g., Conticelli et al., 1999; Ferrando et al., 2008); however, speculation on channelized flow exists in current literature (Roger et al, 1999). It is suggested by some that these heterogeneities are primarily due to mechanisms and processes that are associated with a rising mantle plume (Bedini et al., 1997; Frezzotti et al., 2010) while other suggest theses heterogeneities of the SCLM are primarily driven by a Pan-African event (e.g. Kelemen et al, 1992; Havlin et al., 2013; Rooney et al, 2014a). While it is unclear which of these two mechanisms (if not both) have influenced the SCLM, it is apparent that a significant melt transport event has occurred along the YTVL. The origin of the unique lithologies of the YTVL may be plume-related, a result of reactivation of the YTVL, or a combination of the two. Isotopic data are needed to constrain the origin(s) of the large melt-fractions that strongly influence the physical and chemical properties of the SCLM.   33  7.  Conclusions The SCLM beneath the YTVL exhibits evidence of focused melt-lithosphere interaction. A close examination of the variety of xenoliths present at Nekempte demonstrate that melt beneath the YTVL (a reactivated Precambrian lineament) is strongly channelized. The xenoliths preserved within the Nekempte Suite reveal heterogeneous textures that range from deformed and granular lithologies through to replacement dunites and cumulates.  Similar to other xenolith suites examined in the region, Nekempte exhibits both granular and deformed xenolith varieties. However, we propose a new model that reconciles the presence of other xenolith lithological varieties that are difficult to explain by previously presented mechanisms. We suggest that pervasive melt/rock reaction (also noted elsewhere within the Ethiopian plateau) was initiated by the interaction of basaltic melt with the SCLM resulting in wide-scale recrystallization and the formation of granular xenoliths. The depletion in LREE in orthopyroxene and the absence of clinopyroxene suggests that the mechanism of melt/rock reaction involved substantial flux of melt in order to scavenge the LREEs and destabilize clinopyroxene in the granular xenoliths. As this melt/rock reaction proceeded to completion, dunite xenoliths were formed at the expense of pyroxene. These replacement dunites are characterized by olivine with lower Mg# and spinel with very high Cr# (~0.6). Cumulate xenoliths evident in the Nekempte suite are hypothesized to represent fragments of metasomatizing basaltic melt.  We conclude that shearing associated with the YTVL has influenced the lithospheric mantle such that melt channeling is preferred over chromatographic metasomatism. This conclusion is consistent with previous studies suggesting an enhanced magma supply along the YTVL (Abebe et al., 1998).  Conceptual numerical modeling and observations from the YTVL show that shear-induced porosity and melt focusing along steep topography of the lithosphere-asthenosphere boundary contribute to focused magma supply in the YTVL (Havlin et al., 2013; Rooney et al., 2014a). While previous studies have 34  shown that the crust along the YTVL is deformed during rifting (Keranen et al., 2008), our data suggests that deformation in the form of chemical alteration also extends into the lithospheric mantle. The mechanisms of deformation along structural lineaments is thus, in part, due to the influence of focused magmatic flow through the lithospheric mantle.   35               APPENDICES 36  APPENDIX A TABLES           lmc45  Table 6 Regional thermobarometry  Thermobarometry data from this study and other regional data from Dedessa, Megado/Dilo, Injibara, West Injibara, Kishb, Mer-awi, and Megga locations. Data from this study were calculated with the two-pyroxene method from Purtika, (2008). Standard error for temperature and pressure are 58° C and 2.8 kbar, respectively.   46  APPENDIX B FIGURES 47  Figure 1 East African Rift System (EARS) Location  Location of the East African Rift System (EARS).   48  Figure 2 Main Ethiopian Rift (MER) Location  Location of the Main Ethiopian Rift (MER) and regional xenoliths studies from Injibara,   49  Figure 3 Lithospheric Modification of the YTVL   Modified from Keranen & Klemperer (2008). Figure depicts crustal modification along the YTVL with inferred modification extending into the SCLM. Figure not to scale.   50  Figure 4 Xenolith Petrography  (a,b) Deformed xenoliths in XPL shows two generations of crystals; (c) Granular xenolith in XPL shows more recrystallization; (d) Cumulate xenolith in XPL consists of small euhedral to subhedral crystals. Note: triple junctions between crystals; (e) Dunite replacement xenoliths in XPL show large olivines; (f) fluid inclusion trail in a deformed xenoliths in PPL.   51  Figure 5 Comparison of Xenoliths  Comparison of deformed versus granular xenoliths. Example of granular xenoliths from Nekempte compared to deformed xenolith from Injibara, Ethiopia. Note that the granular xenoliths from Nekempte contain more second generation crystals indicating recrystallization.    52   Olivine major and minor element chemistry for TiO2, CaO2, Sc, Cr, and Ni versus % forsterite (% Fo).  53   Orthopyroxene major and minor element chemistry for CaO2, Cr, and Ni versus % enstatite (% En). 54   (a,b) Spinel major and minor element chemistry for TiO2 and Ni versus Cr# [Cr/(Cr+Al)] in spinel(% Fo); (c) Cr# [Cr/(Cr+Al)] in spinel versus % forsterite (% Fo) in olivine.  55   (a) Chondrite normalized trace element abundances in orthopyroxene for the deformed, granular, and cumulate xenoliths. (b) Primitive mantle (PM) normalized trace element chemistry in orthopyroxene for the deformed, granular, and cumulate xenoliths. Chondrite and primitive mantle values taken from Sun & McDonough, (1989). 56   (a) Chondrite normalized trace element chemistry in clinopyroxene for the deformed and cumulate xenolith. (b) Primitive mantle (PM) normalized trace element chemistry in clinopyroxene for the deformed and cumulate xenolith. Chondite and primitive mantle values taken from Sun & McDonough, 1989.    57  Figure 11 Xenolith Thermobarometry  P T thermobarometry data from this study and other regional data from deformed, granular, transitional, and undifferentiated xenoliths at Dedessa, Megado/Dilo, Injibara, West Injibara, Kishb, Mer-awi, and Megga locations. Thermobarometry data from this study were calculated with the two-pyroxene method from Putirka, (2008) and are shown in Table 6 Regional Thermobarometry. Standard error for temperature and pressure are 58° C and 2.8 kbar, respectively.   58   (a) Zr/Hf versus % Enstatite (% En) in orthopyroxene; and (b) Zr/Hf versus % wollastonite (% Wo) in orthopyroxene. In both cases, the deformed xenoliths display higher Zr/Hf ratio compared to other types.     59  Figure 13 Focused Melt Channel Model  Focused melt channel model shows the relative position of each xenolith type in relation to the focused melt channel. Chemical exchange is allowed in the granular and replacement dunite xenoliths while the deformed xenoliths remains un-reactive with the melt channel. The cumulate xenolith represents a remnant of the melt channel as it cools and crystallizes. 60  APPENDIX C CALIBRATION STANDARDS   62  APPENDIX D STANDARD DEVIATION OF SELECT CALIBRATION STANDARDS   64    .          REFERENCES    65  REFERENCES Abebe, T., Mazzarini, F., & Manetti, P. (1998). "The Yerer-Tullu Wellel Volcanotectonic Lineament: a transtensional structure  in central Ethiopia and the associated magmatic activity." Journal of African Earth Sciences 26: 135-150.  Yerer- Journal of African Earth Sciences. 99: 374-385.  AgoEthiopian Rift and comparison with experimental models: Implications for rift evolution and Nubia-Earth and Planetary Science Letters. 301. 479-492.  Anderson, D. L. (2007). New theory of the Earth. Cambridge University Press.  origins of mantle Zr/ Hf-Contrib. Mineral. Petrol. 162: 1047-1062.  Bailey, D. K. (1982). "Mantle metasomatism: continuing chemical change within the Earth." Nature 296: 525-530.  Beccaluva, L., Bianchini, G., Natali, C., Siena, F. (2009). "Continental flood basalts and mantle plumes: a case study of the Northern Ethiopian Plateau." Journal of Petrology 50 (7): 1377-1403.  gabbros into peridotitic intrusions: North Arm Mountain massif, Bay of Islands ophiolite, Newfoundland, Journal of Geophysical Research 103, B3: 5165-5184.  Bedini, R. M., Bodinier, J. -L., Dautria, J. -M., Morten, L. (1997). "Evolution of LILE-enriched small melt fractions in the lithospheric mantle: a case study from the East Africa Rift." Earth and Planetary Science Letters 153: 67-83.  Bedini, R. M., Bodinier, J. -L. (1999). "Distribution of incompatible trace elements between the constituents of spinel peridotite xenoliths: ICP-MS data from the East African Rift." Geochimica et Cosmochimica Acta 63(22): 3883- 3900.   Africa: Seismic tomography, mantle Geosphere, 5 (1): 23-50.  Journal of Petrology, 42(10), 1927-1945.  -Nubian Massif with special reference to the Sinai Precambrian Research 28: 1-74.  Bianchini, G., Beccaluva, L., Bonadiman, C., Nowell, G., Pearson, G., Siena, F., Wilson, M. (2007). -lherzolite mantle xenoliths from the Iberian plate (Olot, NE Spain): Implications Lithos. 94, 25-45. 66   Bianchini, G., Bryce, J.C., Blichert-Chemical Geology. 386. 49-58.  Bodinier, J. L., Vasseur, G., VerniJournal of Petrology 31(3) 597-628.  tle transition- REE and Sr- Journal of Petrology Special Lherzolites Issue: 191-210.  of the Main Ethiopian Rift  in the frame of Afar and Kenya rifts propagation. Tectonics. 24: TC1007.  Journal of African Earth Sciences 43: 334-378.   of magma in the development of the Afro-arabian Rift System. In: Yirgu, G., Ebinger, C. J., Maguire, P. K. H. (Eds.), The Afar Volcanic Province within the East African Rift System. Geol. Soc. Spec. Publ., 259. 43-54.  Burke, K. (1996). "The African plate." South African Journal of Geology 99(4): 341-409.  -Os systematic of mantle xenoliths from the East Geochimica et Cosmochimica Acta, 63 (7): 1203-1217.  Chorowicz, J. (2005). "The East African rift system." Journal of African Earth Sciences 43: 379-410.  -Journal of Petrology 51 (9): 1913-1940.  Contecelli, S., Sintoni, M. F., Abebe, T., Mazzarini, F., & Manetti, P. (1999). "Petrology and geochemistry of ultramafic xenoliths and host lavas from the Ethiopian volcanic province: an insight into the upper mantle under Eastern Africa." Acta Vulcanologica 11: 143-159.  Corti, G. (2009). "Continental rift evolution: From rift initiation to incipient break-up in the Main Ethiopian Rift, East Africa." Earth-Science Reviews 96: 1-53.  Corti, G. (2012). "Evolution and characteristics of continental rifting: Analog modeling-inspired view and Tectonophysics. 522-523, 1-33.  flood basalts and Earth and Planetary Science Letters. 166. 177-195.  -tectonic evolution of Gondwanaland  GEOLOGICAL SOCIETY OF AUSTRALIA ABSTRACTS (67: 109-109). Geological Society of Australia; 1999.  67  Geophysical Research Letters 21 (2): 145-148.  Danyushevski, L. V., Perf--hosted melt inclusions: evidence from the Contib. Mineral. Petrol. 144: 619-637.   Journal of Petrology 45 (112): 2531-2553.  -element geochemistry of mantle olivine and Chemical Geology 270: 196-215.  Doucet, L. S., Peslier, A. H., Ionov, D. A., Brandon, A. D., Golovin, A. V., Goncharov, A. G., Ashchepkov, I. V. (2014). "High water contents in the Siberian cratonic mantle linked to metasomatism: an FTIR study of Udachnaya peridotite xenoliths." Geochimica et Cosmochimica Acta. doi: http://dx.doi.org/10.1016/j.gca.2014.04.011  Geochimica et Cosmochimica Acta 56: 2417-2423.   29, 527530.  Elkins-Geological Society of America Special Paper. 388: 449-461.  Ferrando, S., Frezzoti, M. L., Neumann, E. -R., De Astis, G., Peccerillo, A., Dereje, A., Gezahegn, Y. ,Teklewold, A. (2008). "Composition and thermal structure of the lithopshere beneath the Ethiopian plateau: evidence from mantle xenoliths in basanites, Injibara, Lake Tana Province." Journal of Mineralogy and Petrology 93: 47-78.  probes into early igneous and mantle melting prEarth and Planetary Science Letters 363: 181-191.  Frey, F. A., 1984. Rare earth element abundances in upper mantle rocks. In: Henderson, P. (ed.). Rare Earth Element Geochemistry. Amsterdam: Elsevier, 153-203.  Frezzotti, M. L., Ferrando, S., Peccerillo, A., Petrellie, M., Tecce, F., Perucchi, A. (2010). "Chlorine-rich metasomatic H2O-CO2 fluids in amphibole-bearing peridotites from Injibara (Lake Tana region, Ethiopian plateau): Nature and evolution of volatiles in the mantle of a region of continental flood basalts." Geochimica et Cosmochimica Acta 74: 3023-3039.  Contrib. Mineral. Petrol. 122: 97-115.  Furman, T. (2007). "Geochemistry of East African Rift basalts: An overview." Journal of African Earth Sciences 48: 147-160.  68  Griffin, W. L., O'Reilly, S. Y., Afonso, J. C., Begg, G. C. (2009). "The composition and evolution of lithospheric mantle: A re-evaluation and its tectonic implications." Journal of Petrology 50: 1185-1204.  Gueydan, F., Precigout, P. (2013). "Modes of continental rifting as a function of ductile strain localization in the lithospheric mantle." Tectonophysics. doi: 10.1016/j.tecto.2013.11.029  Guirard, R., Bosworth, W., Thierry, J., Delplanque, A. (2005). "Phanerozoic geological evolution of Northern and Central Africa: An overview." Journal of African Earth Sciences 43: 83-143.  elt accumulation at the lithosphere-Earth and Planetary Science Letters. 376: 20-28.  as indicators of the extent of melting in mid-ocean-Nature 410: 677-681.  Journal of Petrology 52, 1: 113-146.  Hoffman, C., Courtillot, V., Feraud, G., Rochette, P., Yirgu, G., Ketefo, E., Pik, R. (1997). "Timing of the Ethiopian flood basalt event: Implications for plume birth and global change." Nature 389: 838-841.  Stress-Geochemistry, Geophysics, Geosystems 4 (5): 1-26.  relative importance of asthenospheric doming Journal of Geophysical Research-Solid Earth 106, 1127111291.  -dependent extension, two-stage breakup and cratonic Nature, 473 (7345): 74-78.  peridotite xenoliths from Spitsbergen: implications for trace element signature of mantle carbonate Earth and Planetary Science Letters 119: 283-297.  -bearing mantle peridotite xenoliths from Spitsbergen: phase relationships, mineral compositions and trace-element Contrib. Mineral Petrol. 125: 375-392.  Ionov, D. A. (2007). "Compositional variations and heterogeneity in fertile lithospheric mantle: peridotite xenoliths in basalts from Tariat, Mongolia." Contrib Mineral Petrol 154: 455-477.  Kaeser, B., Kalt, A., & Pettke, T. (2006). "Evolution of the lithospheric mantle beneath the Marsabit Volcanic Field (Northern Kenya): Constraints from textural, P-T and geochemical studies on xenoliths." Journal of Petrology 47: 2149-2184.  Karato, S., Toriumi, M., Fujii, T. (1980). olivine single crystals during high-temperature Geophysical Research Letters. 7: (9), 649-652.  Nature 442: 676- 679. 69   Kawabata, H., Hanyu, T., Chang, Q., Kimura, J., Nichols, A. R. L., Tatsumi, Y. (2011). "The petrology and geochemistry of St. Helena alkali basalts: Evaluation of the oceanic crust-recycling model for HIMU OIB." Journal of Petrology 52(4): 791-838.  Kay, R.WTectonophysics 219, 177189.  earthquakes near the Ethiopian rift induced by magmatic proceGeochem. Geophys. Geosyst. 10. Q0AB02, doi: 10.1029.2010TC002785.   Nature 358: 635-641.  Kelemen, P. B., & Dick, H. JJournal of Geophysical Research 100 (B1): 423-438.  Keranen, K., Klemperer, S.L. (2008). "Discontinuous and diachronous evolution of the Main Ethiopian Rift: Implications for development of continental rifts." Earth and Planetary Science Letters 265: 96-111.  Kieffer, B., Arndt, N., Lapierre, H., Bastien, F., Bosch, D., Pecher, A., Yirgu, G., Ayalew, D., Weis, D., Jerram, D. A., Keller, F., Meugniot, C. (2004). "Flood and shield basalts from Ethiopia: Magmas from the African Superswell." Journal of Petrology 45 (4): 793-834.  Annual Rev. Earth Planet. Sci. 37: 561-593.  ophioliJournal of Geophysical Research 102 (B12:27): 729- 27,749.  -poor Nature, 440: 324-328.   -MgO-SiO2-CO2 at 1 GPa, metasomatic wehrlites,  Contrib. Mineral Petrol. 138: 214-228.  le Roux, A., Class, C. (2014). "Metasomatism of the Pan-African lithospheric mantle beneath the Damara Belt, Namibia, by the Tristan mantle plume: geochemical evidence from mantle xenoliths." Contrib. Mineral. Petrol. 168(1046). doi: 10.1007/s00410-014-1046-y  Lorand, J--group elements and melt percolation Chemical Geology. 196. 57-75.   International Geology Review 57, 15: 1940-1974.  70   Developments in Precambrian Geology, 4: 649-687.   lithospheric mantle beneath Turkana depression in southern Ethiopia (the East Africa Rift): Geochemical and Sr-Nd-Pb isotopic Contrib. Mineral. Petrol. 162: 889-907.  Muntener, O., Manatschal, G., Desmurs, L., Pettke, T. (2010). "Plagioclase peridotites in ocean-continent transitions: Refertilized mantle domains generated by melt stagnation in the shallow mantle lithoshere." Journal of Petrology 51: 255-294.  Nature. 303. 577-584.  Moore, Geophysical Research Letters 26, 10731076.   continental Earth and Planetary Science Letters 202, 303320.  American Mineralogist 78: 1117-1134.  Ottonello, Gdistribution in some spinel peridotite xenoliths from Geochimica et Cosmochimica Acta 44: 1885-1901.  Orlando, A., Abebe, T., Manetti, P., Santo, A. P., Corti, G. (2006). "Petrology of mantle xenoliths from Megado and Dilo, Kenya Rift, Southern Ethiopia." Ofioliti 31(2): 71-87.  Pik, R., Deniel, C., Coulon, C., Yirgu, G., Marty, B. (1999). "Isotopic and trace element signatures of Ethiopian flood basalts: Evidence for plume- lithosphere interactions." Geochimica et Cosmochimica Acta 63(15): 2263-2279.   Earth and Planetary Science Letters 236: 148-166.  Putirka, K. D. (Reviews in Mineralogy & Geochemistry 69: 61-120.  Reisberg, L., Lorand, J-metasomatized continental mantle lithosphere: a case study of Sidamo spinel peridotite xenoliths (East Chemical Geology. 208. 119-140.  Roger, S., Dautria, J.-M., Coulon, C., Pik, R., Yirgu, G., Michard, A., Legros, P., Ayalew, D. (1999). "An insight on the nature, composition and evolution of the lithospheric mantle beneath the north-western Ethiopian plateau: the ultrabasic xenoliths from the Lake Tana Province." Acta Vulcanologica 11(1): 161-198.  Yirgu, G., Ebinger, C. J., Maguire, P. K. H. (Eds.), The Afar Volcanic Province within the East African Rift System. Geol. Soc. Spec. Publ., 259. 77-93.  71  sphere: Geochimica et Cosmochimica Acta. 69. (15). 3889-3910.  Geochemical evidence of lithospheric thinning in the southern Main Ethiopian  Lithos 117: 33-48.  Rooney, T. O., HertzbGeology 40, 1, 27-30.  Rooney, T. O., Bastow, I. D., Keir, D., Mazzarini, F., Movsesian, E., Grosfils, E. B., Zimbelman, J. R., Ramsey, M. S., Ayalew, D., Yirgu, G. (2014arotracted development of focused magmatic Tectonics 33: 875- 897.  Rooney, T. O., Nelson, W. R., Dosso, L., Furman, T., Hanan, B. (2014blithosphere metasomes in the production of HIMU-like magmatism on the northeast African and Arabian Geology 42: 419-422.  Rudnick, R. L., McDonough, WCarbonatite metasomatism in the northern  Earth and Planetary Science Letters 114: 463-475.  Sanfilippo, A., Tribuzio, R., Tiepolo, M. (2014). "Mantle-crust interactions in the oceanic lithosphere: Constraints from minor and trace elements in olivine." Geochimica et Cosmochimica Acta 141: 423-439.  Seyler, M., & -scale melt-rock interaction in lherzolitic mantle in the Earth and Planetary Science Letters 146: 273-287.  Skemer, P., Karato, S. (2008). "Sheared lherzolite xenoliths revisited." Journal of Geophysical Research: Solid Earth 113. doi: 10.1029/2007JB005286  Sobolev, A. V., Hofmann, A. W., Kuzmin, D. V., Yaxley, G. M., Arndt, N. T., Chung, S-L., Danyushevsky, L. V., Elliott, T., Frey, F. A., Garcia, M. O., Gurenko, A. A., Kamenetsky, V. S., Kerr, A.C., Krivolutskaya, N. A., Matveinkov, V. V., Nikogosian, I. K., Rocholl, A., Sigurdsson, I. A., -derived Science 316: 412-417.  Spiegelman, M., & KGeochemistry, Geophysics, Geosystems 4, 7: 1-18.  -Nubian shieNature 382: 773-778.  Annual Rev. Earth Planet Sci. 22: 319- 351.  ntaneous small-scale melt segregation in partial melts undergoing Geophysical Research Letters 16 (9): 1067-1070.  Stosch, H-Geochimica et Cosmochimica Acta 46: 793-811.  72  Journal of Petrology 40 (4): 575-599.  Sun, C., Liang, Y. (2014). "An assessment of subsolidus re-equilibration on REE distribution amoung mantle mineral olivine, orthopyroxene, clinopyroxene, and garnet in peridotites." Chemical Geology 372: 80-91.   Magmatism in the Ocean basins, edited by A. D. Saunders. Geol. Soc. London. Spec. Publ. 42: 313-345.   Takazawa, E., Frey, F. A., Shimizu, N., Obata, M., Bodinier, J. L. (1992). for Nature 359: 55-58.  -assisted strain localization in an orogen-Tectonics 13 (2): 421-437.  Earth and Planetary Science Letters 272: 65-77.  Geophysical Research Letters. 20: (14): 1479-1492.  va-scale convection during continental rifting: evidence Geology 36, 575578.  d Tectonophysics, 296: 61-86.  Contrib. Mineral. Petrol. 162: 675-689.  Warren, J. M., olution of olivine lattice preferred orientation during Earth and Planetary Science Letters 272: 501-512.  Warren, J. M., Hauri, E. H. (2014). "Pyroxenes as tracers of mantle water variations." Journal of Geophysical Research: Solid Earth 119. doi: 10.1002/2013JB010328   Journal of Geophysical Research, 94 (B10): 13, 919-950.  for the differentiation history of the crust-Earth and Planetary Science Letters 205: 309-324.   Journal of Petrology 23: 240-292.  73  Wilson, D. H., Aster, R., West, M., Ni, J., Grand, S., Gao, W., Baldridge, W. S., Semken, S., Patel, P. Letters to Nature, 433: 851-855.  Magmatic Systems 57, 355-379.  Wolfenden, E., Ebinger, C., Yirgu, G., Deino, A., Ayalew, D. (2004). "Evolution of the northern Main Ethiopian rift: birth of a triple junction." Earth and Planetary Science Letters 224: 213-228.  margin: Southern Red Sea, Ethiopia. Geological Society of America Bulletin 117: 846-864.   Journal of Petrology, 39 (11-12), 1917-1930.  Zeyen, H., et alTectonophysics 278, 329352.  Zhang, H--melt reaction: a case of Sino-Earth and Planetary Science Letters 273: 768-780.   Earth Science Reviews 64, 150.