Journal of the Geological Society, London, Vol. 148, 1991, pp. 600-606, 2 figs. Printed in Northern Ireland

Discussion of ophiolites in Northeast and East Africa: implications for Proterozoic crustal growth Journal, Vol. 147, 1990, pp. 41-57

W. R. Church writes: In discussing the Pan-African geology theeastern margin of theEastern . They do not of Northeast andEast Africa, Berhe (1990) affirms that necessarily cross into Saudi Arabia. On the Saudi Arabian ophiolite decorated lineaments in the Arabian-Nubian side of the Red Sea, the AI Wask, and Farri Group rocks of Shield representsuture zones. The difference of opinion the supposed Yanbu suture may represent parts of a similar between Berhe (1990) and Stern et al. (1989) concerning the ophiolitic sheet exposed as a result of out-of-sequence thrust location of the supposed suture zones of northern and faulting, with the820Ma old Iqwaq granodiorite repre- southern Egypt, is a clear illustration of the ambiguity that senting a window of older arc basement. The 780 Ma old may arisethrough the uncritical use of thisparadigm. JabalEss ophiolite (Pallister et al. 1988), whichis little Whereas Berhe considers the Sol Hamed-Wadi Onib deformedand also associated with melange and pebbly ophiolite belt to represent in situ oceanic material located mudstone units, may form the uppermost unit of the nappe along a north-south trendingextension of the Saudi pile. Since the 808 Ma (Stern et al. 1989) Sol Hamid-Wadi Arabian Yanbu-Sol Hamed ‘suture’, Stern et al. (1989) link Onib ophiolites are also upthrust tothe south of the thelatter with the east-west trending Allaqi-Heiani regionally south dipping 741 Ma Gerf ophiolite, the relative ophiolite belt, which they consider to be a major east-west age and disposition of the ophiolites could be taken to imply suture ‘extending well into the interiorof North Africa. This that the ophiolitic rocks represent oceanic material formed important difference in opinion is a clear indication of the along the eastern margin of the Hijaz ocean, which would arbitrary nature of suture selectionbased solely onthe therefore now be buried somewhere beneath the ophiolitic distribution of ophiolitic rocks. Furthermore,neither nappesomewhere to the south of the Yanbu-Sol Hamid explanation may be correct. ophiolitebelt. The location of the ocean further south The distribution of ophiolitic rocks in the southern part within the Pan-African likely lies east of the ophiolite of the and northern Sudan is more likely occurrence at Ingessana, since Shackleton (1988) has argued controlled by the southern Eastern Desert domal culmina- that the Ingessana ophiolite represents an erosional remnant tion, out-of-sequencefaulting, andthe development of of a large thrust sheet. The thrust sheet may have extended zones of intenseductile strain. Within the domal or slid asfar tothe west as the Nuba Hills. Problems culmination, ophiolitic material exposed at lower structural associated with the location of sutures in the Kenya section levels nearAbu Swayel, Gebel Nagy, Um Krush and of the Pan African havebeen discussed by Shackleton perhaps GebelGerf, overlie highly deformed and (1986). metamorphosed pelitic metasedimentscontaining a ‘con- Evenin the Saudi Arabian shield,where terranesare tinental’ Nd isotopesignature (Harris et al. 1984), relatively well defined on the basis of age criteria, and where quartzo-feldspathic gneisses (Umm Tundeiba), hornblende- accretion is most likely to have involved lateral transporta- cummintonite-garnetamphibolites of arcderivation, and tion of arc terranes, it is now apparent that ophiolites are compositionally laminated hornblende-garnet felsic rocks of not the best indicators of terrane boundaries or sense of unknown tectonic affiliation. To the south in Wadi Murra, terrane movement. Taken at face value, the model ages of highly strained mafic schists are succeeded by a southerly- Pallister et al. (1988) indicate thatthe ophiolites of the facing upward-coarsening siltstone-turbidite-pebbly Arabian Nabitah ‘suture’ belong to two different terranes, mudstone-melange sequence, similar in most essential one with an age of c. 850 Ma. (Asir), the other with an age respects tothat described furthernorth in the ophiolitic of c. 750 Ma. (AI Qarah).The Junaynah‘age suture’ nappe pile of the Marsa Alam region of the Eastern Desert therefore appears to transect the Nabitah ‘ophiolite suture’. (El Sharkawi & El Bayoumi 1979; Basta et al. 1986). The In neither case are the ophiolites confined to the terrane clasts of quartzite from which detrital Archean zircons were boundary. first obtained by Dixon (1981) and which have subsequently In the case of the supposed Bir Umq-Thurwah-Nakasib been found elsewhere in the central Eastern desert (Wust et ‘suture’, it is also worth noting that (1) there is clear al. 1987), were taken from pebbly mudstones of the Murra evidence of crustal contamination in the Thurwah ophiolite, melange. The relationship of the melange theto but not in the Bir Umq diorite (Pallister et al. 1988); (2) the discontinuousophiolite belt north of Wadi Allaqi is Thurwah ophiolite could be as young as 810 Ma or as old as unknown, but by analogy with theGhadir region the 870 Ma, and could therefore be 30 Ma older or younger than ophiolites could form an upper ophiolite component of the the Bir Umq diorite; (3) the Thurwah ophiolite lies north of nappe pile. Rather than being a ‘suture’, the Allaqi-Heiani the Labunah thrust, the supposed suture, whereas the Bir ophiolite belt would thereforerepresent a structural level Umq rocks lie south of the supposed suture; (4) rocks in the within the rim of anarched ophiolitic nappe, or evena Rabigh area to the northof the Thurwah ophiolite are as old pinched syncline separatingthe southern Eastern Desert as 945 f 45 Ma (AI-Shanti et al. 1984); and (5) all definitions culmination from a northern Sudan culmination west of the of asuture along a Bir Umq, Thurwah,Nakasib (Sudan) Hamisana shear zone. The Allaqi-Heiani-Gerf ophiolitic line have ignored the presence of serpentinites on the Sudan belt may skirt the southern Eastern Desert culmination to side of the Red Sea southwest of Mohamed Qol at 20”30’, join up with the ophiolite/melange units cropping out along 36’30‘ (Vail Map of N.E.Sudan, unpublished map 600

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/148/3/600/4892200/gsjgs.148.3.0600.pdf by guest on 27 September 2021 DISCUSSION 601

compilation, 1978). It is therefore far from certain that the 1990) and contrasting geology is found on either side of the Thurwah ophiolite defines the boundary of the Asir terrane; sutures. the boundary could lie underthe c. 740-700 Ma old AI If the Jabal Ess ophiolite formed the uppermost unit of Ays-Furayh-Neferdeib volcano-sedimentary successor ba- the AI Wask nappe pile, then we would expect the former to sin of the Hijaz terrane. contain only a limited portion of an ophiolitesequence. At the eastern edgeof the Saudi Arabian Shield, the Abt However both the Jabal Ess (Shanti & Roobol 1979) and Schist, which contains detrital chromite, garnet, carbonate, the A1 Wask complexes display completea ophiolite muscovite and anatase, was deposited in a basin to the rear succession fromserpentinized ultramafic rocks,cumulate of the western leading edge of the obducted Urd-AI Amar and high level gabbros andan upper metavolcanic and ophiolite. It was subsequently overthrust by the Ar Rayn arc metasedimentarysequence. This suggests that theyrepre- rocks along the AI Amar ‘suture’. The main ocean basin and sent distinct complexes. closure suture therefore likely lie to the east of the Ar Rayn The suggestion that the Ingessana ophiolite could have terrane. extended as far west as the Nuba Mountains implies that the Berhe (1990) suggests that geological differences along Nuba and Ingessana ophiolites arepart of the same the length of the Pan African-Mozambique belt reflects the complex. It is true that the Ingessana complex represents an existence of a relatively narrow belt of oceanic crust, and erosional remnant of a large thrust sheet (Shackleton 1988), consequently a low degree of crustalgrowth, within the but it is unlikely thatthe Ingessanaophiolite could have Mozambique portion of thebelt. In contrast, Reymer & been detached and transported for over 200 km without the Schubert (1984) and Pallister et al. (1990) have argued that other mafic-ultramafic complexes in the surrounding regions thearc accretion rate of thenorthern Arabian-Nubian being affected. Besides, the Nubaophiolite (Hirdes & shield was excessively high compared to that of the present Brinkmann 1985; Steiner 1987) and the Ingessana ophiolite day.This antipathetic relationship can partin be (Abdel Rahman 1983) separate high-grade metasediments rationalized,however, if it is assumed thatthe missing from low-grade volcano-sedimentary sequence. This means southern arc terranes have migrated laterally northwards to thatthere is acontrast of geology oneither side of the form part of the Arabian-Nubian amalgamated arc system, postulatedsutures as indicated by a recent study of the in thesame way thatthe North AmericanCordilleran Nuba Mts (Abdelsalam & Dawoud 1991). system has amalgamated by lateral arc accretion. In orderto show the ambiguity of suture selection, Church discusses the Bir Umq-Thurwah-Nakasib belt as a 15 June 1990 case study. The Thurwah and Bir Umq ophiolites do show minor differences in geochemistry along strike (Nassief et al. 1984). Pallister et al. (1988) dated zircons fromgabbroic Seife M. Berbe replies: Churchraises two principal rocks of Thurwah complex and found old ages (1250Ma) objections to my recentpaper: (1) ophiolites are not the from two of the fractions. They explained this result as a best indicators of terrane boundaries or sense of terrane product of assimilation of older materialduring emplace- movement; (2) the Mozambiquebelt incorporates a small ment of the gabbro. Church uses these data to suggest that proportion of oceanic crust implying a low degree of crustal since the Bir Umq diorite does not show evidence of growth in the region. The main thrust of his argument is that contamination, then these complexes cannot define a uncritical use of distribution of ophiolitic rocks to suggest terrane boundary. The fact that Bir Umq diorites do not suture zones could create ambiguity in linking suture zones. show crustal contaminationas comparedto the Thurwah I agreeI thatsuture selectionbased solely onthe complex does not in itself mean that they do not form part distribution of ophioliticrocks is arbitrary.In fact Berhe of the same suture. Crustal contamination can be variable (1990) suggests that in order to define a suture zone three within an intrusion and theseobservations offer no real criteria have to be fulfilled: (a) the presence of convincing constraints on inter-ophiolitecorrelation. Asfaras ophiolite assemblages; (b) structural trends between the differences in age are concerned, ophiolite rocks are difficult ophiolites that align along strike of the suture; (c) a contrast to date (e.g. Desmons 1982) so that apparent age differences in geology on either side of the suture. The important point may merely reflect analytical uncertainties. It is also in suture selection is not so much whether the ophiolites are essential to compare the same rock types within the same close tothe suture itself but whether they represent stratigraphic level which, in most documentedstudies of fragments of oceanic crust. It is true that ophiolites, because ophiolites, hasnot been done.At present the ophiolites of the nature of their emplacement, must have moved from which are most accurately dated are those of Saudi Arabia, their origin and nowhere in the paper does it state that the but even in these areas the analytical errors are so large and ophiolites are found insitu. However it is argued that in the differences in ages of the ophiolites so small that it is at most cases the ophiolites did not move more than a few tens present difficult to differentiate sutures on the basis of age of kilometres. (Harris et al. 1990). As far as structural data are concerned, Responding to Church’s specific comments, I agree that it is not surprising thatthe Thurwah andthe Bir Umq the Allaqi-Heaini ophiolitebelt doesnot necessarily ophiolites lie on opposite sides of the postulated suture continue into Saudi Arabia.The Allaqi-Heainiophiolite because post-emplacementdeformation has offset the was interpreted as forming part of the ophiolitic melange of ophiolites along aleft-lateral strike slip Najd fault. Camp theEastern Desert and hence could not representan (1984), Stoesser & Camp, (1985) and Pallister et al. (1988) E-W-trending suture zone. The most unambiguous suture have camed out detailed studies and suggest that the Bir zones arethe Sol Hamed-Wadi Onib complexes andthe Umqophiolite marks a major suture that crosses to Port Yubdo ophiolite belt because the ophiolites do not form a Sudan, and study of the Port Sudan area indicates that this tectonic melange; they can be traced linearly for hundreds suture continues into NE Africa. of kilometres (Fitches et al. 1983; Hussein et al. 1984; Berhe Church claims that since Reymer & Schubert (1984) and

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/148/3/600/4892200/gsjgs.148.3.0600.pdf by guest on 27 September 2021 602 DISCUSSION

Pallister et al. (1990) have suggested that the arc accretion Reconstruction of the ophiolite-decoratedsutures. Berhe rate of the northern Arabian Shield was excessively high as (1990) suggested two reconstructions for the sutures in the compared to the present day. Based on this, Church argues Sudan: (a) the Sol Hamed-Onib and Nakasib suture joined that the high rate of arc accretion could be explained by are linked to the Bayuda ophiolite, continuing southward to migration of arc terrane from the Mozambique belt to form the Ingessana ophiolite and thence as a single suture into the the Arabian-NubianShield. In a later study,Reymer & Mozambique Belt; (b) the Sol Hamed-Onib suture is linked Schubert (1986) suggest that the increase in crustal accretion with the Bayudaophiolite, and southward tothe Nuba rates could either be relatedto hot spot volcanism and ophiolite. The Nakasib suture is linked tothe Qala En underplating in addition toarc accretion, or that large Nahal ophiolite. These continueas two sutures intothe amounts of pre-existing basementhave gone undetected. Mozambique Belt. Pallister et al. (1990) calculated crustal growth rates an order Berhe (1990) indicated that the Allaqi-Heiani ophiolite of magnitude lower thanthose suggested by Reymer & belt is not a suture due toits E-W trend which implies that Schubert (1984, 1986). Harris et al. (1990) conclude growth it extends perpendicular to the boundary of the craton. rates, although fast, are comparable with crustal growth of We would like to draw attention to thedifferent tectonic the CanadianCordillera. Actual crustal growth ratesare settings of the Sol Hamed-Onib and Nakasib ophiolites on lower than those suggested by Reymer & Schubert (1986) one hand and the Bayuda and Nuba ophiolites on the other. partly because older cratonic material is present in the Afif The first group comprises sutures between Pan-African arc terrane (Stacey & Hedge 1984; Stacey & Agar 1985), and terranes (intraoceanicsutures) while the second group hence estimates based on assuming thatthe whole shield representssutures between the Pan-African juvenile crust consists of upper Proterozoic crust are invalid. andolder cratonic elementsto the west (Abdelsalam & Arc accretion may have contributed to crustal growth of Dawoud 1991). Hence, we suggest thatthe Bayudaand the Arabian-Nubian Shield,but there is no evidence to Nubaophiolites are linkedand extend northward tothe suggest that the source of these arcs lay in the Mozambique Keraf zone (Almond & Ahmed 1987). This configuration is Belt.Indeed the widespreadoccurrence of ophiolites in in agreement with the available geological data (Dawoud Tanzania,Mozambique and Madagascar andthe palaeo- 1980; Ries et al. 1985; Abdelsalam 1987; Almond & Ahmed magnetic data from the region (McWilliams 1981) do not 1987) and geochronological data(Harris et al. 1984; support such a model. Schandelmeier et al. 1988). The Sol Harned-Onib suture is AlthoughChurch has raised severalinteresting points, linked, across the Hamisanashear zone,to the Allaqi- several issues remainunresolved. I believe that until Heiani suture (Stern et al. 1990). The Allaqi-Heiani suture systematic structural, isotopic and palaeomagneticstudies is not well documented in theliterature but published are carried out for NEand East Africa it is difficult to geological maps show serpentinite bodies, elongated along dismiss the reconstruction I suggested. Inthe absence of an E-W trend, that can be followed from the Hamisana reliable data the discussion will remain speculative. shear zone to the Nile. This belt must be considered as a I believe it is amistake to generalize the geological suture, both in terms of the abundance of the ophiolitic evolution of the entire Arabian-Nubian Shield based on a fragments within it and in its lateral extent. We agree with single area or region,studiesas have shown that Berhe (1990) in that thisreconstruction implies thatthe Arabian-Nubian Shield and MozambiqueBelt can be Allaqi-Heiani suture appears toextend perpendicular to the divided intothree distinct ophiolitedomains. (a)The boundary of the Nile craton. This conflict requires ophiolitic melange of the Eastern Desert of Egypt extending modification of the model, not dismissal of this important tothe northwesternpart of theRed Sea Hills (Allaqi- suture. Heaini area); in this area no coherent suture zone can be inferred. (b) The central zone which includes Saudi Arabia, The NW-trendingfault zones. Berhe (1990) remarked on Sudan, Ethiopia and northern Kenya. This domain marks an the presence of numerous sinistral strike-slip faultzones area where the ophiolites are considered not to have moved dominantly trending NW in the Arabian-Nubian Shield and more than a few tens of kilometres and extend linearly for the MozambiqueBelt (Fig. la).He interpreted these as hundreds of kilometres. (c) The Southern ophiolite domain conjugate faults related to continent-continent collision in which includes southern Kenya, Tanzania and Mozambique. the vicinity of the Mozambique Belt. In this area, tracing suture zones represent a major problem The indenter position is important in understanding the as there is evidence for large scale horizontal movements. geometry and distribution of conjugate fault sets related to continent-continent collision. In Fig. 1 we show the I would like to thank R. M. Shackleton, N. B. W. Harris, F. indenter as EastGondwanaland (the African part of it is McDermott and R. Price for their comments and discussion. now exposed as theEastern Kibaran Craton (Key etal. 3 September 1990 1989) colliding with West Gondwanalandalong the Mozambique Belt. This configuration (with the addition of M. G. Abdelsalam & R. J. Stem write: The paper of Berhe India and ) is taken from Burke & Sengor (1986) (1990) represents an important effort to integrate the late and isbasically that advocated by Berhe (1990). This Precambrian orogenic history of East and Northeast Africa configuration shows remarkable similarities to the Cenozoic and Arabia, but we disagree with Berhe regarding: (1) example of India colliding against Asia (compare Figs la the reconstruction of the ophiolite-decorated sutures in and 2a). Also shown in Fig. la are the NW-trending fault the Sudanese sector of the Arabian-Nubian Shield; (2) the zones as shown in fig. 5 of Berhe (1990). Lines representing interpretation that NW-trending strike-slip fault zones of the the predicted conjugate sets of strike-slip faults (Fig. lb) are Arabian-Nubian Shield andthe MozambiqueBelt are superimposedas dashed lines on Fig. la. These lines are conjugatesets related tolate Precambriancontinent- taken from Tapponnier et al. (1982) who conducted plane continent collision. indentation experiments on unilaterally confined blocks of

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/148/3/600/4892200/gsjgs.148.3.0600.pdf by guest on 27 September 2021 DISCUSSION 603

EGONDWANALAND a

Fig. 1. (a) Reconstruction of Africa and Arabia as part of Fig. 2. (a) Schematic map of Cenozoic extrusion tectonics and large Gondwanaland (modified after Burke & Sengor 1986). EKC, faults in eastern Asia. (b) Hand drawing of the unilaterally confined Eastern Kibaran craton; M, Madagascar. NW-trending strike-slip indentation experiment on plasticene. Indenter displacement is faults of east and northeast Africa and Arabia are shown as solid 6.3 cm. Different regions of the experiment labelled are: indenter, lines and are taken from Berhe (1990). Lines representing the India; plastic body, Asia; escaping blocks, Indochina and south predicted conjugate sets of strike-slip faults are superimposed as China (after Tapponnier er al. 1982). dashed lines and are taken from Tapponier et al. (1982). (b) Hand drawing of the unilaterally confined indentation experiment on Plasticine (Tapponnier et al. 1982). Indenter displacement is 3.5 cm. marked. In contrast,the NW-trendingfaults of the Different regions of the experiment labelled are: indenter, East Arabian-Nubian Shield and the Mozambique Belt deviate Gondwanaland; plastic body, West Gondwanaland; escaping block, significantly from that expected from the Tapponnier et al. Nubian-Arabian Shield. (1982) model. Thispart of the discussion points out the argument against theinterpretation of the NW-trending faults of East and Northeast Africa and Arabia as conjugate Plasticine in ordertounderstand finite intracratonic sets related to continent-continent collision. deformation and the evolution of strike-slip faulting in east (a) The strike-slip faultzones of the Arabian-Nubian Asia (Fig. 2). Shield andthe Mozambique Belt consistently trend NW It is difficult to modelexperimentally theintraplate (Berhe 1990). The requisitecomplementary SW-trending deformation related to continent-continent collision. This is fault sets (to considerthem as conjugate sets) are not because we areignorant of the long-term mechanical reported from the region. behaviour of the continentalcrust and lithosphere. (b) The model of Tapponnier et al. (1982) shows that However, the gross resemblancebetween the actual (Fig. the conjugate fault sets generally concentrate in front of the 2a) andthe predictedfaults (Fig. 2b) fromeast Asia is indenter whereas the fault sets near the freeface are parallel

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/148/3/600/4892200/gsjgs.148.3.0600.pdf by guest on 27 September 2021 604 DISCUSSION

to the collisional zone (Figs lb & 2b). Hence, in the case of linked tothe Keraf zone(Almond & Ahmed 1987) that east and northeast Africa and Arabia the conjugate sets are separates high and low-grade rocks in western Sudan. I do expected west of the Mozambique belt. No such fault sets not think there is any problem in connecting the Bayuda and are shown in Berhe (1990). Burke & Sengor (1986) Nuba Mountains tothe Keraf zone provided there is suggested thatthe free faceduring the latePrecambrian confirmation of ophiolitic affinity of the ultramafic rocks of continent-continent collision in east Africa was located the Keraf zone. This reconstruction does not invalidate the southwest of Turkey. Hence, if the Najd fault system was argument thatthe Sol Hamed-Onib complexes can be induced by collision, then it should trend N-S parallel to the linked with the Bayuda complex, because an intra-oceanic Mozambique Belt and not NW. suture can be linked with a suture that separates juvenile (c) Berhe (1990) outlined the NW-trending fault zones Pan-African crust with older cratonicboundary. For of east Africa as extending from the Congo Craton into the example, the Yubdo ophiolite (an intra-oceanic suture) has Eastern Kibaran Craton (Fig. la). Following the analogy of been linked tothe Sekerr ophiolite of Kenya which India, the Eastern Kibaran Craton should be largely free of separates older cratonicmaterial with Pan-African crust conjugate fault sets since it is considered the rigid indenter (Vearncombe 1983; Berhe 1990). These type of during the collision. Instead, there does not appear to be reconstructions are also observed in the Circum-Pacific any significant difference in theabundance of the region (Doutch et al. 1981; Stauffer 1983). NW-trendingstrike-slip faults oneither side of the I have alreadyargued in my reply to Church that the Mozambique Belt at least as these are shown on fig. 5 of Allaqi-Heiani ophiolitebelt could be related tothe Berhe (1990). ophiolite melange of theEastern Desert of Egypt.The The absence of a geneticrelationship betweenthe existence of an E-W suture implies that there are accreted NW-trending strike-slip faultzones in East and Northeast arc terranes connecting the Arabian-Nubian Shield with the Africa andArabia and the collisional event along the Hoggar of Central Africa. At presentsuch evidence is Mozambique Belt is further supported by geochronological lacking. datafrom the region. Thesedata indicate thatthe two tectonic events are notsynchronous. The age of the The NW-trending fault zones. It is true that there are broad granulite facies metamorphism along the Mozambique Belt similarities between the Cenozoic example of India colliding can beused to constrain when crustal thickening (and with Asia,and the collision of Easternand Western by implication,continental collision) occurred. Geo- Gondwanaland. Fleck et al. (1980) and Davies (1984) have chronological data (Maboko et al. 1985; Kroner et al. 1987) suggested conjugatefault sets in Arabia as evidence of suggest that this collisional eventtook place atabout collision. They have also discussed the similarity with the 700-750Ma ago. Unlike the NW-trending strike-slip faults Cenozoic collision of India with SE Asia.However the in East and Northeast Africa, the timing of the Najd fault evolution of the Arabian-Nubian Shield and the system is well constrained. Stern (1985) concluded that the Mozambique Belt is more complex than that of SE Asia. principal Najd movementoccurred during the interval The major difference between the Arabian-Nubian 560-620Ma. Stacey & Agar (1985) suggested that the Najd Shield and the collision in SE Asia is that the NW-trending faulting commenced with a dextral phase 640Ma ago, and faults of the Arabian-Nubian Shield and the Mozambique thatthe system changed to sinistral strike-slipmotion at Belt deviate significantly from that expectedfrom the about 620 Ma ago. This suggests that the initiation of the experimental model of Tapponnier et al. (1982). This led Najd faulting occurred at least 60Ma after the collisional Abdelsalam & Stern to suggest that the NW-trending faults event to the south. could nothave formed asconjugate sets related to Finally, suggesting an alternativemodel is beyond the continent-continent collision. The difference in scope of this discussion. We hope that the above points will interpretation was also due to the fact that complementary be useful towards understanding better the tectonic history sets of SW-trendingfaults and conjugate sets were not of the Arabian-Nubian Shield and Mozambique Belt. shown west of the Mozambique Belt on fig. 5 (Berhe 1990). In this figure the NW trendingfault zones are identified 9 August 1990 because they are the most important fault zones that show displacement of ophiolite belts, but that does not mean that Seife M. Berhe replies: Abdelsalam & Stern raise two there are no other faults in the area (see fig. 1, Berhe 1990). principal objections to my recent paper to which I would However the NW-trendingfaults produce the dominant like to respond. fracture pattern in the region. Detailedstudies have beencarried out in NEand E Reconstruction of the sutures in Sudan. Responding to Africa which show the presence of conjugatesets which Abdelsalam & Stem's specific comments, it is truethat trend NE-SW and NW-SE in the NE Sudan region, and when I considered the reconstruction of suture zones in NE trend NNE-SSW and NW-SE in western Ethiopia (Berhe Sudan, I envisaged two possible scenarios: (i) that the Sol 1986a, 1990). InSE Ethiopia the conjugate sets trend Hamed-Onib andthe Khor Nakasibophiolite complexes NE-SW (045") and NW-SE (150") (Berhe 19866), while in can be aligned with the Bayudacomplex (fig. 1,Berhe the Baragoi area of N Kenya there are four sets of faults 1990); or alternatively (ii) the Sol Hamed-Onib complex trending OlO", OW, 120" and 160" (Berhe & Rothery 1986). can linkedbe theto Bayuda complex while the Based on theoretical considerations and experimental data, Khor-Nakasib may beconnected tothe Qala Nahal it has been established that two deformation episodes most complex. However I mentioned that the first scenario was likely controlled the growth of the wrench faultzones implausible and supported the latter reconstruction (Berhe (Berhe & Rothery 1986). No conjugate sets are shown west 1990). Abdelsalam & Stern suggest a thirdalternative, of the Mozamb'ique Belt because of the absence of major namely thatthe Bayuda and Nuba Mountainscould be conjugate shear zones in the area.

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/148/3/600/4892200/gsjgs.148.3.0600.pdf by guest on 27 September 2021 DISCUSSION 605

The other objection of Abdelsalam & Stern is that the and petrochemistry of Um Gerad granitic rocks (Rabigh area), western NW-trending fault zones extend from theCongo Craton into Saudi Arabia. PanAfrican Crustal Evolution in the Arabian Nubian the coastal areas of Mozambique andTanzania. This is Shield, I.G.C. 164. Faculty of Earth Sciences, King Abdulaziz University, Jeddah, Bulletin 6, 221-232. mainly because the northwesterlyfaults have had an BASTA,E. Z., CHURCH,W. R., HAFEZ, A. M. A. & BASTA,F. F. 1986. extended history andhave been reactivated duringthe Proterozoic ophiolitic melangeand associated rocks of Gebel Ghadir Mesozoic and the Tertiary. area, Eastern Desert, Egypt. International BasementTectonics Theimportant observation raised by Abdelsalam & Association Publication, 5, 115-123. BERHE,S. M. 19%~. Application of remote sensing to tectonicand Stern is the absence of a genetic relationship between the metallogenicstudies in NEAfrica. Proceedings of theFifrh Thematic NW-trending fault zones in NE and E Africa and Arabia. Conference on Remote Seming for Exploration Geology, 1, 383-391. This can be resolved if the collision model of the evolution -1986b. Geologic and geochronologic constraints on the evolution of the of the Pan-African/Mozambique Belt proves more complex Red Sea-Gulf of Aden and Afar Depression. Journal of African Earth Sciences, 5, 101-117. than the Cenozoic collision of India with Asia. I suggest that - 1990. Ophiolites in northeast andeast Africa: Implications for collision was induced from twodirections; from the Proterozoic crustal growth. Journal of the Geological Society, London, northeast (EasternArabia) and from thesoutheast 147, 41-57. (Madagascar). -& ROTHERY,D. A. 1986. Interactive processing of satellite images for structural andlithological mapping in northeast Africa.Geological It is not only important to understand the position of an Magazine, 123,393-403. indentor,but also the sense of movement in orderto BURKE,K. & SENGOR,A. M. C. 1986. Tectonic escape in the evolution of the establish the area1 distribution of conjugate sets of faults in continental crust. American Geophysical Union, Geodynamics Series, 14, thearea. A classical example, (Fig. 1) EastGondwana is 41-53. CAMP,V. E. 1984. Island arcs and their role in the evolution of the western shown to collide with West Gondwanahead on. If we Arabian Shield. Geological Society of America Bulletin, 95, 913-921. consider the collision to have occurred obliquely rather than CHURCH,W. R. 1988. Ophiolites, sutures, andmicroplates of the head on, then the conjugate sets have to be expected further Arabian-NubianShield: a critical comment. In: EL-GABY,S. & to the northwest in Tanzania, Uganda, Kenya and southern GREILING,R. 0. (eds) The Pan-African belt of North-eastAfrica and Adjacentareas. Freidr. Viewingand Sohn, Braunschweig,Wiesbaden, Sudan. In these areas the major structures areNW-trending, 289-316. but N-S- and NE-trending rifts also occur. It is possible that DAVIES,B. F. 1984. Strain analysis of wrench faults and collision tectonics of some of the rifts could be reactivated basement structures the Arabian Shield. Journal of Geology, 82, 37-53. during the Tertiary. DAWOUD,A. S. 1980. Structural and metamorphicevolution of the area Abdelsalam & Stem dismiss the continent-continent southwest of Abu Hamed, Nile province, Sudan. PhD thesis, University of Khartoum. collision model because the tectonic events in Mozambique DESMONS,J. 1982. Radiometric dating of Late Proterozoic ophiolites: and the Najd are not synchronous. This hypothesis is true meaning, correlations and problems. Precambrian Research, 16, A15. only if we assume thatthe collision inducedin the DIXON, T.H. 1981. Age and chemical characteristicsof some pre-Pan African Mozambique Belt andthe Najd is caused by a single rocks in the Egyptian Shield. Precambrian Research, 14, 119-133. DOLITCH.H. F., PACKHAM,G. H., RINEHART,W. A., SIMKIN,T., SIEBERT,L., indenter as shown in Figs 1 & 2. It is at present difficult to MWRE, G. W., GOLOVCHENKO,X., LARSON, L. & PITMAN,W. C. 111. constrain the two collision events,but if we take 1981. Plate tectonic map of the circum-Pacifc region, southwest quadrant: Abdelsalam & Stern’s suggestion that the Najd faulting was Tulsa, Oklahoma, scale 1 : 1OOOOOOO. AmericanAssociation of initiated afterthe collision eventto the south, then it is Petroleum Geologists. plausible to suggest northwestwardmovement of EL SHARKAWI,M. A. & EL BAYOUMI, R.M. 1979. The ophiolites of Wadi Ghadir area, Eastern Desert, Egypt. Annak of the Geological Survey of Madagascar,and westward movement of theAr Rayn Egypt, 9, 125-135. microcontinent of the eastern Arabian Shield (Stoesser et al. FITCHES, W. R., GRAHAM,R. H., HUSSEIN,1. M., RIES,A. C., SHACKLETON, 1984). R. M. & PRICE,R. C. 1983. The late Proterozoic ophiolite of Sol I believe that until systematic structural, geo- Hamed, NE Sudan. Precambrian Research, 19, 385-411. FLECK,R. J., GREENWOOD,W. R.,HADLEY, D. G., ANDERSON,R. E. & chronological and palaeomagnetic studies are carried out in SCHMIDT,D. L.1980. Age and evolution of the southern part of the NE and E Africa, contentious issues will remain. However Arabian Shield. Institute of Applied Geology, Bulletin, 3, 1-19. any model proposed must account for tectonic and HARRIS,N. B. W., GASS,I. G. & HAWKESWORTH,C. J. 1990. A geochemical geodynamical variationsobserved across theentire Pan- approach to allochthonous terranes: a Pan-Africancase study. PhilosophicalTransactions of theRoyal Society, London, A331, African/Mozambique Belt. 533-548. -, HAWKESWORTH,C. J. & RIES, A. C. 1984. Crustal evolution in I wouldlike to thank N. B. W. Harrisand F. McDermott for northeast and east Africa from model Nd ages. Nature, W,773-776. comments and discussion. HIRDES,W. & BRINKMAN,K. 1985. The Kabus and Balula serpentinite and 3 September 1990. metagabbro complexes-a dismembered Proterozoic ophiolite in the north-eastern Nuba Mountains, Sudan. Geologische Jahrbuch, B.%, 3-43. References HUSSEIN,I. M,, KRONER,A. & DURR,ST. 1984. Wadi Onib-a dismembered Pan-African ophiolite in the Red Sea Hills of Sudan. Faculty of Earth ABDEL RAHMAN, E.M. 1983. The geology of mafic-ultramaficmasses and Sciences Bulletin, King Abdulaziz Universiry, Jeddah, 6, 320-327. adjacentrocks south of thelngessana igneous complex, Blue Nile KEY, R. M., CHARSLEY,T. J., HACKMAN,B. D., WILKINSON,A. F. & Province, E. Sudan. M. Phil. thesis, Portsmouth Polytechnic. RUNDLE,C. C. 1989.Superimposed upper Proterozoic collision- ABDELSALAM,M. G. 1987. Geology, structure, and tectonics of northeastern controlled orogenies in the Mozambique Orogenic Belt of Kenya. NubaMountains, Sudan, with special emphasb on the El Bitiera area. Precambrian Research, 44, 197-225. MSc thesis, University of Khartoum. KRONER, A.,STERN, R. J., DAWOUD,A. S., COMPSTON,W. & REISCHMANN, - & DAWOUD,A. S. 1991. The late Precambrian Kabus ophiolitic T. 1987. The Pan-African continental marginof northeastern Africa: melange,Nuba Mountains, Sudan, and its bearing on the western Evidence from a geochronological study of granulite at Sabaloka, Sudan. boundary of the NubianShield. Journal of theGeological Society, Earth and Planetary Science Letters, 85,91-104. London, 148, 83-92. MCWILLIAMS,0. 1981. Palaeomagnetic and Precambrian tectonic evolution ALMOND, D. C.& AHMED,F. 1987. Ductile shear zones in the northern Red of Gondwana. In: KRONER,A. (Ed.) Developments in Precambrian SeaHills, Sudan and their implicationfor crustal collision.Geological Geology: Precambrian Plate Tectonics. Elsevier, Amsterdam, New York, Journal, 22, 175-184. 649-87. AL-SHANTI,A. M. S., ABDEL,A. A. & RADAIN,A. A. 1984.Rb-Sr dating MABOKO,M. A. H., BOELRIJK,N. A. I. M., PRIEM, H. N. A. & VERDURMEN,

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/148/3/600/4892200/gsjgs.148.3.0600.pdf by guest on 27 September 2021 606 DISCUSSION

E. A. TH. 1985. Zircon U-Pb and biotite Rb-Sr dating of the Wami River Jabal Ess in northern Saudi Arabia. Nqture, 279, 488-491. granulites, Eastern Granulites, Tanzania: Evidence forapproximately STACEY,J. S. & AGAR,R. A. 1985. U-Pb isotopic evidence for the accretion 715 Ma old granulite-fades metamorphism and final Pan-African cooling of a continental microplatein the Zalmregion of the Saudi Arabian approximately 475 Ma ago. Precambrian Research, 30, 361-378. Shield. Journal of the Geological Society, London, 142, 1189-1203. MOODY,J . D. 1973. Petroleum exploration aspects of wrench-fault tectonics. -& HEDGE,C. E. 1984. Geochronologic and isotopic evidence for early American Association of Petroleurn Geologists, SI, 449-476. Proterozoic crust in the eastern Arabian Shield. Geology, U,310-313. NASSIEF,M. 0..MACDONALD, R. & GAS, I. G. 1984. The Jebel Thunvah STAUFFER,P. H. 1983. Unravelling the mosaic of Palaeozoic crustal blocks in Upper Proterozoic Ophiolite Complex, western Saudi Arabia. Journal of Southeast Asia. Geologische Rundschau, 72, 1061-1080. the Geological Society, London, 141,537-546. STEINER,L. 1987. The Nuba ophiolite and its geological setting. In: MATHEIS, PALLISTER,J. S., COLE,J. C., STOESSER,D. B. & QUICK, J. E.1990. Use and G. & SCHANDELMEIER,H. (eds) Currentresearch inAfrican earth abuse of crustal accretion calculations. Geology, 18, 35-39. sciences. Balkema, Rotherdam, 101-104. -, STACEY,J. F., FISCHER,L. B. & PREMO,W. R. 1988. Precambrian STEW, R. J. 1985. The Najd fault system, Saudi and Arabia and Egypt: A ophiolites of Arabia: geologic settings, U-Pb geochronology, Pb-isotope late Precambrian rift-related transform system? Tectonics, 4, 497-511. characteristics, andimplications for continental accretion. Precambrian -, KRONER,A., MANTON,W. I., REISCHMANN,T., MANSOUR,M. & Research, 38, 1-54. HUSSEIN,I. M. 1989. Geochronology of the late Precambrian Hamisana REYMER,A. & SCHUBERT,G. 1984. Phaneraoic addition rates to the shear zone, Red Sea Hills, Sudan and Egypt. Journal of the Geological continental crust and crustal growth. Tectonics, 3, 63-77. Society, London, 146, 1017-1030. - & - 1986.Rapid growth of some major segments of continental -, NIELSEN,K. C., BEST,E., SULTAN,M., ARVIDSON,R. E. & KRONER, A. Crust. Geology, 14, 299-m. 1990. Orientation of the late Precambrian sutures in the Arabian-Nubian RIES, A. C., SHACKLETON,R. M. & DAWOUD,A. S. 1985. Geochronology, Shield. Geology, 18, 1103-1106. geochemistryand tectonics of northeast Bayuda Desert, north Sudan: STOESER,D. B. & CAMP,V. E. 1985. Pan African microplate accretion of the Implication for the western margin of the late Proterozoic fold belt of Arabian Shield. Geological Society of America Bulletin, %, 817-26. northeast Africa. Precambrian Research, 30,43-62. -, STACEY,J. S., GREENWOOD,W. R. & FISHER,L. B. 1984. UIPb zircon SCHANDELMEIER,H., DARBYSHIRE, D. P. F., HARMS,U. & RICHTER,A. 1988. geochronology of the southern portion of theNabitah mobile belt and The Eastern Craton: Evidence for pre-Pan-African Crust in Pan-African continental collision in the Saudi Arabian Shield. Ministry of northeast Africa west of the Nile. In: EL GABY,S. & GREILING,R. 0. Petroleum and Mineral Resources, Jiddah. Technical Record USGSTr- (Eds) ThePan-African belt of northeast Africa and adjacentareas. 04-5. Frieder Vieweg and Sohn (Wiesbaden), 69-94. TAPWNNIER, P,, PELTLER,G., LE DAIN,A. Y., ARMIJO,R. & COBBOLD, P. SHACKLETON,R. M. 1986. Precambrian collision tectonics in Africa. In: 1982. Propagating extrusion tectonics in Asia: New insights from simple COWARD,M. P. & RIES, A. C. (eds) Collision Tectonics. Geological experiments with Plasticine. Geology, 10, 611-616. Society, London, Special Publication, 19, 329-349. VEARNCOMBE,J. R. 1983. A proposed continental margin in the Precambrian - 1988. Contrasting structural relationships of Proterozoic ophiolites in of western Kenya. Geologische Rundrchau, 72, 663-670. Northeast and East Africa. In: EL-GABY,S. & GREILING,R. 0.(eds) WUST,H. J., TODT,W. & KRONER,A. 1987. Conventional and single grain The Pan-Afncan belt of North-Easf Africa and Adjacent areas. Freidr. zirconages from metasediments and granite clasts from the Eastern Viewing and Sohn, Braunschweig, Wiesbaden, 183-194. Desert of Egypt:evidence for active continental marginevolution in SHANTI, M.& ROOBOL,M. J. 1979. A late Proterozoic ophiolite complex at Pan-African times. Terra Cognita, 7, 333-334.

W. R. CHURCH,Department of Geology, University of Western Ontario, London, Ontario, Canada N6A 5B7. MOHAMEDG. ABDELSALAM & ROBERTJ. STERN,The University of Texas at Dallas, Programs in Geosciences, PO Box 830688, Richardson, Texas, 75083-0688, USA SEIFEM. BERHE,Department of Earth Sciences, The Open University, Walton Hall, Milton Keynes MK7 6AA, UK

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/148/3/600/4892200/gsjgs.148.3.0600.pdf by guest on 27 September 2021