The Mesoproterozoic Kibaride belt (Katanga, SE D.R. Congo)...
Journal of African Earth Sciences 46 (2006) 1–35 www.elsevier.com/locate/jafrearsci
The Mesoproterozoic Kibaride belt (Katanga, SE D.R. Congo) J.W. Kokonyangi Kokonyangi a,*, A.B. Kampunzu b,z, R. Armstrong c, M. Yoshida d, T. Okudaira e, M. Arima a, D.A. Ngulube f a
Geological Institute, Graduate School of Environmental and Natural Sciences, Yokohama National University, Yokohama 240-8501, Japan b University of Botswana, Department of Geology, Private Bag 0022, Gaborone, Botswana c Research School of Earth Sciences, The Australian National University, Canberra, ACT 0200, Australia d Gondwana Institute for Geology and Environment, 147-2 Hashiramoto, Hashimoto 648-0091, Japan e Department of Geosciences, Faculty of Science, Osaka City University, Osaka 558-8585, Japan f ´ dagogique National (IPN), Kinshasa/Binza, Congo Institut Pe´ dagogique
Received 15 February 2004; accepted 15 January 2006 Available online 17 July 2006
Abstract
Five representative key regions from the NE–SW-trending Mesoproterozoic Kibaride belt of SE Congo are described. Although the present database is still insufficient for a definitive reconstruction of the tectonic setting, available data suggest that the investigated areas experienced a similar geological history. The Kibaran Supergroup comprises four major lithostratigraphic units in SE Congo. The Kiaora Group is the oldest unit. It starts with a basal conglomerate which is overlain by siliciclastic rocks deposited in fluviatile and possibly lacustrine environments. The conglomerate is overlain by abundant metapelites (predominantly black schist) with calc-silicate, volcanosedimentary rocks and minor metachert deposited in shallow marine environments. These units are cut by 1.38 Ga granitoids. The overlying Nzilo Group is composed of coarse-grained, siliciclastic metasedimentary rocks including metaconglomerates, quartzites and minor metapelites and metamorphosed ironstones. Frequent herringbone and wavy ripples suggest tidal flat deposition. The maximum depositional sitional age of this group is given by the 1.38 Ga granitoid granitoidss on which it rests disconforma disconformably bly and by detrital zircons zircons from a quartzite quartzite,, which yields a concordant age of 1360 ± 27 Ma. The above two groups are separated by a disconformity along with the matrix-supported Kataba Conglomerate occurs at the base of the Nzilo Group. Higher in the succession, the Hakansson Group is essentially pelitic with minor quartzites. At the top of the succession, the Lubudi Group is made of (stromatolitic) carbonates, black schists and minor black quartzites and is inferred to record shallow marine deposition. All these metasedimentary rocks were deposited before the emplacement of ca. 1.0–0.95 Ga tin granites and are older than 1.08 Ga, which is the age of the climax of Kibaran deformation in the Mitwaba area. Two major deformational events have been recognized in the study areas. The earliest (D 1) is characterized by ENE-trending asymmetric folds and thrusts showing N to NNW transport directions. These structures occur in the Kiaora Group and predate the deposition of the Nzilo, Hakansson and Lubudi Groups. The second deformation (D 2) marks the climax of the Kibaran orogeny and affects all sedimentary units. It is defined by NW-verging mesoscopic and macroscopic isoclinal folds (F 2) and reverse faults parallel to D2 planar fabrics. M2 metamorphism is characterized by medium-pressure/medium-temperature (MP/MT) mineral parageneses, with preliminary data indicating peak P-T conditions between 740–780 C and 6–6.5 Kb. U–Pb dating of metamorphic zircon in older orthogneisses in the Mitwaba region tentatively constrains the timing of M 2 metamorphism at 1079 ± 14 Ma. The Kiaora Group was intruded by widespread arc-related gabbro-diorite and ca. 1.38 Ga syn-D 1 calcalkaline and strongly peraluminous granitic plutons similar to those documented in the Lachlan and Hercynian belts. Late to post-kinematic granites and related pegmatites and greisens hosting tin-group ore deposits were emplaced at 1.0–0.95 Ga and exhibit geochemical similarities with SE Asian collisional granites. The Kibaran orogenic system was active between 1.4–1.38 (accretionary stage) and 1.0–0.95 Ga (continental collision and post-orogenic exhumation), but the Kiaora Group sedimentary rocks were deposited prior to 1.4–1.38 Ga, which is the igneous crystallization age for the syn-D 1 Kibaran orthogneisses intruding them. Sedimentological data broadly indicate that the Kiaora Group was deposited in shallow marine environments, during the rift-drift stage of the evolving Kibaran basin while the post-D 1 Nzilo
*
Corresponding author. Tel.: +81 45 333 1416.
[email protected] (J.W. Kokonyangi). E-mail address:
[email protected] (J.W.
z
Deceased November 2004.
1464-343X/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.jafrearsci.2006.01.017
2
J.W. Kokonyangi Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
Group was deposited in an intra-arc environment, although more modern data are required before a definite conclusion can be reached. Available structural, sedimentological, geochronological and petrological data support a convergent margin setting for the Kibaride belt and are inconsistent with an intracontinental, rift model. 2006
Elsevier Ltd. All rights reserved.
Keywords: Congo; Kibaride belt; Mesoproterozoic orogeny; Subduction; Continental tectonics
1. Introduction
Available geological, tectonothermal, geochronological and paleomagnetic data from 1.4 to 1.0 Ga orogens and adjace adjacent nt craton cratonic ic blocks blocks around around the world world sugges suggestt that that the Mesoproterozoic was dominated by accretionary and collisional orogenic processes that culminated in the assembly of the Rodinia supercontinent (e.g. Dalziel (e.g. Dalziel et al., 2000
and refere reference ncess therein therein). ). In central central and southe southern rn Africa Africa (Fig. 1, 1, upper inset), Mesoproterozoic orogenic events led to the development of the Kibaran orogenic system (Rob( Robert, 1931; Cahen, 1954; Cahen et al., 1984; Thomas et al., 1994). 1994 ). The segment of the Kibaran system exposed in SE Democratic Republic of Congo (hereafter Congo) is known as the Kibaride belt, and has been defined as the type-area of the the Kibara Kibaran n orog orogen enic ic syst system em of Afric Africaa (e.g. (e.g. Cahen
Fig. 1. The Kibaride belt of SE Congo. Inset shows the distribution of the Kibaran orogenic system in sub-equatorial Africa. C = Congo craton, B = Bangweulu block, BT = Buganda-Toro, K = Kalahari craton, T = Tanzania craton. Compiled from Cahen (1954) and Lepersonne (1974). (1974).
J.W. Kokonyangi Kokonyangi et al. / Journal of African African Earth Sciences 46 (2006) (2006) 1–35
et al., 1984; Kampunzu et al., 1986). 1986). In fact, the original name given to this Mesoproterozoic belt in French was ‘‘syste` me du Kibarien’’. With the progress of mapping in central central and eastern eastern Africa, Africa, it was realize realized d that that severa severall orog orogen enic ic belt beltss in this this regi region on were were corre correla lati tive ve of the the ‘‘sys ‘‘syste te` me du Kibarien’’, including the Burundian, Karagwe–Ankole, and Irumides belts. In the 1960s/early 1970s therefore it was commonplace to use the name Kibaran for the whole Mesopr Mesoprote oteroz rozoic oic orogen orogenic ic system system of Africa Africa.. In order order to avoid potential potential confusion, confusion, Cahen and others introduced introduced in the 1970s, in the Anglophone literature, the name Kibarides for the type area of the Kibaran belt in central Africa and Irumides for the Kibaran segment exposed in Zambia. Zambia. The geographic position of the Kibaride belt makes it a critical area for any correlation between the northern (e.g. Burundian belt of Rwanda, Burundi and northern Congo, Karagwe Karagwe–An –Ankole kolean an belt belt of Uganda Uganda and Tanzan Tanzania) ia) and southern southern (e.g. Namaqua–Sin Namaqua–Sinclair–R clair–Rehobot ehoboth h of Namibia, Namibia, Namaqua–Natal of South Africa) segments of the Mesoproterozoic orogenic system of Africa. The Kibaride orogenic belt is preserved as a NE-trending zone of deformation, metamorphism and magmatism over over 600 km long long and 100 100–30 –3000 km wide, wide, from from Kongol Kongolo o (NE) to Nzilo (SW) in the Katanga province of Congo (Fig. 1). 1). This belt is composed composed of supracrustal supracrustal sedimentary sedimentary and volcano-sedi volcano-sedimentar mentaryy assemblages assemblages intruded intruded by numerous plutons. Along its eastern margin, the Kibaride belt is in tectonic contact with the Palaeoproterozoic Bangweulu block, which was accreted to the Tanzania craton during the Palaeoprote Palaeoproterozoic rozoic (Schandelmeier, ( Schandelmeier, 1981; Andersen and Unrug, 1984; Kabengele et al., 1991). 1991 ). Along its western margin, the Kibaran belt has both unconformable (Kibaran basal conglomerate exposed in Nzilo region; Mortel-
3
mans, 195 mans, 19511) and tecton tectonic ic contac contacts ts (see (see Bukama Bukama region region,, Lataillade and Bielle, 1972) 1972) with the Archaean/Palaeoproterozoic Congo Craton. There are no geological similarities between between Archaean–Pa Archaean–Palaeopr laeoproteroz oterozoic oic rocks exposed exposed on either side of the Kibaride belt. To the south of the Kibarides (Fig. (Fig. 1) 1) occur the rocks defining the Neoproterozoic Katangan belt, which have both unconformable and tectonic contacts with the Kibaride belt. This contribution presents an overview of the geological, geochr geochrono onolog logical ical and tecton tectonic ic evolut evolution ion of the Kibarid Kibaridee belt belt on the the basi basiss of data data from from five five repr repres esen enta tati tive ve key key area areas: s: MitwMitwaba, Mwanza, Bukama, Bia and Nzilo (Fig. (Fig. 1) 1) in the southwestern sector of the Kibaran belt in SE Congo. Existing data data are are of varia variabl blee quali quality ty and and insu insuffic fficie ient nt to prov provid idee a comcomprehensive geological and geochronological reconstruction of the evolution evolution of the entire belt. However, the lithostratilithostratigraphic, graphic, structural structural and magmatic magmatic similarities similarities along with geological continuity of units exposed in these five adjacent regions allow preliminary regional correlations. The Mitwaba region has been used as the reference region because it is the best documented area of the Kibaride belt (e.g. Kam(e.g. Kampunzu et al., 1986; Kokonyangi et al., 2001a,b, 2004). 2004 ). 2. Mitwaba
The Kibara Mountains, the type locality of the Mesoprotero proterozoi zoicc Kibara Kibaran n orogen orogenic ic system system as first first defined defined by Robert (1931), (1931), occur in the Mitwaba region. They define a ca. ca. 100100-km km-lo -long ng,, NE-tre NE-trend ndin ingg moun mounta taino inous us regi region on underlain underlain by greenschist greenschist to amphibolite amphibolite facies metasedimetasedimentary sequences intruded by abundant felsic and mafic plutonic bodies (Fig. (Fig. 2). 2). Excellent exposures allow documentation of geological structures along the NW–SE to
Fig. 2. General General geological geological map of the Mitwaba region. region. Modified after Cahen (1954), Van de Steen (1953b), (1953b), Kokonyangi et al. (2001b), Kampunzu et al. (unpublished (unpublished data). A–B represents represents the cross section shown in Fig. in Fig. 10a. 10a.
4
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
NNW–SSE Mitwaba–Kapia road, which runs across strike and allows a detailed lithostratigraphic succession of the Kibaran metasedimentary units to be established in this area (e.g. Cahen, 1939, 1954; De Magne´ e, 1935a,b; Grosemans, 1948; Mortelmans, 1951; Raynaud, 1952a; Van de Steen, 1953a, 1959; Kampunzu et al., 1986; Kokonyangi, 2001; Kokonyangi et al., 2001a,b, 2002). A schematic lithostratigraphic column for the Kibaran metasedimentary
units in this area is shown in Fig. 3. It is important to note at this stage that due to the grade of metamorphism and the intensity of ductile deformation in some locations, all the thicknesses reported in this paper are structural rather than stratigraphic thicknesses. Three main units can be recognized within the supracrustal sedimentary packages (Table 1): (1) the oldest has a structural thickness up to 4300 m and is composed of muscovite-chlorite schists,
Fig. 3. Lithostratigraphic columns of the five representative key-regions within the Kibaride belt. Main data sources are Cahen (1954), Lepersonne (1974), Byamungu et al. (1979), Cahen et al. (1984) and Kokonyangi (2001).
6
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
chloritoid schists, garnet-biotite schists, subordinate psammitic schists with quartzite intercalations and garnet-biotite-staurolite and garnet-biotite-sillimanite gneisses. Most of the rocks forming this oldest unit in the Kibarides, except the garnet-biotite sillimanite gneisses, are black to dark-green in color, have well developed laminations and contain pyrite. Cahen (1939) and Mortelmans (1951) reported thick lenses/layers of calc-silicate rocks within this unit. This sequence was referred to as the K1 unit by Cahen (1954) and Van de Steen (1959), the Kiaora Group by Cahen et al. (1984) and the Mitwaba Group by Kokonyangi et al. (2001a). In this review, we use the name Kiaora Group as this reflects the location of the best exposures. Within the Kiaora Group, Mortelmans (1951) reported wavy ripples in quartzites intercalated within chloritoid schists exposed for about 6 km along Mitwaba–Kapia Road. The base of the Kiaora Group is not exposed in the Mitwaba area. Field relations indicate that the group is older than the 1.4–1.38 Ga orthogneisses intruding it (Kokonyangi et al., 2004, 2005).
The Kataba Conglomerate overlies the Kiaora Group and spatially associated orthogneisses in the Mitwaba region (e.g. Figs. 3 and 4a). The conglomerate contains detrital zircons as young as 1329 ± 32 Ma (Kokonyangi and Armstrong, unpublished data). It is the basal unit of the Nzilo Group, which has a structural thickness between 3900 and 5500 m (De Magne´e, 1935a; Mortelmans, 1951; Cahen, 1954; Kokonyangi, 2001; Kokonyangi et al., 2001a). The Kataba Conglomerate is 100–300 m thick, matrix-supported, poorly sorted to unsorted and contains ellipsoidal clasts of black schist, quartzite, metapelites, quartz, mafic rocks and granite, set in a black schistose matrix. A disconformity separates this conglomerate from the underlying chloritoid schist of the Kiaora Group (De Magne´e, 1935a; Mortelmans, 1951; Cahen, 1954; Kokonyangi et al., 2001a, 2002). The conglomerate is characterized by an upward decrease in clast size and abundance. It is overlain by coarse-grained quartzites, psammitic schists, conglomerate and minor metapelite intercalations and Fe–Ti-oxide-bearing quartzites (Mortelmans, 1951; Cahen,
Fig. 4. (a) Geological map of the Mitwaba area showing the distribution of the 1.38 Ga felsic and mafic bodies relative to the Kiaora and Nzilo Groups. Modified after Cahen (1954), Van de Steen (1953b), Kampunzu et al. (unpublished data) and Kokonyangi et al. (2004). Note that the Kataba Conglomerate at the base of the Nzilo Group disconformably overlies the Kiaora Group and adjacent orthogneisses. C–D represents the cross section shown in Fig. 10b. (b) Contact between an undeformed, mineralized pegmatite and Kisele monzogranite gneiss. The D2-related fabrics are well developed within the Kisele monzogranite gneiss. (c) Quartz vein cutting the Kisele monzogranite and folded during D2. Hammer for scale. (d) Tin mineralization (black cassiterite crystals) within the undeformed pegmatite shown in (b).
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
7
Fig. 4 (continued )
1954; Kokonyangi et al., 2001a). Sedimentary structures include planar bedding, bi-directional (herringbone) cross bedding, graded bedding, symmetrical ripple marks (wavy ripples) and erosional channels (Mortelmans, 1951; Cahen, 1954). Most of the rocks forming this group are whitecream or pinkish or deep-plum in color, depending on the amount of iron oxide, commonly hematite but rarely magnetite (Mortelmans, 1951; Cahen, 1954). This sedimentary succession corresponds to the Nzilo Group of Mortelmans (1951), the K2 unit of Cahen (1954) and Van de Steen (1959), and the Lufira Group of Cahen and Snelling (1966). The name Nzilo Group, which has priority, is adopted in this review, and the best outcrops of this succession are exposed in the Nzilo area described later in this paper. The youngest age of 1360 ± 27 Ma obtained from a pink quartzite in the Mitwaba area sets the maximum time of deposition of the Nzilo Group in this area. The Nzilo Group in the Mitwaba region is overlain by a succession of slate, black to greenish metapelitic schists and minor quartzite with psammitic intercalations, forming the K3 Goup of Cahen (1954) and Van de Steen (1959) and the Hakansson Group of Cahen et al. (1984). The term Hakansson Group is adopted in this review. The contact between the Nzilo and Hakansson Groups is gradational (Mortelmans, 1951). The Kiaora Group is intruded by gabbro-diorites and strongly peraluminous (SP) granitoids, including the Kifinga–Kisele batholith and coeval smaller bodies, such as the Nyangwa pluton and Kungwe–Kalumengongo stock (Kokonyangi et al., 2004, Fig. 2). The composite Kifinga–
Kisele batholith is the largest granitic body in the region and consists of the Kabonvia granodioritic augen gneiss, Kisele monzogranite gneiss and Fwifwi foliated leucomonzogranite (Kokonyangi et al., 2001b, 2004). The granitoids carry two foliations developed during the two deformational events identified in the region and contain xenoliths of country rocks and abundant, foliated, micasrich restitic enclaves (Kokonyangi et al., 2004, Fig. 4b and c). Both micaceous enclaves and xenoliths are elliptical with long axes aligned parallel to the trend of the oldest foliation. Quartz veins within the Kisele foliated monzogranite gneiss and Kabonvia granodioritic augen gneiss are folded with axial planes trending NE–SW (Fig. 4c). Mafic/intermediate igneous bodies intruding the Kiaora Group sedimentary rocks in the Mitwaba area (Figs. 2 and 4a) include the Lwabwe metagabbro/metadolerite and the Kidilo orthoamphibolites (Van de Steen, 1953a,b; Cahen, 1954; Kampunzu et al., 1986; Kokonyangi et al., 2005). Both deformation fabrics identified in the Mitwaba region affect these mafic bodies. Kokonyangi et al. (2004, 2005) reported a contact aureole in the mafic bodies close to the contact with the Kifinga–Kisele batholith. Grosemans (1948) suggested that the Lwabwe mafic body and the Kifinga–Kisele granitic batholith were coeval. The structural patterns along with field relationships suggest that the granitoids and closely associated mafic bodies intruding the Kiaora Group in the Mitwaba area are broadly coeval and syn-kinematic with the earliest deformation event (Grosemans, 1948; Cahen et al., 1967; Kampunzu et al.,
8
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
1986; Kokonyangi et al., 2001b, 2004). Nowhere in the Mitwaba region have the orthogneisses intruded the Nzilo Group. Near Mwadianvula Falls (Fig. 2), the orthogneisses are disconformably overlain by the Kataba Conglomerate at the base of the Nzilo Group. Preliminary conventional U–Pb zircon isotopic analyses of the Bukena granitic gneiss, representing the northeastern extension of the Nyangwa pluton (Fig. 1), yielded a date of 1306 ± 35 Ma (Cahen et al., 1967, 1972). Recent U–Pb SHRIMP zircon investigations yielded the following igneous crystallization ages for the Mitwaba granitoids (Kokonyangi et al., 2004): 1386.3 ± 8.1 Ma for the Kisele monzogranite gneiss, 1385.5 ± 7.4 Ma for the Kabonvia granodioritic augen gneiss, 1383.3 ± 4.8 Ma for the Nyangwa monzogranite (southern termination of the Bukena massif, Fig. 1), 1377 ± 10 Ma for the Kungwe–Kalumengongo monzogranite (including the Shombio pluton, Fig. 4a) and 1372 ± 9.6 Ma for the Fwifwi foliated leucomonzogranite, including the Mandwe facies (Table 2). Parts of the Mitwaba granitoids have previously yielded younger conventional U–Pb dates, including the Shombio and Mandwe plutons (Fig. 4a) dated at 1050 ± 50 Ma (Eberhardt et al., 1956; Ledent et al., 1956) and the Bukena igneous body (the northern extension of the Nyangwa pluton; Fig. 1) dated at 1306 ± 35 Ma (Cahen et al., 1967, 1972). However, field relations suggest that these plutons are all part of the same magmatic assemblage emplaced at 1381 ± 8 Ma (Kokonyangi et al., 2004). In addition, all the plutons, including those dated by Eberhardt et al. (1956) and Ledent et al. (1956) and Cahen et al. (1967, 1972) are affected by both Kibaran deformational events. Metamorphic zircon from the strongly deformed amphibolite facies Kisele monzogranite gneiss yielded a U–Pb date of 1079 ± 14 Ma, which is indistinguishable from the date of 1050 ± 50 obtained by Eberhardt et al. (1956) and Ledent et al. (1956) (Table 2). This date has been interpreted to record the time of the climax of D 2 regional deformation and related amphibolite facies metamorphism in the Mitwaba area (Kokonyangi et al., 2004). Zircon dating of the Mitwaba mafic intrusive bodies yielded a 207Pb/206Pb date of 1417 ± 1.7 Ma (97% concordant) and an upper intercept age of 1376 ± 13 Ma (MSWD = 0.046) for the Kidilo orthoamphibolites and Lwabwe metagabbro/metadolerite, respectively (Table 2, Kokonyangi et al., 2005). Because the data from the Kidilo orthoamphibolites that yielded the date of 1417 ± 1.7 Ma are 3% discordant, the upper intercept of 1376 ± 13 obtained for the Lwabwe body is taken to represent the best estimate of the crystallization age of the Mitwaba mafic igneous rocks. Zircon crystallization ages therefore support broadly coeval emplacement of felsic and mafic magmas in the Mitwaba area, as previously suggested by Grosemans (1948). Igneous rocks intruding the Nzilo Group sedimentary sequences are rare in the Mitwaba area. The Kisandji pluton (Fig. 2) contains elliptical microgranular mafic enclaves
(pyroxene + amphiboles + biotite + quartz). This intrusion and the Nzilo Group post-date the older deformation event in the Mitwaba region and are only affected by the NE-trending younger foliation. No geochronological and petrological data are available for this pluton. Tin granites and related mineralized aplitic and pegmatitic bodies intrude both the 1.38 Ga orthogneisses and the complete Kibaran Supergoup metasedimentary succession. They host tin, colombite–tantalite, tungsten and tourmaline, and the mineralized bodies are free of pervasive solid-state deformation fabrics (Fig. 4d), indicating that the tin granites and related aplites and pegmatites are late to post-kinematic with respect to the Kibaran deformation events. 3. Mwanza region
The Mwanza region as described in this paper comprises the following areas: Mwanza, Kikonja, Ankoro, Bukena, Manono and the area around the Luvidjo River ( Fig. 1). For the sake of simplicity in this paper, the boundary between the Mwanza (NW) and the Mitwaba (SE) regions is taken to be the Upemba Valley, which is filled by Phanerozoic sedimentary rocks. Previous research includes reconnaissance mapping and lithostratigraphic investigations by a number of workers (Matthieu, 1912; Van Aubel, 1928; Karpoff and Karpoff, 1938; Mortelmans, 1947; Aderca, 1950; Landa et al., 1950; Thoreau, 1950; Raynaud, 1952b; Van de Steen, 1950a, 1953b; Cahen, 1954; Lonchampt, 1971; Lonchampt and Heinry, 1972; Lepersonne, 1974; Bassot and Morio, 1989; Gu¨nther and Ngulube, 1992; Ngulube, 1994) and preliminary, conventional U– Pb and Rb/Sr whole-rock and/or mineral dating (Cahen et al., 1967, 1972, 1984). The Kibaran lithostratigraphic succession in the region comprises four lithostratigraphic units (Fig. 3 and Table 1). The oldest unit comprises chlorite schists, mica schists, sillimanite- and/or kyanite-bearing gneisses and amphibole-clinopyroxene-bearing calc-silicate rocks. This succession corresponds to the Kiaora Group (K 1) of Cahen et al. (1984). Cahen and Lepersonne (1967) reported dark carbonate rocks within the Kiaora Group in this region. These carbonates are separated from the younger units either by the Lubweyi Conglomerate (Cahen, 1954) or by volcanic sequences including metabasalts and metarhyolites (Van de Walle, 1959). Lonchampt (1971) and Lonchampt and Heinry (1972) in addition described para-amphibolites interbedded with a volcano-sedimentary breccia along the Mwimbi River and suggested that these rocks are part of the older sedimentary sequences (Kiaora). Above these Kiaora Group rocks are >1500 m of pinkish quartzite sequences alternating with minor metapelites and subordinate ironstones and conglomerate intercalations, possibly representing the Nzilo (K2 of Cahen et al., 1984) Group. These rocks are overlain by ±1000 m of metapelites, slates and subordinate quartzite (K 3 of Cahen et al., 1984). The uppermost unit consists of marbles, calc-silicates and
10
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
minor black to purplish schists and quartzites (K 4 of Cahen et al., 1984, Fig. 3 and Table 1). Large granite and gabbro-diorite bodies intrude the Kiaora Group supracrustal metasedimentary rocks in the Mwanza region. The main granitic intrusions include the Mwanza, Bukena, Luvidjo and Manono plutons and related quartz veins, aplite and pegmatite bodies (Figs. 1 and 5). Metadiorite-metagabbro-metadolerite bodies containing mainly plagioclase-amphibole ± pyroxene intrude the Kiaora Group metasedimentary sequences around Mwanza (e.g. Lonchampt and Heinry, 1972) and Manono (e.g. Landa et al., 1950; Ngulube, 1994). Orthogneisses exposed in the Manono area contain mafic enclaves and carry two foliations (an older one trending N50–70E cut or crenulated by the younger one oriented N30– 40E, e.g. Figs. 5 and 6a). The orthogneisses exhibit intrusive relationships with the Kiaora Group metapelites, which include chlorite schists and mica schists (Ngulube, 1994). In general, three main intrusive bodies can be dis-
tinguished in the Mwanza region (Cahen, 1954; Van de Walle, 1959; Cahen et al., 1967; Lonchampt, 1971; Lonchampt and Heinry, 1972; Pasteels, 1971; Lepersonne, 1974; Gu¨nther and Ngulube, 1992; Ngulube, 1994): Type (1) hornblende and/or biotite orthogneisses containing titanite as the main accessory mineral; type (2) two-mica granitic orthogneisses and type (3) two-mica tin granites and associated tin-bearing quartz, pegmatites and aplites. Types (1) and (2) orthogneisses carry both Kibaran deformation fabrics and intrude the Kiaora Group (K 1) without cutting the younger overlying groups. They are commonly associated with mafic to intermediate igneous rocks, e.g. the arc-shaped mafic body of Lonchampt and Heinry (1972) surrounding the Mwanza granitic batholith (not shown in Fig. 1 due to the lack of detailed maps for this region). Van de Walle (1959) indicated that the types (1) and (2) orthogneisses are syn-kinematic with respect to earliest Kibaran deformation in the Mwanza region and, by analogy with the Mitwaba region, their igneous
Fig. 5. Simplified geological map of the Manono area showing the distribution of granitic gneisses, tin granites and pegmatites (Source: Thoreau, 1950; Landa et al., 1950; Ngulube, 1994).
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
11
Fig. 6. (a) A NE–SW-trending reverse fault affecting a mafic enclave within the Manono syn-D1 orthogneiss. (b) A NW–SE cross-section of the Kahungu open pit (Manono) showing the relationships between Kiaora Group, orthoamphibolites, Sn–Nb–Ta pegmatite and quartz veins. The pegmatite lacks pervasive solid-state foliation and exhibits sharp intrusive relationships with the adjacent, older, NE-trending mafic igneous rocks, which were affected by contact metamorphism. (c) Sketch drawn from a field photograph of Landa et al. (1950) showing deformed mafic rock xenoliths within the undeformed Manono Sn–Nb–Ta pegmatite. Kinematic indicators suggest a sinistral sense of shear. See text for details.
crystallization ages could be ca. 1.38 Ga. The Lovoi River stock and the Luvidjo River granitic plutons (Fig. 1) could belong to the types (1) and/or type (2) granitoids, although Lonchampt (1971) and Lonchampt and Heinry (1972), regarded these plutons as pre-Kibaran, based on mineralogical similarities with the Paleoproterozoic granitoids exposed near Kamina. In contrast, Lepersonne (1974) considered them to be Kibaran igneous bodies on the basis of field relationships. Precise geochronology is required in order to resolve this controversy. Type (3) tin granites and pegmatites intrude all Kibaran sedimentary units and cut across types (1) and (2) orthogneisses in the Mwanza region (e.g. Karpoff and Karpoff, 1938; Grosemans, 1946a; Cahen, 1954; Cahen et al., 1967, 1984). The Manono pegmatite is the biggest tin-bearing pegmatitic body in the Kibaride belt (Fig. 5). Exposed for >14 km along strike with an average width of ca. 400 m, this pegmatite is among the largest pegmatitic bodies in the world (e.g. Ngulube, 1994). It hosts fabulous tin-group ores, including cassiterite, colombite–tantalite, spodumene and thoreaulite (Landa et al., 1950; Thoreau, 1950). Previous work on this pegmatite includes reconnaissance and descriptive geological studies (Landa et al., 1950; Thoreau, 1950; Bassot and Morio, 1989) and preliminary Rb–Sr and K–Ar geochronological investigations (Cahen et al., 1967, 1972, 1984). A comprehensive review of previous work and a modern petrological study was undertaken by Ngulube (1994), whose results are summarized below. The Manono pegmatite is elongated northeast, parallel to the foliation in the adjacent micaschist (e.g. Landa et al., 1950). It intrudes the Kiaora Group metasedimentary rocks (Figs. 5 and 6b), which include micaschists, garnetstaurolite gneisses and subordinate quartzites in the Man-
ono area. The northeastern segment of the pegmatite body (Fig. 5) exhibits sharp intrusive relationships with the adjacent, older, NE-trending mafic igneous rocks, which were affected by contact metamorphism during the emplacement of the pegmatite (Landa et al., 1950; Thoreau, 1950; Van de Steen, 1953b). The pegmatite contains enclaves of the adjacent mafic igneous rocks and xenoliths of the metasedimentary host rocks (Fig. 6c). Pervasive solid-state fabrics are absent within the pegmatite. However, Landa et al. (1950) described a weak foliation developed at the margins of the pegmatite near its contact with the micaschist of the Kiaora Group, and locally the pegmatite margins exhibit a gneissic facies. Based on field and structural observations within both the host rocks and the pegmatite margins, Ngulube (1994) inferred these marginal gneissic fabrics to record stress generated during pegmatite emplacement. Detailed field observations within the pegmatite are required in order to draw a definite conclusion. The oldest orthogneisses exposed in the Manono area contain mafic enclaves and some of them are affected by reverse faults (Fig. 6a). Similarly, mafic xenoliths observed near the contact between the orthoamphibolites and the pegmatite are sheared and mapping indicates a sinistral sense of movement (Fig. 6c). Geochronological data are scanty in the Mwanza region (Table 2). Cahen et al. (1967, 1984) reported a conventional U–Pb zircon age of 1306 ± 35 Ma and Rb/Sr whole rock and mineral isochron of 1329 ± 55 Ma for the Bukena granitic gneiss. The same authors obtained the following whole-rock Rb–Sr isochron dates in the Mwanza region: 1324 ± 71 Ma for the Mwanza granitic gneiss, 977 ± 18 Ma for the tin granites, 925 ± 28 for the Manono Sn–Nb–Ta–Be–Li-bearing pegmatites and 912 ± 30 Ma for the Mwanza pegmatite (Table 2).
12
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
4. Bukama
The Bukama region (Fig. 1) includes localities scattered between Lubudi and Kamina, including the Hakansson Mountains. The limited amount of data available from this region come from Cahen and Mortelmans (1946), Mortelmans (1947, 1948), Dumont (1950, 1952), Van de Steen (1959), Cahen (1954), Lonchampt (1971), Lonchampt and Heinry (1972) and Lataillade and Bielle (1972). There are no modern studies covering this region, which is dominated by metasedimentary rocks. Minor granitic stocks are exposed south of Bukama (e.g. Van de Steen, 1950b). The Archaean to Palaeoproterozoic Congo craton is exposed to the west, and the Kibaran rocks in the Bia Mountains to the east (Fig. 1). The Kibaride belt rocks and those of the Congo craton are partly covered by undeformed Neoproterozoic and Phanerozoic sedimentary rocks (Fig. 7). Where exposed, the rocks of the Congo craton are composed of ortho and paragneisses intruded by gabbroic and doleritic dykes of unknown age. This basement is in tectonic contact with the Kibaran Supergroup sedimentary succession. The contact is presumably a thrust, but kinematics of the shear zone have not been properly documented (e.g. Lataillade and Bielle, 1972).
The lower part of the Kibaran metasedimentary succession in the Bukama region (Table 1) comprises a 4000–4500-m-thick metamorphic complex including gneisses, phyllites, talc-schists, chloritoid schists, biotite schists and subordinate fine grained quartzite and ironstone intercalations. Above this, a 2000–2200-m-thick package of fine- to coarse-grained quartzites and subordinate Fe–Ti-oxide-bearing quartzites and conglomerates is present. This succession is capped by 1000–1300-m-thick amygdaloidal basaltic and rhyolitic lavas (Mortelmans, 1947; Dumont, 1950, 1952; Cahen, 1954; Lonchampt, 1971; Lonchampt and Heinry, 1972). At the confluence of the Lubudi and Lutembwe Rivers (not shown in Fig. 7 due to the scale), the sedimentary and volcanic sequences are thrust onto the Congo craton (Lataillade and Bielle, 1972). In the Hakansson Mountains and Lubudi areas (parts of the Bukama region), the amygdaloidal basalts and rhyolites are overlain by ±400–500 m of graphitic black schist and metapelites, which are capped by a 1000–1300-m-thick succession of marbles, calc-silicates and stromatolitic marbles with minor black schists corresponding to the K 3 unit of Van de Steen (1959), K3 –K4 of Cahen (1954) and Lubudi Group of Cahen et al. (1984) (Fig. 3 and Table 1).
´ologiques et Minie`res (BRGM, France, Orlean). Fig. 7. Geological map of the Bukama area. Data source: Bureau de Recherches Ge
13
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
No geochronological data are available on the Bukama volcanic rocks. However, field relationships indicate that the volcanic rocks overlie the lower units along an erosional surface (Mortelmans, 1947). These volcanic rocks were originally described as post-Kibaran in age (e.g. Mortelmans, 1947; Dumont, 1950, 1952). However, they show the same deformation fabrics and metamorphic history as observed within the Kibaran metasedimentary rocks, thus cannot represent post-Kibaran metavolcanic rocks (Mortelmans, 1947; Van de Steen, 1950b; Cahen, 1954; Lonchampt, 1971; Lonchampt and Heinry, 1972). South of Bukama (Fig. 7), a Kibaran terrain known as the Kikosa metasedimentary complex forms a basement inlier within the Neoproterozoic sedimentary cover of the Katangan Supergroup. The Kikosa complex is composed of marbles, calc-silicates and stromatolitic marbles with an exposed thickness exceeding 700 m (Grosemans, 1946b; Mortelmans, 1947; Cahen, 1954). The stratigraphic position of the Kikosa metasedimentary complex within the Kibaran Supergroup is controversial because this metasedimentary complex was originally thought to represent Neoproterozoic rocks given that it is surrounded by the Katangan basal Conglomerate (e.g. Grosemans, 1946b; Mortelmans, 1947; Cahen, 1954). The metasedimentary complex shows the same regional structural patterns as the Kibaran metasedimentary rocks, and like them is affected by the two Kibaran deformation events (Cahen, 1954; Cahen et al., 1984). Furthermore, it is intruded and metamorphosed by the oldest Kibaran granitoids (e.g. the Bakalenge Mountains granitic gneisses), which are similar to the early Kibaran orthogneisses described elsewhere in the Kibaride belt (Van de Steen, 1950b), and which yielded U–Pb zircon crystallization ages of 1.38 Ga in the Mitwaba region. These observations support the conclusion of Mortelmans (1947), Cahen (1954) and Cahen et al. (1984) that the Kikosa carbonates are part of the Kiaora Group. Mesoproterozoic orthogneisses and mafic intrusive bodies exposed in this region were suggested to be the source of heat for the generation of the amphibolite facies metamorphism recorded in the adjacent metasedimentary rocks (e.g. Cahen, 1954; Lataillade and Bielle, 1972). Mesoproterozoic orthogneisses and mafic rocks are not well exposed in the Bukama region. They include the Bakalenge orthogneisses to the southeast of Bukama, which intrude the Kikosa metasedimentary complex (Fig. 7). There are no geochronological data available for the igneous rocks exposed in this region.
and their host rocks in the Bia region generally trend NE–SW, parallel to the Kibaride structural grain. The following lithotectonic units can be defined from existing data (from the youngest to the oldest): (1) 1700-m-thick slate and purplish blue quartz–phyllites alternating with minor laminated quartzites, sometimes with metaconglomeratic horizons, with a conglomeratic quartzite at the base; (2) >2500–3000-m-thick quartzites, laminated quartzites with cross-bedding and ripple marks, metaconglomerates, Fe– Ti-oxide-bearing quartzites and subordinate metapelitic intercalations; (3) 3000–4000-m-thick black schist alternating with finely bedded black quartzite, finely laminated phyllites and metapelites and locally thick lenses of calc-silicates, marbles and para-amphibolites (Fig. 3; Cahen and Mortelmans, 1940, 1946; Beugnies, 1950; Cahen, 1954). Intrusive rocks are widespread and dominated by the Bia granitic body and its southern extension, known as the Kalule granitic gneiss, and a granitic pluton exposed NE of the Bia Mountains (Fig. 1). The following individual plutons were described by Beugnies (1950), Cahen (1954) and Cahen et al. (1967): gabbro, hornblende–biotite orthogneisses, the Kakongwe biotite orthogneisses and two-mica granitic gneiss. The granitic gneisses and gabbros intrude the Kiaora Group, including a carbonate sequence equivalent to the Kikosa complex and were affected by the two Kibaran deformation events, showing a stronger mylonitic foliation near the margins and weaker foliations in the centre of the intrusive bodies. The youngest intrusions are late to post-kinematic with respect to the youngest Kibaran deformation and include two-mica + garnet microgranitic and aplitic bodies, the Sofwe and Shienzi pegmatites and related quartz veins (Beugnies, 1950). Beugnies (1950) indicated that the two-mica granites, aplites and pegmatites intrude the older orthogneisses and all the Kibaran metasedimentary units in the region. Two generations of doleritic sills and dykes were identified. The older intrusions cut the Kiaora Group metasedimentary units and have been folded and metamorphosed during the Kibaran orogeny. The younger generation occurs mainly along NEtrending fractures and joints (Mortelmans, 1947, 1948; Cahen, 1954) and is undeformed. There are no reliable geochronological data on intrusive bodies exposed in this region. The following whole-rock Rb–Sr dates were reported by Cahen et al. (1967, 1984) for the Bia granitoids: 1324 ± 71 Ma for orthogneisses, 966 ± 21 Ma for tin-bearing granites and 953 ± 29 Ma for the tin-barren Sofwe and Shienzi pegmatites (Table 2).
6. Nzilo 5. Bia
This region is the southwestern extension of the Kibara Mountains (Fig. 1). Previous studies covered the regional geological framework and defined the Kibaran lithostratigraphic column and regional structural trends in this region (Cornet, 1897; Asselberghs, 1938; Cahen and Mortelmans, 1940, 1946; Beugnies, 1950). The orthogneisses
Located north of Kolwezi (Fig. 1), the Nzilo region hosts the south westernmost exposures of the Kibaride belt in Congo. Metasedimentary rocks are widespread in this region, whereas igneous rocks are scarce (Figs. 1 and 8a). The Kibaran rocks in the Nzilo region are bounded to the west by the Archaean to Paleoproterozoic Congo craton and are overlain to the east by the Neoproterozoic
14
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
Fig. 8. (a) Geological map of the Nzilo area. NEP: Nzilo electric power plant. Map completed after Byamungu et al. (1979). (b) The boundary between Katangan strata and the Kibaride belt in the Nzilo area (Yoshida et al., 2004). (c and d) Hand-specimen and field photographs showing herringbone cross bedding within the Nzilo Group. Battery and hammer for scale respectively. (e) Field photograph showing wavy ripples within the Nzilo Group. (f) Kataba Conglomerate in Nzilo area.
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
15
Fig. 8 (continued )
Katangan Supergroup. The contact between the Congo craton and the Kibaride belt was described along the Lungenda River and 4 km toward the north, away from the confluence between the Katchiza and Tshiezi Rivers (Moureau, 1960). The Archaean to Paleoproterozoic Congo craton is composed of orthogneisses, chlorite schists and quartz phyllites. It is overlain by a 100-m-thick conglomerate at the base of the Kibaran supracrustal rocks ( Moureau, 1960). The conglomerate is a poorly sorted to unsorted and locally strongly deformed. It contains oriented, ellipsoidal quartz and quartzite pebbles set in a sericite schist matrix. The pebbles have flat bases. The lower boundary of the Katangan strata against the Kibaran rocks is marked by a basal conglomerate. The Katangan Supergroup sedimentary rocks are gently dipping to the SE (10–20) and some authors (e.g. Cailteux, 1990) suggested that the Katangan was thrust onto the Kibaran rocks, although Moureau (1960) proposed that this boundary is marked by a major unconformity. This controversy arises from a lack of detailed field descriptions of the boundary. We have recently made lithostratigraphic and structural observations, in the Nzilo area, which show that the contact is a well defined unconformity that can be traced for several km (Fig. 8b, Yoshida et al., 2004). The unconformity runs nearly NE–SW (sub-parallel to the strike of the Kibaride rocks) and is gently dipping (10–20 ) to the SE. An undeformed, pebble-supported conglomerate above the contact defines the base of the Katangan and covers the upper part of the Nzilo Group. Brecciation and associated hematite mineralization affects Kibaran rocks a few meters below the unconformity but is absent within the
basal conglomerate. These data rule out the hypothesis that the Katangan is in thrust contact with the Kibaran rocks in this area (Yoshida et al., 2004). Two main Kibaran lithostratigraphic units were defined in the Nzilo region (Fig. 3, Table 1; Robert, 1931; Cahen, 1954; Moureau, 1960; Byamungu et al., 1979): (1) a basal conglomerate of unknown thickness, which is overlain by a 2000–3000-m unit made of chlorite schists, chloritoid schists, calc-silicates, talc schists, stromatolitic marbles, mica schists and minor lenticular metacherts corresponding to the K1 or Kiaora Group; (2) a >3000-m-thick succession of quartzite and subordinate phyllite-conglomerate alternations. Sedimentary structures within this succession include normal bedding, graded bedding, bi-directional cross stratification or herringbone cross stratification (Fig. 8c and d) and wavy or symmetrical ripple marks (Fig. 8e). This lithological association and related sedimentological features are characteristic of the Nzilo Group (K2) as described elsewhere within the Kibaride belt. The boundary between the Kiaora and the Nzilo Groups is defined by a 100–200-mthick, poorly sorted to unsorted, matrix-supported conglomerate (Fig. 8f) at the base of the Nzilo Group. The conglomerate is composed of clasts of black shale, quartzite, quartz, granite and mafic igneous rock, similar to the Kataba Conglomerate reported in Mitwaba area. These pebbles show flat bases and were mostly derived from the underlying Kiaora Group metasedimentary units and the igneous rocks intruding this group. Igneous rocks are rare in the Nzilo area, but Lepersonne (1974) reported NE–SW-elongated tonalitic bodies intruding the Kiaora Group sedimentary rocks around Nseke and northeast of Mutchacha (Fig. 1). The Nseke pluton
16
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
is affected by both Kibaran deformation events. It lacks primary muscovite and contains hornblende, kinked or bent biotite, plagioclase and quartz with minor alkali feldspar. Zircon, titanite and apatite are the main accessories (Kokonyangi and Roser, unpublished data). This orthogneiss is similar to those described in the Manono area (Ngulube, 1994). Moureau (1960) describes two-mica granitic gneisses intruding the Kiaora Group northeast of the Kolwezi–Mutchatcha railway line. There are no detailed petrological or geochronological data on these granitic gneisses, but they show structural similarities with the 1.38 Ga orthogneisses exposed in the Mitwaba region. Kibaran mafic igneous complexes are exposed west of Kolwezi, along the Lubudi River, and extend a strike length of >50 km (Moureau, 1960). Byamungu et al. (1979) document mafic igneous rocks near the Nzilo electric power plant (Fig. 8a). The rocks include basalts associated with rhyolites at the top of the Nzilo Group (e.g. Fig. 3). In the same area, dolerites and gabbros intrude
quartzites, metapelites and metaconglomerates of the Nzilo Group. The mafic rocks are composed of plagioclase and pyroxene, with well-preserved ophitic texture in the gabbros (Cahen, 1954; Moureau, 1960). Pyroxenes are partially converted into hornblende. Moureau (1960) suggested that the mafic igneous rocks exposed along the Lubudi River could be coeval with adjacent parts of the Nzilo Group. The relationship between these mafic rocks and the Kibaran granitoids has never been described. However, because the mafic rocks intrude the Nzilo Group, they must be younger than the 1.38 Ga granitoids, which predate that group. Geochronological data are needed from the mafic rocks. 7. Structure and metamorphism
Detailed structural and metamorphic data are scarce in the Kibaride belt and come mainly from the Mitwaba area. The regional structural trends (Fig. 9) indicate that all the
Fig. 9. Sketch map showing the regional structural trends in the Kibaride belt. Modified from Cahen (1954), using data of Aderca (1950), Van de Steen (1953b), Lepersonne (1974), Kampunzu et al. (1986) and Kokonyangi et al. (2001b).
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
regions defining the >600-km-long Kibaride belt experienced the same deformation history. The main structural trend is NE–SW, with NW-verging macroscopic isoclinal folds (Figs. 9 and 10a–c). These regional structures are linked to the youngest deformation event that has deformed and/or partly or totally overprinted earlier ENE-trending mesoscopic fabrics and related metamorphic parageneses (Byamungu et al., 1979; Kampunzu et al., 1986; Ngulube, 1994; Kokonyangi et al., 2001a, 2004, 2005). In the Mitwaba region (e.g. Fig. 4a), two main contractional events affected the Kibaran metasedimentary and igneous rocks (Van de Steen, 1959; Kampunzu et al., 1986; Kokonyangi, 2001; Kokonyangi et al., 2002, 2004, 2005). Where preserved, the oldest deformation event is characterized by an ENE-trending foliation/schistosity (N60–90E) and associated mesoscopic asymmetric folds with a steeply dipping (50–60SE) axial plane foliation.
17
Because these oldest fabrics affect the bedding (S 0), they are taken here to represent the D 1 deformation event in the Kibaride belt. This fabric is well developed within upper amphibolite facies rocks of the Kiaora Group and related granitoids (e.g. Figs. 4a and 11a). D1 deformation is also characterized by south-dipping reverse faults and thrusts (Fig. 10a). Syn-D1 granitoid gneisses yielded a SHRIMP U–Pb zircon crystallization age of 1381 ± 8 Ma, which is taken to represent the timing of D1 in the Mitwaba region (Kokonyangi et al., 2004). The S1 fabric is totally absent within metasedimentary rocks of the Nzilo and Hakansson Groups, which are not intruded by syn-D1 orthogneisses (e.g. Figs. 2 and 4a). Near Mwadianvula Falls in the Mitwaba region, the Kataba Conglomerate unconformably overlies the Lwabwe mafic complex. In this outcrop, the Kataba Conglomerate contains clasts from the 1.38 Ga igneous rocks (Kokonyangi et al., 2005). These field relationships indicate
Fig. 10. Cross-sections displaying the structural style within the Kibaride belt: (a) A–B shows characteristic NW-verging F2 isoclinal folds in the NW part of the Mitwaba area (Fig. 2). (b) C–D shows asymmetric F1 folds in central Mitwaba area (Fig. 4a). (c) NW-verging F2 isoclinal folds in the Bia region (map unavailable).
18
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
Fig. 11. (a) Photograph showing the field relations between the Kisele monzogranite gneiss and its host sillimanite-grade metasedimentary rocks of the Kiaora Group. S1 within both the ortho- and paragneisses is parallel to the long axes of metasedimentary xenoliths. (b) Microphotograph showing two stages of garnet growth (G1 and G2) in metapelites located within the contact aureole of the Kifinga–Kisele batholith in the Mitwaba area. See text for further details.
that the Nzilo, Lubudi and Hakansson Groups were deposited after the emplacement of the syn-D1 1.38 Ga igneous bodies (Mortelmans, 1951; Kokonyangi et al., 2001a, 2004). Geochronological data (Kokonyangi and Armstrong, unpublished data) indicate that the Kataba Conglomerate contains zircons from the 1.38 Ga orthogneisses, supporting the post-D1 deposition of the Nzilo, Lubudi and Hakansson Groups. The younger deformation is recorded by regional, NE-trending, isoclinal folds with a related schistosity deforming the earlier (D1) fabrics and reverse faults (e.g. Fig. 6a). These regional structures deform a metamorphic/structural foliation and therefore are linked to the D 2 in this paper. SHRIMP analyses of metamorphic zircon from the strongly deformed and metamorphosed Kisele foliated monzogranite gneiss and inferred to record M2 metamorphism yielded an age of 1079 ± 14 Ma, which was tentatively suggested to date the timing of the D2 contractional event and related M2 metamorphism in the Kibaride belt (Kokonyangi et al., 2002, 2004). The syn-D1 Mitwaba granitoid bodies are surrounded by a metamorphic contact aureole including the assemblages biotite–garnet–hornblende in mafic rocks and garnet–biotite in metapelites (Kokonyangi et al., 2005). Most of the garnet porphyroblasts from the contact aureole exhibit two stages of growth (Fig. 11b). Kokonyangi et al. (2002, 2005) suggested that the cores of these garnets porphyroblasts grew during syn-D1/M1 metamorphism while the rims developed during syn-D 2/M2 regional metamorphism. The mineral assemblages including chlorite, biotite, garnet, staurolite, kyanite and sillimanite and defining the S2 foliation display medium-pressure (MP)-medium temperature (MT) Barrovian metamorphism (Kokonyangi et al., 2001a). There is a general increase in metamorphic grade from the chlorite zone (NW) to sillimanite zone (SE). The preliminary peak P–T conditions for the amphibolite facies mineral assemblages documented SE of Mitwaba were estimated at 740–780 C and 6–6.5 kb (Kokonyangi et al., 2001b).
In the Mwanza region (Manono), Ngulube (1994) defined two foliations within granitic gneisses and their host metapelites (Figs. 5 and 6). S1 trends N50–70 E, parallel to the bedding. This foliation is crosscut or crenulated by S2, which trends N30E and is oblique to the bedding. Cahen (1954) indicated that the large-scale Kibaran structures in the Mwanza region form a succession of NE-trending isoclinal anticlines and synclines verging towards the NW, similar to that identified in the Mitwaba region. The same author reports the existence of thrust sheets in this area with a NW-transport direction, but no detailed study has been conducted on these tectonic units. Although detailed metamorphic data are not available in the Mwanza region, various reports document upper amphibolite conditions, as shown by the assemblage garnet + biotite + sillimanite ± kyanite in metapelitic rocks and tremolite + diopside in calc-silicate rocks (Lonchampt, 1971; Tegyey, 1971; Lonchampt and Heinry, 1972). The mineral assemblages, including garnet–staurolite–kyanite, suggest a MP/MT metamorphism, similar to that reported in the Mitwaba area. In the Mwanza area, Lonchampt (1971) and Lonchampt and Heinry (1972) pointed out that the grade of metamorphism is higher in the Kiaora Group than in the overlying groups. Indeed, in many parts of the Kibaride belt in Katanga, several studies indicate that the older Kiaora Group exhibits higher metamorphic parageneses (up to higher amphibolite facies) whereas the younger groups display greenschist facies parageneses (e.g. Lonchampt, 1971; Lonchampt and Heinry, 1972; Byamungu et al., 1979; Kokonyangi, 2001). In keeping with similar interpretations in orogenic belts around the world (e.g. Barker, 1998) this difference could suggest that the older Kibaran sedimentary units (Kiaora Group) were sited in the middle crust during the onset of the D2/M2 tectonothermal event while the younger units were at shallower crustal levels. However, metamorphic conditions were clearly variable during M2 and reached sillimanite grade in the Mitwaba area, as discussed above.
J.W. Kokonyangi et al. / Journal of African Earth Sciences 46 (2006) 1–35
In the Bukama and Bia Mountains areas, Cahen (1954) reported two deformation events, with predominant mesoscopic and macroscopic NE-trending isoclinal folds verging towards the NW (Fig. 10c). In the Hakansson Mountains and Mwanza areas, Van de Steen (1959) documented thrusts overprinting D1 and D2 Kibaran structures. Lataillade and Bielle (1972) indicated that these areas were affected by a low- to medium-grade metamorphism, and reported local garnet-biotite-staurolite-kyanite assemblages suggestive of MP/MT metamorphism similar to the Mitwaba region. In the Nzilo region, two deformation events were identified within the metasedimentary successions (Byamungu et al., 1979; Kokonyangi and Yoshida, unpublished data). S1 is a flow schistosity, axial planar to F1 folds affecting the bedding S0. This fabric is oblique to the stratification and occurs within the Kiaora Group sedimentary rocks exposed around Nseke and Nzilo (Kokonyangi and Yoshida, unpublished data). Structures related to the second deformation event include NE- to NNE-trending S2 cleavage, axial planar to F 2 macroscopic and mesoscopic isoclinal folds verging to the NW. D2 affects all Kibaran metasedimentary rocks exposed in the Nzilo region (Byamungu et al., 1979). In addition, these authors documented large-scale reverse faults affecting the Nzilo anticline (Fig.8a) and Yoshida et al. (2004) mapped a large-scale zone of brecciation marked by hematite mineralization (Fig. 8b), although the significance of these brittle deformations is as yet not well understood. The grade of regional metamorphism is low (chlorite-biotite zone) in the Nzilo area. The chlorite zone is recorded in the Nzilo Group, whereas the biotite zone is recorded in the Kiaora Group ( Byamungu et al., 1979). 8. Geochemistry of igneous rocks
8.1. Mafic igneous rocks
Mafic igneous rocks exposed in the Kibaride belt can be subdivided into two different groups: (1) syn-D 1, 1.38 Ga igneous bodies intruding the Kiaora Group sedimentary rocks and; (2) mafic igneous rocks of unknown age intruding the Nzilo Group sedimentary rocks and affected by the D2 fabric, as well as post-D2 lavas in the Bukama region. Modern geochemical data are only available on syn-D1 1.38 Ga mafic rocks from Mitwaba (Kokonyangi et al., 2005).
19
associated with granitic gneisses around the N’seke area (Nzilo region). In the Mitwaba area, these mafic complexes are closely associated with the 1.38 Ga Kifinga–Kisele granitic batholith (Figs. 2 and 4a), and they share the same field relations, deformational history and petrographic features with the older mafic igneous bodies exposed elsewhere within the Kibaride belt (e.g. Mwanza, Bia and Manono). All these rocks are tentatively considered to represent the same syn-D1 mafic igneous event in this paper, although more geochronological data are required to confirm this hypothesis. These igneous rocks are affected by both Kibaran deformation events. They are variably metamorphosed and composed of plagioclase, amphibole, biotite and rare relict igneous orthopyroxene. Two types of plagioclase and hornblende were distinguished in the Mitwaba mafic rocks (Kokonyangi et al., 2005). Greater than 30 m away from the contact with the Kifinga–Kisele granitic batholith (Fig. 2), relict magmatic labradorite (An up to 66) is preserved, and the main amphibole is a metamorphic actinolite (X Mg = {Mg/Mg + Fe2+} = 0.71–0.76) or cummingtonite (X Mg = 0.54–0.55), although relict hornblende (X Mg 0.52) is preserved within the Lwabwe complex. Close to the contact (75 up to 81 wt%), Al2O3 (>16, up to 20 wt%), and K2O (0.9–4.22) and are extremely peraluminous (ASI: >2 up to 7) compared to those marked by higher albite and muscovite (70 < SiO2 < 75 wt%, Al2O3: 13.89–17.28 wt%, K2O: 0.52–1.57 wt%, ASI: 0.98–1.1, i.e. a metaluminous to weakly peraluminous composition). In contrast, albite-muscovitebearing samples are richer in Na2O (7.99–8.9 wt%) than those containing spodumene and alkali feldspar (Na2O: 0.35–3.44 wt%). The pegmatite is characterized by Rb contents between 150 and 590 ppm, except two samples with Rb close to 120 ppm, Sr contents of 7–30 ppm and Ba contents
