4.2. Established chronostratigraphic schemes
There is no formal chronostratigraphic schemes for the Eyasi-Wembere and Tanga basins. Here the Miocene-Quaternary correlated chronostratigraphic schemes for the Eyasi-Wembere and offshore Tanga basins (Fig. 5) are established based on core logging, seismic interpretation and the accessed literature and core reports. These chronostratigraphic schemes show vertical variations in rock units that are similar to rock distributions in other rift basins with comparable tectonic-magmatic and structural evolution (e.g. Sakai et al., 2013; Ragon et al., 2019).
4.2.1. Stratigraphy of the Eyasi-Wembere basin: core logging
The Eyasi-Wembere stratigraphy is established based on sedimentary logging of Wembere-1 and Wembere-3 cores. In this work, the Wembere cores were logged just for correlation purpose, further detailed sedimentological logging on these cores will be published in the subsequent studies. Based on these cores, the basal part of the Eyasi-Wembere stratigraphy contains weakly metamorphosed, polymictic breccio-conglomerates (Fig. 6A). Clasts of the conglomerates are composed of reddish-brown to light orange colored fragments that were delivered from a well-rounded, medium-conglomeratic sandstone. Complex folding are present in some of the sandstone clasts making up the lower part of the stratigraphy (Fig. 6B). This interval was followed by deposition of volcaniclastic deposits of variable volumes of volcanic lapilli (Fig. 6C) and ashes (Fig. 6D) that is overlain by successions of fining upward beds characterized by poorly sorted, very coarse-conglomeratic sandstone deposits at their basal parts (Fig. 7A & B). Quartz and feldspar grains are common in this interval (Fig. 7A). Upper part of the Eyasi-Wembere stratigraphy contain volcaniclastic deposits dominated by volcanic ashes and coarse-fine grained clastic deposits. Limestone clasts are common in the lower section of the uppermost part of the stratigraphy (Fig. 8A).The clastic deposits include unconsolidated, coarse sandstone deposits alternating with sandy-mudstone caped by grey-dark grey, red-reddish brown carbonaceous silty-sandy mudstone (Fig. 8B-D).
The weakly metamorphosed, polymictic breccio-conglomerates in the lower part of the Eyasi-Wembere stratigraphy reflect early-rift basin sedimentation (e.g. Sakai et al., 2013) while complex folding is interpreted to be convolute lamination indicating rapid sedimentation and squeezing of wet sediments during deposition of the source beds through which the sandstone fragments were delivered from. Grain size variations displayed by sedimentary successions in the rift basins are mostly linked to water-level change and structural position within the basin (Scholz et al., 1990; Sakai et al., 2013). Similar interpretation is adopted herein to explain observed vertical grain size changes in the Eyasi-Wembere cores (Fig.s 7 & 8). Color differences displayed by the deposits in Fig.s 8 suggest varied depositional oxygen conditions, whereby greyish-reddish dark grey-black deposits are interpreted to have been deposited under oxic and anoxic conditions respectively.
4.2.2. Stratigraphy of the Tanga offshore basin
The Miocene-Quaternary stratigraphy of the Tanga offshore basin is established based on Ras Machuisi North 1 wellbore report, seismic interpretation (see section 4.3 for the results), previous works on the Tanzania coastal basin (e.g. Kapilima, 2003), and field observation, whereby carbonates deposits were observed (e.g. Fig. 9). The Tanga offshore basin experienced continental sedimentation during Miocene-Pliocene period. The Miocene-Pliocene successions are dominated by reddish-brown and minor greenish silty-shale alternating with fine-very coarse quartzose-feldspatic sandstone layers. The Miocene successions of the Tanga offshore basin are barren of dateable fossils just like the Miocene successions of the Eyasi-Wembere basin. Upper part of the basin contain Neogene-Holocene shallow marine clastic sediments and Paleogene-Neogene limestone deposits (e.g. Fig. 9). These clastic sediments are characterized by fine-coarse sand bodies alternating with greenish-grey silty-sandy clays. Seismic features interpreted to represent magmatic bodies (Section 4.3.2) are seen in the Miocene-Quaternary stratigraphy of the Tanga offshore basin.
4.3. Seismic interpretation results
4.3.1. Miocene-Quaternary tectonic features
The studied seismic interval contain wedge shaped sedimentary packages that are expanding toward the bounding faults (Fig. 10). These wedges, characterized by internal strata that are also expanding toward the bounding faults, they first occur just below the Quaternary reflector (Fig. 10). The immediate sedimentary deposits above the Quaternary reflector do not display wedge shaped geometry but they underlie wedge shaped packages with similar features to the wedges below the Quaternary reflector. That is, the observed sedimentary wedges overlie and underlie sedimentary successions characterized by more or less uniform thickness strata (Fig. 10). This sequence repeats itself whereby wedge shaped deposits are seen further up in the stratigraphy of the area (Fig. 10). Based on previous researches from different basins (e.g. Kiswaka & Felix, 2020), the observed sedimentary wedges are interpreted to mark periods of active tectonics, and therefore syn-rift deposits in the Miocene-Holocene stratigraphy of the Tanga offshore basin. Similarly, the sedimentary packages overlying and underlying the wedges are interpreted to mark intervening periods of tectonic quiescence and are named post rift deposits. The syn-rift deposits below the Quaternary reflector (Fig. 10) are assigned the Miocene-Pliocene tentative age while the Pleistocene and Holocene tentative ages are given to the two syn-rift intervals above the Quaternary reflector due to their stratigraphic positions relative to the Quaternary reflector, but also due to regional reports on rift occurrence in East Africa (See Mollel & Swisher, 2012; Courgeon et al., 2018; Mvile et al., 2021).
4.3.1.1. Faults movement: timing of tectonic pulses
The Quaternary surface map shows that the Miocene-Quaternary development of the Tanga offshore basin has been mostly influenced by the EARs which culminated at the development of about 25 km wide sub-basins with a more or less N-S to NE-SW orientation (e.g. Fig. 2; Mvile et al., 2021). Seismic interpretation revealed that the sea bottom has been dissected by several faults that have been active during different periods to recent time. An example of the timing differences in fault activity is shown by the faults marked 1-3 (Fig. 10). Due to presence of wedge shaped deposits separated by uniform thickness sedimentary packages, periods of active extensional tectonics have been separated from the intervening periods of tectonic quiescence that record post rift sedimentation (Fig. 10). These depositional geometries show that movements on faults 1 and 2 occurred simultaneously since they are both bounding age equivalent syn-rift deposits. However, movement on fault 3 occurred after movements on faults 1 and 2 have stopped. That is sedimentary packages characterized by syn-rift wedges relative to fault 1 pass out to post rift deposits relative to faults 1 and 2, and vice versa. These observation indicate occurrence of localized tectonic pulses along the East African Rift System and that there might be more pulses than what has been reported by previous workers. The fact that all of these faults have dissected the sea bottom despite being active in different times suggests that periodic reactivation and consequently faults movement occurred to recent times. The EARs component studied herein is characterized by two distinct fault system: (1) the dominant N-S to NNE-SSW trending major faults which define key orientations and mark the margins of the Quaternary fault bounded sub-basins in the study area, and (2) the more or less E-W trending faults that mark the northern and southern limits of the Quaternary sub-basins (e.g. Fig. 2).
4.3.2. Miocene-Quaternary magmatic features
Seismic profiles into the Tanga offshore basin contain eye-shaped features in some intervals of the studied stratigraphy. The eye-shaped structures have produced localized folds through which high amplitude reflectors are onlapping onto them (Fig.s 11 & 12). The high amplitude reflectors onlapping onto the localized folds are overlain by intervals characterized by weak near parallel seismic reflectors that are laterally continuous (Fig. 11). In some places, these eye-shaped structures occur adjacent to high amplitude discordant reflectors in the Quaternary stratigraphy (Fig. 12). In this work, the eye shaped features are found in the upper part of an interval assigned the Miocene-Pliocene tentative age but also in the Quaternary successions (Fig. 12). Another interesting feature was observed on seismic line TA-08-118 (Fig. 12) whereby an inclined linear feature crosscuts the Pliocene-Pleistocene successions without a noticeable displacement. The concurrent occurrence of eye-shaped features and concordant anomalous reflectors are used to confirm presence of magmatic intrusions in the basin. This interpretation is based on previous works that identified magmatic bodies on seismic images from different places (e.g. Trude, 2004; Hanset et al., 2008; Zhao et al., 2014; Zhao et al., 2016; Eide et al., 2018). Following the presence of eye-shaped features and discordant reflectors indicative of magmatic intrusions, the linear feature crosscutting the Pliocene-Holocene successions (Fig. 12) is interpreted to mark igneous dyke in the basin suggesting a relative much younger volcanic activity which is believed to have occurred during the Holocene. The high amplitude reflectors onlapping onto the localized folds are indicative of coarse grained sediments while interval with weak reflections suggest homogeneous shale-very fine sand deposits due to limited contrast in acoustic impedance (e.g. Armitage et al., 2012; Berlin, 2014). One would link the observed features to salt intrusions/ diapirism but this interpretation is less likely due to absence of salt deposits in the Tanga Basin.
4.3.3. Conformity of tectonic features and magmatic bodies
Stratigraphically, the sedimentary wedges on Fig. 10 conform to intervals with linear features crosscutting other sedimentary layers without noticeable displacement (Fig. 12) and areas containing eye shaped features, localized folds and localized high amplitude discordant seismic reflections (Fig.s 11 & 12). That is, occurrence of the observed magmatic elements coincide to periods of tectonic episodes. This observation suggests simultaneous occurrence of the two.
4.4. Elemental distributions and their link to Eyasi-Wembere basin stratigraphy
Elemental distributions have allowed subdivision of the Wembere-3 core into three major zones marked 1, 2 and 3 in Fig.s 13-17. These zones were marked based on distribution trends that are grouped into four. Basal part of the core begins with relatively high Fe and Fe/Ti (group 1) and Zn and Ti (group 2) values that conform to relatively low Rb/Ti and K/Ti (group 3) values (Fig.s 13-16). The group 3 values display a general upward increasing trend in zone 1 until high values are reached in zone 2 and extend to zone 3 where uniform distribution for these values is attained (Fig.s 13-16). The groups 1 and 2 values exhibit a general upward decreasing trend in zone 1, until more or less, uniform distributed low values are reached at about 115 m core depth where zone 2 begins (Fig.s 13-15). Zone 2 extends to around 40 m core depth where a general upward increasing trend (for group 1) and more or less uniform distributed high values (for group 2) characterizing zone 3 begins. Group 4 (Zr and Zr/Ti) values (Fig.s 13 & 17) display similar distribution trends to Fe, Fe/Ti, Zn and Ti (groups 1 & 2) in zones 1 and 2 except that they show a general upward decreasing trend in zone 3. Despite the observed general trends in zones 1-3, these zones are characterized by several peaks and troughs reflecting local influences in elemental distributions.
Basal part of zone 1 (Fig.s 13-17) conform to oldest interval penetrated by Wembere-3 core. The interval is dominated by conglomeratic sandstone that has been slightly metamorphosed. The conglomeratic clasts are characterized by chemically weathered fine-coarse sandstone fragments, some of which have syn-depositional deformation structures (e.g. Fig. 6A & B) reflecting rapid sedimentation during deposition of the source material. An interval overlying this conglomeratic unit, which display upward decreasing trend for groups 1, 2 & 4 values and upward increasing trend for group 4 values, conform to volcaniclastic successions with variable amounts of grain and matrix supported lappilistone and volcanic ashes (e.g. Fig.s 6 C).
Zone 2, which is marked by minimum-uniform distributed groups 1, 3 and 4 values and Maximum-uniform distributed group 3 values, coincide to sedimentary successions dominated by fining upward beds that contain poorly sorted, very coarse-conglomeratic sandstone deposits at their basal parts (Fig. 7).
Zone 3, which shows upward increasing trend for group 1 values, uniform distribution for groups 2 and 3 values and upward decreasing trend for group 4 values, correspond to sedimentary successions characterized by volcanic tuff, lime sandstone and fining upward clastic deposits (Table 1 & Fig. 8).
Several workers have used elemental ratios to reconstruct provenance areas for clastic deposits whereby a decrease in K/Al and Rb/Al ratios have been used to indicate reduced influence in river inputs (e.g. Wehausen & Brumsack, 1999; Martinez-Ruiz et al., 2003; Sangiorgi et al., 2006; Martinez-Ruiz et al., 2015). Here Al was not measured by the pXRF and that is why Ti was used for calculations of elemental ratios. Both Al and Ti have been interpreted to be of detrital origin (Tribovillard et al., 2006), and thus similar interpretation for elemental ratios is assumed in this work. Generally, zone 1 has high values of Fe, Ti, Zn, Fe/Ti and Zr that correspond to low values of K/Ti and Rb/Ti implying dominance of inputs that were not delivered by rivers into the basin. Core logging has shown that zone 1 is dominated by volcanic products. Overall low, uniform distributed Fe, Ti, Zn, Fe/Ti, Zr and Zr/Ti values, which coincide to high, uniform distributed K/Ti and Rb/Ti values in zone 2 indicate dominance of river inputs into the basin as it is reflected by successions of fining upward beds. Influence of volcanic inputs into the basin, during deposition of zone 3, is reflected by an increase in Zn, Fe and Ti values and decrease in Zr and Zr/Ti values (Fig.s 13, 15 & 17). The fact tha K/Ti and Rb/Ti values of zone 3 (Fig. 16) are more or less similar to that of zone 2 suggest that river sediments were as equally important to basin development when zone 3 was laid down.
Based on core logging (Section 4.2.1), deposition of zones 1-3 (Fig.s 13-17) is known to have been mostly influenced by input of volcanic products into the basin, particularly zones 1 and 3. Therefore, elemental variations in zones 1 and 3 are mostly due to volcanic inputs into the basin while river inputs and fluctuations in water level within the basin were mostly dominant during deposition of zone 2. At this level of understanding, deciphering chemical composition of the volcanic products, which influenced deposition of zones 1 and 3 (Fig.s 13-17), will help to confirm whether the associated volcanism is linked to tectonic rifting or not. This will be done upon correlation to known compositions of volcanic products that originated from tectonic events that influenced basins development within the East African Rift System (e.g. Boccaletti et al., 1999).
Pearce & Norry (1979) used a plot of Zr/Y against ZR to establishing sources of volcanic products whereby they observed an increase in Zr/Y ratio from island arc and mid-ocean ridge to within plate basalts. Similar interpretation was adopted by Hiscott & Gill (1992) when they assessed igneous origin of Oligocene to Quaternary volcaniclastic sands and sandstones from the Izu-Bonin arc by using major and trace element geochemistry. In this work, the relationship between Zr/Y and Zr values (Fig. 18) has been established qualitatively in which most of the values are plotting outside the island arc, mid-oceanic ridge (MORB) and within plate (WPB) basalts ranges of Pearce & Norry (1979) and Hiscott & Gill (1992). An understanding that their WPB values were delivered from the oceanic crust allowed us to link further increase in Zr/Y values beyond the WPB range of Pearce & Norry (1979) to continental basalts since the plotted elemental values were measured from continental rift successions in the Eyasi-Wembere sub-basin. This is interpretation is likely because the Miocene-Quaternary volcanism of the East African Rift System is reported to have been dominated by eruption of basaltic products (Boccaletti et al., 1999). Thousands of kilometers north of the study area, deformation and magmatism linked with the EARs is concentrated along a narrow zone where by Early Pleistocene and Pliocene volcanic products exist (Boccaletti et al., 1999).