Unusual Components Within a Fine-Grained Contourite Deposit: Significance for Interpretation of Provenance and the Contourite Budget

Deep-water contourite muds are an important component of many continental margin systems and are currently the focus of much interest amongst deep-water researchers. One outstanding gap in our knowledge of these systems is to understand and quantify a contourite budget, both at the small (facies) scale and at the larger drift scale. A second problem concerns the establishing of robust criteria for discriminating between contourites and associated deepwater facies—turbidites and hemipelagites. This paper contributes to these topics by detailed examination of sediment composition, with a particular focus on potentially diagnostic components, within contourites and hemipelagites from the same depositional basin. Samples were selected from Pliocene to Quaternary muddy contourites from the Gulf of Cadiz (IODP 339) and examined by scanning electron microscopy. The presence of tunicate spicules, micro-bored shell fragments, and a particular species of coccolithophore, Braarudosphaera biglowii, all indicate derivation from shallow waters and hence lateral off-shelf supply. In contrast, micro-mudclasts and fragmented bioclasts are indicative of alongslope transport in bottom currents. A normal planktic component of the contourite muds shows a significant vertical input from pelagic settling. Such diagnostic components can also help in the discrimination between contourites, turbidites and hemipelagites.


Introduction
Fine-grained deep-water muds are the dominant sediment type of continental margins, covering large swathes of the outer shelf, slope and rise. Three principal facies groups are recognizedcontourites, turbidites and pelagites/hemipelagites-each, to some extent, with its own distinctive characteristics [1,2]. Contourites are a group of closely related deepwater facies deposited under the influence of semi-permanent current action and are commonly referred to as alongslope deposits resulting from semi-continuous depositional processes. This distinguishes them from other deepwater facies that have been deposited either by episodic downslope processes or events (turbidites, debrites, slides, mass-transport deposits and hyperpycnites), or from continuous vertical settling-the so-called background processes (pelagites and hemipelagites) [3][4][5]. However, because these different processes are part of a process continuum in deepwater, the different facies deposited show some overlap and gradational characteristics, which can make them more difficult to distinguish [6]. A synthesis of the most important criteria for their distinction is provided by Stow and Smillie [2].
Whereas recent research has yielded a much improved documentation of bottom-current processes and resultant contourite facies [7][8][9], there is still a major gap in our understanding of the contourite budget-i.e., where the material for a contourite depositional system comes from and in what relative proportions. Stow et al. [10] outlined the principal sources of sediment supplied to a typical contourite drift to include: (a) vertical flux of siliciclastic material from windblown particles and volcanic dust, river plumes and glaciomarine suspension delivered to the sea surface; (b) vertical flux of biogenic material from sea surface primary productivity, including calcareous and siliceous bioclastic debris, and organic material; (c) slow horizontal advection and vertical settling of mixed siliciclastic and bioclastic material by a combination of hemipelagic processes; (d) episodic downslope flux from low-density turbidity currents and hyperpycnal plumes; (e) intermittent downslope flux via spillover processes, bioturbational and shelf-edge current resuspension; (f) erosion of the seafloor and resuspension by bottom currents adjacent to and upstream from the site of drift deposition. They further summarise the typical and varied composition of contourites and outline a potential sediment budget for the Eirik Drift in the North Atlantic.
The primary aim of this study is to contribute to the discussion of the contourite budget, with specific reference to the Gulf of Cadiz contourite depositional system in the NE Atlantic Ocean, which was drilled and extensively cored during IODP Expedition 339 [11]. This work is not intended as an exhaustive comparative study of all facies types and their composition, but is instead a synthesis of a number of interesting observations made during a larger study into the microstructure of finegrained deep-water mudstones reported elsewhere [12,13]. The secondary aim is to stimulate additional in-depth investigations into the type of particulate materials within deep-water muds of all facies types (turbidites, contourites and pelagites/hemipelagites), their significance in terms of origin, additional environmental information, and their use in differentiating between these different facies; herein we look at contourites and hemipelagites.

Contourites, General Characteristics
All of the examined (contourite) samples are of fine-grained contourite muds with moderate bioturbation throughout, which accumulated at average rates of sedimentation of 25-35 cm ka −1 (Figure 1). Most have no visible sedimentary structures, although two samples (1390B 2H4 7-9, 1390B 2H5 94-96) show the preservation of discontinuous silt laminae, within a succession having a much higher rate of sedimentation (~100 cm ka −1 ) (Figure 1). Previous textural study shows that they are all classified as silty clays to clayey silts [12], with a minor sand-sized detrital component present in the silty contourite sample (1387C 48R3 33-35). The general appearance is one of a chaotic microfabric (Figure 2), although detailed analyses by Bankole et al. [13] have demonstrated a mixed microfabric is typical of the Cadiz contourites in general, including semi-random, random and parallel-oriented microfabrics.  Table 1). Note, sample 8 is a silty contourite, identified from cored material, within an otherwise turbidite dominated section. Note, that water depth and sedimentation rates are given below site location data. Based upon http://publications.iodp.org/proceedings/339/EXP_REPT/CHAPTERS/339_101.PDF.

Tunicate Spicules
Spicules of these primitive invertebrate chordates are present in variable amounts within the muddy contourite thin sections examined, with three out of eight having at least 10 tunicate spicules each (Table 1). They are calcareous stellate forms, and highly variable in morphology ( Figure 3), with between 4 and 12 rays per spicule. Rays can be equal length, or variable in length, and some show apparent indications of abrasion, to the extent that a number of samples are broken in half, or have rays missing ( Figure 3K). Individual rays comprise an acicular fabric, some examples show intricate "growth-rings" (Figure 3Q), and in some cases compositional segmentation ( Figure 3B).      [13][14][15]. Key: Y = present, YY= very common, VM = very minor. * includes sand sized particles. § Large number in part due to larger area examined, and amount of time taken. Green = muddy-silty contourites, yellow = hemipelagite. ? denotes uncertainty in generic assignment.

Micro-Mudstone Clasts
These are commonly observed within all the muddy contourites. The majority of micromudstone clasts are in the upper range of silt sized (~50 µm), elongate to equigranular in shape, and angular to subrounded (Figure 4). Although many micro-clasts are approximately parallel to bedding, some are perpendicular or at variable angles to bedding. They contain variable amounts of silty particles (quartz), mica, dolomite and calcareous micro-organisms ( Figure 4J,K).

Coccolithophores (Braarudosphaera biglowii)
The "standard" range of coccolithophores, as described by Balestra et al. [16] and in the shipboard results [11], including Emiliana huxleyi and Gephyrocapsa caribbeanica, are common within all the muddy contourites examined ( Figure 5A-C), whereas Braarudosphaera biglowii occurs within two of the muddy contourites (Table 1), and then only rarely. Braarudosphaera biglowii has a very distinctive morphology, comprising robust pentameral plates (nanoliths) that can be seen to be further subdivided into five ( Figure 5D,E). The overall width is ~15 µm, with individual plates ~5 µm.

Micro-Bored Shells
Micro-bored shells/tests are not straightforward to identify, being easily confused with microporous (recrystallized) calcite. However, clear samples are commonly observed in all of the examined muddy contourite samples ( Figure 6), with 43 identified from three methodically scanned areas (sample 1390B 2H5 94-96, laminated contourite). Micro-borings occur in broken often highly fragmentary shelly materials, and have a variety of distinctive different forms, with both vertical and horizontal elements, but with horizontal apparently dominant, as well as a variety of sizes ( Figure 6). As such the observed micro-borings are likely to have had a range of constructors.

Silty Contourite
The silty contourite is similar in its bioturbation, composition and fabric to that of the finergrained muddy contourites (above), although it is stratigraphically older-early Pliocene. The principal differences are the presence of coarser sandy detrital grains, and larger foraminifera ( Figure  7a). In addition, the sample contains abundant examples of Discoaster coccolithophores (Figure 8), some minor Braarudosphaera biglowii and one ?Pontosphaera spp.

Hemipelagites
In general, the two hemipelagite samples are finer grained than the contourites, i.e., more clayrich than silt-rich, with less obvious micaceous material, and a greater carbonate content (>60%). They show extensive and varied bioturbation and burrowing throughout, and a completely random microfabric. The high amount of biogenic calcareous material includes a variety of coccolithophores (whole and disassociated), planktic and benthic foraminifera ( Figure 7B,C). In addition, the Miocene sample contains abundant Discoaster (Figure 9), and three ?Pontosphaera, whereas the Quaternary specimen contains limited numbers of Discoaster (Figure 10), and one example of Braarudosphaera biglowii. Some small micro-clasts and limited numbers of micro-bored shelly material (Figure 10), as well as detrital quartz filled and pyrite burrows (Figures 7C and 10) are present. There are no definitive tunicate spicules.

Tunicate Spicules
The occurrence of tunicate spicules may well be of significance in terms of environmental interpretation. According to the British Columbia tunicate key [17], stellate calcareous tunicate spicules occur in species of Didemnum and Trididemenum, and are abundant in rocky shallow-water environments, from intertidal to 400 m depth. Other workers have illustrated didemnid ascidian spicules on hard substrates commonly from shallow water (0-50 m), with some deeper-water (1500 m) [18]. Wei [19] records tunicate spicules mainly from less than 400 m, and more rarely down to 900 m, where redistribution downslope is associated with turbidites. It seems most likely, therefore, that the tunicate spicules recorded from the contourites herein originated from shallow-shelf waters, from the colonisation of rocky substrates (or larger shelly material) [17]. Their broken up remains were then transported into deeper water by turbidity currents, slope spillover or other mass flow processes, and subsequently reworked by bottom currents flowing parallel to the continental slope. The present-day water depths for the contourite sites sampled range from 559 m to 992 m, and the highest count (sample 7) is from site 1390 in 992 m water depth. This further supports an off-shelf supply, although an in situ origin for the tunicate spicules cannot be totally excluded. Transportation, or at least a degree of reworking and winnowing, of the tunicate spicules is indicated by the broken and abraded nature of some of the samples ( Figure  3K).

Coccolithophores
The coccolithophore assemblage at all sites is dominated by the "standard" species found in this region and in the time intervals sampled, and clearly represents an abundant supply of pelagic material to the contourite sediments, by vertical settling coupled with alongslope transport in bottom currents. However, the presence of the distinctive coccolithophore species Braarudosphaera biglowii may indicate an additional source. Konno et al. [20] have noted that this species shows a preference for low-salinity nutrient-rich coastal water, often associated with algal blooms. This is also confirmed by other authors [21].
This suggests that those samples recorded from the contourites herein, most likely originated from shallow-water sediments. Although the tests of Braarudosphaera biglowii appear thick and robust, the plates are held together by an organic frame, and as such would be unlikely to fair well if transported over large distances (pers comm. A. Poulton 2019). Given the apparently intact nature of the Braarudosphaera biglowii species observed, and their low occurrence, these could be interpreted as taphonomic survivors of short mass flow transport from shallower waters, or as offshore dispersal of coastal pelagic blooms via wind and near-shore currents. The numbers and distribution currently noted are not enough to resolve this question, although it is flagged here as worthy of further investigation.
Other substantially whole coccolithophores ( Figures 5A,B and 8K) and foraminifera ( Figure 7A) clearly represent pelagic input into the contourite drift. While disarticulated coccolithophores (Figures 2A,B, 5C and 8I) and broken foraminifera (Figure 2A,B), strongly indicate current reworking (winnowing), possible heavy alteration through bioturbation, or a combination of both. Although many other species of coccolithophore observed during the current study are apparently intact ( Figures 5A,B and 8K), the strongly interlocking nature of the coccoliths make these more robust than Braarudosphaera biglowii, and their occurrence does not therefore discount either winnowing or bioturbation.

Micro-Boring
Micro-borings of shelly fragments can be the product of a range of micro-organisms, but are commonly formed by endolithic cyanobacteria, red and green algae and fungi [22,23]. In general, many micro-borings that are predominantly orientated perpendicular to the substrate are considered likely to be the product of photosynthetic organisms (such as algae), whereas more horizontally orientated ones may result from fungal (non-photosynthetic) activity [22]. Microboring orientation has therefore been taken as an indicator of palaeo water-depth, or more literally as the position of the photic zone [22,24,25]. Although, in reality this is an over simplification, the small number of observed samples herein illustrate both horizontal and vertical components (Figure 6), making any clear interpretation of their environmental significance difficult. High-levels of micro-boring, could be taken as an indicator of multiple phases of reworking/winnowing due to strong bottom currents, or through storm activity in shallow waters, with subsequent relocation to deeper water by mass flow transportation. Future work into this area is clearly necessary, and likely to be informative for environmental and process studies.

Micro-Clasts
Mudstone and siltstone micro-clasts are relatively common within the contourite deposits studied here, and have recently become of great interest in sedimentological studies, associated with both turbidite and contourite deposits [26,27].
The shape and zonation of dolomite within some of the micro-mudclasts ( Figure 4G-J) clearly indicate an authigenic origin, as does micro-quartz growth on quartz silt grains within the clasts. In both cases, the latter implies that the micro-mudclasts have undergone a degree of lithification, although at least some clasts were soft enough to have undergone deformation during compaction ( Figure 4C). Energy dispersive X-ray (EDX) mapping of mudclasts illustrates that compositionally the clasts are higher in K, Fe and Mg, compared to the enclosing matrix ( Figure 4L), indicating that the micro-clasts are not from the same source as the background contourite matrix material. All indications suggest that the majority of micro-mudclasts derive from at least a partially lithified source and, therefore, are likely to have been transported some distance before deposition and incorporation within the muddy contourites.

Facies Discrimination
Discrimination between the principal deepwater facies-hemipelagites, contourites and turbidites-can be fraught with difficulty, especially where the deposits are all fine-grained [2]. Different criteria have been applied in various studies, including primary sedimentary structures, bioturbation, textures and microfabric [10,28]. Textural discrimination [29] and microfabric discrimination [30] have been further developed in recent PhD theses by these authors. Alonso et al. [6] applied a combination of attributes to the mixed deepwater facies of the Gulf of Cadiz.
This study has drawn attention to the use of "unusual" components for elucidating potential source areas and hence likely depositional processes. Although the sample numbers used are insufficient to be definitive, there are some notable differences between the examined muddy contourites and hemipelagites. In particular, tunicate spicules, micro-borings and micro-mudclasts appear to be relatively common in the contourites but generally absent to rare in the hemipelagites. Furthermore, the environmentally sensitive coccolithophore Braarudosphaera biglowii is more common in contourites than hemipelagites, although the evidence is not definitive. Some of these components noted are most likely to have been derived from shallow water (shoreline to shelf depths) and would therefore be still more abundant in turbidites. Confirmation of this proposition awaits further study. We would further suggest that these criteria can be applied elsewhere (other geographies and time periods).

Contourite Budget
A still outstanding question in contourite research is to accurately establish a sediment budget for any particular contourite drift or system. The only attempt at quantifying such a budget was for the Eirik Drift, although this was simply a preliminary estimate [10]. The data presented here is quite insufficient on its own to make a similar estimate for either the Faro Drift or the entire Cadiz Contourite Depositional System, but we can make some pertinent observations as follows: 1. The presence of tunicate spicules, micro-borings and the coccolithophore Braarudosphaera biglowii in the Cadiz contourites all suggest an origin in shallow nearshore to shelf waters. This implies that an important part of the contourite budget was derived laterally, via turbidity currents or other off-shelf supply process.

The presence of micro-mudclasts and generally fragmented bio-clasts indicate probable erosion
and transport by alongslope bottom currents as a supply to the contourite drift. 3. The abundance of a normal planktic assemblage of foraminifera and coccolithophores indicates that vertical supply remains a significant part of the contourite budget.
However, if we combine these observations with the more extensive work on composition of turbidites and contourites from the Gulf of Cadiz by Alonso et al. [6], then we can propose an initial budget estimate for the Cadiz drift system. Compositional data on contourites (from [6]) show a relatively high carbonate percentage (18-45%), much of which was of planktonic biogenic origin. A further distinctive component of these contourites is smectite (and inter-layered) clay minerals, almost certainly derived from the Guadalquivir River, re-circulation via the Strait of Gibraltar, and erosion of mud diapirs exposed at the seafloor. These are all well upstream of the drift sites sampled and therefore indicate incorporation into and transport by MOW bottom currents. All the contourites sampled, from a range of different sites spread along the margin, have a well-mixed, uniform composition, which indicates a dominant alongslope supply, rather than varied downslope input locally. Turbidites that are interbedded with contourites near the base of the drift succession (Pliocene age) have a significant proportion of quartz and mica, some heavy minerals and glauconite, less carbonate (mainly land-derived), and little or no smectite clays.
Based on these data, together with our own observations on unusual components, we would infer the Quaternary drift budget to include approximately 30-35% pelagic biogenics and 65-70% land-derived material ( Figure 11). Of the latter, only about 5% arrives at the drift sites directly via hemipelagic diffusion from wind-blown and river plume supply, whereas 60-65% is transported in bottom currents. The downslope supply of tunicate spicules, micro-bored shells and Braarudosphaera biglowii to the bottom current system occurred in an upstream location. The micro-mudclasts were mostly eroded from the seafloor by bottom currents, and some perhaps by turbidity currents, which were then pirated by bottom currents. These also served to fragment bioclasts and homogenise the overall drift composition. Further work to better quantify the distribution of these unusual components, together with that of the bulk components, would help contribute towards establishing a more accurate contourite budget. Red arrows indicate settlement of pelagic material, solid green arrow equals turbidite derived, broken green arrow represents slope spillover, and black arrows local erosion (of the drift) by bottom currents. Large blue arrow largely erosional and transportational processes, leading to influx into drift. Large yellow arrow, outflux from drift area. After Stow et al. [10]. Estimated contourite budget: 30-35% for vertical settling (planktonic) over the drift, 60-65% for bottom-current transport (includes downslope and planktonic supply to the bottom current), and 5% for direct lateral supply near the drift.

Conclusions
Detailed analysis of the composition of contourites can yield important information about sediment provenance, and hence provide evidence for establishing a contourite sediment budget. This concept is illustrated by careful study of the fine-grained particulate material from selected IODP 339 sites, with a particular focus on some of the more unusual components in the contourite sediments. Tunicate spicules and micro-bored fragmentary shelly material are both indicative of derivation from shallow water (coastal to shelf). Furthermore, the occurrence of Braarudosphaera biglowii suggests an origin in low-salinity nutrient-rich coastal waters and the common occurrence of algal blooms. Together these components support the importance of lateral material supply to the contourite drift system. However, when combined with other compositional data, they suggest that much of this lateral supply occurred upstream of the drift sites, and was transported as normal load in the bottom currents. The presence of moderately well lithified micro-mudclasts within the muddy contourites demonstrates that such particulates may have travelled considerable distances, from areas where lithified mudstones were exposed and undergoing active erosion. These are associated with fragmentary bioclastic material and therefore indicate the importance of alongslope supply in bottom currents. The presence of planktic foraminifera and coccolithophores indicates the importance of vertical supply to the contourites. We have made a first attempt at a contourite budget for the Cadiz drift system. More detailed work to quantify all components is needed to establish a more accurate contourite budget.

Funding:
No specific funding of this study, although data has been extracted from materials from a PhD, funded by PTDF, Nigeria.