DISCUSSION

Palaeoenvironmental Implications

Irrespective of behavioral interpretations of the regularly-oriented structures, their biogenic origin, in conjunction with other palaeontolgical evidence, implies that at least episodically, the Early Jurassic southern African palaeoenvironment was capable of supporting a diverse ecosystem. Furthermore, contrasting with previous perceptions of low preservation potential of continental trace fossils (e.g., Ekdale et al. 1984), and in line with more recent reports of occasional high trace fossil diversity in aeolian systems (Loope and Rowe 2003; Ekdale et al. 2007), the ichnological record of Clarens Formation suggests that this Early Jurassic aeolian ecosystem was also suitable to preserve such biogenic sedimentary structures.

Given a biogenic origin for the structures, their occurrence, distribution, and morphology suggest episodic changes in the physico-chemical parameters of their habitat. Some of the environmental changes were most likely related to perturbations in an otherwise arid climate (e.g., changes in the local hydrogeological conditions and nutrient availability). The presence of water and temporal as well as spatial fluctuations in substrate moisture levels are indicated by both organic and inorganic evidence. For instance, considering that the surface ornamentations on the associated ribbed tubes are likely scratches on the outer surface of the burrows, their presence implies that excavation of the burrows occurred in a firm, possibly moist substrate. Lack of lining and horizontal lamination of the fill suggest that the burrows were left open, and subsequently filled by coarse and fine sand laminae, which imply several depositional events with variable energy levels. In addition, the analysis of growth rings in rare large woody taxa testifies to periods during the deposition of the Clarens Formation when good growth rates occurred in a climate with some seasonal variation, but without dramatic water shortage (Meijs1960; Bordy and Catuneanu 2002:309). Furthermore, the vertical distribution of the spheres (i.e., upward increase in abundance - Figure 9.6) and the well-defined, straight, horizontal upper surface in their distribution may reflect vertical moisture level variations along the hydrological profile and the position of the capillary zone of the water table, respectively (cf. Hasiotis et al. 2007). In case of assuming a biogenic origin for these spheres, such vertical distribution may suggest that the activities of their tracemakers were dependent on the vertical changes in substrate moisture levels which in turn limited the vertical dimension of the niche area.

A further evidence for temporal increase in substrate moisture levels is the repeated occurrence of large-scale, dramatically contorted Clarens Formation strata in the Tshipise Basin which also indicate that: (1) the disturbance of the water-saturated strata occurred prior to lithification; and (2) their rather violent disturbance occurred rapidly (cf. Horowitz 1982; Alvarez et al. 1998; Netoff 2002) most likely in an active tectonic setting (Bordy and Catuneanu 2001).

The water needed for the moist, semi-cohesive (i.e., "burrowable") substrate, the growth ring development in woody taxa, and the large-scale soft-sediment deformations, was most likely derived from episodic (seasonal?) precipitation, possibly in the form of low-frequency, high-intensity storms and from groundwater brought to the depositional sites by basinal groundwater migration systems (i.e., discharge areas). Therefore, the occurrence of trace fossils in the climate-sensitive sedimentary succession of the Clarens Formation implies episodic changes in the climate (e.g., increased substrate moisture rates) in this part of Gondwana, and thus the findings may aid the refining of the Early Jurassic global palaeoclimate zones/ patterns.

Spatiotemporal Distribution and Co-occurrence with Other Traces

The regular co-occurrence of the horizontal, regularly-oriented structures with other organism-produced traces (e.g., ribbed tubes) may suggest that they could have formed as related, structural elements of a compound trace fossil (cf. Miller III 2003) or as traces left behind by a contemporaneous biotic community of unrelated organisms. The great abundance of the oriented structures and their recurring association in geographically distant regions of southern Africa, be it as part of a complex trace fossil or as discrete, biologically unaffiliated trace fossil associations, indicate that ecological circumstances (e.g., favourable nutrient amount, palaeohydrologic regime) conducive to the generation of such biogenic sedimentary structures were widespread in southern Gondwana (cf. Hasiotis et al. 2007). More specifically, they suggest that the predominantly aeolian Early Jurassic of this region was at least episodically capable of supporting a diverse ecosystem (cf. Ekdale et al. 2007). Intermittent moist conditions in this Early Jurassic ecological system are also supported by independent indicators of water (e.g., plant and vertebrate fossils, soft-sediment deformations) throughout southern Africa. Furthermore, if the tracemaking activity related to the generation of the oriented structures was depth dependent, it can be inferred that the individual layers of horizontal traces represent consecutive tracemaking events (i.e., horizontally adjacent structures are approximately coeval), during which the tracemakers kept pace with the supposedly lower sedimentation rates. If, indeed, this increase in biological activities occurred when other physical sedimentary processes (sedimentation rates, erosion, etc.) were less dominant, presenting stable environmental niches for the burrowing organisms, then the intensely bioturbated strata may represent climate-linked localized diastems or regional unconformities. Unfortunately, the current stratigraphic resolution of the Clarens Formation is inadequate for the determination of the relative ages of the various sites or direct correlation between the geographically distant exposures. Thus, the question as to whether or not the preserved biotic communities were secluded in time and space, and flourished only in certain favourable landscape mosaics of the more arid Clarens palaeoclimate, remains open for future research. Hence, it is not possible at this time to determine the exact nature (i.e., repetitive progressive changes or near-random episodic events), number, duration and geographical extent (local vs. regional) of these climate fluctuations and related variations in the depositional system.

The Termite Issue

Without comparable modern biogenic structures of similar dimensions and spatial arrangement to the Clarens structures, the identity of the tracemakers as well as why and how these bizarre, Early Jurassic biogenic sedimentary structures were generated, remain to be found. This highlights the pressing need for neoichnological studies in arid environments, for without them, the usefulness of continental ichnotaxa for identifying physico-chemical palaeoenvironmental changes remains limited, despite the fact that such trace fossils are fairly abundant, varied and often well-preserved in lithofacies associations of continental depositional systems (cf. Hasiotis 2002; Hasiotis et al. 2007).

The identification of the tracemakers as termite-like insects is clearly tentative in this study, and it is chiefly based on the similarity of these structures to termite-made galleries in their vast abundance, strict directional orientation as well as size and overall burrow morphology. The profusion of the structures suggests an enormous number of individuals with well-coordinated and oriented behavior, a typical characteristic of social insects, especially that of some modern termites (cf. Jacklyn and Munro 2002). The interpretation of these oriented structures as ichnofossils of termite-like insects, together with the previously interpreted termite nests of the Clarens Formation (Bordy et al. 2004) would have important ramifications for the appearance of sociality in these insects. In particular, the fact that the four earliest (Late Triassic to Late Jurassic) fossil termite nest occurrences are exclusively reported from rocks formed in arid and semi-arid environments (e.g., Hasiotis and Dubiel 1995; Smith and Kitching 1997; Hasiotis 2004; Bordy et al. 2004) may imply that the onset of sociality in these insects was a behavioural response to environmental triggers. More specifically, in such stressful habitats with episodic and irregular nutrient distribution, the appearance of the evolutionary novelty of sociality seems more practical, as it may have assisted organisms in a more successful foraging. Similar explanations (i.e., aridity food-distribution hypothesis) for the evolution of sociality in the African mole-rats were offered by Faulkes et al. (1997).