Palaeoecological aspects of some invertebrate trace fossils from the mid- to Upper Permian Middleton Formation (Adelaide Subgroup, Beaufort Group, Karoo Supergroup), Eastern Cape,

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

http://www.elsevier.com/copyright

Author's personal copy

Palaeoecological aspects of some invertebrate trace fossils from themid- to Upper Permian Middleton Formation (Adelaide Subgroup,Beaufort Group, Karoo Supergroup), Eastern Cape, South Africa

Emese M. Bordy ⇑, Sean Linkermann, Rose PrevecDepartment of Geology, Rhodes University, Grahamstown 6140, South Africa

a r t i c l e i n f o

Article history:Received 3 March 2011Received in revised form 27 June 2011Accepted 29 June 2011Available online 24 July 2011

Keywords:Middle/Late PermianNon-marine trace fossilsImpoverished Mermia ichnofaciesLacustrine ichnological suiteMiddleton FormationMain Karoo BasinSouth Africa

a b s t r a c t

Ichnological and sedimentological analyses in the Eastern Cape allowed the first description of a Cochlich-nus-dominated ichnofossil site from the mid- to Upper Permian Middleton Formation (Karoo Supergroup)in South Africa. The locality is within the uppermost Pristerognathus Assemblage Zone, a biostratigraphicinterval characterized by a low vertebrate biodiversity at the turn of the mid- to Late Permian. Our fielddata indicates that the surficial bioturbation of very fine to fine-grained sand layers resulted from lifeactivities of shallow infaunal and epifaunal invertebrates (possibly annelids, aquatic oligochaetes, nem-atodes, insect larvae) and fish. The morphology of the trails, their relationship to the substrate and thebehaviour inferred from them indicate that the tracemakers developed a strategy that facilitated the opti-mization of low food resources in a permanently submerged freshwater setting. Combined ichnologicaland sedimentological evidence suggests a low-energy, freshwater lacustrine depositional environment,where occasional higher energy currents brought nutrients. Data also imply that colonization of theseerratic event beds by opportunistic sediment-feeders was short-lived and followed by longer intervalsof lower energy deposition under possibly poorly oxygenated conditions. We propose that these eventbeds as well as the sporadic red mudstones of the Middleton Formation may have formed duringshort-term, higher storm-frequency and dryer periods, signalling changes in the otherwise humid climatein this part of the main Karoo Basin during the mid- to Late Permian.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

The Middleton Formation in the Eastern Cape (a partial equiva-lent of the Teekloof Formation in SW South Africa), is essentially anon-marine, siliciclastic succession that was deposited in mean-dering rivers and their extensive overbank environments (flood-plains, lakes and ponds), within the main Karoo Foreland Basinduring the late Middle to early Late Permian, about 260 millionyears ago (Visser and Dukas, 1979; Smith, 1980; Catuneanuet al., 2005; Johnson et al., 2006). Consequently, the Middleton For-mation potentially records the biogeological effects (e.g., markedfaunal changes) of the terrestrial mass extinction event at theend-Capitanian (‘‘end-Guadalupian event’’ – Retallack et al., 2006but contrast this with Lucas, 2009). In addition to vertebrate palae-ontological assessments, detailed sedimentological, ichnologicaland palaeobotanical investigations could contribute greatly to-wards the characterization of these effects, revealing clues on bio-logical adaptation strategies, palaeo-environmental conditions,

and ultimately the reason(s) for a potential Middle Permian biolog-ical disaster.

Herein, we document for the first time sinuous trails deter-mined as Cochlichnus anguineus and Undichna britanica from theupper bedding planes of very fine- to fine-grained sandstoneswithin the upper part of the Middleton Formation. Previously, sim-ilar ichnospecies have only been recorded in the Lower PermianEcca Group, and therefore the current paper describes ichnofossilshitherto unknown from this part of the Karoo Supergroup in SouthAfrica.

2. Lithostratigraphic background

The trace fossils were found in a small outcrop in the upperMiddleton Formation, on the eastern side of the road R344,�20 km south of Adelaide (Eastern Cape, South Africa; Fig. 1).According to Johnson (1976), the Middleton Formation is a�1650–2000 m thick, mudstone-dominated unit with conformablelower and upper boundaries. Fine-grained rock types, comprising�75–80% of the formation, generally form few hundred metrethick, monotonous successions. The grain size of the mudstonesis fine to coarse silt; claystones are extremely rare. Outcrops

1464-343X/$ - see front matter � 2011 Elsevier Ltd. All rights reserved.doi:10.1016/j.jafrearsci.2011.06.002

⇑ Corresponding author.E-mail address: [email protected] (E.M. Bordy).

Journal of African Earth Sciences 61 (2011) 238–244

Contents lists available at ScienceDirect

Journal of African Earth Sciences

journal homepage: www.elsevier .com/locate / ja f rearsc i


exposing fresh rock surfaces of the formation are virtually non-existent in the Eastern Cape, thus the objective assessment of theinternal sedimentary structures of the mudstones is difficult inthe field. Notwithstanding the paucity of such outcrops, most mud-stones appear massive with subordinate horizontally laminated fa-cies. Locally, the mudstones contain <1 m thick, greyish red (5R4/2)or greyish red purple (5R4/2) units that tend to occur in definitezones in the otherwise grey mudstones. Lithostratigraphically,these maroon interbeds are distinctive features, setting the succes-sion aside from the under- and overlying formations which onlycontain greenish grey and medium to dark bluish grey mudstones.It is noteworthy that north of Grahamstown (Fig. 1), the diagnosticmaroon mudstone interbeds are increasingly abundant upwards inthe succession, but from about 550 m below the upper boundary ofthe formation, these thin, hallmark maroon mudstones graduallybecome very rare (Johnson, 1966). The formation is �1500–1650 m in thickness in this region, and �1950 m along the roadR67 (Fig. 1) (Johnson, 1976, 1966; Catuneanu and Bowker, 2001).

The sandstones of the Middleton Formation make up �20–25%of the unit and are light grey or locally medium grey, but turn olivegrey or greenish grey on weathering. The sandstone beds are fine-to very fine-grained, mainly litharenites, usually 0.3–1.5 m thick(average �1 m), lenticular, mostly massive and rarely cross-bed-ded (especially the thicker units). The sandstone beds typicallyform weakly upward-fining successions, which tend to be morecommon, coarser grained and thicker in the lower part of the for-mation at certain localities. These characteristics give an overallupward-fining profile to the Middleton Formation. Along the roadR67 (Fig. 1), Catuneanu and Bowker (2001) identified four majorupward-fining depositional sequences in the Middleton Formation(sequences E–H, see their Fig. 5). These individual upward-finingsequences range from 413 m to 574 m in thickness, comprise�10% sandstone and successively show a very slight decrease inthe overall grain size. The individual upward-fining sequenceswere interpreted as evidence for a change in the depositional stylefrom short-lived, low- to more permanent, high-sinuosity rivers,whereas the overall upward-fining profile to the Middleton Forma-tion was explained by the gradual gradient decrease of the regionalpalaeoslope (Catuneanu and Bowker, 2001).

3. Biostratigraphic background

Based on the spatial and temporal distribution of vertebrate fos-sils, the Beaufort Group of South Africa has been subdivided intoeight vertebrate assemblage zones (AZ), which, in descending ageorder are: Eodicynodon, Tapinocephalus, Pristerognathus, Tropidos-toma, Cistecephalus, Dicynodon, Lystrosaurus and Cynognathus(Rubidge, 1995). The entire Tropidostoma AZ occurs within theMiddleton Formation (Rubidge, 1995, 2005; Catuneanu et al.,2005). Vertebrate fossils characteristic of the Middleton Formationinclude amphibians, anapsids and therapsids (Rubidge, 2005).While anapsid fossil diversity decreases in the Middleton Forma-tion, some therapsid taxa (Dicynodontia, Gorgonopsia) show diver-sification in the Tropidostoma and especially in the CistecephalusAZs (Rubidge, 2005). The Pristerognathus AZ is particularly re-nowned for its relatively low vertebrate biodiversity (Nicolas andRubidge, 2010). Biostratigraphically, the trace fossil site falls with-in the uppermost Pristerognathus AZ.

4. Sedimentology of the study site

4.1. Description

The study location is a roadside drainage ditch along the roadR344 (Figs. 1 and 2) that is surrounded by sparsely vegetated hills

and shallow erosion gullies. The site, which is �0.5–0.8 m high and<10 m wide, shows evidence of recent weathering and erosion pro-cesses in a thick, monotonous succession of grey mudstones(Fig. 2). The mudstones are heavily weathered and thus detectionof primary lamination is not possible. The site also reveals twosmall horizontal surfaces: the first is an �1.2 m2, ripple-markedupper bedding plane of an �10 cm thick, mainly fine-grained sand-stone; the second (�1.2 m above the first) is a smaller, �0.5 m2

pavement on top of an �8 cm thick, very fine-grained, massivesandstone layer (Fig. 2). The ripple-marked sandstone layer is di-vided into an upper, fine-grained portion exhibiting ripple cross-

Fig. 1. Simplified geological map of the region north of Grahamstown (EasternCape, South Africa) showing the locality (white star; GPS: 32�50052.4000S26�2106.0000E), the main towns (black circles) and roads (dashed lines). Regionaldirection of stratigraphic younging: from south to north. The mid- to UpperPermian Middleton Formation is essentially structurally undeformed, and formspart of the Permian Adelaide Subgroup of the Beaufort Group (Karoo Supergroup). �

denotes informal stratigraphic name of the Abrahamskraal Formation in the EasternCape (modified after Johnson, 1976 and the 1:250 000 Geological Map Sheet 3326 ofthe Grahamstown area, 1995, Council for Geoscience, South Africa).

Fig. 2. Photograph of the study site showing the dominance of the apparentlymassive mudstones which lend themselves to low quality outcrops. White arrowsmark the two pavements on which trace fossils are observed; the lower pavement isripple-marked (not observable in this picture). Scale bar is 30 cm long.

E.M. Bordy et al. / Journal of African Earth Sciences 61 (2011) 238–244 239


lamination, and a lower, medium- to fine-grained portion withhorizontal lamination. Bioturbation structures are noted only onthe upper bedding planes and a few mm below them (i.e., veryshallow biogenic structures). In addition to the trace fossils, thereare five well-developed ripple-marks on the lower surface whichhave continuous and very slightly sinuous crests, uniformtrough-to-crest height (average: �1 cm) and fairly regular spacing(average crest-to-crest distance: �15 cm) (Figs. 3, 4A). Because oftheir asymmetric, angular profiles, these ripple-marks resemblesawtooth forms with a steeper, shorter lee side of �6.3 cm and agentler, longer stoss side of �11.4 cm on average. Having a lee tostoss ratio of 0.55 on average, these strongly asymmetrical rip-ple-marks trend roughly east–west (105–285�), with their lee sidesdipping to south.

4.2. Interpretation

The overall grain size and the bedform characteristics of the rip-ple-marks indicate that during deposition the fine sandy substratewas shaped by localized, very gentle, but fairly persistent unidirec-tional currents that were flowing from north to south with averageflow velocities of less than 0.5 m/s (cf. Allen, 1968; High and Picard,1974; Southard and Boguchwal, 1990). The normal grading andhorizontal lamination giving way to ripple cross-lamination inthe lower sandstone bed also suggest that the flow velocity ofthe current progressively diminished from probably >1 m/s to<0.5 m/s over a short period of time (possibly in a few hours) (cf.Allen, 1968; Southard and Boguchwal, 1990; Stow et al., 2009).The preservation of the full morphology of these current ripplesprovides evidence for subsequent, abiding very low-energy condi-tions. The persistence of such calm waters is signalled by the over-whelming proportion of the mudstones at the site. The episodicinterruption of the eventless suspension settling is further demon-strated by the upper, thin massive sandstone layer. Its massive nat-ure is interpreted here as a result of very rapid deposition during

another short-lived, higher energy event (cf. Miall, 1996). Water-depth estimates based solely on the above sedimentary structuresare not possible as current ripples are not water-depth diagnosticfeatures (Allen, 1968; Stow et al., 2009).

The foregoing sedimentation dynamics are in agreement withthose that occur in standing water bodies of various size and depthwithin the overbank areas of large, permanent meandering rivers(cf. Buatois and Mángano, 2002; Miall, 1996). Moreover, such a flu-vio-lacustrine environment resonates well with previous interpre-tations of the depositional setting of the Middleton Formation (e.g.,Johnson, 1976; Johnson et al., 2006).

5. Systematic Ichnology

Two ichnospecies of two ichnogenera are distinguished at thistrace fossil locality in the Middleton Formation. They include Coch-lichnus anguineus Hitchcock (1858) and Undichna britannica Higgs(1988). The illustrated specimens are still in situ, however someCochlichnus anguineus specimens have been collected and arehoused in the Albany Museum, Grahamstown (catalogue number3541).

Ichnogenus Cochlichnus (Hitchcock, 1858).Ichnospecies Cochlichnus anguineus (Hitchcock, 1858).

(Figs. 3, 4)

5.1. Description

All specimens are convex and concave epireliefs, occurring onthe upper bedding planes of sandstones at the study site (Figs. 3and 4). The specimens are horizontal, simple, fairly regularlymeandering trails, resembling imperfect sine waveforms. The chiefcharacteristics of their meandering patterns (average diameter,amplitude and wavelengths) are shown in Table 1. All dimensionsare roughly consistent within individual specimens.

Fig. 3. Photograph (A) and simple annotation (B) of the �1.2 m2, ripple-marked surface which is the upper bedding plane of a �10 cm thick, mainly fine-grained sandstone.Note that the monospecific ichnofossils (Cochlichnus anguineus) are mostly associated with the troughs of the ripples (locally nutrient richer area?). The ripple-marks have (i)continuous and very slightly sinuous crests; (ii) uniform trough-to-crest height (average: �1 cm); (iii) fairly regular spacing (average crest-to-crest distance: �15 cm); (iv)asymmetric (lee-to-stoss ratio �0.55), angular profiles; (v) roughly east–west (105–285�) orientation and vi) southward dipping lee sides. Scale bar is 10 cm long.

240 E.M. Bordy et al. / Journal of African Earth Sciences 61 (2011) 238–244


Two morphotypes are differentiated: Morphotype A is simple,smooth, sinuous endogenic tunnels (endichnia), forming slightlyconvex ridges (�1.7 mm in diameter) on the bedding plane,whereas Morphotype B is simple, exogenic grooves (epichnia) of

�2.2 mm in diameter. Locally, Morphotype A shows a transitionfrom regular sinuous to less sinuous form (Fig. 4B). The preservedlength of the specimens varies from a few cm to �20 cm. Truecross-overs are absent; the cross-over in one Morphotype A trace

Fig. 4. Overview (A) and close-up images (B, C, D, E) of selected trace fossils at the study site. A – Upper bedding plane view of the ripple-marked, lower sandstone layershowing (with white arrows) the distribution of Cochlichnus anguineus Morphotype A and Morphotype B traces. B – Close-up of Morphotype A. C – Close-up of Morphotype B.Note that the Cochlichnus anguineus traces lack annulations or lateral markings. D and its annotation in E show the upper bedding plane view of �0.5 m2, upper sandstonelayer (marked with the upper white arrow in Fig. 2). On this surface, there are Undichna britannica traces (with blue in E) and one Cochlichnus anguineus (Morphotype A) trace(with in yellow in E) and. Scale bar is 10 cm long. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Table 1Ichnometry of the trace fossils. All forms resemble imperfect sinusoidal curves (Fig. 4).

Trace fossil type Average diameter (in mm) Average peak-to-peak amplitude (in mm) Average wavelength (in mm)

Cochlichnus anguineus Morphotype A 1.7 5 10Morphotype B 2.2 7.3 15

Undichna britannica 12 150 350



(Fig. 4B) is only apparent. Specimens usually occur in isolation. Dis-tribution along the bedding plane is variable: Morphotype A iscommon (>10 specimens on a 1 m2 surface), Morphotype B is veryrare (1 specimen on a 1 m2 surface) (Figs. 3, 4B). Morphotype A ismostly associated with the troughs of the ripple-marks.

5.2. Discussion

Specimens are assigned to Cochlichnus anguineus (Hitchcock,1858) which is the type ichnospecies of Cochlichnus. It lacks annu-lations or lateral markings, unlike the other two valid Cochlichnusichnospecies: C. antarticus and C. annulatus (Buatois and Mángano,1993). Detailed ichnotaxonomic treatment of C. anguineus has beenpublished recently (Kim et al., 2005; Gibert and Sáez, 2009; Uch-man et al., 2004, 2009) and it is superfluous to repeat all of thatinformation here. In non-marine settings, Cochlichnus isp. has beeninterpreted as grazing trails produced by mobile deposit feederssuch as annelids, aquatic oligochaetes, nematodes and insect larvae(biting midge and dipteran larvae) (Hitchcock, 1858; Hasiotis,2002; Uchman et al., 2004, 2009). Chironomid dipterous insectswere also considered as potential tracemakers, but this affinityhas been questioned recently (see Uchman et al., 2009 and refer-ences therein). Besides, in our case, dipteran larvae tracemakersare unlikely as undisputable dipteran fossils are unknown priorto the early Middle Triassic (Anisian) (Krzeminski and Krzeminska,2003). In non-marine settings, the presence of Cochlichnus isp. sug-gests perennial freshwater conditions and indicates a low-energyenvironment (Hasiotis, 2002, p. 61; Buatois and Mángano, 2002),however Cochlichnus isp. has also been noted under subaerial con-ditions in wet mud (Hasiotis, 2002).

In South Africa, Cochlichnus anguineus and C. iustus were de-scribed from the Lower Permian Ecca Group. Occurrences at theZak River and Laingsburg localities (Western Cape) were first men-tioned by Anderson (1974, p. 72, 73) together with arthropodtrackways and fish trails, and were reconstructed as trace fossilassemblages of permanent, low-energy, shallow, freshwater lakes.These Ecca Group localities were referred to again by Anderson(1981), Thomas (2001) and Braddy and Briggs (2002). In similarnon-marine, Glossopteris-bearing successions, Cochlichnus antarcti-cus has been previously described from the Permian in Antarctica(Tasch, 1968a, 1968b) and Patagonia (Buatois et al., 1997).

Ichnogenus Undichna Anderson, 1976 (Fig. 4D).Ichnospecies Undichna britannica Higgs, 1988 (Fig. 4E).

5.3. Description

The specimen illustrated in Fig. 4D and E occur on the upperbedding plane of a 0.5 m2 sandstone surface and exhibit trails thatconsist of an intertwined pair of horizontal, simple, out of phase,sinusoidal waves. The chief characteristics of these cross-cutting,meandering patterns (average diameter, amplitude and wave-lengths) are shown in Table 1. All dimensions are roughly consis-tent within the individual trails. The sinusoidal trails formsimple, exogenic, sinuous furrow (epichnia) that are�12 mm wide.Locally, the flat to gently concave main epirelief furrow has a pairof slightly raised levees of �2–3 mm in height (see left hand side ofFig. 4D). The preserved length of the trails is up to 80 cm long.

5.4. Discussion

The traces display the characteristic morphology of Undichnabritanica (Higgs, 1988), which has been interpreted as multiplewaves generated in soft sediment by the ventrally-protruding finsof swimming fish in freshwater settings (Higgs, 1988; Gibert et al.,1999; Trewin, 2000). Our specimen has similar morphology to pre-

viously reported examples of Undichna britanica (e.g., Buatois et al.,1998; Gibert et al., 1999, 2000; Trewin, 2000; Minter and Braddy,2006; Benner et al., 2009), however exactly matching examples arenot reported. Considering that size alone is not regarded as a keyichnotaxobase, the size discrepancies are deemed as part of therange variation within the ichnotaxon. The identification of a spe-cific fish as the Undichna trailmaker is not possible, however byemploying Bainbridge’s (1962) formula (4 �wider wave ampli-tude), the trailmaker’s body length may be roughly estimated. Thissuggests that the trailmaker fish may have been approximately60 cm long (4 � 150 mm).

The Middleton ichnological suite shows similarities to the Mer-mia ichnofacies which characterizes ichnofossil assemblages ofwell-oxygenated, low-energy, permanently subaqueous zones ofboth deep and shallow freshwater lakes (Buatois and Mángano,1998, 2007). This ichnofacies comprises: a moderate to high diver-sity, high abundance, simple, superficial to very shallow, (a) graz-ing; (b) common feeding; and (c) rare locomotion traces. Thefeeding and grazing traces are made by mobile deposit feeders(annelids, nematodes, oligochaetes, nematomorphs, bivalves, gas-tropods), whereas the locomotion traces are mostly fish trailsand amphibian trackways (see full review in Buatois and Mángano,2007). The low abundance and low diversity in the Middletonichnological suite however, suggest that these traces represent asub-variant of the Mermia ichnofacies. These simple trails wereprobably produced by an impoverished, but opportunistic commu-nity of grazers and swimmers that briskly occupied a short-livedniche area in an otherwise stressful ecological setting with lowfood and possibly low oxygen levels. In particular, the horizontalnature and sporadic distributions of the Cochlichnus grazing tracessuggest motion in a structured manner in order to maximize foodcollection in or at the surface of soft sediment. This efficient utili-zation of feeding space was probably controlled by the sporadicdistribution of food in the uppermost part of the depositional sur-face (now bedding plane) (cf. Buatois and Mángano, 2002; Uchmanet al., 2009). Hydrodynamically, it is possible that more nutrientswere preserved in the local depressions represented by the troughsbetween the ripple crests, leading to the increased abundance oftrails preserved here. A similar explanation was provided in caseof Psammichnites isp. distribution by Mángano et al. (2003).

6. Discussion

Combined ichnological and sedimentological evidence suggeststhat the study site represents deposition in a low-energy, perma-nent lacustrine setting in the earliest mid- to Late Permian of SouthAfrica. However, the available ichnological and sedimentologicaldata do not permit any meaningful inference on the depth of theseMiddleton lakes. In general, ichnological reconstruction of lacus-trine palaeobathymetry is limited to large, deep lakes (Buatoisand Mángano, 2007, p. 307).

Factors such as low diversity, low abundance and simplicity ofthe shallow infaunal and epifaunal trace fossils at the study localityare suggestive of a stressful lacustrine ecosystem with variableenvironmental conditions (e.g., fluctuating oxygen, food and sedi-ment supply). Such inhospitable bottom conditions, absent orscarce infaunal organisms and associated impoverished ichnofaunawere apparently characteristic of Palaeozoic freshwater lakes (Gi-bert et al., 1999; Buatois and Mángano, 2002, 2007). Althoughadditional reliable indicators of anoxia in the Middleton lakes, suchas primary lamination of the grey mudstones, are not available inthe currently weathered outcrops, a further indirect evidence forinstability in lake oxygenation levels may be supplied by the hall-mark red mudstone interbeds of the Middleton Formation. The redmudstones have been associated with deposition in permanently



subaerially exposed parts of the floodplain under an overall humidclimate (Johnson, 1976, p. 270). An alternative explanation for thisis that periodic conditions of higher oxygenation levels occurred,largely due to geologically short climatic fluctuations (on 100 kyrscale?) that resulted in more seasonal weather patterns with high-er storm-frequencies. During these periods, the otherwise strati-fied lakes of the Middleton Formation became better oxygenated,with episodic storm-induced (turbidity?) currents providing coar-ser grained sediments, oxygen and nutrients to the distal parts ofthe lake bottom. These events possibly provided favourable cir-cumstances for opportunistic, low diversity, sediment-feedingcommunities, and resulted in subsequent surficial bioturbation ofthese newly deposited and relatively nutrient-rich substrates (Gi-bert et al., 2000, p. 97; Buatois and Mángano, 2007, p. 306). Concur-rently, shallower, proximal lake settings were probably subaeriallyexposed during dryer periods, hence the colouring of the sporadic,hallmark red beds.

The simplicity, low abundance and low diversity of the Middle-ton ichnofossils suggest that the colonization of the event bedsmust have been short-lived. Coinciding with a decline in favour-able ecological conditions (e.g., change in oxygenation of bottomwaters), the short colonization phase was followed by periods ofdeposition under more typical reducing conditions, when the bulkof the chiefly grey Middleton mudstones accumulated from sus-pension fall-out in mostly permanently waterlogged settings underlow-energy conditions. These inferred, episodically recurring peri-ods of anoxia however are also unreliable water depth indicators ofthese Middleton lakes as stratification of the water column mayalso develop in lakes of relatively modest depths (Trewin et al.,2002). In to order test and potentially develop the foregoinghypothesis into a working model of the depositional evolution ofthe lakes of the Middleton Formation, detailed and systematic ich-nological, sedimentological and taphonomic analyses will be re-quired in the future.

7. Conclusions

� The Middleton ichnospecies are simple, low diversity, lowabundance, grazing and locomotion trails of benthic and nek-tonic animals;� The ichnofossils occur on upper bedding planes of erratic event

beds in a lacustrine succession.� Trace fossils identified as Cochlichnus anguineus (invertebrate

grazing traces) and Undichna britannica (fish swimming traces)form part of an impoverished, permanently subaqueous Mermiaichnofacies.� Lakes are tentatively interpreted to have been deep enough for

stratification of the waterbodies to occur, with the developmentof relatively inhospitable bottom conditions (poor in oxygenand nutrients).� Ichnofossil locality is in the uppermost Pristerognathus AZ which

potentially records the mass extinction event hypothesized tohave occurred at the end of the Middle Permian; therefore cur-rent findings provide a further palaeoenvironmental piece tothe overall mosaic of information of this biocrisis.� Additional biogeological investigations needed to refine the pal-

aeoecological setting of the Middleton Formation.

Acknowledgements

We thank Conrad Labandeira for his field assistance and adviceon potential trace-makers and Robert Nagy for help with some ofthe illustrations. Research funds received from the Joint ResearchCommittee of Rhodes University (by EMB), National Research

Foundation of South Africa (by EMB, SL), a DAAD Doctoral Scholar-ship (by SL) and National Research Foundation (African OriginsGrant, UID: 65241) (by RP) are gratefully acknowledged. The fund-ing sources had no other involvement in this research. We are alsovery grateful to the Editor as well as Luis Buatois and an anony-mous reviewer for assisting in improving this manuscript.

References

Allen, J.R.L., 1968. Current ripples: Their Relation to Patterns of Water and SedimentMotion. North Holland Publishing Company, Amsterdam, 433 pp.

Anderson, A.M., 1974. Arthropod trackways and other trace fossils from the EarlyPermian Lower Karroo Beds of South Africa. Unpublished Ph.D. thesis,University of the Witwatersrand, Johannesburg, South Africa. 172 pp.

Anderson, A., 1976. Fish trails from the Early Permian of South Africa. Palaeontology19, 397–409.

Anderson, A.M., 1981. The Umfolozia arthropod trackways in the Permian Dwykaand Ecca series of South Africa. Journal of Paleontology 55, 84–108.

Bainbridge, R., 1962. Caudal fin and body movement in the propulsion of some fish.Journal of Experimental Biology 40, 23–56.

Benner, J., Ridge, J., Knecht, R.J., 2009. Timing of post-glacial reinhabitation andecological development of two New England, USA, drainages based on trace fossilevidence. Palaeogeography, Palaeoclimatology, Palaeoecology 272, 212–231.

Braddy, S.J., Briggs, D.E.G., 2002. New Lower Permian non-marine arthropod tracefossils from New Mexico and South Africa. Journal of Paleontology 76, 546–557.

Buatois, L.A., Mángano, M.G., 1993. Trace fossils from a Carboniferous turbiditiclake: implications for the recognition of additional nonmarine ichnofacies.Ichnos 2, 237–258.

Buatois, L.A., Mángano, G.M., 1998. Trace fossil analysis of lacustrine facies andbasins. Palaeogeography, Palaeoclimatology, Palaeoecology 140, 367–382.

Buatois, L.A., Mángano, M.G., 2002. Trace fossils from Carboniferous floodplaindeposits in western Argentina: complications for ichnofacies models ofcontinental environments. Palaeogeography, Palaeoclimatology, Palaeoecology183, 71–86.

Buatois, L.A., Mángano, M.G., 2007. Invertebrate ichnology of continental freshwaterenvironments. In: Miller, W., III (Ed.), Trace Fossils Concepts, Problems,Prospects. Elsevier, Amsterdam, pp. 285–323.

Buatois, L.A., Jalfin, G., Aceñaloza, F.G., 1997. Permian nonmarine invertebrate tracefossils from southern Patagonia, Argentina: ichnologic signatures of substrateconsolidation and colonization sequences. Journal of Paleontology 71, 324–336.

Buatois, L.A., Mángano, M.G., Maples, C.G., Lanier, W.P., 1998. Ichnology of an UpperCarboniferous fluvio-estuarine paleovalley: the Tonganoxie sandstone, BuildexQuarry, Eastern Kansas, USA. Journal of Paleontology 72, 152–180.

Catuneanu, O., Bowker, D., 2001. Sequence stratigraphy of the Koonap andMiddleton fluvial formations in the Karoo foredeep South Africa. Journal ofAfrican Earth Science 33, 579–595.

Catuneanu, O., Wopfner, H., Eriksson, P.G., Cairncross, B., Rubidge, B.S., Smith,R.M.H., Hancox, P.J., 2005. The Karoo basins of south-central Africa. Journal ofAfrican Earth Sciences 43, 211–253.

de Gibert, J.M., Sáez, A., 2009. Paleohydrological significance of trace fossildistribution in Oligocene fluvial-fan-to-lacustrine systems of the Ebro Basin,Spain. Palaeogeography, Palaeoclimatology, Palaeoecology 272, 162–175.

de Gibert, J.M., Buatois, L.A., Fregenal-Martinez, M.A., Mángano, M.G., Ortega, F.,Poyato-Ariza, F.J., Wenz, S., 1999. The fish trace fossil Undichna from theCretaceous of Spain: Taphonomic and palaeoenvironmental implications for theichnogenus. Palaeontology 42, 409–427.

de Gibert, J.M., Fregenal-Martínez, M.A., Buatois, L.A., Mángano, M.G., 2000. Tracefossils and their palaeoecological significance in Lower Cretaceous lacustrineconservation deposits, El Montsec, Spain. Palaeogeography, Palaeoclimatology,Palaeoecology 156, 89–101.

Hasiotis, S.T., 2002. Continental Trace Fossils. SEPM, Short Course Notes Number 51,Tulsa, Oklahoma. 132 pp.

Higgs, R., 1988. Fish trails in the Upper Carboniferous of southwest England.Palaeontology 31, 255–272.

High, L.R., Picard, M.D., 1974. Reliability of cross-stratification types as paleocurrentindicators in fluvial rocks. Journal of Sedimentary Research 44, 158–168.

Hitchcock, E., 1858. Ichnology of New England. A report on the sandstone of theConnecticut Valley, especially its fossil foot marks. W. White, Boston. 220 pp.

Johnson, M.R., 1966. The stratigraphy of the Cape and Karoo systems in the EasternCape Province. Unpublished M.Sc. Thesis, Rhodes University, Grahamstown,South Africa. 76 pp.

Johnson, M.R., 1976. Stratigraphy and sedimentology of the Cape and Karoosequences in the Eastern Cape. Unpublished Ph.D. Thesis, Rhodes University,Grahamstown, South Africa. 336 pp.

Johnson, M.R., van Vuuren, C.J., Visser, J.N.J., Cole, D.I., Wickens, H.de V., Christie,A.D.M., Roberts, D.L., Brandl, G., 2006. Sedimentary rocks of the KarooSupergroup. In: Johnson, M.R., Anhaeusser, C.R., Thomas, R.J. (Eds.), TheGeology of South Africa. Geological Society of South Africa, Johannesburg,Council for Geoscience, Pretoria, pp. 461–499.

Kim, J.Y., Keighley, D.G., Pickerill, R.K., Hwang, W., Kim, K.-S., 2005. Trace fossilsfrom marginal lacustrine deposits of the Cretaceous Jinju Formation, southerncoast of Korea. Palaeogeography, Palaeoclimatology, Palaeoecology 218, 105–124.



Krzeminski, W., Krzeminska, E., 2003. Triassic Diptera: descriptions, revisions andphylogenetic relations. Acta Zoologica Cracoviensia 46 (suppl.– Fossil Insects),153–184.

Lucas, S.G., 2009. Global Middle Permian reptile mass extinction: the Dinocephalianextinction event. Geological Society of America Abstracts with Programs 41,360.

Mángano, M.G., Buatois, L.A., Rindsberg, A., 2003. Carboniferous Psammichnites:systematic re-evaluation, taphonomy and autecology. Ichnos 9, 1–22.

Miall, A., 1996. The Geology of Fluvial Deposits. Blackwell Scientific Publications,Oxford, 582 pp.

Minter, N.J., Braddy, S.J., 2006. The fish and amphibian swimming traces Undichnaand Lunichnium, with examples from the Lower Permian of New Mexico, USA.Palaeontology 49, 1123–1142.

Nicolas, M., Rubidge, B.S., 2010. Changes in Permo-Triassic terrestrial tetrapodecological representation in the Beaufort Group (Karoo Supergroup) of SouthAfrica. Lethaia 43, 45–59.

Retallack, G.J., Metzger, C.A., Greaver, T., Jahren, A.H., Smith, R.M.H., Sheldon, N.D.,2006. Middle-Late Permian mass extinction on land. Geological Society ofAmerica Bulletin 118, 1398–1411.

Rubidge, B.S., 1995. Biostratigraphy of the Beaufort Group (Karoo Supergroup).Council for Geoscience, Geological Survey of South Africa, SACS BiostratigraphicSeries 1. 46 pp.

Rubidge, B.S., 2005. Re-uniting lost continents – fossil reptiles from the ancientKaroo and their wanderlust. South African Journal of Geology 108, 135–172.

Smith, R.M.H., 1980. The lithology, sedimentology and taphonomy offloodplain deposits of the Lower Beaufort (Adelaide Subgroup) stratanear Beaufort West. Transactions of the Geological Society of South Africa3, 399–402.

Southard, J.B., Boguchwal, L.A., 1990. Bed configurations in steady unidirectionalwater flows, 2: synthesis of flume data. Journal of Sedimentary Petrology 60,658–679.

Stow, D., Hernandez-Molina, F.J., Llave, E., Sayago-Gil, M., Diaz del Rio, V., Branson,A., 2009. Bedform-velocity matrix: the estimation of bottom current velocityfrom bedform observations. Geology 37, 327–330.

Tasch, P., 1968a. Trace fossils from the Permian Polarstar formation Sentinelmountains, Antarctica. Kansas Academy of Science Transactions 71, 184–194.

Tasch, P., 1968b. A Permian trace fossil from the Antarctic Ohio range. KansasAcademy of Science Transactions 71, 33–37.

Thomas, C.G.C., 2001. Sedimentology and Stratigraphy of the Falkland IslandsPermian with comparisons to Gondwana stratigraphy of South Africa and SouthAmerica. Unpublished Ph.D. Thesis, University of Aberdeen [various pages].

Trewin, N.H., 2000. The ichnogenus Undichna, with examples from the Permian ofthe Falkland Islands. Palaeontology 43, 979–997.

Trewin, N.H., MacDonald, D.I.M., Thomas, C.G.C., 2002. Stratigraphy andsedimentology of the Permian of the Falkland Islands: lithostratigraphic andpalaeoenvironmental links with South Africa. Journal of the Geological Society159, 5–19.

Uchman, A., Pika Biolzi, M., Hochuli, P.A., 2004. Oligocene trace fossils fromtemporary fluvial plain ponds: an example from the Freshwater Molasse ofSwitzerland. Eclogae Geologicae Helvetiae 80, 133–148.

Uchman, A., Kazakauskas, V., Gaigalas, A., 2009. Trace fossils from Late Pleistocenevarved lacustrine sediments in eastern Lithuania. Palaeogeography,Palaeoclimatology, Palaeoecology 272, 199–211.

Visser, J.N.J., Dukas, B.A., 1979. Upward-fining fluviatile megacycles in the BeaufortGroup, north of Graaff-Reinet, Cape Province. Transactions of the GeologicalSociety of South Africa 82, 149–154.


Palaeoecological aspects of some invertebrate trace fossils from the mid- to Upper Permian Middleton Formation (Adelaide Subgroup, Beaufort Group, Karoo Supergroup), Eastern Cape,

Documents