ELECTRONIC SUPPLEMENTARY MATERIAL When and how did the terrestrial mid-Permian mass extinction occur? Evidence from the tetrapod record of the Karoo Basin, South Africa Michael O. Day1*, Jahandar Ramezani2, Samuel A. Bowring2, Peter M. Sadler3, Douglas H. Erwin4, Fernando Abdala1 & Bruce S. Rubidge1 1 Evolutionary Studies Institute, School of Geosciences, University of the Witwatersrand, Private Bag 3, Johannesburg 2050, South Africa 2 Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA 3 4 Department of Earth Sciences, University of California, Riverside, California 92521, USA Department of Paleobiology, National Museum of Natural History, Washington, D.C. 200137012, USA Contents: Supplementary Methods Supplementary Discussion Supplementary References Supplementary Datasets Supplementary Figures 1. SUPPLEMENTARY METHODS (a) Calculating biostratigraphic ranges for individual taxa Initially, 1:250 000 geological maps for the relevant regions were consulted (map sheet: Republic of South Africa 3122, Victoria West, 1:250 000 Geological Series 1992; map sheet: Republic of South Africa 3120, Williston, 1:250 000 Geological Series 1992; map sheet: Republic of South Africa 3220, Sutherland, 1:250 000 Geological Series 1983; map sheet: Republic of South Africa 3220, Sutherland, 1:250 000 Metallogenic Series 1997; map sheet: Republic of South Africa 3222, Beaufort West, 1:250 000 Geological Series 1979; map sheet: Republic of South Africa 3224, Graaff-Reinet, 1:250 000 Geological Series 1994; map sheet: Republic of South Africa 3326, Grahamstown, 1:250 000 Geological Series 1996). The outcrop of geological units shown in Figure 2 is primarily taken from these published geological maps, which vary in quality. West of 24°E only the members of the Teekloof Formation have been mapped, while east of this meridian not even the contact of the Middleton and Balfour formations is mapped. A review of the existing sub-formational stratigraphic paradigms for the Abrahamskraal Formation was conducted and the extensive fieldwork undertaken in this study allowed the outcrop area of particular horizons within the Abrahamskraal Formation and Poortjie Member to be identified [1]. Fossil localities were imported into Google Earth from the ArcGIS database constructed by Nicolas [2] and it was therefore possible to determine, to varying degrees of accuracy, the stratigraphic position of individual specimens. The division of the Karoo Basin into sectors was necessary as the lithostratigraphy of the Lower Beaufort Group has long been known to vary laterally in thickness and in architecture [3–5]. The stratigraphic sections used to determine the number of sectors are presented in ref. [1]. (b) Fossil identifications Numerous tetrapod genera have been described from the Tapinocephalus and Pristerognathus AZs and the validity of these had to be addressed to ensure the most reliable analyses of diversity. Of the most common taxa, only the dicynodonts, anteosaurid dinocephalians and lycosuchid therocephalians have received relatively thorough taxonomic attention; however, current projects are reviewing the taxonomy of tapinocephalid dinocephalians (S Guven), titanosuchid dinocephalians (S Jirah), gorgonopsians (C Kammerer), and temnospondyls (C Marsicano). For this study, valid therocephalian genera and the most recent specimen identifications were taken from Abdala et al. [6], Abdala et al. [7], van den Heever [8] and the personal observations of FA. The identification of dicynodont specimens relied primarily on work by Angielczyk [9], Angielczyk and Rubidge [10–12], Angielczyk et al. [13,14] and Kammerer et al. [15] and observations made by K Angielczyk and J Fröbisch of ESI collections. Dinocephalian identifications were mostly unchanged from collection catalogues with the exception of recent collections, though relying on the review of Kammerer for anteosaurids [16]. The identification of pareiasaur material was taken from the collection catalogues except for Embrithosaurus, for which a revised description was provided by Lee [17]: only the referred specimens listed by Lee and one recent specimen are accepted as Embrithosaurus. Eunotosaurus and Broomia identifications were taken from collection catalogues and the literature [18]. Basal synapsid identifications were taken from Botha-Brink and Modesto [19], Modesto et al. [20,21] and Reisz and Modesto [22]. Some museum specimens and those collected during recent fieldwork by the ESI in Johannesburg were identified by MOD, BSR or FA with reference to recent literature. The specimens included in this study are listed in Supplementary Table 1. The taxonomy of gorgonopsians has been reviewed over the last few decades [23–25] but no consensus was reached on the genera from the Tapinocephalus AZ. Kammerer suggested that most previously described gorgonopsian genera from the Tapinocephalus AZ are junior synonyms of Eriphostoma [26] and this has been recently formalised, suggesting the presence of only this one genus within the Tapinocephalus AZ and two (along with Gorgonops) in the Pristerognathus AZ [27]. In the Tropidostoma AZ, there are reports of several gorgonopsian taxa: Cyonosaurus, Lycaenops, Aelurognathus, Aelurosaurus, Aloposaurus, Gorgonops and Scymnognathus. However, as in the older zones few of the specimens attributed to this AZ are mentioned in the review studies. Only one specimen from this zone, of Aelurognathus, is widely agreed to be correctly identified. Because most genera are represented by very few specimens and the current understanding of gorgonopsian systematics has stalled at the level of the alpha taxonomy, no gorgonopsians were included in the CONOP analysis. Only Gorgonops, Aelurognathus and Eriphostoma are considered sufficiently secure to be included in Figure 1. The temnospondyls of the Tapinocephalus and Pristerognathus AZs were also excluded from the analyses as their taxonomy is uncertain; two valid genera from the Tapinocephalus AZ are considered valid, Rhinesuchus and Rhinosuchoides [28], but this is based on the type specimens only. No re-identifications were made and much of the material has been arbitrarily assigned to Rhinesuchus. (c) Uncertainty surrounding ages of biozone boundaries due to variable biostratigraphic resolution The uncertainty in the age assignment of biozones is affected by error associated with uncertainty surrounding the exact lithostratigraphic position of biostratigraphic boundaries in the southeast of the basin, from where the dated horizons of Rubidge et al. [29] are situated. Low biostratigraphic resolution below the Cistecephalus AZ in this area introduces uncertainties in the order of 300–500 meters with respect to the Tapinocephalus–Pristerognathus and Pristerognathus–Tropidostoma assemblage zone boundaries. Time equivalents of this error were calculated by estimating generalised time representation within the lower Beaufort Group in the southeast of the basin using the oldest three dated horizons of Rubidge et al. [29], for which the intervening stratigraphic distance is known (Fig. 4). This suggests that between horizon Bruce-1 (upper Koonap; 261.24 Ma) and horizon K220307-2 (uppermost Koonap; 260.41 Ma), 265 m of rock represents deposition over approximately 0.83 m.y. Between K220307-2 and horizon ‘Tortoise’ (Middleton; 259.26 Ma), 420 m of rock was deposited over 1.15 Myr. This provides an approximate time representation of 3132 years/metre in the upper Koonap Formation and 2738 y/m in the lower Middleton Formation. A 100 m uncertainty in the lithostratigraphic position of a biozone boundary could therefore reflect a corresponding temporal uncertainty of up to 0.274 Myr. Stratigraphic uncertainty calculations for individual boundaries are addressed in the Suplementary Methods. Only two known biozone boundaries are situated at or near a dated horizon: that of the Tapinocephalus–Pristerognathus AZ (this paper) and the Cistecephalus–Dicynodon AZ [29]. Error due to the lithostratigraphic contraints on these biozone boundaries is low (<50 m), which corresponds to an error of between 0.14 to 0.16 m.y. (Fig. 4). Previous to this study the age of the Tapinocephalus–Pristerognathus AZ boundary was also poorly constrained, with an stratigraphic error of perhaps 300 m (<0.94 m.y. based on upper Koonap Formation depostional rate). An age estimate of ~256.75 Ma for the Cistecephalus–Tropidostoma AZ boundary can be extrapolated from an age of 255.24 ± 0.16 Ma for the top of the Cistecephalus AZ and that of 256.25 Ma for the early-middle part of the AZ [29], with an error of around ~100 m (~0.27 m.y.). Uncertainty surrounding the age of the Pristerognathus–Tropidostoma AZ boundary is highest, as no conclusive indicator taxa for the Tropidostoma AZ have been found in the southeast of the basin (see Supplementary Information). Assuming <3.5 Ma for the Pristerognathus and Tropidostoma AZs combined using the ages posited above, an estimate of ~258.5 Ma can be tentatively suggested for the age of the Pristerognathus-Tropidostoma AZ boundary; however, until dates are obtained from better biostratigarphically constrained horizons in the southwest of the basin this comes with high lithostratigraphic error of up to 500 m (~1.37 m.y.) (Fig. 4). (d) CONOP Not all genera present in each sector were included in the CONOP analysis. Genera for which minimum stratigraphic ranges could not be calculated due to uncertainty in their provenance were excluded to avoid the obfuscation of genuine diversity and range patterns. The CONOP runs used a wider list of genera with known minimum ranges from the Eodicynodon, Tapinocephalus, Pristerognathus and Tropidostoma AZs to minimize edge-effects within the Tapinocephalus–Pristerognathus AZ transition (Supplementary Table 2). Specimen occurrences within the Tropidostoma AZ are taken from published data [30, 31]. The exclusion of genera from some sectors leads to a lower generic richness in the fully-resolved composite sequence produced by CONOP, which in turn means our estimates of extinction rates and of severity are conservative. The best-fit composite sequences show an increase in generic richness in the upper Moordenaars Member that reaches a peak in the Karelskraal Member (Fig. 3C). As any diversity pattern reconstructed from fossil remains is susceptible to collecting bias, this was preliminarily assessed by considering the number of specimens recovered from each stratigraphic bin. Because the number of specimens retrieved from each bin increases up to the top of the Abrahamskraal Formation during which the diversity crash begins, we do not consider our results to be a consequence of sampling intensity. Similarly, although the sampling intensity for the lower Poortjie member is lower than that of the upper Abrahamskraal Formation, this is bracketted by the much better sampled mid-Poortjie Member. 2. SUPPLEMENTARY DISCUSSION (a) Age of the Pristerognathus and Tropidostoma AZs Tuff sample K1202-B1 reported here was recovered from the farm Puntkraal in the Sutherland district at coordinates S 32° 22.519, E 21° 05.857. In situ, this tuff appeared as a white layer within blue and maroon overbank fines in a horizon 3.5 m above the basal sandstone bed of the Poortjie Member (7 m above the base of the member itself). This is remarkably close to the traditional lithostratigraphic placement of the Tapinocephalus–Pristerognathus AZ contact and fossil collecting in the vicinity supports this. Previously published geochronology for the lower Beaufort Group in the Eastern Cape Province [29] was based on collecting efforts conducted in the Jansenville district. In this part of the basin there is a paucity of fossils but the presence of Endothiodon in the lower Middleton Formation, close to a horizon dated at 259.262 ± 0.092 Ma, suggested that these strata belonged to the Tropidostoma AZ, from where this genus is most commonly known [31]. However, our recent collecting has demonstrated that Endothiodon is relatively abundant in the upper Poortjie Member (below the range of Tropidostoma and thus within the Pristerognathus AZ) and occurs rarely even in the middle Poortjie Member. Previous work by Day [32] indicated that Endothiodon occurred as low as the uppermost Abrahamskraal Formation, but during preparation of this paper it was found that these specimens had dubious locality data. The presence of Endothiodon is therefore not a reliable indicator for the presence of the Tropidostoma AZ, thus the age of ~259.26 Ma could correspond with the upper Pristerognathus AZ. The biostratigraphically well constrained age of 260.259 ± 0.081 Ma for the position of the Tapinocephalus–Pristerognathus AZ boundary does not cast doubt on the reliability of the existing ages but rather reflects the comparatively lower biostratigraphic resolution of the lower Beaufort Group in the south-east of the Karoo Basin. The acknowledgement of stratigraphic uncertainty surrounding biozone boundaries, particularly in the south-east of the basin, accounts for the difference between the biozone ages presented here and those suggested by Rubidge et al. [29]. SUPPLEMENTARY REFERENCES 1 Day, M. O. & Rubidge, B. S. 2014 A brief lithostratigraphic review of the Abrahamskraal and Koonap formations of the Beaufort Group, South Africa: towards a basin-wide stratigraphic scheme for the Middle Permian Karoo. J. Afr. Earth Sci. 100, 227–242. 2 Nicolas, M. 2007 Tetrapod biodiversity through the Permotriassic Beaufort Group (Karoo Supergroup) of South Africa: Unpublished PhD thesis, University of the Witwatersrand, Johannesburg, 312. (http://wiredspace.wits.ac.za/handle/10539/5391) 3 Johnson, M. R. 1976 Stratigraphy and Sedimentology of the Cape and Karoo Sequences in the Eastern Cape Province: Unpublished PhD Thesis, Rhodes University, Grahamstown, 336. 4 Jordaan, M. J. 1990 Basin Analysis of the Beaufort Group in the Western Part of the Karoo Basin: Unpublished PhD Thesis, University of the Orange Free State, Bloemfontein, 271. 5 Keyser, A. W. & Smith, R. H. M. 1979 Vertebrate biozonation of the Beaufort Group with special reference to the Western Karoo Basin. Ann. Geol. Surv. S. Afr. 12, 1–36. 6 Abdala, F., Rubidge B. S. & Van Den Heever, J. 2008 The oldest therocephalians (Therapsida, Eutheriodontia) and the early diversification of Therapsida. Palaeontology 51(4), 1011–1024. 7 Abdala, N. F, Kammerer, C. F., Day, M. O., Jirah, S. & Rubidge, B. S. 2014 Adult morphology of the therocephalian Simorhinella baini from the Middle Permian of South Africa and the taxonomy, geographic and temporal distribution of the Lycosuchidae. J. Paleontol. 88(6), 1139-1153. 8 Van Den Heever, J. 1987 The comparative and functional cranial morphology of the early Therocephalia (Amniota: Therapsida): Unpublished PhD thesis, University of Stellenbosch, Stellenbosch. 9 Angielczyk, K. D. 2007 New specimens of the tanzanian dicynodont “Cryptocynodon” parringtoni Von Huene, 1942 (Therapsida, Anomodontia), with an expanded analysis of Permian dicynodont phylogeny. J. Vertebra. Paleontol. 27, 116–131. 10 Angielczyk, K. D. & Rubidge, B. S. 2009 The Permian dicynodont Colobodectes cluveri (Therapsida, Anomodontia), with notes on its ontogeny and stratigraphic range in the Karoo Basin, South Africa. J. Vertebra. Paleontol. 29, 1162-1173. 11 Angielczyk, K. D. & Rubidge, B. S. 2010 A new pylaecephalid dicynodont (Therapsida, Anomodontia) from the Tapinocephalus Assemblage Zone, Karoo basin, Middle Permian of South Africa. J. Vertebra. Paleontol. 30, 1396–1409. 12 Angielczyk, K. D. & Rubidge, B. S. 2012 Skeletal morphology, phylogenetic relationships and stratigraphic range of Eosimops newtoni Broom, 1921, a pylaecephalid dicynodont (Therapsida, Anomodontia) from the Middle Permian of South Africa. J. Syst. Palaeontol. 11, 1–41. 13 Angielczyk, K. D., Sidor, C. A., Nesbitt, S. J., Smith, R. M. & Tsuji, L. A. 2009 Taxonomic revision and new observations on the postcranial skeleton, biogeography, and biostratigraphy of the dicynodont genus Dicynodontoides, the senior subjective synonym of Kingoria (Therapsida, Anomodontia). J. Vertebra. Paleontol. 29, 1174–1187. 14 Angielczyk, K. D., J. Fröbisch, and R. M. H. Smith. 2005 On the stratigraphic range of the dicynodont taxon Emydops (Therapsida: Anomodontia) in the Karoo Basin, South Africa. Palaeontologia Africana 41, 23–33. 15 Kammerer, C. F., Angielczyk, K. D. & Fröbisch, J. 2011 A comprehensive taxonomic revision of Dicynodon (Therapsida, Anomodontia) and its implications for dicynodont phylogeny, biogeography, and biostratigraphy. J. Vertebra. Paleontol. 31, 1–158. 16 Kammerer, C. F. 2010 Systematics of the Anteosauria (Therapsida: Dinocephalia). J. Syst. Palaeontol. 9, 261–304. 17 Lee, M. S. Y. 1997 A taxonomic revision of pareiasaurian reptiles: implications for Permian terrestrial palaeoecology. Modern Geology 21, 231–298. 18 Cisneros, J. C., Rubidge, B. S., Mason, R. & Dube, C. 2008 Analysis of millerettid parareptile relationships in the light of new material of Broomia perplexa Watson, 1914, from the Permian of South Africa. J. Syst. Palaeontol. 6, 453–462. 19 Botha-Brink, J. & Modesto, S. P. 2007 A mixed-age classed ‘pelycosaur’aggregation from South Africa: earliest evidence of parental care in amniotes? Proc. R. Soc. B 274, 2829–2834. 20 Modesto, S., Sidor, C. A., Rubidge, B. S. & Welman, J. 2001 A second varanopseid skull from the Upper Permian of South Africa: implications for Late Permian ‘pelycosaur’evolution. Lethaia 34, 249–259. 21 Modesto, S., Smith, R. H., Campione, N. & Reisz, R. 2011 The last “pelycosaur”: a varanopid synapsid from the Pristerognathus Assemblage Zone, Middle Permian of South Africa. Naturwissenschaften 98, 1027–1034. 22 Reisz, R. R. & Modesto, S. P. 2007 Heleosaurus scholtzi from the Permian of South Africa: a varanopid synapsid, not a diapsid reptile. J. Vertebra. Paleontol. 27, 734–739. 23 Sigogneau, D. 1970 Révision systématique des Gorgonopsiens sud-africains. Cah Paléont, 1–416. 24 Sigogneau-Russell, D. 1989 Wellnhofer, P. (ed) Theriodontia I: Encyclopedia of Palaeoherpetology Part 17B: Gustav Fisher, Stuttgart. 25 Gebauer, E. 2007 Phylogeny and Evolution of the Gorgonopsia with a Special Reference to the Skull and Skeleton of GPIT/RE/7113 (‘Aelurognathus?’ parringtoni): Unpublished PhD thesis, Universität Tübingen, 315. 26 Kammerer, C. F. 2014 A Redescription of Eriphostoma microdon Broom, 1911 (Therapsida, Gorgonopsia) from the Tapinocephalus Assemblage Zone of South Africa and a Review of Middle Permian Gorgonopsians In Kammerer, C. F., Angielczyk, K. D. & Fröbisch, J (eds) Early evolutionary history of the Synapsida: Springer Netherlands, Amsterdam, 171–184. 27 Kammerer, C. F., Smith, R. M. H., Day, M. O., & Rubidge, B. S. 2015 New information on the morhpology and stratigraphic range of the mid-Permian gorgonopsian Eriphostoma microdon Broom, 1911. Papers in Palaeontology, 1–21. (doi:10.1002/spp2.1012) 28 Damiani, R. J. Temnospondyls from the Beaufort Group (Karoo Basin) of South Africa and their biostratigraphy. Gondwana Res 7, 165–173 (2004). 29 Rubidge, B. S., Erwin, D. H., Ramezani, J., Bowring, S. A. & De Klerk, W. J. Highprecision temporal calibration of Late Permian vertebrate biostratigraphy: U-Pb zircon constraints from the Karoo Supergroup, South Africa. Geology 41, 363–366 (2013). 30 Rubidge, B. S. 1995 Biostratigraphy of the Beaufort Group (Karoo Supergroup). South African Committee for Stratigraphy Biostratigraphic Series 1: Council for Geoscience Pretoria. 31 Smith, R. H. M., Rubidge, B. S. & Van Der Walt, M. V. M. 2012 Therapsid biodiversity patterns and palaeoenvironments of the Karoo Basin, South Africa In Chinsamy-Turan A (ed) The Forerunners of Mammals: Indiana University Press, Indianapolis, 31–64. 32 Day, M. O. 2013 Middle Permian Continental Biodiversity Change as Reflected in the Beaufort Group of South Africa: a Bio- and Lithostratigraphic Review of the Eodicynodon, Tapinocephalus and Pristerognathus Assemblage Zones: Unpublished PhD Thesis, University of the Witwatersrand, Johannesburg. 33 Jaffey, A. H., Flynn, K. F., Glendenin, L. E., Bentley, W. C. & Essling, A. M. 1971 Precision Measurement of Half-Lives and Specific Activities of 235U and 238U. Physical Review C 4, 1889–1906. SUPPLEMENTARY DATASETS Dataset S1. Stratigraphic position of all fossil specimens used in this study. Dataset S2. Ranges of each genus in each sector of the Main Karoo Basin for CONOP. Dataset S3. U-Pb isotopic data for analyzed zircons. (a) Thermally annealed and pre-treated single zircon. Data used in date calculation are in bold. (b) Total common-Pb in analyses. (c) Measured ratio corrected for spike and fractionation only. (d) Radiogenic Pb. (e) Corrected for fractionation, spike and blank. Also corrected for initial Th/U disequilibrium using radiogenic 208Pb and Th/U[magma] = 2.8. Mass fractionation correction of 0.25%/amu ± 0.04%/amu (atomic mass unit) was applied to single-collector Daly analyses. All common Pb assumed to be laboratory blank. Total procedural blank less than 0.1 pg for U. Blank isotopic composition: 206Pb/204Pb = 18.42 ± 0.35, 207Pb/204Pb =15.36 ± 0.23, 208Pb/204Pb = 37.46 ± 0.74. Corr. coef. = correlation coefficient. Ages calculated using the decay constants λ238 = 1.55125E-10 and λ235 = 9.8485E-10 (Jaffey et al. [33]). Dataset S3 SUPPLEMENTARY FIGURES Figure S1. Date distribution plot of analyzed zircons from tuff sample K1202-B1. Bar heights are proportional to 2σ analytical uncertainty of individual analyses. Horizontal line signify calculated sample date and the width of the shaded band represents internal uncertainty in weighted mean at 95% confidence level. Arrows point to additional analyses (outliers) plotting outside the diagram. See Extended Data Table 1 for complete analytical data and text for details of date uncertainties. MSWD is mean square of weighted deviates. Figure S1