Palaeont. afr., 35,25-40(1999) PERMO-TRIASSIC FOSSIL WOODS FROM THE SOUTH AFRICAN KAROO BASIN by Marion Bamford Bernard Price Institute for Palaeontological Research, University o f the Witwatersrand Private Bag 3, WITS, 2050 Johannesburg South Africa ABSTRACT The Karoo Basin extends over more than half of the South African land surface and incorporates sediments deposited over a period of more than 100 million years, from the Upper Carboniferous to the Lower Jurassic. Biozones have been established on the basis of the abundant vertebrate fauna. Fossil plant deposits are numerous but best represented by the Lower Permian Glossopteris floras and Middle to Upper Triassic Dicroidium floras. Fossil woods occur throughout the sequence. In this paper previously described woods are discussed, newly collected woods are described and an attempt is made to correlate the woods with the Formations and vertebrate biozones. Prototaxoxylon africanum (Walton) Krausel and Dolianiti is common but restricted to the Permian (Ecca and Lower Beaufort Groups). Prototaxoxylon uniseriale Prasad has the same distribution but is rare. Australoxylon teixeirae Marguerier extends from the Ecca to the middle Beaufort. Araucarioxylon occurs throughout the Karoo but there are several species that have different ranges. Araucarioxylon africanum Bamford sp. nov. occurs throughout the Beaufort and into younger deposits. Araucarioxylon karooensis Bamford sp. nov. occurs in the Normandien Formation of the Beaufort Group. Woods with podocarpacean affinities, recognized as Mesembrioxylon, first occur in the uppermost Beaufort and extend into the Cretaceous. The woods can, therefore, be used as broadscale biostratigraphic indicators but further data need to be collected. KEYWORDS: gymnosperm woods, Araucarioxylon, Australoxylon, Mesembrioxylon Prototaxoxylon, biostratigraphy. INTRODUCTION The first good description of Karoo fossil woods was by Warren (1912) on collections in the Natal Museum but woods had been recorded earlier than that. Atherstone (1857) described the “wood beds” of the Uitenhage area (now Kirkwood Formation, Lower Cretaceous) and Mellor & Leslie (1906) described fossil root casts of Stigmaria in the Vaal River near Vereeniging which is in the Ecca Group. Bancroft (1913) described the unusual wood Rhexoxylon africanum from the Molteno Formation. The locality and whereabouts of the holotype are unknown but several Rhexoxylon species have been described from other Gondwana Triassic deposits. A taxonomic revision of the genus is needed as well as a more feasible interpretation of the growth form of the plant. Walton (1923,1925,1956), Rodin (1956), Marguerier (1973) and Erasmus (1976) have also described fossil woods from the Karoo. These woods are summarised in Table 1 together with this author’s interpretation and updating of nomenclature. Dadoxylon is the most common genus in Palaeozoic deposits but this is a most unsatisfactory taxon because there are several interpretations of its diagnostic features and no consensus. Endlicher erected this genus in 1847 and included specimens described by Lindley & Hutton (1831) and Witham (1833). The type species is usually considered to be D. withamii which has a homogeneous pith, but this feature was not noted in the original description. There is considerable nomenclatural confusion and Philippe (1993) also showed that the genus is invalid according to the Botanical Code of Nomenclature because Endlicher (1847) included all the syntypes of Pinites Witham 1833. Dadoxylon is, therefore, a synonym of Pinites but Pinites is not a homoxylous wood. Furthermore the diagnostic features are in contention; species have been described with or without pith and primary xylem, and with or without multiseriate rays. Revisions of Dadoxylon tend to be lengthy because the original diagnosis is too general. Araucarioxylon Kraus is another genus with which there is much confusion and some authors synonymise Araucarioxylon and Dadoxylon or use the names together with one or other genus in parentheses (see Giraud 1991). The present author prefers the use of “Dadoxylon” as a catch-all for woods with araucarian tracheid pitting and pith, which do not fit into the described woods with characteristic or distinctive pith types. With so much new data and more narrowly defined taxa, Dadoxylon is of lim ited biostratigraphic or phylogenetic application as a genus. Perhaps Lepekhina’s (1972) group Dadoxyleae has greater use because she made a clear distinction between woods with secondary xylem only and woods with secondary xylem and associated pith. Araucarioxylon and Agathoxylon Hartig (1848) are useful designations for woods which have secondary xylem with araucarian tracheid pitting and 26 Previously described Karoo woods from South Africa. The right hand column shows the most recent taxonomic status. * specimen is redescribed in this paper. TABLE 1 STRATIGRAPHY NUMBER LOCALITY AUTHOR WOOD COMMENT AM619 Modderpoort Walton 1925 Dadoxylon sclerosum pith MOLTENO Formation SAM 2954 Kromme Spruit Walton 1925 cf Dadoxylon sclerosum -- Dordrecht Walton 1925 Rhexoxylon priestleyi - Empangeni Erasmus 1976 Dadoxylon arberi Australoxylon teixeirae SAM 1089 K bp gat Walton 1925 Dadoxylon arberi Australoxylon teixeirae AM 508 Van Wyk's Vlei Walton 1925 Dadoxylon arberi Australoxylon teixeirae BEAUFORT NM 219 Umkomaas Walton 1925 Dadoxylon arberi Australoxylon teixeirae Group NM 11 Umkomaas Walton 1925 Dadoxylon arberi Australoxylon teixeirae SAM 4354 Weltevreden Walton 1925 ?Dadoxylon arberi ?Australoxylon teixeirae *SAM 2945 Hamsfontein Walton 1925 Spiroxylon africanum NM 160 Natal Warren 1912 Dadoxylon sp.2 Protophyllocladoxylon? AM 2296 Van Wyk's Vlei Walton 1925 Dadoxylon arberi - Tete Marguerier 1973 Australoxylon teixeirae ECCA Group -- Natal Marguerier 1973 Australoxylon natalense SAM 2946 Noro Kei Pan Walton 1925 Dadoxylon sp. SAM 2947 Van Wyk's Vlei Walton 1925 Dadoxylon arberi Australoxylon teixeirae NM (23 specimens) Natal Coal Measures Warren 1912 Dadoxylon sp.l Australoxylon teixeirae Australoxylon natalense Araucarioxylon spp. no additional distinguishing characteristics, such as the ray border cells of Dammaroxylon (Schultze-Motel 1966), clustered pits of Australoxylon (Marguerier 1973) or traumatic resin canals of Brachyoxylon (Hollick & Jeffrey 1909). Philippe (1993) said that Araucarioxylon (also a synonym of Pinites) was invalid and Agathoxylon or Brachyoxylon should be used. This author (and many others) consider that Brachyoxylon should be restricted to taxa with traumatic resin canals. Agathoxylon is rarely used but has abundant axial parenchyma and so could be used for woods with parenchyma which is not confused with resin-filled tracheids. This could mean that Araucarioxylon could be emended for woods with secondary xylem only and little to no axial parenchyma. Pant & Singh (1987) have also reviewed this group of woods but came to no conclusion about their affinities. MATERIALS AND METHODS Fossil woods have been collected from all over South Africa but only those from known biozones will be considered here. The specimens, localities and stratigraphy are summarised in Tables 2 and 3 in the Discussion. Polished thin sections have been made of all the samples using the standard method for the transverse, radial longitudinal and tangential longitudinal sections: the polished surface of a small block was mounted on a petrographic glass slide and the rest of the block was cut using a discoplan. The newly cut surface was then ground and polished to the required thickness, usually 25-80 fim, depending on the cell size and crystal size, and studied under a transmitted light microscope. Most woods are silicified but some are calcified. 27 Fossil woods studied by the author and used for the biostratigraphy. All specimens are housed in the Bernard Price Institute. TABLE 2 Group FORMATION Member LOCALITY SPECIMEN NUMBERS BP/16/ - COLLECTOR FOSSIL WOOD CLARENS Rhebokhoek 351-363 M. Bamford Mesembrioxylon sp. (?) MOLTENO Kannaskop Boesmanshoek Sterkstroom 728-729 727 392 H. Anderson P. J. Hancox Araucarioxylon sp. Araucarioxylon sp. Araucarioxylon sp. Beaufort BURGERSDORP Sterkstroom 490 395 Mesembrioxylon c f rajmahalense Mesembrioxylon sp. N O RMANDIEN Thukela, Natal Harris mith McFie's donga 550 551, 552 311, 312 313-315 316 302 C. Gow C. MacK night T. Mason Araucarioxylon karooensis Araucarioxylon africanum Araucarioxylon africanum Australoxylon teixeirae Araucarioxylon africanum BALFOUR Aberdeen 721 P. Karpeta Australoxylon teixeirae TEEKLOOF Poortjie Member Stellenbos chvlei 303-308 J. Loock Australoxylon teixeirae ABRAHAMSKRAAL Koomplaats Member Wildheuvel Swaers kraal 281, 282a, 283d 284a 298, 299 300 B. Rubidge R Heard Prototaxoxylon africanum Prototaxoxylon uniseriale Australoxylon teixeirae Prototaxoxylon uniseriale Ecca WATERFORD FORT BROWN Laingsburg 387 B. Rubidge Prototaxoxylon africanum KOOKFONTEIN Paardekraal 285, 286f B. Rubidge Prototaxoxylon africanum C OLLIN GHAM Ecca Pass 469 M. Bamford Australoxylon teixeirae The thin sections were studied under a Zeiss Axiophot microscope, measured and photographed. Descriptions of the taxa are given below. All specimens and slides are catalogued and housed in the Bernard Price Institute, Johannesburg. DESCRIPTION OF FOSSIL WOODS Prototaxoxylon africanum (Walton) Krausel & Dolianiti 1958 Specimen and Slide No: BP/16/285 Locality: Paardekraal, Sutherland Stratigraphy: Kookfontein Formation, Ecca Group, Early Permian Collector: Bruce Rubidge Figures: 1-4 The log is streaky grey and has a diameter of 25 x 20 cm. Growth rings are not visible on the prepared slide and there are numerous areas of recrystallised silica that have destroyed the otherwise well preserved cells. In transverse section the tracheids are round, oval or polygonal with an average combined wall thickness of 4-5 mm (two adjacent cells) in the early wood (Figure 1). The mean tangential diameter of earlywood tracheids is 30 |um (range 27-35 jim) and the mean radial diameter is 28 |um (range 22-32 jam). Bordered pitting occurs on the radial walls only, is mostly uniseriate and contiguous, but also sporadically biseriate, alternate and contiguous. The pits are round and have an average diameter of 12 jam . Fine spiral thickening is visible on both radial and tangential walls, less than 1 |um in width, 10-12 |am apart and almost horizontal (Figure 2). Rays are fairly short, 2-9-15 cells high (minimum, average, maximum), mostly uniseriate but some rays have a biseriate portion of 2-3 cells (Figure 3). There are 7 rays per horizontal mm but as the cell shape varies they tend to look irregular in both radial and tangential sections. Tangential ray cell walls are thin and unpitted. There are 2-5 narrowly bordered pits per cross-field, mostly oval, separate, randomly arranged and 7 jum in diameter (Figure 4). Many ray cells have a dark substance which obscures the pits. Axial parenchyma and resin canals are absent. 28 Stratigraphy of the Permian and Triassic Karoo Basin. Ecca stratigraphy from Wickens (1994). Beaufort stratigraphy from Rubidge et al. (1995) and SACS (1980). Ranges of fossil wood taxa from this study. TABLE 3 29 FIGURES 1-4. Prototaxoxylon africanum BP/16/285. 1: TS (transverse section) showing regular tracheids and no growth rings. The dark patches are a preservational artefact. Scale bar =150 pm. 2: RLS (radial longitudinal section) showing the uniseriate and biseriate alternate bordered pits and fine spiral thickenings on the radial walls of the tracheids. Scale bar = 30 pm. 3: TLS (tangential longitudinal section) showing uniseriate rays and fine spiral thickenings on the tracheid walls. Dark deposits in the cells are probably resiniferous. Scale bar = 200 pm. 4: RLS showing narrowly bordered pits in the cross-field. Scale bar = 15 pm. Figures 5-9. Prototaxoxylon uniseriale BP/16/300. 5: TS showing a partial (false) growth ring. Scale bar = 130 jam. 6: RLS with heavy spiral thickening criss-crossing on the tracheid walls, and uniseriate bordered pits. Scale bar = 50 |j.m. 7: TLS uniseriate rays. Scale bar =130 p.m. 8: RLS with single, oblique cross-field pits partially obscured by the spiral thickening. Scale bar = 15 |am. 31 Prototaxoxylon uniseriale Prasad 1982 Specimen No: BP/16/300 Locality: Swaerskraal, Laingsburg Stratigraphy: Koomplaats Member, Abrahamskraal Formation, Beaufort Group, Late Permian Collector: Rob Heard Figures: 5 - 8 The log is 9 x 6 cm in diameter and 4 cm in length, beige and browns in colour. Growth rings are distinct but the earliest earlywood is distorted and crushed, compacted into the latewood, so growth rings are difficult to measure; they range from 1-3 mm. In transverse section (Figure 5) the tracheids have roundish lumina and square to polygonal outlines (but the comers are not thickened). Wall thickness is 3-5 pm. The earlywood tracheid mean tangential diameter is 26 pm(range 17-35ja m) and mean radial diameter is 26 pm (range 20-35 pm). Bordered pits occur on the radial walls only, are uniseriate, contiguous and not compressed, with a diameter of 20 |um. Spiral thickening is biseriate, criss-crossing over the pits, 3-5 pm thick and 7-10 pm apart (Figure 6). The spirals are about 45° to the horizontal. These are definitely spirals; wall degradation was seen in some specimens and the markings are at a steeper angle and appear more broken up and erratic. Rays are uniseriate, homogeneous, 7-10-20 cells high and there are 10 rays per horizontal mm (Figure 7). Cross-field pits are mostly single, bordered and oblique and also covered by the spiral thickening (Figure 8). Very rarely there are two pits per field. The pits are 7 x 17 mm. There are no resin canals or axial parenchyma. Discussion This genus, Prototaxoxylon, was first described as Spiroxylon africanum by Walton from Harmsfontein, Cape (1925, unknown age) but, as the generic name had already been used, Krausel and Dolianiti (1958) transferred it to their new genus Prototaxoxylon. Although Walton’s slides cannot be traced, the type specimen, SAM 2945, was borrowed from the South African Museum and new thin sections made. There is no doubt that the Paardekraal and other specimens (Table 2) belong to the same taxon. Prototaxoxylon africanum - description of holotype SAM 2945 Figures: 9-12 The specimen is black, smooth and measured 8 x 4 x 3 cm but had already been sectioned. In transverse section the tracheids are square to polygonal with fairly thick walls, 7-10 pm. Growth rings are 4-7 mm wide, indistinct, with very little latewood, and the earlywood tracheid mean tangential and radial diameters are 41 and 40 pm respectively (Figure 9). Bordered pits occur on the radial walls, 1-2 seriate, contiguous, alternate and with a diameter of 12-14 pm (Figure 10). The same pitting occurs on the tangential walls but much less frequently. Rays are uniseriate with occasional biseriate portions of 1-3 cells, 2-9-14 cells high (Figure 11). Cross-field pits are round to oval with a thick, poorly preserved border. The aperture is 5-7 pm and the border 1-2 pm (total diameter 6-8 |nm). There are 2-8 of these pits, which are between cupressoid and dacrydioid, per field (Figure 12). Walton described them as “There are two to eight small pits per field. No border can be seen to the pits, though it is possible that a border may have been present.” (1925, p. 19). The spiral thickening is very delicate, less than 1 pm in thickness, 10-14 pm apart, almost at right angles to the vertical tracheid walls and passing between the tracheid pits (Figures 10-12). Emended diagnosis: Secondary xylem wood with 1-2 seriate alternate bordered pitting on radial walls of the tracheids. The pits are 12-14 (am in diameter and also occur on the tangential walls but less frequently. Fine spiral thickening on the tracheids is almost at right angles to the walls, less than 1 jam thick and 10-14 pm apart. Rays are low, uniseriate with occasional biseriate portions and the walls are thin and unpitted. There are 2-8 bordered pits per cross-field, with a diameter of 6-8 pm. Axial parenchyma and resin canals are absent. Prototaxoxylon africanum has been described only from South Africa. Several other species have been documented, mostly from India but also from Russia and Brazil (Prasad 1982). They range in age from the Carboniferous to “Mesozoic”. P. intertrappeum (Prakash & Srivastava 1962) is of questionable Tertiary age. Prototaxoxylon uniseriale was first described from the Upper Permian of India (Prasad 1982) and is quite distinct from P. africanum as it has much heavier spiral thickening. P. africanum is common in the South African Karoo with occurrences in the W aterford, Fort Brown (Ecca) and Abrahamskraal Formations (Lower Beaufort) which are Lower Permian. In South Africa, P. uniseriale has to date only been recorded in the Abrahamskraal Formation. Australoxylon teixeirae Marguerier 1973 Specimen No: BP/16/298 Locality: Swaerskraal, Laingsburg Stratigraphy: Koomplaats Member, Abrahamskraal Formation, Beaufort Group, Late Permian Collector: Rob Heard Figures: 13-16 The log measures 10 x 5 x 6 cm and is beige with orange streaks. Growth rings are distorted. Tracheids are square to rectangular in transverse section and the cell walls are 5-7.5 (im thick (Figure 13). The earlywood tracheid mean tangential diameter is 39 pm (range 35-55 pm) and the mean radial diameter is 32 * * ̂m * f * * * ** *? ; « 11 K ** * * *Z»* *• %n* **# ’ Z**'* » «? ** »*#>* *• :: n :* i *2 ■ , ! : * ; * **1 * * w *** ; i**5 * * i *» ***«*% $ »• *»*»*■ %*, ,* • » ,:**> ;% -* .?♦*. »• : ; ” ?*•-* r *, #» * 5?.•* * ♦ , #*.» *»♦*« ** , »*# * **« * *,** 1** *•*« * Z•** * «*«» * I . *. * >*«»: • * * ** ♦ £ * : „» *■* * * t ; u . * * * # % * *** § » ** ** t* t** n» * <• *» !*•«■» ** ** * * * * * s;*t * «* , *’ *’ »#** I »»*»» i* > • *%* * » * } • • • § • ; ;♦ * • *!*«* • *« j : ; • -« 5 ' t , ;* • »* ; • • « » T «** % •* *♦»' * * ** .* - »t:*r • •' , * ** * *' ; . *#* t . » *« «» »** v » ♦* ISSISliBl ** 1 ** « ' *# « * i % ; »«* *t» * ****** ? **% *» ; * •* ;* *•* ** _ *** *4 ,•<»%% -•.!•• -i ; ?5 : : ; ! • • • • ; ■• »*<*, .* t *»« *. i: * * : • ! : ; n » •,.‘i > "• •.*:* *«•*» # 11< ** « i * . «• »* , * ; • *■ r* 9 *?- a* r»§ ?* ** • i ! #‘ : ;* * ♦ : : » * . m . ' - » * . t * »Z#*%* • N : . : : ** ; - ,♦* t * * * * »**» ! li* ! ,<»•** » * * i# .*T»* ***** » •** #•* * »* * ■* •*,*«#- - I* I* * * V ***** * ** * ; * • : »-•*» * * %* ***#*» * * i*% • ff,* »* = ! • ***** • i« * ’ *^*2! > r >:i*v Figures 9-12. Prototaxoxylon africanum SAM 2945 (type specimen). 9: TS with growth ring. Scale bar = 200 (im. 10: RLS showing uni- and biseriate pitting and fine spiral thickening. Scale bar =30 (im. 11: TLS with uniseriate rays, some with resin deposits, and fine spiral thickening. Scale bar = 60 |im. 12: RLS showing narrowly bordered cross-field pits. Scale bar = 10 (im. 33 Figures 13-16. Australoxylon teixeirae BP/16/298.13: TS with distorted preservation. Scale bar = 200 fj.m. 14: RLS with clusters or stellate groups of pits. Scale bar = 50 (im. 15: TLS with uniseriate rays and occasional biseriate portions. Scale bar = 100 p.m. 16: RLS with narrowly bordered pits in the cross-field. Scale bar = 20 (im. 34 44 |um (range 25-60 |nm). Bordered pits occur on the radial walls only, are mostly 1-2 seriate, very rarely 3- seriate, opposite or alternate, and mostly in distinct clusters (Figure 14). When contiguous, the pits are still round with no compression. The pits are 10 (am in diameter with an aperture of 5 (am. Uniseriate rays predominate and occasionally there are biseriate portions of 1-4 cells. The rays are 3-8-21 cells high, homogeneous and there are 10 rays per horizontal mm (Figure 15). The cross-field pits have narrow borders, are round to oval, 7.5 (am in diameter, and there are 3-7 pits randomly arranged in the cross­ field (Figure 16). There are no resin canals or axial parenchyma. Discussion This species occurs in the Permian of southern Africa. It is very common in the Abrahamskraal Formation and the lower part of the Teekloof Formation (Lower and Middle Beaufort Group). The youngest specimen is from the contact of the Aulacephalodon and Cistecephalus Assemblage Zones in the Aberdeen area. Marguerier (1973) described two species, Australoxylon teixeirae and A. natalense from the Karoo of Mozambique and Ecca of Natal, respectively. They are easily distinguished by the quadriseriate tracheid pitting and presence of crassulae in A. natalense. Twenty one other species of Australoxylon have been described (Giraud 1991, Weaver et al. 1997) and all are from Permian deposits. The specimens of Dadoxylon arberi (Warren 1912) from Natal have definite clustering of the pits and should be placed in Australoxylon teixeirae (M arguerier 1973, and personal observation). Erasmus (1976) also described Dadoxylon arberi but from the Beaufort Series of the Empangeni district in Natal. His material also has the characteristic stellate groups or clusters of tracheid pits of Australoxylon but he overlooked the work of Marguerier. Thus, the restriction of this genus to Permian sediments in southern Africa is to be expected. Araucarioxylon Kraus 1870 Type species: Araucarioxylon carbonaceum (Witham 1833) Kraus Araucarioxylon africanum Bamford sp. nov. Holotype: BP/16/311 Locality: Near Harrismith Stratigraphy: Dicynodon Assemblage Zone, Normandien Formation, Beaufort Group, Late Permian Collector: Chris MacKnight Etymology: named for southern Africa Figures: 17-20 The log i s l 3 x l 0 x l 4 c m and is a streaky beige colour. Growth rings are distinct with 8-9 rows of thick- walled radially compressed tracheids comprising the latewood. The average width of the growth rings is 2mm. In transverse section the tracheids are square (Figure 17). The earlywood tracheid mean tangential diameter is 37 (am (range 30-45 (am) and mean radial diameter is 41 jam (range 27-50 jam). The latewood tracheid mean tangential diameter is 36 (am (range 32-48 |am) and mean radial diameter is 13 (am (range 10-20 (am). Bordered pits occur on the radial walls only, are mostly biseriate and alternate, sporadically uniseriate, contiguous to compressed and hexagonal in shape, with a diameter of 12-15 mm and aperture of 5 |am (Figure 18). Rays are uniseriate only, 2-8-18 cells high and there are 5 rays per horizontal mm (Figure 19). The cross­ field pits are araucarioid with narrow borders, 2-4-7 per field and the pits are round to oval and 6-8 (am in diameter (Figure 20). There are no resin canals or axial parenchyma. Diagnosis Secondary xylem wood with 1-2 seriate bordered pits on the radial walls of the tracheids, contiguous to slightly compressed, alternate, pit diameter 12-15 (am. Rays are uniseriate and have an average height of 15 cells. Cross-field pits are araucarioid (i.e. narrow border), oval shape, 6-8 (am and 2-4 per field, rarely up to 7. No axial parenchyma or resin canals. Resin in varying amounts in rays and/or tracheids. Araucarioxylon karooensis Bamford sp. nov. Holotype: BP/16/313 Locality: Near Harrismith Stratigraphy: Dicynodon Assemblage Zone, Normandien Formation, Beaufort Group, Late Permian Collector: Chris MacKnight Etymology: named after the Karoo Basin and Karoo Supergroup Figures: 21 -24 The specimen is 5 x 6.5 x 4 cm and beige with streaks. Growth rings are distinct, on average 6mm wide and terminated by 5-7 rows of latewood cells (Figure 21). In transverse section the tracheids are roundish and the earlywood tracheid mean tangential diameter is 37 (am (range 27-50 |am) and mean radial diameter is 39 (am (range 27-50 (am). The latewood tracheid mean tangential diameter is 30 |am (range 25-38 (am) and mean radial diameter is 17 |am (range 12-25 (am). Alternate bordered pitting occurs on the radial walls only and is mostly biseriate, but also uni- or triseriate, contiguous to hexagonal (Figure 22). The pits are 12-15 (am in diameter with an aperture of 5-7.5 (am. The rays are uniseriate with sporadic biseriate portions of 1-2 cells, 3-15-25 cells high and there are 5 rays per horizontal mm (Figure 23). The cross-field pits are round to slightly oval with a very narrow border, and there are 2-4 pits per field, 7.5-10 (am in diameter (Figure 24). There are no resin canals or parenchyma. 35 ‘ i !»!♦•!*! i ’ t i * l * f £ * # * *:> ♦ «*V** <•* * • *; % **« «♦ *% !« •;«: * ' i ' s ; * *'t «* » • # *».« S2- *i**n**m*► »* ;'<♦ “ >** ;•> * »*»2;«ti;» **t * • >*#* **#%*! X**Z ** ** * , »%* * * * * * ¥ - # *#*** < ****** ** *T:* *• * * * * * : »*#.** Figures 17-20. Araucarioxylon africanum sp. nov. BP/16/311.17: TS with growth ring. Scale bar=250 (im. 18: RLS with biseriate, alternate and compressed bordered pitting (i.e. araucarian). Scale bar = 10 |im. 19: TLS with uniseriate rays. Scale bar =30 |im. 20: RLS cross-field pits with narrow borders. Scale bar =15 |im. 36 Figures 21-24 Araucarioxylon karooensis sp. nov. BP/16/313. 21: TS with growth ring. Scale bar = 200 jam. 22: RLS showing triseriate araucarian pitting in the central part of the tracheid. Scale bar = 30 fim. 23: TLS with uniseriate rays and occasional low biseriate portions. Scale bar = 200 jam. 24: RLS with narrowly bordered cross-field pits. Scale bar = 30 |am. 37 The pit size varies in some specimens, but this may be an artefact of preservation. Diagnosis Secondary xylem wood with araucarian bordered pitting in the radial walls of the tracheids only. Pitting 1-3 seriate, contiguous, slightly compressed, alternate. Triseriate pitting must occur at least 5% of the time in the central part of the tracheid. Pits 12-15 pm in diameter. Rays predominantly uniseriate with occasional low, biseriate portions, average height 15 cells, maximum around 30 cells. 2- 4 cross-field pits, oval with a narrow border (araucarioid), and a diameter of 7.5-10 pm. No axial parenchyma or resin canals. Resin sometimes present in rays and tracheids. Discussion The two new species are very similar but Araucarioxylon africanum does not have any triseriate pitting whereas Araucarioxylon karooensis does have this feature. The rays in both species are similar in height but they are exclusively uniseriate in A. africanum whereas in A. karooensis there are occasional biseriate portions. Ray features, however, are not usually considered to be as important as pitting. The sizes of the tracheid pits are similar and so are the sizes of the cross-field pits in both species. Their numbers differ: A. africanum tends to have more cross­ field pits than A. karooensis. New species have been erected because the South African woods show some differences from the previously described species. A. africanum is similar to Dadoxylon (Araucarioxylon) dallonii (Duperon- Laudoueneix 1976) from the Cretaceous of Chad. D. dallonii has araucarian bordered pitting which is predominantly uniseriate, also biseriate, with diameters of 16-22 pm, the rays are mostly uniseriate with occasional biseriate portions, 1-29 cells high, and there are 1-2 half bordered pits per cross-field with a diameter of 9-10 pm. The differences are in the larger bordered pits of D. dallonii and the type and number of cross-field pits. A. africanum is also similar to Dadoxylon (Araucarioxylon) kiliani (Batton, 1965) from the Cretaceous of Morocco. D. kiliani has much larger bordered pits which are predominantly uniseriate and araucarian, but the rays and cross-field pits are the same as the South Africa woods. A third species which is similar to A. africanum is Dadoxylon sp. 1 (Lemoigne 1981) from the Permian of Saudi Arabia. The difference is again in the more uniseriate nature of the tracheid pitting. No pit sizes are given in the description. The differences described above are sufficient to warrant a new species designation for the South African wood. Araucarioxylon karooensis is similar to several Indian woods. A. agathioides (Krausel & Jain 1964, Jurassic) has 1-3 seriate araucarian pitting but the pits, 10-23 mm, are larger, and more variable in size than those of A. karooensis. The rays and cross-field pits of both species are virtually identical. D. amraparense (Sah & Jain 1963, Jurassic) and D. kharkhariense (Maithy 1965, Permian) are also very similar but the tracheid pitting is more randomly arranged, with some degree of opposite pitting, unlike the South African specimens. These differences justify the erection of a new species. There are numerous other species of Araucarioxylon and Dadoxylon described from Palaeozoic and Mesozoic sediments all over the world, all with fairly minor differences. The taxonomy and affinities have yet to be sorted out, however, careful description is necessary for distinguishing useful biostratigraphic types. A. africanum occurs in the Normandien Formation and in younger rocks (unpublished internal reports). A. karooensis occurs in the Normandien Formation. The Dwyka woods observed or described in the literature tend to be small, have araucarian tracheid pitting but non-araucarioid cross-field pits, and pith. According to a survey on araucarian woods (Giraud 1991), woods with multiseriate tracheid pitting are older than the biseriate ones and by the end of the Triassic 5-seriate pitting no longer occurred, by the end of the Jurassic 4-seriate pitting no longer occurred, so during the Cretaceous only uniseriate and biseriate araucarian pitting is found. No specimens of A. karooensis have yet been found in the younger sediments of South Africa either, but further collecting will confirm or refute these results and the ranges shown in Table 3 will have to be modified. Mesembrioxylon cf. M. rajmahalense Jain 1964 Specimen No: BP/16/470 Locality: Norwood, Sterkstroom. Stratigraphy: Upper Cynognathus Assemblage Zone, Burgersdorp Formation, Upper Beaufort Group, Middle Triassic Collector: John Hancox Figures: 25 - 28 The piece of wood measured 5.5 x 5.5 x 3cm and is a milky beige. The tracheids are roundish in outline and neatly arranged in radial rows (Figure 25). Growth rings are not visible. Earlywood tracheid mean tangential diameter is 24 pm (range 20-32 pm) and mean radial diameter is 29 pm (range 22-37 pm). Bordered pitting is uniseriate and contiguous but not compressed, confined to radial walls, with the pit diameters 17-20 pm (Figure 26). Rays are uniseriate, 1-8-13 cells high (Figure 27). The cross-field pits are round to oval, mostly single, very rarely 2 pits per field, and have a diameter of 12 mm (Figure 28). They are too poorly preserved to see if narrow borders are present. Other ray cell walls are thin and smooth. No resin canals or axial parenchyma occurs. Discussion Woods with abietinian pitting (uni- to biseriate, opposite, separate to contiguous but not compressed) 38 Figures 25-28. Mesembrioxylon cf. rajmahalense BP/16/470. 25: TS with no growth rings. Scale bar = 50 pm. 26: RLS with uniseriate, contiguous but not compressed bordered pits (abietinian). Scale bar = 20 pm. 27: TLS uniseriate rays. Scale bar = 20 pm. 28: RLS with single, large, simple cross-field pits. Scale bar =120 pm. 39 first occurred in the Upper Triassic (Selmeier & Vogellehner 1968, Mueller-Stoll 1986) as did the Protopinaceae which have “mixed” pitting (araucarian and abietinian). This specimen, BP/16/470, has the features of several abietinian genera, Podocarpoxylon, Phyllodadoxylon and Mesembrioxylon. These woods have affinities with the Podocarpaceae but Mesembrioxylon was erected by Seward (1919) for woods whose cross-field pits are between those of the other two genera. The South African specimen is poorly preserved but is most similar to M. rajmahalense from the Jurassic(?) Rajmahal Hills, Bihar, of India (Jain 1964). The Indian wood has mostly uniseriate and contiguous tracheid pitting, 10 x 12 pm in diameter. The rays are 1-10 cells high and there are 1-2 large, simple pits “eiporen” per cross­ field, 12-14 pm in diameter, but “sometimes broad, 10 to 20 pm in size”. The tracheid pitting in the South African specimen is more regular than the Indian species and a little bigger. Other specimens have been collected from the Molteno which also appear more podocarpacean than araucarian but the preservation is not good enough to identify them confidently. An example is BP/2/395 from the Burgersdorp Formation. More material needs to be collected and studied. Fossil woods have also been studied from the Molteno and Clarens Formations but are too poorly preserved to identify to species level. Wood collected by Heidi Anderson from Boesmanshoek and Kannaskop has been distorted and compressed, possibly recrystallized by heat after preservation. The wood is most probably araucarian. John Hancox collected several other samples from the Burgersdorp Formation which also have the same appearance. I collected wood from the Clarens Formation at Rhebokhoek. This wood is unidentifiable because none of the pitting has been preserved but appears to be more like the Mesembrioxylon type of wood than araucarian. No wood has been collected and studied from the Elliot Formation. DISCUSSION The Karoo Basin vertebrate biostratigraphy has been substantially modified and revised over the last 90 years and in the most recent publication (Rubidge 1995) the ranges of the fish, amphibians and terrestrial vertebrates have been documented within revised Assemblage Zones. No comparable biostratigraphic zonation has been proposed for plant fossils. Although plant material is abundant, most of the material is vegetative which means that the delimitation of taxonomic groups is difficult. In the Glossopteris flora for example, there are 40 to 60 leaf species but only half that number of fructifications. These are obviously difficult to reconcile and correlate. The other taxa present are the sphenopsids of which there are several genera, sterile fern fronds, lycopods, ginkgos and gymnosperms of uncertain affinity. As the leaves have not been preserved attached to the wood, the affiliations of dispersed plant organs will remain uncertain. None of the fossil plant taxa has been thoroughly studied throughout the Karoo of South Africa, and ecological differences also complicate the record, so biostratigraphical correlations have not been made. Fossil woods on the other hand, are more conservative than the leaves, and although they cannot be correlated with any particular plant group, they have a limited distribution through time. The woods, therefore, can be used for a broadscale biostratigraphy. Based on the fossil woods described above from throughout the Karoo sequence, summarised in Table 2, a preliminary biostratigraphic distribution can be drawn up. Prototaxoxylon africanum occurs from the Kookfontein Formation to the Abrahamskraal Formation (Ecca to Lower Beaufort Groups). Prototaxoxylon uniseriale is rare and has only been recorded from the Abrahamskraal Formation. Australoxylon teixeirae occurs from the Collingham Formation to the Normandien Formation and is very common. Australoxylon natalense has been recorded only from the Ecca (Marguerier 1973). Araucarioxylon africanum occurs in the Normandien Formation and in even younger sediments (Bamford, unpublished data). Araucarioxylon karooensis occurs only in the Normandien Formation. This preliminary biostratigraphy (Table 3) is based on approximately 70 identified woods from known stratigraphic horizons but further collections and studies need to be made in order to check the consistency of this correlation. ACKNOWLEDGEMENTS I would like to thank all the geologists and palaeontologists, in particular, Heidi Anderson, Chris Gow, John Hancox, Rob Heard, Paul Karpeta, Johann Loock, Chris MacNight, Tom Mason and Bruce Rubidge, who collected fossil woods from known stratigraphic horizons and allowed me to study them. Richard Lewis is thanked for preparing all the thin sections. Peter Gasson, Kathleen Pigg and an anonymous reviewer are also thanked for their constructive comments on the manuscript. REFERENCES ATHERSTONE, W.G. 1857. Geology of Uitenhage. Eastern Prov. monthly Mag. 1, 518-532. BANCROFT, N. 1913. III. Rhexoxylon africanum, a new medullosean stem. Trans. Linn. Soc. Lond. ser. 2, Bot. 8, 87-103. BATTON, G. 1965. Contribution a l ’etude anatomique et biostratigraphique de la flore du Continentale intercalaire saharien. Publ. Centre Rech. Sur les zones arides. Paris, C.N.R.S., ser. geol. 6, 37-50. DUPERON-LAUDOUENEIX, 1976. Sur la presence de Dadoxylon (A.) dallonii Boureau dans la region de Fardiallah (Tchad). 97° Congr. Nat. Soc. Sav. (Nantes 1972), sci. 4, 678. ENDLICHER, S. 1847. Synopsis Coniferarum. Sangalli Apud Scheitlin & Zollikofer 1-386. 40 ERASMUS, T. 1976. On the anatomy of Dadoxylon arberi Seward with some remarks on the phylogenetical tendencies of its tracheid pits. Palaeont. afr. 19, 127-133. GIRAUD, B. 1991. Les especes du genre Dadoxylon depuis 1962: leur repartition et leur evolution du Permien a la fin du Mesozo'ique. Rev. Palaeobot. Palynol. 67, 13-39. HARTIG, T. 1848. Beitrage zur Geschichte der Pflanzen und zur Kenntnis der norddeutschen Braunkohlen-Flora. Bot. Zeitung 6,166-172. HOLLICK, A. & JEFFREY, E.C. 1909. Studies of coniferous Cretaceous remains from Kreischerville (New York). Mem. NY Bot. Gard. 3, 1-137. JAIN, K.P. 1964. A new species of Mesembrioxylon, M. rajmahalense, from the Rajmahal Hills, Bihar, India. Palaeobotanist 13 153-154. KRAUSEL, R. & JAIN, K.P. 1964. New fossil coniferous woods from the Rajmahal Hills, Bihar. India. Palaeobotanist 12, 59-67. ------& DOLIANITI, E. 1958. Gymnospermenholzer aus dem Palaozoikum Brasiliens. Palae onto graphic a 104B, 115-137. LEMOIGNE, Y. 1981. Flore mixte au Permien superieur en Arabie Saoudite. Geobios 14, 611-635. LEPEKHINA, V.G. 1972. Woods of Palaeozoic pycnoxylic gymnosperms with special reference to North Eurasia representatives. Palaeontographica 138B, 44-106. LINDLEY, J. & HUTTON, W. 1831. The Fossil Flora of Great Britain. London. MAITHY, P.K. 1965. Studies in the Glossopteris flora of India. 19. Two new species of Dadoxylon from the Lower Gondwana of India. Palaeobotanist 13, 89-93. MARGUERIER, J. 1973. Paleoxylologie du Gondwana Africain. Etude et affinites du genre Australoxylon. Palaeont. afr. 16, 37-58. MELLOR, E.T. & LESLIE, T.N. 1906. On a fossil forest recently exposed in the bed of the Vaal River at Vereeniging. Trans, geol. Soc. S. Afr. 9, 125-128. MUELLER-STOLL, W.R. 1986. Evolutionary trends in gymnospermous wood structures during the Mesozoic - Protopinaceous woods in the German Jurassic. Palaeobotanist 35, 233-235. PANT, D.D. & SINGH, V.K. 1987. Xylotomy of some woods from Raniganj Formation (Permian), Raniganj Coalfield, India. Palaeontographica 203B, 1-82. PHILIPPE, M. 1993. Nomenclature generique des tracheidoxyles fossiles mesozoi'ques a champs araucarioi'des. Taxon 42, 74-80. PRAKASH, U. & SRIVASTAVA, S.K. 1962. On a gymnospermous fossil wood from Sitapuri, District Dhar, in Madhya Pradesh. Palaeobotanist 10, 10-18. PRASAD, M.N.V. (1982). An annotated synopsis of Indian Palaeozoic gymnospermous woods. Rev. Palaeobot. Palynol. 38, 119156. RODIN, J. 1951. Petrified forests in South West Africa. J. Palaeontol. 25, 18-20. RUBIDGE, B.S. 1995. (Ed). Biostratigraphy of the Beaufort Group (Karoo Supergroup). Council for Geoscience, Geological Survey of South Africa. South African Committee for Stratigraphy, Biostratigraphic Series No. 1, 46pp. SACS, 1980. South African Committee for Stratigraphy. Handbk geol. Surv. S. Afr. 8, 690pp. SAH, S.C.D. & JAIN, K.P. 1963. Some fossil woods from the Jurassic of Rajmahal Hills, Bihar, India. Palaeobotanist 12, 169-180. SCHULTZE-MOTEL, J. 1966. Gymnospermen-holzer aus den oberkretazischen Umzamba-Schichten von Ost-Pondoland (S-Afrika). Senck. leth. 43, 279-337. SELMEIER, A. & VOGELLEHNER, D. 1968 .Podocarpoxylon triassicum n. sp., einphylogenetisch bedeutsames “modemes” Sekundarholz aus dem Keuper von Franken. N. Jb. Geol. Palaont. Abh. 132, 70-86. SEWARD, A.C. 1919. Fossil Plants, Vol. 4. Cambridge, Cambridge University Press. WALTON, J. 1923. On Rhexoxylon Bancroft, a Triassic genus of plants exhibiting a liane-type of vascular structure. Phil. Trans. R. Soc. Lond.B. 212,79-109. — 1925. On some South African fossil woods. Ann. S. Afr. Mus. 22, 1-26. — 1956. Rhexoxylon and Dadoxylon from the Lower Shire region of Nyasaland and Portuguese East Africa. Colon. Geol. Miner. Resour. 6, 159-168. WARREN, E. 1912. On some specimens of fossil woods in the Natal Museum. Ann. Natal Mus. 2, 345-380. WEAVER, L., McLOUGHLIN, S. & DRINNAN, A. 1997. Fossil woods from the Upper Permian Bainmedart Coal Measures, Northern Prince Charles Mountains, East Antarctica. J. Austral. Geol. Geophys. 16, 655-676. WICKENS, H de V. 1994. Basin floor fan building turbidites of the southwestern Karoo Basin, Permian Ecca Group, South Africa. Unpublished PhD thesis, University of Port Elizabeth, South Africa. WITHAM, H. 1833. Observations on fossil vegetables accompanied by representations of their internal structure as seen through the microscope. Edinb. Journ. Sci. 5, 183.