SIMILARITIES and DIFFERENCES in DICOTYLEDONOUS WOODS of the CRETACEOUS and PALEOCENE. SAN JUAN BASIN, NEW MEXICO, USA E.A. Wheel

IAWA Journal, Vol. 16 (3),1995: 223-254

SIMILARITIES AND DIFFERENCES IN DICOTYLEDONOUS WOODS OF THE CRETACEOUS AND PALEOCENE. SAN JUAN BASIN, NEW MEXICO, USA by

E.A. Wheeler i , J. McCIammer 2 & C.A. LaPasha i

SUMMARY Fossil wood is common in the Late Cretaceous and Early Paleocene of the San Juan Basin, New Mexico. Six types of dicotyledonous wood are recognized: Paraphyllanthoxylon arizonense Bailey, Paraphyllanthoxy• Ion anasazi sp. nov., Plataninium piercei sp. nov., Metcalfeoxylon kirt• landense gen. et sp. nov., Chalkoxylon cretaceum gen. et sp. nov., Carl• quistoxylon nacimientense gen. et sp. nov. Woods with the characteristics of Paraphyllanthoxylon arizonense Bailey are the most common and occur in the Cretaceous Kirtland Shale and the Paleocene Ojo Alamo Sandstone and Nacimiento Formation. This wood type's characteristics are stable from the Cretaceous to the Paleocene. There were no signifi• cant differences in the vessel diameters, vessel densities, ray sizes, or estimated specific gravities of the P. arizonense woods from the Late Cretaceous (Kirtland Shale) and Early Paleocene (Nacimiento Formation and Ojo Alamo Sandstone). Based on the samples examined for this study, dicotyledonous woods were more diverse in the Cretaceous (five types) than in the Paleocene (two types) of the San Juan Basin. Diame• ters of the Cretaceous woods examined ranged from 14-40 cm indicat• ing they were trees rather than shrubs; diameters of the Paleocene woods examined ranged from 10-80 cm. All the woods have generalized struc• ture with combinations of features seen in more than one extant family, order, or subclass. Information from databases for fossil and extant woods indicates that some combinations of features (e. g., solitary narrow vessels, low vessel density and scalariform perforation plates, as seen in Metcalfe• oxylon kirtlandense and Chalkoxylon cretaceum ), while relatively com• mon in the Cretaceous, represent strategies of the hydraulic system that are extremely rare in the Tertihry and at present. None of the dicotyle• donous woods have distinct growth rings, although some samples of Paraphyllanthoxylon arizonense from the Paleocene show variations in vessel density and vessel diameter that may correspond to seasonal vari• ations in water availability. Key words: Fossil wood, Cretaceous, Paleocene, Paraphyllanthoxylon, Plataninium, San Juan Basin, Kirtland Shale, Ojo Alamo Sandstone, Nacimiento Formation, paleobotany.

1) Department of Wood and Paper Science, North Carolina State University, Raleigh, N.C. 27695- 8005, U.S.A. 2) Connecticut Valley Environmental Services, Inc., Charlestown, N.H. 03603, U.S.A.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 224 IAWA Journal, Vol. 16 (3),1995

INTRODUCTION

The San Juan Basin of New Mexico has long been of paleontological and archaeologi• cal interest, containing remains of dinosaurs and early mammals, and pre-historic Indian sites, The occurrence of well preserved woods in the Cretaceous and Paleocene of the San Juan Basin was noted by early geological explorers (e, g., Brown 1910; Sinclair & Granger 1914) and was used to help identify the beds of the Ojo Alamo Sandstone (Baltz et al. 1966). However, the anatomical structure and affinities of the petrified woods have not been detailed. One report mentions an unidentified fossil dicotyledonous wood and illustrates a palm stump that was one of many appearing to be in growth position in the Fruitland Formation near Cuba, New Mexico (Tidwell et al. 1981). An extensive stump field some 10 km southeast of Bisti, New Mexico, in the transitional Fruitland-Kirtland Shale (Late Cretaceous) has been identified by the U. S. Bureau of Land Management as a valuable paleontological resource and was designated a Research Natural Area in the San Juan Basin Wilderness Protection Act. This 'petrified forest' has several hundred stumps and logs and three forest levels; it is reported to be essentially monotypic and comprised oftaxodiaceous conifers (Wolberg et al. 1988). Tidwell et al. (1981) noted that seven species of conifers occur in the Fruitland-Kirtland floras, with one ofthe seven being a type of wood appearing similar to Cupressinoxylon coloradoense which was described from the Vermejo flora of the Raton Basin, New Mexico. Some work has been done on the compression floras of the San Juan Basin. Ac• cording to Tidwell et al. (1981) dicotyledonous angiosperms were the most diverse and most abundant plants of the Fruitland and Kirtland Shale. Robison et al. (1982) also found dicotyledonous leaves the most common at one locality in the Kirtland Shale. McClammer has collected over 7,000 identifiable leaf impressions from the Kirtland, Ojo Alamo, and Nacimiento Formations. These are available for study and are housed in the paleobotanical collections at Yale University. All 29 of these leaf localities are dominated by a diverse dicotyledonous flora. Most paleontological work has concentrated on the fauna (e.g., papers in Lucas et al. 1981) and Lucas's (1981: 348) remark "Without doubt the most poorly understood fossils in the San Juan Basin are those of plants" still applies. The objective of this paper is to contribute to the knowledge of Cretaceous and Paleocene plants of the San Juan Basin, New Mexico, by describing six types (17 sam• ples) of well-preserved dicotyledonous woods from the upper Cretaceous (Late Campanian and Early Maastrichtian) and lower Paleocene (Puercan land mammal stage) and comparing them to other Cretaceous and Paleocene dicotyledonous woods known from western North America. This work continues a study begun by the late Lee Pierce while he was a student at Yale University. There are relatively few reports of Cretaceous dicotyledonous woods; there are less than 150 records worldwide (Wheeler & Baas 1991, 1993), and approximately one• third of these records are small fragments from the Late Cretaceous of California (Page 1979, 1980, 1981). For the Paleocene, there are fewer than 40 dicotyledonous woods known worldwide. Thus, these San Juan woods, all representing mature wood sam-

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 225 pIes, some from large diameter logs, expand our knowledge of the past distribution and diversity of dicotyledonous trees during the Cretaceous and Paleocene.

MATERIALS AND METHODS

Locality and specimens The samples were collected in the west-central portion of the San Juan Basin, south of Farmington, New Mexico, and north of Chaco Canyon National Monument (Fig. 1, 2). All specimens and detailed information on the stratigraphic sections are deposited in the Paleobotanical Collections of the Peabody Museum, Yale University. All sam-

o 500 1 I Miles

Fig. I. Map showing location of San Juan Basin in North America. Redrawn from information in Fassett & Hinds (1971), Baltz et al. (1966).

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 226 IAWA Journal, Vo!' 16 (3), 1995

I .r' 37' ---'---bH-

c( oZ N or: c( I I T

, ~_~_C~_ I Sandoval Co

36 San JuanCo -McKi~ley Co------~c...... ;;;..o __ -

o 10 20 3,0 Miles I I ,

Fig. 2. Index map of San Juan Basin, New Mexico. Redrawn from information in Fassett & Hinds (1971). Numbers refer to measured sections by McClammer. Kkf = Cretaceous Kirtland Shale, T =Tertiary (includes Ojo Alamo Sandstone and Nacimiento Formation). Section 5 includes wood localities W14 (sample YPM 30151), W16 (samples YPM 30153, 30154); Section 8 includes wood locality WI (YPM 30144, 30145); Section 14 includes wood locality W4 (YPM 30150, 30152); Section 20 includes wood localities W6 (YPM 30147,30159), WlO (YPM 30155, 30156); Section 21 includes wood locality W8 (YPM 30149). pIes are cited by their Yale Peabody Museum (YPM) collection numbers. Table 1 lists the dicotyledonous wood specimens in relative stratigraphic order along with the lo• cality number, numbered stratigraphic section, depositional environment and name assigned to the wood. Locality W5 occurs in the lower Kirtland Shale and is correlated with the Late Campanian; localities WlO, W7, W14, WI8, and W6 occur in the middle and upper Kirtland Shale and are correlated with the Early Maastrichtian; localities WI and W16 occur in the Ojo Alamo Sandstone, localities W8, W4, and W2 occur in the Nacimiento Formation, these are from the type area of the Puercan Provincial Stage of the lower Paleocene. The minimum diameters of the source axes of the samples were estimated by using the divergence of the rays and curvature of 'rings' in the hand samples used for sec• tioning. All samples are silicified wood, and ground thin sections of cross, radial, and tangential surfaces were prepared.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 227

Table l. Source and identification of Paleocene and Cretaceous dicotyledonous woods from the San Juan Basin.

Formation Locality Section Collection No. Depositional environment Name

Paleocene Nacimiento W2 11 YPM 30157 flood plain Carlquistoxylon nacimientense W4 14 YPM 30152 flood plain Paraphyllanthoxylon arizonense W4 14 YPM 30150 flood plain P. arizonense W8 21 YPM 30149 flood plain P. arizonense

OjoAlamo WI 8 YPM 30144 alluvial flood plain P. arizonense WI 8 YPM 30145 alluvial flood plain P. arizonense W16 5 YPM 30153 alluvial flood plain P. arizonense W16 5 YPM 30154 alluvial flood plain P. arizonense

Cretaceous Kirtland [upper] WlO 20 YPM 30155 upper channel fill, or levee P. arizonense WlO 20 YPM 30156 upper channel fill, or levee P. arizonense W7 20 YPM 30148 flood plain Chalkoxylon cretaceum W14 5 YPM 30151 flood plain P. arizonense W18 20 YPM 30146 flood plain Plataninium piercei W6 20 YPM 30159 flood plain Paraphyllanthoxylon anasazi W6 20 YPM 30147 flood plain P. anasazi [lower] W5 16 YPM 30158 coastal plain Metcalfeoxylon kirtlandense point bar

Depositional environments - ecological setting The flora of the upper Fruitland Formation and the lower Kirtland Shale grew in a coastal environment, drained by northeastward flowing rivers, along the western edge of the Cretaceous continental (epeiric) sea (Fig. I). A worldwide drop in the sea level and concurrent uplift (beginning ofthe Laramide Orogeny) to the north and northeast (and possibly to the east) resulted in the retreat of the sea from the present-day San Juan Basin. The flora of the upper Kirtland Shale, the Ojo Alamo Sandstone and the Nacimiento Formation grew in a somewhat more inland environment onto which southward flowing rivers deposited sediments. Based on features of the vertical profile, lateral profile, and geometry of sedimen• tary units and on the texture, structure, mineralogy, and faunal composition of the sedi• ments' five successive sedimentary facies are apparent. These are 1) the upper Fruitland Formation and lower Kirtland Shale, 2) the middle and upper Kirtland Shale, 3) the

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 228 IAWA Journal, Vol. 16 (3), 1995

Ojo Alamo Sandstone, 4) Interval A of the Nacimiento Formation, and 5) Interval B of the Nacimiento Formation. Upper Fruitland Formation and lower Kirtland Shale - This interval consists of laterally continuous coals up to 2.5 meters thick; drab coloured carbonaceous shales and siltstone that vertically and laterally grade into sandstone and coal; lenticular cross• stratified medium-grained sandstone with occasional manganese oxide nodules; and terrestrial vertebrates, abundant leaf impressions and petrified trunks in situ. These features suggest a coastal plain facies with sluggish meandering streams and extensive interchannel swamps. The predominant drab gray colours and abundant car• bonaceous material indicate a reducing environment. Locality W 5 in the lower part of the Kirtland Shale occurs in lenticular fine-grained sandstone with low angle cross-beds that suggest a point bar facies. Middle and upper Kirtland Shale - From the bottom to the top of this interval there is a decrease in carbonaceous beds, an increase in grain size, an increase in the brightness (colour) of the fine-grained sediments, and an increase in the lateral conti• nuity of the lenticular sandstone beds. These features suggest a flood plain facies that initially contained a suspended load meandering river facies and a reducing flood basin facies. Up-section, the increase in coarse bed load deposits over fine-grained suspended load deposits suggests higher energy, laterally migrating streams of low sinuosity. The presence of ferric-oxide stains and the absence of carbonaceous deposits suggest a well-drained, oxidizing, flood• basin environment (Lehman 1985). Localities W6, W18, and Wl4 occur in silty mudstone inferred to be flood-basin deposits. Locality W7 occurs in me• dium to coarse-grained sandstone that suggest lower channel-fill deposits. Locality WlO occurs in ripple-lami• nated fine-grained sandstone inter• bedded with planar beds of coarse sandstone that suggests upper chan• nel fill or levee facies. Ojo Alamo Sandstone - The 're• stricted' Ojo Alamo Sandstone con• sists of discontinuous sheet-like, ero• sion-based conglomeratic sandstone. Finer-grained deposits are rare. Where present they consist of fine-grained sandstone and siltstone, and carbon• aceous shale. The sandstones are drab yellow, locally cross-bedded, and con• tain abundant siliceous and volcanic pebbles. They also contain large pieces offossil wood (Fig. 3), and 'cannon• ball' concretions. Fig. 3. Log in the Ojo Alamo Sandstone.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 229

These deposits have been interpreted by Powell (1973) and Sikkink (1987) as a braided stream facies (the distal portion of the alluvial fan or piedmont association of Reineck & Singh 1980). Localities WI and W16 occur in conglomeratic, very coarse• grained, dark yellowish orange sandstone with trough, and tabular cross-beds and lag pebbles 1-2 cm in diameter. Intervals A and B of the Nacimiento Formation - Interval A is characterized by gradual facies changes, laterally continuous sandstone, the absence of abandoned-chan• nels, evidence of extensive pedogenesis, and lower sandstone / siltstone ratios. This stratigraphic interval corresponds to the Puercan Stage. Interval B is characterized by rapid facies changes, lenticular sandstone, abandoned-channels, less evidence of pedo• genesis, well-laminated flood basin deposits, more megaflorallocalities, fewer mam• mal fossils, and higher sandstone/siltstone ratios. This stratigraphic interval correlates with the Torrejonian Stage. Interval A of the Nacimiento Formation. This interval consists of shale, siltstone, and sandstone that appear banded from a distance; lignitic deposits and fossil plants are rare, but concretions, manganese oxide nodules, barite, silcrete, and mammal fos• sils (of both the Ectoconus and Taeniolabis zones) are common. The drab yellowish gray to white sandstone is fine to medium-grained, cross-bedded, and laterally con• tinuous. The siltstones are vary-coloured (brownish gray, olive green, and maroon), lack well-defined bedding, and are nodular and slickensided. Lenticular, laterally continuous sandstone with epsilon cross-beds, the absence of carbonaceous deposits, and the absence of abandoned-channels suggest either 1) a low sinuosity river with suspended load or 2) a low energy meandering river with little basin subsidence where the fine-grained sediments have been reworked. The relatively large amount of fine-grained flood-basin deposits, the absence of pebbles, and the gradual facies changes favor interpretation of deposition by a low energy meandering stream on a broad, nearly featureless flood plain. The mudstones represent flood-basin deposits that have been extensively altered by pedogenic processes. Evidence for paleosol formation includes the tabular geometry, lateral persistence of beds, predomi• nance of silt, negligible organic matter, slickensides, nodular fractures, absence of bed• ding or poor bedding, development of cyclic colour-banded horizons, red beds, col• our-banding independent of grain-size, gley-mottling in colour zones, barite, manga• nese nodules, calcium carbonate nodules, and scattered disarticulated vertebrate fos• sils with bite marks. These soils, many of which are well-differentiated, suggest slow accumulation in the flood basin. This interval contains locality W8 where logs, some over 1.5 meters in length, occur. Interval B ofthe Nacimiento Formation - This interval consists oflenticular, fining• upward, epsilon cross-bedded, medium to fine-grained, soft, yellowish light olive-gray sandstone; tabular, olive-gray, poorly bedded to laminated siltstone; and lenticular, brownish gray, poorly bedded to laminated carbonaceous shales. Concretions, iron spherules, silcrete, fossil plants, turtle and crocodile fossils are common, but mammal fossils are rare. The rapid facies changes, low sandstone/siltstone ratio, and abandoned-channel sequences suggest a meandering river facies. Localities W4 and W2 occur in a flood• basin facies.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 230 IAWA Journal, Vol. 16 (3),1995

Descriptions and 'identifications' Lists of extant species with combinations of characters similar to the fossils were made using the OPCN (Oxford / Princes Risborough / Centre Technique Forestier Tropi• cal/ North Carolina State U ni versity) database for extant woods (5260 entries, Pearson & Wheeler 1981; Wheeler et al. 1986; LaPasha & Wheeler 1987), and comparisons of these woods to other fossil woods begun by reference to a database for fossil dicotyle• donous woods (Wheeler & Baas 1991, 1993). Descriptive terminology of the fossil woods generally conforms to that recommended by an IAWA Committee (1989). The vulnerability indices (mean tangential diameter/number of vessels per square mm, Carlquist 1988) were calculated.

Statistics Pairwise comparisons of means, based on the studentized maximum modulus and Sidak's uncorrelated-t inequality (yielding Hochberg's GT2 method when sample sizes are unequal), were run for MTD (mean tangential diameter of vessels), VMM2 (ves• sels per sq. mm), and RH (ray height) using PC-SAS version 6.04 proc GLM, means/ smm option. This controls the experimentwise error rate of falsely rejecting the null hypothesis (no difference between means) to the alpha level of 0.05. Lack of signifi• cant difference between means at the 0.05 level does not imply that the means are the same, only that they cannot be distinguished statistically based on this sample size. The 95% confidence intervals for each mean were estimated as the mean ± 1.96 times the sample standard deviation.

Estimates of specific gravity The specific gravity (sg) of cell wall substance is similar in unrelated woods (coni• fers and dicotyledons). The sg of swollen cell wall substance is close to 1.05 (com• puted from data in Stamm 1964; C. A. Hart, N. C. State University, personal communi• cation). We are assuming that the cell walls in silicified woods are preserved in a swol• len state because what is known of the silicification process indicates that impreg• nating silica must be water-borne and infiltrate cell walls to produce well-preserved woods (Leo & Barghoom 1976). With an estimate of the percentage of the fossil wood that is cell wall, it is possible to estimate the original sg g (specific gravity based on oven-dry weight and green volume) of the fossil. For example, if 30% of a wood's cross-sectional area is cell wall, then an estimated sg g would be 1.05 x 0.30 = 0.315. A dot grid method was used to determine cell wall percentage from well-preserved areas of cross sections, and was based on a 500 point count per sample.

Nomenclature and Classification There continue to be differences in philosophy about assigning names to fossil plants, particularly to fossil woods with generalized structure seen in more than one extant family or order. Chapman (Chapman & Smellie 1992) provided a useful discussion and considerable argumentation for not using binomials for fossil wood. Chapman and Smellie (1992) described 3 types of Cretaceous angiosperm woods as palaeotaxa. Their terminology emphasized ray structure and they incorporated into the paleotaxon names

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 231 ray terms that are not mutually exclusive: Dicotwood-heterorays, Dicotwood• multiseririlYs, Dicotwood-dumpirays. Heterorays is intended to indicate a wood with heterocellular rays; multiseriray is intended to indicate wood with multiseriate rays. These terms are not mutually exclusive. Most of the known Cretaceous dicotyledonous woods have heterocellular multi seriate rays (Wheeler & Baas 1991), and over 100 of the known Cretaceous woods could be assigned to either palaeotaxon - Dicotwood• heterorays or Dicotwood-multiserirays. Also, ray structure varies considerably with cambial age, and is not as constant a character as other features of dicotyledonous woods (e. g., perforation plate type, intervessel pit type, vessel arrangement, paren• chyma type). Page (1979, 1980, 1981) worked on a large suite of fossil Cretaceous woods (more than 200 samples representing 47 different entities). After considerable comparative work and reference to classic papers on systematic anatomy, she devised a classifi• cation scheme based primarily on vessel element features (perforation plate type, intervessel pitting, vessel groupings), features that have been studied in great detail and have long been considered to be of ecologic and phylogenetic significance (Metcalfe & Chalk 1950; Baas 1986; Carlquist 1988). Thus, if a worker chooses to use palaeotaxa we suggest it would be best to follow Page and reference vessel features or both vessel and ray features, rather than ray features only, in the taxon name. One of the more enticing reasons for using palaeotaxa is, of course, the time it saves for the investigator(s). Creation of biorecords and palaeotaxa, rather than species or genera, involves less comparative work than creating species names; thus the informa• tion on a suite of fossil woods can be published without having to deal with the often frustrating and confusing nomenclatural problems of fossil woods. Many of the names for fossil woods were created early in this century, and it is often difficult, if not impos• sible, to locate the types. The original descriptions are very brief, and apply to many different groupings of extant woods. Most Cretaceous angiosperm woods have gener• alized features and cannot be assigned to a single extant genus, family, or at times order and subclass; thus, for woods of this age, use of paleotaxa seems particularly appealing. Certainly, no one who works with isolated vegetative fragments of Cretaceous and early Tertiary plants considers them equivalent in rank to biological species, but to be morphological entities. And equally certainly, some of the names for fossil woods are unfortunate in the affinities they imply. For example, Ulminium implies affinities to Ulmaceae, but woods of this genus have features of Lauraceae. It is highly probable that some species of the fossil wood genus Paraphyllanthoxylon represent different genera, families, or even orders or subclasses. The genus Paraphyllanthoxylon was es• tablished for a Cretaceous wood with features seen in the phyllanthoid Euphorbiaceae (Bailey 1924), yet Paraphyllanthoxylon marylandense wood was attached to a repro• ductive structure with characteristics of the Lauraceae (Herendeen 1991a). Nonethe• less, a binomial of a fossil wood represents a combination of anatomical features, and in that sense facilitates comparison and developing evolutionary and ecologic infer• ences as much if not more than a biorecord or palaeotaxa name. Data for woods as• signed binomials can be and have been computerized (Wheeler & Baas 1991, 1993) so

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 232 IAWA Journal, Vol. 16 (3), 1995 that comparison of specimens via computer can be done as easily for woods assigned binomials as for palaeotaxa. We have chosen to use binomials as well as to reference Page's groups (Page 1979). Some (3 of 6) of the woods have a combination of characters seen in commonly occur• ring fossil wood genera (Plataninium and Paraphyllanthoxylon). The others have a combination of characters that have not been previously described. We believe the dif• ferences between these woods and previously described genera warrant naming new genera.

DESCRIPTIONS

Metcalfeoxylon kirtlandense gen. et sp. nov. - Fig. 4-9

Generic diagnosis Wood diffuse-porous. Vessels exclusively solitary, mean tangential diameter of vessels 50-200 /JIll; fewer than 20 per sq.mm; perforations exclusively scalariform with fewer than 25 bars; intervessel pits opposite; vessel-ray parenchyma pits similar to intervessel pits. Rays 2-6 cells wide; heterocellular and frequently with more than 10 marginal rows of upright cells at at least one margin, pronounced difference in height between procumbent and upright cells; uniseriate rays numerous and composed exclusively of upright cells. Axial parenchyma apotracheal diffuse. Fibers non-septate with distinctly bordered pits, more common on radial walls. Derivation of generic name - For C. R. Metcalfe, co-author of classic works on the systematic anatomy of the dicotyledons. Species diagnosis Growth rings absent. Wood diffuse-porous. Vessels exclusively solitary; mean tangential diameter 121 (sd = 28) /JIll, range 78-201 /JIll; radial diameter 171 (sd = 38), range 112-257 /JIll; 4 per sq. mm, range 1-5 per sq. mm; perforation plates exclusively scalariform with 10-19 bars, spacing between bars 8-14 /JIll and bars 3-5 /JIll thick; vessel-vessel pits 7-8 /JIll, opposite; vessel-ray parenchyma pits similar to vessel pits. Axial parenchyma apotracheal diffuse. Multiseriate rays 2-4-seriate (32-48 /JIll wide); height of multi seriate portion 609 (sd = 176),327-1051 /JIll; markedly heterocelluiar, often asymmetric with one margi• nal row of upright cells at one end, and long (15-20 cells) uniseriate margins of up• right cells at the other; upright cells nearly twice the vertical height of the procumbent cells. Uniseriate rays present, numerous, composed exclusively of upright cells, and typically over 25 cells and 1500 mm high; 4-10, mostly 6-9, rays per mm. Imperforate elements with distinctly bordered pits (3-5 /JIll) on radial and tangential walls, but more common on radial walls. Vasicentric tracheids apparently present as there are cells with distinctly bordered pits in more than one row that encircle vessels. Etymology - For the Kirtland Formation from which the wood was collected.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 233

Fig. 4-9. Metcalfeoxylon kirtlandense gen. et sp. nov., YPM 30158. - 4: Diffuse-porous wood with exclusively solitary vessels, cross section; scale bar =500 jJl11. - 5: Detail showing diffuse apotracheal parenchyma and collapsed vessel element; scale bar = 100 jJl11. - 6: Scalariform perforation plate; scale bar =50 jJl11. - 7: Opposite intervessel pitting, and vessel-ray parenchyma pitting similar to intervessel pitting; scale bar = 50 jJl11. - 8: Multiseriate rays with uniseriate margins of upright cells, tangential section; scale bar = 50 jJl11. - 9: Heterocellular rays, with up• right cells distinct from procumbent ray parenchyma cells, radial section; scale bar = 100 jJl11.

Holotype: YPM 30158, cylindrical piece of stem, 7 cm diameter, 13 cm long, and three slides numbered 29,30,31. Estimated diameter of22 cm. Source: Locality W5, grey-brown, medium-grained sandstone layer in McClammer Section 16, Betonnie-Tsosie 2. E 1/2, Section 8, T22 N, R 9W. Lower Kirtland Formation. Late Campanian. [Note: dinosaur and mammal bones, leaf impressions and microvertebrates also occur in the wood-bearing unit, Univ. Kansas loco 16.]

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 234 IAWA Journal, Vol. 16 (3), 1995

This wood confonns to Page's (1979) Group IIIB, sect. 4 (Vessels present, scalariform perforations, fewer than 50 bars per perforation, apotracheal parenchyma, intervessel pits opposite-alternate). The general pattern of exclusively solitary vessels, scalarifonn perforation plates, opposite pits, fibers with distinctly bordered pits, rays that are not exclusively uniseriate or markedly large and frequently with over 10 marginal rows of upright/ square cells, and predominantly apotracheal parenchyma occurs in extant spe• cies of 4 of Cronquist's (1988) subclasses: Hamamelidae (Hamamelidales: Hamame• lidaceae; Daphniphyllales: Daphniphyllaceae), Dilleniidae (Dilleniales: Dilleniaceae; Theales: Theaceae; Violales: Stachyuraceae; Ebenales: Symplocaceae), Rosidae (Rosa• les: Escalloniaceae / Grossulariaceae;Myrtales:M yrtaceae;Cornales: Cornaceae; Santales: Olacaceae; Celastrales: Icacinaceae; Linales: Hourniriaceae), and Asteridae (Dipsacales: Caprifoliaceae ). The vessels in this wood had collapsed and often were compressed radially, reduc• ing their tangential diameters; in longitudinal sections, it was apparent that the scala• riform perforation plates kept the vessel from collapsing at that region. Some vessels were filled with fungal hyphae and dark-coloured contents. When quantitative data (mean tangential diameter between 100-200 /JIll, fewer than 20 vessels per sq. mm; fewer than 40 bars per perforation; rays not exclusively uniseriate or more than 1O-seriate) were added to the general description above, a search of the fossil wood database (Wheeler & Baas 1991, 1993) revealed that this particular combination of characters is nearly unique to the Cretaceous. With the ex• ception of the extant genus Hibbertia (Dilleniaceae) and one Eocene wood (Elaeoden• droxylon sp., Gottwald 1992), the features of this wood only occur in other Cretaceous woods (CASG 60118, 60133, 60134, 60135 from the Late Cretaceous of California, Page 1979, 1980; Myrtoxylon pichasquensis Torres & Rallo 1981 from the Late Creta• ceous of southern South America; an unnamed wood from the Late Cretaceous of Baja California, unpublished data from Jay Jones; wood called Potomac Grp II, Herendeen 1991a). Neither intervessel nor vessel-ray pitting was observed in the Eocene Elaeo• dendroxylon and its rays have only a few « 3) marginal rows of upright cells. Nishida et al. (1988) pointed out that the Cretaceous Myrtoxylon diagnosis was incomplete and no details about pits were provided. Only 110 of the 1356 woods in the fossil wood database are Cretaceous in age, so the near restriction of this combination of features to woods of Cretaceous age seems particularly noteworthy. A wood with similar combinations of features also occurs in the Upper Shale Member of the Aguja Fonnation (Campanian) of Big Bend National Park, Texas (Wheeler & Lehman, unpublished data). Rays are shorter and somewhat narrower in the Big Bend material, but other quantitative features (vessel diameter, density) are similar.

Chalkoxylon cretaceum gen. et sp. nov. - Fig. 10-17

Generic diagnosis Wood diffuse-porous. Vessels exclusively solitary; mean tangential diameter 50- 150 /JIll, fewer than 25 vessels per sq.mm; perforation plates exclusively scalariform

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 235

Fig. 10-17. Chalkoxylon cretaceum gen. et sp. nov., YPM 30148. - 10: Diffuse-porous wood with predominantly solitary vessels, cross section; scale bar = 100 /JlIl. - 11: Solitary vessels, diffuse apotracheal parenchyma, and thick-walled fibers; scale bar = 50 /JlIl. - 12: Multiseriate rays, 2-4 cells wide, without elongate uniseriate margins; scale bar = 50 /JlIl. - 13: Detail of ray cells. - 14: Scalariform perforation plate; scale bar = 50 /JlIl. - 15: Radial section showing imperforate elements with distinctly bordered pits, and vessel-parenchyma pits with reduced borders; scale bar =50 /JlIl. - 16: Upright ray cells, probable solitary crystals present. - 17: De• tail of vessel-ray parenchyma pits; scale bar = 50 /JlIl. with fewer than 25 bars per perforation plate; vessel-ray parenchyma pits with reduced borders and throughout the contact area. Axial parenchyma apotracheal diffuse. Multiseriate rays less than 6 cells wide and heterocellular, most frequently with fewer than 4 marginal rows. Fibers with distinctly bordered pits.

Derivation of generic name - For L. Chalk, co-author of classic work on systematic anatomy of the dicotyledons.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 236 IAWA Journal, Vol. 16 (3), 1995

Species diagnosis Growth rings absent. Wood diffuse-porous. Vessels exclusively solitary, mean tangential diameter 82 ~, range 56-100~; mean radial diameter 194~, range 148-256~; 8-12 per sq.mm; perforation plates exclusively scalariform, apparently less than 20 bars per perfora• tion, 9-25 ~ spacing between bars, bars occasionally branched; vessel-vessel pits not observed; vessel-ray parenchyma pits with reduced borders, somewhat horizon• tally elongate, throughout the ray contact; tyloses abundant. Axial parenchyma apotracheal diffuse. Multiseriate rays 2-4-seriate (40-60 ~ wide); mean height of multiseriate por• tion of the ray 357 ~ (sd = 115), range 140-616~, 12-38 cells; heterocellular with 1-4, rarely more, marginal rows of square and upright cells. Uniseriate rays composed exclusively of square and upright cells, 460 ~ (sd = 236) high, range of 140-1120 ~, 3-18 cells, 16 rays per mm. Imperforate elements with distinctly bordered pits on radial and tangential walls; thick-walled. Etymology - For the Cretaceous, the age of this fossil wood. Holotype: Sample YPM 30148, cylinder 6 cm high, 11 cm radius, 7 cm wide, and three slides: 11, 12, 13. Estimated diameter of 40 cm. Source: Locality W7, in medium to coarse-grained buff sandstone layer in McClammer Section 20, Betonnie-Tsosie 6, Rigby Section E. NE 1/4, Section 10, T 22N, R 9 W. Kirtland Shale, Early Maastrichtian.

The combination of exclusively solitary vessels, scalariform perforation plates, fibers with distinctly bordered pits, narrow rays that are not markedly heterocellular occurs in 4 of Cronquist's subclasses: Magnoliidae (Magnoliales: Cannellaceae), Hamamelidae (Myricales: Myricaceae), Dilleniidae (Ericales: Epacridaceae; Ebenales: Symplocaceae; Theales: Theaceae), and Rosidae (Rosales: Cunoniaceae, Eucryphiaceae; Euphorbiales: Buxaceae). As does Metcalfeoxylon kirtlandense, this wood conforms to Page's (1979) Group IIIB, sect. 4 (Vessels present, scalariform perforations, less than 50 bars per perforation, apotracheal parenchyma, intervessel pits opposite-alternate), but it is dis• tinct from M. kirtlandense. Metcalfeoxylon kirtlandense has multi seriate rays with long uniseriate margins of markedly upright cells, Chalkoxylon cretaceum has multi• seriate rays with uniseriate margins that generally have fewer than four cells and the distinction between upright and procumbent cells is not as pronounced. As was true for Metcalfeoxylon kirtlandense, when quantitative features are used (in this case, mean tangential diameter between 50-1 00 ~; fewer than 20 vessels per sq. mm; fewer than 40 bars per perforation; rays not exclusively uniseriate, but less than 10 cells wide; and more than 12 rays per mm), it appears that the particular com• bination of characters of this wood is also nearly unique to the Cretaceous. Again, with the exception of the Eocene Elaeodendroxylon sp. (Gottwald 1992), the particular combination of features of this wood only occurs in other Cretaceous woods (CASG 60128 and 60134 from the Late Cretaceous of California, Page 1979; Myrtoxylon

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 237 pichasquensis Torres & Rallo 1981 from the Late Cretaceous of southern South America; an unnamed wood from the Late Cretaceous of Baja California, unpublished data from Jay Jones). No extant woods have the combination offeatures diagnostic of Chalkoxylon.

? PLATANACEAE

Plataninium piercei sp. nov. - Fig. 18-24

Species diagnosis Growth rings absent. Wood diffuse-porous. Vessels predominantly solitary; mean tangential diameter 48 (sd = 9) /illl; 49-64-71 per sq. mm; perforations scalariform with widely spaced 6-9 bars; spacing between bars 17-22 /illl and a bar thickness of 5-6 /illl; vessel wall pit• ting opposite to sub-alternate. Axial parenchyma predominantly apotracheal, mostly diffuse, occasionally in ag• gregates of 2 or 3 cells. Multiseriate rays mostly 15 to 25 cells (180-280 /illl) wide; height very variable, range of 0.99-7.91 mm, mean 4.23 mm (sd = 2.11 mm); body of ray with procumbent cells, mostly lor 2 marginal rows of square and upright cells, at times with 4-6 mar• ginal rows. Uniseriate-biseriate rays very rare, 15 to 20 cells high composed exclu• sively of upright cells. Sometimes procumbent ray cells in multiseriate rays subdi• vided into short cells with crystals, 1 crystal per cell. Imperforate tracheary elements with distinctly bordered pits on their radial walls. Derivation ofspecies name - For Lee Pierce, who began study of these dicotyledonous woods from San Juan Basin. Holotype: Sample YPM 30146, hand sample 5 x 5 x 6 cm, and three slides: 23, 24, 25. Estimated source sample diameter of 14 cm. Source: Locality WI8, silty clay-shale layer in McClammer Section 20, Upper Shale Member of Kirtland Formation, Early Maastrichtian. Associated with palm wood.

The general pattern of exclusively solitary vessels, scalariform perforation plates, op• posite pits, fibers with distinctly bordered pits, and rays commonly over 10 cells wide occurs in four of Cronquist's subclasses: Magnoliidae (Order Piperales: Chloranthaceae), Hamamelidae (Order Hamamelidales: Eupteleaceae), Dilleniidae (Order Dilleniales: Dilleniaceae; Order Violales: Flacourtiaceae; Order Ericales: Ericaceae, Epacridaceae), and Rosidae (Order Celastrales: Icacinaceae). Cretaceous fossil woods of this type have been assigned to either the genus Icacinoxylon or Plataninium. Generally, Icacinoxylon is used for woods with large markedly heterocellular multi seriate rays accompanied by numerous uniseriate rays and predominantly solitary vessels; Plataninium is used for woods with large multiseriate rays that are not markedly heterocellular and uniseriate rays are absent to rare, some species have vessels in multiples. There is a need for a re-evaluation of all the woods assigned to Icacinoxylon and Plataninium, and the boundaries of the two genera are not clear. The Platanaceae is not among the 'matching' extant families listed above

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 238 IAWA Journal, Vol. 16 (3), 1995

Fig. 18-24. Plataniniumpierce(sp. nov., YPM 30146. -18: Diffuse-porous woods, predominant• ly solitary vessels, wide rays, cross section; scale bar = 100 ~. - 19: Solitary vessels and diffuse apotracheal parenchyma; scale bar = 50 ~. - 20: Imperforate elements with distinctly bordered pits; scale bar = 25 ~. - 21 : Tall and wide muItiseriate rays. Multiseriate rays lack extensive uniseriate margins, and uniseriate rays are few; scale bar = 100 ~. - 22: Radial view of short uniseriate rays with square and upright cells; scale bar = 100~. - 23: Opposite intervessel pits; scale bar = 25 ~ . - 24: Procumbent ray cells from muItiseriate ray, appearing subdivided and with solitary crystals; scale bar = 25 ~ .

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 239

Table 2. Comparison of quantitative features of Cretaceous and Paleocene North American Plataninium woods.

Species MTD VMM2 MVEL Bars MRH MRW

P. piercei 48 (9) 64 ? 6-9 4.2(2.1) 180-290 IIDl San Juan 49-71 15-25

P. platanoides 60-70 ? 502 4-22 up to2 mm up to 15 cells Page 1968

P. califomicum ? ? 968 < 30 2-7mm up to 15-20 cells Page 1968

P. haydenii 88 (17) 59 944 12-24 3.8mm 59-568 IIDl Wheeler 1991 45-79 3-22 cells

MTD =mean tangential diameter of vessels, (sd) standard deviation in parentheses, if known; VMM2 = vessels per sq.mm; MVEL = mean vessel element length; Bars = number of bars per scalariform perforation plate; MRH = multi seriate ray height in mm; MRW = multi seriate ray width, given first in IIDl (if data available) and then in number of cells.

because the fossil has predominantly solitary vessels, while extant Platanus have some vessel multiples. However, a search of the fossil wood database shows that some Cre• taceous and Paleogene Plataninium species have the same combination of features as does the fossil described above; some Cretaceous and Paleogene platanoid woods have predominantly solitary vessels. Because uniseriate rays are rare and multiseriate rays are not markedly heterocellular, the San Juan wood is assigned to Plataninium. In these ray features this wood differs from the other families listed above; woods of those families commonly have uniseriate rays and multi seriate rays are markedly heterocellular. Rays in the Chloranthaceae often are composed exclusively of upright and square cells. Differences between species of Plataninium are in quantitative features (vessel diameter and density, vessel element length, number of bars per scalariform perfora• tion plate, ray width and height, Table 2). The San Juan platanoid has larger rays and fewer bars per perforation (6-9) than any other Cretaceous or early Tertiary platanoid described to date. The differences between the San Juan platanoid and the other Creta• ceous and Paleocene platanoids is more than occurs today between different species of Platanus (Wheeler 1991), and so this wood is assigned a new specific epithet. The preservation of the San Juan sample is not the best, and very few perforation plates were observed, consequently it is possible that only the plates with a few coarse bars were detectable. In the platanoids, if bars are numerous they tend to be fine, and so might be less likely to be preserved. Previously, information about extant Platanus species indicated that Cretaceous and Paleogene platanoid woods differed from extant Platanus in ray width (wider in

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 240 IAWA Journal, Vol. 16 (3), 1995 the fossils) and perforation plate type (exclusively scalariform in the fossils, a mixture of simple and scalariform in the extant species) (Wheeler 1991). There is a tendency for woods of Platanus occidentalis, P orientalis, P racemosa, and P wrightii to have a few tangential vessel groups. Examination of a single sample of mature wood of P kerrii indicated that this species has a higher percentage of solitary vessels and wider rays (often more than 20-25 cells wide) than the other species (Wheeler 1995). Twig wood of P kerrii, native to Viet-Nam and Laos, had a higher percentage of scalariform perforation plates (85%) and longer vessel elements than twig wood of all other extant species (Baas 1969). Mature wood of Platanus kerrii has 53% scalariform perforation plates, a value somewhat higher than that reported for other extant species (Wheeler 1991). Mean vessel element length in the mature wood was 648 !lill (sd = 112, range 443-880), a value comparable to other extant species. Thus in perforation plate type P kerrii differs from the Cretaceous and early Tertiary platanoid woods which have exclusively scalariform perforation plates. However, in vessel grouping and ray width, Platanus kerrii is more similar to the fossil platanoids than are the other extant species. These differences may have some relationship to habitat, as the climatic regimes of Cretaceous and Paleogene platanoids may be more similar to present-day Viet-Nam than to temperate North America, Europe, and Asia. Fossil platanoid leaves and reproductive structures similar to extant Platanaceae are common in the Cretaceous and members of the platanoid complex are among the ear• liest known angiosperms (Crane et al. 1993). Most Cretaceous platanoid inflorescences resemble extant P kerrii more than other extant species. Consequently, it is not sur• prisingly that Cretaceous platanoid woods also resemble P kerrii more than other ex• tant species. Platanoid wood occurs in the Paleocene and Eocene of North America and almost all of these woods have very wide rays (Wheeler 1991) and so resemble P kerrii in that feature.

Paraphyllanthoxylon Bailey 1924 The genus Paraphyllanthoxylon was established for Cretaceous woods with the combination of characters: vessels solitary and in short radial multiples, simple perfo• ration plates, crowded alternate pitting, vessel-ray parenchyma pits with reduced bor• ders, abundant tyloses, axial parenchyma rare to absent, if present scanty paratracheal, septate fibers without distinctly bordered pits, heterocellular multi seriate rays. Species of Paraphyllanthoxylon are known from the Cretaceous through Tertiary (reviewed in Herendeen 1991b; Thayn & Tidwell 1984; Wheeler 1991). Quantitative features of the phyllanthoid woods from the San Juan Basin are sum• marized in Table 3 and Fig. 33 and compared to quantitative features of the type speci• men of Paraphyllanthoxylon arizonense Bailey. Two kinds of phyllanthoid woods are distinguished among the San Juan collections on the basis of differences in vessel diameter and ray size (Tables 3, 4), as well as differences in 1) fiber wall thickness, 2) ray cell wall thickness and 3) tyloses type. The first type (10 samples) is assigned to Paraphyllanthoxylon arizonense Bailey; this type occurs in Nacimiento Formation, OjoAlamo Sandstone, and Kirtland Shale, the second type (2 samples) is designated a new species (diagnosed below); this type was only recovered from the Kirtland Shale.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 241

Table 3, Comparison of quantitative features of Paraphyllanthoxylon woods in the San Juan Basin, New Mexico, and P. arizonense Bailey (type),

Sample MTD VMM2 Vul MRW MRH SGg

P. arizonense Type 175 (32) 3-6-9 29.2 2-6 221-779-1893 0.56 Bailey 1924 Paleocene

Nacimiento P. arizonense YPM 30152 133 (23) 14-20-27 6.7 3-5 361-729-1502 0.53 YPM 30150 122 (16) 11-18-28 6.8 3-5 350-683-1616 0.58 YPM 30149 168 (26) 2-5-8 33.6 2-5 292-611-1469 0.47

OjoAlamo P. arizonense YPM 30153 172 (39) 5-6-8 28.7 3-5 417-747-1446 0.59 YPM 30154 224 (53) 2-3-6 74 3-6 350-825-1796 0.53 YPM 30144 173 (27) 2-5-7 34.6 3-5 341-870-2201 0.55 YPM 30145 182 (31) 3-5-10 36.4 2-4 339-883-1751 0.48 Cretaceous

Kirtland P. arizonense YPM 30155 184 (38) 3-6-10 30.7 2-3 169-809-1692 0.58 YPM 30156 180 (37) 5-14-28 12.9 3-4 273-679-1808 0.42 YPM 30151 143 (26) 3-5-6 28.6 2-3 248-680-1495 0.57 P. anasazi YPM 30159 86 (16) 12-22-31 3.9 2-3 169-294-526 0.59 YPM 30147 84 (15) 18-23-27 3.7 2 56-234-526 0.65

MTD = mean tangential diameter of vessels, Mm (standard deviation in parentheses); VMM2 = vessels per sq.mm, minimum-mean-maximum; Vul = vulnerability index (mean tangential diameter / mean number of vessels per sq. mm); MRW =multi seriate ray width in cell number; MRH =multiseriate ray height, ~, minimum-mean-maximum; SGg = estimated 'green' specific gravity.

Table 4. Comparison of Paraphyllanthoxylon arizonense and P. anasazi.

Fiber walls Ray cells Tyloses type

P. arizonense Thin-medium Thin-walled Abundant and bubble-like P. anasazi Thick-walled Thick-walled Segmenting vessels

The salient characteristics of Paraphyllanthoxylon arizonense (Fig. 25-33) are sum- marized below: Wood diffuse-porous. Vessels solitary and in radial multiples of 2-4; perforations exclusively simple, intervessel pits crowded alternate, hexagonal in outline with in-

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 242 IAWA Journal, Vol. 16 (3), 1995

Fig. 25-30. Paraphyllanthoxylon arizonense Bailey. - 25: Diffuse-porous wood with vessels solitary and in radial multiples of2 and 3, YPM 30153; scale bar =500 j.UIl. - 26: Detail showing variation in vessel density, possibly indicating some seasonality in moisture availability, YPM 30155; scale bar = 500 j.UIl. - 27: Vessels in radial multiples and filled with tyloses, YPM 30154; scale bar = 200 j.UIl. - 28: Crowded alternate intervessel pitting, YPM 30149; scale bar = 25 j.UIl. - 29: Vessel-ray parenchyma pits with reduced borders, but not horizontally enlarged, YPM 30149; scale bar = 25 j.UIl. - 30: Radial section showing rays with procumbent cells and vessel filled with tyloses, YPM 30151; scale bar = 100 j.UIl.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 243

Fig. 31 & 32. Paraphyllanthoxylon arizonense Bailey. - 31: Multiseriate rays commonly more than 4 cells wide, ray cells with thin walls, YPM 30154; scale bar = 100 /JIIl. - 32: Ray cells with thin walls, septate fibers, YPM 30151; scale bar = 50 /JIIl. eluded apertures, large, 12-14 J.Ull; bubble-like tyloses abundant and filling all ves• sels. Vessel-ray parenchyma pits with reduced borders, but not markedly larger than vessel pits. Axial parenchyma absent or extremely rare, with a few cells adjacent to vessels, scanty paratracheal. Multiseriate rays 2-5 (6) cells wide; heterocellular, with a few marginal rows of square and upright cell; ray cells very thin-walled and isodiametric in tangential sec• tion; uniseriate rays rare. Fibers all septate, no pits observed.

Material of ParaphyUanthoxylon arizonense Nacimiento Formation, Paleocene Samples YPM 30151 (slides 38, 39, 40) estimated diameter of 80 cm, YPM 30150 (slides 2, 3,4) estimated source diameter of 40(-60) cm from Locality W4, McClammer Section 14. Sample YPM 30149 (slides 14, 15, 16) from LocalityW8, McClammer Section 21, estimated source diameter of 40 cm. Ojo Alamo Formation, Paleocene Samples YPM 30153 (slides 44, 45, 46) estimated diameter of 80 cm, YPM 30154 (slides 47, 48,49) estimated diameter of 60 -70 cm from Locality W16, McClammer Section 5. Samples YPM 30144 (slides 20, 21, 22) estimated source diameter of 40 cm, YPM 30145 (slides 5, 6, 7) estimated source diameter of 36-40 cm from Locality WI, McClammer Section 8.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 244 IAWA Journal, Vol. 16 (3), 1995

30 - -

E E a 20 ~ g-'"::> b i il c. I '" 11 - '" 10 ~ d j h Y 1 I ~ c f 9 e O~------'I ------'------'------'I ------~ 100 200 300 Mean tangential diameter (J.Un)

Paleocene o Nacimiento Fm a YPM 30152 b YPM 30150 c YPM 30149

• Ojo Alamo SS d YPM 30153 e YPM 30154 f YPM 30144 g YPM 30145

Cretaceous o Upper Kirtland Fm h YPM 30155 YPM 30156 YPM 30151

Fig. 33. Mean tangential diameters and vessels per sq. mm. for samples of Paraphyllanthoxylon arizonense. Line for tangential diameters = 2 x standard deviation; line for vessels per sq. mm = range.

Kirtland Shale, Early Maastrichtian Samples YPM 30155 (slides 26, 27, 28) estimated diameter of 20 em, YPM 30156 (slides 17, 18, 19) from Locality WI 0, McClammer Section 20, estimated diameter of 25 cm. Sample YPM 30151 (slides 35, 36, 37) from Locality W14, McClammer Section 5, estimated source diameter of 30 cm.

Figure 33 shows that the Paraphyllanthoxylon arizonense samples from the Creta• ceous and Paleocene overlap in their quantitative features. Two of the Paraphyllanth• oxylon arizonense samples from locality W4 in the Nacimiento have smaller, more numerous vessels than the others, but this difference is not statistically significant. These two woods are similar to samples from other localities in quantitative ray fea• tures, and all qualitative features, so assigning them to a different taxon is not justified.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 245

Given the abundance of fossil woods of the phyllanthoid type it is possible that in• creased sampling would reveal consistent and statistically meaningful differences in vessel characteristics of the phyllanthoid types from different sites that might indicate environmental differences between those localities. Narrower, more numerous vessels may be considered indicative of decreased water availability (Baas & Schweingruber 1987; Carlquist 1988). Vulnerability indices of all the woods are high relative to ex• tant woods, but typical of other known Cretaceous and Paleocene woods that are large trees (Wheeler et al. 1987; Wheeler 1991). Two woods from locality W4 in the Naci• miento have lower vulnerability indices than woods from other sites. Although no samples had distinct growth rings, the samples from this site had some variation in vessel density and diameter as did some samples from the upper Kirtland Shale, such variation might indicate seasonal variation in water availability. Of the dicotyledonous trees known from the Cretaceous and Paleocene, the phyl• lanthoid type is the most common. None of the phyllanthoid wood samples from the Cretaceous came from in situ stumps or logs, but all were mature wood samples. Some Paraphyllanthoxylon arizonense samples from the Paleocene came from logs in ex• cess of one meter in diameter. It is striking that the vessel diameters and densities, ray sizes, and estimated specific gravity of the Cretaceous (Early Maastrichtian) and Early Paleocene P. arizonense from San Juan are so similar (Fig. 33, Table 3), given the dif• ferences in age and their occurrence in different depositional settings. Moreover, they are similar to the type specimen of P. arizonense. The specific gravities of the woods indicate a tree with medium density wood, not low density as is generalized as charac• teristic of pioneer species.

Paraphyllanthoxylon anasazi, sp. nov. - Fig. 34-38

Species diagnosis Wood diffuse-porous. Vessels solitary and in radial mUltiples of 2 or rarely 3; mean tangential diameter of 84 /lID (sd = 16); 12-37, mostly 20-25 per sq.mm; perforations exclusively simple; vessel element lengths 444-988 /lID; intervessel pitting crowded alternate, hexagonal in outline with included apertures, 8-11 /lID horizontal width; en• larged vessel-ray parenchyma pits with reduced borders, occurring throughout the ray contact area; tyloses present and appearing to segment the vessel. Parenchyma extremely rare, scanty paratracheal. Rays 2- or 3-seriate; mean multi seriate ray height of 234-294 /lID; heterocellular. Uniseriate rays extremely rare. Fibers septate and pits without distinct borders; thick-walled.

Etymology - For the Anazasi, pre-historic Indians of the southwest United States.

Holotype: Sample YPM 30159, slides numbered 41,42,43, estimated source diameter of 20 cm. Paratype: Sample YPM 30147, slides numbered 32, 33, 34, estimated source diameter of 24 (-30) cm. Source: Locality W6, buff brown clay-shale layer in McClammer Section 20, Betonnie-Tsosie 6, Rigby Section E. NE 114, Sectiom 10, T 22N, R 9 W. Kirtland Shale, Early Maastrichtian.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 246 IAWA Journal, Vol. 16 (3), 1995

Fig. 34-38. Paraphyllanthoxylon anasazi sp. nov., YPM 30159. - 34: Diffuse-porous wood with vessels solitary and radial multiples of 2 and 3; scale bar =250 J.UIl. - 35: Radial mUltiples of vessels, and thick-walled fibers; scale bar =50 J.UIl. - 36: Crowded alternate intervessel pitting; scale bar =25 J.UIl. - 37: Vessel-ray parenchyma pits with reduced borders and occurring through• out the ray; scale bar = 25 J.UIl. - 38: Multiseriate rays that are 2, or occasionally 3 cells wide; scale bar = 50 J.UIl.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 247

Differences between the Paraphyllanthoxylon species described to date are primarily in quantitative features, mainly vessel diameter and density (Thayn & Tidwell 1984; Herendeen 1991a; Wheeler 1991). The wood described above has narrower vessels than any other Paraphyllanthoxylon (see Herendeen 1991a; Thayn & Tidwell 1984). It also has smaller rays and thicker fiber walls, characteristics that we believe warrant designating it a new species. The vulnerability indices of the two samples of P. anasazi are lower than those of the P. arizonense from the San Juan Basin, and also those that occur in the Paleocene of Big Bend National Park, Texas (Wheeler 1991). The esti• mated diameters of P. anasazi were smaller (20-24 cm) than for the P. arizonense (30-80 cm), and the specific gravities slightly higher (Table 3).

Carlquistoxylon nacimientense, gen. et sp. nov. - Fig. 39-43

Generic diagnosis Wood diffuse-porous. Vessels solitary and in radial multiples of 2 or 3; mean tan• gential diameter between 50 and 150 j.UIl; fewer than 40 per sq.mm; perforations ex• clusively simple, mean vessel element length between 500 and 800 j.UIl; intervessel pitting crowded alternate; vessel-parenchyma pits with reduced borders; parenchyma rare, if present, scanty paratracheal; fibers non-septate and pits not obvious; rays nar• row less than 4 cells wide; uniseriate rays rare. Derivation of the generic name - For Sherwin Carlquist on the occasion of his retire• ment from Rancho Santa Ana Botanic Garden.

Specific diagnosis Wood diffuse-porous. Vessels solitary and in short radial multiples, mean tangential diameter 106 j.UIl (sd = 21); 4-26 per sq.mm.; perforations exclusively simple; mean vessel element length 616 j.UIl (sd = 124), range 396-832 j.UIl; intervessel pitting crowded alternate, hexagonal in outline with included apertures, 11-14 j.UIl horizontal width; enlarged vessel-ray parenchyma pits with reduced borders to appearing simple, ori• ented vertically, horizontally, and diagonally in upright cells, in procumbent cells pits circular with reduced borders and of similar size to the intervessel pits; tyloses not observed. Parenchyma extremely rare, scanty paratracheal. Vessel-axial parenchyma pits with reduced borders to appearing simple, oriented vertically, horizontally, and diagonally. Rays 2- or 3-seriate; mean multi seriate ray height of 507 j.UIl (sd = 118), range 240- 764 j.UIl; heterocellular. Uniseriate rays extremely rare. Fibers non-septate and pits without distinct borders; thin-walled; with noticeable radial alignment. Derivation of species name - For the Nacimiento Formation.

Holotype: Sample YPM 30157, three slides labelled 8, 9, 10, estimated source diameter of 10 cm. Source: Locality W2, dark grey mudstone layer in McClammer Section 11, Kimbeto West Fork 4. SE 1/4, NE 114, Section 12, T 23 N, R 10 W. Nacimiento Formation, Early Paleocene.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 248 IAWA Journal, Vol. 16 (3), 1995

Fig. 39-43. Carlquistoxylon nacimientense gen. et sp. nov., YPM 30157. - 39: Diffuse-porous wood with vessels solitary and in radial multiples of 2 and 3; scale bar = 100 1lITI. - 40: Radial multiples of vessels, thin-walled fibers; scale bar = 50 1lITI. - 41: Multiseriate rays 2 or 3 cells wide, vessel elements with simple perforation plates and crowded alternate intervascular pits; scale bar = 50 1lITI. - 42: Vessel-ray parenchyma pits with reduced borders; scale bar = 25 1lITI. - 43: Detail of intervessel pits, and vessel-axial parenchyma pits with reduced borders and of variable shape; scale bar = 25 1lITI.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 249

Although the general pattern of Carlquistoxylon is similar to Paraphyllanthoxylon, the non-septate fibers of this wood preclude it from being assigned to that genus, The combination of vessels solitary and in short radial mUltiples, exclusively simple perfo• ration plates, alternate intervessel pitting, vessel-ray parenchyma pitting with much reduced borders, rays that are not exclusively uniseriate or markedly wide (> 1O-seriate) or markedly heterocellular, non-septate fibers without obvious bordered pits, and pre• dominantly scanty paratracheal parenchyma occurs in extant species of only 2 of Cronquist's subclasses: Magnoliidae (Order Laurales: family Lauraceae) and Rosidae (Order Sapindales: Anacardiaceae and Burseraceae; Order Apiales: Araliaceae; Myrtales: Melastomataceae; Order Euphorbiales: Euphorbiaceae), This wood bears resemblance to some woods assigned to the fossil wood genera Ulminium and Euphor• bioxylon. Ulminium includes woods that resemble Lauraceae, these woods have oil cells, this San Juan wood does not. Euphorbioxylon Felix is a genus that needs revi• sion. It includes species with exclusively simple perforation plates, and species with scalariform perforation plates; some species have exclusively uniseriate rays, others have rays in excess of 10 cells wide and over 1 mm high; some species have scanty paratracheal parenchyma and marginal parenchyma, others banded. Although the San Juan wood described above shares some characters (vessels in radial multiples, simple perforation plates, alternate pits, scanty paratracheal parenchyma, non-septate fibres, and 1-3-seriate heterocellular rays) with Euphorbioxylon saggarense Mahabale & Deshpande (1963), given the use of this genus to accommodate woods with very dif• ferent characteristics, we are hesitant to assign the San Juan wood to this genus. We suggest the genus Carlquistoxylon is useful for woods with the generalized structure described above. Mahabale and Deshpande (1963), although they assigned the Paleo• gene Indian wood they studied to Euphorbioxylon, were circumspect in their discus• sions about its relationships to the Euphorbiaceae.

DISCUSSION

The Cretaceous woods described herein, with the exception of Plataninium piercei, have a generalized structure, and are not referable to a single genus, family, order, or subclass. The character combinations (excluding quantitative features) of the San Juan woods (Table 5) are widely distributed among extant taxa, and are likely plesiomorphic, so that the fossil woods cannot be assigned with confidence to an extant taxon at or below the subclass. The San Juan Cretaceous woods resemble those known from other localities in the Northern Hemisphere (Wheeler & Baas 1991, 1993). Phyllanthoid woods are common, the platanoidiicacinoid type is present, and there are woods with combi• nations of features considered primitive in the Bai1eyan sense (two woods with soli• tary vessels, scalariform perforation plates, heterocellular rays, and apotracheal paren• chyma). Woods of Plataninium, /cacinoxylon, and Paraphyllanthoxylon account for 54 of the 110 reports of Cretaceous woods (information from a database for fossil dicotyledonous woods, Wheeler & Baas 1991, 1993). No known North American Cre• taceous woods have elaborate vessel distribution patterns or abundant paratracheal parenchyma.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 250 IAWA Journal, Vol. 16 (3), 1995

Table 5. Comparison of dicotyledonous woods from the San Juan Basin, New Mexico.

Vessel Perforation Intervessel Fibers Parenchyma groups plates pits

Paleocene

Nacimiento Formation Paraphyllanthoxylon arizonense sol & rm si alt sep, nb sc. p Carlquistoxylon nacimientense sol & rm si alt n-sep, nb sc. p

OjoAlamo P. arizonense sol & rm si alt sep, nb sc. p

Cretaceous

Kirtland P. arizonense sol & rm si alt sep,nb sc. p P. anasazi sol & rm si alt sep,nb sc. p Plataninium piercei mostly sol scal ? n-sep, b dif Chalkoxylon cretaceum sol scal opp n-sep, b dif Metcalfeoxylon kirtlandense sol scal opp n-sep, b dif

Vessel groups: sol = solitary, rm =radial multiples - Perforation plates: si = simple, scal = scalariform- Intervessel pits: alt =alternate, opp =opposite - Fibers: sep =septate, n-sep =non-septate, b =with distinct borders, nb = without distinct borders - Parenchyma: sc. p = scanty paratracheal, dif = diffuse.

It is not surprising, given the abundance of leaf and pollen types unique to the Creta• ceous, that the particular combination offeatures (qualitative and quantitative) of two of the San Juan woods (Metcalfeoxylon kirtlandense and Chalkoxylon cretaceum) ap• pear to be nearly unique to the Cretaceous. There are some combinations of features that are extremely rare in extant woods, but which occur relatively commonly in Cre• taceous woods. For example, one Malvalean tree common in the Late Cretaceous of southern Illinois had large diameter vessels (> 200 /JIll mean tangential diameter) and scalariform perforation plates (Wheeler et al. 1987). Of 2070 genera in a database for extant woods (Wheeler et al. 1986; LaPasha & Wheeler 1987), only some species of Dillenia have that combination of features. A survey of the fossil record for dicotyledonous woods found that there was a marked difference in the incidence of selected vessel features (e.g., scalariform perforations) between the Cretaceous and the Tertiary (Wheeler & Baas 1991, 1993). It appears that some strategies of the hy• draulic system were more common in the Cretaceous than in Tertiary, and may be unique to the Cretaceous. Very few Paleocene dicotyledonous woods are known worldwide; only two types have previously been reported for North America (one phyllanthoid and one platanoid type, Wheeler 1991). Carlquistoxylon nacimientense is a variation on the phyllanthoid pattern (radial multiples, simple perforation plates, alternate pitting, scanty paratracheal parenchyma) but lacks septate fibres. The report of this one wood increases by 50% the types of Paleocene dicotyledonous woods described for North America.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 251

Growth rings are either absent or weakly developed in the San Juan Basin dicotyledonous woods. This suggests near-continuous growth throughout the year and implies that the climate was not strongly seasonal. The vulnerability indices of Para• phyllanthoxylon arizonense are high and suggest tropical conditions. These inferences are consistent with those offered by Wolfe and Upchurch (1987) for the western in• terior of North America. Some specimens of Paraphyllanthoxylon arizonense Bailey in the Ojo Alamo Sandstone exceed 30 meters in length and 2 meters in diameter at breast height. This suggests that at least some species of angiosperms were prominent, if not dominant, members of the canopy. Data relating maximum diameter to height in extant trees (Rich et al. 1986) indicate that the height of a 2 meter diameter tree would be over 70 m - by any standards a large tree. The minimum estimated diameters of the Cretaceous San Juan dicotyledonous woods are 14,20,22,25,30, and 40 cm, which, while not as large as the Paleocene San Juan woods, contradict Dodson's (1993: 230) generalization about the size of Cretaceous angiosperms and the ease with which horned dinosaurs could have felled them for food: "Late Cretaceous angiosperm tree trunks were mainly less than 10 cm in diameter and thus were of a size that they could have easily been knocked over by ceratopsids." Species of Paraphyllanthoxylon account for 12 of the 17 angiosperm specimens, and woods of the Paraphyllanthoxylon arizonense type are the most common. This may mean that either 1) a single tree species dominated the Late Cretaceous and early Tertiary floras in the San Juan Basin, 2) this type of wood represented a grade of evolution associated with more than one dicotyledonous clade, or 3) the other dicoty• ledons known to occur in the San Juan Basin were trees and abundant, but their wood was not preserved (because of the susceptibility of the woods to decay or because the trees grew in locations not likely to be preserved). It is remarkable that the Para• phyllanthoxylon arizonense samples from the Early Maastrichtian and Early Paleocene are so similar, and that they are similar to the type specimen from the Middle Creta• ceous of Arizona. Phyllanthoid woods are common at other Cretaceous localities, some• times being the only type present (Alabama, Cahoon 1972) or among the most com• mon (Maryland, Herendeen 1991a, 1991b; southern Illinois, Wheeler et al. 1987). Only woods that appeared to be well-preserved were collected for this study, and stumps and trunks were not systematically sampled. Consequently, conclusions based on this sample are tentative, and would need to be verified by a more detailed survey and inventory. These preliminary data and field observations suggest that in the San Juan Basin, dicotyledonous trees were more diverse in Cretaceous than in the Paleocene.

ACKNOWLEDGMENTS

We thank Leo Hickey and Linda Klise (Peabody Museum, Yale University) for loan of the specimens of San Juan Basin woods, Dmitri Gromyko (Komarov Institute, St. Petersburg) for providing the sample of Platanus kerrii wood, James Ferrigno (Smithsonian Institution) for making available the type specimen of Paraphyllanthoxylon arizonense, and Tom Lehman (Texas Tech), Leo Hickey (Yale), and Gary Upchurch (Southwest Texas State) for their useful suggestions on a draft of this manuscript. This work was supported in part by NSF Grant BSR 9010068 (EW).

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 252 IAWA Journal, Vol. 16 (3), 1995

REFERENCES

Baas, P. 1969. Comparative anatomy of Platanus kerrii Gagnep. Bot. J. Linn. Soc. 62: 413-421. Baas, P. 1986. Ecological patterns in xylem anatomy. In: T. J. Givnish (ed.), On the economy of plant form and function: 327-349. Cambridge University Press, Cambridge. Baas, P. & E H. Schweingruber. 1987. Ecological trends in the wood anatomy of trees, shrubs and climbers from Europe. IAWA Bull. n. s. 8: 245-274. Bailey,I.W. 1924. The problem of identifying the wood of Cretaceous and later dicotyledons: Paraphyllanthoxylon arizonense. Ann. Bot. 38: 439-451. Baltz, E.H., S.R. Ash & R.Y. Anderson. 1966. History of nomenclature and stratigraphy of rock adjacent to the Cretaceous-Tertiary boundary western San Juan Basin, New Mexico. U. S. Geological Survey Prof. Paper 524-D, 23 pp. Brown, B. 1910. The Cretaceous Ojo Alamo beds of New Mexico with description of the new dinosaur genus Kritosaurus. Bull. Amer. Mus. Nat. History 28: 267-274. Cahoon, E.1. 1972. Paraphyllanthoxylon alabamense - a new species of fossil dicotyledonous wood. Amer 1. Bot. 59: 5-11. Carlquist, S. 1988. Comparative wood anatomy. Springer-Verlag, Berlin, New York. Chapman, 1.L. & 1.L. Smellie. 1992. Cretaceous fossil wood and palynomorphs from Williams Point, Livingston Island, Antarctic Peninsula. Rev. Palaeobot. Palynol. 74: 163-192. Crane, P.R., K.R. Pedersen, E.M. Friis &A.N. Drinnan. 1993. Early Cretaceous (Early to Middle Albian) platanoid inflorescences associated with Sapindopsis leaves from the Potomac Group of eastern North America. Systematic Botany 18: 328-344. Cronquist, A. 1988. The evolution and classification of flowering plants. 2nd ed. The New York Botanical Garden, Bronx, NY. Dodson, P. 1993. Comparative craniology of the Certaopsia. Amer. 1. Sci. 293-A: 200-234. Fassett, J.E. & J.S. Hinds. 1971. Geology and fuel resources of the Fruitland Formation and Kirtland Shale of the San Juan Basin, New Mexico. U. S. Geol. Surv. Prof. Paper 676,76 pp. Gottwald, H. 1992. HOlzer aus marinen Sanden des oberne Eozan von Helmstedt (Niedersachsen). PalaeontographicaAbt. B 225: 27-103. Herendeen, P.S. 1991a. Charcoalified angiosperm wood from the Cretaceous of eastern North America and Europe. Rev. Palaeobot. Palynol. 69: 225-239. Herendeen, P.S. 1991b. Lauraceous wood from the mid-Cretaceous Potomac Group of eastern North America: Paraphyllanthoxylon marylandense sp. nov. Rev. Palaeobot. Palynol. 69: 277-290. IAWA Committee. 1989. IAWA list of microscopic features for hardwood identification. lAWA Bull. n.s. \0: 219-332. LaPasha, C.A. & E. A. Wheeler. 1987. A microcomputer based system for computer-aided wood identification. lAWA Bull. n. s. 8: 347-354. Lehman, T.M. 1985. Depositional environments of the Naashoibito Member of the Kirtland Shale, Upper Cretaceous, San Juan Basin, New Mexico. In: Contributions to Late Cretaceous Paleontology and Stratigraphy of New Mexico. Part I (compiled by D.L. Wolberg): 55-79. New Mexico Bureau of Mines & Mineral Resources Circular 195, Socorro, New Mexico. Leo, R.E & E.S. Barghoorn. 1976. Silicification of wood. Bot. Mus. Leafl. Harvard Univ. 25: 1-47. Lucas, S. 1981. Dinosaur communities of the San Juan Basin: A case for lateral variations in the composition of Late Cretaceous dinosaur communities. In: S. Lucas, K. Rigby Jr. & B. Kues (eds.), Advances in San Juan Basin Paleontology: 337-393. University of New Mexico Press, Albuquerque. Lucas, S., K. Rigby Jr. & B. Kues (eds.). 1981. Advances in San Juan Basin Paleontology. Uni• versity of New Mexico Press, Albuquerque. 393 pp.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access Wheeler, McClammer & LaPasha - Fossil woods of Cretaceous and Paleocene 253

Mahabale, T. S. & S.R. Deshpande. 1963. Euphorbioxylon sagarense spec. nov., a fossil wood from Sagar (M.P.) belonging to the family Euphorbiaceae. J. Ind. Bot. Soc. (Maheshwari Comm. Vol.) 42A: 102-109. Metcalfe, C. R. & L. Chalk. 1950. Anatomy of the dicotyledons. Vol. I, II. Clarendon Press, Oxford. Nishida, M., H. Nishida & M. Rancusi. 1988. Notes on the petrified plants from Chile (1). J. Jap. Bot. 63: 39-48. Page, V.M. 1968. Angiosperm wood from the Upper Cretaceous of central California: Part H. Amer. J. Bot. 55: 168-172. Page, v'M. 1979. Dicotyledonous wood from the Upper Cretaceous of central California. I. J. Arnold Arbor. 60: 323-349. Page, v'M. 1980. Dicotyledonous wood from the Upper Cretaceous of central California. II. J.ArnoldArbor. 61: 723-748. Page, v'M. 1981. Dicotyledonous wood from the Upper Cretaceous of central California. III. Conclusions. J. Arnold Arbor. 62: 437-455. Pearson, R.G. & E.A. Wheeler. 1981. Computer identification of hardwood species. IAWA Bull. n. s. 2: 37-40. Powell, lS. 1973. Paleontology and sedimentation models of the Kimbeto Member of the Ojo Alamo Sandstone. In: lE. Fassett (ed.), Cretaceous and Tertiary Rocks of the Southern Colorado Plateau: 111-122. Four Comers Geological Society Memoir. Reinick, H.E. & LB. Singh. 1980. Depositional sedimentary environments with reference to terrigenous clastics, 2d rev. and updated edition. Springer-Verlag, New York. Rich, P.M., K. Helenurm, D. Keams, S.R. Morse & M.W. Palmer. 1986. Height and stem dia• meter relationships for dicotyledonous trees and arborescent palms of Costa Rican tropical wet forest. Bull. Torrey Bot. Club 113: 241-246. Robison, C.R., A. Hunt & D.L. Wolberg. 1982. New Late Cretaceous leaflocality from lower Kirtland Shale member, Bisti area, San Juan Basin, New Mexico. New Mexico Geology (August): 42-45,48. Sikkink, P.G.L. 1987. Lithofacies relationships and depositional environments of the Tertiary Ojo Alamo Sandstone and related strata, San Juan Basin, New Mexico and Colorado. In: J. E. Fassett & J.K. Rigby Jr (eds.), The Cretaceous-Tertiary Boundary in the San Juan and Raton Basins, New Mexico and Colorado: 81-104 . Geol. Soc. Amer. Special Paper 209. Boulder, CO. Sinclair, W.J. & W. Granger. 1914. Paleocene deposits of the San Juan Basin, New Mexico. Bull. Amer. Museum Nat. Hist. 33: 297-316. Stamm, A.J. 1964. Wood and cellulose science. Ronald Press Co., NY. Thayn, G.F. & W. D. Tidwell. 1984. A review of the genus Paraphyllanthoxylon. Rev. Palaeo• bot. Palynol. 43: 321-335. Tidwell, W. D., S. R. Ash & L. R. Parker. 1981. Cretaceous and Tertiary floras of the San Juan Basin. In: S. Lucas, K. Rigby Jr & B. Kues (eds.), Advances in San Juan Basin Paleontol• ogy: 307-332. University of New Mexico Press, Albuquerque. Torres, T. & M. Rallo. 1981. Anatomfa de troncos f6siles del Cretacio Superior de Pichasca, Norte de Chile. Anais H Congo Lat. de Paleontologfa 1: 385-398. Wheeler, E.A. 1991. Paleocene dicotyledonous trees from Big Bend National Park, Texas: Variability in wood types common in the Late Cretaceous and Early Tertiary, and ecological inferences. Amer 1 Bot. 78: 658-671. Wheeler, E.A. 1995. Wood of Platanus kerrii Gagnep. IAWA 1 16: 127-132. Wheeler, E.A. & P. Baas. 1991. A survey of the fossil record for dicotyledonous wood and its significance for evolutionary and ecological wood anatomy. IAWA Bull. n. S. 12: 271-332. Wheeler, E. A. & P. Baas. 1993. The potentials and limitations of dicotyledonous wood anatomy for climatic reconstructions. Paleobiology 19: 486-497_

Downloaded from Brill.com10/09/2021 04:18:38AM via free access 254 lAWA Journal, Vol. 16 (3), 1995

Wheeler, E.F., M. Lee & L.C. Matten. 1987. Dicotyledonous woods from the Upper Cretaceous of Southern Illinois. Bot. J. Linn. Soc. 95: 77-100. Wheeler, E.A., RG. Pearson, C.A. LaPasha, T. Zack & W. Hatley. 1986. Computer-aided wood identification. Reference manual. North Carolina Agricultural Research Service Bulletin 474. 160 pp. Wolberg, D. L., J. P. Hall, D. Bellis, W. X. Chavez, O. Anderson, R Moro & A. Gil. 1988. Regional historic, stratigraphic, and paleontologic framework of the Late Cretaceous (Campanian• Maastrichtian) Fossil Forest locality near Split Lip Flats, San Juan, Basin, San Juan County, New Mexico. In: Contributions to Late Cretaceous paleontology and stratigraphy of New Mexico. Part III (Compiled by Donald L. Wolberg): 7-21. New Mexico Bureau of Mines & Mineral Resources Bull. 122. Socorro, New Mexico. Wolfe, J.A. & G.R Upchurch Jr. 1987. North American nonmarine climates and vegetation during the Late Cretaceous. Palaeogeog., Palaeoclimatol., Palaeoecol. 61: 33-77.

Downloaded from Brill.com10/09/2021 04:18:38AM via free access