IAWA Journal, Vol. 30 (3), 2009: 293–318
NEW LATE CRETACEOUS AND PALEOCENE DICOT WOODS OF
BIG BEND NATIONAL PARK, TEXAS
AND REVIEW OF CRETACOUS WOOD CHARACTERISTICS
E.A. Wheeler1 and T. M. Lehman2
SUMMARY
Three new wood types from the Late Cretaceous and one from the Paleocene of
Big Bend National Park, Texas, U.S.A. add to our knowledge of North American Late Cretaceous and Paleocene plants. Sabinoxylon wicki sp. nov. provides
further evidence of similarities in late Campanian-early Maastrichtian vegetation
of Texas, New Mexico, and northern Mexico. This species is characterized by
mostly solitary vessels, scalariform perforation plates, vessel-ray parenchyma
pits similar to intervessel pits, vasicentric tracheids, and two size classes of rays.
Storage tissue accounts for close to 50% of its wood volume. Another of the
new Cretaceous wood types, referred to as Big Bend Ericalean Wood Type I, has
more than 40 % ray parenchyma. Both Big Bend Ericalean Wood Type I
and Sabinoxylon have a combination of characters that occurs in the order
Ericales (sensu APGII). The third new Cretaceous wood type is from a small
axis (less than 3 cm diameter), and has a combination of features that is the
most common pattern in extant eudicots (vessels solitary and in radial multiples
randomly arranged, simple perforation plates and alternate intervessel pits, and
heterocellular rays). The Paleocene wood (cf. Cunonioxylon sensu Gottwald)
differs from all other North American Paleocene woods and has characteristics
found in the predominantly Southern Hemisphere family Cunoniaceae. The
characteristics of these new Big Bend woods contribute to a database for fossil
angiosperm woods, and allow for comparison of incidences of selected wood
anatomical features in Northern Hemisphere Cretaceous woods from Albian to
Maastrichtian time as well as comparison with extant woods. Cretaceous woods
as a whole differ from Recent woods in having higher incidences of exclusively
solitary vessels, scalariform perforation plates, and wide rays (>10-seriate), and
lower incidences of ring porosity, wide vessels (> 200 µm), vessels in groups,
non-random arrangements of vessels, and marginal parenchyma. The occurrence of relatively high percentages of storage cells (> 40%) in some Cretaceous
trees is noteworthy; the ability to produce wood with varying amounts and
arrangements of parenchyma is likely to be a contributing factor to the success
of angiosperm trees in a wide variety of environments.
Key words: Big Bend National Park, fossil wood, Ericales, paleobotany, Cretaceous, Aguja Formation, Javelina Formation, parenchyma, Paleocene, Black
Peaks Formation.
1)Department of Wood and Paper Science, North Carolina State University, Raleigh, NC 27695-8005,
and NC State Museum of Natural Sciences, 11West Jones Street, Raleigh, NC 27601-1029, U.S.A.
2)Department of Geosciences, Texas Tech University, Lubbock, TX 79409-1053, U. S. A.
294
IAWA Journal, Vol. 30 (3), 2009
INTRODUCTION
We know relatively little of the wood anatomy of Cretaceous and Paleocene angiosperms. In the early 1990s a compilation based on the literature found worldwide only
108 records of Cretaceous dicot woods and 36 records of Paleocene dicot woods whose
anatomy had been described in some detail and for which provenance information was
available (Wheeler & Baas 1991, 1993). All but 12 of those 108 Cretaceous woods
were from the latest Cretaceous (Campanian-Maastrichtian), and approximately half
had characteristics of three genera: Paraphyllanthoxylon, Icacinoxylon, or Plataninium.
Since then there have been new reports of Cretaceous dicotyledonous woods, but the
number is still low (about 200 records worldwide) so that any new reports of Cretaceous
woods contribute to our understanding of the diversity of Cretaceous woody plants.
Big Bend National Park in southwestern Texas preserves a continuous sequence of
Upper Cretaceous (Campanian-Maastrichtian) through Lower Tertiary (Paleocene) continental strata. We have described a diverse assemblage of both dicot and conifer woods
collected from these strata (Wheeler 1991; Wheeler et al. 1994; Lehman & Wheeler
2001; Wheeler & Lehman 2000, 2005). As a result of reviewing recent collections in
preparation for an overview of the Cretaceous and Paleocene woods of Big Bend, we
discovered three additional Cretaceous wood types and one Paleocene wood type. The
objectives of this paper are to describe these new wood types, compare them to other
North American Cretaceous-Paleocene woods, to provide an overview of Northern
Hemisphere Cretaceous woods and compare them with present-day woods.
STRATIGRAPHY AND SEDIMENTARY FACIES
Upper Cretaceous and Paleogene strata in the Big Bend region of Texas record the
gradual withdrawal of the Cretaceous Interior Seaway from North America and preserve
a succession of coastal to inland paleoenvironments (Fig. 1). A series of fossil wood
horizons occur within coastal lowland deposits of the Aguja Formation, and upward
through inland fluvial flood-plain deposits of the Javelina and Black Peaks Formations
(Wheeler & Lehman 2000, 2005).
The Aguja Formation contains two terrestrial intervals referred to informally as the
lower and upper shale members (Lehman & Busbey 2007). Both the lower shale member
and the lowermost part of the upper shale member consist primarily of dark organic-rich
shale, lignite, and coal that accumulated in poorly drained brackish water marshes or
swamps close to the shoreline. These deposits interfinger with coastal strandplain and
marine shelf deposits of the Pen Formation. The upper part of the upper shale member
consists of drab fluvial flood-plain deposits, primarily olive, yellow, and gray mudstone
interbedded with lenticular stream channel sandstone. These sediments accumulated
in better-drained, fresh water alluvial environments some distance landward of the
shoreline. Most of the diverse fossil wood assemblage thus far documented for the
Aguja Formation has been recovered from the upper shale member, including a unique
association of in situ tree stumps (Agujoxylon olacaceoides and Metcalfeoxylon kirtlandense) that demonstrates dicot trees as dominant canopy-forming elements in some
Wheeler & Lehman — Cretaceous and Paleocene woods
295
Late Cretaceous southern latitude tropical forests (Lehman & Wheeler 2001). Fossil
wood is particularly abundant in the lowermost 10 to 20 m of the upper shale member
(the cupressoid/podocarpoid “acme zone” of Wheeler & Lehman 2005) where two of
the woods described in the present study (GS 8 and GS 9.2) were also recovered. Vertebrate biostratigraphy (Rowe et al. 1992) and radiometric age determinations (Befus
et al. 2008) indicate that this wood-bearing interval is middle Campanian in age.
EAst
erosional surface overlain
by Eocene and younger strata
5
4
SDF 8
3
T
K
2
JAVELINA FM
AGUJA FM
200 m
GS 11
GS 9.2
GS 8
paralic
100
1
BLACK PEAKS
FM
marine
PEN
FORMATION
0
West
8
7
6
top of
BOQUILLAS FM
385
0
10
texas
20
Km
118
5
3
1 2
Rio
4
6
8
N
7
103 W
Gr
an
de
29 N
Big Bend
National Park
Figure 1. Upper Cretaceous and Paleocene stratigraphy in Big Bend National Park, southwestern Texas, showing position of Cretaceous-Tertiary (K/T) boundary, and map showing general
outcrop distribution within the Park (stippled), selected measured sections (1 through 8), and
positions of wood collection sites discussed in text near Gano Spring (GS 8, 9.2, and 11) and
near South Dagger Flat (SDF 8).
296
IAWA Journal, Vol. 30 (3), 2009
The Javelina and Black Peaks formations also consist largely of fluvial overbank
deposits. These are more brightly colored than in the Aguja Formation, with prominent
beds of gray, green, purple, and red mudstone and with abundant interstitial calcium
carbonate nodules. The color bands represent well-differentiated buried soil profiles
(paleosols) and indicate that extended periods of soil development occurred on these
well-drained alluvial flood-plains (Lehman 1989, 1990). The flood-plain deposits are
interbedded with thick lenses of conglomeratic stream channel sandstone. The Cretaceous/Tertiary boundary occurs within this succession immediately above the Javelina/
Black Peaks formational contact (Fig. 1).
The lower part of the Javelina Formation preserves a diverse fossil wood assemblage,
including several taxa also found in the underlying Aguja Formation (e.g., Baasoxylon
parenchymatosum; Wheeler & Lehman 2000). One of the new wood types described
in the present report (GS 11) was obtained from this interval. Radiometric dating of
a tuff bed above this interval indicates that the lower Javelina Formation is middle
Maastrichtian in age (Lehman et al. 2006). There is a reduction in the number of wood
types in the upper part of the Javelina Formation, where a single type of tree – the huge
dicot Javelinoxylon multiporosum (Wheeler et al. 1994) dominates. Rare araucariacean
conifers also occur, and perhaps grew as isolated emergents in a tropical evergreen
Javelinoxylon forest (Wheeler & Lehman 2005). The same pattern persists following the
K-T boundary. A prolific fossil wood horizon (known informally as the “log jam bed”)
occurs in the lower part of the Black Peaks Formation, which is dated as early Paleocene
(Torrejonian) based on mammalian biostratigraphy (Lehman & Busbey 2007). Many
thousands of logs are preserved in the “log jam bed” – a thin but widespread stream
channel sandstone. However, of the many logs examined microscopically, all but two
pertain to the same species – Paraphyllanthoxylon abbotti (Wheeler 1991). One of the
woods described in the present study (SDF 8) was recovered from the “log-jam bed” and
represents a unique wood type in this interval. Only rare araucariacean conifer woods
have been recovered from higher in the Black Peaks (Wheeler & Lehman 2005).
MATERIALS AND METHODS
The wood specimens described were collected at sites near Gano Spring (GS) and
South Dagger Flat (SDF) in Big Bend National Park, Texas. Exact locality information for these sites is on file with the specimens at the North Carolina Museum of
Natural Sciences (NCMNS). All woods examined are silicified. A diamond lapidary
saw was used to cut thick sections (wafers) of cross (transverse), tangential, and radial
surfaces. One side of each wafer was smoothed to remove saw marks, and then affixed
to a glass slide using epoxy resin. The sections were then ground until they were thin
enough (c. 30–50 µm) so that anatomical details could be observed with transmitted
light microscopy. Photographs were taken with an Olympus D70 digital camera that
embeds scale bars in the images.
The descriptions of woods given below generally conform to the IAWA Hardwood
List (IAWA Committee 1989). Relationships to extant groups were investigated by using
the InsideWood web site (InsideWood 2004 – onwards) and references found on the
Wheeler & Lehman — Cretaceous and Paleocene woods
297
Kew Micromorphology Bibliography web site. Family and order names are those used
on the Angiosperm Phylogeny Website (Stevens 2001-onwards). A dot grid method was
used to estimate relative percentages of cell types based on a minimum 500-point count.
The samples so measured were ones in hand, well-preserved, and from trunks or logs
more than 30 cm in diameter. Images were opened in Image J (Rasband 1997–2009),
a grid was overlain on the image, and the dots on vessels, fibers, axial or ray parenchyma, respectively, were each counted separately.
Cretaceous woods
DESCRIPTIONS
Sabinoxylon Estrada-Ruiz, Martinez-Cabrera & Cevallos-Ferriz 2007
Sabinoxylon wicki sp. nov. (Fig. 2)
Growth rings indistinct.
Diffuse porous. Vessels predominantly solitary, oval in outline. Mean tangential
diameter 84 (14) µm.; 9–11 per sq. mm. Perforation plates scalariform, most commonly 7–10 bars. Vessel-vessel or vessel-tracheid pits alternate to opposite, 5–8 µm
across. Vessel-ray pits small, without reduced borders, crowded. Vessel element lengths
735–1050 µm.
Axial parenchyma diffuse, diffuse in aggregates, generally 8 cells per strand.
Vasicentric tracheids present. Non-septate fibers with medium-thick to thick walls,
some fibers with distinctly bordered pits on both radial and tangential walls.
Rays of two sizes, 1–4-seriate and more than 20 cells wide, in some regions 6 to 8
cells wide. Heterocellular, narrow rays with intermixed procumbent, square, and upright
cells, wide rays with marginal rows of square and upright cells. Individual wide rays to
more than 4 mm high, wide and tall rays usually in vertical series so that with a hand lens
ray height appears greater. More than 12 rays per mm in regions in between wide rays.
Crystals occasional in chambered upright/square ray parenchyma cells.
Holotype: NCSM 1086 (GS 11).
Etymology: Named for Steve Wick of Terlingua, Texas, who collected this specimen,
and has assisted with our fieldwork in Big Bend.
Horizon: Upper Cretaceous, Maastrichtian, Javelina Formation.
Comparisons to extant plants – The combination of exclusively solitary vessels
(Fig. 2A, B), scalariform perforation plates (Fig. 2C), vessel-ray parenchyma pits that
do not have reduced borders (Fig. 2C), vasicentric tracheids (Fig. 2C), diffuse axial
parenchyma (Fig. 2A, B), two size classes of rays (Fig. 2E, F) with the widest rays being
heterocellular (Fig. 2D) and more than 10 cells wide is found in some extant species of
Buxaceae (Buxales), Dilleniaceae (Dilleniales, core eudicot), Geissolomataceae (Crossomatales, Rosid, EuRosid I), Hydrangeaceae (Cornales, Asterid), and Pentaphylacaceae
(Ericales). When the occurrence of crystals in chambered ray parenchyma cells is added
(Fig. 2C), that combination only occurs in species of the Pentaphylacaceae (Ericales,
Asterid, Euasterid I).
The Pentaphylacaceae comprises 12 genera, with three groups (Pentaphylaceae,
Ternstroemieae, Frezierieae). Fossil flowers of the Ericales are known from the Cretaceous (Schönenberger & Friis 2001).
298
IAWA Journal, Vol. 30 (3), 2009
Wheeler & Lehman — Cretaceous and Paleocene woods
299
Comparisons to fossils –This Big Bend wood fits the generic diagnosis of Sabinoxylon
from the late Campanian-early Maastrichtian Olmos Formation of northern Mexico:
predominantly solitary vessels, alternate to opposite pits, scalariform perforation plates
with fewer than 20 bars, diffuse-in-aggregates axial parenchyma, two size classes
of rays with the wider rays being more than 20 cells wide, and vasicentric tracheids
(Estrada-Ruiz et al. 2007). The Big Bend wood is assigned to a new species because
some of the rays in the narrow ray class are wider (up to 4 cells) than in the Olmos
Formation wood (exclusively uniseriate), some fibers have bordered pits on both radial
and tangential walls, and crystals occur in chambered upright ray cells.
Preservation of Mexican Sabinoxylon pasac was such that vessel-ray parenchyma
pits could not be observed. The Big Bend wood has vessel-ray parenchyma pits that are
similar to intervessel pits. There also are quantitative differences between the Big Bend
wood and the Olmos Formation wood, which has a mean vessel tangential diameter of
84 (14) µm; 3–10 vessels per sq. mm, mean vessel element length 1124 (184) µm, 7–16
bars per perforation, and ray widths to 25 cells. These types of quantitative differences
can be related to differences in local growing conditions and location within the tree.
Big Bend Ericalean Wood Type I (Fig. 3)
Wood diffuse porous. Vessels solitary (66%) and in short radial multiples. Solitary
vessels oval to circular in outline. Mean tangential diameter 70 (13) µm, 40–97 µm;
18–20 vessels per sq. mm. Perforation plates scalariform, usually with around 20 bars.
Intervessel pits crowded alternate, angular in outline, with included apertures, 4–7 µm.
Vessel-ray parenchyma pits similar in size to intervessel pits, or slightly smaller. Vessel
element lengths 452–904 µm. Bubble-like tyloses common in many vessel elements.
Fiber walls thin to medium-thick, pits not observed. Non-septate.
Diffuse axial parenchyma, strands of more than 8 cells.
Rays mostly 4–6-seriate, to 8-seriate, heterocellular, with intermixed procumbent
and square to upright cells in the body of the ray, and uniseriate marginal rows of a
variable number of square/upright cells. Multiseriate ray height averages 888 (532)
µm, range 222–2220 µm, 10–12 per mm.
No crystals observed.
Material: NCSM 1087 (GS 8).
Horizon: Upper Cretaceous, Campanian, Aguja Formation (upper shale member).
Comments – In InsideWood, modern woods that are diffuse porous with vessels
solitary and in radial multiples (Fig. 3A), exclusively scalariform perforation plates
(Fig. 3B, C, D), only alternate intervessel pits (Fig. 3E), vessel-ray parenchyma pits simi-
←
Figure 2. Sabinoxylon wicki sp. nov. – A: Diffuse porous wood, with vessels exclusively solitary,
region without wide rays, XS. – B: Solitary vessels, diffuse parenchyma, XS. – C: Scalariform
perforation plates, vasicentric tracheids, vessel-ray parenchyma pits small and without reduced
borders, crystals in chambered upright ray parenchyma (arrow), RLS. – D: Ray composed of
procumbent, upright and square cells, RLS. – E: Rays of two sizes, TLS. – F: Rays of two sizes,
narrow rays 1–3 cells wide, fibers with distinctly bordered pits on their tangential walls, TLS.
— Scale bar = 200 µm in A, E, 100 µm in B, D, 50 µm in C, F.
300
IAWA Journal, Vol. 30 (3), 2009
Wheeler & Lehman — Cretaceous and Paleocene woods
301
lar to intervessel pits (Fig. 3F), diffuse axial parenchyma (Fig. 3A), and heterocellular
rays (Fig. 3 I) that are more than 4 cells wide (Fig. 3G, H) are found in the Ericaceae
and Styracaceae of the Ericales. Because of the resemblance of this wood to those found
in the order Ericales, we refer to it here as the Big Bend Ericalean Wood Type I
Big Bend Small Axis Campanian Wood (Fig. 4)
Wood diffuse porous. Vessels solitary (8%) and in short radial multiples and occasional clusters. Solitary vessels oval to circular in outline. Mean tangential diameter
49 (11) µm, range 27–69 µm; 28–52 vessels per sq. mm. Perforation plates simple.
Intervessel pits crowded alternate, rounded in outline, with included apertures, small
c. 5 µm. Vessel-ray parenchyma pits with reduced borders, horizontally elongated and
irregular in shape. Vessel element lengths 347–448 µm. Widely spaced thin-walled
tyloses present in all vessels.
Fiber walls thin to medium-thick, pits not observed.
Scanty paratracheal parenchyma, intermixed with vessel groups. Strands of 5–8 cells.
Rays 1–3- (4-)seriate, mostly 3-seriate, heterocellular, body ray cells procumbent
with 1–2 marginal rows of square/upright cells. Procumbent cells not much elongated.
Multiseriate ray height averages 313 (126) µm, range 120–682 µm, 9–11 per mm.
No crystals observed.
Material: NCMS 1088 (GS 9.2).
Horizon: Upper Cretaceous, Campanian, Aguja Formation (upper shale member).
Comments – The appearance of this wood in the hand sample was such that we expected it was another example of the commonly occurring small diameter axes found
at localities RM 28 and CR-2 and which we described as “Baileyan Big Bend Wood,
Type I” (Wheeler & Lehman 2000). However, sections reveal that this is a different
wood type. Vessel multiples are common (Fig. 4A, B), perforation plates are simple
(Fig. 4E), intervessel pitting is alternate (Fig. 4D), and axial parenchyma is predominantly paratracheal (Fig. 4B). All these features distinguish this wood from the commonly occurring “Baileyan Big Bend Wood, Type I” which has predominantly solitary
vessels, scalariform perforation plates, and abundant diffuse to diffuse-in-aggregates
parenchyma.
This sample (NCMS 1088) has pith preserved (Fig. 4C) and is a small diameter axis.
Whether this sample represents a young axis of a species of tree, an erect shrub or a
scandent shrub is not known. The appearance of the axes in the sediments suggests
a scandent form. Ray cellular composition and width differ between juvenile wood
and mature wood. The only ray characteristics that can be used to try to determine the
affinities of the wood are the absence of exclusively uniseriate rays and absence of
←
Figure 3. Big Bend Ericalean Wood. – A: Diffuse porous wood, with vessels solitary and rarely
in radial multiples. – B: Scalariform perforation plate. – C: Multiple perforation plate. – D & E:
Crowded alternate intervessel pits. – F: Vessel-ray parenchyma pits. – G & H. Multiseriate heterocellular rays, axial parenchyma strands. – I: Procumbent and square/upright cells intermixed
in the ray, vessel elements with scalariform perforation plates. — Scale bar = 200 µm in A, G;
100 µm in H, I; 50 µm in B, C, D, F; 20 µm in E.
302
IAWA Journal, Vol. 30 (3), 2009
Wheeler & Lehman — Cretaceous and Paleocene woods
303
heterocellular rays (Fig. 4F) commonly more than 10-seriate (Fig. 4G). Fibers do not
appear to be septate so that this is not a phyllanthoid wood type. The wood pattern of
randomly arranged vessels that are solitary and in multiples, simple perforation plates
and alternate pitting that is not >10 µm across, non-septate fibers without distinctly
bordered pits, predominantly paratracheal parenchyma, and rays that are not exclusively
uniseriate or commonly more than 10-seriate is widespread in present-day angiosperms
occurring in species of at least 27 families of 12 orders of core eudicots (as defined
by Stevens 2001– onwards), Sapindales (Anacardiaceae, Meliaceae, Rutaceae, Simaroubaceae), Gentianales (Apocynaceae), Apiales (Araliaceae), Asterales (Asteraceae),
Rosales (Barbeyaceae, Cannabaceae, Moraceae, Rhamnaceae), Lamiales (Bignoniaceae, Oleaceae, Scrophulariaceae, Verbenaceae), Brassicales (Capparaceae), Malpighiales (Dichapetalaceae, Malpighiaceae, Euphorbiaceae, Ochnaceae, Salicaceae),
Fabales (Leguminosae/Fabaceae), Myrtales (Lythraceae, Melastomataceae), Ericales
(Myrsinaceae, Polemoniaceae), Solanales (Solanaceae).
Paleocene wood
cf. Cunonioxylon sensu Gottwald (Fig. 5)
Growth rings indistinct. Wood diffuse porous. Vessels mostly solitary (87%), and
in radial pairs. Vessels oval to slightly blocky in outline. Mean tangential diameter 81
(11) µm; 12–20 per sq. mm. Perforation plates exclusively scalariform, 5–12 bars, bars
about 5–6 µm thick, and about 11–14 µm apart. Intervessel pitting opposite. Vessel-ray
parenchyma pits horizontally elongate and with reduced borders. Widely spaced tyloses
common.
Axial parenchyma rare.
Fibers with bordered pits on both radial and tangential walls, non-septate.
Rays 1–4-seriate, mostly 2–3-seriate, predominantly of upright and square cells,
with some intermixed procumbent cells, with long uniseriate margins, rays often of
the interconnected type with tall alternating multiseriate and uniseriate portions, total
ray height more than 1 mm, 10–14 per mm.
Crystals not observed.
Material: SDF 8.
Horizon: Paleocene (Torrejonian), lower Black Peaks Formation.
Comparisons to extant plants – The combination of vessels randomly arranged
(Fig. 5A, B), exclusively scalariform perforation plates (Fig. 5C, F), opposite intervessel pits (Fig. 5E, vessel-ray parenchyma pits with reduced borders (Fig. 5D), rare axial
parenchyma (Fig. 5A, B), non-septate fibers with pits on both radial and tangential
walls (Fig. 5C, G), and heterocellular rays occurs in the Cunoniaceae (species of Wein-
←
Figure 4. Big Bend Small Axis Campanian Wood. – A: Diffuse porous wood, with vessels solitary
and in short radial multiples. – B: Vessels in short radial multiples. – C: Pith cells isodiametric,
cells not very variable. – D: Crowded alternate intervessel pits. – E: Simple perforation plate. –
F: Body of rays composed of procumbent cells. – G: Rays mostly 3-seriate. — Scale bar =
200 µm in A, 100 µm in B, F, 50 µm in D, E, 20 µm in C.
304
IAWA Journal, Vol. 30 (3), 2009
Wheeler & Lehman — Cretaceous and Paleocene woods
305
mannia) in the Oxalidales. Today the Cunoniaceae is a Southern Hemisphere family
primarily centered in eastern Australia, New Guinea, and New Caledonia, with a few
genera in southern Africa, Madagascar and the Neotropics, including Mexico (Sharp
1953; Hufford & Dickison 1992).
Comparisons to fossils – Schönenberger et al. (2001) reviewed the fossil record of
Cunoniaceae. The bulk of the records are from the Southern Hemisphere, including
a Cretaceous wood from Antarctica, Weinmannioxylon nordenskjoeldii (Poole et al.
2000). The genus Weinmannioxylon Petriella was diagnosed as having axial parenchyma
diffuse or grouped in small tangential bands (translated diagnosis seen in Poole et al.
2000a), features not seen in this Big Bend wood. There are a few Northern Hemisphere fossils assigned to the Cunoniaceae, e.g., flowers from the Late Cretaceous
(late Santonian–early Campanian) of Sweden (Schönenberger et al. 2001) and two
European fossil woods: Cunonioxylon weinmannioxylon from the Oligocene of Bavaria
(Hofmann 1952) and C. parenchymatosum from the Eocene of Germany (Gottwald
1992). Gottwald (1992) noted that the diagnosis of Cunonioxylon Hofmann was such
that it was somewhat questionable whether the wood was related to Cunoniaceae. Hofmann seems to have relied on the cross-sectional appearance of the Bavarian wood as
the basis for assigning it to the Cunoniaceae. However, her generic diagnosis only mentions simple perforations (“Gefassdurchbrechung einfach”), not scalariform. Gottwald
described the Eocene Cunonioxylon as having scalariform perforation plates with
15–40 bars and agreeing in all details with Cunonia. Evidently, as the wood resembled
Cunonia, it was assigned to Cunonioxylon, although it does not conform to Hofmann’s
diagnosis. Although the Big Bend wood has features seen in Cunoniaceae, we are not
assigning it to Cunonioxylon because the genus diagnosis mentions simple perforations
(“Gefassdurchbrechung einfach”) and axial parenchyma frequently in uniseriate rows,
features not seen in the Big Bend wood.
DISCUSSION
Regional Late Cretaceous vegetation — The presence of Sabinoxylon at Big Bend
provides additional evidence for similarity in Campanian-Maastrichtian woody vegetation of Texas, New Mexico and Mexico. Metcalfeoxylon occurs in the Campanian
Aguja Formation of Big Bend (Wheeler & Lehman 2000), Kirtland Formation in
the San Juan Basin, New Mexico (Wheeler et al. 1995; Hudson 2006), and Crevasse
Canyon Formation of south-central New Mexico (Upchurch et al. 2003). This genus
occurs as in situ stumps at sites in both Big Bend (Lehman & Wheeler 2001) and New
Mexico (Upchurch et al. 2003). Baasoxylon occurs in the upper Campanian-lower
←
Figure 5. cf. Cunonioxylon sensu Gottwald. – A: Diffuse porous wood, with vessels predominantly solitary, radial multiples rare. – B: Solitary vessels slightly angular in outline, fibers with thin
walls. – C: Scalariform perforation plate with thick, widely spaced bars. – D: Vessel-ray parenchyma pits with reduced borders and horizontally elongate. – E: Opposite pits. – F: Vessels with
widely spaced tyloses. – G: Rays mostly 3 cells wide, with long uniseriate portions, some fibers
with pits on tangential walls. — Scale bar = 200 µm in A, 100 µm in B, F, G, 50 µm in C, D, E.
306
IAWA Journal, Vol. 30 (3), 2009
Maastrichtian of Big Bend (Wheeler & Lehman 2000) and the San Juan Basin (Hudson 2006). Javelinoxylon, the most common wood in the Maastrichtian of Big Bend
(Wheeler et al. 1994), also occurs in the Maastrichtian of Mexico (Estrada-Ruiz et al.
2007). Collectively, these occurrences support the belief that the Late Cretaceous
southern latitude flora of North America shared common elements, and differed from
the northern latitude flora (Wheeler & Lehman 2005).
Characteristics of Northern Hemisphere Cretaceous woods — All the Big Bend
Cretaceous woods are diffuse porous and lack elaborate paratracheal parenchyma and
any type of banded parenchyma. The three new types of Cretaceous woods are variations on themes of commonly occurring Cretaceous woods; one has predominantly
solitary vessels, scalariform perforation plates and wide rays (features also found
in Icacinoxylon), the other two have vessels solitary and in radial multiples, simple
perforation plates and alternate pits (features also found in Paraphyllanthoxylon).
Features of Paraphyllanthoxylon and Icacinoxylon and incidences of selected features
in Cretaceous woods are discussed in more detail below.
Paraphyllanthoxylon vs. Icacinoxylon/Plataninium – The earliest known dicotyledonous woods with reliable provenance represent two distinct wood anatomical types,
the phyllanthoid type and the icacinoid/platanoid type (Cedar Mt. Formation, Albian
age, Utah, USA, Thayn et al. 1983, 1985). These two wood types are common throughout the Cretaceous, accounting for more than half of the records to date. The general
pattern of these two wood types occurs today in a number of dicotyledonous families
and orders; both patterns occur in the Eumagnoliids and the Eudicots.
Paraphyllanthoxylon is the more abundant of the two wood types, with large logs,
some more than 1 meter in diameter known from the Cenomanian of Alabama (Cahoon
1972) and Arizona (Bailey 1924). According to Bailey (1924) at the P. arizonense locality “fossil wood was fairly abundant … Some of the trunks are more than a foot in
diameter.” Jack Wolfe and one of the authors (EW) visited the Mogollon Rim region to
look for Bailey’s locality, a “barrow pit” in Show Low, Arizona. We easily found more
than 20 logs, behind the barrow pit (now a rodeo ring) and exposed along road cuts in
the area; most were more than 60 cm in diameter. The anatomy of all these Mogollon
Rim logs is that of Paraphyllanthoxylon arizonense. Paraphyllanthoxylon also occurs in the Albian of Utah (Thayn et al. 1983), Idaho (Spackman 1948) and Texas
(Serlin 1982; Wheeler 1991).
These Albian and Cenomanian Paraphyllanthoxylon species have vessels that
would be considered efficient in water conduction, as they are wide and composed of
elements with simple perforation plates. Characteristics of phyllanthoid wood include
vessels solitary and in radial multiples, simple perforation plates, alternate intervessel pitting, vessel-ray parenchyma pits with reduced borders, axial parenchyma rare
to scanty paratracheal, septate fibers that lack distinctly bordered pits, heterocellular
multiseriate rays, and absence of storied structure. This combination of features occurs
in Lauraceae, Burseraceae, Anacardiaceae, Salicaceae, and Phyllanthaceae. Herendeen
(1991b) found attachment between an axis with anatomy of Paraphyllanthoxylon and
Wheeler & Lehman — Cretaceous and Paleocene woods
307
a fruit of Lauraceae, demonstrating that at least one species of Paraphyllanthoxylon is
Lauraceae. However, it is unlikely that all species assigned to this genus are Lauraceae,
particularly those that have markedly heterocellular rays, which are likely to be Salicaceae or Phyllanthaceae. There are Cretaceous localities with mesofossils where the
opportunity exists for finding attachment between plant parts and allow determining
the affinities of these woods. These early phyllanthoid woods are “problematic” in the
way the “Stopes woods” were (whose age is not yet settled), i.e., they have “advanced”
wood structure, see Stopes & Fujii 1910; Stopes 1912, 1915).
Characteristics of Icacinoxylon include exclusively to predominantly solitary vessels, scalariform perforation plates, wide (more than 4-seriate to 10 or more cells wide)
and tall rays (>1 mm high). This combination of characters occurs in a number of eudicot orders: Crossomatales (Aphloiaceae and Staphyleaceae), Berberidopsidales (Berberidopsidaceae), Aquifoliales (Cardiopteridaceae, Helwingiaceae), Apiales (Pennantiaceae, Griseliniaceae), Asterales (Rousseaceae), Dilleniales (Dilleniaceae), Ericales
(Ericaceae), Ranunculales (Eupteleaceae, Lardizabalaceae), Cornales (Hydrangeaceae),
as well as in the Icacinaceae (unplaced as to order, but near Garryales). There is considerable variation within Icacinoxylon; e.g., there are species with septate fibers and species
with only non-septate fibers. As Herendeen (1991a) pointed out, a review of woods
assigned to Icacinoxylon is needed. Woods assigned to Plataninium or Platanoxylon
usually have some vessel multiples and rays tend to be or are homocellular (e.g., Page
1968; Wheeler et al. 1995; Takahashi & Suzuki 2003).
Icacinoxylon and Paraphyllanthoxylon antedate the earliest pollen and mesofossil
record of angiosperms by tens of millions of years. It is puzzling that two such different wood patterns share the distinction of being the earliest confirmed dicot wood
types. This might imply multiple origins of vessel elements within the angiosperms,
once with simple perforation plates, once with scalariform perforation plates, or could
instead reflect a change that occurred in vessel structure from a single origin. These
Albian-Cenomanian woods do not shed light on the stature of the earliest angiosperms,
but do indicate that dicot trees were an important element in some ecosystems by the
Cenomanian. On the basis of a bibliographic survey and new collections of BerriasianCenomanian woods of Europe, Philippe et al. (2008) suggest that early angiosperms
were small statured and had physically weak wood; large fossil angiosperm trunks
are absent from the European Albian-Cenomanian and the abundance of angiosperm
wood is greater in Cenomanian strata. Cantrill and Poole (2005) found comparable
results for Antarctic fossil wood assemblages; pre-Cenomanian beds did not contain
angiosperms. Another example of change in Albian-Cenomanian abundance and diversity of angiosperm woods is provided by Takahashi and Suzuki (2003). Two (4%)
of their 54 Albian wood samples are angiosperms (Icacinoxylon), while 7 (31%) of
33 Cenomanian wood samples are angiosperms (Icacinoxylon, Paraphyllanthoxylon,
Plataninium, and Hamamelidoxylon).
Comparison of Cretaceous Dicot Woods to Modern Dicot Woods — For purpose of the
following comparison, we divided the Northern Hemisphere Cretaceous wood records
into three time bins; Albian-Cenomanian (16 records) when angiosperm woods first
308
IAWA Journal, Vol. 30 (3), 2009
occur and when other types of fossils show a dramatic increase in angiosperm diversity
(Crane 1987; Butler et al. 2009); Turonian-Santonian (50 records), and CampanianMaastrichtian (134 records). These intervals are similar to those discussed by Butler
et al. in their study of diversity patterns among herbivorous dinosaurs and plants. Each
published report of a fossil wood species or type was treated as a different record; some
‘species’ have more than one record because that ‘species’ was described from more than
one locality or age; for example, Paraphyllanthoxylon arizonense was described from
Porosity
%
%
100
100
90
90
80
80
70
60
50
40
30
RP
50 –100
< 50
50
40
DP
30
20
10
10
0
K2
K3
Recent
Exclusively solitary vessels
%
100 –200
60
SRP
K1
> 200
70
20
A
Mean tangential diameter (µm)
0
B
%
50
K1
K2
K3
Recent
Radial arrangement of vessels
25
45
40
20
35
30
15
25
20
10
15
10
5
5
0
C
K1
K2
K3
Recent
Vessel clusters common
0
D
K1
%
K3
Recent
Perforation plates
%
100
25
K2
90
Scal
80
Simp
70
20
60
50
15
40
10
30
20
5
10
0
E
K1
K2
K3
Recent
0
F
K1
K2
K3
Recent
Figure 6. Incidences (percentage occurrence) of wood anatomical features in the Cretaceous:
Albian-Cenomanian (K1), Turonian-Santonian (K2), Campanian-Maastrichtian (K3), and Recent
(R) flora worldwide. Thin lines above bars of graph = one standard deviation. – A: Porosity. RP =
ring porous, SRP = semi-ring porous, DP = diffuse porous. – B: Mean vessel tangential diameter
by IAWA category. – C: Exclusively solitary vessels. – D: Radial arrangement of vessels. –
E: Vessel clusters common. – F: Perforation plates. – Scal = scalariform perforation plates, Simp =
simple perforation plates.
Wheeler & Lehman — Cretaceous and Paleocene woods
309
the Cenomanian of Arizona (Bailey 1924) and the late Campanian-early Maastrichtian
of New Mexico (Wheeler et al. 1995; Hudson 2006). Appendix I contains the sources
for these Cretaceous records, Appendix II the sources for Paleocene woods, excluding
those references cited in the text and given in the reference section. Most of the Albian
and Cenomanian woods from Europe reported by Philippe et al. (2008) have not yet
been described in detail, and so these woods are not included in this survey.
The incidence of selected wood anatomical features is reported here as a percentage.
The base numbers on which the percentages for the three divisions of the Cretaceous
and the Recent flora are given above, but with some differences by feature as some
descriptions did not provide information on all of the features surveyed. The standard
deviations for the percentages are based on the formula of Steel and Torrie (Steel et al.
1997): estimated SD = square root of [p (1-p) / n], where p = percentage/100 and n =
number of counts for that feature.
Because of the small sample size for Albian-Cenomanian woods, and resultant large
standard deviations it is difficult to establish that there are any statistically significant
differences between woods of that time interval and the Santonian-Turonian or the
Campanian-Maastrichtian.
Porosity (Fig. 6A) – Diffuse porous woods dominate throughout Cretaceous time, as
they do in the Recent flora. Ring porous woods are not common even in the extant flora,
and are primarily a Recent north temperate phenomenon (Gilbert 1940; Wheeler et al.
2007). The few ring porous woods from the latest Cretaceous are from high latitudes
in both Southern (Poole et al. 2000b) and Northern Hemispheres (Takahashi & Suzuki
2003). Ring porosity is correlated with deciduousness (Boura & De Franceschi 2007).
The low incidence of ring porous woods in the latest Cretaceous supports Wolfe’s
(1987) analysis that the terminal Cretaceous event was important for the spread and
diversification of deciduous taxa.
Vessel diameters (Fig. 6B) – Today diffuse porous woods with uniform vessel diameters and mean tangential diameters of more than 200 µm occur only in trees (Wheeler
et al. 2007); such woods occur predominantly in the wet tropics. Although most Cretaceous woods have vessel diameters of <100 µm, woods with wider diameters do occur
and are consistent with some being large trees, particularly phyllanthoid types from
the Albian-Cenomanian (Bailey 1924; Cahoon 1972; Serlin 1982). In the Recent flora
there are higher proportions of both very wide and very narrow vessels, as would be
expected with dicots expanding from Cretaceous time into a wider range of habitats,
from arid to arctic (favoring narrow vessels) and development of the multistratal forests
of the wet tropical lowlands during the Tertiary.
Vessel groupings and arrangement (Fig. 6C, D, E) – Consistent with earlier studies
of the incidence of dicot wood features through time (Wheeler & Baas 1991, 1993),
exclusively solitary vessels are more common in the Cretaceous than at present. Nonrandom vessel arrangements, for example, radial to diagonal arrangement of vessels
and vessel clusters, gradually increased in abundance toward the end of Cretaceous
time, but remained rare compared to Recent time.
Perforation plates (Fig. 6F) – The most striking difference between woods of the
Cretaceous and Recent is the high incidence of scalariform perforation plates during the
310
IAWA Journal, Vol. 30 (3), 2009
Cretaceous (more than 50%). It is somewhat surprising that this is the case, as simple
perforations occur in the earliest dicot woods (species of Paraphyllanthoxylon) and
dominate in the Recent flora. However, here we count reports of wood types, not their
relative abundance. At the localities where Paraphyllanthoxylon occurs it is very common (Bailey 1924; Cahoon 1972; Hudson 2006; Oakley & Falcon-Lang 2008), and so
one could hypothesize that during the Cretaceous trees with simple perforation plates
accounted for more biomass and were more successful than those with scalariform
perforation plates.
Axial parenchyma (Fig. 7A, B) – The incidence of marginal parenchyma is markedly
higher in the Recent flora than in the Cretaceous, most likely indicating a lower incidence
of seasonal climates and cambial dormancy in the Cretaceous. Although correlations
of marginal parenchyma occurrence with cambial dormancy and deciduousness have
not been investigated to the same extent as correlation between ring porosity and de-
%
Apotracheal parenchyma
100
%
Dif
Dif-A
Marg
90
80
70
80
70
60
50
50
40
40
30
30
20
20
10
10
A
%
Ray cellular composition
0
B
%
100
50
90
45
80
40
70
35
60
30
50
25
40
20
30
15
20
10
10
5
Ray width classes
0
0
C
ScP
Vc
Alf
90
60
0
Paratracheal parenchyma
100
Homocellular
Heterocellular
D
1
1-3
4-10
>10
Figure 7. Incidences (percentage occurrence) of wood anatomical features in the Cretaceous:
Albian-Cenomanian (K1), Turonian-Santonian (K2), Campanian-Maastrichtian (K3), and Recent
(R) flora worldwide. Thin lines above bars of graph = one standard deviation. – A: Apotracheal
parenchyma. Dif = diffuse, Dif-A = diffuse-in-aggregates, Marg = marginal. – B: Paratracheal
parenchyma. ScP = scanty paratracheal, Vc = vasicentric, Alf = aliform. – C: Ray cellular composition. Homocellular = rays composed of procumbent cells, heterocellular = rays with both
procumbent and upright /square cells. – D: Ray width classes. 1 = exclusively uniseriate, 1–3 =
rays 1 to 3 cells wide, 4–10 = wider rays commonly reach 4 to 10 cells wide, >10 = wider rays
commonly reach more than 10 cells wide.
Wheeler & Lehman — Cretaceous and Paleocene woods
311
ciduousness (Boura & De Franceschi 2007), presumably they are correlated, but this
needs to be investigated. Aliform axial parenchyma is much rarer in the Cretaceous
than in the Recent flora (Fig. 7B).
Ray parenchyma (Fig. 7C, D) – Heterocellular rays have always been more common than homocellular rays. The differences between ray widths in the Cretaceous and
Recent flora are at the extremes, with uniseriate rays more common in the Recent flora
than in the Cretaceous, and wide rays more common in the Cretaceous than at present.
The higher incidence of wide rays in the Cretaceous reflects the common occurrence
of woods of the Icacinoxylon or Platanus-type in the Cretaceous.
Storage tissue in Cretaceous trees – Some Cretaceous dicot woods with trunk diameters of more than 50 cm have a high percentage of storage cells (Table 1). The most
commonly occurring Cretaceous wood, Paraphyllanthoxylon, lacks axial parenchyma
but its ground tissue is comprised of septate fibers, which in extant plants function
as storage cells. The Cenomanian P. arizonense from the Mogollon Rim of Arizona
(Bailey 1924) and the Maastrichtian P. illinoisense from the McNairy Formation of
Illinois (Wheeler et al. 1987) are comprised of more than 75% storage cells. The other
common tree type in the McNairy Formation (Parabombacaceoxylon magniporosum)
has more than 40% parenchyma. Parenchyma percentages for Big Bend Cretaceous
trees are between 40 and 60% (Table 1).
Table 1. Percentage of ray parenchyma (RP), axial parenchyma (AP), and septate fibers
(SpFib) and total (TOT) percentage of storage tissue in selected Cretaceous trees from
the U.S.A. — Ab = absent; ? = cell type present, but not able to determine percentage;
M = Maastrichtian, Cmp = Campanian; Cen = Cenomanian.
Species
Age
Max RW
% RP
Big Bend Ericalean Wood Cmp
6
41
?
Ab
41+
Sabinoxylon wicki M
20 +
38
?
Ab
38 +
Paraphyllanthoxylon illinoisense M
6
39
Ab
46
85
Parabombaceoxylon magniporosum M
2 (3)
26
17
Ab
43
Metcalfeoxylon kirtlandense Cmp-M
4
37
14
Ab
51
Baasoxylon parenchymatosum Cmp-M
8
55
?
Ab
55 +
Agujoxylon olacaceioides Cmp-M
6
34
10
Ab
44
Cen
6
33
Ab
47
80
GS 8
SIPC 622. 155
SIPC 608.18, SIPC 611.20
USNM 507055 (EDF 2A)
USNM 507057 (PGH 4a)
USNM 507039, 507042
Paraphyllanthoxylon arizonense
% AP % SpFib. TOT
312
IAWA Journal, Vol. 30 (3), 2009
The consequences or advantages (if any) of such high proportions of living cells for
Cretaceous angiosperm trees are therefore topics of interest. Living cells function in
storage, for both starch and water (Holbrook 1995). Large amounts of parenchyma with
stored food material might aid in rapidly replacing cropped foliage, and be valuable
for wound response, an ability that could be useful in environments characterized by
large dinosaurian herbivores. Abundant parenchyma might also affect the ability of a
tree to stump sprout if it were subject to being toppled by large herbivores.
High proportions of storage cells in these Late Cretaceous trees might result in a
condition somewhat analogous to that found in stem succulents. Buried calcareous soil
profiles in Cretaceous wood-bearing deposits (e.g., Lehman 1989, 1990) indicate that
at least some of these trees inhabited semi-arid environments. Periods of low water
availability in Cretaceous environments could have made stored water useful, and
large amounts of stored water might inhibit the production of growth rings, as water
would be available year round and cambial activity might not reflect fluctuations in
external water availability. In Cretaceous deposits, conifer woods with distinct growth
rings co-occur with dicots lacking distinct growth rings (Wheeler & Lehman 2005).
Cretaceous angiosperms could have used stored water to avoid water deficits, thus
bypassing seasonal effects, or in contrast, conifer species may have been genetically
stimulated to produce growth rings even when seasonality was not pronounced.
High proportions of ray parenchyma would also be expected to have mechanical
consequences. Relationships between wood anatomical features and physical properties are complex. Parenchyma cell walls are different from fiber and vessel element
walls. Ray parenchyma is horizontally oriented, while fibers and vessels are vertically
oriented. A study of 12 north temperate species found that ray volume affected modulus
of elasticity (MOE) in the radial direction of the wood (Burgert et al. 2001). Studies
of the properties of isolated rays indicate that rays affect the radial stiffness of wood
(Kawamura 1984; Badel & Perré 1999). Greater radial stiffness could have been advantageous for resisting toppling by large dinosaurian herbivores of the Cretaceous or by
strong storms that are likely to have occurred along the margins of inland Cretaceous
seas (Wheeler & Lehman 2005). There may be correlations between high proportions
of ray parenchyma and tree architecture because of the mechanical consequences of
high proportions of parenchyma. Studies to measure ray parenchyma proportions in
different tree architectural types could test whether any such relationships exist.
There are experimental data that indicate that “primitive” vessel elements did not
confer a hydraulic advantage relative to tracheids (Sperry et al. 2007). The wider diameters of the vessel elements made them more vulnerable to cavitation than tracheids, so
that initially angiosperms with vessels would likely have been confined to wet habitats
(Feild et al. 2004; Feild & Arens 2005, 2007). Sperry et al. (2007) suggest that diversification of cell types, not the development of vessels per se would have contributed to
the success of the angiosperms. Certainly dicots show considerable diversification in
cell types and arrangements, not only in vessel features, but also in parenchyma cells
involved in storage and production of secondary metabolites and wound response.
Changes in secondary xylem differentiation that led to variation in parenchyma abundance and arrangement must also have been important in contributing to the success
of the angiosperms and their variety in architecture.
Wheeler & Lehman — Cretaceous and Paleocene woods
313
Paleocene Woods – The bulk of the Paleocene woods at Big Bend are referable to
Paraphyllanthoxylon abbotti, mostly from an extensive layer of logs of this species.
The cf. Cunonioxylon wood type described in this paper and a Platanus-like wood (cf.
Platanoxylon haydenii, Wheeler 1991) are very rare in the same deposits; each is represented by only one sample. Similarly, Paraphyllanthoxylon arizonense dominates an
extensive log zone in the Paleocene of New Mexico (Ojo Alamo Sandstone; Wheeler
et al. 1995). There are fewer anatomically described fossil dicot woods known from the
Paleocene than for any other epoch of the Tertiary; in 1993, there were 36 records total
for the entire world (Wheeler & Baas 1993), today there are 71 (InsideWood, accessed
31 December 2008). These low numbers could reflect a dramatic reduction in diversity
of wood types that accompanied the terminal Cretaceous extinction event, or the general
rarity of continental deposits of Paleocene age from varied latitudes worldwide.
Features that change noticeably in incidence from the Campanian-Maastrichtian
to Paleocene are distinct growth rings (increases to 29%), exclusively scalariform
perforation plates (decreases to 37 %), mean tangential diameter >200 µm (increases
to 7%).
CONCLUSIONS
The three new Cretaceous woods from Big Bend have features consistent with patterns
seen in other Cretaceous dicot woods. Sabinoxylon wicki provides additional evidence
indicating that the Late Cretaceous floras of Texas, New Mexico, and northern Mexico
were similar. Cretaceous woods as a whole differ from Recent woods in the incidences of
vessel and parenchyma features. The high incidence of scalariform perforation plates in
Cretaceous woods is especially striking. The high proportions of storage cells in Cretaceous dicot woods raises questions about the functional significance of variations in ray
and axial parenchyma and septate fiber abundance. There remains much to be learned
about Cretaceous angiosperm woods and any additional collections of wood from any
interval of the Cretaceous would be important for understanding their past diversity.
ACKNOWLEDGMENTS
We thank the Science and Resource Management staff of Big Bend National Park, in particular
D. Corrick and V. Davila, for granting collecting permits for our research and for their continued assistance with paleobotanical work in the Park. J. Browning prepared thin-sections of our wood samples.
This work was in part supported by NSF Grants BRC 0237368 and DBI / BDI 0518386 to N.C. State
University for work on databases for fossil and Recent dicot wood. Thanks are also due to the NCSU
libraries administration for their support, and to the interlibrary loan services of the NCSU libraries.
REFERENCES
Badel, E. & P. Perré. 1999. Détermination des propriétés élastiques d’éléments individuels du
plan ligneux du chêne par des essais de traction sur micro-éprouvettes. Ann. For. Sci. 56:
467– 478.
Bailey, I.W. 1924. The problem of identifying the wood of Cretaceous and later dicotyledons.
Paraphyllanthoxylon arizonense. Ann. Bot. 38: 439–451.
314
IAWA Journal, Vol. 30 (3), 2009
Befus, K.S., R.E. Hanson, T.M. Lehman & W.R. Griffin. 2008. Cretaceous basaltic phreatomagmatic volcanism in West Texas: maar complex at Pena Mountain, Big Bend National Park.
J. Volcanology and Geothermal Res. 173: 245–264.
Boura, A. & D. De Franceschi. 2007. Is porous wood structure exclusive of deciduous trees?
C. R. Palevol 6: 385–391.
Burgert, I., A. Bernasconi, K. J. Niklas & D. Eckstein. 2001. The influence of rays on the transverse elastic anisotropy of deciduous trees. Holzforschung 55: 449–454.
Butler, R. J., P.M. Barrett, P. Kenrick & M.G. Penn. 2009. Diversity patterns among herbivorous dinosaurs and plants during the Cretaceous: implications for hypotheses of dinosaur/
angiosperm co-evolution. J. Evolutionary Biology 22: 446–459.
Cahoon, E. J. 1972. Paraphyllanthoxylon alabamense – a new species of fossil dicotyledonous
wood. Amer. J. Bot. 59: 5–11.
Cantrill, D. & I. Poole. 2005. Taxonomic turnover and abundance in Cretaceous to Tertiary wood
floras of Antarctica. Palaeogeography, Palaeoclimatology, Palaeoecology 215: 205–219.
Crane, P. R. 1987. Vegetational consequences of the angiosperm diversification. In: E.M. Friis,
W. G. Chaloner & P. R. Crane (eds.), The origin of angiosperms and their biological consequences: 107–144. Cambridge University Press, Cambridge.
Estrada-Ruiz, E., H. I. Martinez-Cabrera & S.R.S. Cevallos-Ferriz. 2007. Fossil woods from
the late Campanian-early Maastrichtian Olmos Formation, Coahuila, Mexico. Review of
Palaeobotany and Palynology 145: 123–133.
Feild, T.S. & N.C. Arens. 2005. Form, function, and environments of the early angiosperms: merging extant phylogeny and ecophysiology with fossils. New Phytologist 166: 383–408.
Feild, T.S. & N.C. Arens. 2007. The ecophysiology of early angiosperms. Plant, Cell and Environment 30: 291–309.
Feild, T. S., N.C. Arens, J.A. Doyle, T.E. Dawson & M.J. Donoghue. 2004. Dark and disturbed:
a new image of early angiosperm ecology. Paleobiology 30: 82–107.
Gilbert, S.G. 1940. Evolutionary significance of ring porosity. Bot. Gaz. 102 (1): 105–120.
Gottwald, H. 1992. Hölzer aus Marinen Sanden des Oberen Eozän von Helmstedt (Niedersachsen). Palaeontographica B 225: 27–103.
Herendeen, P. S. 1991a. Charcoalified angiosperm wood from the Cretaceous of eastern North
America and Europe. Review of Palaeobotany and Palynology 69: 225–239.
Herendeen, P.S. 1991b. Lauraceous wood from the mid-Cretaceous Potomac Group of eastern
North America: Paraphyllanthoxylon marylandense sp. nov. Review of Palaeobotany and
Palynology 69: 277–290.
Hofmann, E. 1952. Pflanzenreste aus dem Phosphoritvorkommen von Prambachkirchen in Oberösterreich. Palaeontographica 92B: 122–183.
Holbrook, N. M. 1995. Stem water storage. In: B. L. Gartner (ed.), Plant stems. Physiology and
functional morphology: 151–174. Academic Press, San Diego.
Hudson, P. J. 2006. Taxonomic and paleoclimatic significance of Late Cretaceous wood from the
San Juan Basin, New Mexico. Unpub. M.S. Thesis, University of Nebraska Omaha. 82 pp.
Hufford, L. & W.C. Dickison. 1992. A phylogenetic analysis of Cunoniaceae. Systematic Botany
17: 181–200.
IAWA Committee. 1989. IAWA list of microscopic features for hardwood identification. IAWA
Bull. n.s. 10: 219–332.
InsideWood. 2004 – onwards. Published on the Internet. http://insidewood/lib.ncsu.edu/search
Kawamura, Y. 1984. Studies on the properties of rays (II). Mean micelle angles and physical
properties of broad rays. Mokuzai Gakkaishi 30: 201–206.
Lehman, T.M. 1989. Upper Cretaceous (Maastrichtian) paleosols in Trans-Pecos Texas. GSA
Bull. 101: 188–203.
Wheeler & Lehman — Cretaceous and Paleocene woods
315
Lehman, T.M. 1990. Paleosols and the Cretaceous/Tertiary transition in the Big Bend region of
Texas. Geology 18: 362–364.
Lehman, T.M. & A.B. Busbey. 2007. Big Bend Field Trip Guidebook. Society of Vertebrate
Paleontology, 67th Annual Meeting, Field Trip Guidebook. 69 pp.
Lehman, T.M., F. McDowell, & J. Connelly. 2006. First isotopic (U-Pb) age for the Late Cretaceous Alamosaurus vertebrate fauna of West Texas, and its significance as a link between
two faunal provinces. J. Vertebrate Paleontology 26: 922–928.
Lehman, T.M. & E.A. Wheeler. 2001. Fossil dicotyledonous forest from the Upper Cretaceous
of Big Bend National Park, Texas. Palaios 16: 102–108.
Oakley, D. & H.J. Falcon-Lang. 2008. Morphometric analysis of Cretaceous (Cenomanian) angiosperm woods from the Czech Republic. Review of Palaeobotany and Palynology 153:
375–385.
Page, V.M. 1968. Angiosperm wood from the Upper Cretaceous of central California: II. Amer.
J. Bot. 55: 168–172.
Philippe, M., B. Gomez, V. Girard, C. Coiffard, V. Daviero-Gomez, F. Thevenard, J-P. BillonBruyat, M. Guiomar, J-L. Latil, J. Le loeuff, D. Néraudeau, D. Olivero & J. Schlögl. 2008.
Woody or not woody? Evidence for early angiosperm habit from the Early Cretaceous fossil
wood record of Europe. Palaeoworld 17: 142–152.
Poole, I., D.J. Cantrill, P. Hayes & J. E. Francis. 2000a. The fossil record of Cunoniaceae: new evidence from Late Cretaceous wood of Antarctica. Rev. Palaeobot. Palynol. 111: 127–144.
Poole, I., H.G. Richter & J.E. Francis. 2000b. Evidence for Gondwanan origins for Sassafras
(Lauraceae)? Late Cretaceous fossil wood of Antarctica. IAWA J. 21: 463–475.
Rasband, W.S. ImageJ, U. S. National Institutes of Health, Bethesda, Maryland, USA, http://
rsb.info.nih.gov/ij/, 1997–2009.
Rowe, T.R., L. Ciflli, T.M. Lehman & A. Weil. 1992. The Campanian Terlingua local fauna,
with a summary of other vertebrates from the Aguja Formation, Trans-Pecos Texas. J. Vert.
Paleontol. 12: 472–493.
Schönenberger, J. & E.M. Friis. 2001. Fossil flowers of ericalean affinities from the Late Cretaceous of southern Sweden. Amer. J. Bot. 88: 467– 480.
Schönenberger, J., E.M. Friis, M.L. Mathews & P. K. Endress. 2001. Cunoniaceae in the Cretaceous of Europe. Evidence from fossil flowers. Annals of Botany 88: 423–437.
Serlin, B.S. 1982. An Early Cretaceous fossil flora from Northwest Texas: Its composition and
implications. Palaeontographica 182B: 52–86.
Sharp, A. J. 1953. Notes on the flora of Mexico: world distribution of the woody dicotyledonous
families and the origin of the modern vegetation. J. Ecology 41: 374–380.
Spackman, W., Jr. 1948. A dicotyledonous wood found associated with the Idaho Tempskyas.
Ann. Missouri Bot. Gard. 35: 107–116.
Sperry, J.S., U.G. Hacke, T.S. Feild, Y. Sano & E.H. Sikkema. 2007. Hydraulic consequences
of vessel evolution in angiosperms. Int. J. Plant Sci. 168 (8): 1127–1139.
Steel, R. G.D., J.H. Torrie & D. A. Dickey. 1997. Principles and procedures of statistics. A biometrical approach. Ed. 3. McGraw-Hill, New York.
Stevens, P. F. 2001-onwards. Angiosperm Phylogeny Website. Version 9, June 2008.
Stopes, M.C. 1912. Petrifactions of the earliest European angiosperms. Phil. Trans. Roy. Soc.
London B 203: 75–100 + pl. 6–8.
Stopes, M.C. 1915. Catalogues of Mesozoic plants in the Natural History Museum. The Cretaceous flora. Part 2. Lower Greensand (Aptian) plants of Britain. British Museum (Natural
History), London. 360 pp. + 32 pl.
Stopes, M.C. & K. Fujii. 1910. Studies on the structure and affinities of Cretaceous plants. Phil.
Trans. Roy. Soc. London B 201: 1–90 + 9 pl.
316
IAWA Journal, Vol. 30 (3), 2009
Takahashi, K. & M. Suzuki. 2003. Dicotyledonous fossil wood flora and early evolution of wood
characters in the Cretaceous of Hokkaido, Japan. IAWA J. 24: 269–309.
Thayn, G.F., W.D. Tidwell & W. L. Stokes. 1983. Flora of the Lower Cretaceous Cedar Mountain Formation of Utah and Colorado. Part I. Paraphyllanthoxylon utahense. Great Basin
Naturalist 43: 394– 402.
Thayn, G.F., W.D. Tidwell & W. L. Stokes. 1985. Flora of the Lower Cretaceous Cedar Mountain Formation of Utah and Colorado. Part III. Icacinoxylon pittiense n. sp. Amer. J. Bot.
72: 175–180.
Upchurch, G.R., G.H. Mack & E.A. Wheeler. 2003. Late Cretaceous fossil forests from southcentral New Mexico. Abstracts Botany 2003, Mobile, Alabama, p. 67.
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. J. Bot. 78: 658–671.
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.
Wheeler, E.A., P. Baas & S. Rodgers. 2007. Variations in dicot wood anatomy. A global analysis.
IAWA J. 28: 229–258.
Wheeler, E.A., 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. & T.M. Lehman. 2000. Late Cretaceous woody dicots from the Aguja and Javelina
Formations, Big Bend National Park, Texas. IAWA J. 21: 83–120.
Wheeler, E. A. & T. M. Lehman. 2005. Cretaceous-Paleocene conifer woods from Big Bend
National Park, Texas. Palaeogeography, Palaeoclimatology, Palaeoecology 226: 233–258.
Wheeler, E. A., T. M. Lehman & P. Gasson. 1994. Javelinoxylon, a new genus of malvalean tree
from the Upper Cretaceous of Big Bend National Park, Texas. Amer. J. Bot. 81: 703–710.
Wheeler, E.A., J. McClammer & C. A. LaPasha. 1995. Similarities and differences in dicotyledonous woods of the Cretaceous and Paleocene of San Juan Basin, New Mexico, USA.
IAWA J. 16: 223–254.
Wolfe, J. A. 1987. Late Cretaceous-Cenozoic history of deciduousness. Paleobiology 13:
215–226.
APPENDIX I — REFERENCES FOR INCIDENCES OF CRETACEOUS WOODS
Cevallos-Ferriz, S. 1983. Descripción de una madera de angiosperma Cretácica de Cananea,
Son., Méx., los xilitos en el estudio del origen de las Angiospermopsida. Anal. Inst. Biol.
Univ. Nal. Autón. México, Ser. Bótanica 54: 97–112.
Cevallos-Ferriz, S.R. S. & R. Weber. 1992. Dicot from Upper Cretaceous of Coahuila. Revista
UNAM Inst. Geol. 10: 65–70.
Crawley, M. 2001. Angiosperm woods from British Lower Cretaceous and Palaeogene Deposits.
Special Papers in Palaentology 66. The Palaeontological Association. 100 pp.
Gottwald, H. 2000. Gymnosperme und dicotyle Hölzer (67) aus den “Aachener Sanden” der Oberen Kreide von NO-Belgien und NW-Deutschland. Documenta Naturae No. 131: 1–65.
Iamandei, E. & S. Iamandei. 1999. Platanoxylon catenatum Süss & Müller-Stoll 1977 in Bacea
area, Metalliferous Mts. Romania. Revue Roumaine de Géologie 43: 65–67.
Iamandei, E. & S. Iamandei. 2000. Securinegoxylon bacense n.sp. in an Upper MaastrichtianLower Palaeocene Formation from the Bacea area, southern Apuseni (Metalliferous) Mts.
Revue Roumaine de Géologie 44: 57–61.
Wheeler & Lehman — Cretaceous and Paleocene woods
317
Jones, J.H. 1990. An evolutionary and paleoecological analysis of fossil woods from Baja California. Abstract. Third international Senckenberg conference, Frankfurt am Main, Federal
Republic of Germany, May 26, 1990. Senckenberg. Naturforsch. Ges., Frankfurt am Main,
Germany.
Lamb, J.P. 2001. Late Cretaceous (Santonian) vegetation from the Gulf Coast. Abstracts with
Programs - Geological Society of America 33 (2): 16.
Meijer, J. J. F. 2000. Fossil woods from the Late Cretaceous Aachen Formation. Review of Palaeobotany and Palynology 112: 297–336.
Page, V.M. 1967. Angiosperm wood from the Upper Cretaceous of central California. Part I.
Amer. J. Bot. 54: 510–514.
Page, V. M. 1970. Angiosperm wood from the Upper Cretaceous of central California. III. Amer.
J. Bot. 57: 1139–1144.
Page, V. M. 1979. Dicotyledonous wood from the Upper Cretaceous of central California.
J. Arnold Arb. 60: 323–349.
Page, V.M. 1980. Dicotyledonous wood from the Upper Cretaceous of central California. II.
J. Arnold Arb. 61: 723–748.
Page, V. M. 1981. Dicotyledonous wood from the Upper Cretaceous of central California. III.
Conclusions. J. Arnold Arb. 62: 437– 455.
Petrescu, I., C. Ianoliu & I. Dragos. 1978. Prezenta genului Paraphyllanthoxylon (Euphorbiaceae)
in cretacicul din Bazinul Hateg. Contr. Bot., Cluj 227–234.
Prakash, U. & D. Brezinova. 1969 (iss. 1970). Wood of Bridelia from the Cretaceous of Bohemia.
The Palaeobotanist 18: 173–176.
Süss, H. 1960. Ein Monimiaceen-Holz aus der oberen Kreide Deutschlands, Hedycaryoxylon
subaffine (Vater) nov. comb. Senck. Leth. 41: 317–330.
Süss, H. 1986. Untersuchungen über fossile Buchenhölzer. Beiträge zu einer Monographie der
Gattung Fagoxylon Stopes & Fujii. Feddes Repert. 97: 61–183 + pl. XVII–XXII.l.
Süss, H. & W.R. Müller-Stoll. 1977. Untersuchungen über fossile Platanenhölzer. Beiträge zu
einer Monographie der Gattung Platanoxylon Andreanszky. Feddes Repert. 88: 1–62 + pl.
1–14.
Suzuki, M., S. Noshiro & H. Tobe. 1996. Wood structure of Japanese Mesozoic dicotyledons. In:
L.A. Donaldson, A.P. Singh, B. G. Butterfield & L. J. Whitehouse (eds.), Recent advances
in wood anatomy: 150–158. New Zealand Forest Research Institute.
Suzuki, M. & H. Ohba. 1991. A revision of fossil wood of Quercus and its allies in Japan. J. Jap.
Bot. 66: 255–274.
APPENDIX II — REFERENCES FOR INCIDENCES OF PALEOCENE WOODS
Blokhina, N.I. & S.A. Snezhkova. 1999. Fossil wood Alnus tschemrylica sp. nov. (the Betulaceae)
from the Paleogene of Tschermunaut Bay (Kamchatka). Palaeontological J. 33: 466–470.
Boureau, E. 1950. Étude paléoxylogique du Sahara (IX): Sur un Myristicoxylon princeps n. gen.,
n. sp., du Danien d’Asselar (Sahara soudanais). Bull. Mus. Natl. Hist. Nat. 2e Sér. 22:
523–528.
Crawley, M. 1988. Palaeocene wood from the Republic of Mali. Bull. Brit. Mus. Nat. Hist.
(Geol.) 44: 3–14.
Crawley, M. 1989. Dicotyledonous wood from the lower Tertiary of Britain. Palaeontology 32:
597–622.
Crawley, M. 2001. Angiosperm woods from British Lower Cretaceous and Palaeogene deposits.
Special Papers in Palaentology 66. The Palaeontological Association. 100 pp.
318
IAWA Journal, Vol. 30 (3), 2009
Iamandei, E. & S. Iamandei. 1999. Platanoxylon catenatum Süss & Müller-Stoll 1977 in Bacea
area, Metalliferous Mts. Romania. Revue Roumaine de Géologie 43: 65–67.
Iamandei, E. & S. Iamandei. 2000. Securinegoxylon bacense n.sp. in an Upper MaastrichtianLower Palaeocene Formation from the Bacea area, southern Apuseni (Metalliferous) Mts.
Revue Roumaine de Géologie 44: 57–61.
Koeniguer, J. C. 1969 (iss. 1970). Sur un bois fossile du Paléocéne du Niger. C. R. 94e Congr.
Natl. Soc. Sav., Pau, Sci., 3: 157–173.
Koeniguer, J.C. 1971. Sur les bois fossiles du Paléocéne de Sessao (Niger). Rev. Palaeobot.
Palynol. 12: 303–323.
Melchior, R.C. 1998. Paleobotany of the Williamsburg Formation (Paleocene) at the Santee
Rediversion site Berkeley County, South Carolina. In: A.E. Sanders (ed.), Paleobiology of
the Williamsburg Formation (Black Mingo Group; Paleocene) of South Carolina, U.S.A.:
49–121. Transactions of the American Philosophical Society, Vol. 4, Pt. 4.
Wheeler, E.A. & T.M. Michalski. 2003. Paleocene and Eocene woods of the Denver Basin,
Colorado. Rocky Mountain Geology 38: 29– 43.