Systematics and biogeography of selected modern and fossil Dipteronia and Acer (Sapindaceae)

Material Information

Systematics and biogeography of selected modern and fossil Dipteronia and Acer (Sapindaceae)
McClain, Amy Marie, 1976- ( Dissertant )
Manchester, Steven R. ( Thesis advisor )
Dilcher David L. ( Reviewer )
Judd, Walter S. ( Reviewer )
Place of Publication:
State University System of Florida
Publication Date:
Copyright Date:


Subjects / Keywords:
Botany ( jstor )
Figs ( jstor )
Flora ( jstor )
Fossils ( jstor )
Fruits ( jstor )
Leaf blade ( jstor )
Leaves ( jstor )
Pericarp ( jstor )
Species ( jstor )
Taxa ( jstor )
Botany thesis, M.S ( lcsh )
Dissertations, Academic -- Botany -- UF ( lcsh )
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
non-fiction ( marcgt )


Dipteronia and its sister genus Acer form the traditional family Aceraceae, now interpreted as a subgroup of the Sapindaceae. These two genera have fossil records that extend back to the Paleocene or Eocene. In this thesis, I evaluate the fossil records of Dipteronia based on fruits and those of Acer section Macrantha based on leaves. Although confined to China today, the record of Dipteronia fruits ranges from the Paleocene to the Oligocene in western North America. Leaves of Acer section Macrantha have a similar pattern of distribution, with fossils from both North America and Asia. Acer section Macrantha has a more limited fossil record than reported in the literature, but can be confirmed from the Eocene of Alaska and the Miocene and Pliocene of Japan.
Thesis (M.S.)--University of Florida, 2000.
Includes bibliographical references (p. 78-82).
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
General Note:
Title from first page of PDF file.
General Note:
Document formatted into pages; contains vii, 83 p.; also contains graphics.
General Note:
Statement of Responsibility:
by Amy Marie McClain.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Amy Marie McClain. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
47682127 ( OCLC )
002678733 ( AlephBibNum )
ANE5960 ( NOTIS )


This item has the following downloads:

Full Text







Copyright 2000




I would like to thank the many people who have helped me throughout the last

few years. My committee chair, Steven R. Manchester, provided continual support and

assistance in helping me become a better researcher. The members of my committee,

David L. Dilcher and Walter S. Judd, have spent much time and effort teaching me in

their areas of expertise. The University of Florida Herbarium (FLAS) staff, including

Kent Perkins and Trudy Lindler, were of great assistance. I also thank the Harvard

Herbarium (A, GH) staff, especially Emily Wood, David Boufford, Kancheepuram

Gandhi, and Timothy Whitfeld, as well as those at the Beijing Herbarium (PE) and

Zhiduan Chen, who helped to arrange my visit to China. I thank David Jarzen for help

with the University of Florida fossil plant collections. I appreciate the access to fossil

specimens provided to Steven Manchester and me by Amanda Ash, Melvin Ashwill,

James Basinger, Lisa Barksdale, Richard Dillhoff, Thomas Dillhoff, Diane Erwin, Leo

Hickey, Kirk Johnson, Linda Klise, Wesley Wehr, and Scott Wing. Thanks go to

Richard and Thomas Dillhoff for providing measurements of additional fossil specimens.

I especially thank my husband, Rob McClain, for his patience, help, and support, and my

parents for their love and encouragement. This work was funded in part by a research

assistantship from the Florida Museum of Natural History.



A C K N O W L E D G M E N T S ......... ................... ...............................................................iii

L IST O F T A B L E S .... ....... ...................................................... .... ........ .. .......... .. v i

A B STR A CT ............ ................... ...................................... .... ......... vii

INTRODUCTION .............. ................................... ..............

D ip teronia ............... ..... ............................................................. 3
A c e r ................... ................... ...................4..........

M ATERIALS AND M ETHODS ............................................................. .................7

D ip teronia ............... ..... ............................................................. 7
A c e r ................... ................... ...................1.........6

D IP T E R O N IA ................................ ......... .. .... ....................... .......................................... 2 6

M orphology....................... ............... ...... .............. 26
M o d e rn ............. ......... .. .............. .. ........................................................ 2 6
F o s sil ............. .......... .. .............. .. ............................................. 2 7
Previous N om enclature .......................................................... ......... ...... .............. 28
Dipteronia brownii, McClain et Manchester, sp. nov....... ................................... 30
Discussion of Dipteronia brownii sp. nov. ........................................ 31
B io g e o g rap h y ............................................ ...... ............................................... 3 4
P h y lo g e n y ................................................................... 3 5

A C E R ................... ..................................................................... 4 3

M orphology....................... ............... ...... .............. 43
Modern ............... .............................................. 43
F o s s il ......................................................................................................... 4 9
B io g e o g ra p h y ................................................................................................... 5 1
P h y lo g en y ................................................................... 5 2

C O N C L U S IO N S ....................................................................................................7 6

L IST O F R E FE R E N C E S ........................................................................ ......................78

BIOGRAPHICAL SKETCH .............................................................. ..................83



Table 1: Localities for Dipteronia fossil fruits in the Northern Hemisphere, giving the
estimated geologic ages, sources of age estimates, geographic coordinates, and
exam ple specim ens cited. ............................................ ............................. 10

Table 2: Species included in Acer section Macrantha by various authors...........................19

Table 3: Acer sections and series as circumscribed by de Jong, 1994 ................................20

Table 4: Acer Section Macrantha as circumscribed by de Jong, 1994, with modern range...23

Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science



Amy Marie McClain

December 2000
Chairman: Steven R. Manchester
Major Department: Botany

Dipteronia and its sister genus Acer form the traditional family Aceraceae, now

interpreted as a subgroup of the Sapindaceae. These two genera have fossil records that

extend back to the Paleocene or Eocene. In this thesis, I evaluate the fossil records of

Dipteronia based on fruits and those of Acer section Macrantha based on leaves.

Although confined to China today, the record of Dipteronia fruits ranges from the

Paleocene to the Oligocene in western North America. Leaves of Acer section

Macrantha have a similar pattern of distribution, with fossils from both North America

and Asia. Acer section Macrantha has a more limited fossil record than reported in the

literature, but can be confirmed from the Eocene of Alaska and the Miocene and Pliocene

of Japan.


Maple trees are common elements in forests across the Northern Hemisphere.

Dipteronia and Acer are both well represented in the fossil record with occurrences in the

Tertiary of North America and Asia. Acer is additionally well represented in the

European Tertiary. Although the two genera have long been considered to be sister taxa,

they have had quite different histories since their divergence. The fossil record of both

consists primarily of fruits and leaves, but Dipteronia and Acer present different

challenges to understanding their fossil record. Dipteronia has distinctive fruits, but the

compound leaflets are difficult to distinguish from various other sapindalean genera.

Previous reports of Dipteronia fossils have included both fruits and equivocal foliage,

which confuses the history. This study was designed to incorporate only those reports

where I am confident that the genus has been identified correctly. Acer is easily

recognized in the fossil record from both leaves and fruits, although the fruits are similar

to some other Sapindaceae and to Malpighiaceae. The main difficulty within the genus

has been determining the infrageneric relationships of the fossils. Because there are over

one hundred modern species, monographers have separated the genus into sections, many

of which have been consistently recognized by various authors over time. However, the

specific leaf characters for these sections have been described only in vague terms. Other

characters such as flowers (de Jong, 1976) or bud scales and wood anatomy (Ogata,

1967) have often been given more attention.

This study comprises two parts. The first is an analysis of extant and fossil

Dipteronia fruits, designed to determine the historical biogeography of the genus as

represented by the distinctive fruits. The second examines the genus Acer, which is

easily identified in the fossil record from both leaves and fruits. Only the leaf

morphology of Acer is studied here, but future work will include fruits. To further

understand the biogeography of Acer, it is necessary not only to recognize the genus, but

also to assess the infrageneric relationships of the fossils. Although previous researchers

(Pax, 1902; Akhmetiev, 1971; Walther, 1972; Tanai and Ozaki, 1977; Wolfe and Tanai,

1987; Ozaki, 1991) have often placed fossil taxa in fossil and modem sections of the

genus, the characters upon which such assignments are based have often not been truly

diagnostic and the fossils, as described, sometimes also match in morphology more than

one section.

Dipteronia and Acer have long been considered sister taxa, as indicated by their

traditional placement in the bigeneric family Aceraceae (Pax, 1902; Pojarkova, 1949).

Both share the characters of opposite, exstipulate leaves, flowers with five sepals, five

nonappendaged petals, a nectariferous disc, eight stamens, tricolporate pollen, two

carpels and two ovules per carpel with only one seed developing per carpel and

schizocarpic winged fruits borne in pairs. Recent phylogenetic investigations continue to

support the close phylogenetic relationship between Dipteronia and Acer (Judd, Sanders,

and Donoghue, 1994; Gadek et al., 1996). However, these analyses also indicate that

these taxa are nested within the Sapindaceae; therefore the traditional recognition of

Aceraceae at the family level renders the Sapindaceae paraphyletic. It is more

appropriate to treat the Dipteronia-Acer clade as a subfamily (Aceroideae) or lower rank

within the Sapindaceae.

The two genera are readily distinguished on morphological grounds. Dipteronia

has distinctive fruits in which the wing fully encircles the seed. It also differs from Acer

in lack of bud scales and much larger inflorescences of over 400 flowers (Ogata, 1967; de

Jong, 1976; van Gelderen, de Jong, and Oterdoom, 1994). The pinnately compound

leaves of Dipteronia differ from the leaves of most Acer species, which are typically

simple and palmately lobed. Acer may also be palmately compound, trifoliolate or, as in

Acer negundo, pinnately compound with only 5-7 leaflets instead of the 7-15 common in


Dipteronia and Acer are both well represented in the fossil record with

occurrences of Dipteronia in the Tertiary of North America and Acer throughout the

Northern Hemisphere. Although the fossil record of Acer is well documented in the

Tertiary of North America (Brown, 1935, 1937; Wolfe and Tanai, 1987), Europe

(Walther, 1972), and Asia (Akhmetiev, 1971; Tanai and Ozaki, 1977; Manchester, 1999),

that of Dipteronia has received relatively little attention and some of the previously

published reports refer to isolated leaflet impressions of equivocal diagnostic value

(Brown, 1935, 1937; MacGinitie, 1974). The fossil record of Dipteronia may

complement that of Acer, providing a better understanding of the geographic origins and

timing of the radiation of the Acer-Dipteronia clade.


Dipteronia Oliver (1889) is an extant genus endemic to southern and central

China with two living species, D. sinensis Oliver (1889) and D. dyerana Henry (1903).

The genus is distributed in broadleaved deciduous forests, along stream margins at

altitudes from 1450 to 2400 m, sometimes as low as 1000 m, in the Chinese provinces

Gansu, Guizhou, Henan, Hubei, Hunan, Shaanxi, Sichuan, and Yunnan (Ying, Zhang,

and Boufford, 1993). In the Dipteronia chapter of this thesis, I review the diagnostic

characters of the fruits of the two extant species as a basis for recognizing fossil remains

from western North America. I attempt to unravel the complicated nomenclature of

North American fossil Dipteronia and Bohlenia. Although fossil fruits of Dipteronia

have been illustrated from the Tertiary of North America, they were always attributed to

species that were based on detached leaves of uncertain affinity to Dipteronia (Brown,

1935, 1937). I recognize a new species, Dipteronia brownii sp. nov., based on

characteristic fruit remains, review its stratigraphic and geographic distribution and

compare its morphology with that of extant species. The significance of fossil remains as

a basis to understand the evolutionary and biogeographic history of the Acer-Dipteronia

clade is discussed.


Acer L. is distributed throughout the Northern Hemisphere, in a variety of

environments. The genus is usually deciduous, but some species in warmer areas are

evergreen. There are at least 120 modern species, more according to some taxonomists.

The first comprehensive monograph of the genus was that of Pax (1885, 1886, 1902).

Systematic treatments of Acer have also been published by Pojarkova (1949), Momotani

(1962), Fang (1966), Ogata (1967), Murray (1970), de Jong (1976), Delendick (1981),

and Mai (1984).

Acer fossils are found throughout the Northern Hemisphere from the Eocene to

the Recent (Walther, 1972; Tanai and Ozaki, 1977; Wolfe and Tanai, 1987), and perhaps

as early as the Paleocene (Brown, 1962; Wolfe and Tanai, 1987; Crane, Manchester, and

Dilcher, 1990). The most complete summary of the fossil record of Acer in North

America is the 1987 paper by Wolfe and Tanai. They described and illustrated 21

previously established Acer species from North America as well as 62 new species, two

new combinations, and six undetermined species, based on leaf and fruit fossils. They

placed these fossils in 21 modem and 12 fossil sections. However, I believe important

work still remains. There has not been sufficient analysis as to the reliability of the

infrageneric assignments of many of these species. The unique character combinations

that define each modem and fossil section have yet to be described. Another problem is

that detached fruits and leaves hypothesized to represent the same species have often

been formally merged and given the same name, which creates difficulties for

paleobotanists. By analyzing a single, easily identifiable section ofAcer, I hope to

provide a more reliable basis for determining whether or not fossil Acer leaves belong to

that section.

I surveyed the modem and fossil Acer literature to determine which modem

sections may be reliably diagnosed based on leaf morphology. I determined that some

species in the section Macrantha had a distinctive leaf type not found in any other

modem section. I surveyed the circumscription of the section Macrantha by the major

monographers of Acer in order to determine which modem species to study. Fruit

morphology also may be used in the future, as fruits are also common Acer fossils.

The section Macrantha was first described by Pax in 1885. Most monographers

have considered the type species to be A. pensylvanicum L., but Delendick (1981)

believes that the first species to be identified as the type following the ICBN may actually

be A. rufinerve Sieb. et Zucc. as published by Fang (1966). I have not yet studied the

nomenclature in enough detail to evaluate this and other nomenclatural problems. In this

thesis I will generally follow de Jong (1994), the most recent classification, which

appears to work fairly well within the section Macrantha. This section has been

recognized by all the monographs since it was first described (Pax, 1885). The group

seems very likely to be monophyletic, based on putatively synapomorphic morphogical

characters such as striped bark, stalked bud scales (also present in section Glabra, series

Glabra), and poorly developed inflorescence bracts. Recent molecular analyses of Acer

show support for the monophyly of the section Macrantha (Hasebe, Ando, and Iwatsuki,

1998), or for at least most of the taxa sampled (Ackerly and Donoghue, 1998). However,

a comprehensive molecular analysis of the genus Acer has not been attempted.

Wolfe and Tanai (1987) include five fossil species in section Macrantha based on

leaves and fruits from North America: A. clarnoense Wolfe et Tanai, A. dettermani Wolfe

et Tanai, A. castorrivularis Wolfe et Tanai, A. latahense Wolfe et Tanai, and A.

palaeorufinerve Tanai et Onoe. To determine if these species were appropriately

assigned to section Macrantha, I surveyed the morphology of modem section Macrantha

leaves in order to determine which characters available in the fossil record could be used

to identify the group.



The winged fruits of Dipteronia occur as impressions in lacustrine shales at

localities where fossil dicotyledonous leaves are abundant. Specimens are usually

recovered in the process of splitting the sediment in search of leaf impressions.

Dipteronia fruits were examined from all known North American localities, from

Colorado to British Columbia (Table 1, Fig. 1). Specimens cited in this paper are housed

at the Florida Museum of Natural History, Gainesville, (UF), Burke Museum in Seattle,

Washington (UWBM), Smithsonian Institution, Washington DC (USNM), Peabody

Museum of Natural History, New Haven, Connecticut (YPM), the University of

California at Berkeley (UCMP), and the University of Saskatoon, Saskatchewan

(USASK). The literature and major museum collections in Europe and Asia were studied

by S. Manchester (cited in Manchester, 1999) in search of Dipteronia fruits.

The geographic distribution of extant and fossil occurrences of Dipteronia (Fig. 1)

were plotted using the "Online Map Creation" web page

( intro.html) application of Generic Mapping

Tools by entering the latitudes and longitudes of individual sites. Table 1 provides the

geographic coordinates and stratigraphic position of localities from which Dipteronia

fossil fruits have been studied and indicates the references upon which the geologic age is

based. In some instances the age determinations are based on stratigraphic proximity to

radiometrically dated units; in other instances the age is simply an approximation based

on floristic correlations (Table 1; localities 1, 9-11, and 13). The earliest known

Dipteronia fruits are two fragmentary specimens collected by Manchester from the

Paleocene Fort Union Formation at Hell's Half Acre, Wyoming. These megafossils were

collected from the same stratigraphic sequence used by Nichols and Ott (1978) as the

basis for establishing a palynological zonation that is now widely used in biostratigraphic

investigations of Paleocene sediments in Wyoming and adjacent states. The Dipteronia

fruits were found at the level of pollen zone P-4 sensu Nichols and Ott (1978), which is

considered to be middle Paleocene (Nichols and Flores, 1993), or 60-63 million years


Middle Eocene localities include West Branch Creek and White Cliffs Sr.,

Oregon; Republic, Washington; Joseph Creek, McAbee, and One Mile Creek, British

Columbia (Table 1: localities 2-7). There are many Late Eocene localities, including

Sumner Spring, Teater Road, White Rock Gulch, and White Cap Knoll in Oregon; Ruby

River Basin, Montana; and Florissant, Colorado (Table 1: localities 8-13). The youngest

specimen is from the early Oligocene shales of Bridge Creek, Oregon (Brown, 1959;

Meyer and Manchester, 1997).

Modem comparative material was examined from herbaria at A, FLAS, and PE.

Fruit length is the longest line parallel to the primary vein leading to the seed. Fruit

width is the longest line perpendicular to the length (Fig. 2). Measurements cited in the

text are for the largest measurement of each fruit, as the fruits are usually very similar in

each dimension.

Fossil Dipteronia fruits were measured to determine whether they could be

divided into two or more groups based on size. There were no gaps along width or length

for the fossils (Fig. 3). Length and width appear to be highly correlated. The modem

fruits are generally larger than most fossils. When fossils were incomplete, size measures

are conservative estimates. Because the wing is fairly symmetrical on either side of the

seed body, these estimates should be reasonably accurate. Because the ratio of length vs.

width was highly correlated, I used the maximum dimension to represent the overall size

of the fruits. I graphed maximum dimension over time (Fig. 4), but because of limited

stratigraphic information, the time axis is poorly resolved.

Table 1: Localities for Dipteronia fossil fruits in the Northern Hemisphere, giving the estimated geologic ages, sources of age
estimates, geographic coordinates, and example specimens cited.
Locality Age Locality age assignment Latitude/ Longitude Illustrations or cited specimens
1. Hell's Half Acre, Late Nichols and Ott, 1978 43000.88'N Fig. 24
Wyoming Paleocene 107004.84'W
2. McAbee, Mid Stockey and Wehr, 1996 50047'N Fig. 33
British Columbia Eocene 121008'W
3. Joseph Creek, Mid Wolfe and Wehr, 1987 -50oN Fig. 26
British Columbia Eocene -120W
4. One Mile Creek, Princeton, Mid Stockey and Wehr, 1996 49021' N Figs. 18, 19, 31, 32;
British Columbia Eocene -120019.98'W Stockey and Wehr, 1996
5. Republic, Mid Wolfe and Wehr, 1987 48039.10'N Figs. 13, 16, 25, 34; Brown, 1935;
Washington Eocene 118044.33'W Wolfe and Wehr, 1987, as
Bohlenia; Wehr, 1995, as Bohlenia
6. White Cliffs Sr., Mid Manchester, 1990 44044.302'N Fig. 29
Oregon Eocene 120028.376'W
7. West Branch Mid Manchester, 1990 4435.44'N Fig. 35
Creek, Oregon Eocene 120015.51'W
8. Whitecap Knoll, Late Manchester, 2000 44050.195'N Manchester, 2000
Oregon Eocene 120025.074'W
9. White Rock Late Manchester estimate 44043.245'N Fig. 27
Gulch, Oregon Eocene 120028.443'W
10. Sumner Spring, Late Smith et al., 1998 44025.833'N Figs. 20, 36, 37;
Oregon Eocene 121007.018'W McFadden, 1986
11. Teater Road, Late Wolfe and Tanai, 1987 4410.039'N Fig. 28
Oregon Eocene as Sheep Rock Creek 12014.856'W

Table 1-continued.
Locality Age Locality age assignment Latitude/ Longitude Illustrations or cited specimens
12. Ruby River Late Wolfe and Wehr, 1987 4513.13'N Becker, 1960, 1961;
Basin, Montana Eocene 11212.50'W Manchester, 1999
13. Florissant, Late Wolfe and Wehr, 1987 380 55'N Brown, 1937;
Colorado Eocene 105016'W MacGinitie, 1953
14. Bridge Creek, Early Meyer and Manchester, 1997; 4438.48'N Brown, 1959; Meyer and
Oregon (Painted Hills) Oligocene Bestland et al., 1999 12017.25'W Manchester, 1997

90 120' 150" 180" -150" -120' -90" -60'

Oligocene ofPainted Hills Oregon. All other sites are middle to late Eocene in age.

_____ _____ I _________ ___ 20
s ~;B"d.~:v


Oligocene of Painted Hills, Oregon. All other sites are middle to late Eocene in age.




Figure 2. Photograph of extant mericarp of Dipteronia sinensis showing the orientations
used for measuring length and width. The length (L) is measured parallel to the primary
vein (B) that diverges from the detachment scar (A). The width is measured
perpendicular to the axis of length (A: B. Bartholomew et al., 1063, Hubei).



50 A


3 30


20 -

10 -

0 -
0 10 20 30 40 50 60
Fruit length (mm)

D. brownii D. sinensis A D. dyerana

Figure 3. Modem and fossil Dipteronia fruits length vs. width.


25 -


1 *
E 15o

E 10-



Mid Eocene

Late Eocene


Hell's half acre U White cliffs Sr. A Republic
Princeton Sumner springs + Teater road
White cap knoll U Ruby A Florissant

x McAbee
* White rock
x Bridge Creek

Figure 4. Maximum dimension of fossil Dipteronia mericarps across time.

A Al
A "

a x ~+

A *


As the majority of Acer leaf fossils reported in the literature are palmately lobed, I

surveyed approximately 60 extant species in the genus that normally produce leaves of

this kind, using herbarium collections at Harvard (A, GH), Beijing (PE), and the

University of Florida (FLAS). I analyzed major morphological characters such as

number of primary veins, number of lobes, characters of secondary veins and marginal

condition to identify sections of Acer with similar leaves. Based on this preliminary

analysis, I determined that Acer section Macrantha could be identified solely on leaf

characters. I then surveyed the literature to compare howAcer section Macrantha has

been characterized by various monographers and workers on the genus, including Pax

(1885, 1886, 1902), Pojarkova (1949), Fang (1939, 1966), Ogata (1965, 1967), Murray

(1970), de Jong (1976, 1994), Delendick (1981, 1990), and Mai (1984). Table 2

summarizes the taxa included in Acer section Macrantha section by various authors,

accounting for nomenclatural changes. The modern distribution of each taxon is also


In order to determine the synapomorphies for Acer section Macrantha and to

establish criteria for the recognition of fossil representatives, variation in the section was

studied using modern leaves. The classification I used in this thesis basically follows that

of de Jong (1994). Table 3 lists the species recognized by de Jong (1994) and the

sections and series in which he places them. Further studies will attempt to clarify the

nomenclature, but here I provide all of the names currently used and some of the

historical names. Many herbaria still catalogue the species according to names used by

Fang (1939) or older monographs, so it is important to know under which names to look.

Table 4 lists the species studied from Acer section Macrantha, as well as some


Additionally, the following species were studied; these have leaves somewhat

similar to those in section Macrantha: Acer nipponicum (section Parviflora, series

Parviflora), A. spicatum and A. caudatum ssp. ukurunduense (section Parviflora, series

Caudata), A. argutum (section Glabra, series Arguta), and A. wardii (section Wardiana).

I analyzed leaves of these species using the general leaf characters described by

Hickey (1973), Wolfe and Tanai (1987), and the Leaf Architecture Working Group

(1999). For those characters specific to palmately lobed leaves, I am following the

terminology of Wolfe and Tanai (1987), which I will clarify here.

The thickest vein of the lamina continuous with the petiole of the leaf is the

medial primary (Fig. 5A). Other primary veins that arise at an angle to the medial

primary are termed lateral primaries (Fig. 5B). Secondary veins that branch from the

medial primary are termed medial secondaries (Fig. 5C) and those that branch from the

lateral primaries are lateral secondaries (Fig. 5D).

The sinus between the lobes is termed the lobal sinus. The origin of the veins that

surround the sinus on either side (lobal sinus bracing) is a character stressed by Wolfe

and Tanai (1987). This character is not emphasized here due to variability within species.

Tooth shape terminology follows the Leaf Architecture Working Group (1999).

The upper and lower shapes of each tooth are determined to be convex (CV), straight

(ST), concave (CC), flexuous (FL), or retroflexed (RT). Flexuous is defined as basally

convex and apically concave and retroflexed is the inverse. The abbreviations are written

with the apical shape first. For example, CC/CV would mean that the apical side of the

tooth was concave and the basal side was convex.

The Leaf Architecture Working Group (1999) describes lobes as indentations

along the margin that reach at least 1 /of the distance to the midvein, measured parallel to

the main vein of the lobe, or the axis of symmetry of the lobe. I have modified this

definition to better describe the degree oflobing. Shallow lobes are indentations up to

1/3 of the distance along a lateral primary to the medial primary. Medium lobing varies

from 1/3 to 2/3 the length and deeply lobed leaves have indentations greater than 2/3 the

length of the lateral primary.

The term location is used, as in Wolfe and Tanai (1987), to describe a very large

tooth or group of teeth that subdivides a major lobe of the leaf.

Higher order venation, as studied by Tanai (1978), is not stressed in this study,

although it may prove useful for differentiating other species of Acer.

Epidermal anatomy may be useful for some sections (Walther, 1972). However,

because fossils with cuticle are most often studied in Europe and Acer section Macrantha

does not appear in the modern European flora or the European fossil record, section

Macrantha has not been studied for cuticular characters. Further work may be useful for

resolving relationships among extant species, although the majority of Acer leaf fossils in

North America do not have cuticle preserved.

Table 2: Species included in Acer section Macrantha by various authors.
Pax Fang Ogata Murray Mai Delen- de
1902 1966 1967 1970 1984 dick Jong
1981 1994
A. pensylvanicum X X X X X X X
A. capillipes X X X X X X X
A. rufinerve X X X X X X X
A. tegmentosum X X X X X X X
A. caudatifolium --1 X XX X X11 X
A. crataegifolium X X X X X X
A. pectinatum X2 X4 X/ X' X10 X1 X
A.rubescens X X X X X
A. sikkimense X X X X X X
A. wardii X X X X
A. micranthum X X X X X X X
A. tschonoskii X X X X X X X
A. davidii X5 X5 X X5 X5 X
A. morifolium -- X X X X
A. metcalfii -- X X X X
A. laisuense -- X X

1 Species described after publication
2 As A. maximowiczii Pax and A. pectinatum
3 As A. kawakamii
4 As A. forrestii, A. laxiflorum, A. maximowiczii Pax, and A. taronense
5 As A. davidii and A. grosseri
6 As A. insulare and A. kawakamii
7 As A. chienii, A. laxiflorum, A. maximowiczii, and A. pectinatum
8 As A. hookeri and A. sikkimense
9 As A. laxiflorum, A. maximowiczii Pax, A. pectinatum, and A. taronense
10 As A. laxiflorum, A. maximowiczii Pax, and A. pectinatum
11 As A. caudatifolium and A. morrisonense
12 As A. forrestii, A. laxiflorum, A. maximowiczii Pax, A. pectinatum

Table 3: Acer sections and series as circumscribed by de Jong, 1994.
Section Parviflora
Series Parviflora
A. nipponicum
Series Distyla
A. distylum
Series Caudata
A. caudatum A. spicatum

Section Palmata
Series Palmata
A. ceriferum
A. duplicatoserratum
A. palmatum
A. pseudosieboldianum
A. robustum
A. sieboldianum
Series Sinensia
A. calcaratum
A. chapaense
A. elegantulum
A. fenzelianum
A. kweilinense
A. linganense
A. miaoshanicum
A. oliverianum
A. shangszeense
A. sunyiense
A. tonkinense
A. wuyuanense
Series Penninervia
A. cordatum
A. erythrantum
A. fabri
A. kiukiangense
A. lucidum
A. sino-oblongum

A. circinatum
A. japonicum
A. pauciflorum
A. pubipalmatum
A. shirasawanum

A. campbellii
A. confertifolium
A. erianthum
A. kuomeii
A. lanpingense
A. mapienense
A. olivaceum
A. schneiderianum
A. sichourense
A. taipuense
A. tutcheri
A. yaoshanicum

A. crassum
A. eucalyptoides
A. hainanense
A. laevigatum
A. oligocarpum
A. yinkunii

Section Wardiana
A. wardii

Table 3-continued.
Section Macrantha
A. capillipes
A. crataegifolium
A. laisuense
A. morifolium
A. pensylvanicum
A. rufinerve
A. tegmentosum
Section Glabra
Series Glabra
A. glabrum
Series Arguta
A. acuminatum
A. barbinerve

A. caudatifolium
A. davidii
A. micranthum
A. pectinatum
A. rubescens
A. sikkimense
A. tschonoskii

A. argutum
A. stachyophyllum

Section Negundo
Series Negundo
A. negundo
Series Cissifolia
A. cissifolium A. henry
Section Indivisa
A. carpinifolium

Section Acer
Series Acer
A. caesium
A. pseudoplatanus
Series Monspessulana
A. hyracanum
A. obtusifolium
A. sempervirens
Series Sacharodendrom
A. saccharum
Section Pentaphylla
Series Pentaphylla
A. pentaphyllum
Series Trifida
A. buergerianum
A. discolor
A. oblongum
A. shihweii
A. wangchii

A. heldreichii
A. velutinum

A. monspessulanum
A. opalus

A. coriaceifolium
A. fengii
A. paxii
A. sycopseoides
A. yuii

Table 3-continued.
Section Trifoliata
Series Grisea
A. griseum
A. triflorum
Series Mandshurica
A. mandshuricum
Section Lithocarpa
Series Lithocarpa
A. diabolicum
A. sinopurpurascens
Series Macrophylla
A. macrophyllum
Section Platanoidea
A. campestre
A. longipes
A. mono
A. platanoides
A. tibetense
Section Pubescentia
A. pentapomicum
Section Ginnala
A. tataricum
Section Rubra
A. pycnanthum
A. saccharinum
Section Hyptiocarpa
A. garrettii

A. maximowiczianum

A. sutchuenense

A. leipoense
A. sterculiaceum

A. cappadocicum
A. miyabei
A. nayongense
A. tenellum
A. truncatum

A. pilosum

A. rubrum

A. laurinum

Table 4: Acer Section Macrantha as circumscribed by de Jong, 1994, with modem range.
A. capillipes Maximowicz (1867) Japan
Synonym A. pensylvanicum ssp. capillipes (Maximowicz) Wesmael (1890)
A. caudatifolium Hayata (1911) Taiwan
Synonyms A. insulare Makino (1910)
A. kawakamii Koidzumi (1911)
A. morrisonenese Hayata (1911)
A. ovatifolium Koidzumi (1911)
A. pectinatum ssp. formosanum Murray (1977)
A. crataegifolium Siebold et Zuccarini 1845 Japan
A. davidii Franchet (1885) ssp. davidii China
Synonym A. cavaleriei Leveille (1912)
A. davidii ssp. grosseri (Pax) de Jong (1994) (not studied) China
Basionym A. grosseri Pax (1902)
Synonyms A. pavolinii Pampanini (1910)
A. hersii Rehder (1922)
A. tegmentosum ssp grosseri (Pax) Murray (1966)
A. laisuense Fang et Hu (1966) (not studied) China
A. metcalfi Rehder 1933 [A. sikkimense ssp. metcalfi (Rehder) de Jong (1994)] China
Synonym A. kiangsiense Fang et Fang f. (1966)
A. micranthum Siebold et Zuccarini (1845) Japan
A. morifolium Koidzumi (1914) Japan
Synonyms A. insulare auct. non Makino (1910)
A. caudatum sensu Ito et Matsumura (1920) non Wallich
A. capillipes ssp. insulare (Makino) Murray (1977)
A. pectinatum Wallich ex Nicholson (1881) ssp. pectinatum China
A. pectinatum ssp. forrestii (Diels) Murray (1977) China
Basionym A.forrestii Diels (1912)
Synonym A. laxiflorum sensu Rehder (1911) non Pax
A. pectinatum ssp. laxiflorum (Pax) Murray (1977) China
Basionym A. laxiflorum Pax (1902)
A. pectinatum ssp. maximowiczii (Pax) Murray (1977) China
Basionym A. maximowiczii Pax (1889)
Synonym A. urophyllum Maximowicz (1889)
A. pectinatum ssp. taronense (Handel-Mazzetti) Murray (1977) China
Basionym A. taronense Handel-Mazzetti (1924)
Synonym A. chienii Hu et Cheng (1948)
A. chloranthum Merrill (1941)
A. pensylvanicum L. (1753) Eastern North America
Synonyms A. striatum Duroi (1771)
A. canadense Marshall (1758)
A. tricuspifolium Stokes (1812)

Table 4-continued.
A. rubescens Hayata (1911) Taiwan
Synonym A. morrisonense sensu Li (1963), non Hayata
A. rufinerve Siebold et Zuccarini (1845) Japan
Synonyms A. pensylvanicum ssp. rufinerve (Siebold et Zuccarini) Wesmael (1890)
A. cucullobracteatum Leveille et Vaniot (1906)
A. sikkimense Miquel (1867) ssp. sikkimense China
Synonym A. hookeri Miquel (1867)
A. tegmentosum Maximowicz (1857) East Asia
Synonym A. pensylvanicum ssp. tegmentosum (Maximowicz) Wesmael (1890)
A. tschonoskii Maximowicz (1886) ssp. tschonoskii East Asia
Synonym A. pellucidobracteatum Leveille et Vaniot (1906)
A. tschonoskii ssp. koreanum Murray (1977) (not studied) East Asia
Synonym A. komarovii Pojarkova (1949)


Figure 5. Diagram of a generalized Acer leaf showing terms used here. (A) Medial
primary vein, (B) lateral primary vein, (C) medial secondary vein, (D) lateral secondary




Dipteronia fruits are distinctive winged schizocarps composed typically of two

subelliptical mericarps (Fig. 6). The mericarps are joined to each other at their proximal

edges along an axis continuous with the pedicel. Each schizocarp produces a single style

that arises apically from the junction of the mericarps (Fig. 10). The perianth and disk

are found at the junction between the pedicel and the fruit (Fig. 6). The schizocarps split

apart along the median axis, typically leaving a flat proximal detachment scar on each

mericarp. A prominent vascular trace extends along the attachment scar (Fig. 2A), then

deflects toward the centrally positioned elliptical seed of each mericarp (Fig. 2B). When

this large primary vein reaches the seed, it ramifies into smaller secondary veins that

form a reticulum covering the pericarp (Fig. 9). These veins give rise to tertiary veins

that radiate out from the seed body area to vascularize the encircling wing. These tertiary

veins dichotomize and anastamose and terminate directly at the margin of the wing.

Quaternary veins form a fine reticulum of polygonal areoles throughout the wing with

few or no free-ending veinlets (Figs. 11, 12).

The two modem species of Dipteronia differ in overall size and in the thickness

of the pericarp around the seed. Dipteronia dyerana (Fig. 8) mericarps average 4.5-6.0

cm in diameter, significantly larger than D. sinensis (Figs. 6, 7) fruits (2.0-2.5 cm). The

general appearance of the fruit is different, because D. dyerana fruits have a noticeably

thicker pericarp layer around the seed body and the wing is thicker. This results in a less

distinct boundary between the pericarp and wing. In the fewD. dyerana herbarium

specimens studied, I observed an increased tendency for the fruits to develop only one

mature mericarp per pair, resulting in a more circular fruit (Fig. 8). Also, the attachment

scar between the two fruits is proportionally smaller than in D. sinensis.


All of the fossil Dipteronia fruits appear to fall within a single species, below

named D. brownii sp. nov. Compared to the two living species, the fossil fruits more

closely resemble D. sinensis. The most obvious difference from either modem species is

their smaller size. The fossil mericarps studied range from 0.8 to 2.4 cm in diameter (Fig.

3), with the majority of specimens between 1 and 2 cm, compared to the larger ranges of

the modem species, as described above. Dipteronia brownii fruits have approximately

the same range of variation in general shape as modem D. sinensis fruits. Fossil fruits are

readily distinguished from D. dyerana, because they are significantly smaller, the

attachment scar is larger in proportion to the body of the wing and the pericarp does not

appear to be as thick.

Mericarps of the modern species almost always occur in pairs, although

tricarpellate fruits have been reported (Hall, 1961). The fruits of the fossil species were

bore both in twos (Figs. 14, 20, 21) and in threes (Figs. 13, 15, 16, 17). Complete

undetached schizocarps of Dipteronia are rarely preserved as fossils--currently eight

schizocarps are known from the Eocene of western North America. Four of them clearly

show the attachment of three mericarps to a shared pedicel (Figs. 13, 15, 16, 17), three

show only two mericarps (Figs. 14, 20, 21), and one is equivocal (Figs.18, 19). Most of

these schizocarps are from the intensively collected locality of Republic, Washington.

I am uncertain whether the two-mericarp specimens were original pairs, or

whether they might represent triplets that had already shed one mericarp prior to

deposition. I favor the interpretation that D. brownii bore both bi- and tricarpellate fruits.

The pattern of venation is the same in the fossils as in the modern, including the

very fine reticulation between the radiating veins of the wing (Fig. 23).

I have been unable to detect any consistent differences between Dipteronia fruits

from different geological localities and ages ranging from the Late Paleocene through the

Oligocene. Morphologically, the species appears not to have changed greatly during this

time in its fruit characters, although it would be interesting to know if the foliage also

remained constant. The youngest fossil fruit, a single specimen from the Oligocene

Bridge Creek flora, is one of the largest of the fossil specimens, measuring 2.38 cm

(Meyer and Manchester, 1997). Based on its large size, it falls close to the average size

of the extant D. sinensis. With only one specimen, I cannot evaluate the population

variability for the Bridge Creek occurrence and my placement of the specimen with the

older fossils of D. brownii is only provisional.

Previous Nomenclature

Roland Brown (1935) was the first to recognize Dipteronia from fruits in the

Tertiary of western North America. Unfortunately, the specific names that he applied to

the fruits have priority for unrelated leaf fossils and the nomenclature has become

confused. In 1935, he described a new species Dipteronia americana Brown from

Republic, Washington and included both leaves and fruits in his concept of the species,

although neither has ever been found attached. Brown figured one fruit and one

unattached leaflet and cited a previously described leaflet from Berry (1929, pl. 50, Fig.

5 [USNM 38094]) in synonymy, but he did not designate a holotype. Wolfe and Wehr

(1987) subsequently designated Berry's leaflet specimen as the lectotype for Brown's

species. At the same time they transferred this species into their new extinct genus

Bohlenia. Wolfe and Wehr (1987) included both leaves and fruits in their concept of

Bohlenia americana (Brown) Wolfe and Wehr, although the fruits have never been found

attachment to the Bohlenia foliage. However, the fruits cannot be distinguished

morphologically from modem Dipteronia, except they are usually smaller and might be

more commonly in threes, as discussed above. The possibility that the fossil population

may have had fruits in threes may be used to differentiate species, but does not constitute

the generic distinction made by Wolfe and Wehr (1987). Because fossil Dipteronia fruits

have never been found attached to any type of foliage, I restrict the name Bohlenia

americana to the foliage upon which it is based.

Brown (1937) also recognized a second fossil species of Dipteronia, D. insignis

(Lesq.) Brown, when he transferred an isolated leaf impression previously called Myrica

insignis by Lesquereux from Florissant, Colorado into the genus Dipteronia. Although

Brown implied that the name should be applied both to fruits and foliage, the epithet

insignis is based on leaf remains of equivocal relationship to Dipteronia. Hence, the

name is inappropriate for accommodating the fossil fruits. Wolfe and Wehr (1987)

established the new combination Bohlenia insignis (Lesquereux) Wolfe & Wehr for this


In my opinion, the name Bohlenia should be applied exclusively to foliage. The

true systematic affinities of this foliage remain elusive because similar kinds of leaflets

are produced by more than one extant genus of sapindales. Accordingly, all fossil fruits

previously associated with the names Dipteronia and Bohlenia foliage are orphaned and

need to be placed into a new fossil fruit species. Although Brown considered the fossil

fruits to represent two species, I regard them all as variants of the single species described


Dipteronia brownii, McClain et Manchester, sp. nov.

Diagnosis. Fruits schizocarpic, composed of 3 (or sometimes 2) mericarps,

usually dispersed singly. Pedicel slender, broadened with disk and perianth at junction

with fruit. One style, extending apically between the mericarps. Each mericarp with a

central flattened elliptical to pyriform pericarp 3-8 mm diameter surrounded by a flat,

subelliptical wing 8-24 mm in diameter. Wing margin entire, rounded, except for a flat

proximal edge representing the detachment scar. Primary vein extending from the

pedicel, 1.6-5.6 mm to the apical end of the attachment scar, then deflecting sharply at

90-135 degrees and extending straight 2-8 mm to the centrally positioned seed of each

mericarp (Fig. 25). Secondary veins forming a coarse reticulum over the pericarp (Figs.

13, 17). Tertiary veins radiating from the seed body, dichotomizing and anastamosing

enroute to the wing margin. Quaternary veins forming a fine reticulum of polygonal

areoles throughout the wing with few or no free-ending veinlets (Fig. 23).

Holotype. UWBM 39729, Fig. 13, from middle Eocene of Republic, Washington

(UWBM locality A0307).

Other specimens. Figure 24 (Paleocene, Hell's Half Acre, WY; UF locality

15740), Fig. 33 (Middle Eocene, McAbee, BC; US 38-9734), Figs. 18, 30, 31, 32 (One

Mile Creek, Princeton BC; UWBM locality 3389), Figs. 13-17, 21-23, 25, 34 (Republic,

WA, UF localities 18048, 18152; UWBM localities A0307, B2737, B4131), Fig. 29

(White Cliffs Sr., OR UF locality 00262), Fig. 35 (West Branch Creek, OR; USGS

locality 8637), Fig. 27 (White Rock Gulch, OR UF locality 00237), Figs. 20, 36, 37

(Sumner Spring, OR; UF 00075, 00283), Fig. 28 (Teater Road, OR; UF locality 00256),

Fig. 26 (Joseph Creek, BC, UCMP 168335); Not figured: YMP 37270 (Florissant, CO);

Not figured: USNM locality 8641 (Bridge Creek, OR).

Number of specimens examined: over 65

Etymology: The epithet acknowledges the work of Roland Brown in describing

the first known Dipteronia fossils from North America.

Discussion of Dipteronia brownii sp. nov.

Dipteronia brownii mericarps have a ratio of length to width that ranges from

0.68 to 1.45, but commonly they are more or less orbicular. If only a few of the

specimens were studied, one might interpret the morphological differences to indicate

that more than one species is represented. However, at localities where numerous

specimens are available, e.g., Republic, Washington (25 specimens), One Mile Creek, BC

(eight specimens) and Sumner Spring (25 specimens), it becomes clear that there is a

wide and continuous range in size and shape of the wing and even the proportion of

pericarp size to wing size. Some specimens have a shallow concavity in the wing on the

basal side adjacent to the attachment scar (Figs. 27, 30, 33, 36), whereas others from the

same localities are only convex along the basal wing margin.

In some specimens the proximal margin of the schizocarp continues straight well

beyond the attachment scar before rounding (Figs. 28, 30); in others the corresponding

margin becomes convex immediately beyond the attachment scar (Figs. 27, 33). Some

specimens show a small circular area within the pericarp (Figs. 27, 30). This could be

interpreted as the immature seed, or possibly the second aborted ovule, although aborted

ovules in Acer are much smaller. It is possible that, similar to some Acer species, the

pericarp may be fully developed although the seed is still growing.

The reticulum of quaternary veins forming a mesh between the radiating tertiary

veins is well preserved in only a few specimens (Figs. 22, 23). It corresponds to that

observed in the modern species (Figs. 11, 12).

No consistent morphological differences were noted in North American fossils

across time or between localities. There are some differences in mean fruit size between

localities, but the ranges overlap. Smaller fruits tend to occur in the populations at One

Mile Creek (9.5 to 15.0, average 11.8 mm, N = 8) and at Sumner Spring (8.0-13.5,

average 11.0 mm, N = 25), but at Republic there is both a greater range of size and

greater mean size (10-18, average 16.0 mm, N = 25). From most localities only one or a

few specimens are available, making it difficult to determine the "typical" fruits of those

populations. For example, at McAbee, the five specimens range from 11.1 to 14.2 mm

(average 12.8 mm).

The range of variation is fruit size does appear to increase through time (Fig. 4).

However, it is difficult to interpret whether the increase in size range throughout time

represents the true evolution or is only an artifact due to small sample size during the

Paleocene. One hypothesis that may be suggested based on these results is that

Dipteronia was more diverse during the Eocene, with either more species or more

morphological variation in fruit size than today. The small fruits from Sumner Spring

may be explained by the hypotheses that they represent a species or population that

became extinct and that only the species or population producing larger fruits migrated

to, or continued to exist in, Asia.

I have adopted a broad circumscription of Dipteronia brownii, which

accommodates the full range of morphological variability seen at the type locality

(Republic) and other sites ranging from Paleocene to Oligocene. It is possible that if the

corresponding foliage were available, I could distinguish more than one species, but the

associated leaflets remain speculative. Fruits in twos vs. threes could possibly have been

used as a basis to recognize two species, but because the schizocarps are very rarely

preserved intact, most fossil mericarps would be unassignable to a species. I have so far

found only eight complete schizocarps, half of which show three mericarps (Figs. 13, 15,

16, 17), three fossils show two mericarps (Figs. 14, 20, 21), and one is equivocal (Figs.

18, 19). In my view, a distinction based on merosity would be artificial, because carpel

number was probably a variable feature in the population; even today fruits of D. sinensis

occasionally develop from tricarpellate ovaries (Hall, 1961). Tricarpellate schizocarps

can also be found on modem Acer trees (Pax, 1885; de Jong, 1976). Wolfe and Wehr

(1987) cited the occurrence of three carpels as a basis for distinguishing a separate genus

and argued that the taxon be moved from Aceraceae (two carpels) to Sapindaceae

(usually three carpels). However, the clade containing Acer and Dipteronia likely

originated within the Sapindaceae, from an ancestor with the primitive sapindalean trait

of three carpels. Therefore, the three-fruited specimens merely call into question the

level of universality of the bicarpellate fruit character. A reduction from three to two

carpels may not have occurred at the base of the Aceroideae clade, but higher up. This

reduction may have occurred gradually over time, so some populations could have been

polymorphic. If this hypothesis is true, combining the two- and three-fruited fossils may

possibly result in a paraphyletic species. However, because this is mostly conjecture and

number of carpels is not usually known in the fossils, I adopt a broad circumscription for

Dipteronia brownii.


The earliest known Dipteronia fruits are two specimens from the Paleocene Fort

Union Formation of Hell's Half Acre, central Wyoming (Fig. 24), 60-63 million years

old. Surprisingly, the distinctive fruits have not been recovered from other Paleocene

localities, which are widely distributed and well collected in Wyoming, Montana and

North Dakota (Brown, 1962; Hickey, 1977; Crane, Manchester, and Dilcher 1990; Wing,

1994) and they have not yet been observed in early Eocene sites. However, they are well

represented at middle and late Eocene localities. Middle Eocene localities are distributed

in Washington, Oregon, and British Columbia. Late Eocene occurrences are known from

Colorado, western Montana, and Oregon. The only locality of certain Oligocene age is

from the Bridge Creek flora of Oregon, 32 million years old.

The Asian fossil record is undoubtedly incomplete and more collections need to

be made. In Europe, however, where paleobotanical study has been more intense,

Dipteronia fruits have not been described in the literature and are not present in major

museum collections (personal communication from H. Walther, Staatliches Museum fuir

Minerologie und Geologie, Dresden and Z. Kvacek, Charles University, Prague to S.

Manchester). Therefore, my current hypothesis assumes the genus has been limited to

Asia and North America throughout its history.

This leads to the supposition that the genus evolved in one of these continents and

emigrated to the other. Taking the fossil record as it is currently known, one might

conclude that the genus first appeared in North America in the Paleocene and migrated to

Asia soon after. At this time, both continents were at higher latitudes, but the climate

was warmer and a land bridge was present (Tiffney, 1994). Although the literature

suggests there are records of Dipteronia in the Miocene of North America (LaMotte,

1952), this information is based on either erroneous taxonomic assignments or inaccurate

age dates. I am not aware of any valid occurrences in North America subsequent to the

lower Oligocene. From this I infer that Dipteronia went extinct in North America by the

end of the Oligocene.

The fossil record of Dipteronia is nearly continuous from the Paleocene through

the Oligocene in North America, but because of the lack of fossil fruits in Asia, it is

uncertain whether the genus inhabited the continent prior to the Recent. My

interpretation is that Chinese fossil fruits have not found because they are rare.

Dipteronia fruits are never very common fossils, except perhaps at Republic, Washington

and Sumner Springs, Oregon.


Judd, Sanders, and Donoghue (1994) suggest that Acer and Dipteronia diverged

from a "winged samaroid clade" of the Sapindaceae including members of the tribes

Thouinineae (Athayana, Bridgesia, Diatenopterys, and Thouinia) and Paulinieae

(Serjania and Thinouia). The divergence of aceroids from other Sapindaceae follows a

pattern of temperate taxa diverging from large tropical families (Judd, Sanders, and

Donoghue, 1994). Judging from the Paleocene occurrences of Acer and Dipteronia, the

divergence of the clade probably occurred during the Late Cretaceous or Early Paleocene.

It is interesting that Dipteronia fruits have remained essentially the same from the

Paleocene to the Recent, perhaps only just recently forming two distinct species.

Although it is unknown if there has been similar stasis in other parts of the plant, clearly

the genus has not diversified as much as its sister taxon, Acer. Although the two genera

have existed for the same amount of time, there are now only two living species of

Dipteronia and over 120 species of Acer (van Gelderen, de Jong, and Oterdoom, 1994).

It is essential that the fossil representatives of Dipteronia be understood from a

"whole plant" perspective so that they can be used in a cladistic analysis of Acer and

Dipteronia. Once Dipteronia is well understood, the modern and fossil representatives

can be included in phylogenetic analyses of the clade and possibly be used as an outgroup

for the genus Acer.

To gain a better understanding of the phylogeny of Dipteronia, it is necessary to

obtain more characters from other organs. It is likely that leaves of Dipteronia are also

present at some of the localities where the fruits occur, but detailed comparative studies

to show how to distinguish Dipteronia leaves from other similar sapindalean genera have

not been published. It would be of interest to know if the leaves from the Eocene of

China more closely resemble those of the extant species, or those of the North American

Eocene representatives. In addition, an understanding of foliage variability within the

genus would provide a better understanding of the evolution of leaf characters in

comparison with other species of both Dipteronia and Acer.

Bohlenia foliage appears to be sapindalean, but does not closely resemble modem

Dipteronia leaflets. The leaflets of Bohlenia are much smaller, have rounder teeth and

may have bohleneoid venation, where alternating secondary veins terminate at a sinus

instead of a tooth. In fact, I believe the Bohlenia leaflets figured by Wolfe and Wehr

(1987, Plate 13, Figs. 1, 3) are more similar to the modem Koelreuteria elegans (Seem.)

A.C. Sm. Koelreuteria is also represented at the same site by fruit remains (Wolfe and

Wehr, 1987).

In this thesis, I have attempted to clarify the fossil record of Dipteronia in North

America. Because Acer leaf and fruit fossils are known from Tertiary localities in North

America, Europe, and Asia, I focused my biogeographic study of Dipteronia in these

areas. However, I did not find any valid Dipteronia specimens from Europe or Asia. I

believe that the lack of European Dipteronia fossils is significant, because Acer fossils

are known from many European localities. However, the Asian plant fossil record is not

as well understood as that of North America and Europe for many plant taxa. Because

Dipteronia lives in China today, it is likely that fossils will be found if localities are

collected further.

6 7 8

9 ' ,_ iirt

Figures 6-12. Extant fruits of Dipteronia. Figs. 6-8, by combined transmitted and
reflected light, scale bar = 1 cm. 6. Typical schizocarp of D. sinensis with two fully
developed mericarps (A: B. Bartholomew et al., 1063, Hubei). 7. Another specimen
from the same collection showing an enlarged schizocarp (left) and an undeveloped one
(right). 8. Dipteronia dyerana showing large fruit size (A: H. T. Tsai, 51524, Yunnan).
9. Dipteronia sinensis detail from Fig. 7, showing secondary veins over the pericarp and
radiating tertiary veins (scale bar = 1 mm). 10. Dipteronia sinensis enlargement from
Fig. 6, showing the pair of recurved styles arising apically from the pair of mericarps
(scale bar = 1 mm). 11. Enlargement from Fig. 6, showing the reticulum of quaternary
veins between the tertiaries (scale bar = 1 mm). 12. Enlargement of specimen in Fig. 7,
showing the higher order venation (scale bar = 1 mm).

Figures 13-23. Schizocarps and mericarps of Dipteronia. brownii sp. nov. from the
Eocene of western Canada and western USA (scale bar = 1 cm). 13. HOLOTYPE, from
Republic, Washington, showing three attached mericarps. Previously illustrated as
Bohlenia americana by Wolfe and Wehr (1987). UWBM 39729 (locality A0307). 14.
Schizocarp with two attached mericarps from Republic, Washington, SR 96-08-07. 15.
Schizocarp with three attached mericarps from Republic, Washington, SR 92-20-3A. 16.
Schizocarp with remnants of perianth at junction of pedicel with fruit. Specimen shows
three mericarps: two in face view and a third with the margin of its wing protruding on
the lower left. Republic, Washington, UWBM 71380 (locality B2737). 17. Schizocarp
with three attached mericarps. Republic, Washington UWBM 26006 (locality B4131).
18. Schizocarp showing two attached mericarps, pedicel expanded to form a disk at
junction with the fruit. One Mile Creek, Princeton, British Columbia, UWBM 56575A
(locality B3389). 19. Enlargement of Fig. 18 showing possibly two mericarps. Note the
margin of seed body is not continuous, indicating one partial mericarp may be appressed
to another underneath. 20. Incompletely detached mericarps from Sumner Spring, Gray
Butte, Oregon, UF 00283-31375. 21. Schizocarp with two attached mericarps from
Republic Washington, UF 18152-32028. 22. Mericarp from the type locality showing
pyriform pericarp, presumably containing one enlarged seed, with the impression of a
small, perhaps immature, seed in the center, UF 18048-30283. 23. Enlargement of Fig.
22 showing venation.

* .a j
- u

', 'L 'r? ., "^ "* *

.^ *-

'1 3" "'. ..


w- *p.. w *

*.. **.H
H: .:' .,

", E: ++ : ,"

tl'''- '-i ~~ L -


. '** -

:~.~: I..

I -,

IfPI Lj~l




'k ~ i/.i ~

Figures 24-37. Dipteronia brownii mericarps from various localities of western North
America and eastern Asia (scale bar = 1 cm). 24. Hell's Half Acre, Wyoming, UF
15740-23086. 25. Republic, Washington, UWBM 55044 (locality B4131). 26. Joseph
Creek, British Columbia, UCMP 168335. 27. White Rock Gulch, Oregon, UF 00237-
30301. 28. Teater Road, Oregon, UF 00256-20862. 29. White Cliffs, Sr., Oregon, UF
00262-30303. 30. One Mile Creek, British Columbia, UWBM 57496A (locality B3389).
31. One Mile Creek, British Columbia, UWBM 97007 (locality B3389). 32. One Mile
Creek, British Columbia, UWBM 57497. 33. McAbee, British Columbia, US 38-9734.
34. Republic, Washington, UWBM. 35. West Branch Creek, Oregon, USNM 509800
(USGS locality 8637). 36. Sumner Spring, Oregon, UF 00283-31376. 37. Sumner
Spring, Oregon, UF 00275-21691.

. q.
, .-

I, "


r ."-

*. *. "

I ^:





The character used consistently by most monographers to unite Acer section

Macrantha is striped bark. The appearance of vertical white stripes on the bark gives the

common name to the group, the snakebark maples. The inflorescence bracts are poorly

developed in this section and the buds are often stalked (Ogata, 1967). The rays in the

wood are very narrow, usually only three cells wide, with a fairly irregular outline

(Ogata, 1967). The group has two pairs of bud scales, leading some researchers to

believe that it is fairly primitive, as the sister genus, Dipteronia, has no bud scales (van

Gelderen, de Jong, and Oterdoom, 1994). Other groups with very few bud scales are

section Parviflora (called Spicata by some authors), series Parviflora, Caudata, and

Distyla, section Wardiana, section Glabra, series Glabra and Arguta, section Negundo,

series Negundo and Cissifolia. Of these, the sections with three to five lobed leaves have

sometimes been placed close to Macrantha. That is, sections Parviflora, Caudatum,

Wardiana, Glabra, and Arguta. These taxa may be related to section Macrantha, but

they have leaf forms distinct from Acer section Macrantha. In this thesis, I will explain

how the three to five lobed leaves of these groups can be distinguished from similar

leaves in the Macrantha group.

Pax (1902) used staminal position as the primary character to delimit his sections.

Species with intermediate staminal insertion (Perigyna) were placed in section

Macrantha while species with an intrastaminal nectar disc (Intrastaminalia) were placed

in section Indivisa, which resulted in an artificial classification. Pojarkova (1949)

divided section Macrantha into four series: Parviflora, Tegmentosa, Macrantha, and

Crataegifolia, based on number of leaf lobes. However, her groupings did not make

sense with the normal range of leaf variation found in the taxa. Momotani (1962)

rearranged the species of section Macrantha based on fruit characters, which are not

evaluated here, but may be useful. Fang (1966) divided section Macrantha into three

series: Crataegifolia, Tegmentosa, and Maximowicziana. Ogata (1967) also

distinguished three series in section Macrantha, series Macrantha, Micrantha, and

Rufinerve based primarily on bud characters. Murray (1970) basically followed

Pojarkova's divisions, although he excluded series Parviflora. Tanai (1978) found no

fine venation characters to help him distinguish any subgroups within section Macrantha,

although he did mention that all the leaves of members of section Macrantha did not

stain well compared with those of other sections. Mai (1984) studied fruit anatomy and

divided section Macrantha into three series, Macrantha, Rufinerve, and Wardiana.

As summarized in Table 4, I studied the following species in Acer section

Macrantha: A. capillipes, A. caudatifolium, A. crataegifolium, A. davidii, A. metcalfi, A.

micranthum, A. morifolium, A. pectinatum, A. pensylvanicum, A. rubescens, A. rufinerve,

A. sikkimense, A. tegmentosum, and A. tschonoskii. From the leaf characters investigated,

I have determined that within Acer section Macrantha, there is a "Core section

Macrantha" consisting of A. pensylvanicum (Figs. 38, 58), A. tegmentosum (Figs. 39, 59),

A. rufinerve (Figs. 40, 60), A. capillipes (Figs. 41, 61), and A. morifolium (Figs. 42, 43,

62, 63). The leaves of these species are difficult to distinguish from one another. The

leaves are usually shallowly 3-5 lobed, with small basal lobes. Acer morifolium normally

has unlobed leaves. However, when it does occasionally produce lobed leaves, they are

very similar to the others in this group.

Leaves of the "Core section Macrantha" differ from those of other section

Macrantha species in their vein and margin characters. Venation is craspedodromous,

with distinct, straight medial secondary veins that do not branch until very close to the

leaf margin. The medial secondaries and lateral primary veins are similar in size, shape,

and spacing. The margin is very finely serrate, usually with 5-15 teeth/cm. The most

common tooth shapes are CC/CV and CC/ST. The leaves have acuminate apices. When

A. morifolium produces lobed leaves, they are very similar, although the teeth are usually

FL/FL. However, normally the leaves are unlobed, which makes them very easy to

distinguish from the other members of the "Core section Macrantha". When unlobed,

the venation is not very different, except that there are usually only three primary veins.

Acer morifolium leaves are distinguishable from other unlobed leaves in the section

Macrantha because the venation is craspedodromous, while the venation is

brochidodromous in other species of section Macrantha that have unlobed leaves.

Acer metcalfi (Figs. 46, 64) and A. rubescens (Figs. 47, 67) have the most similar

leaf types to the "Core section Macrantha". The leaves are usually medium 5-lobed, with

deeper lobal sinuses than the core group. The lateral primary veins are more distinct

from the medial secondary veins in the core group. The venation of medial secondaries

in these two species is craspedodromous, but not very strongly, as the veins curve up as

they get close to the leaf margin and fade as they branch increasingly. Both species have

teeth: CC/CV, ST/CV, and ST/ST, which are much larger and fewer than in the core

group. The leaves usually have only 3-4 teeth/cm.

Leaf morphology supports the retention of A. metcalfi as a species distinct from A.

sikkimense (Figs. 45, 65), whereas de Jong (1994) regards it as a subspecies of the latter.

The two clearly have different leaf forms, where A. sikkimense has unlobed leaves and A.

metcalfi has 3-lobed leaves. In my opinion, the leaves of these two taxa are also

distinguishable based on venation. Acer sikkimense has clear brochidodromous venation

while A. metcalf has weak craspedodromous to weak brochidodromous venation.

The leaves of Acer micranthum (Figs. 48, 68) and A. tschonoskii (Figs. 49, 69)

are very similar to one another. They are distinct from the core group in that they have

many lobations along the margin. The medial secondary veins are fewer in number than

in the core group and are broadly curving upwards. The leaves resemble "Core section

Macrantha" because they are usually medium five lobed, with a finely serrate margin

with teeth of the CC/CV type. Acer tschonoskii leaves tend to have deeper lobations,

although the leaves of the two species are often virtually indistinguishable.

The subspecies of A. pectinatum have leaves very similar to one another, with

overlapping ranges of variation. Therefore, I will focus on A. pectinatum ssp. pectinatum

(Figs. 50, 70) here. The leaves of A. pectinatum, A. crataegifolium (Figs. 52, 71), and A.

caudatifolium (Figs. 51, 72) differ from "Core section Macrantha" in overall shape. The

lateral primaries are significantly smaller than the primary vein, resulting in two small,

shallow lateral lobes. The medial secondaries of these species are more broadly curving

than the core group, with poor craspedodromous to weak brochidodromous venation.

The leaves are similar to the core group in their finely serrate margin and usually three or

five shallowly lobed leaves. Leaves of A. pectinatum have teeth of the CC/CV type;

those of A. crataegifolium have CC/CV, FL/CV, and FL/FL teeth; those of A.

caudatifolium have ST/ST teeth. Of the three species, A. crataegifolium leaves tend to

have venation closer to brochidodromous. In all, venation type is difficult to categorize

because the medial secondary veins become very faint towards the leaf margin.

Therefore, it is important to carefully observe this area.

Leaves ofA. davidii (Figs. 44, 64) and A. sikkimense (Figs. 45, 46) are easily

distinguished from the "Core section Macrantha" by their unlobed leaves. The leaves are

similar to the core group in their finely serrate margin. The medial secondaries, however,

are not as well defined and less straight than as in the core group and venation is


I also studied species of other Acer sections whose leaves may be similar to leaves

of section Macrantha species: A. wardii, A. spicatum, A. caudatum ssp. ukurunduense, A.

nipponicum, and A. argutum.

Acer wardii leaves differ from the leaves of species of section Macrantha in that

they are deeply three-lobed with elliptical shaped lobes. The medial secondary veins are

well defined, but broadly curving, and branch before reaching the margin. The serrate

teeth are CC/CV.

Acer spicatum leaves (Figs. 54, 74) resemble those of members of section

Macrantha in their venation but not margin characters. The straight, fairly well defined

medial secondary veins are similar to those in leaves of members of section Macrantha.

In contrast, the margin of leaves of Acer spicatum has teeth that are larger in proportion

to the leaf size. There are 2-3 teeth/cm of the FL/FL and CC/CV types.

Acer caudatum ssp. ukurunduense (Figs. 55, 75) leaves are similar to leaves of

species in the "Core section Macrantha" because of their distinct, straight medial

secondary veins. However, these veins are fewer in number and more irregularly spaced.

The margin is very different, with fewer, larger teeth, although they are of the same shape

as many in section Macrantha, ST/ST, CC/ST, and CC/CV. The leaves are usually

medium 5-lobed.

Leaves ofA. nipponicum (Figs. 56, 76) are very similar to those of members of

the "Core section Macrantha". They are shallowly 5-lobed, with distinct, straight medial

secondary veins. The margin is finely serrate as in the core group, with 10 teeth/cm.

The teeth are ST/ST and CC/CV. The teeth may be more incised than in the core group,

although the leaves are otherwise indistinguishable from the core group based on margin

and venation characters.

Leaves of A. argutum (Figs. 57, 77) resemble those of A. tschonoskii and A.

micranthum in their overall shape, with five medium lobes and acuminate apices. The

medial secondary veins of leaves of A. argutum resemble those of the "Core section

Macrantha" in their even spacing and distinct appearance. However, they are more

broadly curving than those in the core group. Teeth are ST/ST and CC/CV. The margin

has lobations, although they are not as incised as in A. tschonoskii and A. micranthum.

There is no single leaf character that can be used to place a species in section

Macrantha. However, there are smaller groups within section Macrantha whose species

have combinations of characters that can be used to place a species into the group based

on leaf characters, as I have described. Most of the leaves of species that do not belong

to section Macrantha can be distinguished from the leaves of section Macrantha,

although this might not be possible for A. nipponicum.


In order to assign a fossil leaf to Acer section Macrantha, the margin and the

secondary venation need to be preserved. The leaf does not have to be complete,

although it would be useful to see the lobes and the apices to provide additional support

for the sectional placement. I studied the five fossil species assigned to section

Macrantha by Wolfe and Tanai (1987) in their review of the North American Acer fossil

record. I also reviewed the Asian fossil leaves assigned to the section by Tanai and

Ozaki (1977), Tanai (1983), and Ozaki (1991). I analyzed the leaves for the vein and

margin characters, described in the Materials and Methods chapter, to determine if the

fossils could be placed in the modern section Macrantha.

Leaves ofA. castorrivularis (Fig. 78) are three-lobed with straight, well-defined

medial secondary veins. The margin is not well preserved in the holotype (Wolfe and

Tanai, 1987). In one section of the margin near the petiole, there appears to be only one

or two teeth along an otherwise entire margin. This is markedly different from the

modem species of Macrantha, in which the core group all have a finely serrate margin

with lots of small teeth. Based on the margin characters shown in this fossil, the species

appears closer to section Parviflora, than to section Macrantha.

Acerpalaeorufinerve has only been reported in North America from fossil fruits

(Wolfe and Tanai, 1987). The validity of fossil fruit assignments are not addressed here.

Acerpalaeorufinerve leaf fossils were originally described from the Late Miocene of

Japan (Tanai and Onoe, 1961). As described in Tanai and Ozaki (1977), A.

palaeorufinerve does appear to belong to section Macrantha. The leaf has three large

lobes and two small basal lobes, all shallowly lobed. The medial secondaries are fairly

strong, but the margin does not appear to be well preserved. Ozaki (1991) also placed

fossils from the Pliocene of Honshu, Japan in A. palaeorufinerve. These leaf remains do

seem to belong to section Macrantha, although the teeth seem rather large.

Acer uemurae is suggested by Tanai and Ozaki (1977) to be most closely related

to A. micranthum and A. tschonoskii. Leaves of A. uemurae do resemble the leaves of

these modem species in the five medium lobed leaf shape with lobations. There are few

medial secondary veins that broadly curve upwards. The margin is finely serrate, similar

to the modern species. This species clearly fits into the section Macrantha.

Ozaki (1991) refers one Pliocene fossil leaf to the modern species A.

crataegifolium. Although it is not given a formal name, this leaf does appear to be

closely related to A. crataegifolium, based on its overall leaf shape and finely serrate


Acer koreanum is a fossil fruit species described by Endo (1950) from a single

North Korean specimen. Fruits are not analyzed here, so the placement in section

Macrantha cannot be evaluated at this time.

Leaves ofA. clarnoense (Figs. 80, 83), described by Wolfe and Tanai (1987), do

not fit into the "Core section Macrantha". The leaves are shallowly 3-lobed, with few,

indistinct, broadly curving medial secondary veins. The serrate margin is different from

any species in the modem section Macrantha. There are very few teeth, even in

proportion to the small size of the leaf.

Acer dettermani Wolfe et Tanai (1987) (Fig. 81) does fit into the "Core section

Macrantha", although the specimens of this species have fragmentary preservation.

Medial secondary veins are well defined and fairly straight. Portions of the margin that

are preserved are finely serrate.

Leaves ofA. latahense (Figs. 79, 82), described by Wolfe and Tanai (1987), have

distinct, straight medial secondary veins and a serrate margin, which may warrant the

placement of this species in the "Core section Macrantha". The leaves are shallow to

medium three-lobed. However, the margin has larger teeth than the modern species in

"Core section Macrantha". The teeth are of the CC/CV shape. The large teeth indicate

closer similarity to section Parviflora than to section Macrantha.


The modem species of Acer section Macrantha are concentrated in eastern Asia,

except for one species in eastern North America. The distribution of each species is

given in Table 4.

The leaf fossils that fit into section Macrantha based on vein and margin

characters are A. dettermani from the middle or late Eocene of Alaska, A. uemurae from

the Late Miocene of Japan, A. palaeorufinerve from the Late Miocene and Pliocene of

Japan, and A. crataegifolium from the Pliocene of Japan. Species placed in section

Macrantha by Wolfe and Tanai (1987) but excluded here are A. clarnoense from the

middle to late Eocene of Oregon, A. castorrivularis from the latest Eocene of Montana,

and A. latahense from the early to middle Miocene of Washington and Oregon.

Based on the leaf fossils accepted here, section Macrantha dates back to the

Eocene in Alaska, with younger occurrences in the Miocene of Japan.

Two possible explanations are suggested by the presence of fossils with "Core

section Macrantha" leaf characters in only Asia and Alaska. The core group may have

diverged in eastern Asia and only recently migrated to North America. This could

explain why the modem diversity is higher in Asia in the core group (for example, three

species versus one in North America). It is hypothesized that this section has always

been limited to North America and Asia. I have not seen any reports of European Acer

species that I would place in section Macrantha. Alternatively, the section could have

been present in North America, but represented by species with leaf morphologies

different from the modem taxa. The modem leaf characters of straight medial secondary

veins with a finely serrate margin may be derived, and therefore the fossils could not be

assigned to the modem section Macrantha, based on those characters.


Acer section Macrantha is resolved as monophyletic according to Hasebe, Ando,

and Iwatsuki (1998), in a molecular study using chloroplast DNA restriction fragment

length polymorphisms. However, the resolution of the cladogram is poor because it is the

consensus tree of 398 equally parsimonious trees. A significant problem with the

Hasebe, Ando, and Iwatsuki (1998) study is their choice of an outgroup, A. spicatum and

A. ukurunduense [A. caudatum ssp ukurunduense (Traut. & Meyer) Murray]. Acer

spicatum is nested high in the cladograms in the ITS analysis of Ackerly and Donoghue

(1998), not basally as hypothesized by Hasebe, Ando, and Iwatsuki (1998). Ackerly and

Donoghue (1998) chose Dipteronia sinensis as their outgroup.

Acer section Macrantha is paraphyletic in the consensus tree of two most

parsimonious trees of Ackerly and Donoghue (1998), although the species A.

tegmentosum, A. rufinerve, A. crataegifolium, and A. pensylvanicum form a monophyletic

group. Acer tschonoskii and A. micranthum are phylogenetically adjacent, but section

Macrantha does not form a monophyletic group in their analysis. However, a

comprehensive analysis has not yet been undertaken. Future analyses may help to clarify

these relationships. It is interesting that Ackerly and Donoghue's (1998) analysis of

selected Acer species grouped the section Macrantha species studied (A. tschonoskii, A.

micranthum, A. tegmentosum, A. rufinerve, A. crataegifolium, and A. pensylvanicum)

with the sections Glabra (A. glabrum), Negundo (A. negundo), Parviflora (A. spicatum,

A. nipponicum, and A. distylum), and Ginnala (A. ginnala). With the exception of A.

negundo, these species all produce palmately lobed simple leaves.

Pax (1902) included A. nipponicum [as A. crassipes Pax and A. parviflorum

Franch et Sav.] and A. caudatum ssp. multiserratum [as A. erosum Pax] in his

circumscription of section Macrantha.

Figures 38-41. Modern Acer leaves (scale bar = 1 cm). 38. Acerpensylvanicum, A: I.
W. Anderson, August 16, 1915. 39. Acer tegmentosum, A: V. Komarov, 1053. 40.
Acer rufinerve, A: T. Azuma et al., 201. 41. Acer capillipes, A: Y. Kadota, 2302.

wii __B

Figures 42-45. Modern Acer leaves (scale bar = 1 cm). 42. Acer morifolium, lobed leaf,
A: M. Suzuki et al., E462. 43. Acer morifolium, unlobed leaf, A: T. Yahara et al., 9183.
44. Acer davidii, A: Q. X. Wang and J. L. Sun, 13020. 45. Acer sikkimense, A: A.
Henry, 10640.

"4 i r



Figures 46-49. Modern Acer leaves (scale bar = 1 cm). 46. Acer metcalfii, A: Y. W.
Taam, 593. 47. Acer rubescens, A: W.-H. Hu, 1429. 48. Acer micranthum, A: M.
Furuse, August 21, 1962. 49. Acer tschonoskii, A: C.-S. Chang, 884.



'46 47
. II 47




Figures 50-53. Modern Acer leaves (scale bar = 1 cm). 50. Acerpectinatum, A: T. T.
Yu, 6788. 51. Acer caudatifolium, A: C.-C. Wang, 1052. 52. Acer crataegifolium, A:
T. Fukuda et al., 73. 53. Acer wardii, A: T. T. Yu, 22080.

* F.


Figures 54-57. Modern Acer leaves (scale bar = 1 cm). 54. Acer spicatum, A: E. L.
Palmer, 775. 55. Acer caudatum ssp. ukurunduense, A: M. Furuse, June 23, 1960. 56.
Acer nipponicum, A: S. Suzuki, 398020. 57. Acer argutum, A: Y. Kadota, 2408.


1 a 18 1 '
,13i '



Figures 58-61. Modern Acer leaves (scale bar = 1 cm). 58. Acerpensylvanicum, A: I.
W. Anderson, August 16, 1915. 59. Acer tegmentosum, A: V. Komarov, 1053. 60.
Acer rufinerve, A: T. Azuma et al., 201. 61. Acer capillipes, A: Y. Kadota, 2302.



60 61

Figures 62-65. Modern Acer leaves (scale bar = 1 cm). 62. Acer morifolium, lobed leaf,
A: M. Suzuki et al., E462. 63. Acer morifolium, unlobed leaf, A: T. Yahara et al., 9183.
64. Acer davidii, A: Q. X. Wang and J. L. Sun, 13020. 65. Acer sikkimense, A: A.
Henry, 10640.


6 3


Figures 66-69. Modern Acer leaves (scale bar = 1 cm). 66. Acer metcalfii, A: Y. W.
Taam, 593. 67. Acer rubescens, A: W.-H. Hu, 1429. 68. Acer micranthum, A: M.
Furuse, August 21, 1962. 69. Acer tschonoskii, A: C.-S. Chang, 884.


68 6

Figures 70-73. Modern Acer leaves (scale bar = 1 cm). 70. Acerpectinatum, A: T. T.
Yu, 6788. 71. Acer crataegifolium, A: T. Fukuda et al., 73. 72. Acer caudatifolium, A:
C.-C. Wang, 1052. 73. Acer wardii, A: T. T. Yu, 22080.



Figures 74-77. Modern Acer leaves (scale bar = 1 cm). 74. Acer spicatum, A: E. L.
Palmer, 775. 75. Acer caudatum ssp. ukurunduense, A: M. Furuse, June 23, 1960. 76.
Acer nipponicum, A: S. Suzuki, 398020. 77. Acer argutum, A: Y. Kadota, 2408.


Figs. 78-83. Fossil Acer leaves (scale bar = 1 cm). 78. Acer castorrivularis, UCMP
9310. 79. Acer latahense, USNM 396136. 80. Acer clarnoense, UCMP 9010. 81.
Acer dettermani, USNM 396034. 82. Acer latahense, USNM 396136. 83. Acer
clarnoense, UCMP 9010.

1 I-



The fossil records for Dipteronia and Acer section Macrantha are more limited, in

my opinion, than reported previously in the literature. The fossil record of Dipteronia

may be expanded in the future if leaves are confirmed to be diagnostic for the genus.

The Dipteronia fossil record, based on fruits, extends from the Paleocene to the

Oligocene in North America. Dipteronia may have been present in North America

earlier or later, but today occurs only in China. The Asian fossil record is unfortunately

not as well known. Therefore, the place of origin for Dipteronia, the direction of

intercontinental migration, and the age of its initial appearance in Asia are unknown.

Leaves remain to be analyzed to determine if they can be reliably be assigned to

the genus Dipteronia. The leaves could provide additional evidence for the existence of

the genus in areas where the fruits are not commonly preserved. This could possibly be

important in Asia, where there are no known fossil Dipteronia fruits.

In order to determine if a fossil Acer leaf fits into section Macrantha, the

secondary venation and margin need to be preserved. The medial secondary veins in

section Macrantha are fairly regularly spaced and straight to broadly curved. They

generally do not branch until very close to the margin. The margin in leaves of section

Macrantha is finely serrate, with three or more orders of teeth. The teeth are very small

in relation to the size of the leaf, with 5-15 teeth/cm. The teeth are irregularly spaced and

many are of the CC/CV shape as defined by the Leaf Angiosperm Working Group


The remaining modem sections of Acer still need to be studied and described in

order to determine if these sections are present in the fossil record. Additionally, the

fossil sections of Acer need to be analyzed to determine if they exhibit characters or sets

of characters not present in any modem section. The fossil fruits could also be analyzed

in the same way to provide additional evidence and to determine if different organs show

the same rates of evolution. It is hoped that by providing a character-based recognition

system for leaves of the modem sections, neo- and paleobotanists will be better able to

assign Acer fossils at the sectional level and, thus, to learn more about the biogeography

and evolutionary patterns within the genus.


Ackerly, D. D., and M. J. Donoghue. 1998. Leaf Size, Sapling Allometry, and Comer's
Rules: Phylogeny and Correlated Evolution in Maples (Acer). American
Naturalist 152: 767-791.

Akhmetiev, M.A. 1971. Tertiary maples of eastern Asia. Paleontological Journal 5:

Becker, H. F. 1960. Additions to the Tertiary Ruby Paper Shale flora of Southwestern
Montana. Bulletin of the Torrey Botanical Club 87: 386-396.

Becker, H. F. 1961. Oligocene plants from the Upper Ruby River Basin, Southwestern
Montana. Geological Society of America Memoir 82: 1-127.

Berry, E. W. 1929. A revision of the flora of the Latah Formation. U. S. Geological
Survey Professional Paper 154-H: 225-265.

Bestland, E. A., P. E. Hammond, D. L. S. Blackwell, M. A. Kays, G. J. Retallack, and J.
Stimac. 1999. Geologic framework of the Clamo Unit, John Day Fossil Beds
National Monument, central Oregon. Oregon Geology 61: 3-19.

Boulter, M. C., J. N. Benfield, H. C. Fisher, D. A. Gee, and M. Lhotak. 1996. The
evolution and global migration of the Aceraceae. Philosophical Transactions of
the Royal Society of London B, Biological Sciences 351: 589-603.

Brown, R. W. 1935. Miocene leaves, fruits, and seeds from Idaho, Oregon, and
Washington. Journal of Paleontology 9: 572-587.

Brown, R. W. 1937. Additions to some fossil floras of the Western United States. U. S.
Geological Survey Professional Paper 186-J: 163-206.

Brown, R. W. 1959. A bat and some plants from the Upper Oligocene of Oregon.
Journal of Paleontology 33: 125-129.

Brown, R. W. 1962. Paleocene Flora of the Rocky Mountains and Great Plains. U. S.
Geological Survey Professional Paper 375: 1-119.

Crane, P. R., S. R. Manchester, and D. L. Dilcher. 1990. A preliminary survey of fossil
leaves and well-preserved reproductive structures from the Sentinel Butte

Formation (Paleocene) near Almont, North Dakota. Fieldiana Geology, new
series, 20: 1-63.

de Jong, P. C. 1976. Flowering and sex expression in Acer L. A biosystematic study.
Mededelingen Landbouwhogeschool Wageningen 76: 1-201.

de Jong, P. C. 1994. Chapter 6 In van Gelderen, D. M., P. C. de Jong, and H. J.
Oterdoom. Maples of the world, 69-103. Timber Press, Portland, Oregon, USA.

Delendick, T. J. 1981. A systematic review of the Aceraceae. Ph. D. thesis, City
University of New York, New York.

Delendick, T. J. 1990. A survey of foliar flavonoids in the Aceraceae. Memoirs of the
New York Botanical Garden 54:1-129.

Endo, S. 1950. On the fossil Acer from Japan, Korea, and south Manchuria. Short
Papers, Institute of Geology and Paleontology, Sendai 1: 11-17.

Fang, W.-P. 1939. A monograph of Chinese Aceraceae. Contributions from the
Biological Laboratory of the Chinese Association for the Advancement of
Science, section botany, 11: 1-346.

Fang, W.-P. 1966. Revisio taxorum Aceracearum Sinicarum. Acta Phytotaxonomica
Sinica 11: 139-189.

Gadek, P. A., E. S. Fernando, C. J. Quinn, S. B. Hoot, T. Terrazas, M. C. Sheahan, and
M. W. Chase. 1996. Sapindales: molecular delimitations and infraordinal groups.
American Journal of Botany 83: 802-811.

Hall, B. A. 1961. The floral anatomy of Dipteronia. American Journal of Botany 48:

Hasebe, M., T. Ando, and K. Iwatsuki. 1998. Intrageneric Relationships of maple trees
based on the chloroplast DNA restriction fragment length polymorphisms.
Journal of Plant Research 111: 441-451.

Henry, A. 1903. Chinese maples. Gardeners' Chronicle, series III, 33: 21-22, 62-63,

Hickey, L. J. 1973. Classification of the architecture of dicotyledonous leaves.
American Journal of Botany 60: 17-33.

Hickey, L. J. 1977. Stratigraphy and paleobotany of the Golden Valley Formation (early
Tertiary) of western North Dakota. Geological Society of America Memoir 150:

Judd, W. S., R. W. Sanders, and M. J. Donoghue. 1994. Angiosperm family pairs:
preliminary phylogenetic analyses. Harvard Papers in Botany 1(5): 1-51.

LaMotte, R. S. 1952. Catalogue of the Cenozoic plants of North America through 1950.
Geological Society of America Memoir 51: 1-381.

Leaf Architecture Working Group. 1999. Manual of Leaf Architecture morphological
description and categorization of dicotyledonous and net-veined
monocotyledonous angiosperms. Published privately. Washington, D. C. 1-65.

MacGinitie, H. D. 1953. Fossil plants of the Florissant beds, Colorado. Carnegie
Institution of Washington Publication 599: 1-198.

MacGinitie, H. D. 1974. An early middle Eocene flora from the Yellowstone -
Absaroka volcanic province, northwestern Wind River Basin, Wyoming.
University of California Publications in the Geological Sciences 108: 1-103.

Mai, D. H. 1984. Die Endokarpien bei der Gattung Acer L. (Aceraceae) Eine
biosystematisch Studie. Gleditschia 11: 17-46.

Manchester, S. R. 1990. Eocene to Oligocene floristic changes recorded in the Clarno
and John Day Formations, Oregon, USA. In E. Knobloch and Z. Kvacek [eds.],
Symposium Proceedings, paleofloristic and paleoclimatic changes in the
Cretaceous and Tertiary, 183-187. Geological Survey Press, Prague, Czech

Manchester, S. R. 1999. Biogeographical relationships of North American Tertiary
floras. Annals of the Missouri Botanical Gardens 86: 472-522.

Manchester, S. R. 2000. Late Eocene fossil plants of the John Day Formation, Wheeler
County, Oregon. Oregon Geology 62: 51-63.

McFadden, J. J. 1986. Fossil flora near Gray Butte, Jefferson County, Oregon. Oregon
Geology 48: 51-58.

Meyer, H. W., and S. R. Manchester. 1997. The Oligocene Bridge Creek flora of the
John Day Formation, Oregon. University of California Publications in the
Geological Sciences 141: 1-195.

Momotani, Y. 1962. III. System of Aceraceae. Memoirs of the College of Science;
Kyoto Imperial University, series B, biology 29: 177-189.

Murray, A. E. 1970. A Monograph of the Aceraceae. Ph. D. thesis, Pennsylvania State

Nichols, D. J., and H. L. Ott. 1978. Biostratigraphy and evolution of the Momipites-
Caryapollenites lineage in the early Tertiary in the Wind River Basin, Wyoming.
Palynology 2: 93-112.

Nichols, D. J., and R. M. Flores. 1993. Palynostratigraphic correlation of the Fort Union
Formation (Paleocene) in the Wind River Reservation in the Waltman Area, Wind
River Basin, Wyoming. In W. R. Keefer, W. J. Metzger, and L. H. Godwin
[eds.], Wyoming Geological Association Special Symposium on Oil and Gas and
other resources of the Wind River Basin, Wyoming, 175-189. Pioneer Printing,
Cheyenne, Wyoming, USA.

Ogata, K. 1965. A dendrological study on the Japanese Aceraceae, with special
reference to the geographical distribution. Bulletin of the Tokyo University
Forests 60: 1-99.

Ogata, K. 1967. A systematic study of the Aceraceae. Bulletin of the Tokyo University
Forests 63: 89-206.

Oliver, D. 1889. Dipteronia sinensis. Hooker's Icones Plantarum, Third series, vol. 9.
Plate 1898.

Ozaki, K. 1991. Late Miocene and Pliocene floras in Central Honshu, Japan. Bulletin of
Kanagawa Prefectural Museum Natural Science Special Issue: 1-244.

Pax, F. 1885. Monographie der Gattung Acer. Botanische Jahrbuch 6: 287-347.

Pax, F. 1886. Monographie der Gattung Acer. Botanische Jahrbuch 7: 177-263.

Pax, F. 1902. Aceraceae. In A. Engler [ed.], Das Pflanzenreich, series IV, 163: 8: 1-89.

Pojarkova, A. I. 1949. Aceraceae. Flora USSR 14: 443-476 (English translation by
Israel Program for Scientific Translation, Jerusalem).

Smith, G. A., S. R. Manchester, M. Ashwill, W. C. McIntosh, and R. M. Conrey. 1998.
Late Eocene-early Oligocene tectonism, volcanism, and floristic change near
Gray Butte, central Oregon. Geological Society of America Bulletin 100: 759-

Stockey, R. A., and W. C. Wehr. 1996. Flowering Plants in and around Eocene Lakes of
the Interior. In R. Ludvigsten [ed.], Life in stone, a natural history of British
Columbia's fossils, 234-247. UBC Press, Vancouver, British Columbia, Canada.

Tanai, T. 1978. Taxonomical investigation of the living species of the genus Acer L.,
based on vein architecture of leaves. Journal of the Faculty of Science, Hokkaido
University, series IV, 18: 243-282.

Tanai, T. 1983. Revisions of Tertiary Acer from East Asia. Journal of the Faculty of
Science, Hokkaido University, series IV, 20: 291-390.

Tanai, T., and Onoe, T. 1961. A Mio-Pliocene flora from the Ningyotoge area on the
border between Tottori and Okayama Prefecture, Japan. Geological Survey of
Japan Report 187: 1-62.

Tanai, T., and K. Ozaki. 1977. The Genus Acer from the Upper Miocene in Tottori
Prefecture, Western Japan. Journal of the Faculty of Science Hokkaido
University, series IV, 17: 575-606.

Tiffney, B. H. 1994. An estimate of the early Tertiary paleoclimate of the southern
Arctic. In M. C. Boulter and H. C. Fisher, [eds.], Cenozoic plants and climates of
the Arctic, 267-295. Springer-Verlag, New York, New York, USA.

van Gelderen, D. M., P. C. de Jong, and H. J. Oterdoom. 1994. Maples of the world.
Timber Press, Portland, Oregon, USA.

Walther, H. 1972. Studien tiber tertiare Acer Mitteleuropas. Abhandlungen des
Staatlichen Museums fuir Mineralogie und Geologie 19: 1-309.

Wehr, W. 1995. Early Tertiary flowers, fruits and seeds of Washington State.
Washington Geology 23: 3-16.

Wing, S. L. 1994. Fossil plant localities. In T. M. Bown, K. D. Rose, E. L. Simons, and
S. L. Wing, [eds.], Distribution and stratigraphic correlation of Upper Paleocene
and Lower Eocene fossil mammal and plant localities of the Fort Union,
Willwood, and Tatman Formations, southern Bighorn Basin, Wyoming. U. S.
Geological Survey Professional Paper 1540: 1-103.

Wolfe, J. A., and T. Tanai. 1987. Systematics, phylogeny, and distribution of Acer
(maples) in the Cenozoic of western North America. Journal of the Faculty of
Science of Hokkaido University, series IV, 22: 1-246.

Wolfe, J. A., and W. Wehr. 1987. Middle Eocene dicotyledonous plants from Republic,
Northeastern Washington. U. S. Geological Survey Bulletin 1597: 1-25.

Ying, T.-s., Y.-l. Zhang, and D. E. Boufford. 1993. The endemic genera of seed plants
in China. Science Press, Beijing, P. R. China.


Amy Marie McClain was born in Lansing, Michigan, on July 22, 1976. She

graduated from Michigan State University with a Bachelor of Science in botany and plant

pathology in 1998. Her positive undergraduate research experience and encouragement

from many people in the Department of Botany and Plant Pathology led her to pursue a

master's degree. Her interest in studying evolution through the fossil record brought her

to the University of Florida to work with Steven Manchester and David Dilcher. She will

continue her research on modem and fossil Acer as a doctoral student with Dr. Steven

Manchester in the Department of Botany at the University of Florida.