HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (27) Citing Articles via Google Scholar Google Scholar Articles by Crepet, W. L. Articles by Nixon, K. C. Search for Related Content PubMed Articles by Crepet, W. L. Articles by Nixon, K. C. GeoRef GeoRef Citation Agricola Articles by Crepet, W. L. Articles by Nixon, K. C.

Two new fossil flowers of magnoliid affinity from the Late Cretaceous of New Jersey¹

^a L. H. Bailey Hortorium, 462 Mann Library, Cornell University, Ithaca, New York 14853-4301

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS SYSTEMATICS DISCUSSION REFERENCES

 
Two taxa of cupulate magnoliid fossil flowers, Cronquistiflora and Detrusandra, are described from the Late Cretaceous (Turonian, ~90 million years before present [MYBP]) Raritan (or lower Magothy) Formation of New Jersey. The fossil taxa are represented by flowers at various stages of development, associated fragments of cup-shaped floral receptacles with attached anthers, and isolated anthers. Both taxa have laminar stamens with adaxial thecae and valvate dehiscence. Pollen is boat-shaped and foveolate in anthers associated with Cronquistiflora and spherical with reticulate ornamentation in Detrusandra. Cup-shaped receptacles are externally bracteose in both taxa. The receptacle of Cronquistiflora is broader than the campanulate one of Detrusandra. Cronquistiflora also has more carpels (~50 in a spiral vs. ~5 in a whorl or tight spiral). In Detrusandra the carpels are surrounded by dorsiventrally flattened structures (pistillodes?) that are remote from the attachment of the stamens near the distal rim of the receptacular cupule. Detrusandra stigmas are rounded and bilobed, while those of Cronquistiflora, although bilateral in symmetry, are somewhat peltate. The fossil taxa share prominent characters with extant cupulate magnoliids (e.g., Eupomatia, Calycanthus), but also share characters with other magnoliids including Winteraceae. These fossils represent taxa that are character mosaics relative to currently recognized families. Inclusion of these fossils in existing data matrices and ensuing phylogenetic analyses effect changes in tree topologies consistent with their mosaicism relative to modern taxa. But such analyses do not definitively demonstrate the affinities of the fossils other than illustrating that these fossils are generalized magnoliids. Additional analysis of modern and fossil magnoliids is necessary to fully appreciate the phylogenetic significance and positions of these fossil taxa. However, the results of the phylogenetic analyses do introduce the possibility that extinct taxa of Magnoliales with cupulate floral receptacles were transitional between basal angiosperms and those with tricolpate pollen. The fossils provide insights into the timing of evolution of character complexes now associated with coleopteran pollination.

Key Words: angiosperms • Cretaceous • eudicots • Magnoliales

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS SYSTEMATICS DISCUSSION REFERENCES

 
Exceptionally rich deposits of angiosperm floral, fruit, and leaf remains from the Late Cretaceous of New Jersey are the subject of an ongoing study (e.g., Crepet et al., 1992; Herendeen, Crepet, and Nixon, 1993, 1994; Nixon and Crepet, 1993; Crepet and Nixon, 1994, 1996, 1997; Crepet, 1996; Gandolfo, Nixon, and Crepet, in press a, b; Nixon, Weeks, and Crepet, in press). These deposits include extremely well-preserved fossil flowers, fruits, seeds, and wood preserved by charcoalification or lignification. Many of the fossils retain microscopic details such as trichomes, epidermal characters, ovules/seeds within carpels, stigmatic surfaces, pollen, and in many cases complete preservation of anatomical features such as cell wall form and pitting. Thus far, these deposits have revealed the oldest flowers of hamamelidaceous affinity (Crepet et al., 1992), a suite of dilleniid taxa including the oldest fossil flowers of ericalean affinity (Nixon and Crepet, 1993; Nixon, Weeks, and Crepet, in press), those of capparalean affinity (Gandolfo, Nixon, and Crepet, 1998b) and those with affinities to Clusiaceae (Crepet and Nixon, 1998), a host of rosidean taxa (Crepet, 1996; Crepet and Nixon, 1996; Gandolfo, Nixon, and Crepet, 1998a), and a variety of taxa within Magnoliidae (e.g., Herendeen, Crepet, and Nixon, 1993, 1994; Crepet and Nixon, 1994).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS SYSTEMATICS DISCUSSION REFERENCES

 
Fossils were isolated from silty clay collected at the Old Crossman Clay Pit in Sayreville, New Jersey. The fossil-bearing deposits are in the Raritan or Lower Magothy Formation, and on the basis of pollen analysis are thought to represent the Amboy Fire Clay of Late Turonian Age (~90 MYBP; Christopher, 1979; Grimaldi, Beck, and Boon, 1989; G. J. Brenner, unpublished data). Fossils were isolated by dissolving the clay matrix in water and carefully pouring the fossil-bearing fraction through a series of sieves. After isolation, fossils were cleaned with hydrofluoric acid, washed, and then rinsed with distilled water and allowed to dry. They were then mounted for scanning electron microscopy (SEM). Scanning was accomplished with an AMR 1000 and Hitachi S4500 SEM. One specimen was dissected, and carpels were embedded in acrylic medium and sectioned transversely.

	SYSTEMATICS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS SYSTEMATICS DISCUSSION REFERENCES

 
Genus
Cronquistiflora Crepet and Nixon, gen. nov.

Type species
Cronquistiflora sayrevillensis Crepet and Nixon, sp. nov.

Generic diagnosis
Flowers bisexual, with an enlarged cup-shaped receptacle bearing imbricate bracts (or perianth). Vascular tissue of pedicels with vessels with scalariform perforation plates. Inner surface of receptacle glabrous, striate. Bracts auriculate. Stamens (presumably) attached to rim or receptacle, laminar, fleshy. Anthers tetrathecal, embedded, opening by longitudinal valves. Pollen monosulcate, exine apparently solid with microchannels. Carpels free, numerous, attached to the base of the cupulate receptacle; stigmas peltate, contiguous, forming a platform in the receptacular chamber. Seeds six or more in two marginal rows, distally winged, with a proximal flange or aril.

Description
Inflorescence and vegetative morphology unknown. Flowers on pedicels 0.43–0.87 mm thick; pedicel eustelic with limited secondary wood or radial metaxylem developed in the individual bundles, bundles near the periphery of the axis (cortical?). Vessels with scalariform oblique perforation plates with 10–12 bars and scalariform lateral pitting. Vessel length x width is ~117 x 9.75 µm. Floral receptacle range 1.3–4.3 mm x 1.4–3.0 mm, thickened, cup-like, somewhat campanulate, externally bearing spirally arranged, imbricate bracts (or perianth); inner surface smooth, somewhat striate with longitudinal grooves, glabrous. Bracts (perianth parts) range 0.52–1.8 mm in width (at bases, none whole for length measurement) ovate, auriculate, apices apiculate, adaxial surface glabrous, abaxial surface with tufts of simple trichomes near apex. Androecium attachment uncertain, presumed to be numerous free stamens attached in one or more rows along internal rim of receptacular cup. Stamens: one example with entirely preserved length is 1.1 mm, in other specimens width range is 0.62–1.4 mm, laminar, relatively thick and presumably fleshy, with a short flattened stalk, and a distally expanded ovate or sometimes obovate portion that may be incurved, glabrous except for a few simple trichomes near the apex on the abaxial surface (similar to those of the bracts); thecae four, in two pairs, elongate, embedded (presumably on the adaxial side [based on the direction of stamen curvature]), dehiscence by valves. Pollen boat-shaped (average 29 x 14 µm; range: 13–42 µm x 8.0–22 µm), monosulcate with occasional longitudinal folds, smooth perforate (foveolate), exine apparently solid (in SEM) with microchannels. Gynoecium of 49–55 free carpels with elongate hairs at the bases, attached in a tight spiral to the flattened or slightly concave base of the receptacular cup. Carpels ascidiate, clavate, with an incompletely sealed suture, epidermis pustulate, with large spherical subepidermal cells (observable in section). Stigma (range: 100–360 µm x 167–520 µm) polygonal-cordate/sagittate, flattened in a plane 90° from the longitudinal axis of the carpel and overhanging the carpel walls (peltate). Stigmatic epidermis papillate. A continuation of the carpel suture into the stigmatic surface does not quite bisect the stigma and has different epidermal texture from the remaining stigmatic surface. Ovules/seeds numerous (six or more), borne marginally in two rows, 0.30–0.36 mm long, elliptic, with distal flattened wing; a small flange that may be an aril also present on proximal part of seed. Tissues of stamens, pedicel, and carpels with large spherical cells.

Species
Cronquistiflora sayrevillensis Crepet and Nixon, sp. nov.

Specific diagnosis
as for the genus Cronquistiflora.

Holotype
L. H. Bailey Hortorium Paleobotanical Collection CUPC 1175 (Figs. 1–9).

View larger version (174K): [in this window] [in a new window]   Figs. 1–9. Cronquistiflora sayrevillensis. CUPC 1175. 1. Lateral view of a flower/fruit illustrating the pedicel and ascidiate carpels in a cupulate receptacle. Note the bracts on the pedicel and cupulate receptacle. CUPC 1175. Bar = 892 µm. 2. Top view of specimen CUPC 1175 showing the helical arrangement of the carpels. Bar = 408 µm. 3. Top view of several stigmas showing their bilateral symmetry. CUPC 1175. Bar = 119 µm. 4. Lateral view of carpels. CUPC 1175. Bar = 260 µm. 5. Two carpels illustrating the peltate stigmas and pustulate lateral walls. CUPC 1175. Bar = 119 µm. 6. Carpel bases with elongate hairs. CUPC 1175. Bar = 124 µm. 7. A broken carpel illustrating two rows of winged arillate seeds. CUPC 1175. Bar = 57 µm. 8. Several seeds removed from the carpels illustrating the arils surrounding the funicular ends of the seeds. CUPC 1177. Bar = 71 µm. 9. A high magnification view of part of the stigmatic surface. CUPC 1175. Bar = 18 µm.

Type locality
Old Crossman Clay Pit, Sayreville, New Jersey

Etymology
The fossil is named in honor of Arthur Cronquist.

Remarks
Cronquistiflora
All specimens of Cronquistiflora flowers/fruits (12 specimens) are variously incomplete. In each specimen, portions of the upper receptacular cup are missing, thus no specimens exist with attached stamens. Pollen occasionally found on the carpellate receptacles appears identical to pollen discovered in associated stamens in exine ornamentation, apertural configuration, and size. Also, these stamens were isolated from the same small samples of matrix that produced two of the carpellate receptacles and have not been isolated from other samples despite extensive sampling. For these reasons, and because the stamens bear trichomes similar to those on the cupulate receptacles, the detached stamens and carpel-bearing receptacles are presumed to represent the same taxon.

Cronquistiflora has a robust pedicel. The floral axis has a broad well-defined pith (Figs. 1, 10). It is eustelic with bract traces in the cortex. Vascular bundles may or may not be entirely separate with limited secondary wood development (Fig. 14). Vessels have scalariform lateral pitting and oblique end walls with 10–12 scalariform bars/perforation plate (Fig. 15).

View larger version (164K): [in this window] [in a new window]    Figs. 10–18. Cronquistiflora sayrevillensis. 10. An immature fruit or flower in lateral view. CUPC 1102. Bar = 476 µm. 11. Top view of the specimen illustrated in Fig. 10 . Bar = 357 µm. 12. Top view of several stigmas. CUPC 1102. Bar = 80 µm. 13. Close-up view of pedicel with bract attachment showing elongate hairs on the pedicel and abaxial bract. CUPC 1175. Bar = 143 µm. 14. Pedicel broken transversely to illustrate eustele bundles with limited radial xylem and large spherical cells in both the pith and cortex. CUPC 1171. Bar = 35 µm. 15. Longitudinally broken pedicel illustrating several vessels with scalariform wall thickenings and an oblique perforation plate. CUPC 1171. Bar = 10 µm. 16. Anther in presumed adaxial view illustrating the pollen sacs with valvate dehiscence and elongate hairs on the distal abaxial surface. CUPC 1179. Bar = 130 µm. 17. Another associated stamen illustrating open pollen sacs with rolled valves and distal abaxial elongate hairs. CUPC 1180. Bar = 238 µm. 18. A close-up of a single pollen sac illustrating the rolled valves. CUPC 1180. Bar = 89 µm.

Associated stamens are ovate and auriculate (Figs. 16, 17). There are four adaxial sporangia and these dehisce by longitudinal valves (Figs. 17, 18). The connective extensions are incurved and there are elongate hairs on the distal abaxial parts of the stamens (Figs. 16, 17, 19). These are similar to those found on the pedicel, receptacular bracts and at the bases of the carpels (Figs. 6, 13). Pollen (20 µm in diameter) is sparsely preserved within the anthers and is monosulcate and boat-shaped with occasional longitudinal folds (Figs. 20–23). Pollen exine is smooth foveolate with scabrate/rugulate sculpturing at the ends of the grains (Figs. 21–23). The pollen wall appears atectate in SEM (Fig. 22). Variation in the width of the aperture is suggestive of the phenomenon observed by Ward, Doyle, and Hotton (1989) where in dispersed monosulcate grains, the aperture margin rolls up with the aperture membrane somewhat in the fashion of a window shade (Fig. 23).

View larger version (187K): [in this window] [in a new window]   Figs. 19–23. Cronquistiflora sayrevillensis. 19. Distal abaxial surface of an anther illustrating the elongate hairs concentrated in that area. CUPC 1180. Bar = 95 µm. 20. Pollen grains in situ. CUPC 1179. Bar = 8 µm. 21. A single pollen grain showing the sulcus and exine. CUPC 1179. Bar = 3 µm. 22. Several broken pollen grains showing exine ornamentation (smooth with microperforations), and the atectate (solid) wall structure. CUPC 1181. Bar = 7 µm. 23. A single pollen grain illustrating how the margins of the sulcus are sometimes inrolled. CUPC 1179. Bar = 2 µm. Figs. 24–26. Detrusandra mystagoga. CUPC 1184. 24. Top view of a flower that retains several obovate tepals. Note the adaxial indumentum of elongate hairs. Bar = 408 µm. 25. Lateral view of flower. Note the cupular rim and narrow bases of the obovate tepals. Bar = 571 µm. 26. Distal part of flower split longitudinally showing the overlapping tepals, stamens with adaxial pollen sacs, and stamen attachment near the rim of the cupule. Bar = 571 µm.

Type species
Detrusandra mystagoga Crepet and Nixon

Generic diagnosis
Flowers bisexual, with an enlarged cupulate receptacle bearing imbricate bracts (or perianth). Vascular tissue of pedicels with vessels with scalariform perforation plates. Inner surface of receptacle glabrous, smooth. Bracts auriculate. Several imbricate tepals borne on rim of cupule. Stamens (presumably) attached near rim of receptacle, laminar, fleshy, incurved. Anthers tetrathecal, adaxially embedded, opening by longitudinal valves. Pollen spheroidal without well-defined apertures, but with sulcoid areas, exine tectate columellate, reticulate. Pistillodes (?) present to outside of carpels. Carpels free, usually about five, attached to the base of the cupulate receptacle; lacking associated trichomes, stigmas bilobed, not peltate, not contiguous. Seeds more than six, in two marginal rows, distally winged, with a proximal flange or aril.

Generic description of Detrusandra
Inflorescence and vegetative morphology unknown. Flowers on pedicels 0.31–0.97 mm in diameter. Pedicels bearing spirally arranged bracts. Floral receptacle range: 1.0–2.5 mm x 1.1–3.2 mm, campanulate, externally bearing spirally arranged imbricated bracts, width range = 600–628 µm. The extremity of the cupule bears several expanded ovate tepals, range = 0.9–3 mm x 0.6–2.4 mm, which invest the flower and are hairy on the abaxial surfaces. Upon abscising they leave a cupulate rim. Inner receptacle surface smooth, bearing no appendages between the zone where stamens are attached and the cupular base. Androecium attachment restricted to the upper part of the internal surface of the cupulate receptacle. Stamens (range: 0.68–3.1 mm x 0.28–0.94 mm) sharply incurved, laminar, ovate-lanceolate, spirally arranged, and numerous. Stamens with an extensive connective extension and four adaxial sporangia with valvate dehiscence. Pollen spherical (17 µm average diameter; range: 16–26 µm), with no well-defined aperture, but with occasional sulcoid areas, tectate columellate, micromorphology reticulate. Gynoecium composed of dorsiventrally flattened structures (pistillodes?) (range: 328–832 µm x 100–266 µm) spirally arranged around functional carpels. Carpels (range: 840–926 µm x 185–235 µm) relatively few (five) without elongate hairs at the bases, whorled or arranged in a very low spiral and ascidiate with bilobed distal stigmas (range: 50–180 µm x 90–200 µm). Distal pistillode epidermal cells often having collapsed periclinal walls. Ovules/seeds (range: 116–175 µm x 52–83 µm) more than six, with proximal flanges (arillate), winged, and borne marginally in two rows. Tissues with large spherical cells, presumably ethereal oil cells.

Species
Detrusandra mystagoga Crepet and Nixon, sp. nov.

Etymology
mystagoga, priest who showed strangers the mystery of the temple

Specific diagnosis
as for the genus Detrusandra

Holotype
L. H. Bailey Paleobotanical Collection CUPC 1188 (Fig. 27).

View larger version (163K): [in this window] [in a new window]    Figs. 27–38. Detrusandra mystagoga. CUPC 1188. 27. Lateral view of a flower missing tepals, but illustrating cupule bract scars and stamen attachment. Bar = 571 µm. 28. Top view illustrating the cupular rim with attachment scars of tepals and the incurved laminar stamens. Bar = 571 µm. 29. Specimen CUPC 1188, split longitudinally, illustrating the attachment of the incurved stamens to the cupulate receptacle near the rim and the pistillodes and pistils at the base. Bar = 571 µm. 30. Close-up of cupulate receptacle wall illustrating stamen attachment and pustular stamen tissue. Bar = 286 µm. 31. Part of distal cupulate receptacle with attached anthers. Note the anther shape and adaxial anther tissues. Bar = 408 µm. 32. Close-up of cupular rim illustrating stamen attachment and broken bases of tepals. Bar = 204 µm. 33. Adaxial view of a stamen removed from a flower. Note the elongate pollen sacs and the pustulate stamen tissue. Bar = 357 µm. 34. Close-up of unopened pollen sacs covered by valves and unevenly filled with pollen. Bar = 95 µm. 35. In situ pollen grains showing folding, irregular apertures and tectate columellate exine. Bar = 7 µm. 36. Close-up of a single pollen grain illustrating a weakly developed aperture and perforate tectum. Bar = 0.95 µm (or 950 nm). 37. Longitudinally broken flower illustrating pistillodes and five pistils at the base of the receptacle beneath incurved stamen extremities. Bar = 238 µm. 38. Top view of whorl of five carpels showing four complete bilobed stigmas and part of a fifth stigma (all five can be observed in Fig. 37 at lower magnification). Note dorsiventrally flattened ?pistillodes with collapsed distal periclinal walls surrounding the pistils. Bar = 57 µm.

Type locality
Old Crossman Clay pit, Sayreville, New Jersey

Detrusandra is represented by at least nine fossil flowers/fruits. Some of these are preserved almost in their entirety (Figs. 25, 27) with the exception that there are no examples with completely preserved receptacular bracts. The cupulate receptacle of Detrusandra is narrower basally than that of Cronquistiflora and is campanulate/funnelform in shape (Figs. 27, 29). The receptacle bears helically arranged bract bases on the outer surface (Fig. 27). In contrast with the fossils of Cronquistiflora, there are no hairs on the pedicel or at the bases of the bracts (Fig. 27).

The cupulate receptacle of Detrusandra terminates in a definitive rim that bears a tight spiral of ovate perianth parts (Figs. 24–26, 32). The perianth segments expand dramatically so that each one invests about one-third of the area of the floral surface (Fig. 24). There is a notable abaxial indumentum of elongate hairs on the perianth parts (Fig. 24). Laminar lanceolate stamens are attached internally to the rim of the cupulate receptacle (Figs. 30–32). Stamens incurve sharply with their distal tips inserted between the pistillodes and receptacular wall at the base of the flowers (Figs. 29, 30). There are four adaxial pollen sacs approximately in the middle of the stamen that are closed by valves (Figs. 33, 34). Pollen sacs are sometimes unevenly filled (Fig. 34). Connective tissue of the stamens is replete with large spherical cells suggestive of ethereal oil cells (Fig. 33). Pollen is copiously preserved in some anther sacs but is collapsed, suggesting that it is thin-walled (Fig. 35). Pollen is small (average diameter = 17 µm) with a reticulate exine and apparent tectate columellate infrastructure (Figs. 35, 36). There are no well-defined apertures, but there are occasional smooth areas that appear to be sulcoid in nature (Fig. 36).

There are no appendages between the attachment of the stamens near the rim of the receptacle and the sterile structures surrounding the carpels at the base of the receptacle (Figs. 29, 30). These sterile structures are dorsiventrally flattened and often have a weakly developed abaxial keel (Fig. 37). In the distal epidermis of these structures (? pistillodes) the relatively large epidermal cells are almost uniformly preserved without intact periclinal walls (Fig. 38). Within the pistillodes is a whorl or tight spiral of five carpels. These are conduplicate with bilobed stigmas (Figs. 37, 38). There are two rows of at least six winged seeds borne in each carpel.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS SYSTEMATICS DISCUSSION REFERENCES

 
Affinities
Although superficially similar to the flowers of several modern magnoliid taxa, Cronquistiflora and Detrusandra have unique combinations of features relative to the flowers of living magnoliids, known fossil magnoliids, and to one another. They clearly represent extinct taxa (Table 1; Cronquist, 1981; Kubitzki, Rohwer, and Bittrich, 1993; Takhtajan, 1997). The fossils do share numerous characters with several genera of modern magnoliids having cupulate receptacles (i.e., Eupomatia, Calycanthus, Idiospermum, and genera in the families Himantandraceae and Monimiaceae) as well as sharing characters with noncupulate magnoliids (e.g., some Winteraceae). For example, Cronquistiflora has much in common with flowers of extant Eupomatia bennettii (Eupomatiaceae, Magnoliales sensu Cronquist, 1981; Takhtajan, 1997). This similarity is supported by several characters, including the cup-like receptacle, broad ovate stamens with adaxial sporangia, numerous spirally arranged carpels each with several marginally attached seeds. However, Cronquistiflora had external bracts on its cup-like receptacle in contrast to the naked receptacles of extant Eupomatia (e.g., Cronquist, 1981; Takhtajan, 1997), and Cronquistiflora does not have zonasulculate pollen. Similarities between Cronquistiflora and modern Eupomatia are also supported by the arrangement, shape, and epidermal features (especially in the stigmatic regions) of the carpels. Again, there are differences. In Eupomatia, but not in Cronquistiflora, the carpels are reported to be basally connate and are closely appressed in the fruiting stage, although they separate easily at maturity (direct observation of herbarium specimens). Cronquistiflora also differs from Eupomatia bennettii in not having secretory tufts on the stamens (Endress, 1984).

Because Cronquistiflora and Detrusandra share numerous characters with various families of Magnoliidae, it is difficult to associate the fossils with any particular extant families on the basis of raw character similarity. The fossils show mosaics of characters now distributed primarily, but not exclusively, among different genera and different families of modern Magnoliales and Laurales (sensu Cronquist, 1981; Kubitzki, Rohwer, and Bittrich, 1993; Takhtajan, 1997; Table 1) or, by other authors, all included within Magnoliales (Endress, 1990). Fossils that show mosaics of characters relative to modern taxa present an interesting challenge. As with all such fossils, their mosaicism makes them hard to place taxonomically without a high degree of subjective interpretation. At the same time, they potentially present phylogenetic opportunities and are attractive candidates for phylogenetic analysis. One might hope that such an approach could resolve the difficulties noted above by objectively placing the fossil taxa in a hierarchy with modern taxa based on synapomorphies. Of course, reciprocally, the addition of fossil data to cladistic analyses may change the results of analyses to favor different phylogenetic hypotheses and, thus, fossil data may directly affect our conclusions about phylogenetic history (Patterson, 1981; Donoghue et al., 1989). However, successful taxonomic placement of the fossils through phylogenetic analyses depends on the existence (or creation) of an appropriate data matrix and also depends on having an appropriate number of characters available from the fossil taxon relative to those in any data matrix representing living taxa. Problems inherent in the inclusion of fossil taxa in cladistic analyses include excessive missing values and difficulty in homology assessment (see Nixon, 1996). The fossils presented here are represented only by floral morphology and so lack character data for all vegetative characters, as well as the chemical and cytogenetic characters often included in "morphological" data sets. However, because of the exquisite preservation of the floral parts that are available, homology assessment of the available characters is not hampered to the extent that it usually is in compression and impression fossils, where details of structures are sometimes obscure.

Detrusandra and Cronquistiflora were added as terminals to matrices from three separate published data sets. That of Lammers, Stuessy, and Silva (1986) included a relatively high proportion of floral characters, but proved to be too small a matrix to be very stable or informative when the fossil taxa and attendant missing values were incorporated into the analysis. Another matrix that also included a high number of reproductive characters was that of Donoghue and Doyle (1989). This matrix included some tricolpate pollen-bearing taxa in addition to magnoliids. We also used a slightly modified seed plant data matrix based on the published matrix of Loconte and Stevenson (1991). The computer program NONA (Goloboff, 1993) was used to analyze these augmented published matrices. Searches for most parsimonious trees were implemented using 100 randomized taxon entry sequences and TBR (tree bisection reconnection) branch swapping holding 20 starting trees for each iteration (hold/20; mult*100), followed by complete TBR swapping of all shortest trees found in the initial 100 analyses (max*). A strict consensus was calculated for each set of most parsimonious trees ("nelsen"). Trees were saved and then printed using Clados (Nixon, 1994).

We analyzed the Doyle and Donoghue (1989) matrix, as originally published, and rooted with a hypothetical basal angiosperm (ANC) in order for our results to be comparable. Problems with this approach have been pointed out by various authors (see Nixon and Carpenter, 1993; Nixon et al., 1994; Nixon, 1996). We ran the matrix with one fossil taxon at a time and with both fossil taxa simultaneously. The results of these three analyses differed. With Cronquistiflora only, the analysis produced five most parsimonious trees of 178 steps (CI [Consistency Index] = 38, RI [Retention Index] = 59). It is interesting to note that there were 27 equally parsimonious trees (178 steps; CI 38, RI 58) in the analysis of the data matrix without the fossil (Fig. 39) and that the consensus tree including the fossil Cronquistiflora (Fig. 40) was much more resolved than the consensus tree based only on the Donoghue and Doyle matrix of 1989 (Fig. 39 vs. Fig. 40). In the analysis that includes Cronquistiflora, the fossil is sister group to a clade composed of Eupomatiaceae and 21 other families (Fig. 40). Although Cronquistiflora did not group directly with Eupomatiaceae, and Calycanthaceae, patristically it is relatively close to these two families, suggesting similarity due to retained plesiomorphic features, as manifested in the cup-like floral receptacles of these taxa. This realignment of taxa within Magnoliidae raises the possibility that angiospermous taxa with tricolpate pollen (or tricolpate-derived pollen) might have been derived from an ancestral plexus of taxa that were united by an axial or otherwise derived cup-shaped receptacle.

View larger version (14K): [in this window] [in a new window]   Fig. 39. The consensus tree of 27 equally parsimonious trees based on the Donoghue and Doyle (1989) matrix analyzed without Cronquistiflora or Detrusandra.

View larger version (17K): [in this window] [in a new window]   Fig. 40. The consensus tree of five most parsimonious trees generated by the addition of Cronquistiflora to the Donoghue and Doyle (1989) matrix.

View larger version (15K): [in this window] [in a new window]   Fig. 41. The consensus tree of 16 most parsimonious trees resulting from the inclusion of Detrusandra alone in the Donoghue and Doyle (1989) matrix.

View larger version (16K): [in this window] [in a new window]   Fig. 42. Consensus tree of nine most parsimonious trees generated by including both Cronquistiflora and Detrusandra in the analysis of the Donoghue and Doyle (1989) matrix.

View larger version (36K): [in this window] [in a new window]   Fig. 43. Consensus tree of 12 most parsimonious trees based on the analysis of the Loconte and Stevenson (1991) matrix including Cronquistiflora. See text for details.

View larger version (36K): [in this window] [in a new window]   Fig. 44. Consensus tree of 32 most parsimonious trees based on the analysis of the Loconte and Stevenson (1991) matrix including Detrusandra. See text for details.

View larger version (40K): [in this window] [in a new window]   Fig. 45. Consensus tree of 51 most parsimonious trees based on the analysis of the Loconte and Stevenson (1991) matrix including both Cronquistiflora and Detrusandra. See text for details.

Pollination biology of Eupomatiaceae is complex and involves a coordinated sophisticated set of structural, chemical, and, on the part of the beetle pollinators, complicated and specific, interactions (e.g., Armstrong and Irvine, 1990). Whether or not the fossils described here had similar syndromes of pollination is to some extent a matter of conjecture. The fossils, however, do share many of the structural features found in modern Eupomatiaceae and the presence of apparent ethereal oil cells suggests that there may have been chemical similarities as well. Curculionidae include the pollinators of extant Eupomatia (the genus Elleschodes; Armstrong and Irvine, 1990) and the curculionids are diverse by the Upper Cretaceous, as well as being represented in the same sediments as these fossil flowers (unpublished data). Minimally, the fossil taxa suggest that structural adaptations consistent with beetle pollination had evolved in Magnoliidae at least by the Turonian.

Paleoclimate
The fossil locality in Sayreville, New Jersey, was in southern Laurasia in a low-middle latitudinal position on the north side of the warm Tethyan Sea at the time of deposition of the fossil taxa described here (Parrish, Curtis, and Barron, 1982). Paleoclimatic data (Bowen, 1966), including leaf margin analyses (Upchurch and Wolfe, 1987), suggest that the climate was tropical or subtropical. In contrast, data on the carbon isotope content in Cenomanian to Lower Turonian marine sediments suggest the possibility that this interval was characterized by global climatic cooling (Arthur, Dean, and Pratt, 1988). The interpretation of the paleoclimate as subtropical to tropical is consistent with the affinities of other fossil taxa now known from this locality to modern taxa with largely tropical extant distributions (e.g., Crepet et al., 1992; Herendeen, Crepet, and Nixon, 1993, 1994; Nixon and Crepet, 1993; Crepet and Nixon, 1994, 1996, 1998; Crepet, 1996; Gandolfo et al., 1997; Gandolfo, Nixon, and Crepet, 1998a, b; Nixon, Weeks, and Crepet, in press).

Finally, it should also be noted that a majority of fossils from the New Jersey Turonian site are exceptionally small relative to flowers of related modern taxa. It is not clear whether or not this phenomenon is due to bias for smaller objects imposed by depositional sorting. Certainly, uniform shrinkage is associated with the phenomenon of charcoalification, but only to the extent of 30–50% (Lupia, 1995). Thus, even allowing for estimates of preservation mode associated shrinkage, these fossils probably represent flowers that were extremely small in life. If these general sizes are indeed reflective of the true dimensions of the Cretaceous taxa, they may have some functional implications, perhaps involving pollinator relationships. It is interesting to consider whether increasing size of the flowers through time might also have been related somehow to their evolutionary relationships with pollinators.

	FOOTNOTES

 
¹ The authors thank Jennifer Svitko and John Freudenstein (Kent State University) for SEM and laboratory preparations, and wish to acknowledge thoughtful comments and suggestions from Peter Endress, University of Zurich, Ed Schneider, Santa Barbara Botanic Garden, and Alejandra Gandolfo, Cornell University. Research supported by NSF grant DEB9420512.

² Author for correspondence.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS SYSTEMATICS DISCUSSION REFERENCES

 
Armstrong, J. E., and A. K. Irvine.1990.Functions of staminodia in beetle-pollinated flowers of Eupomatia laurina. Biotropica 22: 429–431.

Arthur, M. A., W. E. Dean, and L. M. Pratt.1988.Geochemical and climatic effects of increased marine organic carbon burial at the Cenomanian/Turonian boundary. Nature 335: 714–717. [CrossRef]

Bowen, R.1966.Paleotemperature analysis. Elsevier, Amsterdam.

Christopher, R. A.1979.Normapolles and triporate pollen assemblages from the Raritan and Magothy Formations (Upper Cretaceous) of New Jersey. Palynology 3: 73–121. [Abstract]

Crane, P. R., and D. L. Dilcher.1984.Lesqueria: an early angiosperm fruiting axis from the mid-Cretaceous. Annals of the Missouri Botanical Garden 71: 384–402. [CrossRef][ISI]

———, E. M. Friis, and K. R. Pedersen.1994.Palaeobotanical evidence on the early radiation of magnoliid angiosperms. In P. K. Endress and E. M. Friis [eds.], Early evolution of flowers. Plant Systematics and Evolution, Supplementum 8: 51–72.

———, ———, and ———.1995.The origin and early diversification of angiosperms. Nature 374: 27–33. [CrossRef]

Crepet, W. L.1996.Timing in the evolution of derived floral characters: Upper Cretaceous (Turonian) taxa with tricolpate and tricolpate-derived pollen. Review of Palaeobotany and Palynology 90: 339–360.

———, and K. C. Nixon.1994.Flowers of Turonian Magnoliidae and their implications. In E. M. Friis and P. K. Endress [eds.], The early fossil record of angiosperm flowers. Plant Systematics and Evolution 8: 73–91.

———, and ———.1996.The fossil history of stamens. In W. G. D'Arcy and Richard C. Keating [eds.], The anther, form, function and phylogeny, 25–57. Cambridge University Press, Cambridge.

———, and ———.1998.Fossil Clusiaceae from the Late Cretaceous (Turonian) of New Jersey and implications regarding the history of bee pollination. American Journal of Botany 85; in press.

———, ———, E. M. Friis, and J. V. Freudenstein.1992.Oldest fossil flowers of hamamelidaceous affinity, from the Late Cretaceous of New Jersey. Proceedings of the National Academy of Sciences, USA 89: 8969–8989.

Cronquist, A.1981.An integrated system of classification of flowering plants. Columbia University Press, New York, NY.

Dilcher, D. L.1979.Early angiosperm reproduction: an introductory report. Review of Palaeobotany and Palynology 27: 291–328. [CrossRef]

———, and P. R. Crane.1984.Archaeanthus: an early angiosperm from the Cenomanian of the Western Interior of North America. Annals of the Missouri Botanical Garden 71: 351–383. [CrossRef][ISI]

Donoghue, M. J., and J. A. Doyle.1989.Phylogenetic analysis of angiosperms and the relationships of the Hamamelididae. In P. R. Crane and S. Blackmore [eds.], Evolution, systematics and fossil history of the Hamamelididae. I. Introduction and "lower" Hamamelididae, 17–45. Clarendon, Oxford.

———, ———, J. Gauthier, A. Kluge, and T. Rowe.1989.The importance of fossils in phylogenetic reconstructions. Annual Review of Ecology and Systematics 20: 431–460.

Endress, P. K.1984.The role of inner staminodes in the floral display of some relict Magnoliales. Plant Systematics and Evolution 146: 269–282. [CrossRef][ISI]

———.1990.Evolution of reproductive structures and functions in primitive angiosperms (Magnoliidae). Memoirs of the New York Botanical Garden 55: 5–34.

Friis, E. M., H. Eklund, K. R. Pedersen, and P. R. Crane.1994.Virgininanthus calycanthoides gen. et sp. nov.—a calycanthaceous flower from the Potomac Group (Early Cretaceous) of eastern North America. International Journal of Plant Sciences 155: 772–785. [CrossRef]

Gandolfo, M. A., K. C. Nixon, and W. L. Crepet.1998a.Tylerianthus crossmanensis gen. et sp. nov. (Rosales) from the Upper Cretaceous of New Jersey. American Journal of Botany 85: 376–386. [Abstract]

———, ———, and ———.1998b.A new fossil flower from the Turonian of New Jersey: Dressiantha bicarpellata gen. et sp. nov. (Capparales). American Journal of Botany 85: 964–974. [Abstract]

———, ———, ———, and G. E. Ratcliffe.1997.A new fossil Gleicheniaceae from Late Cretaceous sediments of New Jersey. American Journal of Botany 84: 483–493. [Abstract]

Goloboff, P.1993.NONA (Parsimony program for IBM compatibles and SUN workstations). Published by the author, Buenos Aires, Argentina.

Grimaldi, D., C. W. Beck, and J. J. Boon.1989.Occurrence, chemical characteristics, and paleontology of the fossil resins from New Jersey. American Museum Novitates 2948: 1–18.

Herendeen, P., W. L. Crepet, and K. C. Nixon.1993.Chloranthus-like stamens from the Upper Cretaceous of New Jersey. American Journal of Botany 80: 865–871.

———, ———, and ———.1994.Fossil flowers and pollen of Lauraceae from the Upper Cretaceous of New Jersey. Plant Systematics and Evolution 189: 29–40. [CrossRef][ISI]

Kubitzki, K., J. G. Rohwer, and V. Bittrich [eds.].1993.The families and genera of flowering plants II. Springer, Berlin.

Lammers, T. G., T. F. Stuessy, and M. Silva.1986.Systematic relationships of Lactoridaceae, an endemic family of the Juan Fernandez Islands, Chile. Plant Systematics and Evolution 152: 243–266. [CrossRef][ISI]

Loconte, H., and D. W. Stevenson.1991.Cladistics of the Magnoliidae. Cladistics 7: 267–296.

Lupia, R.1995.Paleobotanical data from fossil charcoal: an actualistic study of seed plant reproductive structures. Palaios 10: 465–477. [Abstract/Free Full Text]

Nixon, K. C.1994.Clados. Program and Documentation. Published by the author, Trumansburg, NY.

———.1996.Paleobotany in cladistics and cladistics in paleobotany: enlightenment and uncertainty. Review of Palaeobotany and Palynology 90: 361–373.

———, and J. M. Carpenter.1993.On outgroups. Cladistics 9: 413–426. [CrossRef][ISI]

———, and W. L. Crepet.1993.Late Cretaceous fossil flowers of Ericalean affinity. American Journal of Botany 80: 616–623. [CrossRef][ISI]

———, and Q. D. Wheeler.1992.Extinction and the origin of species. In M. J. Novacek and Q. D. Wheeler [eds.], Extinction and phylogeny, 119–143. Columbia University Press, New York, NY.

———, W. L. Crepet, D. W. Stevenson, and E. M. Friis.1994.A reevaluation of seed plant phylogeny. Annals of the Missouri Botanical Garden 81: 484–533. [CrossRef][ISI]

———, A. Weeks, and W. L. Crepet.In press.Teuchestanthes squamata, a new fossil flower with ericalean affinities from the mid-Cretaceous of New Jersey. Plant Systematics and Evolution.

Patterson, C. R.1981.Significance of fossils in determining evolutionary relationships. Annual Review of Ecology and Systematics 12: 195–223.

Parrish, J. T., R. L. Curtis, and E. J. Barron.1982.Atmospheric circulation, upwelling, and organic rich rocks in the Mesozoic and Cenozoic Eras. In J. T. Parrish [ed.], Palaeogeography, palaeoclimatology, palaeoecology, 31–66. Elsevier, Amsterdam.

Takhtajan, A. L.1997.Diversity and classification of flowering plants. Columbia University Press, New York, NY.

Upchurch, G. R., Jr., and J. A. Wolfe.1987.Mid-Cretaceous to Early Tertiary vegetation and climate: evidence from fossil leaves and wood. In E. M. Friis, W. G. Cahloner, and P. R. Crane [eds.], The origins of angiosperms and their biological consequences, 75–106. Cambridge University Press, Cambridge.

Ward, J. V., J. A. Doyle, and C. L. Hotton.1989.Probable granular magnoliid angiosperm pollen from the Early Cretaceous. Pollen et Spores 31: 113–132.

Ying, T., Y. Zhang, and D. E. Bufford.1993.The endemic genera of seed plants of China. Science Press, Beijing.

This article has been cited by other articles:

				W. L. Crepet, K. C. Nixon, and M. A. Gandolfo An extinct calycanthoid taxon, Jerseyanthus calycanthoides , from the Late Cretaceous of New Jersey Am. J. Botany, September 1, 2005; 92(9): 1475 - 1485. [Abstract] [Full Text] [PDF]

				W. L. Crepet, K. C. Nixon, and M. A. Gandolfo Fossil evidence and phylogeny: the age of major angiosperm clades based on mesofossil and macrofossil evidence from Cretaceous deposits Am. J. Botany, October 1, 2004; 91(10): 1666 - 1682. [Abstract] [Full Text] [PDF]

				P. R. Crane, P. Herendeen, and E. M. Friis Fossils and plant phylogeny Am. J. Botany, October 1, 2004; 91(10): 1683 - 1699. [Abstract] [Full Text] [PDF]

				M. A. Gandolfo, K. C. Nixon, and W. L. Crepet Cretaceous flowers of Nymphaeaceae and implications for complex insect entrapment pollination mechanisms in early Angiosperms PNAS, May 25, 2004; 101(21): 8056 - 8060. [Abstract] [Full Text] [PDF]

				M. A. Gandolfo, K. C. Nixon, and W. L. Crepet Triuridaceae fossil flowers from the Upper Cretaceous of New Jersey Am. J. Botany, December 1, 2002; 89(12): 1940 - 1957. [Abstract] [Full Text] [PDF]

				D. Dilcher Toward a new synthesis: Major evolutionary trends in the angiosperm fossil record PNAS, June 20, 2000; 97(13): 7030 - 7036. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (27) Citing Articles via Google Scholar