Abstract

A pollen and spore assemblage of 50 species was recovered from the late Paleocene (pollen zone P5) Almont locality in the Williston Basin, central North Dakota, USA. This palynoflora was extracted from the same layer containing a diverse megaflora preserved in a silicified shale with compressed leaves, and anatomically preserved fruits and seeds. More than 44 megafossil genera assignable to 26 extant plant families thus far have been recognized. The palynomorphs, which are of exceptional preservation, were examined using the same-grain technique with both light microscopy (LM) and scanning electron microscopy (SEM). Additional LM and SEM studies augmented the same grain studies to provide an understanding of sculptural features, and additional, rare taxa. Of particular note are the in situ pollen types known from catkins and pollen cones, allowing for confirmation of the dispersed pollen's systematic position by tying it to its parent plant. Taxa for which in situ pollen is known from Almont include taxodiaceous conifers, Betulaceae, Hamamelidaceae, Juglandaceae, and Platanaceae, and several catkins of uncertain affinities, some with monosulcate grains. This study emphasizes the role of palynology in providing an expanded view of the flora from palynomorphs for comparison with a rich megafossil assemblage.

1. Introduction

The Almont flora of the Williston Basin, central North Dakota, USA is one of the most informative paleobotanical sites of the late Paleocene (Pigg and DeVore 2010). It contains a diverse megafossil assemblage of silicified, compressed leaves along with anatomically preserved fruits, seeds, and inflorescences. The first plant organs described from Almont were fruits of Cyclocarya brownii (Juglandaceae; Manchester and Dilcher 1982). This study was followed by a preliminary treatment of the megafossils in which over 35 types of leaves and reproductive structures were described (Crane et al. 1990). Subsequent collection and ongoing studies have resulted in revision of some of these taxa, as well as recognition and establishment of new forms, such that the megafossil flora now has 44 genera assignable to 29 extant families (Table 1), along with additional unknown forms.

Among formally recognized taxa are representatives of Betulaceae (Manchester et al. 2004), Cornales (Cornaceae/Nyssaceae, Manchester et al. 1999; Manchester 2002a; Manchester and Hickey 2007), Hamamelidaceae (Benedict et al. 2008), Icacinaceae (Pigg et al. 2008a), Juglandaceae (Manchester and Dilcher 1982; Taylor et al. 2007), Myrtaceae (Pigg et al. 1993), Nymphaeaceae (Chen et al. 2004; Taylor et al. 2006), Polygalaceae (Pigg et al. 2008b), Polygonaceae (Manchester and O'Leary 2010), Ranunculaceae (Pigg and DeVore 2005), Sapindaceae (Manchester 2001), Toricelliaceae (Manchester et al. 2009), Trochodendraceae (Crane et al. 1991), and Porosia (Manchester 2002b). Additional megafossils under study include Actinidiaceae (Pigg and DeVore 2003), Cornaceae (Cornus, Xiang et al. 2003), Ochnaceae (Ochna, Pigg et al. 2005), Sapindaceae (Acer, Kittle et al. 2005), Zingiberales (Spirematospermum, Benedict et al. 2007) and other unknown forms (Crane et al. 1990; Pigg and DeVore 2010).

Along with leaves, stems, fruits, cones, seeds, and other reproductive structures, the Almont matrix contains a diverse palynomorph assemblage. Several taxa, including taxodiaceous conifers, Betulaceae, Hamamelidaceae, and Platanaceae are known in situ within megafossil cones and catkins (Crane et al. 1990; Manchester et al. 2004; Benedict et al. 2008). Preliminary studies of the palynomorph flora obtained from the same matrix as the megafossils demonstrated the high diversity and excellent level of preservation of this pollen and spore assemblage (Manchester 1989; McClain and Manchester 1999). In this study we provide a comprehensive survey of the Almont palynoflora. Taxa described in this study establish an important basis for comparison with other floras of similar age and for expanding upon the Almont taxa known from megafossils. In addition, the deposit presents a rare opportunity to detail both the megaflora and microflora, with taxa being known from only one, or the other, or both sources (Table 1).

Earlier treatments of Paleocene palynofloras from the North American Western Interior and Rocky Mountains, based on light micrographs (LM), were particularly focused on stratigraphic questions, for example defining the K/T boundary (Norton and Hall 1969; Leffingwell 1970; Tschudy 1970; Hotton 2002). Several studies emphasized the value of pollen for establishing a biostratigraphic framework for the Paleocene (for example, Leffingwell 1970; Nichols and Ott 1978, 2006; Nichols and Brown 1992; Nichols 1994; Pocknall and Nichols 1996). Recently, Nichols and Ott (2006) formalized the zonation they proposed in 1978. In the current study we are using this previously defined biostratigraphic control (Nichols and Ott 1978, 2006).

Palynofloras also have been used to examine floristic change across the Initial Eocene Thermal Maximum (IETM). The most detailed studies taking a floristic approach are of the Bighorn Basin of Wyoming, USA (Wing and Harrington 2001; Wing et al. 2005). The most distinctive change across the Paleocene–Eocene boundary documented by this study was a decrease in abundance of Caryapollenites spp. and Polyatriopollenites vermontensis (Juglandaceae), and an increase in abundance in palynomorphs with affinities to taxodioid Cupressaceae, Betulaceae and Ulmaceae. Wing and Harrington (2001) found that the combined richness from 31 late Paleocene samples was 73 taxa. The present study documents 50 taxa of pollen and spores from a single site.

The focus of the present study is on the taxonomic diversity and systematic implications of both dispersed palynomorphs and in situ pollen found within pollen cones and catkins present in the Almont Lagerstätte. In this study, the same-grain technique is used to evaluate each individual pollen grain or spore with both light (LM) and scanning electron microscopy (SEM) (Zetter 1989). The additional information available from SEM, in particular provides the potential for additional taxonomic characters that may help in resolving the affinities of forms known primarily from sporae dispersae taxa.

2. Geologic setting

The Almont site is located within the Williston Basin, about 55 km from the eastern margin, in Morton County, central North Dakota, USA (figure 1; Crane et al. 1990; Bluemle 2000). Fossiliferous deposits are recovered from shallowly excavated pits in agricultural fields. The matrix is a silicified nonmarine shale found within the Sentinel Butte Formation of the Fort Union Group. Freshwater molluscs and mammals provide the basis for regional correlation as late Paleocene (Tiffanian 3; Kihm and Hartman 1991). Although the stratigraphic section is not clearly exposed at the Almont site, silicified shales of nearly identical lithology, containing the same suites of megafossils and pollen are exposed in stratigraphic sequence in the Beicegel Creek area, MacKenzie County, North Dakota, 120 km to the west (figure 2; Manchester et al. 2004; Pigg and DeVore 2005, 2010). Both sites fall within the late Paleocene Pollen Zone P5 of Nichols and Ott (Nichols 1998). The preliminary Almont pollen study by McClain and Manchester (1999) found 15–18 morphotypes, with pollen representing two additional families (Buxaceae and Pinaceae) that were unknown at the time from megafossils.

Initially described as lacustrine, perhaps representing a small lake or pond (Crane et al. 1990), more recent study suggests that the Almont site may be part of a larger fluvial system (Pigg and DeVore 2010). The Almont matrix itself shows no large-scale sedimentary structures (e.g. cross-beds) that would indicate a major fluvial source, however, this site could represent an oxbow lake that is part of a meandering stream (Figure 2).

3. Materials and methods

Individual palynomorphs were examined both in transmitted light microscopy (LM) and scanning electron microscopy (SEM; Zetter 1989). Surface sediment was removed to prevent contamination by recent pollen. The rock matrix was pulverized and gently boiled in HF to remove silicates, 3–4 l of water was added, and the solution was left to stand until the solids had settled out. Liquid was decanted and residue was boiled in concentrated HCl for 5 minutes to prevent the formation of calcium fluoride. After settling, the sample was washed in distilled water and centrifuged 3–4 times. Pollen was acetolyzed using standard techniques (Erdtman 1943). In a few cases it was necessary to separate the organic fraction from the inorganic material by heavy liquid separation using a solution of zinc bromide.

In order to prepare for same grain analysis, glycerine was mixed with the organic residue to form a suspension. A drop of this solution was pipetted onto a glass slide. Grains of particular interest were brushed to the edge of the glycerine with a needle with a coarse human hair affixed to its tip. They were then transferred to another glass slide with a fresh drop of glycerine and photographed under LM (Nikon Optiphot-2). Because no cover slip is used, it is possible to photograph the same grain in several orientations (Zetter 1989). In order to examine exine stratification, some grains were fractured with a dissecting needle. These and the intact grains were transferred to an aluminium SEM stub and rinsed in absolute ethanol to remove traces of the glycerine. Because the ethanol may dehydrate the pollen, dimensions of the grains in SEM photographs may vary somewhat from those recorded under LM. The stub was sputter-coated with gold in a BIORAD Sputter Coater for 4 minutes and examined with a Jeol 6400 SEM. Same grain specimens are housed at the Paleontological Institute, University of Vienna.

Additional samples were processed by standard palynological processing techniques outlined in Doher (1980). These samples were not acetolyzed nor subjected to boiling HF. Materials were prepared using Cellosize dispersant on coverslips and then mounted in Permount. These additional samples are deposited in the Plant Fossil Collections, School of Life Sciences, Arizona State University.

4. The Almont palynoflora

The palynoflora recovered from the Almont matrix is unusual for a late Paleocene deposit in several ways:

  1. Fern spores are rare, except for smooth, monolete types that are attributed to Laevigatosporites, a genus that may have affinities with extant Blechnaceae and Thelypteridaceae. Fern megafossils are unknown at Almont, however rhizomes, fertile remains and foliage, including that of Woodwardia (Blechnaceae) have been recovered from the similar Beicegel Creek flora in western North Dakota (Pigg et al. 2006; Matthews et al. 2007). Megaspores of Isoetes are also unknown from Almont but found at Beicegel Creek, as well as several other Paleocene localities including Golden Valley, North Dakota, Ravenscrag, Saskatchewan, and various Colorado sites (Brown 1962; Hickey 1977; McIver and Basinger 1993; Pigg and DeVore 2010).

  2. Several types of bisaccate gymnosperm pollen occur, although megafossils of pinaceous conifers are rare.

  3. The angiosperm component of the flora is dominated by porate pollen grains attributable to Juglandaceae, Betulaceae, and other presumed wind-pollinated taxa.

5. Systematic paleontology

5.1. Organization

This treatment is organized in the following sequence: first pteridophyte spores, then gymnosperm pollen grains, followed by the angiosperms. Angiosperm pollen types with generally accepted taxonomic affinities are listed alphabetically by family, followed by those of uncertain affinities alphabetically by genus.

5.2. Pteridophyte spores

Genus Baculatisporites Pflug & Thompson in Thompson and Pflug 1953

Type species. Baculatisporites primarius (Wolff) Pflug & Thompson in Thompson and Pflug 1953

Remarks. The baculate sculpture readily distinguishes this presumably osmundaceous genus from others, such as Osmundacidites which has more echinate processes.

Baculatisporites sp.

Plate 1, figure 1

Remarks. This spore type strongly resembles the specimen illustrated as Baculatisporites sp. by Pocknall and Nichols (1996).

Affinity.Pocknall and Nichols (1996) suggested affinity of this species with the fern family Osmundaceae, noting that megafossil remains of this family had been reported by Brown (1962).

Genus Laevigatosporites Ibrahim emend. Schopf et al. 1964

Type species. Laevigatosporites vulgaris (Ibrahim) Ibrahim 1933

Remarks. This genus is the major spore type recovered from the Almont matrix.

Laevigatosporites haardtii (Potonié & Venitz) Thompson & Pflug 1953

Plate 1, figures 2–5

Affinity.Srivastava (1971) noted that several families besides Polypodiaceae produce similar looking spores, although most workers attribute affinity for this genus to that family. It is not clear how meaningful this designation might be, given the considerable differences in both past and current classification of filicalean ferns, particularly Polypodiaceae. Monolete grains similar to this are known from Woodwardia (Blechnaceae), a genus known from megafossils in the Beicegel Creek (Pigg et al. 2006; Matthews et al. 2007), Golden Valley (Hickey 1977) and Ravenscrag floras (McIver and Basinger 1993). Spores of Christella dentata (Thelypteridaceae) also show similar characters.

cf. Thelypteridaceae

Genus unknown

Plate 1, figures 6–10

Description. Spore monolete, ellipsoidal, exospore showing characteristic coarse folds (thickenings of the exospore).

Remarks. We refer this material to Thelypteridaceae, a family of filicalean ferns with monolete spores showing characteristic folds.

Family incertae sedis

Spore incertae sedis no. 1

Plate 1, figures 11–13

Description. Spore trilete, outline triangular in polar view, 23 μm in equatorial diameter, surface granulate.

Remarks. This triangular spore is uncommon.

Spore incertae sedis no. 2

Plate 1, figure 14

Remarks. This species has sculpture that precludes its placement in Cyathiidites, and it lacks the three circular marks on its distal surface that would allow it to be placed into Stereisporites.

5.3. Gymnosperm pollen

Ginkgoaceae

Genus Monosulcites Cookson ex Couper 1947

Type species. Monosulcites minimusCookson 1947

Remarks. Elliptical to oblate, boat-shaped, monosulcate grains are characteristic of several taxonomic groups including both the cycads and Ginkgo. According to Ottone and García (1991) it is not possible to distinguish between cycad, Ginkgo and perhaps some other taxa based on LM alone. However, Ginkgo pollen has characteristic rugulate sculptural elements that can be seen with SEM. These are clearly present in the Almont specimens (Plate 1, figures 17, 20).

In addition to the name Monosulcites, Cycadopites can be used for grains with this general morphology, as seen in LM. Ottone et al. (2005) noted that the name Cycadopites has priority over Monosulcites and several other names used for these types of pollen, and suggested that all names should be merged into Cycadopites. However, they did not propose formal revision. Other morphological differences are also sometimes used to distinguish Monosulcites from Cycadopites, namely Monosulcites pollen may have a constriction in the centre of the grain, while Cycadopites grains may appear more spindle-shaped, with tapered, rather than bluntly rounded ends. We use the name Monosulcites because the Almont pollen shows the diagnostic sculptural features of Ginkgo pollen as seen in SEM.

Monosulcites crescentus Norton in Norton and Hall 1969

Plate 1, figures 15–20

Description. Pollen oblate, elliptical in polar view, elliptical acuminate in equatorial view, acute (boat shaped), 30–35 μm in equatorial diameter, monosulcate; exine tectate, sexine thicker than nexine; sculpture in LM psilate, in SEM rugulate.

Remarks. The specimen figured in Plate 1, figures 15–17 is quite similar to material Norton and Hall (1969) illustrated from the Cretaceous–Paleocene of Montana. The surface ornamentation shows a pattern of small rugulae when this species is examined with SEM. The specimen in Plate 1, figures 18–20 illustrates some of the variation seen in this grain type from Almont.

Affinity.Norton and Hall (1969) suggested a gymnospermous affinity. Ginkgo leaves, seeds and stalks bearing seeds are perhaps the most well known components of the Almont flora (Crane et al. 1990) and are prized by collectors worldwide. To date no cycads are known at Almont. These pollen grains show all the characteristics of Ginkgo pollen.

Cupressaceae (including Taxodiaceae)

Genus Taxodiaceaepollenites Kremp 1949 ex Potonié 1958

Type species. Taxodiaceaepollenites hiatus Potonié ex Potonié 1958 (= Taxodiaceae-Pollenites hiatus Potonié in Kremp 1949; Pollenites hiatus Potonié 1931, 1932)

Remarks.Pocknall and Nichols (1996) illustrated specimens quite similar to ours, without assigning them to a species or form genus. Taxodiaceous megafossil remains are common at Almont, including leaves, ovulate cones and pollen cones with in situ pollen (Crane et al. 1990).

Taxodiaceaepollenites sp.

Plate 1, figures 21–25

Description. Pollen oblate to spheroidal, 22–25 μm in equatorial diameter, distal apertural region thinned; exine 1–1.5 μm thick, tectate sexine slightly thicker than nexine; sculpture in LM scabrate, in SEM rugulate to microverrucate with a microechinate suprasculpture.

Remarks. Papillae on taxodiaceous grains recovered are not always evident.

Affinity. Taxodiaceous genera such as Taxodium, Metasequoia and Glyptostrobus.

Araucariaceae

Genus Araucariacites Cookson ex Couper 1953

Type species. Araucariacites australisCookson 1947 ex Couper 1953

Remarks.Couper (1953) validated Cookson's (1947) genus proposed for nonaperturate grains similar in ornamentation to extant Araucariaceae.

Affinity. Araucariaceae

Araucariacites sp.

Plate 2, figures 1–3

Description. Pollen free, disc-shaped, inaperturate; exine thin; sculpture microverrucate.

Remarks. Our specimens are similar to Araucariacites sp. of Nichols and Brown (1992). The same grain SEM examination reveals the crumpled nature of the exine and the densely packed sculptural elements, as well as their great variability in shape. The type species of Araucariapollenites Reyre 1973 was also examined with SEM, and showed the tall sculpture elements. We did not note the ‘circular depressed rim’ in Reyre's type description and so placed our specimen in the more widely used genus Araucariacites.

Sciadopityaceae

Genus Sciadopitipollenites M. Takahashi

Plate 2, figures 4–9

Description. Pollen with leptoma, surface ornamented with large verrucae, 0.6–1.2 μm in width. Verrucae are covered with evenly spaced microechinae, ca. 0.1 μm in height and diameter.

Affinities. As noted by Takahashi (1997), the exine sculpture of Sciadopitipollenites matches that of extant Sciadopitys (cf. Kurmann and Zavada 1994). These grains fall within the size range of the fossil species, S. megaorbiculus M. Takahashi, from the Campanian of Sakhalin.

Gymnosperm incertae sedis

Genus unknown

Plate 2, figures 10–17

Description. Pollen monosaccate, oblate (disc-shaped) 30–40 μm in equatorial diameter, 15–25 μm in polar diameter, corpus proximal thickened, distal polar area (Leptoma) thinned (sunken), saccus folded, distal saccus-folds, more wrinkled, corpus proximal (cappa) rugulate, foveolate, distal leptoma granulate. Affinities unknown.

Remarks. This pollen type has a conspicuous flange, and a superficial resemblance to Dacrydium (Podocarpaceae). This grain may represent an early type of Tsuga.

Pinaceae

Genus Abietineapollenites Potonié 1951 ex Delcourt & Sprumont 1955

Type species. Abietineapollenites microaulatus (Potonié) Delcourt & Sprumont 1955 (nomen nudum in Potonié 1951)

Remarks. We use this genus for bisaccate pollen whose saccal width equals or is less than that of the central body, following the usage in Farabee and Canright (1986). A comparative table for the distinction of extant genera with bisaccate pollen of Pinaceae and Podocarpus is provided by Liu and Basinger (2000, p. 832–833).

Affinity. This genus is typically allied to the Pinaceae. These grains are similar to modern Picea.

Abietineapollenites foveoreticulatus Norton in Norton & Hall 1969

Plate 3, figures 1–4

Description. Bisaccate pollen grain, elliptic in polar view, equatorial diameter 80–90 μm, sacci broadly connected, half-spherical, leptoma slightly rugulate, verrucate, perforate, sacci granulate, perforate.

Remarks.Norton and Hall (1969) noted this species from their transitional assemblage of the Fort Union Group. Pinaceous conifer megafossils are relatively rare at Almont.

Affinity.Norton and Hall (1969) suggested affinity with Pinaceae but according to our SEM study these grains are clearly associated with Picea.

Abietineaepollenites latisulcatus Norton in Norton and Hall 1969

Plate 3, figures 5–7

Description. Bisaccate pollen grain, elliptic in equatorial view, equatorial diameter 60–70 μm, sacci broadly connected to corpus, sacci half-spherical, proximal polar area of corpus verrucate with a granulate suprasculpture.

Affinity. Pinaceae.

Genus Pinuspollenites Raatz 1937

Type species. Pinuspollenites species Raatz 1937 Pinuspollenites elongatus Norton in Norton and Hall 1969

Plate 3, figures 8–9

Description. Bisaccate pollen grain, ± elliptic in polar view, equatorial diameter 70–80 μm, more or less spherical sacci, folded connection area between corpus and sacci.

Remarks. The large pollen body, relative to the sacci, and fine reticulations of the sacci, distinguish this species and place it in Pinuspollenites. An equatorial view detailing the finer gradation of the sacci reticulations suggests a closer affinity of this specimen with Picea than Pinus. Note differences between external ornamentation on the pollen body and saccus.

Affinity.Pinus (Pinaceae) attributed by Norton and Hall (1969).

Genus Cathaya Chun et Kuang 1962

Type species. Cathaya argyrophylla Chun et Kuang 1962

Cathaya sp.

Plate 3, figures 10–12

Description. Bisaccate pollen grain, elliptic in polar view, equatorial diameter 60–65 μm, corpus rhombic in outline, sacci broadly connected to corpus, sacci half-spherical, sculpturing microechinate, perforate in SEM.

Cathaya pollen grains usually have a corpus that is equal in diameter to the width of the sacci (Plate 3, figures 10 and 12).

Remarks. Some pollen grains included herein have a much smaller body relative to their sacci. We interpret these grains as highly degraded Cathaya pollen. Plate 3, figure 11 shows an aberrant Cathaya pollen grain with a degenerated corpus.

Affinity. Fossil pollen of this kind, when known only from LM, is best assigned to the form taxon Pityosporites microalatus (Potonié) Thomson et Pflug (Liu and Basinger 2000), but it has often been placed in the genus Podocarpidites.

Many fossil grains previously considered to represent Podocarpus have been reinvestigated and reassigned to Cathaya (Liu et al. 1997; Liu and Basinger 2000). Pollen of the extant genera Podocarpus and Cathaya is convergent in general form, however by LM extant Cathaya pollen grains can be distinguished from those of Podocarpus by the connection between the saccus and the corpus. In polar view, the sacci of Cathaya grains appear to originate at the margin of the corpus. The most definitive features for identification of Cathaya pollen can be seen only under SEM. Cathaya grains have irregularly spaced spinules and perforations on the surface of both corpus and saccus, except for the leptoma area on the distal surface of the grain (Liu et al. 1997).

Podocarpaceae is often described as a Southern Hemisphere family originating in Gondwana (Hill 1994). As part of the Antarctic flora, presumably only the development of Indonesia allowed migration of the family into Asia, perhaps during the Miocene. However, several papers report the presence of podocarp-type pollen from Mesozoic and younger sediments in the Northern Hemisphere (e.g. Rouse 1957; Norton and Hall 1969; Farabee and Canright 1986). Stanley (1965) made such an assignment, although Norton in Norton and Hall (1969) moved this species to the form genus Podocarpidites without comment. Nichols and Brown (1992) reported this species from their study of the Paleocene in Montana and Wyoming, although they regarded Podocarpidites as a form genus that was most likely not related to Podocarpaceae. Pocknall and Nichols (1996) indicated the bisaccate pollen in their study was assignable to the modern pinaceous genera Pinus and Picea, but did not assign any to Podocarpus/Podocarpidites.

Genus cf. Pinus

Plate 3, figures 13–16

Description. Bisaccate pollen grain, ± elliptic in equatorial view, equatorial diameter 65–70 μm, sacci broadly connected to corpus. sacci half-spherical, sculpturing of cappa rugulate, foveolate, perforate.

Affinity.Pinus (Haploxylon type).

5.4. Angiosperm pollen

Betulaceae

Genus Alnipollenites Potonié 1931 (May)

Type species. Alnipollenites verus (Potonié) ex Potonié 1931 (May) (=Pollenites verus Potonié 1931 April).

Remarks.Farabee and Canright (1986) outlined their use of this genus. Features that distinguish it from the morphological genus Ulmipollenites include the presence of arci and a microechinate sculpture in Alnipollenites in contrast to the verrucate sculpture of Ulmipollenites.

Alnipollenites scoticus (Simpson) Pocknall & Nichols 1996

Plate 4, figures 1–3

Description. Pollen oblate, equatorial diameter 23–25 μm, stephano (8)-porate, arcuate, pori elongated, vestibulate (atrium) neighbouring pori connected by nexinous thickenings (arci); exine 1–1.5 μm, aperature lolongate, tectate, sexine thicker than nexine; sculpture in LM psilate, in SEM microechinate, distal polar area with ring-like nexinous thickening.

Remarks. This species displays seven or eight pores with distinct arci connecting adjacent pores. A central ring is formed by nexinal thickening is seen in distal polar view in this species. Pocknall and Nichols (1996) also included five-pored specimens and did not observe any eight-pored grains. Pocknall and Nichols (1996) used the presence of circular exinal thickenings at the poles to distinguish A. scoticus from other fossil Alnus grains, such as A. verus Potonié and A. speciipites (Wodehouse) Pocknall & Nichols.

Affinity. The organization of microechinate on low ridges revealed by SEM confirms the betulaceous affinity of Alnipollenites scoticus. The large number of pores, and presence of arci indicates closest affinity with Alnus (Betulaceae). Pocknall and Nichols (1996) noted that several extant species of Alnus, all restricted to Japan, possess the annular polar exinal thickening of Alnipollenites scoticus.

Alnipollenites verus Potonié 1931

Plate 4, figures 4–6

Remarks.Alnipollenites verus includes pollen grains with four or five pores connected by distinct arci, with the grains lacking verrucate sculpture (a distinction from Ulmaceae) and appearing psilate under LM. Alnipollenites speciipites (Wodehouse) Pocknall & Nichols 1996 is reported to have a larger size. Both A. verus and A. speciipites lack the circular polar thickening characterizing A. scoticus.

Affinity. Pollen placed in this genus (such as in Norton and Hall 1969; Farabee and Canright 1986) typically is considered to be from plants in the genus Alnus (Betulaceae). Megafossils of Alnus are not known from Almont. An alternative name for this entity is Polyvestibulopollenites verus (R. Potonié) Thomson et Pflug 1953. Polyvestibulopollenites Thomson & Pflug was validly published prior to the validification of Alnipollenites by Potonié (1960) (see Srivastava 1972 for synonymy).

Genus Triporopollenites Pflug & Thompson 1953

Type species. Triporopollenites coryloides Pflug in Thompson and Pflug 1953

Remarks. Oblate, triporate pollen lacking distinct atria are placed in this form genus. Some species of Momipites (notably those described by Leffingwell (1970) as Maceopolipollenites) bear superficial similarity to this genus, but they are usually noticeably smaller than Triporopollenites species and easily distinguished at the SEM level by the distinct spinule arrangements.

Affinity.Pocknall and Nichols (1996) noted affinities of several species of this genus to Betulaceae and Myricaceae. With SEM, it is clear that the specimens considered below are assignable to Betulaceae and not Myricaceae because of the arrangement of spinules in rows. However, it is possible that some of the other reports in the Northern Hemisphere based only on LM may represent Myricaceae.

Triporopollenites coryloides Pflug 1953

Plate 4, figures 7–25

Remarks. This species is one of the most common and variable in the Almont matrix. As suggested by the generic designation, most of the grains are triporate (Plate 4, figures 7–12), but some four- and five-porate grains with otherwise identical morphology and ornamentation were recovered (Plate 4, figures 13–18, 23–25). Tetrads are also found (Plate 4, figures 19–22).

Betulaceous catkins associated with Palaeocarpinus dakotensis (Betulaceae, Coryloideae) from Almont had only triporate grains (Manchester et al. 2004). However, single anthers from catkins associated with Palaeocarpinus aspinosa in the Paleocene of Wyoming contain three-, four- and five-porate grains (Manchester and Chen 1998). Variation in pore structure is also noted, and may need to be re-evaluated to determine the morphological limits of this species.

Affinity.Norton and Hall (1969) noted this species had affinities with the extant family Betulaceae. Grains of this type were observed in situ by Manchester et al. (2004) in catkins associated with Palaeocarpinus dakotensis, a common plant at Almont. Pollen fitting this description also occurs in stamens of catkins associated with the extinct betulaceous infructescences and fruits known as Cranea from the Paleocene of Wyoming (figures 22–29 in Manchester and Chen 1998).

Buxaceae

Genus Erdtmanipollis Krutzsch 1962

Type species. Erdtmanipollis pachysandroides Krutzsch 1962

Remarks. Pantoporate, spheroidal grains presumably belonging to the Buxaceae historically have been placed in this genus in previous studies of Cretaceous–Tertiary palynology (i.e. Samoilovich 1961; Stanley 1965; Norton and Hall 1969; Srivastava 1969a, 1969b). The genus ranges from the Cretaceous to Tertiary throughout the areas in western North America and into eastern Asia. There are two species of this genus in the Almont.

Erdtmanipollis cretaceus (Stanley) Norton in Norton and Hall 1969

Plate 5, figures 1–5

Remarks. This species, as illustrated in Stanley (1965), has more tightly packed sculptural elements than E. pachysandroides. This pollen resembles that of some Euphorbiaceae as well as Buxaceae.

Erdtmanipollis pachysandroides Krutzsch 1962

Plate 5, figures 6–12

Remarks. This species is more commonly encountered, and is distinguished by a looser sculpture with some elements ringing the numerous pores, but still has some spaces between the pores with no pore visible. This species occurs in the region from Cretaceous to Paleocene and younger sediments.

Affinity.Pachysandra and Sarcococca of the Buxaceae. Megafossils of Buxaceae are not known from Almont.

Cercidiphyllaceae

Genus Cercidiphyllum sp.

Plate 5, figures 13–17

Description. Pollen 4-poroidate, equatorial diameter ca. 28 μm wide, polar axis 30–35 μm long, semitectate, sculpturing microreticulate with a granulate suprasculpture in SEM.

Remarks. This pollen compares favourably with that recovered in situ from stamens of inflorescences named AlasiaKrassilov & Kodrul (2008) from the Paleocene of Amur Province Russia. The same type of inflorescences, associated with leaves referred to Cercidiphyllum or Trochodendroides and fruits of Nyssidium occur at Almont (see Crane et al. 1990, figure 10 D–F).

Cornaceae

Genus Cornus sp.

Plate 6, figures 1–3

Description. Pollen prolate, tricolporate with ‘H-shaped’ thickening (a pair of linear thickenings on either side of each colpus), equatorial diameter 18 μm, polar diameter 24 μm; sculpture in LM psilate or apparently scabrate, in SEM, two sizes of spinules, the larger 0.2 μm in diameter and height, and the smaller 0.05 μm in diameter.

Remarks. The combination of tricolporate grains with H-shaped endoapertures (thinning of the endexine with lamellation (Ferguson 1977) and irregularly spaced microechinae of variable size, are characteristic of Cornus. Extant Cornaceae pollen grains have been illustrated with both LM and SEM by Ferguson (1977), with SEMs of additional species reported by Adams and Morton (1976). Ferguson noted that in the cornaceous genus Afrocrania (now treated as a member of Cornus, subg. Cornus) there are two types of columellar spines, one short, the other very prominent, 1.5–3 μm long. In comparison the spinules of the Almont grain are relatively short, conforming to the size in all other species of Cornus.

Affinity. Affinities seem to lie securely with the extant genus Cornus. Cornus subgenus Cornus is also known from the Almont locality based on fruits with distinctive cavities (Xiang et al. 2003) and from other Paleocene localities in North America and Asia based on leaves with diagnostic trichomes and venation (Manchester 2009).

Ericaceae

Genus EricipitesWodehouse 1933

Type species. Ericipites longisulcatusWodehouse 1933

Remarks. This genus contains obligate tetrads whose supposed affinites are with Ericaceae. Grains of this type, when found singly can be seen to have tricolporate grains. While Ericipites tetrads and tetrads of extant Ericaceae are tricolporate, this feature is not easy to observe with tetrads.

Ericipites rallus (Stanley) Farabee & Canright 1986

Plate 6, figures 4–6

Remarks. Our specimens resemble the material Stanley (1965) illustrated, as well as pollen illustrated as Ericipites sp. 3 in Pocknall and Nichols (1996). Nichols and Brown (1992) transferred this species to the genus Simpliciepollis Harris, a taxon containing obligate tetrads with each grain having triporate apertures. We reject their transfer, since Stanley's (1965) description noted the apertures as weakly tricolpate to inaperturate, but made no mention of triporates. Nichols and Brown (1992) noted the genus Stanley (1965) had originally used to house this species (Ericacaeoipollenites) was a junior synonym of Ericipites. Farabee and Canright (1986) transferred Stanley's species to Ericipites for that reason.

Affinity. Ericaceae. Affinities with ‘Kalmia?’ were suggested by Stanley (1965), however the Almont fossils differ in sculpturing pattern and lack viscin threads, a feature that is characteristic of extant Kalmia and other members of subfamily Rhododendroideae (Zetter and Hesse 1996).

Ericipites sp. A

Plate 6, figures 7–9

Remarks. This tetrad differs from that treated above by lacking the colpal cleavages aligned between adjacent grains of the tetrad. In addition, the granular ornamentation is finer. Each grain 25 μm in diameter, and the tetrad is 32 μm across.

Eucommiaceae

Genus Eucommia Oliver 1891

Type species. Eucommia ulmoides Oliver 1891

Remarks. Several studies of Paleocene strata reported pollen sharing enough distinctive characters with extant Eucommia to warrant the use of the genus in the Paleocene (Leopold 1974; Frederiksen et al. 1983; Pocknall and Nichols 1996). Call and Dilcher (1997) accepted the North American dispersed pollen records of this genus from pre-Eocene deposits. Pocknall and Nichols (1996) provided a detailed discussion of why this type of pollen was not assignable to other palynomorph genera and why it should be placed in the extant genus.

Affinity.Pocknall and Nichols (1996) attributed this pollen type to the extant genus Eucommia (Eucommiaceae). Eucommia is well documented by fruits from the Eocene of North America, Eocene to Pliocene of Asia, and Oligocene to Pliocene of Europe (Manchester et al. 2009). Specimens designated as ‘elliptical biwinged fruits’ from Almont that were tentatively compared with Eucommia (Crane et al. 1990, figure 23E) are now assigned to the Polygonaceae (Manchester and O'Leary 2010). Leaves referred to Leaf Type VI (Crane et al. 1990) were compared with the family, but have since been attributed to the extinct nyssaceous genus, Browniea (Manchester and Hickey 2007). As yet, no megafossil remains of the Eucommia have been recovered from Almont.

Eucommia? leopoldaeFrederiksen 1983

Plate 6, figures 10–13

Description. Pollen prolate, in polar view circular, in equatorial view elliptical, polar diameter 20–22 μm, equatorial diameter 14–16 μm, tricolporoidate; exine tectate, sexine thicker than nexine; sculpture in LM psilate, in SEM granulate with irregularly scattered microechini.

Remarks.Eucommia? leopoldae is distinguished by its relatively thin exine, smooth-scabrate surface and poorly developed pores within straight colpi. We retain the question mark that was included in the original designation of the binomial by Frederiksen (in Frederiksen et al. 1983), but both LM and SEM support the identification to extant Eucommia.

Hamamelidaceae

Genus RetitrescolpitesSah 1967

Type species. Retitrescolpites typicus Sah 1947

Remarks. The genus Retitrescolpites is used for tricolpate dispersed pollen with reticulate lumina larger than 1 μm. Thus, affinities must be made at the species level. Tricolpites is used for tricolpate reticulate pollen with lumina smaller than 1 μm. Rousea has lumen gradations that occur finer at the poles, unlike what we observe in these specimens.

Retitrescolpites catenatusPocknall & Nichols 1996

Plate 6, figures 14–20

Description. Pollen spheroidal to slightly oblate, in polar view circular (trilobate), in equatorial view circular to elliptical, tricolpate, colpi broad, semitectate, sexine thicker than nexine; sculpture reticulate heterobrochate, brochipolygonal, larger lumina often accompanied by finer (smaller) lumina.

Remarks. This species is distinguished by its lumina grading finer approaching the smooth sided colpal margins as well as some lumina being smaller while others are larger.

Affinity. The pollen is known in situ from catkins associated with Hamawilsonia infructescences (Benedict et al. 2008). Hamawilsonia is an extinct genus known from anatomically preserved fruits, infructesences, and seeds that combines features of Sinowilsonia, Hamamelis and several other extant genera of Hamamelidaceae. Among the 29 extant genera of Hamamelidaceae for which SEMs of the pollen grains are known, this pollen most closely resembles that of Sinowilsonia (Bogle and Philbrick 1980). Shared features include the pattern of the coarse reticulum, showing numerous very small lumina scattered among the larger lumina, and a decrease in size of the larger lumina at the poles.

Retitrescolpites anguloluminosus (Anderson) Frederiksen 1979

Plate 6, figures 21–23

Description. Pollen prolate, in equatorial view elliptical, in polar view circular, polar diameter 20–23 μm, equatorial diameter 15–17 μm; tricolpate; exine 1.8–2 μm, semitectate, sexine thicker than nexine; sculpture retriculate, columellae as long or slightly shorter than muri depth, colpi sometimes accompanied by small luminae within the bordering rim of tectum.

Remarks.Pocknall and Nichols (1996) noted the similarity of this species to their R. catenatus. The two species can be distinguished at higher magnifications by R. catenatus having heterobrochate muri that are absent in R. anguloluminosus. In comparing the same grain LM and SEM of each species however, we note a great deal of similarity of muri height and geometry of the reticulum, but the interspersed smaller luminae are lacking in R. anguloluminosus. Nichols and Brown (1992) summarized the geographic and stratigraphic range of the species.

Affinity.Norton and Hall (1969) suggested affinity with Salix. However, the resemblance of R. catenatus with pollen of Hamamelidaceae is strong, and considering the similarity between the SEMs of these two species, we suggest Hamamelidaceae, e.g. Corylopsis (see Bogle and Philbrick 1980).

Retitrescolpites sp.

Plate 6, figures 24–26

Description. Pollen spheroidal to slightly oblate, in polar view circular (trilobate), in equatorial view circular to elliptical, tricolpate, colpi broad, equatorial diameter 23–25 μm; exine 1.8–2.5 μm, semitectate, sexine thicker than nexine; sculpture reticulate, homobrochate, brochipolygonal, muri deep, thin, borne on short columellae.

Remarks. This species is similar in size to Retitrescolpites catenatus Pocknall & Nichols, but the lack of heterobrochate lumina in Retitrescolpites sp. distinguishes it. This species lacks the fining gradation toward the poles of Rousea spp., and it has lumen sizes greater than 1 μm, thus placing it into Retitrescolpites.

Juglandaceae

Genus Momipites (Wodehouse) emend. Frederiksen & Christopher 1978

Type species. Momipites coryloidesWodehouse 1933

Remarks.Momipites is useful as a morphotype genus to accomodate triporate grains of juglandaceous affinity, but the concept of this genus has varied among authors. The type species, Momipites coryloidesWodehouse 1933 was first described from the Green River Formation (Wodehouse 1933). Stanley (1965) later examined material from this area (although not the type specimen) and reported that M. coryloides was abundant. Leffingwell (1970) proposed the genus Maceopolipollenites to include pollen with polar thinnings in the form of triangular areas, and described several species. The presence of polar thinnings and ‘islands’ apparently was used by Leffingwell (1970) to separate Maceopolipollenites from Momipites. This feature was not mentioned by Stanley (1965).

Nichols (1973) recognized MaceopolipollenitesLeffingwell 1970, Engelhardtiapollenites Raatz 1937, Engelhardtioipollenites Potonié 1951, and Engelhardtioides? Potonié, Thompson, & Thiergart 1950 and several other genera as junior synonyms to Momipites, and emended Momipites to include pollen with morphology typical of the modern genera Engelhardtia and Alfaroa, and other species with certain features not found among those living genera (Nichols 1973, p. 106). This emendation thus includes species with specialized exinous thin areas for which Leffingwell (1970) had erected Maceopolipollenites. Using a narrower concept for the genus, Frederiksen and Christopher (1978) emended Momipites to specifically exclude ‘multiple thin spots and other elongate thin streaks, and triradiate folds and thickenings’. Jansonius and Hills (1979 supplement, card 3576) note that this concept differs from Nichols' 1973 emendation in excluding the M. triradiatus and M. triorbicularis groups. However, Frederiksen and Christopher's (1978) emendation would allow for species with circular thickenings as in Momipites annelus.

Manchester (1989) briefly reviewed the nomenclature for Paleogene juglandaceous pollen and retained Maceopolipollites as defined by Leffingwell to include triporate pollen with polar thinnings in the form of triangular areas or circumpolar circles, on a single pole. Maceopolipollites thus can be distinguished from Platycarya and Plicatopollis, which have thinnings on both surfaces. Manchester and Dilcher (1997) documented this pollen in situ from catkins co-occurring with the extinct juglandaceous fruit genus, Polyptera. Pocknall and Nichols (1996) continued to favour the broader concept of Momipites (including Maceopolipollenites). Momipites in this broader sense thus includes pollen grains that likely correspond to more than one natural genus.

Affinities. The affinities of these grains with Juglandaceae are confirmed by the distinctive even arrangement of fine spinules on porate grains (Stone and Broome 1975). Juglandaceous pollen has been an important biostratigraphic indicator in the Paleogene, however the biological affinities of these fossil grains and their degree of intraspecific variability is not well resolved. Whereas extant genera of this family are distinguished mainly on the number and distribution patterns of pores, and the presence or absence of peculiar patterns of thinning in the exine, the same combinations of these characters that define extant genera are not found among the fossils. Study with LM, as is typical for biostratigraphic purposes, may not always resolve important diagnostic features for the family that are best seen by SEM.

Based on the co-occurrence with fruit megafossils, Manchester (1989) and Manchester and Dilcher (1982, 1997) have suggested the possible affinities of several juglandaceous pollen types. For example, Momipites was first described from the Eocene Green River Formation, suggesting that the type species, M. coryloides, may have affinity to the engelhardioid fruits that are present in the same formation (MacGinitie 1969).

The presence of all of these different juglandaceous pollen types at Almont is intriguing, especially since we recognize only a single juglandaceous fruit at this locality, Cyclocarya brownii. Newly recognized megafossil remains of Cyclocarya, including infructesences and putative leaves and pollen catkins, are currently under study (Taylor et al. 2007). It might be expected that the corresponding pollen type for this plant would be among the sporae dispersae grains. Pollen of extant Cyclocarya is typically four-pored (with a smaller percentage of triporate and pentaporate grains), and somewhat heteropolar (Stone and Broome 1975; Taylor et al. 2007). Another interesting feature of dessicated extant Cyclocarya pollen is the superficial appearance of arci-like structures (Taylor et al. 2007). Tetraporate juglandaceous pollen is rare in the Almont assemblage, suggesting that Cyclocarya brownii pollen may have been triporate, perhaps of the M. coryloides and/or M. triorbicularis types, forms previously associated with engelhardiod rather than juglandoid forms.

Momipites annelusNichols & Ott 1978

Plate 7, figures 1–2

Remarks. This species is distinguished from other Momipites species by the presence of a circular ring of exinal thinning on one pole, and from otherwise similar species in Caryapollenites by its equatorial pores, which all fall in a single plane.

Affinity. Grains of this type occur in situ within catkins that consistently co-occur with the extinct genus Polyptera (Manchester and Dilcher 1982, 1997; Manchester 1989). Although Nichols and Ott (1978) assigned the species to Momipites under the broader concept of the genus mentioned above, Manchester (1989) argued to maintain Leffingwell's distinction of a separate genus Maceopolipollenites. Grains of this kind have a Carya-like circular thin area on one face, and truly equatorial pores, a feature not found in extant Carya. However, the type species of Maceopolipollenites, M. triorbiularis appears to be generically distinct, so a new generic name is probably warranted for juglandaceous pollen with isoporate grains having the circular exinous thinning at one pole. For the present, we apply Nichols' broader concept of Momipites, which accommodates both of these kinds of grains as well as those that lack exinous thinning.

Momipites triorbicularis (Leffingwell) Nichols 1973

Plate 7, figures 3–8

Remarks. This species is distinct from others by its thinning exine approximately half-way to each pole. This produces grains with three circular thinnings alternating with the pores (as in the micrograph of Leffingwell's 1970 holotype) or a polar y-shaped ‘hump’ in the specimen illustrated here. Leffingwell (1970) noted the occurrence of this species in the upper parts of his sections of the Fort Union Formation (Paleocene).

Affinity. The three pores and finely scabrate exine (appearing almost smooth under LM, but with small, evenly distributed coni in the SEMs) are consistent with placement of this species in the Juglandaceae.

Genus Caryapollenites Raatz 1937 ex Potonié 1960 emend. Krutzsch 1961

Type species. Caryapollenites simplex (f. communis) Raatz 1937

Remarks. This genus of juglandaceous fossil pollen with presumptive affinity to extant Carya has triporate grains with one or more of the aperatures offset from the equator of the grain, making these grains heteropolar. Because of their short polar axis they are typically found in polar view. This orientation also facilitates observation of the frequent polar thinning on the polar surface on the opposite surface with pores (known to be proximal pole in extant Carya). SEM examination reveals a surface sculpture characteristic for the family. Although morphologically most similar to extant Carya, many authors have noted that Paleocene Caryapollenites species have smaller grains than those of extant Carya. Based upon fruits, Carya is first known from the late Eocene, but those of Juglandicarya simplicarpa Manchester co-occur with Caryapollenites pollen at several North American Paleocene sites, including the section from which Nichols and Ott (1978) recognized five species of Caryapollenites.

Pocknall (1987) noted that this genus characterizes the Paleocene of the Wind River Basin, and that as floral change occurred at the Paleocene–Eocene boundary, the Momipites/Caryapollenites-dominated palynofloras were replaced by those dominated by Platycarya pollen. The Almont deposit does not yield any mega- or microfossils assignable to Platycarya.

Caryapollenites imparalisNichols & Ott 1978

Plate 7, figures 9–11

Remarks.Pocknall and Nichols (1996) noted this species could be separated from similar forms by its having only two pores in subequatorial position, with the other at the equator and the lack of any polar thinning. Caryapollenites species usually, but not always, exhibit both heteropolar pore position and some degree of polar thinning at one pole.

Nyssaceae

Genus Nyssapollenites Thiergart 1937 ex Potonié 1960 Type species. Nyssapollenites thompsonianus (Traverse 1955) Potonié 1960

Plate 7, figures 12–18

Description. Pollen subprolate to spheroidal, in equatorial view elliptical, in polar view triangular, tricolporate, colpi long, constricted towards endoaperture, ends of the colpi often wider with rounded ends, endoaperture often slightly wider than colpus (costate endospore), in polar length 30 μm, sculpture rugulate, foveolate, perforate.

Affinity.Nyssa, Nyssaceae. Although no fruits of Nyssa are confirmed from Almont or other Paleocene localities, the extinct nyssaceous infructescence Amersinia, and probable foliage, Beringiaphyllum cupanioides, are common at Almont and other Paleocene sites in North America.

Genus CaprifoliipitesWodehouse 1933

Type species. Caprifoliipites viridifluminisWodehouse 1933.

Remarks. This genus is used as a catch-all for tricolporate, reticulate, isopolar pollen grains, which may or may not have affinity to the same extant family.

Affinity.Pocknall and Nichols (1996) noted Wodehouse's (1933) assignment of the type species of this genus to the extant genera Viburnum and Sambucus (both extant genera of Caprifoliaceae).

Caprifoliipites paleocenicusPocknall and Nichols 1996

Plate 7, figures 19–22

Description. Pollen subprolate to spheroidal, in equatorial view elliptic to circular, in polar view circular, polar length 30 μm, equatorial diameter 22–25 μm, tricolporate, colpus membrane microverrucate endopori (endoapertures) characteristically elongate, slitlike; exine thickness 1.8–2 μm, sexine thicker than nexine, semitectate, sculpture reticulate to microreticulate. Heterobrochate reticulum in the central part of the mesocolpilum with mostly polygonal lumina, becoming microreticulate, perforate towards the apertures.

Remarks. This species has tricolporate apertures with reticulate sculpture. Pocknall and Nichols (1996) noted the reticulum was thin and loose, often torn in some specimens. Our specimens conform to the overall size of the species and are reticulate, although not the loose reticulate type noted by Pocknall and Nichols (1996).

Affinity.Pocknall and Nichols (1996) noted that this genus represented a ‘generalized morphologic type’, but did not definitely state an affinity. This type of pollen was documented in situ in anthers of flowers of the nyssaceous megafossil genus Browniea in the Paleocene of Montana (Manchester and Hickey 2007).

Caprifoliipites sp.

Plate 7, figures 23–26

This pollen grain differs from C. paleocenicus by its larger size, much thicker pollen wall. Both species show rectangular endoaperatures with a thickening perpendicular to the polar axis. This form is quite common at Almont.

Ranunculaceae

Genus Echitricolpites Da Silva Pares Regali, Uesugui, & Da Silva Santos 1974

Remarks. The echinate sculpture distinguishes grains of this genus from other tricolpate forms.

Echitricolpites supraechinatusPocknall & Nichols 1996

Plate 8, figures 1–4

Description. Pollen spheroidal to suboblate, in polar view circular (lobate), tricolpate; equatorial diameter 12–14 μm; exine tectate; sculpture echinate, echinii about 1 μm long.

Remarks.Pocknall and Nichols (1996) described this species from the Tongue River Member of the Fort Union Formation, which is slightly older than the Almont. The specimens from Almont differ slightly from the type description, so we include our own description.

Affinity. Dicot angiosperm indeterminate, possibly Ranuculaceae. Paleoactaea nageli fruits (Ranunculaceae) are present at Almont (Pigg and DeVore 2005).

Sapindaceae

Genus AesculipollisPocknall & Nichols 1996

Type species. Aesculipollis wyomingensisPocknall & Nichols 1996

Remarks.Pocknall and Nichols (1996) proposed the genus Aesculipollis to accommodate pollen with granulate colpal membranes and striate ornamentation parallel to the pollen equatorial plane. This form has presumed affinity to extant Aesculus (Sapindaceae). The presence of granulate colpi membranes separates this genus from the similar Aesculiidites Elsik 1968. Both genera have striate ornamentation paralleling the equatorial plane of the pollen grain, thus separating them from similar sized striate Acer-like pollen in which the striae parallel the polar axis.

The presence of a significant megafossil record of Aesculus (Manchester 2001) and Acer-like samaras (Crane et al. 1990; Kittle et al. 2005) at Almont and other sites in North Dakota, Montana, and Wyoming strongly suggests affinities of these striate pollen types with Sapindaceae. However, similar striate forms are found also in several other families including Anacardiaceae, Loasaceae, Rosaceae, and Rutaceae (Jones et al. 1995). A more comprehensive knowledge of ultrastructural features of striately ornamented grains may be helpful in resolving their affinities.

Aesculipollis wyomingensisPocknall & Nichols 1996

Plate 8, figures 5–7

Description. Pollen prolate to subprolate, polar diameter 13–15 μm, equatorial diameter 11–13 μm, tricolporate, colpi long, narrow, pori circular, colpus membrane echinate; exine tectate, sexine thicker than nexine; sculpture striate with striae radiating from the apertures such that they are approximately horizontal at the equator and longitudinally oriented at the polar ends.

Remarks. The granulate membranes on colpi described based on LM by Pocknall and Nichols (1996) form an echinate colpus membrane (Plate 8, figure 7).

Pozhidaev (1995) recognized four pollen morphological groups among extant species of Aesculus and the related genus Billia. The first group, which includes all the New World species of Aesculus except A. californica, and the two examined species of the New World genus Billia, have smooth or tubercled colpus membrane, while the other three groups have spinous colpus membranes, similar to what is seen in our material. The ornamentation of the fossil pollen resembles most closely that of Pozhidaev's Groups 2 and 3. Based on the shape of the colpus endings, the fossil pollen corresponds to Group 2, which includes the extant Eurasian species A. hipposatanum, A. turbinata, and the western North American species, A. californica. The Almont pollen is smaller (12 × 11 μm) than the extant species studied by Pozhidaev (1995) which range from 16–23 μm in equatorial and 24–37 μm in polar diameter.

Affinity.Aesculus, Sapindaceae (Pocknall and Nichols 1996). A palmately compound leaf of Aesculus hickeyi has been documented from Almont (figure 1D in Manchester 2001) and the species is represented by fruits and foliage from several Paleocene localities in Wyoming and North Dakota (Manchester 2001). Manchester (2001) concluded that this fossil species, with its echinate colpus membrane, corresponds well to that of Aesculus section Pavia (e.g. A. glabra), as well as to that of the related genus Billia (Pozhidaev 1995).

Genus Striatopollis Krutzsch 1959

Type species. Striatopollis sarstedtensis Krutzsch 1959.

Remarks. This genus contains striate, tricolpate pollen with striae parallel to the polar axis. Simpsonipollis Srivastava 1975 is likewise sculptured but is tricolporate.

Affinity.Farabee and Canright (1986) noted the affinity of the genus could reside with several families, such as Aceraceae sensu stricto (now Sapindales, Judd and Manchester 1997), Solanaceae (Sah 1967), and Fabaceae (Dettmann 1973). This grain also shows some similarities with Anacardiaceae.

Striatopollis tectatusLeffingwell 1970

Plate 8, figures 8–10

Description. Pollen prolate, tricolporate, tectum ornamentation striate, colpus membrane smooth.

Remarks. This tricolporate grain has striae parallel to the polar axis of the grain. Aesculipollis is also striate and tricolporate, but has striae mostly perpendicular to the polar axis.

Affinity.Pocknall and Nichols (1996) suggested Aceraceae sensu stricto, but also noted Hippocastanaceae sensu stricto and Simaroubaceae have similar pollen. Aceraceae and Hippocastanaceae are now recognized within Sapindaceae (Judd and Manchester 1997).

Ulmaceae

Genus Ulmipollenites (Wolff, 1934) emend. Srivastava 1969a

Type species. Ulmipollenites undulosus Wolff 1934

Remarks. We follow Srivastava's (1969a) suggestion regarding the priority of Ulmipollenites over Ulmoideipites.

Ulmipollenites tricostatus (Anderson) Farabee & Canright 1986

Plate 8, figures 11–14

Remarks.Farabee and Canright (1986) noted the essentially Upper Cretaceous distribution of this species (with Anderson, 1960, having a Paleocene occurrence for U. tricostatus). Ulmaceae megafossils are unknown at Almont.

Affinity. Ulmaceae.

Ulmipollenites sp.

Plate 8, figures 15–16

Remarks. This four-pored grain differs from U. tricostatus. The three-pored type of elm pollen grain has more robust sculpture than the four-pored grains.

cf. Bignoniaceae genus and species unknown

Plate 8, figures 17–23

Description. Pollen oblate, polar view circular, equatorial view elliptical, equatorial diameter 20–25 μm, tricolpate, copli long, broad, often with truncate ends, colpus membrane microrugulate to microverrucate; colpi bordered by a thin margo; exine 2.5–3 μm, semitectate, sexine thicker than nexine; sculpture reticulate, muri with smooth rounded surfaces, varying remarkably in thickness, multicolumellate; columellae numerous and thin, lumina with densely packed freestanding columellae.

Remarks. The reticulum formed in this species is unlike anything else seen in Almont, and distinguishes this form from all others. Muri in this species do not have a constant thickness as they do in Retirescolpites sp. Muri appear like wax flowing down a candle, thin at one end and thickening several times the original diameter in the middle. This produces a lumen size that is irregular, with some lumina being larger than 1 μm in size, but with no apparent gradation as in the genus Rousea. This pollen is sufficiently distinctive to warrant a new genus and species.

Affinity. Similar pollen morphological characters can be observed in extant genus Nyctocalos (Bignoniaceae; Burman 1977) but those grains differ by their very large size. At 100 μm, they are four to five times larger than the Almont fossils.

Genus ChenopodipollisKrutzsch 1966

Type species. Chenopodipollis

Genus ChenopodipollisKrutzsch 1966

Chenopodipollis sp.

Plate 9, figures 1–4

Description. Pollen spheroidal, globose, multiporate with ca. 100 pores, equatorial diameter 23 μm; sculpture with widely scattered verrucae.

Remarks. This pollen type resembles grains placed in Periporopollenites Pflug & Thompson.

Affinity.Chenopodipollis has been attributed to the Amaranthaceae (including Chenopodiaceae). No megafossils of these affinities are known at Almont. The similar pollen of Periporopollenites was originally not attributed to any specific family (Thompson and Pflug 1953), however, the genus was emended by Krutzsch (1966) to attempt restricting it to pollen assignable to Liquidambar. Pocknall and Nichols (1996) followed this original diagnosis. No megafossils assignable to Liquidambar have been thus far recovered from Almont, or other Paleocene sites in the Rocky Mountains and Great Plains regions.

Genus Fraxinoipollenites Potonié 1951 (Wien) ex Potonié 1960

Type species. Fraxinoipollenites pudicus Potonié 1951 ex Potonié 1960

Remarks. This genus includes tricolpate, reticulate pollen that is not assignable to Tricolpites Cookson ex Couper, Retitrescolpites Sah, or Rousea Srivastava.

Fraxinoipollenites pachyexinousLeffingwell 1970

Plate 9, figures 5–7

Remarks. The margins along the colpi that Leffingwell (1970) mentioned in his species description are not clearly visible in our specimens, although the lumen sizes do grade smaller approaching the colpi margins. The SEMs show a decrease in lumen size but the lumina remain open all the way to the colpi margins.

Affinity.Leffingwell (1970) did not suggest an affinity. One possibility is Platanaceae, for which megafossils are present at Almont.

Genus Intratriporopollenites Pflug & Thompson in Thompson and Pflug 1953 emend. Mai 1961

Type species. Intratriporopollenites instructus (Potonié) Thompson and Pflug 1953

Remarks.Mai's (1961) emendation of this genus restricted it to pollen with tiliaceous characters. Pocknall and Nichols (1996) noted the specimens they placed in the genus had a variable aperture number as compared with the tricolporate nature of all extant Tilia.

Intratriporopollenites sp. cf. Tilia tetraforaminipitesWodehouse 1933

Plate 9, figures 8–13

Description. Pollen oblate, in polar view, tri-, tetra-, pentagonal–convex, in equatorial view, elliptical, 3–5 colporate, colpi short, endopori circular (brevicolporate); exine tectate, sexine thicker than nexine; sculpture in LM psilate, in SEM verrucate to rugulate, foveolate perforate.

Remarks. Our specimens assignable to this taxon were typically tri- or tetra-aperturate forms, unlike the specimens of Pocknall and Nichols (1996) that had up to five apertures. This pollen type is one of the most common angiosperm elements in the Almont microflora. Grains are often found in clumps.

Affinity.Pocknall and Nichols (1996) discussed the affinity of this species and concluded they could not ascribe this species to extant Tilia, but rather to some early, extinct member of Tiliaceae.

Genus Pistillipollenites Rouse 1962

Type species. Pistillipollenites macgregori Rouse 1962

Affinity. Uncertain. Rouse and Srivastava (1970) suggested affinity of this genus as angiosperm but of uncertain placement. Pollen of this type was found in the anthers of a fossil flower from the Eocene of Texas that conformed with the family Gentianaceae (Crepet and Daghlian 1981). The type material of the Cretaceous Pistillipollenites was reexamined and the authors concluded that the Cretaceous forms had a different aperture and thus belonged in a different genus, although they chose not to designate one (Crepet and Daghlian 1981). Pollen of the Pistillipollenites type was also found in a second type of flower from the Eocene of British Columbia that lacked features of Gentianceae (Stockey and Manchester 1988). Flowers similar to those from British Columbia are present in the Almont megafossil assemblage (e.g. Crane et al. 1990, figure 25K). Pollen with this morphology may belong to more than one taxonomic group.

Pistillipollenites macgregori Rouse 1962

Plate 9, figures 14–18

Description. Pollen, oblate to spheroidal, elliptic in equatorial view, triangular to convex triangular in polar view, tricolpate, colpi short, exine tectate, sexine thicker than nexine, sculpturing verrucate to gemmate in LM, mainly microverrucate in SEM, bigger verrucae and gemmae irregulary distributed, around apertures a higher density of verrucae and gemmae is visible.

Remarks. Pollen assigned to this morphological taxon is quite long-ranging, occurring from the Cretaceous to Eocene of North America, Europe and Japan. Frederiksen (1980) accepting the suggestion of gentianaceous affinity of this pollen, inferred that it was produced by a mainly herbaceous family. However, this type of pollen occurs in at least two different types of flowers and cannot be exclusively tied with the gentians (Stockey and Manchester 1988).

Genus Rousea Srivastava 1969

Type species. Rousea subtilis Srivastava 1969

Remarks. This genus was proposed to include tricolpate, reticulate pollen with mesh sizes grading finer toward the poles as well as colpi margins (Srivastava 1969a, 1969b). The genus Retitrescolpites is used for tricolpate pollen with mesh sizes greater than 1 μm without appreciable mesh size gradation at the poles, but rather exclusively toward the colpi margins.

Rousea crassimurinaPocknall & Nichols 1996

Plate 9, figures 19–22

Remarks. As noted by Pocknall and Nichols (1996), this species is quite distinct from other Paleocene tricolpate species. In their description, Pocknall and Nichols (1996) mention this species as duplibaculate where the lumina were larger in mid mesocolpial areas; a feature not evident on our specimens from Almont.

Affinity. None suggested at species level, although some authors suggest affinity for other species of Rousea with Salicaceae.

Genus SiltariaTraverse 1955

Type species. Siltaria scabriextimaTraverse 1955

Remarks. This genus includes tricolporate pollen with a much longer polar axis than an equatorial axis (prolate). Traverse's (1955) original diagnosis for the genus, based on pollen from the Miocene of Vermont, noted that it had features between those of Quercus and Castanea.

Siltaria hanleyiPocknall & Nichols 1996

Plate 10, figures 1–7

Description. Pollen prolate, tricolporate, equatorial diameter 15 μm, polar diameter 22 μm, sculpture in LM psilate, in SEM tectum is ornamented with interwoven short rugulate to microrugulate elements.

Remarks. Our specimens resemble those illustrated by Pocknall and Nichols (1996).

Affinity.Pocknall and Nichols (1996) noted the similarity of this species to the extant genus Castanea (Fagaceae).

Genus Tricolpites Cookson ex Couper 1953 emend. Potonié 1960

Type species. Tricolpites reticulatusCookson 1947

Remarks. The form genus Tricolpites is used here for tricolpate grains with lumen sizes less than 1 μm, following the suggestion of Srivastava (1969b). Various botanical affinities for this genus have been suggested and are better addressed at the species level.

Tricolpites hiansStanley 1965

Plate 10, figures 8–11

Remarks. The gaping nature of this species colpi when viewed in polar or oblique view make it easy to recognize, as noted in Farabee and Canright (1986).

Affinity. Dicot angiosperm indeterminate.

Genus and Species Unknown

Type 1

Plate 10, figures 12–15

Remarks. These striate, tricolporate grains resemble those of Rosaceae.

Type 2

Plate 10, figures 16–19

Remarks. These pollen grains have a general morphological similarity to the Type 1 grains above, and have distinctive flap-like structures extending over the colpi in a manner similar to some Rosaceae. The sculpturing pattern differs from that of the Type 1 striate grains, that of the Type 1 grains being cross-hatched, rather than prominently striate.

Type 3

Genus Plate 10, figures 20–22

Description. Prolate grain, tricolporate, equatorial diameter 16 μm, polar diameter 24 μm, colpi elongate (more than 90% of polar length). Tectum ornamented with conspicuous evenly spaced spinules 0.6–0.8 μm; spinules smooth, area between spines granulate.

Remarks. This type of grain is known also from a mesophytic flora of the middle Oligocene of Austria (Zetter, unpublished data), the Late Eocene to Oligocene in Kazahkstan (Zaklinskaya 1957, p. 22, 28), and the Upper Gancaigou Formation (Oligocene) of Qaidam Basin, China (Rowley et al. 1981).

6. Comments and conclusions

With the present study, we now have a basis for comparing the Almont megafossil flora with the palynomorph assemblage from the same rock matrix. The combination of anatomically preserved fruits and seeds, coupled with same-grain pollen studies provides an increase in both the breadth and depth of our understanding of this late Paleocene site (Table 1). Some taxa are represented by both pollen and megafossils (e.g. Ginkgo, Cornus, Crane et al. 1990), and some even have in situ grains that demonstrate the variability within an individual (e.g. taxodiaceous conifers, Palaeocarpinus, Hamawilsonia, Crane et al. 1990; Manchester et al. 2004; Benedict et al. 2008). Other families that are rare as megafossils are known in greater abundance from pollen. For example, saccate pollen is common while megafossils of Pinaceae are rare. Pollen of both Hamamelidaceae and Juglandaceae is more diverse than the corresponding record of megafossils. However, while we know that Retitrescolpites catenatus pollen occurs in situ in catkins of Hamawilsonia, pollen of Juglandaceae recovered at Almont is known in situ in taxa of a different subfamily (Engelhardioideae) than the megafossil present (Cyclocarya, Juglandoideae). Ranunculaceae is known from both pollen and the fruit Paleoactaea, however whether this pollen type belonged to the plant that bore that fruit is unclear. Families such as Ericaceae, Eucommiaceae, Ulmaceae, and possibly Bignoniaceae are known only from pollen; Actidiniaceae, Icacinaceae, Meliosmataceae/Sabiaceae, Myrtaceae, Nymphaeaceae, Ochnaceae, and Polygalaceae are known only from megafossils (Table 1).

The same grain studies presented here demonstrate the value of scanning electron microscopy to help resolve some issues of identity. For example, monocolpate pollen assigned to Monosulcites, Cycadopites and several other genera is thought to belong to cycads, Ginkgo or possibly some other plants. Using the LM, pollen of extant cycads and Ginkgo is indistinguishable (Ottone and García 1991), however, SEM photomicrographs show that Ginkgo has a characteristic ornamentation pattern. Same grain analysis of Monosulcites from Almont documents the Ginkgo pattern of ornamentation, a result consistent with the fact that cycads are unknown in the megaflora. While the expense and time involved in same-grain analysis prohibit its broadscale use in stratigraphic palynology, these studies provide a basis for knowing which particular grains might need to be studied with SEM for taxonomic resolution.

Some previous studies of Paleocene palynomorph assemblages (e.g. Farley 1989) have focused on the influence of lithofacies distribution effects on the accuracy of biostratigraphic interpretations. In his study of the Bighorn Basin, Farley (1989) found distinctive lenticular channel fills that contained a greater diversity of palynomorph types than those found in the tabular units. Within the Williston Basin, the Almont and Beicegel Creek floras occur in lenticular channel fills that are considerably smaller than those in the Bighorn Basin, in some cases no more than 5 m in width. These deposits contain silicified shales that provide an extraordinary source for pollen that can help to clarify the taxonomic affinities of both anatomically preserved megafossils and palynomorphs. In situ pollen studies provide an understanding of potential affinities of sporae dispersae grains, and in the case of highly variable types, can demonstrate that more than one palynomorph form may occur within an individual. By carefully looking at the easily recognized and highly diverse palynoflora of the Almont and Beicegel Creek silicified shales, we can fine tune the systematics and expand the documentation for some pollen types.

Acknowledgements

This research was funded in part by NSF EAR-0345838, a Faculty Grant-in-Aid, College of Liberal Arts & Sciences, ASU, a Research Incentive Award, ASU; to KBP; NSF EAR-0345569, a Faculty Research and Development Award, GC&SU, to MLD, and by NSF funding to SRM. Michael Nowak's work was supported by NSF funding to David L. Dilcher.

We acknowledge Terry Lott, John C. Benedict and Witt Taylor for technical assistance. Michael J. Farabee was supported by a sabbatical grant from the Maricopa County Community College District and Estrella Mountain Community College.

We dedicate this paper to the memory of Douglas J. Nichols (1942–2010) with appreciation for his many contributions of Tertiary palynology. His own work on Paleocene palynofloras of western North America contributed greatly to the knowledge base that facilitated the present study.

Author biographies

REINHARD ZETTER, palynologist, is at the Institute of Palaeontology, Geocentre, University of Vienna, Vienna, Austria.

MICHAEL J. FARABEE, palynologist, is Professor of the Division of Science at Estrella Mountain Community College, Avondale, AZ, USA.

KATHLEEN B. PIGG, paleobotanist, is a Professor of the School of Life Sciences, Arizona State University, Tempe, AZ, USA.

STEVEN R. MANCHESTER, paleobotanist, is the Curator of Paleobotany at the Florida Museum of Natural History, University of Florida, Gainesville, FL, USA.

MELANIE L. DEVORE, paleobotanist, is a Professor of the Department of Biological and Environmental Sciences, Georgia College and State University, Milledgeville, GA, USA.

MICHAEL D. NOWAK is a graduate student in evolution and systematics at Duke University, Durham, NC, USA.