The Yukon-Koyukuk Region
|Introduction||Geology||Hollick Localities||Spicer Localities||Locality Relationships||Wider Correlations|
Scott and Smiley (1979) were of the opinion that during mid-Cretaceous time the floral province which extended from the MacKenzie River area in the east to west of the Lena River, and from the Arctic Ocean to the Pacific Ocean to south of the Okhotsk Sea in Northeastern Eurasia mostly possessed species that appeared to be unique to the area. We are not convinced of this, particularly in regard to the angiosperms. It has already been pointed out that the species limits of many Cretaceous angiosperm leaves are uncertain, but within the range of variation exhibited by the various forms, similar entities can be recognized in assemblages in various parts of North America.
The assemblages reported here have a number of elements which suggest particular affinities (on the basis of shared forms) with the Cenomanian Dunvegan Formation, the Albian to Cenomanian assemblages of the Dakota Group, the upper Cretaceous sections of the Potomac Group and some localities on Nelson Island, Western Alaska which may be as young as Turonian to Santonian but no older than Cenomanian.
According to Bell (1963) (p. 9) the age of the Dunvegan Formation is well established not only on marine fossils from overlying and underlying formations but on marine invertebrates from the Dunvegan itself. In addition Bell's study of the plant megafossils also suggests a Cenomanian age in that "the largest number of common species occur in floras of the Dakota, Woodbine, Tuscaloosa, Raritan and Magothy".
A maximum age for the Woodbine Formation has been established as Cenomanian on the basis of marine fossils in the underlying Washita Group, although the time interval represented by the depositional hiatus between the Washita and Woodbine is unknown (Bell, 1963).
In view of the secure age for the Dunvegan flora, its more northerly position, and floristic composition, the similarity between it and the assemblages from the Yukon-Koyukuk Province is particularly important. The significance of the similarity between the Dunvegan assemblages and those of the Yukon-Koyukuk Province was not lost on Bell (1963) when he suggested that the Albian age for the Melozi Formation was inconsistent with his analysis, and Bell's conclusions regarding the Cenomanian age for Hollick's Yukon material are not invalidated by analysis of the more recent collections.
Correlation with the Greenland floras described by Heer (1883) is difficult because of his inadequate descriptions and illustrations, the poor labelling of his specimens, and the results of more recent stratigraphic work relating to the rock units which yielded his material. According to Koch (1964), (p. 543) "Heer's correlation of the Atane Formation with the Cenomanian must be regarded with reservation" and quotes Heer as merely assigning the flora to the time interva l" *** between Gault and Senonian *** " (Heer, 1883) (p. 186) and that it was only "probably" of Cenomanian age.
Koch (1964) was of the opinion that plants in the Atane Formation are Senonian and that of all the material reportedly from the formation only those specimens from the "Bregnelaget" (fern bed) and "Liriodendronlaget" ('Liriodendron' bed) are sufficiently well labelled so as to allow them to be stratigraphically fixed. In spite of this probable Senonian age the 'Liriodendron' bed assemblage is floristically similar to that of locality 11556 and as the name suggests, one of the main components of the assemblage is a magnoliid form with an emarginate apex which Heer (1883) considered akin to the extent genus Liriodendron Linnaeus. Other magnoliid-like leaves are also present including Leguminosites ovalifolius Heer (1883) (Plate 27, Fig. 14) which is very similar to form MA63 except that Heer's specimens has less pronounced venation, and 'Myrsine' borealis Heer (1883) (Plate 24, Fig. 8, not Fig. 7b) which has identical venation to MA63 but is smaller.
The genus Podozamites Braun is also represented in the 'Liriodendron' bed, as is a species of Pseudocycas Nathorst (Cycas dicksoni Heer, 1883; Plate 16, Fig. 73). If the Atane strata are Senonian as Koch (1964) suggests then the older elements such as Podozamites apparently survived longer in Greenland than on the Arctic Slope of Alaska (Smiley, 1966, 1969a).
The upper Cenomanian New Jersey Raritan Formation (Zone IV of Doyle and Hickey, 1976) and the younger Magothy Formation (Santonian to Campanian; Doyle, 1969; Wolfe and Pakiser, 1971) have predominantly magnoliid-like forms or generalized 'species' of wide stratigraphic range (for example cf. Thyrsopteris decurrens Fontaine) in common with the Dunvegan and Yukon River localities. The generalized character suites typical of magnoliid-like leaves together with their somewhat conservative nature reduces the significance of correlations with the lower latitude younger rocks of the Magothy. In the Magothy the magnoliid-like forms may well represent a more conservative component of the flora predominantly found in swamps of lacustrine facies. The only platanoids common to the Dunvegan, the Yukon-Koyukuk Province, and the Atlantic Coastal Plain are forms referrable to the loosely defined, widespread, and stratigraphically long ranging 'Platanus' newberryana Heer. On the whole correlations based on shared forms between the Yukon-Koyukuk Province and the later Cretaceous successions of the Atlantic Coastal Plain are of doubtful value.
Correlations Based on Leaf Architecture
Whereas the Potomac Group cannot provide reliable correlations with the Yukon-Koyukuk Province assemblages on the basis of shared forms, it is of critical importance in the context of comparative leaf architectural studies largely because of the unique pioneering study of Doyle and Hickey (1976). The good palynological control, which is missing in many of the Alaskan successions, provides a rigorous stratigraphic framework into which temporal changes in leaf architecture may be placed. In their studies of plant fossil assemblages from the Patuxent Formation, Arundel Clay, Patapsco, and so-called 'Maryland Raritan' (Elk Neck locality of Patapsco Formation of Wolfe and Pakiser; 1971) Formations, Doyle and Hickey (1976) document sequential changes in the increasing degrees of vein organization, including the differentiation of veins into discreet orders, and consistency of vein courses.
In the lower Albian(?) Fredericksburg and Drewrys Bluff localities, Virginia, (Zone I) entire margined pinnately veined forms occur with poorly organized venation. Anastomosing veins enclose elongated rhombic areas of lamina (for example Rogersia angustifolia Fontaine, Celastrophyllum latifolium Fontaine).
In middle Subzone IIB (middle to upper Albian) palmately veined forms occur for the first time. In view of the range of morphology found in later platanoid forms it is perhaps significant that palmately veined peltate and lobate-cordate-ovate leaves intergrade with each other with respect to shape, venation, and margin characters. The palmately veined forms also show more regularity in their vein courses than Zone I leaves and such forms as Menispermites virginiensis Fontaine possess tertiary veins comprising an irregular reticulum, but with a tendency to form convex transverse bridges between the pectinal veins.
The middle of Subzone IIB also yields leaves with doubly convex glandular serrations as well as both pinnately and palmately lobed forms. Of interest with respect to later platanoids is the occurrence of palmately lobed forms (Araliaephyllum obtusilobum Fontaine) with palinactinodromous venation and a tendency for the slightly irregularly spaced secondary veins to be looped. The tertiary veins at the leaf base are irregularly percurrent becoming weak and irregular above. This type of leaf is apparently restricted to coarser sediments and shares both magnoliid and platanoid characters.
Upper Subzone IIB (upper Albian) sees a possible elaboration of the cordate-reniform complex. Leaves such as Menispermites potomacensis Berry have an ovate or shallowly lobate shape, entire margins, and approximately six radiating veins of primary order which divide and loop near the margin. Other leaves such as Populophyllum reniforme Fontaine from upper Subzone IIB also have radiating major veins which irregularly divide and rejoin forming poorly organized loops but are irregularly lobed. As with RAPE11, RAPE12, and RAPE13 the venation shows a tendency to be bilaterally arranged about a single midvein but overall the vein organization is poor compared to the Yukon specimens. That the Menispermites-type leaves of the Yukon-Koyukuk Province are evolved from the same complex as the Potomac forms cannot, as yet, be discounted.
Truly pinnately compound leaves first appear in upper Subzone IIB of the Potomac Group. These Sapindopsis-type leaves have leaflets with pinnate festooned brochidodromous venation and secondary and tertiary vein orders are readily distinguishable. With the possible exception of MA59 pinnately compound leaves are missing from the Yukon successions.
Palmately lobed leaves are also present in upper Subzone IIB (middle to upper Albian) and become dominant in Subzone IIC (uppermost Albian to lower Cenomanian). Confined to the coarser facies they are typified by 'Sassafras' potomacensis Berry. The looped secondary venation of these leaves is far less regular than anything seen in the Yukon material and the tertiary veins are weak, closely spaced, and irregularly and obliquely percurrent. This form continues into Subzone IIC by which time it has one of the most poorly organized venation systems of any of the lobed 'platanoids'.
The Subzone IIC 'platanoids' possess strong, rigidly organized, tertiary veins perpendicularly crossing broad intercostal areas, with subparallel transverse quaternary veins. The secondary veins of such leaves as Araliopsoides cretacea (Newberry) Berry are still brochidodromous and the tertiary veins less regular than seen in all but one of the Yukon platanoids. This particular form has a decurrent base although peltate bases are also common. Protophyllum multinerve Lesquereux (Subzone IIC - ? Zone III), which Doyle and Hickey (1976) consider a possible unlobed member of the same complex, does, however, display third and fourth order venation as well organized as any in the Yukon-Koyukuk Province.
The level of vein organization seen in the specimens from the Yukon-Koyukuk Basin is not achieved in the Potomac Group until the lowermost part of the Cenomanian. Clear differentiation of vein orders is seen in most of the Alaskan platanoid leaves, and is even displayed in the well preserved Menispermites-type leaves (RAPE11, RAPE12, and RAPE13) which posses lower vein orders with somewhat irregular vein courses. Furthermore the lack of any palmately veined forms with poorly organized tertiary and higher order venation must place the Yukon assemblages developmentally above even the Potomac Zone III assemblages.
The platanoid leaf from the Yukon-Koyukuk Province with the least number of advanced features is HAPL27. Here the pectinal abmedial veins are irregularly looped, inter-abmedial veins are present, and the tertiary venation is somewhat irregular. Fourth order veins display minimal tendency to run percurrently between the tertiary veins and are poorly differentiated from higher order veins. The teeth are apparently non-glandular with the medial vein terminating, without thickening, at the tooth apex.
The similarity in architecture between form HAPLD27 and Araliopsoides cretacea suggests that HAPLD27 is derived from the Araliopsoides complex. The level of vein organization in HAPLD27 is slightly lower than that seen in A. cretacea and HAPLD27 could easily be a relictual form.
A significant differences between the forms is the development of semicraspedodromous veins supplying the apparently non-glandular teeth of HAPLD27 in contrast to the entire margins of Araliopsoides Berry. Entire margined palmately veined leaves are absent in the Yukon assemblages; a fact which may be associated with cooler climatic conditions. The tertiary veins running between the basal superior secondary veins and the pectinal veins are distinct and convex percurrent in HAPLD27, whereas in Araliopsoides they have a tendency to be irregular or sinuous in course and irregularly branched. On the other hand the tertiary veins running between the abmedial veins of HAPLD27 are poorly organized compared to the regularly spaced, seldom branched, percurrent tertiary veins of Araliopsoides. The pectinal veins of HAPLD27 are weak and there is no evidence of lobing.
Lobed leaves from the Yukon-Koyukuk Province assemblages that are also clearly similar to those of the Araliopsoides complex are represented by form HAPLD2. In this form both the secondary and abmedial veins are craspedodromous and one of the abmedial veins is strengthened in a return to the palinactinodromous condition seen in Araliaephyllum obtusilobum of the Potomac middle Subzone IIB.
While clearly similar in gross venation, specimens USGS 11558.4 and USGS 11557.2 (HAPLD2) differ in a number of respects. The teeth of USGS 11558.4 are considerably coarser then those of USGS 11557.2 and the tertiary veins are more irregular in course and branching. As far as can be determined the poorly preserved specimen USGS 11558.1 is intermediate between these two leaves. The significance of these apparently more primitive characters of USGS 11558.4 cannot yet be evaluated, but in view of the usual variation found in such leaves they have here been included, perhaps erroneously, in a single form.
By the beginning of the Cenomanian the finer-grained facies of the Potomac Group yield predominantly simple, pinnately veined, second rank (sensu Hickey, 1971) leaves with poorly organized but broadly camptodrome venation. These leaves evidently possessed venation systems even less well organized than those of similar type found in the equivalent facies of locality 11556, which again indicates that the Yukon fossils are no older than early Cenomanian even from this locality which Hollick (1930) suggested might pre-date the others.