Website Banner


The Yukon-Koyukuk Region

Introduction Geology Hollick Localities Spicer Localities Locality Relationships Wider Correlations  



The Yukon-Koyukuk Basin occupies an area of approximately one sixth of the total land area of Alaska and is filled with ~ 8km of Valanginian to Santonian sediments. It is bounded to the west, north, and southeast by older and more highly metamorphosed rocks of the Seward Peninsula (Seward terrrane), Brooks Range (Arctic Alaska terrane), and Ruby geanticline (Ruby terrane) respectively (Jones et al., 1987), and extends southwestward under the modern Yukon River delta.

The Yukon-Koyukuk basin can be divided into two sub-basins, the Kobuk-Koyukuk to the north and east and the Lower Yukon to the south and west (Patton, 1973), separated by Middle Jurassic to Lower Cretaceous marine volcanic and volcaniclastic rocks generally referred to as the Hogatza trend or Koyukuk terrane (Jones et al., 1987; Box and Patton, 1989). This volcanic pile is intruded by Early to mid-Cretaceous plutons (Miller, 1989). Movement along the Kaltag Fault has divided and offset these sub-basins.

Schematic map of the geological areas within the Yukon-Koyukuk Basin and surrounding areas   Map showing the location of the Yukon-Koyukuk basin and sub-basins, surrounding terranes, some major faults, and key geographic localities (modified from Beikman (1978), Box et al. (1985) and Nilsen (1989). Abbreviations of villages: A, Aniak; B, Bettles; H, Hughes; K, Kaltag; R, Ruby; U, Unalakleet. Abbreviations of faults: ATCP, AniakThompson Creek fault; CF, Chiroskey fault; SFM, South Fork-Malemute fault. A more deatiled geological map of the area can be accessed here.


Plant fossils and mollusc remains from the sandstone cliffs on the west bank of the Yukon River near Nulato were first reported by Dall (1868) and considered to be Miocene in age (Dall, 1882).  The rocks were subsequently referred to as the Nulato "sandstones" and Kenai Group by Dall and Harris (1892)Spurr (1900) included a report on the plants by Knowlton and considered the Kenai "series" and Nulato "sandstone" to be Tertiary in age. In 1889 Schrader collected marine molluscs, which were subsequently identified by Stanton to be Late Cretaceous in age, from the Nulato "sandstone" between Nulato and the mouth of the Koyukuk River (Schrader, 1900) but a Late Cretaceous age was not assigned to the Nulato Sandstone. 

In 1904 Schrader published a list of species with a description of their occurrence in a discussion of the correlation of the Nanushuk "series" of the Arctic Coast. Brooks (1902) considered the Nulato "sandstone" to be closely associated with Upper Cretaceous beds, but it was following the collection of marine brackish or fresh water invertebrates and plants by Collier in 1902 that the Late Cretaceous age for some of these beds was firmly established.  A few of the plants were thought to be Tertiary and some Early Cretaceous in age, but Hollick (1930) later considered them all to be Late Cretaceous. 

Collier (1903) in a preliminary report stated that the coal-bearing beds of the Lower Yukon consisted of "sandstones, shales and conglomerates".  He regarded these as forming an uninterrupted sedimentary series, ranging in age from middle Cretaceous to the Upper Eocene.  Collier (1903) eventually erected three divisions:  freshwater cycad-bearing beds of probable Early Cretaceous age; marine and freshwater coal-bearing sandstones, conglomerates and shales yielding plants determined by Knowlton (Collier, 1903) to be of Late Cretaceous age; and the apparently conformable plant-bearing Kenai "series of supposed late Eocene age". 

Knowlton (1904), studying Collier's original plant collection together with additional material obtained by Hollick in 1903, was doubtful of the Tertiary age.  He saw strong similarities with the mid- and Late Cretaceous floras of Bohemia.

Smith and Eakin (1911) examined cliff exposures along the Yukon River between Melozi and Kaltag and divided the rocks into two named units:  the Ungalik Conglomerate and the overlying, now-abandoned, Shaktolik Group consisting of sandstones and shales.  The Shaktolik was subdivided into two unnamed units; the lower having a preponderance of sandstone over shales, and the upper being mainly shale.

Martin (1926; and in Hollick, 1930) also recognized the Ungalik Conglomerate and stated that it consisted of about 300 m of conglomerate, sandstone and sandy shale yielding fossils of "undetermined marine pelecypods, worm tubes, trails and indeterminate vegetable detritus" (Hollick, 1930, p. 23).  This was supposedly overlain, apparently conformably, by the Melozi Formation which consisted of freshwater shale and sandstone, 300 m or more in thickness, and contained plant fossils and freshwater invertebrates.

According to Martin (1926) and Hollick (1930) the Nulato Formation overlies the Melozi Formation, again apparently conformably, and is composed of marine sandstone and shale about 300 m thick, almost devoid of plants but yielding marine invertebrates.

The coal-bearing Kaltag Formation was considered to overlie conformably the Nulato Formation.  In addition to coal beds, the Kaltag Formation, at least 240 m thick, consists of freshwater sandstone and shale and has yielded mainly freshwater invertebrates and plant fossils.

In his study of the Late Cretaceous Dunvegan, Bad Heart, and Milk River floras of Western Canada, Bell (1963) attempted to compare his plant fossils with those described by Hollick (1930) from the Melozi, Nulato, and Kaltag Formations.  Bell (1963) was forced to conclude that the Melozi and Kaltag units were of Cenomanian age.  The Nulato Formation, supposedly overlying the Melozi, had however been shown to yield abundant shallow-water molluscs of Albian age including Inoceramus altifluminus McLearn and Gastroplites kingi McLearn (Patton and Bickel, 1956; Patton, 1973).  This inconsistency led Bell to suggest that the original stratigraphic sequence was in error and that the Melozi "may be found eventually to belong to a group that includes the Kaltag and be stratigraphically above the Nulato" (Bell, 1963) (p. 10).

In mapping the area, Patton and Bickel (1956) departed from the earlier stratigraphic divisions.  A border facies succession and an interior facies succession were recognized, each of which was subdivided into unnamed map-units. Patton (1973) subsequently abandoned the Shaktolik Group and instead recognized the following four broadly gradational lithologic units within the Cretaceous sedimentary rocks of the Yukon-Koyukuk Province:-

1. During Early Cretaceous (Albian) time, large quantities of marine turbidites (~ 6 km thick) consisting of volcanic graywacke and mudstone accumulated in two basins separated by a ridge of Neocomian andesitic rocks.  This volcanic high extended from Kotzebue Sound and the Seward Peninsula to the Koyukuk River valley.

2. To the west the turbidites are overlain gradationally by calcareous graywacke and mudstone up to 1500 m in thickness.  These calcareous strata were deposited in a broad north-south trough along the western margin of the Yukon-Koyukuk Province, and at the western edge grade into non-marine conglomeratic coal-bearing strata.  On the basis of rare shallow-water molluscs of probable Albian age (Patton and Miller, 1968) and the presence of both Early and Late Cretaceous pollen in the coal-bearing facies, the calcareous graywacke and mudstones are dated as Albian to early Late Cretaceous.  Non-marine deposits which overlie the calcareous assemblage in the Waring Mountains have been radiometrically dated as 84 Ma (Patton and Miller, 1968).

3. The calcareous graywacke and mudstone unit is apparently correlative with shallow-marine and non-marine beds of sandstone, siltstone, shale, and coal which overlay the volcanic graywacke and mudstone unit to the south east.  Approximately 3000 m thick, these coal-bearing beds represent a regressive succession grading upward from marine sandstone into non-marine shale, siltstone, sandstone, and coal with locally interfingering strand line sandstone and quartz conglomerate.  These beds were apparently deposited in a broad north-south trough on the eastern margin of the Yukon-Koyukuk Province with a source area in the Kaiyuh Mountains and Kokrines-Hodzana Highlands to the south east.  The exposure of these coal-bearing beds along the Yukon represents the Melozi, Nulato, and Kaltag Formations of earlier workers and are the main plant-bearing horizons of Hollick (1930).

4. The fourth lithologic unit recognized by Patton (1973) is a narrow band of quartz conglomerate along the northern and north-eastern margins of the Province yielding poorly preserved plant fossils, and a biotite from an interbedded ash-fall tuff has been radiometrically dated as 83.4 ± .2 Ma (Santonian/Campanian) (Patton and Miller, 1968).

Although widely used no basin-wide map showing the distribution of these four units has been published and the nature of the boundaries between the units has not been defined. Nevertheless Nilsen and Patton (1984), based on examination of the Lower Yukon sub-basin north of the Kaltag fault, concluded that:

(1) the axial volcanic graywacke and mudstone unit consists of northeast-transported turbidites that grade northeastward from middle-fan to outer-fan deposits;

(2) the western calcareous graywacke and mudstone unit consists of east transported alluvial fan, fan delta, and shelf deposits; and

(3) the eastern sandstone, siltstone, and shale unit consists of northwest transported deltaic deposits that include nonmarine delta plain and marine delta front and prodelta deposits.

Deltaic facies are most abundant in the southern part of the Kobuk-Koyukuk sub-basin where they crop out along the Yukon River between Ruby and Kaltag. Paleocurrent directions are variable but overall the deltaic sediments predominantly show progradation towards the northwest (Nilsen, 1989) and are provenanced from the Seward, Arctic Alaska and Ruby terranes.

Patton (1973), referring to Patton and Bickel (1956) and Bell (1963) suggests that the Melozi probably represents a marginal-facies equivalent of the Kaltag but points out the age of the plant-bearing beds as a whole remains in doubt.  The occurrence of the Cenomanian molluscs in the underlying marine beds on the Anvik River (D. L. Jones, written communication referred to in Patton, 1973) means that at least some of the plant-bearing beds are of that age or later. Comparison of the plant megafossils from the Melozi, Nulato, and Keltag Formations (Patton and Miller's unnamed map unit 3) with those from elsewhere in N. America, Greenland and N.E. Russia, comparative vein architecture, the marine biostratigraphic constrains and the radiometric constraint of Patton and Miller (1968) referred to above suggest an age bracket of Cenomanian to Santonian, with most of the material being Cenomanian and Turonian in age.