Eoconodon is unknown in Lancian faunas but the genus is a member of Puercan and early Torrejonian faunas of the Western Interior and possibly Texas. The samples of these faunas come from sites with different depositional and, most likely, taphonomic histories. They represent the fruits of different collecting techniques ranging from surface collecting to underwater screening. As emphasized by
Lofgren et al. (2005) these differences in sampling must be considered in any paleobiogeographic study.
Lancian North American Land Mammal Age. Biochronologically the boundary between the Lancian and Puercan is defined by the first appearance of the archaic ungulate Protungulatum and approximates but is not always coeval with the Cretaceous/Tertiary boundary. In contrast, the base of the Lancian is not well defined (Cifelli et al. 2004).
Wilson (2005) suggested that the ages of the Lancian local faunas in northeastern Montana fall within the last 1.8 m.y. of the Cretaceous. Lancian faunas from other areas are not as well constrained temporally and might well include some that are older (Cifelli et al. 2004).
Geographically, large samples of Lancian local faunas have been made from sites in both overbank and channel lag deposits discovered in Colorado, Wyoming, Montana, North and South Dakota, and northward into Saskatchewan and Alberta (Figure 1 and,
Cifelli et al. 2004.,
Figure 2.2). To the south a few hard won Lancian fossils have been recovered in the Dragon Canyon area, Utah, (Cifelli et al. 1999) and the San Juan Basin, New Mexico (Weil and Williamson 2000).
Remains of Eoconodon have not been discovered in any Lancian vertebrate fossil locality within the Western Interior. Some of these samples amassed through underwater screening at localities in the northern Western Interior are very large. For example UCMP collections from the two primary collecting localities in channel lag deposits within the Lance Formation, Wyoming (UCMP V-5620 and V-5711) include 3,249 cataloged mammalian specimens (UCMP online database). UCMP collections from Lancian localities in the Hell Creek Formation, Montana, include 1,471 cataloged mammalian specimens obtained from both overbank and channel lag deposits (Wilson 2005). Many of these collections include molars of Didelphodon vorax, which are only slightly smaller than the morphologically similar molars of the smallest known species of Eoconodon. Although non-tribosphenic in morphology, isolated molars of Lancian multituberculates rival the size of those of the smaller species of Eoconodon. It is unlikely that the absence of cheek teeth of Eoconodon from these collections is a product of hydrodynamic sorting favoring smaller fossils. In summary, given their geographic distribution, derivation from both overbank and channel deposits, and large size of some of the currently available samples, the lack of specimens of Eoconodon probably documents its absence from at least the northern Western Interior during the Lancian.
Early Puercan, Pu1 interval zone. The beginning of the Pu1 interval zone is defined by the first occurrence of Protungulatum; its end by the first occurrence of Ectoconus (Lofgren et al. 2004). At localities where the magnetostratigraphy can be clearly determined, all Pu1 local faunas are preserved in sediments deposited during magnetic polarity chron C29r. Radiometric age determinations of deposits in the Western Interior provide an age of 65.58 ± 0.04 m.y. for the Cretaceous/Tertiary boundary, which is within chron 29r; the age of the boundary between chrons C29r and C29n is taken at 65.16 m.y. (Wilson 2005). Thus the duration of the Pu1 interval zone is on the order of 420,000 years.
The geographic distribution of fossil localities yielding Pu1 mammalian local faunas is heavily biased in favor of the northern Western Interior. Pu1 local faunas have yet to be discovered in either the San Juan Basin, New Mexico, or the Dragon Canyon area, Utah. Currently, the southernmost Pu1 local fauna is the Littleton local fauna from the Denver Basin, Colorado (Middleton and Dewar 2004). Here and at the Pu1 Mantua locality, Big Horn Basin, Wyoming, most of the material was obtained through surface collecting and hand quarrying. Pu1 localities in the Hanna Basin, Wyoming, were sampled with these and screen-washing techniques. Two localities in Saskatchewan, Frenchman 1, and Long Fall have yielded samples of local faunas that appear to contain a mixture of Lancian and Pu1 mammals. Provisionally they have been interpreted as being of Pu1 age (Lofgren et al. 2004).
The largest samples of Pu1 local faunas and time-averaged (Lancian and Pu1) Bug Creek Assemblages come from the uppermost Hell Creek and basal Tullock Formations in northeastern Montana.
Archibald's (1982) study of the Pu1 Hell's Hollow local fauna in Garfield County, Montana, was based on 717 mammalian specimens (UCMP online database) obtained primarily by screen washing. The Bug Creek Assemblage, known primarily from large channel fillings in western McCone County, Montana, is a time-averaged accumulation of both Lancian and Pu1 vertebrate fossils deposited during the Pu1 interval zone (Lofgren 1995). The exuberantly productive Bug Creek Anthills and other sites in the valley of Bug Creek have yielded thousands of mammalian specimens.
Sloan and Van Valen (1965), for example, reported that after 10 weeks of field work about 27,000 mammalian teeth and fragmentary jaws had been collected at the Bug Creek Anthills locality, and they presented an analysis of taxonomic diversity based on 6,000 specimens. Sites in the McGuire Creek area to the south of the valley of Bug Creek yielding the Bug Creek Assemblage produced approximately 3,000 mammalian specimens that were analyzed by
Eoconodon is not represented in any of the large samples of Pu1 faunas or Bug Creek Assemblage localities found in northeastern Montana. Given the immense size of these collections the lack of specimens referable to Eoconodon probably reflects its absence from the area. As currently known, occurrences of Eoconodon in Pu1 local faunas are limited to two sites in Wyoming. Eoconodon copanus is typified on an isolated molar in the sample of the Mantua local fauna, Polecat Bench Formation (Van Valen 1978).
Eberle and Lillegraven (1998b) reported the presence of Eoconodon sp. in the Hanna Basin (Ferris Formation), Wyoming, based on an isolated upper molar that most likely is not referable to E. copanus.
It remains an open question if the lack of records of Eoconodon in smaller samples of Pu1 local faunas in other areas, e.g., Littleton Fauna, Denver Basin, Colorado, or the purported Pu1 faunas of Saskatchewan, is a product of small sample size or absence of the genus from these areas. In either case, the limited distribution of Pu1 occurrences of the genus suggests Eoconodon was an immigrant that dispersed into the Western Interior sometime during this interval zone. Presence of what appear to be two distinct Pu1 species of Eoconodon raises the question of the timing of their differentiation. Did it occur during the Pu1 interval zone or was it a product of a Cretaceous diversification in some other as yet un-sampled area?
Middle Puercan, Pu2 interval zone. Pu2 and Pu3 were defined in the San Juan Basin, New Mexico, and they have been clearly identified as far to the north as the Hanna Basin, Wyoming, and, possibly, in Makoshika Park, southeastern Montana. Attempts to recognize them farther northward and westward have been unsuccessful, probably because of paleobiogeographic provinciality. Because of the lack of suitable sediments in the San Juan Basin, the durations of the Pu2 and Pu3 interval zones have yet to be determined by radiometric age determinations. The Pu2 and Pu3 local faunas of this area are preserved in sediments deposited during magnetic polarity chron C29n (Williamson 1996). The boundaries between chrons C29r and C29n, 65.16 m.y., and C29n and C28r, 64.39 m.y. (Wilson 2005), include an interval of approximately 770,000 years.
In the San Juan Basin, faunas of the Pu2 interval zone are known from large samples obtained primarily by surface collecting and include records of three species of Eoconodon – E. coryphaeus, E. gaudrianus, and E. ginibitohia (Williamson 1996, Clemens and Williamson 2005). Surface collecting in a series of exposures of the North Horn Formation in the Dragon Canyon area, Utah, has yielded a small sample of the Gas Tank local fauna that is tentatively regarded as Pu2 (Lofgren et al. 2005). Eoconodon is not represented in this sample or in the surface collections from the sparsely fossiliferous Corral Bluffs and West Bijou Creek -1 localities in the Denver Basin, Colorado, which could be of either Pu2 or Pu3 age. A single fragment of an upper molar records the presence of the genus within the Pu2 interval zone in the Hanna Basin, Wyoming, where both surface collecting and screen washing techniques were employed. More material is needed to determine whether this specimen represents one of the species of Eoconodon known from the San Juan Basin, one of the species known from northern Western Interior sites, or a new species.
Late Puercan, Pu3 interval zone. Faunas that can be definitely assigned a Pu3 age are limited to the San Juan Basin, New Mexico; Dragon Canyon, Utah; and the Hanna Basin, Wyoming. The local faunas in the San Juan Basin are known from samples obtained by surface collecting. Two species of Eoconodon, E. coryphaeus, and E. gaudrianus, are documented but these collections are smaller than those of Pu2 local faunas (Williamson 1996,
2005). The sample of the Wagonroad fauna was derived by surface collecting of strata of the North Horn Formation exposed in Dragon Canyon, central Utah, that record the transition from magnetic polarity chron C29n to C28r. This fauna was assigned to the Pu3 interval zone by
Lofgren et al. (2004). It has yet to yield a record of Eoconodon but includes the type specimen of the "triisodontid" Goniacodon hiawathae Van Valen (1978). Eoconodon is lacking from the large Pu3 samples in the Hanna Basin.
Undifferentiated Pu2/Pu3 interval zones. The major collections of this interval are from sites in northeastern Montana, particularly Garfield County, where localities in the Garbani Channel deposits are very productive (Clemens 2002). Where the magnetostratigraphy has been determined, all of the collecting sites fall within magnetic polarity chron C29n. Radiometric age determinations obtained in northeastern Montana bracket them in an interval between the boundary between C29r and C29n – 65.16 m.y.– and an age determination of 64.52 ± 0.02 m.y. for a tephra preserved in the W coal, which overlies the Garbani Channel deposits, giving a duration of approximately 640,000 years (Wilson 2005). The Garbani Channel localities have been worked with a combination of hand quarrying followed by underwater screening. To date approximately 4,500 mammalian fossils have been cataloged (UCMP online database), but when fully curated the sample of this Pu2/Pu3 fauna probably will be at least three to four times as large.
Both Eoconodon hutchisoni and E. nidhoggi are represented in the Garbani Channel deposits, and E. nidhoggi is typified by a specimen from Purgatory Hill. Currently E. hutchisoni is only known from northeastern Montana. Isolated teeth from other areas have been tentatively referred to E. nidhoggi: Eoconodon sp. cf. E. nidhoggi from the Simpson Quarry, Bear Formation, south central Montana of Pu2/Pu3 undifferentiated age (Buckley 1994). A fragmentary lower molar from the Hiatt South locality, Makoshika State Park, southeastern Montana was tentatively referred to E. nidhoggi by
Hunter et al. (1997) who argued for a Pu2, possibly early Pu2, age assignment. Finally
Johnston and Fox (1984) identified three fragmentary teeth found at Rav W-1, Saskatchewan, as ?Eoconodon nidhoggi.
Fox (1997) and
Lofgren et al. (2004) tentatively assigned a Pu2 age to this locality. These occurrences suggest that E. nidhoggi ranged over the northern Western Interior during the Pu2 and Pu3 interval zones. The absence of records of E. hutchisoni in areas outside northeastern Montana could well be a product of small collection sizes.
Early Torrejonian, To1 interval zone. As noted above,
1995) suggested that Eoconodon might be present at a locality in the Tornillo Formation, Texas, which, on the basis of faunal composition,
Williamson (1996, p. 66-67) regarded as probably Torrejonian in age. A few isolated teeth from two early Torrejonian local faunas in northeastern Montana, the Farrand Channel and Horsethief Canyon local faunas, document the presence of two "triisodontids" (Clemens and Wilson 2009); one might represent a new species of Eoconodon. These records, added to To1 occurrences of species of Goniacodon and Triisodon, document the increased taxonomic diversity of "triisodontids" in the early Torrejonian.