Paleontological Arguments for
a Lower Cenozoic Flood Model

Given the importance that a Flood Model holds within creation science, the debate over the Flood/post-Flood boundary placement in the geological record is critical. That placement directly affects how we view the Genesis Flood narrative (specifically what the ‘clean’ and ‘unclean’ kinds represented), how we understand post-Flood diversification and dispersal of animals, and how we interpret the stratigraphic fossil record.

Geological features are commonly discussed in the boundary debate (see, for example, Whitmore and Garner [2008]), but a number of arguments have been made using fossils to support the Flood/post-Flood boundary at different positions in the stratigraphic record. The validity of an argument rests on the reliability of the evidence (the premise of an argument), transparency of all assumptions, and whether the conclusion logically follows. There have been several cases, for example, where authors have asserted that fossil data fits within their Flood Model, thereby ‘confirming’ their Flood Model—but ignoring the fact that the fossil data also fits within competing Flood Models. That is poor argumentation.

Here I review and summarize paleontological arguments that argue for a Lower Cenozoic Flood Model or against an Upper Cenozoic Flood Model. In a Lower Cenozoic Flood Model, the Flood/post-Flood boundary may vary a bit by location, but is often indicated as near the K-Pg boundary in most terrestrial deposits. Grouped arguments may share a similar structure, but focus on different specific elements to make a case.

Arguments relating to the Ark Kinds as baraminic lineages

1. If two or more living members of the same unclean terrestrial kind are found in the same fossil deposit, that fossil deposit represents a post-Flood environment.

In my first paper related to this topic, I noted (Arment 2014) that a Late Miocene fossil site in Knox County, Nebraska, included fossils of five living North American snakes: Pantherophis guttatus (cornsnake), Pantherophis obsoletus (ratsnake), Lampropeltis getula (kingsnake), Lampropeltis triangulum (milksnake), and Pituophis melanoleucus (bullsnake). These three genera are all capable of hybridization, so we know that all five species are part of the same Created Kind. If the Genesis Flood narrative is to be taken as historically accurate, we must take seriously the multiple statements within those passages that only one pair of every unclean terrestrial kind was taken on the Ark. This means that only one pair of this colubrid kind (however broadly it may encompass living colubrid snakes) on the Ark. There were not two separate pairs of different Pantherophis species, there were not two separate pairs of different Lampropeltis species, and there were not separate pairs of all three of these genera. There was only one pair of a colubrid snake which was ancestral to all three of these genera. This means that any fossil site with more than one of these species must represent a post-Flood environment.

This argument can be made with any unclean terrestrial Created Kind where multiple living species are found in a fossil deposit (Arment 2020b). It is most effective when there is available hybridization data to securely place different species or genera in the same Created Kind.

2. If two or more members of the same unclean terrestrial kind are found on both sides of a proposed Flood/post-Flood boundary, that boundary is refuted.

This is very similar to Argument 1, except that it does not require the animals in question to be extant, or to be in the same location. Fossil animals that are reasonably asserted to be within the same Created Kind (supported by statistical baraminology or other evidence) may be used (Arment 2020b). As Lynx, Puma, Felis, and Caracal are all known from Pliocene deposits around the world, and are still living today, this argues against a Pliocene-Pleistocene boundary.

Arguments relating to the continental placement of fossils

3. The extremely high percentage (a statistical improbability) of animals which would have had to migrate from their original landforms to the Ark, then back to the location where their ancestors were buried in Flood deposits, argues against an Upper Cenozoic Flood Boundary.

This original argument was made by Ross (2012a), noting that 27 North American mammal families demonstrated Pliocene-Pleistocene boundary crossing, with 70 genera being found both below and above the boundary. Ross noted that “these data are more naturally interpreted as representing time-sequential recolonization of the post-Flood world by diversifying terrestrial mammal baramins.”

Closely following the argument above, in my investigation of the Australian marsupial fossils, I examined the likelihood of a Flood/post-Flood boundary on that continent at three separate stratigraphic locations (Arment 2020a). Numerous boundary-crossing genera were found for each. Within any Flood model, the original organisms discovered in Flood sediments could have originated on different pre-Flood landforms which may or may not still be in existence. So, the marsupial fossils found in Australia may not have originated in Australia, if they were carried there by Flood sediment. But why would so many living marsupial genera now be living solely on the continent where their pre-Flood relations were buried? For the most commonly suggested Upper Cenozoic boundary (Pliocene-Pleistocene), the probability of the same marsupial genera showing up above and below the boundary in Australia, was calculated to be 4.42x10^-45. That is a ridiculously low probability, but Australian marsupials are not alone in this issue.

The same low probability arises when we look at many different Ark kinds around the world. For example, I looked at endemic North American terrestrial reptiles, and found at least 29 genera that are found both above and below the Pliocene-Pleistocene boundary (Arment 2020b). In looking more broadly at terrestrial mammals around the globe, endemic Pliocene-Pleistocene boundary crossers included Africa (96 genera), Asia (35 genera), Australia (66, including bats and monotremes), Europe (30 genera), North America (60 genera), and South America (55 genera) (Arment 2025). So at least 342 mammal genera crossed that boundary, being found only on a single continent. That is incredibly implausible.

4. The very low percentage of animals from Old World Neogene deposits showing up in North American Pleistocene deposits, is evidence against a Pliocene-Pleistocene boundary.

This argument emerges from Ross’ (2012a) evaluation of North American boundary-crossing mammals, being the ‘flip side’ of the previous argument. In a ‘sweepstakes’ route within post-Flood dispersal, there would be no reason to expect that only animals with near relations in the North American Paleogene-Neogene record would show up in North America. We can, of course, broaden this argument to include both South America and Australia, as they also rely on ‘sweepstakes’ routes rather than ‘open bridges’.

As I noted (Arment 2025), “If it is just random chance that positions a genus solely above its ancestors of the same genus, on the same continent, then we should also see many other genera solely on a single continent after the Flood, while all their pre-Flood buried ancestors of the same genera are solely on a different continent. Where are all those genera?”

Arguments relating to boundary placement

in the stratigraphic record

5. The sheer number of genera within the same families (putative Created Kinds, or at least monobaramins) found both above and below Upper Cenozoic boundaries make those boundaries highly improbable for a Flood model.

An excessive number of genera crossing the Pliocene-Pleistocene (also known as the Neogene-Quaternary) boundary explicitly demonstrates the unlikelihood of that boundary for Flood models. Within the Canidae, at least ten genera cross that boundary, while at least thirteen genera of the Felidae do so (Arment 2020b). This means that there would have been at least ten pairs of canines and thirteen pairs of felines on the Ark. That turns the Ark’s ‘clean kinds’ and ‘unclean kinds’ into meaningless phrases—why bother saying that only one pair of every unclean kind was on the Ark, when Noah had to deal with ten pairs of wolf- and fox-like canids? Beyond that, extrapolation of genus-survivorship to all the other Ark Kinds would require an armada of Arks to save all those animals. Note, this particular argument does not rely on stratigraphic ordering of the fossils themselves, beyond placement above or below a proposed boundary.

6. A biostratigraphic break should be expected if a given stratigraphic position is to be considered for the Flood/post-Flood boundary placement.

A proposed boundary lacking a distinctive biostratigraphic break is probably incorrect. Ross (2012b) evaluated forty-three North American tetrapod families (247 genera) at the K-Pg boundary, noting 14.2% survivorship (though these were not strictly terrestrial animals). In comparison, a higher percentage of survivorship was noted in mammal families at the Pliocene-Pleistocene boundary. Ross noted that this was a preliminary analysis, but it has potential for further investigation.

7. The fossil record in the immediate aftermath of the Flood/post-Flood boundary deposition is likely to demonstrate higher diversity in families and lower diversity in genera and species.

This argument derives from Whitmore and Wise (2008), and their observations in the Green River Formation. Within this Eocene formation, which would be very near the boundary in a Lower Cenozoic Flood Model, there are some 120-154 species in 85 genera of terrestrial mammals, in 41 families. This averages to 2.1 genera/family, 1.4-1.8 species/genus. If this fossil site represents a post-Flood dispersal site within decades of the Flood, that suggests ‘low diversity’ dispersion initially, with higher speciation further along the dispersal trajectory, or as the authors suggest “at the termini of post-Flood dispersion paths.” This suggests a possible test for boundary placement: do comparable Paleogene fossil sites demonstrate similar levels of diversity in family, genus, and species?

Arguments relating to the stratigraphic order of fossils

8. The Post-Flood Fossil Continuity Criterion (PFCC) posits that “high preservability baramins should have a continuous fossil record back at least to the Flood, and high preservability groups with a first-appearance in the fossil record substantially after the Flood are sub-baraminic.”

Wise (2008) proposed the PFCC as a way to determine the Flood/post-Flood boundary, noting that “if mammal preservability was constant following the Flood,” this suggests a boundary between the middle Eocene and lower Oligocene, while “if (as is more likely) mammal preservability rose with population following the Flood, the boundary is lower”. Wise suggested that many mammalian holobaramins might be identifiable with superfamilies or even higher groups.

9. The continuity of boundary-crossing species at each delineated stage can be used to suggest pre-Flood/Flood and Flood/post-Flood boundaries.

Wise and Richardson (2023) evaluated overall patterns and percentages of boundary-crossing species from the Precambrian through the Cenozoic. The high boundary-crossing values derived from this research indicate that the Flood/post-Flood boundary is best placed in the early Cenozoic.

10. The heavy bias in survivorship (crossing the Pliocene-Pleistocene boundary) among genera in the uppermost stratigraphic layers of the Cenozoic, argues against an Upper Cenozoic Flood Model.

When I looked at terrestrial mammal presence across Cenozoic epochs, 640 out of 5,586 genera crossed the Pliocene-Pleistocene boundary (Arment 2024). Of those boundary crossers, 0% were found in Paleocene deposits, 0.8% in Eocene deposits, 4.1% in Oligocene deposits, 52.8% in Miocene deposits, and 91.6% in Pliocene deposits. As all Cenozoic genera should have been on the Ark (within an Upper Cenozoic Flood Model), there is no reason we should see such a high survivorship bias by the fossil genera in the uppermost layers. Rather, the pattern of mammal presence reflects post-Flood diversification and dispersal over time throughout the Cenozoic.

Arguments relating to post-Flood markers

11. In situ fossilization in a Cenozoic layer where the process requires a time period greater than the length of the Flood, did not occur during the Flood.

The best example of this are the giant Miocene ichnofossils now known to be the fossil burrows of a burrowing beaver, Palaeocastor (Arment 2023). Fossilization of these burrows only occurred due to silicification after complete infiltration of the burrow walls by grassland plant roots. The size of these burrows, upright helical burrows up to 9 feet high, ending in a sloping, nearly horizontal tube almost 15 feet in length, dictates a very long period of time for grassland plants to reach and infiltrate the walls. In burrows that were active, roots remained in the walls, but in abandoned burrows the roots infilled the entire burrow. Today, grassland root growth to that extent would take 3-5 years. These fossil burrows are found over hundreds of square miles in western Nebraska and eastern Wyoming, with some burrows staggered above each other in different layers of sediment. Clearly these were not formed during ‘brief exposure’ periods during the Flood, as some creationists argue.

12. Fossil material that is discovered within cave sediments and cave flowstone, rather than original host-rock, will be post-Flood.

A similar argument placed Neanderthal material found in cave sediments in a post-Flood context (Wood 2022). For an argument to specifically support a Lower Cenozoic Flood Model, however, the fossil material must be in strata below a suggested Upper Cenozoic Flood Model’s proposed boundary. Pliocene or Miocene material would be necessary to argue against a Pliocene-Pleistocene boundary, for example.

Fossils found in caves include animals which were permanent residents, temporary residents (including use for den sites), seasonal hibernators, prey dragged in by predators, those randomly killed by falling into trap caves, and those transported after death by flooding or other means (Lundelius 2006; Plotnick et al. 2015). Cave sites include recognizable caves, as well as cave deposits in ‘unroofed caves’, where the top sediments have been removed through karst denudation (Arena et al. 2014). (Karst is a landscape formation in which soluble carbonate bedrock has been dissolved by acidic water, leaving behind caves, sinkholes, underground rivers, and similar features.)

A middle Miocene cave deposit in Queensland, Australia, has been interpreted as a pitfall trap (Arena et al. 2014). This is an ‘unroofed cave’, recognized by a delineated lithology, speleothem features, and abundant bat fossils. A late Miocene cave site in Spain has also been interpreted as a pitfall trap (Domingo et al. 2013), pointing to a very high abundance of carnivores and very few herbivores. As the cave host rock in each case was fully lithified for cave formation, this points to the fossils in those cave deposits being part of a post-Flood biota.

Care must be taken in evaluating paleokarst sites, to properly distinguish between actual cave deposits and infill from the dissolution and transport of bedrock fossils. Generally, cave fossils are found back to Miocene deposits, while earlier paleokarst fossils are mostly ‘fissure fills’ (Plotnick et al. 2015). Post-Flood caves can act as natural traps for eroding surface sediments, including those that include Flood fossils (e.g. Carboniferous paleokarst ending up trapped as cave fill in north-central Illinois [Plotnick et al. 2009]).

Further Discussion

Reactions by proponents of other Flood Models to these varied arguments have been to claim that 1) the Ark Kinds do not represent Created Kinds, but included multiple lineages within each Created Kind; 2) each single Ark Kind pair quickly produced offspring that were virtually identical to a wide range of pre-Flood relations via hyper-convergence; 3) the strata that incorporate multiple living members of the same kind may actually be post-Flood, and are simply misidentified as being below the Pleistocene; 4) the geological column is an evolutionary fabrication; 5) fossil identification is arbitrary and fraudulent.

Rebuttals to these assertions are discussed by Ross (2014a; 2014b) and Arment (2020b; 2022).