How Many Were There?

A Predicted Kind Boundary from Calibrated Genetic Drift

Part Three of the Diversification Series

Disclaimer: This paper was developed collaboratively between Claude (Anthropic) and D. L. White. It builds on the framework established in “When Did the Wolves Start Howling?” and should be read as a companion to that paper. The drift model, calibration data, and convergence results referenced here are documented with full appendices in the companion paper.

The Question the Wolves Raised

The second paper in this series demonstrated that a standard genetic drift model, calibrated against known-age dog breed data and validated across canids, equids, bovids, and Drosophila, produces diversification timescales of roughly 4,000 to 8,000 years for the origin of modern species from common ancestors. It closed by asking: how many genetically distinct founding kinds does it take to produce the full roster of species alive today? And does that number fit inside anything that floats?

This paper answers both questions. But first, it must resolve a prior one: where does one ancestor end and another begin? If wolves and coyotes diversified from a common canid ancestor, and horses and donkeys from a common equid ancestor, how do we know the canid ancestor and the equid ancestor were different starting points rather than branches of an even deeper common ancestor?

This is the “kind” boundary problem. Creation scientists have been working on it for decades under the name baraminology, using hybridization data, visual similarity, and statistical trait analysis. Their results have been productive — roughly 137 mammalian kinds and 196 bird kinds by the most recent comprehensive estimates. But the methods are observational. They describe where the boundaries appear to fall. They do not predict where the boundaries must fall from first principles.

The drift model from the companion paper offers a prediction. If the diversification window is finite — roughly 5,000 years — then there is a maximum amount of genetic differentiation that drift can produce in that time. That maximum is the kind boundary. Not because someone decided to draw a line there, but because the math runs out of room.

Deriving the Boundary

The drift equation that governs genetic differentiation between isolated populations is:

FST(t) = 1 − (1 − 1/(2Nₑ))^t

where Nₑ is the effective population size (roughly, the number of breeding individuals averaged over a population's history — typically much smaller than the total headcount) and t is the number of generations since the populations separated.

For a fixed diversification window of 5,000 years, the number of generations depends on the generation time: roughly 1,667 for canids (3-year generations), 1,000 for bovids (5-year), and 625 for equids (8-year). The maximum achievable FST depends on both the generation count and the effective population size — smaller populations drift faster.

But population size cannot be arbitrarily small. Populations below about 50 effective breeders are at serious risk of extinction from inbreeding and demographic stochasticity (random bad luck in small populations — a few bad breeding seasons in a row can end the line). Conservation biology calls this the minimum viable population. It represents the floor below which a founding lineage cannot sustain itself. A clarification is warranted here. The minimum viable population of 50 is derived from modern populations carrying thousands of generations of accumulated copying errors — deleterious recessive alleles that become lethal when exposed through inbreeding. Founding populations with undegraded genomes do not carry this mutational load, and inbreeding between maximally heterozygous individuals does not produce the same penalty. The effective floor for pristine founders is substantially lower than 50 — potentially as low as a single breeding pair. Using Ne=50 as the floor throughout this analysis is therefore a conservative assumption. It understates the diversification capacity of the model, which means the kind boundary derived here is narrower than the actual boundary may have been. The math has more room than it claims. Even so, very small founding populations remain subject to demographic stochasticity regardless of genome quality — random fluctuations in births, deaths, and sex ratios can end a lineage through sheer bad luck. Some founding kinds may have been lost this way, which means the original kind count may have been slightly higher than what the modern species roster implies.

It is worth noting what the single-pair case implies for starting diversity. Two diploid individuals can carry up to four unique alleles per locus. If both founders are fully heterozygous with non-overlapping alleles, the expected heterozygosity of that founding pair is approximately 0.75 — substantially higher than the H₀ ≈ 0.40 to 0.50 assumed in the companion paper’s drift model. The model’s assumed starting diversity is therefore conservative not only in its population size floor but in its starting heterozygosity. No parameter in the analysis is changed by this observation. The math throughout uses Ne=50 and H₀ ≈ 0.40 to 0.50. The point is that the framework has margin in both directions: the actual starting conditions implied by the text permit more diversification capacity than the model claims, not less.

Additionally, the initial small population size is not the liability it would be for a modern bottleneck. With empty ecological niches and no competition, population growth is explosive. Within a dozen generations, effective population sizes reach hundreds, then thousands. The early rapid drift at small Ne is precisely the mechanism that sorts the original variation into distinct lineages quickly — it is the engine of diversification, not a threat to it. The initial bottleneck drives speciation rather than endangering it.

The following table shows the maximum FST achievable in 5,000 years across the full range of plausible effective population sizes and generation times. The reader can examine exactly where any particular threshold falls within the predicted range.

Maximum FST achievable in 5,000 years:

Several patterns emerge from this table. At the minimum viable population size (Ne=50), nearly any generation time produces FST approaching 1.0 — essentially complete differentiation. At Ne=500, which represents a moderate founding population, FST ranges from 0.28 to 0.92 depending on generation time. At Ne=1000, the range is 0.15 to 0.71.

The critical observation is not a single threshold but a zone. For mammals with typical generation times of 3 to 8 years, at the biologically realistic Ne range of 50 to 200 for early post-founding populations, the maximum achievable FST clusters between approximately 0.79 and 1.00. At Ne=100 — a reasonable estimate for a kind in its first centuries of post-founding growth — the range is 0.96 (at 8-year generations) to 1.00 (at 3-year generations).

The predicted kind boundary is not a razor. It is the zone where drift runs out of room — and populations that have drifted past this zone can no longer find their way back to reproductive compatibility.

The Empirical Check

In December 2025, a meta-analysis of genomic data from hundreds of sister lineages of large mammals was published, testing whether genetic distance thresholds could predict taxonomic species status. The researchers found two empirical thresholds.

The species boundary — where taxonomists consistently draw the line between species — falls at an FST of approximately 0.26. Below this, populations are usually classified as the same species. Above it, they are usually classified as different species.

The hybridization-failure boundary — where Haldane’s Rule applies (hybrid offspring of one sex are infertile or inviable) — falls at an FST of approximately 0.55. Above this threshold, species can no longer produce fully functional hybrid offspring.

These thresholds were measured empirically from hundreds of mammalian species pairs. They were not derived from any model. The researchers were not studying created kinds or biblical timelines. They were doing conventional mammalian taxonomy.

Our drift model, calibrated against dog breeds and validated on wolves, independently predicts that the maximum achievable differentiation in the available timeframe clusters around FST ~0.55 at minimum viable population sizes. The predicted boundary from the model and the observed hybridization-failure threshold from the meta-analysis land on the same number.

Two roads. Same destination. The kind boundary is not arbitrary. It is the point where drift runs out of room, which is the same point where hybridization runs out of compatibility. The mechanism explains the observation.

Validation on Known Kinds

The predicted threshold should correctly separate populations within the same kind (FST below ~0.55) from populations belonging to different kinds (FST above ~0.55). We can test this on the three families validated in the companion paper. A methodological note: the ~0.55 threshold was derived from the drift equation applied to the time window and the minimum viable population floor. It was not fitted to any of the FST values used in this validation. The observed FST values below are independent measurements, making this a genuine out-of-sample check.

Within-kind FST values (should be below threshold):

Wolf-Coyote: FST ~0.40. Below 0.55. These are the most divergent members of the canid kind. They still hybridize. The model correctly places them within one kind.

Italian wolves vs Iberian wolves: FST = 0.293. Well below. Same kind, same species, different populations. Correct.

Dog vs Wolf: FST = 0.165. Well below. Same kind. Correct.

Horse breed maximum (Clydesdale vs Mangalarga Paulista): FST = 0.254. Below. Same kind. Correct.

Cattle breed global average: FST = 0.100. Well below. Same kind. Correct.

The threshold correctly classifies every validated case. Within-kind differentiation falls below the boundary in every instance where hybridization data confirms the populations belong together.

Between-kind FST values (should be above threshold):

The other side of the boundary must also hold. If the threshold correctly defines a kind, then populations belonging to different kinds should show FST values above the threshold — and should be unable to hybridize.

Direct FST measurements between different mammalian families (for example, Canidae vs Felidae, or Equidae vs Bovidae) are rarely published because at those levels of divergence, FST saturates — it approaches 1.0 and ceases to be a useful metric. Researchers use different distance measures for comparisons that deep. The absence of published inter-family FST values is itself informative: the distances are so large that the standard measure breaks down.

More to the point, no cross between members of different families has ever produced offspring of any kind. Dogs and cats cannot hybridize. Horses and cattle cannot hybridize. The idea is not merely untested — it is biologically absurd, because the genetic distance is so far beyond the hybridization-failure zone that reproductive machinery cannot engage at all. The 2025 meta-analysis found that hybridization begins to fail at FST ~0.55. Between-family distances are well past that threshold, past the point of partial fertility, past the point of embryonic viability, and into the zone where gametes (reproductive cells) cannot even recognize each other. At these distances, even the species-specific surface proteins that must match before fertilization can initiate are too divergent to engage—the molecular equivalent of a key that no longer fits any lock in the building.

The threshold separates correctly in both directions. Everything below it can hybridize to some degree. Everything well above it cannot hybridize at all. The boundary works.

The one potential exception is the river buffalo vs swamp buffalo comparison, where one study reported FST values up to 0.68 between Egyptian and Indonesian populations. However, this comparison used a SNP panel designed for river buffalo, which introduces ascertainment bias that inflates apparent differentiation. Within-region comparisons using the same panel produce FST values of 0.003 to 0.05 — consistent with one kind. The high value likely reflects measurement artifact rather than genuine kind-level separation.

From Threshold to Count

With the kind boundary established, the question becomes whether existing kind-count estimates are consistent with the model’s predictions. We did not independently derive the kind groupings from genetic distance data for every animal family on Earth. What we did is demonstrate, on the three families where we have detailed genetic data, that the drift model’s predicted boundary aligns with the observed hybridization-failure threshold. The kind count itself comes from existing taxonomy, not from our model. The model provides a framework to evaluate those counts, not to adopt them.

The reason this is defensible rather than circular is that three independent methods — our drift-based threshold, Lightner’s hybridization-based baraminology, and conventional family-level taxonomy — all converge on similar groupings without coordinating with each other. Taxonomists grouped animals into families based on morphological similarity. Baraminologists grouped them by hybridization capacity. Our model predicts a genetic distance ceiling that lands at the same level. When three different methods, using three different criteria, draw the same lines, the correspondence is not an assumption being recycled. It is a signal being detected from multiple angles.

Our specific contribution is the mechanistic explanation for why the family level is the right one: it is the level at which genetic distances approach the maximum achievable by drift in the available timeframe. The boundary is a consequence of the physics of population genetics operating over a finite window.

Starting from the published family-level taxonomy and filtering for land-dwelling, air-breathing vertebrates as specified in Genesis 7:15 and 7:22:

Mammals: 167 recognized families, minus approximately 15 marine families (cetaceans, sirenians, pinnipeds), leaving roughly 152 terrestrial families. Hybridization data indicates some families should be lumped — multiple canid subfamilies into one kind, for example — bringing the estimate to approximately 130 to 150 mammalian kinds. Lightner’s independent baraminology estimate of 137 mammalian kinds falls squarely within this range.

Birds: Approximately 249 recognized families. Hybridization data, particularly in the passerines where interfamily crosses are documented, indicates substantial lumping is warranted. Estimated bird kinds: 175 to 250. Lightner’s estimate of 196 avian kinds falls within this range.

Land reptiles: Approximately 75 to 80 families after excluding marine species. Estimated reptile kinds: 60 to 75.

Amphibians: Approximately 75 families. Whether amphibians require ark passage is debated on textual grounds, as many could survive in aquatic environments. If included: 60 to 75 kinds. If excluded: zero.

Insects and other invertebrates: Excluded. Genesis 7:22 specifies creatures “in whose nostrils was the breath of the spirit of life.” Insects respire through spiracles, not nostrils. Most creation scientists exclude them from the passenger manifest, and most insect species could survive the flood on floating debris, as eggs, or in larval forms.

Total estimated kinds: 425 to 550 for extant land vertebrates, depending on lumping decisions and whether amphibians are included.

This estimate addresses only extant kinds. Extinct kinds known from the fossil record — including dinosaurs, pterosaurs, and various synapsid groups — would add to the count. Several existing estimates can be evaluated against this framework: Lightner’s baraminology total of approximately 1,400 kinds, the Ark Encounter’s similar figure, and Woodmorappe’s earlier, more aggressive estimate of roughly 8,000 kinds (which used extensive splitting and extinct-kind inclusion).

The Passenger Count

The animal count follows from the kind count and the boarding rule. Genesis specifies two of each unclean kind and seven of each clean kind. Clean animals in the biblical context are a small subset — primarily livestock and sacrificial animals, perhaps 3 to 5 percent of the total kinds.

At 1,400 total kinds (the Ark Encounter’s comprehensive estimate including extinct forms): approximately 3,000 individual animals.

At 550 kinds (our extant-only upper estimate): approximately 1,200 individual animals.

At 425 kinds (our extant-only lower estimate): approximately 900 individual animals.

The feasibility of housing, feeding, and watering these numbers within the ark’s specified dimensions has been analyzed in detail by Woodmorappe in “Noah’s Ark: A Feasibility Study” (1996) and by the Ark Encounter research team. Their analyses account for space requirements, feed storage, water supply, waste management, ventilation, and animal husbandry logistics for a 371-day voyage. This paper does not reproduce that analysis. The interested reader can evaluate their methods and conclusions directly.

What This Paper Adds

The baraminologists defined kinds by observation — hybridization, appearance, statistical trait analysis. Their kind counts have been consistent across researchers, but the methods are empirical rather than predictive.

This paper provides the theoretical underpinning for why their counts come out where they do. The kind boundary falls at the family level not by coincidence or convenience but because the family level is where genetic distances approach the maximum achievable by drift in the available timeframe. The boundary is a consequence of the physics of population genetics operating over a finite window.

The predicted boundary (FST ~0.55 at minimum viable population sizes in 5,000 years) converges with the empirically observed hybridization-failure threshold (FST ~0.55 from the 2025 meta-analysis of hundreds of mammalian species pairs). This convergence was not designed into the model. It fell out of the same drift calibration used in the companion paper on canid and equid diversification.

The framework connects three previously separate lines of work:

First, the diversification timeline from the companion paper — which established that modern species within each kind diversified in the range of 4,000 to 8,000 years.

Second, the baraminology research that counted kinds through observational methods — arriving at approximately 1,400 kinds including extinct forms.

Third, the ark feasibility studies that demonstrated the physical logistics of housing the estimated animal count within the specified vessel dimensions.

What was missing was the link between the timeline and the kind count — a mechanism that explains why the kind boundary falls where it does and confirms that the diversification rate is consistent with producing modern species counts from the estimated kind count in the available time. The drift model provides that link. The system is self-consistent: timeline, boundary, kind count, modern species count, and vessel capacity all interlock without requiring any parameter to be forced.

What This Paper Does Not Claim

This paper does not claim to have precisely determined the number of kinds. The estimate range of 425 to 550 for extant land vertebrates carries uncertainty from lumping decisions, amphibian inclusion, and the inherent imprecision of mapping a continuous genetic distance metric onto a discrete kind boundary. The true count could be somewhat higher or lower.

This paper does not claim that the FST threshold of ~0.55 is an exact, sharp line. Biological boundaries are gradients, not walls. Some kind pairs may fall slightly above or below the threshold due to selection effects, gene flow, or ascertainment bias in the genetic data. The threshold is approximate and should be treated as a zone rather than a razor.

This paper does not claim that all extinct kinds have been identified. The fossil record is incomplete. Some kinds may have left no fossil trace. The total kind count including extinct forms is necessarily less certain than the extant-only estimate.

This paper does not reproduce the ark feasibility analysis. That work exists and can be evaluated on its own terms.

This paper does not claim to have resolved the mechanistic question of how coordinated adaptive differences between species — the morphological, physiological, and ecological distinctions that make a wolf different from a coyote, or a horse different from a donkey — were assembled from a common ancestor. The drift model addresses neutral genome-wide divergence, not the specific allelic combinations underlying functional traits. One observation is relevant here, however. If each kind’s founding genome was not merely diverse but architecturally complete — carrying not just raw allelic variation but the linkage relationships, regulatory elements, and epistatic interactions that produce distinct body plans when expressed in different combinations — then selection in different environments does not build adaptive combinations from scratch. It reveals combinations that were already present, preserving those that work and allowing drift to erode those that do not. The functional speciation question then becomes a question about the information content of the founding genome, which this paper explicitly leaves as someone else’s problem. But it is worth noting that the direction of the evidence — the staircase from the companion paper, where every descendant is a reduction of its ancestor — is consistent with an architecture that was front-loaded rather than gradually assembled.

The Connection

The second paper in this series ended with a convergence: fourteen populations from three unrelated animal families, plus a validation on fruit flies, all pointing to a common diversification onset in the range of 4,000 to 8,000 years ago. The paper noted the convergence and left the implications to the reader.

This paper takes the next step. If the convergence window is real, there is a maximum genetic distance achievable within it. That maximum predicts a kind boundary. The predicted boundary matches the observed hybridization-failure threshold. The kind count at that boundary matches the independent estimates from baraminology. The animal count at that kind count fits within the vessel dimensions specified in Genesis.

Each piece was derived or validated independently. The timeline came from breed-calibrated drift modeling. The boundary came from the same model applied to the timeframe. The threshold was confirmed by a conventional meta-analysis that had no interest in created kinds. The kind count was estimated by baraminologists using different methods entirely. The vessel feasibility was analyzed by engineers using the Genesis dimensions.

One further observation. The Genesis text specifies the boarding rule as pairs — two of each unclean kind, seven of each clean kind. The analysis throughout this series uses conservative population parameters: Ne=50, H₀ ≈ 0.40 to 0.50. The actual starting condition implied by the text is a single pair per kind with maximally diverse, undegraded genomes — a starting heterozygosity potentially near 0.75, with no inbreeding depression because no deleterious recessives have yet accumulated. Every parameter in the model has more room than it claims. The math was not adjusted to fit the text. The text describes conditions that give the math more room than it asked for.

Nothing was tuned to fit. The pieces interlock because they describe the same system from different angles.

The first paper in this series examined the physical event itself — the geological and geophysical conditions consistent with the simultaneous initiation of diversification across all kinds. That paper begins with a rhinoceros.