White Sands footprints

What happened

1. 2019NPS and Bournemouth University researchers excavate human footprints on several buried surfaces at Locality 2.
2. 2021Bennett et al. (Science) report 61 human prints bracketed by Ruppia seed layers dated ~23,000–21,000 cal BP.
3. 2022Madsen et al. warn that hard-water effects may make the seed dates thousands of years too old; the team replies in Science.
4. 2022Haynes proposes the prints may be Clovis-age (~13,000 years); the team replies that this is highly unlikely.
5. 2022Oviatt et al. report a 1947 spring-grown Ruppia specimen that dates to ~7,400 cal BP.
6. 2022Rachal et al. propose that storm seiches carried the seeds in from deeper water.
7. 2023Pigati et al. report terrestrial-pollen 14C and OSL ages that agree with the seed dates.
8. 2024Rachal et al. use modern Ruppia to argue the prints may be at least ~7,500 years younger.
9. 2024Rhode et al. ('Unresolved') and Pigati et al. ('…is Resolved') trade critiques in PaleoAmerica.
10. 2025Holliday et al. publish 26 dates from two outside labs on linked lake and wetland beds that fit the LGM age.
11. 2025Michael Waters (Texas A&M) says he remains concerned about the White Sands radiocarbon ages.
12. 2025Rachal and Dello-Russo (J. Archaeol. Sci.) argue the LGM claim is flawed and call for more dating.

Objections 1 of 8 proved wrong

• ?
  Hard-water (reservoir) effect: the aquatic Ruppia seeds may have taken up old carbon, making the dates thousands of years too old.

  Dating or samples · Madsen, Davis, Rhode & Oviatt · 2022

  Unresolved — Pollen 14C and OSL (2023) and mud 14C from other labs (2025) agree with the seed dates; critics still argued for large offsets in 2024–25.
• ?
  A 1947 Ruppia specimen from a nearby spring dated to ~7,400 cal BP, showing local reservoir offsets can be very large.

  Dating or samples · Oviatt, Madsen, Rhode & Davis · 2022

  Unresolved — Holliday et al. (2025) argue spring-fed modern water is a poor analog for the runoff-fed Pleistocene lake margin.
• ?
  The seeds may have been reworked from older deposits or driven ashore from deeper water by storm seiches, so they give only maximum ages.

  Layers or association · Haynes; Rachal, Mead, Dello-Russo & Cuba · 2022

  Unresolved — Proponents report seeds in thin layers, often on their stems; Holliday et al. (2025) saw no evidence of turbulent water at dated lenses, though undated seed balls occur in one nearby trench.
• ✕
  The prints could be Holocene tracks made by people crossing much older sediment.

  Layers or association · C. Vance Haynes Jr. · 2022

  Proved wrong — Pleistocene megafauna tracks are interspersed with the human prints; later critics, including Haynes, argue for Pleistocene rather than Holocene ages.
• ?
  Modern Ruppia from Salt Creek shows a δ13C–age offset that implies the prints could be at least ~7,500 years younger.

  Dating or samples · Rachal, Dello-Russo & Cuba · 2024

  Unresolved — Holliday et al. (2025) reply that the hard-water effect tracks alkalinity, not salinity, and that spring-fed analogs do not fit the runoff-fed margin.
• ?
  The OSL ages are likely maximum ages (partial bleaching, water content) from one level below most prints; the pollen may be older reworked grains.

  Dating or samples · Rhode, Neudorf, Rachal, Davis, Madsen & Dello-Russo · 2024

  Unresolved — Pigati et al. (2024) reject each point; the 2023 OSL came from one level, and critics want OSL from more track layers.
• ?
  The 2025 bulk-mud dates may share the old-carbon problem; the origin of the dated carbon is unclear.

  Dating or samples · Michael Waters · 2025

  Unresolved — Holliday et al. (2025) report mud δ13C unlike the seeds and 55 dates consistent across both teams' work; no published reply yet on where the mud carbon originates.
• ?
  No artifacts or settlements left by the track-makers have been found.

  Contradicts the accepted model · Critics since 2021, as reported by Holliday · 2021

  Unresolved — Holliday calls it a fair question without a peer-reviewed answer; he argues brief crossings need not leave artifacts.

Evidence at the time within ~5 years of the claim

• Unambiguous evidenceYes — Prints show heel impressions, arches and toe pads consistent with modern humans; the critiques target the age, not the human origin.
• Documented contextYes — In situ prints on several excavated surfaces plus a track-bearing package, in a trenched sequence interbedded with seed layers; context of the dated seeds is contested (obj. 3).
• Independent dating methodsPartly — 2021: 14C on aquatic seeds only (2021 alone: no). 2023: pollen 14C and a one-level OSL minimum (>21.5 ka) agreed with the seeds — partial, since the OSL covers one level only.
• More than one labYes — 2021 seed dates came from one lab (2021 alone: no). By 2023 (LLNL) and 2025 (Univ. of Arizona, NAU), labs outside the original dating gave concordant results, within the 5-year window used here.
• Professional recoveryYes — Excavated by NPS, USGS and university scientists under an NPS permit.
• Peer-reviewed reportYes — Science, 2021, with a USGS data release of the dates.
• Open to criticsPartly — Dating data and 3D print models released; exact location withheld by law but available from NPS to legitimate researchers.
• Independent teamsPartly — Print dates came from one NPS–USGS–Bournemouth team (2021, 2023). A 2025 group led by original co-author Holliday added independent lake-bed dates.
• Outside stakeNo — No commercial, religious or nationalist motive documented; the authors declared no competing interests.

What would settle it for a field partner

1. OSL-date sediment directly above and below several track horizons (single-grain quartz or feldspar), blind, at a lab outside both camps, with water-content history modelled.Medium cost
   Would show: Ages of ~21,000–23,000 years above the upper prints would confirm the LGM age; ages near or below 16,000 would support a younger chronology.
2. Re-date conifer pollen from the seed layers with larger samples and a site-specific procedural blank at a second AMS lab, to cut the ±2,300–2,500-year errors.Medium cost
   Would show: Precise pollen ages matching the seeds would rule out a multi-millennial offset; clearly younger pollen ages would confirm one.
3. Thin-section micromorphology of archived seed-layer blocks to test whether the dated seeds grew in place or arrived as storm-carried seed balls.Low cost
   Would show: In-place growth in very shallow water would weaken the transport model; transported seed balls would support reading the dates as maximum ages.
4. Under an NPS permit, sample a fresh cut through the track horizons jointly with critics, splitting seed, pollen and OSL samples between labs chosen by each side.Medium cost
   Would show: Concordant results from both sides' labs would settle the age question either way.

What would change this record

OSL or terrestrial-carbon dates from several track horizons, run by labs outside the original team: ages below ~20,000 years would refute the LGM age (below ~16,000 would remove it entirely); concordant ~21,000–23,000-year results would move the case to vindicated.

Notes

• Status. All direct dates (seeds, pollen, OSL, mud; several labs) fit ~23-21 ka and later work builds on it, but two critic groups and M. Waters argued in 2024-25 that each method could read too old. An OSL test on more track layers remains undone.
• Accepted view then. People reached the Americas after the LGM: pre-Clovis sites of up to ~16,500 years were argued, and early post-LGM or older sites were questioned.
• Dating basis. AMS 14C on 11 aliquots of aquatic Ruppia cirrhosa seeds bracketing the track horizons, calibrated with IntCal20: 22.86 ± 0.32 to 21.13 ± 0.25 ka.
• Correction. The starter brief (catalog/briefs.json) says the 2023 pollen/OSL dating 'is widely seen as settling the age' and expected status 'vindicated.' 2024–25 critiques (Rachal, Rhode, Waters) mean the case is recorded as open, not vindicated.
• Correction. CLAUDE.md's starter list calls this 'largely vindicated'; project chat said it overturned 'Clovis-first.' But pre-Clovis sites to ~16,500 years were accepted by 2021; this claim challenges a later, LGM-era limit, not Clovis-first itself.
• Correction. The 2023 OSL result is a single minimum age (>21.5 ± 1.9 ka) from a single level within the footprint sequence (critics say below most trackways), not an age for each track horizon.
• Correction. Pigati et al. 2023 used independent methods, materials and an outside AMS lab, but came from the same core team. The partly independent study is Holliday et al. 2025, led by a 2021 co-author who did not take part in the excavation.
• Correction. Critiques go beyond the hard-water effect: storm transport of seeds (2022), OSL bleaching and pollen reworking (2024), bulk-mud carbon (2025), and the lack of artifacts.
• Independent-methods count = seed 14C, terrestrial-pollen 14C, OSL, and lake-mud 14C. The 2021 data release also lists 14 U-series ages on gypsum sand (24.8–100.6 ka, ±9–23 ka) that the age model did not use. Loren Davis said OSL ages from more footprint layers would 'very likely' resolve the dispute; the team says most layers are too thin to sample. The exact site location is withheld under US law. Features are coded for 2021–26 because the new dating fell within five years of the claim. Seed and mud 14C both date aquatic carbon and share the hard-water bias Waters raises; in 2025 only 21 of 26 new dates were on mud, 5 again on seeds, so the four methods are not fully independent here.

People and evidence

• Proponents. Matthew R. Bennett (Bournemouth University), David Bustos (National Park Service), Jeffrey S. Pigati (US Geological Survey), Kathleen B. Springer (US Geological Survey), Daniel Odess (National Park Service), Vance T. Holliday (University of Arizona)
• Critics. David B. Madsen, Loren G. Davis (Oregon State University), David Rhode, David M. Rachal, C. Vance Haynes Jr., Michael Waters (Texas A&M University)
• Evidence types. footprints, radiocarbon, osl, stratigraphy, plant remains, geology · 4 independent lines
• Record confidence. medium — Facts are well sourced (key papers read in full or as abstracts), but open vs vindicated is a judgment call, and several critiques were read only as abstracts.