At Zachełmie, feet arrive before bodies—but the prints remain on trial

Two researchers examine the exposed dolomite surface at Zachełmie. The IUGS page identifies its shallow impressions as tetrapod footprints; a close reading must keep the physical traces separate from that biological interpretation. Photograph by M. Hodbod, via IUGS.[8]

A footprint is a fossil organized around an absence. There is no foot at Zachełmie Quarry, no skeleton lying at the end of the trail, not even a named animal that can be fitted neatly into the impressions. There are hollows, rims of displaced sediment, repeated left-right patterns and, in the best examples, short projections interpreted as digits. The body that made them has vanished so completely that the marks must testify for it.

The lead photograph keeps that uncertainty visible: two researchers, a measuring tape and a broad rock surface whose shallow marks do not resolve themselves into an animal at first glance.[8] The image is documentary evidence of the site, not a substitute for the morphological case.

That testimony carries an extraordinary claim. In 2010, Grzegorz Niedźwiedzki, Piotr Szrek, Katarzyna and Marek Narkiewicz, and Per Ahlberg described the Zachełmie traces as the footprints of limbed vertebrates from the early Eifelian Age of the Middle Devonian, close to 390 million years ago. On the timescale used in the original paper, they preceded the earliest known tetrapod body fossils by about 18 million years and the oldest then-known elpistostegalians—the fish group containing Panderichthys and Tiktaalik—by about 10 million.[1] Later calibrations have altered the numerical gaps, but not the central mismatch: if the identification is right, feet entered the record long before bodies did.[5][6]

The age and the identity are different questions. The bedding interval can be placed in the early Middle Devonian even if its impressions turn out to have been made by fish. Conversely, a shape can look foot-like without carrying its own date. Zachełmie became consequential because a stratigraphic argument and an anatomical argument met on the same slabs. Reading the find well means testing each without allowing the drama of the conclusion to stand in for evidence.

First read the surface, not the family tree

The track-bearing beds belong to the Wojciechowice Formation in Poland's Holy Cross Mountains. They are fine, laminated dolomitic sediments crossed by signs of shallow water and periodic exposure: microbial surfaces, desiccation features, small fossils and burrows. The traces occur at three horizons. The 2010 supplementary record catalogued roughly 220 individual marks and nine proposed trackways, while judging only about 20 impressions informative enough for detailed ichnological work.[1]

That ratio is the first useful restraint. A bedding plane crowded with ovals is not automatically a walking surface. Soft sediment records pressure imperfectly: the same foot can make unlike prints as water content, mat thickness, entry angle and the firmness of underlying mud change. A true surface impression may later be filled and preserved as a raised natural cast; pressure can also deform layers below the one actually touched. Even the authors who identified tetrapods divided the material into several preservation types rather than treating every hollow as an equally clear foot.[1]

The strongest specimens carry more than an outline. Some have a depressed central area, a compacted zone beneath it and a rim where mud appears to have been pushed outward. One large isolated impression preserves a fan of digit-like marks. Repeated depressions form trackways rather than random clusters, including a short sequence interpreted as a lateral-sequence walk and others read as a more nearly symmetrical gait. The proposed manus and pes impressions differ in placement and size. No continuous belly or tail groove accompanies the quadrupedal sequences.[1][5][6]

None of those observations names the maker by itself. Displacement demonstrates contact with soft substrate. Repetition implies movement or repeated behavior. Digit-like projections make a foot with toes more plausible. The tetrapod case is cumulative: several properties occur together in ways the describers argue are most economically explained by paired, digit-bearing limbs.

A trackway is an action, not an identity card

Body fossils and trace fossils fail in complementary ways. A skeleton can preserve joints, teeth and fin rays, yet the carcass may have drifted away from the habitat where the animal lived. A trackway fixes an organism in a place while it was alive, but it usually withholds nearly every feature used to name a species. Zachełmie therefore records a behavior with no securely associated body.

The 2010 interpretation goes furthest where the prints are most coherent. The largest impressions were about twice the linear dimensions of the best-known Ichthyostega foot, leading to a rough maximum body-length estimate near 2.5 metres. Short, triangular digit impressions were compared with an anatomically reconstructed Ichthyostega hind foot. Other marks were interpreted as kicks or slips made in shallow water.[1][5] These comparisons establish an envelope of possible makers, not a portrait. There is no basis for assigning the tracks to Ichthyostega, and the named animal lived much later.

The missing skeleton also blocks a seductive claim: Zachełmie's trackmaker cannot be installed as the direct ancestor of later tetrapods. If genuine, the tracks require an early branch of limbed vertebrates somewhere in the Eifelian. They do not reveal whether that branch led toward the tetrapod crown, ended without descendants, or belonged to a body plan unlike the Late Devonian forms currently known.

This is why the find does not make Tiktaalik obsolete. Tiktaalik remains anatomical evidence for a particular combination of fish and tetrapod-like features—a mobile neck, robust paired appendages and fin rays among them. Zachełmie, by contrast, would push the origin of limbs with digits deeper in time and turn the familiar elpistostegalians into sampled relatives that overlapped with an older, unseen history. Evolutionary trees branch; the youngest transitional-looking fossil never had to be the last stop before all tetrapods.[1][6]

The fish alternative raises the standard

Trace fossils are vulnerable to convincing impostors because unrelated actions can disturb mud in similar ways. Spencer Lucas's 2015 review applied three tests to proposed Devonian tetrapod tracks: foot morphology compatible with known anatomy, a smaller manus than pes, and an alternating pattern expected from quadrupedal walking. He rejected Zachełmie under those criteria and proposed that at least some structures could be fish feeding or nesting traces, including forms assigned to Piscichnus.[2]

Experiments make the general caution harder to dismiss. Living lungfish can move over exposed mud by planting the head and pivoting the trunk, leaving alternating jaw impressions while most of the body and fins leave little trace. Peter Falkingham and Angela Horner showed that such locomotion leaves sequences of impressions with features capable of inviting comparison to ancient trackways. Their result did not demonstrate that a lungfish made the Zachełmie marks; it demonstrated that “fish” and “ordered trail” are not mutually exclusive categories.[3]

Ahlberg's 2018 review answered that challenge at the level where it belongs: morphology. He argued that experimental lungfish trails retain body contact or lack the regular print form and digit impressions present in the Polish material. Against Lucas's mixed set of feeding marks and other fish traces, he emphasized Zachełmie's deep, repeated digit impressions and recognizable trackway organization.[6] The disagreement is not between evidence and denial. It is between competing accounts of how one surface acquired several recurring shapes.

That dispute suggests a fair burden of proof. A fish alternative must reproduce the whole association—not merely one oval depression—including the order, spacing, paired pattern, sediment displacement and best digit-like marks. The tetrapod interpretation must show that these features are genuinely anatomical rather than a selection of the clearest shapes from a variable trace assemblage. Neither side gets to win from the age alone: being surprisingly old is not evidence against a fossil, but surprise is a reason to demand unusually explicit comparisons.

The landscape changed beneath the footprints

The environmental story has also been revised. The original report placed the tracks on a marine tidal flat.[1] A detailed 2015 sedimentological study instead reconstructed very shallow, restricted dolomitic lagoons surrounded by low, sparsely vegetated islands and spits.[4] In 2018, a further analysis of the beds, microfossils and geochemistry argued for short-lived, largely non-marine lakes on a coastal plain, interrupted by limited marine influence. That study described stressed, low-diversity waters dominated by microbial mats and charophytes, with signs that the basins periodically dried into soils.[5]

These models share a physical core—flat terrain, shallow muddy water, microbial surfaces and repeated drying—but disagree about how directly the basin communicated with the sea. The distinction matters because the original tidal-flat reading encouraged a shoreline origin story for tetrapods. The ephemeral-lake model instead led its authors to infer that large predators could not have fed continuously in the impoverished pools and may have travelled over land to reach other waters or feeding grounds.[5][6]

That is an ecological inference, not a footprint preserved on dry ground. The 2018 study interpreted the Zachełmie tracks themselves as subaqueous. The argument for terrestrial competence joins the track anatomy to a reconstruction of the surrounding landscape: digit-bearing limbs, a lake that could shrink, scarce local prey and other habitats beyond it.[5] Change the environmental model and the behavioral envelope changes with it, even while the depressions stay exactly where they are.

What survives the dispute

Zachełmie now supports a ladder of claims with very different confidence. The early Eifelian age of the track-bearing interval is strongly constrained. The surfaces unquestionably preserve repeated biological or sedimentary disturbances. Their identification as tetrapod footprints has a detailed positive case and a published alternative. The maker's exact anatomy, lineage and lifetime habitat remain unknown. A responsible caption should not flatten that ladder into “the moment vertebrates walked onto land.”

The tetrapod reading nevertheless remains active in current research. A 2025 Nature study on far younger amniote tracks treated the Eifelian Zachełmie record as the earliest appearance of limbed stem tetrapods in the ichnological record, while using the mismatch between tracks and bones to show how severely early vertebrate diversity is undersampled.[7] That is evidence of continuing scientific acceptance, not a vote that erases Lucas's objection.

The deepest lesson is methodological. A track can outrun the body-fossil record because feet repeatedly touch environments in which skeletons rarely become fossils. It can also mislead because mud stores motion as ambiguous geometry. Zachełmie deserves attention for both reasons. Its slabs may preserve a few seconds from a limbed vertebrate lineage otherwise missing for millions of years. The claim is strongest when the absent animal is allowed to remain absent—constrained by marks, challenged by alternatives, and never reconstructed beyond what the surface can bear.

cronfeed.work