Editor’s note: We are delighted to present a series of posts by paleontologist Günter Bechly on the Ediacaran organism Kimberella. If identified as an animal, it would “predate the Cambrian explosion of bilaterian animal phyla as a kind of ‘advance guard.’” The question is of interest for debates about evolution and arguments about intelligent design raised by Stephen Meyer, among others. Find the full series about Kimberella here.

The association of Kimberella and its traces was assumed but not proven until the discovery of a lot of fossils that show these traces together with the putative trace-maker Kimberella caught in the act. The body fossils are generally positioned at the focal points of the fan-shaped scratch marks (Ivantsov 2013, Gehling et al. 2014), where “one end of such a trail is in contact with the head side of the soft body imprint, as if continuing or ‘flowing’ from the body” (Ivantsov 2010b, 2012). The first such specimen was an illegally collected fossil from the White Sea area revealed by Mikhail Fedonkin at the Museum of Natural History in Milan (Italy). It was later returned to Russia by the museum’s director Giorgio Teruzzi (Fedonkin 2001, Fedonkin & Vickers-Rich 2007, Ivantsov 2010b, 2012, 2013), who became the name patron for the new trace fossil species described by Ivantsov (2013). Numerous images of an association of Kimberella fossils with Kimberichnus traces have been provided in the published literature, i.e., by Gehling (1996: fig. 5.13), Seilacher (1999: fig. 5), Fedonkin (2003: fig. 16), Fedonkin et al. (2007b: fig. 20), Trusler et al. (2007: figs 3-4), Peterson et al. (2008: fig. 2a), Seilacher & Hagadorn (2010: fig. 1), Ivantsov (2009, 2010b: pl. 1 fig. 6, 2012: fig. 12 2013), Gehling et al. (2014: figs 5-7), Scheltema (2014: fig. 1B), Vinther (2015: fig. 3C,D,F), Parkhaev (2017: fig. 2b), and Giribet & Edgecombe (2020: fig. 46.6B). All these specimens leave hardly any doubt that Kimberella indeed was the maker of the Kimberichnus traces.

Several authors (Buatois et al. 2014, Gehling et al. 2014, Tarhan et al. 2015) have convincingly argued that the taxonomic association of the traces, as well as their association with sand pellets (see below), also excludes the fringe hypothesis by Retallack (2013), who suggested that these Ediacaran sediments are of terrestrial origin and that the fan-shaped arrays are “too straight and sharp to be molluscan radular scratches” and rather represent “casts of needle ice” crystals. Nevertheless, this example shows how vastly the views of specialists can differ, even though they are all working with the same empirical data. Lay people are often unaware how much of paleobiology indeed is relatively uncertain interpretation of poor evidence or sometimes even mere speculation. In the best case it is an inference to the best explanation, using circumstantial evidence to reconstruct an unknown past. And we all know how fallible such a procedure can be.

Pellets

Clusters of small round tubercles or pits (1.5-5 mm in diameter, occasionally even up to 10 mm) on or near the scratch marks have been interpreted as either fecal pellets or as globular sand casts (either or prey or of sediment) from the feeding activity (Fedonkin & Waggoner 1997, Fedonkin et al. 2007b, Ivantsov 2010b, 2012,2013, Gehling et al. 2014, Vinther 2015, Coutts et al. 2016, Coutts 2019), because they never occur beyond the Kimberichnus fans. However, Budd & Jensen (2017) raised questions concerning the origin of these pellets and asked if they might even be only casts of pyrite crystals. In my view this is highly unlikely, as it would not explain the strict association with the feeding traces. Nevertheless, again we find the experts to wildly disagree. Imagine this level of disagreement among expert witnesses in a courtroom. What do you think a judge or jury would have to conclude? But we are not done yet.

Reconstructing the Feeding Mode

As remarked by Ivantsov (2017), “the study of the morphology of body and trace fossils makes it possible to reconstruct the feeding behavior of this animal in details.” Here we go:

Gehling (1996), Gehling et al. (1995, 2005), Seilacher et al. (2005), and Seilacher & Hagadorn (2010) claimed that the traces are clearly scratches rasped by a bilateral symmetric structure of two rows of teeth as in modern radulas, and that “Kimberella did not move-on while it was browsing, but used a long proboscis that swung out more widely the further it extended away from the stationary body.” Seilacher & Hagadorn (2010: fig. 2) even proposed a detailed hypothetical reconstruction of the feeding apparatus and grazing process.

Seilacher & Hagadorn (2010) also remarked that “This stationary feeding is quite different from the continuous grazing modes of modern gastropods and chitons and even those of Paleozoic mollusks.” Likewise, Droser & Gehling (2015) cautioned that “the suggestions of molluscan affinities for Kimberella … should not overlook some important differences from extant gastropod grazing behaviors.”

Seilacher & Hagadorn (2010) suggested that the stationary feeding strategy of Kimberella disappeared with the Cambrian explosion and its onset of active benthic macropredators, who would have easily bitten off the long proboscis. This hypothesis is of course quite questionable, as none of the experts reconstructed Kimberella with such a long elephant-trunk-like proboscis, which is also not seen in any of the numerous fossils.

Like a Mechanical Shovel Excavator?

Mainly based on the 16 specimens at the South Australian Museum in Adelaide, Gehling et al. (2014) provided the most elaborate study of the Kimberella feeding traces and reconstructed the feeding style as similar to the scratching of a mechanical shovel excavator: 

… each pair of grooves is interpreted as corresponding to a single excavation, produced by a rasping action on the substrate. Due to a pendulum-like movement of the presumed proboscis, the scratch marks in each arcuate set are radially oriented and ascribed to a single focal point on the concave side of each set. … The rasping motion appears to have been directed back toward the focal point … they fan in a manner that suggests repeated probing by a single organ with a pair of ‘‘teeth’’ … The trace fossil arrays of Kimberichnus teruzzii are interpreted as rasping excavations by benthic organisms feeding on biomats … The K. teruzzii scratch fans appear to have been executed by an extensible organ capable of three components of movement. Each pair of scratches involved a radially directed scratch stroke, like that in living gastropods. Arcuate sets of scratches were made by stepwise lateral shifts between strokes. The arcuate swings began with sets close to the anterior end of the animal. At the end of each set the head end extended to make the next set, with a wider arc. The fan was completed when the proboscis reached its limit of distal extension. Having grazed a pie-shaped area of the biomat, the overlapping of scratch fans shows a variety of behaviors. From specimens with evidence of Kimberella in place (Figs. 5–7), the animal appears to have first advanced over its previous excavations to a new stationary locus and then retreated, or rotated in place, before grazing a new sector of mat. … Kimberichnus teruzzii is interpreted as the grazing trace of Kimberella quadrata armed with a proboscis bearing two ‘‘teeth’’. … Kimberichnus teruzzii and Kimberella quadrata represent the first record of systematic grazing by an epifaunal metazoan of bilaterian grade in the geological record.

The British mollusk specialist Jakob Vinther suggested that the scratch marks were generated by a more complex feeding apparatus similar to a molluscan radula (Vinther 2015). He said that 

specimens are found in great abundance on surfaces that also preserve distinct scraping traces named Kimberichnus Ivantsov, 2013 (Fig. 3B), which is also well known from Australia (Gehling et al. 2014). These traces suggest a mode of feeding on the microbial mats involving two larger teeth (forming paired grooves), and a series of smaller denticles (as evidenced by the presence of relief outside the tooth marks indicating feeding over a larger area by a more elaborate feeding apparatus than the visible paired teeth marks; Fig. 3B).

Among all Ediacaran organisms “only Kimberella seems to have actively pushed sediments during self-powered movement” (Gehling et al. 2005, Fedonkin et al. 2007b, Xiao & Laflamme 2009). The type and pattern of preservation of the feeding trace fossils suggests that “Kimberella could possibly scratch the mat through its entire thickness or even tear off pieces from it” (Sperling & Vinther 2010, Ivantsov 2012, 2013). Ivantsov said that “by stretching and contracting its body, Kimberella was able to pull and push its head part bearing teeth in order to scratch the substrate covered by microbial mats on which it, probably, fed” (Ivantsov 2017), and that “the animal is believed to have repeatedly passed over the microbial mat during food gathering” (Ivantsov et al. 2019).

Next, “Kimberella — Locomotory Tracks.”