TRACKWAYS PRODUCED BY FISH: IMPLICATIONS OF SUBSTRATE-CONTACT BEHAVIOR BY FISH RECORDED IN TRACE FOSSILS FROM LATE PLEISTOCENE GLACIAL VARVES, NEW ENGLAND, USA

Trace fossils from late Pleistocene varved glacial lake beds of the Connecticut and Merrimack Valleys of New England, USA reveal substrate-contact behaviors of fish that lived from 15.6-12.7 (cal) kyr BP. The trace fossils are the earliest postglacial record of freshwater fish in New England, and they show that fish were present in the glacial lakes within a century of deglaciation. Placed within the New England Varve Chronology, the trace fossils provide a high-resolution timeline for post-glacial reinhabitation by fish from Atlantic coastal refugia.

Undichna trails and a trackway that can be referred to Broomichnium (Kuhn 1958) are present in the glacial varves. Broomichnium appears to represent repeated ventral impressions of a benthic fish, resembling those made by modern Cottus cognatus, the slimy sculpin. The trace fossil is a small, bilaterally symmetrical impression composed of two pairs of nested linear or curvilinear imprints diverging to the posterior, and occasionally a longer median imprint at the posterior. Broomichnium is always preserved on bedding planes as negative epireliefs or as positive hyporeliefs.

Laboratory testing supports a link between Broomichnium and freshwater sculpin. We observed slimy sculpin during feeding and after being induced to swim. At rest, the pectoral, pelvic and anal fins are in contact with the substrate. During fast swimming (as in predator avoidance), the large fan-like pectoral fins are “flapped” to help launch the fish off of the substrate. During slower substrate-contact behaviors (as in orienting to prey), pectoral fins are used to “hop” into position. Pelvic fins, which are in a thoracic position, “prop” up the fish while the posterior portion of the body remains in contact with the substrate.

Bedding plane trackways produced by fish can be added to the suite of trace fossils found in laminated glacial sediments and possibly aid in the interpretation of other subaqueous paleoenvironments. Occurrences of Broomichnium previously attributed to tetrapods must be re-evaluated within their paleoenvironmental contexts. In particular, Quadrispinichna parvia (Anderson 1974) (=Broomichnium permianum Kuhn) from the Permian of South Africa may have been produced by fish, which would reconcile a long-term controversy about the origin of these deep water trackways.