Four neochoristoderan vertebral centra are described from the latest Cretaceous of New Jersey. One specimen was recovered from the basal transgressive lag of the Navesink Formation in the area of Holmdel Park, New Jersey, and two others were recovered nearby and likely were derived from the same horizon. The fourth was recovered from the Marshalltown sequence in the vicinity of the Ellisdale Dinosaur Site. These vertebrae expand the geographic range of Late Cretaceous neochoristoderes in North America by over 2000 km further east, and represent the first neochoristoderan remains from the Atlantic coastal plain. To discern whether neochoristodere remains are to be expected in New Jersey, and elucidate why neochoristoderes are apparently so rare in Appalachia, we implemented ecological niche modeling to predict the range of suitable habitat for Champsosaurus, the only known genus of Late Cretaceous neochoristoderes. We found that in Appalachia, the ideal habitat of Champsosaurus likely existed slightly further north and west than the Atlantic coastal plain, and New Jersey is likely on or near the margin of this suitable habitat space. These results suggest that the occurrence of neochoristoderes in New Jersey is consistent with the habitat requirements of known Late Cretaceous neochoristoderes. These vertebrae therefore may represent the southern margin of a population of neochoristoderes that lived further inland, where latest Cretaceous sediments are not preserved. The continued recovery of material from Late Cretaceous deposits along the Atlantic coast, and review of existing collections, is encouraged to clarify the true distribution of neochoristoderes in Appalachia.

The occurrence datasets from the Palaeobiology Database (PBDB) were downloaded on the 18th of February 2020. These are: (1) occurrence datasets for Champsosaurus from the Campanian and Maastrichtian; and (2) fossil occurrences from the Campanian Maastrichtian (for the reconstruction of sedimentary outcrop). Please refer to the details listed at the top of each document for a breakdown of how the data were searched, and what variables they contain.