Discriminant analysis loadings of variables for log10-transformed and mean removed data comparing Limiavipedidae to tracks of Cretaceous theropods and tracks of similarly-sized Mesozoic, Cenozoic, and extant avians.

Variable loadings for discriminant analysis comparing log10-transformed and mean removed linear data (footprint length, FL; footprint width, FW; digit II length, DLII; digit IV length, DLIV; pace length, PL; stride length, SL) and mean removed angular data (total divarication, DIVTOT; pace angulation, PA) of Limiavipedidae (Wupus agilis, Limiavipes curriei) to small- (Irenichnites isp.) medium- (Columbosauripus isp., Magnoavipes isp.) and large-sized (Irenesauripus isp.) theropod ichnotaxa, and Mesozoic (Archaeornithipes isp., Sarjeantopus isp.) and Cenozoic (Leptoptilostipus isp., Culcipeda isp., Gruipeda isp., Fuscinapeda isp., Pavoformipes isp.) avian ichnotaxa, and data collected from the tracks of extant large avians (Ardea herodias, juvenile Branta canadensis). Fig 7 shows the axes (axes 1 and 2) along which maximal separation of the grouped data occur. Axis 1 is interpreted as the effect of size on the variation present in the dataset; however, as size and pace angulation increase (as the size of the trackmaker increases and as the trackway narrows), total divarication decreases. This is consistent with the observations of theropods having a smaller total divarication, as well as a larger size and narrower trackway. Theropods group positively along Axis 1, while birds, with their smaller size and higher total divarication, and more “toed-in” footprints, group negatively along Axis 1. Axis 2 is interpreted as the relationship between FW and the lengths of the lateral digits to FL, PL, and SL; footprints with longer lateral digits (DII, DIV) are shorter in length, and are found in trackways with shorter PL and SL. Limiavipedidae group with Cretaceous avian traces and Cenozoic avian traces (Fig 7). Avian prints are interpreted to have subequal lateral digits and a higher L/W ratio, and the discriminant analysis correlates with the interpretation that avian prints belong to trackways with a relatively shorter pace length (Table 4).
Discriminant analysis loadings of variables for log10-transformed and mean removed data comparing Limiavipedidae to tracks of Cretaceous theropods and tracks of similarly-sized Mesozoic, Cenozoic, and extant avians.