The Proceedings of the Eighth International Conference on Creationism (2018)

the ceratopsids and the non-ceratopsid neoceratopsians. Classical MDS results show a similar division, including placement of Psittacosaurus and Chaoyangsaurus with the outgroup taxa (Fig. 89). In agreement with the other methods, PCA suggests separation between Ceratopsidae and other ceratopsians and between neoceratopsians and the outgroup + Psittacosaurus (Fig. 90). “Basal” Ceratopsia for Han et al.’s (2017) data matrix reveals separation between Ceratopsia and all outgroups, including pachycephalosaurs, for PC 1 (Fig. 91). Both PC 2 and PC 3 separate basal Ceratopsia ( e.g., Psittacosaurus and Yinlong ) and Neoceratopsia (Fig. 92). In PC 2, The ceratopsian taxa are also aligned stratigraphically with the Upper Jurassic “basalmost” ceratopsian, Yinlong , on the bottom and the more “derived” forms ( Protoceratops and Leptoceratops ) found in Upper Cretaceous rocks at the top. However, Leptoceratops , a Maastrichtian form, is found below Protoceratops , a Campanian form, in the series. I. Ceratopsidae The BDC results for Dodson et al . ’s (2004) data matrix, in Weishampel et al . (2004), show three blocks of positive correlation: 1) non-ceratopsid outgroup, 2) Chasmosaurinae, and 3) Centrosaurinae (Fig. 93). Almost all of the ceratopsid taxa (Chasmosaurinae and Centrosaurinae together) share negative correlation with the non-ceratopsid outgroup taxa. There is only one instance of shared positive correlation between the two ceratopsid blocks ( Triceratops and Avaceratops ) and no instances of shared negative correlation. Classical MDS results show separation between Ceratopsidae and the outgroup taxa, as well as separation between the two ceratopsid subfamilies (Fig. 94). PCA results likewise show separation between the three groups, however it is also revealed that Zuniceratops is separated from Protoceratops and the hypothetical outgroup by PC 2 (Fig. 95). PCA for Fry (2015) separates the Chasmosaurinae from the non-chasmosaurine taxa (centrosaurines and non-ceratopsids), with a suggestion of two parallel morphoseries. Triceratopsini forms a series within Chasmosaurinae with Arrhinoceratops and Bravoceratops clustering more closely with Triceratopsini members (Fig. 96). Aparallel series is revealed in non-triceratopsin chasmosaurines. PC 3 groups taxa from all groups but reveals a separation for the three included centrosaurine species (Fig. 97). DISCUSSION These results answer a number of interesting questions regarding not only dinosaurs and the relationships of birds to dinosaurs, but also for methodologies aiding baraminology. Close analysis of the PCA results further suggest details for Flood-related burial of dinosaur assemblages. We have inferred continuity between dinosaur taxa based on shared positive correlation (BDC) and close clustering/trajectories (MDS and PCA), whereas discontinuity was inferred for shared negative correlation (BDC) and distant spacing in MDS and PCA. Because of these patterns discussed in our Results section, we arrived at various conclusions on the baraminic status of various dinosaur taxa (Table 1, Appendix Table 2). PCAanalyses yielded unexpected insights. BDC analyses identified dinosaur holobaramins, often near the family level, as groups of taxa with significant similarity and dissimilarity. PCA revealed morphospatial relationships of dinosaurs across wider taxonomic ranges than BDC was designed for. Component plots provided additional details on the nature of holobaramins. For PCA, holobaramins are taxa confined to spatial clusters or, more often, linear series. Dinosaur holobaramins visualized in multivariate space could be defined as discontinuity-bounded morphospatial series of related taxa . That is, dinosaur holobaramins were frequently ordered in short series of taxa that were spatially-isolated from other series. The series of closely-spaced taxa in multivariate space represent trajectories through limited morphological space. Morphospatial series yielded additional potential insights. Using conventional radiometric ages as relative chronometers, series can sometimes be identified as stratomorphic. Stratomorphic morphospatial series are defined as sequential morphospatial series whose morphological sequence appearance matches its fossil record first appearance order (Wise 1995). Of the 42 lineages, seven had strong suggestive stratomorphic overprints (“Yes” and “Nearly” in Table 1). These results suggest that dinosaurs, like many other taxonomic groups examined through baraminology, frequently show discontinuity at or near the family level. Of 27 lineages examined in the most recent matrices which suggest holobaramins or near- holobaramins, six are at the family level, four are subfamilies, and fourteen are at suborder or between suborder or family (Table 1). A further inference drawn from these observations raised questions about how dinosaur baramins were distributed within the Mesozoic. For example, all four of the proposed subfamily- level holobaramins were found in the Upper Cretaceous. Four of the seven stratomorphic holobaramins were entirely confined to the Cretaceous, and the remaining three included Cretaceous taxa. The results also address a number of questions regarding dinosaur biosystematics and bird-dinosaur relationships. 1. Baraminology A. Dinosaur Groups The Dinosauria contain three large, morphologically-disparate groups. Baron’s (2018) Dinosauriamatrix provided themost diverse set of taxa for comparison. Analysis of the complete data with the ordination-based systematics employed here leads to a conclusion similar to Baron’s phylogenetic analysis. That is, the complete data matrix suggests an association of ornithischians and theropods within an Ornithoscelida group (Figs. 1 and 2). From the standpoint of a creation model approach to systematics, the existence of an Ornithoscelida cluster is a neutral question; biological creation includes hierarchies of higher-order associations. However, after removing taxa with the highest proportions of missing data ( i.e ., taxa with 50% or more missing data) a different ordination emerged: theropods, ornithischians, and sauropodomorphs formed three equally distant spatial associations (Figs. 3 and 4). This was consistent with ordinations of basal Saurischia (Fig. 8) that likewise divided sauropods and theropods. We propose that the ordination resulting in an Ornithoscelida cluster ( i.e ., Fig. 1) was generated, in part, as a result of missing data. Taxa with large amounts of missing data likely function as “noise” that obscures underlying group spatial relationships; removal of problematic taxa results in different spatial geometries. Alternatively, another possibility is Doran et al. ◀ Dinosaur baraminology ▶ 2018 ICC 440