A combined approach to identify isolated theropod teeth from the Cenomanian Kem Kem Group of Morocco: cladistic, discriminant, and machine learning analyses

The Kem Kem Group of Southeastern Morocco, North Africa, is well known for theropod remains, especially isolated teeth. Here, a collection of isolated theropod teeth is assessed for diversity using a combination of linear discriminant, phylogenetic, and machine learning analyses for the first time. The results confirm earlier studies on Kem Kem theropod diversity, with teeth referred to Abelisauridae, Spinosaurinae, and Carcharodontosauridae. A single tooth is ascribed to a non-abelisauroid ceratosaur or a megaraptoran and may represent the enigmatic averostran <i>Deltadromeus</i>. Spinosaurine teeth are clearly differentiated by all three methodologies, whereas abelisaurid and carcharodontosaurid teeth could only be distinguished by the machine learning and phylogenetic analyses. This study shows that a combination of independent methods is most effective at providing strong evidence on theropod dental diversity in a particular assemblage, and that cladistic and machine learning analyses are the most reliable approaches to identify isolated dinosaur teeth. The methodology used here is likely to yield results in other dinosaur assemblages where isolated teeth are more abundant than body fossils.

北非摩洛哥东南部的Kem Kem群（Kem Kem Group）以产出兽脚类（theropod）恐龙遗骸而闻名，其中尤以孤立牙齿标本最为典型。本研究首次结合线性判别分析（linear discriminant analysis）、系统发育分析（phylogenetic analysis）与机器学习分析（machine learning analysis），对一批孤立兽脚类牙齿的物种多样性展开评估。研究结果验证了此前针对Kem Kem群兽脚类多样性的相关研究结论，所涉牙齿分别隶属于阿贝力龙科（Abelisauridae）、棘龙亚科（Spinosaurinae）与鲨齿龙科（Carcharodontosauridae）。另有一枚牙齿被归为非阿贝力龙形类角鼻龙类（non-abelisauroid ceratosaur）或巨盗龙类（megaraptoran），其可能代表神秘的鸟兽脚类（averostran）物种<em>Deltadromeus</em>。三种分析方法均可清晰区分棘龙亚科的牙齿标本，而阿贝力龙科与鲨齿龙科的牙齿仅能通过机器学习分析与系统发育分析加以区分。本研究证实，联合多种独立分析方法，可为特定化石组合中的兽脚类牙齿多样性提供强有力的证据；同时，分支系统学分析与机器学习分析是鉴定孤立恐龙牙齿最可靠的技术路径。本次研究所采用的分析方法，同样有望在其他以孤立牙齿标本相较于躯体化石更为丰富的恐龙化石组合中获得有效研究成果。