Morphology delimits more species than molecular genetic clusters of invasive Pilosella

• Published in: American journal of botany Vol. 102; no. 7; pp. 1145 - 1159
• Main Authors: Moffat, Chandra E., Ensing, David J., Gaskin, John F., De Clerck-Floate, Rosemarie A., Pither, Jason
• Format: Journal Article
• Language: English
• Published: United States Botanical Society of America, Inc 01.07.2015; Botanical Society of America
• Subjects: Adaptation; amplified fragment length polymorphism; Asteraceae; Asteraceae - anatomy & histology; Asteraceae - classification; Asteraceae - genetics; Biodiversity; Botany; British Columbia; Cluster Analysis; Ecological studies; Ecology; ecosystems; Flowers & plants; Flowers - anatomy & histology; Flowers - classification; Flowers - genetics; genetic cluster; Genetic Variation; Genetics; Genotype; Geography; Hieracium; Hybridization, Genetic; interspecific hybridization; Introduced Species; introgression; invasive; Invasive species; morphological molecular congruence; Morphology; morphometric; morphometry; Phenotype; Pilosella; plant morphology; randomForest; Reproducibility of Results; structure; Variance analysis; British Columbia; genetic cluster; morphological molecular congruence; morphometric; invasive; morphology; Asteraceae; randomForest; Hieracium; Pilosella; structure
• ISSN: 0002-9122; 1537-2197
• DOI: 10.3732/ajb.1400466

PREMISE OF THE STUDY: Accurate assessments of biodiversity are paramount for understanding ecosystem processes and adaptation to change. Invasive species often contribute substantially to local biodiversity; correctly identifying and distinguishing invaders is thus necessary to assess their potential impacts. We compared the reliability of morphology and molecular sequences to discriminate six putative species of invasive Pilosella hawkweeds (syn. Hieracium, Asteraceae), known for unreliable identifications and historical introgression. We asked (1) which morphological traits dependably discriminate putative species, (2) if genetic clusters supported morphological species, and (3) if novel hybridizations occur in the invaded range. METHODS: We assessed 33 morphometric characters for their discriminatory power using the randomForest classifier and, using AFLPs, evaluated genetic clustering with the program STRUCTURE and subsequently with an AMOVA. The strength of the association between morphological and genotypic dissimilarity was assessed with a Mantel test. KEY RESULTS: Morphometric analyses delimited six species while genetic analyses defined only four clusters. Specifically, we found (1) eight morphological traits could reliably distinguish species, (2) STRUCTURE suggested strong genetic differentiation but for only four putative species clusters, and (3) genetic data suggest both novel hybridizations and multiple introductions have occurred. CONCLUSIONS: (1) Traditional floristic techniques may resolve more species than molecular analyses in taxonomie groups subject to introgression. (2) Even within complexes of closely related species, relatively few but highly discerning morphological characters can reliably discriminate species. (3) By clarifying patterns of morphological and genotypic variation of invasive Pilosella, we lay foundations for further ecological study and mitigation.