| Abstract Detail
Hybrids and Hybridization Jolman, Devani [1], Wallace, Lisa [2]. Usability of Biodiversity Databases to Answer Questions About Plant Hybridization. In recent years, natural plant hybridization has increasingly been seen less as evolutionary noise and more as an important evolutionary mechanism that promotes speciation and diversity. With increasing effects of ecological disturbance and climate change, it is essential for plant scientists to understand the potential impacts of hybridization on life at all spatial scales. Mapping the distribution of hybrid plants could provide a means of studying the impacts of hybridization in a multitude of ways, including hybrid zone movement, localities of hybridization events, conditions that facilitate hybridization, and areas of concern for plant conservation. Occurrence data available through floras, digitized herbaria, and other sources are a starting point for mapping hybrid distributions, but many issues, such as the difficulty of hybrid identification, bias in the field, and incompleteness of digitized herbarium records, can leave large holes in reconstructed distributions of hybrid taxa. For studying the geography of hybridization, occurrence data were compared to assess the usability of several commonly used biodiversity databases: GBIF, iDigBio, SERNEC, vPlants, and BONAP. Occurrence data by county were extracted from each of the databases for Georgia, North Carolina, South Carolina, and Virginia for three hybrids of the Fabaceae, chosen for their distinct nativity and commonality: Lespedeza x nuttallii, Baptisia x serenae, and Wisteria x formosa. Presence was marked if the database either had a county occurrence point (BONAP) or a representative herbarium sheet (GBIF, iDigBio, SERNEC, vPlants). For each of the three hybrids, BONAP has the highest number of occurrence points, iDigBio second, and GBIF has the lowest. Cronbach’s Alpha consistency tests were used to test for consistency of occurrence data between the databases. Tests show that the databases have varying levels of consistency; consistency is high between the herbarium-based databases, while BONAP has little to no consistency with the herbarium-based databases both individually and combined. GBIF, while showing acceptable consistency with the other herbarium-based databases, lacks many essential herbarium records, resulting in notably lower numbers of occurrence points for all three hybrid species. BONAP, while reporting the highest number of occurrence points, is based on a vast array of unique data references, including published floras and literature, checklists, herbarium records, and personal observations. With significant inconsistencies between county occurrences across all three hybrids, BONAP is notably dissimilar from the herbarium-based databases. Overall, the results suggest that there are discrepancies between databases that warrant investigation before use in hybrid distribution mapping. Considerations should include how databases handle various names for a single hybrid, what sources of data are being used, and the number of herbaria from which the vouchers are included. Databases that include greater numbers of herbarium sheets may be better suited for hybrid studies due to the ability to check vouchered specimens for location and accurate identification. Our study also suggests that continued efforts in herbarium digitization and the development of public databases will benefit research on natural hybridization greatly by allowing specimen data to be easily acquired, compared, and used in analyses.
1 - 7742 Enfield Ave., Apt. 201, Norfolk, VA, 23505, United States
2 - Old Dominion University, Biological Sciences, Mills Godwin Building Rm. 110, Norfolk, VA, 23529, United States
Keywords:
hybridization
Biodiversity Informatics
Herbarium Digitization
Distribution Mapping
Biodiversity Database.
Presentation Type: Poster
Number: PHH002
Abstract ID:224
Candidate for Awards:None |
