GSA Annual Meeting in Seattle, Washington, USA - 2017

Paper No. 91-6
Presentation Time: 9:15 AM

ASSESSMENT OF DATA RESOURCES FOR THE CONSTRUCTION OF LATE ORDOVICIAN SHALLOW MARINE FOOD WEBS: DATA COMPILATIONS VERSUS MUSEUM COLLECTIONS


CASTRO, Ian O.1, KEMPF, Hannah L.1, TYLER, Carrie L.1, DINEEN, Ashley A.2 and ROOPNARINE, Peter D.2, (1)Department of Geology and Environmental Earth Science, Miami University, Oxford, OH 45056, (2)Department of Invertebrate Zoology and Geology, California Academy of Sciences, San Francisco, CA 94118, castroio@miamioh.edu

Our current understanding of the effects of invasive species is largely based on changes in biodiversity, which are unable to capture the number or types of interactions between organisms. Food webs provide an alternative approach, representing trophic interactions and energy flow within a community. However, food webs remain underutilized in paleoecology, in part due to the logistical challenges of assembling paleocommunity data across trophic levels. Data compilations, such as the Paleobiology Database (PBDB), may provide information appropriate for food web reconstructions, however, the reliability of such data compilations for food web reconstructions have not yet been directly assessed.

Here we compare Late Ordovician food webs reconstructed using the PBDB with specimen based reconstructions. Specimen based data included 9 museum collections (n=836) and 20 field localities (n=5,517). The PBDB data consists of 357 collections (n=2,814). To determine whether both data types consistently reproduce structural changes in the face of ecological reorganization, networks were examined before and after the Richmondian Invasion (RI; Cincinnati Arch, USA). The post-RI PBDB network had fewer guilds (28 pre-RI to 25 post-RI), links (81 to 70), and omnivores (50% to 48%), with comparable complexity (0.10 to 0.11), and higher modularity (0.23 to 0.28) when compared to PBDB pre-RI network. Similarly, the post-RI specimen-based network consisted of identical numbers of guilds (26), fewer links (75 to 72) and omnivores (54% to 46%), with identical complexity (0.11), and higher modularity (0.22 to 0.27) in comparison with the specimen based pre-RI network. Both data sources generate comparable numbers of guilds and links, and produce consistent structural changes, despite substantially higher numbers of species in both specimen-based networks relative to the PBDB networks.

Given that both data types produced similar numbers of guilds and links despite differences in richness, these results suggest that the PBDB could serve as an accessible and reliable data resource suitable for food web reconstructions. This has important implications for assessing the role of biodiversity in ecosystem functioning as well as determining the long-term consequences of ecosystem response to perturbation.