A virtual symposium held on five consecutive Fridays beginning March 5, 2021.

REGISTRATION required for Zoom entry. Registrants will receive the Zoom link and passcode via email. See links this page to register and for more information.

Session III (Moderator: Ben Nicholas)

1 pm Alexis Mychajliw

1:30 pm Daniel Park

2 pm Alex White

2:30 pm Panel discussion: Alexis Mychajliw, Daniel Park, Alex White

Alexis Mychajliw
Research Associate, La Brea Tar Pits and Museum
Assistant Professor, Middlebury College

Talk title: Conflicts in context: natural history collections as archives of human-carnivore interactions through time

Abstract
Human-wildlife conflict in both rural and urban areas is a persistent threat to the continued existence of many predators. Historically, mammalian carnivores have faced severe range contractions, but, in some places, they are naturally re-expanding or being reintroduced, accompanied by a range of political dilemmas and public misconceptions. In combination with traditional conservation studies of extant populations, I provide historic and pre-European baselines to anticipate these dilemmas and distinguish between novel behaviors versus the return of normal, pre-contraction variation. Museum collections (natural history, archaeological, and paleontological can provide unexpected spaces for policy-relevant dialogues among stakeholders and be at the center of interdisciplinary working groups including ecologists, paleontologists, historians, practitioners and social scientists. Employing a diverse range of geochemical, morphological and archival techniques in a conservation paleobiology framework, I have used museum collections to reframe public discourse on the reintroduction of grizzly bears to California, discover the first successful case of North American “Pleistocene rewilding” for carnivores from the 1930s, and evaluate how culturally important Japanese red foxes have responded to urbanization with attendant disease-transmission consequences. Given the cultural biases present in historical narratives of carnivore conflict, museum specimens provide an irreplaceable mechanism for separating perception from reality to guide real-world policy decisions, and museum exhibits themselves can be vehicles for communicating this new understanding.

Daniel Park
Assistant Professor, Department of Biological Sciences, Purdue University

Talk title: Herbarium collections reveal wide variation in plant phenological responses to climate

Abstract
Changes in phenology–the timing of life history events–are among the most dramatic biological responses to climate change. Herbarium specimens represent snapshots of phenology (i.e. flowering and fruiting) at a specific place and time, and have tremendous promise to increase the spatial, temporal and taxonomic resolution of phenological data. However, difficulties in extracting useful information from specimens efficiently have limited efforts to apply collections-based approaches to large-scale phenological research. Here we present two contrasting approaches for this purpose; crowdsourcing and machine learning; and discuss the promises and opportunities of applying specimen-derived data to phenological research. Using these approaches, we examined tens of thousands of specimens and uncovered substantial and unexpected variation in phenological sensitivity across species ranges. We also observe patterns of temporal convergence among closely related species when they co-occur. These results suggest that phenological responses to climate change will be heterogeneous within communities and across regions, with large amounts of regional variability driven by local adaptation, phenotypic plasticity and differences in species assemblages. We thus demonstrate the utility of natural history collections in revealing large-scale patterns within assemblages and across continents that ultimately can improve forecasts of climatic change impacts on the structure and function of ecosystems.

Alex White
Machine Learning Postdoctoral Fellow, Data Science Lab, National Museum of Natural History, Smithsonian Institution

Talk title: Biogeography of fern shapes as revealed by deep learning

Abstract
With digitized herbarium specimens and associated metadata accumulating rapidly in open access repositories, we are now able to exploit data-hungry computer vision techniques in order to evaluate fundamental questions in plant evolution. High among the list of unknowns are the roles that ecological factors and morphological similarity play in mediating biogeographic patterns of taxonomic and phylogenetic diversity. Here, I integrate deep convolutional neural networks (CNNs) into a biogeographic study of morphological, taxonomic, and phylogenetic diversity in ferns and lycophytes. I show how CNNs and digitized specimens can be used to extract quantitative estimates of morphospace occupation, and I use these techniques to evaluate diversity-disparity relationships within ferns across latitudes. I also discuss how CNNs can be used to overcome logistical obstacles arising from modern specimen based workflows involving millions of images.

Read more, including about the speakers and their talks, on the ECSS website: https://sites.lsa.umich.edu/ecss/

REGISTER: https://sites.lsa.umich.edu/ecss/home/register/