This session explores how GIS, remote sensing, and spatial analysis are being used to monitor plant populations and inform conservation across California. Presentations will focus on data-driven approaches to vegetation mapping, species distribution, and ecological modeling.
SESSION CHAIRS Dr. Matthew Clark1, Dr. Brooke Rose2
1Department of Applied Human and Environmental Science and Center for Interdisciplinary Geospatial Analysis (CIGA), Sonoma State University, Rohnert Park, CA, United States. 2Department of Geography and The Center for Open Geographical Sciences, San Diego State University, San Diego, CA, United States
Dr. Matthew Clark
Sonoma State University
Dr. Matthew Clark is a professor in the Applied Human and Environmental Science department at Sonoma State University. He has a PhD in Geography from University of California, Santa Barbara and a MSc in Conservation and Ecosystem Analysis from University of Washington. He teaches classes in geographic information systems (GIS), remote sensing, and ecology. His research is focused on using novel forms of remote sensing, including satellites, airplane sensors and drones, for monitoring biodiversity, assessing land change, and helping conservation and land management.
Dr. Brooke Rose
San Diego State University
Brooke Rose is a Research Scientist and Adjunct Professor in the Department of Geography and Center for Open Geographical Sciences at San Diego State University. Her work integrates ecological modeling and remote sensing to better understand how California’s plants are responding to global change. She holds a PhD in Plant Biology from the University of California, Riverside and has worked on a variety of projects, including modeling the future distributions of culturally important plant species in partnership with Tribal groups in southern California and mapping oak woodland resilience with The Nature Conservancy.
22.1 From Seeds to Systems: Harnessing Geography for Biodiversity Action
Sunny Fleming
Esri, Asheville, NC and Redlands, CA, United States
Description Numerous entities have a role to play in protecting biodiversity and preventing species extinction. Botanical gardens and related institutions safeguard seed sources and plant material critical for restoration efforts. Land managers prioritize, implement, and monitor conservation initiatives, while research scientists provide insights that shape our understanding and methods.
Today, the urgency to accelerate our work has never been greater. As we face unprecedented ecological challenges, our community must strengthen interdisciplinary collaboration and amplify storytelling to inspire action. Geographic science and spatial tools are central to this effort. They empower us to integrate diverse data, visualize complex patterns, and make informed decisions that drive conservation forward.
In this session, we will explore how GIS supports our unified goal of stewarding the planet; helping us plan and prioritize conservation actions, track restoration progress, monitor species and ecosystems, and engage both existing and new stakeholders. Together, we can harness the power of geography to transform insights into impact.
Presenter Bios
Sunny Fleming
Esri
Sunny Fleming is Esri’s industry director for the domains of environment, conservation, and natural resources. Throughout her career, she has applied GIS concepts and technology to environmental policy, conservation, and natural resources, from monitoring species in the field to helping state parks manage assets and assess their economic impacts. She continues to pursue her passion for the environment by helping industry leaders streamline and improve their work with GIS technology, whether in the field or in the office, and whether setting policy or managing wildlife and resources.
22.2 Never a Straight Line: The Iterative Process behind Creating California's Important Plant Areas and the Role of Automated GIS
Dariya Draganova
GreenInfo Network, Oakland, CA, United States
Description With California’s 30x30 initiative in full swing and the dire need to conserve and protect more of California’s native plant habitats, California Native Plant Society (CNPS) is in the throes of utilizing GIS to find and delineate Important Plant Areas in California. The process of delineating Important Plant Areas in California has proven to be anything but a straight line and the role of automated GIS has been crucial in this iterative process. Instead of approaching this manually, we have created a script system – an automated process that takes us from A to Z. As it goes, not everything can be foreseen and planned for when dealing with a complex GIS process and multiple sets of data. Come hear about some of the pitfalls and obstacles we faced, about methodology changes, about inevitable human error, and how we handled these unforeseen events in an automated system. This gargantuan effort – lead by CNPS and masterminded by a team of expert scientists, is in partnership with GreenInfo Network, a non-profit GIS consultancy.
Presenter Bios
Dariya Draganova
GreenInfo Network
Dariya Draganova is a GIS Analyst and Developer at GreenInfo Network, a non-profit GIS consultancy based in Oakland, California. Dariya specilizes in creating systems of GIS processes to solve problems, analyze data, and create geospatial products. She has been working with the California Native Plant Society on their Important Plant Areas project for the last year and a half.
22.3 Targeted Surveys Informed by Species Distribution Modeling Reveal Unrecorded Populations of Cirsium fontinale var. obispoense, an Endangered Serpentine Seep Endemic
Mike Butler1, Dr. Andrew Fricker1, Dr. Nishi Rajakaruna1,2
1California Polytechnic State University, San Luis Obispo, CA, United States. 2North-West University, Potchefstroom, South Africa
Description Chorro Creek bog thistle (Cirsium fontinale var. obispoense, Asteraceae) is a listed endangered species narrowly distributed within only San Luis Obispo (SLO) County, California. This biennial to short-lived perennial is both a serpentine endemic and restricted to areas with seeps and springs. This narrow abiotic niche makes potentially suitable habitat uncommon, but in theory, predictable. The U.S. Fish and Wildlife Service stated in their 2022 5-Year Review that additional populations were likely to exist, and that surveys ought to be done to find them.
Following the agency’s recommendations, we developed a Species Distribution Model (SDM) for the purpose of guiding our targeted surveys. We used Maxent to correlate Chorro Creek bog thistle occurrence data with environmental predictors such as serpentine extent, distance to mapped spring, elevation, topographic position, precipitation, temperature seasonality, and climatic water deficit. The resulting SDM (AUC=0.99, CBI=0.88) was a map of potentially suitable habitat across SLO County.
Our targeted surveys took place from May-November of 2025 and focused on the overlap between accessible land and potentially suitable habitat. These surveys resulted in the discovery of two unrecorded populations, totaling approximately 1600 individuals. Our SDM proved useful for targeting surveys, but only to the extent that the environmental predictors used were accurate, a well-documented limitation of SDM for edaphic endemics. These efforts aid the conservation of Chorro Creek bog thistle as climate change threatens it. Serving as an umbrella species, the conservation of this taxon and its serpentine seep habitat benefits both wildlife and other rare plant taxa.
Presenter Bios
Mike Butler
California Polytechnic State University, San Luis Obispo
Mike Butler is a Biological Sciences master's student at Cal Poly San Luis Obispo, where he is a member of the Geoecology Lab. His research focuses on geospatial approaches to rare plant conservation, combining remote sensing with boots on the ground. He has previously worked as a restoration ecologist for Chambers Group Inc., and as a botany biotechnician for the National Park Service. He holds an undergraduate degree in Environmental Science and Policy from Cal State Long Beach.
22.4 Opportunities and Challenges for Mapping Fine-Scale Vegetation Change Across California with Deep Learning
Dr. Lauren E Gillespie1,2, Dr. Sherrie Wang1, Dr. Sara Beery1
1Massachusetts Institute of Technology, Cambridge, MA, United States. 2University of Michigan, Ann Arbor, MI, United States
Description From longer droughts to more intense fire regimes, anthropogenic climate and land use change threatens biodiversity globally. Systematically measuring vegetation change across these twin drivers at the community level is challenging, especially at fine spatial and temporal scales. Deep learning models have shown recent promise at detecting community-level vegetation change from remote sensing imagery. However, opportunistic iNaturalist citizen science data used to train these models contains many systemic biases and limits their reliability in undisturbed regions and hard-to-access wilderness where vegetation communities are often at elevated risk to rapid climate warming. As one of the fastest-warming states in the continental U.S. but also one of the most comprehensively sampled through vegetation surveying efforts driven by the California Native Plant Society (CNPS), California serves as the ideal testbed for improving these deep learning-based vegetation monitoring approaches for tracking anthropogenic environmental change.
In this talk, I will overview our ongoing efforts to improve these change predictions for California's native vegetation in collaboration with CNPS and California's Department of Fish and Wildlife (CDFW). First I will overview existing tools that use deep learning to map rapid vegetation change across California. Then I'll cover California-specific challenges introduced by the modeling process and training data. Lastly, I will overview our ongoing work to leverage CNPS' Rapid Assessment and Relevé survey data to improve these deep learning-based estimates of vegetation change using a variety of statistical tools.
Presenter Bios
Dr. Lauren E Gillespie
Massachusetts Institute of Technology
Coming from a background in both computer science and biology, Dr. Gillespie's interdisciplinary research develops new AI-integrated approaches for monitoring ecosystems at scale. Her work develops foundation models, AI models that can rapidly make sense of large-scale but noisy data with little guidance, and aims to uncover the effects of rapid environmental change on species to improve our ecological forecasting of the natural world. By leveraging diverse and widely available data from sources including remote sensing and citizen + community science, her research aims to create models of biodiversity that are accurate and useful for conservation decision-makers around the world.
Dr. Alyssa R. Phillips1, Erin Carroll1, Dr. Meixi Lin1, Dr. Roxanne Cruz-de Hoyos1, Dr. Benjamin W Blonder1, Dr. Moises Exposito-Alonso1,2
1University of California, Berkeley, Berkeley, CA, United States. 1Howard Hughes Medical Institute, Berkeley, CA, United States
Description Quaking aspen (Populus tremuloides) is the most widespread tree in North America and known for its colorful autumn foliage. Aspen stands provide critical and unique ecosystem services such as having a high water storage capacity and serving as a fire break. In California, aspen is important for meadow, riparian, and beaver habitat restoration. Unfortunately, aspen is declining range-wide due to drought stress, increases in herbivore pressure, disease, and fire exclusion. Further, aspen stands in California are smaller and have low connectivity compared to other parts of the species range, increasing their vulnerability to inbreeding. Thus, we hypothesized California aspen stands have lower genetic diversity and greater population structure than highly contiguous populations. Comprehensive genome sequencing of California aspen populations is expensive and time consuming, limiting the ability to monitor genetic diversity. In this project, we aimed to (1) investigate genetic diversity of California aspen populations and (2) evaluate whether remote sensing data can be applied to monitor genetic diversity. We leveraged sequencing data for 2,000 aspen (124 in California), a novel reference genome, and NEON Airborne Observation Platform data. Our preliminary population genomic analyses suggest some California aspen populations have genetic signatures of inbreeding. In assessing whether remote sensing data can be applied to monitor genetic diversity, we found tree genotype and ploidy level had a significant effect on leaf spectral reflectance. Additionally, spectral reflectance had high heritability at multiple wavelengths. These results suggest remote sensing data can be used to map genetic diversity across aspen populations.
Presenter Bios
Dr. Alyssa R. Phillips
University of California, Berkeley
Dr. Alyssa Phillips is a plant evolutionary biologist interested in how polyploidy and genome size variation contribute to or inhibit adaptation. As a postdoc co-advised by Dr. Ben Blonder and Dr. Moises Exposito-Alonso, Alyssa is developing evolution-informed models of quaking aspen mortality for forest management. She completed a bachelor’s from Appalachian State University in Biology and Ph.D. in Plant Biology from UC Davis with Dr. Jeffrey Ross-Ibarra, where her Ph.D. research integrated population genetics and ecophysiology to study the origins and persistence of mixed-ploidy in Andropogon gerardi, an ecologically dominant prairie grass.
Thank you to our 2026 sponsors!
Giant Sequoia
Special thanks to Esri, our 3-time Giant Sequoia Sponsor, for making this movement stronger than ever!
The mission of the California Native Plant Society is to protect California’s native plants and their natural habitats, today and into the future, through science, education, stewardship, gardening, and advocacy.
DateFriday, February 6
