HMEI Awards $1.2 million to Support Biodiversity Research
High Meadows Environmental Institute (HMEI) has selected eight faculty-led research projects to receive funding through the Thomas A. and Currie C. Barron Family Biodiversity Research Challenge Fund, also known as the Biodiversity Challenge. One of several research cooperatives within HMEI’s Grand Challenges program, the Biodiversity Challenge supports research, teaching, and mentorship to enhance understanding of biodiversity; examine the effects of environmental change on biodiversity; and explore methods for supporting biodiversity, now and in the future.
The Grand Challenges program prioritizes the funding of cross-disciplinary collaborations that represent new directions in faculty work, create innovations in undergraduate teaching and mentorship, and pave the way for external funding. Princeton faculty from multiple academic units, including the departments of ecology and evolutionary biology, geosciences, civil and environmental engineering, history, American studies, computer sciences, and mechanical and aerospace engineering, were awarded funding as the outcome of the campus-wide call for proposal.
Like all Grand Challenges programs, the Biodiversity Challenge capitalizes on funded activities to enhance undergraduate teaching programs including the introduction of new courses and course modules that incorporate research advances and by providing immersive experiences for undergraduate students in the junior and senior years and through summer internships.
The newly funded projects explore the following biodiversity topics:
- Investigating whether “ghost genes” in red wolf-coyote hybrids could bolster genetic diversity in red wolves as a conservation strategy;
- Balancing the global benefit of solar farms with concerns for local biodiversity loss due to habitat disruption and using this knowledge to inform the placement of solar farms;
- Exploring the use of machine learning for biodiversity monitoring to accelerate biodiversity assessments and to guide timely conservation action.
- Using machine learning methods to map historical vegetation changes from archival aerial photographs to more accurately assess the extent of biodiversity loss during the 20th century and predict future changes.
- Reconstructing Triassic food webs from fossilized nitrogen isotopes to advance understanding of how food webs and ecosystems are structured and constrained.
- Examining genomic diversity in Hawaiian fruit flies to understand how biodiversity evolves in the face of rapid habitat loss; and to inform the conservation of Hawaiian Drosophilidae and insects more broadly.
- Investigating why savanna plants respond to environmental fluctuations in unpredictable ways; and by understanding this phenomenon, consider approaches to stem the loss of biodiversity in savanna ecosystems.
- Using South African kelp forests as a case study for creating multimedia narratives that raise awareness of ecological change in local settings and to aid in conservation efforts.
The following provides additional detail on the awarded projects:
Reviving “Ghost” Genomic Variation in the Endangered Red Wolf: A Precedent for Unconventional Species Conservation
PI: Bridgett vonHoldt (Professor of Ecology and Evolutionary Biology)
The red wolf (Canis rufus) is one of North America’s most critically endangered canids. Human persecution drove the species to the brink of extinction during the 20th century, and red wolves now exist only in captivity and small reintroduced wild populations. However, because this captive-bred population descends from only 14 ancestors, it is severely lacking in genetic diversity, which makes the wolves less able to adapt to environmental variation. But some wild canids may hold other extant genes. That’s because, as red wolf populations plummeted, some interbred with coyotes (Canis latrans), and the genetic “ghosts” of this interbreeding likely exist in coyotes today. To aid red wolf conservation efforts, vonHoldt will investigate whether additional genetic diversity exists in red wolf-coyote hybrids that roam the Gulf Coast. Ghost genes are unobservable in an animal’s outer appearance because they are suppressed by molecular regulatory mechanisms, but their presence can be revealed through genomic sequencing. To identify ghost genes and understand their molecular regulation, vonHoldt’s team will sequence the genomes of captive-bred red wolves, red wolf-coyote hybrids, and coyotes. By identifying and characterizing red wolf ghost genes, this project will establish a framework for monitoring red wolf genetic diversity and inform conservation strategies aimed at reviving lost ancestral variation in red wolves and other critically endangered species.
The Biodiversity Impacts of Solar Power: Balancing Local Harm With Global Benefits
PIs: Jonathan Levine (J.N. Allison Professor in Environmental Studies and Professor of Ecology and Evolutionary Biology) and Jesse Jenkins (Associate Professor of Mechanical and Aerospace Engineering and the Andlinger Center for Energy and the Environment)
Renewable energy initiatives present a positive force for biodiversity globally by helping to mitigate climate change, but at the local scale, solar and wind farms can cause habitat loss. It is unclear whether solar farms have a net positive impact on biodiversity, because prior studies have investigated global and local impacts on biodiversity using different metrics. To simultaneously evaluate the biodiversity impacts of building solar farms both locally and globally, Levine and Jenkins will develop a model that estimates the impacts of solar farm locations on local habitat loss and global reductions in climate-related stressors. Then, they will use this model’s findings to simulate solar farm deployment under a range of U.S. energy policy scenarios—from current domestic energy policies to alternatives that would transition the U.S. to net-zero greenhouse gas emissions by 2050—and evaluate the resulting biodiversity outcomes. Finally, the team will use their findings to create a free, interactive online dashboard to allow land managers, policymakers and stakeholders to explore the biodiversity consequences of deploying solar panels in different locations and landscapes. By balancing local and global impacts of solar farm placement, this project will help inform renewable energy policies so that they can have the best possible impact.
Bridging the Expert Validation Gap in Biodiversity Monitoring with Inference-Aware Vendi Information Gain
PIs: Adji Bousso Dieng (Assistant Professor of Computer Science), Lydia T. Liu (Assistant Professor of Computer Science)
Collaborators: Mary C. Stoddard (Professor of Ecology and Evolutionary Biology) and David S. Wilcove (Henry W. Putnam Professor of Ecology and Evolutionary Biology and HMEI; Professor of Public Affairs)
Monitoring how climate change and other environmental disruptions impact biodiversity is essential for conservation, but monitoring often generates enormous, unwieldy datasets that make timely conservation action difficult. Machine learning systems could make biodiversity monitoring easier and more efficient by automating many aspects, for example by identifying species from photos, videos, or audio recordings. However, ecologists currently lack a system for evaluating the quality of machine learning predictions’ ability to infer ecological patterns, such as changes in species distributions, population trends, and interactions between species. To fill this gap, Dieng and Liu will develop a framework for evaluating the ability of machine learning predictions to inform ecological inferences. They will use this framework to test machine learning predictions based on video monitoring footage of hummingbird-flower interactions at the Rocky Mountain Biological Laboratory and acoustic recordings of bird calls from threatened biodiversity hotspots. By providing a framework for ecologists and conservation managers to more easily monitor biodiversity, this project will help accelerate biodiversity assessments and guide timely conservation action.
Measuring the Anthropocene: Machine-Reading Historical Aerial Photographs to Assess Long-Term Biodiversity Loss During the Great Acceleration
PIs: Emmanuel Krieke (Norman B. Tomlinson ’16 and ’48 Professor of War and Society; Professor of History), Xiaoyan Li (Senior Lecturer of Computer Science), Tsering Wangyal Shawa (Geographic Information Systems and Map Librarian), and Bill Guthe (Senior Geographical Information Systems Visualization Analyst)
Starting in the 1940’s, biodiversity loss accelerated due to industrialization, population growth, and agricultural expansion. To help map these changes, Krieke, Li, Shawa, and Guthe will train machine learning systems to extract vegetation and land use information from archived aerial photographs. Most climate and biodiversity models rely on vegetation data derived from satellite imagery, but satellite images collected before the 2000’s were too low resolution for this purpose. Aerial photographs captured between the 1940’s and 1990’s could be leveraged to understand how land use and biodiversity changed over that period. However, current machine learning systems are unable to read aerial photographs, and extracting this data manually is tedious and time consuming. To enable data to be efficiently extracted from aerial photographs, the team will collect and process aerial photographs from several case study areas, including regions of Namibia, South Africa, Nigeria, Colombia, Cuba, Sri Lanka, and Vietnam. They will then use these case study photographs to train machine learning systems to identify and extract data from aerial photographs. By unlocking the potential of the tens of millions of aerial photographs stored in archives and depositories around the world, this project will empower researchers to more accurately assess the extent of biodiversity loss during the 20th century and predict future changes.
What Sets Ecosystem Trophic Structure?: Insights From the Application of New Measurements to Vertebrate Fossils
PIs: Christopher Griffin (Assistant Professor of Geosciences) and Daniel Sigman (Dusenbury Professor of Geological and Geophysical Sciences)
Modern marine ecosystems are structured differently from terrestrial ecosystems. While terrestrial food webs often have only three levels—plants, herbivores, and carnivores—marine food webs can have up to five levels, including carnivore-eating-carnivores. It’s unclear whether this is a modern phenomenon, or if ancient terrestrial food webs were similarly tri-layered. Griffin and Sigman will investigate the structure of terrestrial food webs during the Late Triassic Period (235–201 million years ago), an era known to have had an unusually large number of carnivore species relative to herbivores, similar to the carnivore-dominant food webs of modern marine ecosystems. By analyzing nitrogen isotopes in fossilized teeth, the team will determine how food webs were structured during a 30-million-year window that involved several large climatic shifts and dramatic changes in biodiversity. The study will focus on fossils collected from the Petrified Forest National Park in Arizona and several small, newly discovered fossil sites in Texas and will examine teeth from a range of species, including lungfish, sharks, invertebrates, amphibians, crocodilians, and dinosaurs. By untangling food web structure during this unique period, the project will provide insight into how food webs and ecosystems are structured and constrained.
Broad-scale Genome Sequencing Across Hawaiian Drosophilidae to Understand the Evolution of Complex Biodiversity in the Face of Rapid Habitat Loss and Extinction
PIs: Kim Bernard (Assistant Professor of Ecology and Evolutionary Biology) and Jacqueline Robinson (Associate Researcher of Ecology and Evolutionary Biology)
Insect species are declining globally due to habitat loss, insecticide use, invasive species, and climate change. Though insects are far more biodiverse than vertebrates, insect conservation lags far behind. In particular, little is known about the genomic diversity of insects, especially how it has been shaped by evolution and how it is being impacted by environmental threats such as habitat destruction and climate change. Bernard and Robinson will study the genomic diversity of the Hawaiian Drosophilidae, a group of more than 1,000 fly species that originated from a single colonization event in the archipelago around 25 million years ago. These flies are enormously diverse in appearance, behavior, and habitat use, but this project is the first attempt to map their genomic diversity. By collecting and sequencing 1,600 flies from around 100 species from the major Hawaiian Islands, the team will shed light on the evolutionary drivers that caused Hawaiian Drosophilidae to evolve into such a dramatically diverse set of species. They will also investigate how recent host and habitat declines have impacted the genomic diversity of Hawaiian Drosophilidae and develop genomic indicators to estimate the extinction risk of different species. By establishing a framework for insect conservation genomics, this project will contribute to the conservation of Hawaiian Drosophilidae and insects more broadly.
Environmental Predictability as a Driver of Savanna Plant Trait Variation
PIs: Carla Staver (Professor of Ecology and Evolutionary Biology) and Gabriele Villarini (Professor of Civil and Environmental Engineering and HMEI)
Most plants adapt their growth to match predictable seasonal and yearly environmental shifts, but savanna plants respond to environmental fluctuations in unpredictable ways, which has downstream implications for savanna biodiversity. Staver and Villarini will investigate a possible explanation for this unpredictability: that savanna plants are unable to adapt to environmental fluctuations in predictable ways because the environmental conditions they experience are so unpredictable. Savanna rainfall can be highly variable both within and across years, and plants can be unexpectedly wiped out by chance fires. This makes it difficult for plants to “choose” a strategy, for example in terms of how deep and wide to spread their roots, whether to invest in spines to ward off herbivores, and how much energy to invest in leaves and bark. To test their hypothesis, the team plans to develop a model that predicts plant growth based on rainfall, soil, and landscape features for two savanna ecosystems: Kruger National Park in South Africa, which has one annual wet season, and Mpala Research Station in Kenya. They will then evaluate the model’s predictions using remote sensing and field collected data from Kruger and Mpala. By advancing our understanding of biodiversity in savannas, which historically has received limited research attention, this project will help stem the loss of biodiversity in savanna ecosystems.
Ocean Biodiversity Narratives Beyond Charismatic Megafauna: A Case Study in South African Kelp Forests
PIs: Allison Carruth (ENV Program Director and Professor of American Studies and HMEI) and Laure Resplandy (Associate Professor of Geosciences and HMEI)
Kelp forests, which support marine biodiversity and coastal livelihoods, are declining globally. The Great African Seaforest, which extends along the west coast of South Africa from Cape Town to Namibia, is the only kelp forest to have expanded over the past four decades. But despite this expansion, two species within the Great African Seaforest are at risk of extinction due to local environmental stressors and overfishing: the South African abalone (Haliotis midae) and the West Coast rock lobster (Jasus Ialandii). Both species are important for the kelp forest’s overall health—while abalone help prevent algal blooms, rock lobsters keep ecologically destructive sea urchins in check—but neither is the typical poster child for conservation. To aid abalone and rock lobster conservation, Carruth and Resplandy will create a series of narrative-based multimedia stories that interweave scientific research with community interviews and archival materials to paint portraits of the abalone and rock lobster within the context of the changing kelp forests. These stories will take a variety of forms, including videos, podcasts, photography collections, and social media campaigns, which the team will compare and assess through test group viewings with community members. This project will raise local awareness for the abalone and rock lobster and create a blueprint for developing storytelling approaches that aid the conservation of other species that are easily overlooked but ecologically essential.