Purdue University


A central question this research group poses is: how can we meet the needs of people while sustaining the diversity of plants, animals and microbes that we depend on. Recent reports about the status and trends in biodiversity are alarming: one in eight species on Earth are listed as endangered or threatened, the number of birds in North America has declined by over 29% in the last 50 years, pollinators of crops are being lost at unprecedented rates (one in five bites of food are provided by pollinators), and iconic animals such as penguins, tropical tree frogs, and primates may not be here by the end of the current century.

There is also an urgent need for the public to become more aware of the biodiversity crisis. With over half of society now uncoupled from nature because they live in urban or semi-urban areas, new ways to engage the general public in biodiversity protection are needed. The Biodiversity SRA focuses on several aspects of biodiversity, including its relationship with:

  1. Artificial intelligence
  2. Environmental Sensing
  3. Food Security
  4. Outreach and Education

Biodiversity Projects

Multi-sensor Biodiversity Framework Developed from Bioacoustic and Space-based Sensor Platforms

PI: Bryan Pijanowski

Co-PIs: Jinha Jung, Jingjing Liang

Funding: NASA

The vision for this work is to use three ISS sensor platforms (GEDI, DESIS and ECOSTRESS), several space-based remote sensing platforms (e.g., MODIS, Landsat, ICESat 1/2), in situ acoustic sensor data, and "silent" in situ data (field surveys, national and regional forest inventory data, UAV data, and meteorological data) to build a multi-sensor biodiversity framework that is applied to major terrestrial global biomes. This three-year, hypotheses-driven project will be accomplished through implementation across five major integrative tasks: prepare data, build and validate habitat maps, develop habitat-biodiversity models, test hypotheses, and create a Biodiversity Framework. The team is leveraging two global biodiversity databases (a 4 million+ audio recording database from Purdue's Center for Global Soundscape and the 1.2+ million plot global database of tree species inventories from Purdue's Global Forest Biodiversity Initiative) to build a Biodiversity Framework using three dimensions of habitats at landscape levels (structural, compositional, and dynamic) that will allow scientists to examine how climate change and habitat alteration affects animal and plant biodiversity across the globe. The team is testing phases of the model development using advanced machine learning techniques (i.e., convolutional neural network, random forest, discriminant analysis, hierarchical clustering) and multivariate statistical methods (i.e., parametric and non-parametric) for temporal and spatial trend analysis (i.e., time-series, spatial auto-correlation, ARIMA models). This work is filling a gap in current understanding of the dynamics of landscapes and animal presence and activities over time by developing phenology models for plants and animals at major terrestrial biomes on Earth. Overall, this project will advance NASA's Biodiversity A.7 initiative by integrating a massive library of recordings from over 11 different ecosystems around the world to build models using space-based remote sensing and other in situ data to understand how plant and animal diversity varies over time, space and disturbances.

Hardwood Ecosystem Experiment

We speak for the trees. This could be the slogan for the long-term Indiana-based research project known as The Hardwood Ecosystem Experiment (HEE), a 100-year collaborative study of the effects of forest management on animals and plants in Indiana. While many may not think of Indiana as a ‘forest’ state, it has, particularly in the southern third of the state, oak and hickory forests that have been adversely affected by fire, grazing and timber harvesting. HEE brings together academic, non-profit and state organizations including, among others, Purdue, the Indiana Department of Natural Resources and the Ruffed Grouse Society, a conservation group focused on improving wildlife habitats and forest health. Together, these organizations and their researchers have articulated several objectives including:

  • Develop a proven system of forest management prescriptions to maintain desired populations of native plant and animal species and important communities such as those dominated by oak species;

  • Understand the response of targeted native wildlife, plant species, and microbes to forest management, in order to identify the positive effects and mitigate the potential negative effects on species of conservation concern;

  • Assess public attitudes towards forest management to develop new approaches for education of the general public and private landowners; and to engage various interest groups in a discussion of proper land management;

  • Identify direct and indirect benefits of specific forest management practices to local and regional communities, and understand the impact of forest management practices in community development.

Several Center affiliates conduct research in the HEE, including Barny Dunning (FNR), who studies songbirds and owls; Mike Jenkins (FNR), who studies understory vegetation; and Catherine Aime (Botany and Plant Pathology), who focuses on fungi.

Record the Earth

What does the Earth sound like today? This citizen science project gives individuals the opportunity to join with researchers involved in one of the newest scientific disciplines – Soundscape Ecology – as they map the sounds of our planet, and what they tell us about the health of the world we live in.

The Record the Earth app and website allows anyone to record their soundscape – the way the Earth sounds where they are – and upload it to a database with over 6,000 others, adding to an ever-growing map of recordings from around the world.

A Global Sustainable Soundscape Network

PI: Bryan Pijanowski

Funding: NSF

In recent years, ecologists, cognitive psychologists, and scholars in the humanities have begun to study soundscapes from their own natural, social, and creative disciplinary perspective. Work in the U.S., Canada, and Europe is moving forward in these three camps with little to no coordination. Even within the ecological community, no formal group has been formed that allows research to be integrated and compared. To facilitate interaction among these communities of scholars and to invite scholars from other fields to examine sounds as measures of and perspectives into human and natural system interactions, this project has created the Global Sustainable Soundscape Network as a National Science Foundation Research Coordination Network (RCN).