Collaborative research: using AI-powered non-lethal sampling to inform evidence-based forest management practices and enhance pollinator conservation

Mississippi State University received a $797,762 Project Grant from the National Science Foundation (NSF) Biological Sciences program (CFDA 47.074) awarded September 1, 2026, with a completion date of August 31, 2030. The grant supports collaborative research to develop, test, and implement non-lethal monitoring methods for pollinators in regenerating forests using artificial intelligence (AI)-powered acoustic and camera-based technologies. The primary deliverable is a real-time AI system capable of automatically identifying and tracking diverse pollinator species in forest environments, thereby providing scientists and conservationists with actionable data on pollinator activity patterns and population dynamics without the destructive impacts of traditional lethal trapping methods. The software system will learn and adapt over time through cutting-edge AI techniques, producing easy-to-use tools that integrate species identification data into forest management practices to support evidence-based conservation of native bee populations.

The project is executed as a collaborative multi-institutional research effort. Kansas State University (sub-awardee) focuses on developing the real-time AI identification system under the leadership of Dr. Samee U. Khan and collaborating researcher Dr. Khoa Luu from the University of Arkansas at Fayetteville. The University of Massachusetts-Amherst (sub-awardee), through co-Investigator Laura Figueroa, provides ecological oversight including Northeast field site selection, technician training on audio recorder deployment, and graduate student mentorship on data analysis. The National Council for Air and Stream Improvement (NCASI) Foundation (sub-awardee) manages research technician administration, coordinates cross-institutional collaboration, and integrates collected data into NCASI's Habplan harvest scheduling software to communicate simulation results to forest management stakeholders and member companies. Together, these partners deliver an integrated monitoring system and supporting analytical tools to advance sustainable pollinator conservation in managed forest ecosystems.

Abstract

The united states is home to nearly 4,000 species of native bees, which are important for the ecosystem. Unfortunately, declines in many economically important species have been documented in past years. Recent studies found regenerating forests that are managed for timber in the united states can be refuges for wild and native pollinators, including rare and economically important species of bees. However, despite this knowledge, there remains a lack of sustainable management practices for conservation of wild bees in managed forests. Moreover, monitoring BEE pollinators in forests is currently very difficult and unfortunately destructive in nature, as IT requires lethal trapping of individuals, which are then identified in a laboratory. Lethal trapping methods can have negative impacts on pollinator populations and are labor-intensive and inefficient. Furthermore, pollinators may be shifting their activity based on changes in average temperature, and we currently do not have effective ways to track these changes. The primary goal of this project is to develop, test, and implement non-lethal methods for monitoring pollinators in forests using acoustics and camera-based artificial intelligence (AI). Native BEE species in the unites states contribute as much as 3.5 billion dollars annually to agricultural pollination, but bees are on the decline. This project will develop AI technology that can be deployed in a field setting to automatically identify pollinator species in real time, thereby tracking patterns of activity. By combining different cutting-edge AI techniques, the system will learn and adapt over time, making IT more accurate and user-friendly. The goal is to create easy-to-use software that can help track pollinators in the wild, giving scientists and conservationists valuable insights into how structural changes affect these important species. The new technology will enable assessment of the status of pollinators across forests in the southeastern and northeastern united states in real-time, tracking of changes in pollinator activity, and determination of how changes in the forest landscape may affect pollinator abundance and diversity. This information will then be integrated into a harvest scheduling program that forest companies can use to help them in conservation planning, which is part of sustainable forest certification programs. This research will inform conservation strategies, helping protect pollinators and the ecosystems that depend on them. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the foundation's intellectual merit and broader impacts review criteria.- subawards are not planned for this award.