Science & Technology, Australia (Commonwealth Union) – Researchers from The University of New South Wales (UNSW) have stated that machine learning can play a role in extracting vital information from the large number of plant specimens stored in herbaria.
In a world-first, researchers from UNSW together with the Botanic Gardens of Sydney have trained AI to show data from millions of plant specimens stored in herbaria worldwide, for evaluating and tackling the effects of climate change on flora.
The lead author for the study, Associate Professor Will Cornwell indicated that herbarium collections are perfect items to measure the past of plant specimens. He also indicated that every year more than 8000 specimens are added to just the National Herbarium of New South Wales, so it is not feasible to look into them manually anymore.
Utilizing an innovative machine learning algorithm, the research team examined over 3000 leaf samples and made a surprising discovery. Contrary to commonly observed trends across different species, the study revealed that leaf size does not increase in warmer climates within a single plant species.
This groundbreaking research, featured in the American Journal of Botany, not only challenges the prevailing belief that climate is the sole determinant of leaf size variation within a species but also highlights the transformative potential of artificial intelligence (AI) in dynamically documenting the impacts of climate change.
By harnessing the power of AI, the study effectively revolutionizes the analysis of static specimen collections, enabling swift and comprehensive documentation of the effects of climate change on various plant characteristics. This advancement paves the way for more efficient and accurate investigations into the factors influencing leaf size and offers new insights into the intricate relationship between plants and their environments.
Herbaria have been known to serve as scientific libraries for plant specimens that were there since at least the sixteenth century.
“Historically, a valuable scientific effort was to go out, collect plants, and then keep them in a herbarium. Every record has a time and a place and a collector and a putative species ID,” explained Associate Professor Cornwell, a researcher at the School of BEES who is also a member of UNSW Data Science Hub.
Roughly 2 years back to assist in carrying out scientific collaboration, there was a movement to transfer the collections online.
“The herbarium collections were locked in small boxes in particular places, but the world is very digital now. So to get the information about all of the incredible specimens to the scientists who are now scattered across the world, there was an effort to scan the specimens to produce high-resolution digital copies of them.”
The Botanic Gardens of Sydney embarked on an ambitious endeavor, known as the largest herbarium imaging project, which involved the conversion of over 1 million plant specimens from the National Herbarium of New South Wales into high-resolution digital images.
Over the course of more than two years, the dedicated team meticulously carried out the digitization project. Upon its successful completion, Dr. Jason Bragg, one of the researchers involved, reached out to me from the Botanic Gardens of Sydney. He expressed a keen interest in exploring the integration of machine learning techniques with the extensive collection of high-resolution digital images from the Herbarium.
Dr. Bragg’s enthusiasm was palpable as he collaborated with A/Prof. Cornwell to develop models capable of detecting leaves in plant images. These models would then be utilized to analyze vast datasets, enabling the examination of relationships between leaf size and climate factors.
This indicates a promising partnership between Dr. Bragg and A/Prof. Cornwell held great potential for advancing our understanding of the intricate connections between plant morphology and environmental conditions, fueled by the application of cutting-edge machine learning technology to the invaluable repository of digitized Herbarium specimens.
