Machine learning reveals how often fish schooling evolved—and repeatedly disappeared
BiologyEarth Science
THE AI ANGLE
Classifying behavioral descriptions and measuring fish alignment in imagesResearchers used machine learning to map schooling behavior across nearly three thousand fish species over 350 million years. They discovered that collective swimming evolved about 60 times and disappeared roughly 320 times. The findings show that flexible animal behaviors can be quantified and tracked across large evolutionary trees.
THE TEACHING ANGLE
Consider why fish lineages repeatedly abandon schooling behavior even though schooling groups produce new species at higher rates.Read the original at phys.org Generate teaching or study materials