Leveraging machine learning and citizen science data to describe flowering phenology across South Africa

Summary: • Phenology — the timing of recurring life history events—is strongly linked to climate. Shifts in phenology have important implications for trophic interactions, ecosystem functioning and community ecology. However, data on plant phenology can be time consuming to collect and current records are biased across space and taxonomy. • Here, we explore the performance of Convolutional Neural Networks (CNN) for classifying flowering phenology on a very large and taxonomically diverse dataset of citizen science images. We analyse >1.8 million iNaturalist records for plants listed in the National Botanical Gardens within South Africa, a country famed for its floristic diversity (~21,000 species) but poorly represented in phenological databases. • We were able to correctly classify images with >90% accuracy. Using metadata associated with each image, we then reconstructed the timing of peak flower production and length of the flowering season for the 6,986 species with >5 iNaturalist records. • Our analysis illustrates how machine learning tools can leverage the vast wealth of citizen science biodiversity data to describe large-scale phenological dynamics. We suggest such approaches may be particularly valuable where data on plant phenology is currently lacking.

研究概要：
• 物候学（Phenology）——即周期性生命史事件的发生时间——与气候密切相关。物候期的转变对营养级相互作用、生态系统功能以及群落生态学均具有重要意义。然而，植物物候数据的采集往往耗时费力，且现有记录在空间分布和类群覆盖上均存在偏差。
• 本研究针对公民科学图像构建的大规模、类群多样化数据集，探索了卷积神经网络（Convolutional Neural Networks，CNN）在开花物候分类任务中的表现。我们分析了南非国家植物园收录植物的超180万条iNaturalist记录；南非以其约21000个物种的植物区系多样性闻名，但其相关数据在物候数据库中代表性不足。
• 本模型实现了超90%的图像分类准确率。借助每张图像附带的元数据，我们为拥有超5条iNaturalist记录的6986个物种重建了其开花盛期时间与花期时长。
• 本研究表明，机器学习工具可借助海量公民科学生物多样性数据，揭示大规模物候动态。我们认为，此类方法在当前植物物候数据匮乏的场景中具备尤为重要的应用价值。