Herbaria as Big Data Sources of Plant Traits

Abstract

Herbarium specimens have long been a cornerstone of taxonomic research but are only recently being recognized for their potential as a source of spatially and temporally extensive data on plant functional traits. Many researchers in trait-based disciplines, including functional ecologists and evolutionary biologists alike, remain surprisingly unaware of this powerful use of herbaria. This review brings together disparate studies to synthesize the past, current, and potential future uses of specimens as functional trait data sources, answering the following questions. First, what insights using specimens (including traits measured, approaches used, and research questions answered) have been made to date? Second, what new trait-based insights (including potential contributions to global trait databases and recent advancements such as machine learning and high-throughput phenotyping) can be made from herbarium specimens? And third, how can inherent limitations and collection biases be addressed when using specimens in unanticipated ways (including new analytical approaches and improved methods for collecting)? I conclude by identifying what is needed to foster the future of herbaria as big data sources of plant traits. Most notably, plant collecting must be continued, expanded beyond predominantly systematists, and intentionally revised with downstream trait measurements in mind. Furthermore, community-wide standards are needed to integrate otherwise disconnected data and directly link newly derived measurements back to specimen records. Specimens serve as reliable (and necessary) sources of phenotypic data, enabling us to answer questions across phylogenetic, temporal, and spatial dimensions that are otherwise not possible to answer. Herbaria should be embraced as centers for functional trait research.