Forest wildflowers in Europe are flowering much earlier. The study, by Franziska Willems and colleagues and published in New Phytologist, found that, on average, plants now flowered almost a week earlier than in 1900. The study used herbarium specimens not only to track plant changes over time but also over geographical space.
Phenology, the study of the timing of life-cycle events, like first budding, flowering or fruit development, is a powerful tool for studying the effects of climate change. But these changes are best understood over a long time frame, and an interest in how the climate is changing is comparatively recent. Herbaria, collections of preserved plant specimens, are valuable sources of historical plant data, but it’s not simply a matter of going to a drawer and grabbing the material to study it.
Willems and colleagues write in their article that there’s often a strong geographical bias in herbarium collections. “[S]pecimens are more frequently collected where collectors live, and around academic institutions… However, when modelling across a spatial range, standard methods such as linear regression ignore the spatial dependency between sampling locations and treat all data points as independent. This assumption is very likely to be not correct, as the proximity of spatial locations is usually related to their environmental similarity… Ignoring spatial dependency, therefore, results in pseudoreplication and it can strongly bias model results.”
The botanists analysed trends in the flowering time of twenty common forest understory wildflowers, and their relationships with climate change, across Europe. This involved examining over six thousand herbarium specimens collected over more than a century. The choice of forest wildflowers was because of the environment that the wildflowers live in.
“We focused on early-flowering understory plants, because they have a very distinct phenology, with a critical time window for flowering before the leaf-out of deciduous trees. Because of this, they may be particularly sensitive to climate change and phenology shifts,” write Willems and colleagues.
Research, such as a recent study on Thoreau’s observations in his book Wild Fruits, has shown that the herbarium records are likely to be very similar to field observations carried out at the same time. As long as the records are accurately labelled so that botanists know where the samples came from, they can be used to build a record of how plants adapt to climate. Understanding the spatial element is critical to success, write Willems and colleagues.
“Herbarium data from large geographic ranges are particularly powerful, but they also come with challenges, and we showed that accounting for spatial autocorrelation significantly improved model fits and parameter estimates. Phenology as well as phenological responses to climate change can vary substantially across large scales, and failing to account for this might draw a biased picture of how climate change affects plants and their associated communities and ecosystems. Future studies should more frequently employ spatial modelling when analysing large-scale phenology variation and its different drivers, ideally across multiple climatic regions.”
Willems, F.M., Scheepens, J.F. and Bossdorf, O. (2022) “Forest wildflowers bloom earlier as Europe warms: lessons from herbaria and spatial modelling,” New Phytologist. https://doi.org/10.1111/nph.18124