Abstract

AbstractOrganisms across the globe are experiencing shifts in phenological events as a result of ongoing climate change. Recently, a variety of novel methods have been applied in order to fill gaps in the phenological data set, in which records often have a patchy temporal, spatial, and/or taxonomic resolution. Here, I tested whether changes in flowering phenology could be detected through the months of flowering stated in 11 guides to the Swedish flora published over a period of 220 yr (1798–2018), focussing on 241 plant species (approximately 8% of the Swedish flora), and accounting for the large increase in herbarium records that have occurred over the same period. Despite the coarse, monthly scale of flowering times reported, historical floras and wildflower guides may hold potential to fill temporal and taxonomic gaps in the plant phenological data set. However, factors other than climate may also influence any apparent phenological shifts over time. Here, flowering was found to start earlier (0.49 d/decade), end later (0.71 d/decade), and carry on longer (1.19 d/decade), with flowering length also associated with increases in the regional temperature anomaly during the 20th century (0.11 months/°C). First flowering occurring earlier in 71% of species (14% showing a significant negative trend), 68% of species ceased flowering later (20%), and 80% flowered for longer (29%). Detected phenological shifts also appeared to be related to species’ flowering seasonality. Later‐flowering species were found to flower later and for longer, while increasing temperatures appeared to drive stronger responses both in flowering onset in early‐flowering species and in flowering cessation in later‐flowering species. Although potential issues exist regarding the largely unknown ways by which authors have determined flowering times and the coarseness of the data, historical floras may be a useful resource in phenological and climate change research, with the potential to both identify and compare the broad climatic responses of a region’s entire flora over long time periods, as well as filling gaps in an otherwise patchy data set.

Highlights

  • Together with shifts in species distributions (Chen et al 2011, Freeman et al 2018) and the rearrangement of species within communities (Devictor et al 2012, Auffret and Thomas 2019), shifts in phenological events are one of the main fingerprints of anthropogenic climate change on the natural world (Parmesan and Yohe 2003, Root et al 2003, Poloczanska et al 2013)

  • Considering all species together, the flowering times stated in Swedish floras have gradually become earlier over time, in line with a warming climate (Fig. 1, Appendix S1: Table S2)

  • Based on the parameter estimate of the linear mixedeffects model, the shift in start of flowering occurred at a rate of 0.49 d per decade (95% confidence intervals 0.31–0.65), or more than one and a half weeks across the 220-yr time period

Read more

Summary

Introduction

Together with shifts in species distributions (Chen et al 2011, Freeman et al 2018) and the rearrangement of species within communities (Devictor et al 2012, Auffret and Thomas 2019), shifts in phenological events are one of the main fingerprints of anthropogenic climate change on the natural world (Parmesan and Yohe 2003, Root et al 2003, Poloczanska et al 2013). Shifts in the timings of life-history events have been linked to species responses to climate change in terms of distributional shifts (Amano et al 2014, Macgregor et al 2019). It is clear that phenological data are a valuable resource for understanding ecological responses to a changing climate at both the species and the community level. Such data sets are generally quite patchy. Org.uk/), and its Swedish equivalent (https:// www.naturenskalender.se) are constrained to a more recent time period and a specific set of study species and events that are recognizable to the wider public. Recent advances in machine-learning technology provide further potential for herbariabased phenology research by automating the processing of specimens (Pearson et al 2020)

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call