Harnessing herbaria to advance plant phenology research under global change

Phenology, which refers to the timing of recurring biological events, is rapidly shifting under global change and has thus been a central topic in ecology. In an article published in this issue of New Phytologist, Park et al. (2023; 2153–2165) compiled a large phenological dataset comprising over 70 000 digitized herbarium specimens to investigate the impact of urbanization on the timing of plant reproductive events and the susceptibility of flowers to frost damage. This dataset, which comprehensively represented flowering and fruiting phenology over 120 yr in the eastern US, provided valuable insights into the complex interaction between urbanization and climate in shaping plant reproductive phenology. Their findings revealed that urbanization had diverse effects on plant phenology depending on the regional climatic conditions. Specifically, it advanced flowering in colder and wetter regions while delaying fruiting in wetter regions. Furthermore, the study identified that urbanization led to changes in the timing of spring frost and flowering, thereby increasing the risk of frost damage in areas with colder and wetter springs. These findings help anticipate potential changes in plant phenology in human-dominated landscapes under climate change.

Herbarium data offer unique strengths for studying shifting phenology under global changes (Willis et al., 2017; Fig. 1a), as Park et al. (2023) have demonstrated. Long-term records from herbarium specimens provide a direct way to establish a pre-climate change baseline for the relationship between phenology and environmental factors. These historical data cannot be replaced by ongoing efforts on phenological observation from remote sensing or crowd-sourcing. Furthermore, herbarium specimens are collected on a global scale, which enables phenological studies that transcend local and regional scales. Such temporal and spatial scales allow for investigations of the heterogeneity in phenological responses across the globe, including under-studied areas such as the global south and the tropics. With herbarium data, researchers can obtain a reliable and comprehensive understanding of how phenological shifts are occurring and their implications for ecological systems.

Herbarium data have emerged as some of the most compelling evidence for the impacts of climate change on phenology (Willis et al., 2017), providing researchers with crucial insights into the mechanisms and magnitude of shifts in the timing of biological events. For example, studies conducted on herbarium specimens from the Boston area have shown that flowering times have shifted earlier in response to warming between 1885 and 2002 (Primack et al., 2004). Larger datasets spanning wider geographic regions have also shown that flowering phenology advances under warming conditions, while revealing nuanced differences in response between plant functional types (Calinger et al., 2013). Conversely, warming may delay flowering in late-flowering species in warm regions, indicating that phenological responses to climate change can be complex and multifaceted (Park & Schwartz, 2015). In addition to tracking plant flowering phenology, herbarium data have been used to investigate a range of other biological events, including plant leafing, fungal fruiting (Meineke et al., 2018), and even animal phenology (Weaver & Mallinger, 2022). These studies showcase the diverse applications of herbarium data in detecting an unequivocal fingerprint of climate change on biological systems while elucidating complex mechanisms.

Adding to these studies, Park et al. (2023) conducted a comprehensive investigation by utilizing one of the most extensive herbarium datasets available. Their analysis included 71 278 plant specimens encompassing broad spatial, temporal, and taxonomic ranges, featuring 200 angiosperm species from 780 counties in the eastern US from 1895 to 2018. The considerable scale of the dataset employed by Park et al. (2023) enabled the examination of two intricate questions about phenological changes under global change related to the interplay between climate change and urbanization.

First, Park et al. (2023) addressed the context-dependent effects of anthropogenic stressors on phenology. It is known that urbanization can amplify, dampen, or offset the effects of climate change on phenology. Most current knowledge of this interaction is based on remote sensing data, which has broad spatial coverage but a short time span. Longer-term herbarium data have also been used to study the impact of urbanization on phenology, but these studies were limited to local scales (Neil et al., 2010). The work of Park et al. (2023) adds broader-scale evidence to support the previous finding of urbanization advancing plant phenology in colder regions (Li et al., 2019), and additionally further highlights the role of both temperature and precipitation as key components of the climatic background. We believe that future research using large herbarium datasets could move beyond linear models to identify nonlinear higher-order interactions.

Park et al. (2023) also tackled a second pressing question regarding the consequences of shifting phenology. Despite qualitative knowledge of the drivers of phenological shifts, there is a need for more accurate predictions of ecological consequences. Several assessments have been conducted on the changing time lag between frost and plant development under climate change, yielding inconsistent conclusions on the frost risk (Richardson et al., 2018; Park et al., 2021), depending on species and location. This complex topic was further investigated by Park et al. (2023), incorporating the impact of urbanization to refine predictions of spatially heterogeneous changes in frost risk. Focusing on the risk of frost to flowers, this study is part of a larger effort to examine the changing ecological synchrony in the Anthropocene, where herbarium data can provide realistic predictions given their extended time spans and broad environmental gradients.

The growing herbarium data, coupled with new methods, present remarkable opportunities for global change biology research (Meineke et al., 2018). Although the number of herbarium specimen records has increased considerably, their availability online lags behind contemporary observational data (Davis, 2023). A substantial proportion of herbarium specimens have yet to be made available for scientific research. For instance, at the University of Michigan Herbarium, c. 30% of the 1.7 million specimens are yet to be digitized and published (B. R. Ruhfel, pers. comm.). Intense digitization efforts are underway to capitalize on the vast potential of herbarium data (Fig. 1b). In addition, innovative techniques such as crowd-sourcing (Park et al., 2023) and machine learning are revolutionizing data extraction from herbarium specimens. These opportunities afforded by the ‘herbarium of the future’ (Davis, 2023) are expected to drive phenology research forward in the Anthropocene.

To stimulate herbarium research interest, we have identified a number of open questions regarding phenology in the context of global change. One such question is the magnitude and consequence of phenological mismatch on a large scale. The absence of a clearly defined baseline has resulted in controversy in the field, but herbarium data have shown the potential to address this issue by quantifying phenological mismatch with the abiotic and biotic environment (Weaver & Mallinger, 2022; Song et al., 2023). Furthermore, measurements of functional traits, which can be efficiently obtained from digitized herbarium specimens, can help explain and predict phenological responses (Perez et al., 2020). Another avenue of research is exploring the evolution of phenology and its coevolution with other traits with phylogenetic information in herbarium specimens (Meineke et al., 2018). Despite its potential, using herbarium data shares methodological challenges with contemporary phenological data collection efforts, such as sampling bias in crowd-sourced data and spatiotemporal uncertainty in social media data. Advances in the processing of herbarium data can enhance the use of phenological data from multiple sources. Studying the future of phenology under global change requires the examination of specimens in herbaria, which represent a powerful yet under-utilized data source.

KZ and YS contributed equally to this work.