COS 94-5 - Disorder or a new order: how climate change affects phenological variability

Tuesday, August 16, 2022

2:30 PM – 2:45 PM EDT

Presenting Author(s)

MS

Michael Stemkovski

Utah State University
Logan, Utah, United States

Coauthor(s)

JB

James Bell

Rothamsted Research Institute, United Kingdom
EE

Elizabeth Ellwood

Natural History Museum of Los Angeles County, United States
BI

Brian D. Inouye, PhD

Florida State University
Tallahassee, Florida, United States
HK

Hiromi Koboro

Tokyo City University, Japan
SL

Sang Don Lee

Ewha Womans University, Republic of Korea
TL

Trevor Lloyd-Evans

Manomet Observatory, United States
RP

Richard B. Primack

Boston University, United States
BT

Barbara Templ

data-analysis OG, Austria
WP

William D. Pearse

Imperial College London, United Kingdom

Background/Question/Methods

Advancing spring phenology is a well-documented consequence of anthropogenic climate change, but it is not well understood how climate change will affect the variability of phenology year-to-year. Species’ phenological timings reflect adaptation to a broad suite of abiotic needs (e.g. thermal energy) and biotic interactions (e.g. predation and pollination), and changes in patterns of variability may disrupt those adaptations and interactions. We analyzed a collection of phenology datasets from across the northern hemisphere to detect patterns of phenological shifts, temperature sensitivity, and changes in the inter-annual variability of spring onset events using quantile regression.

Results/Conclusions

Our analysis included nearly 10,000 long-term time-series representing over 1,000 species of plants, insects, and birds, and spanned a wide range of climatic conditions, allowing us to investigate the regional drivers of phenological changes. We found that early-season species in colder and less seasonal regions were the most sensitive to temperature change and had the least variable phenologies. The timings of leaf-out, flowering, insect first-occurrence, and bird arrival have all shifted earlier (on average 1.63 days/decade, p < 0.001) and tend to be less variable in warmer years (on average -0.21 days/°C). This has led leaf-out and flower phenology to become moderately but significantly less variable over time (leaf-out by -0.28 days/decade, p < 0.001; flowering by -0.26 days/decade, p < 0.001). These simultaneous changes in phenological averages and the variation around them have the potential to influence mismatches among interacting species that are difficult to anticipate if shifts in average are studied in isolation.