PHENOLOGY & CLIMATE

[PUBLICATION] Temperature controls phenology in continuously flowering Protea species of subtropical Africa.

Protea species in South Africa

Overview

This research, published in Applications in Plant Sciences (2019), investigates how climate influences flowering patterns in the iconic Protea genus of subtropical Africa. While most phenological research has focused on temperate plants with discrete flowering seasons, this study examined how plants with year-round flowering respond to climate variation, finding that even in these aseasonal species, temperature is a primary driver of flowering timing.

Research Approach

In this study led by Barnabas Daru, we analyzed phenological patterns for species with year-round flowering, using a database of herbarium specimens collected between 1950 and 2011. I developed the novel circular statistical methods for analyzing aseasonal phenology and its relationships with spatial and temporal climate variation. Our analysis:

Characterized seasonal and geographic flowering patterns across 25 Protea species
Evaluated how site-to-site and year-to-year variation in temperature and precipitation influenced flowering dates
Tested for phylogenetic conservatism in climate-driven phenological responses
Validated herbarium-based flowering estimates against independent field observations

Key Findings

Our study revealed that:

Protea species advance flowering by 3-5 days per degree Celsius of warming, both across spatial gradients and between warmer and cooler years
This temperature sensitivity is remarkably similar to that found in better-studied northern temperate plant species, suggesting a generality across biomes
Precipitation had no significant effect on flowering timing, highlighting temperature as the primary driver
Climate responses are phylogenetically conserved, with closely related species showing similar shifts in flowering time with warming

Methodological Innovation

The sliding window approach developed for this study represents an important advancement in phenological research methodology. By re-centering each species' observations on periods of maximum and minimum flowering activity, we were able to apply linear regression methods previously used only in highly seasonal temperate ecosystems to plants with more complex, year-round flowering patterns.

This approach was validated by a strong correlation (r = 0.93) between our herbarium-derived flowering estimates and independent expert field observations, demonstrating the utility of herbarium collections for phenological research in understudied tropical and subtropical regions.

Model results