Butterfly flight phenology in response to urban heat island effects

Background/Question/Methods

Urban ecology can inform our predictions of how organisms will respond to climate change. City dwellers are selected not only for their ability to survive in degraded habitat, but also for their adaptation to increased temperatures. Due to the urban heat island effect, cities are warmer and drier on average than the surrounding rural environment. Phenology shifts are one mechanism by which organisms can track suitable thermal conditions. By observing which species do not respond well to elevated temperatures in urban areas, we may be able to predict which species will decline as the global climate warms. We measured the summertime emergence dates of 38 species with weekly transects in 15 sites representing five habitat classifications along an urban gradient in Toronto, ON, Canada. We extracted the daytime land surface temperature of each site using Landsat Thematic Mapping imagery. For each site, the timing of a species’ first flight was subtracted from the species-wide average to quantify accelerated or delayed phenology. These differences were regressed against year, habitat, species, site location, surface temperature, and butterfly abundance to evaluate which factors affect urban phenology.

Results/Conclusions

The best model ranked by AIC included site temperature, site longitude, and butterfly abundance. Higher daytime temperatures at a site significantly accelerated the emergence of butterflies. Although nighttime temperatures are elevated in urban heat islands, we did not measure them directly. Sites near the western boundary of the urban area had significantly delayed phenology, likely due to cooler temperature in the upwind rural areas. Eastern sites should be warmer at night as impervious surfaces in the city radiate heat that raise the downwind air temperature. A common criticism of phenological studies is that extreme events are more likely to be observed with larger sample sizes. We confirmed this sampling effect; higher butterfly abundance at a site was significantly correlated with earlier emergence. Although the habitat classification (golf course, park, residential, industrial, or nature preserve) did not have a significant effect on emergence date, it plays a role in determining the species composition of the site.