Urbanization advanced plant phenology in cold, but not warm, areas

Analysis of an integrated database of in situ phenological observations across two continents reveals that high human population densities advance plant flowering and leaf-out dates in cold areas but delay them in warm areas. Urban heat island effects only explained part of this pattern.

Ecosystems in cities are different from surrounding wildland and rural areas, and one of the most noticeable examples is that cities often have extended growing seasons compared to adjoining areas. Perhaps this comes as no surprise since cities are relatively warmer during spring due to the urban heat island effect, and warm conditions are critical for spurring plants to leaf out and flower. But does this pattern holds for all cities with different average temperatures? For example, if plants in New York city flower and leaf-out earlier than its surrounding rural areas, do plants in Orlando show the same pattern? Our work examined this question and found the surprising result that urbanization does not always advance plant leaf out and flowering. In the most southern cities in the Northern hemisphere, plant season timing, also called phenology, may actually be delayed.

Blooming flowers in the city of Madison, Wisconsin, USA

In order to ask questions about phenology across large spatial areas, monitoring data of plant phenology at continenetal to global scale scales are needed. Our study would be impossible without open phenological data shared by phenological networks from both the United States (USA-NPN and NEON) and Europe (PEP725). The PEP725 network includes 32 European meteorological services and serves data assembled from the late 1800’s to present. The end result is more than 12 million phenology observations covering 100s of species. The NPN was founded in 2007 and through its partnerships and networks brings together historical and current data resources about plant and animal phenology. Both PEP725 and NPN are critically important resources for phenology, but one downside is that integrating the data in those networks into a larger whole is challenging. 

The critical reason why integrating data from NPN/NEON and PEP725 is so challenging is that these different data sources have different ways of reporting phenology. Put simply, NPN and PEP725 don’t speak the same phenological language. For example, NPN and PEP725 divide plant development up into different phenological “stages,” and they use different terminology to describe those stages. They also differ in the methods they use to observe and record phenological information. These incompatibilities make it quite difficult for end users to meaningfully (and correctly!) combine the information in the NPN and PEP725 datasets. To address this, we have developed machine-reasoning based systems that are able to intelligently -- and automatically -- combine the data from all three data sources. The results are available at a web portal that makes it easy to simultaneously access all of these data sources (plus a few more).

Leaf-out of a Tulip tree (Liriodendron tulipifera)

Using the standardized data from this portal, we found that urbanization considerably advanced both flowering and leaf-out in relatively cold areas, such as New York city. But surprisingly, urbanization delayed flowering and leaf-out in warm areas, such as Orlando. We also found that this effect is not entirely due to the effects of the “heat islands” created by cities. Even after accounting for the urban heat island effect, we found significant changes in leaf-out and flowering times of plants in more urbanized areas, which suggests that other environmental changes caused by urbanization must be to blame. Although our study covered a large geographical area, it was still limited to the United States and western Europe. Do the patterns we found also hold in more tropical areas? What about in the southern hemisphere, or in developing countries where urbanization is occurring at a faster pace? Answering these questions will require even more phenological data.

To move the field forward, long-term, large scale monitoring programs and open data are the keys.

Please sign in or register for FREE

If you are a registered user on Nature Portfolio Ecology & Evolution Community , please sign in