Exploring Geographically Variant Climatic Requirements of Plant Phenology through Common Garden Observations

A more detailed understanding of population-level variations in plant phenology and the corresponding environmental drivers is crucial for monitoring and predicting geographically different phenological responses to climate change. Here I report an on-going project observing spring and fall phenology of an important native forest species Fraxinus americana (white ash) in a common garden/plantation in Kentucky, U. S. A total of 41 populations from across the species' distribution range are represented at this site. Weekly visual phenology surveys were conducted in spring and fall seasons of 2013, and spring season of 2014 (with the fall 2014 observation anticipated). Concurrent temperature variations have been recorded at the common garden with automatic data loggers. Preliminary results showed that in 2013 the southern populations demonstrated earlier spring leaf bud burst and later autumn leaf coloration and leaf fall than the northern populations. In spring 2014, with a delayed phenology overall, the southern populations obscurely appeared to show later leaf bud burst, and with frost damage observed on some of the trees from these populations. The 2012-2013 winter was warmer than normal, and the 2013-2014 winter had record low temperatures followed by a cold spring.

I hypothesize that the spring phenology timing of white ash is controlled by a strong chilling requirement with a less influential warming requirement, and the northern populations require more chilling (and less warming) than the southern populations. This is supported by the delayed phenology with the northern populations after a relatively warm winter in spring 2013. When the chilling requirement was fulfilled very early in spring 2014, the effect of warming was more manifested, leading to earlier phenology for the northern populations. Due to the quick response to warmth after the abundant chilling for some of the southern populations in 2014, frost injury occurred and may have confounded the phenological observations for some individuals. Moreover, the differential photoperiod requirement is also useful for explaining the geographic patterns in 2013, which may serve as an alternative hypothesis. However, given the observed interannual variations, these environmental factors are unlikely to operate alone, but are most possibly coupled with one another. More in-depth study of both the spring and autumn phenology (with additional data to be collected) of this particular species with the environmental drivers may provide important insight on and facilitate building a modeling framework for assessing geographically explicit climate change impact on plant phenology.