Autumn leaf phenology: A search for patterns using 1000 species at four botanical gardens

Background/Question/Methods

A warming climate is increasing the length of the growing season in many temperature ecosystems, due to both earlier leaf out in the spring and later leaf senescence in the autumn. Spring leaf out is relatively easy to measure, well-studied, and driven largely by temperature. Autumn leaf senescence, on the other hand, is less well studied, more difficult to measure – as it involves both leaf color change and leaf drop – and is driven by multiple environmental cues.  We monitored 1000 woody tree, shrub and vine species once a week at four botanical gardens in North America (Ottawa and two in Boston) and Germany (Berlin) to address three questions: (1) Are there growth habit differences in leaf senescence times? (2) Are leaf senescence times correlated across botanical gardens and between years? (3) Are autumn leaf senescence times related with spring leaf out times? A species’ leaf senescence date was recorded when at least 50 % of the leaves on a plant had changed color or had fallen off; only one date was recorded per species. 

Results/Conclusions

Certain species began leaf senescence in August, and others not until December or later. Some species drop their leaves while still green, other species’ leaves change color while on the tree, and most species have a combination of leaf color change and drop. There was a tendency for tree species to undergo leaf senescence before shrubs at one site, but not at the other sites. Senescence times were strongly correlated between years at individual sites. However, there is little correlation of leaf senescence times between the sites, suggesting species are responding in unique ways to site conditions and climatic variation, or that methods of evaluation are not sufficiently standardized. Autumn leaf senescence times were not correlated with spring leaf out times. In conclusion, autumn leaf senescence represents challenges to climate change studies due to problems with evaluation, lack of clear patterns associated with growth habit, and lack of correlation among sites. Future work should focus on evaluating variability of additional species and sites in senescence times, developing and standardizing a method of evaluating leaf senescence that captures more of the process of senescence, and investigating the physiological mechanisms that affect leaf senescence in the field.