OOS 25-8 - The discovery of decadal cycles in Rocky Mountain flowering, facilitated by LTREB funding from NSF

Wednesday, August 9, 2017: 10:30 AM
D136, Oregon Convention Center
David W. Inouye, Rocky Mountain Biological Laboratory, Crested Butte, CO

Cyclic patterns of abundance occur in ecology, and we have some understanding of what generates them. The best examples are from animal populations, and the question arises whether there are cycles in plant population dynamics. There are many examples of tremendous variation among years, such as the flowering of desert annuals in response to unusually heavy rains. Mast flowering of some trees and of bamboo may be examples of plant cycles, and there are a few other examples that may be appropriate matches to animal examples. In 1973 I initiated a study of the abundance and phenology of flowering by 120 species of wildflowers at the Rocky Mountain Biological Laboratory. The is significant variation among years in these parameters, and we know the environmental variables that generate much of it. Winter snow depth and the timing of snowmelt in the spring explain most of the variation in phenology and influence the abundance of flowering for some species, while summer precipitation can also play a role. One of the benefits of having decades-long data is that they can be used to look for variation on the scale of decades, and there is now evidence of such a pattern in my data.


By plotting six-year running subsets of data from 1935 – 2016 on peak runoff in the East River at Almont, CO, which is strongly correlated with the amount of winter snowfall and with snowmelt date in my study plots, I identified peaks of late snowmelt in 1953, 1970, 1982, 1994, and 2006 (intervals of 17, 12, 12, and 12 years). These are good matches with peaks in precipitation anomalies identified in the Rocky Mountains by Ault and St. George (2010).

Delphinium nuttallianum (Ranunculaceae) is a common spring-flowering wildflower at RMBL, whose abundance and phenology are strongly influenced by the previous winter’s snowpack and spring snowmelt date; flowering is much more abundant following years with more snow and later snowmelt. A survey of flowering abundance for Helianthella quinquenervis, carried out since 1974, identified three cycles with peaks at about 1982, 1995, and 2009 (intervals of 13 and 14 years). The length of these cycles mean that even a 10-yr LTREB award from NSF is insufficient to identify them, yet given the importance of wildflowers as resources for pollinators and seed predators, knowledge of these cycles makes it possible to forecast variation in the resources and the implications for plant demography.