No matter the temperature, I don't consider it to be really spring until I spot the first spring beauties of the year. These sweet whitish/pinkish mid-Atlantic florets (Claytonia virginica) are among the first to stretch out of the mud and leaf litter to add a spritely touch to an otherwise brown woodscape. When I see them for the first time each year, I linger a moment to admire their bright clean white petals standing out against the drab background and do a little dance—spring is finally here.

This appreciation for first blooms is not limited to the first flower of spring; naturalists have long recorded the first blooms of many species over the course of the year. This work is part of the field of phenology, which focuses on yearly or otherwise periodic events in nature and how they are influenced by climate. Some of the longest records in phenology are the first blooms of flowers, likely because of that same joy I feel with the spring beauties each year: First blooms are notable and thrilling to discover, a perfect combination for fun fieldwork for amateur and professional phenologists alike.

Identifying other flowering events—like peak flowering (when the most blooms of a particular species are open) and last flowering—is more difficult. They require noting a flower's absence rather than its presence, and being sure of absence is never easy. But by ignoring peak and last flowerings, no matter how difficult they are to keep track of, we risk missing changes in flowering patterns caused by climate change.

This serves to highlight a unique study out of Colorado’s Rocky Mountain Biological Laboratory, published yesterday in the Proceedings of the National Academy of Sciences. Researchers at the long-term research site (at 9,500 feet elevation) have been counting the flowers of 60 species for 39 years (since 1974), totaling more than 2 million flowers. "I was a graduate student studying hummingbirds and bumble bees, and I wanted to know what flower nectar resources are available for them, so I started counting flowers," paper author and University of Maryland phenologist David Inouye said in a press release. Their records include not only first flowering, but also peak and last flowering dates for each species.

By the end of those 39 years, the climate in the area had changed: temperatures at the mountain meadow were around 2.7°F (1.5°C) higher and snow melted 12 days earlier. And the flowers have noticed. The full flowering season, from first flower to last, stretched 36 days longer than when the researchers started taking notes on flowering times four decades ago; it used to run from late May to early September, and now lasts from late April to late September. The first spring flower bloomed about 24 days earlier*, the spring peak occurred 20 days earlier, and the last flower of fall closed about 12 days later.

But thinking of this as a whole ecosystem change—flowering season now a month longer than four decades ago!—obscures the true complexity of this mountain meadow. Each flower species is dancing on its own, blooming earlier or later depending on its preferred conditions. For the one-third of flower species that didn’t react to climate change in this study, their growing season is just as long as before. Just under one-fifth of the flowers shifted forward entirely, blooming, peaking and closing earlier while their flowering season remained the same length. For a quarter of species, the three dates didn’t shift in synchrony, and their growing seasons were shortened or lengthened. And illustrating the need to look beyond first flower, another quarter had significant changes in their peak or last flowering date with no change in the first.

For some species, the longer blooming season will mean more growth and a population boom; for others, the warmth is less than ideal, causing those species to migrate to higher or lower elevations to a more comfortable temperature. Each of these species’ shifts, mostly imperceptible to us, will alter the composition of the meadow—and the resources available to pollinators and other animals in the area. This isn't necessarily a bad thing; most animals can adapt to changes in their resource distribution and availability. But over time, we might see more significant changes in the ecosystem.

Numbers are very nice and all, but they don't quite convey the complexity of the changes. So here are some individual species and their changes in flowering times to help us imagine it better.

Heartleaf bittercress (Cardamine cordifolia) now blooms 28 days earlier, peaks 25 days earlier, and closes 18 days earlier, for a longer growing season by 10 days.**

Twolobe larkspur (Delphinium nuttallianum) first blooms 14 days earlier and peaks 12 days earlier for a longer flowering season.

Sulphur buckwheat (Eriogonum umbellatum) has its last bloom 41 days earlier—but its first bloom has stayed the same. Its abundance has decreased by 88 percent.

Yarrow (Achillea millefolium) blooms linger for 17 more days into the fall.

Creeping mahonia (Mahonia repens) peaks 16 days earlier.

Wild strawberry (Fragaria virginiana) now blooms 11 days earlier and is 36 percent more common.

*These numbers are all calculated from the average change in days per decade * 4 (39 years ~ 4 decades). Basic info: (1) First flower bloomed 6 days earlier per decade on average; (2) Spring peaked moved up 5 per decade on average; (3) Last flower bloomed 3 days later per decade on average; (4) Site temperature increased 0.4°C per decade on average; (5) Snowmelt occurred 3.5 days earlier per decade on average

** All individual species flowering times calculated from the average change per year * # of years recorded; see additional paper data.

CaraDonna P.J., Iler A.M. & Inouye D.W. Shifts in flowering phenology reshape a subalpine plant community, Proceedings of the National Academy of Sciences, DOI: