At what times of year are phenological events across species sensitive to climatic variables, and how sensitive are they? Answers to these questions emerge from the analysis of a wealth of long-term data sets. See Article p.241
Spring is an exciting time of year, with flowers coming into bloom, bees emerging from their winter rest and migrant birds returning to their breeding grounds. In the United Kingdom, all such phenological events have been monitored for decades — across the seasons — by professionals and citizen scientists alike (Fig. 1). And although many of these schemes started well before we recognized that the world is warming, these long time series now provide a valuable infrastructure to document the impact of global climate change on nature. On page 241, Thackeray et al.1 use no fewer than 10,003 phenological data sets, all collected over periods of at least 20 years, to quantify the unequal rates at which phenology has shifted in different species groups. Climate change thus leads to disruption of the phenological match between species, and so to ecological relationship problems.
Climate change leads to unequal shifts in phenology across species2. To understand why, Thackeray and colleagues studied the sensitivity of phenological events to two climate variables in a wide range of species. Specifically, for each of their 10,003 phenological time series, they asked at what times of year species are sensitive to changes in temperature and precipitation, and how sensitive events in their phenological cycles are to these variables. Both questions are relevant, because climate has changed — and will continue to change — in a non-uniform manner over the course of the year. Some periods warm faster than others, and two species with equal temperature sensitivities but at different times may thus shift their phenological events at different rates.
On the whole, the authors found that species at different trophic levels (positions) of the food chain did not differ in the times of year at which they were sensitive to annual variations in temperature, but did vary in how sensitive they were. The phenology of species high up the food chain (the secondary consumers) is much less sensitive to temperature than is that of species at the base of the chain — the primary producers and primary consumers, which are twice as sensitive to temperature. Secondary consumers were also less sensitive to variations in precipitation. When combining the species sensitivities with climate scenarios, Thackeray et al. forecast that, by 2050, primary consumers will have shifted the timing of their phenological events twice as much as will species at other trophic levels.
But why do secondary consumer species respond more weakly to temperature compared with primary consumers or producers? After all, both predators and prey will use temperature as a cue to time their phenology.
The reason is that species at different trophic levels do not rely on exactly the same temperature cue. Different species respond to temperature at different times of year. Take, for instance, a winter moth egg of 2 millimetres in diameter, which has been lying dormant on a 40-metre-tall oak tree for months. It manages to hatch within days of its host's bud burst. Obviously, these two species will have radically different physiological mechanisms underlying their phenological events and will use different cues to determine the timing of these events. So, although the two sets of cues will be correlated, the cue used by the consumer (the moth in this case) will always be, to some extent, unreliable. Theoretical work3 shows that this imperfect cue reliability means consumers will evolve a less temperature-sensitive phenology than will the species at the trophic level they rely on.
As for understanding the consequences of the phenological mismatches caused by climate change, it is notable that almost all previous research has focused on simple two-species interactions of a predator and its prey, or a herbivore and its host plant. But despite the clear effects of the phenological mismatches on population numbers, such effects can be buffered by ecological mechanisms such as density dependence4 — whereby the success rate of an individual increases when the number of individuals declines.
A more overarching, but also more challenging, approach would be to turn away from these two-species interactions and to look at the effects on the entire food web5. How are the strengths of the links in a food web affected by phenological mismatches? What happens if the phenology of species at one trophic level shifts more than that of species at another? Does this lead to the loss of some links and the formation of others? Does this destabilize the web? Such analyses would be a stepping stone from studying the phenological shifts of species to understanding the effects of climate change on ecosystem function6.
But, to complicate matters further, species' climate sensitivity is not fixed. The phenological mismatches lead to selection on the timing of phenological events. And, because phenology is often heritable, this leads to genetic change in sensitivity7.
It will be a major challenge to combine genetic change with a food-web approach, and to include the necessary detailed climatological projections. But it is one that must be undertaken to forecast the effects of climate change, through phenological responses, on ecosystem function. What is clear is that long time series, such as the 10,003 analysed in the present paper, are essential for this. Therefore, professional and citizen scientists — who together made the 379,000 individual phenological observations on which these time series are based — need to be encouraged and facilitated to keep up their good work. The additional advantage is that observing phenological shifts in, sometimes literally, your own backyard drives the message of global climate change home.
About this article
Variations in the temperature sensitivity of spring leaf phenology from 1978 to 2014 in Mudanjiang, China
International Journal of Biometeorology (2017)