Phenology in a warming world

azandis@gmail.com

I’m thrilled to announce that the first of my dissertation chapters has just been published in Ecography.

Update (Nov. 2020): And, I’m especially thrilled that our piece will be the cover article for the journal, featuring a pair of breeding wood frogs from our population! My hands nearly froze trying to get this underwater shot, so I’m glad it was worth the effort!

Over the past 20+ years, our lab has been monitoring over 50 populations of wood frogs at Yale Myers Forest. Each year in early spring, we listen for the duck-like clucks of the male frogs which means that they have emerged from under the snow and moved into the breeding ponds. Shortly afterward, we head out into the freezing ponds to count the egg masses as a way to monitor population density over time.

Here, I am wading into one of the ponds to count egg masses. Wood frogs are remarkable in the cold temperatures that they can function.

 

In this study, we looked at how the oviposition date (the day on which frogs deposited eggs) has changed over time. As climates warm, we usually expect for the timing life-history events (like oviposition, emergence from hibernation, flowering time, etc.) called ‘phenology’ to advance in the year as winters get shorter. That’s just what most species do. And the trend of advancing phenology is strongest for amphibians.

This slide from my presentation at the World Herpetological Congress shows that, in three major metaanalyses, amphibians show some of the strongest advances in phenology compared to other species.

Given that annual temperatures at our field site have increased by almost 0.6 C in the past two decades, we expected frogs to breed and lay eggs earlier. If our frogs were like other amphibians, we might expect oviposition to come around 6 days earlier.

Surprisingly, we found the opposite. Our frogs seem to be breeding 3 days LATER.

To figure out what might be going on with our frogs, we decided to look more closely at climate across the season, not just annual averages. It turns out that most of the increase in annual temperatures are felt later in the summer, but relatively less when frogs are breeding. Snowpack, on the other hand, is actually accumulating later and lasting longer. In the figure (Fig. 3 from the paper) below, you can see these trends. On the left are the comparisons between temperature, precipitation, and snowpack between 1980 and 2018. On the right, we plot only the difference in trends over time. At the top-right, we plot the oviposition dates to show how seasonal changes in climate line up with frog breeding.

Figure 3 from the paper. Seasonal trends in daily temperature (a), precipitation (square root scale) (b), and snow water equivalent (c) from 1980 (blue) to 2018 (red) as predicted by generalized additive model with interaction between Year and penalized spline smooth on day-of-year with 95% confidence intervals. Points represent daily values (N = 13,869 for all models). Annual mean oviposition dates (2000-2019) (d) in comparison to relative, seasonal change in temperature (e), precipitation (f), and snow water equivalent (SWE) (g) between 1980 and 2018. Seasonal change is the difference in daily values fit by generalized additive models for between 1980 and 2018. All differences are scaled to the standard deviation between annual averages for each variable in order to compare relative magnitude of change that coincides with the oviposition window (dotted lines). Dark bands indicate significant difference between 95% confidence intervals. Light bands indicate total difference. All meteorological observations from Daymet data between 1980 and 2018.

 

We also looked at how the timing of oviposition correlated with climate across the season. We found that breeding occurs later when there is more snow at the beginning of the breeding window. Also colder temperatures just before breeding correlate with delayed oviposition (which makes sense if colder temps mean more snow and less melting).

So, we think that frogs may be kind of stuck. Persistent snowpack might be keeping them from breeding earlier. But at the same time, warmer summer temperatures might be drying up their ponds faster. If so, this could be a big problem for tadpoles that need to maximize their time for development. The figure below shows that frogs tend to breed earlier when winter and early spring air temperature are high. As we’d expect, more snowpack correlates with later breeding. High precipitation during the spring delays breeding (probably because it is falling as snow).

Figure S2 from the paper’s supporting information. The correlation between 10-, 20-, 30-, and 40-day rolling averages of daily mean temperature (b), precipitation (c), and snow water equivalent (d) between 2000 and 2018 with oviposition timing (annual averages 2000-2019, 3-day bin width)(a). Dotted lines indicate 95% confidence interval (+/- 0.45) for Pearson’s correlation for n = 20 pairs and 18 degrees of freedom. Light grey bands indicate non-overlapping windows of greatest correlation.

 

Twenty years is a long time to be collecting ecological data, but it is a pretty short window into the evolutionary history of wood frogs. And, we don’t know how long snow and temperatures may have been working against these frogs. So, as a final piece of our analysis, we used a machine learning technique called a random forest to predict oviposition dates backwards in time an additional 20 years. It doesn’t seem like much has changed over the past half-century or so. In one way, that could be good news in that at least things don’t seem to be getting any worse.

The big question is, how will frogs cope with these climate changes? If tadpoles are faced with an ever-shrinking window of time to develop into frogs, will they be able to keep up? Or, will they lose the race and end up as tadpole-shaped raisins in our ponds?

I won’t give away any spoilers, but I’m looking at our long-term larval datasets to ask that question next.

This male wood frog is learning why it doesn’t pay to get to the breeding ponds too early. His pond is still frozen and he is waiting for the ice to, literally, thaw out from under him.