Advancing flowering is altering the dates of some cultural events

Soon I’ll be looking out for signs of bluebells. There’s an Iron Age fort near me that I’d like to visit to photograph that has a carpet of them when they’re in bloom. They’re also around when they’re not in bloom, but they look much less impressive then. It’s a matter of catching them at the right time. A recent tweet shows the right time is changing.

The headline is that this year was the earliest date for Kyoto cherry blossoms, but is that really a signal of climate change? Could it be a freak year? Thanks to another tweet, we can see the year is, in context, part of a long-term trend.  

The cherry blossom dates are certainly consistent with the climate warming. But there could be other confounding variables. Maybe this is local to Kyoto or a freak of cherry blossom. One of the attractions of studies like this is that it is something you could do yourself. If you can find records of an event that relies on flowering, then examining that should give you a proxy measurement for the date of flowering for specific plants. And those plants don’t have to be cherry trees.

The daffodils of Thriplow

One of my favourite climate papers is this one: Local-scale adaptation to climate change: the village flower festival. In it, Tim Sparks examines an unusual climate record held at Thriplow.

Thriplow is a small village in southern Cambridgeshire. It’s a typical English country village with a small shop, a community-owned pub and a lot of farmland. A quick look at Google Street View shows lots of open space and plenty of daffodils in the roadside verges. They’re proud of their daffodils in Thriplow. You can even see one in the header of the parish’s homepage. It’s these daffodils that cause concern.

In 1969 the people of Thriplow had a problem. Their church roof needed repairing. To raise the money, they held Thriplow Daffodil Weekend. They raised £206, which might not sound like a lot, but would be the equivalent of over £2900 in 2021. This was a lot of money for a small village. So they held the event again, and again, and again. But to hold the event, the village needed daffodils. If the daffodils were late, they’d have to delay the weekend. If the daffodils were early, they would bring the weekend forward. Tim Sparks realised that the dates of the festival were a record of daffodil phenology.

Phenology is the study of seasonal changes in plants and animals from year to year. In the case of plants, you could have their first leaf, or first bud, flowering, or harvest. Or you could record the latest dates. Phenology is something particularly suitable for citizen science. While we might not be experts on identifying specific plants or animals, we can recognise when we see a crocus appear or hear the first bee buzz of the year. By holding a flower festival, the people of Thriplow were keeping a record of daffodil phenology, though it has its problems.

It’s not a perfect record of phenology. One of the limits of a daffodil weekend is that it has to be on a weekend. So if the daffodils always peaked that the same time of year, then a weekend that starts April 19 this year will be April 18 next year, then April 17 and so on, at least till you got more than four days away from peak daffodil. April 15 would be further away from the peak on April 19 than April 22, so the weekend would leap back, then carry on advancing a day at a time.

Reality is messier. Leap years muck up the pattern, and the daffodils would come earlier or later from year to year. If you were to graph the flower festival dates, you’d expect shallow downward slopes punctuated by leaps back up or occasional dips down around the peak daffodil date. If the peak flowering day isn’t changing, the plots of the date would be scattered in a horizontal band on the graph. Here’s the graph. You’re looking for the black line.

Over the past forty years, there has been a gentle downward pressure. The result is that the Thriplow Daffodil Weekend is now, on average, about two weeks earlier than it was when it started.

The paper goes on to look at local temperature records and finds a correlation between the dates of the festival and the temperatures two years before the festival. Why this difference? That’s because the festival’s date is set 18 months in advance, adding a little more noise to the data. Nonetheless, the pattern over time is clear. The advancing daffodils are a warning.

I like this paper because it asks a simple question: if rising temperatures affect plants, can we see that in seasonal events? We can at Thriplow, but there are other similar events elsewhere. If they’ve been going on for a few decades, you could examine the dates and look for a pattern.

Another reason I like this paper is it’s an excellent example of how science works in the real world. A lot of people have an idea that scientists only keep ideas when they can’t disprove them. It’s a simplistic interpretation of Karl Popper’s idea of falsifiability if you want to get into the philosophy. In reality, scientists like to draw on multiple lines of inquiry. If you have multiple independent ways of investigating something and all come to similar conclusions, you can be more confident in your findings.

In this case, if you were going to investigate if climate change were real, you wouldn’t start in Thriplow. But, if climate change is real, you’d expect to see cultural events adapting to it. And this is what we see from independent evidence.

Flower tourism elsewhere

Flower tourism is likely to be affected by climate change as it becomes more apparent, making it a rich source of data. But you don’t need to attend the festivals to see the effect phenology has. Lu Wang and colleagues have another way of measuring peoples interaction with flowering phenology.

The team looked at the changing dates of people’s visits in Beijing to view ornamental plants. They picked five locations and set out to see if visits to these places peaked with plants’ flowering there. The easiest way to get the number of visitors would be to look at ticket sales. Unfortunately, that wasn’t possible, so instead, they turned to Sina Weibo.

Sina Weibo is the Chinese equivalent of Twitter. Like Twitter, when people visit somewhere and take a photo, they post it to Sina Weibo. Wang and colleagues sifted data from 2010 to 2014 to see when people visited sites. At Yuyuantan Park, people would go to see the cherry blossom. At Jingshan Park, they’d view the peonies. City Wall Relics park has flowering crab apples. They also included the Summer Palace and the Beijing Botanical Garden, which had a variety of plants.

What they found is that people visited sites when there was something to see. That’s not surprising, but it has a cultural impact. Somewhere like the botanical garden can have a stable peak season because there’ll always be something to see when you visit. You might see different flowers if you visit April 15 each year, but you will see something. For sites with a single star species, matters are different.

“For example, in 2010, the peony blossom festival of Jingshan Park started 13 d earlier than the FFD [first flowering date] of peony,” write Wang and colleagues. “Only a few people visited the park at that time, but more came after the peony FFD when the high season occurred. The difference between the festival period and high season influenced the number of visitors participating in the opening ceremony of the blossom festival and related cultural and recreational activities, which may further impact the commercial benefits of the events.”

Across the other side of the globe, Jones and colleagues have come to similar conclusions for the Canadian Tulip Festival. “The Festival is vulnerable to climate as well because climate influences the timing of tulip development and the relationship with the dates of the Festival,” they write. “In 1997, tulips in the NCR [National Capital Region] reached their peak late (at the end of the Festival), due largely to cold, wet conditions in April that delayed bud formation… By comparison, the 1998 and 1999 Festivals experienced a shortened period of bloom because above normal temperatures in April and May resulted in the peak bloom occurring early in the Festival, with many of the tulips wilting in the unseasonably warm temperatures before the Festival ended…”

More than an inconvenience

For human events, these shifts are a nuisance but not a show-stopper. At worst, a visit to Thriplow could be disappointing if the daffodils are out of sync with the festival. For the plants, however, things are more complicated.

While years are, on average, getting warmer, individual years might not be. Plants budding for spring could be the victim of a late frost. The earlier they bud, the more time there is for a late frost to happen. There is also the danger that plants may not be reacting to rising temperatures the same way as their pollinators.

Complaining about the advance of flower festivals might seem like an appeal to nostalgia, but it’s a very visible sign that we need to be prepared for change.