Was it really such an exceptional year?
Our lives this year have been dominated by graphs and charts. And, worryingly, many of the lines on these graphs are now heading in the wrong direction again. Infection rates are rising, the ‘rule of six’ prevails (except where it doesn’t), and local lockdowns are starting to proliferate—though, thankfully, not in Somerset yet. One secondary school in Taunton has had two year-group ‘bubbles’ self-isolating within 10 days of the start of the new school year. That’s more than 400 pupils, apparently. The test and trace system seems to be creaking under the strain. This doesn’t bode well, and right now it’s hard to see where it will all end.
We are in for a tough winter. From today’s vantage point, perched on the autumn equinox, the notion that any time soon we might be able to meet in a room together to talk about plants seems fanciful. Let’s hope that by the time of the next equinox we’ll be able to meet again. Presumably some of the field meetings cancelled this year could be rolled forward to 2021: maybe we’ll hold that early-April meeting in Orchard Wood after all, just twelve months later than anticipated.
But what about first flowerings? Has it really been such an exceptional year? This isn’t the easiest question to answer, and it may have to involve a few charts or graphs, which I’m sorry about because we’re probably all sick to the back teeth of such things. At least these graphs have nothing, absolutely nothing, to do with viral pandemics. So, let’s try to think of them as light relief, if that’s possible?
Walter Watson. Let’s start with Walter Watson who, as you know, in the early decades of the last century kept detailed records of first flowering dates (FFDs). His paper, published in 1949, included a 19-page Table of average FFDs and ‘flowering periods’ for no less than 843 species. A total that puts the rest of us to shame, frankly. Watson lived in Taunton, and much of the fieldwork he did was in the south and west of the county. For the bulk of species his FFDs were based on at least 10 years’ records, which we think were made in the 1920s and early 1930s[1]. The man’s energy was extraordinary. Don’t forget that during this time he was teaching Biology at Taunton School, while also actively involved in the Somerset Archaeological and Natural History Society, as well as being an internationally-renowned lichenologist, a more-than-competent bryologist and mycologist, and (as it happens) a man with an interest in plant galls. As if that weren’t enough, he was an avid watcher of cricket. For most of his adult life he was a member of Somerset County Cricket Club; at the time of his death, in 1960, he was the Club’s longest-serving member. He proved, if proof were needed, that one can combine natural history with a love of cricket. A relief to us all, I’m sure you’ll agree.
But that’s not the point. The point, really, is that Walter Watson bequeathed to his successors this amazing baseline of information, allowing us to compare FFDs today with those of roughly a century ago. Of his 843 species, this year we’ve recorded first dates for 526 of them. Up until mid-March the FFDs were mine, of course, but from the start of lockdown they could have been anyone’s. Plotting our earliest FFD for each species in 2020 against Watson’s FFDs gives us the graph in Fig. 1. There are 526 dots, each one a species.
Fig. 1 First flowering dates (FFDs) for 529 species in 2020, plotted against ‘average first flowering times’ given by Watson. Dates are shown as day no. (1 January = day 1). The diagonal line marks the line along which the data-points would lie if 2020 FFDs were identical to Watson’s; above the line the 2020 date is later than Watson’s date, below the line is earlier
You can see immediately that the data-points lie mainly below the line, indicating that our dates were, on the whole, earlier than Watson’s. Indeed, only 19 species (3.6%) had later dates than Watson’s, and some of these were probably only ‘late’ because during lockdown the places in which we might find them were hard for us to visit. If you don’t like graphs, let’s summarise these findings with a single statistic: overall, our FFDs were, on average, 31 days earlier than the dates listed in Watson’s big Table.
Really? Watson, I think, would be spinning in his grave. Mind you, his dates were average FFDs over more than a decade, so obviously in some years his dates would have been earlier than average while in others they’d have been later[2]. But a whole month earlier? He’d doubtless protest that his dates were based on one pair of eyes operating across a fairly restricted geographical area, whereas ours had the benefit of many pairs of eyes scattered across the entire county. And, to be fair, he’d be right to insist that the latter would produce an earlier crop of dates than the former. In essence, then, he’d say that we had an unfair advantage.
So, maybe a fairer comparison would be to test our dates against his by restricting the analysis to just one person’s FFDs. So let’s try that. We’ll take mine, if that’s okay with everyone, since not only does that give us the largest pool of species to work with, it also comprises a set of records drawn from a geographical area that is roughly similar to Watson’s. The first thing to be said, though, is that as soon as you use only one person’s records you lose an awful lot of information, as the pool of species becomes substantially reduced—despite the fact that the individual concerned was being spurred on each week by everyone else’s recording! I recorded FFDs for 406 species, just over three-quarters of the number recorded by the group as a whole. The results for these are shown in in the graph in Fig. 2.
Fig. 2 FFDs for 406 species recorded by SJL in 2020, plotted against ‘average first flowering times’ given by Watson. Dates are shown as day no. (1 January = day 1). The diagonal line marks the line along which the data-points would lie if 2020 FFDs were identical to Watson’s; above the line the 2020 date is later than Watson’s date, below the line is earlier
I’m not altogether happy with my own records for this spring. In late March and April, for instance, there were places I couldn’t get to, either because they were beyond the limits of daily permitted exercise or because Ben had gone off with the car. So some dates may actually be a little later than they would have been had we not been in lockdown. But bear in mind that Watson, too, was living through a pandemic—in his case the Spanish ‘flu—and he’d probably be keen to stress that his own mobility would have been pretty limited[3], and his records were made mostly while he was still working full-time, while I’ve had the advantage of being retired and with time on my hands. We could have argued the toss late into the night, him and me, but none of that would alter the fact that my own FFDs for 2020 are still strikingly early—below the line—in comparison with his. Again, for those who dislike graphs, my dates for these 406 species were, on average, 28 days earlier than Watson’s.
Interestingly, if we do the calculation again, but this time using the group’s earliest FFDs for these species rather than my own, the figure is 34 days earlier. Which illustrates well the added value of having many pairs of eyes, the group’s FFDs being earlier, on average, by almost a week compared with my own dates. Actually, the difference would have been greater than this, since first-flowerers before lockdown (i.e. between January and mid-March) were only recorded by me, even though others in the group would doubtless have conjured up earlier records had they been involved from the outset.
So, okay: our FFDs in 2020 were exceptionally early in comparison with the sorts of dates that Watson was getting about a century ago. But, you might say, this is barely newsworthy: climate change is happening, the evidence is all around us, and frankly it would have been surprising had our dates not been much earlier than Watson’s. And yet many of us still felt—at least during lockdown—that the spring of 2020 was unusual, even when assessed against today’s ‘new normal’. Was it, perhaps, that the season’s gallop, its speed of advancement, accelerated just at the moment our own movements were suddenly curtailed? In which case, could our perceptionof ‘earliness’ really just have been an artefact of lockdown? Was it simply that we were keeping a closer eye on spring that we would normally do, and therefore noticing things that we might otherwise have overlooked?
Well, the weather this spring was certainly unusual. It was, after all, one of the driest and sunniest springs on record. In fact it was the sunniest since records began, in 1929. Also, it followed an exceptionally mild and relatively frost-free winter. Nationally, it was the sixth-warmest January since 1884, and the warmest in the 13 years that I’ve been recording FFDs, i.e. since 2008. February was also relatively mild, being the second-warmest during that same period (only 2019 was warmer). March wasn’t especially mild, but this was followed by the second-warmest April and May since 2008. Summer was marked by prolonged periods of mainly dull weather; yet while June and July were relatively cool by today’s standards, August 2020 was, nationally, the warmest on record. In SW England, seven of the nine months from December to August had mean daily temperatures at least 1°C warmer than the long term (1961-1990) average.
Given the weather, then, one might anticipate that spring 2020 would indeed have been ‘early’, even in comparison with other recent springs. We can test this by looking at the FFDs of species in 2020 against their 2008-2017 ‘decadal average’ FFDs. This reduces the pool of species still further, since there are only 339 species for which we have a decade’s-worth of FFDs. In 2020 I failed to record FFDs for ten of these, but results for the other 329 are shown in the graph in Fig. 3. Note that the x-axis has changed: it now shows my own 2008-17 average FFDs, rather than Watson’s from back in the day.
Fig. 3 FFDs for 329 species recorded by SJL in 2020, plotted against 2008-17 decadal average FFDs recorded by same observer. Dates are shown as day no. (1 January = day 1). The diagonal line marks the line along which the data-points would lie if 2020 FFDs were identical to the decadal average; above the line the 2020 date is later than the decadal average, below the line is earlier
Again, most data-points sit well below the line, indicating that FFDs in 2020 were for most species earlier than their decadal average. Look closely, though, and you’ll see a little group of dots sitting well above the line around day 100 on the y-axis; these ‘late’ FFDs were at the end of March/beginning of April when lockdown brought me to a temporary standstill and some usual haunts like Thurlbear Quarrylands and Orchard Wood suddenly became off limits. Nevertheless, taking all species combined, FFDs in 2020 were, on average, 15 days earlier than the decadal average. More than that, though, it turns out that they were 4 days earlier than even the earliest set of dates during that decade, in 2014.
In summary, then: FFDs in 2020 were 28-31 days earlier (on average) than in Watson’s day, 15 days earlier (on average) than in the decade 2008-17, and 4 days earlier (on average) than even the earliest spring of that decade. So, to answer the original question: yes, it has been an exceptional year.
* * *
Today, 21st September, has been sunny “from the word GO!”, as weather forecasters like to put it. A cloudless sky to mirror perfectly the weather we were experiencing in Week 1 of lockdown. A lot warmer today though. The spring equinox was marked by a ground frost and a daytime maximum temperature of 14°C. Today’s maximum is 25°C. We’ve been sitting in the garden enjoying Red Admirals, Vanessa atalanta, and Speckled Woods, Pararge aegeria,and a Hummingbird Hawk-moth, Macroglossum stellatarum, has been busily working its way across the patch of ground we optimistically call ‘the flower border’.
Several of you have been reporting second first flowerings. An Indian summer can sometimes bring with it a nod to spring, and so it’s with a sense of déjà vu that Gill, Helena, and Margaret—and maybe others—have been noting flowers on trees of Holly, Ilex aquifolium, otherwise laden with berries. Ann Fells and I have noted Wayfaring-tree, Viburnum lantana, blooming again in the Taunton area—Ann has seen flowering Spindle, Euonymus europaeus, too—while Hester’s Greater Chickweed, Stellaria neglecta, was soon followed by Grass-leaved Vetchling, Lanthyus nissolia, on a grassy bank in Longrun Meadow, as well as Goat’s-beard, Tragopogon pratensis and its startlingly beautiful hybrid with Salsify, T. porrifolius, T. x mirabilis, which hadn’t been seen flowering since the end of May. There have also been records of Dogwood, Cornus sanguineus, and Apple, Malus domestica,flowering a second time, while one or two of you have mentioned unseasonal sightings of Cowslip, Primula veris, and Primrose, P. vulgaris. Margaret claimed possibly the unlikeliest record of the last few weeks when she saw several newly-flowering Yellow Iris, Iris pseudacorus, in a ditch near Chew Valley Lake on 15th September.Blooming ridiculous, if you’ll excuse the pun!
This morning at Thurlbear I’ve had a wide variety of still-flowering plants like Wild Basil, Clinopodium vulgare, Hawkweed Oxtongue, Picris hieracioides, Wild Thyme, Thymus drucei, Musk Mallow, Malva moschata, Agrimony, Agrimonia eupatoria, and Traveller’s-joy, Clematis vitalba. We are used to the flowering periods of such species continuing long into the autumn, but these records of spring/early-summer species are much more surprising. It’s like we’re viewing the world through a distorting mirror, the seasons temporarily jumbled in an early-autumn heat haze. We claimed that spring had finally ended with the first blossom of Ivy, Hedera helix, but now there’s this echo, a faint reprise of better days—like a second spring, of sorts. But it won’t last, it never does, and as the temperatures tumble so these flowers will begin to falter, like little lights going out. The Ivy, though, will see us through. You can always rely on the Ivy.
21st September
[1] We don’t know precisely which years, unfortunately.
[2] Unfortunately, the raw data from which his average FFDs were calculated cannot be found, so we have no idea the range of dates he recorded.
[3] There is much we don’t know about Walter Watson. Did he own a car, for instance? Or did he have to rely on public transport to get around the county? We’ve got no idea…