In order to estimate the size of the challenge of restoring numbers of Common Swifts Apus apus to ‘historical levels’, it would be useful to know how many pairs have actually been lost. The rate of the decline is rather straightforward to work out – since 1994, the BTO/JNCC/RSPB Breeding Bird Survey (BBS; www.bto.org/our-science/projects/breeding-bird-survey/bbs-publications/bbs-reports) has provided an annual population index – but working out the numbers of birds lost requires a knowledge of how many pairs were breeding in Britain when records first began. Then, we can use these figures and the BBS population index to calculate the number of birds lost in between. 

However, Common Swifts (hereafter referred to simply as Swifts) were not monitored before the inception of the BBS and the BTO recognises that ‘their monitoring is complicated by the difficulty of finding occupied nests, by the weather-dependent and sometimes extraordinary distances from the nest at which breeding adults may forage, and by the often-substantial midsummer influx of non-breeding individuals to the vicinity of breeding colonies’ (www.bto.org/understanding-birds/birdfacts/swift). Although breeding Swifts (usually fourth-calendar-year (4CY) or older) arrive back in the UK from African wintering grounds between late April and mid May, numbers are boosted from the latter half of May into June by non-breeding 3CY+ birds returning to look for a mate and nest site to use the following year. These birds may build a nest and roost in the new site but generally do not breed. Numbers again increase later in June with the return of 2CY birds, which interact with breeding birds but stay just a few weeks. 

Musgrove et al. (2013) published estimates of the populations of birds in Great Britain and the United Kingdom in 2009. The estimate for the Swift, a Green-listed species at the time, was 87,000 (63,000–111,000) pairs. The methodology used by the BBS, which requires surveyors to count all adult birds seen or heard on the transect, means that non-breeders were included within the estimate (since they cannot be distinguished in appearance from breeding birds), so the actual number of breeding pairs would be smaller. Additionally, the paper noted that the reliability of the estimate was considered ‘poor’ and that the approach used was likely to have underestimated breeding numbers.

bb-eye 001 swifts.jpg

73. Common Swifts Apus apus, Cromer, Norfolk, July 2019.

David Tipling

The same paper included estimates for two at-the-time (and still) Red-listed species: House Sparrow Passer domesticus, at 5.3 million pairs, and Common Starling Sturnus vulgaris, at 1.9 million pairs. Thus, there was one pair of Swifts for every 61 pairs of House Sparrows and for every 22 pairs of Starlings. 

Looking at it another way, with 62 million people in the country in 2009, there was one pair of Swifts for every 716 people, or the equivalent of one pair of Swifts for every 300 households. While Swifts are scarce in many towns and villages, in others, which provide suitable nesting opportunities, they are frequently encountered, and the occurrence ratios given above do not appear to match observations. 

In a stable population, the percentage of young birds entering the population must equal the annual mortality of older birds. Perrins (1971) estimated the annual mortality of adult Swifts to be 21%, corroborated by Finch et al. (2023). Thus, to keep the population stable, for every 100 adult birds in a given year, 21 of these are expected to die; thus, 21 young birds must be available to replace them. 

However, the Swift population in Britain is not stable; over the last ten years, BBS data have shown that Swifts have declined at an average rate of 5% per annum. Therefore, a population of 100 birds becomes 95 in the following year. If we assume that this decline is not due to an increase in adult mortality but rather a lack of recruitment of young birds to the population, then we can continue to assume that 79 of those 100 adults will return the following year. 

Based on the methodology used to count Swift populations and on knowledge that all age classes visit the breeding areas, these 79 returning birds will be a mix of ages from 3CY and older; so we can say that 16 2CY birds need to be recruited into the population as future breeders to bring the population up to the 95 birds. The percentage of 2CY within a population in any given year is thus, assuming all of the above variables are true, 17% of the population. The same assumptions give a calculated percentage of 13% 3CY birds in the same population.

Assuming that Swifts start breeding in their third summer (4CY), about a third of the overall population counted using the methods in Musgrove et al. (2013) will be birds that are too young to breed. Removing this non-breeding third from the population estimate of 87,000 pairs gives 58,000 pairs in 2009. Is the number of breeding Swifts in Britain really so low? 

The term ‘pairs’ is ambiguous, as the methodology used to count the population leads to the inclusion of non-breeding birds – both young birds, as described above, and unpaired adults. Therefore, we will talk in terms of birds from here on, taking the Musgrove et al. (2013) estimate of 87,000 pairs in 2009 to equal 174,000 individual Swifts.

East-coast movements
Almost every year, large movements of Swifts are reported from the east coast of England in late June or early July, most notably at sites along the English east coast, such as Spurn, Yorkshire (Roadhouse 2016), and Gibraltar Point, Lincolnshire (see below). These movements can be spectacular and, significantly in the context of this discussion, they occur before Swift chicks have fledged, and thus before adults have finished breeding. They typically occur when the wind is from the southwest with squally showers, and the birds are only ever noted passing along the coast, not coming in off the sea.

The timing of these movements coincides with a temporary reduction in activity around colonies in northern and central England. One might therefore expect the birds involved in these east-coast movements to include a mix of 2CY, 3CY and non-breeding adult birds. However, the majority caught by ringers on the east coast during such movements are 2CY birds (Roadhouse 2016). From this, we surmise that, if birds are not breeding until they’re in their 4CY, then 3CY birds and non-breeding adults (4CY or older) remain focused on establishing a future nest site and thus do not participate in these movements. 

We have concluded, given the southwesterly winds and the lack of observations of birds coming in off the sea, that the individuals involved are all birds from Britain that have been pushed eastwards to the coast by frontal weather systems. Our own observations suggest that breeding birds, even in bad weather, stay close to their nest sites.

Gibraltar Point June 2020
In 2020, across just two days in late June, over 65,000 Swifts (19,418 on 28th and 46,026 on 29th) were counted moving south over Gibraltar Point (Ward 2020). We presume they originated from the northern half of England, including much of the Midlands and believe it unlikely that birds from the west and south would be represented (though one could argue that the southwesterly winds may have brought some birds from farther south).

We also believe that it is unlikely that the observers were able to count every bird moving on those days, nor that all non-breeding Swifts in the Midlands and the north of England were pushed to the coast. Since they were observed across a two-day period and the count was higher on the second day, we believe there is little likelihood that any of the birds looped around and came past a second time. Thus, we consider that the number of birds moving along the east coast was a true representation – or even an underestimate – of the number of individuals present.

Assuming that most of these were 2CY birds, and continuing with the earlier-calculated figure that 17% of a given population as a whole are 2CY birds, then the birds passing Gibraltar Point in June 2020 would represent a population of 382,000 Swifts! 

Assuming that Swift populations are limited in distribution by the availability of nest sites rather than food, and since most Swift nests are in people’s houses, we have used human population density as a proxy for the distribution of breeding population density of Swifts in Britain. The number of people living north of a line from the Wash to the Wirral is about one-third of the British population. This, combined with the assumption that birds moving south along the east coast had come from farther north, would imply that birds were originating from one-third of the British population. It is therefore tempting to multiply the original number by three to derive the entire British population based on these figures; but, given some uncertainty as to whether southwesterly winds may also have pushed birds to the coast from elsewhere, we have decided to be conservative and apply a factor of two, instead of three. Our estimate of the number of Swifts in Britain in 2020 is therefore 764,000.

In contrast, BBS data from 2020, the year of the Gibraltar Point movements, showed the Swift population to be at 58% of the 2009 population, meaning that the ‘official’ estimate of the population in 2020 would have been just over 100,000 Swifts in Britain, including non-breeders – nearly an order of magnitude of discrepancy from our calculated figures. And this, we believe, may still underestimate the numbers a little.

Even with a more conservative approach, assuming that the Gibraltar Point counts included all 3CYs, and that mortality of 2CY birds is greater than that of older birds (say, for argument’s sake, double, at 42%), our calculations still give a large discrepancy of nearly a factor of four. If we based the estimate on just the one-day count of 46,000 Swifts on 29th June, the discrepancy would drop further to a factor of three – but this still represents a massive underestimate by the official statistics. 

The scale of the problem
How many Swifts have we lost?
While we may, on the face of it, be pleased that we may have far more Swifts in Britain than we thought, this means we have grossly underestimated the numbers lost and the size of the problem of restoring their numbers. This number lost is important, because if we are to reverse the decline of Swifts and get back to at least the numbers in 1994, when BBS started, we need to know the scale of the problem. 

Using BBS index values (fig. 1), we can estimate how many Swifts there were in 1994, and their numbers in any year since then. Based on the 2009 anchor point of 87,000 ‘pairs’ (174,000 Swifts) given in Musgrove et al. (2013), Britain had lost nearly 180,000 Swifts by 2022 according to official statistics. However, using the numbers we have calculated based on the movements through Gibraltar Point, that number could be as high as 1.37 million Swifts lost! And, of course, this does not include any losses prior to 1994, before BBS started to monitor Swifts. 

Screenshot 2024-02-29 at 16.06.20.png

Fig. 1. Index values for Common Swift Apus apus based on BTO/JNCC/RSPB Breeding Bird Survey data, 1994–2022, annotated with population estimates based on data from 2009 in Musgrove et al. (2013) and our own population estimates based on counts at Gibraltar Point, Lincolnshire, in June 2020.

Can we be sure that nest-site loss is a significant factor?
Before we assert that the solution is the provision of more nestboxes, we need to be sure that loss of nest sites is the main cause of the problem. There are other potential causes: fewer insects, increasing losses on migration and in the winter quarters, a rise in predators and poorer summer weather in the breeding season (Finch et al. 2023).

BB-eye 002 swifts.jpg

74. Adult Common Swift leaving a Swift nestbox, Norfolk, July 2021.

David Tipling

However, if any of these other causes are significant, one would expect established colonies to also reduce in size, but there are few, if any, colonies where this happens – with the exception of large colonies that attract the attention of a predator, such as a Eurasian Sparrowhawk Accipiter nisus, Common Kestrel Falco tinnunculus or Magpie Pica pica. However, in the years after the predator moves on or loses interest, the colony can recover (https://oumnh.ox.ac.uk/learn-swifts-diary). Fortunately, persistent predator attacks on large Swift colonies are relatively rare.

On the other hand, unless done with care, the damage caused by building modifications and ‘improvements’ results in the permanent removal of whole colonies, and thus a reduction in the number of colonies, not a reduction in average colony size.

Of course, occasionally Swifts will not return to an established nest site; this is a chance event because some Swifts die between breeding seasons. An isolated nest site may take time to be occupied again, but small groups of nest sites will attract replacement birds – one of the advantages of the Swift’s social behaviour.

Undisturbed colonies, not destroyed by roof renovations, the fitting of PVC soffits or external wall insulation, can maintain their numbers for decades. Of course, it could be that colonies suffering losses are being replenished by birds displaced from colonies that are destroyed, thus hiding the effects of other causes but, in our opinion, the loss of nest sites is a significant, if not the most significant, factor in the decline of Swifts in Britain.

Are Swift nestboxes effective?
In 2022, the official estimate of the Swift breeding population (BBS figures with the removal of non-breeders from the figures) works out at 30,000 breeding pairs. In the same year, there were 67 million people in the UK: one pair of breeding Swifts to 2,200 people; or, more likely, if our above calculations are correct, one pair to 289 people.

Landbeach, near Cambridge, Cambridgeshire, has a population of 800 people. It should, based on the figure of one pair of Swifts to 289 people, and assuming a uniform spread of nesting birds, have two to three breeding pairs of Swifts – which is about what there were 20 years ago. Instead, in 2022, there were over 70 pairs of Swifts breeding in nestboxes in Landbeach. There are many other examples of nestbox schemes raising the local Swift breeding population to levels far higher than the national average density (Newell & Willis 2023). 

Where mitigation boxes are provided in the near vicinity of nest sites that are lost, they are often quickly occupied by Swifts. A prime example is at Fulbourn, also near Cambridge, where a new estate containing 276 Swift boxes now houses over 100 pairs of Swifts, replacing 70 nest sites lost when 1960s flat-roofed housing was demolished.

In addition, where nestboxes have been installed and new colonies established, increasing the number of boxes can lead to a growth in the size of the colony. This is illustrated by the expansion of a colony in TC’s house in Rutland, where the addition of new nestboxes has resulted in an increase in the number of breeding pairs (fig. 2).

Screenshot 2024-02-29 at 16.08.38.png

Fig. 2. Growth of a Swift nestbox colony at North Luffenham, Rutland.

Conclusion
Many tens of thousands of nestboxes for Swifts have already been installed but, thus far, the inexorable decline of Swifts, by about 5% per annum, continues unabated. Trying to recover 180,000 lost Swifts is enough of a challenge, let alone 1.37 million. Could it be that the lack of any detectable effect on BBS index trends is that nestboxing efforts so far are a mere drop in the ocean when the actual target is so high?

At a time when the provision of ‘swift bricks’ is the subject of parliamentary debate following a government petition that attracted over 100,000 signatures (https://petition.parliament.uk/petitions/626737), and given the opportunity presented by major government house-building programmes, it is important that the data on which conservation decisions are based are correct.

BB-eye 003 swifts.jpg

75. Two Common Swift chicks in an ‘S Brick’ Swift brick, Oxfordshire, July 2023.

Amanda Miller

Our conclusion is that it is likely that there are far more Swifts in the UK than officially recognised, with the corollary that we have lost far, far more than we realised – perhaps as many as 1.37 million since 1994. 

Since a major driver of these losses is the loss of nest sites, there is an even greater imperative for the installation of large numbers of nestboxes on existing housing and Swift bricks in new-build housing.

Acknowledgments
We would like to thank Chris Hewson and Rob Robinson, BTO, for comments on an earlier draft.

References
Finch, T., Bell, J. R., Robinson, R. A., & Peach, W. J. 2023. Demography of Common Swifts (Apus apus) breeding in the UK associated with local weather but not aphid biomass. Ibis 165: 420–435.

Musgrove A., et al. 2013. Population estimates of birds in Great Britain and the United Kingdom. Brit. Birds 106: 64–100.

Newell, D., & Willis, J. 2023. Are nest boxes for Swifts a good idea? Brit. Wildl. 34: 585–591.

Perrins, C. 1971. Age of first breeding and adult survival rate in the Swift. Bird Study 18: 61–70.

Roadhouse, A. 2016. The Birds of Spurn. Spurn Bird Observatory, Kilnsea.

Ward, B. 2020. A British record-day for Common Swift passage www.birdguides.com/articles/general-birding/a-british-record-day-for-common-swift-passage

Dick Newell , Old Beach Farm, 91 Green End, Landbeach, Cambridge CB25 9FD; e-mail [email protected]

Tim Collins, 6 Sycamore Road, North Luffenham, Oakham, Rutland LE15 8JL; e-mail [email protected]

Jonathan Pomroy, 4 Pottergate, Gilling East, North Yorkshire YO62 4JJ; e-mail [email protected]

Editorial comment This BB eye offers substantial food for thought on the difficult task of calculating populations. A lot of assumptions have been required to come up with these revised population figures for Swifts in Britain; some of these assumptions are nothing short of a necessity, since data is severely lacking in some areas and unlikely to become available any time soon. However, some areas of research, such as the continued deployment of data loggers onto Swifts, may result in data that challenges some of the assumptions made about movements along the east coast of England and the proportion of non-breeding birds at colonies. Views will no doubt differ on how correct many of the assumptions made in this BB eye are – but challenging the status quo when it comes to long-accepted population data, as the authors have done here, will undoubtedly help to provoke thought and drive forward further research. 

Volume: 
Issue 3
Start Page: 
114
Authors: 
Dick Newell, Tim Collins and Jonathan Pomroy
Display Image: 
Article Series: 
BB eye
Subject: 

Stay at the forefront of British birding by taking out a subscription to British Birds.

Subscribe Now