Beijing Swift project preliminary results

Action for Swifts is delighted to have been involved in the Beijing Swift project where we attached 31 geolocators to Swifts at the Summer Palace in 2014 (see story). On 24th May 2015 we retrapped 13 of these birds, and downloaded all of the data successfully.

by Dick

Prior to this project, the journey made by these iconic birds - synonymous with Beijing since 1417 when they made their nests in the original city gatehouses, was a matter of speculation. We now know that they arrive in Beijing in April and, after breeding, begin their long journey to Africa in late July, taking a route that first leads them west-northwest into Mongolia, from where they pass north of the Tianshan mountains, then south through Iran and central Arabia into tropical Africa, before spending 3 months of the winter in Namibia and the Western Cape.

They begin the return journey in February, retracing a similar route on the way back, arriving in Beijing in mid-April, a journey that sees them cross about 20 borders. 

For the full story, read the Press release

Also see piece on Birding Beijing
The Guardian
Xinhua article - largest media agency in China
Birdguides webzine
British Birds
Birding Frontiers

One track of 13 geolocators recovered (click map to enlarge)

Lyndon Kearsley releasing a Swift with geolocator Photo Zhang Weimin

Terry Townshend, right, releasing a Swift Photo Zhang Weimin

#beijing

The winter movements of a pair of Swifts (part 3)

Contributed by Lyndon Kearsley

Getting into position. A further look at the winter timing of a pair of Belgian swifts in 2012. This is the third instalment of our story about where the Swifts were in December and February.

March is a difficult month for tracking birds using light level dataloggers. From about 15 days either side of the spring equinox on the 20th, the day length is about the same across all latitudes and so no north / south fix is possible. However the timing of local midday is not affected, so longitude (how far east or west) can be calculated. 

In the following map I've plotted the locations during March (shown in closed or open black dots) by using the longitude combined with a fabricated latitude derived by simply dividing the first and last fix before and after the equinox by the number of days. The February and April positions are also included to show how this ties in before and after the equinox

You can see that the female leaves East Africa crossing rapidly to a position which coincides with her December location on the lower Congo River. She remained there at about that longitude for the last two weeks of the month and then departed towards West Africa at the beginning of April.

The male moved very little in this period probably remaining above the lower Congo River wetlands in the Democratic Republic of Congo.

                   Red dots are female positions in February and in April

                   Black dots are female positions in March

                   Blue dots are male positions in February and in April

                   Open dots are male positions in March
                  [Click map to enlarge]

The April tracks show how this story continues and marries up with this theory. The male indeed remains in the lower Congo until about the 20th of April when he too makes his move out to West Africa (Sierra Leone).

The female had already been in West Africa for quite a while by that time having crossed the Gulf of Guinea in early April, and her positions show her to have been first in Ivory coast, then Liberia, Sierra Leone and Guinea. Bear in mind that geolocator fixes are at best approximate and the tracks shown are smoothed 3 day rolling averages. Never the less at this scale one gets a clear idea of the birds timing and strategy.

On 29th and 30th April both birds are shown to be somewhere in Sierra Leone and closer to each other than they have been all winter, nicely placed for the dash back to Belgium.

Recently a  paper discussing a statistical track estimation solution for the latitude problem during both equinoxes was published by Wahlstrom et al (2012). 

Wahlstrom, N., Gustafsson, F., & Akesson, S. (2012, July). A voyage to Africa by Mr Swift. In Information Fusion (FUSION), 2012 15th International Conference on (pp. 808-815). IEEE.

The winter movements of a pair of Swifts (part 2)

This is an update to our story about where the Swifts were in December and January, here.

Contributed by Lyndon Kearsley

It is February 2013 now and some common swifts have been seen in Malta and Israel over the last two weeks and a number of House Martins are in Lisbon (and Malta) along with lots of Swallows. That is normal timing for Southern and Mediterranean Europe, but what are "our" swifts doing. Maybe, they are in habitat like this:

Orange River region in Northern Cape Province / S Namibia in 2011. Photo F. Ambrogio.

Not possessing a crystal ball we can't of course answer that, but here is what the Belgian pair discussed in my last blog entry were doing in February 2012 last year.

February movements (click to enlarge)

The female moved around an area between Lake Malawi and the headwaters of the River Congo in south east Congo Republic (RDC) in February.  This range is about 1000 km in a SE to NW direction. It is very consistent with a second Belgian female who covered a similar range size that month, but a little to the NW in Central Congo.

The male partner which as you remember spent January in a rather small part of South Africa near Johannesburg moved slowly north into southern Zimbabwe and then headed WSW swinging across the south of Botswana in to western Namibia by mid month. It remained thereabouts for 12 days before heading quickly due north into the middle of the Congo Basin. As the crow flies about 4000 km, and always keeping between 1000 and 1500 km south and west of his mate which it neatly bypassed by month end.

Kalahari rain clouds. Photo Oompie

Male track superimposed on SABAP density map
(Click to enlarge)

You'll probably be thinking what on earth made it head over to Namibia? The east coast southern rains form a separate weather system and move due west across the Kalahari through January and February, greening it and triggering insect eruptions, particularly of various termite species. When the track is plotted on the South African Bird Atlas (SABAP) smoothed map of Common Swift observations one sees some correlation, although I have not had access to the dates of the field observations.

Please bear in mind that the geolocator data is only precise to within about 150 km and that Common Swifts are thought to spend the whole of their winter period flying. The dawn and dusk fixes give then only approximate midday and midnight averages. Those plotted here are smoothed averages taking into consideration the movements the day before and day after. The swifts themselves fly a huge multiple of these distances as they dart around going about their business. Since all geolocators are light level archival data-loggers, a constantly flying aerial platform (a swift) is quite ideal, even if not stationary, and the data surprisingly good. The calculated plots when considered on an African scale certainly reveal broad distribution, strategies and timing adequately. 

The winter movements of a pair of Swifts

It is New Year and our swifts are very very far away. It’s the middle of winter here in the north and of course mid-summer in Southern Africa. Swifts are usually faithful to both nest site and partner, but do they stay together outside the short northern breeding season? Here’s one couple's answer.

Contributed by Lyndon Kearsley

These are maps of one of my geolocator pairs showing movements in December 2011 and January 2012. They nested in a nest box above a busy school playground in Hechtel, Limburg Province, Belgium. The coloured dots are smoothed average 24 hour locations, but remember that swifts keep flying so the accuracy is not great and the light level data weather dependent. Hopefully they are doing well and quite likely in the same regions at the moment. In 2012 they both received new geolocators and with luck we’ll know if they are equally faithful to their preferred wintering grounds come summer.

Click map to enlarge

December 2011

The female is the red track and she spent September, October and November above the Republic of Congo (Congo Brazzaville) and Gabon, leaving there on 6th December to head directly ESE to the lower Zambezi river area in East Africa. This region between Lake Malawi and the Zambezi Delta is an important wetland area with flood plains and coastal mangrove. Two large dam projects higher up-river in Zambia have reduced the impact of the summer rains but it is still a draw for west European Swifts.

The male in blue left west Congo much earlier and was, by this time, already in the south and spent the first half of the month in south east Botswana before heading into South Africa (RSA), close to the Kruger National Park, for Christmas.

As you can see they did not spend their winter together.

Click map to enlarge

January 2012

The lady goes “fly-about”. First up to the SE Tanzanian coast on 5th January where a second Belgian female breeder happened to be wintering, so an interesting area to keep an eye on. Then she flew inland towards the Lower Zambesi National Park in Zambia around 12th January, and finally south towards the Mozambique capital Maputo on 26th January returning to central Mozambique at month’s end.

The male stayed in South Africa (RSA) between Johannesburg and Swaziland in the coastal highlands and remained restricted to a fairly small area in comparison to the female.

In early February they gradually moved north, getting into position for the dash north in early May back to Belgium.

More about that later.

Swift ringing recoveries and geolocator tracks compared

It is fascinating how 50 years of ringing recoveries had already given us a clue as to where our Swifts spend the winter, but it is also surprising that some important destinations were never detected.

Contributed by Lyndon Kearsley

The BTO recently posted a piece on their Demog Blog about more ringing recovery maps on their Online Ringing Reports. 

When one checks Common Swift, the selection shows totals per country, a general map and a listing of the most interesting movements with full details.

If we look at the country totals for British ringed swifts that have been found abroad and filter this to show either locations due south (France and Iberia) or in Africa, one gets the following:

BTO foreign recoveries of Swifts Apus apus

I put these more or less in a N:S order and split into Europe, N Africa, Central Africa or East and South Africa. There were no recoveries in West Africa at all.

Unfortunately the BTO does not list full details for all recoveries and not for counties with more than a handful of them. The result is a lack of recovery dates for clumps of locations in for instance France, Spain, Morocco, DRC (Congo) and Malawi.

For recoveries where ringing and finding details are listed, I was able to annotate the BTO map as follows:

Foreign locations of birds ringed or recovered in Britain & Ireland

     Purple: ringed in Great Britain & Ireland, recovered here

     Yellow: ringed here, recovered in Great Britain & Ireland

You will note that the finding dates (often many years after initial ringing) are very consistent with the periods that Dick's Landbeach geolocator bird was in that part of Africa. Of note too is that Algerian and Tunisian recoveries are all in May (Spring) and that, although there is a string of recoveries in nearby Malawi, Mozambique only has one recovery where one would expect a great many more. The same goes for West Africa; not one recovery in that part of the continent although we now know that it is so important for a large part of the European population particularly in Spring.

Since all the points on this map are the ringing or finding locations of single individuals recovered over a span of 50 years, it is quite sobering that one tracked swift from Cambridge can join the dots. On the other hand it's just that connection that increases the value of the ringing recoveries. Conversely the ringing recoveries certainly help to confirm the quality of the geolocator results.

What an exciting time we live in. Hopefully we can use this new knowledge to lever a better time for our Swift friends too

Results from the Landbeach Geolocator Swift

by Dick Newell

We previously reported on the behaviour of one of the Landbeach Swifts, Swift 396, in its nest-box on the day that it was fitted with a geolocator, 21st July 2011.  We recorded it again on its last day, 28th July 2011, having successfully raised 2 chicks. We were therefore delighted when it reappeared in the box on 12th May 2012. It settled down to breed and we retrapped it to retrieve the geolocator on 21st June. About a week later, the chicks died, probably through lack of food, but both birds continued to visit the nest-box until 28th July, the same date as the previous year.
Since that time, the BTO has published the results of one of the Fowlmere birds retrapped in 2011, and results from birds fitted with geolocators in Sweden have also been published. So now, we can tell the story of the Landbeach geolocator swift.

click image to enlarge

The route taken by Swift 396. The red line marks its journey south and green the return journey.

The dotted green section is around the spring equinox, when positions are unreliable. 

The bird was in the Congo for the autumn equinox.

The journey followed by Swift 396 resembles quite closely a number of other British Swifts. The following is a paraphrase of Chris Hewson's analysis and account of Swift 396 (thank you Chris):

"The positions (yellow symbols) are averages for day & night positions for three day periods, this smooths the data and also compensates for movement between successive dawn or dusk events.  Points within 20 days  of each equinox are removed, because the determination of latitude cannot be achieved reliably. After 21 April, daily 3-day running averages are shown (i.e. the three day periods overlap by two days) to give a clearer idea of the migration from central through to west Africa. It probably arrived back in Landbeach sometime on 12 May but didn't go back into the box until after dark. The first time one can be sure it was back in the box was 0728 on 13 May but it was probably there in the evening of 12 May. 

It left on 28 July 2011. It then spent about 8 days in southern Spain but still managed to be in the Congo in less than 4 weeks. There is a nice cluster in the western part of the Democratic Republic of Congo, where it spent the time from around 24th August until it left on its winter 'excursion' to Mozambique around 10th December.  After the spring equinox 'blackout', it looks like it is in northern Angola before moving into Gabon then across the Gulf of Guinea to Ivory Coast / Liberia. This bird seems to cross the Gulf of Guinea from near the equator with the first positions in West Africa off Ivory Coast. While at first sight, this may be hard to believe, tracks of other birds also show birds arriving off Liberia from the south-east, which indicates an oversea approach and gives confidence that these birds fly over the Gulf as does the fact that no tracks have been seen with a bird hanging around further east than the Ivory Coast."

As if a 2000km journey across the ocean was not enough, the bird refuelled for a week in Ivory Coast and Liberia until 5th May, and then it was back in its box in Landbeach on 12th May, that is over 5000 km in 7 days!

Geolocator Swift returns to Landbeach

Contributed by Dick

This swift with a geolocator returned to its nestbox on 13th May, following its mate that returned on 11th May. It was fitted with its geolocator on 21st July 2011, and then continued returning to the nest-box until the chicks fledged on 28th July.
The video shows the bird with the geolocator returning first, followed over 20 minutes later by its mate bringing the first feather in of 2012. The initial greeting looks more like a fight, but they soon settle down for a night of mutual preening.

New light-weight geolocator for Swifts from Migrate Technology

Although we still wait to see published results, geolocators have already proved their effectiveness in tracking Swifts all the way from Europe to Africa and back. Particularly pleasing is that return rates are no worse than from ringed birds, showing that they have negligible impact on survival.

Now, Migrate Technology Ltd has announced a new range of devices, the Intigeo-W65 and Intigeo-W55, weighing 0.65gm and 0.55gm respectively. Especially attractive is the low price of these devices at £95 ex VAT.

Swift with Integeo-W55B1J geolocator
© Lyndon Kearsley

Novel in many ways, the manufacturer claims that this extremely energy efficient yet powerful miniature device will be the smallest geolocator available in the world to record near full range light level for an entire annual cycle. The device may be particularly useful for wader (shorebird) researchers or those studying highly aerial birds such as swifts, martins and swallows.

These new devices record, not only light levels for the determination of dawn and dusk for location, but also temperature and conductivity. Could temperature give an indication of the altitude that Swifts fly, given that temperature drops about 6.5°C for every 1km increase in altitude? Conductivity is unlikely to be useful with Swifts, but it could tell what kind of wet habitat a shorebird or waterbird is occupying.

This technology is in the process of causing a revolution in our knowledge of where Swifts and other small migrants spend their time in Africa.

A Swift's last 3 hours in the nest

Contributed by Dick

The first Swift fledged early in the morning of 28th July. So, I left the camera running for 3 hours in the hope of capturing the 2nd chick leaving. This is the result showing the last feed by the parent Swift with geolocator, a cameo appearance by a flatfly Crataerina pallida, the Swift's last exercises, and finally leaving. We were in the garden outside having lunch, and witnessed its first flight - magic!

Swift behaviour with geolocator

Contributed by Dick

Swifts are declining, and one of the reasons could be something happening on their migration or in winter quarters on top of the loss of nest-sites. The advent of light-sensitive geolocators has provided great potential to fill in the gaps of where they go, and where they linger when they leave our shores.

There is bound to be apprehension at fitting anything to a small bird, particularly a Swift with its extreme fitness requirements, however, return rates of birds in previous years are no worse than one would expect from normal mortality, especially for birds caught leaving the nest, rather than birds caught on the nest.

On 21st July, one Swift was caught leaving my camera box, and fitted with a geolocator by Chris Hewson and Phil Atkinson from the BTO. That night, and since, the bird returned to the nest-box. Here is some video: