Curlews at pre-season roosts

Pre-season roosts

Under the subject of Roosting, “The Birds of the Western Palearctic” (Volume 2, page 506) distinguished between roosts outside the breeding season (whether at high tide or at night) and in the breeding season. In the latter period it notes: ‘After moving inland to breeding grounds but before breeding starts, most birds continue to roost communally (typically 25-50 birds together, but sometimes more); arrival time determined by light intensity’. Rather little attention seems to have been paid to this behaviour, though birds have been observed acting in exactly this way for many years along the Severn in Gloucestershire (at the Gloucestershire Wildlife Trust reserve at Coombe Hill Meadows) and along the Avon above Tewkesbury in Worcestershire. The birds generally congregate in the evening round a water body or on an island, where they are safe from nocturnal predators such as foxes or badgers, staying until the follow morning when they depart early. This behaviour has been noted in spring by Tony Cross higher up the Severn on the Montgomeryshire Wildlife Trust reserve at Dolydd Hafren (‘Severn Meadow’), an island site where he has mist-netted and colour-ringed adults over the years. Another roost was recorded at the BTO Nunnery Lakes reserve at Thetford in spring 2019, when some birds were caught and colour-ringed; one of them was a bird from the Severn estuary catches in Gloucestershire between 2010 and 2013 and already known to be nesting close to Thetford (Samantha Franks pers. comm).

Such roosts have also been noted at Belmont Reservoir in Lancashire by Stephen Martin and his team. Steve notes: “We have a lot of data (roosts and breeding) going back several decades on Curlew in the West Pennine Moors and I have always considered it an ideal site for further studies mainly because of its accessibility and the fact we still have good numbers breeding.”

The above table shows that a shallow decline in pre-breeding roosting numbers is evident since the late 1980s. However numbers, albeit subject to wide annual fluctuations, have been fairly level since 2005. The establishment of the huge Black-headed Gull colony in the last 10-15 years on the preferred roost site could also influence roosting numbers: early in the pre-breeding roost cycle (up to early March), the gulls are still roosting on the water, leaving the main Curlew roost site (on the northwest water’s edge of the main island) free. Thereafter the gulls remain on the island overnight, leaving very little room for Curlews. Whilst once the gull colony was probably a positive for the Curlews (as it gave added warning of predators) the huge increase in recent years (up to 11,500 pairs) has meant that the roost site is now tightly-packed with gulls when the Curlews arrive.

Steve quotes the following summary of data prepared for the 2019 Lancashire Bird Report:

  • First two Curlews back at Belmont Reservoir on 16.02.19
  • Pre-breeding counts at Belmont Reservoir: 14 on 19.02.19; 62 on 24.02.19; 177 on 27.02.20; 110 on 01.03.20; 140 on 03.03.20; 172 on 08.03.20; 166 on 10.03.20; 55 on 17.03.20.

In 2020 the peak was of 114 Curlews on 6 March, though 110 remained until 21 March, but only 56 birds came in on 23 March, some birds now presumably staying on territory

The table below shows the dates of peak counts at Belmont Reservoir.

When the birds first arrive at Belmont in February/early March they normally feed on a few favoured fields and come to the roost in flocks, whereas from mid-March, most of the birds come into the roost in pairs, from all directions, having presumably been visiting their territories during the day.

After the breeding season, the roost builds up again at Belmont – but that’s another story!

At another site in Lancashire, Alston Wetland (a former reservoir, now disused), Gavin Thomas and his team of observers have noted similar pre-season roosts. Gavin writes: “We have been increasing efforts in recent years to do similar pre-breeding roost counts around Preston – the Grimsargh Wetlands (another former reservoir) and Brockholes, a  Lancashire Wildlife Trust Nature Reserve  – to get a handle on numbers and timings. By no means as exhaustive as the excellent data gathered around Belmont, but nonetheless useful. Most of the counts are logged on the East Lancashire Ornithological Club’s database. The roost picture at Alston Reservoirs/Wetland, is similar to Belmont with late February or March peaks, but birds are still around into April. Post-breeding roosts are smaller in number but also start early, due to adults having lost nests and chicks.”

Local breeders or migrants in these pre-season roosts?

The question then arises of the status of the birds in these pre-season spring roosts: are they birds which are going to breed within a relatively short distance of the roost? This seems to be the general consensus. Or do they also include some birds which are passing through, on their way to breeding grounds further to the northeast?

Of the Belmont birds, Stephen Martin says: “It is not conclusively known that all birds using this roost are local breeders, however study of groups/pairs entering the roost would suggest that this is highly likely.”

At the Coombe Hill roost in Gloucestershire up to 20 birds were recorded from 16 February to 16 March 2019 by Andy Jayne and Mike Smart, though observations became difficult in mid-March because of a Severn flood. The birds arrived late in the evening, and left early in the morning; one of the birds was a male, colour-ringed on the Severn estuary in September 2010, and known to have attempted to nest every year since at a site about three or four miles from the roost; thus at least one was a local breeder. Meanwhile along the Avon in Worcestershire up to 32 birds were noted coming to a roost on several dates in March 2019.

The situation in the Severn and Avon Vales in 2020 has been complicated by very severe flooding. At Coombe Hill and along the Avon, the roost sites have been under water, sometimes two or three metres deep, for much of the first three months of the year, so the roosts have not been accessible either to birds or to observers. And indeed the flooding has extended over many of the meadows in which the birds nest, so that the birds did not need to fly off to roosts near water, and instead seem to have stayed to roost on the spot: thus along the Avon a flock of up to 36 was present in late February and early March, sitting on the narrow green top of the flood-bank, the only area to emerge from the flood; there seems no reason for these birds to fly off in the evening to another roost. One of the birds was colour-ringed and is strongly suspected (though its ring number has not yet been definitely confirmed) of being a breeding bird ringed only a mile away in 2019. Do these 36 birds represent the number of birds likely to nest in the immediate area?

Similarly on 6 March 2020, Chris Wells recorded a group of Curlews on the Lower Lugg Meadow in Herefordshire, a Lammas meadow, in a fully organic flood plain, where fertilisers have not been used.  It is an SSSI and is closed for public access during the nesting period under the CROW Act. Curlew have historically bred there in such numbers that they were not counted but, in line with the international species decline, numbers have tumbled to  near extinction level. In the last three years Curlew have nested but failed to produce any young. The meadow covers in excess of one hundred  hectares, and has been under water since October 2019. Peak numbers on the Lugg were 33, all unringed, on 6 March; numbers remained fairly constant for a week, then gradually reduced to 21 on 17  March, 12 on 19 March and only three on 22 March. Two of these were feeding in a nesting area used in previous years. None of these Curlews were observed on the roost site used during the nesting period. Presumably, they found a safe haven on the small area of the meadow that was not covered in water. The vast majority of the meadow is now clear of water. Also on 6 March Chris Wells recorded 29 Curlews feeding on short grass, across the Wye in Powys, one with an unread ring from Dolydd Hafren; this area had also suffered flooding during the past weeks. The Powys site was still holding about 30 birds on 21 March. Were these Lugg and Wye birds mainly concentrations of local birds preparing to breed locally? Chris Robinson comments “If the suggestion that the early influx at Lugg Meadows were birds that were preparing to breed elsewhere in the county is correct, I’d like to know where they all are now! How nice it would be to have a few marked birds!

In East Anglia too, floods affect the roosts; Sam Franks comments: “I wouldn’t say the roost formed again in 2020 in the same way. There was a maximum of seven  birds seen at any one time, and then not necessarily in an evening roost as per last year, and the maximum was only ever seen once – numbers were in the ones, twos and threes. Rather, I think birds may congregate at any time of day around water bodies (as mentioned, a flock of 12 were seen feeding together along the River Lark on Cavenham Heath in early March this year) to feed, rest and preen. But there is a lot more water around this year in Breckland, whereas in 2019, the Nunnery Reserve had one of the only decent water bodies in the area.”

Evidence from colour-ringed birds

It certainly seems that some migrant birds join these pre-season roosts. Gavin Thomas notes at Alston: “included in the roost in recent years has been a Finnish-ringed bird (CJ6) which often stays around for a fair few weeks, seemingly in no hurry to move on, so it seems not all of the Scandinavian breeders move through at speed and not all birds around in early spring are local breeders. She (inferred from bill length) first appeared on 8 March 2020 was still present on 24 March (see picture).”

Previous posts on this website have reported on recoveries of colour-ringed Curlews ringed by a BTO/WWT team on the Usk estuary near Newport in winter 2015/16 (see post of 30.11.19). Forty-one Curlews were marked with five colour rings: two colour rings, orange above white, on the left tarsus (this is the ”marker”, common to all birds ringed on the Usk), one colour ring on the left tibia, and two colour rings on the right tibia. Curlews are well known to be highly faithful to their wintering grounds, so it is no surprise that many of them have been re-sighted back on the Usk in later winters (though, intriguingly, one of them was observed from August to October 2019 along the coast of South Wales at the WWT centre at Llanelli, an unusual change of wintering ground – see post of 04.10.19). Several have provided distant recoveries on their breeding grounds, confirming that they are also faithful to breeding areas: one in Poland in spring 2018 and again in 2019 (posts of 19.09.18 and 03.04.19), one in Finland in both 2018 and 2019, and a third in Sweden in summer 2016.

What routes are taken by these Usk birds to reach their distant breeding grounds? As noted in the post of 30.11.19, one Usk Curlew was noted a little further up the Severn estuary in Gloucestershire by John Sanders on 19 April 2016, the only time he has ever seen any Usk Curlew in ten years of observations there; this must have been a migrant en route to breeding grounds further northeast; and (also mentioned in that same post) on 30 March 2019 Gavin Thomas recorded an Usk-ringed bird at Alston Reservoir in Lancashire. This bird was seen on only one occasion at a well-watched site, and the verdict at the time was that it was probably a passing migrant.

Recently another Usk-ringed bird has been recorded in Lancashire: Stephen Martin saw Orange Orange Green in the evening roost at Belmont Reservoir at Lancashire on 8 March 2020; it appeared on only one evening, so is likely to have been a migrant en route to a breeding area further off.

In addition to these adult Curlews colour-ringed in winter on the Usk, some adult Curlews have been colour-ringed at Dolydd Hafren by Tony Cross and the Mid-Wales Ringing Group. One of these birds was seen by John Sanders on the Severn in Gloucestershire: on 13 March 2017 John saw one of Tony Cross’s colour-ringed birds with a yellow ring in the high tide roost at Guscar Rocks (without being able to read the ring code); on 20 and 21 March 2018 he again saw one of these birds and read the ring as yellow AJ (originally ringed as an adult male at Dolydd Hafren on 22 March 2015); then on 18 March 2019 he saw yellow AJ again at nearby Wibdon Warth, where it stayed until 1 April; finally on 14 March 2020 yellow AJ was back on Wibdon Warth, and was at nearby Guscar on 20 March, on the shore at low tide, then feeding in the field, so it is hanging around. John is certain that yellow AJ does not winter on the Severn, and only appears in mid-March, sometimes staying for several days. It was very probably the bird he saw in 2017. Curlews are known to be exceptionally site faithful to their nesting sites and winter quarters, but these observations on the Severn (as well as the observations of Finnish CJ6 at Alston) suggest that they follow the same migration routes and use the same stop-over sites each year.

Another of the birds ringed by Tony Cross at Dolydd Hafren was in the flock of twenty-nine seen along the Wye in Powys on 6 March by Chris Wells. There was one colour-ringed bird, ring not read due to the distance, but ringed by Tony Cross: orange ring on right tibia and an inscribed yellow ring on the left tibia. Tony Cross comments “Think they will largely be local breeders having a pre-season gathering, as they do at Dolydd Hafren, but we know from ringing birds there that some will be migrant birds breeding further north.” Tony added “65 Curlews reported from Dolydd Hafren on the Severn near Welshpool on 4 March 2020 by Evelyn Bowles with at least twelve of them colour-ringed, unfortunately too far off for her to identify any of them.” Furthermore, says Tony, ”On 8 March Paul Leafe had 58 Curlews on the River Severn at Caersws, including two colour-ringed the previous spring in the Curlew Country project area on the Montgomeryshire/ Shropshire border; later the same evening Paul, Silvia Cojocaru and I caught and colour-ringed twelve more birds from the same flock. Colour-ringing attempts at Dolydd Hafren in 2020 have been impossible due to high river levels and windy weather.”


As always: ‘more research needed’ on these pre-season roosts and their composition. In some cases  (Coombe Hill, Avon, Dolydd Hafren, Nunnery) the roosts contain birds known to be local breeders; the number of the birds roosting (sometimes surprisingly large) may give an indication of the size of the local population, difficult to assess when the birds disperse to their nesting territories. In other cases (Alston, Belmont, Powys), passing migrants may occur briefly, often just for a single night in a roost. And the observations of individuals on the Severn estuary (AJ) and Alston (CJ6) raise the intriguing possibility that migrant Curlews are faithful not only to breeding and wintering sites, but to migration stop-overs too.

Many thanks to all those who have contributed information to this round-up: Tony Cross, Samantha Franks, Steve Martin, Chris Robinson, John Sanders, Gavin Thomas and Chris Wells.

Mike Smart


Leave a comment