The degree to which juvenile waders congregate, either in distinct flocks or in subgroups within flocks, is pertinent to the accuracy of productivity estimates (e.g. Rogers et al. 2005). While the potential for flocks to contain non-randomly distributed birds is not a new topic for discussion (e.g. Weston 1992), observations on how (or even whether) birds actually do ‘bunch’ in age-classes have been few (Harrington 2004, Rogers et al. in press). The lack of detailed information on flocking behaviour may be due to difficulties distinguishing juveniles from adults in the field. Juvenile Bar-tailed Godwits (Limosa lapponica) are, however, readily identified in the field due to their distinctive plumage and recent observations have shown clearly how juveniles can congregate within mixed-age roosting flocks.
On 8 September 2004, Martin Green (University of Lund, Sweden) and I were observing Bar-tailed Godwits at a high-tide roost near the Opagyarak River, on the Yukon-Kuskokwim Delta, Alaska, U.S.A. (60°05'N 165°30'W). Initially, 80 birds (from an earlier flock of 101) were preening and roosting on the tidal flats in a flock that contained some juveniles. When this flock flew, only 35 birds landed close to us, of which 34 were juvenile. About half an hour later these birds were joined by a large flock that had flown in from several kilometres away. Counts showed that only 46 of the 2,410 birds were juveniles (so the age breakdown of the second flock was a maximum of 12 juveniles to 2,318 adults).
The newly formed flock was flighty, regularly lifting off and circling around, with waves of birds dropping out and settling back onto the sand flat. Almost every time this happened, the juveniles were the first birds to resettle, and they often had their bills tucked into their feathers in a sleeping pose before the second wave of birds had even landed. Successive waves of birds did not start to settle until they flew over the first birds in the flock, with the end result that most of the juveniles were grouped together at the rear end of the flock.
In the Firth of Thames, New Zealand (37°11'S, 175°19'E), I have seen birds distributed in a similar way. For example, on 29 October 2004, godwits flew into a northerly breeze to land on a shallow pond adjacent to the mudflats to roost. Of the 173 juveniles, most were at the southern end of the flock, having been apparently the first to land.
It is also clear in October/November (when juveniles are most obvious) in the Firth of Thames, that if some godwits are still feeding as the tide-line approaches the shore (while the majority is pre-roosting on the flats), the feeding group will contain a disproportionate number of juveniles. On 14 October 2004, at least 84 juveniles were present in a flock of over 2,000 birds; 40 of these were feeding together as the tide came in.
On a larger scale, it is clear that juveniles may not be randomly distributed between feeding or roosting sites. In Alaska, a sandbar only about 10 km from the Opagyarak roost held 870 godwits at high tide on 5 September 2004. Of 738 birds formally scanned for an age ratio, 606, or 82.1%, were juvenile; the corresponding figure for the Opagyarak roost was 1.95%.
These observations illustrate two mechanisms which can lead to a non-random distribution of juveniles in roosting flocks: the settling before adults of juveniles at a roost and a tendency for juveniles to make up a disproportionate part of ‘late-feeding’ birds and presumably later to group together within a roosting flock. The mechanisms have been observed in Bar-tailed Godwits and presumably also occur in other species of wader. Only in species such as the Bar-tailed Godwit in which juveniles are readily distinguishable can the mechanisms be observed. Why juveniles distribute themselves non-randomly over larger scales is a much larger research question, but the fact that they do stresses the importance of sampling widely when assessing age ratios, either by cannon-net catches or visual sampling.
Thanks to Sue Moore for comments, and to Bob Gill and Brian McCaffery for superb logistical organisation in Alaska. My work is supported by the Foundation for Research, Science and Technology, New Zealand.
Harrington, B. 2004. Use care in determining age-ratios in shorebirds: they may differ relative to flock position, flock location and behaviour. Wader Study Group Bull. 104: 92-93.
Rogers, D., K.G. Rogers & M.A. Barter. In press. Measuring recruitment of shorebirds with telescopes: a pilot study of age ratios on Australian non-breeding grounds. In: Status and Conservation of Shorebirds in the East Asian-Australasian Flyway. Eds.: P. Straw and D. Milton, International Wader Studies and Wetlands International.
Rogers, K.G., Minton, C. and Rogers, D.I. 2005. Some sampling considerations relevant to estimating the first year percentage. The Stilt 47: 4-9.
Weston, M. 1992. A potential methodological problem with determining the reproductive success of Palearctic waders from the proportion of juveniles in cannon-netted samples. The Stilt 20: 18-19.