Aves, p�le ornithologique de Natagora

Consultez les articles du Bulletin Aves !


Retour à la liste des articles

Aves, 43/1 | 2006 | 1-18

  Emergence and development of a winter roost of Goosanders (Mergus merganser) : the case of Coo (Belgian Ardennes)
Bronne, L. & Brunin, Ch.

Article is available in pdf Download the French article in pdf (4.21 MB)

Article summary

The lake of Coo was created in the early 1970s in an abandoned loop of the river Amblève as part of an hydroelectric plant. Coo is now used as a winter roost by Goosanders. The often poor visibility conditions and the late arrivals of some of the birds allow only 37 out of the 98 evening counts made to be considered as good condition counts. Out of these, 26 counts were carried out during the winters of 2002-03 and 2003-04. We derived two models (one for the brownheads, one for the drakes) from these data, using a regression based on least squares method. We used winter day number (J and J2) and a temperature function (f) as variables. A large number of different temperature functions were tested. In order to select the most appropriate one we defined the "extended standard error of estimate", that takes into account not only the differences between the regression curve and the samples used in the regression but also the differences between the curve and the bad condition counts that are above the curve. As a matter of fact, these counts have a constant feature: they are always underestimated, and should then be under the curve. In the past Goosanders were only short-staying and solitary visitors at Coo. From February, 19th,1996 groups have been regularly seen on the lake. The recorded numbers of brownheads fit very well the model in all subsequent winters (see Fig. 2). For the drakes, the model provides too high numbers in the winters before 2002-2003 and too low numbers after that. That model can be correctedby a multiplicative coefficient (Fig. 7) that can be written as a linear function of year. Fig. 3 shows the corrected model along with the counts. We thus propose an explanation for the emergence and development of the roost. The rigorous month of February in 1996 forced Goosanders to leave their traditional wintering areas located northeast of Coo. Some birds discovered a new set of sites, including Coo, between which they move during the winter according to local weather conditions. We call this set of sites the "complex of Coo". On the next autumn when they had to "decide" for a wintering area, Goosanders preferred this complex over the one they used before February 1996, possibly because of overcrowding of the later. As the water level of the lake of Coo is constantly varying, it never freezes. Thus temperatures cold enough to force Goosanders to leave the lake for southwestern areas are unlikely to ever occur. Therefore Coo appears as the last remaining site at the end of the complex. This explains why the number of Goosanders only rises with lowering temperature. In the first winter, almost only brownheads (full-grown females and first-year males and females) were present. On the next year, the previous winter first-year male came back and could then be identified as drakes. And so on with new first-year birds joining the complex in similar proportion each year. This can explain the linearly increasing number of drakes (see Fig. 4). We feel confident that this scheme of wintering habit development could be exported to other wintering areas. Unfortunately other Goosander counts in Belgium have so far only been carried out during the day. A look at observations made by day on the graphs is enough to convince oneself of the poor significance of these data.

Articles on the same species

Harle bièvre
Mergus merganser

Consultez la fiche de l'espèce