|Habitat and Ecology:
Behaviour Most of this species is fully migratory although it may only travel short distances (Kear 2005b), and certain populations in the north-west of Europe may also be sedentary (del Hoyo et al. 1992). The species breeds from April in solitary pairs (del Hoyo et al. 1992), after which it undertakes short northerly moult migrations to coastal areas, large lakes and rivers (Kear 2005b) to undergo a period of flightless moult lasting 3-4 weeks (males leave for this moult migration first while females are still incubating) (Scott and Rose 1996). Large moult gatherings are common during this period, males arriving at such gatherings in early-June and numbers peaking in late-August when adult females arrive (Scott and Rose 1996). The southward autumn migration begins in late-August with most arriving in the winter quarters by early-December (Scott and Rose 1996). Females tend to migrate further than males (Madge and Burn 1988, Scott and Rose 1996) and juveniles migrate further than adults (Scott and Rose 1996). The return migration to the breeding areas occurs as early as mid-February (Scott and Rose 1996), the species timing its arrival to coincide with the thawing and appearance of open water (Kear 2005b). Non-breeders may also oversummer on wintering grounds (Madge and Burn 1988). The species is gregarious outside of the breeding season (Snow and Perrins 1998, Kear 2005b) usually being observed in small scattered groups (Scott and Rose 1996) or in small flocks on migration (Kear 2005b). Several hundred individuals may roost together (Snow and Perrins 1998) and large flocks often gather to feed at sewage outfalls (del Hoyo et al. 1992) during the winter, although the species rarely occurs in very large flocks (Scott and Rose 1996). Habitat The species is restricted to water close to the shore and less than 10 m deep (Scott and Rose 1996) (showing a preference for waters 4 m deep) (Snow and Perrins 1998). Breeding When breeding the species shows a preference for oligotrophic lakes devoid of fish (Kear 2005b) but with abundant invertebrate life (Johnsgard 1978), and requires tree-holes (or artificial nestboxes) for nesting (del Hoyo et al. 1992). Suitable habitats include freshwater lakes, pools, rivers (del Hoyo et al. 1992) and deep marshes (Johnsgard 1978) surrounded by coniferous forest (del Hoyo et al. 1992). Non-breeding The species winters mainly at sea (Scott and Rose 1996) on inshore waters (del Hoyo et al. 1992), shallow bays (Kear 2005b), estuaries (del Hoyo et al. 1992) and coastal lagoons (del Hoyo et al. 1992, Scott and Rose 1996), especially in the vicinity of sewage outfalls (del Hoyo et al. 1992). Further to the south and on migration the species may also frequent large rivers, lakes (Scott and Rose 1996, Kear 2005b) and reservoirs (Scott and Rose 1996). Diet The diet of the species consists predominantly of aquatic invertebrates such as molluscs, worms, crustaceans, aquatic insects and insect larvae (del Hoyo et al. 1992) (e.g. dragonflies, damsel flies and may flies) (Johnsgard 1978), as well as amphibians, small fish (del Hoyo et al. 1992) and some plant material (mainly in the autumn) such as seeds, roots and the vegetative parts of aquatic plants (del Hoyo et al. 1992). Breeding site The species nests in hollows of mature trees (del Hoyo et al. 1992) (e.g. aspen, spruce or oak) (Flint et al. 1984) formed by woodpeckers or by bacterial or fungal heart-rot invasions (Kear 2005b) that have internal cavity diameters of c.20 cm (although the height of the hollow does not seem to be important) (Johnsgard 1978). The species will preferentially nest in trees in open stands near water (Johnsgard 1978, Madge and Burn 1988) or solitary trees on the edges of marshes (Johnsgard 1978), rather than in trees in dense stands (to increase the ease of entry by flying) (Johnsgard 1978). The species will also nest in artificial nestboxes (del Hoyo et al. 1992). Management information In Scotland, UK the introduction of a sewage treatment scheme in the Firth of Forth (a large marine bay) resulted in a considerable reduction in the abundance of the species, with feeding flocks only remaining at outfalls where sewage continued to be discharged in large quantities (Campbell 1984). It was unclear whether the changes in the species's distribution were due to reductions in the number of food items borne in the sewage or to reductions in aquatic invertebrate abundance as a result of the new treatment system (Campbell 1984). In some areas nestbox erection programmes have been shown to cause significant range expansions and population increases (Dennis 1987, del Hoyo et al. 1992), although an experiment in southern Finland found that even though nestbox provision increased breeding numbers of the species there was a negative density-dependent effect on reproductive output (i.e. the number of fledged young did not increase despite an increase in breeding pairs) (Poysa and Poysa 2002). Nesting habitats in general may also benefit from a more extended rotation of timber harvesting (Kear 2005b).