Behaviour In central and north-west Europe this species is mainly sedentary (del Hoyo et al. 1992) although other populations are chiefly migratory (Madge and Burn 1988). Migratory populations arrive on their breeding grounds from late-April (Kear 2005a) where they breed in single pairs or loose groups (del Hoyo et al. 1992) with hundreds often nesting at the same site (although not colonially) (Kear 2005a). Once incubation has commenced (Madge and Burn 1988) the males gather in flocks (Madge and Burn 1988, Kear 2005a) to moult (moult migrations occurring in some areas) between late-June and early-September when they become flightless for 3-4 weeks (Scott and Rose 1996) (females moult their flight feather 1-2 months later) (Scott and Rose 1996). The autumn migration begins in September (Kear 2005a), with a segregation of the sexes occurring in autumn and winter due to the difference in the timing of the moult (Scott and Rose 1996). The return spring migration begins in late-February (Kear 2005a). During the winter the species is highly gregarious and may occur in flocks of several thousand individuals (Kear 2005a). Habitat Breeding The species breeds in lowland regions and shows a preference for eutrophic waters 3-5 m deep (avoiding lakes deeper than 15 m) (Kear 2005a) with open water, islands for breeding (del Hoyo et al. 1992) and abundant marginal and emergent vegetation (Kear 2005a). It is common on large, freshwater lakes, ponds, reservoirs (del Hoyo et al. 1992, Kear 2005a), gravel-pits (Kear 2005a) and quiet stretches (Kear 2005a) of wide slow-flowing rivers during this season (del Hoyo et al. 1992, Kear 2005a). Non-breeding During the winter the species frequents large freshwater lakes (Kear 2005a), reservoirs (Snow and Perrins 1998) and sheltered coastal locations (Kear 2005a) such as brackish lagoons (del Hoyo et al. 1992), brackish inland seas (Kear 2005a) (e.g. Caspian Sea) (Scott and Rose 1996), tidal bays (del Hoyo et al. 1992) and estuaries (Madge and Burn 1988, Snow and Perrins 1998) although it avoids strong wave action and very exposed maritime conditions unless all inland freshwaters become frozen (Snow and Perrins 1998). Diet The species is omnivorous (Kear 2005a) a major part of its diet consisting of molluscs (especially Mytillus and Cardium spp., gastropods (del Hoyo et al. 1992) and zebra mussels Dreissena polymorph (Kear 2005a)), crustaceans and aquatic insects (del Hoyo et al. 1992), as well as grain (Kear 2005a) and the seeds and vegetative parts of aquatic plants (del Hoyo et al. 1992). Breeding site The nest is constructed of vegetation and is placed in water (Kear 2005a) on floating mats or islets (Flint et al. 1984), or on the ground on islands (del Hoyo et al. 1992) in rush or grass tussocks (Snow and Perrins 1998), under bushes (Kear 2005a) or in the open amidst marsh-nesting gull or tern colonies (for protection against predators) (Madge and Burn 1988, Kear 2005a). Nests are usually placed within 20 m of water (Kear 2005a) although on islands they may be placed up to 150 m away (Snow and Perrins 1998). Although the species is not colonial hundreds may nest on the same site with neighbouring nests spaced 7-11 m apart (within gull or tern colonies nests are spaced 2-3 m apart) (Kear 2005a). Management information In the UK (Salford docks, Manchester) the species prefers to feed in waters heavily polluted with sewage that are devoid of aquatic vegetation but hold high densities of oligocheates and other pollution-tolerant organisms (Marsden and Bellamy 2000). The species may therefore suffer from plans to improve water quality in the docks (e.g. modernising sewage treatment systems and oxygenating water) (Marsden and Bellamy 2000). Experimental removal (extermination) of the nest predator American mink Neovison vison in the outer archipelago of south-west Finland also resulted in an increase in the breeding density of this species (Nordstrom et al. 2002). The species competes for food (e.g. chironomid larvae and molluscs) with fish species such as roach Rutilus rutilus (Winfield and Winfield 1994). A study in Czech ia found that fish ponds with a fish stock density of less than 400 kg ha¹, water transparency of more than 50 cm, mixed fish stocks (e.g. tench and pike or perch) rather than monospecific stocks (e.g. of carp), and systems that include ponds with fish fry (to provide areas with low fish competition and high invertebrate availability) are more successful in supporting breeding pairs of this species (Musil 2006). The cyclical removal of adult fish from an artificial waterbody (gravel pit) in the UK resulted in an increase in invertebrate food availability and increases in the growth of submerged aquatic macrophytes, which in turn led to a increase in brood use of the habitat by the species and increased species survival (Giles 1994). The removed fish (dead or alive) were sold to generate funds (Giles 1994).

Conservation Actions Underway
CMS Appendix II. EU Birds Directive Annex II and III. The following information refers to the species's European range only: Experimental removal of the nest predator American Mink (Neovison vison) in the outer archipelago of south-west Finland resulted in an increase in the breeding density of this species (Nordstrom et al. 2002). The cyclical removal of adult fish from an artificial waterbody (gravel pit) in the U.K. resulted in an increase in invertebrate food availability and increases in the growth of submerged aquatic macrophytes, which in turn led to an increase in brood use of the habitat by the species and increased species survival (Giles 1994). The removed fish (dead or alive) were sold to generate funds (Giles 1994).

Conservation Actions Proposed
The following information refers to the species's European range only: Large scale conservation measures should look to protect key breeding sites from habitat alteration and look to promote low-intensity farming. Hunting should be monitored to ensure numbers harvested are sustainable. Predator control should be considered in breeding areas.