163 Diet and habitat use of scoters Melanitta in the Western Palearctic - a brief overview A. D. Fox Department of Wildlife Ecology and Biodiversity, National Environmental Research Institute, Kalø, Grenåvej 12, DK-8410 Rønde, Denmark. E-mail: tfo@dmu. dk If patterns of scoter distribution and abundance are to be understood, there is a need to know upon which prey items these birds feed, how they obtain these prey items and the habitats from which these food items are most easily harvested. Dietary studies and descriptions of habitats exploited by Common and Velvet Scoter in the non-breeding season are reviewed. The existing literature strongly suggests that, outside of the breeding season, these species forage mainly upon marine bivalve mol­ luscs (especially those less than 4cm long) that live on the surface or within the upper 3cm of clean, coarse, sandy substrates in waters less than 20m deep. Although there is a large energetic cost to diving, han­ dling and crushing such prey prior to digestion, such sedentary prey items often occur in very high densities, offering a locally abundant and predictable feeding resource. Since single species often dominate the diet, but dominant food items differ between feeding areas, it seem s likely that scoters simply take whatever prey is locally available in suffi­ cient abundance to fulfil nutritional needs. Large differences in docu­ mented prey size frequency distributions suggest that scoters may not select for specific prey size classes below an upper digestive limit. However, in the absence of any precise understanding of how scoters obtain their prey, nor any simultaneous studies of available benthic food abundance and size class distributions in scoter diets, it is not possible to confirm if differences simply reflect differences in profitability between different prey at different sites at different times of the year. There remains a considerable need to study the basic feeding ecology and the behaviour of scoters and their prey if their patterns of distribu­ tion and abundance are to be better understood. Key Words: feeding ecology, Common Scoter, Velvet Scoter, bivalve molluscs, diving ducks © Wildfowl & Wetlands Trust Wildfowl (2003) 54: 163-182 164 Diet and habitat use of scoters Of all the factors determining the cient profitability to the predators. In distribution of organisms, the extent this case, sufficient profitability' is and availability of food is assumed to be defined as prey that yields sufficient of central importance. For Common energy/nutrient gain per unit handing M elanitta nigra and Velvet Scoters M. time to ensure profitable foraging fusca, which spend the majority of their (sensu Zwarts et al. 1992). This is espe­ time in shallow marine waters, food cially important am ongst organism s supply is likely to play a major role in that crush the tough outer shells of determining how birds are distributed ingested molluscs in a muscular giz­ in time and space. Although little is zard to obtain nutrients and energy. In known about the potential predators of the Comm on Eider Somateria moltissi­ these birds, it seem s likely that the ma, Nehls (1995) showed that 50-60% response to risk of natural predation of the ingested energy was expended in plays at best a secondary role in deter­ capturing, handling, processing and mining how scoters distribute them­ digesting prey. In birds that dive to feed selves at sea. Certainly, the role of on benthic organisms in the open sea, quasi-predator stimuli (such as human physiological limits on the depths to disturbance) apparently can play a which predators can dive may further major modifying role in the distribution affect the availability and profitability of of scoters locally. In order to under­ prey in deeper water. Energetic models stand why scoters occur where they do, and empirical observations suggest when they do, it is important to under­ that diving ducks show strong prefer­ stand what features of the food supply ences for foraging in the shallowest characterise those areas where scoters waters, but that prey depletion may are especially abundant or remain for force birds to shift to deeper waters as prolonged periods of time. So what pre­ a season progresses (Lovvorn & Jones cisely are the major features that char­ 1991; van Eerden ef at. 1997; de Leeuw acterise moulting and wintering habi­ 1997, 1999; de Leeuw et at. 1999). tats used by scoters in the Western Hence, selection of rich prey resources Palearctic? may change in time and space because This question needs to be of changes in water depth and deple­ addressed in a number of ways. Firstly, tion of the food base. Furthermore, prey there is a need to be able to determine items may be too cryptic, lie too deep in the diet and link the abundance of the sediment or dig themselves too dietary prey species to the habitats that deep to be easily accessible or available support the greatest densities of those to the predators (Piersma 1994; Kube prey items. Secondly, there is a need to 1996). Hence, an apparently suitable understand the way in which the birds rich benthos exhibiting high biomass exploit their prey, since conditions may may effectively lie beyond the reach of restrict the availability of prey of suffi­ the feeding apparatus of foraging scot- Diet and habitat use of scoters 165 ers. Even when abundant and available degree (Schricke 1993; Aulert & to predators, the size of prey (eg too Sylvand 1997). Crustaceans (isopods, small to be profitable or too large to be amphipods and small crabs), annelids, physically ingested) may affect the echinoderms and fish may also figure profitability of predation. Prey biomass in the oesophagus contents (Cramp & alone may thus not reflect the suitabil­ Sim m ons 1977; Stempniewicz 1986; ity of a given area to a given predator. Schricke 1993; Byrkjedal et al. 1997; Indeed, in one of the larger investiga­ Frengen & Thingstad 2002; Zydelis tions of benthos communities and scot­ 2002). Several observers report a more er distributions, Degraer et al. (1999) varied diet amongst Velvet Scoter and found no good correlation between consider that this is because they tend bivalve and scoter densities, perhaps to forage closer to the coastline than because water depth played such a sig­ Common Scoter, resulting in a more nificant role in determining the distri­ diverse diet and more prey items typical bution of the predators. For this reason, of rocky or stony shores (eg Madsen this synthesis draws on unpublished 1954; Cram p & Sim m o ns 1977). grey and published literature to briefly Nevertheless, in both scoter species, consider not merely the favoured prey molluscs comprised more than 95% of species of scoters, based on existing the diet by frequency and 80% by vol­ information, but also their abundance ume, so it is clear that these represent and habitat use, and those factors like­ the dominant prey items. ly to affect the suitability of the avail­ The composition of prey items able prey. varies considerably between studies - for instance in Lithuania, Common Diet Scoter was found to have consumed most frequently Macoma balthica, with Studies show that both Common relatively high proportions of poly- and Velvet Scoter consum e many chaetes and isopods, whereas Velvet bivalve molluscs (Madsen 1954; Cramp Scoter took mainly Mya arenaria & Simmons 1977; Stempniewicz 1986; (Zydelis 2002). In contrast, in the Gulf of Bräger & Nehls 1987; M eissner & Gdansk, Poland, Stempniewicz (1986) Bräger 1990; Durinck et al. 1993; found that Mya arenaria, Macoma baltica Schricke 1993; Bräger et al. 1995; and Cardium lamarcki were the most Leopold et al. 1995; Leopold & van der important items for both scoters. This Land 1996; Hughes et al. 1997; Degraer contrasts with the dominance of Spisula et al. 1999; Zydelis 2002). The most fre­ in North Sea coasts of Denmark and quently reported species are those list­ the Netherlands (Offringa 1991; ed in Tables 1 and 2. In addition to Durinck et al. 1993; Leopold et at. 1995; bivalve molluscs, gastropods (such as Leopold 1996). Analyses of the whelks Nassarius and periwinkles oesophageal content of shot scoters Littorina} are also taken to a lesser 166 Diet and habitat use of scoters Table 1. Commonly reported mollusc species present in the diet of Common Scoter, giving typical sediment type used by each organism (based on McMillan 1973 and other sources), the origin of the study and the source reference. Species Substrate Region Macoma balthica Mud/sand Belgium', Britain7, Danish Baltic4, Sweden5, Lithuania*, German Baltic12, Danish Baltic13 Cardium edule Sand Danish Wadden Sea3, Britain', Poland1, German Baltic1', Danish Baltic13 Mytilus edulis Various Danish Baltic4, Britain7, White Sea1", German Baltic1’ Mya arenaria Sand Lithuania'1, Poland", German Baltic1', Danish Baltic11 Spisula subtruncata Sand Belgium', Netherlands2, 8, Danish Wadden Sea3 Cyprina islandica Sand Geman Baltic1*, Danish Baltic'3, Danish Wadden Sea13 Donax vittatus Sand Belgium1, Britain7 Tellina tenuis Sand Belgium', Britain7, Littorina spp. Various Danish Baltic4, Britain7 Venus gallina Sand Danish Baltic4, Britain7 Mactra corallina Sand Netherlands7, Britain7 Pharus legumen Sand Britain7 Spisula elliptica Sand Britain7 Lutraria lutraria Mud/sand Britain Scrobicularia plana Estuarine mud Britain7 Gibbula cineraria Various Britain7 Cardium echinatum Sand Britain7 Abra alba Mud/sand Belgium' Tellina fibula Sand Belgium' Barnea candida Wood/peat/day Belgium1 Nassarius reticulatus Various Danish Baltic4 Cardium scarbrum Sand Danish Baltic4 Nucella sp. Various Danish Baltic4 Neretina fluviatilis Brackish Danish Baltic4 Musculus nigra Sand Danish Baltic4 Mya truncata Sand German Baltic'3 Cardium lamarcki Sand Poland" Ensis sp. Sand Danish Wadden Sea13 Taxonomy follows McMillan (1973]. Source references: 'Degraer ef al. (19991, 'Leopold et al. (19951, ‘Durinck et al. (1993) "Madsen (1954), 5Nilsson (1972), “Zydelis I2002), ’Hughes et al. (1997), “Offinga (1991), 'Bottelier 11938), "'Bianki (1992), "Stemniewicz (1986), 1 Meissner & Bräger (1990), "see text.