Conservation biologists have used surrogate species as a shortcut to monitor or solve conservation problems. Indicator species have been used to assess the magnitude of anthropogenic disturbance, to monitor population trends in other species, and to locate areas of high regional biodiversity. Umbrella species have been used to delineate the type of habitat or size of area for protection, and flagship species have been employed to attract public attention. Unfortunately, there has been considerable confusion over these terms, and several have been applied loosely and interchangeably. We attempt to provide some clarification and guidelines for the application of these different terms. For each type of surrogate, we briefly describe the way it has been used in conservation biology and then examine the criteria that managers and researchers use in selecting appropriate surrogate species. By juxtaposing these concepts, it becomes clear that both the goals and selection criteria of different surrogate classes differ substantially, indicating that they should not be conflated. This can be facilitated by first outlining the goals of a conservation study, explicitly stating the criteria involved in selecting a surrogate species, identifying a species according to these criteria, and then performing a pilot study to check whether the choice of species was appropriate before addressing the conservation problem itself. Surrogate species need to be used with greater care if they are to remain useful in conservation biology.

ABSTRACT. Female common eiders (Somateria mollissima v-nigrum) generally select nest sites in areas with driftwood cover. Previous studies of common eiders have shown a positive relationship between nest success and driftwood cover. Our observations led us to hypothesize that cover does not enhance nest success when mammalian predators are present. To evaluate nest cover selection in common eiders, we examined five years of nesting data to determine the interactions between the probability of nest activity and the amount of driftwood cover in the presence of avian versus mammalian predators. Most common eider nests were surrounded by low (40%) or moderate (38%) driftwood cover. Nest failure rates were high (32% -95%), and arctic foxes (Alopex lagopus), alone or with polar bears (Ursus maritimus), appeared to be more destructive than glaucous gulls (Larus hyperboreus) to eider nests. Logistic regression was used to model common eider nest activity associated with driftwood cover and predators. When glaucous gulls were the only predators, more driftwood cover consistently increased the probability of nest activity. But when foxes were present, nest activity consistently decreased with increasing cover. Our models support our observations that nest cover was beneficial to eiders when glaucous gulls alone were predators. Driftwood cover may be most important for the thermal and structural protection it offers, rather than for the camouflage it provides. The energetic benefit provided by driftwood windbreaks coupled with the common eider's behavioral response of decreased nest attendance, or increased exposure to avian depredation of nests as energy reserves are depleted during incubation, provides an explanatory mechanism for our model results.Key words: arctic fox, Alopex lagopus, driftwood habitat, egg depredation, glaucous gull, Larus hyperboreus, polar bear, Ursus maritimus RÉSUMÉ. L'eider à duvet femelle (Somateria mollissima v-nigrum) choisit en général son site de nidification dans des zones ayant un couvert de bois flotté. Des études précédentes sur les eiders à duvet ont révélé qu'il existe une relation positive entre le succès de la couvée et le couvert de bois flotté. Nos observations nous ont amenés à émettre l'hypothèse que le couvert n'augmente pas le succès de la couvée en présence de prédateurs mammifères. Afin d'évaluer le choix de couvert du nid chez l'eider à duvet, nous avons examiné des données de nidification obtenues sur cinq années, en vue de dégager les interactions entre la probabilité d'activité au nid et la quantité de couvert de bois flotté en présence de prédateurs aviens par opposition aux prédateurs mammifères. La plupart des nids de l'eider à duvet étaient entourés par un faible couvert de bois flotté (40 %) ou un couvert modéré (38 %). Les taux d'insuccès étaient élevés (32 à 95 %) et le renard arctique (Alopex lagopus), seul ou avec l'ours polaire (Ursus maritimus), semblait plus destructeur pour les nids de l'eider que le goéland bourgmestre (Larus hyperboreus). On a utilisé la ré...

No abstract

(Somateria spp.) and scoters (Melanitta spp.) were documented in three barrier island-lagoon systems in the central Alaska Beaufort Sea. Concentration areas for each species were determined during 16 aerial surveys. Kernel density procedures were used to delineate 75% and 50% "activity" or concentration areas for all three species. Long-tailed Ducks were 13 times more numerous than eiders and 38 times more numerous than scoters. The Long-tailed Duck 75% activity area encompassed all three lagoon systems and was three times as large as the eider activity area and one-third larger than the scoter activity area. Eider activity areas were located only in the eastern lagoon, and scoter activity areas were located only in the western lagoon. Density contours showed patterns of repeated habitat use for sea ducks over the four years of sampling and improve our understanding of sea duck habitat use within Beaufort Sea barrier island-lagoon habitats.