ABSTRACT1. The Don˜ana wetland complex (SW Spain) holds more wintering waterfowl than any other wetland in Europe.2. This study focused on the use made by 12 common waterbirds (eight ducks and four waders) of the natural seasonal marshes in Don˜ana National Park (DNP) and the adjacent Veta la Palma (VLP) fish ponds created in the early 1990s. Data used were from aerial and terrestrial surveys collected between October and February during six consecutive winters from 1998/99 to 2003/04. Changes in distribution of each bird taxon were related to changes in the extent of flooded marshes within DNP. Up to 295 000 ducks were counted in VLP during dry periods, and up to 770 000 in DNP when it was flooded.3. The timing and extent of flooding in DNP was highly variable, but there was a consistent pattern in which ducks concentrated in VLP during dry months and winters but redistributed to DNP as more of it was flooded. This refuge effect was also strong for black-tailed godwits Limosa limosa, but much less so for other waders. Waders feed mainly on invertebrates, and invertebrate biomass in VLP was found to be higher than in DNP. Ducks feed mainly on seeds and plant material, which are more abundant in DNP when flooded.4. When water levels in DNP were stable over the course of a winter, or controlled for in multivariate models, the numbers of ducks at VLP declined over time, probably due to reduced availability of plant foods. In contrast, numbers of waders at VLP were more stable, and their invertebrate prey became more abundant over time, at least in the winter 2003/4. 5. In this extremely important wetland complex, the value of natural and artificial wetlands for wintering waterbirds are complementary, providing suitable habitat for different species and for different conditions in a highly variable Mediterranean environment.

No abstract

Size-structured interspecific interactions can shift between predation and competition, depending on ontogenetic changes in size relationships. I examined the effects of common carp (Cyprinus carpio), an omnivorous fish, on the reproductive success of the red-necked grebe (Podiceps grisegena), an avian gape-limited predator, along a fish size gradient created by stocking distinct age-cohorts in seminatural ponds. Young-of-the-year (0+) carp were an essential food source for young grebes. Only adult birds were able to consume 1-year-old (1+) fish, while 2-year-old (2+) fish attained a size refuge from grebes. Amphibian larvae were the principal alternative prey to fish, followed by macroinvertebrates, but the abundance of both dramatically decreased along the carp size gradient. Fledging success was 2.8 times greater in ponds with 0+ versus 1+ carp; in ponds with 1+ carp, chicks received on average 2.6–3 times less prey biomass from their parents, and over 1/3 of broods suffered total failure. Breeding birds avoided settling on 2+ ponds. These results show that changes in prey fish size structure can account for shifts from positive trophic effects on the avian predator to a negative impact on the predator’s alternative resources. However, competition did not fully explain the decrease in grebe food resources in the presence of large fish, as carp and grebes overlapped little in diet. In experimental cages, 1+ carp totally eliminated young larvae of amphibians palatable to fish. In field conditions, breeding adults of palatable taxa avoided ponds with 1+ and older carp. Non-trophic interactions such as habitat selection by amphibians or macroinvertebrates to avoid large fish may provide an indirect mechanism strengthening the adverse bottom-up effects of fish on birds.Electronic supplementary materialThe online version of this article (doi:10.1007/s00442-010-1862-3) contains supplementary material, which is available to authorized users.

Wildlife occurring at aquacultural sites can ignite conflicts over resources managed by humans. A telephone survey concerning nuisance wildlife occurrence, perceptions of inflicted damage, and use of preventive measures at pond fisheries was conducted in 2003-2004 in eastern Poland. Significant economic losses to wildlife were reported by 80% of the respondents, with 41% claiming severe losses to more than one species. Serious damage was attributed primarily to otters Lutra lutra (56% of farms), cormorants Phalacrocorax carbo (26%), grey herons Ardea cinerea (23%), and beavers Castor fiber (21%). Two alien species, mink Neovison vison and muskrat Ondatra zibethicus, were widespread, but rarely blamed for causing substantial harm. Lethal controls were the most popular means of damage reduction and were more likely to be deployed at fish farms reporting significant losses. Cormorants, shot at half of the farms visited, were the most widely persecuted. Both otters and beavers were culled, notwithstanding their protection status and compensation payments available for beaver damage to ponds. The survey indicated poor cooperation between conservation authorities and fish farmers in managing wildlife concerns. Monitoring conflict interactions, i.e., wildlife occurrence and alleged damage at fish farms vs damage management processes, is an essential step toward conflict alleviation. Considering the wide range of species interacting with fisheries, adoption of more flexible policies to address the conflicts than a singlespecies conservation approach is recommended.

In anthropogenic landscapes animals may be lured into low‐quality habitats where they survive or reproduce poorly (‘ecological traps’). I investigated breeding habitat selection in relation to intra‐seasonal changes in food availability and reproductive output in red‐necked grebes Podiceps grisegena, a size‐limited predator, of common carp Cyprinus carpio ponds. Carp farms constitute highly heterogeneous habitat mosaics due to separate stocking of different age/size fish. Pond features significant for grebe settling decisions, i.e. hydroperiod and emergent vegetation cover, had no obvious effects on prey abundance for chicks and on fledging success. Breeding grebes avoided ponds containing fish too large for them to ingest but exhibited little preference between ponds with medium‐sized one‐year‐old carp that could be exploited by pre‐laying birds, and ponds designated for young‐of‐the‐year carp, where only invertebrates and amphibians were available as prey in early spring. Red‐necked grebes settling on ponds with medium‐sized fish failed to predict future shifts in interactions with carp stocks; carp exceeded the prey‐size threshold of chicks and adversely affected their non‐fish prey levels. The resulting food shortage led to severe egg‐to‐fledging mortality rates compared to fishless ponds or those containing young‐of‐the‐year fish. This study shows that waterbirds vulnerable to competition from fish can risk maladaptive habitat selection due to unrecognised spatial and temporal variation in food resources caused by fish stocking practices. Ecological traps created by perturbations to trophic interactions may be common but difficult to detect because altered dynamics of trophic resources can affect wildlife indirectly. As with other types of ecological traps, manipulation of habitat features identified as attractive cues for settling animals, but not related to critical food resources, may help to reduce perceptual pitfalls. For example, wetland management to mitigate trap effects driven by commercially stocked fish should preserve abundant emergent vegetation in habitats with weak fish impact and extend their hydroperiod.