Artificial lights at night cause high mortality of seabirds, one of the most endangered groups of birds globally. Fledglings of burrow-nesting seabirds, and to a lesser extent adults, are attracted to and then grounded (i.e., forced to land) by lights when they fly at night. We reviewed the current state of knowledge of seabird attraction to light to identify information gaps and propose measures to address the problem. Although species in families such as Alcidae and Anatidae can be grounded by artificial light, the most affected seabirds are petrels and shearwaters (Procellariiformes). At least 56 species of Procellariiformes, more than one-third of them (24) threatened, are subject to grounding by lights. Seabirds grounded by lights have been found worldwide, mainly on oceanic islands but also at some continental locations. Petrel breeding grounds confined to formerly uninhabited islands are particularly at risk from light pollution due to tourism and urban sprawl. Where it is impractical to ban external lights, rescue programs of grounded birds offer the most immediate and employed mitigation to reduce the rate of light-induced mortality and save thousands of birds every year. These programs also provide useful information for seabird management. However, these data are typically fragmentary, biased, and uncertain and can lead to inaccurate impact estimates and poor understanding of the phenomenon of seabird attraction to lights. We believe the most urgently needed actions to mitigate and understand light-induced mortality of seabirds are estimation of mortality and effects on populations; determination of threshold light levels and safe distances from light sources; documentation of the fate of rescued birds; improvement of rescue campaigns, particularly in terms of increasing recovery rates and level of care; and research on seabird-friendly lights to reduce attraction.

The transport of nutrients by migratory animals across ecosystem boundaries can significantly enrich recipient food webs, thereby shaping the ecosystems’ structure and function. To illustrate the potential role of islands in enabling the transfer of matter across ecosystem boundaries to be gauged, we investigated the influence of seabirds on nitrogen input on islands. Basing our study on four widely differing islands in terms of their biogeography and ecological characteristics, sampled at different spatial and temporal intervals, we analyzed the nitrogen isotopic values of the main terrestrial ecosystem compartments (vascular plants, arthropods, lizards and rodents) and their relationship to seabird values. For each island, the isotopic values of the ecosystem were driven by those of seabirds, which ultimately corresponded to changes in their marine prey. First, terrestrial compartments sampled within seabird colonies were the most enriched in δ15N compared with those collected at various distances outside colonies. Second, isotopic values of the whole terrestrial ecosystems changed over time, reflecting the values of seabirds and their prey, showing a fast turnover throughout the ecosystems. Our results demonstrate that seabird-derived nutrients not only spread across the terrestrial ecosystems and trophic webs, but also modulate their isotopic values locally and temporally on these islands. The wealth of experimental possibilities in insular ecosystems justifies greater use of these model systems to further our understanding of the modalities of trans-boundary nutrient transfers.

We studied the foraging ecology of a tropical seabird community in 2 islands of the Seychelles from 2005 to 2007. Chick dietary samples were used to compare feeding habits among species and assess inter-annual and seasonal variations in diet. Fish prey dominated the diet of the community (68 to 100% of prey consumed), although cephalopods were present in 61.3, 40.0, 27.1 and 32.9% of the food samples from white-tailed tropicbirds, sooty terns, brown noddies and wedgetailed shearwaters, respectively. We found high diet overlap between species (Mullidae fish being the first prey consumed [30 to 90%] for all species except for the white-tailed tropicbird [< 2%]) but some segregation in prey length. Other important prey were Exocoetidae, Carangidae, Scombridae and Clupeidae mainly for white-tailed tropicbirds, sooty terns and brown noddies, Hemiramphidae and Coryphaenidae for white-tailed tropicbirds, and Engraulidae and fish larvae for lesser noddies and Audubon's shearwaters. Despite some inter-annual and seasonal variations in diet, these were not consistent within the community. Responses of the seabird community to an environmental perturbation that negatively affected chick growth and breeding success of lesser noddies emphasised the higher vulnerability of species with smaller foraging ranges and/or with lower ability to switch diet (lesser noddy, roseate tern) compared to less range-restricted and/or more opportunistic ones (white-tailed tropicbird, brown noddy, sooty tern, white tern). Although situated in a tropical region, the food availability in the Seychelles seems to be predictable at a large (annual) temporal scale, but highly unpredictable at a small (intra-seasonal or daily) temporal scale.

Feral cat Felis catus predation on seabirds has been well documented; however, details regarding shifts in feral cat diet in relation to seabird availability, seabird predation rate and impact on seabird population dynamics are scarce. Here, we present data documenting a seasonal shift in feral cat diet at Juan de Nova Island, Mozambique Channel. We also quantify sooty tern Sterna fuscata predation by feral cats and examine the impact on sooty terns over both the short term (by removing individual cats from sub-colonies) and over the longer term by highlighting their influence on population growth rate (l) using a deterministic matrix model. Cat diet shifted dramatically from insects, rats and mice outside the tern breeding season to primarily terns when terns were breeding. The predation rate of sooty terns at Juan de Nova was estimated at 5.94 terns cat À1 day À1 , with a proportion of these (22%) being killed without being consumed ('surplus kills'). When only one cat was removed from each sub-colony, tern predation declined tenfold in the short term. From our matrix model, the annual growth rate for sooty terns was 1.01 in the absence of cat predation. It remained above one until a predation impact equivalent to approximately three times the estimated cat density (12.04 per km 2 ) was incorporated. Our results demonstrate that cats preferentially predate and have an impact on breeding sooty terns at Juan de Nova, and that an increase in cat density could lead to negative effects on population growth, despite the large breeding tern population.

No abstract