The study of prey and prey size selection helps us better understand ecological interactions between predator and prey, and their possible influence on the structuring of food webs and prey communities (Pokharel 2020). The decision of what and where to forage is guided theoretically by the strategy of optimizing energy acquisition by unit of time, as analysed by optimal foraging theory (Fargallo et al. 2020, Strandmark 2024). However, fitness factors other than the balance between energy acquisition and costs may also play a role, as will be discussed below.

Oystercatchers are known to be quite plastic regarding prey type choice, consuming the most abundant prey at each coastal site (Tjørve and Tjørve 2010). In fact, when they colonize inland territories far from the coast, their diet can shift to terrestrial prey (van de Pol et al. 2014). Additionally, the selection of prey items by oystercatchers may also be influenced by parameters that determine the cost/benefit ratio, such as size, thickness of the valve, pulp content, presence of barnacles on the valves and seasonal differences in available sizes (Hilgerloh and Pfeifer 2002, Rossignol et al. 2011). Many studies have approached prey size selection by the Eurasian oystercatcher (see e.g. Cayford and Goss-Custard 1990, Zwarts et al. 1996), reporting a wide range of results depending on the physical and ecological characteristics of each study site. However, Hilgerloh and Pfeifer (2002) suggested the existence of a preferred mussel size for oystercatchers, so that if the size class which seemingly optimizes the benefit/cost relationship is not available, oystercatchers would then try to forage on mussels as close in size to it as possible, if costs or time constraints are not too demanding.

The Iberian Peninsula represents the southernmost limit of the distribution of Eurasian oystercatchers (Haematopus ostralegus) in the western Palearctic, along with the populations of the Greek and Turkish peninsulas. Populations located at the limit of their distribution range may have different vital rates and foraging ecology to those of populations located closer to the centre of the distribution range of the species, tending to be more vulnerable to environmental stochasticity due to their small size, and to local extinction due to a poor rescue effect when distance to population cores is great (Van Schmidt and Beissinger 2020). Hence, the study of edge populations should be considered a conservation priority (see e.g. Martínez-Abraín et al. 2019A, 2023) despite the overall conservation status of the species, which in Europe is Near Threatened in the case of H. ostralegus (BirdLife International 2019).

In the Iberian Peninsula, oystercatcher populations have experienced a decline in the northeast (Ebro Delta), showing a 4% annual decrease during the period 2001-2020. On the other hand, northwestern Iberian populations (Galicia) have progressively increased (Rías Baixas) or remained stable (Lugo coast) over the last few decades (Mouriño et al. 2021). The oldest colonies in NW Spain are located in abrupt coastal islets where nesting oystercatchers have found an ecological refuge in the last few decades (Mouriño et al. 2021). However, human rural flight during the last 60-70 years has allowed oystercatchers to colonize or recolonize small flat islets (Fig. 1). These sites are easily accessible by humans and remained unoccupied by gulls and oystercatchers until recently due to permanent human disturbance (see Martínez-Abraín et al. 2019b, 2021). They constitute a suboptimal alternative to beaches and dunes, which are locally occupied by people and thus largely unavailable to wildlife, unlike the Ebro Delta colony in NE Spain where beaches and dunes are protected (Fig. 1). This forced nest site selection (anthropogenic forcing) is expected a priori to influence foraging behaviour due to factors unrelated to foraging preference or edge-of-distribution-related effects (ecological forcing), a fact that is often overlooked in studies of prey and prey size selection.

Here we analyse for the first time whether the expanding breeding populations of Eurasian oystercatcher from NW Spain show a preference for any prey type or prey size (for chick feeding) on three small flat islets in which the species has nested syntopically with yellow-legged gulls (Larus michahellis) during the last few decades. Based on current knowledge, our a priori expectations were that oystercatchers would show high foraging plasticity, with no selection of prey type but a tendency to select larger, more profitable prey sizes.

Based on our results, we cannot conclude the existence of clear positive or negative selection of one prey type over another. Although the values of the Savage index pointed to some preference for foraging on mussels over limpets, the null selection hypothesis was only statistically rejected for limpets on Pedregoso (negative selection). These results coincided with what was expected a priori, because previous studies found that oystercatchers choose the prey types that are most abundant within breeding territories (e.g. Tjørve and Tjørve 2010). On Farallóns, we were unable to study prey selection because mussels were not available in the colony or present in shell piles. All consumed prey were limpets. Since shells remain in the shell piles for a long time, it is unlikely that the absence of valves in the foraging ground means that oystercatchers had already consumed all mussels available and that our results have validity only for a short time window. Additionally, we recorded the presence of periwinkles (Littorina sp.) on Farallóns (in seven out of the ten squares sampled in the foraging grounds), but this mollusc was not found in shell piles, suggesting a negative selection of this species, which would be consistent with the findings reported by Tjørve and Tjørve (2010) in Norway. However, this species was not included in our study because it was not present in shell piles or sampling squares on Areoso and Pedregoso that were sampled prior to Farallóns.

In summary, the results of our study showed heterogeneity among sites in prey type, supporting the idea that oystercatchers may act as facultative specialists, foraging on the most common prey at each site without showing a clear preference for one prey type or another. Only the negative selection of limpets on Pedregoso suggested that oystercatchers may forage on limpets in a lower proportion than expected by its abundance in the local environment. This foraging plasticity has applied conservation consequences because it is easier for facultative specialists to expand and colonize new sites, as this is actually happening with this species in Atlantic Iberia.

Regarding prey size selection, our analyses detected selection for larger sizes in most cases: mussels and limpets on Areoso and limpets on Farallóns, consistently with the findings of most studies on prey size selection by the Eurasian oystercatcher (Drinnan 1958, Ens et al. 1992). However, Leopold et al. (1996) found that selection for the most profitable prey could increase chick predation risk, because adults had to remain longer away from colonies and, hence, they did not select for large prey size unless they had to travel long distances. Our study oystercatchers nested associated with yellow-legged gulls (265 and 88 gull pairs on Pedregoso and Areoso, respectively, in 2022; Dirección Xeral de Patrimonio Natural 2024) and great black-backed gulls (Larus marinus) (1 and 4 pairs on Pedregoso and Areoso, respectively, in 2023; J. Mouriño own unpublished data), all of them facultative oystercatcher predators (Tjørve and Tjørve 2010). Nevertheless, we found that oystercatchers selected for larger prey to feed chicks, meaning that their foraging grounds had to be close to nesting sites (i.e. in exposed rocks at low tide around nesting areas, as previously observed by us), and thus that they were away from the proximity of chicks for short time periods. Moreover, our field observations of birds foraging in the Coelleira, Ansarón and Gaveira de Viveiro colonies (X.M. unpublished) support the need to minimize the risk of chick predation because the two adults of each nesting pair do not forage at the same time, but one of them remains in the vicinity of chicks until the second brings food for the chicks.

Additionally, we also detected selection for small size of limpets on Pedregoso but were unable to show selection for size of mussels there. The selection for smaller limpets could be due to lack of independence of the limpets measured, as a large number of them came from the same quadrat (for example, individuals could be affected by density dependence or be closely related to each other). Alternatively, this result could also be a consequence of adult oystercatchers foraging farther from colonies than usual, trying to minimize the time spent away from chicks. The lack of selection for mussels could simply be an artefact caused by low sample size (n=17 mussels consumed and n=8 available). Selection for smaller prey has only been previously found in studies in which the Eurasian oystercatcher’s target prey was the common cockle (Cerastoderma edulis). However, the closely related African black oystercatcher (H. moquini) showed lack of discrimination of clam (Donax serra) size; there was temporal segregation of small and large clams but, even when large clams were abundant they took both small and large clams (Ward 1991). Moreover, if oystercatchers forage mainly on mussels from the upper limit of their size class distribution, they could also influence the composition of mussel size classes available in the long run, so only smaller size classes would be available (i.e. self-trophic downgrading regarding prey size occurred; Hamilton 2000). A similar effect is known to be caused by the anthropogenic preference for large prey in hunting and fishing (Soga and Gaston 2018). Prey of different sizes present different profitability for the predator, depending on the time required for prey handling and the meat content. As stated in the introduction, Hilgerloh and Pfeifer (2002) suggested that oystercatchers had an optimal median mussel size of 51 mm, and that when the mussels available were smaller than the ideal size, they selected the largest ones available. Interestingly, in our study the overall median length of consumed mussels was 46.86 mm (48.14±7.85; arithmetic mean±SD), close to 51.

As the study oystercatcher population is currently growing, with new colonies being established every year, further studies should be carried out in the future to confirm or amend these preliminary findings. Likely human influence on mussel availability due to collection of juvenile mussels for industrial mussel growth should also be explored within the framework of future plans for the conservation of this expanding oystercatcher population. This problem would be less important if oystercatchers could have access to protected mainland beaches and dune fields in the future for nesting and forage on different prey types (e.g. soft prey from sandy substrates).