INTRODUCTION

Large-scale variations in seabird distributions are mainly caused by environmental heterogeneity resulting from physical oceanographic processes (Haney 1987) and pervasive anthropogenic disturbance (Jackson et al. 2001, Camphuysen 2005). However, patchiness in seabird distribution may also occur at smaller scales due to species-specific responses to the environment (e.g. Garthe 1997, Wanless et al. 1997, Boyd et al. 2006). Likewise, the presence of surface features (e.g. floating wood, debris, seaweed, buoys, and fronts) may determine seabird distribution as they provide resting places and can temporarily increase available food sources, to which seabirds can quickly respond. Haney (1987), for example, described patchiness due to the visible surface manifestations (slicks alternating with ripples of rough water) of internal waves. The seabird patchiness was probably due to qualitative differences in prey composition as most birds were actively foraging or feeding. Similar observations were done at fronts (e.g. Haney 1985, Balance & Pitman 1999, Spear et al. 2001), due to the locally elevated level of prey biomass. Especially the distribution of phalaropes (mostly feeding or resting) has been linked to the presence of large and persistent oceanic fronts and other oceanographic features that concentrate Zooplankton at the surface (e.g. Haney 1985 & 1986, Lee 1987, Brown & Gaskin 1988).

On an even smaller scale, a wide variety of floating objects have been reported to attract seabirds: Cadée (2002) reports on peckmarks on and ingestion of debris like plastic, styrofoam and cuttlebones in the North Sea; Arcos (2000) observed an alternative feeding strategy of Balearic Shearwaters Puffinus mauretanicus involving capture of fish under floating objects; and floating seaweeds in the Bay of Fundy (Canada) are shown to attract seabirds such as phalaropes, gulls and terns (Parsons 1986, Huettmann pers. comm.).

The influence of floating seaweed patches on the distribution and behaviour of seabirds formed the key issue of the present study. Floating seaweeds, both the permanently floating Sargassum and ephemeral patches composed of different species, are shown to teem with small marine animals seeking food and refuge, including crustaceans and fish (e.g. Fine 1970, Tully & O'Ceidigh 1986, Coston-Clements et al. 1991, Ingolfsson 1995, Kingsford 1995, Vandendriessche et al. 2006a, Vandendriessche et al., 2007). The increased biomass in invertebrates and fish compared to the surrounding water column may constitute an important, more or less predictable source of extra food, although probably exploited in an opportunistic way (Arcos 2000). In the Sea of Okhotsk, for example, Dunlins Calidris alpina were seen feeding on rafts of floating seaweed, probably taking snails and insects (Huettmann pers. comm.). Furthermore, floating seaweeds could play an important role by signalling suitable feeding areas to birds since they tend to accumulate in biologically rich water masses (Arcos 2000).

From the observations listed above, it is clear that seabirds are attracted to surface phenomena like floating seaweeds. Other than a few studies concerning Sargassum and some sporadic notes, however, few investigations have been done on the topic of seabirds associated with floating seaweeds. As a result, the present study aims to examine the seaweed's possible attractions for seabirds in Northeast Atlantic waters, based on the European Seabirds At Sea database (ESAS). The main research questions of this study are: “Are there seabirds that are frequently seen associated with ephemeral patches of floating seaweed?”, and “Are these associations feeding mode-dependent?”.

MATERIALS AND METHODS

The seabird data used in the analyses were extracted from the European Seabirds At Sea database, which is composed of seabird observations collected and coded using standardised survey techniques (Tasker et al. 1984, Camphuysen et al. 2004). The database was established in the early 1980s and contains results of ship-based and aerial seabird surveys in Northwest European waters (Camphuysen & Garthe 2004). Of special interest for this study is the use of standardised coding of behaviour types in the database. The coding system, for example, introduced specific coding of associations of birds with certain surface phenomena and emphasises on feeding behaviour and foraging interactions. Codes have been devised for birds associating with near-surface fish shoals or marine mammals, with floating objects such as wood, rubbish, oil slicks and seaweeds, and with fronts, buoys, markers, vessels, offshore installations, sea-ice or land. The birds are further described according to behaviour, such as flying towards the surface phenomenon, scavenging, searching for prey, feeding, or resting. Furthermore, the description of the foraging behaviour is detailed and distinguishes between 20 behavioural codes like holding fish, aerial pursuit, scavenging at fishing vessel, surface pecking, and actively searching. The great benefit of detailed behavioural coding is that it provides insight in potential correlations between seabird presence and oceanographic or other factors driving prey, and that it allows discrimination between real associations and coincidental observations (Camphuysen & Garthe 2004).

The used data originated from the period 1979–2000. Detailed association codes were only recorded in < 1% of all records. Because the coding system was not always used at the same level of detail (e.g. a large proportion of the records only distinguished between ‘associated with fish shoals’, ‘approaching observation base’ and ‘pattering’), a quantitative analysis of the importance of floating seaweeds was biased. The lack of metadata concerning the level of detail in the use of the coding system prohibited a straightforward restriction of the data. However, as the bias depended on the observers and circumstances per trip, it was assumed that the bias was straightforwardly applied to the entire dataset, enabling comparisons between species. Consequently, the numbers resulting from the analyses can be used for interspecies comparison, but they do not give reliable estimates of real seaweed-associated densities.

Although the database contained data about all Northwest European waters, the data about seaweed-associations in seabirds were concentrated in the North Sea. Therefore, only the seabird data from that region (59°17′–51°26′N and 2°47′W–6°58′E, see Fig. 1) were extracted and used in the analyses. Only data of common seabirds were included; rare seabird species (less than 300 records) and terrestrial birds were excluded. All abundances of seabirds were expressed as frequencies of occurrences to correct for aggregation behaviour.

Seabird species were grouped a priori according to foraging behaviour based on literature (Cramp et al. 1978–1997). The non-parametric Kruskal-Wallis test was applied to test for differences between multiple independent groups. Pairwise tests were done using the multiple comparisons procedure (Conover 1971).

RESULTS

Floating seaweed accounted for 2% of all observations of surface phenomena (Fig. 2A; large manmade objects like buoys, platforms and vessels not taken into account) and for 4% of all seabird counts in association with these phenomena. These percentages, however, are likely to be underestimated because floating seaweed is often an important constituent of patches of floating matter and lines in sea, and because floating seaweeds often converge at fronts.

Figure 1.

Map indicating the observation points (n = 212, dots partly hidden) of seabirds associated with floating seaweed and the delimitation of the study area.

The most common visitors of floating seaweed patches were Guillemots Uria aalge, Common Scoters Melanitta nigra, Lesser Black-backed Gulls Larus fuscus and Kittiwakes Rissa tridactyla, Gannets Sula bassana, Razorbills Alca torda, Eiders Somateria mollisima, Sandwich Terns Sterna sandvicensis and Common Terns Sterna hirundo (Fig. 2B). Occasional visitors included other gulls (Herring Gulls L. argentatus, Common Gulls L. canus, Greater Black-backed Gulls L. marinus, Black-headed Gulls L. ridibundus) and Arctic Terns S. paradisaea, Fulmars Fulmarus glacialis, Greater Skuas Stercorarius skua, Cormorants Phalacrocorax carbo and Red-breasted Mergansers Mergus serrator.

Figure 2.

(A) Pie chart of relative importance of surface phenomena, based on frequencies of occurrences (n = 24 845). (B) Pie chart showing the top-10 observation frequencies of seabirds in association with floating seaweed (n = 212).

Figure 3.

Bar chart showing the percentage of the occurrences during which seabirds were associated with floating seaweed. Arrows indicate foraging habits of the five groups. A Kruskal-Wallis test showed significant differences between the groups (df = 4, P = 0.03); a Multiple Comparisons test indicated differences between groups 2 and 3, and 2 and 5 (P < 0.05).

In order to quantify the importance of floating seaweeds to seabirds, the percentages of observations that seabirds were observed with floating seaweed were calculated per species (Fig. 3). These percentages differed among species, depending on the mode of foraging (groups indicated in Fig. 3). The results indicate that surface feeding species that make shallow dives (terns and Red-breasted Mergansers) benefit most from the presence of floating seaweeds and their associated macro- and ichthyofauna (on average 0.49% of occurrences associated with seaweed). Species hunting for pelagic and bottom-dwelling prey (divers, Guillemots, Razorbills, Puffins Fratercula arctica, Gannets and Cormorants; together 0.03%), and especially benthos feeders (Common Scoters and Eiders; 0.21%) were frequently seen in association with floating seaweeds, while opportunists and scavengers like gulls and skuas were recorded on few occasions (0.02%). Finally, petrels and shearwaters (surface-seizing, pursuitplunging, pursuit-diving) were seldomly seen in association with floating seaweeds (<0.001%).

The most common behavioural activities of the birds associated with floating seaweed were found to be surface pecking and actively searching (mainly gulls and terns), and pursuit plunging (mainly Cormorants) (Table 1). At the few occasions that activities of Razorbills, Fulmars, seaducks and Guillemots were recorded, they were mostly pursuit-plunging or actively searching.

Table 1.

Behavioural activities of birds associated with floating seaweeds, expressed as percentage of total observations per species.

Because foraging behaviour in terns was regularly described, we compared their behaviour around seaweeds to their behaviour outside seaweed patches (Fig. 4). The three tern species were most commonly associated with fish shoals (up to 20%) and their top-3 foraging activities varied when comparing seaweed-associated birds with the rest of the observed birds. Arctic Tern was mainly seen dipping and surface pecking in the vicinity of floating seaweeds, which is similar to other cases, in which they were mainly seen surface pecking, actively searching and dipping. Foraging behaviour was quite similar in Sandwich Tern as well: in both cases the main activities were actively searching and deep plunging. Considerable differences, however, could be observed in Common Tern, which was mainly seen surface pecking and dipping around floating seaweeds, but was actively searching, pursuit diving or scavenging in most other cases. These results indicate that, especially in the case of Common Tern, the presence of floating seaweed patches may influence foraging behaviour and therefore also prey choice.

Figure 4.

Comparison of foraging behaviour at floating seaweed vs. outside floating seaweed in three tern species.

DISCUSSION