Food Habits

Main Foods Taken

Variety of fish, particularly (along East Coast) killifish (Fundulus spp.), herrings, pipefish (Sygnathus sp.) (Bent 1921, Burger and Gochfeld 1990). Also may take small crustaceans (Tomkins 1933, Leavitt 1957).

Microhabitat For Foraging

Mainly tidal waters of bays, estuaries, lagoons, rivers, and salt marsh pools, creeks, and ditches; such habitats concentrate small fish (Valiela 1984). In Virginia, 88% fed in salt marsh tidal pools (Erwin 1977b). Species feeds preferentially under low wind conditions when water surface smooth. On wintering grounds in Florida, regularly forages in water < 20 cm deep within 2 m of land (Black and Harris 1983) and sometimes in water < 2.5 cm deep (Tomkins 1951).

Foraging distance from colonies variable. On Long Island, NY, main feeding areas ≤ 8 km from colony. In Georgia, fed 5.2 km from a colony (Tomkins 1951). Colony sites often near inlets which may reflect access to feeding areas as well as suitable substrate.

Food Capture And Consumption

Reportedly a nocturnal forager (Bent 1921, Erwin 1977a, Burger and Gochfeld 1990), but feeds regularly in daytime at appropriate tide cycle, especially when feeding young. In Florida, nocturnal feeding often not prominent (Black and Harris 1981).

A tactile feeder. Although Wilson accurately described the feeding behavior in 1810 (Brewer 1839), and Bent (1921) described the capture of fish by skimming, some authors (see Murphy 1936) have considered skimming unrelated to feeding. Skimmers typically glide low over the water, usually on motionless wings, with lower mandible tip slicing surface of water (Fig. 2). When the mandible contacts a fish or other object, the maxilla clamps down while the head and neck tuck downward, securing the fish, which is then turned and swallowed headfirst or carried crosswise in the bill to the nest; see Zusi (1962) for detailed analysis of the functional morphology and adaptations of the bill, head, and neck. Reports of feeding while standing or plunging bill into water (Arthur 1921, Stone 1921) are either rare or erroneous. No other method of feeding documented.

Usually do not feed in flocks; > 60% of Virginia observations are of a solitary bird; only 6% had ≥ 4 birds (Erwin 1977a, b). In New Jersey, 2 or more birds often feed in tandem (MG, JB) (Fig. 2). Diurnal feeding cycle varies with tide and moon cycles (Bent 1921, Burger 1982b). Most fish caught at low tide and again 2 h before high tide, without the obvious early morning feeding peak shown by terns in the same colonies (Burger and Gochfeld 1990). Black and Harris (1981) describe “tidal feeding” when birds forage briefly over a wide area, mainly about 2 h after low tide, while “flurry feeding” occurs when a mid-ebb tide coincides with dusk, and many birds concentrate in a short area for nearly an hour. Erwin (1977b) found a low coefficient of variation in the temporal distribution of feedings at the nest, suggesting that skimmers feed on a spatially uniform food resource.

Fishing rates at Cayo Atascosa, TX, ranged from 1 fish/50 s skimming time (Zusi 1959) to 1/6.2 min (Davis 1951); in Virginia 1/5.0 min (Erwin 1977b).

Piracy

On Long Island, NY, skimmers rarely attempt aerial piracy, but they are occasionally subject to piracy by terns. Seasonality and fish size influence piracy attempts; at Cedar Beach, NY, 0.1% of fish carried by skimmers were pirated compared with 16% of fish carried by Common Terns (Sterna hirundo) at same time (Burger and Gochfeld 1990). Piracy by terns sometimes occurs when skimmers attempt to feed very young chicks. In Virginia, Forster’s Terns (S. forsteri) sometimes attempt piracy (Bent 1921). At Laguna Atascosa, TX, skimmers lost about 20% of the fish they caught to persistently piratic Laughing Gulls (Larus atricilla) (Zusi 1958).

Major Food Items

Remarkably little specific information; mostly small fish in range of 3–12 cm length (rarely up to 20 cm). Mean weight of 41 fish captured: 4.3 g (Erwin 1977b). In Guyana, nonbreeding skimmers (probably cinerascens) took fish mainly between 67% and 133% of bill length (Erwin 1990). In Florida, fish ranged from 2.1–5.5 cm (Leavitt 1957). In Florida, 12 mullet brought to chicks averaged 8.9 cm and 8.4 g; 7 Fundulus, 7.7 cm and 7.3 g.

Most authors (beginning with Bent 1921) mention shrimp as occasional prey, but we find no primary documentation for this in the ne. U.S.; perhaps confined to the Southeast. Young on Long Island, NY, fed mainly on Fundulus spp., although some parents bring back mainly pipefish, which are often rejected by chicks. Wilson (Brewer 1839) found mainly “silversides” (presumably Menidia sp.) in stomachs. In Virginia, fed small fish of which over 90% were Menidia spp. or Fundulus spp.; other prey include Anchoa mitchelli, Leistomus xanthurus, and Pomatomus saltatrix (Erwin 1977a). In Georgia, shrimp (Penaeus sp.), Mugil spp., and Brevoortia (Tomkins 1933). In Florida, 67 items regurgitated by chicks included Mugil cephalus (73%), Fundulus heteroclitus, Brevoortia tyrannus, Paralichthys spp., Elops saurus, and Echgeneis naucrates (Loftin 1982). In Louisiana young fed only fish including Cynoscion nothus, Atherina sp., Mugil sp., Scomberomorus maculatus, Pomatomus saltatrix (Arthur 1921). Salton Sea skimmers brought back sailfin mollies (Poecilia latipinna) (Grant and Hogg 1976).

Quantitative Analysis

Of 10 birds collected at Merritt Is., FL (Feb 1954), 6 stomachs contained fish and shrimp (Palaemonetes sp.), 4 contained only fish including Lutjanus, Fundulus, and a needlefish (Tylosaurus sp.) (Leavitt 1957). In Alabama, 4 of 10 stomachs contained small fish; the rest held nothing (Howell 1928). In Texas, King (1989) identified 500 food items, including 18 fish and 2 invertebrate species, by ligaturing chicks and finding discarded fish at nest; tidewater silversides (Menidia peninsulae) made up 25% of sample by frequency.

Modal size of fish about 8 cm (range 4–12 cm). Fish < 2 cm fed to newly hatched chicks (Burger and Gochfeld 1990). Males and females have different-sized bills, but only Quinn (1990) has documented differential prey selection. Probably take fish on as-available basis, depending on where they are foraging and what is within 3 cm of surface. Compared with co-nesting terns, bring back many more pipefish and more large killifish.

No data from wild birds. Ground effect decreases drag and provides lift to birds skimming close to the water surface (Withers and Timko 1977, Salcedo and Blake 1988). Blake (1985) provides models for skimmer flight energetics assuming that skimmers consume average prey of 5 g (at 1.49 kcal/g) yields an energetic value of 30 kJ. Estimated energy content of a skimmer egg 500 kJ.

No data. Lipid content of carcasses 4–14% (Blake 1985); for thermoregulatory behavior see Behavior: self-maintenance below.

No quantitative data. Adults drink periodically while feeding. Young obtain water from food. Adults walk away from nest to defecate. Young often defecate when handled. Pellets very rarely found in Northeast (JB, MG).

Anterior end of the esophagus is highly muscularized and forms a pseudosphincter that very likely prevents the swallowing of water while birds are skimming (MG unpubl. data). Unlike terns and gulls, captured young virtually never regurgitate, perhaps owing to this sphincter.