68 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 124. No. 1. March 2012 
n = 2), Caspian Tern (Hydroprogne caspia, n = 2), 
Forster’s Tem ( S.forsteri , n — 3), Gull-billed Tern 
(Gelochelidon nilotion, n — 3), Least Tem (Sternala 
cintillarnm. n = 1). Sandwich Tem (Thalctsseus 
sandvicensis , n = 8), Sooty Tem (n = II), and 
Brown Noddy (Anous stolid us. n = 2). 
Four species of seabirds had frequencies of 
Sargassum-associated prey >25% and were 
considered Sargassum specialists (Table I). The 
single Sabine’s Gull (Xenia sabini) sampled 
contained a seahorse (Hippocampus sp.), which 
is a Sargassutn dweller. These gulls were 
observed following Sargassutn drift lines during 
migration but. sample size did not support 
including this gull in the Sargassutn specialists 
category. The four specialists contained almost 
exclusively Sargassutn -associated prey, as evi¬ 
denced by high volumes of identifiable prey items 
in their digestive tracts (Table 2). Sargassutn 
users, birds that contained Sargassutn prey less 
frequently (7 species), also Contained high 
volumes ot Sargownw-associated prey. 
Most birds feeding in Sargassutn contained 
small Sargassutn -associated fishes (Table 2). The 
only exceptions were the two species of phala- 
ropes, which had consumed Sargassum shrimp 
(Latreutes fucorunt) and the Sargassutn- associat¬ 
ed gastropod. Litiopa inelanosioma. This gastro¬ 
pod was also found in Cory’s Shearwaters 
(Calonectris diomedea). Specialists fed on a 
minimum of seven Sargassum-dssocialed fish 
species, and most measurable fish prey were 
<50 mm standard length (Table 2). However. 
Royal Terns (Thalasseus maximus) generally 
consumed slightly larger fish (range^ = 40- 
105 mm) than the other birds we examined (range 
- 6-75 mm). Filefish (Monacanthus sp.) occurred 
with the highest frequency in Audubon’s Shear¬ 
waters (Puffinns Iherminieri ), Bridled Terns 
(Onychoprion anaeihetus), and Royal Terns, but 
numerical abundance of filefish was highest only 
in Audubon's Shearwaters and Royal Terns 
(Fig. 1A. C). The mean number of filefish per 
bird was five. Bridled Terns (Fig. IB) had the 
highest number of tetradontids (mean number of 
puffers/bird = 6) and also consumed large 
numbers of ostraciids (mean number of trunk- 
fish/bird = 4). 
DISCUSSION 
Gut content analysis identified four Sareasv, 
spec.ahsts, while visual observations indicat 
ha. several add.tional species target this habi 
for feeding. Audubon’s Shearwater. Royal Tem. 
Bridled Tern, and Red-necked Phalarope iPhahtr- 
opus lobatus) contained relatively high levels a 
Sargassutn -associated prey. The single Sabine- 
Gull examined contained a 5argaw«m-associaied 
prey item and this species was observed to follov 
Sargassutn drift lines. Visual observations indi¬ 
cated Bridled Terns regularly associated with 
Sargassutn patches and tended to use the mats and 
associated flotsam as roosting sites (Duncan and 
Harvard 1980, Haney 1986). Our observational 
data also indicated that Masked Boobies I Sulo 
dactylatra) and the two species of tropicbinh 
target Sargassutn patches while feeding as also 
reported by Haney et al. (1999). 
Diet analysis underestimated prey from Sargas- 
sum habitat and excluded some seabird species. For 
example, the digestive tracts of Bridled Terns in our 
study contained insects of terrestrial origin, which 
we did not consider to be Sargassutn associates (5 
consumed Lepidoptera, 6 ate Coleoptera. 2 ale 
Hymenoptera. and 6 contained unidentified in¬ 
sects). Haney ef al. (1999) reported insects were (lie 
second most common food item in Bridled Temv 
These insects may have been resting on Sargassutn 
mats when ingested, We commonly observed both 
species of Atlantic tropiebirds feeding around 
Sargassutn reefs. Flying fish were recovered from 
their digestive tracts, but this was not direct 
evidence of Sargassum use. as flying fish regularly 
occur where Sargassum is absent (Casa/za and 
Ross 2008). We observed tropiebirds as they 
plunged near and sometimes under the Sargassum 
The same was true of Masked Boobies, a species 
seen infrequently off the Carolinas but usually tn 
association with Sargassum. 
Some seabird species may not feed directly over 
Sargassum , but the alga is critical habitat Ion- 
certain life stages of their fish prey. For example 
(lying fish can represent >50% of the total diet 
ot the two tropiebirds collected off North Carolina 
(Lee et al. 1981, Lee and Irvin 1983 1 and Hying fish 
are important prey in other parts of their range (Lee 
and Walsh-McGchee 1998). Flying fish use 
Sargassum for spawning and rearing, and Sargas- 
sum is essential habitat for these and other fish 
species (Casa/za and Ross 2008). Thus. Sargassum 
contributes indirectly to the fitness of tropiebirds 
and other seabirds, notably Sooty Terns and 
Masked Boobies. 
Sargassum specialists used a variety of foraging 
modes, including surface-seizing, plunging (aerial 
diving), pursuit plunging/diving, pattering, and 
