350 
Fishery Bulletin 106(4) 
and fishes that swam near the vessel were collected by 
five or six crew members using small mesh (6.4-mm 
mesh) dip nets. Each 30-minute segment of time during 
the drift represented a station. During the nightlight- 
ing sampling, the presence or absence of Sargassum 
within the field of view was recorded, and if present, 
whether the Sargassum was collected in dip nets was 
recorded. Fishes were also opportunistically collected 
with dip nets during daylight when dense aggregations 
of Sargassum were encountered. Limited hook-and-line 
sampling occurred in both Sargassum and open-water 
habitats during the day and at night, and each sam- 
pling period (station) lasted from 15 to 160 minutes. 
One longline set was made in the Cape Hatteras study 
area. The line was about 366-m long and contained 104 
baited hooks that fished within 1-2 m of the surface. 
The set was made at night, lasted for 501 min, and 
drifted for 30 km through open-water habitat. 
In 1999, underwater video was recorded under a large 
Sargassum weedline at two stations off Cape Hatteras, 
North Carolina (Fig. 1). Snorkelers using a handheld 
color camcorder (SONY model DCR-TRV900, New York, 
NY) in a waterproof case swam at and just below (<3 m) 
the surface along the edge of and under the weedline. A 
total of 62 minutes of video footage was recorded dur- 
ing the two stations. Analyses of the underwater video 
footage included identification of species, documentation 
of behaviors, and placement of fishes within or below 
the weedline. 
Specimens were preserved at sea in 10% formalin- 
seawater solution and later stored in 40% isopropanol. 
Larval fishes had been collected in previous Sargassum 
studies, and this fact implied an association with this 
habitat. However, because distributions of pelagic fish 
larvae are highly influenced by currents and they gener- 
ally lack affinity for drift algae (Kingsford and Choat, 
1985), their presence in Sargassum collections (Settle, 
1993; Wells and Rooker, 2004) is probably coincidental. 
For this reason and because the neuston net mesh size 
was inappropriate for sampling larvae, larval fishes 
(classified according to Richards, 2006) were excluded 
from this study. Fishes were identified to the lowest pos- 
sible taxon, counted, measured to the nearest mm for 
standard length (SL), and weighed (wet weight) to the 
nearest 0.1 g. Damage to some fishes precluded identi- 
fication to species and SL measurements. When more 
than 500 individuals of the same species were collected 
in a tow, a subsample (approximately 10% of the catch) 
was measured for SL and wet weight. 
Data analysis 
Fish catches from neuston nets were analyzed statisti- 
cally to assess differences in fish community structure 
between habitats, and diurnal differences in community 
structure. Neuston tows without Sargassum were des- 
ignated as open-water (OW). Because clumps of algae 
as small as 0.005 kg could influence the distribution 
and abundance of fishes (Kingsford and Choat, 1985), 
samples were classified as Sargassum (S) if algae were 
collected, regardless of the quantity. The number of 
individuals and number of species collected from Sargas- 
sum and open-water habitats were log (x+1) transformed 
before analysis to correct for heterogeneity of variance, 
to reduce the influence of abundant species, and to 
enhance the contribution of rare species. If the assump- 
tions of homogeneity of variance and normality were not 
satisfied after data transformation, a nonparametric 
Mann-Whitney test was applied to determine whether 
there were differences in the number of individuals 
and species in Sargassum versus open-water habitat. A 
Kruskal-Wallis test was used to compare day and night 
fish catches from neuston nets within and across station 
types (i.e., S versus OW), and a Dunn’s multiple compari- 
son test was used to determine where significant differ- 
ences occurred. The relationship of fish abundance and 
species richness to the quantity of Sargassum collected 
by neuston nets was evaluated with regression analysis. 
Length-frequency distributions for dominant species col- 
lected from Sargassum habitat were compared to the size 
structures of the same species collected from open-water 
habitat by using a Kolmogorov-Smirnov test. 
Habitat type sampled (S versus OW) was also des- 
ignated for the supplemental methods. If Sargassum 
was collected by dip net, the station was designated 
as S; otherwise it was OW. Likewise, if the hook-and- 
line gear was placed in Sargassum (S), the catch was 
designated as S; if the gear was placed in unvegetated 
habitat, the catch was designated as OW. 
Results 
Catch composition 
For all methods and cruises combined, most fishes were 
collected in samples containing Sargassum habitat. A 
total of 18,799 fishes, representing 80 species from 28 
families, were collected in 162 Sargassum samples, and 
a total of 2706 fishes, representing 60 species from 23 
families, were collected in 80 open-water samples (Fig. 1; 
Table 1). Both Sargassum and open-water collections 
were dominated by the families Monacanthidae (75% of 
S, 45% of OW), Carangidae (13%, 21%), and Exocoetidae 
(6%, 19%). Individuals of nine species represented 93% 
of the total Sargassum catch (in decreasing order of 
abundance): Stephanolepis hispidus (planehead filefish), 
Caranx crysos (blue runner), Cheilopogon melanurus 
(Atlantic flyingfish), Balistes capriscus (gray triggerfish), 
Seriola rivoliana (almaco jack), Parexocoetus brachyp- 
terus (sailfin flyingfish), Monacanthus ciliatus (fringed 
filefish), Decapterus punctatus (round scad), and Cory- 
phaena hippurus (dolphinfish). Individuals of 10 spe- 
cies represented 92% of the total open-water catch (in 
decreasing order of abundance): S. hispidus, C. crysos, 
Clupea harengus (Atlantic herring) (all from a single 
station), C. melanurus, P. brachypterus, D. punctatus, 
Prognichthys occidentalis (bluntnose flyingfish), Oxypor- 
hamphus micropterus (smallwing flyingfish), Istiophorus 
platypterus (sailfish), and C. hippurus. For all methods 
