Casazza and Ross: Fishes associated with pelagic Sargassum and open water off North Carolina 
36) 
mesopelagic fishes representing three families (Gonosto- 
matidae, Sternoptychidae, Myctophidae) collected with 
Sargassum likely resulted from combinations of their 
upward diel migrations, upwelling (reported from the 
Cape Hatteras study area: Lohrenz et al., 2002; Thomas 
et ah, 2002), or convergent currents bringing them 
into contact with Sargassum, rather than the result 
of attraction to the algae. The associations of many 
fish species with Sargassum appears to be facultative 
(Dooley, 1972; Wells and Rooker, 2004), and all stud- 
ies to date have recorded fishes incidentally associated 
with Sargassum that are normally not considered to 
be structure-associated species. It remains difficult to 
determine exactly why or how some species use this 
habitat and the degree to which it influences their life 
histories. 
Despite methodological differences between the stud- 
ies, patterns of abundance for dominant species collected 
from Sargassum habitat were comparable to those from 
previous collections off North Carolina, Florida, and in 
the Gulf of Mexico. Stephanolepis hispidus dominated 
all Sargassum collections in all areas, followed closely 
by C. cry sos and B. capriscus (Dooley, 1972; Wells and 
Rooker, 2004). Histrio histrio was abundant in Gulf of 
Mexico (Bortone et al., 1977; Wells and Rooker, 2004) 
and Florida east coast (Dooley, 1972) collections with 
Sargassum but was not abundant in collections off 
North Carolina (Dooley, 1972; Moser et al., 1998; this 
study). Because the majority of H. histrio are found in 
the Sargasso Sea and Caribbean Basin (Adams, 1960), 
their lower abundance off North Carolina may repre- 
sent a winnowing of the population with northward 
or westward drift. Dooley (1972) suggested a progres- 
sive decrease in fish species richness from Florida to 
North Carolina and across the Atlantic to the Azores. 
This decrease in species richness may be an artifact of 
limited collections off North Carolina and the Azores 
because our more extensive sampling produced Sargas- 
sam-related species richness exceeding that reported in 
other areas. Although based on limited sampling, data 
indicate that fewer fish species are associated with 
Sargassum habitat in the Sargasso Sea compared with 
Sargassum habitat in the Gulf of Mexico or the Gulf 
Stream (Fine, 1970; Stoner and Greening, 1984). The 
great difference between the Sargasso Sea and U.S. 
continental shelf collections indicates that the majority 
of the fish fauna recruits to Sargassum habitat after 
the algae are entrained into the Loop Current (Gulf of 
Mexico) and the Florida and Gulf Stream currents. 
The structural complexity of habitats strongly affects 
fish assemblages. Our open-water samples contained 
fewer fishes compared with samples containing Sar- 
gassum habitat. Clearly, fishes that use Sargassum 
habitat also are found in open water without Sargas- 
sum, but abundance is heavily skewed toward floating 
structured habitat (Kingsford, 1993). Stephanopelis 
hispidus dominated both habitats but was two orders 
of magnitude more abundant in Sargassum collections. 
Considering this, and that S. hispidus usually occupies 
structured habitat, it seems likely that the S. hispi- 
dus collected from open-water habitat may have been 
displaced by physical disturbance to Sargassum mats, 
or they may have been caught in open water because 
they had strayed away from the preferred habitat. If 
so, an even larger difference exists between open-water 
and Sargassum fish communities. The strong fidelity of 
fishes to floating Sargassum habitat is also illustrated 
by the distinct boundary observed between open-wa- 
ter and Sargassum habitats (Fig. 5F). The open water 
adjacent to the edge of the weedline was unpopulated, 
compared with the area immediately underneath and 
within the Sargassum where a high density of juvenile 
fishes was evident. Higher abundances and diversity of 
fishes in vegetated (versus unvegetated) habitats is a 
common theme (Weinstein et ah, 1977; Orth and Heck, 
1980) that results from increased structural complexity 
(Stoner, 1983). Although fundamental differences exist 
between Sargassum and seagrass ecosystems, fish com- 
munities use the two habitats in similar ways. Both 
habitats are nursery areas for juvenile fishes and sup- 
port diverse and abundant fish communities. Addition- 
ally, the abundance of fishes increases with increasing 
seagrass density (Orth and Heck, 1980; Thayer and 
Chester, 1989) and Sargassum abundance (Moser et al., 
1998; Wells and Rooker, 2004). 
Juvenile fishes may seek drifting objects to improve 
future benthic settlement opportunities (Dempster and 
Kingsford, 2004), thus facilitating early survival and 
eventual recruitment to adult populations. Most of the 
juvenile fishes using Sargassum are species that ulti- 
mately occupy either inshore benthic reef or complex, 
structured habitats (demersal) or the open ocean (pe- 
lagic). However, the length of time juvenile fishes reside 
in Sargassum and the fates of juvenile fishes after leav- 
ing this habitat are unknown. Some fishes remain in 
the Sargassum longer than expected, perhaps because 
they missed a settlement opportunity. This appears to 
be the case for some unusually large juveniles (e.g., 
Hippocampus spp., Mulloidichthys martinicus, Kyphosus 
spp., A. saxatilis, balistids, monacanthids) collected in 
the present study. Caribbean damselfishes, including A. 
saxatilis, settle between 10 and 12 mm SL (Robertson 
et al., 1993), yet A. saxatilis collected in our study from 
Sargassum habitat were 16-29 mm SL. The dominant 
Sargassum -associated fish, S. hispidus, may settle into 
North Carolina estuarine seagrass beds at 11-40 mm 
(Adams, 1976; Ross and Epperly, 1985), well below the 
sizes of some individuals collected offshore in this study. 
The movement of large quantities of Sargassum habi- 
tat across the continental shelf as far as the estuaries 
transports vast numbers of associated juvenile fishes 
toward other habitats (e.g., seagrass beds, reefs) and 
probably facilitates recruitment to adult populations. 
Young fishes entrained in the Gulf Stream that ulti- 
mately have demersal populations, including species us- 
ing Sargassum, have a more uncertain future once they 
drift north of Cape Hatteras where the Gulf Stream 
moves offshore (McBride and Able, 1998; Ross et al., 
2007). Juveniles of demersal species that do not move 
from Sargassum before reaching the Cape Hatteras area 
