262 
Fishery Bulletin 109(3) 
Woude et al., 2006). Blooms in this region often occur 
at cape-and-bay spatial scales and are ephemeral, typi- 
cally lasting several days after an upwelling wind event 
(Largier et al., 2006). 
Coastal topographical features also affect the dis- 
tribution of zooplankton and larval invertebrates in 
this region (Ebert and Russell, 1988; Largier, 2004). 
Localized differences in zooplankton density and com- 
munity structure have been measured across distances 
of 10 km or less in the vicinity of headlands and their 
upwelling shadows (Graham et al., 1992; Wing et al., 
1998; Mace and Morgan, 2006), and across-shelf varia- 
tion in zooplankton (Morgan et al., 2003) and ichthyo- 
plankton (Auth, 2008) community structure is well 
described. A complex field of mesoscale eddies and 
fronts off coastal Oregon in the summer of 2000 cor- 
related spatially with four or five different zooplank- 
ton assemblages, and dynamic water-mass attributes 
influenced by recent strong upwelling and advection 
appear to be the principal factors driving horizontal 
planktonic distribution (Keister et al., 2009). Seasonal 
zooplankton shifts have been linked to intensity of up- 
welling and variation in the timing and delivery rate of 
subarctic and subtropical water onto the shelf (Peterson 
and Miller, 1977; Roesler and Chelton, 1987), and in- 
terannual patterns appear related to El Nino-South- 
ern Oscillation (ENSO) events and other multiyear 
climate cycles that impact the ocean on basin-scales 
(Rebstock, 2003). The physical processes that promote 
rapid growth and patchiness among plankton have 
additional downcurrent trophic effects among krill, 
fishes, seabirds, and marine mammals (Ainley, 1990; 
Croll et al., 2005; Ware and Thomson, 2005; Jahncke 
et al., 2008). 
Assemblage patterns of epipelagic fishes in the CC 
are known mostly from studies in the northern portion. 
Orsi et al. (2007) summarized broad-scale species as- 
sociations from surface trawls in both the California 
and the Alaska Current, in a region spanning 1100 
km of coastline and identified three or four spatially 
distinct assemblages. In a study covering about 400 
km of coastline in Washington and Oregon, Brodeur 
et al. (2005) sampled across-shelf transects and de- 
scribed seven different fish assemblages, some of which 
were spatially distinct. At the finest scales yet exam- 
ined for epipelagic fishes in this region, Emmett et 
al. (2006) found significant differences in abundance 
among years, months, and stations for several common 
predator and forage species in a region influenced heav- 
ily by the Columbia River plume. These and several 
prior studies (summarized in Brodeur et al., 2003) 
highlight the scales of variation found among epipelagic 
nekton in the northern CC, but the area south of the 
Oregon-California border was sampled in only a few 
of these studies. 
The central CC supports vast schools of clupeiform 
fishes that are essential prey items for fish and avian 
predators. Food-web models constructed for the CC eco- 
system are an essential component of ecosystem-based 
fishery management; these models require data from 
surveys of forage fishes and their associates, details of 
the spatial and temporal structure of their populations, 
and measurement of the oceanographic variables that 
drive local abundance (Field et al., 2006; Samhouri 
et al., 2009). This area is also home to several large 
state and federal marine sanctuaries and is one of the 
primary testing grounds for Marine Protected Areas 
(MPAs) in the United States. Although these MPAs 
are designed primarily for the conservation of demersal 
fish and invertebrates, knowledge of pelagic fish habitat 
and community structure could also be used to inform 
decisions on size and placement of MPAs (Reese and 
Brodeur, 2006). 
Our study covered a strip of coastal ocean run- 
ning north-south above the inner continental shelf 
in a region of strong seasonal upwelling off northern 
and central California. The region encompasses sev- 
eral prominent headlands and bays, a small group of 
offshore islands, and the outflow of the largest river 
in the state through the San Francisco Bay. Detailed 
patterns of spatial and temporal community structure 
of epipelagic fishes in this portion of the CC are un- 
described. The objectives of this study were 1) to test 
for regional, seasonal, and interannual patterns of fish 
abundance in catch data from six years of summer and 
fall surface trawl surveys; 2) to identify dominant spe- 
cies associated with patterns of community structure, 
and correlate biotic patterns to a suite of water proper- 
ties that may be influencing fish distribution; and 3) to 
provide a detailed baseline record of species abundance 
and distribution for the region, against which future 
change may be measured, and to provide primary data 
for ecosystem-based management for the California 
Current ecosystem. 
Materials and methods 
The study area encompassed a 185-km strip of coastal 
ocean between Point Arena (38°55'N) and Point San 
Pedro (37°35'N) in northern and central California 
(Fig. 1). Ocean sampling stations were located from 1 to 
39 km offshore at a depth of 18-141 m mostly over the 
inner portion of the continental shelf and largely within 
the boundaries of three national marine sanctuaries 
(Cordell Bank National Marine Sanctuary [NMS], Gulf 
of the Farallones NMS, and Monterey Bay NMS). Fish 
obtained for this study were collected as part of a more 
detailed examination of Chinook salmon (Oncorhynchus 
tshawytscha) growth and energy status during their 
first ocean year (MacFarlane, 2010). Because Chinook 
salmon are thought to remain very close to shore during 
this life-history stage, we chose to extend our nearshore 
north-south coverage as much a possible. This spatial 
arrangement of stations restricted our analysis to tests 
of primarily along-shelf latitudinal patterns, with less 
emphasis on onshore-offshore gradients. 
We divided our study area into two geographic re- 
gions, the north coast (NC) and the Gulf of the Faral- 
lones (GF), separated by a prominent headland at Point 
