274 
et al., 1991). Forage and predatory fishes, seabirds, and 
marine mammals actively seek out and concentrate at 
fronts in order to feed (Olson et al., 1994; Sims and 
Quayle, 1998). Nearshore frontal probability, measured 
in fine-scale (5 km) bins, during the upwelling season in 
2006 was much higher in the GF than elsewhere along 
the northern California coast (Woodson et al. 3 ) — fur- 
ther evidence of the unique oceanographic properties 
of the GF and a possible reason for the much higher 
density of clupeiform fishes that we recorded there, 
than along the NC. Wing et al. (1998) described the 
spatial pattern of Gulf mesoscale fronts in 1994 and 
1995 and collected several taxa of larval crabs and 
rockfish whose distributions in the GF and along the 
north side of Pt. Reyes were strongly correlated with 
specific water masses, water movement and locations 
of fronts. The highest abundance of larval crabs was 
found in the northern GF retention zone and the high- 
est abundance of rockfish larvae along fronts farther 
offshore. The catch variability and higher average fish 
densities that we observed in six years of trawling in 
the GF is consistent with this view of the region as a 
complex frontal mixing and retention zone downcurrent 
of a major headland and also indicates that adjacent 
coastal regions may be less complex hydrographically 
and thus more predictable biologically. With different 
water masses impinging on the Gulf from all four sides 
and shifting frontal boundaries in different years and 
seasons, a more variable pelagic community would be 
expected here than elsewhere. 
Conclusions 
In this study we provide the first detailed baseline data 
on epipelagic fish abundance in a highly dynamic yet 
less studied portion of the CC. As such it can be used 
to evaluate the effects of future perturbations, such as 
climate-induced oceanographic changes or variation in 
fishing pressure, on fish communities. The analytical 
techniques employed offer a powerful method to reveal 
community structure and relationships with environ- 
mental conditions, and their use in other systems is 
encouraged. The analyses reveal spatial and temporal 
gradients in community structure between two adjacent 
but oceanographically different regions. Our results 
indicate that waters within the GF and nearshore to the 
estuary exit are oceanographically complex transition 
zones in contrast to more uniform but separate coastal 
communities; hence they reveal the importance of the 
GF and similar zones elsewhere as connectors or corri- 
dors between coastal and estuarine communities, where 
3 Woodson, C. B., M. A. McManus, J. A. Tyburczy, J. A. Barth, 
L. Washburn, J. E. Caselle, M. H. Carr, D. P. Malone, P. T. 
Raimondi, B. A. Menge, and S. R. Palumbi. 2011. Coastal 
fronts set recruitment and connectivity patterns across mul- 
tiple taxa. Unpubl. manuscript, 13 p. [Available from C.B. 
Woodson, Environmental Fluid Mechanics Lab., Dept. Civil 
and Environmental Engineering, Standord Univ., Stanford, 
CA 94305-4020. 
Fishery Bulletin 109(3) 
more closely spaced biological and physical sampling is 
required to untangle the inherent complexity of such 
areas compared to open coastlines (Schwing et al., 1991). 
In order to move incrementally toward ecosystem-based 
management, the Pacific Fisheries Management Council 
may consider adopting a more regional “nested approach” 
to spatial management of living marine resources in the 
CC, in which smaller segments of the ecosystem could 
be defined for management purposes on cape-and-bay 
scales that better match the scales of variation in com- 
munity and habitat structure (Ecosystem Plan Develop- 
ment Team 2 ). 
The gradual implementation of ecosystem-based man- 
agement will continue to involve regular ship-based 
oceanographic and biological sampling. It will be neces- 
sary to measure species diversity within smaller cohe- 
sive regions, to account for multispecies patterns of dis- 
tribution, to identify and describe essential fish habitat, 
and to model and understand food-web dynamics includ- 
ing predator-prey relationships, interspecific competi- 
tion, and guild membership and biomass, among other 
things (Francis et al., 2007). Quantitative indicators for 
describing fish communities and for tracking ecosystem 
status are used in current modeling (Rice and Rochet, 
2005). The best indicators are easily measured and are 
reliable proxies for a suite of desirable ecosystem attri- 
butes, and often include species that are not themselves 
the targets of any fishery and would likely have been 
ignored in previous single-species management plans. 
Indicators of ecosystem status often consist of species 
with common properties such as foraging guild member- 
ship, spatial distribution, ecotype, or some combination 
of these (e.g., small planktivorous fish, migratory me- 
sopelagic fish, all sharks) (Fulton et al., 2005). Among 
the best performing indicators for a set of north Pacific 
ecosystem models are biomass groups consisting of de- 
tritivores, flatfish, and zooplanktivorous fish, as well as 
some surprising compound metrics such as the ratio of 
forage fish to jellyfish biomass (Samhouri et al., 2009). 
The use of indicators such as these to track and evalu- 
ate ecosystem attributes is a central part of the inte- 
grated ecosystem assessment process and an important 
policy objective of the NMFS (Levin et al., 2009). 
Acknowledgments 
We thank the officers and crew of NOAA ship David 
Starr Jordan and the fishing crews of FV Irene’s Way, 
FV Frosti, FV Cassandra Anne, and RV Shana Rae. We 
especially thank Captains J. Christmann and J. Pennisi. 
Nets were skillfully mended by D. King (NMFS) and S. 
Patterson (NET Systems). Field support was provided 
by B. Jarvis, C. Royer, M. Fuller, R. Barnett-Johnson, 
A. Milloy, S. Painter, S. Campbell, M. Bond, C. Hanson, 
E. Norton, J. Alonzo, and many others. Statistical advice 
from E. J. Dick and P. Raimondi, discussions with J. 
Ciancio, J. Field, S. Hayes, and B. Wells, and comments 
from four anonymous reviewers greatly improved the 
manuscript. 
