442 
Fishery Bulletin 109(4) 
seine used during 1981-85 was 457-496 m long, had 
32-mm or finer mesh, fished to a depth of 30-65 m, and 
sampled about 20,000 m 2 (Pearcy and Fisher, 1988). 
The midwater trawl deployed during 1998-2008 was a 
Nordic rope trawl that fished near the surface mainly 
during the day with a mouth opening 30 m wide x 20 m 
deep and had a 0.8-cm fine mesh liner in the codend 
(Brodeur et al., 2005). 
Shad catches and sizes were available from both com- 
mercial midwater and bottom trawling, 1997-2009, off 
British Columbia (Davidson and Fargo 5 ), from com- 
mercial landings from bottom and midwater trawls and 
from gill nets, set nets, dip nets, and hook-and-line gear 
fished in the Columbia River and the Oregon coast, 
1978-2009 (Karnowski and Hurtado 6 ) and California 
(Larinto 7 ). Catches in all regions were highly variable 
and shad discarded as bycatch were not reported. In 
addition, observer data on shad catches in the limited- 
entry trawl groundfish fishery in Washington, Oregon, 
and California for “summer” (April-October) and “win- 
ter” (November-March) seasons, 2002-2009 were also 
examined (Majewski and Bellman 8 ; Olson 9 ). 
Shad, a schooling pelagic fish, undertake diel vertical 
migrations in the Atlantic (Neves and Depres, 1979). 
Such migrations are not known, however, for shad in 
the Pacific Ocean. We assumed that they would be more 
susceptible to capture during the daytime in bottom 
tows, or as the net descended or ascended to surface 
waters than in surface waters. Because schooling be- 
havior may result in a few extremely large catches and 
many zero catches, we restricted our analyses mainly 
to log 10 transformed numbers for fish caught and pres- 
ence-absence data to deemphasize the rare catches of 
large numbers of shad. Although bottom trawls are de- 
signed to capture demersal species, catches may reflect 
major changes in abundance or availability of pelagic 
species, such as shad (Neves and Depres, 1979). In ad- 
dition, shad migrate into estuaries and freshwater to 
spawn during May, June, and July along the Pacific 
coast (Hamman, 1981; Petersen et ah, 2003) and hence 
adults and some juveniles were not available during 
the early months of ocean sampling. Also, small and 
young-of-the-year shad are unlikely to be retained by 
all sampling nets. 
5 Davidson, J., and J. Fargo. 2010. Personal commun. 
Department of Fisheries and Oceans, 200-401 Burrand St., 
Vancouver, British Columbia, Canada V6C 3S4. 
6 Karnowski, M., and N. Hurtado. 2010. Personal commun. 
Oregon Department of Fish and Wildlife, 2040 Southeast 
Marine Science Dr., Newport, OR 97365, and 3406 Cherry 
Ave NE, Salem, OR 97305. 
7 Larinto, T. 2010. Personal commun. California Depart- 
ment of Fish and Game, 4665 Lampson Ave Los Alamintos, 
CA 90720. 
8 Majewski, J., and M. Bellman. 2010. Personal commun. 
Northwest Fisheries Science Center, 2725 Montlake Blvd. 
E„ Seattle, WA 98112. 
9 Olson, J. 2010. Personal commun. Pacific States Marine 
Fisheries Commission, 7600 Sand Point Way, Seattle, WA 
98115. 
Results 
Shad were caught in 1178 of the 5612 tows by the AFSC 
(frequency of occurrence, FO=21%), and in 403 of 3762 
tows by the NWFSC (FO=ll%). Highest log 10 catches 
were noted along the continental shelf off Washing- 
ton and Vancouver Island and off San Francisco Bay 
during the AFSC cruises (Fig. 1A). Catches in NWFSC 
early season and late season tows were more uniformly 
distributed along the coast from northern Washington 
to northern California than were AFSC catches, but 
again with highest catches off Washington and lower 
catches to the south, with a cluster of catches off San 
Francisco (Fig 1, B and C). Large catches appeared to 
shift from Oregon to off Washington up to Vancouver 
Island between the early and late NWFSC cruises. This 
shift was consistent with the high catches off Vancouver 
Island also in late summer during the AFSC cruises 
(see also Fig. 2). 
When the log 10 (catc/i + l) and FO data were pooled 
across years, clusters of high shad catches became evi- 
dent off the Washington coast (45-49°N lat.) and along 
the central California coast (37-38 °N lat.) during the 
AFSC sampling (Fig. 2A). Similar latitudinal trends in 
abundance were shown with the late-May to late-July 
NWFSC sampling (Fig. 2B), and this northward shift in 
abundance was also documented with the late-August 
to late-October NWFSC sampling (Fig. 2C). In addition 
to the two latitudinal centers of abundance seen in the 
AFSC sampling, the NWFSC sampling indicates a third 
center of abundance in northern California (41-42°N 
lat.). Despite interannual variations in the distribution 
of catches along the coast among years of sampling, this 
fairly consistent distributional pattern emerged. Note 
that catches in the AFSC tows were sometimes orders 
of magnitude higher than those in the NWFSC tows 
(Fig. 2), a difference related to the faster tow speeds, 
longer tow durations on the bottom, and the higher net 
opening of the AFSC tows. 
Shad were also collected in purse seine surveys off 
Oregon and Washington during cruises conducted by 
Oregon State University from 1981 through 1985 (sum- 
marized by month in Fig. 3). Over 1100 shad were 
caught in 29 sets. Catches had a restricted distri- 
bution mainly near the Columbia River plume and 
close to shore. The largest catch comprised 883 shad 
in one purse seine set off Cape Disappointment in 
August 1981. In the daytime surface trawls by the 
NWFSC, only 139 shad were captured from 43 tows 
out of a total of 1536 tows (FO = 2.8%) from central 
Oregon and the Washington coast during 1998-2008 
(Fig. 4). These numbers and frequencies of occurrence 
were much lower than those seen during the demersal 
sampling (Fig. 2), supporting the observations in the 
Atlantic that shad undertake diel vertical migrations 
and are more available in subsurface than in surface 
waters by day. 
An inshore-offshore gradient in abundance of shad 
was significant; most shad were caught in AFSC and 
NWFSC tows on the continental shelf (<200 m) (Fig. 5, 
