444 
Fishery Bulletin 109(4) 
Latitude (N) 
Figure 2 
Log 10 -transformed catch of American shad (Alosa sapidissima) and 
percent frequency of occurrence (FO) out to the 200-m isobath by 
one degree latitude ranges during (A) the 1977-2004 Alaska Fish- 
eries Science Center triennial surveys, (B) the 2003-08 Northwest 
Fisheries Science Center early season (May-July) sampling, and 
(C) late season ( August-October) sampling. Box plots show the 
50th-75th percentile (shaded columns), the 10th-90th percentiles 
(whiskers), and outliers (dots). Small squares indicate FO of shad 
in tows. Kruskal-Wallace tests showed that catch differed by lati- 
tude (P<0.05) for all three studies. 
a decline in size with year was still apparent 
(Fig. 7B). Shad caught in the surface trawls in 
1998 were larger than those in the four sub- 
sequent years (2001, 2004, 2007, and 2008) 
when more than 10 fish were caught (Kruskal- 
Wallace test, P=0.001) and therefore also in- 
dicated a possible decrease in average weight 
of shad during this decade. 
During the AFSC sampling survey over 
the shelf (<200 m), shad were caught in tows 
over a wide range of sea surface tempera- 
tures (SSTs), from about 9° to 18°C, with the 
10th and 90th percentiles of log 10 (catc/i+l) 
occurring at 11.2° and 16.5°C, respectively 
(Fig. 8A). Cumulative frequency curves of 
log 10 (catc/i+l) vs. SST and presence-absence 
vs. SST closely followed that of sampling, in- 
dicating that shad were widespread across 
most sea surface temperatures sampled on 
the shelf. However, the raw catch of shad 
(numbers per tow) indicated that the largest 
catches tended to occur where SSTs were from 
13° to 17°C (Fig. 8A). This finding indicates 
that large schools of shad may be more abun- 
dant in areas where the SST is above 13°C. 
Conversely, the largest catches of shad also 
tended to occur where the bottom temperature 
(where gear was situated) was colder than 
that found in most other areas sampled; 80% 
of the raw catch occurred between 6.4°C and 
8.0°C, whereas only 56% of sampling was in 
water that cold (Fig. 8, B and C). During both 
AFSC and NWFSC sampling surveys, bottom 
temperature was strongly negatively corre- 
lated with latitude (f? = 0.78 and 0.86, respec- 
tively). SST was weakly positively correlated 
with latitude (J? = 0.36). Therefore, the shad 
abundance vs SST patterns seen in Figure 8 
(largest catches where bottom temperature is 
cool and SST is warm) are consistent with the 
generally larger catches seen north of 44°N 
and the smaller catches seen off central Cali- 
fornia (Fig. 2). 
Discussion 
Shad undertake long distance seasonal migra- 
tions along the Atlantic coast of the United 
States, swimming thousands of kilometers 
north in the summer and south in the winter 
(Talbot and Sykes, 1958; Walburg and Nichols, 
1967; Leggett and Whitney, 1972; Leggett, 
1973). Although little is known about shad 
migrations in the Pacific Ocean (Moyle, 1976), 
Leggett and Whitney (1972) speculated that 
shad in the Pacific Ocean migrate long dis- 
tances within their preferred SST range of 
13-18°C as they do in the Atlantic — migrat- 
ing south of Point Conception into southern 
