446 
Fishery Bulletin 1 10(4) 
in 2007-10 (1.8 million; FPS database). By contrast, 
Johnsen and Sims (1973) report that in 1969, 27% of all 
fish caught were American shad, at a time when only 
317,000 adults were counted crossing Bonneville Dam; 
the reason for this apparent discrepancy is unknown. 
Given the rapid expansion of American shad within 
the Columbia River (Hasselman et ah, 2012a) and in 
marine waters along the west coast of North America 
(Pearcy and Fisher, 2011), there is increasing concern 
about the potential impacts of this exotic species on 
riverine, estuarine, and marine ecosystems (Pearcy and 
Fisher, 2011; Hasselman et ah, 2012b). 
As in our study, purse seines were used in 2 previous 
studies conducted in 1969 (Johnsen and Sims, 1973) 
and in 1978-80 (Dawley et al. 2 ’ 7 ) to sample the fish 
assemblage in open waters of the lower Columbia River 
estuary. However, these earlier studies used different- 
size nets (in length and depth but not mesh size) that 
were towed for 10 min before being closed. This meth- 
odological difference makes abundance comparisons 
among studies problematical, although we assume 
that species composition would be less affected (see 
discussion later in this section). These studies used 
the North Channel sampling location, eliminating 
possible site-specific differences. Comparison of the 
overall fish assemblage in the 3 time periods (1969, 
1978-80, and 2007-10) indicates that the relative 
abundances of threespine stickleback and juvenile 
salmon are quite different between the earlier years 
and the most recent years (Fig. 8). Specifically, ju- 
venile salmon in our study made up a small fraction 
(2%) of the overall catch, but historically they contrib- 
uted between 43% and 57% of the total number of fish 
caught. In contrast, threespine stickleback were much 
more abundant in our study (averaging 88% of fish 
caught in April to 22% in June) than in earlier surveys 
(35% in April, <1% in either May or June) (Fig. 8A). 
Such changes in this estuarine fish community, if 
true, seem surprising at first glance but really are 
not unexpected given the substantial changes to fish 
assemblages documented farther upstream (Hughes et 
ah, 2005; Maret and Mebane, 2005). 
The relatively high proportion of juvenile salmon in 
estuarine catches in the late 1970s compared with that 
found in our study in 2007-10 could be a result of ei- 
ther lower juvenile salmon abundances, higher abun- 
dances of nonsalmon forage fish, or some combination 
of the two. Comparisons of hatchery production and 
total adult run size (hatchery and wild; size estimates 
for wild runs are not available basinwide) between the 
2 time periods indicate that the latter — an increase 
in nonsalmon abundances — is largely responsible. For 
example, compared with the annual hatchery produc- 
tion levels (~140 million) in our survey years, hatchery 
production was similar (141 million) in 1969 and only 
moderately higher in 1978-80 (164-187 million) (Hil- 
born and Hare 8 ). In addition, improvements to fish pas- 
sage through mainstem dams have increased in-river 
survival (Williams et al., 2001), and, therefore, more 
juvenile salmon should currently survive downstream 
migration to the estuary than they did in earlier times. 
Similarly, a comparison of total (hatchery and wild) 
adult run sizes within the Columbia River between the 
2 time periods indicates that total adult-run sizes were 
higher during our study than they were in the early 
1980s. For example, annual counts of sockeye salmon 
and steelhead over Bonneville Dam in 1980-83 aver- 
aged 66,000 and 166,000 fish, respectively, but recent 
(2008-10) counts averaged 259,000 and 457,000 fish, 
respectively (FPC database). Similarly, estimated total 
in-river run size (which does not include catch in ocean 
fisheries that target fall [subyearling] Chinook and coho 
salmon) in 2008-10 for Chinook salmon (mean=843,000) 
and coho salmon (555,000) is higher now than in the 
early 1980s (556,000 Chinook and 294,000 coho salmon) 
(WDFW and ODFW 9 ; PFMC, 2011). Assuming that 
predators, disease, and other sources of mortality are 
similar between the 2 time periods, these estimates 
of juvenile and adult abundances indicate that juve- 
nile salmon should be as abundant — if not more so — 
in the Columbia River estuary during our study than 
they were 3 decades ago. This conclusion indicates that 
changes in the fish assemblage are most likely the re- 
sult of increased abundance of nonsalmonids. 
Methodological differences also may have contributed 
to these observed patterns because in the earlier stud- 
ies the net was towed instead of being set in a circle as 
happens in round hauls. We examined whether this dif- 
ference might influence catch composition by comparing 
the ratios of juvenile salmon or threespine stickleback 
to all fish caught in our round hauls (n = 188) and tows 
(« = 81) during cruises in which both were conducted. 
Although statistically significant differences were ob- 
served between tows and round hauls for the ratio of 
juvenile salmon to all fish (Bonferroni multiple compari- 
son test, P<0.05), the differences were quite small (0.10 
in round hauls, 0.12 in round hauls), and there were no 
statistically significant differences for the ratio of three- 
spine stickleback to total catch between round hauls 
and tows (0.34 and 0.36, respectively). These results 
suggest that method of net deployment was unlikely to 
cause the observed differences. Although our abundance 
estimates are not directly comparable to the earlier 
data because of differences in techniques, our results 
certainly suggest that the abundance of nonsalmonids, 
especially of the threespine stickleback, has increased 
in recent years. 
If these changes to the fish assemblage are indeed 
true, they likely reflect changes in environmental condi- 
tions in the Columbia River estuary. Over the last 150 
8 Hilborn, R, and S. R. Hare. 1992. Hatchery and wild fish 
production of anadromous salmon in the Columbia River 
basin. Fisheries Research Institute, Univ. Washington, 
Seattle, WA, Tech. Rep. FRI-UW-9207. [Available from 
http://www.fish.washington.edu.] 
9 WDFW and ODFW (Washington Department of Fish and 
Wildlife and Oregon Department of Fish and Wildlife 
ODFW). 2002. Status report: Columbia River fish runs 
and fisheries, 1938-2000, 324 p. [Available from WDFW, 
600 Capitol Way N., Olympia, WA 98501-1091.] 
