812 
Abstract .—To characterize the im- 
pact of spring floods on the survival of 
juvenile chinook salmon in the un- 
stable, braided rivers on the east coast 
of New Zealand’s South Island, I exam- 
ined correlations between spring and 
summer flows in the mainstem of the 
Rakaia River and fry-to-adult survival 
for chinook salmon spawning in a head- 
water tributary. Flow parameters that 
were investigated included mean flow, 
maximum flow, and the ratio of mean 
to median flow (an index of flow vari- 
ability), calculated during peak down- 
river migration of ocean-type juveniles 
(August to January). Survival was 
uncorrelated with mean or maximum 
flow but was positively correlated with 
the ratio of mean to median flow dur- 
ing spring (October and November). 
The correlation suggests that pulses of 
freshwater entering the ocean during 
floods may buffer the transition of fin- 
gerlings from fresh to saline waters and 
thus partly compensate for the lack of 
an estuary on the Rakaia River. A posi- 
tive correlation between spring flow 
variability and the proportion of ocean- 
type chinook in relation to stream-type 
chinook is also consistent with this hy- 
pothesis. All correlations were rela- 
tively weak, reinforcing earlier results 
that production is primarily controlled 
by marine influences. These findings 
further demonstrate the considerable 
ability of chinook salmon to adapt to 
new habitats. 
Manuscript accepted 7 May 1997. 
Fishery Bulletin 95:812-825 (1997). 
Survival of chinook salmon, 
Oncorhynchus tshawytscha, from a 
spawning tributary of the Rakaia River, 
New Zealand, in relation to spring and 
summer mainstem flows 
Martin J. Unwin 
National Institute of Water and Atmospheric Research (NIWA) 
PO Box 8602, Christchurch, New Zealand 
E-mail address: m.unwin@niwa.cri.nz 
To understand the population dy- 
namics of anadromous Pacific 
salmonids ( Oncorhynchus spp.), it 
is important to isolate and charac- 
terize the influence of varying en- 
vironmental factors on annual pro- 
duction. In the course of their life 
cycle salmon inhabit a succession of 
freshwater and marine environ- 
ments, where prospects for survival 
depend on prevailing conditions. 
Spawning and incubation success 
may be adversely affected by sub- 
strate disturbance during floods; 
the suitability of riverine waters as 
habitat for rearing juveniles is de- 
pendent on both flow and tempera- 
ture and may be reduced by flows 
that are too low or too high; and 
adult survival within the marine 
environment is at least partly de- 
termined by environmentally con- 
trolled factors such as oceanic wa- 
ter masses and the availability of 
suitable prey. Numerous studies 
have demonstrated significant cor- 
relations between environmental 
variables and indices of survival 
and growth, at scales ranging from 
local to global. Although correlation 
analysis in fisheries studies has 
been criticized for its potential for 
misuse and for a propensity to pro- 
duce weak results of little practical 
value (Walters and Collie, 1988), 
other authors have noted that pro- 
vided the method is used with dis- 
cretion, biologically meaningful re- 
sults can be derived (Kope and Bots- 
ford, 1990). 
Despite the importance of in- 
stream habitats for rearing juvenile 
chinook salmon (O. tshawytscha), 
the relation between flow and brood 
year survival has received compara- 
tively little attention. Interannual 
trends in the abundance of chinook 
salmon in the Fraser River, British 
Columbia, have been linked to flow 
variations in the mainstem (Beamish 
et al., 1994 ) and in the Nechako River 
tributary (Bradford, 1994). In the 
former study, annual production 
was inversely related to mean an- 
nual discharge, whereas in the lat- 
ter study, juvenile survival in the 
upper Nechako appeared to decline 
as a result of flow diversion for hy- 
droelectric generation, and the pro- 
portion of spawning fish using the 
upper river appeared to be nega- 
tively correlated with August flows. 
However, in the Nechako River 
study, as in some other studies link- 
ing downriver migration to river 
flows (e.g. Kjelson et al., 1982), low 
flows were often associated with 
increased water temperatures, 
making it difficult to differentiate 
between flow-related and tempera- 
ture-related effects. Williams and 
Matthews (1995) found that sur- 
vival of Snake River spring and 
summer chinook salmon juveniles 
was reduced during low flow condi- 
tions but concluded that these 
