Unwin: Survival of Chinook salmon in relation to spring and summer mainstem flows of the Rakaia River, New Zealand 
815 
for these fingerlings is consistent with the back-cal- 
culated mean FL at seawater entry for ocean-type 
adults of Rakaia origin (Unwin and Lucas, 1993), 
confirming the importance of springtime mainstem 
rearing for ocean-type fry. Freshwater residence pat- 
terns of juvenile stream-type fish are less well un- 
derstood, but there is some evidence that an initial 
period of tributary rearing lasting 3-6 mo is followed 
by mainstem rearing for the remainder of the first 
year (Unwin, 1986). 
In the absence of a commercial marine fishery for 
salmon, very little is known about the marine phase 
of the salmon life cycle. Adult chinook are piscivo- 
rous and appear to feed opportunistically within the 
pelagic zone, although prey diversity is low and food 
availability is potentially limited by annual fluctua- 
tions in prey abundance (James and Unwin, 1996). 
Brood year survival rates for both naturally and 
hatchery-produced chinook of Rakaia origin vary by 
up to two orders of magnitude and appear to be pre- 
dominantly related to marine influences (Unwin, in 
press). However, these results do not preclude the 
possibility that survival may also be partly influenced 
by conditions within the freshwater environment. 
Data sources and methods 
The Rakaia River 
The Rakaia River is a large, braided, glacier-fed river 
draining a 2,910 km 2 catchment that spans 70 km of 
the Southern Alps and rises to 2,800 m. Apart from a 
gentle 5-km gorge where the flow is briefly confined 
to a single channel, the river occupies an unstable, 
highly braided shingle bed up to 5 km wide (see Fig. 
2 of Glova and Duncan, 1985). After collecting water 
from two major headwater tributaries (the Mathias 
and Wilberforce), the lower section (90 km) of the 
river flows directly into the Pacific Ocean with no 
significant tributary input. All major salmon spawn- 
ing waters are located upstream of the Wilberforce 
confluence (Fig. 1). River gradient is virtually con- 
stant below this point, averaging 4.5 m/km, and the 
river discharges into the ocean via a small freshwa- 
ter lagoon extending inland from the open sea for 
less than 100 m. The term “lagoon” is used in prefer- 
ence to “estuary” because there is no ebb and flow of 
the tide (although there is a tidal backup of fresh 
water) and because the area supports few, if any, 
predominantly estuarine life forms. 2 A detailed de- 
2 Eldon, G. A., and A. J. Greager. 1983. Fishes of the Rakaia 
Lagoon. Fisheries Environmental Report 30. N. Z. Ministry 
of Agriculture and Fisheries, Fisheries Res. Div., Christchurch, 
65 p. [Copy held by NIWA, Christchurch, New Zealand.] 
scription of the river and its catchment is given by 
Bowden. 3 4 5 6 7 
Continuous flow data for the Rakaia River have 
been collected since 1959 by means of recorders at 
the downstream end of the gorge, 62 km above the 
mouth. For the purposes of this study, all flow sta- 
tistics were calculated from the daily mean discharge 
(Q). Discharge (annual mean 200 m 3 /s) shows a mod- 
erately seasonal pattern, monthly means varying 
from 127 m 3 /s in July to 265 m 3 /s in November. 3 The 
mean annual flood (the average of the annual maxi- 
mum flow) is 1 448 m 3 /s, with a peak instantaneous 
discharge of 5,600 m 3 /s (estimated to have a return 
period of 60 yr) recorded in January 1994. The bank- 
full discharge (the instantaneous flow which results 
in complete inundation of the river bed as individual 
braids coalesce) ranges from 800 m 3 /s just below the 
gorge to about 2,500 m 3 /s in the lower river. Flood 
waters move rapidly downriver, typically reaching 
the mouth 8-12 h after passing through the gorge, 
although peak velocity increases with flood inten- 
sity, and travel times as short as 3.5 h over 40 km 
have been observed. 4,5 Over the months relevant to 
this study, daily mean water temperatures in the 
lower river (23 km above the mouth) range from 6°C 
in August to 16°C in January (Unwin, 1986). 
Glenariffe Stream spawning runs 
Glenariffe Stream is a spring-fed tributary joining 
the Rakaia River 100 km above its mouth at an alti- 
tude of 430 m (Fig. 1). The flow regime is exception- 
ally stable, with a mean discharge of 3.4 m 3 /s and a 
maximum recorded discharge (over a seven-year pe- 
riod) of 16 m 3 /s. Chinook salmon spawning runs in 
Glenariffe Stream have been monitored annually by 
means of a counting fence installed in 1965 (Quinn 
and Unwin, 1993; Plain 1 ). The modal age-at-matu- 
rity is three years, with smaller numbers of 2-year 
and 4-year-olds and very few 5-year-old fish. The 
angler interception rate varies little between years, 
typically ranging from 30% to 40%. 6,7 Since 1980, 
3 Bowden, M. J. 1983. The Rakaia River and catchment — a 
resource survey, vol. 2. North Canterbury Catchment Board 
and Regional Water Board, Christchurch, New Zealand, 101 p. 
[Copy held by NIWA, Christchurch, New Zealand.] 
4 1997. Unpubl. data, NIWA, Christchurch, New Zealand. 
5 Horrel, G. 1997. Canterbury Regional Council, PO Box 345, 
Christchurch, New Zealand. Personal commun. 
6 Unwin, M. J., and S. F. Davis. 1983. Recreational fisheries 
of the Rakaia River. Fisheries Environmenal Report 30. New 
Zealand Ministry of Agriculture and Fisheries, Fisheries Re- 
search Division, Christchurch, New Zealand. [Copy held by 
NIWA, Christchurch, New Zealand.] 
7 Millichamp, R. 1997. North Canterbury Fish and Game 
Council, 3 Horatio Street, Christchurch, New Zealand. Per- 
sonal commun. 
