282 
Fishery Bulletin 116(3-4) 
majority overwinter in the river before outmigrating 
to the marine environment during the following spring 
(Ruggles, 1980). For those fish that overwinter in-river, 
survival to outmigration is >50% (Gray et al., 1987; 
Downton and Reddin 2 ). Across systems (North Ameri¬ 
can and European rivers) and species (Atlantic salmon 
and steelhead, Oncorhynchus my kiss), downstream kelt 
migrations in spring are generally finished by July 
(Halttunen et ah, 2010; Marston et ah, 2012; Lacroix, 
2013). Once at sea, the time needed to rebuild energy 
stores varies according to several factors: the amount 
of time previously spent in freshwater (Niemela et ah, 
2006), migratory distance to marine feeding areas (La¬ 
croix, 2013), and fish condition after spawning (Jons- 
son et ah, 1991). 
The ability of kelts to survive 1) a return migra¬ 
tion to the sea, 2) a year or more of reconditioning, 
and 3) a return migration to spawning grounds affects 
the proportion of repeat spawners, which varies from 
0 to 26% for Atlantic salmon populations (Fleming and 
Reynolds, 2004; Niemela et ah, 2006). However, spawn¬ 
ing runs composed of >40% repeat spawners have been 
reported in undammed systems in Canada that are 
not subject to commercial exploitation (Jessop, 1976; 
Saunders and Schom, 1985). The number of broods a 
fish has is also variable in salmonines. In steelhead, 
nearly all iteroparous individuals spawn no more than 
twice in their lifetime (Seamons and Quinn, 2010). At¬ 
lantic salmon have been documented as returning for 
as many as eleven spawning runs during their lives 
(Ducharme, 1969; Hutchings and Morris, 1985; Saun¬ 
ders and Schom, 1985), but with at-sea kelt survival 
<10% (Jonsson and Jonsson, 2004), multiple returns is 
likely the exception, rather than the rule. 
It is unknown how recent changes to the Penob¬ 
scot River (Opperman et ah, 2011) may have affected 
migration of naturally spawned kelts from the upper 
reaches of the river, particularly because the down¬ 
stream bypasses at all dams are closed for much of 
the winter, and therefore may not be operational dur¬ 
ing the postspawning period. Kelts outmigrating down 
the upper mainstem of the Penobscot River must pass 
two dams to reach the middle portion of the river (up¬ 
stream from Milford Dam, and downstream from the 
Piscataquis River confluence [Fig. 1]), whereas kelts 
moving through the Piscataquis River must pass three 
dams to reach the same location. Once reaching the 
confluence of the Piscataquis and Penobscot rivers, 
kelts may travel downstream through the main stem 
of the Penobscot River to the estuary (passing Milford 
Dam), or migrate through the Stillwater Branch of the 
river through Gilman Falls Dam, Stillwater Dam, and 
Orono Dam (Fig. 1). 
The Penobscot River has been the focus of hatch¬ 
ery supplementation based on sea-run broodstock since 
Map of the study area that included the Penobscot 
River from Orrington, Maine, at river kilometer (rkm) 
33, to the West Enfield Dam. Black rectangles repre¬ 
sent functioning hydroelectric dams that have no oper¬ 
able fish passage upstream from 15 November to 15 
April. Open rectangles represent water control dams 
that are passable under certain conditions. Gray tri¬ 
angles represent the locations of stationary radio re¬ 
ceivers, and circles indicate the sites where tagged 
kelt Atlantic salmon (Salmo salar ) were released on 17 
November 2015. River kilometer zero is located at the 
southwestern tip of Verona Island (516414E 4927755N, 
UTM Zone 19N). The area surrounded by the Milford, 
Orono, Stillwater, and Gilman Falls Dams is Marsh 
Island. 
2 Downton, P. R., and D. G. Reddin. 2004. Status of Atlantic 
salmon (Salmo salar L.) in Campbellton River, Notre Dame 
Bay (SFA 4), Newfoundland in 2003. Can. Sci. Adv. Seer., 
Res. Doc. 2004/043, 64 p. [Available from website.] 
