Maynard et al.: Movement and mortality of Atlantic salmon kelts (Salmo solar) released into the Penobscot River, Maine 
283 
the late 1800s (Spidle et al., 2001). Current broodstock 
management practice requires the collection of ma¬ 
ture adults from a trap at the terminus of the first 
fishway on the river (currently, Milford Dam) by the 
Maine Department of Marine Resources. These fish are 
transported to the U.S. Fish and Wildlife Service Craig 
Brook National Fish Hatchery in Orland, Maine, where 
they are held for up to five months before spawning is 
artificially induced in late October or early November. 
After spawning, kelts are released into the Penobscot 
River just downstream of the head of tide (at river ki¬ 
lometer [rkm] 43.5). This release point is thought to 
minimize spatial and temporal overlap with naturally 
spawning salmon and maximize survival by expediting 
their return to the sea (Bartron et al. 3 ); however, nei¬ 
ther of these assumptions have been tested. Because 
of record low returns in 2013 and 2014, 56% (18-98%) 
on average of all salmon returning to the Penobscot 
River have been taken to the hatchery as broodstock 
(USASAC 4 ). Coinciding with decreased returns to the 
Penobscot River there has been a reduction in the per¬ 
centage of the run of repeat spawners from an average 
of 1.7% in the 1980s to an average of 0.6% in recent 
years (Maynard et al., 2017). 
Although the total number of repeat spawners in 
the Penobscot River has been low for several decades, 
their contribution to the total number of eggs pro¬ 
duced in the system was likely important, historically. 
A review of salmon egg production in New England 
from the early 1980s to 2011 indicated that, on aver¬ 
age, repeat spawners produced 2300 more eggs per fe¬ 
male than virgin spawners in the Connecticut River, 
3000 more eggs/female in the Sheepscot River, and 
3100 more eggs/female in the Merrimack River (USA¬ 
SAC 3 ). Similarly, research from the River Teno, Fin¬ 
land, found that although iteroparous Atlantic salmon 
numerically represented an average of 5% of the run, 
they accounted for an average of 7% of the total mass 
of salmon caught in the river (Niemela et al., 2006). 
In the river systems of the Pacific Northwest, female 
iteroparous steelhead have a lifetime reproductive suc¬ 
cess that is 1.3 times higher than that of semelparous 
individuals, and iteroparous males have a lifetime re¬ 
productive success that is 2.8 times higher than that of 
semelparous individuals (Seamons and Quinn, 2010), 
indicating that this life history strategy allows some 
salmonines to produce more offspring than semelpar- 
3 Bartron, M. L., D. Buckley, T. King, M. T. Kinnison, G. Mack¬ 
ey, T. F. Sheehan, K. F. Beland and J. Marancik. 2006. Cap¬ 
tive broodstock management plan for Atlantic salmon at 
Craig Brook National Fish Hatchery, 133 p. Report to the 
Maine Technical Advisory Committee. [Available from Pro¬ 
tected Species Branch, Northeast Fish. Sci. Cent., Natl. Mar. 
Fish. Serv., NOAA, 166 Water St., Woods Hole, MA 02543- 
1026.] 
4 USASAC (U.S. Atlantic Salmon Assessment Committee). 
2015. Annual report of the U.S. Atlantic Salmon Assess¬ 
ment Committee. Report No. 7—2014 activities. Kittery, 
Maine, February 9-12, 2015, 228 p. Prepared for U.S. Sec¬ 
tion to the North Atlantic Salmon Conservation Organiza¬ 
tion. [Available from website.] 
ity would. Previously spawned individuals may also 
provide an important buffer against years with low es¬ 
capement (Saunders and Schom, 1985) because of their 
high fidelity to their natal rivers (Hansen and Jonsson, 
1994) and high reproductive potential. 
Research on kelt movements and survival is rela¬ 
tively sparse, compared with research on other life his¬ 
tory stages of salmonines (Hubley et al., 2008). This 
sparsity of research is especially true in the context 
of individuals that are artificially induced to spawn at 
hatcheries. Because nearly all kelts in the Penobscot 
River system are the product of artificial spawning at 
hatcheries (i.e. the fish are anesthetized and stripped 
of gametes before release), a better understanding of 
postrelease movements may allow a targeted selection 
of release sites to match the need for suitable over¬ 
wintering habitat, and therefore improve survival to 
outmigration. 
Our goals were 1) to compare the postrelease move¬ 
ment and exit timing of kelts released at two sites in 
the Penobscot River, -head of tide and upstream of the 
Marsh Island hydropower complex, which comprises 
the 4 dams that surround Marsh Island (Milford, Oro- 
no, Stillwater, and Gilman Falls), and 2) to evaluate 
differences in outmigration success between fish re¬ 
leased upstream of the hydropower complex and those 
released downstream, in the free-flowing estuary. 
Materials and methods 
Study area 
The Penobscot River, Maine, is the second largest wa¬ 
tershed in New England (with an area of 22,000 km 2 ) 
and has been dammed in multiple locations since the 
late 1800s (Opperman et al., 2011). It also currently 
has the largest remaining run of Atlantic salmon in the 
United States and is dependent on hatchery stocking. 
Since the 1970s, an average of 82% of returning adult 
Atlantic salmon have originated from hatchery-reared 
smolts (USASAC 3 ), and in recent years, the number is 
>95%. 
Our study area (Fig. 1) included a 66-km section of 
the Penobscot River from Orrington, Maine (rkm 33), 
to the West Enfield Dam (rkm 99), as well as a 67-km 
stretch of the Piscataquis River from Howland Dam 
(rkm 98) to Brown’s Mill Dam (rkm 165). There are 
8 dams in the study area. Howland Dam has a newly 
constructed bypass channel that is operational in the 
winter months. Stillwater Dam and Gilman Falls Dam 
have no dedicated upstream passage facilities, and the 
remaining five dams (Orono, Milford, West Enfield, 
Pumpkin Hill, and Brown’s Mill) have upstream pas¬ 
sage facilities that are not operated between November 
15 and April 15. Downstream passage facilities at Still¬ 
water Dam, Orono Dam, and Milford Dam are closed 
between December 31 and April 1 because of ice, but 
passage may still be possible through the spillways or 
when water overflows the dams. 
