Maynard et al.: Movement and mortality of Atlantic salmon kelts (Salmo solar) released into the Penobscot River, Maine 
285 
Figure 2 
Representative paths of 4 tagged kelt Atlantic salmon (Salmo solar) tracked 
from mid-November 2015 to min-June 2016 on the Penobscot River, Maine. Pie 
charts indicate the proportion of fish that exhibited 1 of 2 movement patterns or 
1 of 2 mortality types. The gray pie charts provide percentages for fish released 
at the upriver site at river kilometer (rkm) 92, and the black pie charts provide 
percentages for fish released at the downriver site at rkm 43.5. The tracks indi¬ 
cate fish that (A) overwintered in freshwater and then migrated to the estuary 
in the spring, (B) had an apparent overwinter mortality, (C) outmigrated rapidly 
(within 1 month) after release, and (D) had an apparent initial mortality. The 
dotted lines represent locations of hydroelectric dams on the main stem of the 
river. Because our lowest radio receiver in the Penobscot River system was lo¬ 
cated at rkm 30, detection at that point was considered successful outmigration. 
ter fish (i.e. they spent >2 years maturing at sea), and 
because no grilse were tagged, only a limited number 
of males were available for the project. Total lengths 
ranged from 720 mm to 850 mm (median=77Q mm). 
Fish masses ranged from 3.3 kg to 5.9 kg (median=4.4 
kg), and body condition index ranged from -0.33 to 
0.16 (median=0.01). Of the 55 fish released, 30 were 
detected at least once on the stationary radio array, 
and 49 were detected at least once during mobile track¬ 
ing. Only three fish were never detected. Over half of 
all fish (56%) were tracked for more than 100 days 
(median=46 days). On the basis of detection histories 
(as described in the Materials and methods section) 
19 fish from the upriver release and 20 fish from the 
downriver were included in the survival analysis. 
Broadly considered, movement could be classified 
into 2 major groups overwintering in-system or rapid 
(within one month) downstream movement to the es¬ 
tuary (Fig. 2). Most of the tagged kelts (46 individu¬ 
als) overwintered in the Penobscot River, and only 6 
rapidly exited the system (Fig. 3). Of the 6 kelts that 
rapidly exited the system, 4 were initially released at 
the downriver site (rkm 43.5) and two were released 
at the upriver site (rkm 92). Many of the fish (36 in¬ 
dividuals) that remained in the river over the winter 
made directed upstream movements (Fig. 4). Fish that 
overwintered in the system did so between rkm 33 and 
rkm 99, depending on where they were stocked (i.e. 
under most river conditions Milford Dam at rkm 61 
impeded movement upstream). Of these, 15.4% of the 
fish were located downstream of the head of tide (~rkm 
48), 34.6% were located between the head of tide and 
Milford Dam, and the remainder of our tagged fish 
(50%) overwintered between Milford Dam and the Pis¬ 
cataquis River confluence within the main stem of the 
river. 
A total of 23 fish moved to the next dam upstream 
from their release location. Of these fish, 9 made re¬ 
peated approaches to the dams over a period of days 
or weeks, indicating that the dams likely impeded their 
