Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, Montana

Scientific Disciplines
Biological Sciences - Aquatic
Environmental science
PPL Montana
madison river
Montana State University
rainbow trout
Veterinary parasitology
myxobolus cerebralis
fish diseases
whirling disease
missouri river
wildlife parks
oily fish
holter dam
brown trout
Great Falls
rainbow, missouri
Missouri River, North Dakota
Missouri, Ontario
Wildlife Service
Lake Washington
Leathe S.A.
Missouri River-Holter Dam
Craig Bridge
rainbow trout spawning
prickly pear creek
radio telemetry
Craig, Montana
Dearborn River
sentinel fish
Dick Vincent
Wolf Creek, Montana
Lyons Creek, Montana
Sheep Creek, Montana
Ron Pierce
population estimate
pelican point
Andy Taylor
Volume 18, No. 1-4

Rainbow trout spawning characteristics and 
Relation to the parasite myxobolus cerebralis in 
The missouri river, Montana
Grant G. Grisak, Montana Fish, Wildlife & Parks, 4600 Giant Springs Road, Great Falls, Montana 
1Adam C. Strainer, Montana Fish, Wildlife & Parks, 4600 Giant Springs Road, Great Falls, Montana 
Brad B. Tribby, Montana Fish, Wildlife & Parks, 4600 Giant Springs Road, Great Falls, Montana 59405
The myxosporean parasite Myxobolus cerebralis is responsible for significant declines 
of rainbow trout (Oncorhynchus mykiss) populations in several western states, including 
Montana. Despite a high prevalence of the parasite in Montana’s Missouri River, there have 
been no apparent impacts to the  rainbow trout population. This study examined long-term M. 
cerebralis monitoring data from the Missouri River system below Holter Dam and evaluated 
rainbow trout spawning characteristics such as migration distance, spawning location, site 
fidelity and amount of spawning in the Missouri River and tributaries over three years in 
an attempt to explain why the population has not declined in the presence of M. cerebralis
Over 13 years of monitoring, a mean 5.3 percent of rainbow trout handled during population 
estimates had clinical signs of M. cerebralis infection. In experiments using sentinel fish 
53 percent of the spawning habitat had high severity of M. cerebralis, 38 percent had low 
to moderate severity, and 9 percent had no infection. Radio telemetry showed spawning 
locations varied among years and tagged fish lacked spawning site fidelity. The distance that 
radio-tagged rainbow trout migrated to spawning locations was significantly different among 
river sections of the study area. Twenty-eight percent of the spawning redds were found in the 
Missouri River and 72 percent in the tributaries. Relative to previous studies, we found less 
tributary spawning and an increase in Missouri River spawning, where M. cerebralis infection 
severity is lower. Our findings suggest that diverse spawning behaviors may contribute to 
rainbow trout population stability by spreading the risk of M. cerebralis impact over spawning 
locations that have a broad range of infection severity. 
Key Words: rainbow trout, whirling disease, Myxobolus cerebralis, radio telemetry, redds, 
is approximately 83 percent rainbow trout 
(Oncorhynchus mykiss) and 17 percent 
The Missouri River-Holter Dam tail-
brown trout (Salmo trutta). Brown trout 
water fishery is one of the most productive 
and rainbow trout stocking began in 1928 
trout fisheries in Montana. Out of 1,200 
and 1933, respectively, and continued 
fisheries monitored for angler use statistics 
intermittently though 1973. At that time, 
in Montana it consistently ranks as one 
Montana instituted a statewide wild fish 
of the top four most heavily fished waters 
management policy and discontinued 
(Montana Fish, Wildlife and Parks (MFWP) 
stocking trout in most rivers and streams. 
unpublished data). From 1995 through 2009 
This trout fishery has been sustained by 
anglers spent an average 97,430 (range 
wild reproduction ever since. From 1980 
75,000-123,000) days per year fishing 
through 2010 MFWP has conducted annual 
this reach of river. The trout assemblage 
boat-mounted electrofishing mark-recapture 
1 Montana Fish, Wildlife & Parks, 930 Custer Avenue, 
population estimates of trout greater than 
Helena, Montana 59620
25 cm long in two sections of this river. 

Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     7

Median estimates for number of trout 25 cm 
become recruitment limited, resulting in 
and longer per km in the 8.9 km-long Craig 
population declines (McMahon et al. 2001, 
section are 1,793 (range =1,096-3,185) for 
Leathe et al. 2002, Munro 2004). Despite 
rainbow trout and 306 (range =91-827) for 
15 years of M. cerebralis presence in the 
brown trout (Fig. 1: Grisak et al. 2010). In 
most productive spawning tributaries, there 
the 6.6 km-long Pelican Point section median 
has been no measurable reduction of adult 
estimates are 885 (range =389-2,360) for 
rainbow trout in the Missouri River (Grisak 
rainbow trout and 193 (range =84-427) for 
2010). We hypothesized that rainbow trout 
brown trout (Fig. 1: Grisak et al. 2010). 
spawning behavior may be responsible for 
In 1995, Myxobolus cerebralis, a 
the stable populations.
myxosporean parasite that causes whirling 
In this study we (1) describe annual 
disease in rainbow trout, was discovered in 
timing of spawning migrations and 
Little Prickly Pear Creek; one of the most 
locations to test the hypothesis of localized 
productive trout spawning tributaries of the 
spawning behavior, (2) estimate the amount 
Missouri River below Holter Dam (Leathe 
of spawning that occurs in the Missouri 
et al. 2002, Munro 2004). Within three 
River compared to its tributaries, (3) 
years, clinical signs of whirling disease, 
describe spawning migration behavior over 
such as black tail and whirling behavior, 
multiple years, (4) describe the magnitude 
were observed in wild age-0 rainbow trout 
of spawning migrations over multiple 
in Little Prickly Pear Creek (Grisak 1999). 
years, (5) estimate the scale at which repeat 
The parasite’s range expanded into the 
spawners return to previous spawning sites 
Missouri River and some additional spawning  as a measure of spawning site fidelity, and 
tributaries over the next four years (Grisak 
(6) summarize the  M. cerebralis monitoring 
1999, Leathe 2001). This development 
data collected over the past 13 years. 
raised great concern among fishery managers 
because from 1991 to 1994 the Madison River,  Study area
Montana and the Colorado River, Colorado 
The study area was located in central 
experienced sharp declines of rainbow trout, 
Montana and consisted of a 41.6-km reach 
approaching 90 percent, due to M. cerebralis 
of the Missouri River and its tributaries 
(Vincent 1996, Nehring and Walker 1996). 
extending from Holter Dam downstream to 
Fearing similar population declines in the 
the Pelican Point Fishing Access Site  
Missouri River, fishery managers investigated  (Fig. 1). At the lower end of the study area, 
the relative recruitment of rainbow trout from  the Missouri River drains 47, 187 km2 
tributaries where the parasite was established 
and had a gradient of 0.88m/km. Three 
to total rainbow trout production (Leathe 
upstream hydroelectric dams (Canyon Ferry, 
2001, Munro 2004). Monitoring M. cerebralis  Hauser and Holter) regulate flows. Mean 
in the Missouri River and its tributaries began  annual discharge measured below Holter 
in 1997 to evaluate the spread of the parasite 
Dam from 1946 to 2010 ranged from 85 
and to develop population risk assessments. 
to 241 m3/s (USGS unpublished data for 
This monitoring has continued for 13 years.
station 06066500) and mean annual water 
Historically, fishery managers 
temperature was 8.4oC (SE: 0.17). The 
hypothesized that rainbow trout life history in  principal spawning tributaries in the study 
this area was locally adapted, believing that 
area were Little Prickly Pear Creek (which 
fish in the upstream portion of this section 
included Lyons Creek and Wolf Creek), the 
spawned in Little Prickly Pear Creek, and fish  Dearborn River, and Sheep Creek (Fig. 1). 
in the downstream portion spawned in the 
Little Prickly Pear Creek originates near 
Dearborn River (Horton et al. 2003, Horton 
the Continental Divide and flows easterly 
and Tews 2005, Horton and Hamlin 2006). 
from the Rocky Mountains for 57 km 
The presence of M. cerebralis in both of 
before entering the Missouri River 3.8 km 
these spawning tributaries fueled speculation 
downstream of Holter Dam. Little Prickly 
that the adult populations would ultimately 
Pear Creek basin drains 1,026 km2. Mean 
Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     7

Figure 1. Missouri River-Holter Dam tail-water study area with population estimate sections, 
M. cerebralis
 monitoring sites, and M. cerebralis infection severity.
annual discharge from 1963 to 2010 ranged 
from 1.25 to 5.06 m3/s (USGS unpublished 
data for station 6071300) and mean annual 
Radio Telemetry
water temperature was 7.7oC (SE: 1.03). 
We used radio telemetry to describe 
The Dearborn River originates on the east 
rainbow trout spawning migrations from 
slope of the Rocky Mountain front, near the 
2008 through 2010 and estimate spawning 
Continental Divide, and flows easterly for 
site fidelity among years. We divided the 
107 km before entering the Missouri River 
Missouri River portion of the study area 
25.3 km downstream of Holter Dam. This 
into five sections (each 8.3-km long) and 
basin drains 1,418 km2 and the mean annual 
dedicated an equal number of transmitters 
discharge from 1946 to 2010 ranged from 
for each section. We tested two hypotheses 
1.7 to 10.3 m3/s (USGS unpublished data 
relating to the scale at which trout migrate 
for station 6073500). Mean annual water 
to spawn. The first hypothesis, derived from 
temperature was 7.7oC (SE: 0.18). The North 
previous managers, was whether fish from 
and South forks of Sheep Creek originate in 
the upper half (sections 1 and 2) of the study 
the Big Belt Mountains and each flows for 
area would preferentially spawn in Little 
approximately 10 km before forming Sheep 
Prickly Pear Creek and fish from the lower 
Creek. Sheep Creek flows westerly for 3.3 
half (sections 3, 4 and 5) would spawn in 
km before entering the Missouri River 37.8 
the Dearborn River. The second hypothesis 
km downstream of Holter Dam. The Sheep 
we tested was whether all fish would spawn 
Creek basin drains 96 km2. Temperature data 
within the 8.3 km-long section of the 
were not available for Sheep Creek.
Missouri River in which they were tagged. 
8          Grisak et al.
Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     9

Three of the sections (1,3,5) had tributaries 
and times by documenting a fish’s furthest 
that were considered part of the section.
movement during the spawning season and 
Fish were captured using boat 
assuming it spawned in the area at that time 
electrofishing (smooth DC, 300 V, 4 A). We 
(Grisak 1999, Burrell et al. 2000, Henderson 
selected individual rainbow trout ranging 
et al. 2000, Pierce et al. 2009).  
from 40 to 60 cm total length (TL) to ensure 
Locations of tagged fish were recorded 
fish  were sexually mature, young enough 
on a map with longitudinal distances of the 
to spawn in the next two seasons, and large 
Missouri River and its tributaries marked to 
enough  to carry the radio transmitter (<6.7 
the nearest 0.1 km. The beginning reference 
% total body weight) without influencing 
point (river kilometer [rkm] 0.0) in the 
behavior (Brown et al. 1999). We surgically 
Missouri River was Holter Dam, located on 
implanted radio transmitters (Lotek MCFT-
the upstream end of the study area (Fig. 1). 
3A, 148 MHz, 6.7g in-water weight) in 
For tributaries, the beginning reference point 
the abdomen of six trout in each section 
(rkm 0.0) was the mouth of each stream 
following the external antennae procedure 
and longitudinal measurements continued 
described by Cooke and Bunt (2001). Fish 
were tagged approximately six months 
Redd Counts
before the spawning season to allow them 
to acclimate to transmitters. Because of 
We counted spawning redds in the 
predation by osprey (Pandion haliaetus), 
tributaries and the mainstem Missouri 
bald eagle (Heliaeetus leucocephalus), 
River from 2007 to 2010 to estimate the 
and great blue heron (Ardea herodias) and 
spatial distribution and relative magnitude 
angler transmitter returns, four radios were 
of rainbow trout spawning  each year. Redd 
recovered and re-implanted in new fish. We 
counts were conducted weekly by surveying 
tagged 23 female and 7 male rainbow trout 
index sections of tributary streams starting 
(range 42.4-51.1 cm TL). In the upper half 
on 15 March (Fig 1).  The peak of the 
of the study area we tagged 14 female and 4 
spawning run for each stream was calculated 
male fish and in the lower half we tagged 9 
using the mean value of all redds counted 
female and 3 male fish. Mean body weight 
in the index sections and determining the 
loading of the transmitters was 1.65 percent 
dates of the peak.  At the point when few 
(SE: 0.08). 
additional redds and few or no spawning 
Tagged fish were detected using a 
fish were observed in the index sections, we 
mobile radio-telemetry receiver (Lotek SRX 
conducted an expanded basin-wide count in 
400 W5) and a truck, jet boat, or airplane.  
each stream within one week (Grisak 1999). 
Stationary radio receivers (Lotek SRX 400 
Redd counts on the Missouri and Dearborn 
W7AS) were positioned at the mouths of 
rivers were conducted using a Bell OH-58 
Sheep Creek, Dearborn River and Little 
helicopter with the doors removed flying 
Prickly Pear Creek, and at the Craig Bridge 
≈75 meters above the water’s surface.  Two 
(Fig. 1).  Receivers at tributary mouths 
observers wearing polarizing eyeglasses 
provided precise dates and times when fish 
counted redds and recorded the locations 
entered or exited the tributaries.
with a GPS unit. The spatial distribution of 
Mobile surveillance was increased 
redds in the rivers was determined using 
during spawning seasons in order to locate 
ArcGIS. Two-way communication between 
spawning sites. We had a priori knowledge 
the observers reduced the likelihood double-
from previous spawning studies in this area 
counting redds in the center of the rivers. 
that  radio-tagged rainbow trout ascended 
M. Cerebralis 
tributaries and stayed at redd sites for a 
mean 4.7 days (range 1-22) before moving 
Monitoring of M. cerebralis was 
downstream (Grisak 1999). Given the 
conducted according to protocols described 
difficulty in determining exact spawning 
by Baldwin et al. (2000) and Vincent (2002). 
sites, we estimated spawning locations 
Caged sentinel fish (60 age-0 rainbow trout) 
Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     9

were placed in streams for 10-d periods 
starting point was the section where each 
when water temperatures were conducive to 
fish was located on 1 January. We used 
promote infections in rainbow trout (9-17oC: 
chi-square goodness-of-fit tests (α<0.05) to 
Vincent 1999). An Onset temperature data 
determine whether the predicted number of 
recorder attached to the cage recorded water 
fish that spawned within their annual starting  
temperatures on 30 minute intervals during 
section or the upper and lower half of the 
the exposure period. In order to measure the 
study area was significantly different from 
peak and range of infection severity at many 
that observed. 
of the sites, multiple 10d sequential sentinel 
We hypothesized that the distance 
fish experiments were conducted when mean 
rainbow trout moved to spawning sites 
daily water temperature ranged from 9 to 
would be similar among the five groups. 
17oC. After exposure, the fish were moved 
Migration distances from annual starting 
to a laboratory facility where the parasite 
sections to spawning locations were 
was allowed to mature in infected fish over 
evaluated using single factor analysis of 
a 90-d period. The fish were euthanized with 
variance (ANOVA: α<0.05) to determine 
tricane methanesulfonate (MS-222) and 
whether magnitudes of movements were 
prepared for histological analyses by first 
significantly different among sections. 
removing and then preserving the head and 
Histopathology and water temperature 
gill arches in formalin. Histological analyses 
data were analyzed using logistic regression 
were performed on median plane sections 
(Peng et al. 2002). Mean histology scores 
of fish heads, which involved quantifying 
and mean daily water temperature were the 
the abundance and severity of M. cerebralis 
ranked categories. Histology scores were 
lesions in cranial tissue. 
ranked according to risk of population 
The severity of infection was rated on 
impacts (>2.75 = high, <2.74 = low). A 
a 0-5 scale for each fish and histological 
maximum likelihood estimate was used to 
scores were averaged for each lot (Hedrick 
determine if there was a significant (α<0.05) 
et al. 1999, Baldwin et al. 2000, Vincent 
relationship between water temperature and 
2002). Population risk assessments were 
risk category and an odds ratio estimate was 
developed based on reports that mean 
used to determine the odds of sentinel fish 
histological score of 2.75 and higher 
test lots being in one of the risk categories 
usually cause declines in wild rainbow trout 
and what influence temperature had on the 
populations (Vincent 2002). From this we 
odds ratio. Tests were performed using the 
developed three risk categories based on 
proc logistic procedure in SAS. 
mean histology score (>2.75 = high, < 2.74 
= moderate, 0.00 = no risk).
We also measured clinical signs of 
Radio Telemetry
whirling disease such as deformed mandible/
maxilla, shortened operculum, dolphin head 
We relocated the 30 radio-tagged 
and spinal deformity (Vincent 2002) while 
rainbow trout a total of 1,577 times and 
handling fish during annual rainbow trout 
mean relocations per fish was 34 (range 
population estimates.  
Rainbow trout spawning was not 
Data Analysis
localized to the half of the study area 
Because we evaluated fish movements 
in which they were tagged (χ2=9.76, 
over multiple years, it was necessary to 
df=1, P=0.009). We found no significant 
define a starting point for fish movements 
difference between the number of fish 
each year. For the first year, the starting 
that spawned within or outside of the 
point was the 8.3 km-long section of 
five sections (χ2=0.08, df=1, P=0.39). 
the Missouri River where each fish was 
The distance that rainbow trout migrated 
collected and implanted with a radio 
to spawning locations was significantly 
transmitter. For each successive year the 
different among the five sections (ANOVA: 
10          Grisak et al.
Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     11

F=3.44; df=3,20; P=0.03). Fish from 
7.1 (SE: 2.7) km to spawning sites. 
the uppermost section (Section 1) of the 
Overall, radio-tagged rainbow trout 
Missouri River moved the greatest mean 
were presumed to have spawned, based on 
distance (28.9 km, SE: 9.0) to spawning 
their movements, over a 79 d period from 3 
locations. Fish from each section entered 
March to 5 May. Over the three-year period, 
the Dearborn River to spawn, yet 36 
69 percent of rainbow trout spawned only 
percent of the Dearborn spawners migrated 
in the Missouri River, 31 percent spawned 
downstream from section 1. 
in a tributary at least one time, 12  percent 
In 2008, 88 percent of the radio tagged 
spawned in tributaries multiple times, 73 
rainbow trout made spawning migrations and 
percent spawned in consecutive years, 27 
62 percent of these moved outside of their 
percent spawned only once and, 15 percent 
starting section to spawn. Of the tagged fish 
spawned in a different river or stream. 
that made spawning migrations, 73 percent 
Tagged fish traveled a mean 16.4 (SE: 3.7) 
spawned in the Missouri River and 27 percent 
km to reach spawning sites (Fig 2).
spawned in the Dearborn River.  Tagged fish 
After spawning, 21 percent of the 
migrated a mean 17.0 (SE: 4.0) km to reach 
rainbow trout returned to within 1.6 km 
spawning sites (Table 1). 
of their starting point for that year, but the 
In 2009, 81 percent of the radio-tagged 
remaining fish moved to new locations 
rainbow trout migrated to spawning sites 
after spawning. Only 8 percent of the 
and only 20 percent of these moved outside 
rainbow trout displayed repeat tributary 
of their starting section to spawn. The mean 
spawning in two successive years. The mean 
distance fish traveled to reach spawning sites 
distance between spawning locations for 
was 13.1 (SE: 6.3) km (Table 1). Only one 
repeat tributary spawners was 6.5 (SE: 0.2) 
fish from each section traveled outside of their 
km. Missouri River mainstem spawners 
respective section to spawn in 2009. Fish from 
accounted for over half (58%) of the sample.  
sections 2 and 4 moved the greatest distance 
Mean distance between spawning locations 
to spawning sites (81.6 km and 59.6 km, 
for repeat mainstem spawners was 6.1 (SE: 
2.1) km. The average total distance rainbow 
In 2010 only half of the radio tags were 
trout traveled over the three year period 
transmitting and only 20 percent of the fish 
was 68.5 km (SE: 17.4). Only 11 percent 
with active tags moved outside of their home 
traveled more than 160 km in any one 
section to spawn. The mean distance fish 
calendar year. Radio- tagged rainbow trout 
traveled to reach spawning sites was 20.8 (SE: 
spawned in each of the five sections in the 
13.9) km (Table 1), but this mean estimate 
Missouri River and in the lower 52.8 km of 
is heavily influenced by one fish that left the 
Dearborn River (Fig 2.). None of the radio 
study area in late 2009 and resided near the 
tagged fish entered the Little Prickly Pear 
town of Great Falls over the winter. In early 
drainage (Little Prickly Pear Creek, Lyons 
2010, it traveled 116.8 km from its starting 
Creek, Wolf Creek) or Sheep Creek during 
section to spawn in the Dearborn River for the 
the study. Over half (52%) of rainbow trout 
second time in three years. Excluding this fish, 
traveled downstream to reach spawning 
the remaining fish traveled a mean distance of 
Table 1. Spawning migration characteristics of radio-tagged rainbow trout in the Missouri 
River, Montana, 2008-10.
% Of Tagged Fish
Mean Km Traveled
Standard Error
That Spawned
To Spawn
  All years 
Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     11

Figure 2. Distance radio-tagged rainbow trout traveled to spawning sites upstream (+km) 
and downstream (-km) by section, Missouri River, Montana 2008-2010. Gradient filled bar 
represents a fish that left the study area and then returned to spawn.
Redd Counts
11 redds at rkm 14.2, 2 redds at rkm 16.9, 2 
In April 2007, the Missouri River 
redds at rkm 18.4 and 6 redds at rkm 22.7. 
portion of the study area was surveyed to 
From 2008-10 the spatial distribution of 
identify the spatial distribution of redds. The 
redds spanned 122 km of river and stream 
mean distance between redds in the Missouri 
annually. Overall, 28 percent of the mean 
River was 1.89 (SE: 0.21) km. Spawning 
number of annual redds occurred in the 
redds were distributed throughout the entire 
Missouri River and 72 percent occurred 
longitudinal reach of the Missouri River.  
in tributaries (Table 2). The mean annual 
During the Spring of 2010 weather 
number of redds for the entire system was 
conditions (wind, overcast, rain) and flow 
4,071 (SE: 378) (Table 2).  Little Prickly 
(122 m3/s) in the Missouri River were 
Pear Creek had 44-47 percent of the total 
favorable for observing redds from the air 
annual redds, the highest mean number of 
and 39  percent (1,644) of the redds we 
redds and the highest mean density (92/km) 
counted occurred in the Missouri River 
of redds. The Dearborn River had the lowest 
(Table 2). We observed redds in deep 
mean density at 21/km (Table 2). 
water (> 2 m deep) at six locations in the 
M. Cerebralis 
Missouri River. The greatest number of 
deep water redds was 47 which occurred at 
Little Prickly Pear Creek had the 
rkm 3.5, followed by 24 redds at rkm 4.9, 
highest M. cerebralis infection scores in 
11 of 13 years (Table 3). Other streams 
12          Grisak et al.
Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     13

Table 2. Rainbow trout redds counted during the basin wide surveys and redd density (redds/
km) for the Missouri River and tributaries, Montana 2007-10.
Number Of Redds (redds Per Km)
Dist (km)  2007
Rel %
Missouri River 

1644  (40)  1644  (40) 
Dearborn River 

632  (21) 
632  (21) 
Lit le Prickly Pear Cr.  19.5  2125 (109)  1461  (75) 

  1793  (92) 
Lyons Creek 
11.2    847  (76)    897  (80) 

386  (34) 
710 (63) 
Wolf Creek 
12.5  1289 (103)    678  (54) 

 1451 (116)  1139  (91) 
Sheep Creek 
3.2    282  (88)    286  (89) 

234  (73) 
267 (83) 


b - high turbid flows precluded counting redds
with multiple high (> 2.75) scores over 
There was a significant relationship 
the monitoring period were Wolf Creek 
between histopathology rank and 
(58%) and the Dearborn River (50%). In 
temperature (maximum likelihood estimate, 
the Missouri River, the Craig monitoring 
P=0.02). The mean daily water temperature 
site had high scores in 6 of 13 (46%) years 
in test lots with infection scores 2.75 and 
and was the only Missouri River site with 
greater was 11.6oC (range = 8.7-16.5), 
multiple scores greater than 2.75.  Overall, 
whereas it was 12.9oC (range = 9.6-18.1) 
53 percent of the habitat used by rainbow 
for test lots with scores 2.74 and less.  The 
trout for spawning had a high severity of 
odds of sentinel fish test lots being classified 
M. cerebralis, 38 percent had low to 
as low risk was higher by 1.097 for each 
moderate severity, and 9 percent had no 
infection, as detectable by sentinel fish. 
degree temperature increase and the odds 
Lyons Creek remains the only stream in our 
increased by a factor of 2.519 for every 10 
study area where the parasite has not been 
degree increase in temperature.
From 1997 to 2009, 5.3  percent 
Discussion and 
(range 1.7-8.9%) of the rainbow trout 
sampled during the annual population 
estimates displayed clinical signs of 
A variety of factors may influence the 
M. cerebralis infection (jaw deformity, 
relative impact that M. cerebralis has on 
shortened operculum, dolphin head and 
rainbow trout populations. Among these 
spinal deformity: Table 4). Dolphin head 
are life history and spawning behaviors 
was the deformity that occurred the most. 
(Downing et al. 2001, Pierce et al. 2009), 
The highest numbers of deformed fish 
differential infection susceptibility among 
occurred in 2001, when 8.9 percent of the 
strains of fish (Vincent 2002), habitat 
fish handled during the population estimate 
conditions (Granath et al. 2007) and water 
displayed clinical signs of M. cerebralis 
temperature at time of emergence and during 
early-life rearing (Baldwin et al. 2000). 
Myxobolus cerebralis infection in 
Where native, rainbow trout have 
sentinel fish occurred from 6 May to 3 July 
demonstrated discrete spawning behaviors 
(median = 31 May). The dates of exposure 
and site fidelity for spawning, feeding and 
for test lots with mean infection scores of 
overwintering (Meka et al. 2003). The 
2.75 and greater ranged from 15 May to 3 
data from this study did not support the 
July (median = 30 May) and the dates of 
hypotheses that radio-tagged fish would 
exposure for tests lots with mean infection 
spawn within their annual starting section 
scores of 2.74 and less ranged from 5 May 
or that spawning was localized to either half 
to 3 July (median = 1 July).  
of the study area. The location of rainbow 
Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     13

trout at the point of tagging or on 1 January 
was not an indicator of where they spawned. 
4.70 --- 3.94 0.00 0.85 1.44 1.56 4.13 ---
The mean distance and maximum distance 
that fish traveled to spawning locations 
was greater than has been reported for 
1.96  0.00  2.70  0.00  4.02  ---  0.04  4.27  .34  0
other rainbow trout populations (Adams 
1996, Lisac 1996, Meka et al. 2003). In 
many instances fish traveled over suitable 
4.72    - 4.58  ---  2.02  1.96  1.80  .00  --- 
habitat, as evidenced by existing redds, to 
reach eventual spawning destinations. Such 
behavior suggests a determination to reach a 
0.11  0.00  0.15  ---  0.15  0.13  0.00  4.13  0.00 
specific area for spawning, but the majority 
of fish lacked spawning site fidelity between 
at select sites in the Missouri River below 
4.97  0.00  4.42 
3.78  3.00  1.66  4.72  0.08 
Low site fidelity may be an important 
life history characteristic that spreads the 
risk posed by M. cerebralis over spawning 
4.87  0.00  4.80 
3.33  2.00  0.97  4.59 
locations with a broad range of infection 
M. cer
severity. There is some speculation of 
differing virulence of M. cerebralis strains 
4.92  0.00  3.98  0.00  1.76  0.34  0.12  2.58  0.00 
based on geographical location (Modin 
1998). Other studies in our area showed 
a high degree of spatial variation and M. 
4.96  0.00  3.96  0.20  2.22  0.19  0.10  1.18  0.00 
cerebralis infection intensity in the primary 
host Tubifex tubifex (McMahon et al. 
2001). The variable infection severity in 
4.43  0.00  4.47  0.16  3.47  1.60  .26  0.07  0.00 
T. tubifex
 may contribute to the variable 
levels of disease severity seen among stream 
salmonid populations (Stevens et al. 2001). 
4.76  0.00  0.78  0.14  2.82   0 ---  0.00  0.00 
Relating this to infection in fish or fish 
population declines has proven difficult in 
other areas. For example, despite variable 
0.20  0.28  3.88 
 2 0.00  0.00 
M. cerebralis
 infection in T. tubifex among 
varied habitats in the Rock Creek drainage 
of Montana, lower infections in this host did 
3.98  0.00  1.24  0.00  1.83  0.00  0.02  0.00  0.11 
not necessarily translate to lower infections 
in fish because the infectious stage of the 
parasite (triactinomyxon) spreads rapidly in 
2.94  ---  ---  0.00  0.66  ---  ---  ---  0.00 
flowing water (Granath et al. 2007). 
The low spawning site fidelity observed 
Values represent the highest mean infection severity measured for any 10 d period for each year
in our study and the broad spatial scale over 
which fish redistributed after spawning 
irth d
were in contrast to other research but are 
on  oint  87 
behaviors that could facilitate avoidance 
reek - W
of unfavorable habitat conditions (Adams, 
olter D
wy 2
raig  id C elican P
1996, Adams 1999, Meeka et al. 2003, 
ear C
Pierce et al. 2009). Whether the difference  
iver -  H iver - C iver - M iver - P iver - H
Mean cranial lesion histopathology scores (0-5) from sentinel rainbow trout exposed to 
represent behavioral changes in response 
rickly P reek - m
reek - m
reek - m
to M. cerebralis, or differences in behavior 
born R
have been longstanding components of the 
Able 3. 
olf C issouri R issouri R issouri R issouri R
Holter Dam and its tributaries. 
Litle P Lyons C W M M M M Dear Sheep C
trout life history, remains unknown. 
14          Grisak et al.
Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     15

Table 4. Relative occurrence of clinical signs of M. Cerebralis infection in wild rainbow trout 
sampled during population estimates in the Craig section of the Missouri River, Montana 

Total Rainbow  % Of Deformed
Deformed  trout Handled  in Total Catch
1997 45  19  18  10 
1998 23  15 107  8 
1999 54  20 116  40 
2000 87  19  111  60 
2001 176  16  158  46 
2002 121  15  143  58 
2003 68  33  84  74 
2004 67  19  97  40 
2005 58  47  80  57 
2006 34  24  71  48 
2007 19  37  52  44 
2008 20  35  55  53 
2009 14  24  55  73 
  BB – bent back/spinal deformity
  SOP – shortened operculum
Pear, Lyons, Wolf creeks: Grisak 1999), 77 
  DH – dolphin head
percent decrease in the Dearborn River and 
  JAW – deformed mandible/maxil a
64 percent increase in the Missouri River 
After spawning, the majority (79%) of 
(Leathe et al. 1988). Such changes may 
tagged fish moved to sites other than their 
contribute to the persistence of rainbow trout 
pre-spawn locations. In other areas where 
in the Missouri River-Holter Dam tailwater 
M. cerebralis has caused reductions in 
fishery because M. cerebralis infections 
rainbow trout populations, 75 to 76 percent 
measured in the Missouri River are less 
of spawners returned to within 1.5 to 2.0 
severe than in its tributaries. By comparison, 
km of their pre spawn location (Downing 
spawning by rainbow trout in Montana’s 
et al. 20032, Pierce et al. 2009).  The fact 
Madison River system occurs mostly in the 
that some fish spawned only once during the 
mainstem and is concentrated in the upper 
study suggests the possibility of alternate 
reaches where the majority of emerging 
year spawning by a low percentage (27%) of 
young are susceptible to a high prevalence 
rainbow trout in the Missouri River system. 
of M. cerebralis (Downing et al. 2002). 
Alternate year spawning has been reported 
Other research has shown that life histories 
in female rainbow trout in tributaries to the 
of rainbow trout could lead to higher M. 
King Salmon River in Alaska (Adams 1999) 
cerebralis infection if young fish have to 
and is typical of fish with slow metabolism 
migrate through areas with a high density of 
that cannot store the proper amount of 
infected T. tubifex (Sandell et al. 2001). 
energy necessary to develop eggs and make 
Over the 13-year monitoring period 
migrations to spawning areas (Palstra et al. 
the adult rainbow trout population in the 
Missouri River remains near the long term 
Comparing these results with the 
median despite high (>2.75 histopathology 
limited historic rainbow trout life history 
score) M. cerebralis severity measured 
information for this area suggest a higher 
in sentinel fish in over half (53%) of the 
use of the Missouri River for spawning 
available spawning habitat (Table 2: Grisak 
(Munro 2004, Leathe et al. 1988, McMahon 
et al. 2010). Three of the four Missouri 
et al. 2001). Changes in the number of 
River monitoring sites (Holter, Mid Canon, 
redds included a 27 to 37 percent decrease 
Pelican Point) had relatively low infection 
in the tributary streams (Little Prickly 
severities compared to the Craig site. The 
Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     15

high infection severity at the Craig site 
uncertain contribution from 12 different 
may be related more to its proximity to 
strains there is no way of knowing which 
Little Prickly Pear Creek, located 7.4 rkm 
strains have succeeded or persisted over 
upstream, than to the infection severity in 
time.  The consequences of such a stocking 
the Missouri River. Granath et al. (2007) 
history may be a population that has a 
reported that infections in sentinel fish may 
high level of gene flow and lacks selective 
be the result of triactinomyxons produced 
pressures for distinct behavior patterns and 
hundreds or thousands of meters upstream.  
spawning site fidelity (Allendorf and Phelps 
In essence, localized heavily degraded 
1981, Slatkin 1985, Slatkin 1987), as our 
habitats, suitable to infected T. tubifex, may 
results show. These concepts have been 
serve as a major source of infection at a 
studied extensively in Pacific salmon species 
broader scale than would be assumed from 
when evaluating the interaction of wild and 
sampling the distribution of T. tubifex worms 
hatchery-reared fish and potential changes 
or from the results of sentinel fish studies. 
in local genetic adaptations (Hendry et al. 
The monitoring site that had the second 
1996, Quinn 1993, Quinn et al. 2006).         
highest infection severity scores was the 
The 79-d rainbow trout spawning 
Mid Canon site, which is located 2.4 rkm 
period observed in our study indicates 
downstream of the Dearborn River.     
the fry emergence period would be broad 
Histopathology rank was correlated 
enough to span the entire thermal range of 
to water temperature. Mean water 
M. cerebralis infection susceptibility. As 
temperature at sites with high severity 
such, fish of each cohort could be either 
infections was slightly less than at sites 
unaffected or exposed to the parasite at 
with low to moderate severity infections 
levels which they likely survive. With such 
suggesting infection severity peaks as 
a broad spawning period, fry from early and 
water temperature rises and then drops as 
late spawning fish could emerge at times 
water temperature increases above 12oC. 
when water temperatures are outside of the 
Other field experiments with rainbow trout 
range of  M. cerebralis infection (Vincent 
showed the highest infection intensities 
2002), or when high spring flows dilute the 
occurred when water temperatures were 
concentration of triactinomyxons and flush 
between 12 and 16oC then declined rapidly 
fry out of tributaries into the Missouri River 
as mean daily water temperatures decreased 
where the infection severity is less. The 
or increased from these optimum water 
timing of outmigration of young rainbow 
temperatures (Vincent 2002). 
trout can influence M. cerebralis infection. 
There is evidence of differing infection 
Infection can occur for a minimum of 60 
susceptibility among rainbow trout strains 
to 65 days post hatch (Baldwin et al. 2000, 
(Vincent 2002). However, attributing the 
Vincent 2002, Downing et al. 2002). Fish 
stability of rainbow trout populations in the 
that migrate out of natal streams at age-0 
Missouri River to a particular strain would 
have been shown to have a higher infection 
be difficult because rainbow trout are not 
severity than fish that migrated at age-1 
native to this region and the population 
(Leathe et al. 2002).  
was founded by over 40 years of stocking 
This study found that the only 
undesignated strains. Since 1973, the 
observable impact of M. cerebralis infection 
population  has been influenced by at least 
to the adult population was the small number 
12 strains of rainbow trout (Eagle Lake, 
(<9%) of adult rainbow trout that we 
Arlee, Erwin, DeSmet, Madison River, 
encountered during population estimates that 
Shasta, McConaughy, Bozeman, Lewistown, 
had clinical signs of M. cerebralis infection. 
Winthrop, Beula, Wild) stocked in the 
Despite no outbreak of whirling disease, 
upstream reservoirs (Holter, Hauser, Canyon 
a small segment of the rainbow trout 
Ferry: MFWP unpublished data).  Given the 
population persists with a mild infection. 
unknown origin of the majority of rainbow 
The spawning behaviors we have described 
trout stocked in this area, as well as the 
likely contribute to the stability of the 
16          Grisak et al.
Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     17

rainbow trout populations in the Missouri 
Baldwin,T.J.,  E. R.Vincent, R.M. Silflow, 
River, even after 15 years of M. cerebralis 
D. Stanek. 2000. Myxobolus cerebralis 
presence in the study area.
infection in rainbow trout (Oncorhynchus 
) and brown trout (Salmo trutta
exposed under natural stream conditions. 
Eli McCord, Lance Breen and Kelly 
Journal of Veternary Diagnostic 
Smith assisted with field and lab work. We 
Investigation. 12:312 321.
are grateful for the highly skilled pilots of 
Brown, R.S., S.J. Cooke, W.G. Anderson 
Montana Fish, Wildlife & Parks; Jim Lowe, 
and R.S. McKinley. 1999. Evidence to 
Joe Rahn, Bill Winninghoff, Jim Heppner 
challenge the 2% rule for biotelemetry. 
and Greg Thielman who flew many hours in 
North American Journal of Fisheries 
the most challenging of weather and terrain 
Management. 19:867-871. 
conditions. Thanks to pilots Andy Taylor 
Burrell, K.H., J.J. Isely, D.B. Bunnell, 
and Scott Snider for their aviation services. 
D.H. Van Lear and C.A. Dolloff. 2000.  
Montana Fish, Wildlife & Parks-Information 
Seasonal movement of brown trout in a 
Services provided maps and -Whirling 
southern Appalachian River. Transactions 
Disease Research Program provided 
of the American Fisheries Society 
monitoring data for our study area. We thank 
the Lord Ranch, Sterling Ranch, Chesnutt 
Valley irrigators, Josephine Lahti and Robert 
Cooke, S.J. and C.M. Bunt. 2001. 
Neary for providing access to the Missouri 
Assessment of internal and external 
River. Dr. J. Richard Alldredge assisted 
antenna configurations of radio 
with analysis of the histopathology data. We 
transmitters implanted in smallmouth 
thank the following people for their critical 
bass. North American Journal of 
reviews of earlier drafts; Dr. Bill Gould, 
Fisheries Management 21:236–241.
Steve Leathe, Dr. Tom McMahon, and 
Downing, D.C., T.E. McMahon, B.L. 
five anonymous reviewers. This study was 
Kerans and E.R. Vincent. 2002. Relation 
funded by PPL-Montana as part of FERC 
of spawning and rearing life history 
license 2188.   
of rainbow trout and susceptibility to 
Myxobolus cerebralis infection in the 
Literature cited
Madison River, Montana. Journal of 
Adams, F. J. 1996. Status of rainbow trout 
Aquatic Animal Health 14:191–203.
in the Kanektok River, Togiak National 
Granath, W.O.,  M. A. Gilbert, E.J. Wyatt-
Wildlife Refuge, Alaska, 1993–1994. 
Pescador and E. R. Vincent. 2007. 
U.S. Fish and Wildlife Service, Alaska 
Epizootiology of Myxobolus cerebralis
Fisheries Technical Report 39, King 
the causative agent of salmonid whirling 
disease. Journal of Parasitology. Vol 93 
Adams, F. J. 1999. Status of rainbow trout 
No. 104-119.
in tributaries of the upper King Salmon 
Grisak, G. G. 1999. Missouri River rainbow 
River, Becharof National Wildlife 
trout spawning study, Missouri River 
Refuge, Alaska, 1990–1992. U.S. Fish 
young of the year trout investigations II 
and Wildlife Service, Alaska Fisheries 
and Montana statewide whirling disease 
Technical Report 53, King Salmon.
investigations. Progress report. Montana 
Allendorf, F.W. and S.R. Phelps.1981. 
Fish, Wildlife & Parks, Bozeman.
Use of allelic frequencies to describe 
Grisak, G.G., A.C. Strainer and P. D. 
population structure. Canadian Journal of 
Hamlin. 2010. Montana statewide 
Fisheries and Aquatic Science :18:1507-
fisheries management. Fisheries 
division federal aid job progress report 
F-113-R10. Montana Department of 
Fish, Wildlife & Parks. Great Falls.
Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     17

Hedrick, R.P., T.S. McDowell, M. Gay, 
Leathe, S.A., W.J. Hill and A. Wipperman. 
G.D. Marty, M.P. Georgiadis and E. 
1988. North central Montana trout 
MacConnel1. 1999. Comparative 
stream investigations. Job Progress 
susceptibility of rainbow trout 
Report, Federal Aid to Fish and Wildlife 
Oncorhynchus mykiss and brown trout 
Restoration Act, Montana Project No. 
Salmo trutta to Myxobolus cerebralis
F-46-R-1, Job 1-g. 
the cause of salmonid whirling disease. 
Leathe, S.A. 2001. An evaluation of juvenile 
Diseases of aquatic organisms. 37: 173-
trout production of Little Prickly Pear 
Creek and Dearborn River, Montana 
Henderson, R., J.L. Kershner and C.A. 
using rotary screw traps. Montana Fish, 
Toline. 2000. Timing and location of 
Wildlife & Parks, Great Falls.
spawning by nonnative wild rainbow 
Leathe, S., G. Liknes, T. McMahon, A. 
trout and native cutthroat trout in 
Munro, E.R. Vincent, and A. Zale. 2002. 
the South Fork Snake River, Idaho, 
Epidemiology of whirling disease in the 
with implications for hybridization. 
Missouri River, Montana: disease spread 
North American Journal of Fisheries 
and trout population response five years 
Management 20:584–596.
after discovery. Pages 19-20 in Eighth 
Hendry, A.P., T.P. Quinn and F.M. Utter. 
Whirling Disease Symposium, Denver, 
1996. Genetic evidence for the 
persistence and divergence of native 
Lisac, M.J. 1996. Length frequency, age 
and introduced sockeye salmon 
distribution and movements of rainbow 
(Oncorhynchus nerka) within Lake 
trout in the Negukthlik and Ungalikthluk 
Washington, Washington. Canadian 
Rivers, Togiak National Wildlife Refuge, 
Journal of Fisheries and Aquatic Science. 
Alaska, 1989-90. U.S. Fish and Wildlife 
53: 823–832.
Service, Alaska Fisheries Technical 
Horton, T. B. Tews. A. E., G. A. Liknes, 
Report Number 34. Dillingham.
W. Hill and D. Yerk.  2003. Statewide 
McMahon, T., A. Munro, A. Zale, B. Kerans, 
Fisheries Investigations. North central 
M. Gangloff, D. Vincent, G. Grisak, S. 
Montana Coldwater Stream Ecosystems.  
Leathe and G. Liknes. 2001. Relation 
Federal Aid Job Progress Report Federal 
of life history to whirling disease 
Aid Project Number: F-113-R1 and 
susceptibility in Missouri River rainbow 
F-113-R2.Montana, Fish, Wildlife and 
trout. Final report- year two. Montana 
Parks. Fisheries Division, Helena, MT.
State University, Bozeman.
Horton, T. B., and A. E. Tews.  2005.  
Meka, J.M., E.E. Knudsen, D.C.Douglas and 
Statewide Fisheries Investigations.  2003 
R.B. Benter. 2003. Variable migratory 
Annual Report.  Northcentral Montana 
patterns of different adult rainbow trout 
Coldwater Stream Ecosystems.   Federal 
life history types in a Southwest Alaska 
Aid Job Progress Report. Federal Aid 
watershed. Transactions of the American 
Project Number:  F-113-R3 & F-113-R4. 
Fisheries Society 132:717–732.
Montana Fish, Wildlife & Parks. 
Fisheries Division, Helena, MT.
Modin, J. 1998. Whirling Disease in 
California: A Review of Its History, 
Horton, T. B., and P. D. Hamlin.  2006.  
Distribution, and Impacts, 1965–1997, 
Statewide Fisheries Investigations.  
Journal of Aquatic Animal Health, 10:2, 
2004 Annual   Report.  Great Falls 
Management Area. Federal Aid Job 
Progress Report. Federal Aid Project 
Munro, A.R. 2004. Identification of life 
Numbers:  F-113-R4, F-113-R5. 
history variation in salmonids using 
Montana Fish, Wildlife & Parks. Helena, 
otolith microchemistry and scale 
patterns: implications for illegal 
introductions and for whirling disease in 
18          Grisak et al.
Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     19

Missouri River rainbow trout. Doctoral 
Sandell, T.A., H.V. Lorz, D.G. Stevens, 
dissertation. Montana State University, 
and J.L. Bartholomew. 2001. Dynamics 
Bozeman. 218 pp.
of  Myxobolus cerebralis in the Lostine 
Nehring, B.R. and P.G. Walker. 1996. 
River, Oregon: implications for resident 
Whirling disease in the wild: the new 
and anadromous salmonids. Journal of 
reality in the the Intermountain West. 
Aquatic Animal Health 13:142–150.
Fisheries 21(6):28-30.
Slatkin, M. 1985. Gene Flow in Natural 
Palstra, A.P., D. Crespo, G. E. E. J. M. 
Populations. Annual Review of Ecology 
van den Thillart and J. V.  Planas. 
and Systematics, Vol. 16, pp. 393-430.
2010. Saving energy to fuel exercise: 
Slatkin, M. 1987. Gene flow and the 
swimming suppresses oocyte 
geographic structure of natural 
development and downregulates ovarian 
populations. Science. 236:787-792.
transcriptomic response of rainbow 
Stevens, R., B. L. Kerans, J. C. Lemmon, 
trout Oncorhynchus mykiss American 
and C. Rasmussen. 2001. The effects 
Journal of Physiology - Regulatory, 
of Myxobolus cerebralis dose on 
Integrative and Comparative Physiology. 
Triactinmyxon production and biology 
of Tubifex tubifex from two geographic 
Peng, C.J., K.L. Lee and G. M. Ingersoll. 
regions. Journal of Parasitology, 87(2) 
2002. An introduction to logistic 
regression analysis and reporting. Journal 
Vincent, E. R. 1996. Whirling disease and 
of Educational Research. Vol. 96 No. 1, 
wild trout: the Montana experience. 
Fisheries 21(6):32–33.
Pierce, R., C. Podner, M. Davidson and E.R. 
Vincent, E. R. 1999. The relationship of 
Vincent. 2009. Correlation of fluvial 
time, temperature, fish life histories to 
rainbow trout spawning life history 
whirling disease infections. Whirling 
with severity of infection by Myxobolus 
Disease Symposium. Research in 
cerebralis in the Blackfoot River basin, 
Progress. Pp. 31-32
Montana. Transactions of the American 
Fisheries Society 138:251–263.
Vincent. E.R. 2002. Relative susceptibility 
of various salmonids to whirling disease 
Quinn, T. P. 1993. A review of homing and 
with emphasis on rainbow and cutthroat 
straying of wild and hatchery-produced 
trout. American Fisheries Society 
salmon. Fisheries Research 18:29–44.
Symposium 29:109-115.
Quinn, T.P., I.J. Stewart, and C.P. Boatright. 
2006. Experimental evidence of homing 
to site of incubation by mature sockeye 
Received January 9, 2012
salmon, Oncorhynchus nerka. Animal 
Accepted August 2, 2012
Behaviour 72: 941-949.
Rainbow Trout Spawning Characteristics and Relation to the Parasite Myxobolus cerebralis in the Missouri River, MT     19