Hydrobiologia (2009) 635:237–249
DOI 10.1007/s10750-009-9917-3
PRIMARY RESEARCH PAPER
What is ‘‘fallback’’?: metrics needed to assess telemetry tag
effects on anadromous fish behavior
Holly J. Frank Æ Martha E. Mather Æ
Joseph M. Smith Æ Robert M. Muth Æ
John T. Finn Æ Stephen D. McCormick
Received: 14 March 2009 / Revised: 31 July 2009 / Accepted: 2 August 2009 / Published online: 20 August 2009
Ó Springer Science+Business Media B.V. 2009
Abstract Telemetry has allowed researchers to
document the upstream migrations of anadromous fish
in freshwater. In many anadromous alosine telemetry
studies, researchers use downstream movements
(‘‘fallback’’) as a behavioral field bioassay for adverse
tag effects. However, these downstream movements
have not been uniformly reported or interpreted. We
quantified movement trajectories of radio-tagged
anadromous alewives (Alosa pseudoharengus) in the
Handling editor: K. Martens
Use of brand names does not confer endorsement by the U.S.
government.
H. J. Frank J. M. Smith
Massachusetts Cooperative Fish and Wildlife Research
Unit, Department of Natural Resources Conservation,
University of Massachusetts, Amherst, MA 01003, USA
M. E. Mather (&)
U. S. Geological Survey, Massachusetts Cooperative Fish
and Wildlife Research Unit, Department of Natural
Resources Conservation, University of Massachusetts,
Amherst, MA 01003, USA
e-mail: mather@nrc.umass.edu
H. J. Frank M. E. Mather J. M. Smith
R. M. Muth J. T. Finn
Department of Natural Resources Conservation,
University of Massachusetts, Amherst, MA 01003, USA
S. D. McCormick
Conte Anadromous Fish Research Center, U. S.
Geological Survey, Turners Falls, MA 01376, USA
Ipswich River, Massachusetts (USA) and tested blood
chemistry of tagged and untagged fish held 24 h. A
diverse repertoire of movements was observed, which
could be quantified using (a) direction of initial
movements, (b) timing, and (c) characteristics of bouts
of coupled upstream and downstream movements
(e.g., direction, distance, duration, and speed). Because
downstream movements of individual fish were almost
always made in combination with upstream movements, these should be examined together. Several of
the movement patterns described here could fall under
the traditional definition of ‘‘fallback’’ but were not
necessarily aberrant. Because superficially similar
movements could have quite different interpretations,
post-tagging trajectories need more precise definitions.
The set of metrics we propose here will help quantify
tag effects in the field, and provide the basis for a
conceptual framework that helps define the complicated behaviors seen in telemetry studies on alewives
and other fish in the field.
Keywords Alosine Tag effect
Movement Behavior
Introduction
Telemetry research using radio, acoustic, and passive
integrated transponder (PIT) tags is important for
anadromous fish research and management (McDowall,
1999; Lucas & Baras, 2000; Lassalle et al., 2008).
123
238
Tagging studies, however, are only useful if the tag does
not alter fish behavior compared to untagged fish
(Bridger & Booth, 2003; Rogers & White, 2007; Cooke
et al., 2008). Identifying tag effects in the field is
difficult because untagged fish cannot be tracked, and
consequently the complex movements of untagged and
tagged fish are difficult to compare. Although examining physiology and behavior of tagged and untagged
fish in the laboratory is possible, such studies are time
consuming, and confining migratory fish can cause
additional stress.
In the northeastern United States, the closely related
alewife (Alosa pseudoharengus) and blueback herring
(A. aestivalis), collectively referred to as river herring,
have historically and ecologically been an important
component of coastal rivers. Tracking within-river
movements of spawning adults of the genus Alosa is of
special interest to many researchers and management
agencies. Because anadromous shad and herring are
sensitive to handling, researchers often use ‘‘fallback’’
(i.e., downstream movement of an upstream migrating
anadromous fish following tagging) as a behavioral
bioassay to document adverse tag effects on alosines
(Beasley & Hightower, 2000; Hightower & Sparks,
2003; Bailey et al., 2004; Olney et al., 2006). Here, we
use a literature review to show that the present
language describing downstream movements is not
standardized. Then, we illustrate the diversity of
possible downstream movements. Finally, we propose
a standardized series of metrics for quantifying
post-tagging movements.
In the literature, many features of existing tagging
studies on anadromous shad and herring are similar
(Dodson et al., 1972; Bell & Kynard, 1985; Barry &
Kynard, 1986; Chappelear & Cooke, 1994; Beasley &
Hightower, 2000; Moser et al., 2000; Hightower &
Sparks, 2003; Acolas et al., 2004; Bailey et al., 2004;
Sprankle, 2005; Olney et al., 2006; Table 1). Of these,
81.8% focused on American shad (Alosa sapidissima),
9.1% examined blueback herring (A. aestivalis), and
9.1% used Allis shad (A. alosa). These studies either
quantified fish passage (54.5%) or sought to understand migratory behavior (45.5%). All these studies
used upstream migrating adult fish captured during the
spawning run. All such studies were undertaken in
river systems and most used fish obtained from fish
passage structures (45.5%) or in-river capture (45.5%;
no information provided 9.1%). Radio (‘‘R’’, 54.5%)
or acoustic (‘‘A’’, 45.5%) tags were gastrically
123
Hydrobiologia (2009) 635:237–249
implanted (100%). With one exception, these studies
were conducted without anesthetic (Acolas et al.,
2004). Fish were typically released at the capture site
(66.7%) or downstream of it (25.0%; no information
provided, 8.3%).
Although the conditions of the studies were similar,
individual researchers reported very different information about downstream movements post-tagging.
The number of fish tagged (N) ranged from 7 to 110
(Table 1). All these studies reported some proportion
of the study population to ‘‘fallback’’ (range, n =
1–87 individual fish; 8.6–100% of the tagged fish in
each study). While all these studies describe downstream movement, uniform terms were not used to
quantify this behavior. The term ‘‘fallback’’ was used
in 18.2% of the studies. This type of movement was
also described by phrases such as ‘‘swam or passively
drifted,’’ ‘‘moved,’’ ‘‘migrated,’’ and ‘‘drifted’’ downstream, as well as ‘‘dropback’’ and ‘‘left the study
area.’’ In studies where quantitiative measures were
reported, fish were listed as ‘‘falling back’’ when they
moved downstream at times ranging from\1 to 168 h
(7 days) post-release. In addition to the temporal
frame of reference, the spatial focus of ‘‘fallback’’
activity was highly variable. The distance that fish
moved downstream post-tagging in ‘‘fallback’’ activities ranged from \1 to 30 km. Of these studies,
45.5%, did not report a specific distance. While the
majority of researchers (63.6%) included ‘‘fallback’’
fish in the data analyses as long as the fish returned
upstream, 27.3% excluded ‘‘fallback’’ fish from analysis (no data reported, 9.2%). Hence, although the
concept of downstream movement was embraced by
most studies as a field diagnostic of adverse tag
effects, how researchers quantified this behavior
relative to time frame, spatial scale, and data analysis
was variable, preventing comparisons across studies.
The interpretation of downstream movements of
upstream migrating fish after tagging is an important
issue and has significant consequences for field
research, data analysis, and management. Unfortunately, our examination of the literature has shown that
there is little consistency in how ‘‘fallback’’ is reported.
Here, we use movement trajectories from our own field
research on alosines to construct a conceptual framework for organizing the diversity of possible downstream movements. Specifically, we ask: (1) What types
of downstream movements occur in upstream migrating
anadromous alewives post-tagging? (2) Were tagged
Citation
Species
Purpose
Tag
Release
in relation
to capture
Acolas et al. (2004)
Allis Shad
Behavior
A
Same or down
Bailey et al. (2004)
Am. Shad
Passage-up
R
NR
Barry & Kynard (1986)
Am. Shad
Passage-up
R
Down
Hydrobiologia (2009) 635:237–249
Table 1 A review of alosine telemetry studies showing authors of study, species, purpose of the study, tag type, location of release site in relation to capture site, N released, n
fallback, language used in the text to describe fallback, time period during which fish moved downstream (h), distance fish moved downstream (km), and whether ‘‘fallback’’ fish
were excluded from analysis
‘‘Fallback’’
N
n
Language
Time
period (h)
Distance
(km)
Excluded from
analysis
Yes (mortality)
23
2
‘‘Moved downstream’’
B24
B1
110
87
‘‘Downstream movement’’
B168
1.3 to [30
No
34
34
‘‘Drop back’’
\1
1–8
No
Beasley & Hightower (2000)
Am. Shad
Passage-up
A
Same
25
‘‘Several’’
‘‘Fallback’’
B10
NR
No
Bell & Kynard (1985)
Am. Shad
Passage-down
R
Down
36
28
‘‘Swam or passively drifted
downstream’’
B8
0.7–16.5
No
Chappelear & Cooke (1994)
Blueback
Passage-up
R
Same
45
8
‘‘Left the study area and
never returned’’
B24
NR
Yes
Dodson et al. (1972)
Am. Shad
Behavior
A
Same
7
1
‘‘Migrated downstream’’
B10
NR
–
Hightower & Sparks (2003)
Am. Shad
Behavior
R
Same
17
‘‘Most’’
‘‘Movement downstream’’
NR
NR
No
Moser et al. (2000)
Am. Shad
Passage-up
A
Same
86
‘‘Most’’
‘‘Drifted downstream’’
B24
NR
No
Olney et al. (2006)
Am. Shad
Behavior
A
Same
29
13
‘‘Unexpected movement
downstream’’
NR
C7.4
No
Sprankle (2005)
Am. Shad
Behavior
R
Same
72
7
‘‘Fallback’’
72
B1
Yes
Am. shad is American shad, blueback is blueback herring. Passage studies examined either upstream or downstream passage. Tags are represented by a single letter: acoustic or
ultrasonic (A) or radio (R). NR indicates no explicit reporting of value. Dashes indicate a measure that is not applicable to that study
239
123
240
Hydrobiologia (2009) 635:237–249
A. Study Area
Ipswich River B. Ipswich River
Atlantic
Ocean
Massachusetts
To Atlantic
Ocean
Willowdale
Dam
6 (13.4)
2 (5.8)
5 (12.6)
7 (16.3)
3 (6.8)
4 (9.8)
Nemasket River
9 (26.2)
1 (5.1)
Ipswich
Mills
Dam
Great Wenham Swamp
8 (21.0)
flow
Bostick
Finley
Dam
2007 Receiver ID (River kilometer)
Release Site
Dam
Fig. 1 A Map of the Nemasket and Ipswich Rivers in
Massachusetts, USA. B Adult alewives volitionally migrating
upstream in the Ipswich River were obtained, tagged, and
released near the Ipswich Mills Dam [river kilometer (rkm) 5.9]
and tracked through nine stationary receivers (rkm 5.1–26.2).
Black dots indicate receivers. Text indicates receiver number
and rkm in parentheses. The star indicates where fish were
tagged and released at the Ipswich Mills Dam (rkm 5.8). The
largest available spawning area is thought to be Great Wenham
Swamp between receivers 7–8
fish more stressed than untagged fish? (3) What
standardized metrics should be reported in future studies
to allow comparisons across systems and fish? (4) Do
downstream movements necessarily have adverse
consequences?
covering 6.47 km2. Temperatures during the period of
alewife migration, averaged 13.5°C (SE = 1.3, range
6.1–16.1) and mean discharge was 21.9 m3 s -1 (SE =
3.4, range 13.1–38.9).
For tagging, adult alewives were captured in the
Ipswich Mills Dam fishway (rkm 5.9) using a box
trap placed at the upstream fishway exit. The trap
(61 cm high, 61 cm wide, 122 cm long) was checked
at least once per day during the spring when it was
fishing (55 fishing days in 2007, April 2–June 15).
Only fish that appeared healthy and uninjured were
tagged, and only those that recovered quickly from
the tagging process were released.
Alewives (n = 21, mean TL = 267 mm, SE =
3.6) were tagged during April 23–27, 2007. After fish
were obtained from the box trap, they were placed into
a rectangular tank (31 cm wide, 64 cm long, 20 cm
water depth) where they were gently caught by hand
for tagging. We used radio tags (Model NTC-6-1
transmitters, Lotek Wireless, Newmarket, ON) that
were 22.4 mm long, 9.1 mm diameter, weighing 2.8 g
in air, with a calculated operational life of 124 days.
On average, alewives weighed 175.0 g; hence, tags
were less than the recommended maximum of 2% of
Materials and methods
The Ipswich River is a 72.4-km, fifth-order river in
northeastern Massachusetts that empties into the
Atlantic Ocean through Plum Island Sound (Fig. 1).
Three low-head dams (1.4–2.0 m spillway height)
with varying degrees of passage are present in the
mainstem. Ipswich Mills Dam [river kilometer (rkm)
5.9] has a Denil fish ladder and Willowdale Dam
(rkm 13.7) has a notched weir–pool fishway. BostikFinley Dam (rkm 41.2) has no passage and represents
the upper limit of anadromous fish range in the river.
Historically, river herring spawned in the 0.9 km2
Wenham Lake, now a municipal water supply that is
inaccessible to fish. At present, the largest available
alewife spawning habitat is Great Wenham Swamp,
an extensive wetlands upstream of Willowdale Dam
123
Hydrobiologia (2009) 635:237–249
the body weight (Winter, 1996). Tags were individually coded and assigned to one of five frequencies.
Within 30 s, tags were implanted gastrically, without
the use of anesthetics. Using a hollow plastic insertion
tool (12.3 cm long and 8 mm diameter tapering from
8 to 5 mm) the tag was inserted following the
procedure described by Smith et al. (2009). Fish were
then released at the capture location (rkm 5.9).
Nine stationary radio telemetry receivers (Model
SRX_400, Lotek Wireless, Newmarket, ON) were
located in the Ipswich River at rkm 5.1, 5.8, 6.8, 9.8,
12.6, 13.4, 16.3, 21.0, and 26.2 (Fig. 1B). Receiver
gain was changed as needed during the study season.
Receivers scanned all the frequencies in 5.5 s. Ranges
for each receiver were determined prior to and after
the release of tagged fish. The linear range extended up
and downstream from 42 to 298 m. Receiver efficiency, the relationship between the detections of a
tagged fish at a specific receiver divided by number of
times that fish was detected by adjacent receivers
above and below the target, was 88.7–100.0%.
Receivers were downloaded two to four times per
week. Data on fish movements were recorded from
April 23 to June 5, 2007 (43 days).
We also examined plasma cortisol, glucose, and
chloride ions of tagged and untagged alewives from the
Nemasket River (Fig. 1) using the same protocol. For
this physiological assay, fish were dipnetted from the
Wareham Street Dam fishway (rkm 12.1) on April 30,
2007 (n = 20). One to two fish were captured from the
fishway at a time. After netting, fish were placed in a
cylindrical holding tank (113 l; 0.6-m diameter, 0.6-m
height) filled with ambient river water. From this tank,
individual fish were collected one at a time with a
smaller net. Blood was drawn from each fish’s caudal
blood vessels using a heparinized syringe. After blood
collection, fish were euthanized. The entire blood
collection process was completed within 5 min of the
time each individual fish was captured in the fishway.
Blood samples were kept on ice until all the fish were
sampled, then samples were centrifuged at 2,000g for
5 min. Plasma was decanted and frozen on dry ice and
stored at -80°C until it could be analyzed in the
laboratory.
In order to measure stress in response to tagging, we
inserted dummy tags (22.4 mm long and 9.1 mm in
diameter, weight in air = 2.8 g) into 10 alewives
using the methods described above. An additional 10
alewives were removed from the river and handled but
241
not tagged. Two pairs of tagged and untagged fish
were placed in each of five cylindrical mesh net pens
(61 cm in diameter, 61 cm deep, 64 cm mesh)
anchored in a still section of the Nemasket River at a
depth of 1.5 m. After fish had been in the pens for
24 h, we assessed survival and took blood samples. At
24 h, all fish were removed from a single pen using a
dipnet and placed in a cylindrical holding tank (113 l;
0.6-m diameter and 0.6-m height) filled with ambient
river water. Blood was drawn from each fish as
described above within 5 min of the initial disturbance
of each pen. Each pen was processed sequentially.
Plasma cortisol, glucose and chloride ions were
analyzed at the USGS Conte Anadromous Fish
Research Center (Turners Falls, MA, USA). Plasma
cortisol was measured by direct enzyme immunoassay
(Carey & McCormick, 1998) which has been validated
for use in alosines (Shrimpton et al., 2001). Glucose
was measured by the hexokinase and glucose6-phosphate dehydrogenase enzymatic method using
external standards (Stein, 1963). Plasma chloride was
analyzed by the silver titration method using a
Buchler-Cotlove digital chloridometer and external
standards. One tagged fish did not yield enough blood
to analyze the sample for chloride ions. We used
multiple, nonparametric Mann-Whitney procedures
(PROC NPAR1WAY, SAS 9.1) to test for differences
in plasma cortisol, glucose and chlorides (1) between
tagged and untagged alewives held 24 h, and (2)
between all the unheld fish sampled initially and all the
handled fish (tagged and untagged).
Results
Movements
We describe below movement trajectories of tagged
alewives to illustrate the diversity of possible movement patterns. In general, we cite each trajectory for a
single type of movement but most natural trajectories
are complex combinations of multiple movements.
We show all the 21 fish tagged in 2007. Our goal,
however, was not to provide a quantitative analysis of
fish movements, but to use individual fish trajectories
to demonstrate the array of movements that may be
encountered in the field.
Anadromous alewives differed in timing and
direction of the initial movements (Fig. 2A–H), the
123
242
Hydrobiologia (2009) 635:237–249
Fig. 2 Individual locations (river kilometer) and detection
times (days after release) recorded for anadromous alewives
caught and tagged during their spring upriver spawning
migration in the Ipswich River, 2007. Individual fish are shown
below to indicate real movement trajectories. A Short pause then
downstream movement (Fish 19); B Pause lasting [24 h
followed by downstream movement (Fish 20); C Short distance
(\1 km) down, long duration ([24 h) (Fish 6); D, E No pause,
up then down (Fish 1, 15); F, G No pause, up only (fish does not
return downstream) (Fish 4, 9); H Short pause (\24 h), then up
(Fish 7); I, J Short distance up for a short and moderate duration
(Fish 21, 17); K Moderate distance up for a short duration (Fish
18); L Moderate distance up, moderate duration (Fish 11); M
Long distance up, moderate duration (Fish 12); N Fast
movement over a long distance (Fish 3, 14); O Slow movement,
moderate distance, with initial forays (Fish 16). P Intermediate
forays (Fish 5); Q, R Delayed forays, following a long distance
migration (Fish 2, 8); S, T Complex forays (Fish 10, 13). Dotted
lines indicate that locations between receivers are unknown. All
upstream migrating fish tagged in 2007 are shown
duration, distance, and speed of initial movements
(Fig. 2I–O), and patterns of forays or movement
reversals (Fig. 2P–T). Most of the movement trajectories of individual fish included both upstream and
downstream movements. The timing and direction of
the first movement following release varied, with fish
moving from the tagging site both downstream
(Fig. 2A–C) and upstream (Fig. 2D–H). We observed
downstream movement after a short pause (16.8 h
after release, Fig. 2A) and also after a longer pause
(55.2 h after release, Fig. 2B). Some fish that initially
moved downstream stayed within our array briefly
and then exited the receiver array for the duration of
the study (Fig. 2A–B). Others that moved downstream stayed within the lower part of the array for
a prolonged time (Fig. 2C). We also observed
upstream directed movements immediately postrelease (Fig. 2D–G), as well as following a short
pause (\11.0 h, Fig. 2H–J). In most cases (Fig. 2D,
E, H), but not all (Fig. 2F–G), initial upstream
movements were followed by a return downstream.
We also examined the distance (km traveled in a
single direction, i.e., Y-axis trajectory), duration (how
long the fish was heard by a single receiver, i.e.,
123
Hydrobiologia (2009) 635:237–249
X-axis trajectory), and speed (i.e., slope) of subsequent movements. Because we had data only for time
at a receiver, we hypothesized interim movements by
connecting these detections with a dotted line. We
defined a movement event as the trajectory resulting
from coupled bouts of adjacent upstream and downstream movements. Fish initially traveling upstream
post-release moved short (Fig. 2I–J), moderate (Fig. 2
K–L), and long distances upstream (Fig. 2M–N). For
fish traveling a short distance upstream (\1 km from
release site receiver; Fig. 2I–J), the duration of time
spent upstream was both short (0.1 h) and moderately
long (9.0 h, Fig. 2I–J). Similarly, fish traveling a
moderate distance upstream (Fig. 2K–L), stayed short
(Fig. 2K) and moderately long (Fig. 2L) durations at
the upstream site. Other fish travelled long distances
upstream (15.1 km from release site) and remained
for a moderate period of time before returning
downstream (Fig. 2M). Fish traveled upstream both
at faster (1.02 km/h, Fig. 2N), and slower speeds
(Fig. 2O).
Fish often reversed the direction of their movements
(Fig. 2O–T). These forays occurred at different periods of the migration. Repetitive short distance
upstream and downstream forays preceded longerdistance migrations upstream (Fig. 2O), occurred in
the middle (Fig. 2P), or occurred at the end of the
migration (Fig. 2Q–R). Other fish made multiple longdistance directional bouts of movement during their
migration combining many directions and distances
(Fig. 2S–T).
The above descriptions characterize select components of individual movement trajectories. Each
complete trajectory was composed of three parts: (a)
timing and direction of first initial movements; (b)
distance, duration, and speed of subsequent bouts of
coupled upstream and downstream movements, and
(c) multidirectional forays (Table 2). Some similarities and some differences occurred when these
components were summarized across trajectories.
Out of 21 trajectories recorded in 2007, three had
only a downstream component (Fig. 2A–C), three
had only an upstream component (Fig. 2F, G, N), but
the majority (n = 15; 71.4%) had both an upstream
and downstream component associated with each
bout of coupled upstream–downstream movements.
Of these, three (Fig. 2D, E, L) had a single upstream–
downstream trajectory whereas all the others had
multiple bouts of upstream and downstream
243
Fig. 3 Plasma A cortisol (ng/ml), B glucose (mM), and C
chloride (mM) ion responses of unheld, untagged (U), and
tagged (T) fish. Unheld, pre-tag levels were obtained before
any activity occurred. The tagged and untagged fish were
sampled at 24 h. NS indicates no significant difference
between tagged and untagged fish. Data are mean ± 1 SE.
Statistics are shown in Table 3
movements (Fig. 2H–K, M, O–T). In some of these
multiple-maxima trajectories, distance traveled in
both bouts of upstream and downstream movements
were approximately equal in distance but differed in
duration (Fig. 2P, S). In others, the distance and
duration in one bout of movements were greater than
the other (Fig. 2H, K, T). Still others included
multiple, small, short forays (Fig. 2I–K, O–R). While
we cannot neatly group fish that exhibited identical
behaviors to characterize ‘‘normal’’ in the entire
tagged group, use of these metrics would enable us
start to systematically describe complicated fish
movements (Table 2).
123
244
123
Table 2 Trajectories of each anadromous alewife tagged in 2007 broken down by three components: (a) timing and direction of initial movement indicating whether there was no,
a short, or a longer pause in both downstream (down) and upstream (up) directions; (b) distance, duration, and speed of subsequent post-release movements in both the up and
downstream directions including distance moved (short, moderate, or long), duration of time spent at extreme locations (short, moderate, long), and speed of movement (fast or
slow); (c) forays or repeated short duration movements (initial or delayed). Components a and b are analyzed for all major peaks. The related figure panel is also indicated
Fish a. Timing and direction of first post-release b. Distance, duration, and speed of subsequent postrelease movements
movement
Down
Up
Up
Duration
Pause
Pause
Distance
Speed
Down
X
20
6
X
Figure
Initial Delayed
Short Moderate Longer Distance
None Short Longer None Short Longer Short Moderate Long Fast Slow
19
c. Forays
Speed
Short Moderate Long Fast Slow
X
X
2A
X
X
2B
X
X
1
15
X
X
X
X
X
X
X
4
X
X
X
X
9
X
X
X
X
X
X
X
X
X
2C
X
X
X
X
X
2D
2E
2F
2G
7
X
21
X
X
X
X
X
X
X
17
X
X
X
X
X
X
X
X
2J
X
X
X
X
X
2K
X
X
X
X
X
X
18
X
11
X
12
X
X
X
X
X
X
X
X
X
X
X
14
X
X
X
X
16
X
X
5
X
X
X
2
X
X
X
8
X
X
X
10
13
X
X
X
X
X
X
X
X
X
X
X
X
X
X
2H
2I
2L
X
2M
2N
2N
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
2O
X
2P
X
X
2Q
X
X
2R
X
2S
2T
Hydrobiologia (2009) 635:237–249
3
X
Hydrobiologia (2009) 635:237–249
245
Table 3 Mann–Whitney test results for the effect of tagging
on plasma cortisol, glucose, and chloride ion concentrations
(N = 20)
Chemical
n
df
F value
P
Cortisol
20
18
3.25
0.09
Glucose
20
18
0.94
0.35
Chloride
19
17
0.01
0.92
Results indicate no difference in plasma concentrations
between tagged and untagged fish held 24 h in the Nemasket
River
Physiology
All tagged and untagged fish held in net pens were alive
at 24 h. Plasma cortisol (Fig. 3A), glucose (Fig. 3B),
and chloride ions (Fig. 3C; Table 3) did not differ
between tagged and untagged fish. However, handling
and confinement, whether associated with tagging or
not, altered plasma cortisol, glucose, and chloride for
both untagged and tagged fish compared to unheld, pretagging values (Fig. 3; P \ 0.001 for cortisol, glucose,
and chloride ions).
Discussion
We observed a diverse repertoire of downstream
movements. Some fish moved downstream immediately; others moved downstream after a considerable
period of upstream activity. Some fish moved at fast
speeds; others moved more slowly. Some fish moved
downstream and stayed there for a considerable period
of time; others moved both downstream and upstream
after both short and long pauses. Movements typically
occurred in combination with upstream movements,
so bouts of adjacent upstream and downstream
movements should be examined together. Several of
the movement patterns described here could fall under
the traditional definition of ‘‘fallback’’ but were not
necessarily an adverse reaction to tagging, and may, in
fact, not have undesirable consequences. These
patterns may represent the normal diversity of movements in pre-spawning river herring.
All trajectories can be quantified by (a) direction and
time of initial movements, (b) distance traveled, speed,
and time spent at a specific location for each coupled
bout of adjacent upstream and downstream movements, and (c) number, type, and sequence of
movement events. Because numerous possible explanations exist for this wide range of movements, the
following issues should be addressed in future telemetry studies. First, researchers should note the context
of the fish prior to capture. For all of the alosine
telemetry studies under review in our study, anadromous fish were actively moving upstream prior to
capture and tagging. When this is not the case, different
interpretations of upstream and downstream movements may exist. Second, the location of the release site
should be specified relative to the capture site.
Researchers should provide a distance from the river
mouth for both capture and release sites and should
consider the role of upstream or downstream displacement (Makinen et al., 2000). Past telemetry studies on
anadromous alosines have focused on fish passage, so
that fish capture and release sites were typically the
same. As the behavior of spawning anadromous fish is
evaluated for river restoration, this may not always be
true. For example, to evaluate stocking as a way to
supplement depleted populations, tagged fish may be
released directly into upstream spawning areas, a
strategy that could have radical implications for the
interpretation of telemetry data. Third, we suggest
that researchers report where spawning habitat is
located relative to the release site. If fish are released
directly into an appropriate spawning habitat, fish may
not need to move until they are ready to emigrate
following spawning. In this case, interpretations of
movements would be quite different than if a fish is
required to swim a distance upstream to access
spawning habitat.
Fourth, it will also be valuable to report all the
metrics usually associated with traditional definitions
of ‘‘fallback,’’ including time to and direction of
initial movement following tagging. The timing and
direction of initial movements can aid in interpreting
behaviors. For example, chinook salmon (Oncorhynchus tshawytscha) were classified as ‘‘motivated’’ or
‘‘hesitant’’ based on the initial direction of movement
following release (Bernard et al., 1999). Immediate
upstream movement may indicate that the urge to
spawn overrides other considerations. Immediate
downstream movement may indicate altered migratory behavior (Olney et al., 2006). Fifth, distance,
duration, and speed of movements following release
should be reported. These metrics are often recorded
in telemetry studies, but ‘‘normal’’ distances and times
have not been identified. As examples, American shad
123
246
with limited upstream movement within 72 h were
classified as ‘‘non-viable’’ (Sprankle, 2005), and sea
lamprey (Petromyzon marinus) with brief upstream
forays (\1 km) punctuated by long stationary periods
(several weeks) were termed ‘‘atypical’’ (Andrade
et al., 2007). In order to best interpret whether these
movements were aberrant, more information is
needed on patterns and mechanisms associated with
pre-spawning fish behavior.
Sixth, if fish move downstream and then later
return upstream, the time required to return to the
tagging location should be documented. Often, as in
our study, emphasis is placed on the upstream
migration, and receivers are distributed upstream of
the release location. However, this can limit a
researcher’s ability to assess and document downstream behaviors, whether normal or abnormal. If
field interpretation of the tag effect depends on
downstream movement, future telemetry studies
should allocate receivers specifically to quantify
downstream behavior. Seventh, the occurrence of
short distance forays (\2 km) also should be reported,
as this indicates active swimming behavior. The
timing of these movements may indicate exploration
(Keefer et al., 2008), the drive to spawn (Acolas et al.,
2004), or a reaction to the environment (Dodson et al.,
1972). If possible, these post-tagging movements
should be linked to known information about success
of spawning.
Finally, authors should clearly define and justify
their reasons for excluding ‘‘fallback’’ fish from
analyses. Varying methods have been used to determine whether ‘‘fallback’’ fish will be included in data
analyses including eventual return upstream (Beasley
& Hightower, 2000; Moser et al., 2000; Hightower &
Sparks, 2003; Bailey et al., 2004; Olney et al., 2006)
or movement within a specified time period (Chappelear & Cooke, 1994). The majority of alosine telemetry
literature include fish in the analyses that eventually
resume upstream migration after initial downstream
movements. However, the criterion of limited
upstream movement following tagging has also been
used to exclude fish from analyses (Sprankle, 2005)
and to identify altered migratory behavior (Olney
et al., 2006). Researchers should report whether the
entirety of the telemetry record is used, or if data are
only collected once a fish resumes migration or moves
a specified distance upstream (Bernard et al., 1999;
Beasley & Hightower, 2000; Keefer et al., 2004). If
123
Hydrobiologia (2009) 635:237–249
researchers provide information on all of these
parameters in future telemetry field studies, a body of
literature will emerge on which to base tagging
protocols, and from which much can be learned about
spawning behavior in the field.
The high variability in downstream movement
metrics has resulted in inconsistent interpretations of
these movements. In the alosine telemetry literature,
there is no agreement regarding the role of sex, age, or
timing on post-tagging downstream movements.
Males may be more affected than females because of
their smaller size (Moser et al., 2000) or females may
be more sensitive to the handling process due to their
higher condition factor (Acolas et al., 2004). Young or
virgin spawners of either sex may be more affected
than older or repeat spawners (Hightower & Sparks,
2003) or there may be no link between ‘‘fallback’’
behavior, sex (Bailey et al., 2004), and age (Olney
et al., 2006). Researchers have suggested that later
migrants may respond rather differently to tagging and
handling than early migrants (Glebe & Leggett, 1981;
Bailey et al., 2004; Sprankle, 2005), but no consensus
exists.
Although ‘‘fallback’’ in the alosine literature is
defined as unnatural downstream movement related to
tag effect and handling, salmonid telemetry studies
rarely link ‘‘fallback’’ to tag effects or handling
(Bernard et al., 1999; Makinen et al. 2000; Holbrook
et al., 2009). Often the downstream movements of
upstream migrating salmon are described as a purposeful behavior in response to the environment,
obstacles, or a mechanism of homing (Keefer et al.,
2006). These complex behaviors include overshooting
of natal systems (Naughton et al., 2006), exploratory
movements (Keefer et al., 2008), seeking alternate
routes, waiting for appropriate conditions (Thorstad
et al., 2005; Holbrook et al., 2009), disorientation in
certain hydraulic conditions (Naughton et al., 2006),
being swept over dams (Matter & Sandford, 2003), or
varying sensitivity in distinct migratory phases
(Makinen et al., 2000; Jokikokko, 2002). Aberrant
movement in salmonids has not been explicitly related
to ‘‘fallback’’ or tag effect (Young et al., 2006); for
example, when radio tags were used to examine the
effect of catch-and-release on adult Atlantic salmon
(Salmo salar), uncharacteristic up and downstream
movements of radiotagged fish observed post-release
were attributed to angling (Thorstad et al., 2003). This
marked difference in how ‘‘fallback’’ behavior is
Hydrobiologia (2009) 635:237–249
interpreted across fish taxa may be because little is
understood about the migrations of non-salmonid
anadromous fishes. As the body of telemetry literature
on other anadromous species grows, we anticipate the
emergence of alternative hypotheses to explain the
range of upstream and downstream movements in
alosines.
Downstream movements post-tagging should be
viewed on a continuum of potential consequences.
‘‘Fallback’’ may result in increased likelihood of
injury or death during downstream movement,
potential re-exposure to a fishery, reduced likelihood
of reaching spawning grounds, migratory delay, and
energy expenditure to re-gain lost ground (Bernard
et al., 1999; Boggs et al., 2004; Scruton et al., 2007).
From a management perspective, ‘‘fallback’’ may
also result in inflated fishway counts (Naughton et al.,
2006) or incorrect estimates of exploitation and
fishing mortality rates (Olney et al., 2006). Migration
abandonment is a severe consequence of ‘‘fallback,’’
in which fish never resume upstream migration
following ‘‘fallback’’ (Hightower & Sparks, 2003;
Olney et al., 2006). However, as we have suggested,
downstream movements following tagging need not
be abnormal or have adverse consequences. Neither
‘‘fallback’’ nor abandonment precludes the possibility
of spawning (Beasley & Hightower, 2000) if fish can
use secondary spawning habitats (Acolas et al., 2004;
Jepsen et al., 2005; Lopez et al., 2007; Maes et al.,
2008). Furthermore, up and downstream movements
may be part of normal pre-spawning migration,
exploration, and habitat selection.
The trajectories of tagged alewives we used here to
illustrate the range of possible movements would not
be instructive if these fish movements were caused by
tagging. We took exceptional care tagging our fish and
used a detailed protocol that involved a limited
number of designated taggers and several training
sessions before the actual tagging. Increased plasma
cortisol is part of a fish’s primary response to stress,
and the magnitude of corticosteroid response typically
indicates the severity of the stressor (Barton & Iwama,
1991). Secondary responses to stress include changes
in plasma glucose and the major ions, sodium and
chloride (Close et al., 2003). The tagged fish held
24 h in our experiment did not exhibit significant
differences in plasma cortisol, glucose, or chlorides
relative to untagged fish held for this period. Handling
and holding fish, whether associated with tagging or
247
not, however, resulted in higher plasma cortisol and
glucose and lower plasma chloride compared to initial
levels of unheld fish. However, cortisol, glucose, and
chloride levels of our anadromous alewives were not
extreme. In a related study that transported anadromous alewives for 2 h, cortisol and glucose were
much higher and chlorides much lower than the fish
that were handled and held but not transported in this
study [transported means: cortisol = 1047 ng/ml;
glucose = 14.5 mM; chlorides = 55.1 mM (Frank,
2009); untransported means, (tagged and untagged
combined): cortisol = 512.4 ng/ml; glucose = 9.9
mM; chlorides = 69.1 mM (this study)]. Thus, this
lack of a difference between tagged and untagged
alewives was not because both groups of fish were
maximally stressed. Quantifying how tagging affects
fish in the wild is a challenge because it is very
difficult to measure fish behavior, physiology, or
stress in the field without either tagging them or
holding them in confinement. Stress related to
handling occurs in virtually all animals in the wild,
making this problem an inherent difficulty in studies
of the behavior and physiology of wild animals.
Certainly a tag may affect other aspects of fish
performance (e.g., swim speed, searching behavior,
depth in water column) besides what we measured.
With advances in technology such as tags that can
assess physiological condition, future researchers may
be able to more precisely separate out tagging,
handling, and confinement stress. With the increasing
number of research studies on telemetry, understanding these physiological and behavioral tag effects in
the field is both a critically important and extremely
challenging area in which future research is needed.
Conclusion
In summary, we encourage other researchers to report
the following data relative to post-tagging movements: the number of fish that move downstream;
context of capture; time to initial movements; direction of and time to initial movements; characteristics
of movements from the release site including distance,
speed and duration at each location; changes in
direction and associated distance, duration, and speed;
and whether or not all fish are included in the analysis.
Information on sex, age, and migration timing related
to the incidence of ‘‘fallback’’ and other movements
123
248
will also be useful to better understand which fish are
more likely to exhibit this behavior. Physiological
assessments combined with behavioral studies will
provide better information on how stressors, both
human and natural, affect migratory behavior. With
this information, we can start to sort out the complicated behaviors seen in most fish telemetry studies
including river herring and other alosines.
Acknowledgments This project was administered through the
Massachusetts Cooperative Fish and Wildlife Research Unit. The
Massachusetts Cooperative Fish and Wildlife Research Unit is a
collaboration of the University of Massachusetts, the U. S. Geological Survey, the Massachusetts Division of Marine Fisheries,
the Massachusetts Division of Fisheries and Wildlife and the
Wildlife Management Institute. We thank the Massachusetts
Division of Marine Fisheries, Ebsco Publishing, New England
BioLabs, Ipswich Bay Fly Fishing Derby, numerous alewife
adopters, Northeast Utilities, and U.S. Fish and Wildlife Service
for their respective supports. A. Silberzweig, S. Turner, and
M. Burak provided field assistance. We thank Amy Regish for
analyzing plasma samples. Comments from Joe Hightower, Ken
Sprankle, and two anonymous reviewers improved the
manuscript.
References
Acolas, M. L., M. L. B. Anras, V. Veron, H. Jourdan, M. R.
Sabatie & J. L. Bagliniere, 2004. An assessment of the
upstream migration and reproductive behaviour of Allis
shad (Alosa alosa L.) using acoustic tracking. ICES
Journal of Marine Science 61: 1291–1304.
Andrade, N. O., B. R. Quintella, J. Ferreira, S. Pinela, I. Povoa,
S. Pedro & P. R. Almeida, 2007. Sea lamprey (Petromyzon marinus L.) spawning migration in the Vouga river
basin (Portugal): poaching impact, preferential resting
sites and spawning grounds. Hydrobiologia 582: 121–132.
Bailey, M. M., J. J. Isely & W. C. Bridges, 2004. Movement
and population size of American shad near a low-head
lock and dam. Transactions of the American Fisheries
Society 133: 300–308.
Barry, T. & B. Kynard, 1986. Attraction of adult American shad
to fish lifts at Holyoke Dam, Connecticut River. North
American Journal of Fisheries Management 6: 233–241.
Barton, B. A. & G. K. Iwama, 1991. Physiological changes in
fish from stress in aquaculture with emphasis on the
response and effects of corticosteroids. Annual Review of
Fish Diseases 1: 3–26.
Beasley, C. A. & J. E. Hightower, 2000. Effects of a low-head
dam on the distribution and characteristics of spawning
habitat used by striped bass and American shad. Transactions of the American Fisheries Society 129: 1316–1330.
Bell, C. E. & B. Kynard, 1985. Mortality of adult American
shad passing through a 17-Megawatt Kaplan turbine at a
low-head hydroelectric dam. North American Journal of
Fisheries Management 5: 33–38.
123
Hydrobiologia (2009) 635:237–249
Bernard, D. R., J. J. Hasbrouck & S. J. Fleischman, 1999. Handling-induced delay and downstream movement of adult
chinook salmon in rivers. Fisheries Research 44: 37–46.
Boggs, C. T., M. L. Keefer, C. A. Peery, T. C. Bjornn & L. C.
Stuehrenberg, 2004. Fallback, reascension, and adjusted
fishway escapement estimates for adult Chinook salmon
and steelhead at Columbia and Snake River dams. Transactions of the American Fisheries Society 133: 932–949.
Bridger, C. J. & R. K. Booth, 2003. The effects of biotelemetry
transmitter presence and attachment procedures on fish
physiology and behavior. Reviews in Fisheries Science
11: 13–34.
Carey, J. B. & S. D. McCormick, 1998. Atlantic salmon smolts
are more responsive to an acute handling and confinement
stress than parr. Aquaculture 168: 237–253.
Chappelear, S. J. & D. W. Cooke. 1994. Blueback herring
behavior in the tailrace of the St. Stephan Dam and fish
lock. In Cooper, J. E., R. T. Eades, R. J. Klauda, & J. G.
Loesch (eds), Anadromous Alosa Symposium. Tidewater
Chapter. American Fisheries Society, Bethesda, MD:
108–112
Close, D. A., M. S. Fitzpatrick, C. M. Lorion, H. W. Li & C. B.
Schreck, 2003. Effects of intraperitoneally implanted
radio transmitters on the swimming performance and
physiology of Pacific lamprey. North American Journal of
Fisheries Management 23: 1184–1192.
Cooke, S. J., S. G. Hinch, A. P. Farrell, D. A. Patterson, K.
Miller-Saunders, D. W. Welch, M. R. Donaldson, K. C.
Hanson, G. T. Crossin, M. T. Mathes, A. G. Lotto, K. A.
Hruska, I. C. Olsson, G. N. Wagner, R. Thomson, R.
Hourston, K. K. English, S. Larsson, J. M. Shrimpton &
G. Van der Kraak, 2008. Developing a mechanistic
understanding of fish migrations by linking telemetry with
physiology, behavior, genomics and experimental biology: an interdisciplinary case study on adult Fraser River
sockeye salmon. Fisheries 33: 321–338.
Dodson, J. J., R. A. Jones & W. C. Leggett, 1972. Behavior of
adult American shad (Alosa sapidissima) during migration
from salt to fresh water as observed by ultrasonic tracking
techniques. Journal of the Fisheries Research Board of
Canada 29: 1445–1449.
Frank, H., 2009. Evaluation of pre-spawning movements of
anadromous alewives in the Ipswich River using radiotelemetry. M.S. Thesis, University of Massachusetts,
Amherst, MA.
Glebe, B. D. & W. C. Leggett, 1981. Temporal, intrapopulation
differences in energy allocation and use by American shad
(Alosa sapidissima) during the spawning migration.
Canadian Journal of Fisheries and Aquatic Sciences 38:
795–805.
Hightower, J. E. & K. L. Sparks, 2003. Migration and
spawning habitat of American shad in the Roanoke River,
N.C. In Limburg, K. E. & J. R. Waldman (eds), Biodiversity, Status, and Conservation of the World’s Shads.
American Fisheries Society, Bethesda, MD: 193–199.
Holbrook, C. M., J. Zydlewski, D. Gorsky, S. L. Shepard & M.
T. Kinnison, 2009. Movements of prespawn adult Atlantic
salmon near hydroelectric dams in the lower Penobscot
River, Maine. North American Journal of Fisheries
Management 29: 495–505.
Hydrobiologia (2009) 635:237–249
Jepsen, N., E. E. Nielsen & M. Deacon, 2005. Linking individual migratory behavior of Atlantic salmon to their
genetic origin. In Spedicato, M. T., G. Lembo &
G. Marmulla (eds), Aquatic Telemetry: Advances and
Applications. FAO/COISPA, Ustica, Italy: 45–51.
Jokikokko, E., 2002. Migration of wild and reared Atlantic
salmon (Salmo salar L.) in the river Simojoki, northern
Finland. Fisheries Research 58: 15–23.
Keefer, M. L., C. A. Peery, M. A. Jepson & L. C. Stuehrenberg,
2004. Upstream migration rates of radio-tagged adult
Chinook salmon in riverine habitats of the Columbia
River basin. Journal of Fish Biology 65: 1126–1141.
Keefer, M. L., C. A. Peery & C. C. Caudill, 2006. Long-distance downstream movements by homing adult Chinook
salmon. Journal of Fish Biology 68: 944–950.
Keefer, M. L., C. C. Caudill, C. A. Peery & C. T. Boggs, 2008.
Non-direct homing behaviours by adult Chinook salmon
in a large, multi-stock river system. Journal of Fish
Biology 72: 27–44.
Lassalle, G., M. Beguer, L. Beaulaton & E. Rochard, 2008.
Diadromous fish conservation plans need to consider
global warming issues: an approach using biogeographical
models. Biological Conservation 141: 1105–1118.
Lopez, M. A., N. Gazquez, J. M. Olmo-Vidal, M. W. Aprahamian & E. Gisbert, 2007. The presence of anadromous
twaite shad (Alosa fallax) in the Ebro River (western
Mediterranean, Spain): an indicator of the population’s
recovery? Journal of Applied Ichthyology 23: 163–166.
Lucas, M. C. & E. Baras, 2000. Methods for studying spatial
behaviour of freshwater fishes in the natural environment.
Fish and Fisheries 1: 283–316.
Maes, J., M. Stevens & J. Breine, 2008. Poor water quality
constrains the distribution and movements of twaite shad
Alosa fallax fallax (Lacepede, 1803) in the watershed of
river Scheldt. Hydrobiologia 602: 129–143.
Makinen, T. S., E. Niemela, K. Moen & R. Lindstrom, 2000.
Behavior of gill-net and rod-captured Atlantic salmon
(Salmo salar L.) during upstream migration and following
radio tagging. Fisheries Research 45: 117–127.
Matter, A. L. & B. P. Sandford, 2003. A comparison of
migration rates of radio- and PIT-tagged adult Snake
River Chinook salmon through the Columbia River
hydropower system. North American Journal of Fisheries
Management 23: 967–973.
McDowall, R. M., 1999. Different kinds of diadromy: different
kinds of conservation problems. ICES Journal of Marine
Science 56: 410–413.
Moser, M. L., A. M. Darazsdi & J. R. Hall, 2000. Improving
passage efficiency of adult American shad at low-elevation dams with navigation locks. North American Journal
of Fisheries Management 20: 376–385.
Naughton, G. P., C. C. Caudill, M. L. Keefer, T. C. Bjornn, C.
A. Peery & L. C. Stuehrenberg, 2006. Fallback by adult
sockeye salmon at Columbia River dams. North American
Journal of Fisheries Management 26: 380–390.
249
Olney, J. E., R. J. Latour, B. E. Watkins & D. G. Clarke, 2006.
Migratory behavior of American shad in the York River,
Virginia, with implications for estimating in-river
exploitation from tag recovery data. Transactions of the
American Fisheries Society 135: 889–896.
Rogers, K. B. & G. C. White, 2007. Analysis of movement and
habitat use from telemetry data. In Guy, C. S. & M. L.
Brown (eds), Analysis and Interpretation of Freshwater
Fisheries Data. American Fisheries Society, Bethesda,
MD: 625–676.
Scruton, D. A., R. K. Booth, C. J. Pennell, F. Cubitt, R. S.
McKinley & K. D. Clarke, 2007. Conventional and EMG
telemetry studies of upstream migration and tailrace
attraction of adult Atlantic salmon at a hydroelectric
installation on the Exploits River, Newfoundland, Canada.
Hydrobiologia 582: 67–79.
Shrimpton, J. M., J. D. Zydlewski & S. D. McCormick, 2001.
The stress response of juvenile American shad to handling
and confinement is greater during migration in freshwater
than in seawater. Transactions of the American Fisheries
Society 130: 1203–1210.
Smith, J. M., M. E. Mather, H. J. Frank, R. M. Muth, J. T. Finn
& S. D. McCormick, 2009. Evaluation of a gastric radio
tag insertion technique for anadromous alewives. North
American Journal of Fisheries Management 29: 367–377.
Sprankle, K., 2005. Interdam movements and passage attraction of American shad in the lower Merrimack River main
stem. North American Journal of Fisheries Management
25: 1456–1466.
Stein, M. W., 1963. D-glucose determination with hexokinase
and glucose-6-phosphatedehydrogenase. In Bergmeyer,
H. U. (ed.), Methods in Enzymatic Analysis. Academic
Press, New York: 117–122.
Thorstad, E. B., T. F. Næsje, P. Fiske & B. Finstad, 2003.
Effects of catch and release on Atlantic salmon in the River
Alta, northern Norway. Fisheries Research 60: 293–307.
Thorstad, E. B., P. Fiske, K. Aarestrup, N. A. Hvidsten, K.
Harsaker, T. G. Heggberget & F. Okland, 2005. Upstream
migration of Atlantic salmon in three regulated rivers. In
Spedicato, M. T., G. Lembo & G. Marmulla (eds), Aquatic
Telemetry: Advances and Applications. FAO/COISPA,
Ustica, Italy: 111–121.
Winter, J. D., 1996. Advances in underwater biotelemetry. In
Murphy, B. R. & D. W. Willis (eds), Fisheries Techniques, 2nd ed. American Fisheries Society, Bethesda,
MD: 555–590.
Young, J. L., S. J. Cooke, S. G. Hinch, G. T. Crossin, D. A.
Patterson, A. P. Farrell, G. Van Der Kraak, A. G. Lotto, A.
Lister, M. C. Healey & K. K. English, 2006. Physiological
and energetic correlates of en route mortality for abnormally early migrating adult sockeye salmon in the
Thompson River, British Columbia. Canadian Journal of
Fisheries and Aquatic Sciences 63: 1067–1077.
123