Discriminating natal origin of spawning adult sea lamprey (Petromyzon marinus)... Lake Champlain using statolith elemental signatures

Discriminating natal origin of spawning adult sea lamprey (Petromyzon marinus)... Lake Champlain using statolith elemental signatures

JGLR-00557; No. of pages: 8; 4C:
Journal of Great Lakes Research xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Journal of Great Lakes Research
journal homepage: www.elsevier.com/locate/jglr
Discriminating natal origin of spawning adult sea lamprey (Petromyzon marinus) in
Lake Champlain using statolith elemental signatures
Eric A. Howe a, 1, 2, Aude Lochet a,⁎, Carrol P. Hand b, 3, Stuart A. Ludsin c, 4, J. Ellen Marsden a, 2, Brian J. Fryer b, 3
a
b
c
Rubenstein School of Environment and Natural Resources, University of Vermont, 81 Carrigan Drive, Burlington, VT 05405, USA
Great Lakes Institute for Environmental Research, University of Windsor, 401 Sunset Avenue, Windsor, Canada ON N9B 3P4
Aquatic Ecology Laboratory, Department of Evolution, Ecology, and Organismal Biology, The Ohio State University, 1314 Kinnear Road, Columbus, OH 43212 USA
a r t i c l e
i n f o
Article history:
Received 3 August 2012
Accepted 6 January 2013
Available online xxxx
Communicated by Thomas Pratt
Keywords:
Sea lamprey
Petromyzon marinus
Statolith microchemistry
Natal origin
Lake Champlain
a b s t r a c t
Sea lamprey (Petromyzon marinus) is a nuisance species in the Great Lakes and Lake Champlain. Information
about tributary contributions to the spawning adult phase is critical for appropriate allocation of efforts to
control this species. We examined the accuracy of statolith elemental composition to identify the natal origin
(i.e., individual rivers or clusters of rivers) of 33 known-origin adults from the Lake Champlain basin. To do
so, we first established natal origin chemical signatures using the statoliths of 238 larvae from the same
basin. Using laser-ablation inductively coupled plasma mass spectrometry, the 238 larvae originating from
12 streams and one delta were discriminated with a classification accuracy of 57% (range: 25–80%) and
70% (range: 29–80%) when individual streams and groups of streams were considered respectively,
highlighting the potential of statolith microchemistry to identify natal origins. However, the assignment of
natal origin for adults was overwhelmingly incorrect. Using a maximum likelihood procedure, 88% of the
adults were assigned to a cluster of three streams and one delta, while only 3% of these individuals were
known to originate from this particular cluster. More research is required to understand the low classification
accuracy of sea lamprey adults and validate the use of statolith microchemistry to identify sea lamprey
natal origin.
© 2013 International Association for Great Lakes Research. Published by Elsevier B.V. All rights reserved.
Introduction
The sea lamprey (Petromyzon marinus) is a nuisance species in the
Laurentian Great Lakes and Lake Champlain and has caused major
damage to economically and ecologically important fishes, especially
lake trout (Salvelinus namaycush) and other salmonines (Smith and
Tibbles, 1980). Sea lamprey larvae are filter feeders that inhabit
burrows in the soft sediments of their natal streams for four to six
years (Moore and Mallat, 1980; Potter, 1980). At the end of the larval
phase, individuals transform into parasitic-phase sea lampreys that
migrate into the open lake to feed on large-bodied fishes (Farmer,
1980; Swink, 2003). After 12 to 18 months as parasites, adults
migrate to streams for spawning. Sea lamprey population control efforts in the Great Lakes and Lake Champlain have relied heavily on
⁎ Corresponding author at: Rubenstein School of Environment and Natural Resources,
University of Vermont, 81 Carrigan Drive, Burlington, VT 05405, USA. Tel.: +1 802 656
4280.
E-mail addresses: [email protected] (E.A. Howe), [email protected] (A. Lochet),
[email protected] (C.P. Hand), [email protected] (S.A. Ludsin),
[email protected] (J.E. Marsden), [email protected] (B.J. Fryer).
1
Present address: Lake Champlain Basin Program, 54 West Shore Rd., Grand Isle, VT,
05458, USA.
2
Tel.: +1 802 656 4280.
3
Tel.: +1 519 253 3000x3750.
4
Tel.: +1 614 292 8088.
the periodic treatment of streams with chemical lampricides to kill
larvae (Christie et al., 2003; Marsden et al., 2003; Smith and Tibbles,
1980).
More streams contain sea lamprey larvae than can be treated, due
to limited resources. Streams are selected for lampricide treatment
based on (i) the assessment of larval population and size structure,
(ii) the prediction of the proportion of sea lamprey larvae within a
given stream that is likely to undergo metamorphosis into fish parasites in the following year and (iii) treatment cost (Christie et al.,
2003; Fenichel and Hansen, 2010; Treble et al., 2008). However, larval
abundance estimates do not translate into numbers of spawningphase sea lampreys produced because of complex in-stream processes and differences in the survival of newly transformed parasiticphase lampreys. Factors such as larval density may affect survival
and growth in streams (Dawson and Jones, 2009; Morman, 1987;
Rodríguez-Muñoz et al., 2003) so that not all larvae metamorphose
into parasitic-phase sea lampreys. Similarly, not all parasitic-phase
sea lampreys will contribute to the spawning adult population,
although the differences in open lake survival of parasitic-phase sea
lampreys from different streams are not well known (Jones, 2007).
Thus, treating a stream with high larval densities that will not be
translated into parasitic-phase lampreys is a poor use of scarce resources. A better understanding of the tributary production of adult
sea lampreys could improve the control program.
0380-1330/$ – see front matter © 2013 International Association for Great Lakes Research. Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jglr.2013.02.006
Please cite this article as: Howe, E.A., et al., Discriminating natal origin of spawning adult sea lamprey (Petromyzon marinus) in Lake Champlain
using statolith elemental signatu..., J Great Lakes Res (2013), http://dx.doi.org/10.1016/j.jglr.2013.02.006
2
E.A. Howe et al. / Journal of Great Lakes Research xxx (2013) xxx–xxx
Material and methods
dish floating on Milli-Q ultrapure water in an ULTRAsonik cleaner
(model 57X; Ney Dental, Inc., Bloomfield, Connecticut). Statoliths
were then transferred with a glass probe to a clean Petri dish where
they were rinsed three times in Milli-Q water. All laboratory apparatus
in contact with the statoliths were acid-washed prior to use (Ludsin et
al., 2006).
The method for statolith preparation differed depending on the
sea lamprey life stage. Because larvae were collected from their
natal streams, the stream chemical signature is represented by the
entire statolith. Larval statoliths were mounted on their base using
Scotch double-sided tape (3 M, St. Paul, Minnesota) on a petrographic
microscope slide. Larval statoliths were ablated by traversing their
entire width, from the apex to the base on the opposite side.
For the statoliths of adults, only the portion of the statoliths deposited during the larval stage was of interest. Statoliths grow in a conical
shape, with the oldest material found at the apex of the statoliths and
the most recently deposited material found at the base (Carlström,
1963; Lychakov, 1995). Thus, the post-larval stage material is expected
to be found at the base of the statoliths of adults. To properly extract
data only from the larval portion, a mid-sagittal section of the statoliths
of adults was prepared. Specifically, statoliths were mounted in crystal
bond (Structure Probe, Inc., West Chester, PA) on a strip of transparency
film, with the median plane of the statolith parallel to the film. Mounted
statoliths were then ground perpendicular to their base until their
banding patterns were exposed. Prepared statoliths were then placed
on a clean glass slide using Scotch double-sided tape. Statolith sections
were ablated along a transect from the apex to the base. The statolith
material along this transect was deposited during stream residency as
larvae and during the residency in Lake Champlain as fish parasites.
Because statolith size increases at a decreasing rate, with limited statolith growth after the larval stage (Brothers, 1987; Meeuwig and Bayer,
2005), the information recorded from the apex to three-quarters of
the way down to the base was assumed to be an accurate representation of the material deposited during stream residency at the larval
stage. Consequently, only this portion of the statoliths of adults were
considered.
Sea lamprey collection
Statolith analysis
Sea lamprey sampling focused on Lake Champlain, which lies between New York and Vermont, USA, and Quebec, Canada (Fig. 1). Ten
to 56 sea lamprey larvae were collected from each of 12 tributaries to
Lake Champlain and one tributary delta during lampricide treatments
or by electrofishing. Because larvae were collected as part of different
research projects, larvae used in this study were collected in summer
2002 through summer 2005 (Table 1, Fig. 1) and specimens were preserved in two ways. Among the 238 larvae collected, 137 were immediately frozen and 101 were preserved in 95% ethanol (Table 1).
We acquired known-origin adult sea lampreys from a tagging
study conducted by Howe et al. (2006). In that study, recently metamorphosed sea lampreys were captured in the fall of 2001 and 2002
from five tributaries to Lake Champlain: Lewis Creek, Malletts Creek,
Pike River, Morpion Stream and Saranac River (Fig. 1). All lampreys
were marked with coded wire tags and released back into their
stream of collection. Thirty-three tagged lampreys were recaptured
in 2003 and 2004 as spawning-phase sea lamprey during their upstream migration or from nests during spawning. One lamprey each
originated from Morpion Stream and Saranac River, two from Pike
River, nine from Malletts Creek, and 20 from Lewis Creek. All adults
were frozen after collection.
Statoliths were analyzed for a suite of elements using an inductively coupled plasma-mass spectrometry ICP-MS (Thermo Elemental
X7; Thermo Fisher Scientific Inc., Waltham, Mass.) coupled with a
Continuum® Surelite® solid-state Nd:YAG laser (wavelength =
266 nm, maximum power = 40 mJ, pulse rate = 20 Hz, primary
beam width = 6 mm; Continuum Inc., Santa Clara, Calif.) following
the techniques outlined by Hand et al. (2008).
A typical acquisition consisted of a 60 s measurement of the gas
blank before the laser was switched on, followed by 100 s of measurement with the laser on and statolith material being ablated. Outputs from laser-ICP-MS were counts per second. After ablation, we
chose the time intervals over which to integrate the background
(measured as gas blank) and the statolith ablation count rates. For
larvae, statolith data integration was started when the laser hit the
statolith and was terminated when the laser started to sample the
Scotch tape. For adult statoliths, data were integrated over the first
three-quarters of the time interval between the hit of the statolith
and the hit of the crystal bond, to ensure that only material deposited
at the larval stage was integrated in the signal. Calcium was used as
an internal standard to account for ablation-yield differences. A
Microsoft excel™ macro was then used to calculate backgroundcorrected signals, average the data down to one value per isotope
per statolith, and convert the counts per second into concentrations.
To calibrate analytical sensitivity, estimate measurement precision, and to account for instrumental drift, a reference standard
(National Institute of Standards and Technology [NIST] 610) was
run in pairs prior to and after every ten statoliths. A coefficient of
Natural geochemical tags in calcified structures, especially otoliths, have been widely used to track fish migration and assess natal
origin (Campana, 1999; Elsdon et al., 2008). Otoliths are calciumcarbonate concretions in the teleost sensory system that grow continuously incorporating elements from the surrounding waters in the
process and that are metabolically inert (Campana and Thorrold,
2001). Consequently, fish growing in chemically distinct waters will
record unique signatures in their otoliths that reflect those habitats.
Statoliths in sea lampreys are considered as primitive otoliths
(Gauldie, 1996; Lychakov, 1995). Elemental composition of statoliths
in sea lamprey larvae varies geographically (Brothers and Thresher,
2004; Hand et al., 2008). Accuracy of classifying larvae to their tributaries reached 88.9% for larvae from five different source locations in
the Lake Huron watershed (Brothers and Thresher, 2004), and averaged 82% among larvae from 13 streams located in lakes Michigan,
Huron and Superior with individual stream accuracies ranging from
31% to 100% (Hand et al., 2008). However, the potential for using
statolith elemental signatures to identify the natal tributaries of sea
lamprey adults is not well known, mostly because having adults of
known-origin is rare. Only one study to date has tested the use of
statolith microchemistry to identify the natal origin of known-origin
adult sea lampreys from the Lake Huron watershed (Brothers and
Thresher, 2004). In that study, 18 adult sea lampreys were assigned
to their natal river with limited success (44%) despite high assignment accuracy for the larvae from the same river. Herein, we further
examine the use of statolith microchemistry as a tool to identify natal
origins of adult sea lampreys by extending the study of Brothers and
Thresher (2004) to a different system (i.e., Lake Champlain), with
larger larval and adult sample sizes collected from a greater number
of natal streams. The elemental composition of statoliths in larvae
from tributaries of Lake Champlain was analyzed to test whether statolith chemistry varies geographically. Then, samples from known-origin
adults were used to determine the accuracy of using statolith elemental
fingerprints to assign sea lamprey adults to their natal origin.
Statolith preparation
Using the methods described by Hand et al. (2008), sagittal statoliths were dissected from the left and right otic sacs of each individual
in a Class-100 clean room and sonicated for five minutes in a Petri
Please cite this article as: Howe, E.A., et al., Discriminating natal origin of spawning adult sea lamprey (Petromyzon marinus) in Lake Champlain
using statolith elemental signatu..., J Great Lakes Res (2013), http://dx.doi.org/10.1016/j.jglr.2013.02.006
E.A. Howe et al. / Journal of Great Lakes Research xxx (2013) xxx–xxx
74°W
3
73°W
Morpion
Quebec
45°N
Pike
45°N
MISSISQUOI
BAY
Great Chazy
Missisquoi
Saranac
Salmon
Malletts
Ausable
New York
Lewis
Vermont
75°W
44°N
44°N
Mill
70°W
Lake Champlain Basin
45°N
45°N
Mt Hope
Poultney
N
01
75°W
0
kilometers
70°W
74° W
02
73° W
Fig. 1. Map of Lake Champlain and tributaries sampled for sea lamprey. The triangle denotes the single tributary delta from which statolith samples were collected.
variation (CV = standard deviation/mean × 100) was calculated for
each element in each run. Among the 10 elements analyzed, only
those with 90% or more of the samples above the limits of detection
and with an average coefficient of variation less than 10% were used
in this study (Ludsin et al., 2006). Six elements met these criteria:
magnesium (Mg), manganese (Mn), zinc (Zn), rubidium (Rb), strontium (Sr) and barium (Ba) (Table 2). All elemental concentrations
were natural log transformed to normalize the data.
Outlier identification and data corrections due to the effects of
preservation on Rb
Any single larval data point (an elemental concentration for an individual sea lamprey) that was greater than three standard deviations
from the mean for its respective tributary was considered to be an
outlier. Less than 2% of the data were outliers. We evaluated the influence of outliers by running the analysis described in this study twice:
using a dataset that included the original outlier values, and using a
dataset where outliers were replaced with a random value generated
using a normal distribution from the mean and standard deviation of
the element for that tributary, a method known for artificially increasing the precision of the measurements (Quinn and Keough,
2002). Because the two approaches led to similar conclusions, the influence of outlier values was considered negligible and the original
outlier values were kept in the analysis.
As some larvae were preserved in 95% ethanol while others were
frozen (Table 1), the effects of the mode of preservation on statolith
chemistry were tested using the larvae collected from Lewis Creek
VT in summer 2002, as Lewis Creek had the most abundant supply
of larvae at the time (Howe, 2006). Among the 56 larvae from Lewis
Creek, 29 were randomly selected and immediately preserved in
95% ethanol while the remaining 27 were immediately frozen. Only
Rb concentrations were affected by the mode of preservation
(Howe, 2006), in agreement with Hand et al. (2008). In both studies,
Rb concentrations were higher for larvae preserved in ethanol than
those that were frozen.
To remove the effect of preservation method, an analysis of variance
(ANOVA) was performed using the natural log of Rb concentration of all
individuals (whatever their stage or their natal stream) as the dependent variable and the mode of preservation (ethanol versus frozen) as
Please cite this article as: Howe, E.A., et al., Discriminating natal origin of spawning adult sea lamprey (Petromyzon marinus) in Lake Champlain
using statolith elemental signatu..., J Great Lakes Res (2013), http://dx.doi.org/10.1016/j.jglr.2013.02.006
4
E.A. Howe et al. / Journal of Great Lakes Research xxx (2013) xxx–xxx
Table 1
Sea lamprey larval sample sizes for Lake Champlain tributaries, with tributary groupings
used for discriminant analysis. The collection year, preservation method (F = frozen,
E = ethanol), and the total number of individuals (N) are shown for each sample site.
Groupings (A = individual stream level, B = streams grouped based on geographic
proximity and similar geologic drainages) are indicated by reference under a column
header. The number of larvae frozen and preserved in ethanol is indicated in parentheses
for tributaries where both methods were used. Larvae collected from Great Chazy River in
2003 and 2004 were grouped.
Site
Collection year
Preservation
N
Great Chazy River, NY
Ausable River, NY
Saranac River delta, NY
Saranac River, NY
Salmon River, NY
Mill Brook, NY
Mount Hope Brook, NY
Poultney River, NY, VT
Lewis Creek, VT
Malletts Creek, VT
Missisquoi River, VT
Morpion Stream, QUE
Pike River, QUE
2003–2004
2002
2004
2005
2002
2004
2004
2004
2002
2003
2003
2004
2004
F(6) + E(15)
F
F(5) + E(5)
F
F
F(15) + E(3)
F
E
F(27) + E(29)
E
E
F(14) + E(4)
F
21
15
10
16
15
18
10
17
56
16
12
18
14
Grouping
A
B
A1
A2
A3
A4
A5
A6
A7
A8
A9
A10
A11
A12
A13
B1
B2
B2
B2
B2
B3
B3
B4
B5
B6
B7
B7
B7
the independent variable. The aim of this ANOVA was to partition the
total variation into portions associated with the explanatory factor
(here, the mode of preservation) and residuals (the variation not
explained by the mode of preservation). Variation due to stream differences was expressed in the residuals. Thus, in this study, statolith
elemental signatures from different streams were characterized using
the residuals of the ANOVA rather than the natural log of Rb concentrations. Low statolith Rb concentrations were represented by highly
negative residuals. High Rb concentrations were represented by highly
positive residuals. Such a statistical approach has been used in other
studies (Burge, 2004).
Natal origin assignment for larvae
A quadratic discriminant function analysis (DFA) was used to determine the accuracy with which sea lamprey larvae could be assigned to
their natal origin. A quadratic DFA was appropriate because this procedure does not assume homogeneity of variance–covariance matrices
(McGarigal et al., 2000). The DFA uses a jackknife cross-validation procedure to determine classification accuracy. Two discriminant analyses
were conducted to test the effects of grouping rivers on classification accuracy. A first discriminant analysis was performed using larvae from all
12 tributaries and the single delta kept separate (grouping A). A second
discriminant analysis was performed using clustered tributaries of geographic proximity and similar geologic drainages (grouping B). The
Lake Champlain watershed is divided by areas of different geology
(Doolan, 1996; Lake Champlain Basin Program, 2004). Much of the
western side of the Lake Champlain watershed (New York) drains
the Adirondack Mountains, while most of the eastern side (Vermont)
drains the Green Mountains. The northern tributaries of the
Adirondacks – Salmon, Ausable, Saranac rivers and the Saranac delta –
were clustered together based on geographic proximity (cluster B2).
The southern tributaries of the Adirondacks – Mill and Mount Hope
brooks – were grouped together (cluster B3). Finally, Morpion Stream,
Pike River and Missisquoi River, all flowing into Missisquoi Bay at the
northeastern end of Lake Champlain, were assigned to the same group
(cluster B7) (Fig. 1, Table 1). Geographic differences in elemental signatures among rivers and groups of rivers were visualized using canonical
discriminant analysis. Canonical variate coefficients were used to assess
the relative importance of each variable to the observed separation
among rivers and groups of rivers.
Statistical analyses were performed using R software (R Development
Core Team, 2010). The packages MASS and ade4 were used to perform
the discriminant analysis and the canonical discriminant analysis, respectively (Dray and Dufour, 2007; Venables and Ripley, 2002).
Validation of natal origin assignments of adults
A maximum likelihood estimation (MLE) procedure (HISEA; Millar,
1987) was used to determine the natal origin of the 33 known-origin
adult sea lampreys. In this procedure, the adults were treated as the
stock mixture of unknown origin. The baseline was the statolith
elemental signatures of the 238 larvae from the seven reference populations presented in grouping B. Stock mixtures and associated standard
deviations were calculated in bootstrap mode by resampling the baseline 500 times with replacement. The MLE algorithm does not identify
origins of individual fish but estimates the proportions of each reference
population in the unknown mixture. The predicted proportions of
adults from each reference population were compared to the known
proportions.
To test the stability of statolith composition between larvae and the
larval portion of the statoliths of adults from the same stream, twosample t-tests were used to compare elemental concentrations
between the two groups. We did not apply a Bonferroni correction as
each river and each element were considered independently (Cabin
and Mitchell, 2000). The comparisons were restricted to Lewis Creek
and Malletts Creek due to the very low number of adults coming from
the three other streams. The level of statistical significance was at
α = 0.05.
Results
When tributaries were kept separate, larvae were assigned to their
natal origin with an average accuracy of 57.1% (range: 25.0%–80.4%)
(Table 3a). Clustering tributaries by geographic proximity and similar
geologic drainages (grouping B) improved the average classification
accuracy to 70.2% (range: 29.4%–80.4%) (Table 3b). Lewis Creek was
the best discriminated stream regardless of the grouping (Table 3a,
b). For grouping B, the separation of chemical signatures was primarily driven by variations in Rb along the first canonical variate and
by variations in Sr along the second canonical variate (Table 4). Canonical variate 1 discriminated mostly between Lewis Creek (low
Rb) and cluster B3 (Mill and Mount Hope brooks, high Rb). Canonical
variate 2 mostly separated cluster B7 (Morpion Stream, Pike and
Missisquoi rivers, high Sr) from the Great Chazy River and cluster
B2 (northern New York tributaries, low Sr) (Fig. 2, Table 5).
Classification accuracy of the known-origin adults was extremely
poor. The MLE procedure assigned 88% of the sea lamprey adults to
cluster B2 (northern New York tributaries), 11% to cluster B3 (Mill
Table 2
Elements quantified (isotope measured is indicated) in sea lamprey using laser-ablation inductively coupled plasma mass spectrometry (LA-ICP-MS). Mean limits of detection were
calculated based on all LA-ICP-MS runs. The coefficient of variation CV, determined from a reference standard (National Institute of Standards and Technology [NIST] 610), is the
average for all runs. The percentage of samples greater than detection limits (% > LOD) are shown. Elements in bold type met our criteria for inclusion in the analyses.
Element
7Li
25Mg
55Mn
66Zn
85Rb
86Sr
138Ba
140Ce
208Pb
238U
LOD (ppm)
CV (%)
% > LOD
0.98
7.94
10%
13.27
3.10
100%
0.68
2.91
100%
0.92
8.06
92%
0.32
5.70
94%
3.93
6.35
100%
0.17
6.99
100%
0.04
5.66
25%
0.14
8.53
70%
0.06
6.66
46%
Please cite this article as: Howe, E.A., et al., Discriminating natal origin of spawning adult sea lamprey (Petromyzon marinus) in Lake Champlain
using statolith elemental signatu..., J Great Lakes Res (2013), http://dx.doi.org/10.1016/j.jglr.2013.02.006
E.A. Howe et al. / Journal of Great Lakes Research xxx (2013) xxx–xxx
5
Table 3
Cross-validation summary from the quadratic discrimination function analysis for classifying Lake Champlain sea lamprey larvae to their natal origin for groupings A (panel a) and B
(panel b). Rows represent the known streams or clusters of origin, columns are predicted streams or clusters of origin. Reported values are percent classification with the number of
individuals in each classification (in parentheses). Accurate classifications are shown in bold, on the diagonal.
a.
Great Chazy
R.
Ausable
R.
Saranac R.
delta
Saranac
R.
Salmon
R.
Mill
Br.
Mount Hope
Br.
Poultney
R.
Lewis
Cr.
Malletts
Cr.
Missisquoi
R.
Morpion
St.
Pike
R.
Great Chazy R.
(21)
Ausable R. (15)
Saranac R. Delta
(10)
Saranac R. (16)
Salmon R. (15)
Mill Br. (18)
Mount Hope Br.
(10)
Poultney R. (17)
Lewis Cr. (56)
Malletts Cr. (16)
Missiquoi R. (12)
Morpion Str.
(18)
Pike R. (14)
Overall accuracy
71.4
0.0
0.0
14.3
4.8
0.0
0.0
0.0
9.5
0.0
0.0
0.0
0.0
0.0
0.0
26.7
0.0
0.0
40.0
13.3
20.0
0.0
10.0
13.3
0.0
0.0
0.0
6.7
0.0
40.0
30.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
6.3
0.0
5.6
0.0
6.3
0.0
11.1
0.0
0.0
0.0
0.0
0.0
62.5
6.7
5.6
0.0
18.8
40.0
0.0
0.0
0.0
13.3
55.6
40.0
0.0
0.0
5.6
60.0
0.0
6.7
5.6
0.0
6.3
20.0
0.0
0.0
0.0
6.7
0.0
0.0
0.0
6.7
0.0
0.0
0.0
0.0
11.1
0.0
0.0
0.0
0.0
0.0
5.9
1.8
0.0
0.0
0.0
5.9
5.4
12.5
8.3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.8
0.0
0.0
0.0
0.0
3.6
6.3
8.3
0.0
11.8
0.0
6.3
0.0
22.2
0.0
0.0
0.0
0.0
0.0
29.4
1.8
18.8
25.0
0.0
11.8
80.4
0.0
16.7
0.0
5.9
0.0
56.3
8.3
0.0
29.4
5.4
0.0
25.0
0.0
0.0
0.0
0.0
8.3
66.7
0.0
0.0
0.0
0.0
11.1
0.0
0.0
0.0
0.0
0.0
7.1
0.0
14.3
0.0
0.0
7.1
21.4
50.0
57.1
b.
Great Chazy R.
B2
B3
Poulney R.
Lewis Cr.
Malletts Cr.
B7
Great Chazy R. (21)
B2 (56)
B3 (28)
Poultney R. (17)
Lewis Cr. (56)
Malletts Cr. (16)
B7 (44)
Overall accuracy
66.7
0.0
0.0
5.9
1.8
0.0
0.0
23.8
75.0
10.7
11.8
10.7
18.8
6.8
0.0
7.1
78.6
0.0
0.0
12.5
9.1
0.0
1.8
0.0
29.4
1.8
12.5
9.1
9.5
14.3
0.0
5.9
80.4
0.0
4.5
0.0
0.0
3.6
5.9
0.0
50.0
0.0
0.0
1.8
7.1
41.2
5.4
6.3
70.5
70.2
and Mount Hope brooks), and the remaining 1% to Great Chazy River
(Fig. 3). Based on the tagging study (Howe et al., 2006), the majority
of adults originated from Lewis Creek (62%) and Malletts Creek (26%)
in Vermont. Only 3% of the mixed stock originated from cluster B2
and none from cluster B3 and Great Chazy River.
For both Lewis Creek and Malletts Creek, elemental concentrations
varied between larvae and adult stages. For all elements except Sr, a
significant difference was detected between larvae and adults originating from Lewis Creek (Table 6). The differences were higher for
Rb and Zn, compared to all other elements. The statoliths of adults
exhibited higher concentrations of Rb and Zn than larvae (Table 5).
For Malletts Creek, significant differences were found for Mn, Rb
and Ba (Table 6). The difference was particularly high for Rb and to
a lesser extent for Ba, where higher concentrations were found in
the statoliths of adults (Table 5).
proximity and geologic similarities (grouping B). Thus, statolith
elemental signature can be used to discriminate natal origins of larvae
in the Lake Champlain watershed but at the cost of losing resolution
at the individual tributary level. Clustering streams did not mask
any stream with a highly specific signature, as the grouping involved
streams that were not uniquely identifiable at the individual level;
the single exception was Morpion Stream, for which the classification
accuracy was 66.7% before clustering. Regardless of the grouping, our
correct classification rate of larvae was lower than in similar studies
performed in different systems. Sea lamprey larvae from 13 streams
located in lakes Michigan, Huron and Superior were discriminated
with 82% classification accuracy (Hand et al., 2008). The classification
Discussion
The correct classification rate of larvae reached 57.1% when all the
streams were considered individually (grouping A) and it increased
to 70.2% when combining sea lamprey from tributaries by geographic
Table 4
Coefficients for canonical discriminant analysis performed on the natural log concentrations in Lake Champlain sea lamprey larval statoliths collected from the sources
presented in grouping B.
Ln (Mg)
Ln (Mn)
Ln (Zn)
Residuals Ln (Rb)
Ln (Sr)
Ln (Ba)
Cumulative proportion
Canonical
Canonical
Variate 1
Variate 2
−0.37
−0.17
0.48
0.62
0.40
−0.17
0.38
0.38
−0.04
−0.03
0.42
−0.96
0.48
0.71
Fig. 2. Canonical discriminant analysis of sea lamprey larvae statolith signatures from
the seven Lake Champlain sources (rivers and cluster of rivers) presented in grouping
B. Source symbols are as follows: cluster B2 (●), cluster B3 (△), cluster B7 (✳), Great
Chazy River (○), Lewis Creek (w), Malletts Creek (■), Poultney River (×). Tributaries
in each cluster (B2, B3 and B7) are listed in Table 1.
Please cite this article as: Howe, E.A., et al., Discriminating natal origin of spawning adult sea lamprey (Petromyzon marinus) in Lake Champlain
using statolith elemental signatu..., J Great Lakes Res (2013), http://dx.doi.org/10.1016/j.jglr.2013.02.006
6
E.A. Howe et al. / Journal of Great Lakes Research xxx (2013) xxx–xxx
Table 5
Sea lamprey statolith elemental signatures for larvae (L) and adults (A) from the seven Lake Champlain sources of grouping B. Median (min–max) concentrations in ppm are
presented for all elements except Rb for which the residuals of the ANOVA between Ln(Rb) and the mode of preservation, as explained in the Material and methods section, are
shown.
Natal origin
Mg × 10²
Mn
Zn
Rb residuals
Sr
Ba
Great Chazy
River (L)
B2 (L)
B2 (A)
B3 (L)
Poultney R. (L)
Lewis C. (L)
Lewis C. (A)
Malletts C. (L)
Malletts C. (A)
B7 (L)
B7 (A)
45.3 (32.7–243.5)
48.8 (34.3–274.4)
19.3 (0.4–77.3)
0.1 (−0.5–0.7)
341.3 (263.0–1169.3)
23.9 (12.6–70.2)
56.3 (36.0–76.5)
47.6
34.7 (21.9–73.6)
59.30 (36.4–105.9)
55.3 (42.8–70.2)
74.1 (54.6–89.5)
54.8 (31.3–335.7)
63.2 (51.1–90.7)
47.4 (33.1–169.1)
64.4 (62.4–65.0)
63.0 (25.9–250.7)
114.6
35.4 (16.9–74.2)
75.4 (40.8–112.0)
54.8 (34.4–105.0)
61.6 (46.7–110.3)
48.1 (32.9–64.3)
76.4 (50.3–182.2)
53.7 (16.9–218.6)
74.3 (49.8–99.5)
6.7 (0.7–428.5)
18.2
25.2 (2.0–75.4)
11.0 (1.4–64.5)
1.9 (0.2–33.6)
46.8 (1.9–259.7)
21.9 (1.0–66.7)
64.1 (2.9–222.8)
18.6 (0.7–123.3)
414.3 (121.9–954.4)
0.5 (−1.0–1.8)
1.2
0.6 (0.0–2.3)
−0.4 (−0.9–0.3)
−1.1 (−2.4–0.5)
1.1 (0.6–1.5)
0.0 (−0.6–0.7)
1.4 (0.9–1.6)
−0.6 (−2.0–0.7)
0.7 (0.6–1.2)
582.0 (329.2–3194.0)
482.4
679.0 (368.9–1645.5)
676.0 (537.3–1760.8)
619.6 (411.8–1256.0)
652.9 (474.0–1387.9)
629.3 (346.6–1825.6)
693.5 (613.3–1033.0)
1063.5 (462.7–1853.8)
1580.6 (916.7–1677.3)
24.9 (10.0–142.4)
42.6
21.9 (5.0–429.9)
19.8 (5.0–41.6)
24.8 (9.3–191.1)
35.6 (18.3–120.2)
8.9 (4.5–46.3)
19.8 (14.5–48.2)
20.8 (7.8–63.7)
62.0 (26.5–96.6)
accuracy reached 88.9% for larvae from five different source locations
in Lake Huron watershed (Brothers and Thresher, 2004). Hand et al.
(2008) recommend a minimum of 15 larvae per stream to characterize stream-specific chemical signatures, a criterion that was not met
for all the sites sampled in our study. However, the number of larvae
per stream in our study falls into the ranges used by Brothers and
Thresher (2004) and Hand et al. (2008), which were 4–19 and
10–30 larvae per stream, respectively. In addition, 70% of the sites
in our study were described by 15 individuals or more, while this percentage was 25% and 54% for the study of Brothers and Thresher
(2004) and Hand et al. (2008), respectively. Consequently, our
lower classification success is not related to larval sample size. Rather,
we suspect that streams in Lake Champlain watershed are not as
chemically different as streams in the Great Lakes.
Despite the 70.2% success of larval discrimination, assignment of
adults to their stream of origin was highly inaccurate. A similar conclusion was drawn by Brothers and Thresher (2004) although their classification success was higher than ours with 44% of adult sea lampreys
tracked back to their river of origin in the Lake Huron basin. The performance of classification using MLE is affected by factors such as sample
size in the baseline (e.g., larvae from each stream), sample size in the
group mixture (e.g., adults), and the number of classification variables
(e.g., elements) (Millar, 1987, 1990; Wood et al., 1989). But more
importantly, accurate prediction of natal origin using natural tags
strongly relies on three criteria: (i) each group in the baseline should
be characterized by specific and reproducible markers; (ii) all possible
groups contributing to the group mixture should be characterized;
and (iii) group-specific markers should be stable over the interval
between characterization (baseline) and mixing (group mixture)
(Campana et al., 2000).
According to the first criterion defined by Campana et al. (2000),
statolith chemistry among larvae from different tributaries should
be characterized by specific and reproducible markers. The wide
range of larval stream assignment accuracy (from around 25 to 80%)
indicates that in the Lake Champlain watershed some streams present
more specific and reproducible markers than others, which shows
that the statolith microchemistry approach is not necessarily appropriate to depict the chemical characteristics for all streams. Interestingly, even adults from streams with highly specific signature
showed poor success in natal origin assignment. This discrepancy is
well illustrated by Lewis Creek. With a classification accuracy of
80%, larvae from Lewis Creek present specific markers. They mostly
differ from the larvae of other streams by their relatively low statolith
Rb and Zn concentrations. However, none of the adults originating
from Lewis Creek were successfully assigned to this river. Consequently, factors other than the ability to discriminate among the
geographic locations in the reference groups (larvae) explain the
unsuccessful natal origin identification for adult sea lamprey.
According to the second criterion defined by Campana et al.
(2000), all the groups contributing to the group mixture should be
characterized in the baseline. Uncharacterized groups of fish present
in the stock mixture could be mistakenly interpreted and assigned
to a group that was characterized in the baseline (Gillanders, 2005).
Consequently, an exhaustive identification of potential sources is critical (Waldman and Fabrizio, 1994). Our study meets this criterion:
although we did not sample all streams that contain sea lamprey
larvae, the five streams from which the adults originated were
known and characterized in the baseline.
Table 6
Probabilities associated with the t-test comparisons of statolith chemistry between sea
lamprey larvae and adults originating from Lewis Creek and Malletts Creek. A t-test
was performed for each element and each river using the log-natural transformations
for all elements except for Rb where the residuals of the ANOVA between Ln(Rb) and
the mode of preservation were considered. Significant effects (P b 0.05) are indicated
by asterisks.
Element
Fig. 3. Percent composition estimates derived from mixed-stock analysis (grey bars)
and percent composition of known-origin sea lamprey adults from Lake Champlain
(white bars). B2 = cluster B2, B3 = cluster B3, B7 = cluster B7, GCR = Great Chazy
River, LEW = Lewis Creek, MAL = Malletts Creek, POUL = Poultney River.
Mg
Mn
Zn
Rb residuals
Sr
Ba
Stream of origin
Lewis Creek
Malletts Creek
b0.001*
0.015*
b0.001*
b0.001*
0.786
0.001*
0.826
b0.001*
0.231
b0.001*
0.483
0.001*
Please cite this article as: Howe, E.A., et al., Discriminating natal origin of spawning adult sea lamprey (Petromyzon marinus) in Lake Champlain
using statolith elemental signatu..., J Great Lakes Res (2013), http://dx.doi.org/10.1016/j.jglr.2013.02.006
E.A. Howe et al. / Journal of Great Lakes Research xxx (2013) xxx–xxx
The third criterion of Campana et al. (2000) requires that
group-specific markers remain stable over the interval between the
larval stage (used for group characterization) and the adult stage
(for which natal origin has to be assigned). The comparison of statolith chemistry between larvae and adults revealed that the concentrations of most elements vary between the two life stages, with the
most notable differences reported for Rb and to a lesser extent for
Zn and Ba. Therefore, the criterion of stability was not met for all of
the group-specific markers. In their study, Brothers and Thresher
(2004) reported higher variability in Rb and Zn concentrations for
adults than for larvae from the Black Mallard river (Lake Huron tributary), although the differences between the two life-stages were
not significant. It is important to emphasize that the higher Rb concentrations in adults are not an artifact of preservation method.
Frozen samples exhibit lower Rb concentrations than alcoholpreserved samples (Hand et al., 2008; Howe, 2006). Adults were
frozen so their higher Rb concentrations cannot be an effect of the
preservation method. Conversely, Sr did not differ between the two
life stages from the same streams, suggesting that the stability criterion was met for this element. The effects of the non-stability of
group-specific markers on our ability to identify natal origin of adult
sea lamprey depend on the amplitude of the variation and the
power of elements to discriminate among lamprey from different
streams. Rubidium, which shows the most notable change between
larvae and adults for Lewis Creek and Mallets Creek, is among the
most important elements for larval discrimination in our study as
well as others (Brothers and Thresher, 2004; Hand et al., 2008).
The stability of group-specific markers might be affected in at least
two ways. First, investigations into the natural tag properties of
otoliths revealed that elemental composition of individuals from the
same stream may vary among years, in relation to annual changes
in water chemistry (Gillanders, 2002b; Schaffler and Winkelman,
2008). Such temporal variations might complicate natal origin identification. As an example, natal origin discrimination success of one
year-class juvenile striped bass Morone saxatilis from two tributaries
of Lake Texoma (Oklahoma-Texas) was lower when the discriminant
functions were based on individuals from another year-class compared
to individuals from the same year-class (Schaffler and Winkelman,
2008). To account for temporal variability, it is often recommended to
match cohorts between fish used in the baseline to characterize the
potential natal origins and fish of unknown origin (Gillanders, 2002a;
Walther and Thorrold, 2009). Because sea lamprey age estimation is
still problematic (Dawson et al., 2009), the sea lamprey larvae and
adults from our study could not be assigned to any cohort. However,
we can offer an estimation of their stream residency time. The adults
from our study were first captured and tagged in fall 2001 and fall
2002, when they were in their stream of origin (Howe et al., 2006).
Assuming that the larval phase in Lake Champlain sea lamprey lasts
for four years before metamorphosis (Marsden et al., 2003), the adult
sea lampreys were in their natal streams as larvae from 1997 to 2001
or from 1998 to 2002. The larvae from this study were collected from
tributaries in 2002, 2003, 2004, and 2005 and they had probably spent
a few years (less than four) in these streams before collection. Consequently, the adults might have been in streams as larvae at the same
time as the larvae collected in this study. However, the duration of
this overlap remains unknown and could be limited. The uncertainties
associated with the temporal variability of stream-specific signatures
are also amplified by our approach itself. Stream chemical signatures
were characterized using measurements in the entire statoliths of sea
lamprey larvae, integrating material deposited over the several years
of larval stream residency. The extent to which the statolith chemical
variability integrated over these years accurately depicts the variability
occurring over a longer time period is unknown. In addition, as statolith
size increases at a decreasing rate (Meeuwig and Bayer, 2005), it is very
likely that earlier years contributed more to stream-specific chemical
signatures than older years. Consequently, the extent to which temporal
7
variability affects our ability to identify natal origin of adult sea lamprey
remains unclear.
Second, group-specific markers can also be affected by ontogenic
effects. Unlike otoliths (Campana and Neilson, 1985), the absence of
reworking of previously deposited statolith material has never been
demonstrated. Two lines of evidence suggest the potential for chemical
reworking of the statoliths. First, statoliths are made of apatite, a basic
form of calcium phosphate (Carlström, 1963). The same mineral is
found in teeth, bones and scales (Ikoma et al., 2003; Pasteris et al.,
2008), for which resorption has been shown (Kacem et al., 1998;
Witten et al., 2000). Second, Barker et al. (1997) reported cases where
statoliths were found in larvae but they were not systematically present
in metamorphosing lampreys from the same stream, suggesting that
statolith resorption might be related to stream calcium content. Resorption in bones or scales is often associated with events of intense physiological stress (Kacem et al., 1998). For sea lamprey, metamorphosis and
sexual maturation are events of intense physiological changes (Larsen,
1980; Youson, 2003) that might induce restructuring of the statolith.
The results of this study invite further research regarding statolith
chemical stability through ontogeny and, at this time, we do not recommend using statolith microchemistry to classify adults to a natal site.
Because the sea lamprey control program in Lake Champlain and the
Great Lakes would greatly benefit from the natal origin identification
of sea lamprey adult spawners, by improving the prioritization of
streams to be treated with chemical lampricides for example, we
strongly recommend pursuing the efforts to validate this approach.
Acknowledgments
This study was funded by the Lake Champlain Basin Program,
USFWS, and the Great Lakes Fishery Commission. We thank C. Martin,
USFWS, for his assistance in obtaining funding, W. Bouffard, S. J.
Smith, and the staff of the USFWS Lake Champlain Fish and Wildlife Resources Office, staff of the New York State Department of Environmental
Conservation, J. Fricke, and A. Zerrenner for their assistance in collecting
sea lamprey, C. Martin and W. Bouffard, for the help with the logistics of
this study in its early stages, P. Smith, J. Gagnon, J. Barrette, and Z. Yang
for their assistance with the LA-ICP-MS at the Great Lakes Institute for
Environmental Research in Windsor, Ontario, and C. J. Goodnight
for his statistical assistance. We are also thankful to the anonymous
reviewers for their constructive comments on the paper.
References
Barker, L.A., Morrison, B.J., Wicks, B.J., Beamish, F.W.H., 1997. Age discrimination and
statolith diversity in sea lamprey from streams with varying alkalinity. Trans.
Am. Fish. Soc. 126, 1021–1026.
Brothers, E.B., 1987. Elemental composition of statoliths of sea lamprey (Petromyzon
marinus). Great Lakes Fishery Commission Project Completion Report. Great
Lakes Fishery Commission, Ann Arbor, Mich (71 pp.).
Brothers, E., Thresher, R., 2004. Statolith chemical analysis as a means of identifying
stream origins of lampreys in Lake Huron. Trans. Am. Fish. Soc. 133, 1107–1116.
Burge, L.M., 2004. Testing links between river patterns and in-channel characteristics
using MRPP and ANOVA. Geomorphology 63, 115–130.
Cabin, R., Mitchell, R., 2000. To Bonferroni or not to Bonferroni: when and how are the
questions. Bull. Ecol. Soc. Am. 81, 246–248.
Campana, S.E., 1999. Chemistry and composition of fish otoliths : pathways, mechanisms and applications. Mar. Ecol. Prog. Ser. 188, 263–297.
Campana, S.E., Neilson, J.D., 1985. Microstructure of fish otoliths. Can. J. Fish. Aquat. Sci.
42, 1014–1032.
Campana, S.E., Thorrold, S.R., 2001. Otoliths, increments, and elements: keys to a comprehensive understanding of fish populations? Can. J. Fish. Aquat. Sci. 58, 30–38.
Campana, S.E., Chouinard, G.A., Hanson, J.M., Frechet, A., Brattey, J., 2000. Otolith elemental fingerprints as biological tracers of fish stocks. Fish. Res. 46, 343–357.
Carlström, D., 1963. A crystallographic study of vertebrate otoliths. Biol. Bull. 441–63.
Christie, G.C., Adams, J.V., Steeves, T.B., Slade, J.W., Cuddy, D.W., Fodale, M.F., Young,
R.J., Kuc, M., Jones, M.L., 2003. Selecting Great Lakes streams for lampricide treatment based on larval sea lamprey surveys. J. Great Lakes Res. 29, 152–160.
Dawson, H.A., Jones, M.L., 2009. Factors affecting recruitment dynamics of Great Lakes
sea lamprey (Petromyzon marinus) populations. J. Great Lakes Res. 35, 353–360.
Dawson, H.A., Jones, M.L., Scribner, K.T., Gilmore, S.A., 2009. An assessment of age determination methods for Great Lakes larval sea lampreys. N. Am. J. Fish Manag.
29, 914–927.
Please cite this article as: Howe, E.A., et al., Discriminating natal origin of spawning adult sea lamprey (Petromyzon marinus) in Lake Champlain
using statolith elemental signatu..., J Great Lakes Res (2013), http://dx.doi.org/10.1016/j.jglr.2013.02.006
8
E.A. Howe et al. / Journal of Great Lakes Research xxx (2013) xxx–xxx
Doolan, B., 1996. The geology of Vermont. Rocks Miner. 71, 218–225.
Dray, S., Dufour, A.B., 2007. The ade4 package: implementing the duality diagram for
ecologists. J. Stat. Softw. 22, 1–20.
Elsdon, T.S., Wells, B.K., Campana, S.E., Gillanders, B.M., Jones, C., Limburg, K.E., Secor,
D.H., Thorrold, S.R., Walther, B.D., 2008. Otolith chemistry to describe movements
and life-history parameters of fishes: hypotheses, assumptions, limitations and inferences. Oceanogr. Mar. Biol. Annu. Rev. 46, 297–330.
Farmer, G.J., 1980. Biology and physiology of feeding in adult lampreys. Can. J. Fish.
Aquat. Sci. 37, 1751–1761.
Fenichel, E.P., Hansen, G.J.A., 2010. The opportunity cost of information: an economic
framework for understanding the balance between assessment and control in sea
lamprey (Petromyzon marinus) management. Can. J. Fish. Aquat. Sci. 67, 209–216.
Gauldie, R.W., 1996. Fusion of otoconia: a stage in the development of the otolith in the
evolution of fishes. Acta Zool. 77, 1–23.
Gillanders, B.M., 2002a. Connectivity between juvenile and adult fish populations: do
adults remain near their recruitment estuaries? Mar. Ecol. Prog. Ser. 240, 215–223.
Gillanders, B.M., 2002b. Temporal and spatial variability in elemental composition of
otoliths : implications for determining stock identity and connectivity of populations. Can. J. Fish. Aquat. Sci. 59, 669–679.
Gillanders, B.M., 2005. Using elemental chemistry of fish otoliths to determine connectivity between estuarine and coastal habitats. Estuar. Coast. Shelf Sci. 64, 47–57.
Hand, C.P., Ludsin, S.A., Fryer, B.J., Marsden, J.E., 2008. Statolith microchemistry as a technique for discriminating among Great Lakes sea lamprey (Petromyzon marinus)
spawning tributaries. Can. J. Fish. Aquat. Sci. 65, 1153–1164.
Howe, E.A., 2006. A life cycle approach to modeling sea lamprey in the Lake Champlain
basin, Ph.D. thesis, The University of Vermont, Department of Natural Resources,
Burlington, VT.
Howe, E.A., Marsden, J.E., Bouffard, W., 2006. Movement of sea lamprey in the Lake
Champlain basin. J. Great Lakes Res. 32, 776–787.
Ikoma, T., Kobayashi, H., Tanaka, J., Walsh, D., Mann, S., 2003. Microstructure, mechanical, and biomimetic properties of fish scales from Pagrus major. J. Struct. Biol. 142,
327–333.
Jones, M.L., 2007. Toward improved assessment of sea lamprey population dynamics in
support of cost-effective sea lamprey management. J. Great Lakes Res. 33, 35–47.
Kacem, A., Meunier, F.J., Baglinière, J.L., 1998. A quantitative study of morphological
and histological changes in the skeleton of Salmo salar during its anadromous migration. J. Fish Biol. 53, 1096–1109.
Lake Champlain Basin Program, 2004. Lake Champlain Basin Atlas. Available from
http://www.lcbp.org/Atlas/HTML/nat_geology.htm (accessed 20 January 2012).
Larsen, L.O., 1980. Physiology of adult lampreys, with special regard to natural starvation,
reproduction, and death after spawning. Can. J. Fish. Aquat. Sci. 37, 1762–1779.
Ludsin, S.A., Fryer, B.J., Gagnon, J.E., 2006. Comparison of solution-based versus laser
ablation inductively coupled plasma mass spectrometry for analysis of larval fish
otolith microelemental composition. Trans. Am. Fish. Soc. 135, 218–231.
Lychakov, D.V., 1995. Study on structure of the otolith membrane in the lamprey Lampetra
fluviatilis in the context of otolith and otoconium evolution. J. Evol. Biochem. Physiol.
31, 90–97.
Marsden, J.E., Chipman, B.D., Nashett, L.J., Anderson, J.K., Bouffard, W., Durfey, L.,
Gersmehl, J.E., Schoch, W.F., Staats, N.R., Zerrenner, A., 2003. Sea lamprey control
in Lake Champlain. J. Great Lakes Res. 29, 655–676.
McGarigal, K., Cushman, S., Stafford, S., 2000. Multivariate Statistics for Wildlife and
Ecology Research. Springer, New York.
Meeuwig, M.H., Bayer, J.M., 2005. Morphology and aging precision of statoliths from
larvae of Columbia River basin lampreys. N. Am. J. Fish Manag. 25, 38–48.
Millar, R.B., 1987. Maximum likelihood estimation of mixed stock fishery composition.
Can. J. Fish. Aquat. Sci. 44, 583–590.
Millar, R.B., 1990. Comparison of methods for estimating mixed stock fishery composition. Can. J. Fish. Aquat. Sci. 47, 2235–2241.
Moore, J.W., Mallat, J.M., 1980. Feeding of larval lamprey. Can. J. Fish. Aquat. Sci. 37,
1658–1664.
Morman, R.H., 1987. Relationship of density to growth and metamorphosis of caged larval
sea lampreys, Petromyzon marinus Linnaeus, in Michigan streams. J. Fish Biol. 30,
173–181.
Pasteris, J.D., Wopenka, B., Valsami-Jones, E., 2008. Bone and tooth mineralization: why
apatite? Elements 4, 97–104.
Potter, I.C., 1980. Ecology of larval and metamorphosing lampreys. Can. J. Fish. Aquat.
Sci. 37, 1641–1657.
Quinn, G.P., Keough, M.J., 2002. Experimental Design and Data Analysis for Biologists.
Cambridge University Press, Cambridge, UK.
R Development Core Team, 2010. R: a Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria3-900051-07-0
(URL http://www.R-project.org).
Rodríguez-Muñoz, R., Nicieza, A.G., Braña, F., 2003. Density-dependent growth of sea
lamprey larvae: evidence for chemical interference. Funct. Ecol. 17, 403–408.
Schaffler, J.J., Winkelman, D.L., 2008. Temporal and spatial variability in otolith traceelement signatures of juvenile striped bass from spawning locations in Lake
Texoma, Oklahoma-Texas. Trans. Am. Fish. Soc. 137, 818–829.
Smith, B.R., Tibbles, J.J., 1980. Sea lamprey (Petromyzon marinus) in lakes Huron, Michigan, and Superior: history of invasion and control, 1936–78. Can. J. Fish. Aquat. Sci.
37, 1780–1801.
Swink, W.D., 2003. Host selection and lethality of attacks by sea lampreys (Petromyzon
marinus) in laboratory studies. J. Great Lakes Res. 29, 307–319.
Treble, A.J., Jones, M.L., Steeves, T.B., 2008. Development and evaluation of a new predictive model for metamorphosis of Great Lakes larval sea lamprey (Petromyzon
marinus) populations. J. Great Lakes Res. 34, 404–417.
Venables, W.N., Ripley, B.D., 2002. Modern Applied Statistics With S, fourth ed. Springer, New York.
Waldman, J.R., Fabrizio, M.C., 1994. Problems of stock definition in estimating relative contributions of Atlantic striped bass to the coastal fishery. Trans. Am. Fish. Soc. 123,
766–778.
Walther, B.D., Thorrold, S.R., 2009. Inter-annual variability in isotope and elemental ratios recorded in otoliths of an anadromous fish. J. Geochem. Explor. 102, 181–186.
Witten, P.E., Villwock, W., Peters, N., Hall, B.K., 2000. Bone resorption and bone
remodelling in juvenile carp, Cyprinus carpio L. J. Appl. Ichthyol. 16, 254–261.
Wood, C.C., Rutherford, D.T., McKinnell, S., 1989. Identification of Sockeye Salmon
(Oncorhynchus nerka) stocks in mixed-stock fisheries in British Columbia and
Southeast Alaska using biological markers. Can. J. Fish. Aquat. Sci. 46, 2108–2120.
Youson, J.H., 2003. The biology of metamorphosis in sea lampreys: endocrine, environmental, and physiological cues and events, and their potential application to lamprey control. J. Great Lakes Res. 29, 26–49.
Please cite this article as: Howe, E.A., et al., Discriminating natal origin of spawning adult sea lamprey (Petromyzon marinus) in Lake Champlain
using statolith elemental signatu..., J Great Lakes Res (2013), http://dx.doi.org/10.1016/j.jglr.2013.02.006