Using Otoliths to Evaluate a Novel Measuring Technique ......Using Otoliths to Evaluate a Novel...
Transcript of Using Otoliths to Evaluate a Novel Measuring Technique ......Using Otoliths to Evaluate a Novel...
Using Otoliths to Evaluate a Novel Measuring Technique,
Identify Spawning Years, and Compare Somatic Growth
Patterns in Lake Trout, Salvelinus namaycush
By
Danielle Jean Eisner
A thesis submitted in conformity with the requirements
for the degree of Master’s of Science
Graduate Department of Ecology and Evolutionary Biology
University of Toronto
© Copyright by Danielle Jean Eisner 2014
ii
Using Otoliths to Evaluate a Novel Measuring Technique, Identify
Spawning Years, and Compare Somatic Growth Patterns in Lake
Trout, Salvelinus namaycush
Danielle Jean Eisner
Master’s of Science
Graduate Department of Ecology and Evolutionary Biology
2014
Abstract
I evaluated a novel, simple-to-use, efficient method to estimate age and examine reproductive
history of lake trout, Salvelinus namaycush, using sagittal otolith sections. Age estimations were
made using a novel otolith section measuring technique that uses a series of parallel lines and
approximates the difficult to locate nucleus with a straight line. Previous measuring techniques
used straight lines or a segmented line to measure from nucleus to edge. Reproductive history
was examined by investigating the prevalence of post-maturation checks; I provide evidence that
these are good indicators of spawning. Additionally, mature females were significantly more
likely to skip spawning than mature males (p=0.0375). Additionally, I compared the growth
trajectories of mature tagged fish to mature free-living fish and did not find a statistically
significant effect of tagging on the growth of either sex (male: F1=2.783, N1=61, N2=108,
p=0.0967; female: F1=0.129, N1=79, N2=64, p>0.05).
iii
Acknowledgements
To my supervisor, Brian Shuter, and committee; Helen Rodd, and Ken Minns, I could not have
done this work without your help, support, encouragement, and most importantly, your
enthusiasm towards my project. The support and collaboration that I received from the students
and staff in the Brian Shuter and Don Jackson labs during this work helped me gain a greater
understanding of aquatic systems, form great friendships, and kept me excited for future
research.
A big thanks to Harkness Laboratory of Fisheries Research and Mark Ridgeway for letting me
use the previously collected samples and facilities. Gary Ridout was instrumental in my learning
of the proper aging procedure. And the students and staff of Harkness who made my summers
amazing in ways I will always treasure.
The academic curriculum offered by my undergraduate degree at the University of Guelph
sparked my interest in pursuing aquatic research. Thank you Cortland Griswold and Karl
Cottenie for encouraging me to do independent projects during my undergraduate degree.
And, of course, a huge thanks to my wonderful friends and family who supported me through
this process.
vi
Table of Contents
Abstract ................................................................................................................................... ii
Acknowledgements ................................................................................................................. iii
Table of Contents ................................................................................................................... vi
List of Tables ........................................................................................................................ viii
List of Figures ......................................................................................................................... ix
List of Appendices .................................................................................................................. x
1 Introduction ......................................................................................................................... 1
1.1 Current research .......................................................................................................... 1
1.2 Why otoliths are a good aging structure ....................................................................... 3
1.3 Otolith structure and function........................................................................................ 4
1.4 Otolith growth ............................................................................................................... 5
1.5 Otolith markings ........................................................................................................... 6
1.6 Skipped spawning ........................................................................................................ 7
1.7 Aging ............................................................................................................................ 8
1.8 Fish growth .................................................................................................................. 8
1.9 Measuring technique .................................................................................................... 8
1.10 Tagging uses ............................................................................................................. 10
1.11 Study species ............................................................................................................. 11
1.12 Study objectives ......................................................................................................... 12
2 Methods ............................................................................................................................ 13
2.1 Collection ................................................................................................................... 13
2.2 Otolith extraction ........................................................................................................ 13
2.3 Otolith preparation ...................................................................................................... 13
2.4 Imaging ...................................................................................................................... 15
2.5 Measuring techniques ................................................................................................ 15
2.6 Aging, annuli and checks ........................................................................................... 21
2.7 Tagged fish growth ..................................................................................................... 23
2.8 ‘Spawning’ checks ...................................................................................................... 24
2.9 Statistical methods ..................................................................................................... 25
3 Results .............................................................................................................................. 28
3.1 Measuring technique .................................................................................................. 28
3.2 ‘Spawning’ checks ...................................................................................................... 28
3.3 Tagged fish ................................................................................................................ 34
vii
4 Discussion ......................................................................................................................... 42
4.1 Measuring technique .................................................................................................. 42
4.2 ‘Spawning’ checks ...................................................................................................... 45
4.3 Tagged fish ................................................................................................................ 49
5 Conclusions ...................................................................................................................... 53
6 References ........................................................................................................................ 54
Appendices .............................................................................................................................. 62
viii
List of Tables
Table 1. Parameters and correlation coefficients for the otolith-age relationship of immature and
mature fish. . ....................................................................................................................32
Table 2. Parameters and correlation coefficients for the otolith length-fork length relationship of
tagged and free-living mature fish. . ................................................................................37
Table 3. Von Bertalanffy parameters for free-living mature and tagged fish and the fit associated
with the curve created by the parameters. . ......................................................................39
ix
List of Figures
Figure 1. Straight line measuring technique. . .............................................................................17
Figure 2. Line segments measuring technique. . ..........................................................................18
Figure 3. Parallel lines (novel) measuring technique. . ................................................................19
Figure 4. Identification of the topmost part of the sulcus for the otolith presented in Figures 1-3. .
............................................................................................................................................20
Figure 5. The visual difference between an annulus and a ‘spawning’ check. . ..........................22
Figure 6. The relationship between otolith length and fork length for each of the three otolith
measuring techniques; parallel lines, line segments, and straight line. . ..........................29
Figure 7. The relationship between otolith length and fork length for each sex using the parallel
lines measuring technique. . .............................................................................................30
Figure 8. Age-otolith relationship for both immature and mature lake trout. . ............................31
Figure 9. The proportion of tagged fish that had a ‘spawning’ check the year they were tagged,
two years before they were tagged, and two years after they were tagged. . ...................33
Figure 10. The relationship between fork length and otolith length for male tagged and free-
living mature fish. . ..........................................................................................................35
Figure 11. The relationship between fork length and otolith length for female tagged and free-
living mature fish. . ..........................................................................................................36
Figure 12. The relationship between fork length and age for male tagged and free-living mature
fish. . .................................................................................................................................40
Figure 13. The relationship between fork length and age for female tagged and free-living mature
fish. . .................................................................................................................................41
x
List of Appendices
Figure S1. Distribution of gonadal somatic index for all individuals caught at varying times
between May and October (1994-2011, inclusive) that had their sex, gonad weight, and
fish weight recorded. . ......................................................................................................62
Figure S2. The proportion of the creel lake trout population that were mature between 1994 and
2011, inclusive. . ..............................................................................................................63
Figure S3. The relationship between fork length and otolith length for free-living mature males
and females. . ...................................................................................................................64
Figure S4. The relationship between fork length and age for free-living mature males and
females. . ........................................................................................................................65
1
1 Introduction
Otoliths are one of the few non-skeletal calcified structures that have been shown to continually
grow throughout the life of a fish (Campana and Thorrold 2001). In teleosts, otoliths are
generally the first calcified structure to form (Campana and Neilson 1985). Therefore a large
amount of information can be recovered about fish from their otoliths such as: ages, growth
patterns, habitat usage, and movement patterns. Such information can be used to address
questions about life histories at the species and/or population level (Shuter et al. 1998; Begg et
al. 2005; Elsdon et al. 2008). Fisheries scientists suggest otoliths are one of the most useful
biological structures, especially for shedding light on management issues, because sampling
otoliths can provide much useful information on wild populations without the need to closely
monitor them (Gauldie and Nelson 1990; Dwyer et al. 2003; Begg et al. 2005). In addition to
fisheries management, otoliths are used in neuroscience to better understand hearing and
orientation in fish (Popper et al. 2005). They are also used in paleontology to determine the
species of fish because otoliths are often the only structure found that can be used to identify
species (Nolf 1985).
Otoliths have been shown to reflect the environment a fish resides in, more than its genetic
makeup (Dunlop et al. 2007; McDermid et al. 2007; Halden and Friedrich 2008). This implies
that otolith shape, size, and patterning are species and population specific (Nolf 1985; Secor and
Dean 1989; Popper et al. 2005). Otolith shape alone can reflect the environment where the fish
resides (Campana and Casselman 1993; Brazner et al. 2004; Campana 2005), and the spawning
group to which the fish belongs (Campana and Casselman 1993; Galley et al. 2006). The
interaction between organic and inorganic components of otoliths determines their shape
(Gauldie and Nelson 1990).
1.1 Current research
Recent studies have emphasized the study of fish life-history characteristics through otolith
research (Begg et al. 2005; Elsdon et al. 2008). Life history characteristics can be discerned by
investigating the chemical composition of fish otoliths to show their habitat and/or diet (Radtke
et al. 1985; Campana 1999; Begg et al. 2005; Elsdon et al. 2008). The external environment and
the physiological state of a fish at otolith formation has been shown to be reflected in their otolith
2
composition (Kalish 1989). Currently, Strontium-Calcium ratios are used to infer the
environmental composition of fish habitat (Kalish 1989; Campana 1999, 2005), which can be
used to determine the stock identity of fish (Campana and Casselman 1993; Campana and
Thorrold 2001), their migratory patterns (Munro et al. 2006; Halden and Friedrich 2008), and
their spawning grounds (Brazner et al. 2004). A solid understanding of the environmental
conditions and possible food sources is required to interpret the results of chemical analyses
(Campana 1999; Elsdon et al. 2008).
Fisheries, both marine and freshwater, require information on the age, growth, and maturity of
their stocks in order to ensure that those stocks can persist while supporting sustainable harvests
(Six and Horton 1977; Francis and Campana 2004; Conover et al. 2005). Marine fisheries have
been more frequently researched to provide stock assessment information, and possibly, to
indirectly infer historical temperature and environmental data (Campana 1999; Campana and
Thorrold 2001; Francis and Campana 2004; Begg et al. 2005; Halden and Friedrich 2008).
Fisheries management uses fish age information to determine the safe and allowable yield for the
fishery (Mugiya 1966; Gauldie and Nelson 1990). Otoliths are currently the main structure used
for age determination in fisheries management due to their accuracy in long-lived fish
(Casselman 1990). Not only are otoliths useful for estimating age, but they also store
information on the condition of the fishery; therefore analysis of the otolith chemical
composition could provide information on the possible presence of pollutants (Halden and
Friedrich 2008). The proper identification of age rings (annuli) from other markings (checks) on
otoliths is essential for proper age determination (Dwyer et al. 2003). Juveniles have also been
emphasized in research (mainly in laboratory settings) because they show the best representation
of daily otolith markings (Meekan et al. 1998; Begg et al. 2005).
Back-calculation utilizes the relationship between the length or age of an aging structure (scales
or otoliths) and fish length to estimate previous lengths of fish (Casselman 1990). This is useful
to determine the growth of fish in the fishery over time or to determine the approximate average
size of fish in each age cohort (Mosegaard et al. 1988). Also, back-calculation can be used to
determine size selective mortality for different life stages (Campana 2005). This information can
be used in new management strategies to ensure the population is exploited sustainably (Gauldie
and Nelson 1990; Dwyer et al. 2003). Back-calculation validation requires knowing the previous
and present lengths of a fish and lengths or ages of an aging structure (Casselman 1990; Neilson
3
and Campana 2008); to ensure proper later identification, physical or chemical tags are often
applied to fish (Mosegaard et al. 1988; Campana 2005). The length of the aging structure is also
used to compare the growths between subsequent catch years (Mosegaard et al. 1988; Klumb et
al. 2001). Back-calculation methods have mainly been used with scales and otoliths although
models based on scale growth are usually more accurate than those based on otolith growth
(Klumb et al. 2001). Bias in back-calculated lengths based on otolith size should be expected
(Campana 1990; Klumb et al. 2001) since otolith growth is less related to the somatic growth rate
than scale growth (Campana 1990; Casselman 1990).
1.2 Why otoliths are a good aging structure
Fish growth and small changes in growth are appropriately reflected by the relative size of
calcified structures (Casselman 1990). Structures that have been used for aging are the opercle,
otolith, scale, vertebra, and anal fin pterygiophore (Six and Horton 1977). Otoliths and scales are
mainly used for age and growth analysis (Campana and Neilson 1985; Campana 1990). Scale
size and body size have been shown to be typically highly correlated (Campana and Neilson
1985; Klumb et al. 2001). Conversely, otolith size is not as highly correlated with body size
because otoliths continue to grow during periods of stress (temperature stress, low food
availability, etc.) when scale and somatic growth cease (Campana and Neilson 1985; Reznick et
al. 1989).
Although scales reflect somatic growth more accurately, there are important difficulties that
make their use problematic: they are easily lost, suffer from abrasion, can be resorbed if the fish
is stressed enough to affect somatic structure growth (Campana and Neilson 1985; Campana and
Thorrold 2001), and typically cease to exhibit clearly delineated age markers in later life, when
somatic growth slows or ceases (Casselman 1990). The ages determined from otoliths are
generally more consistent than those of other aging structures (Six and Horton 1977; Sharp and
Bernard 1988). Additionally, otolith length has been shown to be more correlated with fish age
than with body size (Boehlert 1985). In old or slow growing fish, otoliths grow more quickly
compared to other structures and continue to record seasonal growth and age (Casselman 1990);
this is shown by slower growing fish who often have proportionately larger otoliths (Reznick et
al. 1989; Lychakov and Rebane 2000). Additionally, otoliths can be used for fish that do not
have scales or where scale growth is irregular (Campana and Neilson 1985).
4
1.3 Otolith structure and function
Otoliths are crystalline structures composed mainly of calcium carbonate (Gauldie and Nelson
1990). Otolith crystals are usually in the aragonite form (Degens et al. 1969; Gauldie and Nelson
1990) most likely because of the high concentration of magnesium in the endolymph (the sac
surrounding the otolith) (Mugiya 1966). Otoliths have a central role in hearing, balance, and
orientation (Degens et al. 1969; Campana and Neilson 1985; Schirripa and Goodyear 1997;
Campana 2005; Popper et al. 2005; Munro et al. 2006). Most fish have three otolith pairs;
sagittal, lapillar, and asteriscal that are approximately three times more dense than the rest of the
body (Rogers et al. 1988; Popper and Lu 2000). These pairs reside in otolithic end organs
(endolymphatic sacs) called sacculus, utriculus, and lagena, respectively (Nolf 1985; Lychakov
and Rebane 2000), and are located lateral to or below the hind brain in the cranial cavity (Popper
et al. 2005). The sagittal and asterstical otoliths function mainly in sound reception and
production whereas the lapillar otoliths and the semicircular canals function mainly in balance
and orientation (Nolf 1985). Fish that are more specialized in sound production/reception are
likely to have larger specialized sagittal otoliths (Nolf 1985; Popper and Lu 2000; Lychakov and
Rebane 2000).
The left and right otolith pairs of most fish species have similar growth patterns (Klumb et al.
2001; Mc Dougall 2004; Galley et al. 2006), with the exception of flat fishes (Nolf 1985;
Armsworthy and Campana 2010). The sagittal otolith is the largest of the three in most fish
species and most commonly used for analysis (Nolf 1985); if it is not the largest (such as in
Cyprinids), the larger of the lapillar or asteriscal otoliths are used (Campana and Neilson 1985;
Brown et al. 2004). Sagittae have also been shown to be the most closely aligned with fish
growth and to be the most sensitive to sound of the three otolith pairs (Lychakov and Rebane
2000; Popper et al. 2005). The saccular epithelium consists of three epithelia: sensory,
transitional, and squamous, which contribute to the sensitivity of sagittae (Parmentier et al.
2007). The most sensitive of these layers is the sensory epithelia which consists of high prismatic
hair cells, many blood vessels and nerve endings (Parmentier et al. 2007). The sensory
epithelium is located in the sulcus acusticus (referred to as the sulcus) of the otolith (Pannella
1980; Popper and Lu 2000). Because the overall density of a fish is approximately equal to the
surrounding water, the denser otoliths move in response to sounds and bend the hair cells in the
5
sensory epithelia allowing the fish to hear (Popper and Lu 2000). The term ‘otolith’ in this study
will refer only to the sagittal otolith unless otherwise specified.
1.4 Otolith growth
Calcium and other minerals are continually deposited on otoliths via food consumption and the
environment (Radtke et al. 1985; Schirripa and Goodyear 1997; Campana 1999; Brown et al.
2004; Brazner et al. 2004; Munro et al. 2006). Secor and Dean (1989) have suggested that the
accretion of calcium carbonate and organic compounds on otoliths continues throughout the
lifetime of a fish due to an endogenous rhythm. Additionally, fish age and the time of year could
be used to explain the varying amount of accretion throughout each year (Secor and Dean 1989;
Hoff and Fuiman 1993). Because otoliths are acellular and metabolically inert (Campana and
Neilson 1985; Brazner et al. 2004), unlike other bones of fish which are composed of apatite
crystalline material that is continuously resorbed and remineralized (Gauldie and Nelson 1990),
the accreted otolith materials remain where they were deposited for the duration of the fish’s life.
Additionally, Campana (1983) showed that in stressful situations fish will not resorb otolith
calcium. Although otoliths continually grow, the accretion of compounds is correlated with food
consumption and metabolic activity (Pannella 1980; Gauldie and Nelson 1990), therefore less is
deposited during fasting periods; this creates daily and annular rings and checks (Pannella 1980;
Gauldie 1988; Payan et al. 2004).
The endolymphatic fluids that bathe the otoliths are alkaline bicarbonate buffers; if the buffer
was a different composition, accretion on otoliths would not occur (Gauldie and Nelson 1990;
Campana 1999). This fluid is mainly composed of potassium, sodium, strontium, calcium, and
carbonate ions, and proteins (otolin, collagen, and proteoglycan) whose concentration influences
the amount of deposition of organic or inorganic materials onto the otolith (Mugiya 1966;
Degens et al. 1969; Kalish 1989; Gauldie and Nelson 1990; Borelli et al. 2003). The organic
layer (or organic matrix) of the otolith begins deposition at the start of evening when the
concentration of proteins is at its maximum in the endolymph and continues through the night,
whereas the inorganic layer begins deposition in the morning when the concentrations of calcium
and carbonate ions are at their maximum and continues through the day (Borelli et al. 2003). The
organic matrix is composed of a complex network of macromolecules that have calcium binding
capacity and can express anticalcifying activity which could affect the amount of calcium
6
carbonate accreted (Mugiya 1966, 1986; Borelli et al. 2003). The protein in the organic matrix is
characterized by acidic amino acids in high abundance (Degens et al. 1969). In otolith-forming
cells, calmodulin has been suggested to affect the amount of accretion by mediating calcium
function (Mugiya 1986). The inorganic layer is composed of calcium carbonate which can only
be formed when its constituent ions have reached saturation in the endolymph and then
precipitate out of the endolymph solution and accrete to the protein matrix (Gauldie and Nelson
1990; Borelli et al. 2003). The daily formation of inorganic and organic layers has been
demonstrated in many fish species (Campana 1984; Meekan et al. 1998; Campana and Thorrold
2001), but not all fish species form these layers in the same temporal pattern. In Carapus
boraborensis, a fish that lives within other organisms, the organic layer develops during the day
and the inorganic layer develops during the night (Parmentier et al. 2007).
1.5 Otolith markings
During the growth of fish, the protein matrix (organic) and the calcium carbonate crystals
(inorganic) of the otolith interact to form structures that are used for age identification (Mugiya
1966; Campana 1983; Gauldie and Nelson 1990). The structures found in otoliths are opaque
zones, hyaline zones, check rings, annular rings and daily growth rings (Gauldie and Nelson
1990). Opaque zones identify periods of fish growth and are often associated with the summer;
hyaline zones identify periods when growth has ceased and are often associated with the winter
(Victor and Brothers 1982; Gauldie and Nelson 1990). Hyaline zones, generally referred to as
checks, are used for aging and noting stressful situations (Campana 1983; Payan et al. 2004).
Checks and daily growth rings are more prominent in juvenile fish otoliths. Checks indicate areas
of halted growth due to stressful life events (identified by incomplete rings) (Pannella 1980;
Campana 1983; Gauldie 1988; Walker and McCormick 2004). Checks have been shown to form
after the stressful event in food deprivation experiments (Campana 1983). Daily growth rings are
viewed with higher magnification microscopy and indicate the age in days of a juvenile fish
(Campana and Neilson 1985; Gauldie and Nelson 1990).
During stressful situations, calcium uptake is limited, the proportion of calcium carbonate to
protein accretion is altered; there is a noticeable reduction in calcium deposition when checks are
formed (Campana 1983; Payan et al. 2004). Therefore the hyaline zones, that form ‘checks’, are
mainly composed of protein and are visually distinct from the opaque zones which are mainly
7
composed of calcium carbonate (Campana 1983; Gauldie 1988; Payan et al. 2004). Protein is
continually incorporated into the otolith, even during stressful events, but etched away by acid
during these stressful events; the degree of stress is reflected in the size of the check mark
(Campana 1983). Also, fish that experience seasonality or migration will have more pronounced
annuli due to varying environmental conditions (Campana and Neilson 1985; Brown et al. 2004;
Begg et al. 2005). Reproduction is also considered a stressful life stage because energy is
allocated to gonad development rather than growth (Lester et al. 2004; Rideout et al. 2005).
Pannella (1974, 1980) suggested that reproduction (or spawning) checks can be viewed on
otoliths, but has not validated this claim.
1.6 Skipped spawning
Individuals that are mature and do not reproduce during a spawning season are considered to
have avoided that spawning season. This phenomenon is referred to as ‘skipped spawning’
(Rideout et al. 2005). Skipped spawning has often been considered to be due to a lack of ideal
environmental conditions and food resources and an anomaly in iteroparous fishes (Engelhard
and Heino 2005). In the past decade, however, there has been growing interest in the hypothesis
that skipping spawning is a life-history strategy for increasing investment in future reproduction
(Rideout and Tomkiewicz 2011). Newly matured individuals have been noted to skip spawning
more than older individuals to invest in growth rather than reproduction (Engelhard and Heino
2005; Jørgensen et al. 2006). This phenomenon has mainly been identified in species that
migrate for reproduction (Rideout et al. 2005; Loher and Seitz 2008). Tags can be used to
indicate when fish migrate to the spawning area to determine which fish use energy stores for
migration and reproduction versus growth (Warner 1971; Rideout et al. 2005; Loher and Seitz
2008). Identifying if fish have spawned in any given season is difficult especially when the fish
are not under direct observation (Rideout et al. 2005).
Methods of identifying reproductive status are more extensively studied in females than in males
(Rideout et al. 2005), and are based on liver mass (Rideout et al. 2000; Loher and Seitz 2008),
fish condition (Loher and Seitz 2008), scale interpretation (Engelhard and Heino 2005), and
gonadal observation (Rideout et al. 2000; Loher and Seitz 2008). Scale interpretation uses both
the width between annuli (Engelhard and Heino 2005) and the presence of spawning checks
8
(Rideout et al. 2005; Rideout and Tomkiewicz 2011). To my knowledge, the use of otoliths to
identify the years a fish has spawned has yet to be conducted.
1.7 Aging
The active parts of age and growth studies are the identification of otolith markings and aging
structures (Campana and Neilson 1985; Begg et al. 2005) and the validation of aging structures
(Campana 2001, 2005; Brown et al. 2004; Begg et al. 2005). For example, otoliths have been
used to estimate (i) age (Six and Horton 1977; Campana and Moksness 1991; Oxenford et al.
1994), (ii) age at maturity (Campana and Thorrold 2001; McDermid et al. 2007), (iii) growth rate
(Kozłowski 1996; Schirripa and Goodyear 1997; Dunlop et al. 2005), (iv) yearly (Campana and
Thorrold 2001) and (v) daily growth (Campana 1984; Begg et al. 2005). The identification of
otolith annuli is usually not well described for a number of species utilised in aging studies. It
has been suggested, but poorly executed that an aging structure example for each species should
be put in a database for reference (Beamish and McFarlane 1987). This poor execution leads to
aging difficulty encountered by novice agers, and also hinders experienced agers when studying
different fish species (Boehlert 1985; Campana and Moksness 1991).
1.8 Fish growth
Fish growth has been shown to reflect an approximate linear relationship between length and age
prior to maturity, and a von Bertalanffy growth curve post maturity (Quince et al. 2008a). This
pattern has been recorded in a variety of fish species (Allen 1966; Kozłowski 1996; Lester et al.
2004; Shuter et al. 2005; Quince et al. 2008b). Prior to maturity most of a fish’s energy is
allocated to growth, to ensure becoming bigger faster, which is an advantage when evading
predators (Kozłowski 1996; Lester et al. 2004; Arnott et al. 2006). Once maturity is reached,
energy allocated to reproduction will increase progressively with age (Lester et al. 2004; Quince
et al. 2008b). The relative partitioning of energy between reproduction and growth will differ
between the sexes, the specific differences will vary between species depending on their
reproductive patterns and styles (Charnov et al. 2001; Lester et al. 2004).
1.9 Measuring technique
Generally, it is difficult to measure otoliths because they are curved and the curve is fairly
species and/or population specific (Nolf 1985; Secor and Dean 1989; Popper and Fay 2011).
9
Additionally, the otolith-measuring community uses at least two different methods for
measuring, which can create problems when trying to compare the growth of fish inferred from
newly measured otoliths to previously measured ones within the same species and/or population.
Otoliths are generally measured on the dorsal side, from nucleus (the point where a fish began to
grow) to annulus/leading edge (Boehlert 1985; Dwyer et al. 2003). One of the current methods
for measuring otoliths is using a straight line (Boehlert 1985; Wright et al. 1990; Secor and Dean
1992), which is inconsistent because the measurement to each annulus will not necessarily be on
the furthest point of growth. The other common method is using line segments (usually used by
fisheries technicians) that measure from nucleus to each subsequent annulus at their furthest
point of growth, this method creates a segmented curve. Although this method can account for
the varying curvature of otoliths, additive mistakes can be encountered if annuli are
misidentified.
The starting point of otolith measurements is usually the nucleus (Pannella 1980). Nuclei are
sometimes hard to identify, especially for a novice ager (Pannella 1980). Additionally, it is
possible the nucleus is not visible even to an expert if the otolith section preparation was not
done well (Pannella 1980; Niva et al. 2005). The novel measuring technique, introduced in this
study, negates the need to find this point and focuses on approximating it. There are few studies
that actually describe, in enough detail to replicate, the manner in which annular measurements
were obtained (Boehlert 1985; Secor and Dean 1992). Many studies specify using ‘otolith
radius’ or ‘otolith diameter’ as a measurement, however they do not specify the exact method in
which they employ these measurements (Victor and Brothers 1982; Wright et al. 1990; Schirripa
and Goodyear 1997). Additionally, there is little agreement, in the fish-aging community, of how
to interpret age structures (Six and Horton 1977; Gauldie and Nelson 1990).
Not only is the means of measuring not well described, it is also easier for seasoned agers to
measure and age otoliths (Boehlert 1985). Those new to aging have to spend time learning the
location of the nucleus and annuli, and where on these structures measurements should be taken
(Boehlert 1985; Campana and Moksness 1991). This lengthy process could be accelerated by
implementing an otolith measuring protocol that uses landmarking techniques to identify both a
consistent index location for the nucleus and consistent locations on each annulus to which
annular measurements are to be made. Therefore, a new method for measuring otoliths that
approximates the position of the nucleus instead of estimating it is proposed. The current method
10
of estimating the nucleus involves placing a point as the starting measurement (Boehlert 1985),
and if that point is not precisely on the nucleus then measurement error is encountered. The new
method will use a line to approximate the nucleus instead of a point like that of the current
methods. As the nucleus is approximated with a line, the current measuring techniques cannot be
used (which both identify the nucleus with a point); therefore otoliths will be measured from
nucleus to edge with parallel lines in sequence. Additionally, the current methods assume in each
case that the nucleus is readily visible; the new method bypasses this assumption by using
landmarks that are always present on otolith sections.
1.10 Tagging uses
Laboratory work with otoliths emphasizes the different ways to identify fish that are less
invasive than physical tagging. For example, in fisheries supported by stocking, temperature
(Bergstedt et al. 1989) and/or chemical ‘tags’ (Klumb et al. 2001; Campana 2005) can be
impressed on the otoliths of young fish during their time in the stocking hatchery. In both of
these methods, young fish are subjected to varying environmental conditions (enough to affect
otolith patterning) in the hatchery and then are released back into the fishery (Bergstedt et al.
1989; Mc Dougall 2004). The main reason for tagging fish is for later identification within a
stock (i.e. to see which fish survived to adulthood after being released from the hatchery) (Otterå
et al. 1998) or for back-calculation of growth over the interval from hatchery release to later
recapture (Klumb et al. 2001).
An application of appropriate age identification and measuring techniques is back-calculation
using tag-recapture information (Klumb et al. 2001). Tag-recapture studies have been effective in
investigating the growth of individual fish (Campana 2005). There are numerous ways in which
fish have been tagged; temperature or chemical markers (as previously mentioned), fin clips,
passive integrative transponder (PIT) tags, and anchor tags (Carline and Bynildson 1972;
Tranquilli and Childers 1982; McAllister et al. 1992). In each of these cases, fish have to be
handled and this handling can affect their growth (Bergstedt et al. 1989). The latter three
techniques involve piercing or cutting the integument, which may start an infection, or stress the
fish while it heals (McAllister et al. 1992). This effect is difficult to discern from the decoupling
of the otolith-fish length relationship (Klumb et al. 2001).
11
Many studies have noted this decoupling, showing that lengths back-calculated using otoliths are
usually underestimated (Campana 1990; Campana and Moksness 1991; Dwyer et al. 2003) and
have speculated that tagging may be a possible reason for the discrepancy (Klumb et al. 2001).
Laboratory experiments have shown that decoupling of the otolith-fish length relationship can be
counteracted with adequate temperatures and food availability not found naturally (Mosegaard et
al. 1988). Additionally, back-calculation is more accurate for scales than otoliths (Klumb et al.
2001): this could be the result of decoupling, or it could arise from basing back-calculated sizes
on a relationship (i.e. linear) between body size and otolith size that does not reflect the ‘true’
relationship (Campana 1990).
1.11 Study species
Lake trout, Salvelinus namaycush, are long-lived, indeterminate growing, cold-water fish that are
broadcast spawners and reproduce in October (McDermid et al. 2007; Morbey et al. 2010). They
are usually top predators in their respective lake ecosystems and are found in oligotrophic lakes
in northern America (Shuter et al. 1998). The lake trout for this study were caught by anglers in
Lake Opeongo in Algonquin Provincial Park, Ontario. This population is piscivorous, mainly
consuming lake herring (Coregonus artedii) – a species that was introduced in 1948 to improve
the quality of fish produced by the fishery (Kerr 1971; Martin and Fry 1972). Individuals from
this population exhibit higher condition values and higher individual growth rates than those
from a nearby lake with a planktonic feeding S. namaycush population (Morbey et al. 2010).
Algonquin Park fish experience significant seasonal variation in the conditions that support
individual growth – they pause growth during the winter and resume growth in the spring, and
this is clearly shown on their otoliths (Morbey et al. 2010). Mature lake trout show a decrease in
somatic growth rate in early to late summer, when they allocate much of their energy to produce
gonads (McDermid et al. 2007; Quince et al. 2008b; Morbey et al. 2010). Considering lake trout
are broadcast spawners, then any energy put into reproduction will be used for producing and
storing gonads and not parental care (Rennie et al. 2008).
Lake trout are often assumed to reproduce every year after they mature, which is around age
eight for both sexes in Lake Opeongo (McDermid et al. 2007). Although lake trout may invest in
reproducing every year, they could potentially resorb their gonads with little energy loss
(Morbey et al. 2010). Some studies have speculated that noted checks on otoliths are ‘spawning’
12
checks (Pannella 1974, 1980; Gauldie 1988), although such markings have not been
systematically validated. Since checks have been noted on many wild fish species it is imperative
that researchers understand how and why they occur (Campana 1983). This study attempts to
validate a method for identifying ‘spawning’ checks on otoliths and assesses the presence of
‘spawning’ checks that would serve as an indicator of spawning frequency after maturity. If such
marks are present, it would make identification of maturity possible without assessing the
gonads.
1.12 Study objectives
This study will focus on the gaps in current otolith methodology; focusing on otolith
measurement, annuli description and identification, and how these can be applied to studying
somatic growth patterns. The objectives of this study were as follows:
(i) To develop a new protocol for otolith measurement that could be consistently applied
without requiring extensive training and direct experience;
(ii) To compare results from this new measurement protocol with results from methods
currently in use;
(iii) To identify and describe markings on otoliths that indicate the first year and every
subsequent year fish have spawned and how they differ from annuli;
(iv) Employing (i) and (iii) to compare the growth of free-living fish to those that were tagged
(with an anchor tag) and released; and,
(v) To compare males and females for (i), (iii), and (iv) to assess differences between sexes
in growth and reproduction.
13
2 Methods
2.1 Collection
All of the otoliths for this study came from S. namaycush caught by anglers in Lake Opeongo,
Algonquin Provincial Park, Ontario, Canada. Lake Opeongo is 51.8 m at its deepest (mean depth
of 14.8 m) and has a surface area of 58.6 km2. It is the largest lake in the park (King et al. 1999).
Once a fish is caught in this lake by an angler its information is recorded (length, weight,
stomach contents, sex, etc.) and/or collected (otoliths and scales) by trained technicians, then the
fish is returned to the angler for consumption. The fishing season for S. namaycush is from the
end of April to the end of September. This lake was chosen for this study because of the data on
lake trout (tagged and free-living) available from the Harkness Laboratory of Fisheries Research.
Additionally, the only limits on catching fish for anglers are the amount of fish (maximum of two
per day), there are no size limits, and therefore any size fish can be caught and sampled.
2.2 Otolith extraction
Typically fish were euthanized by the angler before they arrived at the angler monitoring station.
Fish that were still alive, were euthanized by applying a sharp blow to the head with a club. For
each fish, length and weight were recorded and then the fish was held ventral side up and an
incision was made laterally through the isthmus, cutting upward through the gill arches. The gills
were removed by twisting and pulling to expose the roof of the mouth. A bone cutter was used to
cut into the parashenoid bone to remove the sagittal otoliths. The endolymphatic sacs containing
the otoliths were removed with tweezers. Once both sacs were removed, the sacs were broken,
and the otoliths were allowed to dry. The otoliths were then stored dry in a vial until sectioning.
Once the rest of the fish was cleaned (stomach contents examined, scales collected, liver
weighed, gonads weighed and examined), it was returned to the angler for consumption.
2.3 Otolith preparation
The fish chosen for otolith sectioning by Harkness technicians were randomly chosen within the
catch season based on their sample number (assigned to each group of anglers interviewed
during that year). The collection technicians attempted to remove both sagittal otoliths and store
them both in a 5 mL dry vial prior to sectioning. In the case where both otoliths were collected,
14
the right otolith was sectioned and the left one was stored in a dry vial. When only one otolith
was gathered from the fish, it was sectioned.
The otolith was removed from the vial and examined under a dissecting microscope at 40X
magnification to find the nucleus which was identified with a marker. The otolith was placed in a
well, filled with epoxy resin (type: mirror coat; manufacturer: system 3) making sure the otolith
was totally encased. The epoxy resin was allowed to harden and then it was removed from the
well. The dried resin was placed in the chuck arm of an Isomet saw and aligned in such a way to
ensure that the saw blades would bisect the nucleus of the otolith to create a transverse section.
The section was cut with the diamond grit blades positioned 400 μm apart. After the section was
completed, it was removed from between the saw blades and rinsed in isopropyl alcohol to
remove any excess saw dust.
The section was sanded using fine grit sand paper (1000-1600 grit) to remove excess saw cuts.
After sanding, the section was rinsed with distilled water and placed sanded side down on a glass
slide. A drop of epoxy resin was placed beside the section and the sanded side of the section was
pressed down in the middle of the epoxy pool until totally encased and was allowed to dry. The
dried otolith slide was wet sanded by hand or with a sanding disk using 1200-1600 grit sand
paper. During the sanding process the otolith was examined under a dissection microscope to
determine whether the edges of the otolith were exposed and also to note the appearance of the
nucleus and/or the first annulus. Continual sanding and checking occurred until both the nucleus
and edge were exposed. Once these structures were exposed, the otolith surface was rinsed with
distilled water. Well-worn 1600 grit wet sand paper was used to remove any final scratches and
to make the surface of the otolith mirror-like.
As these fish are long-lived, it is difficult to identify annuli near the edge using whole otolith
sections, especially for older fish (Boehlert 1985), therefore acetate replicates were made for
examination under a compound microscope. Additionally, acetate replicates allow for the
inspection of ‘spawning’ checks not usually visible on whole otoliths (Pannella 1971; Campana
1983; Gauldie 1988). Acid etching was performed on the whole otolith section using 1.5%
hydrochloric acid for 10-30 sec, depending on the size of the otolith. A drop of acetone was
placed on the etched surface followed by an acetate slip which was pressed down on the acetone
ensuring all air bubbles were pressed out. The acetate slip was allowed to dry for 3-5 min and
15
then was removed from the surface of the otolith with tweezers. The acetate slip was placed on a
slide etched-side up and then sandwiched with the slide containing the whole otolith section.
Both sides were then taped together. The acetate replicate of the otolith was examined for this
study and will be referred to as an otolith section. Otoliths have only been collected from 1990
onward in this lake, and the usage of acetate replicates has been employed from 1994 onward,
therefore the data that involves otolith sections comes from fish caught between 1994 and 2011,
inclusively.
2.4 Imaging
An image was taken of each otolith section using the program Leica Application Suite V4.3 (©
2013 Leica Microsystems (Switzerland) Limited). The images were taken using a 10X objective
lens and a 6.3X magnified photo-tube, mounted on a compound microscope. A 2 mm slide
micrometer was used to calibrate the camera to the microscope to ensure all images had accurate
measurements. Each otolith section was viewed with transmitted light, and the camera effects
were adjusted to ensure easy annuli identification. Because otolith sections were larger than the
field of view of the camera, each otolith section image was made up of many images stitched
together. Images of otoliths are provided to start the trend of providing ageing structure examples
for species.
2.5 Measuring techniques
The three measuring techniques (parallel lines (novel), straight line and line segments (current))
were compared using 50 female lake trout otolith sections. Each section was measured with each
technique without knowledge of the otolith length for the other two techniques. Due to the way
otoliths rest on the otic cleft, the ventral side experiences restricted growth (Nolf 1985; Gauldie
and Nelson 1990). Therefore the dorsal side of all otoliths was measured.
The otoliths collected for this part of the study were from fish caught between 2000 and 2005,
inclusively, whose lengths were evenly distributed through that time period. Additionally, the
novel method was evaluated on male lake trout (n=25) caught during the same time period and
also evenly distributed by fork length. Fork length is considered to be the measure from the tip of
the snout to the centre of the fork in the tail. Both mature and immature fish were used to assess
the utility of the novel measuring technique. For all measurements in this study, the place where
16
the measurement was taken, on the edge or on the annuli, was to the furthest point of growth
(largest distance from nucleus to structure) at the beginning of the dark zone. For this
comparison only measurements from the nucleus to the edge were used to assess the utility of the
approximation method. All otolith section measurements were taken in millimetres, to three
decimal places.
For the estimation (current) methods, the location of the nucleus was identified with a point. The
point was drawn by viewing the sulcus as an arrow pointing to the nucleus (Fig. 1 and 2). The
otolith texture in the ‘pointed-to’ region was visually examined to locate the nucleus, looking for
a circular pattern or where the calcium carbonate crystals radiate from a point that would indicate
where the otolith began to grow. The same nucleus identification point was used for both
estimation methods. The straight line method involved drawing a straight line from the estimated
nucleus to the edge (Fig. 1), with no consistency of line angle for each otolith section. The line
segment method involved drawing a segmented line from nucleus to edge using each annulus as
a vertex (Fig. 2). The measurement was the length of the segmented line.
The approximation (novel) method, introduced in this study, involves using parallel lines in
sequence: the distances measured were the distances from a baseline (starting line) to one or
more structures using parallel lines. The baseline was drawn in such a way that it would
theoretically pass through the nucleus, thus approximating the nucleus, by using the sulcus as a
landmark (Fig. 3). On the images it is possible to zoom in to see the precise point where the
sulcus begins (Fig. 4), that is where the first point of the baseline is placed. The second point is
placed so as to divide the ventral and dorsal lobes of the first annulus (or the second if the first
cannot be identified on both the dorsal and ventral sides) through the sulcus. This division is
made by visually approximating a perpendicular angle between the baseline and the anti-sulcul
(opposite of the sulcus) side while maintaining an approximately even division of the sulcus.
After the baseline is created, the program automatically produces another line parallel to the first
which can then be placed on the subsequent annuli or edge. For this section, the measurement
was the perpendicular distance between the baseline and the line identifying the edge.
Perez and Munch (2013) compared the accuracy of linear and allometric relationships for otolith
length-fish length relationships and concluded that they did not differ appreciably. They decided
to use the linear model because it had a lower Akaike information criterion value.
17
Fig
ure
1. S
trai
ght
line
mea
suri
ng t
echniq
ue.
The
das
hed
lin
es d
enote
the
top o
f th
e su
lcus
(see
Fig
. 4
) to
be
use
d a
s an
arr
ow
hea
d t
hat
poin
ts t
o t
he
nucl
eus.
The
nu
cleu
s is
iden
tifi
ed b
y
a ci
rcula
r pat
tern
or
wher
e th
e ca
lciu
m c
arbon
ate
nee
dle
s ra
dia
te f
rom
a p
oin
t. T
he
assu
med
nuce
lus
is i
ndic
ated
by t
he
circ
le. T
he
mea
sure
men
t use
d i
s th
e le
ngth
of
the
stra
ight
line.
18
Fig
ure
2.
Lin
e se
gm
ents
mea
suri
ng t
echniq
ue.
Th
e das
hed
lin
es d
enote
the
top o
f th
e su
lcus
(see
Fig
. 4
) to
be
use
d a
s an
arr
ow
that
poin
ts t
o t
he
nucl
eus.
The
nucl
eus
is i
den
tifi
ed b
y a
circ
ula
r p
atte
rn o
r w
her
e th
e ca
lciu
m c
arbonat
e n
eedle
s ra
dia
te f
rom
a p
oin
t. T
he
assu
med
nuce
lus
is i
ndic
ated
by t
he
circ
le. T
he
nu
cleu
s an
d s
ulc
us
are
the
sam
e as
in F
ig.
1. T
he
segm
ente
d l
ine
is d
raw
n f
rom
the
nucl
eus
to e
dge
usi
ng t
he
furt
hes
t poin
t of
gro
wth
of
each
annuli
as
ver
tice
s. T
he
mea
sure
men
t use
d i
s th
e le
ngth
of
the
segm
ente
d
line.
19
Fig
ure
3.
Par
alle
l li
nes
(n
ovel
) m
easu
rin
g t
echniq
ue.
The
das
hed
shap
e d
enote
s th
e dors
al (
right)
and v
entr
al (
left
) lo
bes
of
the
firs
t an
nulu
s. T
he
circ
le i
ndic
ates
th
e to
pm
ost
par
t o
f th
e
sulc
us
(see
Fig
. 4
), w
her
e th
e st
arti
ng p
oin
t of
the
bas
elin
e w
as d
raw
n. T
he
seco
nd p
oin
t
of
the
bas
elin
e is
dra
wn i
n s
uch
a w
ay t
hat
div
ides
the
ven
tral
and d
ors
al l
obes
of
the
firs
t
annulu
s th
rough t
he
indic
ated
sulc
us.
Div
idin
g t
he
sulc
us
involv
es v
isual
ly
appro
xim
atin
g a
bas
elin
e per
pen
dic
ula
r to
the
anti
-sulc
us
(opposi
te s
ide
of
sulc
us)
whil
st
stil
l pas
sing t
hro
ugh t
he
appro
xim
ate
centr
e of
the
tria
ngle
cre
ated
by t
he
sulc
us.
The
mea
sure
men
t use
d i
s th
e per
pen
dic
ula
r dis
tance
bet
wee
n t
he
two p
aral
lel
lin
es.
20
Figure 4. Identification of the topmost part of the sulcus for the otolith presented in Figures 1-3.
In Figures 1 and 2 this is represented by the tip of the dashed arrow head, and in Figure 3
it is indicated with a circle. Using a digital program it is possible to zoom in as far as
possible to identify the topmost part of the sulcus (approximately identified by the
arrow). Identification is done by following the triangular sulcus inward from the sulcul
side and looking at the otolith patterning until a ‘Ʌ’ shape is seen (lower right quadrant of
figure). The location of this does not have to be precise because the parallel lines method
assumes the nucleus will fall somewhere along the baseline.
21
Therefore fitted straight lines were used to represent the otolith length-fork length relationship in
this study, where otolith length was a function of fork length. Otolith length, here, was chosen as
the dependent variable because fork length measurements were assumed to have no error and the
accuracy of the three otolith measuring techniques was being tested.
2.6 Aging, annuli and checks
Otoliths were aged by counting the annuli inward from the edge. Counting annuli inward is more
effective in long-lived fish because the first few annuli are difficult to identify and the later
annuli are much easier to identify (Dwyer et al. 2003). The annuli on S. namaycush otoliths are
dark bands that represent slower winter growth and less accretion. Annuli closer to the edge are
more defined (darker and thinner) than those closer to the nucleus. Near the edge, there are thin
white checks that are noticeable right before the annuli (Fig. 5) (also seen by Gauldie (1988) on
otolith sections of the New Zealand snapper, Chrysophrys auratus); these are suspected to
indicate spawning. Counting from the outside also reminds the ager that checks are not annuli
and should not be used for aging purposes. Closer to the nucleus annuli are harder to identify
(since they are less defined), therefore the anti-sulcus was viewed to see depressions and
followed them inward to identify the annuli. The first annulus is the hardest to identify, however
it is approximately 0.5 mm toward the edge from the nucleus, therefore this was used as a
guideline for identifying it.
Annuli were identified using points, which allows the ager to click on each annulus and when all
annuli are identified the program counts the points and relays the number. Since all free-living
fish in this study (except those used for measurement comparison) were caught late in the
growing season, the number of all annuli identified is equivalent to the fish age. Fish caught in
April and sometimes May do not show the annulus closest to the edge because summer growth
has not begun, and thus that annulus is likely not identified on otolith section images.
Additionally, this method allows for less miscounting of annuli and allows a novice ager to
review their results with an experienced ager to improve their annuli identification methods.
To show the relationship between age and otolith length, exponential curves were fit to mature
fish (nmale=61, nfemale=79) and linear curves to immature fish (nmale=108, nfemale=96). The growth
curves were separated by maturity because the residual sum of squares was minimized with two
curves instead of one. This was expected due to the biphasic nature of the relationship between
22
Figure 5. The visual difference between an annulus and a ‘spawning’ check. Annuli are darker
thicker lines and ‘spawning’ checks are thinner and lighter lines. ‘Spawning’ checks are
only seen nearing the edge of a mature otolith. The age of the fish associated with the
‘spawning’ check is from the annulus before the check. The measurement from the
nucleus to the nth annulus reflects otolith size at the end of the nth year of life; it also
reflects otolith size at the beginning of the (n+1)th year of life.
Annulus
‘Spawning’ check
23
fish size and age (Quince et al. 2008b) which should be similarly reflected in the otolith size-age
relationship. The exponential curves were of the form
𝐴𝑔𝑒 = 𝑏𝑒𝑎∗𝑂𝑡 (1)
where age is a function of otolith length (𝑂𝑡) and 𝑎 and 𝑏 are constants. The linear relationship
shows age as a function of otolith length. This set of relationships can be used for easy age
identification with proper otolith measurement.
All fish were caught between 1994 and 2011, inclusively. The maturity of fish was determined
by looking at the gonadal somatic index (GSI=gonad weight/fish weight) of all fish (separated by
sex) caught between 1994 and 2011, inclusive (referred to as the ‘creel population’). The
relationship where GSI is a function of date-caught (independent of year) was visually inspected
to determine which GSI threshold values indicated maturity for each sex (Fig. S1). The
thresholds for mature females and males were determined to be >0.008 and >0.004, respectively.
Slight changes in these values did not appreciably change the proportion of mature individuals in
the population for each age group (all fish otoliths were aged by an experienced technician) (Fig.
S2). These values were used to determine the maturity status of fish used for this study caught in
August or September. There were two exceptional individuals with borderline GSI values, one
from each sex: a GSI slightly smaller than the threshold for a fish caught in early August and a
GSI slightly larger than the threshold for a fish caught in late September.
2.7 Tagged fish growth
Fish from this lake have been tagged every fall since 1975 with anchor tags (t-tags) (mainly) and
sometimes PIT tags as well. Fish are t-tagged on the left side of their body near their dorsal fin
and PIT tagged abdominally. Other than tagging, the fish were also measured (length and
sometimes weight), sexed, and then released. Each tagged fish used for this study was chosen to
meet the following criteria: (i) it was caught (and killed) by anglers sometime between 1994 and
2011, inclusively, (ii) it experienced at least five years of liberty between the date when it was
tagged and the date when it was killed by angling; this ensured there was noticeable growth
between tagging and death; (iii) its otoliths were properly extracted and suitable for sectioning,
aging, and measuring. The tagged fish (nmale=54 and nfemale=32) were measured with the novel
technique in the same fashion as mentioned above (measuring from nucleus to edge at furthest
24
point of growth), however they were also measured to the annulus that represents the year they
were initially tagged (measured to right before the dark zone). To ensure otolith measurements
were more related to both fish size and age, measurements were taken to the annuli after the age
of tagging (Fig. 5). This is because otoliths and fish grow after their ‘birthday’, so the fish would
have already grown for that given year by the time it was tagged. Lake trout were tagged after
being caught on spawning shoals during their spawning season. They were netted using four
panels of 2 inch or 2.5 inch mesh (50 feet per panel, 6 feet height). The nets were set for 10
minutes in 2-3 m water parallel to the shore on cobble sized rubble between 18:30 and 21:00.
Fish were caught by their maxillaries instead of gills to ensure they survive the netting process.
Because all tagged fish were caught during spawning season on a spawning shoal; all tagged fish
were assumed mature.
Physical tags may also affect each sex differently because stressful situations are handled
differently between sexes (Lester et al. 2004), thus the growth of males and females were kept
separate when performing analyses and fitting relationships. The relationship between otolith
length and fork length for tagged fish was compared to that of free-living mature fish (identified
by GSI), separated by sex. Both the tagged fish and the mature free-living fish were fitted to a
straight line where fork length was a function of otolith length. This form was chosen because
the objective was to generate a relationship that would permit estimating fork length from otolith
length. The relationship between age and fork length for each sex was also investigated. The
growth of free-living mature fish was compared to the growth of tagged fish using the von
Bertalanffy curve:
𝐿 = 𝐿∞(1 − 𝑒−𝐾(𝑡−𝑡0)) (2)
where fork length (𝐿) is a function of age (𝑡) and where 𝐿∞ is the asymptotic fork length, 𝐾 is the
growth constant and 𝑡0 is the age at length zero (Allen 1966).
2.8 ‘Spawning’ checks
After aging and measuring, the otoliths of tagged individuals were examined for the presence of
‘spawning’ checks. This was a blind assessment, conducted without reference to the year of
tagging, the year of capture, nor the number of years between tagging and capture. The annuli
that were accompanied by these markings were identified (Fig. 5). The age associated with each
25
‘spawning’ check was based on the annulus before the check. Checks were only identified on
otolith sections that were relatively easy to age. The prevalence of ‘spawning’ checks was
recorded for: (i) the year of tagging, (ii) two years before tagging, and (iii) two years after
tagging for analysis. Given that all tagged fish were assumed mature, a ‘spawning’ check was
expected to be observed in the year of tagging. The tagged fish used in this section were the same
fish as those used in the previous section.
Using the previously determined GSI thresholds and fish ages from the creel population, the
majority of males were found to mature by age 7 and the majority of females by age 8 (Fig. S2).
To validate the ‘spawning’ checks as a reliable indication of maturity, I assumed that most fish
had just matured if tagged on or before age 9 for females and age 8 for males. Using these
recently matured fish that had a ‘spawning’ check at tagging, I identified which fish possessed
‘spawning’ checks two years prior to being tagged. If spawning was indicated by these checks
then the proportion of fish with checks at this age should be significantly less than the proportion
identified at tagging.
I investigated the prevalence of skipped spawning, using only fish that had a ‘spawning’ check
identified on their tag year: I identified which of these fish also possessed a spawning check two
years after tagging. If skipped spawning persisted the proportion of check-identified fish should
decrease significantly from the proportion identified at tagging (which is 100%). The interval of
two years after tagging was used because the fish in this study were from a variety of cohorts
therefore any instances of biannually skipped spawning (Engelhard and Heino 2005; Jørgensen
et al. 2006) will be hidden by their various ages. Additionally, the prevalence of skipped
spawning typically decreases with age (Jørgensen et al. 2006), therefore evaluating skipped
spawning at more than two years could hide years of skipped spawning when the fish were
younger.
2.9 Statistical methods
Linear regressions were performed on all three measuring techniques and their fits were
represented by coefficients of determination (R2s). The three measuring techniques were
compared using an analysis of covariance (ANCOVA) to determine if any of the methods made
significantly different measurements (p<0.05). Any significance found by the ANCOVA was
further analysed by Tukey’s post hoc test. To justify the usage of the new measuring technique
26
on male fish of the same species, the coefficients of determination for both male and female
regression analyses were compared and an R2 of over 0.80 was considered as high enough to
justify the use of this method for both sexes. The male and female relationships were compared
with an ANCOVA. The otolith length-age relationship was also compared between the sexes.
The mature fish were compared using the method of analysis of residual sum of squares (ARSS)
presented in Chen et al. (1992) for comparing curves. The immature fish were compared using an
ANCOVA. The fits for the linear regressions were assessed with R2 values and the fits for the
exponential curves were assessed with the following quantity:
1 −𝑅𝑆𝑆
𝑇𝑆𝑆 (3)
where 𝑅𝑆𝑆 is the residual sum of squares and 𝑇𝑆𝑆 is the total sum of squares, which creates an
approximate coefficient of determination (Perez and Munch 2013).
The presence of ‘spawning’ checks was assessed as a useful indicator of maturity if (i) the
proportion of tagged fish with checks in the year of tagging was ‘high’ (>75%), and (ii) the
proportion of the tagged fish with checks two years prior to the year of tagging was reduced
by >50%. The number of fish with ‘spawning’ checks identified in the year of tagging and two
years before were compared using Fisher’s exact test (FET). FET is similar to a chi-squared test,
but it accounts for small sample sizes (Williams 1993); at least one cell in each chi-squared
contingency table had an expected value under five. This test was also used to assess the
prevalence of skipped spawning by using only those fish that were identified to have ‘spawning’
checks in their tag year. These fish, representing 100% ‘spawning’ check presence, were
compared to those same fish two years after tagging using FET. Additionally, the number of
‘spawning’ checks found for the year of tagging, two years before, and two years after were
compared between the sexes using FET.
Since only mature fish were used to compare tagged fish to free-living fish, the 𝑡0 value of the
von Bertalanffy equation would be difficult to estimate (Pardo et al. 2013), therefore the
possibility of fixing this term was assessed. First, the ARSS method was used to compare the
growth curves generated by the three-term equations for both tagged and free-living fish
(separated by sex). If significant differences were uncovered, I investigated if they were caused
by the 𝑡0 term using the comparing parameter estimates (CPE) methods of Ratkowsky (1983). If
27
differences were caused by this term, then it could not be fixed. However if the significant
differences were not caused by this term, then I felt justified in fixing it in a reasonable fashion to
obtain more robust estimates for the remaining two parameters. The free-living fish data were
used to estimate two common 𝑡0 parameter values, one for each sex, since this data range
encompassed the whole data range of the tagged fish. These 𝑡0 estimates were used in estimating
the remaining parameters of both the free-living and tagged von Bertalanffy curves.
The otolith length-fork length relationship of tagged fish and free-living fish of each sex were
compared using an ANCOVA. The ANCOVA, for this section, only focused on the significant
differences between the slopes since the intercepts are more affected by smaller fish than larger
ones. The linear regression fits were represented with R2 values. After fitting the von Bertalanffy
curves to the free-living and tagged fish data for the fork length-age relationship, the curves were
compared and the fit was assessed using the same methods as for the exponential curves.
28
3 Results
3.1 Measuring technique
The three transect measuring techniques were not found to be significantly different from each
other (ANCOVA: F2=1.209, p=0.302) (Fig. 6). The R2 values for all three relationships were
0.922, 0.918, and 0.884 for parallel lines, line segments, and straight line, respectively. Since the
parallel lines technique had the highest R2 value when measuring females, males were measured
with that technique to justify the usage of parallel lines for both sexes. Figure 7 shows male and
female otolith length-fork length relationships, which were found to differ significantly only by
their intercepts (ANCOVA: F1=5.916, p=0.0175). The female measurements used were the same
as those in Figure 6 using the parallel lines technique. The R2 for males measured with the
parallel lines technique was 0.897.
We then investigated the otolith-age relationship from a sample of fish caught between 1994 and
2011, inclusively, to show how this relationship can be used to estimate age. The residual sum of
squares was minimized when using a linear curve and an exponential curve to represent mature
fish, respectively, for both sexes (ARSSfemale: F2,171=10.919, p<0.001; ARSSmale: F3,165=13.537,
p<0.001). Using the GSI to determine maturity, the exponential curves created by the otolith-age
relationship were compared for mature fish of each sex, which were found to be not significantly
different from each other (ARSS: F2,200=0.883, p>0.05). Additionally, the immature fish that
produced a linear relationship were compared between the sexes and found to be not
significantly different (ANCOVA: F1=0.054, p>0.05). Because both relationships were not
significantly different between sexes the otolith-age relationship was presented as one
exponential curve and one linear curve for simple age identification from otolith length (Fig. 8).
The parameters for the curves produced by males and females pooled are presented in Table 1
for immature and mature fish.
3.2 ‘Spawning’ checks
‘Spawning’ checks were generally present the year the fish was tagged (Fig. 9). One tagged
female fish was removed from this section because the otolith was not a good enough cut to see
‘spawning’ checks. ‘Spawning’ checks of male fish were not significantly more often identified
the year of tagging than female fish (FET: p>0.05).
29
Figure 6. The relationship between otolith length and fork length for each of the three otolith
measuring techniques; parallel lines, line segments, and straight line. The R2s for each
relationship are 0.922, 0.918, and 0.884, respectively.
30
Figure 7. The relationship between otolith length and fork length for each sex using the parallel
lines measuring technique. Female and male growth are shown together to justify the
parallel lines measuring technique for males. The R2s for the relationships are 0.922 for
females and 0.897 for males.
31
Figure 8. Age-otolith relationship for both immature and mature lake trout. Males and females
were combined in these relationships. Those individuals that are between the ages 6 and 8
could be either mature or immature, thus both curves cover those regions. This plot can
estimate the age of an Opeongo lake trout based on its otolith size up to age 30.
32
Table 1. Parameters and correlation coefficients for the otolith-age relationship of immature and
mature fish. The immature fish were fit to straight lines (𝐴𝑔𝑒 = 𝑎 ∗ 𝑂𝑡 + 𝑏) and the
mature fish were fit to exponential curves (𝐴𝑔𝑒 = 𝑏𝑒𝑎∗𝑂𝑡). Fit refers to R2 for the linear
relationships and 1 −𝑅𝑆𝑆
𝑇𝑆𝑆 for nonlinear relationships (see text). Combined refers to the
male and female relationships being combined. SEx is the standard error for parameter x.
The distribution for the combined relationships are presented in Figure 8.
a SEa b SEb Fit
Immature Male 5.866 0.818 -1.742 1.028 0.456
Female 6.084 0.628 -2.050 0.796 0.544
Combined 5.990 0.497 -1.915 0.628 0.509
Mature Male 1.356 0.061 1.040 0.116 0.799
Female 1.227 0.078 1.322 0.196 0.670
Combined 1.286 0.049 1.183 0.108 0.736
33
Figure 9. The proportion of tagged fish that had a ‘spawning’ check (SC in figure) the year they
were tagged (nmale=54, nfemale=31), two years before they were tagged (nmale=33,
nfemale=15), and two years after they were tagged (nmale=50, nfemale=25). ‘Spawning’
checks noted before tagging were used to justify ‘spawning’ checks as an indicator of
maturity by using fish that had checks identified at tag and were assumed to have just
matured (aged ≤9 and ≤8 for females and males, respectively). The asterisks (*) denote
the pair of columns that were significantly different using Fisher’s exact test. Those fish
that had ‘spawning’ checks noted at tag were used to investigate skipped spawning by
observing which of those fish no longer had a ‘spawning’ check two years after tagging.
The lone asterisk denotes the borderline significance (p=0.0502) found when comparing
those fish identified with ‘spawning’ checks at tag (100%) to those same fish that were
also identified with ‘spawning’ checks two years after tagging.
*
*
*
* P
roport
ion o
f ‘S
pa
wn
ing’ C
he
cks P
resent
34
To verify these marks as spawning identifiers I used the assumption that most fish are tagged
during their first year on the spawning shoal therefore only fish aged 9 and under for females and
aged 8 and under for males were assessed for ‘spawning’ checks two years prior to being
tagged. Females that possessed a ‘spawning’ check two years before tagging were not
significantly different from the males (FET: p>0.05) (Fig. 9). Females that had ‘spawning’
checks at tagging were significantly different from those identified with checks two years before
tagging (FET: p<0.001). Similarly, the males also showed a significant difference between the
number of checks identified at tagging and two years before tagging (FET: p<0.001). A 60.6%
decrease was found in the proportion of ‘spawning’ checks identified two years prior to tagging
for females and a 54.5% decrease for males.
Skipped spawning was investigated by looking at the prevalence of ‘spawning’ checks two years
after the tag year. Only fish that were identified with a ‘spawning’ check the year of tagging
were used to see if a mark persisted two years after tagging. The prevalence of skipped spawning
was shown to be significantly different between the sexes (FET: p=0.0375). Females identified
to have ‘spawning’ checks two years after tagging were borderline significantly different from
those same fish previously identified to have ‘spawning’ checks at tag (FET: p=0.0502).
Conversely, males identified to have ‘spawning’ checks two years after tagging were not
significantly different from those same fish previously identified to have ‘spawning’ checks at
tag (FET: p>0.05). There was a 20% decrease in mark presence in females and a 4% decrease in
males (Fig. 9).
3.3 Tagged fish
The growth of free-living mature fish was compared to tagged fish by investigating the fork
length-otolith length relationship for each sex. Male free-living fish were found to not have a
significantly different growth pattern from those that were tagged (ANCOVA: F1=2.783,
p=0.0967) (Fig. 10). When the female tagged fish were compared to the free-living mature
females, their growth patterns were not significantly different (ANCOVA: F1=0.129, p>0.05)
(Fig. 11). The parameters and R2 values for free-living mature and tagged fish are presented in
Table 2.
35
Figure 10. The relationship between fork length and otolith length for male tagged and free-
living mature fish. The dashed lines indicate the approximate individual growth curves
for each tagged fish. The parameters and correlation coefficients of the free-living curve
are provided in Table 2. The distribution of free-living mature fish is provided in Figure
S3.
36
Figure 11. The relationship between fork length and otolith length for female tagged and free-
living mature fish. The dashed lines indicate the approximate individual growth curves
for each tagged fish. The parameters and correlation coefficients of the free-living curve
are provided in Table 2. The distribution of free-living mature fish is provided in Figure
S3.
37
Table 2. Parameters and correlation coefficients for the otolith length-fork length relationship of
tagged and free-living mature fish. The straight lines fit to the data were of the form (Fl =
a ∗ Ot + b) where fork length (Fl) is a function of otolith length (Ot). SEx is the standard
error for parameter x. Figures showing the distribution of the free-living mature
individuals are: 10, 11, S3.
a SEa b SEb R2
Male Free-living 280.85 19.93 75.67 32.79 0.649
Tagged 237.77 16.51 138.79 28.57 0.658
Female Free-living 260.41 21.59 124.68 37.32 0.603
Tagged 248.56 25.74 161.87 47.05 0.594
38
I also investigated how the fork length-age relationship may differ between free-living and
tagged fish. The three-parameter von Bertalanffy curve was fit to the tagged and the free-living
data for each sex. The three-parameter von Bertalanffy curves were compared between tagged
and free-living fish for the females and found to not be significantly different (ARSS:
F3,154=0.958, p>0.05) but were found to be significantly different for males (ARSS: F3,154=4.419,
p<0.01). Because males were found to have significantly different relationships, Ratkowsky’s
(1983) method was used to determine where the differences lie. The 𝑡0 value was not
significantly different between the two groups (CPE: F1,210=3.239, p=0.0733), therefore I decided
to use the two-parameter von Bertalanffy model, with a fixed 𝑡0 (Table 3). The new two-
parameter curves were compared to see if they differed significantly. The male curves were
found to differ significantly (ARSS: F2,212=5.615, p<0.01) (Fig. 12), however the female curves
were found to not be significantly different (ARSS: F2,156=0.689, p>0.05) (Fig. 13). Further use
of Ratkowsky’s (1983) method shows that the 𝐿∞ terms are borderline significantly different
from one another (CPE: F1,210=3.705, p=0.0556) for the male free-living and tagged
relationships.
39
Table 3. Von Bertalanffy parameters for free-living mature and tagged fish and the fit associated
with the curve created by the parameters. The L∞ and K parameters for the tagged
relationships were estimated using the respective fixed t0 values estimated from the free-
living fish relationship for each sex. Thus there is no SEt0 for tagged fish because the
value of t0 was taken from the free-living curves. The fit for the curves was calculated as
1 −𝑅𝑆𝑆
𝑇𝑆𝑆 (see text). SEx is the standard error for the parameter x. Figures showing the
distribution of the free-living mature individuals are: 12, 13, S4.
L∞ 𝐒𝐄𝐋∞ K 𝐒𝐄𝐊 t0 𝐒𝐄𝐭𝟎 Fit
Male Free-living 739.2 58.62 0.099 0.033 -3.617 2.12 0.693
Tagged 697.5 23.14 0.104 0.008 -3.617 - 0.580
Female Free-living 777.6 94.00 0.072 0.040 -7.658 5.43 0.469
Tagged 826.1 51.95 0.066 0.009 -7.658 - 0.496
40
Figure 12. The relationship between fork length and age for male tagged and free-living mature
fish. The grey dashed lines represent the approximate individual growth curve of each
tagged fish. The parameters and correlation coefficients of the free-living curve are
provided in Table 3. The distribution of free-living mature fish is provided in Figure S4.
41
Figure 13. The relationship between fork length and age for female tagged and free-living mature
fish. The grey dashed lines represent the approximate individual growth curve of each
tagged fish. The parameters and correlation coefficients of the free-living curve are
provided in Table 3. The distribution of free-living mature fish is provided in Figure S4.
42
4 Discussion
4.1 Measuring technique
The three measuring techniques had similar and uniformly high coefficients of determination.
This leads to the conclusion that empirically the three measuring techniques do not differ enough
to suggest one is superior to another. Theoretically it can be argued that the approximation
(novel) method is more accurate than the estimation (current) methods. Using a point to estimate
growth versus a line is useful when the nucleus can be readily viewed, but in many cases,
especially with novice agers, nuclei are difficult to identify (Pannella 1980; Niva et al. 2005;
Parmentier et al. 2007). Additionally, if the otolith is not cut in precisely the right position the
otolith section may not even show the true nucleus (Niva et al. 2005). Using a line to
approximate the nucleus overcomes the difficulty in identifying the true nucleus.
The novel method uses the sulcus as a land mark, most otoliths possess a sulcus (Nolf 1985)
making this a useful landmark to use. Although the sulcus persists in a variety of otoliths, it is
rarely exactly the same shape (Popper and Lu 2000). The sulcus of the sagittal otolith houses the
sensory epithelium used for hearing, therefore it should be more pronounced on fish that have
better hearing (Popper and Lu 2000). Due to the functional significance of the sulcus for fish
hearing (Pannella 1980; Popper and Lu 2000) it is an appropriate landmark for measuring otolith
sections. It is assumed that fish inhabiting the same environment of the same species should have
similar hearing capabilities (Popper and Fay 2011), therefore there should be limited variation in
the sulcus of the otoliths used for this study. The shape of the sulcus also depends on the quality
of the otolith section cut (Pannella 1980; Niva et al. 2005). Since the sulcus can always be
viewed the tip of the sulcus should also always be visible, therefore the first point of the baseline
can theoretically always be drawn. Drawing the baseline does depend somewhat on the
magnification limitations of the microscope, if a good image of the otolith thin section is not
possible, placing the first point of the baseline will be difficult.
Dividing the dorsal and ventral lobes of the first annulus through the sulcus should aid in the
consistent placement of the baseline, visually it is easy to divide this shape. The baseline can be
drawn at a few different angles while still dividing the lobes through the sulcus, therefore it is
assumed that these are not appreciable differences. This is because the measurements were
between parallel lines, and the methods described imply that the measurement to each annulus
43
will be at the same point of furthest growth. It is essential to keep a consistent method through all
measurements when placing the baseline through the sulcus. Additionally, when investigating
some otolith sections, a sulcus-like structure on the anti-sulcul side was observed, that appeared
as a faint indentation. If this structure is present, it can be used as a guide in dividing the lobes.
During preliminary annuli identification, this structure was seen more prominently on Coregonus
artedii otolith thin sections. A possible reason for this could be that C. artedii otoliths are more
sensitive to sound than S. namaycush otoliths, and therefore the squamous epithelium of the
sagittae is more sensitive to sound. Conversely, this could be an artefact of C. artedii sagittae
size, being much larger than those of similar-sized S. namaycush.
The curvature of otoliths varies within and among populations (Nolf 1985; Secor and Dean 1989;
Popper and Fay 2011), therefore it is problematic to use a straight line to estimate the growth of a
curved structure. Furthermore, it is difficult to keep a common angle (or curve) when measuring
otoliths and many measuring programs are not capable of this. Additionally, if measurements are
taken to each annulus, multiple lines from nucleus to annuli (along the same axis) must be drawn
which is time consuming and increases measuring error. Many studies that use otolith length
measurements in their calculations use the ‘otolith radius’ measurement but do not describe how
this measurement was taken (Wright et al. 1990; Secor and Dean 1992; Schirripa and Goodyear
1997). This lack of specification implies that the method of measurement is not important
however this study indicates that it does make a difference in the correlation.
Line segments are the most common measuring technique used by technicians, however for them
to be applied properly each annulus has to be correctly identified. If all annuli are not identified
or some are improperly identified, measurement error will be encountered because each
measurement relies on the accuracy of the one preceding it. For an experienced ager, this may
not pose a problem, however for a novice ager it is difficult to identify all annuli accurately
(Boehlert 1985; Campana and Moksness 1991). Also for this technique, the furthest point of
growth on each annulus must be visually estimated, but when using parallel lines the annuli
identification line is moved to the furthest point of growth that touches the line. The need for
correctly identifying each annulus is eliminated when using parallel lines because each annular
measurement is independent of the next. Therefore the ager is less likely to accumulate error
using the parallel lines technique. This technique is also easier for novice agers to employ. It is
recommended that this technique be justified with other species and populations before being
44
utilized. It is also recommended that, in future studies, the measuring technique used be clearly
specified. Increased specification of otolith measuring techniques would allow growth
comparisons between studies.
The significance found between the intercepts when comparing male and female growth is likely
attributed to the differences in growth between sexes at small sizes. Although it has been noted
that male and female fish grow differently (Dwyer et al. 2003; Rennie et al. 2008; Armsworthy
and Campana 2010), female and male lake trout in this lake have been shown to not differ
greatly in length at age (McDermid et al. 2007). Therefore it is likely that young males grow
slower than young females as male otoliths are larger than females at the same fork length for
small fish (Reznick et al. 1989). Both relationships showed high coefficients of determination
therefore the parallel lines measuring technique could be employed for both sexes. The high
correlation coefficients found for both sexes suggest that decoupling of the otolith length-fish
length relationship is not extreme enough in this species to alter the curve from linear.
Inferring age from otolith length using a subset of the population can be more cost effective and
time saving than aging each otolith by counting annuli (Francis and Campana 2004). It is
possible to estimate the age of a fish from its otolith length (Radtke et al. 1985; Secor and Dean
1989). Using the novel measuring technique, novices can verify the age of a fish in the absence
of an experienced ager. This will reduce the large learning curve associated with aging fish
(Boehlert 1985). The methods involved in creating the relationship between otolith length and
age can be applied to many other fish species. The curve presented here, however, should only be
used to estimate the age of S. namaycush from Lake Opeongo since otolith shape and size vary
between populations. Additionally, caution is advised when estimating the age of otoliths larger
than 2.25 mm since only a few otolith sections this large (or larger) were used to create the
curve.
Significant differences were not found between the male and female curves when investigating
the otolith length-age relationship. This could be due to male and female growth being fairly
similar throughout life (McDermid et al. 2007). One of the only documented differences in this
species is reproductive timing, where, on average, males mature at an earlier age (McDermid et
al. 2007). Otolith length-age relationships typically exhibit high R2 values (Lychakov and
Rebane 2000) which implies male and female otolith length-age relationships should not differ
45
greatly and therefore supports combining these relationships. Combining the male and female
curves was also done to simplify the curve for novice agers. Separating for maturity in this
relationship was done because the residual sum of squares was minimized with a linear
relationship and an exponential curve as opposed to an exponential curve for the whole
relationship. It has been documented that otolith length and age are linearly related in juveniles
(Secor and Dean 1989; Wright et al. 1990). Kirkwood (1983) used fish length to estimate age,
however fish length is less correlated with age than otolith length (Boehlert 1985), therefore it is
better to estimate age from otolith length when possible.
The relationship between fish length and otolith length is usually assumed to be linear (Campana
1990), here a linear relationship was used because visually the relationship most resembled a
linear one. Additionally, Perez and Munch (2013) found that linear relationships were the most
parsimonious model for describing this relationship. Therefore, decoupling of the fish length-
otolith length relationship is not extreme enough to alter the relationship from linear.
Experimentally, decoupling of this relationship can be affected by varying environmental
conditions (Hoff and Fuiman 1993; Baumann et al. 2005). Most decoupling observations concern
juveniles and daily growth patterns (Munk and Smikrud 2002).
The aging method used in this study has many advantages over traditional methods. Because it
uses digital images, a novice ager can learn more quickly and efficiently from an experienced
ager through joint discussion and permanent annotation of sampled images. In addition, although
the software used in this study cannot automatically count annuli, identifying each annulus and
having the program count them is more efficient and it makes age verification easier. When
annuli are not clearly identified, miscounting could occur and verification would require
recounting, which is less efficient than looking at the location of the points on an image.
Additionally, all annulus indicators can be hidden and later recalled – this permits other agers to
independently age the same otolith section without bias and facilitates comparison of the results.
4.2 ‘Spawning’ checks
Checks on otoliths have been noted on a number of species although the reasons they occur have
yet to be fully determined (Campana 1983). Checks occur due to stress where the calcium
carbonate is not accreted onto the protein matrix (Payan et al. 2004), however any previously
accreted calcium is not resorbed (Campana 1983). In younger fish, temperature and food stress
46
may be large contributors to the formation of checks (Campana 1983; Miller and Simenstad
1994), however these conditions are less stressful for adult fish. Experienced reproducers invest
more in reproduction than recently matured fish (Lester et al. 2004), therefore reproduction may
be considered one of the more stressful events of the growing season for an adult fish.
Additionally, the trade-off between growth and reproduction are different between male and
female lake trout (Lester et al. 2004).
Checks, mainly composed of proteins, similar to the ones observed in this study were also
viewed by Gauldie (1988) on otoliths of the New Zealand snapper, Chrysophrys auratus,
however their significance was not specified as more than an indication of stress. Pannella (1974,
1980) observed ‘spawning’ checks on otoliths of a variety of species, but only stated they were
likely an indication of spawning and did not validate them. The relatively high prevalence of
‘spawning’ checks in the year of tagging for both sexes suggests that the check could be an
indicator of spawning. The proportion of fish with ‘spawning’ checks the year they were tagged
was higher in males than in females, likely due to a higher concentration of protein in the
endolymph in males than in females (Payan et al. 2004). The protein concentration has been
shown to increase during artificial chemical stress, which may reduce the concentration of free
calcium available for accretion (Payan et al. 2004). Warner (1971) identified ‘spawning’ checks
on scales and concluded that the scale margin is more intensely resorbed in males than in
females. Pannella (1980) suggested that during reproduction, protein-bound calcium is resorbed
from wherever possible, for egg production. This leads to the assumption that protein-bound
calcium is also taken from the endolymph fluid for egg production and not enough is left for
otolith accretion. Therefore it is likely that protein-bound calcium is also resorbed from the
endolymph, which could cause zones of little to no accretion; here observed as ‘spawning’
checks. The mechanism behind ‘spawning’ check formation in males is less clear (Pannella
1980). Studies of male fish are not often completed as it is easier to monitor the reproductive
status of females who have more variation in gonad and gamete size (Rideout and Tomkiewicz
2011).
The significant drop in ‘spawning’ check presence two years prior to tagging in both sexes
suggests that this mark is indeed an indicator of spawning. ‘Spawning’ checks have only been
noted thus far on scales, where researchers have concluded the calcium deposits on the scale
were resorbed for gonad production (Warner 1971; Pannella 1980; Rideout et al. 2005).
47
Unfortunately, ‘spawning’ checks on scales more than two years prior to catch can be lost over
time due to the resorption of calcium from opaque zones (Warner 1971). However on otoliths,
calcium resorption has only been noted experimentally as a sign of an oxygen depleted
environment (Mugiya and Uchimura 1989), therefore any ‘spawning’ checks formed on otoliths
will remain there for the duration of the fish’s life.
Salvelinus namaycush in this population usually mature around eight years (McDermid et al.
2007) which was similar to observations of the creel population within the study years. However,
the majority of males were found to mature by seven, which could be an artefact of differing
maturity identification methods. Therefore using fish aged ≤9 for females and ≤8 for males to
determine if the ‘spawning’ check prevalence was linked to maturity would more likely lead to
demonstrating a change in maturity status between the two years. This was because each sex was
assumed to have matured within those last two years. The proportion of mature fish did not
decrease as much in males as it did in females probably because male S. namaycush, on average,
mature before females and thus would have been mature two years prior to tagging.
Skipped spawning may occur in S. namaycush females since there was a borderline significant
decrease in the prevalence of ‘spawning’ checks between the tagged year and two years post
tagging. A non-significant decrease in S. namaycush males suggests that skipped spawning likely
does not occur in this sex. In temperate environments, switching between growth and
reproduction is optimal in the years after maturity (Kozłowski 1996). Female S. namaycush may
employ this trade-off between reproduction and growth more often than males. Additionally,
recently matured fish have been observed to skip reproduction in their second potential spawning
season, likely due to investing in growth rather than reproduction (Engelhard and Heino 2005,
2006). S. namaycush may skip the first spawning season after their first reproductive season,
however I did not assess this possibility in this study. Females are more likely to skip spawning
because there is more energy involved in egg production than sperm production (Rideout et al.
2005). This could lead to females spending more years investing in future reproduction than
males (Rideout and Tomkiewicz 2011). I hypothesize the reason for skipping spawning in this
species is likely due to investing in future reproductive success.
Skipped spawning has been shown to not differ between sexes and that both sexes skip the
second reproductive season in herring, Clupea harengus (Engelhard and Heino 2006). This
48
contrasts with the present study in there was a significant difference found between male and
female skipped spawning. Often the instance of skipped spawning is not investigated in males
(Rideout et al. 2005). Female fish are more often noted to skip spawning than males by
investigation of gonads and scale ‘spawning’ checks (Rideout et al. 2005; Jørgensen et al. 2006).
This could also be an artefact of more defined ‘spawning’ checks on the scales of males than
females (Warner 1971). Skipped spawning has also been observed when the ratio of males to
females is less than ideal (Engelhard and Heino 2006).
The prevalence of ‘spawning’ checks two years after tagging could be biased by poor otolith
preparation methods which would make checks difficult to identify especially in otoliths
prepared closer to 1994. Although there is evidence that supports males do not skip spawning
and females do, since borderline significance was found, this pattern should be confirmed with a
study that uses a larger sample size. Furthermore, future studies should look into the prevalence
of skipped spawning within cohorts and spawning years to determine the reproductive patterns
that persist in this species and others. This study shows how the age at maturity and years of
spawning can be determined from otoliths for all sampled fish without the need to closely
monitor the population. Additionally, the use of otoliths in this sense can show historically (using
archived otoliths) which fish reproduced in which years and possibly indicate sex-, species-, or
population-specific reproduction patterns or how reproductive patterns have changed over time.
In general, it was easier to identify ‘spawning’ checks on males than on females, this could be
because more male otolith sections were cut better than the female sections. Male otolith sections
are likely to have better cuts than those of females since the majority of male otoliths were
caught and prepared later in the time series, when the otolith sectioning technique was better
employed. This pattern was also observed by Warner (1971) on the scales of Atlantic salmon,
Salmo salar, who concluded it was due to differential calcium resorption of the scale periphery.
Pannella (1980) observed ‘spawning’ checks in males to be more prevalent than females and
concluded it was due to the amount of protein-bound calcium in the fish. When female fish
decide to reproduce the amount of protein-bound calcium within the body increases whereas the
amount of ionic calcium remains consistent and the ionic calcium is assumed to be the form of
calcium that is incorporated into the otolith (Pannella 1980). Recall that protein is increased in
the endolymph during stressful events (Payan et al. 2004), if reproduction is also considered a
stressful event, due to the lack of food intake (Jørgensen et al. 2006), the protein concentration
49
would increase during gonad production. Since protein-bound calcium is increased within the
whole female body (likely more than in males) (Pannella 1980), the relative amount of protein-
bound calcium in the endolymph would be higher in females during reproduction than in males.
Therefore females will have proportionally less protein being continually incorporated and
etched away during reproduction than males (Campana 1983), this would create a more defined
check on male otoliths as compared to females.
The absence of a ‘spawning’ check in the year of tagging does not necessarily mean that the
‘spawning’ check is an unreliable indicator of spawning: immature fish may frequent the
spawning shoals or mature fish that are ‘skipping’ spawning may also still venture to the
spawning shoals. Opeongo S. namaycush is a non-migratory species, unlike most species used
for ‘spawning’ check studies (Rideout et al. 2005), therefore they do not experience the cost of
migration as a trade-off for reproduction (Rideout and Tomkiewicz 2011). Thus it is possible that
some immature (or skipping) individuals might happen to be at the spawning shoal at the time of
tagging. Since the prevalence of ‘spawning’ checks should not be related to the cost of migration
in this population, it is likely related to the cost of reproduction. In migratory species, such as the
cod Gadus morhua, eating is highly reduced during migration to save energy stores for
reproduction (Jørgensen et al. 2006). In S. namaycush, saving energy stores for reproduction
could also occur and could be another source of the definition of the checks. Therefore, on S.
namaycush otoliths the accretion zones between ‘spawning’ checks and annuli could indicate the
time a fish feasts after reproduction prior to entering the stressful winter months.
4.3 Tagged fish
Stress is one of the reasons fish can show a reduction in their growth (Payan et al. 2004;
Baumann et al. 2005). Stress brought on by the insertion of physical tags for scientific studies
could alter the results of those studies (Tranquilli and Childers 1982; Otterå et al. 1998).
Therefore all populations of tagged fish should be monitored to ensure they do not differ from
the rest of the population. Using otoliths to model the effects of tags on growth has yet to be
conducted. Since otoliths are not affected by stress that normally affects somatic growth they can
show differences in growth between stressed and unstressed fish (Campana and Neilson 1985;
Baumann et al. 2005). Baumann et al. (2005) used juvenile fish to investigate how otolith growth
is affected by limited food intake since juveniles are more sensitive to changes in the
50
environment that affect growth, and they found otoliths continue to grow even in the presence of
stressors.
Otoliths are not often used when investigating tag effects, usually the fish length, weight and
condition are used (Otterå et al. 1998; Baras et al. 2000). Additionally, most studies that focus on
the effects of tagging study juveniles, and are rarely conducted for a time period greater than two
years (Carline and Bynildson 1972; Baras et al. 2000). Using otoliths for mature long-lived fish
for back-calculation could be more accurate than using scales since scales do not reflect the
relationship between fish length and scale length as accurately for mature long-lived fish
(Boehlert 1985). Scales often suffer from abrasion and are lost on long-lived fish making it
difficult to gauge the growth between the annuli of mature scales (Campana and Neilson 1985;
Campana and Thorrold 2001). Also for all the tagged fish studied there was an increase in fish
length from tag to catch date which may not be reflected as accurately by scales as by otoliths.
This study did have a limited number of tagged individuals, since the number collected depended
on those caught by anglers, therefore it is suggested that future studies add to this current data to
confirm these conclusions.
The fork length-otolith length relationship for mature free-living male fish was elevated over the
growth trajectory for tagged male fish. This was the expected result if tags affected growth,
however, these relationships were not significantly different from one another. Conversely, the
fork length-otolith length relationship for tagged female fish was elevated over that of free-living
fish suggesting that tagging actually accelerates somatic growth. This result could arise if the
limited number of tagged females exhibited a relationship that does not reflect the tagged
population as a whole. The elevated relationship could suggest that larger females are more often
tagged on the spawning shoals than smaller females. In conclusion, this study did not detect a
statistically significant effect of tagging on the relationship linking fish length to otolith length.
Similarly, Tranquilli and Childers (1982) found that Floy anchor tags do not affect the growth of
largemouth bass, Micropterus salmoides, despite the presence of sores. Since decoupling is not
extreme in this species, small effects on growth will likely not be detectable when comparing
mature free-living to tagged S. namaycush.
The relationship between fork length and age for mature male free-living fish was significantly
different from the tagged fish relationship. This was to be expected if tags affect the growth of
51
fish. It could be reasoned that t-tags affect the growth of males; however, the sample of free-
living mature curve created to reflect the male population had only two individuals over the age
of 18 whereas the tagged relationship had six. Thus the free-living curve is less accurate after
that age, implying that the 𝐿∞ value would change with a repeated study that incorporated more
individuals older than 18. Therefore evidence that t-tags affect the growth of males is weak and
the significant effect that was found may be attributed to a sample that improperly reflected the
true population. Additionally, the borderline significance found between the two 𝐿∞ terms
supports that this term is likely the main contributor to the significance found between the male
tagged and free-living relationships. Since other studies involving mature tagged fish did not
show a reduction in growth (Tranquilli and Childers 1982; Otterå et al. 1998) it is likely the
significance encountered here is due to the limited number of large male S. namaycush
misrepresenting the population. The female relationships show again that tagged fish grow faster
than mature free-living fish, however this result is not significant. Again, likely due to the limited
number of tagged individuals falsely representing the tagged population. Therefore the evidence
developed in this study does not suggest that t-tags have a large effect on the growth of S.
namaycush for either sex, however since this study involved a limited number of individuals,
another study is recommended to confirm the results.
Both the presented fork length-otolith length and fork length-age relationships for the mature
free-living fish of each sex can be used for back-calculation purposes. It is advised that back-
calculation in this capacity focuses on mature S. namaycush individuals. The methods involved
in creating these curves can be applied to many other fish species. Klumb et al. (2001) claimed
there is no otolith axis that will provide the most accurate back-calculation length, however in
long-lived fish, where scales make back-calculation difficult, the novel measuring and annuli
identification techniques can overcome this difficulty.
Finding a curve to represent the relationship between fish age and fish length is more difficult
than finding one for otolith length and fish age (Lychakov and Rebane 2000). Therefore a two-
parameter curve was employed instead of a three-parameter one. Using the two-parameter curve
may have decreased the fit slightly but, since it is reasonable to assume that prior to being tagged
both tagged and mature free-living fish would have been on the same growth trajectory, then it is
also reasonable to assume that length at age zero would have been the same for both (Pardo et al.
2013). The mature free-living individuals were used to estimate 𝑡0 for each sex to avoid biasing
52
the term with the lengths of tagged fish. Additionally, the other two-parameter estimates were
not significant when 𝑡0 was not fixed especially for the tagged fish relationships. Fixing this
value created better estimates for the other two terms. The limited number of individuals added
to the difficulty in finding a good fitting three-parameter curve.
Most of these fish were tagged twice; once with a t-tag and once with a PIT tag. McAllister et al.
(1992) looked at nine different commercially available external tag effects on rainbow trout,
Oncorhynchus mykiss, and found that t-tags were the least harmful to fish. T-tags were also
found to be retained by fish more often than all the other tags, and they showed the least effect
on growth (McAllister et al. 1992). PIT tags on average take two weeks to heal (Baras et al.
2000). External tags can take longer to heal since they are exposed and the frictional forces of
swimming can cause them to move and open the fish epithelium and cause a lesion (Carline and
Bynildson 1972; Tranquilli and Childers 1982; Otterå et al. 1998). Lesions can be the entrance
pathway for parasites and pathogens which could affect the health, and subsequently the growth,
of a fish (McAllister et al. 1992). Carline and Bynildson (1972) speculated that the lag in growth
they observed between tagged and free-living juvenile brook trout, Salvelinus fontinalis, was
caused by tagging lesions. However, other studies usually report tag wounds, that have difficulty
healing, but are not reported to affect growth (Tranquilli and Childers 1982; Otterå et al. 1998).
PIT tags have been used in a number of species and the only effect on growth occurs when
smaller fish are implanted with larger tags (Otterå et al. 1998; Baras et al. 2000). The liver has
been noted to grow around internal tags, which has no observed effect on growth or survival
(Otterå et al. 1998). Since all tagged fish in this study are assumed to be mature, and thus large,
fish will likely not be affected by the mass of their tag. Therefore if growth was found to be
affected by tags in this study, the effects would likely be caused by t-tags.
53
5 Conclusions
The novel otolith measuring technique, presented here, once validated, can be used for a number
of different species and be accurately employed by novice agers. Additionally, novice agers can
utilize the presented otolith images as guidelines for recognising annuli and checks to reduce the
learning curve associated with aging. The post-maturation checks often viewed in older fish
otoliths can be used for identifying spawning years or maturity in S. namaycush. Furthermore,
identifying ‘spawning’ checks on otoliths can be applied to other fish species to determine their
reproductive history. The novel measuring and annuli identification techniques presented here
can be applied to assess somatic growth patterns. I concluded that, any effects that t-tags may
have on the growth of Lake Opeongo S. namaycush were below the detection limits of the
methods used in this study.
54
6 References
Allen, K.R. 1966. A method of fitting growth curves of the von Bertalanffy type to observed
data. J. Fish. Res. Board Canada 23: 163–179.
Armsworthy, S.L., and Campana, S.E. 2010. Age determination, bomb-radiocarbon validation
and growth of Atlantic halibut (Hippoglossus hippoglossus) from the Northwest Atlantic.
Environ. Biol. Fishes 89: 279–295. doi: 10.1007/s10641-010-9696-8.
Arnott, S.A., Chiba, S., and Conover, D.O. 2006. Evolution of intrinsic growth rate: metabolic
costs drive trade-offs between growth and swimming performance in Menidia menidia.
Evolution (N. Y). 60: 1269–1278. Blackwell Publishing Ltd.
Baras, E., Malbrouck, C., Houbart, M., Kestemont, P., and Mélard, C. 2000. The effect of PIT
tags on growth and physiology of age-0 cultured Eurasian perch Perca fluviatilis of variable
size. Aquaculture 185: 159–173.
Baumann, H., Peck, M.A., and Herrmann, J.-P. 2005. Short-term decoupling of otolith and
somatic growth induced by food level changes in postlarval Baltic sprat, Sprattus sprattus.
Mar. Freshw. Res. 56: 539–547. doi: 10.1071/MF04140.
Beamish, R.J., and McFarlane, G.A. 1987. Current trends in age determination methodology. In
Age and Growth of Fish. Edited by R.C. Summerfelt and G.E. Hall. Iowa State University
Press, Ames, Iowa. pp. 15–42.
Begg, G.A., Campana, S.E., Fowler, A.J., and Suthers, I.M. 2005. Otolith research and
application: current directions in innovation and implementation. Mar. Freshw. Res. 56:
477–483. CSIRO. doi: 10.1071/MF05111.
Bergstedt, R.A., Eshenroder, R.L., Bowen, C., Seelye, J.G., and Locke, J.C. 1989. Studies of
homing and reproductive biology of lake trout (part 3 - rearing and stocking of early life
stages). Great Lakes Fishery Commision Research Completion Report.
Boehlert, G.W. 1985. Using objective criteria and multiple regression models for age
determination in fishes. Fish. Bull. 83: 103–117.
Borelli, G., Guibbolini, M.E., Mayer-Gostan, N., Priouzeau, F., De Pontual, H., Allemand, D.,
Puverel, S., Tambutte, E., and Payan, P. 2003. Daily variations of endolymph composition:
relationship with the otolith calcification process in trout. J. Exp. Biol. 206: 2685–2692. doi:
10.1242/jeb.00479.
Brazner, J.C., Campana, S.E., Tanner, D.K., and Schram, S.T. 2004. Reconstructing habitat use
and wetland nursery origin of yellow perch from Lake Superior using otolith elemental
analysis. J. Great Lakes Res. 30: 492–507. doi: 10.1016/S0380-1330(04)70365-2.
55
Brown, P., Green, C., Sivakumaran, K.P., Stoessel, D., and Giles, A. 2004. Validating otolith
annuli for annual age determination of common carp. Trans. Am. Fish. Soc. 133: 190–196.
doi: 10.1577/T02-148.
Campana, S.E. 1983. Calcium deposition and otolith check formation during periods of stress in
coho salmon, Oncorhynchus kisutch. Comp. Biochem. Physiol. Part A Mol. \& Integr.
Physiol. 75: 215–220.
Campana, S.E. 1984. Interactive effects of age and environmental modifiers on the production of
daily growth increments in otoliths of plainfin midshipman, Porichthys notatus. Fish. Bull.
82: 165–177.
Campana, S.E. 1990. How reliable are growth back-calculations based on otoliths? Can. J. Fish.
Aquat. Sci. 47: 2219–2227.
Campana, S.E. 1999. Chemistry and composition of fish otoliths: pathways, mechanisms and
applications. Mar. Ecol. Prog. Ser. 188: 263–297.
Campana, S.E. 2001. Accuracy, precision and quality control in age determination, including a
review of the use and abuse of age validation methods. J. Fish Biol. 59: 197–242. doi:
10.1006/jfbi.2001.1668.
Campana, S.E. 2005. Otolith science entering the 21st century. Mar. Freshw. Res. 56: 485–495.
doi: 10.1071/MF04147.
Campana, S.E., and Casselman, J.M. 1993. Stock discrimination using otolith shape analysis.
Can. J. Fish. Aquat. Sci. 50: 1062–1083.
Campana, S.E., and Moksness, E. 1991. Accuracy and precision of age and hatch date estimates
from otolith microstructure examination. ICES J. Mar. Sci. 48: 303–316.
Campana, S.E., and Neilson, J.D. 1985. Microstructure of fish otoliths. Can. J. Fish. Aquat. Sci.
42: 1014–1032.
Campana, S.E., and Thorrold, S.R. 2001. Otoliths, increments, and elements: keys to a
comprehensive understanding of fish populations? Can. J. Fish. Aquat. Sci. 58: 30–38. doi:
10.1139/cjfas-58-1-30.
Carline, R.F., and Bynildson, O.M. 1972. Effects of the Floy anchor tag on the growth and
survival of brook trout (Salvelinus fontinalis). J. Fish. Res. Board Canada 29: 458–460.
Casselman, J.M. 1990. Growth and relative size of calcified structures of fish. Trans. Am. Fish.
Soc. 119: 673–688. doi: 10.1577/1548-8659(1990)119<0673.
Charnov, E.L., Turner, T.F., and Winemiller, K.O. 2001. Reproductive constraints and the
evolution of life histories with indeterminate growth. Proc. Natl. Acad. Sci. U. S. A. 98:
9460–9464. doi: 10.1073/pnas.161294498.
56
Chen, Y., Jackson, D.A., and Harvey, H.H. 1992. A comparison of von Bertalanffy and
polynomial functions in modelling fish growth data. Can. J. Fish. Aquat. Sci. 49: 1228–
1235.
Conover, D.O., Arnott, S.A., Walsh, M.R., and Munch, S.B. 2005. Darwinian fishery science:
lessons from the Atlantic silverside (Menidia menidia). Can. J. Fish. Aquat. Sci. 62: 730–
737. doi: 10.1139/F05-069.
Degens, E.T., Deuser, W.G., and Haedrich, R.L. 1969. Molecular structure and composition of
fish otoliths. Mar. Biol. 2: 105–113.
Dunlop, E.S., Shuter, B.J., and Dieckmann, U. 2007. Demographic and evolutionary
consequences of selective mortality: predictions from an eco-genetic model for smallmouth
bass. Trans. Am. Fish. Soc. 136: 749–765. doi: 10.1577/T06-126.1.
Dunlop, E.S., Shuter, B.J., and Ridgway, M.S. 2005. Isolating the influence of growth rate on
maturation patterns in the smallmouth bass (Micropterus dolomieu). Can. J. Fish. Aquat.
Sci. 62: 844–853. doi: 10.1139/F05-045.
Dwyer, K.S., Walsh, S.J., and Campana, S.E. 2003. Age determination, validation and growth of
Grand Bank yellowtail flounder (Limanda ferruginea). ICES J. Mar. Sci. 60: 1123–1138.
doi: 10.1016/S1054.
Elsdon, T.S., Wells, B.K., Campana, S.E., Gillanders, B.M., Jones, C.M., Limburg, K.E., Secor,
D.H., Thorrold, S.R., and Walther, B.D. 2008. Otolith chemistry to describe movements and
life-history parameters of fishes: hypothesses, assumptions, limitations and inferences.
Oceanogr. Mar. Biol. 46: 297–330.
Engelhard, G.H., and Heino, M. 2005. Scale analysis suggests frequent skipping of the second
reproductive season in Atlantic herring. Biol. Lett. 1: 172–175. doi:
10.1098/rsbl.2004.0290.
Engelhard, G.H., and Heino, M. 2006. Climate change and condition of herring (Clupea
harengus) explain long-term trends in extent of skipped reproduction. Oecologia 149: 593–
603. doi: 10.1007/s00442-006-0483-3.
Francis, R.I.C.C., and Campana, S.E. 2004. Inferring age from otolith measurements: a review
and a new approach. Can. J. Fish. Aquat. Sci. 61: 1269–1284. doi: 10.1139/F04-063.
Galley, E.A., Wright, P.J., and Gibb, F.M. 2006. Combined methods of otolith shape analysis
improve identification of spawning areas of Atlantic cod. ICES J. Mar. Sci. 63: 1710–1717.
doi: 10.1016/j.icesjms.2006.06.014.
Gauldie, R.W. 1988. Similarities in fine structure of annual, and non-annual, check rings in the
otolith of the New Zealand snapper (Chrysophrys auratus). New Zeal. J. Mar. Freshw. Res.
22: 273–278.
57
Gauldie, R.W., and Nelson, D.G.A. 1990. Otolith growth in fishes. Comp. Biochem. Physiol.
Part A Mol. \& Integr. Physiol. 97: 119–135.
Halden, N.M., and Friedrich, L.A. 2008. Trace-element distributions in fish otoliths: natural
markers of life histories, environmental conditions and exposure to tailings effluence.
Mineral. Mag. 72: 593–605. doi: 10.1180/minmag.2008.072.2.593.
Hoff, G.R., and Fuiman, L.A. 1993. Morphometry and composition of red drum otoliths: changes
associated with temperature, somatic growth rate, and age. Comp. Biochem. Physiol. Part A
Mol. \& Integr. Physiol. 106: 209–219. doi: 10.1016/0300-9629(93)90502-U.
Jørgensen, C., Ernande, B., Fiksen, Ø., and Dieckmann, U. 2006. The logic of skipped spawning
in fish. Can. J. Fish. Aquat. Sci. 63: 200–211. doi: 10.1139/F05-210.
Kalish, J.M. 1989. Otolith microchemistry: validation of the effects of physiology, age and
environment on otolith composition. J. Exp. Mar. Bio. Ecol. 132: 151–178. doi:
10.1016/0022-0981(89)90126-3.
Kerr, S.R. 1971. A simulation model of lake trout growth. J. Fish. Res. Board Canada 28: 815–
819.
King, J.R., Shuter, B.J., and Zimmerman, A.P. 1999. Signals of climate trends and extreme
events in the thermal stratification pattern of multibasin Lake Opeongo, Ontario. Can. J.
Fish. Aquat. Sci. 56: 847–852. doi: 10.1139/f99-020.
Kirkwood, G.P. 1983. Estimation of von Bertalanffy growth curve parameters using both length
increment and age-length data. Can. J. Fish. Aquat. Sci. 40: 1405–1411.
Klumb, R.A., Bozek, M.A., and Frie, R. V. 2001. Validation of three back-calculation models by
using multiple oxytetracycline marks formed in the otoliths and scales of bluegill × green
sunfish hybrids. Can. J. Fish. Aquat. Sci. 58: 352–364. doi: 10.1139/cjfas-58-2-352.
Kozłowski, J. 1996. Optimal allocation of resources explains interspecific life-history patterns in
animals with indeterminate growth. Proc. R. Soc. London Ser. B 263: 559–566.
Lester, N.P., Shuter, B.J., and Abrams, P.A. 2004. Interpreting the von Bertalanffy model of
somatic growth in fishes: the cost of reproduction. Proc. R. Soc. London Ser. B 271: 1625–
1631. doi: 10.1098/rspb.2004.2778.
Loher, T., and Seitz, A.C. 2008. Characterization of active spawning season and depth for
eastern Pacific halibut (Hippoglossus stenolepis), and evidence of probable skipped
spawning. J. Northwest Atl. Fish. Sci. 41: 23–36. doi: 10.2960/J.v41.m617.
Lychakov, D. V, and Rebane, Y.T. 2000. Otolith regularities. Hear. Res. 143: 83–102. Available
from http://www.ncbi.nlm.nih.gov/pubmed/10771186.
Martin, N. V, and Fry, F.E.J. 1972. Lake Opeongo: effects of exploitation and introductions on
the salmonid community. J. Fish. Res. Board Canada 29: 795–805.
58
Mc Dougall, A. 2004. Assessing the use of sectioned otoliths and other methods to determine the
age of the centropomid fish, barramundi (Lates calcarifer) (Bloch), using known-age fish.
Fish. Res. 67: 129–141. doi: 10.1016/j.fishres.2003.09.044.
McAllister, K.W., McAllister, P.E., Simon, R.C., and Werner, J.K. 1992. Performance of nine
external tags on hatchery-reared rainbow trout. Trans. Am. Fish. Soc. 121: 192–198. doi:
10.1577/1548-8659(1992)121<0192.
McDermid, J.L., Ihssen, P.E., Sloan, W.N., and Shuter, B.J. 2007. Genetic and environmental
influences on life history traits in lake trout. Trans. Am. Fish. Soc. 136: 1018–1029. doi:
10.1577/T06-189.1.
Meekan, M.G., Dodson, J.J., Good, S.P., and Ryan, D.A.J. 1998. Otolith and fish size
relationships, measurement error, and size-selective mortality during the early life of
Atlantic salmon (Salmo salar). Can. J. Fish. Aquat. Sci. 55: 1663–1673. doi: 10.1139/f98-
047.
Miller, J.A., and Simenstad, C.A. 1994. Otolith microstructure preparation, analysis, and
interpretation: procedures for a potential habitat assessment methodology. Seattle,
Washington.
Morbey, Y.E., Couture, P., Busby, P., and Shuter, B.J. 2010. Physiological correlates of seasonal
growth patterns in lake trout Salvelinus namaycush. J. Fish Biol. 77: 2298–2314. doi:
10.1111/j.1095-8649.2010.02804.x.
Mosegaard, H., Svedäng, H., and Taberman, K. 1988. Uncoupling of somatic and otolith growth
rates in Arctic char (Salvelinus alpinus) as an effect of differences in temperature response.
Can. J. Fish. Aquat. Sci. 45: 1514–1524.
Mugiya, Y. 1966. Calcification in fish and shell-fish - VI. seasonal change in calcium and
megnesium concentrations of the otolith fluid in some fish, with special reference to the
zone formation of their otolith. Bull. Japanese Soc. Sci. Fish. 32: 549–557.
Mugiya, Y. 1986. Effects of calmodulin inhibitors and other metabolic modulators on in vitro
otolith formation in the rainbow trout, Salmo gairdnerii. Comp. Biochem. Physiol. Part A
Physiol. 84: 57–60. doi: 10.1016/0300-9629(86)90042-3.
Mugiya, Y., and Uchimura, T. 1989. Otolith resorption induced by anaerobic stress in the
goldfish, Carassius auratus. J. Fish Biol. 35: 813–818.
Munk, K.M., and Smikrud, K.M. 2002. Relationships of otolith size to fish size and otolith ages
for yelloweye Sebastes ruberrimus and quillback S. maliger rockfishes. Juneau, Alaska.
Munro, A.R., McMahon, T.E., and Ruzycki, J.R. 2006. Where did they come from? Natural
chemical markers identify source and date of lake trout introduction in Yellowstone Lake.
Yellowstone Sci. 14: 4–12.
59
Neilson, J.D., and Campana, S.E. 2008. A validated description of age and growth of western
Atlantic bluefin tuna (Thunnus thynnus). Can. J. Fish. Aquat. Sci. 65: 1523–1527. doi:
10.1139/F08-127.
Niva, T., Keränen, P., Raitaniemi, J., and Berger, H.M. 2005. Improved interpretation of labelled
fish otoliths: a cost-effective tool in sustainable fisheries management. Mar. Freshw. Res.
56: 705–711.
Nolf, D. 1985. Otolithi piscium . Edited by H.-P. Schultze. Gustav Fischer Verlag, Stuttgart, New
York.
Otterå, H., Kristiansen, T.S., and Svåsand, T. 1998. Evaluation of anchor tags used in sea-
ranching experiments with atlantic cod (Gadus morhua L.). Fish. Res. 35: 237–246.
Oxenford, H.A., Hunte, W., Deane, R., and Campana, S.E. 1994. Otolith age validation and
growth-rate variation in flyingfish (Hirundichthys affinis) from the eastern Caribbean. Mar.
Biol. 118: 585–592.
Pannella, G. 1971. Fish otoliths: daily growth layers and periodical patterns. Science 173: 1124–
1127.
Pannella, G. 1974. Otolith growth patterns: an aid in age determination in temperate and tropical
fishes. In The ageing of fish. Edited by T.B. Bagenal. Unwin Brothers Limited, Surrey,
England. pp. 28–39.
Pannella, G. 1980. Growth patterns in fish sagittae. In Skeletal growth of aquatic organisms.
Edited by D.C. Rhoads and R.A. Lutz. Plenum Press, New York. pp. 519–560.
Pardo, S.A., Cooper, A.B., and Dulvy, N.K. 2013. Avoiding fishy growth curves. Methods Ecol.
Evol. 4: 353–360. doi: 10.1111/2041-210x.12020.
Parmentier, E., Cloots, R., Warin, R., and Henrist, C. 2007. Otolith crystals (in Carapidae):
growth and habit. J. Struct. Biol. 159: 462–473. doi: 10.1016/j.jsb.2007.05.006.
Payan, P., De Pontual, H., Edeyer, A., Borelli, G., Boeuf, G., and Mayer-Gostan, N. 2004.
Effects of stress on plasma homeostasis, endolymph chemistry, and check formation during
otolith growth in rainbow trout (Oncorhynchus mykiss). Can. J. Fish. Aquat. Sci. 61: 1247–
1255. doi: 10.1139/F04-059.
Perez, K.O., and Munch, S.B. 2013. Validating back-calculation models using population data.
Trans. Am. Fish. Soc. 142: 82–94. doi: 10.1080/00028487.2012.728161.
Popper, A.N., and Fay, R.R. 2011. Rethinking sound detection by fishes. Hear. Res. 273: 25–36.
doi: 10.1016/j.heares.2009.12.023.
Popper, A.N., and Lu, Z. 2000. Structure-function relationships in fish otolith organs. Fish. Res.
46: 15–25.
60
Popper, A.N., Ramcharitar, J., and Campana, S.E. 2005. Why otoliths? Insights from inner ear
physiology and fisheries biology. Mar. Freshw. Res. 56: 497–504. doi: 10.1071/MF04267.
Quince, C., Abrams, P.A., Shuter, B.J., and Lester, N.P. 2008a. Biphasic growth in fish I:
theoretical foundations. J. Theor. Biol. 254: 197–206. doi: 10.1016/j.jtbi.2008.05.029.
Quince, C., Shuter, B.J., Abrams, P.A., and Lester, N.P. 2008b. Biphasic growth in fish II:
empirical assessment. J. Theor. Biol. 254: 207–214. doi: 10.1016/j.jtbi.2008.05.029.
Radtke, R.L., Fine, M.L., and Bell, J. 1985. Somatic and otolith growth in the oyster toadfish
(Opsanus tau L.). J. Exp. Mar. Bio. Ecol. 90: 259–275.
Ratkowsky, D.A. 1983. Nonlinear regression modeling. Marcel Dekker, Inc., New York, NY.
Rennie, M.D., Purchase, C.F., Lester, N.P., Collins, N.C., Shuter, B.J., and Abrams, P.A. 2008.
Lazy males? Bioenergetic differences in energy acquisition and metabolism help to explain
sexual size dimorphism in percids. J. Anim. Ecol. 77: 916–926. doi: 10.1111/j.1365-
2656.2008.01412.x.
Reznick, D., Lindbeck, E., and Bryga, H. 1989. Slower growth results in larger otoiths: an
exprimental test with guppies (Poecilia reticulata). Can. J. Fish. Aquat. Sci. 46: 108–112.
Rideout, R.M., Burton, M.P.M., and A, R.G. 2000. Observations on mass atresia and skipped
spawning in northern Atlantic cod, from Smith Sound, Newfoundland. J. Fish Biol. 57:
1429–1440. doi: 10.1006/jfbi.2000.1405.
Rideout, R.M., Rose, G.A., and Burton, M.P.M. 2005. Skipped spawning in female iteroparous
fishes. Fish Fish. 6: 50–72. doi: 10.1111/j.1467-2679.2005.00174.x.
Rideout, R.M., and Tomkiewicz, J. 2011. Skipped spawning in fishes: more common than you
might think. Mar. Coast. Fish. 3: 176–189. doi: 10.1080/19425120.2011.556943.
Rogers, P.H., Popper, A.N., Hastings, M.C., and Saidel, W.M. 1988. Processing of acoustic
signals in the auditory system of bony fish. J. Acoust. Soc. Am. 83: 338–349. Available
from http://www.ncbi.nlm.nih.gov/pubmed/3343448.
Schirripa, M.J., and Goodyear, C.P. 1997. Simulation of alternative assumptions of fish otolith–
somatic growth with a bioenergetics model. Ecol. Modell. 102: 209–223. doi:
10.1016/S0304-3800(97)00057-4.
Secor, D.H., and Dean, J.M. 1989. Somatic growth effects on the otolith – fish size relationship
in young pond-reared striped bass, Morone saxatilis. Can. J. Fish. Aquat. Sci. 46: 113–121.
doi: 10.1139/f89-015.
Secor, D.H., and Dean, J.M. 1992. Comparison of otolith-based back-calculation methods to
determine individual growth histories of larval striped bass, Morone saxath. Can. J. Fish.
Aquat. Sci. 49: 1439–1454.
61
Sharp, D., and Bernard, D.R. 1988. Precision of estimated ages of lake trout from five calcified
structures. North Am. J. Fish. Manag. 8: 367–372. doi: 10.1577/1548-
8675(1988)008<0367.
Shuter, B.J., Jones, M.L., Korver, R.M., and Lester, N.P. 1998. A general, life history based
model for regional management of fish stocks: the inland lake trout (Salvelinus namaycush)
fisheries of Ontario. Can. J. Fish. Aquat. Sci. 55: 2161–2177.
Shuter, B.J., Lester, N.P., LaRose, J., Purchase, C.F., Vascotto, K., Morgan, G., Collins, N.C.,
and Abrams, P.A. 2005. Optimal life histories and food web position: linkages among
somatic growth, reproductive investment, and mortality. Can. J. Fish. Aquat. Sci. 62: 738–
746. doi: 10.1139/F05-070.
Six, L.D., and Horton, H.F. 1977. Analysis of age determination methods for yellowtail rockfish,
canary rockfish, and black rockfish off Oregon. Fish. Bull. 75: 405–414.
Tranquilli, J.A., and Childers, W.F. 1982. Growth and survival of largemouth bass tagged with
Floy anchor tags. North Am. J. Fish. Manag. 2: 184–187. doi: 10.1577/1548-
8659(1982)2<184.
Victor, B.C., and Brothers, E.B. 1982. Age and growth of the fallfish Semotilus corporalis with
daily otolith increments as a method of annulus verification. Can. J. Zool. 60: 2543–2550.
Walker, S.P.W., and McCormick, M.I. 2004. Otolith-check formation and accelerated growth
associated with sex change in an annual protogynous tropical fish. Mar. Ecol. Prog. Ser.
266: 201–212. doi: 10.3354/meps266201.
Warner, K. 1971. Effects of jaw tagging on growth and scale characteristics of landlocked
Atlantic salmon, Salmo salar. J. Fish. Res. Board Canada 28: 537–542.
Williams, B. 1993. Biostatistics. Thompson Press (India) Ltd., New Delhi.
Wright, P.J., Metcalfe, N.B., and Thorpe, J.E. 1990. Otolith and somatic growth rates in Atlantic
salmon parr, Salmo salar L: evidence against coupling. J. Fish Biol. 36: 241–249.
62
Appendices
a) b)
Figure S1. Distribution of gonadal somatic index (GSI) for all individuals caught at varying
times between May (5) and October (10) (1994-2011, inclusive) that had their sex, gonad
weight, and fish weight recorded. The dashed line represents the GSI threshold that
classifies a fish as mature or immature. Those individuals on or below the line are
considered to be immature. The threshold for (a) males and (b) females were visually
determined to be 0.008 and 0.004, respectively.
63
Figure S2. The proportion of the creel lake trout population that were mature between 1994 and
2011, inclusive. The proportion mature refers to the proportion of lake trout that had GSI
values above the thresholds: 0.008 for females and 0.004 for males (see Fig. S1). All fish
used for this figure had their otoliths aged and their sex, gonad weight, and fish weight
recorded. Fish were aged by an experienced technician using sagittal otolith sections. The
large increase in mature proportion between ages 5 and 7 for males, and 6 and 8 for
females indicates that male fish likely mature a year before females. Furthermore, ~50%
of the creel population is mature at age 6 for males and 7 for females, and ~80% is
mature at age 7 for males and 8 for females.
64
a) b)
Figure S3. The relationship between fork length and otolith length for free-living mature (a)
males and (b) females. The distribution of individuals used to create the free-living
curves presented in Figures 10 and 11, respectively, are shown. These relationships were
found to differ significantly only by their intercepts (ANCOVA: F1=4.581, p=0.0335).
The parameter values and correlation coefficients are presented in Table 2.
65
a) b)
Figure S4. The relationship between fork length and age for free-living mature (a) males and (b)
females. The distribution of individuals used to create the free-living curves presented in
Figures 12 and 13, respectively, are shown. These relationships were found to be not
significantly different from each other (ARSS: F1=1.95, p>0.05). The parameter values and
correlation coefficients are presented in Table 3.