ICES Journal of Marine Science: Journal du Conseil Advance Access originally published online on January 8, 2007
ICES Journal of Marine Science: Journal du Conseil 2007 64(2):377-385; doi:10.1093/icesjms/fsl036
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Application and validation of otolith microstructure as a stock identification method in mixed Atlantic herring (Clupea harengus) stocks in the North Sea and western Baltic
1 Danish Institute for Fisheries Research, Department of Marine Fisheries, Charlottenlund Castle, DK-2920 Charlottenlund, Denmark
2 Danish Institute for Fisheries Research, Department of Inland Fisheries, 8600 Silkeborg, Denmark
3 FRS Marine Laboratory Aberdeen, PO Box 101, Victoria Road, Aberdeen AB11 9DB, Scotland, UK
Correspondence to L. A. W. Clausen: tel: +45 33 963364; fax: +45 33 96 3333; e-mail: law{at}dfu.min.dk
Clausen, L. A. W., Bekkevold, D., Hatifield, E. M. C., and Mosegard, H. 2007. Application and validation of otolith microstructure as a stock identification method in mixed Atlantic herring (Clupea harengus) stocks in the North Sea and western Baltic. – ICES Journal of Marine Science, 64: 377–385.Herring (Clupea harengus) populations with different spawning times mix in ICES Division IIIa. For stock assessment, otolith microstructure analysis is used to determine the hatching season of individuals, classifying them into hatch type spring, autumn, or winter. The currently applied method uses visual inspection of season-specific daily increment pattern for the larval period. With this method, variability in individual microstructure and a lack of correspondence between hatch and spawning time may lead to classification error. We validate the visual inspection procedure in relation to these potential sources of error. Otoliths from spawning herring were first classified blindly and the results compared with spawning season. In all, 91% of classifications corresponded, and errors represented misclassifications mainly between autumn and winter spawners. However, the estimates may be biased if hatch and spawning times differ, and an objective method of hatch-time estimation based on linear modelling was employed, enumerating unbroken series of daily increments in 0-group herring hatched in different seasons. Visual inspection and objective estimation agreed in 89% of cases, and most of the errors were explained by overlapping hatch periods. Results show that herring older than the 0-group can be classified using multiple linear regression of hatch time on median increment width.
Keywords: fisheries management, herring assessment, otolith microstructure, stock identification, strayers
Received 18 January 2006; accepted 27 November 2006; advance access publication 8 January 2007.
 |
Introduction |
|---|
 Top Introduction Material and methodsResultsDiscussionReferences |
|---|
Atlantic herring (Clupea harengus) population dynamics are complex, and different stocks often display variation in life history and spawning season (Jennings and Beverton, 1991; McQuinn, 1997a) as well as genetic structuring (Bekkevold et al., 2005; Mariani et al., 2005). Herring perform extensive seasonal migrations between spawning, feeding, and wintering areas (Slotte, 1998), and different stock components often mix on feeding and wintering grounds (Rosenberg and Palmén, 1981; Wheeler and Winters, 1984; Husebø et al., 2005; Ruzzante et al., 2006). Estimation of individual population contributions to these mixed stocks has attracted considerable interest for management purposes (ICES, 2005), because the preservation of complex stock structures necessitates knowledge of how migratory components of various stocks overlap spatially and seasonally.
In the North Sea, management currently recognizes two main stocks: North Sea autumn spawners (NSAS) and winter-spawning Downs herring (ICES, 2004). These populations mix on nursery and feeding grounds in the North Sea as well as in ICES Division IIIa (Cushing, 1967; Rosenberg and Palmén, 1981; Hulme, 1995; Ruzzante et al., 2006). Although meristic characters such as vertebral counts and otolith microstructure to some extent differ between the two groups (Cushing and Bridger, 1966; Hulme, 1995; Mosegaard and Madsen, 1996), little genetic differentiation has been identified between stock components (Mariani et al., 2005). In contrast to the North Sea stock, western Baltic spring spawners (WBSS) comprise several genetically distinct populations spawning in Divisions IIIa, IIIb, and IIIc (Bekkevold et al., 2005), of which the Rügen spawning component is assumed to be the largest and other components of relatively lesser significance. Although there may be population differences (Ruzzante et al., 2006), these three stocks are collectively highly migratory, and NSAS and Downs juveniles as well as adults of WBSS origin migrate into Division IIIa, where they feed in mixed stocks. The large Norwegian spring-spawning (NSS) stock spawning along the west coast of Norway also migrates extensively (Slotte, 1998). However, an extensive literature search has not produced evidence for its migration into Division IIIa (Dragesund et al., 1997; Slotte, 1998; Kvamme et al., 2003; Husebø et al., 2005).
Analysis of otolith microstructure is a powerful tool for determining life history trajectories, determination of hatch season and larval ambient environment being the key proxies for individual population affiliation. Differences in otolith growth trajectories between herring larvae experiencing different temperature and feeding regimes have been identified in both field and laboratory studies (Moksness, 1992; Fossum and Moksness, 1993; Stenevik et al., 1996; Folkvord et al., 1997). Herring larvae hatched at different times of the year in the wild, experiencing different temperature and feeding regimes, display different patterns of primary increments in their otolith. Otolith microstructure has been used to identify larvae from NSAS and NSS stocks (Moksness and Fossum, 1991), and differences in the larval otolith microstructure have been identified too in adult herring (Zhang and Moksness, 1993), and used successfully to separate adult herring from NSAS, Downs, and WBSS spawning stocks at an individual level (Mosegaard and Madsen, 1996; ICES, 2004). For Division IIIa, the ICES Herring Assessment Working Group (HAWG) for the area south of 62°N has applied splitting keys to catches to separate NSAS and Downs herring from WBSS herring. Before 1996, the splitting key used by the HAWG was calculated from a sample-based mean vertebral count. In the period 1996–2001, splitting keys were constructed using information from a combination of vertebral count and otolith microstructure methods (ICES, 2001). From 2001 on, the splitting keys have been constructed solely using otolith microstructure methods (ICES, 2004).
Otolith-based assessment of stock affiliation (Mosegaard and Madsen, 1996) is based on the assumption that herring hatched in a specific season also spawn in that season (known both as the "pure stock concept" and "spawning time fidelity"). However, observations of autumn spawning on traditional spring-spawning sites and the sympatric existence of herring with different spawning times (Brophy and Danilowicz, 2002, 2003; Husebø et al., 2005; Bekkevold et al., in press) make a revisit of these assumptions appropriate now. The current study was initiated to analyse variability in the otolith microstructure pattern in post-larval 0-group herring, hatched during different seasons, to achieve a validation method independent of the assumptions behind the pure stock concept. Formation of the first annual translucent ring in herring otoliths coincides with winter stagnation of growth, and the cessation of daily increment formation (Arneri et al., 1998). The 0-group herring, caught during their first growth period, were chosen according to the assumption that they would exhibit an assessable unbroken series of daily increments from the period after hatching until capture. We use information from larval increment patterns to develop an independent objective validation method that combines backtracking of the date of formation of the first primary increment with measurements of microstructure increment patterns and visual inspection of the larval otolith. Primary increments formed during the larval stage in herring are daily in Norwegian spring spawners (Moksness, 1992), so allowing back-calculation of hatch date by counting the daily increments in a fish from the edge to the centre and adding an estimated initial period with no daily increments (Moksness, 1992). However, it is unknown whether this procedure is valid across populations (Geffen, 1982; Folkvord et al., 2000; Fox et al., 2004).
Our study assesses the currently employed routine of identifying herring from different spawning stocks at an individual level by visually inspecting the larval otolith microstructure in both 0-group and adult spawning herring, using two approaches: (i) evaluating the accuracy of visual inspection, by assessing the extent to which hatch-type classification by visually inspecting otoliths from spawning herring collected in the North Sea (representing NSAS), English Channel (Downs), and the western Baltic (WBSS) corresponds with the respective spawning season of the individual fish, and (ii) evaluating the correspondence between visual hatch-type classification and backtracked hatch date in 0-group herring sampled from a mixed stock in Division IIIa, based on a linear modelling approach that uses objectively measured and enumerated larval otolith microstructure data.
We discuss the application of the results of visually inspecting otolith microstructure as a valid stock separation method in the light of the results from a quantitative objective validation method. We also infer the accuracy of the visual inspection and natural variability of the otolith microstructure methods in terms of the ability to indicate violated assumptions of, for instance, spawning time fidelity.
|
Material and methods |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Validation by visual inspection of spawning herring assuming spawning time fidelity
Ripe-and-running (maturity stage 6) herring were sampled from collections from spawning sites in the North Sea, English Channel, and western Baltic (Table 1, Figure 1). We assumed there were no strays from extant populations with divergent hatching and spawning times in our sample. Otoliths were mounted with the sulcus side up in thermoplastic resin (Buehler Thermoplastic Cement no. 40–8100) at 150°C to facilitate grinding and polishing of both sides. The identity of each individual was coded in such a manner that readers were unable to detect from which population the fish originated. The order of the otoliths was set so that the three possible hatch types (spring, autumn, and winter) appeared randomly. The otoliths were polished using a series of grinding and polishing films with decreasing grain size from 30 to 0.3 µm, to optimize the visual resolution at a focal plane through the otolith's nucleus and a transect from this to the edge.
|
|
Hatch type was estimated for all 697 herring examined. Visual inspection was performed by two experienced readers using a LeicaTM DMLB compound light microscope with objective lenses of 20x and 40 x magnification and a long distance between focus and lens to facilitate viewing of the otolith microstructure through a microscope slide 1.5 mm thick. The readers assigned a hatch type of either spring, autumn, or winter to all otoliths by visual inspection, following the normal laboratory guidelines presented in Table 2. Through experience, the readers have been able to calibrate their perception of distances in the view field of the microscope, such that they know (approximately) the scale of measurement at the two magnifications used. The zone where incremental widths increased from <2 to >2.5 µm was used as a marker for the onset of increased spring growth conditions (Mosegaard et al., 2001). Otoliths considered as unreadable by one or both readers were disregarded for that comparison. The accuracy of the visual inspection of hatch type was calculated as the correspondence between the assigned hatch type and the season in which the fish spawned, assuming that all individuals exhibited spawning time fidelity.
|
Validation by image analysis of otolith microstructure pattern in 0-group herring
A search in the DIFRES database on 0-group herring (fish caught before the onset of the first annual otolith winter ring) from Sub-divisions IIIaN and IIIaS and the transfer area in the North Sea (Figure 1) between 2001 and 2003 was made, and 108 fish were selected from different locations within the area (Table 1, Figure 1).
The otoliths of these herring were retrieved from archives, remounted, and inspected visually following the same procedure as described earlier for otoliths from the spawning populations. After preparation, all otoliths could be classified as autumn, winter, or spring hatch types. Following visual inspection, images of 0-group herring sagittae (n = 108) were digitized; each otolith was analysed by taking several pictures following the longest axis along the postrostrum. Measurements of otolith microstructure were made with a Leica 350 F digital camera and ImageProTM 5.0 image-analysis package for WindowsTM. Increment widths were measured automatically using the Caliper tool in ImagePro, along a profile of grey values and using a profile bandwidth of 10 µm. The Caliper tool was set to identify the onset of an increment as the point at which the grey values changed towards lower values at the fastest rate. The process was monitored by an expert reader, and if the program produced obviously erroneous increments (e.g. because of cracks in the otolith), these were altered manually to fit the real increments. In cases where the increments were not sufficiently clear to be identified by the Caliper tool or by eye, the distance from the last visible increment to the next visible increment was measured. A minimum acceptable increment width was set at 0.5 µm, to filter out the segments where false or no daily rings were visible along the measurement axis. All measurements were transferred to an MS ExcelTM spreadsheet. Areas with no detectable ring structures were occasionally found in the trajectory from the otolith centre to the edge. As these areas would appear as abnormally broad increments but only represent 1 d, a running median value mi = MED(wi–2, ... , wi+2) was applied as a smoother to yield a robust estimate of increment width at distance from centre. This median was then used to estimate duration in days (di) between observed increments wi, as di = wi/mi, independent of whether these were true daily increments or just zones with several unreadable daily increments. A median over five successive increments was enough to screen out all unreadable areas. This was indicated by the fact that no median increment exceeded 7 µm in the first 200 µm from the centre, and no median at all was more than 14 µm wide. Further, only five successive pairs of medians out of 28 300 had more than a 50% change in width between them.
Initial otolith increment position after hatch in herring larvae subjected to suboptimal growth conditions (as in autumn and winter) is sometimes not discernible. Therefore, measurements closer than 10 µm from the centre were disregarded following the findings in Folkvord et al. (2004). In 94% of otoliths, the first measurable increment was formed at a radius (R1) less than 25 µm from the centre. To estimate the number of increments formed in the zone between 10 µm from the centre and R1, otolith initial growth rate was analysed as a function of day of formation. As a measure of early otolith growth rate, the median increment width of the first six measured increments (m) was regressed vs. Julian day number minus 200 at 25 µm from the centre (J) in a quadratic relationship: m = 1.67 + 0.0081J + 0.000037J2 (r2 = 0.75, p < 0.0001, n = 108). The value of –200 is applied to achieve the most monotonic quadratic function and gives the best display of the different seasons' growth patterns. Assuming that increments are daily, this relationship was then used to extrapolate the number of days in the unreadable zone from 10 µm to the first measurable increment [number of days = (R1–10)/m]. The total estimated age at catch was then subtracted from the Julian day of the catch to obtain the Julian day of first possible ring formation (10 µm from the centre), which was then used as a proxy for hatch date, neglecting the possible initial period of very slow otolith growth after hatch (Folkvord et al., 2004). This estimated hatch date, based on counts plus additional zones with an estimated number of increments, is hereafter referred to as the back-calculated hatch date h.
As this method is only applicable to fish with a potentially unbroken series of daily increments, e.g. 0-group herring, we analysed how shorter series (segments) of measured increment widths would estimate h. Owing to variable resolution in the segment 0–15 µm from the centre, this area was neglected in the analysis.
Starting from a distance of 15 µm from the centre, the otolith trajectory was divided into k segments of 30 µm width, and each k segment's median increment width (mz) was used as independent variable in a multiple regression analysis (z = 1 for 15–45 µm; z = 2 for 45–75 µm, etc.). Both original and natural log-transformed values [ln(mz)] were explored. Stepwise regression analysis was performed to obtain a selection of significant coefficients from the total array of coefficients corresponding to all k segments (a, b1, b2, ... , bk). The estimation of hatch date from the multiple regression may be expressed as h = a + 
(bj x f(mj)), where f is the untransformed or ln-transformed median increment width and j is an index of the subset of measured segments giving a significant linear combination for the estimation of hatch date h.
Summer, when very few herring have hatched, constitutes a natural separation between fish hatched in spring and fish hatched in autumn. A good separation was found by letting summer start at Julian date 189–365 = –176. The distribution of h was then analysed using the cumulative frequency distribution from h = –176 to 189. A plot of the data for the 108 herring (Figure 2) suggested the existence of three major aggregations in time (from approximately –150 to –70; –25 to 70; and 90 to 150). Assuming normal distributions of the three clusters, the number of individuals (Nk), the mean (µk) hatch date, and its standard deviation (
k) for each cluster (k) were estimated by the minimum sum of squares (SSQ) method:
|
|
is the cumulative normal distribution with estimated mean µk and standard deviation k, and Nk is the estimated number of individual fish in the kth of the three hatch groups (autumn, winter, or spring).
|
A knife-edge separation of individual hatch season (Saut = 9, Swin = 12, Sspr = 16) was calculated by the highest probability of belonging to a specific period. The ability of f(mz), from the different measured segments, to estimate hatch season Sk was explored using stepwise linear regression, with
= 0.05 for parameters staying in the model. Classification success was compared between analyses of segments in two different otolith areas, dependent on experience in routine preparation of the larval otolith centre in adult herring: (i) when the centre remains intact with visible segments from 15 to 225 µm (z = 1, 2, ... , 7) and (ii) in cases of overgrinding, where usually only increments >300 µm from the centre are visible. |
Results |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Validation by visual inspection of spawning herring assuming spawning time fidelity
The accuracy of visual inspection of hatch type is presented in Table 3. Assuming that herring exhibited spawning time fidelity to hatch date, the accuracy of visual inspection was high with an overall correct classification of 91%. Herring collected ripe-and-running in winter (November–December) were most difficult to classify, with a misclassification rate of 32%, whereas individuals collected in spring (March–June) were of a hatch type determined with the lowest misclassification rate of 3%. The most apparent pattern in the misclassification was that spawning herring from autumn and winter were most frequently confused with each other, whereas spring spawners were assigned equally often to winter or autumn hatch type when misclassified (Table 3).
|
Validation by image analysis of otolith microstructure pattern in 0-group herring
Most otoliths had whole unbroken transects of daily increments from the start of measurement (at an average 20 µm from the centre) to the edge of the otolith (95% of the otoliths had 92% of the transect complete without interruption). The distribution of back-calculated hatch dates is shown in Figure 3, in which the smooth curves are the fitted normal distributions of hatch dates based on the backtracked number of days from catch. The back-calculated hatch dates fell within three well separated groups, winter, spring, and autumn. However, some overlap between groups was evident, especially between the autumn and winter hatch date groups. The periods for the hatching seasons were defined by normal distribution, spring being from 18 February to 9 July, autumn from 9 July to 5 November, and winter from 5 November to 18 February.
|
Visual inspection of 0-group otoliths gave an overall correct classification of 89% when the classification by visual inspection was compared with the back-calculated hatch season of individual fish (Table 4). The misclassification pattern repeated the pattern seen in the pure stock samples, the most frequently confused hatch types being those of autumn and winter. Autumn-hatched herring were classified as winter-hatched in 10% of the fish analysed, and winter-hatched herring were classified as autumn-hatched by visual inspection in 17% of those analysed.
|
The large overlap between autumn and winter hatch seasons and the poor fit of the later autumn-hatched herring (Figure 3), together with the pattern of greater misclassification by visual inspection between autumn- and winter-hatched fish than between either of these two and spring-hatched fish (Table 4) led to further examination of the division between autumn and winter hatching seasons. The seasons were subjectively forced to fixed periods so that the classical start of winter was applied, categorizing winter hatch as from 1 December to 18 February, spring from 18 February to 9 July, autumn from 9 July and 5 November, and late autumn from 5 November and 1 December. Using these four categories, the visual inspection results were re-analysed, revealing that 6% of the misclassified winter hatch types fell within the period late autumn (Table 5). The subjective categorization of hatch season did not affect classification of spring hatch types, whereas classification success of winter hatch types increased.
|
Although three well separated hatch date groups were found, there was a significant within-group difference in mean hatch date for both autumn (p = 0.04) and winter (p = 0.0018) groups among the three sampling years (2001–2003). The development of increment width with distance from the otolith centre is shown for the four hatch types (as determined by increment counts to winter, spring, autumn, and late autumn) in Figure 4. The spring-hatched herring clearly separated from the remaining hatch types by exhibiting increments wider than 2 µm from the beginning of measurements, and the increments continued to increase in width over the whole measurement transect, levelling out at approximately 6 µm at a distance 400 µm from the centre. The increment width development in autumn-, late-autumn-, and winter-hatched herring overlapped, especially in the first part of the measurement transect. At a distance 150 µm from the centre, the three types can be separated, the late-autumn hatch type being on the line of gradual change from autumn to winter type.
|
To aid hatch type determination by visual inspection in these hatch types that are more difficult to separate, a series of segments along a transect from the otolith centre towards the edge was selected, and the separation ability of the increment widths in these segments were tested using a multiple regression analysis.
When median increment widths, mz, from segments 1–7 of the otolith (i.e. the area from 15 to 225 µm from the centre) were analysed, the linear combination Sk = 7.8 + 1.3 x ln(m1) + 1.6 x ln(m5) + 4.8 x ln(m6)–0.74 x m6 (r2 = 0.88) exhibited the best fit (Figure 5). However, a large number of other segment combinations also gave a good prediction of hatch season, with segment 5 (135–165 µm from the centre) often showing up as the major influence. When segments with k > 10 were analysed, the best combination was Sk = 1.6 + 2.6 x ln(m11)+3.0 x ln(m14) + 0.45 x m20 (r2 = 0.76), with segment 11 (315–345 µm from the centre) generally having the highest influence in different combinations.
|
|
Discussion |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
To ensure conservation of herring population diversity in the North Atlantic, all stock components and their natural migration patterns should be considered in compiling scientific advice on the fishery (Stephenson, 2001). Therefore, obtaining high levels of precision in the input data to the assessment of mixed stocks is warranted. The good agreement between the hatch type assigned by visual inspection and the sampling season of spawning herring observed in this study confirms visual inspection of larval otolith microstructure in spawning herring as a valid method of discriminating between hatch types. However, despite the high level of correspondence between assigned hatch type and spawning season, some variation was observed among the readers' classification results. A lack of correspondence between estimated hatch and spawning season was mainly discriminating between autumn- and winter-spawned herring. An explanation for the misclassifications may be found in a possible straying of fish not exhibiting spawning time fidelity (McQuinn, 1997b; Slotte, 1998, 2001). For studies of population structuring, the identification of fish straying among populations with different spawning seasons is of focal interest, but their potential existence also presents a problem concerning validation of the otolith microstructure method. The otolith microstructure of straying fish may not be detected because the apparent variability of the otolith microstructure may be too high to allow detection of the phenomenon. If, on the other hand, a specific hatch period gives rise to highly variable otolith microstructure, some fish may be falsely identified as originating from populations exhibiting different spawning seasons. The misclassification rates observed in this work were higher when gonad stage was used as an indicator of spawning time, than when using backtracked hatch dates (Tables 3 and 4). This could suggest some spawning season straying. However, it could also be related to a natural variability in the larval otolith microstructure formed after hatch, as well as potentially overlapping spawning seasons.
The variation between reader-assigned hatch type and gonad stage indicated that spawning time may be affected by both within- and among-reader variation, influenced perhaps by insufficient training and/or lack of quality control during the reading process. This issue was not investigated in our study. However, both readers in this study have been tested already for reader consistency and exhibit good correspondence between hatch type determination (ICES, 2005).
The quantitative objective classification method of hatch types in 0-group herring developed here provides an opportunity to calibrate the visual inspection of hatch types in all herring. Inclusion of possible variation in environmental influence on otolith microstructure allows for variability in the pattern within each hatch type. The underlying assumption for this approach is that primary increments are sufficiently close to being daily for back-calculation of hatch season in 0-group herring to be possible. A further development of the method assumes that increment measurements along radii at specific distances from the otolith core reflect season- and area-specific environmental conditions during the larval growth phase, so permitting the use of otolith microstructure patterns to identify offspring from different spawning populations.
Although it was not intended to estimate the absolute hatch date, but rather to estimate the hatch season with reasonable accuracy, the validation technique presented here has two important prerequisites: knowledge of the timing of formation of the first daily increment and of the successive daily deposition of micro-increments in the larval otolith. The formation of the first discernible daily increment in herring larvae coincides with the onset of first feeding at the start of post-yolk-sac growth (Moksness, 1992; Høie et al., 1999). This takes place in herring around 10–19 d from hatching, depending on the population in question (Fox et al., 2004). However, growth rate and temperature strongly influence the formation of the first discernible increment (Høie et al., 1997; Folkvord et al., 2000; Pavlov et al., 2000; Fox et al., 2004). Folkvord et al. (2004) found no increase in size of sagittae from herring larvae reared at 4°C up to 30 d, whereas herring reared at 12°C showed sagittal growth after 9 d. We calculated initial undetectable increment widths by a general curvilinear relationship from the fish with clearest otolith patterns. However, otolith no-growth under abnormal environmental conditions could not be detected using our methods. Adding a variable number of days to the counts of daily increments, depending on some uncertain environmental forcing, would make calculation of absolute age more uncertain than necessary for our purpose. In this study, it is likely that the ages of the winter hatch type herring were underestimated, because these fish would have experienced the lowest post-hatch temperatures of the three hatch types. However, estimates for fish hatched during autumn could also have been lower than the actual ages, depending on specific hatch time and annual variation in temperature.
The formation of daily increments in embryonic stages of herring has not been confirmed (McGurk, 1984; Moksness et al., 1987). However, for stages following the absorption of the yolk sac, otolith micro-increments are formed on a daily basis (McGurk, 1987; Moksness and Wespestad, 1989; Moksness and Fossum, 1991; Moksness, 1992). Growth rate, however, seems to influence the deposition of daily increments. Several studies have demonstrated non-daily increment deposition in herring larvae with a growth rate <0.4 mm d–1 (Geffen, 1982; McGurk, 1984; Folkvord et al., 2000; Pavlov et al., 2000; Fox et al., 2004). Autumn-spawned herring larvae may exhibit growth rates below this value (Munk and Christensen, 1990; Johannesen et al., 2000), and it is therefore likely that non-daily rates of micro-increment formation in these fish could lead to underestimating the absolute age, as seen in Feet et al. (2002) and Fox et al. (2004). Notwithstanding these uncertainties, the prerequisites of the validation approach in this study are fulfilled to the extent necessary for hatch type estimation, because the intention was to place the herring within a spawning season, not to estimate their precise hatch date. Despite the possibility of underestimation of absolute age in the 0-group herring in this study, the back-calculated hatch date distributions confirm the trimodal distribution of the peak periods of spawning in winter, spring, and autumn (Figure 3).
The overlapping seasons of the autumn- and winter-spawning herring (Zijlstra, 1969; Burd and Howlett, 1974) and the gradual change in otolith microstructure from autumn hatch type through late-autumn hatch type to winter hatch type identified here appeared to result in classification of late-hatched autumn spawners as winter hatch types by visual inspection, following the guidelines currently applied. Comparing the visual inspection results in Tables 4 and 5, it is clear that most autumn-hatched herring misclassified using the classic division of hatch seasons (Table 4) were represented by fish hatched late in autumn. Therefore, visual inspection of hatch types may fail to classify individuals hatched in the periods of overlapping spawning seasons. This has potential consequences for historically splitting catches between winter and autumn hatch types for herring caught in the Skagerrak and Kattegat. However, recent efforts to separate the winter-spawning Downs component from autumn-spawning components in the North Sea (ICES, 2005) may not be affected by this problem. Downs herring otolith microstructure has been reported to be >2 µm wide at a distance of 100 µm from the nucleus (Mosegaard and Madsen, 1996), more than observed for winter-hatched herring in the present study. Consequently, it is possible that winter-hatched herring analysed in this study (i.e. the archived 0-group herring from Sub-area IIIa) did not originate from the Downs population, but from a different population of winter spawners, e.g. from the western Baltic (Bekkevold et al.in press).
NSS herring daily increments start at a mean increment width of 1.5 µm at the centre and gradually increase to 2.5 µm around 100 µm from the centre (Figure 2 in Husebø et al., 2005). This pattern is different from that observed for spring spawners in the present study, for which the mean increment width was 3 µm at 25 µm from the centre, increasing to >4 µm at 100 µm from the centre (Figure 4). The autumn spawners analysed by Husebø et al. (2005) also showed an increment development pattern different from that in this study.
The appearance of the otolith microstructure is much influenced by the environmental conditions, such as temperature (Folkvord et al., 2004) and food availability (Johannessen et al., 2000), experienced in the first larval phase, so caution is necessary if environmental regimes in the spawning areas change over time. The observed separation of hatch type in the present study was performed on 0-group herring sampled from year classes 2001, 2002, and 2003, and because the sampling year does have an effect on the pattern of otolith microstructure within each hatch type, a more comprehensive analysis is needed before extrapolation to other year classes can be made. This suggests the need for an annual analysis of 0-group herring otolith microstructure to update the calibration criteria for separation of hatch type by visual inspection of otolith microstructure.
In addition to the otolith microstructure patterns formed during the larval period, otolith increments formed during the juvenile growth phase may also be used to identify offspring from different spawning seasons. When the calibration sample is sufficiently large, it is possible to select a subset of segments and apply those to classify hatch types, applying a multiple linear regression model as demonstrated in the present study (Figure 5). Measurements of such defined segments provide quality checks during routine visual inspection and can aid as an additional tool to visual inspection when overgrinding of an otolith precludes application of the routine method.
The objective separation method based on median increment width of segments of otolith microstructure in the juvenile growth phase validates the use of visual inspection for hatch type separation of both juvenile and adult herring. However, it is an improvement on the visual inspection method in two ways. First, the objectivity increases the reliability of hatch type estimation of readers regardless of experience and precision level. Second, the dampening down of the inherent natural variability in otolith microstructure patterns within each hatch type by using median measurements in segments reduces misclassification errors.
As the quality and precision of hatch type estimates determined by visual inspection has depended on individual skills and experience, the need for standardization, objective control, and statistical evaluation is obvious in improving the reliability of the output. The method developed here facilitates an objective determination of hatch type, which makes standardization and quality assurance and quality control less complicated.
|
Acknowledgements |
|---|
We thank Meinhard Poulsen, Mikael van Deurs, Agnete Hedegaard, and Stina Bjørk Stenersen Hansen for their help with the laboratory work. The project was partly funded by the EU project HERGEN (QLRT–2000–01370).
|
References |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
-
Arneri E., Mosegaard H., Wright P. J., Morales-Nin B. (1998) Microstructural validation of annual increments. In Wright P. J. (Ed.). The Present Status of Otolith Research and Applications pp. 54–59 EFAN report1/98.
Bekkevold D., Andre C., Dahlgren T. G., Clausen L. A. W., Torstensen E., Mosegaard H., Carvalho G. R., et al. (2005) Environmental correlates of population differentiation in Atlantic herring. Evolution 59:2656–2668.[CrossRef][Web of Science][Medline]
Bekkevold D., Clausen L. A. W., Mariani S., André C., Christensen T. B., Mosegaard H. Divergent origins of sympatric herring population components determined using genetic mixture analysis. Marine Ecology Progress Series in press.
Brophy D. and Danilowicz B. S. (2002) Tracing populations of Atlantic herring (Clupea harengus L.) in the Irish and Celtic Seas using otolith microstructure. ICES Journal of Marine Science 59:1305–1313.
Brophy D. and Danilowicz B. S. (2003) The influence of pre-recruitment growth on subsequent growth and age at first spawning in Atlantic herring (Clupea harengus L.). ICES Journal of Marine Science 60:1103–1113.
Burd A. C. and Howlett G. J. (1974) Fecundity studies on North Sea herring. Journal du Conseil International pour l'Exploration de la Mer 35:107–120.
Cushing D. H. (1967) The grouping of herring populations. Journal of the Marine Biological Association of the UK 47:193–208.
Cushing D. H. and Bridger J. P. (1966) The stock of herring in the North Sea, and changes due to fishing. Fishery Investigations London 123 Series 2, 25.
Dragesund O., Johannesen A., Ulltang Ø. (1997) Variation in migration and abundance of Norwegian spring spawning herring (Clupea harengus L.). Sarsia 82:97–105.[Web of Science]
Feet P. Ø., Ugland K. I., Moksness E. (2002) Accuracy of age estimates in spring spawning herring (Clupea harengus L.) reared under different prey densities. Fisheries Research 56:59–67.[CrossRef]
Folkvord A., Blom G., Johannessen A., Moksness E. (2000) Growth-dependent age estimation in herring (Clupea harengus L.) larvae. Fisheries Research 46:91–103.[Medline]
Folkvord A., Johannessen A., Moksness E. (2004) Temperature-dependent otolith growth in Norwegian spring-spawning herring (Clupea harengus L.) larvae. Sarsia 89:297–310.[CrossRef][Web of Science]
Folkvord A., Rukan K., Johannessen A., Moksness E. (1997) Early life history of herring larvae in contrasting feeding environments determined by otolith microstructure analysis. Journal of Fish Biology 51:Suppl. A, 250–263.[CrossRef]
Fossum P. and Moksness E. (1993) A study of spring-and autumn-spawned herring (Clupea harengus L.) larvae in the Norwegian coastal current during spring 1990. Fisheries Oceanography 2:73–81.
Fox C. J., Folkvord A., Geffen A. J. (2004) Otolith micro-increment formation in herring Clupea harengus larvae in relation to growth rate. Marine Ecology Progress Series 264:83–94.
Geffen A. J. (1982) Otolith ring deposition in relation to growth rate in herring (Clupea harengus) and turbot (Scophthalmus maximus) larvae. Marine Biology 71:317–326.[CrossRef]
Hulme T. J. (1995) The use of vertebral counts to discriminate between North Sea herring stocks. ICES Journal of Marine Science 52:775–779.
Husebø Å., Slotte A., Clausen L. A. W., Mosegaard H. (2005) Mixing of populations or year class twinning in Norwegian spring spawning herring? Marine and Freshwater Research 56:763–772.[CrossRef]
Høie H., Folkvord A., Johannessen A. (1999) Maternal, paternal and temperature effects on otolith size of young herring (Clupea harengus L.) larvae. Journal of Experimental Marine Biology and Ecology 234:167–184.[CrossRef]
ICES. (2001) Report of the Herring Assessment Working Group for the Area South of 62°N (HAWG). ICES Document CM 2001/ACFM: 12:546.
ICES. (2004) Report of the Herring Assessment Working Group for the Area South of 62°N (HAWG). ICES Document CM 2004/ACFM: 18:548.
ICES. (2005) Report of the Herring Assessment Working Group for the Area South of 62°N (HAWG). ICES Document CM 2005/ACFM: 16:548.
Jennings S. and Beverton R. J. H. (1991) Intraspecific variation in the life history tactics of Atlantic herring (Clupea harengus L.). ICES Journal of Marine Science 48:117–125.
Johannessen A., Blom G., Folkvord A. (2000) Differences in growth between spring and autumn spawned herring (Clupea harengus L.) larvae. Sarsia 85:461–466.
Kvamme C., Noettestad L., Fernø A., Misund O. A., Dommasnes A., Axelsen B. E., Dalpadado P., et al. (2003) Migration patterns in Norwegian spring-spawning herring: why young fish swim away from the wintering area in late summer. Marine Ecology Progress Series 247:197–210.
McGurk M. D. (1984) Ring deposition in the otoliths of larval pacific herring, Clupea harengus pallasi. Fishery Bulletin US 82:113–120.
McGurk M. D. (1987) Age and growth of Pacific herring larvae based on length-frequency analysis and otolith ring number. Environmental Biology of Fishes 20:33–47.
McQuinn I. H. (1997a) Metapopulations and the Atlantic herring. Reviews in Fish Biology and Fisheries 7:297–329.
McQuinn I. H. (1997b) Year-class twinning in sympatric seasonal spawning populations of Atlantic herring. Clupea harengus 95:126–136 Fishery Bulletin US.
Mariani S., Hutchinson W. F., Hatfield E. M. C., Ruzzante D. E., Simmonds E. J., Dahlgren T. G., Andre C., et al. (2005) North Sea herring population structure as revealed by microsatellite analysis. Marine Ecology Progress Series 303:245–257.
Moksness E. (1992) Validation of daily increments in the otolith microstructure of Norwegian spring-spawning herring (Clupea harengus L). ICES Journal of Marine Science 49:231–235.
Moksness E., Butler J., Radtke R. L. (1987) Estimation of age and growth rate in Norwegian spring spawning herring (Clupea harengus L.) larvae and juveniles. Sarsia 72:341–342.
Moksness E. and Fossum P. (1991) Distinguishing spring- and autumn-spawned larvae (Clupea harengus L.) by otolith microstructure. ICES Journal of Marine Science 48:61–66.
Moksness E. and Wespestad V. (1989) Ageing and back-calculating growth rates of Pacific herring, Clupea pallasii, larvae by reading daily otolith increments. Fishery Bulletin US 87:509–513.
Mosegaard H., Clausen L. A. W., Lindberg M. (2001) Manual on herring otolith microstructure preparation and interpretation for stock identification. Kattegat and SW Baltic8 DIFRES manual produced under the EC study 98026: a new sampling regime for resource assessment of herring in the Skagerrak.
Mosegaard H. and Madsen K. P. (1996) Discrimination of mixed herring stocks in the North Sea using vertebral counts and otolith microstructure. ICES Document CM 1996/H: 17.
Munk P. and Christensen V. (1990) Larval growth and drift pattern and the separation of herring spawning groups in the North Sea. Journal of Fish Biology 37:135–148.[CrossRef]
Pavlov D. A., Moksness E., Burmenski V. A. (2000) Otolith microstructure characteristics in White Sea spring-spawning herring (Clupea pallasi marisalbi Berg) larvae. ICES Journal of Marine Science 57:1069–1076.
Rosenberg R. and Palmén L. E. (1981) Composition of herring stocks in the Skagerrak–Kattegat and the relation of these stocks with those of the North Sea and Adjacent waters. Fisheries Research 1:83–104.
Ruzzante D. E., Mariani S., Bekkevold D., André C., Mosegaard H., Clausen L. A. W., Dahlgren T. G., et al. (2006) Biocomplexity in a highly migratory pelagic marine fish, Atlantic herring. Proceedings of the Royal Society of London 273: pp. 1459–1464 Series B.
Slotte A. (1998) Spawning migration of Norwegian spring spawning herring (Clupea harengus L.) in relation to population structure. (Department of Fisheries and Marine Biology, University of Bergen, Norway)186 DSc thesis.
Slotte A. (2001) Factors influencing location and time of spawning in Norwegian spring spawning herring: an evaluation of different hypotheses. In Funk F., Blackburn J., Hay D., Paul A. J., Stephenson R., Toresen R., Witherell D. (Eds.). Herring: Expectations for a New MillenniumUniversity of Alaska Sea Grant College Program pp. 255–278 AK–SG01–04. 789pp.
Stenevik E. K, Fossum P., Johannessen A., Folkvord A. (1996) Identification of Norwegian spring spawning herring (Clupea harengus L.) larvae from spawning grounds off western Norway applying otolith microstructure analysis. Sarsia 80:285–292.
Wheeler J. P. and Winters G. E. (1984) Homing of Atlantic herring (Clupea harengus harengus) in Newfoundland waters as indicated by tagging data. Canadian Journal of Fisheries and Aquatic Sciences 41:108–117.
Zhang Z. and Moksness E. (1993) A chemical way of thinning otolith of adult Atlantic herring (Clupea harengus) to expose the microstructure in the nucleus region. ICES Journal of Marine Science 50:213–217.
Zijlstra J. J. (1969) On the racial structure of North Sea Autumn spawning herring. Journal du Conseil International pour l'Exploration de la Mer 33:67–80.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. J. Geffen Advances in herring biology: from simple to complex, coping with plasticity and adaptability ICES J. Mar. Sci., September 1, 2009; 66(8): 1688 - 1695. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Folkvord, H. Hoie, A. Johannessen, and T. Solbakken Effects of prey concentration, light regime, and parental origin on growth and survival of herring larvae under controlled experimental conditions ICES J. Mar. Sci., September 1, 2009; 66(8): 1702 - 1709. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Burke, D. Brophy, P.-J. Schon, and P. A. King Temporal trends in stock origin and abundance of juvenile herring (Clupea harengus) in the Irish Sea ICES J. Mar. Sci., September 1, 2009; 66(8): 1749 - 1753. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Johannessen, L. Nottestad, A. Ferno, L. Langard, and G. Skaret Two components of Northeast Atlantic herring within the same school during spawning: support for the existence of a metapopulation? ICES J. Mar. Sci., September 1, 2009; 66(8): 1740 - 1748. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. Payne, L. W. Clausen, and H. Mosegaard Finding the signal in the noise: objective data-selection criteria improve the assessment of western Baltic spring-spawning herring ICES J. Mar. Sci., September 1, 2009; 66(8): 1673 - 1680. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Burke, D. Brophy, and P. A. King Otolith shape analysis: its application for discriminating between stocks of Irish Sea and Celtic Sea herring (Clupea harengus) in the Irish Sea ICES J. Mar. Sci., December 1, 2008; 65(9): 1670 - 1675. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||