Embed Size (px)
Citation preview
This article was downloaded by: [Temple University Libraries]On: 21 November 2014, At: 19:20Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
New Zealand Journal of Marine andFreshwater ResearchPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/tnzm20
Age validation, growth and populationcharacteristics of greenback flounder(Rhombosolea tapirina) in a largetemperate estuaryJason Earlab, Anthony J Fowlerb, Qifeng Yeb & Sabine Dittmanna
a School of Biological Sciences, Flinders University, Adelaide,Australiab South Australian Research and Development Institute (AquaticSciences), Adelaide, AustraliaPublished online: 16 Jan 2014.
To cite this article: Jason Earl, Anthony J Fowler, Qifeng Ye & Sabine Dittmann (2014) Agevalidation, growth and population characteristics of greenback flounder (Rhombosolea tapirina) ina large temperate estuary, New Zealand Journal of Marine and Freshwater Research, 48:2, 229-244,DOI: 10.1080/00288330.2013.875928
To link to this article: http://dx.doi.org/10.1080/00288330.2013.875928
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.
This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
RESEARCH ARTICLE
Age validation, growth and population characteristics of greenback flounder(Rhombosolea tapirina) in a large temperate estuary
Jason Earla,b*, Anthony J Fowlerb, Qifeng Yeb and Sabine Dittmanna
aSchool of Biological Sciences, Flinders University, Adelaide, Australia; bSouth Australian Research andDevelopment Institute (Aquatic Sciences), Adelaide, Australia
(Received 22 July 2013; accepted 3 December 2013)
The greenback flounder (Rhombosolea tapirina) is an important fishery species in southern Australiaand New Zealand, whose demographic processes are poorly understood. This study developed anageing protocol based on otolith interpretation and provided estimates of age-based data for an ex-ploited population in the estuary of Australia’s largest river system. The otoliths fulfilled three criteriathat established their usefulness for ageing. Estimates of size and age were used to generate vonBertalanffy growth curves for male and female fish sampled from two areas within the estuary. Therewere no differences in growth between the sexes, despite the strong bias in number towards femalefish. Spatial differences in growth were marginal. Populations involved only a few young age classes.Such truncation may relate to the fishery, to movement of older fish out of the system, or to recentpoor environmental conditions limiting spawning and early life survivorship.
Keywords: Pleuronectidae; flatfish; demography; estuary; otoliths; ageing; validation; Coorong;Murray River
Introduction
Fish population dynamics are driven by a range ofdemographic processes that aremost effectivelymeas-ured using fish age as a time reference (Campana2001). Themost convenient and effective method forestimating fish age is through the direct interpretationof hard, calcified anatomical structures that demons-trate an incremental structure (Beckman & Wilson1995; Campana 2001). Fish otoliths have proven tobe the most effective structures for this purpose(Campana & Thorrold 2001). An otolith grows thro-ugh the regular deposition of newmaterial to its grow-ing surface, forming an optical pattern that reflectsseasonal metabolic processes of the fish which is notsubsequently resorbed or reworked (Campana 1999).Whilst otolith validation studies for a broad range oftaxa from different environments have demonstratedannual periodicity of increment formation (Choatet al. 2009), the method of otolith examination can
influence the accuracy of the interpretation and theresulting estimate of age (Fowler 2009). Furthermore,since the clarity of increments varies amongst taxaand because otolith interpretation is a skill that is sub-ject to human error (Morison et al. 2005), it isrecommended that considerable effort be spent inestablishing a fish ageing protocol, particularly whenthe target species is vulnerable to over-exploitation byfishing (Choat et al. 2009).
For a fish species to be suitable for otolith-based age determination, its otoliths must fulfilthree criteria (Fowler 1990): 1. they must displayan incremental structure that can be interpreted vis-ually; 2. growth increments must be formed on aregular, determinable time scale; and 3. otolithsmust continue to grow throughout the lives of thefish. Prior to the 1980s, most ageing studies werebased on interpreting the structure of wholeotoliths (Beamish 1992). Since then, thin sections
*Corresponding author. Email: [email protected]
New Zealand Journal of Marine and Freshwater Research, 2014Vol. 48, No. 2, 229–244, http://dx.doi.org/10.1080/00288330.2013.875928
© 2014 The Royal Society of New Zealand
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
of otoliths have been used because differentiatingclosely spaced increments near the outer edge ofwhole otoliths can be difficult due to the wayotoliths grow (Mann-Lang & Buxton 1996; Fowler2009). Nonetheless, studies for some species thathave compared the usefulness of both methodsindicated that whole otoliths provided reliable ageestimates (Williams et al. 2003, 2005). As such,determining the most appropriate ageing protocolfor a particular species requires undertaking acomparison of the interpretability of whole andsectioned otoliths.
Validated age and growth information for tem-perate flatfish is generally limited to species in thenorthern hemisphere (Nash & Geffen 2005). Thegreenback flounder (Rhombosolea tapirina) isthe most common pleuronectid in Australian waters(van den Enden et al. 2000). This species is distri-buted in estuaries and coastal waters to depths of100 m along the southern coast of Australia andin New Zealand (Gomon et al. 2008), where itsupports commercial and recreational fisheries(Kailola et al. 1993; Froese & Pauly 2011). InSouth Australia, almost all catches are taken fromthe Murray River estuary and Coorong (hereafterreferred to as the Coorong estuary) (Ferguson2007). This large estuarine system lies at theterminus of Australia’s largest river system, theMurray-Darling. Long-term statistics for thisfishery indicate extreme interannual and spatialvariation in population biomass and abundance(Ferguson 2007). The processes driving this vari-ation remain poorly understood due to a superfi-cial understanding of the population dynamics ofthe species.
Information on the population dynamics of R.tapirina is limited. In southeastern Australia, R.tapirina attain sexual maturity at approximately203 mm total length (Ye et al. 2013), whilst repro-duction occurs during the cooler months fromMarch to October in both estuaries and offshoreenvironments (Kurth 1957; Crawford 1984; Bar-nett & Pankhurst 1999). Post-larval recruits settleout into shallow, unvegetated sandy habitats bet-ween July and October (May & Jenkins 1992).However, there is no understanding of the demo-graphic processes that operate beyond these early
life stages. Whilst unvalidated estimates of age,based on counts of opaque zones in the otoliths ofadult R. tapirina, suggested a maximum age of 10years, several nearshore populations in New Zealandand Tasmania consisted mostly of fish less than3 years of age (Kurth 1957; Sutton et al. 2010).There is uncertainty about these age estimatesbecause of the lack of a validated ageing protocol.The development of such a protocol and providingestimates of age and growth for R. tapirina are es-sential to better understand the population dynam-ics and demography of the species.
The present study addressed the lack of demo-graphic information for R. tapirina by providingan assessment of the usefulness of its otoliths forageing and estimates of age-based data for anexploited population in the Coorong estuary inSouth Australia. The specific objectives addressedwere: 1. to assess the usefulness of whole andtransverse sections of the otoliths of R. tapirinafor ageing by determining whether they fulfil thethree criteria described above, and to develop analgorithm for age determination in months; 2. tocompare growth between male and female fish,and to determine whether growth varies spatially;and 3. to describe the population characteristics interms of size and age structures from two areas inthe Coorong estuary, and to determine how thesestructures varied over a 3-year period.
Materials and methods
Study area
This study was done in the Coorong estuary, SouthAustralia (Fig. 1). This estuarine system is a long(140 km), narrow lagoon that extends southeastfrom the river mouth and represents one of thelargest estuarine habitats in Australia. Due todrought conditions, freshwater inflows to the sys-tem were minimal from 2002 to 2010, with noflows between 2007 and 2009. The lack of flowscontributed to an extensive longitudinal salinitygradient from the mouth of the estuary (Webster2010). Typically, salinity ranged from marine (30–36 g L −1) near the mouth, to hypersaline (> 120 gL −1) in the southern parts of the Coorong
230 J Earl et al.
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
(Kingsford et al. 2011). In late 2010, freshwaterinflows recommenced and continued throughout2011. These flows reduced salinity to mostly fresh(< 4 g L −1) in the vicinity of the mouth, whilehypersalinity (> 90 g L −1) persisted in the south-ern Coorong. The estuarine system supports shal-low (< 8 m), sandy habitats, with soft muddysediments in deeper areas (Gilbertson 1981). Amultispecies, commercial fishery operates through-out the system and targets several finfish spec-ies, including R. tapirina.
Sample collection and processing
In total, 851 R. tapirina were collected betweenDecember 2008 and November 2011 from twostudy areas. One study area, the Estuary, was loc-ated near the Murray mouth and included allwaters downstream of the Goolwa Barrage to theEwe Island Barrage (Fig. 1). The other, NorthCoorong, extended from the Ewe Island Barrage
to Long Point. Fish were collected from the com-mercial fishing sector which targets R. tapirinawith 4.75′′ mesh gill nets. Such nets are size-selective and rarely capture fish < 220 mm and >360 mm total length (TL). Scientific sampling wasalso done using 3.5′′ mesh haul nets to capture abroader size range of fish (i.e. > 150 mm TL).Sampling effort and catch rates varied throughoutthe study period (Table 1), but fish were collectedin most months. For the Estuary, fish weresampled regularly between December 2008 andMay 2011, while for North Coorong samplingoccurred between June 2009 and November 2011(Table 1).
Most fish were processed within 8 h of cap-ture. Each fish was measured for TL to the nearestmm and weighed to the nearest 0.01 g. The go-nads were examined to determine the sex of thefish. The sagittae (i.e. the largest pair of otoliths)were removed, cleaned, dried and stored. Later,the left otolith from each fish was prepared for
Figure 1 Map of the Coorong estuary, showing the two study areas: the Estuary and North Coorong. Inset showslocation of study region relative to Australia. The locations of the barrages that separate the estuarine system fromLake Alexandrina are indicated by black lines.
Age validation, growth and population characteristics of greenback flounder 231
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
ageing. It was embedded in polyester resin andthen a transverse section (TS) of approximately400 µm thickness that incorporated the core wasremoved using a low speed diamond saw (StruersAccutom) (Fowler & McGlennon 2011). The sec-tion was glued to a labelled glass microscope slide.
Usefulness of the otoliths of R. tapirina forageing
Internal structure of otoliths
To determine the presence and interpretability ofincremental structure in the otoliths (criterion 1),and to develop a protocol for ageing this species,whole and TS of otoliths from the same fish wereexamined independently. The TS of each otolithwas smeared with immersion oil (Scharlau ©) andexamined under transmitted light using a dissect-ing microscope. The opaque zones surroundingthe core were counted. On a different day, thewhole right otolith from the same fish wereimmersed in water, with the distal surface up-wards, and examined with a dissecting microscopeusing reflected light, and the opaque zones werecounted. Counts from whole and sectioned oto-liths were compared to assess the interpretabilityof the otolith structure using both methods and toidentify the preferred method for age determina-tion. Counts were compared by constructing anage-bias plot, whereby the mean count fromwhole otoliths was plotted for groups of otoliths
that contained the same number of opaque zonesin the TS. Counts were also compared by calcu-lating the average percent error (APE) for eachtechnique (Beamish & Fournier 1981). Low val-ues of APE indicated higher precision. Otolithsthat did not display an interpretable structure werenot used for comparison, or later for age deter-mination. Overall, the otolith pairs from 627 fishwere examined, of which 21 (3.4%) were rejectedas unreadable.
Timing of opaque zone formation
To determine if the structure in the otoliths ofR. tapirina formed on a regular, determinable timescale (criterion 2), the timing of formation ofopaque zones was investigated. In both 2009 and2010, young-of-the-year R. tapirina were col-lected monthly. Their otoliths were examined todetermine when the first opaque zone formed,using edge analysis. The TS of these otoliths wereexamined with a dissecting microscope using trans-mitted light to differentiate those otoliths with:1. a translucent edge; 2. an opaque edge; and 3. acompleted opaque zone with recently formedtranslucent material on the edge. The percentageoccurrence for each edge type was calculated foreach month.
For the otoliths of older R. tapirina, marginalincrement analysis was done. Where possible, fishwere collected in each month between January
Table 1 Summary of the characteristics of the samples of R. tapirina from the Estuary and North Coorong that wereconsidered for age determination and for analysis of population characteristics (n = 851).
Area Sampling period Sample size (n) Size range (mm) Mean TL (± SD) (mm)
Estuary Dec 08–May 09 104 162–359 276 (± 56)Jun 09–Nov 09 124 157–385 286 (± 34)Dec 09–May 10 142 171–336 262 (± 46)Jun 10–Nov 10 177 185–350 247 (± 28)Dec 10–May 11 22 231–313 267 (± 21)
North Coorong Jun 09–Nov 09 24 271–357 314 (± 26)Dec 09–May 10 127 186–327 271 (± 33)Jun 10–Nov 10 77 157–321 228 (± 45)Dec 10–May 11 24 154–331 225 (± 41)Jun 11–Nov 11 30 291–336 307 (± 10)
232 J Earl et al.
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
2009 and January 2011 and were pooled acrossthe two study areas. Marginal increment analysiscould only be applied to formation of the secondopaque zone, as older fish were not sampled insufficient numbers. Each otolith was examinedwith a dissecting microscope using transmittedlight, and the distances between the outer edgeof the opaque core and the outer edge of the firstopaque zone, and then to the edge of the otolith,were measured towards the proximal surface alongthe ventral side of the sulcal groove (Fig. 2). Themarginal increment for each otolith was expressedas a proportion of the immediately precedingannulus and these were plotted for each monthof the year. The appearance of the otolith marginwas recorded as either opaque or translucent. Allotoliths were read without reference to fish sizeand the time of year the fish was captured.
Growth of otoliths
To determine if otoliths grew continually through-out the lives of the fish (criterion 3), two measuresof otolith size were recorded and related to thenumber of opaque zones. Whole otoliths from 280fish were weighed to the nearest 0.01 mg and thensectioned. For each TS, the otolith thickness (i.e.the distance from the centre of the otolith core tothe proximal surface) was measured (Fig. 2). The
estimates of otolith size were compared to the num-ber of opaque zones in the TS using linear regres-sion. The statistical program SPSS V. 19 was usedfor these statistical procedures.
Development of an ageing protocol
To provide an estimate of age in months for eachfish, an algorithm was developed that consideredthe periodicity of opaque zone formation, the dateof capture, and the universal birthdate of 1 June(i.e. the middle of the reproductive season; Crawford1984; Barnett & Pankhurst 1999). The algorithmwasbased on the results of edge analysis and marginalincrement analysis which identified the timing offormation of the first and second opaque zones.
Growth
Estimates of length-at-age for male and female R.tapirina were used to describe growth using vonBertalanffy growth curves: Lt = L∞ [1− e−K(t−t0)],where Lt = length at age, L8 = theoretical max-imum or asymptotic length that fish would reachif they lived indefinitely, K = curvature parameter,and t0 = the theoretical age at zero length (vonBertalanffy 1938; Beverton & Holt 1957). Thegrowth parameters were estimated using a non-linear least squares von Bertalanffy growth modelin the statistical package R (R Development Core
Figure 2 Comparison of a whole otolith (left) and a transverse section (right) of an otolith of R. tapirina with twoopaque zones. Whole otolith illuminated with reflected light with the distal surface facing upwards. Black dotsindicate opaque zones that were interpreted as annuli. TS of the corresponding otolith from the same fish illuminatedwith transmitted light. While capped line indicates the axis along which measurements were made for marginalincrement analyses and otolith thickness. White dots indicate the opaque zones that were interpreted as annuli. Scalebars = 1 mm.
Age validation, growth and population characteristics of greenback flounder 233
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
Team 2010). The parameter t0 was constrained tozero to provide a better description of early growth(Williams et al. 2003). Univariate analysis of vari-ance (ANOVA) was used to perform step-wisemaximum likelihood ratio tests to statistically com-pare growth parameters K and L∞ between malesand females (Kimura 1980; Francis 1988b). Thisprocedure also allowed growth relationships to becompared graphically using estimates of 95%confidence ellipses generated around least squaresestimates of K and L∞ (Kimura 1980). For thespatial comparison, growth curves were comparedbetween the two study areas using size-at-age datathat were pooled across sexes.
Population characteristics
Estimates of size-at-age of fish collected from thetwo study areas were used to produce area-specificpopulation size and age structures for each 6-monthperiod of the study. The sex ratios of the popula-tions in both areas for each sampling period werealso determined. A chi-square test for goodness-of-fit was used to assess whether the relativefrequency of males and females for each 6-monthperiod conformed to a 1:1 ratio.
Results
Usefulness of the otoliths of R. tapirina forageing
Interpretability of otolith structure
Whole otoliths contained an opaque core that wassurrounded by a translucent zone of variable thick-ness (Fig. 2). Beyond this, larger otoliths display-ed alternating opaque and translucent zones whichwere often difficult to distinguish towards theedge. TS of otoliths were essentially translucentwith relatively distinct opaque zones. In TS of lar-ger otoliths, an alternating sequence of opaque andtranslucent zones extended towards the dorsal andventral edges and proximal surface (Fig. 2). Zoneswere most pronounced along the ventral axis fromthe core to the proximal surface.
In order to compare the interpretability of theirincremental structures, counts of opaque zones were
compared between whole and sectioned otolithsfrom the same fish (Fig. 3). Many otoliths displayedbetween one and three opaque zones, whilst for alarge proportion, no opaque zones were evident. For91% of the otolith pairs, counts were the samebetween whole and sectioned otoliths, suggestingthat features were generally consistent between thepair. However, for otoliths with two or three opaquezones, the count from the whole otolith was oftenlower than for the TS (Fig. 3). The APE was 2.3%for otoliths with no opaque zones, which increasedto 5.7%, 10.8% and 13.5% for those with one, twoand three zones, respectively. This indicates that theincremental structure of the whole otoliths becamemore difficult to interpret accurately as they gotlarger, resulting in counts from such otoliths under-estimating the true count. As such, the use of TSwas determined as the preferred method for agedetermination.
Timing of opaque zone formation
TS of otoliths from young-of-the-year R. tapirinafrom 2009 and 2010 were examined to determinewhen the first opaque zone was formed. In bothyears, the otoliths from fish collected betweenMarch and July had translucent edges whilstotoliths with opaque edges were found from
Figure 3 Age-bias plot comparing the mean count ofopaque zones from whole otoliths (± SD) to the countsfrom the TS of otoliths, for groups of otoliths thatcontained the same number of opaque zones in the TS.Sample sizes for each group are shown in brackets.
234 J Earl et al.
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
August–December (Fig. 4). Otoliths with a com-pleted opaque zone and a translucent edge were firstobserved in September, and increased in frequencyuntil all otoliths were like this by January. Theseresults indicated that the formation of the firstopaque zone was initiated in mid-to-late winter inthe fish’s second year and was completed by mid-summer.
Those fish with one complete opaque zonewere considered for marginal increment analysis.In 2009 and 2010, all otoliths collected betweenJanuary and June had a translucent edge, with themarginal increment increasing through this period(Fig. 5). Some collected in July and most fromAugust to November had opaque margins, indic-ating that the formation of the second opaque zoneoccurred from July onwards in both years. ByDecember, otoliths had translucent margins andthe formation of the new opaque zone had beencompleted (Fig. 5). These results from marginalincrement and edge analysis over a 2-year periodwere consistent with the annual deposition of a
single opaque zone during late winter which wascompleted by early summer.
Growth of otoliths
The size of otoliths increased with the number ofopaque zones in the TS for both measures consid-ered. The two measures of otolith size, otolithweight (OW = 2.318*no. of opaque zones +5.097, P < 0.001, n = 279) and otolith thickness(OT = 2.214*no. of opaque zones + 0.8606, P <0.001, n = 279) increased proportionally with thenumber of opaque zones counted in the TS andthese linear relationships accounted for 53.2% and54.7% of the variation in otolith size, respectively.These results indicate that the otoliths continued togrow throughout an individual’s life. As such,their incremental structure should reflect their age.
Development of an ageing algorithm
A protocol was developed to provide age esti-mates for R. tapirina from the interpreted otolithstructure. The algorithm was:
Figure 4 Monthly time series showing the results of edge analysis from TS of otoliths from young-of-the-year R.tapirina collected in 2009 (non-thatched bars) and 2010 (thatched bars) (n = 486). No samples were available forDecember 2009.
Age validation, growth and population characteristics of greenback flounder 235
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
Am ¼ ðN � 12Þ þ mb þ mc;
where Am = age in months, N = number of opaquezones, mb = number of months from the birth dateto the end of the year (7 months for R. tapirina,i.e. June–December), and mc = number of monthsfrom the start of year to the month of capture.
Growth
Von Bertalanffy growth curves were fitted to thesize-at-age data for males and females to comparegrowth between sexes for the two study areas.Maximum likelihood tests indicated no significantdifference between sexes for the estimates of Kand L∞ for the Estuary (F1,552 = 0.002, p = 0.965)and North Coorong (F1,272 = 0.062, p = 0.804)(Fig. 6a–b). As such, data were pooled across thesexes to examine the spatial variation. The max-imum likelihood test indicated that L∞ was mar-ginally higher for the Estuary (L∞ = 348 mm TL)compared to North Coorong (L∞ = 340 mm TL)(F1,831 = 0.002, p = 0.014), though differences
were relatively small (Fig. 6c). The growth para-meter K was the same for each area (K = 0.09)(F1,830 = 1.1097, p = 0.293). Differences weresupported by non-overlapping 95% confidenceellipses around the estimates of K and L∞ (Fig. 6c).Consequently, there were slight differences in size-at-age of R. tapirina between areas. On average, fishgrew to 231 and 220 mm TL in the Estuary andNorth Coorong, respectively, in their first 12 monthsof life (Fig. 6). From then, growth rate decreased,but was similar in both areas, with size gains of 77,27 and 8 mm in their second, third and fourth years,respectively.
Population characteristics
Size and age structures
For the Estuary, there were considerable differ-ences in the size structures across consecutive 6-month periods (Fig. 7). Fish ranged in size from157 to 385 mm TL. Size distributions generally
Figure 5 Monthly time series showing the marginal increment ratio (proportion of the immediately precedingannulus) for TS of otoliths from R. tapirina that displayed one fully formed opaque zone (i.e. 1+ age class) for the2007 (circle), 2008 (triangle) and 2009 (diamond) year classes (n = 346). The edge type (e.g. translucent edge [opensymbols] or opaque edge [closed symbols]) is also shown.
236 J Earl et al.
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
Figure 6 Size-at-age data for female (open diamonds) and male (closed diamonds) R. tapirina from A, theEstuary; and B, North Coorong with fitted von Bertalanffy growth curves. C, Site-specific growth curves are shownfor the spatial comparison whereby data were pooled across sexes for the Estuary (E) and North Coorong (NC).Also displayed are 95% confidence ellipses for K and L∞ for each sex from the two study areas, and for each area.
Age validation, growth and population characteristics of greenback flounder 237
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
consisted of a single, dominant mode. One excep-tion was the bimodal distribution for the period ofDecember 2009 to May 2010, which involved asmall mode at 210 mm TL and a large mode at310 mm TL. The small mode persisted throughlate 2010 and into 2011. However, the mode oflarger fish was no longer evident in the sizestructures in subsequent periods, with fish > 330mm TL being rare in 2010 and 2011.
The 6-monthly age distributions for the Estu-ary involved only a small number of age classesand were generally dominated by the 1+ age class(Fig. 7). For the period of December 2008 to May2009, this age class accounted for 55% of thesample, with 0+, 2+ 3+ fish accounting for 30%,14% and 1%, respectively. In 2010 and 2011,older fish were rare, with the 1+ age class ac-counting for > 83% of fish in the two 6-month agedistributions from June 2010 to May 2011. In eachyear, the contribution of 0+ fish was highest fromDecember to May. For each 6-month period fem-ales accounted for 85%–99% of the fish sampled.On each occasion, the chi-square goodness-of-fittests indicated a significant difference in sex ratioat the 0.0001 level of significance.
For North Coorong, there were substantialdifferences in size distributions among 6-monthperiods. Fish ranged in size from 154–357 mm TL(Fig. 8). Gill net sampling limited the size range offish collected in June–November 2009; thus, therewas a narrow size distribution around the mode of310 mm TL. In early 2010, these large fish beganto drop out of the size structures, and by mid-2010were absent with very few fish > 280 mm TL. Asa result, small fish dominated the size distributionin late 2010. This cohort persisted in subsequentsize structures and contributed to a dominantmode of larger fish 1 year later.
Overall, age distributions for North Cooronginvolved four age classes (i.e. 0+ to 3+ ), but weredominated by either 1+ or 2+ fish (Fig. 8). In2009, 58% were 2+ fish, while most of theremainder were 1+. In 2010, the age structureswere dominated by 1+ fish, whilst 2+ fish werepresent in relatively low numbers. The highproportion of 1+ fish at that time was no longerevident as 2+ fish the next year. In contrast, the
high proportion of 1+ fish in the population inearly 2011 was still evident as 2+ later that year.In each year, the contribution of 0+ fish to theadult population was highest from December toMay. Females accounted for between 93%–100%of the fish sampled in each 6-month period. Oneach occasion, chi-square goodness-of-fit testsindicated a strong bias in sex ratio at the 0.0001level of significance.
Discussion
Usefulness of the otoliths of R. tapirina forageing
The first criterion for otoliths of R. tapirina to beuseful for ageing is that they must display aninternal structure that can be interpreted visually.Here, whole and TS of otoliths displayed an incre-mental structure that consisted of alternatingopaque and translucent zones. This macrostructurewas clearest in the TS. The use of whole otolithsfor age determination remains a common practicefor the youngest age classes of some flatfishspecies (Dwyer et al. 2003). Here, whole otolithsfrom larger fish that contained two or moreopaque zones were thick, which restricted theirtransmittance of light, obstructed the view of theirinternal structure and complicated their interpreta-tion. This resulted in counts of opaque zones fromwhole otoliths often being lower than those fromthe sectioned otolith from the same fish. There-fore, the preferred method to apply in an ageingprotocol for R. tapirina was to prepare TS ofotoliths in order to most clearly reveal theincremental structure and thereby ensure the high-est accuracy in age determination.
The second criterion for the usefulness of theotoliths of R. tapirina is that the incrementalstructure must be formed on a regular and deter-minable timescale. In this study, edge analysisdetermined that the timing of formation of the firstopaque zone was between July/August and Dec-ember (i.e. late winter–summer) during the fish’ssecond year of life. Completion of the first opaquezone by December corresponded to an age of 19months. Furthermore, edge analysis and marginal
238 J Earl et al.
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
Figure 7 Size and age structures of R. tapirina from the Estuary for the period December 2008–May 2011. Left-hand graphs show the size structures in consecutive 6-month periods. Grey bars represent samples collected fromthe commercial fishery and white bars represent samples collected by scientific sampling. Right-hand graphs showthe age structures for the same consecutive periods. The progression of the 2008 year class is also shown.
Age validation, growth and population characteristics of greenback flounder 239
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
increment analysis indicated that the second opa-que zone was formed between July and December1 year after the first zone was formed. Therefore,
these results substantiate an annual periodicity tothe formation of the otolith macrostructure throughthe first 2–3 years of the lives of the fish.
Figure 8 Size and age structures of R. tapirina from North Coorong for the period June 2009–November 2011.Left-hand graphs show the size structures in consecutive 6-month periods. Grey bars represent samples collectedfrom the commercial fishery and white bars represent samples collected by scientific sampling. Right-hand graphsshow the age structures for the same consecutive periods. The progression of the 2008 year class is also shown.
240 J Earl et al.
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
While the periodicity of formation of opaquezones subsequent to the second one remains unval-idated for now, there is evidence that the thirdzone is also formed annually. First, it is similar inappearance to the second zone. Second, the depo-sition of translucent and opaque material that com-prised the third increment conforms to a similartemporal cycle to that of the previous zone. Ver-ification of the periodicity of formation across theentire age range of interest is the preferred out-come of any validation study; however, as Cam-pana (2001) advised, this is often unrealistic. At aminimum, validation for the youngest age classesis imperative (Choat et al. 2009).
The timing of zone formation described herefor R. tapirina is characteristic of numerous tem-perate and subtropical marine fishes (Beckman &Wilson 1995; Fowler 2009). For some species, theformation of opaque zones coincides with periodsof rapid growth (Beckman & Wilson 1995; Potts& Cowley 2005). This is unlikely for R. tapirinaas water temperatures across southern Australiaduring spring are relatively low. Furthermore, thisis the spawning season for R. tapirina (Crawford1984; Barnett & Pankhurst 1999), during whichhigher demands for energy for gonad developmentare likely to limit growth (Rijnsdorp 1990; Nash& Geffen 2005). As such, opaque zones in theotoliths are likely to correspond to slow otolithgrowth and to narrower widths of daily increments(Fowler 2009; Mann-Lang & Buxton 1996).Nonetheless, the process of opaque zone forma-tion is complex and likely involves numerousexogenous (Fowler & Doherty 1992), endogenousand physiological (Fowler 1995) factors.
Finally, the third criterion for otoliths to beuseful is that they must continually grow through-out the lives of the fish. For R. tapirina, otolithsize increased proportionately through the first 4years of life. This suggests that relatively consist-ent amounts of otolith material were deposited onthe growing surface of the otoliths during eachyear of the first 4 years of the lives of the fish.
In summary, this study has demonstrated thatthe incremental structure in the TS of R. tapirinaotoliths is formed annually and therefore can beinterpreted in terms of fish age. As such, an ageing
protocol was developed to convert counts ofopaque zones in the TS of otoliths into estimatesof fish age. To develop an informative growth mod-el for fast-growing, short-lived species such asR. tapirina, it was advantageous to reduce the tem-poral resolution of size-at-age estimates from yearsto months.
Growth
Estimates of size and age in months were used todescribe growth for R. tapirina. Since the growthrate decreased exponentially with age, von Berta-lanffy growth curves were used to describe andcompare growth between sexes. For the Estuaryand North Coorong, there were no differences ingrowth parameters between males and females.For many flatfish species, sexual dimorphism ingrowth is characteristic, with females typicallyattaining larger sizes and living longer than males(Dwyer et al. 2003; Fischer & Thompson 2004;Nash & Geffen 2005). However, for R. tapirina,there were no such differences detectable for thefirst 4 years of life. This may relate, in part, to thefew male fish and the young age classes availableto this study. Such gender-based differences, how-ever, could possibly become evident in older fishfor which energy has been diverted to reproduc-tion for a longer part of their lives.
Growth parameters were also compared bet-ween the Estuary and North Coorong. Fish in theEstuary attained a higher asymptotic length thanthose from North Coorong. However, the area-specific growth curves indicate that the spatialdifference for L∞ was small (8 mm TL) in relationto the estimated maximum size and lifespan of thespecies. The non-explicit spatial difference sug-gests that either environmental conditions, includ-ing the availability of resources for R. tapirina,were similar among areas, and/or there was someexchange of individuals between the two areas.Nonetheless, area-specific growth curves indicatedthat R. tapirina achieved approximately 66% (231mm TL) and 64% (220 mm TL) of their respectiveL∞ after 12 months of life in the Estuary and NorthCoorong, respectively. Whilst rapid growth is char-acteristic of the genus Rhombosolea, the growth
Age validation, growth and population characteristics of greenback flounder 241
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
of R. tapirina in its first year is much greater thanfor its conspecifics (Coleman 1974; Francis 1988a;Paul 1992).
Population characteristics
Size and age characteristics of R. tapirina wereexamined to inform about population structure anddemographic processes likely to account for fluc-tuations in population size. From December 2008to November 2011, size structures were comprisedmainly of fish < 340 mm TL and the age struc-tures were generally dominated by 1+ and 2+ fish,while older fish were rare. Given the potential forthis species to attain 10+ years of age (Sutton et al.2010), the biomass of R. tapirina in the Coorongestuary essentially consisted of fish in the young-est age classes.
The dominance of only one or two young ageclasses observed here is similar to other popula-tions of R. tapirina in Tasmania and New Zealand.Kurth (1957) indicated that fished populations offthe north coast of Tasmania consisted mainly of1+ and 2+ fish. Similarly, exploited inshore popula-tions off the south coast of New Zealand mainlyinvolved 2+ and 3+ fish (Sutton et al. 2010).However, the latter population also included fishthat were up to 10 years of age, thus demonstrat-ing the potential for this species to achieve consid-erably older ages than were found in the presentstudy. As such, the population in the Coorongestuary appears to be considerably truncated inage, as there is minimal biomass stored in the ageclasses of 3 years of age and older.
There are a number of potential hypotheses toaccount for the observed population truncation.Exploited fish populations typically experiencesome degree of age truncation through the rem-oval of larger, older fish by size-selective fishing(Longhurst 1998). For R. tapirina, historically lowexploitation rates by the commercial fishery in theyears prior to (Ferguson 2007), and during thisstudy suggest that other processes contributed tothe restricted age structure of the population. An-other hypothesis is that the older fish migrate fromthe estuary to offshore habitats, possibly forspawning. Such behaviour occurs for numerous
other flatfish species such as Pleuronectes platessaand Atheresthes stomias (Rickey 1995; Rijnsdorp1990). Perhaps during the drought, the poorenvironmental conditions in the Coorong estuaryforced the adults to leave the estuary for the marineenvironment due to a lack of food or appropriateconditions for reproduction. As yet, it is notpossible to differentiate amongst these alternativehypotheses to account for the truncated populations.
A further complication for the population char-acteristics of this species was the change in agestructure that followed the commencement ofdrought-breaking freshwater flows to the estuaryin late 2010. Flows rapidly reduced salinitythroughout most of the system and improved habi-tat quality for R. tapirina. As a consequence, the1+ fish which dominated the population in NorthCoorong during the final year of the droughtperiod remained in the population and contributeda rare dominant mode of 2+ fish in 2011. Onehypothesis is that the retention of these fish in thesystem for a further year occurred in response toincreased food availability (Grange et al. 2000)and the re-opening of suitable habitat for R. tapi-rina in the Coorong. The question remains as to theinfluence of environmental variation associated withfreshwater flows on demographic processes thatmay contribute to fluctuations in population sizeand structure.
Despite the age truncation, there were slightvariations in population structure over the 3-yearperiod. The size and age structures were relativelysimilar between 6-month periods for both studyareas, due to a lack of modal progression of 1+fish across years. Recruitment of 0+ fish occurredin summer/autumn of each year and older indivi-duals in their third or fourth years of life wererarely detected at that time. Moreover, there werestrong spatial similarities in population structurebetween the Estuary and North Coorong, whichsuggests that either mixing occurred between thetwo areas and/or that similar demographic pro-cesses were operating over this spatial scale.
This study has established the usefulness ofsectioned otoliths for ageing R. tapirina and hascontributed important information on the popula-tion biology of this species. The truncated nature
242 J Earl et al.
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
of the populations in the Coorong estuary indicatethat the previously documented variation in fish-ery productivity for this species is likely driven bydemographic events and/or processes that occurduring the first 2 years of the life history. This islikely to reflect interannual variation in recruit-ment or variable post-settlement survivorship inthis environmentally complex system.
AcknowledgementsWe gratefully acknowledge the assistance of numerousvolunteers and colleagues for their assistance with thecollection of samples. These include several commercialfishermen from the Lakes and Coorong Fishery whowere extremely generous and forthcoming with theirtime. We would also like to thank David Short, NeilWellman, Tom McNamara and Luke Silvestor for theirassistance with fieldwork and the processing of samplesin the laboratory; Paul Burch for his assistance withstatistical procedures and Luciana Bucater for her assis-tance in preparing the map. This project was supportedby Nature Foundation SA. This work was part of JasonEarl’s PhD study for which he was supported by theA. J. and I. M. Naylon Postgraduate Scholarship. Thisproject was approved by the Flinders University AnimalWelfare Committee, Project no. E340.
ReferencesBarnett CW, Pankhurst NW 1999. Reproductive biology
and endocrinology of greenback flounder Rhombo-solea tapirina (Günther 1862). Marine & Fresh-water Research 50: 35–42.
Beamish RJ 1992. The importance of accurate ages infisheries science. In: Hancock DA ed. Proceedingsof the Australian Society for Fish Biology work-shop on the measurement of age and growth in fishand shellfish. Canberra, Bureau of Rural Resou-rces, Australian Government Publishing Service.No.12. Pp. 8–22.
Beamish RJ, Fournier DA 1981. A method for compar-ing the precision of a set of age determinations.Canadian Journal of Fisheries and Aquatic Scien-ces 38: 982–983.
Beckman D, Wilson C 1995. Seasonal timing of opaquezone formation in fish otoliths. In: Secor DH, DeanJM, Campana SE eds. Recent developments in fishotolith research. Columbia, USA, University ofSouth Carolina Press. Pp. 27–43.
Beverton RJH, Holt SJ 1957. On the dynamics of exploitedfish populations. London, Chapman and Hall.
Campana SE 1999. Chemistry and composition of fishotoliths: pathways, mechanisms and applications.Marine Ecology Progress Series 188: 263–297.
Campana SE 2001. Accuracy, precision and quality con-trol in age determination, including a review of theuse and abuse of age validation methods. Journalof Fish Biology 59: 197–242.
Campana SE, Thorrold SR 2001. Otoliths, increments,and elements: keys to a comprehensive understand-ing of fish populations? Canadian Journal of Fish-eries and Aquatic Sciences 58: 30–38.
Choat JH, Kritzer JP, Ackerman JL 2009. Ageing incoral reef fishes: do we need to validate the period-icity of increment formation for every species offish for which we collect age-based demographicdata? In: Green BS, Mapstone G, Carlos G, BeggGA eds. Tropical fish otoliths: information forassessment, management and ecology. New York,USA, Springer. Pp. 23–54.
Coleman JA 1974. Growth of two species of flounders inthe Hauraki Gulf, New Zealand. New Zealand Journalof Marine & Freshwater Research 8: 351–370.
Crawford CM 1984. An ecological study of Tasmanianflounder. Unpublished PhD thesis. Australia, Uni-versity of Tasmania.
Dwyer KS, Walsh SJ, Campana SE 2003. Age deter-mination, validation and growth of Grand Bank ye-llowtail flounder (Limanda ferruginea). ICESMarine Science 60: 1123–1138.
Ferguson GJ 2007. The South Australian greenbackflounder (Rhombosolea tapirina) fishery. SARDIPublication No. F2007/000315-1. Adelaide, SouthAustralian Research and Development Institute(SARDI) (Aquatic Sciences).
Fischer AJ, Thompson BA 2004. The age and growth ofsouthern flounder (Paralichthys lethostigma), fromLouisiana estuarine and offshore waters. Bulletinof Marine Science 75: 63–77.
Fowler AJ 1990. Validation of annual growth incre-ments in the otoliths of a small, tropical coral reeffish. Marine Ecology Progress Series 64: 25–38.
Fowler AJ 1995. Annulus formation in otoliths of coralreef fish—a review. In: Secor DH, Dean JM,Campana SE eds. Recent developments in fish oto-lith research. Columbia, USA, University of SouthCarolina Press. Pp. 45–63.
Fowler AJ 2009. Age in years from otoliths of adulttropical fish. In: Green BS, Mapstone G, Carlos G,Begg GA eds. Tropical fish otoliths: information forassessment, management and ecology. New York,USA, Springer. Pp. 55–92.
Fowler AJ, Doherty PJ 1992. Validation of annual gro-wth increments in the otoliths of two species ofdamselfish from the Southern Great Barrier Reef.Australian Journal of Marine & Freshwater Res-earch 43: 1057–1068.
Age validation, growth and population characteristics of greenback flounder 243
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
Fowler AJ, McGlennon D 2011. Variation in productiv-ity of a key snapper, Chrysosphrys auratus, fisheryrelated to recruitment and fleet dynamics. FisheriesManagement and Ecology 18: 411–423.
Francis RC 1988a. Recalculated growth rates for sandflounder, Rhombosolea plebeian from taggingexperiments in Canterbury, New Zealand. NewZealand Journal of Marine & Freshwater Research22: 53–56.
Francis, RICC 1988b. Maximum likelihood estimationof growth and growth variability from taggingdata. New Zealand Journal of Marine & FreshwaterResearch 22: 43–51.
Froese R, Pauly D 2011. Fishbase. http://www.fishbase.org (accessed 15 March 2013).
Gilbertson DD 1981. The impact of past and presentland use on a major coastal barrier system. AppliedGeography 1: 97–119.
Gomon MF, Bray DJ, Kuiter RH, eds 2008. Fishes ofAustralia’s Southern Coast’. Sydney, Reed NewHolland.
Grange N, Whitfield AK, De Villiers CJ, Allanson BR2000. The response of two South African east coastestuaries to altered river flow regimes. AquaticConservation 10: 155–177.
Kailola PJ, Williams MJ, Stewart PC, Reichelt RE,McNee A, Greive C, eds 1993. Australian fisheriesresources. Canberra, Australia, Bureau of ResourceSciences and Fisheries Research and DevelopmentCorporation.
Kimura DK 1980. Likelihood methods for the vonBertalanffy growth curve. Fisheries Bulletin 77:765–776.
Kingsford RT, Walker KF, Lester RE, Young WJ,Fairweather PG, Sammut J, et al. 2011. A Ramsarwetland in crisis – the Coorong, Lower Lakes andMurray Mouth, Australia. Marine & FreshwaterResearch 62: 255–265.
Kurth D 1957. An investigation of the greenbackflounder, Rhombosolea tapirina Günther. Unpub-lished PhD thesis. Australia, University of Tasmania.
Longhurst A 1998. Cod: perhaps if we all stood back abit? Fisheries Research 38: 101–108.
Mann-Lang JB, Buxton CD 1996. Growth character-istics in the otoliths of selected South African spar-id fish. South African Journal of Marine Science17: 205–216.
May HMA, Jenkins GP 1992. Patterns of settlement andgrowth of juvenile flounder Rhombosolea tapirinadetermined from otolith microstructure. MarineEcology Progress Series 79: 203–214.
Morison AK, Burnett J, McCurdy WJ, Moksness E 2005.Quality issues in the use of otoliths for age estima-tion. Marine and Freshwater Research 56: 773–782.
Nash RDM, Geffen AJ 2005. Age and growth. In:Gibson RN ed. Flatfishes: biology and exploitation.
Oxford, England, Blackwell Publishing Company.Pp. 138–163.
Paul LJ 1992. Age and growth studies of New Zealandmarine fishes, 1921–90: a review and bibliography.New Zealand Journal of Marine & Freshwater Re-search 43: 879–912.
Potts WM, Cowley PD 2005. Validation of the period-icity of opaque zone formation in the otoliths offour temperate reef fish from South Africa. AfricanJournal of Marine Science 27: 659–669.
R Development Core Team 2010. R: A language andenvironment for statistical computing. Vienna,Austria, R Foundation for Statistical Computing.http://www.R-project.org (accessed 1 December2012).
Rickey MH 1995. Maturity, spawning, and seasonalmovement of arrowtooth flounder, Atheresthesstomias, off Washington. Fisheries Bulletin 93:127–138.
Rijnsdorp AD 1990. The mechanism of energy alloca-tion over reproduction and somatic growth infemale North Sea plaice, Pleuronectes platessa.Netherlands Journal of Sea Research 25: 279–290.
Sutton CP, MacGibbon DJ, Stevens DW 2010. Age andgrowth of greenback flounder (Rhombosolea tapir-ina) from southern New Zealand. New ZealandFisheries Assessment Report 2010/48. Wellington,Ministry of Fisheries.
van den Enden T, White RWG, Elliott NG 2000.Genetic variation in the greenback flounder Rhom-bosolea tapirina Günther (Teleostei, Pleuronecti-dae) and the implications for aquaculture. Austra-lian Journal of Marine & Freshwater Research 51:21–33.
von Bertalanffy L 1938. A quantitative theory of org-anic growth: Inquiries on growth laws II. HumanBiology 10: 181–213.
Webster IT 2010. The hydrodynamics and salinityregime of a coastal lagoon – the Coorong, Austra-lia – seasonal to multi-decadal timescales. Estuar-ine, Coastal & Shelf Science 90: 264–274.
Williams AJ, Davies CR, Mapstone BD 2005. Variationin the periodicity and timing of increment formationin red throat emperor (Lethrinus miniatus) otoliths.Marine & Freshwater Research 56: 539–538.
Williams AJ, Davies CR, Mapstone BD, Russ GR 2003.Scales of spatial variation in the demography of alarge coral reef fish: an exception to the typicalmodel? Fisheries Bulletin 101: 673–683.
Ye Q, Earl J, Bucater L, Cheshire K, McNeil D, NoellC, et al. 2013. Flow related fish and fisheries eco-logy in the Coorong, South Australia. Final Report.SARDI Publication No. F2009/000014-2. SARDIResearch Report Series No. 698. Adelaide, SouthAustralian Research and Development Institute(SARDI) (Aquatic Sciences). 84 p.
244 J Earl et al.
Dow
nloa
ded
by [
Tem
ple
Uni
vers
ity L
ibra
ries
] at
19:
20 2
1 N
ovem
ber
2014
