, , LC.E.S C~M.J 995 " C.M.1995/K:39 - ' '. ' . ' ' ' ' , 1 AGE DETERMINATION OF HOMARUS AMERICANUS USING AN INDEX OF L1POFUSCIN A8UNDANCE by Ä.de Kerios (1) G.Conan (2) M.R.J. Sheehy (3) (1) Universlte de Moneton, Departement de ehimie-biöchimie, Marieton, N.8., E1 A 3E9, Canada , , (2) Dep~utmerit of Fisheries and Oeeans, Gulf Region, Moneten, N.8., E1C 986, Canada ,,' . , (3) University of Queenslarid; Zoology Departmerit, Australia arid University of Leicester, Zoology DeparÜnent, England • ~<" ABSTRACT A lipofusein abundanee index is used for age determiriations (Sheehy, M.R.i, 1989) of Homaius americanus. LipofusCins' are identified on histological preparatiens of. the brain by fluorescence microscopy, arid histochemic~1 staining. Lipofuscin. abundance is quaritified by semi:autornated measurerri.ents of its yellowautofluörescence under U.V. excitation. The f1uofescence is. measüred either by digital, image analysis of the preparatioris or by analogue phetometry of fluorescence intensity. The between-age-groups , vadability is much greaÜJr than the between-individuals variability, allowing precise age discriminaticiris. • .. A preliminary calibration on preparations fram cuitured lobsters of known age (Couftesy Bodega Bay; U.C. Davis team) is coridLicted. Lipofuscin provides apreeise age index (R2=.95). No significant differences are found between lipofusein aecumulations in individuals of a same age group. A 100% disei-imination of individuals into cor'rect 'age groups could be achieved. Over the range of ages studied, a positive exponeritial, cir an allometrie model adequately represent Iipefuscin accumulationvs age.Th~ Lipofuscin/age relatioriship should not be ealibrated on 19bsters reared at constarit temperature for age determimitions of wild lobsters· beeause the lipofusein contents can differ considenibly. .Calibrations are attempted on wild lobsters uslng age . " ..... determinations based on modal analysis of younQ individuäls and life 1 •• cycle information. The he~erogeneity of Iipofuscin abundance reveals that ages in sinall size modal groups may, already be quite variable. The life history arid genetic charactedstics of each individual might also affsct Iipofuscin metabolism. The waters of the Gulf' of 5t Lawrence are strongly stratified in summer (~1 oe to 22°C) and provide a yery heterogeneous habitat Individuals Iife historiE~s may widely differ. Notwithstanding the difficulties encountered for calibration, our best working hypothesis suggested an age of 5 years for a 63 mm carapace length (minimal legal size) males arid 8 years for one female of the same size. _. INTRODUCTION ~ After the recerit decline of the cod fishery, the lobster fishery now provides the most important catch in vfllue withiri the Gulf of 5t Lawrence. The forecasting of lobster recruitment to the fishery several years in advance is of growing importance for management as weil as economic planning; Unfortunately, the estimatiori of the abundance of un-recruited age groups is hindered by the present difficulty to estimate the aga of lobsters. For age determinatioris one must rely exclusively on sizes. The molt and growth schedules may vary considerably between individuals and between locations. Only the very first mcidesin the size-frequency distributions of the youngest benthic stages may be attributad to age groups. Large modal sizes are. artefacts created by a combination of harVesting regulations, and molt groups. Lobsters having molted the same nuinber of times may belong t6 very different age groups. This is partially due to the fact that lobsters may molt less than once· a year as they grow older. Females foi' instance will alternate molling arid spawning in such a way that they will· normally skip a molt while incubating eggs. Attempts to age lobster by other means than size have not been fully successful. For instance, the number of antennular segments 0 r the number of cirrimaHdia have proved to be beuer correlated w i t h size than with aga (Henocque, 1987; 5h.elton,1991). We attempt to adapt a technique developed by 5heehy (1989) for age determinations of ci crayfish; Cherax cuspidatus. The. accumulation of lipofuscins in tissues during aging is foimd in a wide range of .animals, from arthropods to mammals. The biochemistry of the . > accumulation in the human eye has been described (Eldred and Lasky, 2 e 1993). It appears thai IlpofusCin accurriulatian is actually the consequence of an incomplete destruction of metabölites during the process,of rejuvenaticin of the cells. The rejuvenation results from a destructiön(lysis) and rebuilding of cellular structures, tiut,the process "of .Iysiswould become incomplete with, seriescence. Tissues in wh ich thecells are thought t6 have stopped dividing (postmitotic tissues) should, contain an arrlOunt of lipofuscin wh ich is a funCtiöri of age. By callbrating the lipofuscin/age relationship arid fitting' an adequate model, it should be possible to read age as a function of amount öf lipofusCin deposited. o. • • MATERIAL AND METHODS • Bioiogical material Thirty four löbsters from early berithic to commercial sizes were collected from the wild in Malpeque Bay (Prince Edward Islarid) by trawling (Conan eil., ,1994). The Jobsters were selected fra'm modal groups ranging in size. from 30 to 80 mm. The iridividuals were measured, weighed and sexed. Caution was taken. not to chose lobsters with missing legs.er sigris of regeneration that might· have hindered growth. The sampling had been designed to be , simultarieously used for surveying and mapping the aburidarice of all slze caiegories of lobsters (Conari et al., this meeting). It is feit that the lobsters selected weil represerit average characteristics of the Malpeque Bay stock. et EighteEm lobsters of known, age were obtained from an aquaculture facility (Courtesy Bodega, Baylaboratory; University of California Davis). Three sets öf six iridividLials were. provided ,each fram three hatc~ing. everits which hild occurred on 11, August, 1991, 19 February . 1992, and 6 August 1992. At the time of the tissue sampling these lobsh3rs, were respectively 42; 37 and 31 month old. on After tdals the heart and ether· tissues, the oifactory lobe of the brain was seiectedas im expected post-mitotic .tissue for measurement of. Iipofuscin abundance. Prior to dissection the . lobsters were ariesthetized at, cold, temperature. The brains, of the animals from the wild were dissected in, vivo. The cephalothöraxes of the. a,nimals from t~e aquaculture, facility were fixed in toto immediately after sacrifice. The fixations were made for 24 hours in 10% formalin in seawater. The lobsters ,from the aquaculture" facility 'ware shipped arid remained in the fixative rrieqiurri for 3 .daYs. 3 Lipofuscin abundance index LipofusCin .deposits were identified arid. quantified .on histological serial histological sections of the ol!actory lobe byselective histochemical staining (Pearse, 1985) and by U.V. induced ,autofluorescence. We had initially considered extraction arid titration by standard biochemicai mettlods, out we were aeterredfrom this procedure hy the fact that "Iipofuscins" a complex mixture of products which; so hir, cannot be easily detected in solution. Lipofuscin grariules conversely have adefinite structure and can be identified in histological preparations oy auto-fluorescence or staining. The Iipofuscin deposits were located by comparing sections stained by reactants specific to different biochemical characteristics of lipofuscin, and by obserVation under U.V; light within a range of wavelengths specific for. Iipofuscin autö-fluorescence. Histological preparations were made from 6Jlm serial sections in. paraffin wax using standard mounting procedures. Lipofuscin-specific staining was achieved using either Sudan, black periodic acid-Schiff, or haematoxylin-eosin (Pearse, 1985). Unstained preparations were observed under a f1uorescerice microscope, using 330 to 380 nm UV excitation light. ' The abundance of Iipofuscins in the deposits within a section was quantified by measuring the auto-fluorescence they produce either by image analysis of a digitized colour picture (IADCP) of the section, or by direct analogue ,measurement of the light emitted (DAMLE) by a specific area of the section. ,Initial trials for directly digitizing the image of the slide' under microscope, using a digital camera; and feeding the data direCtly into a computer did not provide high enough a resolution in coiours and a sufficiEnlt pixel definition. A higher definition, both in colour resqlution and iri definition was obtained by taking a ,colour photographic arid digitizing it using a standard page scariner. In DAMLE, the amourit of light emitted by the Iipofuscin was measured oya photometer mourited on a Leitz fluorescence microscope. Ttle power of resolution ,of the photometer was enhariced tiy a monochromator arid a photomultiplier. The position and size· of llle field measured was adjusted manually. We proceeded by measuririg the 'light emitted bcith by llle preparation in toto, and by 3 conspicuous lipofuscin graiiules. lri the IADCP, the scanned digital data was processed using the software "Adobe Photoshop". Colours, were scanried into three .components red, green, and. blue (RGB) for display on three colour '._. plane monitors. Tne intensities of the red, green, and blue 4 • .\ components for each pixel were classified into frequency histograms which were analyzed ,separately. Since aach, pieture was a different photographie framewith its own. scale of colours, each picture needed to be calibrated separately. For calibratio,n; the range of intensities in red, green,' and blue is measured on histograms . Ülken fram fields on the digitized photograph which appear to 'contain Iipofuscin exclusively. The modal sub-components . charactt3ristic of lipofuscins were theri identified. and estimated on the digitized. in toto, photograph; In ttlis, process the frequencies of the ,intensities within, the range identified .. as characteristic of Iipofuscin ware simply summed. After initial trials, for si~plicity, the cinalyses were conducted only on the green component which contained most of the· information on the characteristic yellow . . fluorescence oflipofuscin. The index of Iipofuscin is rf3presanted by the number of fluorescent pixels or by the ratio of the number of fluorescent pixels rapresenting· the 'organ., ' In the DAMLE, the measures of light emitted we automatically processed by the MPV-STAT software provided by Leitz. Direct measurements of lightintensity. emitted within the selected range of wavelengths either by the lobe in toto or by the seleeted field, were provided as output. ....... RESULTS Location • arid' aspect of Iipofusciri concenthitions. Lipofuscin appeared to be rarido~ly distributed wlthin the olfactory lobe in the form of structured granules of sizes rariging' from 2 to 6 . Jl., Jhe gr~nules fluoiesced in bright yellow (560 to 620 nm) unaer 330 to 380 nm UV light. The. granules are conspicuously revealed by Sudan Black, periodic acia-Schiff reaction, and Haematoxyliri-eosin. The distribution Of lipofuscins along a sequence series of serial sections is presented in figiües. 1. and 2. The data is proviaea both as counts of f1uorescent pixels (1) and as the. rätios of fluorescerit pixels. to the total number. 6f pixels within the se6tio~ of the. lobe (2). Figure 1. shows a quadrati6. relcitionship between counts. eif .'. , fluorescent pixels arid theirserial (Spearman' rank correlatiein P=.47) locationwhi6h was satisfactor,ily cOrrected .by taking the, ratio index (Flg 2). SUbsequently cill further stätistics' were calculated for ra.tio indices. .A ~istogram of the frequencies ..of counts or ratios of fluorescent ...-. pixels on a complete series of 74,serial sections eif an olfactory 5 lobe (figure 3) showed a symetrical distribution which was approximated by anormal curve. A Student test shows that a random selection of 12 sectiens weuld prövide. an estimate of mean ratio within +/- 10% of the true mean in 950/0 ef cases. For practical purposes only 12 sections were fully analyzed for each lobster in further werk. Relationship between lipofu·scin abundance IADCP and age in reared lobsters measured by The best fit was obtained for an exponential or allometric models (R2 = .95) rather than a simple linear (figure 4, 5, and table 1). , There were no significant difference ( P:>.05) between lipofuscin abundance indices of different iridividuals within the same age group (ANOVA and Kruskall-Wallis). The average lipofuscin abundance indices for individuals to different age groups did not overlap. The between-age-group difference was significant (P<.0001) when tested by ANOVA. Within each age group there was lihle correiation between the lipofuscin abundance and the weight of the individuals (Fig. 6; R2 < .25). It was the age independently from the weight of the individual which appeared to determine its Iipofuscin conterit. Relationship between lipöfuscin abundance IADCP and age of wild lebsters measured by The lipofuscin contentsof lobsters .taken from the wild in Malpeque Bay were much lower than those of reared lobsters of comparable size. However, the wild individuals were thought to be much older than the reared lobsters, based on previous, tagging arid modal analyses of size frequency distributions.. The lipofuscin content vs age relatioriship of wilo indiviauals could not be calibrated from data on reared iridividuals. Alternatively age were estimated from modal analysis and previous knowledge on the molting seasons arid molting frequencies of lobsters iri Malpeque Bay' (Moriyasu, 1984). Figure 7 and table 2 illustrate the relationship between lipofuscin abundance and age as defined from the size of the individuals studied and 'the position of the first 6 modes in the size frequency distributions. Ttiere was a definite relationship between Iipofuscin coriterit and presumed age, and a discrimination betWeen presumed, äges was achievablE3 for most individuals on the basis of their lipofuscin content However, there were significarit differences between the Iipofuscin contents . 'of individuals within .any of the modal groups at the 5% level; when 6 tested by ANOVA. The maan lipofuscin eontents of ihe different , modal groups did not diffe,r significantly wl1en compared by ANOVA (Table 3). Measuremerits of Iipofuscin abundance by' DAMLE and age The DAMLE results were eomparable to the IADCP results; but the disci'imination .between .age groups was not as good, and there were signifieant differenees (P <:.05) between individui:lIs within' the same groups. Mecisurements on individual granules did not provide a beUer resolution than measurements on the lobe in toto. The DAMLE was a faster but lass selective pracedure than the IADCP. Photometrie measures were possibly corrupted by light emissions fram sources other than Iipofusein. DISCUSSION .' e Our experimenÜll results ean be explained in the eontext of the ' current interpreÜltions of Iipofuscinmetabolism. Within an age graup; the lobsters of knöwn age obtained trom ci rearing. experiment all came trom " the same brciöd; and their indices of Iipofuscin ",.' . accumulation did not' differ significantly. This may l1ave resulted trom the fact that they all ware genetically similar,' that they häd been reared under similar conditiöns, but also that tl1ey had precisely the same age. Conversely lobsters ,taken from the wild, and attributed to the saine age group had significantly different indices of lipofuscin accurTlUlation. Their actual bii1.h dates must have been spread over ä whole sp'awning season of about montl1 since they most Iikely oi'iginated trom distind broöds. They had experienced different Iife conditions in an environment which is hlghly seasomilly arid spatially variable, 'ranging in temperatures trom -1 to 20 2C; and in saliriity fram.20 to 33 per mil. a The ciriinials obtained from a rearing experiment provide simplified information which is usetuifor tasting models öf lipotusein accumulaÜon vs age. The relationship appeared t6 ba curvilinear arid monotonie with a slope iricreäsing with age (over· the, age interval studied). An exponential or an allometi'ic model provided ci good fit. This is similar to observ.ations o~, Cherax for the earty part of the Iifespan only (sheehy, et ,al., 1994). Althöugh the inadequaey of using reared animals to caiibrate ihe age/lipofuscin abundance relationship is understandable, it is not .fullY explained. We have' no information on the ,influanee of geneties or on environmerital tactors on the procass of lipofuscin 7 accumulation in lobsters. The variability of lipofuscin contents in animals taken trom the wild arid atii-ibuted to the same age group increases considerably with presumed aga. This may reveals an artetact due to imprecise age determination by modal analysis starting trom the' smallest modal sizes, and increasing to the largest. The number of molts and the time interval between molts is known to diverge between individuals mostly after onset of maturity which is reached at a size varying considerably between individuals. Although very provisional, the age/lipoh.iscin content model, we produced can be temtatively used for estimating the age of lobsters captured in Malpeque Bay. We recorTunend however, cautious interpretations since we do not have information on the year to year variability in Iipofusciri accumulations, or on environmerital effects.. The most efficient way of calihraÜng the age/lipofusciri content for lobster in their natural, environment would involve in tagging and recapture of early benthic stages .of the lobster. It would then be possible to analyze the between-brood, the between-Iocation and between-year variability. The experiment should be run over a minimum of, teri years to allow for a comprehensive coverage of the presumed life span of commercially harvested lobsters. This may seem a tedious task, but in the present context of total lack of , precise information on the age structure of naturallobster populations, and of tlle commercial catch; tlle investment would be highly profitable. . • , ACKNOLEDGMENTS The authors wish to thank br. Sharon Me Gladery and Pierre Mallet for the histological advices, Eimer Wade for collecting and processing the data, Dr. Mikio Moryasu for the use of the facilities in Moncton and Guy Robichaud for his assistance on the field. 8 - e REFERENCES Conan, G.Y., M. Comeau, C. Gosset, G. Robichaud, and C. GaraTcoechea. 1994. The Bigouden Nephrops trawl, and the Devismes trawl, two otter trawls efficiently catching benthic stages of Snow Crab (Chionoecetes opilio),. and American lobster (Homarus americanus). Can. tech. rep. Fish Aqual. Sei. 1992, 27 pp. Eldred, G.E. and M.R. Lasky, 1993. Retinal age pigments generated by seit-assembling Iysosomotropic detergents, Nature, 361,p 724-726 Henocque, Y., 1987. Observations des antennules et determination de I'äge chez le homard, Homanis gammarus, La Mer, 25, p 1,12. Moryasu, M., 1984, Molting season and growth at molt of lobsters in Malpeque Bay, P.E.L, Canada. , LC.E.S., Shelfish Commitee, 17 p. Pearse, A., 1985, Histochemistry, Theoretical and Applied, Churchill Eds, London, 245 pp. Sheehy, M.R.J., 1989. Crustacean brain lipofuscin: an examination the morphological pigment in the treshwater crayfish Cherax cuspidatus (Parastacidae), J. Crust. BioL, 9(3), p 387-391. ot Sheehy, M.R.J., Greenwood, J.R, Zielder, D.R, 1994, More acurate age -determination of crustaceans trom field situations using the physiological age marker, lipofuscin, Mar. BioL, 121 (2), p-237-245 Shelton, P.M.J., RG.J Shelton and P.R. Richards, 1991. Eye . development in relation to moult stage in European lobster Homarus gammarus (L.), ICES CM 1991/K:33, 9 pp. 9 . ~ 4000 -f-' ~ 3800 ~ 3600 ~ 3400 V) ~ 3200 3000 ~ 2800 ~ 2600 ~ 2400 ~ 2200 L.....&.. &......L.. ....L.....< -t- I....I... 3 -f--',1"""'T'"""T"'""r"""T'"""T"'""I"""T""""T"""'I"""'T'""'T"""'I"""'T'"........,I"""'T'"""T"""'I"""T'"""T"""'I"""'T'""'T"""I"""'T'"........,r-r-+- 0 Z 10 20 30 40 50 60 70 NUMBER OF THE SECTION IN THE SERIAL SERIES Rgure 1- Lipofuscin deposits along 74 histological 6 Jlm serial sections of the' olfactive lobe of lobster. The deposits appear' sm aller towards the extremities due to the lobular shape of the organ. 7.5 -f-'.................L.....&..........&......L..~.............L.....<....................I....I...~.............L.....<............L...L....L............L....L.-+7 o 6.5 ~ 6 5.5 u 5 z ~ 4.5 ~ 4 c::: 3.5 Li 3 2.5 LU g -f--',1"""'T'"""T"'""1"""'T'"""T"'""r""'T'"""T"""'I"""'T'""'T"""'I"""'T'"-.-,I"""'T""""T"""'lr"""T"""T"""'lI"""'T'""'T"""'I"""'T'".......-,I'""""I"'"-+- o ,0 20 30 40 SO 60 NUMBER OF THE SECTION IN THE SERIAL SERIES 70 Figure 2- Lipofuscin deposits along 74 histological. 6 Jlm serial sections of the olfactive lobe of lobster. The fluorescence ratio is the ratio of the number of fluorescent pixels to the total number of pixels representing the organ. This type of index provides similar results independently of the position of the section. 10 • FREQUENCIES 16 14 12 10 8 6 4 2 o 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 FLUORESCENCE RATIO Figure 3- Frequency distribution of the fluorescence ratios in 74 serial sections of the olfacory lobe of a lobster. The distribution can be approximated by a normal curve.Twelve random sampies would provide an estimate within· +/10% of the mean in 95% of cases. 11 45 _--------------=---r 40 0 f= 35 ~30 ~25 a:i 20 ~ 15 ~ 10 5 g ~O ......"T-------r-------r-----L. 31 37 42 AGE IN MONTHS Y .. -75.388 + 2.483 * X; RA 2 ...894 • o ~ 3.5 3 u a:i 2.5 ~ 2 UJ UJ ~ 1.5 ::::> Li 1 3 .5 -..,.-;~-----~-----~----'31 42 37 AGE IN MONTHS Y = -5.21 + 0.207 o 4 * X; RA 2 .. 0.951 _--------------=--r ~3.5 UJ 3 u a:i 2.5 ~ 2 UJ ~ 1.5 cx: ::::> Li 3 1 .5 ........---,----------,------r'3.3 3.6 3.7 LN AGE IN MONTHS Y - -24.516 + 7.481 Figure 4-Reared lobsters. fluorescence * X; RA2 Linear, exponential - .955 and a1lometric models tor ratio vs age data. A1llobsters within a group had identical ages although the points were shifted in abscissa tor clarity ot the display. 12 Table 1- Fit of models to Fluorescence ratio vs age relationship for reared lobsters. Residual variance for untransformed variables. 16.7 . LI NEAR MOOEL EXPONENTI AL MOOEL 9.0 ALLOMETRI C MOOEL 9.1 13 LINEAR MODEL + ...................... 10 ...1.-..........................&..+ • • ...I.-...............~...............~........I....................L...i..-l ............... 8 - 6 4 - • • 2 V) <i! •• o. u.J Cl:: · · • 2S -2 V) • "" : - -4•• • -6 • -8 -I-T"""'I.......-r-"TI~"T""" ......-r--rl_~.,....,...,..-r-"T""""r-"T"""" 1 .......~"T"""r--rI_""T"""-r+ o 10 15 AGE IN MONTHS 5 20 25 30 EXPONENTIAL MODEL .4 -+--.:::-'-_........-_....... 1-...1 1 . - _.... 1_--'1 .3 .2 - • 1E·l V) -l « ::> Cl Vi u.J . Cl:: 1L..-_.l....-_....... 1_...I.-_+ • • •• "" • • "" r-----------__+_ • o -f------------.. . •• -.1 • -.2 • -.3 -.4 -.5 -.6 1 • • . . 1.25 1.5 1.75 "" rrr- I • 2 2.25 2.5 2.75 LN AGE IN MONTHS 3 3.25 3.5 3.75 ALLOMETRIC MODEL .5 +-_......_--'.__.......__"--._"""'.__.1....-._.. . . . _ _-'-'_......._-+ .4- .3 .2 - • "" • • • .1 - • "" • r• • • o +-.------------;:...----------+ ~ -1 E-l -.2 Cl Vi -.3- :5 u.J Cl:: -.4- -.5 • • •• """" • II- • 1 1 • 1.25 1.5 1.75 1 1 1 I 2 2.25 2.5 2.75 LN AGE IN MONTHS 1 1 3 3.25 Figure 5-Residual plots for the linear, exponential and allometric models of fluorescence ratio vs age for untransformed variables. The allometric model provides the most balanced residuals and the exponential model is the next best. The linear model is not satisfactory. 14 3.5 HATCHED 6 AUGUST 1991 , I • , , , T 1 I III~ ~ I~ I I I 240 200 I I · · · · , ~ I 320 280 360 WEIGHT IN GRAMS Y = 9467.044 + 16.421 * X; RA2 = .057 (J) ~ HATCHED 19 FEBRUARY 1992 Ci: 9000 -f-L..........................L.L..........J...L.........L..L..........L...L..a...&..J'-'-'l....oI..I-'-'"""-J- I § 8000 ~ 7000 c::: o 6000 I ~ :3 5000 o 4000 c::: ~ 3000 ~ 2000 ooh""T"T""""''''''''''''T""T''T'T'''T'"T"T''T'"T""T"TT"r-T''"T'"r''''''''''I'"'T'"'1''::;''''r-r+ z 100 120 140 160 180 200 220 240 260 WEIGHT IN GRAMS Y = 7719.206 - 14.548 * X; RA2 = .251 ~ >< LU HATCHED 11 AUGUST 1992 Ci: !z ~ (J) ~ o :3 LI.. LI.. o c::: LU ~ ~ I 3500 3000 2500 2000 1500 1000 500 , I · ~ ~ 1 · · ~ ~ ~ I 0 95 100 105 ~ • 110 ' I 115 120 125 130 WEIGHT IN GRAMS Y = 2032.828 + 1.825 * X; RA2 = 6.539E-4 Figure 6-There was no apparent relationship between f1uorescence ratio and weight of lobsters within a same age group. It appeared to be the age and not the weight of the lobsters which determined the tluorescence ratio. If the number of fluorescent pixels is independet trom weight, the fluorescence ratio 'should be more independent from the weight of the individuals. 15 16 ...f-a.........w..........l..L.I................................"..I-I-.......&..I...I~~.................. I 14 o ~ 12 ~ 10 as 8 ~ 6 u CI) o :3 4 LL. 2 O+r"'l""T',.+,""""""....................,..,..,........."l'"T"1.................rT"'I'"'T"'I'.........I""'I""lr+ o 1 2 3 AGE ATTRIBUTED TO THE MODEL GROUP IN YEARS Y = 1.123 + 1.1 17 * X; R"2 = .076 3+O-........a..I...I........~.........l..L.I.................J................L.L..&;..a....&....L.L.........~ I 2.5 o 2 0:: 1.5 ~ tj 1 as u .5 ~ -.5 3 -1.5 -2 1ß 0+-~--..,.,.......:::::::.-.-=-----.::.=-----4- ~ -1 r+ -+-r.,...,."'r"'T"'O"'T""'I""T"'lI"'T""l""T"'l'"T'T.............,I"'T"'lr-T"T"T'T.......'O"'T""...... o 1 2 3 AGE ATTRIBUTED TO THE MODAL GROUPS IN YEARS Y = -.81 5 + .769 * X; R"2 = .308 34--.L---I---I--..L..-..L--L.---J.--L.-4- I 2.5 o ~ 0:: tj z LU .5 ~ :3 -.5 ~ I 2 1.5 1 O+:!!:..---~:ooo"fII::::.--__r---+---+ I -1 ~ -1.5 ...J _2+--~___,....-_+-_,..-""T'""--r--___,....-___r'-+ -.75 -.5 -.25 0 .25 .5 .75 1 1.25 1.5 LN AGE AlTRIBUTED TO THE MODAL GROUPS IN YEARS Y = .063 :t 1.172 * X; R"2 = .336· Figure 7-Lobsters trom the wild. Linear, exponential and allometric models tor f1uorescence ratio vs presumed age. 16 Table 2- Fitting models to Fluorescence ratio vs age of wild lobsters. Residual variance for untransformed variables. LI NEAR MOOEL 11.36 EXPONENTIAL MOOEL 12.99 ALLOMETRIC MOOEL 12.34 Table 3- Analysis of Variance in fluorescence ratio between individuals within the first 6 modal groups presumed to be age groups. The differences are not significant. Regressions on a linear exponential or allometric model do show that a correlation exists, however. 51 GNI FI CANCE FLUORE5CENCE RATIO .29 LN FLUORESCENCE RATIO .17 17