Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Life history of marine threespine stickleback in Oyster Lagoon, British Columbia Saimoto, Regina Karin 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1993_fall_saimoto_regina.pdf [ 3.79MB ]
JSON: 831-1.0086293.json
JSON-LD: 831-1.0086293-ld.json
RDF/XML (Pretty): 831-1.0086293-rdf.xml
RDF/JSON: 831-1.0086293-rdf.json
Turtle: 831-1.0086293-turtle.txt
N-Triples: 831-1.0086293-rdf-ntriples.txt
Original Record: 831-1.0086293-source.json
Full Text

Full Text

LIFE HISTORY OF MARINE THREESPINE STICKLEBACKIN OYSTER LAGOON, BRITISH COLUMBIAbyRegina Karin SaimotoB.Sc., University of Winnipeg, 1989A THESIS SUBMITTED IN PARTIAL FULFILMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIES(Department of Zoology)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAJuly 1993© Regina Karin Saimoto, 1993In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.Department of  -Z.-00W C, 1The University of British ColumbiaVancouver, CanadaDate  4 ( la ),S -{ in-fir) 19 CI (Signature)DE-6 (2/88)ABSTRACTMarine populations of threespine sticklebacks (Gasterosteus aculeatus) are possibleancestors of the many of the divergent freshwater populations of this species found along theglaciated coast of British Columbia. Until this study, the life history and reproductive behavioursof marine population along the Pacific coast of North America were unknown. Data wascollected from September 1990 to August 1992 at Oyster Lagoon, Pender Harbour, BritishColumbia. Morphology, migration and growth, reproduction, and nesting success were examinedfor this purely marine population.Morphology analysis of 11 characters on 80 males and 80 females from Oyster Lagoonindicate that this marine population is similar to many populations of threespine sticklebacks, andis intermediate to the most extreme morphological states reported in the literature. The OysterLagoon population is sexually dimorphic in post orbital length, mouth width, first dorsal spinelength, and anal fin ray number. The population is also polymorphic for lateral plates. Themorphology of 35 low and partially plated fish was compared to 160 high plated fish. Low andpartially plated fish have fewer and shorter gill rakers than high plated fish, and females of thisgroup have wider mouths than high plated females. Morphologically, the low and partially platedfish are more similar to stream resident sticklebacks than are their high plated counterparts.Migration and growth data for Oyster Lagoon fish indicate sexual maturity at one year,and that the population is iteroparous. There are two well defined peaks of out-migrations ofyoung for this population. Some young migrate in the summer of their birth (summer migrators),while other young overwinter in the lagoon (overwinterers) and migrate out the following spring.Summer migrators return as mature adults early next spring, whereas most overwinterers do notiireturn to the lagoon to breed until mid-summer. The prolonged breeding season of thispopulation lasts from March/April to September.Reproductive behaviours of both sexes are similar to those reported for other populations;however, the characteristic Zig Zag dance is absent from this population. Cannibalism is highin Oyster Lagoon, and all three types of diversionary displays (nest defence behaviours) wereobserved. Female fecundity and egg morphology is more similar to anadromous populations thanto freshwater populations.Oyster Lagoon males tend to prefer to nest under rocks, in shallow areas near theshoreline. Nests under rock were less likely to be abandoned before the end of courtship, andcontained more eggs at the end of courtship, than nests in open areas. After the end of courtshipthere was no significant difference between the two nesting microhabitats in the probability ofnest survival.In summary, the Oyster Lagoon marine population is similar to other, mainly freshwater,populations of the species. This life history study supports the notion of postglacial marineancestry for freshwater populations on this coast. When compared, the better studied east coastmarine populations, there are some differences. The question of relationship between anadromousand marine sticklebacks is discussed, but not resolved.iiiTABLE OF CONTENTSABSTRACT ^  iiTABLE OF CONTENTS ^  ivLIST OF TABLES  viLIST OF FIGURES ^  viiACKNOWLEDGMENTS  ixGENERAL INTRODUCTION ^  1STUDY SITE ^  6MORPHOLOGY ^  11Introduction  ^11Materials and Methods  ^11Results  ^13Sexual dimorphism  ^13Plate polymorphism  ^13Discussion  ^14MIGRATION AND GROWTH ^  23Introduction  ^23Materials and Methods  ^23Migration Samples  ^23Lagoon Samples  ^24Sampling Regime  ^25Population Size and Growth  ^25spine clipping control  ^26marking regime  ^26Results  ^30Spine clipping control  ^30Migration of Adults  ^30Growth of Adults  ^30Migration of Young  ^31Growth of Young  ^42Age Structure  ^42Discussion  ^51ivREPRODUCTION ^  55Introduction  ^55Materials and Methods  ^55Nest site selection  ^55Breeding behaviour  ^56Batch fecundity  ^56Results  ^57Nest site selection  ^57Breeding behaviour  ^57territoriality  ^57nest construction  ^58courtship  ^58parental care  ^65diversionary displays  ^66Batch fecundity  ^66Discussion  ^69NESTING SUCCESS ^  73Introduction  ^73Materials and Methods  ^73Sampling locations  ^73Sampling regime  ^74Collection Procedure  ^74Results  ^75Preliminary analysis  ^75Probability of nest survival to end of courtship  ^75Nesting success after courtship stage  ^76Courtship success  ^77Probability of nest survival to three days after courtship^77Discussion  ^82GENERAL DISCUSSION ^  86LITERATURE CITED ^  90LIST OF TABLESTable 3.1. Summary of morphology of Oyster Lagoon males and females. Variables known tobe size dependent were normalized to 50 mm standard length. * indicates significance at P <or = 0.00454 for Bonferoni correction  15Table 3.2. Summary of morphological comparisons of high and low/partial plated adults.Variables known to be size dependent were normalized to 50 mm standard length. * indicatessignificance at P < or = to 0.00454 for Bonferoni adjustment  17Table 4.1. A summary of the different marks used to identify different groups of fish.28Table 4.2. A summary of standard length data for the spine clip control ^32Table 4.3. Adult population size estimates for Oyster Lagoon. Numbers of marked fish in thepopulation are adjusted for numbers of marked fish out-migrating throughout the summer. Table4.3a summarizes population size estimates for left pelvic spine clipped adults (marked in earlysummer). The second population size estimates (late summer, l g dorsal spine clipped adults) aresummarized in Table 4.3b. Standard deviation for population size estimates was calculatedaccording to Bagenal (1978, p.139)  43Table 6.1. Analysis of variance for the relationship between collection day (DAY) and nestmicrohabitat (HABITAT) on egg numbers^  80Table 6.2. Analysis of variance for the relationship between collection day (DAY) and nestmicrohabitat (HABITAT) for ranked egg numbers in nests^ 80Table 6.3. Analysis of variance fro the relationship between collection day (DAY) and nestmicrohabitat (HABITAT) and male standard length ^80viLIST OF FIGURESFigure 1.1. The three different plate phenotypes of threespine sticklebacks, including low, partialand high plate morphs ^2Figure 2.1. Map of the study site (Oyster Lagoon), including the two other marine breeding sitesof threespine sticklebacks in the Pender Harbour area (Salt Lagoon and Gerrans Bay). The insetillustrates the location of Pender Harbour with reference to the Strait ofG rgia  7Figure 4.1. Changes in standard lengths (means and standard deviation) of adults caught inOyster Lagoon for 1991 and 1992. Bar graph indicates sample size ^34Figure 4.2. Sex ratios of in-migrating adults to Oyster Lagoon for the 1992 breeding season.The open circles on this graph represent the proportion of in-migrants that are male. Samplesizes are shown in the bar graph. Sex ratios were evaluated for a cumulative collection of in-migrants over seven consecutive days on a rising tide, except February 1992, which consists ofa sample for one day only  36Figure 4.3. Changes in standard lengths (means and standard deviation) of adult in- and out-migrants ^38Figure 4.4. Standard lengths changes (mean and standard deviation) for male and femalethreespine stickleback caught in Oyster Lagoon. These data are the same as for Figure 4.2, butare separated by sex  40Figure 4.5. Growth of young in Oyster Lagoon. The open circles represent adults caught atrandom in Oyster Lagoon (same as Figure 4.2). The closed circles represent young of year 1990,which were marked on their out-migration from the Lagoon in the spring of 1991 (* indicatesrecaptures of these fish). Note that fish are two years old at the end of the study and that fishreturn to the Lagoon to breed at both one and two years of age. Also note the increase instandard length within and between breeding seasons.45vi iFigure 4.6. Growth of young of year overwintering in the lagoon. Open circles represent thestandard lengths of adults caught in Oyster Lagoon at random (same as Figure 4.2) for reference.Closed circles and closed triangles represent young of year born in 1990 and 1991 respectively(* indicates recaptures of these groups). Fish were marked as out-migrants in the spring of 1991and 1992 respectively. The data point of young of year 1991 from September 1991 is the samesample as the data point of young of year 1991 summer migrators for the same month in Figure4.7 47Figure 4.7. Growth of young of year out-migrating out of the lagoon during the summer of theyear they are born. Open circles represent standard lengths of adults caught randomly in OysterLagoon (same as Figure 4.2) for reference. Closed triangles represent young of year born in1991 (* indicate recaptures of this group). Fish were marked during their out-migration in thesummer/fall of the year they were born. The data point represented by the closed triangle inSeptember 1991 is the same point as the one for the same month in Figure4.6 49Figure 5.1. Graph illustrating a predominance of nests under rocks for both the 1991 and the1992 breeding seasons ^59Figure 5.2. Nest sites shift to deeper water as the breeding season progresses for both 1991 and1992 ^61Figure 5.3. Nest sites shift off shore as the breeding season progresses, as illustrated in thisgraph for the 1992 breeding season. The same is true for the 1991 breeding season.63Figure 5.4. Relationship between log clutch size and log female standard length (R 2 = 0.644).Data were analyzed by John Bakker (U. of Arkansas)^  67Figure 6.1. Summary of data obtained for the evaluation of nesting success in different nestingmicrohabitats. Figure 6.1a shows the egg number at different nest locations and days since thecompletion of the courtship phase. Nests under rocks contain more eggs both at the end andthree days after the end of courtship than open area nests. Nests at 0 days after courtship containmore eggs than nests at 3 days after the end of courtship for both under rock and open area nests.Figure 6.1b indicates the probability of nest survival after the end of courtship with different nestlocati ns  78viiiACKNOWLEDGMENTSThis thesis would not have been possible without the help of many individuals. Inparticular, I wish to thank my supervisor, J.D. McPhail for his support, discussions andsuggestions. I also wish to thank everyone who had a chance to visit and help in beautifulPender Harbour, especially Greg Ryskie, Greg Tamblyn, Roger Kanno, Denise Derksen, RobGoepel, Sumi Aota, the Kwan twins, Stephanie, Matthew, Debbie Du, Linda Lu, Byron, hunkyrocket Roy Boy and Scooter (our loyal companion). The use of the Paton property at OysterLagoon greatly facilitated my research, and I wish to thank Richard, Gail, and mostly Julie,Shannon and Ginger for their entertainment. The support of all the people of Madeira Park wasmore than generous. Dick Repasky gave statistical advice, and Gordon Haas, Sumi Aota andFinnbarr Horgan had many helpful comments on my thesis. Dolph Schluter and Al Lewis greatlyhelped in the design of my research and gave constant encouragement. I also wish to thank theSaimoto family for the comfortable and free accommodations in Madeira Park, and Valentin andAnnemie Schiffer for their support through these three long years. Mostly, however, I wish tothank my husband and partner, Ron Saimoto. You kept me reasonably sane through the longdays in Pender.ixGENERAL INTRODUCTIONThe threespine stickleback, Gasterosteus aculeatus L., is a small (40-70 mm standardlength), teleost fish of the family Gasterosteidae, order Gasterosteiformes (Wootton 1976; Paepke1983). Marine, freshwater and anadromous populations of this species are widely distributedthroughout coastal areas of the northern hemisphere (McPhail & Lindsey 1970; Wootton 11976;Paepke 1983). The most distinguishing morphological characteristic of this fish is the presenceof bony, protective plates along the sides of the body, and the absence of scales. Three differentplate morphs are distinguished between in this species: high plated, low plated and partiallyplated (Figure 1.1).Because of its complex behavioral repertoire and apparently rapid evolution anddivergence, the threespine stickleback has been the subject of intensive research (for reviews seeWootton 1976; Paepke 1983; Russell and Russell 1984). Within the species there is a wide rangeof behaviorally, ecologically and morphologically divergent populations (McPhail 1969, 1984;Moodie & Reimchen 1976; McPhail & Hay 1983; Reimchen et al. 1985; Blouw & Hagen 1990),and many authors consider the threespine sticklebacks to be a complex of species (McPhail &Lindsey 1970; Bell 1976; Haglund et al. 1992). This diversity is characteristic of glaciated areasand is thought to stem from multiple colonizations of freshwater by a relatively homogenousmarine/anadromous population into freshwater following the retreat of the last glaciation(McPhail & Lindsey 1970; Bell 1976, 1977, 1984, 1993; Withler 1980; Ziuganov 1983; Lavin& McPhail 1985; Reimchen et al. 1985; Francis et al. 1986; McPhail 1993). Although thishypothetical origin of diverse freshwater populations is generally accepted, it has never beencritically examined.1General IntroductionFigure 1.1. The three different plate phenotypes of threespine sticklebacks, including low, partialand high plate morphs.2General IntroductionA. High plated morphB. Partially plated morphC. Low plated morph3General IntroductionDetailed studies of the freshwater populations are numerous and extensive; however, littleis known about the anadromous form (Cowen et al. 1991), and there is no study of marinepopulations on the West Coast of North America. Some marine populations on the East Coastof Canada have been studied (FitzGerald & Dutil 1981; FitzGerald 1983; FitzGerald &Whoriskey 1984; Whoriskey & FitzGerald 1984; Blouw & Hagen 1990) and Blouw and Hagen(1990) recently described sympatric marine populations of white sticklebacks and G. aculeatuson the Nova Scotia Coast. This suggests that the marine ancestors of freshwater populations maynot have been as homogenous as is usually supposed.Indeed there is evidence of differences between Atlantic and Pacific marine threespinesticklebacks. Orti et al. (in press) indicate that there are at least three distinct clades ofthreespine sticklebacks. The presence of distinct Atlantic, Pacific and Japanese clades is inagreement with a recent electrophoretic study by Haglund et al. (1992) and it is possible thatthese clades also exhibit morphological and behavioral differences. This diversity in marinesticklebacks argues that it is necessary to study the marine form in specific geographic areas toexplain the evolutionary history of freshwater populations. Information on the ecology,behaviour, and morphology of local marine populations is essential to an understanding of theevolutionary mechanisms in the diversification of freshwater populations.In particular, the West Coast of North America is notorious for a high degree ofdiversification in freshwater populations (Wootton 1976; McPhail & Hay 1983). This makes thearea a challenging site for evolutionary studies. Fossil evidence indicates the presence on thePacific coast of a marine/anadromous form, from 10 million years ago, similar to living marineforms (Bell 1977, 1993). Thus, the marine form in glaciated areas may represent a relatively old4General IntroductionGlade that has recently diversified into many isolated freshwater populations. Clearly, a studyof this marine lineage should aid in our understanding of the origins of freshwater populations.The present study investigates the life history of the marine threespine sticklebacks that breedin Oyster Lagoon, British Columbia. The Oyster Lagoon population is the first totally marinebreeding population to be studied on the Pacific Coast of Canada. Migration, growth,reproductive behaviour, fecundity and morphology were evaluated. Nest site selection and itseffect on nesting success is analyzed in detail. This general study of the life history of themarine population should serve as a comparison to other studies on freshwater and anadromouspopulations.5STUDY SITEOyster Lagoon is located in Pender Harbour, British Columbia (Fig. 2.1). Pender Harbour(lat. 49°32' long. 124°2') lies about 80km north of Vancouver on the Sechelt Peninsula whichborders the mainland side of the Strait of Georgia. Pender Harbour is connected to Georgia Straitby two entrance channels separated by Francis Peninsula. Three small creeks drain into theharbour but they provide little freshwater influence and salinity in the harbour is stable (28 to 32ppt). Three breeding sites of marine threespine sticklebacks were found in Pender Harbour(Figure 2.1). Of these, one is in Gerrans Bay and the other two are lagoon sites. Gerrans Baywas not chosen for the life history study since the population density appears to be low, and thebreeding site is not well defined. In contrast, the two lagoon sites (Salt Lagoon and OysterLagoon) are well defined sites with high densities of sticklebacks. The two lagoons are locatedwithin 1 km2 of one another. Salt Lagoon is larger than Oyster Lagoon and has two tidalchannels. This made Salt Lagoon less feasible as a site to evaluate migration. Consequently,since Oyster Lagoon is smaller and has only one entrance, it was chosen as the site for this lifehistory study. In addition, Oyster Lagoon is a manageable size, and the lagoon is relativelyisolated from Bargain Bay and Georgia Strait, but retains reasonable daily access for migratingfish. Most importantly, Oyster Lagoon is a true marine breeding site, with no freshwater inflow(other than precipitation).Oyster Lagoon connects to Bargain Bay by way of a tidal channel which is about 200 mlong (Fig. 2.1). Tidal flooding occurs on 4.0 m or higher tides on most days. Access to thelagoon is established at 3.8 m tides. The lagoon has a surface area of ca. 2 ha and a maximumdepth of ca. 2.5 m. Most of the lagoon is less than 1.5 m deep. The salinity in the lagoonranges from 30 ppt in the summer to 20 ppt in winter. A thin ice layer (ca. 5 cm) formed on the6Study SiteFigure 2.1. Map of the study site (Oyster Lagoon), including the two other marine breeding sitesof threespine sticklebacks in the Pender Harbour area (Salt Lagoon and Gerrans Bay). The insetillustrates the location of Pender Harbour with reference to the Strait of Georgia.7Study Site8Study Sitecentre of the lagoon in January 1991, but only theshore was covered by ice (< 1 cm) in January1992. The temperature in the lagoon ranges from 4 to 28°C. Zostera (eelgrass) beds coveredmore than 50% of the lagoon during the breeding seasons in 1991 and 1992. Green and brownfilamentous algae were also common in the lagoon during both field seasons. The substrate inOyster Lagoon is primarily soft with many large rocks lining much of the shoreline. Large rocksalso predominate near the tidal channel. Lagoon Road borders the south eastern shore of OysterLagoon. The lagoon is highly eutrophic and algal blooms are common during the summer. Theabiotic and biotic characteristics make Oyster Lagoon a site in which marine threespinesticklebacks occur at high densities.The vertebrate and invertebrate fauna of Oyster Lagoon clearly indicate that it is a marinecommunity. There are nudibranchs, oysters, mussels, clams, mud shrimps, ghost shrimps,tubeworms, sea urchins, starfish, barnacles, several species of sculpins, arrow gobies, sand lance,shiner perches, occasionally salmon smolts, river otters, mink, raccoons, Pacific garter snakes,kingfishers, great blue herons, common and hooded mergansers, common goldeneye, buffleheads,surfscooters, Canada geese and mallards. Some of these species are known predators onsticklebacks. Sculpins feed on adult and juvenile sticklebacks in some lakes (Foster 1988) andjuvenile sticklebacks, and a single adult, have been found in sculpin stomachs in Oyster Lagoon.Birds are probably the primary piscivores in the lagoon. Mergansers and kingfishers commonlyeat sticklebacks (Reimchen & Douglas 1984; Raven 1986) and great blue herons have beenobserved taking sticklebacks during in- and out- migration to Oyster Lagoon and Salt Lagoon.Mammals that are potential predators on sticklebacks were river otters, mink and raccoons.Garter snakes also occasionally feed on threespine sticklebacks (McPhail pers. corn.). The9Study Sitevertebrate and invertebrate fauna of Oyster Lagoon clearly indicate that it is a marine community.The presence of a wide variety of potential predators on adult and juvenile threespine sticklebackmay play an important role in the life history exhibited by this population.10MORPHOLOGYIntroductionThreespine sticklebacks are often referred to as a species complex (McPhail & Lindsey1970; Bell 1976; McPhail 1993), and freshwater populations exhibit a remarkable variability(McPhail 1993). Since many freshwater populations differ in their morphology, it is importantto describe the morphology of the putative ancestor. In this section, I evaluated the generalmorphology of the Oyster Lagoon populations.Materials and MethodsTo describe the Oyster Lagoon population, morphological analysis was performed on asample of 86 males and 85 females. The fish were collected by minnow trapping (meshsize=7mm) in the lagoon on June 22', 1992. To evaluate sexual dimorphism, data for the twosexes were compared. Since morphometric traits are size dependent, all measurements werestandardized to 50 mm standard length (McPhail 1984). These included post orbital length,mouth width, body depth, gill raker length and the lengths of the two dorsal spines. Standardizedresults were compared using the t-test with P; < or = 0.00454 for Bonferoni adjustment. Platepolymorphism (as illustrated in Figure 1.1) was evaluated by comparing the general morphologysample to a sample of 38 partially (with large gaps between plates, but possessing a keel;generally 9-25 plates), and low plated (with only anterior plates and no keel; fewer than 9 plates)fish. In total, 11 morphological characters were evaluated (Table 3.1). All of these charactersare commonly used to describe populations of Gasterosteus aculeatus. Fish were stained withalizarin red in 10% KOH. To avoid problems due to incomplete plate development, mostresearchers recommend that only individuals larger than 30 mm standard length be used for11Morphologymorphological analysis (Hagen & McPhail 1970; Hagen & Moodie 1982; Banbura 1989). Onlyadults were used for morphological analysis in this study, all of which were larger than 30 mm(min. size = 40.5 mm). Standard length, postorbital length, mouth width and body depth weremeasured to the nearest 0.1 mm with vernier callipers. Standard length was measured from thetip of the snout to the end of the hypural. Postorbital length is the distance from the posteriormargin of the eye to the end of the hypural. Mouth width was taken as the distance between theposterior margins of the mouth, where the upper jaw articulates with the lower jaw. Body depthwas measured at the point where the body was deepest. Gill raker length, and 1" and 2n d dorsalspine length were measured with an ocular micrometer to the nearest 0.01 mm. Gill raker lengthwas measured on the left anterior most gill arch unless the isopod .gill parasite Rocinellaangustata was attached. In this case, the right gill arch was used to measure gill raker length.1" and 2nd dorsal spine length were only recorded for fish with complete spines (ie. not brokenor clipped). A Wild dissecting scope (M5-48017) was used for all meristic counts. Platenumbers were counted on the left side of the body. All lateral plates were counted, includingplates in the keel. Gill raker number was counted on the left anterior most gill arch and includedall gill rakers (ie. of both upper and lower limb). The right anterior gill arch was used for fishwith R. angustata attached to the left gill. Fin rays were counted in both the anal and the dorsalfin. Fish were blotted dry before weighing on a Sartorius scale (1219MP). Morphological resultsfor males and females were kept separate, since some of the above traits are thought to besexually dimorphic (Blouw & Hagen 1990). The t-test with Bonferoni adjustment was used toanalyze the data for sexual dimorphism and differences between plate phenotypes.12MorphologyResultsSexual dimorphismSeveral researchers have noted that threespine sticklebacks exhibit sexual dimorphism inmany morphological traits (Blouw & Hagen 1990; McPhail 1992). Results of the t-test analysison standardized morphology data for the Oyster Lagoon population are given in Table 3.1. Malesand females did not differ significantly in lengths and there are no significant differences betweensexes in lateral plate numbers (left side), gill raker number and dorsal fin ray number. Also,body depth, gill raker length and 2n d dorsal spine length is similar between the sexes. However,males and females differ significantly in post orbital length with females having greater postorbital lengths than males. There is a significant difference in the number of anal fin raysbetween the sexes, with females having significantly fewer anal fin rays than males. Males havesignificantly wider mouths, and longer 1 3t dorsal spines than females. Gravid females probablyhave deeper bodies than males, however since fully gravid females were excluded from theanalysis, this difference was not observed.Plate polymorphismDuring the summer of 1992, I noticed some low and partially plated adults migrating intothe lagoon. These low and partially plated individuals were relatively infrequent (9 low and 29partially plated fish in 2779 examined). These individuals were not strays from adjacentfreshwater systems, since some of them were spine clipped recaptures of natal homers (e.g.female right pelvic recapture with 6 lateral plates; see Migration and Growth section for detailson spine clipping). During April and May 1992, I preserved 39 low and partially plated in-migrants for morphological analysis. The fish preserved in this sample were analyzed in the13Morphologysame manner as the morphological sample collected June 22n d , 1992. Since the low/partiallyplated sample was collected one to two months before the morphological sample, size dependentmorphological measurements were adjusted to 50 mm standard length (see above). Thestandardized morphological data are given in Table 3.2. Because of the small sample size, lowand partially plated fish were combined for comparison with fully plated fish. For sexuallydimorphic traits (Table 3.1), the sexes were analyzed separately. Comparisons between the twoplate morphs indicate that fully plated and combined low and partial plated fish have similarstandard length, post orbital length for males and females, body depth and anal and dorsal fin raynumber. There is no significant difference between 1s t dorsal spine for either sex, or rd dorsalspine length. High plated individuals have significantly more plates than low/partial individuals,as expected. Low and partially plated fish have significantly fewer and shorter gill rakers thanfully plated fish. There is no significant difference between the groups in mouth width of males,but in females the mouth width of low and partially plated fish is significantly larger than in highplated fish. This absence of a difference between the groups in mouth width in males may bedue to the small sample size of low and partially plated males.DiscussionThe morphology of the Oyster Lagoon population is similar to other populations ofthreespine sticklebacks (Miller and Hubbs 1969; Hagen & Gilbertson 1972; Wootton 1976; Gross1977; Paepke 1983; Francis et al. 1986; Kedney et al. 1987; Ziuganov et al. 1987; Snyder &Dingle 1989; Schluter & McPhail 1992). On the Pacific Coast, the most extreme morphologicalmeasurements are associated with sympatric pairs of threespine sticklebacks. The marine14MorphologyTable 3.1. Summary of morphology of Oyster Lagoon males and females. Variables known tobe size dependent were normalized to 50 mm standard length. * indicates significance at P <or = 0.00454 for Bonferoni correction.15VariableMales Femalest % cliff.N Mean S.D. N Mean S.D.Standard Length 87 51.47 4.570 84 54.71 6.242 -0.481 -5.70Post Orbital Length 85 41.79 0.584 83 43.06 1.026 -9.846 -3.04^*Mouth Width 85 3.48 0.457 83 3.03 0.395 6.820 12.93^*Body Depth 85 11.47 0.528 83 11.25 0.722 2.257 1.92Left Plate Number 87 31.73 1.434 84 31.51 2.946 1.231 0.69Gill Raker Number 87 20.78 1.482 84 20.82 1.544 -0.835 -0.19Gill Raker Length 85 0.86 0.126 83 0.81 0.105 2.786 5.811 5t Dorsal Spine Length 85 3.97 0.128 83 3.76 0.495 3.017 5.29^*2"d Dorsal Spine Length 85 4.30 0.407 83 4.18 0.486 1.736 2.79Anal Fin Ray Number 87 9.11 0.703 84 8.71 0.784 3.506 4.39^*Dorsal Fin Ray Number 87 11.74 0.723 84 11.57 0.731 1.610 1.45MorphologyTable 3.2. Summary of morphological comparisons of high and low/partial plated adults.Variables known to be size dependent were normalized to 50 mm standard length. * indicatessignificance at P < or = to 0.00454 for Bonferoni adjustment.17VariableHigh plated Low platedt % diff.Sex N Mean S.D. N Mean S.D.Standard Length Both 167 53.18 5.174 39 51.44 3.613 1.814 3.30Post Orbital Length Male 82 41.79 0.592 13 42.00 0.375 -1.259 -0.50Female 80 43.10 1.036 25 43.37 2.026 -0.883 -0.63Mouth Width Male 82 3.48 0.465 13 3.73 0.279 -1.881 -7.18Female 80 3.05 0.400 25 3.66 0.332 -7.231 -17.18^*Body Depth Both 162 11.39 0.650 38 11.20 0.620 1.636 1.67Left Plate Number Both 167 31.93 1.123 39 15.54 6.129 28.813 51.33^*Gill Raker Number Both 167 20.83 1.500 38' 19.95 1.874 3.123 4.22^*Gill Raker Length Both 162 0.86 0.182 38 0.75 0.130 3.523 12.79^*l" Dorsal Spine Length Male 82 3.97 0.404 13 3.68 0.363 2.436 7.30Female 80 3.75 0.498 25 3.47 0.324 2.637 7.092"d Dorsal Spine Length Both 162 4.25 0.457 38 4.03 0.352 2.778 5.18Anal Fin Ray Number Male 84 9.12 0.701 14 9.14 0.770 -0.117 -0.22Female 82 8.73 0.771 25 8.80 0.707 0.573 -0.80Dorsal Fin Ray Number Both 167 11.66 0.735 39 11.59 0.549 0.558 0.60Morphologypopulation has characteristics which are similar to, and others that are intermediate to, charactersreported for sympatric benthic and limnetic species pairs (McPhail 1984,1993). Gill rakernumbers in Oyster Lagoon are intermediate to those found in benthic and limnetic sticklebacks(Bentzen & McPhail 1984; McPhail 1984, 1993), and are similar to the anadromous populationdescribed by Ziuganov et al. (1987). Body depth in Oyster Lagoon sticklebacks is greater thaneither member of the species pair in Enos lake (McPhail 1984). Rather than recording fin raynumbers, McPhail (1984) recorded pterygiophores, but there is usually a one to onecorrespondence of fin ray number to pterygiophores (Gross 1977). Anal fin ray numbers areintermediate between the anal pterygiophores of benthics and limnetics, but dorsal fin raynumbers are smaller than those of either the benthics or the limnetics of Enos Lake (McPhail1984) and intermediate to the species pair in Paxton Lake (McPhail 1993).Sexual dimorphism is normal in threespine stickleback populations and many of thecharacters found to be sexually dimorphic in Oyster Lagoon population are often sexuallydimorphic in other populations. Typically, Gasterosteus aculeatus shows sexual dimorphism inpost orbital length / head length (McPhail 1993; McPhail pers. corn.) but gill raker number isgenerally not sexually dimorphic in this species (Gross & Anderson 1984; McPhail 1992). Incontrast, gill raker length is usually sexually dimorphic (McPhail 1992; McPhail pers. coin). Inthe Oyster Lagoon population, neither gill raker number or gill raker length were sexuallydimorphic. Dorsal spine length may (McPhail ), or may not (Crivelli & Britton 1987),be sexually dimorphic in other populations, but Oyster Lagoon males have longer ls t dorsal spinesthan females. Fin ray numbers in both anal and dorsal fins are sexually dimorphic for somepopulations (Mori 1990; McPhail 1992), but in Oyster Lagoon only anal fin ray number is19Morphologysexually dimorphic. Commonly, males smaller than females (Moodie 1970; Giles 1987; Lavin& McPhail 1993), but not in the Oyster Lagoon. The lack of a significant difference in bodydepth between the two sexes is unexpected, since this character has been found to be sexuallydimorphic in other populations (McPhail 1992, Lavin & McPhail 1993); however, since I did notuse gravid females in any morphological analysis, sexual dimorphism in body depth due toreproductive state may have been missed.The relatively large number of morphological differences between the full and low andpartial plated fish was surprising, especially given the small sample size. Low and partiallyplated individuals appear to be more like stream resident fish than high plated fish. Manyfreshwater populations are morphologically adapted to a more benthic way of life than areanadromous populations. Low plated fish (leiurus) have fewer and stouter gill rakers than highplated ones (tracchurus) (Wootton 1976; Gross & Anderson 1984) and generally are found morelittoral adapted than anadromous or marine sticklebacks. It is surprising that low and partial platemorphs occur in Oyster Lagoon, since the loss of plates is associated with freshwater life (Taylor& McPhail 1986; Baumgartner 1992; Reimchen 1992; McPhail 1993). Low plated fish, however,are capable of breeding in slightly saline water, although they generally have low tolerance tosalt water during the breeding season (Wootton 1976). Even though Hagen (1967) stated that"...the freshwater form is unknown from the sea [ie. anadromous and marine]", platepolymorphism has been described for many brackish water populations on the East Coast ofNorth America (Hagen & Moodie 1982), Norway (Klepaker in prep.) and the White Sea(Ziuganov 1983), although no low plated individuals have been reported for these populations(Hagen & Moodie 1982; Ziuganov 1983; Cowen et al. 1991). Plate polymorphism has also been20Morphologydocumented for marine and brackish water populations of the Vancouver area where thefrequency of partially and low plated individuals was found to be inversely related to salinity(Hay unpubl. data). Horseshoe Bay, with a similar salinity to Oyster Lagoon (25 ppt) had a platefrequency of 0:1:99 of low:partial:high respectively (Hay unpubl. data). This is similar to theratio found in Oyster Lagoon (0.32 low : 1.05 partial: 98.63 high).The low and partially plated individuals from Oyster Lagoon are unlikely to be strayersfrom nearby freshwater systems, since at least one of the low plated individuals (6 plates) wasa natal homer marked in the fall of 1991 as a young of year out-migrant. In addition, Heuts(1947) demonstrated that in "mixed" freshwater populations, high plated fish are better able tosurvive immersion in saline water than low plated ones. Therefore low plated strays offreshwater origin should be selected against in a marine environment. Since these different plateforms are a consistent part of the Oyster Lagoon population, this population, and other WestCoast marine populations, are polymorphic for plates. If so, the traditional method of identifyingstream hybrids between resident and anadromous sticklebacks by plate counts (eg. Hagen 1967;Hay & McPhail 1975; Wootton 1976; Gross & Anderson 1984) may be misleading. Clearly, notall partially plated fish are hybrids in the sense of crosses between different gene pools. In somecases, the extent of hybridization between resident and anadromous sticklebacks may have beenoverestimated. The Oyster Lagoon population also argues that plate polymorphism is notconfined to freshwater populations. Hay (unpubl. data) suggests that low plated marineindividuals survive longer in freshwater than their high plated counterparts, indicating thatfreshwater invasion of a polymorphic marine population would simply favour the low platedform. Thus, the selection of low plated fish from a founding population with mixed plate types21Morphologymay be a more likely explanation of the predominance of low plated fish in freshwater than theevolution of low plated fish from a homogenous fully plated ancestor.22MIGRATION AND GROWTHIntroductionThe breeding sticklebacks in Oyster Lagoon are not permanent residents of the lagoon.Both adults and juveniles migrate to and from Oyster Lagoon, and these migrations form animportant part of their life history. For this study, the primary purpose of monitoring migrationwas to document the beginning and end of the breeding migration, and the peak of juvenile in-and out-migration.Materials and MethodsThe monitoring of migration and growth of the Oyster Lagoon population included severalsampling regimes. Migrants (both in- and out-migrants) were sampled daily during the 1991 and1992 breeding seasons, and lagoon samples were taken monthly throughout both years (Sept.1990 - Sept. 1992). By sampling the lagoon population monthly, growth of juvenile fish wasmonitored during their residence in the lagoon. Some fish were marked by spine clipping toestimate population size and to supplement growth data. From recapture data on marked lagoonfish over the two breeding seasons, I also obtained information on the growth of adults andyoung during their feeding migrations outside the lagoon.Migration SamplesTo evaluate migration, I constructed a fence across the tidal channel in April 1991. Fishwere trapped in plexiglass traps (30 x 30 x 60 cm) on both sides of the fence. Captures on theBargain Bay side of the fence were recorded as in-migrants, and those on the lagoon side as out-migrants. Fish were counted and sexed every one to two days, and released on the opposite sideof the fence. Once each month, the standard lengths of the in- and out-migrants were measured23Migration and Growthto the nearest millimetre. On May 17 th, 1991 the head of water building up on the Bargain Bayside of the fence collapsed the fence. Consequently, on subsequent tides of 5 m or higher, thefence was opened, allowing fish to migrate freely between Bargain Bay and Oyster Lagoon.When the fence was open, I attempted to catch migrants in plexiglass traps, but this was notsuccessful until the middle of July 1991 when the current around the plexiglass trap was slowedby erecting a barricade of rocks. This barricade caused the sticklebacks to congregate around thetrap and many of them were caught. Even though catches of migrating fish were increased inthis way, the estimates of daily and seasonal migration when the fence was down are veryconservative. The fence was removed at the end of August 1991.In March 1992, a large funnel shaped fence was erected. The funnel fence withstood alltides, since there was no build up of water during flooding. A larger plexiglass trap (100cm x90cm x 60cm) was set permanently on the Bargain Bay side of the funnel. This trap allowed themonitoring of in-migration to Oyster Lagoon. In-migrants were counted and sexed every one totwo days and released into the lagoon. Out-migrants were not monitored every day during thesummer of 1992; however, monthly samples of approximately 200 in- and out- migrants weretaken. Observations of migratory behaviour indicated that the majority of fish were trapped, butthat some still managed to go around the funnel.Lagoon SamplesSampling in Oyster Lagoon used a combination of dipnetting (dipnet size = 40 x 50 cm;mesh size = 1 mm) and minnow trapping. Dipnetting has been used in other studies tosupplement trapping data (van Mullen 1967; Larson 1976). In the lagoon, fish were captured byovernight minnow trapping once a month for both years of the study. During the non-breeding24Migration and Growthseason, overwintering juvenile fish in the lagoon were caught by dipnetting from shore in thesame sampling areas used for trapping. Small mesh minnow traps (mesh size = 4 mm) capturedjuveniles above 17 mm standard length and dipnetting enabled sampling of fish less than 17 mmin length. The smallest fish caught were 8 mm. To standardize the dipnetting samples, Idipnetted fish until about 100-200 fish were caught, but for no more than 10 dips per sample.To avoid excessive disturbance of nesting habitat, no dipnet samples were taken during thebreeding season.Sampling RegimeIn the summer months, lagoon samples were taken on the same days as monthly in- andout-migration samples. Except for the 1992 breeding season (April - August), all fish werereleased after sexing and length measurement. From April 1992 - August 1992, monthly samplesof lagoon and migrating fish were collected and preserved in 10% Formalin. Standard length andsex were recorded. Juvenile fish in the lagoon were only sampled in the non-breeding season.Young of year caught by minnow trapping and dipnetting were counted, measured and released.These samples of Oyster Lagoon fish allowed monitoring of the winter growth of juveniles inthe lagoon, and comparison of standard lengths between adults from the different sample times.Population Size and GrowthThe sampling regime described above does not allow estimation of adult growth duringthe breeding season, because adults continually migrate in and out of the lagoon. The growthof juveniles who migrate out of the lagoon to Bargain Bay or Georgia Strait in spring andsummer also can not be described by this sampling regime. To estimate these characteristics ofthe Oyster Lagoon population, I marked both adult and juvenile fish. Fish were marked by25Migration and Growthclipping one or more of their pelvic or 1 5t or 2' dorsal spines.SPINE CLIPPING CONTROLSpine clipping may have had some effect on the survival of marked fish, but it is a widelyused technique for sticklebacks (Blouw pers. com.; McPhail pers. corn; Moodie pers. corn.). Toestimate the effect of marking fish by spine clipping, in 1991 65 young of the year were takento the lab and kept in an 80 1 aquarium filled with 30 ppt seawater. Fish were fed a mixture oflive Artemia sp. and frozen Tubifex every one to two days. After allowing the fish to acclimatizeto lab conditions for 14 days, their standard lengths were recorded and their right pelvic spineswere clipped. The fish were returned to the tank immediately after marking. The tank waschecked daily for mortalities. Dead fish were removed and their standard lengths recorded. OnMarch 10th 1992, the fish were again measured and their spines checked for persistence of themark.MARKING REGIMEDifferent groups of fish in the Oyster Lagoon population were marked with different spineclips. The summary of the marks used for the different groups of fish marked is given in Table4.1. Because mortalities due to handling and spine clipping were high on small fish, onlyjuveniles above 25 mm were marked. Juvenile fish born in the summer of 1990 that migratedout of the lagoon in the spring of 1991 were marked by clipping their 2nd dorsal spine. Youngof the 1991 year migrating out of the lagoon in August 1991 were marked by clipping their rightpelvic spine. Young of the 1991 year migrating out of the lagoon during the spring of 1992 weremarked by clipping the 2nd dorsal spine. Marking and monitoring of both groups of young ofyear 1991 coincided with Ron Saimoto's natal homing experiment (Saimoto 1993). In-migration26Migration and Growthsamples were examined for the return of these fish throughout the 1991 and 1992 breedingseasons.To estimate adult population size, two separate mark-recapture experiments wereperformed using two different spine clips. For the first estimate, fish were marked from March29 th to April 1991 by clipping their left pelvic spine. In-migrating adults were marked byclipping their left pelvic spine between April 19th and June 21" 1991. The number of out- andin-migrating marked fish were recorded, and the total number of marked fish in the lagoon wasadjusted by these known changes. Population size was evaluated monthly throughout the 1991breeding season (April - August 1991). The second mark-recapture experiment was conductedat the end of July by clipping the 1st dorsal spine (July 23r d - August 1", 1991). Fish caughtduring the second marking experiment that were already clipped (ie. 2n d dorsal returns or leftpelvic recaptures) were clipped again with the new mark. Recaptures were evaluated one day,and three days, after the marking was finished. The number of marked fish in the lagoon wasadjusted for known in- and out- migration of marked fish. Returns of adults marked in 1991were also monitored during the 1992 breeding season.27Migration and GrowthTable 4.1. A summary of the different marks used to identify different groups of fish.28Migration and GrowthGroup of Fish Type of Mark Number markedYoung of Year 1990spring'91 out-migrants rd dorsal spine 11 057Young of Year 1991fall'91 out-migrantsspring'92 out-migrantsright pelvic spine 7 4762nd dorsal spine 10 025Adult Population Estimatesearly summer 1991late summer 1991left pelvic spine 9 9301" dorsal spine 6 58529Migration and GrowthResultsThe data collected over the two years of the study give consistent estimates of the timingof in and out-migration of both young and adults. Growth data for young and adults is alsoconsistent for both years of the study.Spine clipping controlPossible detrimental effects due to spine clipping on juvenile threespine sticklebacks wereassessed in the lab. Mortalities due to marking occur shortly after the spines are clipped: 87 %of the fish that died, died in the first day. Also, smaller fish (SL < 25 mm) are affected morethan larger ones (Table 4.2). The marks were clearly identifiable on surviving fish at the end ofthe experiment (ie. 179 days after marking). The spine clip control results indicate that spineclipping probably had a minor affect on the size group of fish marked, and that identification ofthe mark is reliable, even after several months growth.Migration of AdultsMature adults begin entering into the lagoon between January and February (Figure 4.1).Initially, males arrive in larger numbers than females, but females soon form more than half ofthe in-migration catches (Figure 4.2). Adults arrived at the lagoon earlier in 1992 (Figure 4.1),probably because of the mild winter temperatures in that year. Adults migrate into, and out of,the lagoon throughout the entire breeding season. Most of the adults leave the lagoon bySeptember, although the timing of the out-migration may vary among years.Growth of AdultsThe standard lengths of adults in Oyster Lagoon tend to increase over the summer (Figure4.1), and the size of in- and out-migrants also increases in a similar fashion (Figure 4.3). Both30Migration and Growthmales and females exhibit this trend (Figure 4.4). For recaptured adults (who are two years old)the mean standard length is higher than that of the population at large (Figure 4.5). In contrast,the standard length of one year old fish generally overlaps with the standard length of randomlysampled lagoon fish (Figures 4.6 and 4.7).Population sizeBecause of low recapture rates population size estimates in the lagoon were veryinaccurate (Table 4.3). Mark-recapture population size estimates for the lagoon are unreliablebecause two of the three of its assumptions are violated. The population is not closed (there isextensive in- and out- migration) and not all members of the population have an equal chanceof being caught (sexes exhibit different behaviours). Territorial males are more restricted in theirmovements than either females or non-territorial males. They are associated with nests andtherefore less likely to enter traps. Recapture rates remain relatively high for two to four weeks,indicating that marked adults remain in the lagoon for about two to four weeks after marking,and that the majority of adults do not return in the same breeding season after leaving the lagoon.Migration of YoungMigration behaviour of the young is more complicated than was first expected, since thereare two 'types' of young of the year in the population. Many of the young emigrate in thesummer or fall of the year in which they are born. For convenience, I call these fish "summermigrators". They tend to be the larger sized individuals, and since the larger young of the yearleave Oyster Lagoon, the mean standard length of young in the lagoon decreased from Septemberto October in both years (Figure 4.6). Another large group of young of the year overwinter inthe lagoon and migrates the following spring. I call these fish "overwinterers". Out-migration31Migration and GrowthTable 4.2. A summary of standard length data for the spine clip control.32Migration and GrowthNumber of Daysafter markingStandard Length (mm)N Min. Max. Mean S.D.Total marked 0 61 15.0 44.0 25.82 6.373Mortalities 1 35 15.0 30.0 22.38 3.5344 2 25.0 31.0 28.0 4.2346 2 24.0 38.0 31.0 9.89013 1 21.0 21.0 21.0Total alive^after 179 days 21 31.0 54.0 40.7 6.33533Migration and GrowthFigure 4.1. Changes in standard lengths (means and standard deviation) of adults caught inOyster Lagoon for 1991 and 1992. Bar graph indicates sample size.34807570656055504540353025201510Migration and Growth500400 •--Co300 N"(7)200 0a100 EaN0Mi l li! _LI^1 1 L _L . „, P^Rp p -0 0 0 0 ••.1 vl •—I •••1 •■I Ir4 •..1 •■I vA •••1 •..1 .40INC1202NOZNO2CO CD CO CD 03 03 CA CD CO CO CO CO CO CO CO CO 03 CD CD CO CO CO CO COCO CO 03 CD CD CD 03 CO CO 03 03 03 CD CO 0) CO CO 03 CD CO CO CO CO 031.4 •••4 ,r4 ,e1 ,r4^4 wl .—I ,v1 gr4 grA .r.1 ....1 rI gr4 .ri vl^4 1■1 - v■I w■4 1.4 Irl111 F > c) Z IA 14 14 >4 Z •-4 0 C14 E■ ›. o Z CA 14 r4 >4 Z ►4 E.1rzl 0 0 rzl d r4 ^a. ^0^0 rzl E.) 0 r.T4 •Ng rzl 4C C14 44 0^0ril 0 Z al '-* C.T.4^d^P-1 " .4 rn 0 Z A " ca. >1 d^" 041 35Migration and GrowthFigure 4.2. Sex ratios of in-migrating adults to Oyster Lagoon for the 1992 breeding season.The open circles on this graph represent the proportion of in-migrants that are male. Samplesizes are shown in the bar graph. Sex ratios were evaluated for a cumulative collection of in-migrants over seven consecutive days on a rising tide, except February 1992, which consists ofa sample for one day only.360^Proportion malesMalesFemales03LL1LL>-Q z CD70006000 c7-'-5000a,"4000 .E-30002000 -171100001992Migration and Growth1. 0.4a. and GrowthFigure 4.3. Changes in standard lengths (means and standard deviation) of adult in- and out-migrants.38Standard Length (mm)^tv N GI cA 4, 4, Ut cn cn cn^^1 Co001001001001001001001011111111^III^illSEP 1990OCT 1990NOV 1990DEC 1990JAN 1991FEB 1991MAR 1991APR 1991MAY 1991JUN 1991JUL 1991AUG 1991SEP 1991OCT 1991NOV 1991DEC 1991JAN 1992FEB 1992MAR 1992APR 1992MAY 1992JUN 1992JUL 1992AUG 19924o55.Migration and GrowthFigure 4.4. Standard lengths changes (mean and standard deviation) for male and femalethreespine sticklebacks caught in Oyster Lagoon. These data are the same as for Figure 4.2, butare separated by sex.40DEC 1990JAN 1991FEB 1991MAR 1991APR 1991MAY 1991JUN 1991JUL 1991AUG 1991SEP 1991OCT 1991NOV 1991DEC 1991JAN 1992FEB 1992MAR 1992APR 1992MAY 1992JUN 1992JUL 1992AUG 19920U)Standard length (mm)" N IQ CA CA 46 -A. U1 01 0) 0) ■I ■,1 CO0 01 0 01 0 CA 0 cn o cn o U1 o th oISEP 1990OCT 1990NOV 1990Migration and Growthwas not monitored during the winter months, so it is not clear if there is a steady amount ofmigration out of the lagoon during the winter. Summer migrators return to the lagoon asreproductively mature adults in the spring or early summer of the year after they are born (Figure4.7). In contrast, overwinterers return as reproductively mature adults a few months after theirout-migration (Figures 4.5 and 4.6).Growth of YoungTo evaluate the growth of young of year inside the lagoon but outside the breedingseason, standard length data obtained through the different capture techniques were treatedtogether. As fish grow out of the range of sizes caught by the dipnet, they grow into the sizerange caught by small mesh minnow traps.Most of the growth of the summer migrators occurs outside the lagoon, primarily duringfall and spring and possibly during winter (Figure 4.7); whereas overwinterers grow primarilyduring spring and early summer (Figure 4.6).Age StructureThe Oyster Lagoon population reaches reproductive maturity within one year of life. Thisis supported by the growth of marked fish in the lab (Table 4.2), some of which became sexuallymature in March 1992. In addition, the standard length of one year old fish completely overlapthe standard length of lagoon fish randomly sampled (Figures 4.6 and 4.7), suggesting that oneyear old adults form the majority of the breeding population. Fish, in the wild, however, are ableto reproduce for at least two years (Figure 4.5). Of the 11 057 young of year fish marked whilemigrating out of the lagoon in the spring of 1991, 483 (4.4 %) returned as one year olds in thesummer of 1991 and 35 (0.31%) returned in the summer of 1992 (two year olds). In addition,42Migration and GrowthTable 4.3. Adult population size estimates for Oyster Lagoon. Numbers of marked fish in thepopulation are adjusted for numbers of marked fish out-migrating throughout the summer. Table4.3a summarizes population size estimates for left pelvic spine clipped adults (marked in earlysummer). The second population size estimates (late summer, 1st dorsal spine clipped adults) aresummarized in Table 4.3b. Standard deviation for population size estimates was calculatedaccording to Bagenal (1978, p.139).43Migration and GrowthDate(1991)Total # markedin population # Fish caught % RecapturePopulationsize S.D.Apr. 18 th 2 897 1 287 13.29 21 804 1490.7May 3"' 3 684 956 14.80 25 080 1906.9May 21" 5 577 1 191 13.35 41 700 2988.1May 30th 6 025 1 474 21.37 28 194 1371.1July 4th 6 251 811 8.51 73 455 8382.7July 26 th 6 327 1 228 4.30 147 139 20000.0Aug. 1" 6 323 1 163 3.18 198 836 31622.8Aug. 24th 6 300 273 2.17 290 323 118321.6Date(1991)Total # markedin population # Fish caught % RecapturePopulationsize S.D.July 26th 3 612 1 228 15.96 22 632 1449.1July 27' 4 644 1 342 26.3 17 658 774.6Aug. 1" 5 621 1 163 17.28 35 259 2377.4Aug. 24' 6 583 273 1.1 598 455 346410.244Migration and GrowthFigure 4.5. Growth of young in Oyster Lagoon. The open circles represent adults caught atrandom in Oyster Lagoon (same as Figure 4.2). The closed circles represent young of year 1990,which were marked on their out-migration from the Lagoon in the spring of 1991 (* indicaterecaptures of these fish). Note that fish are two years old at the end of the study and that fishreturn to the Lagoon to breed at both one and two years of age. Also note the increase instandard length within and between breeding seasons.45Standard Length (mm)N^IQ^GI^GISEP1990OCT1990NOV1990DEC1990JAN1991FEB1991MAR1991APRMAY199119911_4)JUN1991JUL1991AUG1991SEP1991OCT1991NOV1991DEC1991JAN1992FEB1992MAR1992APR1992MAY1992JUN1992JUL1992AUG1992• 0 >o O.C C7 ;l0 0 -4'Co co ID 0..t• -It, 01 01 rn cn^•••.10 C31 0 CJ1 0 U1 0 Ul 0111111111111110 C.11 0 CJI 0 01Migration and GrowthFigure 4.6. Growth of young of year overwintering in the lagoon. Open circles represent thestandard lengths of adults caught in Oyster Lagoon at random (same as Figure 4.2) for reference.Closed circles and closed triangles represent young of year born in 1990 and 1991 respectively(* indicate recaptures of these groups). Fish were marked as out-migrants in the spring of 1991and 1992 respectively. The data point of young of year 1991 from September 1991 is the samesample as the data point of young of year 1991 summer migrators for the same month in Figure4.7.47Standard Length (mm)N.1^W^GI^4b.^-Ph01^0^LT1^0^at01 CM 0) 01 %.41 '4 03o cn o c.n o^o•-• N.30 01 0SEP 1990OCT 1990NOV 1990DEC 1990JAN 1991FEB 1991MAR 1991APR 1991MAY 1991JUN 1991JUL 199100 AUG 1991SEP 1991OCT 1991NOV 1991DEC 1991JAN 1992FEB 1992MAR 1992APR 1992MAY 1992JUN 1992JUL 1992AUG 1992► • 0O0 ><< Q.CD CD C*It•^74:7 7/10 rt.ID 0CD CD •-■CA CA■11 =AICO Cipca (0-a 0 Migration and GrowthFigure 4.7. Growth of young of year out-migrating out of the lagoon during the summer of theyear they are born. Open circles represent standard lengths of adults caught randomly in OysterLagoon (same as Figure 4.2) for reference. Closed triangles represent young of year born in1991 (* indicate recaptures of this group). Fish were marked during their out-migration in thesummer/fall of the year they were born. The data point represented by the closed triangle inSeptember 1991 is the same point as the one for the same month in Figure 4.6.49Standard Length (mm)—• rs.) 1■3 CA CA -r• A CO CA 0) 0) ■4^J CO0 CJI 0 CJI 0 01 0 C.71 0 fit 0 CM 0 (11 011111111111111SEP 1990OCT 1990NOV 1990DEC 1990JAN 1991FEB 1991MAR 1991APR 1991MAY 1991JUN 1991o JUL 1991AUG 1991SEP 1991OCT 1991NOV 1991DEC 1991JAN 1992FEB 1992MAR 1992APR 1992MAY 1992JUN 1992JUL 1992AUG 1992■ 0Na c 3 E3 6'.CD 3 0 CaMigration and Growth9930 breeding adults were marked in the lagoon from April to May 1991 and 22 (0.22%) ofthese adults were recaptured in breeding condition the following summer.DiscussionThe movement of adults to and from breeding grounds is more important in the lifehistories of anadromous and marine populations than for many freshwater populations. Whencomparing the Oyster lagoon population to other ocean migrating populations however, they aresimilar. The prolonged breeding season of the Oyster Lagoon population encompasses mostbreeding seasons reported for freshwater and anadromous populations (van Mullen 1967; Wootton1976; Kynard 1978; Mori 1984; Crivelli & Britton 1987) including marine populations studiedon the east coast of North America (Rowland 1983, 1989c; Whoriskey & FitzGerald 1985a,1985b; Williams & Delbeek 1989).Temperature is known to influence the onset and duration of the breeding season. Inspring fed environments in Japan (constant temp. = 15°C) threespine sticklebacks reproduce yeararound (Mori 1984). The ability of the Oyster Lagoon population to reproduce over a 5 to 6months period may be due to limited tidal flooding, small size, and shallowness of the lagoon.These characters allow the lagoon to warm much faster in the spring and thus triggers an earlyonset to the breeding season (March/April). In the adjacent Salt Lagoon, which is much largerand deeper than Oyster Lagoon, adults do not appear until May. Adults caught in early May1992 in Salt Lagoon were not gravid or as coloured as adults in Oyster Lagoon, caught at thesame time, suggesting that the start of the breeding season is delayed in Salt Lagoon. Inaddition, since Oyster Lagoon is under strong tidal influence, it is cooled on most days and51Migration and Growthtemperatures rarely stay at 28°C (depth= 60cm) for more than one or two hours. Adults andyoung are able to survive the short daily periods of warm temperatures, and thus breedingcontinues over the hottest summer months.The timing of adult in- and out-migration seems variable for the Oyster Lagoonpopulation. Breeding adults remained in the lagoon until later in the year in 1991 than in 1990and reproductively mature fish arrived at the lagoon earlier in 1992 than in 1991. The winterof 1991/1992 was unusually mild due to an El Nino event. The milder winter temperaturesprobably are a major influence on the timing of migration. Gonad development in threespinesticklebacks is influenced by environmental conditions, including temperature (Baggerman 1985).An early onset of the breeding season was also observed in other threespine sticklebackpopulations along this coast (Day pers. corn.).Where the Oyster Lagoon population goes once the fish leave the lagoon is unknown.Sticklebacks have been caught far to sea in the open Pacific (Quinn & Light 1989); however theorigin of these fish is not known. Juvenile threespine sticklebacks also occur up to 110 km offshore in the New York Bight (Atlantic Coast of North America), although densities in winter ofthem are highest between 3 and 75 km from shore (Williams & Delbeek 1989). Since largenumbers of sticklebacks occur near the docks and marinas throughout the Strait of Georgia, itis unlikely that the Oyster Lagoon population migrates to the open Pacific.The striking shift from predominantly male to predominantly female in the in-migrationis not unusual. Kynard (1978) reported that, in Wapato Lake, as the breeding season progresses,females became much more numerous than males. This is contrary to findings from anadromouspopulations in Cow Pasture Creek, British Columbia, where ovulated females arrive before males52Migration and Growth(McLennan 1988).Reproductive maturity after one year of age is common in populations of G. aculeatus,however, in such populations individuals usually die shortly after breeding (Wootton 1976;Paepke 1983; Crivelli & Britton 1987). Since I have found both one and two year old fishreproducing in Oyster Lagoon, at least some individuals are iteroparous. Still, it is clear that oneyear old fish make up the overwhelming majority of the breeding population. In fact, only 35two year old recaptures were found from a total of 11057 fish marked as overwintering youngof year of 1990. In addition to the fish known to be born in 1990, I also recaptured some fishmarked as breeding adults in the spring and summer of 1991. Of this group, only 22 fish wererecaptured (9930 marked) in the summer of 1992 (two year olds). The observation that one yearold adults have a mean standard length similar to that of the randomly sampled breedingpopulation also supports the argument that one year olds form the bulk of the reproducing adults.A similar age structure has been reported for threespine stickleback populations in the Bay ofFundy (FitzGerald et al. 1989; Williams & Delbeek 1989); however, Craig & FitzGerald (1982)have reported that fecund females are two years old in their salt marsh population.Most researchers do not follow populations outside of the breeding season (Picard et al.1990). Thus comparisons of the two strategies for migration and growth of the young of yearin Oyster Lagoon is difficult. Young of anadromous populations (Crivelli & Britton 1987) andsome marine populations (Craig & FitzGerald 1982; FitzGerald 1983; William & Delbeek 1989;Ziuganov et al. 1987; Scott & Scott 1988) usually migrate at the end of the breeding season. Itis unclear, however, whether any young remain at the breeding site between seasons.Observations in Crivelli & Britton's (1987) study indicated that, if any fish remain, the number53Migration and Growthis negligible. I have demonstrated that at least some young overwinter in Oyster Lagoon. Mydata are insufficient to indicate whether young continue to migrate out of the lagoon throughoutthe winter; however, periodic observations during the winter suggest that they do not. Inaddition, it is unclear what distinguishes overwinterers from summer migrators. My data indicatethat size is a factor, but whether it is the main one is unknown.Growth is similar to some freshwater populations, although many populations ofthreespine stickleback exhibit a slower growth rate (Wootton 1976; Ryan 1984). Notsurprisingly, growth of overwintering young was slow during the winter. Growth rate appearsto be higher outside the lagoon than inside. This may reflect increased food availability and/ordecreased competition out side of Oyster Lagoon. Generally, temperate marine environmentsprobably provide better feeding grounds for fish than fresh water (Gross et al. 1988), and in thiscase feeding opportunities may be severely reduced in the lagoon during the winter.Furthermore, winter temperatures in the lagoon are lower than those in Georgia Strait. This mayalso influence growth rates, especially for summer emigrating young and for adults betweenbreeding seasons.In terms of growth and migration, it appears that the Oyster Lagoon population is in someways similar to freshwater and anadromous populations. Also, reproductive maturity is reachedat a similar age to that of other populations. Growth rate is higher for the Oyster Lagoonpopulation than for freshwater populations, but is similar to anadromous populations. Thepresence of overwintering young in Oyster Lagoon may be unique to this location. The breedingseason in Oyster Lagoon is relatively long, but this is not the case for other marine breedingsites.54REPRODUCTIONIntroductionThreespine sticklebacks exhibit a complex repertoire of breeding behaviours, and bothsexes are involved in mate choice (McPhail 1993). Some differences in courtship and nestingbehaviour have been reported for freshwater and anadromous populations (McPhail & Hay 1983)and these differences, particularly those in courtship behaviour, may help restrict gene flowbetween sympatric (Ridgeway & McPhail 1984; Hagen & Blouw 1990; Jamieson et al. 1992) andparapatric (Hay & McPhail 1975; Lavin & McPhail 1993) populations. Site selection, breedingbehaviour and parental care was evaluated for the Oyster Lagoon population. In addition tobreeding behaviour, batch fecundity was evaluated for one sample taken in June 1991.Materials and MethodsNest site selectionTo determine whether males have preferred nest sites, I selected 24 shoreline plots. The24 locations were approximately 3 m x 3 m; dimensions for each location was measured to thenearest dm. The shoreline locations were selected to represent the different kinds of substratesin the littoral zone of the lagoon. To evaluate the effect of the road on breeding behaviour andsite selection, 9 of the 24 shoreline locations were located along the road side of the lagoon. Noindication of such an effect was found, and the road side plots were omitted from observationsin 1992. Consequently, data on nest site selection refers only to the 15 plots used throughoutboth summers. Each location was observed monthly, with observations lasting for a minimumof 10 minutes. If no adults were seen in the location, the location was recorded as having nonests. If adults were seen in the location, the observations lasted for a minimum of 30 minutes55Reproductionand did not end until 5 minutes after the last nest had been recorded. Most of the observationswere done in the afternoon, but in some locations observations were conducted in the morningfor better visibility. For site selection, the following information was recorded: location of nest,type of cover at the nest, distance to cover for the male, number of males in the location,approximate depth, and approximate distance from shore for each nest. The exact depth for nestsrecorded throughout the 1991 and 1992 breeding seasons were measured on August 25 th, 1991and August 18 th, 1992, respectively. Measuring the depth of nests for 1991 and 1992 on one dayfor each year reduces any minor differences in tide height during the different days ofobservations which could affect relative depth of the nests.Breeding behaviourFor many freshwater populations breeding behaviours are recorded in distinct phases(Wootton 1976; Paepke 1983). The phases used in Oyster Lagoon were: territoriality, nestconstruction, courtship, parental care and diversionary displays. Observations on courtship,nesting and parental behaviour were conducted during evaluation of nest site selection in thesummers of 1991 and 1992.Batch fecundityEgg number was estimated on one sample of 41 females caught by minnow trapping inJune 1991. Batch fecundity was estimated for females with mature (not fully grown, but easilycountable in follicles) and ripe (fully grown) eggs before, and after, absorption of water. Onlyripe eggs that had absorbed water were used for egg diameter estimates. Ripe eggs before andafter the absorption of water were used for weight estimates. Batch fecundity refers to thenumber of mature eggs in both ovaries. For ripe eggs only, data were analyzed with Pearson56ReproductionCorrelation and Linear regression for ripe eggs after the absorption of water only (N=28).ResultsNest site selectionShoreline locations chosen for the analysis of nest sites included different substrate types,tidal influences and depth ranges. The most striking evidence for nest site selection for bothbreeding seasons was the predominance of nests under rocks (Figure 5.1). In. addition, nests werefound in very shallow water in both 1991 and 1992, especially during spring and early summer(Figure 5.2). Nest sites were found at greater depth towards the end of the breeding season(Figure 5.2), and distance from shore also increased with the progression of the breeding season(Figure 5.3). Only data collected during the 1992 breeding season are shown in Figure 5.3, butthe trend is identical in the 1991 breeding season. The increase in distance from shore is relatedto nest sites shifting to deeper water later in the season.Breeding behaviourFive categories of breeding behaviours were observed during the study: nest building,territoriality, courtship, parental care and diversionary displays. As the breeding seasonprogressed, males remained inactive for most of the day. Many of them remained hidden underrocks when temperatures were warmest and were active in early morning and late afternoon. Atthese times their nest areas were shaded. This behaviour may allow the Oyster Lagoonpopulation to reproduce even during the hottest summer months.TERRITORIALITYMales in the Oyster Lagoon population establish territories and defend them with57Reproductionaggressive behaviours. Aggressive behaviours include threat displays and physical attacks onintruders. Males will charge intruders and/or threaten them by displaying their spines.Neighbouring males may swim side by side along the edge of their territories. Upon intrusions,territorial males may physically attack the intruder by biting, charging, or grabbing the intruderby his spines.NEST CONSTRUCTIONNests collected in Oyster Lagoon consist of mud, eelgrass, filamentous algae, small sticks,shell pieces and douglas fir needles. Males often swim several meters to collect suitablematerials for nest construction. Males form an excavation in the substrate, deposit some nestingmaterial and spit mud over the nest. They glue (produced by kidneys) the nest material togetherand poke at the nest to form an oval tunnel with two small openings. More nesting material isthen stuffed into the nest, and more glue added. The nest building phase usually lasts one tothree days. After the nest is almost complete, the male creeps through the nest and periodicallyglues the nest and spits more mud over it.COURTSHIPTerritories are usually relatively large (territory radius is often > 20 cm at the onset ofcourtship phase) until the end of the courtship phase, but as the male completes his nest, hebecomes less territorial towards females. In the Oyster Lagoon population, either males orfemales can initiate a courtship. No Zig-Zag dances were observed. The two individuals simplyapproach each other, with the male swimming slightly above and at a right angle to the female.A responsive female usually gives the head-up response and the male then turns and swimstowards his nest. During courtships males often are very aggressive to intruders, and receptive58ReproductionFigure 5.1. Graph illustrating a predominance of nests under rocks for both the 1991 and the1991 breeding seasons.59 \\I UNDER ROCKSOPEN AREASp < 0.05p < 0.0011991**Reproduction7570656003 550 5045403507V 3025201510507570656055.n0 5007 45•..- 403507 3025446 20151050* *<>-^z* *1992UNDER ROCKSOPEN AREASp < 0.05**^p < 0.00160ReproductionFigure 5.2. Nest sites shift to deeper water as the breeding season progresses for both 1991 and1992.61depth (cm)APR 1991MAY 1991JUN 1991JUL 1991AUG 1991   1     APR 1992MAY 1992JUN 1992JUL 1992AUG 199211111i0 N3^0) 03 •-• "0 0 0 0 0 N30 c)Number of nestsReproductionFigure 5.3. Nest sites shift off shore as the breeding season progresses, as illustrated in thisgraph for the 1992 breeding season. The same is true for the 1991 breeding,L_ 100OL..4—ora 80CC.«.,co6 604020100 as...,80 U343C60 w..040 L-e20 EmzAPR^MAY^JUN^JUL^AUG199264Reproductionfemales wait in the head-up position until the male chases off the intruder. Males swim eitherdirectly towards the nest, or in a wide circle around and down towards the nest. In either case,the male stops periodically and checks whether the female is following. Once the male reachesthe nest, he fans, pokes the nest and glues. He then swims rapidly back to the female and thenreturns to the nest. If, when he reaches the nest, the female close behind he shows her the nestopening by lying on his side, with his snout pointed at the nest opening. The female inspectsthe nest and enters. Once the female is in the nest, the male pokes, and sometimes bites, at hercaudal peduncle. The female leaves after depositing the eggs and the male chases her from histerritory. He then rapidly returns to his nest and fertilizes the eggs. A male may obtain severalbatches of eggs during a reproductive cycle. Up to 4 batches (min=0, mean= about 2; N=28)were found in nests collected from Oyster Lagoon. Males and females frequently break offcourtships at any time.PARENTAL CAREParental care includes territorial behaviour, fanning of the eggs, and care of young. Afterthe end of the courtship phase, the territory defended by the male decreases in size (territoryradius approx. 5 cm). Males fan the eggs by creating a current towards the nest entrance withtheir pectoral fins. Because of my sampling and the low survival rates of nests, I was unable tofollow any individual male through an entire reproductive cycle. Eggs reared in the laboratoryat 17°C took about 6 to 7 days to hatch. Some males guarding fry where observed in the lagoon,indicating post-hatching parental care. Some of the nests collected from the lagoon at three daysafter the end of courtship (see Nesting Success Section) contained both well developed eggs andsome hatched individuals.65ReproductionDIVERSIONARY DISPLAYSCannibalism is common in Oyster Lagoon. Schools of females and non-territorial males(up to about 100 individuals) cannibalize eggs in nests. Also, territorial males without eggs intheir nests often join nearby cannibalistic schools and forage with them for short periods. Theprimary defence of males with eggs against these foraging schools is to divert them from the nestwith a diversionary display. All three types of diversionary displays described in Foster (1988),including snout tapping, shimmering and rooting were observed in the lagoon. In addition,parental males frequently join a school until it is past their nest. The frequency of attacks onnests is not uniform across the breeding season: more attacks are observed in May and June thanduring other months.Batch fecundityA sample of 41 gravid females from Oyster Lagoon were analyzed for egg number byJohn Baker (U. of Arkansas). Only ripe females (with ovulated eggs) were used for the analysis.Log egg number in the Oyster Lagoon population is strongly correlated with log standard length(SL) (Figure 5.4), with an adjusted r 2 = 0.644 for 28 females. The regression equation for thisgraph is:logo egg # = 2.8300 (log io SL) - 2.7039There is little variability in egg diameter (mean = 1.786 mm, SD = 0.050 mm), and egg dryweight (mean = 5.45x10' g, SD = 5.33x10 -5 g). Egg diameter and egg dry weight are stronglycorrelated (Pearson correlation coefficient: r=0.80), indicating that the amount of egg swellingbefore ovulation is proportional to its dry weight (Baker pers. com .).66ReproductionFigure 5.4. Relationship between log egg number and log female standard length (R2 = 0.644).Data were analyzed by John Bakker (U. of Arkansas).67Reproduction0 OOOO 0OO2.30.o^E 2.25^ 03z O 00,• 2.200^002.150-J 00^02.10^ 000S.%2.452.402.352.052.00001.951.60 1.62 1.64 1.66 1.68 1.70 1.72 1.74 1.76 1.78 1.80Log(10) S.L. (mm)68ReproductionDiscussionThe Oyster Lagoon population is similar to many freshwater and anadromous sticklebackpopulations with respect to their reproductive biology and behaviour. Nest site selection,reproductive behaviours, and batch fecundity are similar to those described for many anadromousand freshwater populations, and intermediate to the extremes described for the species.In other populations of the species nest site selection is often dependent on depth and theavailability of cover (Tschanz & Scharf 1971; Kynard 1978; Sargent & Gebler 1980; Sargent1982; FitzGerald & Whoriskey 1984; Giles & Huntingford 1984). In some lakes, males maymove their nests to deeper water as temperature increases and water level decreases (McPhailpers. corn.). This occurs in Oyster Lagoon where nest depth and distance from shore increaseas the season progresses. The increase in distance from shore is probably an incidental result dueto the selection for deeper and cooler water as temperature increases. FitzGerald & Whoriskey(1984) and Kedney et al. (1987) found a male preference for nest sites closer to pool shores, andattribute this to increased protection from nest predators. The disproportional number of nestsfound under rocks in Oyster Lagoon also suggests a preference for cover by males. Thispreference fornesting near, or in cover is not universal. Male sticklebacks in some populationstakeadvantage of available cover (Moodie 1970, 1972; Wootton 1976; Sargent 1982; Paepke1983; Giles & Huntingford 1984), whereas in other populations they do not (Worgan &FitzGerald 1981; Kedney et al. 1987; Zyuganov et al. 1987).Most breeding behaviours recorded for other populations of threespine stickleback occurin the Oyster Lagoon population. They are territorial, build nests, and fan eggs duringdevelopment. The number of clutches obtained by a male (mean ca. 2, min=0, max=4; N=28)69Reproductionalso is similar to several other populations (Wootton 1976). Observations indicate that males arethe initiators of courtship early in the breeding season; however, females initiate most courtshipslater in the season. Apparently, this shift is closely related to the ratios of reproductive males tofemales (see Migration and Growth section), and is a natural consequence of the asymmetrybetween the sexes in the length of their reproductive cycles: females can produce a new batchof eggs in 4-10 days, while males take at least 2 weeks to complete a cycle (Wooten 1976).All behaviours used in courtship are common to other populations, although the usual Zig-Zag dance was not observed at Oyster Lagoon. There are at least two explanations for theabsence of the Zig-Zag dance, which are not necessarily exclusive. Oyster Lagoon is shallowand nests are often situated in shallow water. Perhaps the shallowness of the lagoon precludesextensive Zig-Zags. Alternatively, the absence of the Zig-Zag dance may be due to the highlevel of nest predation in Oyster Lagoon. Foster (1990) has suggested that the frequency of Zig-Zag dances is inversely related to the extent of cannibalism. Giles and Huntingford (1984)suggest that it may be associated with predation risk to adults. In many freshwater populations,Zig-Zag dances are greatly reduced (McPhail & Lindsey 1970), and even absent (Foster 1990).Therefore, the absence of this behaviour is not unique to Oyster Lagoon. Indeed, fish frompopulations that do not exhibit Zig-Zags often perform Zig-Zag displays when brought to the lab(McPhail pers. corn.). This suggests that this conspicuous courtship behaviour may depend oncontext.The parental phase is similar to that commonly observed in freshwater. Fanning of nestsand defence of young is ubiquitos in the threespine stickleback; however, these behaviours areabsent in the closely related white stickleback found on the East Coast of North America. The70Reproductiondistraction displays used by the Oyster Lagoon population also are similar to those used byfreshwater populations that encounter nest raiding (Foster 1988, in press; Whoriskey 1991). InOyster Lagoon, cannibalism appears to increase as the breeding season progresses and reachesa maximum in May and June. This seasonal trend probably reflects the changing abundance ofnon-gravid females in the lagoon. At the onset of the breeding season, males predominate in thelagoon, and the number of in-migrating females initially is low (see Migration section); however,as the season progresses, the number of in-migrating females soon outstrips the number of in-migrating males. The number of females in the lagoon increases rapidly from March/April toMay/June. During July and August, most of the in-migrants are again male, and the proportionof lagoon fish that are females decreases. During these months, there is a decrease incannibalism. Since the majority of cannibals are females, female dynamics probably play a majorrole in the frequency of cannibalism.Again, egg number is similar to that has been observed in other threespine sticklebackpopulations. Batch fecundity is variable and clearly dependent on female body size (Moodie1970; Wootton 1976; Paepke 1983; Snyder & Dingle 1989). Mean batch fecundity in the OysterLagoon population is intermediate to that of other populations (Moodie 1970; Wootton 1976);however, it is larger than many freshwater populations (McPhail & Peacock 1983) and smallerthan some marine populations from the East Coast (Coad & Power 1983; Craig & FitzGerald1982; FitzGerald 1983). This is not surprising since female lengths were larger for thesepopulations than for the Oyster Lagoon population. Egg diameter (Snyder & Dingle 1989), dryweight and the correlation between diameter and weight are similar to those of many otherlocations examined by John Baker (pers. corn.). However, egg diameter is larger than those71Reproductionreported by Craig & FitzGerald (1982) for estuarine/marine populations in the St. Lawrence.There are no striking differences between batch fecundity data obtained for the Oyster Lagoonpopulation and those obtained for many other populations. I was unable to estimate fecunditydirectly since I was unable to estimate inter-clutch intervals. Reiffers (1984 as in McLennan1988) found an inter-spawning interval of about 19 days whereas Wootton (1976) reports theinter-clutch interval to be 4-10 days. If females remain in the lagoon for approximately 1 month(see Migration and Growth section), they could spawn 3-4 times with such a spawning interval.With a mean clutch size of 161 eggs, female fecundity would be around 482-644 eggs. Thiscrude estimate is similar to the mean fecundity of approximately 500 eggs for females of thesaltmarsh populations studied by Coad & Power (1973).In summary, the reproductive ecology of the Oyster Lagoon population is similar to manypopulations of threespine sticklebacks from a variety of different habitats. Batch fecundity ismore similar to anadromous and East Coast marine population than to freshwater populations.Excepting the Zig-Zag dance, this population exhibits most of the reproductive behavioursdescribed for other populations. Nest site location is similar to some, and different from otherdescribed populations. In general, the data presented on the reproductive behaviour and ecologyof this marine population indicates that behaviourally similar marine populations could havefounded the diverse freshwater populations that now occur along this coast. The few differencesbetween this marine population and described anadromous populations open the possibility thatthe marine and/or anadromous colonizers of freshwater may not have been homogenous.72NESTING SUCCESSIntroductionWhile documenting nest locations, a disproportionate number of nests were found underrocks as opposed to other microhabitats. This could be due to a preference for such nest sites,or to increased survival of nests under rocks. Males nesting under rocks may be more successfulin defending their nests than males in the open, but less successful at courting females, sincethese nests are less conspicuous. Thus, there may be a "trade-off' involved in male nest siteselection. Conversely, females may prefer nests under rocks over those in other microhabitats.In this case, nesting sites under rocks should be preferred by all males, and therefore occupiedfirst.Materials and MethodsTo evaluate the affect of nest site on male reproductive success, I collected nests at twodifferent stages in the reproductive cycle: at the end of the courtship phase and 3 days after theend of courtship. The purpose was to evaluate differences in both courtship success and nestsurvival. The frequency of nest abandonment before the end of courtship was also recorded forboth, nests in the open and nests under rocks.Sampling locationsFive sampling locations along the shoreline of the lagoon were chosen. These locationswere close to each other and did not include localities used for collecting other nest site andreproductive behaviour data. All locations were in an area with high nest densities.73Nesting SuccessSampling regimeMales with nests under construction or males in early courtship phase were selected.They were monitored daily, and observations on behaviour, nest location, approximate territorysize and reproductive phase were recorded. Either on the first day of parental care (male shiftsto aggression towards females), or three days after the onset of parental care, males werecollected. Nests were assigned a collection day (0 or 3 days after the end of courtship) whenthey were first recorded. I purposely biased my samples to be larger for 3 days after the end ofcourtship, since variability in egg number was expectedly large for these samples.Collection procedureI attempted to capture all males with their nests. This was done by dipnetting from shore,but some males abandoned their nests before their assigned collection date and thus they werenot collected. All such abandoned nests contained 0 eggs, and had probably been raided bypredators or cannibalized by the male. Some males were difficult to catch, and this reducedsample sizes of males and their nests. Since it quickly became apparent that nesting success waslow, all instances of nest abandonment were recorded. Also depth, distance from shore, distancefrom nearest neighbour, approximate territory size, and distance to nearest cover was recordedfor all nests at the time of collection. Captured males were weighed and measured in the lab andegg number and stage of development were recorded for all collected nests. Dead eggs in nestswere included in the total number of eggs per nest for analysis. Males and nests were preservedin 10% Formalin.74Nesting SuccessResultsPreliminary analysisTo test the effect of nest site on nesting success, other factors influencing nesting successmust be accounted for. For example, "better" males may obtain "better" nest sites, thusconfounding the effects of male quality and nest site quality. Other factors, such as depth,distance from shore and distance from nearest neighbour also may influence nesting success.Multiple regression (Zar 1984) of all measured factors was performed on egg number. Since eggnumbers in nests collected 3 days after the end of courtship were not normally distributed,multiple regression analysis was not performed on these nests. For both nests under rocks andnests in the open, there was no detectable effect of male size, depth, distance from shore, territorysize, distance from nearest neighbour or collection date on the number of eggs in the nest at theend of courtship phase. This suggests that any differences between egg number in the two typesof nests is due to differences in nest microhabitat.Probability of nest survival to end of courtshipDuring data collection, it became apparent that many males never reach the end ofcourtship phase. These abandoned nests are indicative of nesting success. The differences in theprobability of a nest surviving to the end of courtship phase were analyzed using log-likelihoodanalysis with a Yates correction. All nests abandoned before the beginning of the parental phasewere included in the analysis. Of 74 nest found under rocks, 60 survived to the end of courtshipphase. Of 82 nests observed in open areas, 36 survived to the end of courtship phase. Clearly,the survival of nests in the open is significantly lower than those under rocks(0=21.768;p<0.001).75Nesting SuccessNesting success after courtship stageDifferences in egg number between the two nesting habitats are illustrated in Figure 6.1a.To estimate the effect of nest site on nesting success immediately after and three days after, theend of courtship, a two-way ANOVA was performed on the data. Only males that completedthe courtship phase were used in this analysis. Of the 156 males with that were nests initiallyrecorded, 96 reached the end of courtship phase. Of these 96 nests, 30 were collected at the endof courtship (12 in the open and 18 under rocks). Of the remaining 66 nests, 27 were in theopen and 39 under rocks at three days after the end of courtship. Two-way ANOVA on this dataindicates that habitat, day of collection, and interaction between habitat and day all havesignificant effects on egg number (Table 6.1). However, egg number at 3 days after the end ofcourtship is not normally distributed. ANOVA on ranked data (Conover & Iman 1981) wasperformed to alleviate this problem. Results of the ANOVA on the ranked data are presentedin Table 6.2. Habitat and day of collection still had a significant effect on egg number in nest,but the interaction between habitat and day is not significant. For nests collected 3 days afterthe end of courtship, the ranked data for egg number in nests also violated the assumption ofnormality for nests collected at 3 days after the end of courtship. Thus, the results of the twoANOVAs performed on the data should be viewed with caution; however, both results suggestthat nests under rocks have higher egg numbers than nests in the open. In addition, nestscollected at the end of courtship have more eggs than those collected three days later. This resultis expected, since males are unlikely to gain eggs once they switch to the parental phase, but theycan lose eggs to cannibalism and other forms of nest predation. Even though multiple regressionanalysis showed no significant effect of male size on egg number, this result also should be76Nesting Successviewed with caution. ANOVA for differences in males size between nesting habitats andcollection date indicate that male size is significantly different between the two nesting habitats:males nesting under rocks are larger than in the open (Table 6.3). No such difference was foundbetween male size and collection date.Courtship successAn increase in nesting success for males nesting under rocks as opposed to those nestingin the open, in part could be achieved by the former obtaining more eggs during courtship. Thusa comparison of egg numbers in nests immediately following the end of courtship should be agood indicator of courtship success. Egg numbers in nests under rocks and in the open werecompared using a two tailed t-test. Males with nests under rock obtained significantly more eggsthan those in the open (Figure 6.1a), t=7.89, p<0.05.Probability of nest survival to three days after courtshipDifferences in nesting success between the two habitats could also stem from differentialsurvival of nests under rocks and nests in the open. Such differences were found for nests beforethe end of courtship, with males nesting under rocks being more likely to reach the end ofcourtship phase than males nesting in the open. Two of 28 open area nests, and 6 of 40 nestsunder rocks survived to three days after the end of courtship. This difference is not significant(G=0.16, p>0.05; Figure 6.1b).77Nesting SuccessFigure 6.1. Summary of data obtained for the evaluation of nesting success in different nestingmicrohabitats. Figure 6.1 a shows the egg number at different nest locations and days since thecompletion of the courtship phase. Nests under rocks contain more eggs both at the end andthree days after the end of courtship than open area nests. Nests at 0 days after courtship containmore eggs than nests at 3 days after the end of courtship for both under rock and open area nests.Figure 6.1b indicates the probability of nest survival after the end of courtship with different nestlocations.780 Days^3 DaysTime since cessation of courtshipunder rockopenNest microhabitat79B) SuccessNesting SuccessTable 6.1. Analysis of variance for the relationship between collection day (DAY) and nestmicrohabitat (HABITAT) on egg numbers.Table 6.2. Analysis of variance for the relationship between collection day (DAY) and nestmicrohabitat (HABITAT) for ranked egg numbers in nests.Table 6.3. Analysis of variance for the relationship between collection day (DAY) and nestmicrohabitat (HABITAT) and male standard length.80Nesting SuccessEGG^NUMBERSource SS df MS F-ratio ProbabiliyDay 4 118 560.862 1 4 118 560.862 135.614 0.000Habitat 282 483.449 1 282 483.449 9.301 0.003Day * Habitat 163 962.032 1 163 962.032 5.399 0.022Error 2 885 130.411 92 30 369.794RANKED^EGG^NUMBERSource SS df MS F-ratio ProbabiliyDay 31 966.978 1 31 966.978 235.690 0.000Habitat 678.710 1 678.710 4.084 0.046Day * Habitat 102.784 1 102.784 0.619 0.434Error 15 454.753 92 166.180STANDARD^LENGTHSource SS df MS F-ratio ProbabiliyDay 0.831 1 0.831 0.035 0.853Habitat 123.746 1 123.746 5.152 0.026Day * Habitat 64.565 1 64.565 2.688 0.105Error 1 753.229 73 24.01781Nesting SuccessDiscussionThe nest collection data clearly demonstrates that nest site influences reproductive success.This is not surprising, since the predominance of nests under rocks suggests that rock sites arebetter than open nest sites. Nesting habitat appears to affect the probability that a nest willsurvive to the end of courtship and males with nests under rocks obtain more eggs than nests inthe open. The mechanisms by which courtship success and nest viability are increased when thenest is under a rock are not clear.Differences in nesting success between under rock sites and open sites could be due toother factors associated with the nest microhabitat; however, analysis of other characters on eggnumbers in the nests collected at the end of courtship phase were not significant. This indicatesthat the differences between nesting microhabitats probably are due to nest position (under rocksor in the open). This is somewhat surprising since other researchers have found that territory size(van Mullen 1967; Black 1971; Goldschmidt & Bakker 1990) and depth (Kynard 1978, 1979;FitzGerald & Whoriskey 1984) affect male reproductive success; however, there are suggestionsthat territory size is less important than cover (Kynard 1978; Sargent & Gebler 1980; Sargent1982; FitzGerald 1983; Whoriskey & FitzGerald 1985). Male behaviour, particularly aggression,also influences nesting success (van dem Assem 1967; Berg 1985; Ward & FitzGerald 1987), butthis was not evaluated in my study. The significant difference in male size between the twonesting microhabitats, however, may confound some of my conclusions. Male size is frequentlycorrelated with male quality in other species (Rowland 1983) and this relationship has beensuggested for threespine sticklebacks (Rowland 1989b), but Kynard (1972) and Sargent (1982)found no effect of male size on the number of eggs in nests. Also in marine populations on the82Nesting SuccessEast Coast of North America, male quality is only poorly correlated with reproductive success(Whoriskey 1985). Once a territory is established, there is no correlation between male size andability to defend a nest site (Sargent & Gebler 1980). In addition, Wootton (1976) found sizeto be unimportant in intraspecific aggression in Gasterosteus aculeatus. Still, Rowland's (1989a)data suggests that female threespine sticklebacks prefer larger males over smaller males andconsequently possible differences in male size may have inflated differences in nesting successbetween the two types of nest sites.The differences in male size between the two nesting microhabitats is an interesting resultin itself. Larger males are sometimes more aggressive (Rowland 1989b) and, if under rock sitesare preferred, large males may obtain more of these nest sites than small ones. However, therelevance of male size to reproductive success and the ability to defend a territory remainsunclear.Data on the probability of nest survival to the end of courtship phase suggests that underrock nests are more likely to survive than open nests. Other researchers examining nestingsuccess in the field, usually collect nests after the courtship phase (ie. when males are fanning),and do not follow nests from nest construction through to the parental phase. It is unclearwhether the low probability of nest survival to the parental phase is unique to Oyster Lagoon,but my data suggest that the lack of correlation between nest site and reproductive success inother field studies may be due to the sampling technique. Most field studies miss this stage (eg.Kynard 1978; FitzGerald 1983), where I found some of the most striking differences betweennesting microhabitats. The differences in the probability of a nest surviving to the end ofcourtship indicate that part of the predominance of nests under rocks may be explained by their83Nesting Successlonger persistence.Nesting success after courtship is also influenced by nest site in Oyster Lagoon. ANOVAdemonstrates show that nest site influences the number of eggs in the nest. In addition, eggnumber decreases with increased time since the end of courtship. This egg loss may be due tocannibalism, nest predation, or nest destruction. There is also a possible interaction between nestsite and the number of days since cessation of courtship on egg number. This interaction isunclear, since it is only significant for un-ranked data. Ranked data showed no significantinteraction between collection day and nest site. The lack of a definite interaction betweencollection day and nest site indicates that, after the end of courtship, the effect of site on nestsurvival is weaker than before the end of courtship.The effects of nest site on egg number, as shown by the ANOVA, are at least partiallydue to differences in courtship success. The 2-way t-test indicates that nests under rockscontained more eggs at the end of courtship than nests in the open. In part, this could be dueto females selecting larger males who preferentially nest under rocks, or due to females selectingnests on the basis of nest sites. Territory quality (in this case, nest concealment) was the primaryfactor affecting female choice in Sargent's (1982) experiment; however, females did notpreferentially court males nesting under rock. Males nesting under rocks had fewer intrusionsduring courtship and this produced a higher probability of a successful courtship (Sargent 1982).Sargent's finding is supported by several other studies which correlate frequency of intrusionswith both courtship (van dem Assem 1967; Sargent 1984; Borg 1985; Ward & FitzGerald 1987)and parental success (van dem Assem 1967; Sargent 1985). In contrast, Whoriskey (1984 as inWard & FitzGerald 1987) found no difference in nest site quality between nests with, or without,84Nesting Successeggs after courtship.There was no significant difference in nest survival from 0 to 3 days after the end ofcourtship between the two nesting microhabitats. Some researchers have found that nest site caninfluence nest survival (Moodie 1970, 1972; Kynard 1978; Sargent & Gebler 1980). Whoriskey(1985) suggests that differential survival between nest sites is not dependent on nest site per se,but that the amount of nest cover may influence the male's ability to obtain eggs and defend hisnest. Since few nests survived to the end of courtship stage, the lack of a significant differencein this aspect in my study may be due to small sample size.In addition to a lower probability of egg predation, and a higher probability of courtshipsuccess, males nesting under rocks may increase the quantity and quality of their parental care.Since the nest is hidden, these males may spend more time on fanning eggs and less time onperforming diversionary displays or chasing intruders. Also the daily temperature under rocksprobably fluctuate less, and the maximum daily temperatures probably is lower than in nests inthe open. Unfortunately, I did not record temperatures under rocks and in the open in OysterLagoon.The Oyster Lagoon data clearly indicate that nest site influences nesting success in at leasttwo ways: nests under rocks are more likely to reach the end of courtship and nests under rockshave more eggs than nests in the open. The influences of male quality and nest site qualityremains confounded, but in Oyster Lagoon it is clear that males nesting under rocks do betterthan males that nest in the open.85GENERAL DISCUSSIONThe life history of the marine threespine stickleback population breeding in OysterLagoon appears similar to most known populations, and less extreme than some. Several authors(e.g. McPhail and Lindsey 1970; Bell 1976) have suggested that marine or anadromoussticklebacks postglacially colonized fresh water and gave rise to the diverse freshwaterpopulations now found along the Pacific Coast of North America. Curiously, however, there isno study of the life history of a marine population in western North America. Since marinethreespine sticklebacks are the putative ancestors of many freshwater populations, a study ofmarine threespine sticklebacks is basic to an understanding of the evolutionary mechanisms thatgave rise to freshwater populations. It is not known whether the Oyster Lagoon population isrepresentative of other marine populations, but the Oyster Lagoon population is polymorphic forlateral plates. Hay (unpubl. data) also found plate polymorphism in marine populations nearVancouver. If such polymorphism is common in marine sticklebacks, perhaps the putativemarine ancestor was not as homogenous, as previously proposed (McPhail & Lindsey 1970; Bell1976, 1977, 1984, 1993; Withler 1980; Ziuganov 1983; Lavin & McPhail 1985; Reimchen et al.1985; Francis et al. 1986; McPhail 1993). Aside from lateral plates there may be other,ecologically more important differences in morphology, behaviour and ecology among marinepopulations. Apparently, gene flow between geographically separate marine populations probablyis limited (Saimoto 1993) and this may allow for divergence between marine populations.One problem concerning the putative ancestor of freshwater populations needsclarification. There is confusion in the threespine stickleback literature between anadromous and"pure marine" populations. Many researchers only distinguish between two life history forms,86General Discussionthe resident freshwater form and the anadromous or migratory form (Hagen 1967; Miller &Hubbs 1969; Bakker & Sevenster 1988). Anadromous populations migrate between the sea andfreshwater. Some authors have ignored the existence of pure marine populations (Hagen 1967;Miller & Hubbs 1969; Gross & Anderson 1984), or lumped them together with the anadromousform (McPhail & Hay 1983). Anadromous populations are often referred to as marine, eventhough they clearly breed in freshwater (Hagen 1967; Miller & Hubbs 1969; Gross & Anderson1984; McPhail & Hay 1984). There is a need for clarification of the three distinct life historytypes found in threespine sticklebacks. Differences between marine and anadromous populationsindicates that the proposed marine/anadromous ancestor is genetically heterogenous. Thepresence of true marine populations needs to be acknowledged in the literature, and their role inthe evolutionary history of the Gasterosteus aculeatus species complex should be considered andinvestigated further.There are similarities and differences between the Oyster Lagoon population and someof the marine populations studied on the East Coast of North America. On the West Coast,breeding sites, at least, in the Pender Harbour area are located either in lagoons (Oyster and SaltLagoon) or in shallow, protected bays (Gerrans Bay). As the breeding season progresses, somenests can be observed under docks, in shallow and protected areas that exhibit complete tidalfluctuations (pers. observ.). On the East Coast of North America, the best studied marinepopulations breed in tidal pools which flood approximately twice every month (Dutil &FitzGerald 1981; FitzGerald 1983; FitzGerald & Whoriskey 1984; Whoriskey & FitzGerald1985b) or in other protected areas (Worgan & FitzGerald 1981; Borg 1985; Whoriskey et al.1986; FitzGerald et al. 1989; Blouw & Hagen 1990). Oyster and Salt Lagoon usually flood once87General Discussionevery day. This creates frequent access to the breeding sites and strongly influences the physicaland biological environment. Thus, the frequency of flooding could lead to differences in breedingecology. For example, East Coast tidal pool populations appear to breed synchronously(Whoriskey & FitzGerald 1985b), and new pulses of males and females enter the pools aboutevery two weeks (Whoriskey et al. 1986; Whoriskey & FitzGerald 1987). In contrast, OysterLagoon is accessible from the sea almost every day and breeding appears to be asynchronous.Males and females at all reproductive stages are present throughout the protracted breedingseason. Tidal flooding also influences the physical environment. Oxygen levels, temperaturesand salinities are expected to be more extreme for East Coast tidal pool populations than inOyster Lagoon. In addition, tidal pools apparently do not appear to form an overwinteringground for juvenile sticklebacks (Picard et al. 1990) on the East Coast. This is also true in SaltLagoon, but a large proportion of the young hatched in Oyster Lagoon remain in the lagoonthrough out the winter.One major biotic difference between Oyster Lagoon and the East Coast tidal pools is thepresence of other sticklebacks on the East Coast. Gasterosteus aculeatus share tidal areas andestuaries on the East Coast with four other species of sticklebacks, two of which are congeneric(G. wheatlandi and an unnamed species, the white stickleback). Presumably, the presence ofother, closely related species alters both the ecological and reproductive environments of EastCoast threespine sticklebacks. Competitive interactions for food and nest sites must be sharpenedrelative to the Pacific Coast, and the potential for reproductive errors must be larger. Therefore,it is a little surprising that there appear to be only slight differences in morphology andreproductive behaviour between marine populations on both coasts.88General DiscussionWhite sticklebacks are not present on the Pacific Coast of North America, even thoughvegetation and other environmental factors appear to be suitable for this form of stickleback(Blouw pers. corn.). Extensive searches for white sticklebacks around Oyster Lagoon (pers.observ., Blouw pers. corn.), and elsewhere along the Pacific Coast (Blouw pers. corn.) indicatethat this form is absent from the Pacific Coast. The lack of allozyme differences betweensympatric populations of white and threespine sticklebacks from the Atlantic Coast implies thatwhite sticklebacks only recently diverged from G. aculeatus (Haglund et al. 1990). The absenceof white sticklebacks from the Pacific Coast indicates that this form evolved after the Pacific andAtlantic clades of G. aculeatus were separated. The discovery of a sympatric, recently divergedspecies of Gasterosteus on the East Coast of North America indicates that all marine populationsare not necessarily homogenous. Perhaps reproductively isolated marine sticklebacks also existon this coast. Saimoto (1993) indicates that there is some isolation between lagoon populations,and considerable isolation between Oyster Lagoon sticklebacks and adjacent anadromouspopulations. Again, these observations suggest that marine and anadromous threespinesticklebacks are not necessarily panmictic.In summary, this general life history study establishes some differences and manysimilarities between the Oyster Lagoon population and other freshwater, anadromous and marinepopulations. The need for further studies on Pacific Coast marine populations is clear. Thisdescription of the Oyster Lagoon population is only a beginning. If colonists from the seas areancestral to the bewildering array of divergent freshwater populations found along the glaciatedWest Coast of North America, then an understanding of marine sticklebacks is fundamental tounderstanding how this remarkable radiation came to be.89LITERATURE CITEDvan dem Assem, J. 1967. Territoriality in the three-spined stickleback, Gasterosteus aculeatusL., an experimental study in intra-specific aggression. Behay. Suppl. 16: 1-164.Baggerman, B. 1985. The role of biological rhythms in the photoperiodic regulation of seasonalbreeding in the stickleback, Gasterosteus aculeatus L.. Netherlands J. Zool. 35 (1,2): 14-31.Bakker, Th.C. and P. Sevenster. 1988. Plate polymorphism of Gasterosteus aculeatus Linnaeus(Pisces: Gasterosteidae): comments on terminology. Copeia 1988 (3): 659-663.Banbura, J. 1989. Lateral plate number development in the complete morph of the three-spinedstickleback, Gasterosteus aculeatus L.. Zool. Scripta. 18 (1): 157-159.Baumgartner, J.V. 1992. Spatial variation of morphology in a freshwater population of threespinestickleback, Gasterosteus aucleatus. Can. J. Zool. 70: 1140-1148.Bell, M.A. 1976. Evolution and phenotypic diversity in Gasterosteus aculeatus superspecies onthe Pacific Coast of North America. Syst. Zool. 25: 211-227.. 1977. A late miocene marine threespine stickleback, Gasterosteus aculeatus aculeatusand its zoogeographic and evolutionary significance. Copeia 1977: 277-282.. 1984. Evolutionary phenetics and genetics: the threespine stickleback, Gasterosteusaculeatus, and related species. In: Evolutionary genetics of fishes. Ed. B.J. Turner.Pleneum Publishing Corp. New York, pp. 431-528.. 1993. Paleobiology of the threespine stickleback. In:  The evolutionary biology of thethreespine stickleback. Bell, M.A. and S.A. Foster eds. Oxford University Press, Oxford.Bentzen, P. and J.D. McPhail. 1984. Ecology and evolution of sympatric sticklebacks(Gasterosteus): specialization for alternative trophic niches in the Enos Lake species pair.Can. J. Zool 62 (11): 2280-2286.Black, R. 1971. Hatching success in the three-spined stickleback (Gasterosteus aculeatus) inrelation to changes in behaviour during the parental phase. Anim. Behay. 19: 532-541.Blouw, D.M. and D.W. Hagen. 1990. Breeding ecology and evidence of reproducitve isolationof a widespread stickleback fish (Gasterosteidae) in Nova Scotia, Canada. Bio. J. Lin.Soc. 39: 195-217.Borg, B. 1985. Field studies on three-spined sticklebacks in the Baltic. Behaviour 93: 153-157.90Coad, B.W. and G. Power. 1973. Observations on the ecology and phenotypic variation of thethreespine stickleback Gasterosteus aculeatus L. 1758 and the blackspotted sticklebackG. wheatlandi Putnam. 1867. (Osteichthytes: Gasterosteidae) in Amory Cove, Quebec.Can. Field Nat. 87: 113-122.Conover, W.J. and R.L.Iman. 1981. Rank transformation as a bridge between parametric andnonparametric statistics. The American Statistician 35 (3): 124-133.Cowen, R.K., L.A. Chiarella, C.J. Gomez and M.A. Bell. 1991. Offshore distribution, size, ageand lateral plate variation of late larval/early juvenile sticklebacks (Gasterosteus) off theAtlantic coast of New Jersey and New York. Can. J. Fsih. Aquat. Sci. 48: 1679-1684.Craig, D. and G.J. FitzGerald. 1982. Reproductive tactics of four sympatric sticklebacks(Gasterosteidae). Env. Biol. Fish. 7 (4): 369-375.Crivelli, A.J. and R.H. Britton. 1987. Life history adaptations of Gasterosteus aculeatus in aMediterranean wetland. Envir. Biol. Fish. 18: 109-125.FitzGerald, G.J. 1983. The reproductvie ecology and behaviour of three sympatric sticklebacks(Gasterosteidae) in a saltmarsh. Biol Behay. 8: 67-69.and F.G. Whoriskey. 1984. The effects of interspecific interactions upon male reproductivesuccess in two sympatric sticklebacks, Gastetrosteus aculeatus and G. wheatlandi.Behaviour 93: 112-126., H. Guderley and P. Picard. 1989. Hidden reproductive costs in the three-spinedstickleback (Gasterosteus aculeatus). Exp. Biol. 48 (5): 295-300.and J.-D. Dutil. 1981. Evidence for differential predation on an estuarine sticklebackcommunity. Can. J. Zool. 59: 2394-2395.Foster, S.A. 1988. Diversionary displays of paternal stickleback: defenses against cannibalisticgroups. Behay. Ecol. Sociobiol. 22: 335-341.^ , S.A. 1990. Courting disaster in cannibal territory. Natural History 11: 53-61.^ . in press. Inference of evolutionary pattern: diversionary displays of threespine stickleback.Francis, R.C., J.V. Baumgartner, A.C. Havens and M.A. Bell. 1986. Historical and ecologicalsources of variation among lake populations of threespine stickelbacks, Gasterosteusaculeatus, near Cook Island, Alaska. Can. J. Zool. 64: 2257-2265.Giles, N. 1987. Population biology of three-spined stickelbacks, Gasterosteus aculeatus, inScottland. J. Zool. Lond. 212: 225-265.91. and F.A. Huntingford. 1984. variability in breeding biology of three-spinedsticklebacks (Gasterosteus aculeatus): problems with measuring populationdifferences in aggression. Behay. 93: 57-68.Goldschmidt, T. and T.C.M. Bakker. 1990. Determinants of reproductvie success of malesticklebacks in the field and in the laboratory. Netherlands J. Zool. 40 (4): 664-687.Gross, H.P. 1977. Adaptive trends of environmentally sensitive traits in the three-spinedstickleback, Gasterosteus aculeatus L.. Z. zool. Syst. Evolut.-forsh. 15: 252-278.and J.M. Anderson. 1984. Geographic variation in the gilirakers and diet of Europeanthreespine sticklebacks, Gasterosteus aculeatus. Copeia 1984 (1): 87-97.Gross, M.R., R.M. Coleman and R.M.McDonald. 1988. Aquatic productivity and the evolutionof diadromous fish migration. Science 239: 1291-1293.Hagen, D.W. 1967. Isolating mechanisms in threespine sticklebacks (Gasterosteus). J. Fish. Res.Bd. Can. 24 (8): 1638-1692.and G.E.E. Moodie. 1982. Polymorphism for plate morphs in Gasterosteus aculeatus ofthe East Coast of Canada and an hypothesis for their global distribution. Can. J. Zool. 60:1032-1042.and G.E.E. Moodie. 1982. Polymorphism for plate morphs in Gasterosteus aculeatus onthe Pacific coast of North America. Can J. Zool. 60: 1032-1042.and L.G. Gilbertson. 1972. Geographic variation and environmental selection inGasterosteus aculeatus L. in the Pacific Northwest, America. Evol. 26: 32-51.and J.D. McPhail. 1970. The species problem within Gasterosteus aculeatus on the PacificCoast of North America. J. Fish. Res. Bd. Can. 27: 147-155.Haglund, T.R., D.G. Buth and D.M. Blouw. 1990. Allozyme variation and the recognition of thewhite stickleback. Biochem. Syst. Ecol. 18: 559-563., D.G. Buth and R. Lawson. 1992. Allozyme variation and phylogenetic relationship ofAsian, North American. and European populations of threespine stickleback, Gasterosteusaculeatus. Copeia 1992 (2): 432-443.and J.D. McPhail. 1975. Mate selection in threespine sticklebacks (Gasterosteus). Can. JZool. 53: 441-450.Heuts, M.J. 1947. Experimental studies on adaptive evolution in Gasterosteus aculeatus L..Evolution 1: 80-102.92Jamieson, I.G., D.M. Blouw, and P.W. Colgan. 1992. Field observations on the reproductivebiology of a newly discovered stickleback (Gasterosteus). Can. J. Zool. 70: 1057-1063.Kedney, G.I., V. Boule, and G.J. FitzGerald. 1987. The reproductive ecology of threespinesticklebacks breeding in fresh and brackish water. American Fisheries Society symposium1: 151-161.Klepaker, T. (in prep). Morphological changes in a marine population of threespined sticklebacks,Gasterosteus aculeatus, recently isolated in fresh water.Kynard, B.E. 1972. Male breeding behavior and lateral plate phenotypes in threesepinesticklebacks (Gasterosteus aculeatus L.). Unpubl. Ph.D. thesis. U. of Washington.. 1978. Breeding behavior of a lacustrine population of threespine sticklebacks(Gasterosteus aucleatus L.) Behaviour 62: 178-207.. 1979. Nest habitat preference of low plte morphs of threespine sticklebacks(Gasterosteus aculeatus L.). Copeia 1979: 525-528.Larson, G.L. 1976. Social behavior and feeding ability of two phenotypes of Gasterosteusaculeatus in relation of their spatial and trophic segregation in a temperate lake.Can. J. Zool. 54 (2): 107-121.Lavin, P.A. and J.D. McPhail. 1985. The evolution of freshwater diversity in threespinestickleback (Gasterosteus aculeatus): site specific differentiation of trophic morphology.Can. J. Zool. 63: 2632-2638.and J.D. McPhail. 1993. Parapatric lake and stream sticklebacks on northern VancouverIsland: disjunct distribution or parallel evolution? Can. J. Zool. 71: 11-17.McLennan, D. 1988. Phylogenetic and experimental studies of the function of sexually dimorphicnuptial colouration in the three-spined stickleback (Gasterosteus aculeatus L.). M.Sc.thesis, University of British Columbia, Vancouver, British Columbia, Canada.McPhail, J.D. 1969. Predation and evolution of a sticklebakc (Gasterosteus). J. Fish. Res. Bd.Can. 26: 3183-3208.. 1984. Ecology and evolution of sympatric sticklebacks .(Gasterosteus): morphology andgenetic evidence for a species pair in Enos Lake, British Columbia. Can. J. Zool. 62 (7):1402-1408.. 1989. Status of the Enos Lake stickleback species pair, Gasterosteus spp.. Canadian FieldNaturalist. 103 (2): 216-219.. 1992. Ecology and evolution of sympatric sticklebacks (Gasterosteus): evidence for aspecies-pair in Paxton Lake, Texada Island, British Columbia. Can J. Zool. 70: 361-369.93. 1993. Speciation and the evolution of reproductive isolation in sticklebacks (Gasterosteus)of south-western British Columbia. In: The evolutionary biology of the threespinestickleback. Bell, M.A. and S.A. Foster eds. Oxford University Press, Oxford.and C.C. Lindsey. 1970. Freshwater fishes of northwestern Canada and Alaska. Bull. Fish.Res. Bd. of Can. 173: 1-381.. and D.E. Hay. 1983. Differences in male courtship in freshwater and marine sticklebacks(Gasterosteus aculeatus). Can. J. Zool. 61: 292-297.and S.D. Peacock. 1983. Some effects of the cestode (Schistocephalus solidus) onreproduction in the threespine stickleback (Gasterosteus aculeatus): evolutionary aspectsof a host-parasite interaction. Can J. Zool. 61: 901-908.Miller, R.R. and C.L. Hubbs. 1969. Systematics of Gasterosteus aculeatus, with particularreference to intergredation and introgression along the Pacific Coast of North America:a commentary on a recent contribution. Copeia 1969 (1): 52-69.Moodie, G.E.E. 1970. Predation as a mechanism in the evolution of an unusual population ofsticklebacks in the Queen Charlotte Islands, Canada (Pisces: Gasterosteidae). Ph.D. thesis,Dept. of Zoology, University of Alberta, Calgary, Alberta.. 1972. Morphology, life history and ecology of an unusual stickleback (Gasterosteusaculeatus) in the Queen Charlotte Islands, Canada. Can. J. Zool. 50: 721-732.and T.E. Reimchen. 1976. Phenetic variation and habitat differences in Gasterosteuspopulations of the Queen Charlotte Islands. Systemat. Zool. 25: 49-61.Mori, S. 1984. Reproductive behaviour of the landlocked three-spined stickleback,Gasterosteus aculeatus microcephalus, in Japan: I. The year-long prolongation of thebreeding period in waterbodies with springs. Behaviour 93: 21-35.. 1990. Two morphological types in the reproductive stock of three-spined stickleback,Gasterosteus aculeatus, in Lake Harutori, Hokkaido Island. Env. Biol. Fish. 27: 21-31.van Mullen, P.J. 1967. On synchronization in the reproduction of the stickleback (Gasterosteusaculeatus L. forma leiura cuv.). Archives Neerlandaises de Zoologie 17:258-274.Orti, G., M.A. Bell, T.E. Reimchen and A. Meyer (in press). Gobal survey of mitochondrialDNA sequences in the threespine stickleback: evidence for recent colonizations.Evolution. In press.Paepke, H.-J. 1983. Die Gattung Gasterosteus Linne 1758. In: Die Stichlinge (Gasterosteidae).A. Ziemsen Verlag. Wittenburg Lutherstadt. pp. 67-144.94Picard, P. jr., J.J. Dodson and G.J. FitsGerald. 1990. Habitat segregation among the age groupsof Gasterosteus aculeatus (Pisces: Gasterosteidae) in the middle St. Lawrence estuary,Canada. Can. J. Zool. 68: 1202-1208.Quinn, T.P. and J.T. Light. 1989. Occurence of threespine sticklebacks (Gasterosteus aculeatus)in the open North Pacific Ocean: Migration or drift? Can. J. Zool. 71 (11): 2850-2852.Raven, P. 1986. The size of minnow prey in the diet of young kingfishers Alcedo atthis. Birdstudy 33: 6-11.Reiffers, C. 1984. Reproduction et comportement des femelles de trois especes d'epinoche(Gasterosteidae) sympatrique en milieu natruel. M.Sc. thesis, Laval University, Quebec,Canada.. 1992. Injuries on sticklebacks by a toothed predator (Oncorhynchus) and implication forthe evolution of lateral plates. Evolution 46: 1224-1230.Reimchen, T.E. and S. Douglas. 1984. Seasonal and diurnal abundance of aquatic birds on theDrizzle Lake reserve, Queen Charlotte Islands, British Cloumbia. Canadian FieldNaturalist 98 (1): 22-28., E.M. Stinson and J.S. Nelson. 1985. Multivariate differentiation of parapatric andallopatric populations of threespine stickleback in the Sangan watershed, QueenCharlotte Islands. Can. J. Zool. 63: 2944-2951.Ridgeway, M.S. and J.D. McPhail. 1984. Ecology and evolution of sumpatric sticklebacks(Gasterosteus): mate choice and reproductive isolation in the Enos Lake species pair.Can. J. Zool. 62: 1813-1818.Rowland, W.J. 1983. Interspecific aggression and dominance in Gasterosteus. Env. Biol. Fish.8 (3/4): 269-277.. 1989a. Mate choice and the supernormality effect in female sticklebacks (Gasterosteusaculeatus). Behay. Ecol. Sociobiol. 24: 433-438.. 1989b. The effects of body size, aggression and nuptial coloration on competition forterritories in male threespine sticklebacks Gasterosteus aculeatus. Anim. Behay. 37: 282-289.. 1989c. The ethological basis of mate choice in male threespine sticklebacks, Gasterosteusaculeatus. Anim. B eh. 38: 112-120.Russell, W.M.S. and C. Russell. 1984. Sticklebacks and ethology. Behaviour 93: 184-193.95Ryan, P.M. 1984. Age, growth, and food of threespine sticklebacks (Gasterosteus aculeatus) inthree lakes of central Newfoundland, Canada. Can. Tech. Rep. Fish. Aquat. Sci. 1325:iv+16 p.Saimoto, R.S. 1993. Reproductive and natal homing of marine threespine sticklebacks(Gasterosteus aculeatus). M.Sc. thesis, U. of B.C., Vancouver, B.C.Sargent, R.C. 1982. Territory quality, male quality, courtship intrusions and female nest-choicein the threespine stickleback, Gasterosteus aculeatus. Anim. Behay. 30: 364-374.. 1985. Territoriality and reproductive tradeoffs in the threespine stickleback, Gasterosteusaculeatus. Behaviour 93: 217-226.and J.B. Gebbler. 1980. Effects of nest site concealment in hatching success, reproductivesuccess and parental behavior of the threespine stickleback, Gasterosteus aculeatus.Behay. Ecol. Sociobiol. 7: 131-142.Schluter, D. and J.D. McPhail. 1992. Ecological character displacement and speciation insticklebacks. Am. Nat. 140 (1): 85-108.Scott, W.B. and E.J. Crossman. 1973. Freshwater fishes of Canada. Bulletin 184. Fish. Res. Bd.Can.: 665-668.and M.G. Scott. 1988. Atlantic fishes of Canada. Can. Bull. Fish. Aquat. Sci. No. 219.Snyder, R.J. and H. Dingle. 1989. Adaptive, genetically based differences in life history betweenestuary and freshwater threespine sticklebacks (Gasterosteus aculeatus L.). Can. J. Zool.67: 2448-2454.Taylor, E.B. and J.D. McPhail. 1986. Prolonged and burst swimming in anadromous andfreshwater threespine stickleback, Gasterosteus aculeatus. Can. J. Zool. 64 (2): 416-420.Tschanz, B. and M. Scharf. 1971. Nestortwahl and Orienteirung zum Nestort beim DreistchaligenStichling. Revue Suisee de Zoologie 78: 715-721.Ward, G. and G.J. FitzGerald. 1987. Male aggression and female choice in the threespinedstickleback, Gasterosteus aculeatus L.. J. Fish Biol. 30: 679-690.Whoriskey, F.G. 1984. Le role des facteurs choisis, biotiques et abiotiques dans lastructuration d'une communaute d'epinoche (Pisces: Gasterosteidae). These de doctorat,l'Universite Laval, Quebec.. 1991. Stickleback distraction displays: sexual or foraging deception against eggcannibalism? Anim. Behay. 41: 989-995.96and G.J. FitzGerald. 1985a. Nest sites of the threespine stickleback: Can site charactersalone protect the nest against egg predators and are nest sites a limiting resource? Can.J. Zool. 63: 1991-1994.and G.J. FitzGerald. 1985b. Sex, cannibalism and sticklebacks. Behay. Ecol. Sociobiol. 18:15-18.and G.J. FitzGerald. 1987. Intraspecific competition in sticklebacks (Gasterosteidae:Pisces): does mother nature concur? J. Anim. Ecol. 56: 939-947., G.J. FitzGerald and G.J. Reebs. 1986. The breeding season population structure of threesympatric territorial sticklebacks (Pisces: Gasterosteidae). J. Fsih. Biol. 29: 635-648.William, D.D. and J.C. Delbeek. 1989. Biology of the threespine stickleback, Gasterosteusaculeatus, and the blackspotted stickleback, G. wheatlandi, during their marinepelagic phase in the Bay of Fundy, Canada. Env. Biol. Fish. 24 (1): 33-41.Withler, R.E. 1980. Genetic relationship among threespine sticklebacks, Gasterosteus aculeatus.M.Sc. thesis. University of British Columbia, Vancouver, British Columbia.Wootton, R.J. 1976. The biology of the sticklebacks. Academic Press. San Francisco. 387 pp.Worgan, J.P and G.J FitzGerald. 1981. Diel activity and diet of three sympatric sticklebacks intidal salt marsh pools. Can. J. Zool. 59: 2375-2379.Zar, J.H. 1984. Biostatistical analysis. 2nd ed. Prentice Hall Inc. Toronto. 718 pp.Ziuganov, V.V. 1983. Genetics of osteal plate polymorphism and micorevolution of threespinestickleback (Gasterosteus aculeatus L.). Theor. Appl. Genet. 65: 239-246., G.J Golovatjuk, K.A. Savvaitova and V.F. Bugaev. 1987. Genetically isolated sympatricforms of threespine stickleback, Gasterosteus aculeatus, in Lake Azabachije (Kamchatka-peninsula, USSR). Env. Biol. Fish. 18 (4): 241-247.97


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items