UBC Faculty Research and Publications

Functional basis of ecological divergence in sympatric stickleback McGee, Matthew D; Schluter, Dolph; Wainwright, Peter C 2013

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

Item Metadata


52383-12862_2013_Article_2508.pdf [ 353.92kB ]
JSON: 52383-1.0223354.json
JSON-LD: 52383-1.0223354-ld.json
RDF/XML (Pretty): 52383-1.0223354-rdf.xml
RDF/JSON: 52383-1.0223354-rdf.json
Turtle: 52383-1.0223354-turtle.txt
N-Triples: 52383-1.0223354-rdf-ntriples.txt
Original Record: 52383-1.0223354-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessFunctional basis of ecological divergence insympatric sticklebackMatthew D McGee1*, Dolph Schluter2 and Peter C Wainwright1AbstractBackground: The evolution of ecological divergence in closely related species is a key component of adaptiveradiation. However, in most examples of adaptive radiation the mechanistic basis of ecological divergence remainsunclear. A classic example is seen in the young benthic and limnetic stickleback species pairs of British Columbia. Ineach pair the benthic species feeds on littoral macroinvertebrates whereas the limnetic feeds on pelagiczooplankton. Previous studies indicate that in both short-term feeding trials and long-term enclosure studies,benthics and limnetics exhibit enhanced performance on their own resource but fare more poorly on the otherspecies’ resource. We examined the functional basis of ecological divergence in the stickleback species pair fromPaxton Lake, BC, using biomechanical models of fish feeding applied to morphological traits. We examined theconsequences of morphological differences using high speed video of feeding fish.Results: Benthic stickleback possess morphological traits that predict high suction generation capacity, includinggreatly hypertrophied epaxial musculature. In contrast, limnetic stickleback possess traits thought to enhancecapture of evasive planktonic prey, including greater jaw protrusion than benthics and greater displacementadvantage in both the lower jaw-opening lever system and the opercular four-bar linkage. Kinematic data supportthe expectations from the morphological analysis that limnetic stickleback exhibit faster strikes and greater jawprotrusion than benthic fish, whereas benthics exert greater suction force on attached prey.Conclusions: We reveal a previously unknown suite of complex morphological traits that affect rapid ecologicaldivergence in sympatric stickleback. These results indicate that postglacial divergence in stickleback involves manyfunctional systems and shows the value of investigating the functional consequences of phenotypic divergence inadaptive radiation.Keywords: Gasterosteus aculeatus, Functional morphology, Suction feeding, Postglacial fishes, Ecological speciationBackgroundImproving our understanding of the process of adaptiveradiation requires a more complete understanding of theorigin and maintenance of ecological divergence betweenclosely related species [1-4]. The four key properties ofadaptive radiation are common ancestry, rapid speciation,phenotype-environment correlations, and trait utility.Shared ancestry is the most commonly tested criterion,typically using a phylogeny with sampling both within theradiation and in its close relatives [5]. Testing for elevatedrates of speciation requires temporal information, typicallyage estimates for newly invaded regions and estimatesof divergence times in the phylogeny [6,7]. It is alsoimportant to establish the existence of a correlationbetween the phenotypic traits of species within the ra-diation and the environments they are found in [8-10].However, phenotype-environment correlations in theabsence of performance data do not necessarily indicatethat trait differences play an important ecological role, asmeasured trait differences may result from correlationswith other traits under selection or as a consequence ofdevelopmental constraints unrelated to ecology [11-13].Here we address the final and crucial criterion for adap-tive radiation, trait utility, “evidence that traits are usefulwhere they are employed” [4]. Trait utility provides thecritical link between phenotype and performance and isrequired to strengthen inferences of the role of naturalselection in producing the radiation. One way to assess* Correspondence: mcgee.matthew@gmail.com1Department of Evolution and Ecology, University of California Davis, 1Shields Avenue, Davis, CA 95616, USAFull list of author information is available at the end of the article© 2013 McGee et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited. The Creative Commons Public Domain Dedicationwaiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwisestated.McGee et al. BMC Evolutionary Biology 2013, 13:277http://www.biomedcentral.com/1471-2148/13/277trait utility is to carry out manipulative experiments ontraits to linking a feature of the phenotype directly to arelevant performance character [14]. Another way, whichwe adopt here, involves the use of functional models offeeding performance developed in other phylogeneticallyand morphologically similar species, which allow the cal-culation of performance from phenotypic data [13,15].The stickleback species pairs offer an excellent systemto test the importance of trait utility in adaptive radiation.In a series of British Columbia lakes created within the last10,000 years by retreating glaciers, threespine stickleback,Gasterosteus aculeatus, have repeatedly diverged into aplanktivorous (hereafter “limnetic”) species and a benthic-feeding (hereafter “benthic”) species [16-19]. In mostspecies-pair lakes, benthic and limnetic stickleback aremore closely related to each other than they are to eco-logically similar forms in nearby lakes [20,21]. Limneticsfeed mostly on evasive pelagic calanoid copepods withlong strain-sensitive antennae capable of detectingincoming predator attacks, giving the copepod time toescape [22,23]. Benthics feed mostly on non-evasiveburied and attached littoral macro-invertebrates thatmust be detected, then forcibly extracted from theirhiding places. In short-term feeding trials, individuals ofeach species experienced higher prey capture successfeeding on their preferred prey than when feeding onthe other species’ prey [24,25], suggesting that thetrophic apparatus plays a direct role in dietary diver-gence. In enclosure experiments in native lakes, individ-uals of each species raised in the appropriate habitatgrew faster than when raised in the other species’ habi-tat [26], and limnetic-benthic hybrids exhibit signs oflower fitness than either parental form in nature [27].Structure and performance associated with the prey cap-ture mechanism may help clarify the functional basis ofecological divergence in the species pairs (Figure 1). Likemany teleosts, stickleback are suction-feeding predatorsthat capture prey by expanding the buccal cavity to drawprey items into the mouth. Limnetic stickleback feed onevasive strain-sensitive copepods, so we might reasonablyexpect limnetics to possess morphology associated withrapid prey capture kinematics [23]. Suction is required todislodge the buried and attached invertebrates that makeup the bulk of benthic diets, suggesting that benthics maypossess functional systems adapted to exert higher forceon attached prey items [28].The biomechanics of suction feeding can be quantifiedusing a series of functional models that treat craniofacialbones and muscles as sets of complex levers and linkages.The suction index model predicts the relative morphologypotential to produce suction pressure in fish species thatuse cranial rotation to expand the buccal cavity [15,29].The opercular four-bar linkage predicts the magnitude ofrotation in the articular, the output link, for a set amountof rotation by the interopercule, the input link [30,31].The jaw lever system predicts the amount of rotation inthe fish’s jaw for a given amount of input rotation in thearticular [32]. Jaw protrusion refers to the anterior excur-sion of the ascending process of the premaxilla duringmouth opening [33]. In other fishes, these models have ac-curately predicted patterns of prey use as well as prey cap-ture kinematics in vivo. [14,15,30,32].In this study, we evaluate trait utility by using morpho-logical data and functional models of fish feeding to predictkinematic patterns, then test these predictions by analyzinghigh-speed films of feeding behavior in limnetic and benthicfish to generate both kinematic and simulated performancedata. We then discuss how component trait divergence inthe four functional systems affects ecological divergencein the species pair. Our approach deepens our under-standing of the mechanisms of adaptive divergence.ResultsWe uncovered substantial functional and kinematicdifferences between the two stickleback species. Pax-ton Lake benthic and limnetic stickleback differ in allfour of the functional systems examined in this study:suction index, transmission coefficient of the opercularfour-bar, lower jaw opening displacement advantage,and jaw protrusion (Figure 2, Table 1). Jaw protrusionis higher in limnetics, as are opercular four-bar trans-mission coefficent and lower jaw opening displacementadvantage (Table 1). Suction index is higher in benthics(0.017 vs 0.010 in 50 mm fish, p < 0.001). These differ-ences imply that benthics have the capacity to generatehigher suction pressure than limnetics, whereas limneticswill have faster jaw movements and greater jaw protrusionduring the strike.Five of the 11 morphological variables were signifi-cantly different between species: epaxial height, epaxialwidth, output link, input link, and opening jaw inleverFigure 1 Limnetic/benthic stickleback diets and associatedfunctional predictions. Photographs are stills from high-speed feedingkinematics.McGee et al. BMC Evolutionary Biology 2013, 13:277 Page 2 of 10http://www.biomedcentral.com/1471-2148/13/277Figure 2 Morphological components of four functional systems associated with prey capture in percomorph fishes. Landmarks:(1) anteriormost extent of premaxilla; (2) anteriormost extent of dentary; (3) point of articulation between the supracleithrum and post-temporal;(4) dorsalmost extent of epaxial, measured dorsal to landmark 3; (5) point of articulation between supracleithrum and post-temporal on oppositeside of fish, measured in the frontal plane; (6) posteriormost extent of buccal cavity, measured between landmarks 1 and 3; (7) anteriodorsalextent of maxilla; (8) quadrate-articular jaw joint; (9) insertion of the interopercular-articular ligament; (10) opercular joint; (11) posterioventralextent of interopercule. Bone names: pmx = premaxilla, max =maxilla, art = articular, quad = quadrate, pop = preopercule, iop = interopercule,sop = subopercule, op = opercule, pt = post-temporal, scl = supracleithrum, cl = cleithrum, nc = neurocranium.Table 1 Functional feeding systems and component traits of limnetic and benthic sticklebackFunctional system Landmarks (Figure 2) W (rank-sum test) Mean ± SE, Limnetic1 Mean ± SE, Benthic1Suction Index 1,2,3,4,5,6 523*** 0.010 ± 0.001 0.017 ± 0.001Disp. adv., jaw opening 2,8,9 13*** 5.96 ± 0.01 4.82 ± 0.01Opercular four-bar KT 8,9,10,11 98*** 5.75 ± 0.13 5.15 ± 0.13Jaw protrusion 1,7 99*** 1.76 ± 0.07 mm 1.48 ± 0.07 mmMorphological component traits Landmarks (Figure 2) W (rank-sum test) Mean ± SE, Limnetic1 Mean ± SE, Benthic1Suction Index:Gape 1,2 343 5.15 ± 0.13 mm 5.33 ± 0.13 mmBuccal length 1,6 204 15.05 ± 0.21 mm 14.66 ± 0.21 mmNeurocranium outlever 1,3 292 16.55 ± 0.20 mm 16.47 ± 0.20 mmEpaxial height 3,4 493*** 1.95 ± 0.08 mm 2.35 ± 0.08 mmEpaxial width 3, 5 572*** 5.46 ± 0.09 mm 6.38 ± 0.09 mmDisp. adv., jaw opening:Jaw opening outlever 2,8 248 5.60 ± 0.11 mm 5.54 ± 0.11 mmJaw opening inlever2 8,9 568*** 0.94 ± 0.02 mm 1.15 ± 0.02 mmOpercular four-bar:Coupler link 9,11 262 6.83 ± 0.10 mm 6.86 ± 0.10 mmFixed link 8,10 231 9.56 ± 0.12 mm 9.44 ± 0.12 mmInput link 10,11 512 *** 5.27 ± 0.08 mm 5.71 ± 0.08 mmOutput link2 8,9 568 *** 0.94 ± 0.02 mm 1.15 ± 0.02 mm1Trait values calculated using linear regression of SL (set to 50 mm) and ecomorph.2Jaw opening inlever = output link of the opercular four-bar.***p < 0.001.McGee et al. BMC Evolutionary Biology 2013, 13:277 Page 3 of 10http://www.biomedcentral.com/1471-2148/13/277(Table 1). The greater epaxial width and epaxial heightof benthics and the smaller opening jaw inlever of lim-netics are consistent with observed functional diver-gence, because epaxial cross-sectional area increasesSuction Index and a smaller opening jaw inlever in-creases displacement advantage of jaw opening. Also,the input and output link of the opercular four-bar dif-fer between limnetics and benthics (Table 1), in a waythat improves force transmission in benthics and velocitytransmission in limnetics. Slopes of the relationshipsbetween the opercular four-bar fixed link (p < 0.05)and opercular four-bar coupler link (p < 0.05) differedsignificantly.As expected from the morphological measurements, ourlinear mixed model analysis of kinematic data revealed thatlimnetics exhibit greater jaw protrusion than benthics, andthey have shorter times to peak gape, peak lower jaw rota-tion, and prey capture (Table 2). SL had a significant effecton some of the kinematic variables, including maximumgape, time to peak gape, time to peak cranial rotation, andtime to prey capture. By including it as a covariate in ourmodel, the kinematic differences recorded are correctedfor size effects. Our mixed-model analysis using SuctionInduced Force Field (SIFF) data also indicated that benthicswould exert higher maximum force on a simulatedattached prey than limnetics (Table 2).DiscussionOur results reveal a previously unknown suite of com-plex morphological traits involved in rapid sympatricecological divergence in a species pair of postglacial fish.Kinematic predictions derived from functional analysesof these morphological traits match observations of highspeed prey capture attempts in the plankton-feeding lim-netic and the littoral macroinvertebrate-feeding benthic.These results show the value of investigating trait utilityfor understanding the performance consequences ofphenotypic divergence in adaptive radiation.Predicted functional differencesBased on our morphological analysis, we predicted largedifferences between the ecologically differentiated formsin their functional performance when feeding. Few ofthese differences had been anticipated in previous workon the ecology and morphology of this system. Thesedifferences likely contribute to divergent feeding successand growth rate in transplant experiments in the nativelakes [26,27,34], and show the value of a functional ana-lysis of morphological differences between species.Benthics have the potential to generate greater suctionpressure and therefore generate greater suction flowspeed. Higher suction index values lead to increasedsuction flow speeds and have been shown to improveperformance in computational models of suction feedingon buried and attached prey items [28]. Similar highersuction index values are also observed in benthic stickle-back populations in other, independently derived speciespairs [35]. The increased suction index values of Paxtonbenthics are driven mainly by two epaxial traits thatdiffer between limnetics and benthics (Table 1). Thesehypertrophied epaxial muscles give benthics their distinct-ive “humped” phenotype [16] and contribute to increasedbody depth. We suggest that the body depth variationTable 2 Kinematic divergence in a stickleback species pairTrait pMCMC (SL) pMCMC (ecomorph) Limnetic value† Benthic value†Excursions:Gape 0.0004*** 0.53 2.91 mm 2.81 mmJaw protrusion 0.87 0.008** 1.35 mm 1.01 mmCranial rotation 0.12 0.21 8.72 deg 7.12 degLower jaw rotation 0.18 0.22 25.06 deg 23.31 degStrike distance 0.11 0.88 2.87 mm 2.92 mmTimings:Gape 0.0034** 0.01* 4.6 ms 8 msJaw protrusion 0.10 0.13 7.8 ms 11.9 msCranial rotation 0.06 0.31 7.3 ms 9.6 msLower jaw rotation 0.04* 0.009** 5.3 ms 9.8 msPrey capture 0.04* 0.049* 6.3 ms 9.9 msForces:Attached prey 0.63 0.045* 2.3 × 10-4 N 3.1 × 10-4 N†The ecomorph values were calculated using the fixed effect of SL (set for a 40 mm fish) and the fixed effect of ecomorph from a mixed-effect model for eachkinematic trait.*p < 0.05, **p < 0.01, ***p < 0.001.McGee et al. BMC Evolutionary Biology 2013, 13:277 Page 4 of 10http://www.biomedcentral.com/1471-2148/13/277commonly observed between lake-stream sticklebackand in recently deglaciated areas is likely connected tovariation in the size of the epaxial muscles [36-38].Oral jaw traits also show a strong pattern of divergencebetween benthics and limnetics. When the neurocraniumis elevated in preserved fish, limnetics exhibit more jawprotrusion than benthics. Zooplanktivorous teleosts oftenpossess high jaw protrusion, which is thought to aid in thecapture of strain-sensitive planktonic prey, particularlycalanoid copepods [22,23]. The increased morphologicaljaw protrusion of limnetic fish sets up clear kinematicpredictions: limnetics should be able to project theiroral jaws farther than benthics during the strike. Theopening jaw lever system indicates that limnetics possessmore displacement advantage when opening the lowerjaw. Assuming equal input velocity, output velocity will beproportional to displacement advantage, implying thatlimnetics should rotate the lower jaw and open the mouthmore rapidly than benthics.Divergence in the transmission coefficient of the opercu-lar four-bar mirrors divergence in the opening jaw leversystem, with limnetics exhibiting a higher transmission co-efficient than benthics. This similarity between the oper-cular four-bar and opening jaw lever likely occurs becauseboth systems share a component trait, the output link/opening jaw inlever (Figure 2, Table 1). In limnetics, an in-crease in this component trait increases velocity transmis-sion of the opercular four-bar while simultaneouslyincreasing displacement advantage of jaw opening. Diver-gence in the opercular four-bar transmission coefficient isalso driven by an increase in the input link due to dorso-ventral expansion of the opercular series in benthicstickleback. Recent work on stickleback opercle shape sug-gests that dorsoventral variation in the Paxton species pairand across populations is connected to a developmentalmodule that is likely under selection [39,40]. It is likelythat recent stickleback opercle shape evolution is a conse-quence of selection on the opercular four-bar transmissioncoefficient.In other teleosts, the opercular four-bar has been lesspredictive of kinematics than the anterior jaw linkage[30,32], though it clearly is involved in jaw depressionsince fish with a severed opercular four-bar linkage exhibitdisrupted feeding kinematics [41]. Kinematic implicationsof the differences in four-bar mechanics suggest a similarpattern as the opening jaw lever. The higher transmissioncoefficient of limnetics predicts that more output rotationis produced for a given input rotation, which should allowlimnetic stickleback to open their jaws more rapidly thanbenthics during a strike.Complex functional systems, including often diverge intheir component traits while converging in their func-tional outputs, a phenomenon called many-to-one map-ping [42]. For example, benthic stickleback from Alaskaand British Columbia have independently evolved an in-creased suction index by modifying different componentsof the system in each population, resulting in a nearlythreefold increase in morphological diversity relative totheir anadromous common ancestor [35]. We suggest thatfuture studies of morphological evolution in postglacialfishes are likely to reveal functional solutions similar tothose seen in Paxton Lake, even if the individual traitscomprising these solutions vary.KinematicsKinematic data support many of the predictions de-rived from functional morphology, implying a strongrelationship between form and function in this youngradiation.Limnetics have higher maximum jaw protrusion, shortertime to peak gape, shorter time to peak lower jaw opening,and shorter time to prey capture than benthics. All ofthese traits are expected to improve performance onstrain-sensitive prey like calanoid copepods and othercrustacean zooplankton, according to simulation studiesand live trials with suction-feeding fish species [23,28].In limnetics, higher speeds of jaw opening and rapidprojection of the flow field towards the prey via jaw pro-trusion both minimize the window of time in whichattacked copepods can sense the incoming flow fieldwhile simultaneously exposing the prey to a more rapidincrease in suction flow speed.Phenotypic plasticity is thought to play a major role inevolution, and adaptive plasticity has been documentedin stickleback and other postglacial fishes [43,44]. Ourmorphological dataset used wild-caught benthic andlimnetic fish, while our kinematic dataset used F1 ben-thics and limnetics raised in outdoor experimentalponds designed to mimic the natural habitat of PaxtonLake. Phenotypic variation in our morphological datasetis affected by both genetics and environment, whereasfish from the kinematic dataset would have been less in-fluenced by environment. However, despite a potentialreduction in environmental influences on phenotype inour kinematic dataset, we see clear differences betweenbenthics and limnetics.Our simulations suggest that benthics exert higherforces on attached prey items than limnetics do. Manycommon benthic prey items, such as chironomid larvae,burrow in the substrate or within aquatic plants andmust be forcibly extracted once located [45]. Other ben-thic prey items, like amphipods, can grip or cling to ob-jects in the littoral zone, requiring the predator to dislodgethem [46]. Enhanced force generation via increased suc-tion pressure is thus likely to increase the ability of benthicstickleback to capture littoral macroinvertebrates byincreasing the proportion of successful strikes producedby the fish.McGee et al. BMC Evolutionary Biology 2013, 13:277 Page 5 of 10http://www.biomedcentral.com/1471-2148/13/277Understanding the functional consequences of pheno-typic divergence is centrally important to studies of adap-tation [47]. A careful examination of trait utility can helpto separate functionally relevant traits from less relevantshape differences. For example, the distinctive “hump-backed” phenotype of benthic stickleback represents oneof the largest shape differences between species [16,17],but the functional consequences have not previously beenunderstood. This phenotype is caused by hypertrophiedepaxial muscle posterior to the neurocranium. These en-larged muscles increase the physiological cross-sectionalarea and therefore force generation of the muscles elevat-ing the neurocranium during a prey capture attempt[15,29]. All else being equal, more forceful epaxial inputwill result in stronger suction pressure. Enhanced suctionpressure is strongly connected with increased performanceon attached prey, suggesting that these enlarged musclesmay help benthics forage on littoral macro-invertebrates[28]. These results mesh well with previous morphometricwork indicating that fish from lakes with a high littoral areaand therefore more benthic prey tend to have shapes con-sistent with hypertrophied epaxial musculature [48-55].Interestingly, a similar pattern exists in Darwin’s finches,with muscle traits strongly contributing to divergent biteforces between closely-related species [56]. In both finchesand stickleback, variation associated with the cross-sectional area of cranial muscles plays a pivotal role, sug-gesting that variation in the sizes and shapes of musclescan be as important as changes in the structure of hardbony elements [57-60].Morphological and kinematic gape data indicate that,contrary to previous studies, size of the open mouthdiffers little between benthics and limnetics after bodysize correction, particularly when compared to changesobserved in the epaxial muscles. Benthic and limneticstickleback were previously thought to differ in mouthsize, with benthics possessing a larger closed-mouthgape width [17]. Paxton benthics and limnetics do differin the width of the closed mouth, but teleost mouths arehighly kinetic and change shape over the course of aprey capture attempt [29,61]. Our results indicating alack of divergence in mouth size make sense in the contextof benthic suction feeding, as a larger mouth wouldincrease the area of the fish’s buccal cavity, reducingthe suction pressure it could exert on attached prey.Studies of trophic morphology in postglacial radiationshave mostly focused on the gill rakers, which are thoughtto enable zooplanktivorous limnetic ecomorphs to retainsmall prey items obtained through suction feeding [62,63].Our kinematic and morphological results are consistentwith divergence in gill raker morphology, with limneticfish using rapid strikes to ingest small zooplankton, thenusing the rakers to prevent escape from the fish’s buccalcavity. The function of large raker spacing in benthicstickleback has yet to be established, and could be associ-ated with the need to sort food items from benthic debrisafter a strike, or it could simply be a function of the largerprey sizes consumed by benthics [16,64]. Though gillrakers are certainly a functionally important trait [65,66],experimental studies of gill raker function (eg. surgical re-moval of the rakers) have focused on specialized phyto-planktivorous oreochromine cichlids, rather than specieswith zooplanktivorous diets similar to postglacial limneticfishes [67,68].Postglacial radiations also differ in many ecologicallyimportant traits aside from trophic morphology [69,70].Limnetic stickleback are often more exposed to preda-tion than benthics, favoring divergence in cryptic color-ation and defensive weaponry [71-73]. Differences instructural complexity of the habitat can lead to diver-gence in maneuverability, sustained swimming, andspatial processing [74-78]. Traits related to searching forprey can also differ, including vision and neuromast pat-terning [79,80]. The large number of potential pheno-typic differences emerging between young sticklebackspecies pairs suggests that further study of integration[81] in the genetic and phenotypic architecture of post-glacial radiation is likely to prove fruitful.ConclusionBenthic and limnetic stickleback differ in many morpho-logical traits affecting suction feeding, and this functionalvariation is associated with divergent performance on at-tached and evasive prey. Ecological divergence in sympatricstickleback involves the evolution of functional divergencevia multiple phenotypic traits, and suggests that examiningtrait utility can provide a fundamental contribution to stud-ies of adaptive radiation.MethodsCollections and photographyWe used previously-collected samples from Paxton Lakebenthic and limnetic species (Paxton Lake, Texada Island,British Columbia, [17]). A total of 48 fish (benthic n = 23,limnetic n = 25) were used. Fish had been previously fixedin formalin and stored in ethanol; we cleared them in atrypsin solution and stained bones with alizarin red, thenplaced the specimens into glycerin for measurement[82]. Clearing with trypsin restores a more natural rangeof motion to the muscles and ligaments than is presentin formalin-preserved fish, allowing us to manipulatethe head and jaws more effectively.Photographs of the fish were taken using a Sony DSC-717 5MP camera attached to a dissecting microscope witha Scopetronix microscope adapter. Three photographswere taken of each fish: one of the head in dorsal view,one lateral head shot with the fish’s jaws closed, and oneMcGee et al. BMC Evolutionary Biology 2013, 13:277 Page 6 of 10http://www.biomedcentral.com/1471-2148/13/277lateral head shot with the jaws fully protruded and headelevated. Each fish’s mouth was opened using a combin-ation of forceps squeezing the fish’s epaxial and hypaxialmuscles, which are involved in opening the mouth duringa feeding event, and using a small metal rod inserted intothe buccal cavity to press dorsally against the ventralsurface of the neurocranium, which rotates upward toopen the jaws in life. Applying force to the neurocra-nium rather than to the jaws directly reduces the abilityof the investigator to open the jaws farther than theywould move in a live specimen. Each photograph alsocontained a ruler for scale.Functional morphologyWe used eleven landmarks to measure the morpho-logical components of four functional systems associatedwith prey capture: the suction index model, the opercu-lar four-bar linkage model, the opening jaw lever system,and jaw protrusion. Ten of the distances between land-marks (hereafter, “component traits”) are then used tocalculate 4 key performance traits of the four functionalsystems using the formulas in [29,32,33,61]. Landmarkswere digitized using the MATLAB program DLtdv3 [83],from which linear distances could be calculated betweenpairs of x and y coordinates. Epaxial width landmarkswere measured from the dorsal photographs. Epaxialheight, four-bar input link, coupler link, and fixed linklandmarks were measured from the closed-mouth lateralphotographs. Gape, buccal length, neurocranium outle-ver, and jaw opening outlever landmarks were measuredfrom the open-mouth lateral photographs. Calipers wereused to measure standard length (SL), defined as the dis-tance from the anterior-most point of the closed upperjaw to the posterior-most point of the vertebral column.The distance between the insertion of the interopercular-articular ligament and the point of articulation betweenthe quadrate and articular (landmarks 8 and 9, Table 1),which is used to calculate both opening inlever and theoutput link of the opercular four-bar, is not in plane in alateral photograph. We measured this distance by hand inall fish using a dissecting microscope at 50× magnificationwith an ocular micrometer (r2 = ?).From these measurements we calculated suction index,the displacement advantage (the ratio of output to inputdisplacement) of lower jaw opening, the transmission co-efficient of the opercular four-bar for a five-degree inputrotation, and jaw protrusion (Figure 2). The transmissioncoefficient refers to the amount of rotation produced bythe output link for a set amount of rotation in the inputlink [30]. We tested for divergence in these functionaltraits using Wilcoxon signed-rank tests (Table 1). We alsotested for limnetic-benthic differences in the 11 compo-nent traits used to calculate the functional traits.Before analyzing the individual component traits usedto derive our functional indices, we corrected each traitfor size with a log-log regression on standard length(SL). We chose standard length over other possible sizetraits (eg. centroid size), because SL is less affected by thefunctionally important hypertrophied epaxial musculatureof benthic stickleback. We used standardized major axisregression in the R package ‘smatr’ to verify there wereno statistically significant interactions between speciesand SL (at alpha = 0.05). In order to size correct ourtraits, we calculated residuals from a log-log linear regres-sion of each trait on SL and species, then calculated eachtrait at a common SL of 50 mm. We report the results oftests on these adjusted traits, but tests on the residualsgive equivalent results.KinematicsAll protocols for animal use and treatment were reviewedand approved by the University of British Columbia AnimalCare Committee and were in compliance with the guide-lines of the Canadian Council on Animal Care, applicationnumber A07-0293. Live fish used in the kinematic analysiscame from two experimental ponds at the University ofBritish Columbia. Each pond had been stocked with eitherwild adult benthic or limnetic fish from Paxton Lake, Brit-ish Columbia during the previous summer, and fish wereallowed to reproduce naturally. Juvenile stickleback weretrapped using unbaited minnow traps and transferred to110 L aquaria. Each fish was then placed singly in a20×10×9 cm plexiglass container attached to the topedge of the tank. Sex is known to affect sticklebackkinematics [84], so we only filmed non-sexually di-morphic juvenile fish. Fish were filmed using a NACMemrecam ci digital system (Tokyo, Japan) at 500 Hz. Wefilmed feeding strikes on live cladocerans (Daphniamagna), as cladocerans occur in both littoral habitat andopen water, benthic and limnetic stickleback both con-sume cladocerans in the wild, and both species deplete cla-doceran populations in mesocosm studies [85]. Prey wereintroduced to the aquarium singly with a pipette. Wefilmed until we obtained at least eight full-effort lateralstrikes per individual in benthics (n = 5) and limnetics(n = 5). After filming, each fish was euthanized with anoverdose of MS-222.We used a custom modification of the DLTdv3MATLAB package [83] to digitize and analyze eachstrike. We tracked ten landmarks on the head and usedthem to calculate excursion and timing variables forgape, jaw protrusion, cranial rotation, lower jaw rotation,and strike distance as described in Oufiero et al. [86].Excursion variables record the maximum value of a dis-tance variable, whereas timing variables indicate the timeit takes for the fish to reach its maximum for the appro-priate excursion variable. We excluded film sequences inMcGee et al. BMC Evolutionary Biology 2013, 13:277 Page 7 of 10http://www.biomedcentral.com/1471-2148/13/277which fish exhibited low effort on the strike, defined as amaximum gape less than 75% of the maximum gaperecorded for that individual. Once those sequenceswere excluded, we retained the three sequences foreach individual with the fastest time to peak gape, de-fined as the time in milliseconds between 20% of peakgape and 95% of peak gape. To ensure that sequencesfilmed from the same individual were not treated asstatistically independent, we used linear mixed modelsto compare kinematics between species. We treatedspecies and SL as fixed effects and each individual fishas a random effect. Including SL as a fixed effect allowsus to control for the expected effect of body size onteleost kinematics [87]. We used ‘pvals.fnc’ from the‘languageR’ R package to perform an MCMC permutationtest to estimate p-values and effect size for our fixedeffects [88]. 10,000 MCMC samples were generated foreach mixed model analysis, and p-values for species wereexamined for each of the kinematic variables (Table 2).We calculated the hydrodynamic (suction) force exertedon a simulated attached prey using the Suction InducedForce Field model, SIFF [28]. We parametrized SIFF usingour previously-described kinematic data in the same man-ner as [28], combined with Suction Index measurements.We then used SIFF to calculate the maximum force thatwould be exerted on a circular 2 mm prey item duringeach strike, retaining the three highest-force strikesper individual for analysis using the linear mixed-modelapproach described above.AbbreviationsSIFF: Suction induced force field; SL: Standard length.Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsMDM, DS, and PCW designed the study. MDM made all measurements andfilmed all fish. MDM wrote the first draft paper, with DS and PCW providingassistance for following drafts. All authors have read and approved the finalmanuscript.AcknowledgementsWe would like to thank our associate editor and two anonymous reviewersfor comments that substantially improved the scope of the manuscript. Thiswork was funded through NSF grant IOS-0924489. Correspondence andrequests for materials should be addressed to M.D.M. All data has beenarchived in Dryad.Author details1Department of Evolution and Ecology, University of California Davis, 1Shields Avenue, Davis, CA 95616, USA. 2Biodiversity Research Centre andZoology Department, University of British Columbia, 2329 W MallVancouverBC V6T 1Z4, Canada.Received: 14 September 2013 Accepted: 4 December 2013Published: 31 December 2013References1. Grant PR, Grant B: How and Why Species Multiply: The Radiation of Darwin’sFinches. New Jersey: Princeton University Press; 2011.2. Losos JB: Lizards in an Evolutionary Tree: Ecology and Adaptive Radiation ofAnoles. Berkeley: University of California Press; 2009.3. Losos JB, Mahler DL: Adaptive Radiation: the Interaction of EcologicalOpportunity, Adaptation, and Speciation, Evolution since Darwin: the first 150years. Massachusetts: Sinauer Associates; 2010:381–420.4. Schluter D: The Ecology of Adaptive Radiation. New York: Oxford UniversityPress; 2000.5. Schliewen UK, Klee B: Reticulate sympatric speciation in Camerooniancrater lake cichlids. Front Zool 2004, 1:5.6. Seehausen O: African cichlid fish: a model system in adaptive radiationresearch. Proc R Soc Lond B 2006, 27:1987–1998.7. Obbard DJ, Maclennan J, Kim KW, Rambaut A, O’Grady PM, Jiggins FM:Estimating divergence dates and substitution rates in the Drosophilaphylogeny. Mol Biol Evol 2012, 29(11):3459–3473.8. Kahilaninen KK, Malinen T, Tuomaala A, Alajarvi E, Tolonen A, Lehtonen H:Empirical evaluation of phenotype–environment correlation and traitutility with allopatric and sympatric whitefish, Coregonus lavaretus (L.),populations in subarctic lakes. Biol J Linn Soc 2007, 92:561–572.9. Langerhans RB, Chapman LJ, Dewitt TJ: Complex phenotype–environmentassociations revealed in an East African cyprinid. J Evol Biol 2007,20:1171–1181.10. Pfaender J, Schliewen UK, Herder F: Phenotypic traits meet patterns ofresource use in the radiation of “sharpfin” sailfin silverside fish in LakeMatano. Evol Ecol 2010, 24:957–974.11. Gould SJ, Lewontin RC: The Spandrels of San Marco and the PanglossianParadigm: A Critique of the Adaptationist Programme. Proc R Soc Lond B1979, 205:581–598.12. Wainwright PC: Functional Morphology as a Tool in Ecological Research.Ecological Morphology: Integrative Organismal Biology. Chicago: Univ.Chicago Press; 1994:42–59.13. Wainwright PC: Ecological explanation through functional morphology:the feeding biology of sunfishes. Ecology 1996, 77:1336–1343.14. Wainwright PC, Shaw SS: Morphological basis of kinematic diversity infeeding sunfishes. J Exp Biol 1999, 202:3101–3110.15. Carroll AM, Wainwright PC, Huskey SH, Collar DC, Turingan RG: Morphologypredicts suction feeding performance in centrarchid fishes. J Exp Biol2004, 207:3873–3881.16. McPhail JD: Ecology and evolution of sympatric sticklebacks(Gasterosteus): evidence for a species-pair in Paxton Lake, Texada Island,British Columbia. Can J Zool 1992, 70:361–369.17. Schluter D, McPhail JD: Ecological character displacement and speciationin sticklebacks. Am Nat 1992, 140:85–108.18. McPhail JD: Ecology and evolution of sympatric sticklebacks(Gasterosteus): origin of the species pairs. Can J Zool 1993, 71(3):515–523.19. McPhail JD: Speciation and the Evolution Of Reproductive Isolation in theSticklebacks (Gasterosteus) of South-western British Columbia, TheEvolutionary Biology of the Threespine Stickleback. USA: Oxford UniversityPress; 1994:399–437.20. Taylor EB, McPhail JD: Evolutionary history of an adaptive radiation inspecies pairs of threespine stickleabcks (Gasterosteus): insights frommitochondrial DNA. Biol J Linn Soc 1999, 66:271–291.21. Taylor EB, McPhail JD: Historical contingency and determinism interact toprime speciation in sticklebacks. Proc Roy Soc Lond B 2000, 271:2375–2384.22. Kiørboe T, Saiz E, Visser A: Hydrodynamic signal perception in thecopepod Acartia tonsa. Mar Ecol Prog Ser 1999, 1999(179):97–111.23. Holzman R, Wainwright PC: How to surprise a copepod: strike kinematicsreduce hydrodynamic disturbance and increase stealth of suction-feeding fish. Limnol Oceangr 2009, 54:2201–2212.24. Bentzen P, McPhail JD: Ecology and evolution of sympatric sticklebacks(Gasterosteus): specialization for alternative trophic niches in the EnosLake species pair. Can J Zool 1984, 62(11):2280–2286.25. Schluter D: Adaptive radiation in sticklebacks: size, shape, and habitatuse efficiency. Ecology 1992, 1992(74):699–709.26. Schluter D: Adaptive radiation in sticklebacks: trade-offs in feedingperformance and growth. Ecology 1995, 76:82–90.27. Gow JL, Peichel CL, Taylor EB: Ecological selection against hybrids innatural populations of sympatric threespine sticklebacks. J Evol Biol 2007,20(6):2173–2180.28. Holzman R, Collar DC, Mehta RS, Wainwright PC: An integrative modelingapproach to elucidate suction feeding performance. J Exp Biol 2012,215:1–13.McGee et al. BMC Evolutionary Biology 2013, 13:277 Page 8 of 10http://www.biomedcentral.com/1471-2148/13/27729. Wainwright PC, Carroll AM, Collar DC, Day SW, Higham TE, Holzman R:Suction feeding mechanics, performance, and diversity in fishes. IntegrComp Biol 2007, 47:96–106.30. Westneat MW: Feeding mechanics of teleost fishes (labridae; perciformes): atest of four-bar linkage models. J Morph 1990, 205:269–295.31. Anker GC: Morphology and kinetics of the head of the stickleback,Gasterosteus aculeatus. Trans Zool Soc Lond 1974, 32:311–416.32. Westneat MW: Transmission of force and velocity in the feedingmechanisms of labrid fishes (Teleostei, Perciformes). Zoomorphology1994, 114:103–118.33. Holzman R, Day SW, Mehta RS, Wainwright PC: Jaw protrusion enhancesforces exerted on prey by suction-feeding fishes. J Roy Soc Int 2008,5:1445–1457.34. Rundle HD: A test of ecologically dependent postmating isolationbetween sympatric sticklebacks. Evolution 2002, 56(2):322–329.35. McGee MD, Wainwright PC: Convergent evolution as a generator ofdiversity in threespine stickleback. Evolution 2013, 67:1204–1208.36. Hendry AP, Taylor EB: How much of the variation in adaptive divergencecan be explained by gene flow? An evaluation using lake-streamstickleback pairs. Evolution 2004, 58:2319–2331.37. Ravinet M, Prodöhl PA, Harrod C: Parallel and nonparallel ecological,morphological and genetic divergence in lake–stream stickleback from asingle catchment. J Evol Biol 2013, 26:186–204.38. Reid DT, Peichel CL: Perspectives on the genetic architecture of divergencein body shape in sticklebacks. Integr Comp Biol 2010, 50:1057–1066.39. Kimmel CB, Cresko WA, Phillips PC, Ullmann B, Currey M, Von Hippel F,Kristjánsson BK, Gelmond O, McGuigan K: Independent axes of geneticvariation and parallel evolutionary divergence of opercule bone shape inthreespine stickleback. Evolution 2012, 66:419–434.40. Kimmel CB, Hohenlohe PA, Ullmann B, Currey M, Cresko WA:Developmental dissociation in morphological evolution of thestickleback opercle. Evol Dev 2012, 14:326–337.41. Durie CJ, Turingen RG: The effects of opercular linkage disruption onprey-capture kinematics in the teleost fish Sarotherodon melanotheron.J Exp Zool 2004, 301:642–653.42. Wainwright PC, Alfaro ME, Bolnick DI, Hulsey CD: Many-to-one mapping ofform to function: a general principle in organismal design? Integ CompBiol 2005, 45:256–262.43. Day T, McPhail JD: The effect of behavioural and morphological plasticityon foraging efficiency in the threespine stickleback (Gasterosteus sp.).Oecologia 1996, 108:380–388.44. Wund MA, Baker JA, Clancy B, Golub JL, Foster SA: A test of the “flexiblestem” model of evolution: ancestral plasticity, genetic accommodation,and morphological divergence in the threespine stickleback radiation.Am Nat 2008, 172:449–462.45. Heinis F, Sweerts JP, Loopik E: Micro-environment of chironomid larvae inthe littoral and profundal zones of Lake Maarsseveen I, The Netherlands.Arch Hydrobiol 1994, 130(1):53–67.46. Limén H, Van Overdijk CDA, MacIsaac HJB: Food partitioning between theamphipods Echinogammarus ischnus, Gammarus fasciatus, and Hyalellaazteca as revealed by stable isotopes. J Great Lakes Res 2005, 31:97–104.47. Lauder GV: Historical biology and the problem of design. J Theor Biol1982, 97:57–67.48. Spoljaric MA, Reimchen TE: 10,000 years later: evolution of body shape inHaida Gwaii three-spined stickleback. J Fish Bio 2007, 70:1484–1503.49. Walker JA: Ecological morphology of lacustrine threespine sticklebackGasterosteus aculeatus L. (Gasterosteidae) body shape. Biol J Linn Soc1997, 61:3–50.50. Lavin PA, McPhail JD: The evolution of freshwater diversity in threespinestickleback (Gasterosteus acueatus): site-specific differentiation of trophicmorphology. Can J Zool 1985, 63:2632–2638.51. Reimchen TE, Stinson EM, Nelson JS: Multivariate differentiation ofparapatric and allopatric populations of threespine stickleback in theSangan River watershed, Queen Charlotte Islands. Can J Zool 1985,63:2944–2951.52. Aguirre WE: Microgeographical diversification of threespine stickleback:body shape–habitat correlations in a small, ecologically diverse Alaskandrainage. Biol J Linn Soc 2009, 98:139–151.53. Ravinet M, Prodöhl PA, Harrod C: On Irish sticklebacks: morphologicaldiversification in a secondary contact zone. Evol Ecol Res 2013,15:271–294.54. Reimchen TE, Bergstrom C, Nosil P: Natural selection and the adaptiveradiation of Haida Gwaii stickleback. Evol Ecol Res 2013, 15:241–269.55. Walker JA, Bell MA: Net evolutionary trajectories of body shape evolutionwithin a microgeographic radiation of threespine sticklebacks(Gasterosteus aculeatus). J Zool 2000, 252:293–302.56. Herrel A, Podos J, Huber K, Hendry AP: Evolution of bite force in Darwin’sfinches: a key role for head width. J Evol Biol 2005, 18:669–675.57. Wainwright PC, Osenberg CW, Mittelbach GG: Trophic polymorphism inthe pumpkinseed sunfish (Lepomis gibbosus Linnaeus): effects ofenvironment on ontogeny. Func Ecol 1991, 5(1):40–55.58. Stedman HH, Kozyak BW, Nelson A, Thesier DM, Su LT, Low DW, Bridges CR,Shrager JB, Minugh-Purvis N, Mitchell MA: Myosin gene mutation correlateswith anatomical changes in the human lineage. Nature 2004, 428:415–418.59. Herrel A, Van Wassenbergh S, Wouters S, Adriaens D, Aerts P: A functionalmorphological approach to the scaling of the feeding system in theAfrican catfish, Clarias gariepinus. J Exp Biol 2005, 208:2091–2102.60. Herrel A, O’Reilly JC: Ontogenetic scaling of bite force in lizards andturtles. Phys Biochem Zool 2006, 79(1):31–42.61. Westneat MW: Evolution of levers and linkages in the feedingmechanisms of fishes. Integr Comp Biol 2004, 44:378–389.62. Hagen DW, Gilbertson LG: Geographical variation and environmentalselection in gasterosteus aculeatus L. In the Pacific Northwest, America.Evolution 1972, 26:32–51.63. Gross HP, Anderson JM: Geographic Variation in the Gill Rakers and Diet ofEuropean Threespine Sticklebacks. Copeia: Gasterosteus aculeatus; 1984:87–97.64. Magnuson JJ, Heitz JG: Gill raker apparatus and food selectivity amongmackerels, tunas, and dolphins. Fish Bull 1971, 69:361–370.65. Lundsgaard‐Hansen B, Matthews B, Vonlanthen P, Taverna A, Seehausen O:Adaptive plasticity and genetic divergence in feeding efficiency duringparallel adaptive radiation of whitefish (Coregonus spp.). J Evol Biol 2013,26:483–498.66. Roesch C, Lundsgaard-Hansen B, Vonlanthen P, Taverna A, Seehausen O:Experimental evidence for trait utility of gill raker number in adaptiveradiation of a north temperate fish. J Evol Biol 2013, 26:1578–1587.67. Drenner RW, Vinyard GL, Hambright KD, Gophen M: Particle ingestion byTilapia galilaea is not affected by removal of gill rakers andmicrobranchiospines. Trans Am Fish Soc 1987, 116(2):272–276.68. Sanderson SL, Stebar MC, Ackermann KL, Jones SH, Batjakas IE, Kaufman L:Mucus entrapment of particles by a suspension-feeding tilapia (Pisces:Cichlidae). J Exp Biol 1996, 199:1743–1756.69. Robinson BW, Wilson DS: Character release and displacement in fishes: aneglected literature. Am Nat 1994, 144(4):596–627.70. Bell MA, Andrews CA: Evolutionary Consequences of Colonization of FreshWater by Primitively Anadromous Fishes, Evolutionary ecology of freshwateranimals, current trends and case studies. Birkhäuser Basel: EvolutionaryEcology of Freshwater Animals; 1997:323–363.71. Reimchen TE: Spine deficiency and polymorphism in a population ofgasterosteus aculeatus: an adaptation to predators? Can J Zool 1980,58:1232–1244.72. Vamosi SM, Schluter D: Character shifts in the defensive armor ofsympatric sticklebacks. Evolution 2004, 58:376–385.73. Clarke JM, Schluter D: Colour plasticity and background matching in athreespine stickleback species pair. Biol J Linn Soc 2011, 102:902–914.74. Blake RW, Law TC, Chan KHS, Li JFZ: Comparison of the prolongedswimming performances of closely related, morphologically distinctthree-spined sticklebacks gasterosteus spp. J Fish Biol 2005,67:834–848.75. Odling-Smee LC, Boughman JW, Braithwaite VA: Sympatric species ofthreespine stickleback differ in their performance in a spatial learningtask. Behav Ecol Soc 2008, 62(12):1935–1945.76. Hendry AP, Hudson K, Walker JA, Räsänen K, Chapman LJ: Geneticdivergence in morphology–performance mapping between MistyLake and inlet stickleback. J Evol Biol 2011, 24(1):23–35.77. Park PJ, Bell MA: Variation of telencephalon morphology of thethreespine stickleback (Gasterosteus aculeatus) in relation to inferredecology. J Evol Biol 2010, 23:1261–1277.78. Park P: Spatial learning ability of the threespine stickleback (Gasterosteusaculeatus) in rlation to inferred ecology and ancestry. Evol Ecol Res 2013,15:213–239.79. Boughman JW: Divergent sexual selection enhances reproductiveisolation in stickleback. Nature 2001, 411:944–948.McGee et al. BMC Evolutionary Biology 2013, 13:277 Page 9 of 10http://www.biomedcentral.com/1471-2148/13/27780. Wark AR, Peichel CL: Lateral line diversity among ecologically divergentthreespine stickleback populations. J Exp Biol 2010, 213:108–117.81. Bell MA, Foster SA: Introduction to the Evolutionary Biology of the ThreespineStickleback, Evolutionary Biology of the Threespine Stickleback. USA:Evolutionary Bioloyg of the Threespine Stickleback" Publisher is OxfordUniversity Press; 1994:1–27.82. Dingerkus G, Uhler LD: Enyzme clearing of alcian blue stained wholevertebrates for demonstration of cartilage. Stain Tech 1977, 52(4):229–232.83. Hedrick TL: Software techniques for two- and three-dimensionalkinematic measurements of biological and biomimetic systems. BioinspirBiomim 2008, 3:034001.84. McGee MD, Wainwright PC: Sexual dimorphism in the feeding kinematicsof threespine stickleback. J Exp Biol 2013, 216:835–840.85. Harmon LJ, Matthews B, Des Roches S, Chase JM, Shurin JB, Schluter D:Evolutionary diversification in stickleback affects ecosystem functioning.Nature 2009, 458:1167–1170.86. Oufiero CE, Holzman RA, Young FA, Wainwright PC: New insights fromserranid fishes on the role of trade-offs in suction feeding diversification.J Exp Biol 2012, 215:3845–3855.87. Wainwright PC, Richard BA: Predicting patterns of prey use frommorphology of fishes. Env Biol Fish 1995, 44:97–113.88. Baayen RH: Analyzing Linguistic Data. Cambridge, UK: Cambridge UniversityPress; 2008.doi:10.1186/1471-2148-13-277Cite this article as: McGee et al.: Functional basis of ecologicaldivergence in sympatric stickleback. BMC Evolutionary Biology2013 13:277.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitMcGee et al. BMC Evolutionary Biology 2013, 13:277 Page 10 of 10http://www.biomedcentral.com/1471-2148/13/277


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