USF Libraries
USF Digital Collections

The early ontogeny of feeding in two shark species

MISSING IMAGE

Material Information

Title:
The early ontogeny of feeding in two shark species developmental aspects of morphology, behavior, and performance
Physical Description:
Book
Language:
English
Creator:
Lowry, David C
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Ecomorphology
Aquatic feeding
Modulation
Variability
Developmental trajectory
Dissertations, Academic -- Biology -- Doctoral -- USF
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: Early ontogeny is a time of rapid anatomical and behavioral development in most organisms. The degree of synchrony between form and function during this period, and the concomitant performance consequences, can strongly impact individual survival. Understanding the development of feeding during early ontogeny is important because nutrient acquisition universally influences organismal biology. A one-year, longitudinal feeding study was conducted for two elasmobranch species that were selected for their disparate morphology, behavior, and habitat: the whitespotted bambooshark Chiloscyllium plagiosum and the leopard shark Triakis semifasciata. To quantify changes in cranial morphology, external attributes of the feeding apparatus were measured weekly. Additionally, specimens were dissected to examine trends in the growth of select muscles and the volume of the buccal cavity. To quantify feeding behavior, individuals were observed weekly using high-speed digital cameras^ as they consumed various food types. Suction performance was evaluated using particle image velocimetry and direct measurements of suction pressure. The cranial morphology of C. plagiosum exhibited primarily isometric growth while the cranial morphology of T. semifasciata was dominated by allometric growth. Allometric increases were noted in the cross-sectional area of every muscle examined in both species, though the primary hyoid depressor, the coracohyoideus, hypertrophied to a greater degree in C. plagiosum. Although intra-individual differences throughout ontogeny complicated comparison, modulation in response to food attributes was clearly evident in T. semifasciata but broadly absent in C. plagiosum. Over ontogeny C. plagiosum generated allometrically greater suction while T. semifasciata generated relatively less. The shape of the parcel of water ingested during feeding did not change over ontogeny in either species. The capacity to perform diverse feeding behaviors thro ughout ontogeny is not constrained in T. semifasciata but tends to be stereotyped and accompanied by enhanced performance in C. plagiosum. A functionally generalized feeding apparatus and repertoire may benefit T. semifasciata by allowing the use of diverse feeding behaviors in variable environments, such as estuaries, over ontogeny. Morphological and behavioral conservation of the feeding apparatus throughout ontogeny, however, may allow C. plagiosum to exploit taxonomically varied crevice-dwelling reef organisms using a single specialized behavior.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2005.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by David C. Lowry.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 113 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001912520
oclc - 173959356
usfldc doi - E14-SFE0001312
usfldc handle - e14.1312
System ID:
SFS0025633:00001


This item is only available as the following downloads:


Full Text

PAGE 1

The Early Ontogeny of Feeding in Two Shark Species: Developmental Aspects of Mor phology, Behavior, and Performance by David C. Lowry A dissertation submitted in partial fulfillment Of the requirements for the degree of Doctor of Philosophy Department of Biology College of Arts and Sciences University of South Florida Major Professor: Philip Motta, Ph.D. Robert Hueter, Ph.D. Susan Bell, Ph.D. Florence Thomas, Ph.D. Date of Approval: September 19, 2005 Keywords: ecomorphology, aquatic feeding, modulation, variability, de velopmental trajectory Copyright 2005, David C. Lowry

PAGE 2

Dedication This work is dedicated to researchers everywhere that seek knowledge for its inherent value and strive for its practical application every day.

PAGE 3

Acknowledgements I am deeply appreciative of the suppor t provided by my family and friends over the course of this study and throughout my lif e. Volunteer laborat ory assistance during various portions of this work was graciously provided by Andrey Castro, Coral Gehrke, Carey Heinrich, Daniel Hube r, Sarah Koh, Sonia Lowry, Kyle Mara, Michael Matott, Ramani Rasile, and Alpa Wintzer. Animal husbandry advice and practices were supplied by Kristen Mathews, Jack Morris, Chris Sc hreiber, and Michael Schulte. Specimens were obtained through Mote Ma rine Laboratory, Sarasota, FL and SeaWorld Adventure Park, Orlando, FL. Philip Motta, Robert Hu eter, Susan Bell, and Florence Thomas contributed to the content and presentation of the manuscript. This research was funded in part by Mote Marine Laboratory, a Tharpe Endowed Scholars Grant through the University of South Florida, and the PADI Project A.W.A.R.E. Foundation Grant Program. All animal maintena nce and experimental handling procedures were approved by the University of South Florida Institutional Animal Care and Use Committee under protocol numbers 1709 and 2299.

PAGE 4

i Table of Contents List of Tables................................................................................................................. ....iii List of Figures................................................................................................................ ......v Abstract....................................................................................................................... ......vii Chapter 1: Introduction....................................................................................................... 1 Aquatic Feeding.......................................................................................................2 The Ontogeny of Feeding........................................................................................4 Goals........................................................................................................................6 Significance..............................................................................................................7 Chapter 2: The ontogeny of feeding behavior and cranial morphology in the leopard shark Triakis semifasciata (Girard 1854): a longitudinal perspective .............9 Abstract....................................................................................................................9 Introduction............................................................................................................11 Methods and Materials...........................................................................................15 Experimental Animals ................................................................................15 Filming Techniques ....................................................................................15 Morphological Measurements ...................................................................19 Statistical Analyses ....................................................................................20 Results ....................................................................................................................2 3 Scaling of Kinematics ................................................................................25 Kinematic Trends .......................................................................................26 Scaling of Morphology ...............................................................................35 Discussion..............................................................................................................38 Chapter 3: The ontogeny of feeding behavior and cranial morphology in the whitespotted bambooshark Chiloscyllium plagiosum (Bennett 1830): a longitudinal perspective...............................................................................................45 Abstract..................................................................................................................45 Introduction............................................................................................................47 Methods and Materials...........................................................................................50 Experimental Animals ................................................................................50 Filming Techniques ....................................................................................50 Morphological Measurements ...................................................................53 Statistical Analyses ....................................................................................54 Results ....................................................................................................................57 Scaling of Kinematics ................................................................................58 Kinematic Trends .......................................................................................58

PAGE 5

ii Scaling of Morphology ...............................................................................61 Discussion..............................................................................................................68 Chapter 4: The comparative ontogeny of feeding performance in two sharks: the leopard shark Triakis semifasciata and the whitespotted bambooshark Chiloscyllium plagiosum ..............................................................................................72 Abstract..................................................................................................................72 Introduction............................................................................................................74 Methods and Materials...........................................................................................77 Experimental Animals ................................................................................77 Experimental Techniques ...........................................................................78 Statistical Analyses ....................................................................................83 Results ....................................................................................................................8 7 Scaling of Performance ..............................................................................89 Behavioral Effects on Performance ...........................................................91 Discussion..............................................................................................................96 Ecological and Evolu tionary Implications ..............................................101 References..................................................................................................................... ...104 About the Author...................................................................................................End Page

PAGE 6

iii List of Tables Table 1: Regression para meters for kinematic variables versus total length (cm) for Triakis semifasciata .........................................................................................25 Table 2: Principal component loadings of kinematic variables associated with capture sequences in Triakis semifasciata .......................................................27 Table 3: Results of RM ANOVAs performed separately on each principal component from a PCA of prey capture kinematics in Triakis semifasciata (N=4)................................................................................................................28 Table 4: Parameters for log-log regre ssions against total length (cm) of external morphological variable data measured on live specimens of Triakis semifasciata ......................................................................................................36 Table 5: Parameters for log-log re gressions against total length (cm) of morphological variable data take n from dead specimens only of Triakis semifasciata ......................................................................................................37 Table 6: Regression para meters for kinematic variables versus total length (cm) for Chiloscyllium plagiosum ..................................................................................59 Table 7: Principal component loadings of kinematic variables associated with capture sequences in Chiloscyllium plagiosum ................................................60 Table 8: Results of RM ANOVAs performed separately on each principal component from a PCA of prey capture kinematics in Chiloscyllium plagiosum (N=5)..............................................................................................63 Table 9: Parameters for log-log regre ssions against total length (cm) of external morphological variable data measured on live specimens of Chiloscyllium plagiosum .........................................................................................................66 Table 10: Parameters for log-log regr essions against total length (cm) of morphological variable data take n from dead specimens only of Chiloscyllium plagiosum ..................................................................................67 Table 11: Comparison between capture ev ents with food catheterized and without, and with particles present and abse nt, during each time segment for both species via MANOVA based on princi pal component loading scores............88

PAGE 7

iv Table 12: Regression parameters for perf ormance variables versus total length and tests of isometry for both species.....................................................................90 Table 13: Regression parameters for perf ormance variables against age in days for both species irrespective of size.......................................................................92 Table 14: Kinematic variables contribut ing to performance differences over time irrespective of size in Triakis semifasciata ......................................................93 Table 15: Principal component loadi ngs after Equamax rotation of kinematic variables exclusively associated with motion of the shark during capture events in Triakis semifasciata ..........................................................................94 Table 16: Partial regression coefficien ts from the backward stepwise multiple regressions of principal components (P C) that contribute to performance differences over time irrespective of size in Triakis semifasciata ...................95 Table 17: Kinematic variables contribut ing to performance differences over time irrespective of size in Chiloscyllium plagiosum ..............................................96

PAGE 8

v List of Figures Figure 1: Composite photogr aphic series of representati ve prey capture sequences for three of the five food types utilized............................................................24 Figure 2: Comparison of principal component 1 (PC1) (general timing) and PC2 (cranial timing/displacement) scores among individuals irrespective of time for Triakis semifasciata ...........................................................................29 Figure 3: Comparison of average principal component (PC) scores for food types irrespective of individua l and time for the components upon which overall differences between foods were found............................................................30 Figure 4: Trends in general timing variables (PC1) over time by food type in Triakis semifasciata ......................................................................................................32 Figure 5: Trends in hyoid timing va riables (PC4) over time by food type for Triakis semifasciata ......................................................................................................33 Figure 6: Trends in predator motion (PC5) over time by food type for Triakis semifasciata ......................................................................................................34 Figure 7: The trend in pr edator motion (PC5) over time compared to the number of attempts required to capture live shrimp in Triakis semifasciata ....................35 Figure 8: Scaled drawi ng of external morphology of Triakis semifasciata at age 1 day (light gray lines) and age 365 days (black lines).......................................36 Figure 9: Scaling of buccal vo lume relative to total length in Triakis semifasciata .......37 Figure 10: Composite photograp hic series of a representa tive food capture sequence for Chiloscyllium plagiosum ............................................................................57 Figure 11: Comparison of pr incipal component 1 (PC1) (general timing) and PC2 (predator motion) scores among indi viduals irrespective of time for Chiloscyllium plagiosum ..................................................................................62 Figure 12: Comparison of pr incipal component 1 (PC1) (general timing) and PC2 (predator motion) scores for large, sm all, and live food i rrespective of time for Chiloscyllium plagiosum ............................................................................64

PAGE 9

vi Figure 13: Trends in predator mo tion (PC2) over time by food type in Chiloscyllium plagiosum .........................................................................................................65 Figure 14: Scaling of buccal volu me relative to total length in Chiloscyllium plagiosum .........................................................................................................67 Figure 15: Illustrative examples of the range of relative size and shape of the parcel of water ingested by Chiloscyllium plagiosum (top) and Triakis semifasciata (bottom) during capture events...................................................83 Figure 16: Values of maximum subambient pressures generated during food capture events in Chiloscyllium plagiosum and Triakis semifasciata relative to total length (TL)...............................................................................................88 Figure 17: Aspect ratio values for parc els of water ingested during food capture events in Chiloscyllium plagiosum and Triakis semifasciata relative to total length (TL)...............................................................................................89

PAGE 10

vii The Early Ontogeny of Feeding in Two Shark Species: Developmental Aspects of Mor phology, Behavior, and Performance David C. Lowry ABSTRACT Early ontogeny is a time of rapid anatomi cal and behavioral development in most organisms. The degree of synchrony between form and function during this period, and the concomitant performance consequences, can strongly impact indi vidual survival. Understanding the development of feeding dur ing early ontogeny is important because nutrient acquisition universally influences organismal biology. A one-year, longitudinal feeding study was conducted for two elasmobran ch species that were selected for their disparate morphology, behavior, and ha bitat: the whitespotted bambooshark Chiloscyllium plagiosum and the leopard shark Triakis semifasciata To quantify changes in cranial morphology, external attr ibutes of the feeding apparatus were measured weekly. Additionally, specimens we re dissected to examine trends in the growth of select muscles and the volume of the buccal cavity. To quantify feeding behavior, individuals were obs erved weekly using high-spee d digital cameras as they consumed various food types. Suction perf ormance was evaluated using particle image velocimetry and direct measur ements of suction pressure. The cranial morphology of C. plagiosum exhibited primarily isometric grow th while the cranial morphology of T. semifasciata was dominated by allometric growth. Allometric increases were noted in

PAGE 11

viii the cross-sectional area of ev ery muscle examined in bot h species, though the primary hyoid depressor, the coracohyoideus, hype rtrophied to a greater degree in C. plagiosum Although intra-individual differences throughout ontogeny complicated comparison, modulation in response to food attr ibutes was clearly evident in T. semifasciata but broadly absent in C. plagiosum Over ontogeny C. plagiosum generated allometrically greater suction while T. semifasciata generated relatively less. The shape of the parcel of water ingested during feeding did not cha nge over ontogeny in either species. The capacity to perform diverse feeding behavi ors throughout ontogeny is not constrained in T. semifasciata but tends to be stereotyped and a ccompanied by enhanced performance in C. plagiosum A functionally generalized feeding apparatus and repertoire may benefit T. semifasciata by allowing the use of diverse feedi ng behaviors in variable environments, such as estuaries, over ontogeny. Morphol ogical and behavioral conservation of the feeding apparatus throughout ontogeny, however, may allow C. plagiosum to exploit taxonomically varied crevice-dwelling reef organisms using a single specialized behavior. Word Count: 342

PAGE 12

1 Chapter 1: Introduction For every organism, obtaining nutrients from resources available in the environment is paramount to survival. In gna thostomes, feeding is a crucial aspect of interaction with the external environment that has been f acilitated by the novel evolution of jaws supported by a rigid endoskelet on (Kent and Carr, 200 1; Benton, 2004). Modification of the anatomy, morphology, mechanical function, and utilization of the jaws over evolutionary time represents a major axis of phylogenetic and ecological diversification that has contribu ted to the radiation of this diverse lineage (Lauder, 1985a; 1985b; Liem, 1990; Schwenk, 2000; Wilga, et al., 2000; Ferry-Graham and Lauder, 2001). The long-term effects of feeding within the life of an indivi dual are numerous and diverse. Generation of tissue for growth, repair, and development can be drastically affected by nutrient availability. Not only is growth rate affected by gross nutrient uptake, but also by the type and quality of nut rients consumed. Because size, both of the entire organism and of select anatomical f eatures, has been shown to affect mechanical capacity, diverse aspects of performance, and behavioral interactions with other organisms, the rate at which size increases (i.e growth rate) is critical in determining the niche of an individual (Werne r and Gilliam, 1984; Galis, et al., 1994; Arendt and Wilson, 1997). Reproductive state and capacit y are also influenced by th e availability of nutrients and nutrient reserves, and the e xpense of energy associated w ith the act of mating can be

PAGE 13

2 immense. As such, the fitness of an or ganism is contingent upon feeding, making elucidating the roots of this behavior key to understanding the ecol ogy and evolution of a species. Aquatic Feeding The interaction between the anatomy, mechanical function, and behavioral application of the feeding apparatus has been broadly investigated in aquatic gnathostomes over the last forty years (Ale xander, 1967; Muller and Osse, 1984; Reilly and Lauder, 1988; Motta, et al. 1997; 2002; Ferry-Graham, 1998a; 1998b; Ferry-Graham and Lauder, 2001; Wainwright, et al., 2001a; 2001b). A prominent conclusion of this research is that physical attr ibutes of the aquatic medium substantially influence the capacity of these organisms to capture prey (Lauder and Clark, 1984; Muller and Osse, 1984; Wainwright, et al., 2001b; Higham, et al., 2005}. The principle of continuity dictates that expansion of th e buccopharyngeal cavity during feeding will cause an influx of water into the cavity as a consequence of the development of subambient pressure. During inertial suction feeding, this flow of water entrains prey or food items allowing them to be ingested by the predator. Iner tial suction feeding is the dominant, and presumably ancestral, mode of prey procur ement in the teleost fishes (Lauder, 1985a; 1985b) and early research on oste ichthian fishes identified a highly conservative feeding sequence associated with this prey capture m ode. In this sequence each bite consists of four phases: preparatory, expansive, co mpressive, and recovery (Liem, 1978). Subsequent studies have confirmed this feeding sequence is found among taxonomically distant species of aquatic gnathostomes, in cluding elasmobranchs (Motta, et al., 1997;

PAGE 14

3 Ferry-Graham, 1998a; 1998b; Wilga and Mott a, 1998a; 1998b; Robinson and Motta, 2002), amphibians (Reilly and Lauder, 1989; Reilly, 1995; Deban, 1997), and reptiles (Lauder and Prendergast, 1992; Van Damme and Aerts, 1997; Summers, et al., 1998b), with variations from the model primarily occu rring in the relative duration of each phase. Inertial suction feeding may or may not be associated with rapid forward motion of the predator (Norton and Brainerd, 1993). Simultaneous with the generation of subambient buccopharyngeal pressure the pred ator may overtake its prey or food and consume it completely or in discrete portions termed ram feeding (Norton and Brainerd, 1993). A single feeding strike can be descri bed as employing a combination of the two modalities (Norton and Brainerd, 1993). Despit e the apparently strict conservation of the medium-constrained, four-stage feedi ng sequence among aquatic gnathostomes, considerable research has shown that prey mobility, size, hardness, and type can all elicit modulation in teleosts, turtles, and elasm obranchs (Wainwright and Lauder, 1986; Lauder and Prendergast, 1992; Nemeth, 1997; Ferry -Graham, 1998a; Ferry-Graham, et al., 2001b). Modulation is the ability of an individual organism to consistently alter the motor and kinematic patterns of prey cap ture in response to changes in feeding conditions, including prey size and/or type (Chu, 1989; Sanderson, 1990). Another important type of deviation within the four-stage aquatic feeding model is termed variation and deals with the intrinsic unpredic tability that exists both between bites from a single organism and among the bites of cons pecifics employing the same kinematic and motor patterns to feed on similar prey. Variat ion differs from modulation in that it is not a predictable and consistent change in kine matics but, rather, a random fluctuation about

PAGE 15

4 the mean within an established kinema tic pattern (Wainwright and Lauder, 1986; Sanderson, 1990; Motta, et al., 1997). Although an organism may be capable of behaviorally modulating its feeding in response to prey size or type anatomical specializations that enhance the capacity to perform suction feeding have been noted. Notable specializations include a deep pharyngeal region, a protrusible upper jaw, a small, laterally-enclosed gape, reduced dentition, and a wide skull fo r the insertion of well-devel oped epaxialis muscles (Muller and Osse, 1984; Liem, 1993; Carroll et al. 2004). In short, th ese anatomical attributes enhance the capacity of the organism to genera te rapid or volumetric ally large influxes of water into the buccopharyngeal cavity. It is the interplay among anatomical features of the feeding apparatus and the behaviors for which they are employed that dictate the niche occupied by an organism. The Ontogeny of Feeding Ontogenetic changes are vari ations in the ecology or morphology of an organism that occur during development. Ontogeneti c changes in diet (Stoner and Livingston 1984; Lowe, et al., 1996; Brickle, et al., 2003) prey capture success (Reilly and Lauder, 1988; Coughlin, 1991), and various other behavioral and morphological parameters associated with feeding have long been noted in aquatic gnathostomes (Skulason, et al., 1989; Osse, 1990; Reilly and Lauder, 1990; El lis and Shackley, 1995; Reilly, 1995; 1996; Richard and Wainwright, 1995) Ontogenetic changes in the mechanical feeding elements and size of teleost fishes have been shown to greatly influence feeding performance and modality (Skulason, et al., 1989; Osse, 1990; Coughlin, 1991; Richard

PAGE 16

5 and Wainwright, 1995; Cook, 1996; Van Wassenbergh, et al., 2005). In lineages that metamorphose between larval and juvenile life stages, such as teleost fishes and amphibians, anatomical and behavioral ch anges concomitant with metamorphosis can drastically influence feeding (Coughlin, 1991; Reilly, 1995; 1996; Hernandez, et al., 2002; Krebs and Turingan, 2003). A crucial attr ibute of ontogenetic changes in feeding ability is that the capacity to feed must be maintained throughout development if the individual is to survive (Galis, 1990; 1993; Galis, et al., 1994). Understanding the contribution of changes in be havior and morphology to devi ations in performance over the course of ontogeny, allows prediction of changes in functiona l capacity and ecology over development. The ontogeny of feeding morphology and mechanisms in elasmobranchs, which do not undergo metamorphosis, has not been as thoroughly studied as in other aquatic gnathostomes. Ferry-Graham (1998b) suggested that the feeding mode of the swellshark Cephaloscyllium ventriosum might change between the hatc hling and juvenile life stages due to absolute consequences of size and the resultant inability of small sharks to generate suction, or to the higher leve l of swimming activit y displayed by younger sharks. Robinson and Motta ( 2002), by contrast, determined that feeding kinematics are isometric and conservative over ontogeny in the obligate suction feeding nurse shark Ginglymostoma cirratum Based on observations of feedi ng behavior for one individual both with and without food, Robinson and Motta (2002) suggests that feeding behavior in G. cirratum is, in fact, highly stereotyped. While sharks are widely recognized as toplevel predators (Cortez, 1999), the ecologica l role of young sharks as they undergoing

PAGE 17

6 ontogenetic dietary, morphologica l, and behavioral changes is poorly known and requires further study to make estimations of the ecological impact of early age classes. Goals The ultimate goal of this study was to quantify and find causation for ontogenetic changes in the feeding morphology, kinematics, and modalities of two disparate species of shark that could be used to predict ecological change during development. The whitespotted bambooshark Chiloscyllium plagiosum and the leopard shark Triakis semifasciata were the species selected because they differ substantially in cranial morphology, feeding behavior, and hab itat (Talent, 1976; Compagno, 1984a; 1984b, Ferry-Graham, 1998a; Wu, 1993; Kao, 2000). In or der to elucidate ont ogenetic trends at both the individual and species level, a longitudinal study design was used to follow several individuals of each species ov er the first year of life. Yearling T. semifasciata are known to employ a mixture of suction and ra m-feeding components in initial capture bites based on the type of prey offered, a lthough ram tends to dominate (Ferry-Graham 1998a). Adult C. plagiosum are known to primarily empl oy suction-dominated bites and this species belongs to a cl ade of bottom-associated a nd morphologically specialized suction feeders (Wu, 1993; 1994; Motta and Wilga, 2001; Motta, 2004). The six questions examined in this dissertation are: 1) Do food capture kinematics differ between C. plagiosum and T. semifasciata and, if so, in what functionally relevant ways? 2) Are initial capture bites of hatchling/neonatal (<3 month old) C. plagiosum and T. semifasciata suctionor ram-dominated?

PAGE 18

7 3) Does the feeding modality and/or kinema tic pattern exhibited by these species change ontogenetically and what are the effects of these changes on feeding performance? 3) Do these species modulate their feedi ng modality and/or kinematics based on food type through ontogeny? 4) Is cranial growth in thes e species isometric or allometr ic? What are the effects of changes in size and/or shape on feeding modality and/or kinematics? 5) Does the magnitude of sub-ambient bucc opharyngeal pressure and pattern of water flow into the buccopharyngeal cavity change in each species over the first year of life? 6) Are changes in size, morphology, behavior, and performance associated with changes in foraging strategy or prey preference as indicated by gut content analysis and existing behavioral studies? Significance The value of this research lies in its examination of the ont ogeny of food capture in two elasmobranch species that employ diffe rent feeding strategies utilizing similar anatomical mechanisms. The longitudinal design of this study allows collection of data from the same individuals at intervals th roughout early developmen t and correlation of changes in performance with changes in mo rphological and behavior al parameters. The importance of examining ecomorphological questi ons during a variety of life stages in a single species in order to present as comp lete an analysis as possible has been underscored by Barel, et al. ( 1989), Motta and Kotrschal (19 92), Galis, et al. (1994), and Liem and Summers (2000). Since the feeding kinematics of adult C. plagiosum and T. semifasciata have already been previously invest igated (Wu 1993; Ferry-Graham 1998a),

PAGE 19

8 the information gained from this study will supplement previous studies and expand their pertinence through ontogeny. Comparison of these findings with other studies of elasmobranch feeding will allow a more thorough understanding of the relationship between feeding structure and function both over ontogeny and through evolutionary time. Additionally, because the functional demands imposed by the aquatic medium on the feeding of elasmobranchs are also a pplicable to other aquatically feeding gnathostomes (Lauder, 1980; Muller and Osse, 1980; Liem, 1990), these findings will enhance understanding of the evolution of th e feeding apparatus of all gnathostomes.

PAGE 20

9 Chapter 2: The ontogeny of feeding be havior and cranial morphology in the leopard shark Triakis semifasciata (Girard 1854): a longitudinal perspective Abstract The relationship between form, function, and biological role in determining the niche occupied by an organism depends on diverse factors. Understanding this relationship is further complicated by c onsidering the interp lay among factors over ontogeny. While much is known about the ecological and functional morphology of feeding in lower vertebrates, studies of elasmobranch f eeding morphology and behavior over ontogeny are broadly lacking. In this study, the ontogeny of fe eding behavior and morphology was investigated in neonatal and young-of-the-year leopard sharks Triakis semifasciata using morphometric measurements of growth and high-speed videography in a longitudinal study. Five food types were used during filming sessions to facilitate differentiation of modulation and variation over ontogeny. F unctional aspects of muscle and buccal volume scaling were investigated through dissection. Size was shown to influence several kinematic variables and intraand inter-individual variability was the dominant factor contributing to variability in feeding behavior. Modulation of feeding behavior based on food size and elusivity was present for timing variables and predator motion during the strike, but not for food mo tion or the relative extent of buccal expansion. Allometric growth occurred in all aspects of external cranial morphology measured, resulting in a shallower head prof ile, anterior displacement of the mouth, and

PAGE 21

10 relatively larger jaw musculature over ont ogeny. While the degree to which morphology constrains or enhances behavi or can not be directly quantif ied, variability in behavior greatly exceeds variability in morphology over early ontogeny. Maintenance of a behaviorally and morphologically versatil e feeding apparatus throughout ontogeny is proposed to enhance the exploi tation of resources and fac ilitate a diverse diet in T. semifasciata under variable environmental conditions.

PAGE 22

11 Introduction A central tenet of functional morphology is that a reciprocal interaction between morphology and behavior (i.e. form and function) exists and affects the capacity of an organism to exploit resources (Bock, 1980; Wainwright, 1994). Ty ing this capacity to the ecology and evolutionary fitness of a species requi res quantification of the performance of the form-function complex (Arnold, 1983; Lauder, 1990) as well as its biological role ( Bock and von Wahlert, 1965; Bock, 1980). In the context of evolutionary biology, the functi onal and ecological morphology of feeding is of interest because the ability to obtain nutrients dire ctly influences individual survival, lifetime reproductive capacity, and fitness ( Schwenk, 2000; Ferry-Graham, et al., 2002). Despite being subjected to diverse and variable selective forces over both ontogeny and phylogeny, maintenance of the utility of the form-function complex throughout development is necessary (Galis, et al., 1994) to ensure the competitive capacity of an individual or species. For aquatically feeding vertebrates, a prey capture paradigm characterized by a posteriorly directed wave of sequential bu ccopharyngeal expansion has been developed based on studies of numerous clades ( Lauder and Lanyon, 1980; Motta, et al., 1991; Lauder and Prendergast, 1992; Reilly and Lauder, 1992). Muller and Osse (1984) and Lauder (1985) noted that functional conve rgence in aquatic feeding behavior and morphology can largely be attributed to c onstraints imposed by physical properties of water, such as density and viscosity. In or der to capture prey, organisms may exploit the density and viscosity of the flui d by entraining prey within a pa rcel of water that is drawn into the mouth (inertial suction feeding), or overtake and consume their prey (ram

PAGE 23

12 feeding) (Liem, 1980b; Norton and Brainerd, 1993). Although ram-feeding and suctionfeeding specialists exist, most aquatic ve rtebrates use a combination of these two modalities to exploit a broad taxonomic and fu nctional diversity of prey types (Liem, 1990). Morphological specializa tions correlated with a pred isposition towards use of a specific feeding mode include those facil itating generation of a broad, unobstructed buccal aperture, in the case of ram feeding, or a narrow buccal aperture leading into an expansible buccopharyngeal cavity, in the ca se of suction feeding (Muller and Osse, 1984; Liem, 1993). Despite the overall conservation of the aquatic feeding paradigm and its associated morphology across taxa systematic behavioral modul ation in response to prey elusivity, size, hardness, and type has been described in many aquatically feeding vertebrates (Liem, 1978; Lauder and Prende rgast, 1992; Anderson, 1993; Ferry-Graham, 1998a). Existing simultaneously with modulatio n, intrinsic variability in kinematic and motor patterns is nearly universal both in traand inter-individually among conspecifics feeding on the same prey or food (Shaffe r and Lauder, 1985a;b; Sanderson, 1990; Cook, 1996). In many cases inter-individual variability in feeding behavior is so extensive that it overwhelms inter-specific differences and complicates comparison among species (Wainwright and Lauder, 1986; Rei lly and Lauder, 1989; Norton, 1991). Though much is known about modulation and variation in prey capture behavior within the adult life stages of aquatically feeding verteb rates, less is known about the effects of these factors over the course of ontogeny (Cook, 1996). Several aspects of a predator change over ontogeny, directly aff ecting its interactio ns with prey and, consequently, its diet. Dramatic changes can occur in the size and shape of a predator

PAGE 24

13 during early stages of development, especi ally in metamorphic species, when growth rates are often at their highest (Grossm an, et al., 1980; Osse, 1990; Coughlin, 1991; Reilly, 1996). Additionally, small incremental changes in morphology and anatomy can accrue over ontogeny, enhancing the ability of the predator to exploit existing or novel prey types (Cook, 1996; Hernandez and Mott a, 1997; Cutwa and Turingan, 2000;). Exposure to prey items early in developmen t can cause behavioral preferences in a predator that lead to dietary specialization as a result of experience (W erner, et al., 1981). Additionally, when a predator repeatedly intera cts with a particular prey type it can apply the results of its experience to future enc ounters, facilitating fa ster, more efficient handling (Coughlin, 1991). Elasmobranchs are a model clade in wh ich to study the relationship between feeding morphology and behavior over ontogeny for a number of reasons. In contrast to osteichthian fishes, the elasmobranch f eeding mechanism comprises relatively few structural elements. Anatomical complexity can often be employed as a predictor of functional potential, especially when considered in a phyloge netic context (Lauder, 1981; Friel and Wainwright, 1999). Despite the comp arable lack of anatomical complexity in elasmobranch feeding systems, however, prey capture behavior, diet and ecological role differ greatly among species (Cortes, et al., 1996; Lowe, et al., 1996; Motta and Wilga, 2001). While the feeding morphology and behavior of sub-adult and adult elasmobranchs is well-studied (Wu, 1994; Motta and Wilga, 2001; Wilga, 2001; Robinson and Motta, 2002), little is known about the interplay betwee n these factors over early ontogeny. Elasmobranchs exhibit con tinuous, non-metamorphic growth and while size variation can be vast within a species (Compagno, 1984a ;b), shape variation of

PAGE 25

14 cranial elements over ontogeny is comparably small for those species for which it has been studied (Ellis and Shackley, 1995; Ferry-Graham, 1998b; Robinson and Motta, 2002). The objective of this study was to qua ntify changes in cranial morphology and feeding behavior over early on togeny in the leopard shark Triakis semifasciata (Girard 1854) (Triakidae). Identifying changes in mor phology and behavior is the first step to isolating the contribution of each of these f actors to feeding performance and ecology. Triakis semifasciata was chosen because this specie s is documented as a dietary generalist (Russo, 1975; Talent, 1976), poten tially indicating a di verse prey capture repertoire, and the feeding kinematics and m odality of early stage juveniles have been previously quantified (Ferry-Graham, 1998a). A longitudinal experimental design was employed to allow discrimination among i ndividuals and food type effects across ontogeny. The specific goals of this study were to: 1) assess the role of individual variabili ty and modulation of feeding behavior in response to food type over the course of ontogeny; 2) assess patterns of size and shap e variation in cranial morphology over early ontogeny and their relationship to feeding behavior; and 3) identify form-function complexes that vary through ontogeny and have the capacity to influence prey capture perfor mance and feeding ecology in neonatal and young-of-the-year (YOY) T. semifasciata

PAGE 26

15 Methods and Materials Experimental Animals Triakis semifasciata is a common neritic, deme rsal species found along the Pacific coast of North America and nor thern Central America (Compagno, 1984b). Triakis semifasciata is aplacental viviparous, giving birth yearly in May or June to litters of 7-36 young (Ackerman, 1971). Pups are appr oximately 20-26 cm at birth and grow between 2 and 4 cm/year (Compagno, 1984b; Ku sher, et al., 1992). Adults can reach a maximum size of 1.98 m TL (Miller and L ea, 1972), though most do not exceed 1.60 m TL (Compagno, 1984b). Triakis semifasciata is an opportunistic generalist that feeds on a broad diversity of prey including benthic invertebrate s (Ackerman, 1971; Russo, 1975; Talent, 1976; Kao, 2000). In Elkhorn Slough, California the di et of individuals between 40 and 70 cm TL was dominated by small grap sid crabs and other mobile crustaceans, although fishes, clam siphons, fish eggs, and echiuroid worm s increased in relative importance with maturation (Talent, 1976). Filming Techniques Four neonatal T. semifasciata (average TL 25.85 cm) were obtained through a commercial aquarium collector and raised for 52 weeks at Mote Marine Laboratory, Sarasota, Florida. Animals were main tained in a 2.4-meter diameter, 1400-liter semicircular communal tank at 27 1C and 322 ppt salinit y. During experimental sessions, individuals were transferred into a 150-liter filming tank filled with water from

PAGE 27

16 the communal holding tank. The temperatur e of the filming tank was maintained throughout experimental sessions at the above temperature. Animals were fed a maintenance diet c onsisting of 3-4% of their body weight in frozen market squid Loligo opalescens frozen Atlantic threadfin herring Opisthonema oglinum, and live and frozen grass shrimp Palaemonetes pugio ad libitum three times per week. Experimental sessions began within tw o weeks of arrival of the sharks at Mote Marine Laboratory, when individuals were de termined to be no more than three weeks old. Five food types were offered during experi mental sessions: 1) cut squid scaled to one half mouth width (MW); 2) cut squid scaled to MW; 3) cut herring scaled to one half MW; 4) cut herring scaled to MW; and 5) liv e shrimp scaled to MW in carapace length. These food types compose a set intended to examine modulation in food capture kinematics in response to small versus larg e and elusive versus non-elusive items. The number of failed attempts occurring prior to food capture was recorded for each capture event throughout the study. During filming sessions, which occurred weekly over the one-year experimental period, food items were presented in a haphazardly determined order until the individual approached satia tion, as evidenced by a decrease in overall activity level and an unwillingness to fee d. Although a filming session often comprised as many as ten capture sequences for a give n individual, only the first five were considered for analysis in order to avoi d potentially confounding e ffects of satiation (Sass and Motta, 2002). To obtain recordings of feeding even ts, a Redlake PCI 1000 high-speed digital camera (Redlake, San Diego, CA, USA) was placed perpendicular to the aquarium providing a lateral view. Recordings we re made at 250 fps and illumination was

PAGE 28

17 provided by two, 500-Watt quartz-ha logen lights. Animals we re trained to feed under illumination prior to the experiment and we re allowed a 20-minute acclimation period prior to each feeding session. A Plexigla s false bottom divided the tank vertically, allowing a ventral view to be obtained by pl acing a mirror beneath the shark. A ruler beside the shark provided dist ance measure and only orthogona l views were retained for analysis. Kinematic data were obtained from recordings using Redlake MotionScope PCI software version 2.21.1 (Redlake, San Diego, CA, USA) and SigmaScan Pro version 4 (SPSS Inc.) The variables measured were sel ected for their availability over the course of ontogeny, functional relevance, and pr ior employment in other studies of elasmobranch feeding (Motta, et al., 1997; Ferry-Graham, 1998a; Wilga and Motta, 1998b). The 52-week experimental time period was broken into four equal segments to facilitate statistical analysis A total of five capture se quences per food type (N=5) per individual (N=4) per time segment (N=4) were recorded, for a tota l of 400 sequences. Sequences were not always obtained weekly for each food type/individual combination due to satiation. From the onset of mandi ble depression (time 0 ms), the following kinematic variables were quantified, with de scriptions provided for uncommon measures: 1) strike distance, from the closest point on th e food to the lower jaw of the shark (cm) at time 0 ms; 2) maximum gape (cm); 3) time to maximum gape (ms); 4) maximum cranial elevation angle (degrees), measured relative to resting head positi on by calculating the difference in the angle formed between the dorsum of the head and the body with its vertex at the inflection poin t of the head and the body; 5) time to maximum cranial elevation angle (ms); 6) time to onset of crania l elevation (ms); 7) time to offset of cranial elevation (ms); 8) duration of cranial eleva tion (ms); 9) time to onset of hyoid depression

PAGE 29

18 (ms); 10) maximum hyoid depression (cm), meas ured relative to resti ng head position at the position of the hyoid; 11) time to maxi mum hyoid depression (ms); 12) time to hyoid retraction (ms); 13) duration of hyoid depressi on (ms); 14) total strike duration (ms), from onset of lower jaw depression to ja w closure on the food; 15) duration of food movement (ms), from the field at which the food began to move until the food entered the mouth; 16) distance moved by the food (cm), ov er the duration of th e prior variable; 17) velocity of the food (cm s-1), over the course of the move ment measured in the prior variable; 18) distance moved by the pred ator (cm), during the duration of food movement; and 19) velocity of the predator (cm s-1), over the course of the movement measured in the prior variable. No meas ures were made of upper jaw protrusion excursion or timing because protru sion was not present in all cases. Variables (16) and (18) were used to calculate the ram-suction index (RSI) (Norton and Brainerd, 1993). The RSI is calculated as (DPREDATOR–DPREY)/ (DPREDATOR+DPREY), where D is the distance moved by either the predator or prey, and indicates the relative contribution of forwar d motion of the predator and motion of the prey to a given capture event. An RSI valu e of 1 indicates a purely ram-based bite and a value of –1 indicates a purely suction-based bi te. Because of inherent problems with the RSI (Van Damme and Aerts, 1997), including the rapidly decreasing effect of suction with increasing distance from the mouth of the predator (Lauder and Clark, 1984; Muller and Osse, 1984), it is used here primarily fo r comparison with othe r studies of fish feeding kinematics and as a generalized indicator of overall feeding modality.

PAGE 30

19 Morphological Measurements To chronicle changes in individual shape over ontogeny, bot h the lateral and ventral camera views were used to measure a suite of external morphological variables. The definitions of these variables followed Compagno (1984b) and c onsisted of: 1) total length; 2) mouth width; 3) mouth length; 4) preoral le ngth; 5) preorbital length; 6) prebranchial length; 7) branchial length; 8) prepectoral length; 9) anterior pectoral fin base to lower jaw; 10) head length; 11) head width; and 12) head de pth at the position of the hyoid. To verify that these measurements were accurate and not influenced by error associated with taking them from video imag es, individuals were removed from the tank every 7 weeks and the same measurements taken. As the measurement error associated with these variables was always less that 2.8%, measurements take n from the recorded footage were deemed valid for describing growth trends. To obtain measurements of muscle mass, muscle cross-sectional area, and buccal volume that could not be obtained from liv e individuals, fresh-dead specimens (N=8) spanning the entire range of total lengths of live specimens used in this study were obtained from commercial fishers. The pr eviously described mo rphological variables measured on live sharks were also measured for dead specimens. Wet muscle mass and cross-sectional area were obtained by excisi ng key muscles whose role in feeding is known (Motta, et al., 1991; Wu, 1994; Motta and Wilga, 1995; 1999). The coracomandibularis, coracohyoideus, and coracoarcu alis were selected due to their role as jaw abductors and hyoid depressors, and the quadratomandibularis was selected for its role in jaw adduction. The quadratomandibul aris was separated into its constituent divisions (anterior, posterior, superficial, and ventral) for measurement because it is a

PAGE 31

20 complex muscle, and Huber and Motta (2004) fo und that treating its di visions separately produced the most accurate estimate of bite forc e. The cross-sectional area from the four divisions was then added toge ther to obtain the morphological cross-sectional area of the muscle as a whole. SigmaScan Pro vers ion 4 (SPSS Inc.) was used to determine anatomical cross-sectional area from digi tal photographs taken with a Nikon Coolpix 4300. The palatoquadrate and Meckel’s cartila ges were also excised and their combined weight recorded because they are the skel etal elements upon which the jaw-abducting and -adducting musculature acts and their allometric growth could influence the velocity and force of feeding motions. Measures of buccal volume were obtaine d by injecting silicone sealant into the mouth of each dead specimen (N=8) and allo wing it to cure for 36 hours before removal and weighing. Using the mass-density of silicone reported by Cook (1996) (0.06 g ml-1), this weight was converted into a volume. Casts were made of the buccal cavity in a resting position and in a maximally expanded position based on kinematic footage. The difference between these two volumes was term ed the buccal reserve volume and used to indicate the maximum potential change in the volume of the buccal cavity during a capture event at disc rete points over ontogeny. Statistical Analyses All kinematic data were log-transformed and checked for normality and homogeneity of variance, employing Kolm ogorov-Smirnov and Levene Median tests, respectively, using SigmaStat Pro version 3.1 (SPSS Inc.). A Spearman Rank Order Correlation test was performed, as several va riables exhibited skew ed distributions, to

PAGE 32

21 establish the relationship between each variable and total length. Several variables were found to be highly correlated (P <0.01) with total length. To assess the nature of these relationships, Model II linear regressions we re performed using the average value for each individual for which kinematic data were available at a given total length (Richard and Wainwright, 1995). Model II regression s were appropriate because both of the variables used in these comparisons contai ned measurement error. Regressions were performed for all individuals combined and then for each individual separately to compare regression coefficients among indi viduals. Differences among individuals within a given kinematic variable were asse ssed using a modified St udent’s t-test (Zar, 1999). To standardize the statistical handling of all data for further analysis, all variables that did not show a relationship to total leng th were regressed agai nst total length using Model II linear regressions and the stude ntized residuals obta ined (Quinn and Keough, 2002). Size-corrected, studentized residuals for all kinematic variables were then used in a correlation matrix-based Principal Com ponents Analysis (PCA) that reduced the expansive kinematic data set to a few, or thogonally oriented com posite variables whose contribution to overall patterns of variation in the data could be elucidated (Quinn and Keough, 2002). An Equamax rotation was used because it produced higher loadings than any other rotation, enhancing data interpretation. Variables that loaded above an absolute value of 0.5 were identified as contributing heav ily to the variability within the respective principal component (PC). Principal com ponents with an eigenvalue greater than 1.0 were retained for further analysis. The f actor loading scores for each capture sequence on each principal component were then us ed in a two-way, mixed-model repeated

PAGE 33

22 measures MANOVA to identify differences be tween clusters of capture sequences in multivariate space over the repeated measure tim e. Individual was a random factor and food type was a fixed factor tested over the interaction term. Significance was assessed using Pillai’s trace because it is robust to multivariate deviance from normality (Zar, 1999). To further investigate differences identified in multivariate space by the RM MANOVA, a two-way, mixed-model RM ANO VA was performed for each principal component separately using individual as a ra ndom factor and food type as a fixed factor tested over the interaction term. Statistical significance among diffe rences detected by the RM ANOVA was evaluated usi ng Fisher’s LSD with an level of 0.05. To determine the order of the be st-fit equation describing tre nds over the repeated measure time, single degree of freedom polynomial contra sts were utilized. B onferroni corrections were not applied to the level of any tests due to their tendency to increase the rate of type II errors (Cabin and Mitchell, 2000; Moran, 2003). The PCA, MANOVA, and RM ANOVA tests were performed using Systat 11 (SPSS Inc.). Morphological variables measured on liv ing specimens were log-transformed and regressed against the logarithm of total length using Model II linear regressions in order to investigate changes in the relative dimensions of these variables over time. If the pattern of growth was isomet ric for a given variable, a sl ope of one was predicted for these regressions. Significant deviations from a slope of one were te sted using Student’s t (Zar, 1999) and indicated a relative retardation or acceleration in growth rate for the given feature. Regressions were performed for each individual separately and then 95% confidence intervals were determined for all live individuals comb ined. For variables that were measured on both live and dead spec imens, the data taken from dead specimens

PAGE 34

23 were compared against the confidence intervals for the live individuals to determine if the data fell within this range, indicating that it could have been drawn from the same population. Comparing live and dead scaling data via this method allowed application of data collected exclusively on dead specimens to live specimens within the appropriate size range. Morphological variab les that were measured exclusively on dead specimens were log-transformed and regr essed against the logarithm of total length using Model II linear regressions. The slope of the regres sion equation for each variable was tested against the expected slope using Student’s t. The expected slope of these regressions depended on the dimensionality of the variab le being considered, with a slope of one expected for linear variables, a slope of tw o expected for planar variables (e.g. muscle cross section), and a slope of three expected for cubic variables (e.g. muscle mass, buccal volume). Results Individuals generally swam slowly around the tank until a piece of food was introduced. The shark would then orient to the food within ~2-5 seconds and approach rapidly. Food was typically engulfed on the fi rst pass, except in the case of live shrimp (see ‘Kinematic Trends’ below). Food captu re kinematics began with either cranial elevation or lower jaw depressi on and continued in a posteriorly directed fashion (Fig. 1). The temporal sequence of kinematic events was conserved over the period of ontogeny studied, with the exception of maximum ga pe and maximum cranial elevation, which were occasionally transposed. When upper jaw protrusion was apparent it occurred after maximum gape and reached its maximum duri ng jaw closure. The absolute distance

PAGE 35

24 from the lower jaw to the food at time zero (strike distance) did not change over ontogeny (P=0.261), regardless of individual (P=0.588) or food type (P=0.264), and averaged 0.940.46 cm. The relative distance to the f ood as a percent of head length averaged 11.05.8% and did not exhibit a trend over ontogeny (P=0.230). Figure 1: Composite photographic series of representativ e prey capture sequences for three of the five food types utilized. The subset of food types se lected illustrates maximum differences in capture kinematics. The left column depicts capture of a large piece of fish, the center column depicts capture of a small piece of squid, and the right co lumn depicts capture of a live shrimp The point on the food that is farthest from the shark’s mouth is indicated by a white dot. This was the point used in calculating the RSI. Time is shown in the upper right corner of each field. Time 0 ms corresponds to the beginning of rapid jaw opening. A indicates maximum gape; B indicates maximum cranial elevation; C indicates maximum hyoid depression; D indicates food capture; E indicates jaw closure on the food. Note that the sequence of kinematic events is largely conserved but that absolute timing of events varies among food types.

PAGE 36

25 Scaling of Kinematics Approximately half of the kinematic va riables measured exhibited a relationship with total length when individuals were co mbined, with displacement variables tending toward smaller values, timing and velocity variables tending toward longer and slower values, and the RSI tending toward ram domi nance over ontogeny (Table 1). Within a given variable, however, signi ficant differences among indivi dual scaling coefficients existed over ontogeny for nearly three-quarters of the variable s (Table 1). For variables quantifying time to and extent of maxi mum linear and angular displacements, relationships with total length were identifie d for measures describing motion of the jaws and cranium but not hyoid depression. Table 1: Regression parameters for kinematic variables versus total length (cm) for Triakis semifasciata Results shown are for all individuals combined (N=4). Where differences among individuals were detected using a modified Student’s t-test, the magnitude of these differences is indicated. Max=maximum; Depress=depression; Elev=elevation. Individuals Combined Individuals Separated Variable Slope y-int StErr P Range of Slopes Max Gape -0.56 1.01 0.15 <0.001* -1.25 -0.59 Time to Max Gape 0.35 1.18 0.23 0.021* 0.17* 0.54* Max Cranial Elev Angle -2.08 3.54 0.60 <0.001* -3.82** 0.70** Time to Max Cranial Elev Angle ---0.269 0.10** 1.10** Time to Onset of Cranial Elev ---0.641 -0.88** 0.64** Time to Offset of Cranial Elev -0.42 2.60 0.21 0.044* -0.49* 0.22* Duration of Cranial Elev -1.26 3.78 0.48 0.010* -2.29** -0.35** Time to Onset of Hyoid Depress ---0.077 -0.02* 1.01* Max Hyoid Depress ---0.853 -0.30 0.80 Time to Max Hyoid Depress ---0.497 -0.71 0.10 Time to Hyoid Retraction ---0.314 -1.27** 0.24** Duration of Hyoid Depress ---0.190 -1.45** 0.13** Total Strike Duration ---0.130 -0.49* 0.62* Distance Moved by Food -0.95 1.46 0.29 0.002* -1.87** -0.26** Duration of Food Movement 0.56 0.88 0.17 <0.001* 0.20* 0.99* Velocity of Food -1.51 3.58 0.33 <0.001* -2.55** -0.45** Distance Moved by Predator ---0.141 -0.21** 1.75** Velocity of Predator ---0.305 -0.89 0.66 RSI 0.46 -0.69 0.14 <0.001* 0.32 1.00 Strike Distance ---0.143 -1.25 1.01

PAGE 37

26 Kinematic Trends Principal components analys is reduced the kinematic data into six components that accounted for 78.11% of the overall variabil ity in the data set (Table 2). Timing and duration variables loaded heavily on princi pal components (PC) 1, 2, and 4. Variables describing cranial elevation and depression lo aded heavily on PC2, variables describing hyoid retraction loaded heavily on PC4, and the remaining timing variables loaded heavily on PC1. Though not a timing variab le, maximum cranial elevation angle also loaded on PC2. Together the three timi ng components accounted for 46.20% of the overall variability in the data set (Table 2) PC3 and PC5 were characterized by variables quantifying forward motion of the food and pred ator, respectively, over the course of the strike. RSI loaded on both PC3 and PC5, but in opposite directions as would be expected from the nature of this index. Lastly, PC 6 was characterized by measures of maximum hyoid and gape displacement distance, or bucca l excursion. Strike di stance did not load heavily on any component. The RM MANOVA conducted simultaneou sly on all six principal components indicated that differences existed over ontogeny (Pillai Trace df=3; F=5.37; P=0.001), among individuals over ontogeny (df=9; F=4.45; P<0.001), among food types over ontogeny (df=12; F=1.83; P=0.044), and among f ood types for a given individual over ontogeny (df=36; F=1.61; P=0.020). Despite a hi gh degree of overlap, when each PC was examined with a separate RM ANOVA fo r individual differences irrespective of time and food type, differences were found on all components (Table 3, Fig. 2). Food type differences, independent of time and individual, were present on all components except PC3 (food motion) and PC6 (buccal excurs ion) (Table 3). Bites on large fish

PAGE 38

27Table 2: Principal component loadings of kinematic va riables associated with capture sequences in Triakis semifasciata Bold face values indicate variables determin ed to load heavily on the respective component (loading scores>|0.5 ) (N=4). Together the six components expl ain 78.11% of the overall variability in the data set. For clarity, all load ings<|0.25| are replaced by 0. Max=maximum; Depress=depression; Elev=elevation. Variable PC 1 PC 2 PC 3 PC 4 PC 5 PC 6 Time to Max Gape 0.852 0 0 0 0 0 Duration of Food Movement 0.838 0 0 0 0 0 Time to Max Hyoid Depress 0.828 0 0 0 0 0 Total Strike Duration 0.823 0 0 0 0 0 Time to Onset of Hyoid Depress 0.718 0 0 0 0 0 Time to Max Cranial Elev Angle 0.508 0 0 0 0 0.446 Duration of Cranial Elev 0 0.945 0 0 0 0 Max Cranial Elev Angle 0 0.869 0 0 0 0 Time to Offset of Cranial Elev 0.458 0.608 0 0 0 0.39 Time to Onset of Cranial Elev 0 -0.604 0 0.364 0 0.344 Distance Moved by Food 0 0 0.954 0 0 0 Velocity of Food -0.332 0 0.920 0 0 0 RSI 0 0 -0.699 0 0.668 0 Duration of Hyoid Depress 0 0 0 0.937 0 0 Time to Hyoid Retraction 0 0 0 0.926 0 0 Velocity of Predator -0.278 0 0 0 0.921 0 Distance Moved by Predator 0.389 0 0 0 0.878 0 Max Hyoid Depress 0 0 0 0 0 0.789 Max Gape 0 0 0 0.487 0 0.502 Strike Distance 0 0.328 0.306 -0.318 0 0.381 Eigenvalue 4.282 2.605 2.467 2.353 2.198 1.715 Percent Variance Explained 21.408 13.027 12.336 11.766 10.990 8.575 tended to have the largest values on all com ponents, followed by bites on small fish. This trend was reversed on PC1 (general timing va riables), with bites on large squid having high values and bites on both sizes of fish having small values. Bites on live shrimp tended to have the lowest values on PCs 1, 4, and 5 (general and hyoid timing variables and predator motion), indicating that these strikes were characte rized by brief, fast cephalic motions, and a smaller ram component (Fig. 3). Differences among sharks for a given food (i.e. effect interactions) were onl y found on PC6 (buccal excursion), with the values for one individual being consistently high and the values for two individuals

PAGE 39

28 fluctuating greatly based on food type. In summary, while a high degree of intraand inter-individual variability was present in all aspects of food capture over ontogeny, individuals modulated timing a nd modality aspects of food cap ture, but not the extent of buccal excursion, in a consistent manner in response to food type. Table 3: Results of RM ANOVAs performed separately on each principal component from a PCA of prey capture kinematics in Triakis semifasciata (N=4). df=Degrees of Free dom; TS=Time Segment. For all between subjects comparisons the error degrees of freedom are 80, while for all within subjects comparisons the error degr ees of freedom are 240. PC1 General Timing PC2 Cranial Timing, Displacement df F P df F P Between Subjects: Shark 3 15.278 <0.001* 3 22.147 <0.001* Food 4 7.700 0.004* 4 7.316 0.005* Shark*Food 12 0.625 0.815 12 1.517 0.135 Within Subjects: Time Segment 3 5.452 0.004* 3 0.623 0.671 TS*Shark 9 0.791 0.625 9 2.936 0.003* TS*Food 12 3.767 <0.001* 12 3.281 <0.001* TS*Shark*Food 36 0.94 0.571 36 1.338 0.105 PC3 Food Motion PC4 Hyoid Timing df F P df F P Between Subjects: Shark 3 14.544 <0.001* 3 34.266 <0.001* Food 4 1.278 0.132 4 61.700 <0.001* Shark*Food 12 0.819 0.630 12 1.400 0.183 Within Subjects: Time Segment 3 2.077 0.136 3 19.964 <0.001* TS*Shark 9 1.991 0.041* 9 3.219 0.001* TS*Food 12 0.435 0.948 12 3.026 0.001* TS*Shark*Food 36 0.991 0.489 36 1.543 0.031* PC5 Predator Motion PC6 – Buccal Excursion df F P df F P Between Subjects: Shark 3 10.398 <0.001* 3 11.948 <0.001* Food 4 8.533 0.002* 4 2.85 0.075 Shark*Food 12 1.494 0.144 12 2.214 0.018* Within Subjects: Time Segment 3 1.394 0.256 3 6.202 0.001* TS*Shark 9 3.626 <0.001* 9 3.254 0.001* TS*Food 12 1.985 0.026* 12 1.417 0.158 TS*Shark*Food 36 1.761 0.007* 36 1.549 0.030*

PAGE 40

29 -4 -2 0 2 4 -4-2024 PC 1-General TimingPC 2-Cranial Timing/Displacement Shark 1 Shark 2 Shark 3 Shark 4 Figure 2: Comparison of principal component 1 (PC1) (general timing) and PC2 (cranial timing/displacement) scores among individuals irrespective of time for Triakis semifasciata Together the variability explained by these two com ponents is 34.4% of the overall variability in the kinematic data set. Polygons delimit the region of the plot occupied by bites from each individual. Note the high degree of intra-individual variability and the large degree of ov erlap among individuals, which is also a prevailing trend on the remaining PCs. On both components Shark 4 is significantly different (P=0.004 and 0.005, respectively) from all other individuals, and on PC2 Shark 3 is significantly different (P=0.041) from all other individuals. Differences over ontogeny regardless of indi vidual or food type occurred on PCs 1, 4, and 6 (Table 3), indicating si gnificant changes in timing an d excursion aspects of food capture with growth. The trend along each of these components was a gradual linear decrease in factor loading scores over th e first 6-9 months. This was followed by a quadratic increase in values over the following 3-6 months for general timing variables (PC 1) and buccal excursion variables (PC 6) Directional trends were not found over ontogeny within cranial timing variables (PC2) or the motion of the food or predator (PCs 3 and 5) (Table 3). Differences in indi vidual trajectories over ontogeny, however, were

PAGE 41

30 Figure 3: Comparison of average principal component (PC) scores for food types irrespective of individual and time for the components upon which overall differences between foods were found. Error bars indicate one standard error. Bars la beled with different letters differ at the =0.05 level. Note that bites on live shrimp are characterized by rapid, brief cranial movements and slow, short predator movement relative to other food types. identified on all PCs except PC1 (general timing variables). Thus within the general trends described over ontogeny there were di stinct inter-individual differences among cranial and hyoid timing, food and predat or motion, and excursion variables. The trajectory over ontogeny differed by food type on PCs 1, 2, 4, and 5 (Table 3). General timing variables (PC1) decreased linearly over ontogeny for large fish and live shrimp, but increased linearly for both si zes of squid and small fish (Fig. 4). Differences in cranial timing variables (PC2) we re attributable to f ood type rather than size, with bites on fish, squid, and shrimp e xhibiting different patterns over time. Bites on fish showed an initial increase in crania l timing variable values before decreasing during the last 6 months. Bites on squid exhibited the opposite trend, while cranial -1.00 -0.75 -0.50 -0.25 0.00 0.25 0.50 0.75 1.00Ave PC Factor Score Live Shrimp Small Fish Small Squid Large Fish Large SquidPC1-General Timing PC4-Hyoid Timing PC5-Predator Motion A B A A B A C D B

PAGE 42

31 timing variables from bites on shrimp decr eased nearly linearly over time. These differences due to food type may be obfuscated by size effects because fish pieces tended to be thicker than squid pieces in their sma llest dimension and may have required greater cranial elevation to accommodate their girt h. Differences in hyoid timing variables (PC4) were attributable to food size rather than type, w ith large food of both types eliciting more brief hyoid motions over ti me and small food eliciting prolonged hyoid motion during the first 3 months followed by mo re brief motions during the last 9 months (Fig. 5). Live shrimp, which were comparable in size to large food, exhibited the same pattern of deviation over time as large food. The pattern of cha nge over time was not significantly different for food types with re gard to buccal excursion variables (PC6), indicating that similar relative buccal expans ion occurred regardless of trends in the duration of hyoid motion. The trend over ti me for predator motion (PC5) differed between all types of dead f ood and live shrimp. For dead food, predator motion either increased slightly or remained constant ove r the first 9 months and then increased or remained constant during the last 3-month segment (Fig. 6). For live shrimp, predator motion decreased over the first 3 months befo re increasing markedly and then declining slightly. The number of atte mpts it took individuals to cap ture live shrimp remained constant at ~2.5 during the first 9 months and only began to decline when predator aspects of modality reached a peak (Fig. 7). As predator motion declined past this point so did the number of attempts needed to cap ture shrimp, eventually reaching an average of 1.5. On PCs 4-6 interactions between shar k and food type over time exist (Table 3), indicating that the pattern of deviation observed is influe nced partially by non-additive effects of individual and food type. While this weakens the ability to generalize about the

PAGE 43

32 trends described above for these factors in isolation, the majority of additive effects include small squid, the food type characterized by the greatest variability in food capture kinematics, and various individuals. -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC1 Score Large Fish, A -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC1 Score Large Squid, B -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC1 Score Small Fish, B -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC1 Score Small Squid, B -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC1 Score Live Shrimp, A Figure 4: Trends in general timing variables (PC1) over time by food type in Triakis semifasciata Model II regressions are based on all bites from all individuals but, for clarity, only daily averages are shown. Graphs labeled with differen t letters differ in the trend displayed over time at the =0.05 level. Over time bites on small food and large squid get longer and involve slower movements, while bites on large fish and live shrimp become more brief and faster. The regressi on equations for each food ty pe are: Large fish y=-0.001x+0.383; Large squid y=0.004x–0.485; Small fish y=0.002x–0.407; Small squid y=0.002x–0.203; Live Shrimp y=-0.003x+0.318.

PAGE 44

33 -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC4 Score Large Fish, A -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC4 Score Large Squid, A -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC4 Score Small Fish, B -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC4 Score Small Squid, B -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC4 Score Live Shrimp, A Figure 5: Trends in hyoid timing variables (PC4) over time by food type for Triakis semifasciata Best fit polynomial trend lines are based on all bites from all individuals but, for clarity, only daily averages are shown. Graphs labeled with different letters differ in the trend displayed over time at the =0.05 level. Over time bites on both types of small food showed a similar pattern of initial increase in duration before exhibiting a decrease. Bites on both types of large food and live shrimp generally exhibited a decrease in hyoid timing variables over time. Changes in hyoid timing variables over time are primarily attributable to food size as live shrimp were selected to be ~1 mouth width in carapace length throughout the study.

PAGE 45

34 -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC5 Score Large Fish, A -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC5 Score Large Squid, A -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC5 Score Small Fish, A -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC5 Score Small Squid, A -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC5 Score Live Shrimp, B Figure 6: Trends in predator motion (PC5) over time by food type for Triakis semifasciata Best fit polynomial trend lines are based on all bites from all individuals but, for clarity, only daily averages are shown. Graphs labeled with different letters differ in the trend displayed over time at the =0.05 level. The pattern of size-corrected forward motion of the predat or over time is similar for all dead food types. For live shrimp, bites during the first six months involve short, slow movement of the predator, whereas during the last six months the motion of the predator more closely resembled that observed for bites on dead food.

PAGE 46

35 -3 -2 -1 0 1 2 3 050100150200250300350400 Time (days)PC5 Score 0 1 2 3 4Number of Attempts Live Shrimp Capture Attempts Figure 7: The trend in predator motion (PC5) over time compared to the number of attempts required to capture live shrimp in Triakis semifasciata Best fit polynomial trend lines are based on all bites on live shrimp from all individuals but, for clarity, only daily av erages are shown. Note that an increase in capture success coincides with an increase in the forward velocity and distance covered by the predator during a strike. Scaling of Morphology The majority of external morphological variables scaled with negative allometry relative to total length, however pre-branchial length and pectoral fi n base to lower jaw length scaled positively (Fig. 8, Table 4), creating changes in overall head shape through ontogeny. Closed and open buccal volume s caled with negative allometry but the coefficient for open volume was larger, produc ing isometric growth in buccal reserve volume (Fig. 9, Table 5). All measures of mu scle mass and cross-sectional area scaled with positive allometry (Table 5), with th e mandibular abductor, the coracomandibularis, exhibiting the greatest scali ng coefficient. Combined jaw weight, however, scaled isometrically.

PAGE 47

36 Figure 8: Scaled drawing of ex ternal morphology of Triakis semifasciata at age 1 day (light gray lines) and age 365 days (black lines). Scaling is based upon the Model II regression line for all live individuals (N=4) and is indicative of growth trends exhibited by all individuals. Note that older individuals tend to have a relatively narrower, more anteriorly placed m outh and a shallower, broade r head profile from the rostrum to the position of the hyoid. Table 4: Parameters for log-log regressions against total length (cm) of external morphological variable data measured on live specimens of Triakis semifasciata Results shown are averages for individuals (N=4). The expected slope for is ometry is one in all cases. t0.05(1), 33=1.692. Variable Slope y-intercept r2 St Error t P Mouth Width 0.790 -0.86 0.99 2.8E-4 -875.39 <0.001* Mouth length 0.879 -1.53 0.99 4.1E-4 -299.27 <0.001* Pre-Oral Length 0.772 -0.87 0.88 7.9E-4 -293.61 <0.001* Pre-Orbital Length 0.935 -1.02 0.94 8.4E-4 -107.32 <0.001* Pre-Branchial Length 1.090 -0.77 0.96 9.2E-4 29.11 <0.001* Pre-Pectoral Length 0.981 -0.70 0.99 5.4E-4 -61.63 <0.001* Pectoral Fin Base to LJ 1.153 -1.14 0.99 5.8E-4 257.98 <0.001* Head Length 0.982 -0.63 0.97 7.7E-4 -120.61 <0.001* Branchial Length 1.011 -1.37 0.99 5.2E-4 6.63 <0.001* Head Depth at Hyoid 0.830 -0.87 0.97 6.6E-4 -364.21 <0.001* Head Width 0.983 -0.91 0.91 7.4E-4 -71.85 <0.001*

PAGE 48

37 Figure 9: Scaling of buccal volume relative to total length in Triakis semifasciata Measurements calculated from the weight of silicone casts obtained from dead specimens placed in resting and maximally expanded positions based on kinematic footage. Reserve volume is the difference between the open and closed position volume for a specimen. Inset shows lateral view of closed (top) and open (bottom) silicone casts from a representative individual. The expected slope for isometry is 3 in all cases and only Buccal Reserve Volume scaled isometrical ly. Model II regre ssion equations are: Buccal Volume Closed y=2.050x-1.453; Buccal Volume Open y=2.600x–2.043; Buccal Re serve Volume y=3.007x–3.003. Table 5: Parameters for log-log regressions against total length (cm) of morphological variable data taken from dead specimens only of Triakis semifasciata The expected slope for isom etry is given in the column labeled SlEXP. t0.05(1), 6=1.943. Variable SlEXP Slope y-intercept r2 St Error P Buccal Volume Closed (ml) 3 2. 050 -1.45 0.99 4. 6E-3 <0.001* Buccal Volume Open (ml) 3 2. 600 -2.04 0.99 5. 0E-3 <0.001* Buccal Reserve Volume (ml) 3 3. 007 -3.00 0.98 1. 0E-2 0.265 Combined Jaw Weight (g) 3 3.020 -4.68 0.96 1.5E-2 0.130 Quadratomandibularis Area (mm2) 2 2.108 -3.72 0.97 9.4E-3 <0.001* Coracohyoideus Area (mm2) 2 2.356 -4.83 0.96 1.0E-2 <0.001* Coracomandibularis Area (mm2) 2 2.535 -5.61 0.95 1.5E-2 <0.001* Coracoarcualis Area (mm2) 2 2.109 -3.95 0.98 7.6E-3 <0.001* Quadratomandibularis Weight (g) 3 3.340 -8.90 0.98 1.2E-2 <0.001* Coracohyoideus Weight (g) 3 3.069 -8.52 0.99 7.5E-3 <0.001* Coracomandibularis Weight (g) 3 4.273 -11.84 1.00 6.5E-3 <0.001* Coracoarcualis Weight (g) 3 3.310 -8.94 0.98 9.9E-3 <0.001*

PAGE 49

38 Discussion Differences in the feeding behavior of T. semifasciata exist irrespective of as well as over the course of ontogeny th at can be attributed to abso lute consequences of size, individual variability, and modulation in res ponse to food attributes. These differences occur concomitant with allometric changes in cranial morphology that have the potential to influence feeding performance by affec ting the generation of sub-ambient buccal pressure and bite force during feeding. Variability in feeding kinematics among individuals was great er than modulation among food types, except in hyoid timing variables. Individual variability was, in fact, the only factor that contributed to over all differences in food motion and buccal excursion variables. The predominance of indi vidual variability in feeding behavior is consistent with other studies of feeding in aquatic vertebrates including sharks (Shaffer and Lauder, 1985a; Summers, et al., 1998b; Wilga and Motta, 1998b). Among individuals, predator motion (i.e. the ram component of th e strike) exhibited the least variability and may reflect the similar degree of experience with particular food types possessed by individuals at a given age. Th e return of the hyoid to its resting position (hyoid retraction), however, exhibited the greate st variability. Successful entrainment of prey, especially elusive pre y, via suction feeding requires temporally coordinated, lowvariability expansion of the buccal cavity and depression of the hyoid (Muller and Osse, 1984; Wainwright, et al., 2001a; Svanback, et al., 2002). Once the food is secured within the buccal cavity, however, the recovery of the hyoid to its resting position does not appear to have a functionally relevant temporal component and occurs slowly with high variability.

PAGE 50

39 Individual differences in the developmen tal trajectory of feeding behavior over the course of early ontogeny have not previ ously been described for elasmobranchs, a fact attributable to the lack of studies examining feeding during the neonatal/hatchling stage (but see Ferry-Graham 1997; 1998a; 1998b). Subtle differences among individuals in morphology, growth trajectories, and experi ence can directly impact the development and execution of behavior (Bryan and Larki n, 1972; Morse, 1980), as well as establish a base of variation upon which natural selec tion can act. That individuals within a population utilize different resources and ar e not necessarily eco logical equivalent, especially during different life stages, ha s long been recognized (Wainwright, 1994; Bolnick, et al., 2003;). In T. semifasciata developmental differences in aspects of cranial and hyoid timing, predator and food mo tion, and buccal expansion exist among individuals that may translate to differences in prey capture performance resulting in differences in diet. The magnitude of behavi oral differences over ont ogeny is likely to be amplified in natural settings due to greater variability in pred ator experience and heterogeneity of available resources (Morse 1980). An ontogenetic dietary study that distinguishes differences in i ndividual exploitation of prey is needed before a definitive conclusion can be reached regarding the long-te rm ecological implications of individual dietary specialization in T. semifasciata The effect of growth on the timing, ve locity, and subsequent performance of animal motion has been reported for di verse behaviors (Ferry-Graham, 1998b; Verwaijen, et al., 2002; Elswort h, et al., 2003). The scaling model of feeding kinematics proposed by Richard and Wainwri ght (1995), based on largemouth bass Micropterus salmoides predicts an isometric increase in linear displacement and maximum linear

PAGE 51

40 velocity over ontogeny. This model also predic ts a scaling coefficient of 0.3 for angular displacement, angular velocity, and time to reach peak displacements based on decreases in per sarcomere contraction velocity over ontogeny (Richard and Wainwright, 1995). While this model and others that examin e relationships between organism size and behavioral response typically assume isom etric growth (Hill, 1950; O'Reilly, et al., 1993), allometric growth occurred in every quantified attribute of T. semifasciata cranial morphology (Table 4). Most notable were an anteriorly directed shift of the mouth, accompanied by a slight narrowing, and the deve lopment of a shallower head profile at the level of the hyoid. Despite this allometry, the scaling coefficient of the time to reach maximum gape for M. salmoides (0.31) (Richard and Wainwr ight, 1995), the nurse shark Ginglymostoma cirratum (0.33) (Robinson and Motta, 2002), and T. semifasciata (0.35) are very similar, suggesting that a physiologi cal limitation to sarcomere contraction rate may generally exist in fishes even under conditions of allometric growth. This relationship may be ecologically important b ecause growth, and associated changes in nutritional needs, can drive variation in resource use (Olson, 1996; Lima-Junior and Goitein, 2003). As the ability to generate sub-ambient buccal pr essures is partially dependent on the rate of buccal expansion (Muller and Osse, 1984; Svanback, et al., 2002), prey capture efficiency may be reduced by an increase in th e time needed to open the mouth, which could provoke ontogenetic dietary shifts. If prey capture performance is nega tively impacted by a decrease in mouth opening speed, it is possible th at behavioral modulation can compensate by affecting factors other than mouth opening speed that determine the magnitude of buccal pressure, such as acceleration, impulse, and magnitude of the volume entrained (Muller and Osse,

PAGE 52

41 1984; Carroll, et al., 2004). One notable de parture from the Richard and Wainwright (1995) model in T. semifasciata is the scaling coefficient of maximum gape distance over ontogeny (-0.555). While the absolu te size of the mouth increased over time, the size of the oral aperture during feeding was behaviora lly regulated to a relatively smaller size. Additionally, when feeding on live, elusive sh rimp all timing variab les decreased over the course of ontogeny and the forward motion of sh arks during a strike tended to increase. Thus, while the time needed to open th e mouth to a given gape size increased, presumably limiting suction generation capacity, individuals were able to behaviorally compensate by employing a smaller oral aper ture and expanding the buccal cavity more rapidly, presumably increasing suction. The e nd product of learning to regulate feeding behavior in this way was an increase in the ca pture efficiency of live, elusive shrimp over ontogeny (Fig. 7). In addition to behavi oral modulation, the positively allometric relationships between jaw muscul ature size, in concert with the isometric relationship for jaw weight, indicate the potenti al for relatively greater forces to act on relatively smaller skeletal elements over the course of ontogeny. Provided that change s in jaw opening and closing lever mechanics are not occurring (Barel, 1983; West neat, 1994), this pattern of growth would produce the capacity for more rapid jaw opening and more forceful jaw closing, thereby increasing suc tion feeding ability as well as maximum bite force (Huber and Motta, 2004). Both suc tion and biting are employed by T. semifasciata in the wild (Russo, 1975; Talent, 1976) and the enhancemen t of both behaviors may be facilitated by allometric growth of cranial components. The only behavioral factors not modula ted in response to differing food types were food motion and the extent of buccal ex pansion. The lack of differences in food

PAGE 53

42 motion among types could represent a hydrodyna mic constraint acting on the generation of effective suction that was not overcome by behavior (Ferry-Graham, et al., 2001a), especially considering buccal re serve volume scaled with isometry and the relative size of food items was kept constant. Both abso lute and relative strike distance were conservative throughout ontogeny, perhaps repr esenting the maximum distance that T. semifasciata can displace food using suction under these experimental conditions. The force generated by a flow of water that is capable of imparting momentum to a food particle decays as the cube of distance fr om its source (Muller, et al., 1982; Muller and Osse, 1984; Svanback, et al., 2002). Provided that the entire buccal reserve volume is being employed, which it appears to be based upon the constant relative degree of buccal expansion, a cubic increase in buccal reserv e volume would generate constant relative food displacement. The lack of apparent change in absolute food velocity and displacement could be a result of this exponential relations hip, which requires that relatively large differences in buccal pressu re be generated to produce differences in water velocity that would cr eate a noticeable effect on f ood (Wainwright, et al., 2001a). In order to determine how much force is being imparted to the food item by the flow generated during suction feeding it is necessary to describe the pattern of flow directly in front of and into the oral aperture. While this pattern has been described for various osteichthian fishes with terminal mouths (Lauder and Clark, 1 984; Muller and Osse, 1984; Ferry-Graham, et al., 2003), the cont ribution of a subterminal mouth and a prominent rostrum to flow pattern remains undescribed and could va ry significantly over ontogeny as a result of behavi oral and morphological changes such as those detected in T. semifasciata

PAGE 54

43 The detection of modulation in resp onse to food size and elusivity in T. semifasciata in this study is in contrast to th e findings of Ferry-Graham (1998a), who described juvenile T. semifasciata as typically employing a feeding modality intermediate between ram and suction that was unaffect ed by food type. The reason suggested by Ferry-Graham (1998a) for the apparent lack of modulation was that the offered live food, mud shrimp, did not behave elusively. Ru sso (1975) and Talent (1976) proposed that T. semifasciata was capable of performing suction-domi nated bites based on the presence of whole clams and burrowing worms in the gut of several individuals. The capacity to modulate feeding behavior and employ a grea ter ram component when feeding on elusive prey may play an important role in the early development of T. semifasciata by facilitating exploitation of a broader prey ba se. While ontogenetic dietary shifts have been noted for T. semifasciata larger than those used in this study (Russo, 1975; Talent, 1976; Kao, 2000), dietary data for neonates a nd YOY are lacking. Based on the versatile feeding repertoire demonstrated in this study, it is predicted th at neonates and YOY, while exhibiting individual dietary specialization and pr eferential usage of select resources, will behave as opportunistic generalists and a shift toward more elusive prey will occur with increases in predator experience and size. In conclusion, to elucidate changes in feeding performance, diet, and ecological niche through ontogeny, longitudi nal studies must begin by un derstanding va riability and modulation in the form-function complex of th e feeding apparatus. Over early ontogeny the feeding behavior and cranial morphology of T. semifasciata change considerably. Despite extensive inter-individual variabili ty in capture kinematics, behavioral modulation occurs in response to food attri butes, most notably elusivity and size, and

PAGE 55

44 consists of changes in the timings of crania l movements as well as the extent of overall predator motion during the strike. Live shrimp were captured using rapid buccal expansion and forward motion of the pred ator, while non-elusive food items were captured using slower motions. Though buccal reserve volume increases isometrically across ontogeny, movement of the food via in ertial suction appears to be limited by hydrodynamic constraints, as evidenced by a l ack of difference in food movement during the strike over ontogeny. Gr owth of cranial elements pr oduces a morphology tailored for both rapid buccal expansion, through positively allometric growth of jaw-abducting musculature and a relative narrowing of th e mouth, and forceful biting, via positively allometric growth of the jaw-adducting muscul ature. Utilization of a broad behavioral repertoire enhanced by morphological de velopment is hypothesized to permit exploitation of a broad prey base in T .semifasciata over early ontogeny.

PAGE 56

45 Chapter 3: The ontogeny of feeding behavior and cranial morphology in the whitespotted bambooshark Chiloscyllium plagiosum (Bennett 1830): a longitudinal perspective Abstract The morphological and behavi oral development of the f eeding apparatus over early ontogeny can profoundly affect the ability of an organism to obtain nourishment, ultimately impacting survival. The interpla y between morphology and behavior over the first year of life was studied in the whitespotted bambooshark Chiloscyllium plagiosum using high-speed videography and dissection. Ex ternally measured variables describing cranial growth, and jaw weight, scaled at or near isometry while jaw and hyoid musculature, especially the coracohy oideus, demonstrated considerable hypertrophication. The difference between th e volume of the buccal cavity when open and closed scaled with positive allometry while the time to reach maximum jaw and hyoid abduction showed no trend with size, i ndicating the capacity for more rapid and greater volumetric intake during feeding. In addition, the relative forward motion of the predator during a strike decreased over ontogeny and the feeding modality became more suction-dominated. Kinematic variables exhibi ted little variability and the primary aspect of food capture that was modulated in res ponse to food type was the forward motion of the predator. An increase in capture success was noted for live, elusive shrimp over ontogeny indicating that morphologi cal and behavioral changes have direct consequences for prey acquisition. Conservation of head shape coupled with a narrow behavioral

PAGE 57

46 repertoire is hypothesized to increase prey capture success in the wild over ontogeny as individuals become more proficient in the execution of a single, lo w-variability, suctiondominated capture behavior.

PAGE 58

47 Introduction A ‘specialist’ is, in its most inclusive se nse, an organism that exhibits limited breadth in one or more aspects of its nich e. Specializations may include behavioral, physiological, anatomical, and morphological as pects of organisms and have the capacity to both greatly restrict and expand the realized niche of a species (K iltie, 1982; Herrel, et al., 1999; Bohn and Amundsen, 2001; Sibbing and Nagelkerke, 2001). On the individual level, specialization can lead to variation in performance, the ability to perform a task, which can lead to differential survival a nd fitness among members of a population (Liem, 1980a; Ferry-Graham, et al., 2002). By virtue of their impact on fitness, performance differences are an underlying force driving adaptation over phylogenetic time (Lauder, 1981; Arnold, 1983; Norton, et al., 1995). Obtaining nutrients is crucial to survival dictating that the f eeding apparatus must be under selective pressure to perform at every stage of development in which nutrients are obtained from external sources (Ga lis, 1990; Galis, et al., 1994). However, aquatically feeding animals are subjected to additional functional constraints because of the density and viscosity of water (Lauder, 1980; Liem, 1990; 1993; Lauder and Shaffer, 1993). If feeding on unattached prey, aquatic organisms may either use the medium to entrain the prey (suction feed) or overcome the medium and engulf the prey (ram feed) (Liem, 1980b). Inertial suction feeding i nvolves the generation of sub-ambient buccal pressures, and maximizing the capacity to ge nerate this pressure has been linked to numerous behavioral and mor phological specializations (Lie m, 1993; Wainwright, et al., 2001a; Sanford and Wainwright, 2002; Svanb ack, et al., 2002; Carroll, et al., 2004). These functional specializations in turn ma y lead to a more stenophagous ( Sanderson,

PAGE 59

48 1991; Motta, et al., 1995a; Hernandez and Mott a, 1997) or catholic diet (Castro, 2000), the difference in part being attributable to the degree of versatility and performance capacity of the feeding a pparatus. Employing a special ized function (e.g. generating sub-ambient buccal pressure) to perform versa tile biological roles (e.g. capturing elusive prey in the water column and benthic prey in crevices) may lead to broader exploitation of the available food base th an if the function were empl oyed only for a single biological role. Qualitatively large differences in feeding behavior and morphology among aquatically feeding vertebrate species are easily recognized a nd have been widely studied (Muller and Osse, 1984; Motta, et al., 1995b; Liem and Summ ers, 2000; Wilga, et al., 2000), as have more subtle differences among populations of a singl e species (Cutwa and Turingan, 2000; Huskey and Turingan, 2001) The interaction between feeding morphology and behavior during ontogeny ha s received less atten tion, but variation among individuals over ontogeny is a key eco logical and evolutionary precept (Van Valen, 1965; Bolnick, et al., 2003). Early ontog eny is often a period of rapid learning and morphological development and provides an opportunity to study rapid changes in feeding behavior and morphology. If capture performance of the f eeding apparatus is enhanced by the generation of greater subambient buccal pressure, then directional development of the functionally relevant behavioral and morphological aspects that enhance suction is expected over ontogeny in species that primarily employ this mode of prey capture. In short, in dividuals should become more proficient (Sanderson, 1991) at suction feeding over ontogeny as their mo rphology and behavior are incrementally tailored to meet the requirements for the generation of grea ter suction forces.

PAGE 60

49 Sharks are exemplary organisms within which to investigate the development of feeding over early ontogeny because they feed readily in captivity, and possess a mechanically simple feeding apparatus co mposed of comparatively few structural elements. Furthermore, unlike teleost fi shes and amphibians that undergo drastic metamorphic changes in thei r feeding apparatus (Reilly, 1995; Hunt von Herbing, 2001; Hernandez, et al., 2002), sharks exhibit nonmetamorphic, continuous growth allowing the study of developmental traj ectories without these conf ounding effects. Additionally, much is known about the feeding of sub-adul t and adult sharks bot h in laboratory and natural settings (Cortes, 1999; Fouts and Nelson, 1999; Mo tta and Wilga, 2001; Huber and Motta, 2004; Motta, 2004) but little is known about the development of feeding morphology and behavior over early ontogeny (b ut see Ferry-Graham, 1997; 1998). To assess ontogenetic shifts in feeding and how they affect performance, we must first describe and understand the pa tterns of change in feed ing morphology and behavior. The whitespotted bambooshark Chiloscyllium plagiosum (Hemiscylliidae) was selected as the subject for this study because specimens are readily available from captive breeders, making it possible to study individuals from first feeding through the first year of life. This species was expected a priori to be an obligate suction feeder based on studies of the feeding appa ratus and behavior of orect olobid sharks (Wu, 1993; 1994; Motta and Wilga, 1999; Motta, et al., 2002; Robinson and Motta, 2002; Matott, et al., 2005). The goals of this study were: 1) to describe ontogenetic ch anges in the feeding morphology and behavior of Chiloscyllium plagiosum feeding on a variety of food types and sizes; 2) to determine the contribution of behavioral differences among individuals, among food types (modulation), and among food t ypes within an indi vidual, throughout

PAGE 61

50 ontogeny; 3) to compare the ontogeny of prey capture behavior and morphology to that of a behavioral and morphological generalist, the leopard shark Triakis semifasciata (Chapter 2); and 4) to identify form-function complexes that change over early ontogeny and may influence feeding performance among sh arks with respect to the generation of subambient buccal pressure. Methods and Materials Experimental Animals Chiloscyllium plagiosum is a common reef-dwelling, benthic species of the IndoWest Pacific and eastern coast of Southern Asia (Compagno, 1984b). Chiloscyllium plagiosum is oviparous, hatching at 12-20 cm TL (A. Cornish, pers. comm.; Tullis and Peterson, 2000). Adult females attain a maxi mum size of ~1 m TL, while males typically reach ~0.7 m. Chiloscyllium plagiosum is an opportunistic generalist that feeds primarily on benthic invertebrates and occasionally fi sh (A. Cornish, pers. comm.). Wu (1993) determined that adult C. plagiosum primarily use suction to capture prey. Filming Techniques Five C. plagiosum (average TL 15.76 cm at hatching) were hatched at SeaWorld, Orlando, Florida and raised for 52 weeks, atta ining an average TL of 43.64 cm. Animals were maintained in a 340-liter communal tank at 26 1C and 332 ppt salinity. During experimental sessions, individuals were is olated within a 100-l iter subsection of the holding tank so their behavior would not be influenced by interactions with other individuals. Animals were fed a maintena nce diet consisting of 3-4% of their body

PAGE 62

51 weight in frozen krill Euphausia superba frozen clam Mercenaria mercenaria, and live and frozen grass shrimp Palaemonetes pugio ad libitum twice per week. This feeding frequency was maintained throughout the study but for feedings im mediately preceding experimental sessions the ration was cut to 2% to encourage active feeding during filming. Experimental sessions began within two weeks of hatching and included firstfeeding in three of the five individuals. To investigate directional changes of f eeding behavior in response to food size, type, and elusivity (modulat ion), five food types were offered during experimental sessions: 1) chopped krill scal ed to one half mouth width (MW); 2) chopped krill scaled to MW; 3) chopped clam scaled to one half MW; 4) chopped clam (shelled) scaled to MW; and 5) live shrimp scaled to MW in carapace length. Filming sessions occurred weekly over the entire one-year experimental period. During filming sessions food items were presented in a haphazard order until the individual approached satiation, as evidenced by a decrease in feeding activity. Though a filming session often comprised as many as ten capture sequences for a given indi vidual, only the first fi ve were considered for analysis to avoid confounding effect s of satiation (Sass and Motta, 2002). The methods for obtaining kinematic and morphological data are described in detail elsewhere (Chapter 2). In brief, a Redlake PCI 1000 high-speed digital camera (Redlake, San Diego, CA, USA) was placed perpendicular to the aquarium providing both a lateral view and, via a mirror beneath the shark, a vent ral view. Only orthogonal views were retained for analysis. Recordings were made at 250 fps and illumination was provided by two, 500-Watt quartz-ha logen lights. Animals we re trained to feed under illumination and were allowed a 20-minute acclimation period prior to each feeding

PAGE 63

52 session. The 52-week experimental time period was broken into four even segments. A total of five capture sequences per food type (N=5) per individual (N=5) per time segment (N=4) were recorded, for a total of 500 sequences. Sequences were not obtained weekly for each food type/indi vidual combination due to satiation. Kinematic data were obtained using Redlake MotionScope PCI so ftware version 2.21.1 (Redlake, San Diego, CA, USA) and SigmaScan Pro version 4 (SPSS Inc.) The variables measured parallel those of Chapter 2. From the onset of ma ndible depression (time 0 ms), the following kinematic variables were quantified: 1) stri ke distance (cm) at time 0 ms; 2) maximum gape (cm); 3) time to maximum gape (ms); 3) maximum cranial elevation angle (degrees); 4) time to maximum cranial eleva tion angle (ms); 5) time to onset of cranial elevation (ms); 6) time to offset of cranial el evation (ms); 7) durati on of cranial elevation (ms); 8) time to onset of hyoid depression (ms); 9) maximum hyoid depression (cm); 10) time to maximum hyoid depressi on (ms); 11) time to hyoid re traction (ms); 12) duration of hyoid depression (ms); 13) total strike dur ation (ms); 14) duration of food movement (ms); 15) distance moved by the food (c m); 16) velocity of the food (cm s-1); 17) distance moved by the predator (cm) during the durati on of food movement; and 18) velocity of the predator (cm s-1). No measures were made of the extent or timing of upper jaw protrusion because protrusion was often obscure d by movement of the labial cartilages. Variables (15) and (17) were used to calcu late the ram-suction i ndex (RSI) (Norton and Brainerd, 1993). The RSI is calculated as (DPREDATOR–DPREY)/(DPREDATOR+DPREY), where D is the distance moved by either the predat or or prey, and indi cates the relative contribution of forward motion of the predator and motion of the prey to a given capture event. An RSI value of 1 indicates a purely ram-based bite and a value of –1 indicates a

PAGE 64

53 purely suction-based bite. The number of fa iled attempts occurring prior to food capture was recorded for each capture event throughout the study. Morphological Measurements To chronicle ontogenetic changes in morphology, frames from the digital recordings were used to measure several external mo rphological variables. The definitions of these variables follow Compa gno (1984b) and consisted of: 1) total length; 2) mouth width; 3) mouth length; 4) preoral length; 5) preorb ital length; 6) prebranchial length; 7) branchial length; 8) prepectoral length; 9) anterior pectoral fin base to lower jaw; 10) head length; 11) head width; and 12) head depth at the location of the hyoid. To verify that these measurements were accurate, individuals were removed from the tank every 4 weeks and the same measurements taken. Measurement error was always less that 2.3%. To acquire measurements of muscle mass, muscle cross-sectional area, and buccal volume that could not be obtained from liv e individuals, fresh-dead specimens (N=9) spanning the total length range of live specimens used in th is study were dissected. The morphological variables measured on live sharks were also measured for dead specimens. Wet muscle mass and cross-sectional area we re obtained by excising muscles involved in feeding, specifically the coracomandibular is, coracohyoideus, co racoarcualis, and quadratomandibularis (Motta et al., 1991; Wu, 1994; Mo tta and Wilga, 1995; 1999). SigmaScan Pro version 4 (SPSS Inc.) was used to determine anatomical cross-sectional area from digital photographs taken with a Nikon Coolpix 4300. The palatoquadrate (upper jaw) and Meckel’s carti lage (lower jaw) were also excised and their combined

PAGE 65

54 weight recorded. These skelet al elements were considered because their growth could influence the velocity and force of feeding motions. Measures of buccal volume were obtaine d by injecting silicone into the buccal cavity of each dead specimen and allowing it to cure for 36 hours before removal and weighing. Using the mass-density of silicone reported by Cook (1996) (0.06 g ml-1), this weight was converted into a volume. Casts we re made of the buccal cavity in a resting position and in a maximally expanded position based on kinematic footage. The difference between these two volumes, the bucca l reserve volume, was used to indicate the maximum potential change in the volume of the buccal cavity du ring a capture event. Statistical Analyses All kinematic data were log-transformed and checked for normality and homogeneity of variance via the Kolm ogorov-Smirnov and Levene Median tests, respectively, using SigmaStat Pro version 3.1 (S PSS Inc.). As several variables exhibited skewed distributions, a Spearman Rank Order Co rrelation test was performed to establish the relationship between each variable and to tal length. Several variables were highly correlated (P<0.01) with total length, so Mode l II linear regressions were performed to describe the nature of these relationships. Model II regressions we re appropriate because both variables used in these comparisons c ontained error (McGowa n, 1988). Regressions were performed for all individuals combined then for each individual separately to compare regression coefficients among individuals. Differences among individuals were assessed using a modified St udent’s t-test (Zar, 1999).

PAGE 66

55 To standardize statis tical handling of all data for furt her analysis, all variables that lacked a relationship to total length were re gressed against total length using Model II linear regressions and the studentized re siduals obtained (Quinn and Keough, 2002). Size-corrected, studentized residuals for all kinematic variables were then used in a correlation matrix-based Principal Compone nts Analysis (PCA) that reduced the expansive kinematic data set to a few, ort hogonally oriented composite variables. An Equamax rotation was used because it produced higher loadings than any other rotation, enhancing data interpretation. Variables lo ading above an absolute value of 0.5 were considered to contribute heavily to the va riability within the respective principal component (PC). Principal components with an eigenvalue greater than 1.0 were retained for further analysis. Factor loading scores for each capture seque nce on each principal component were then used in a two-way, mixed-model repeated measures MANOVA to identify differences in multivariate space over the repeated measure time. Individual was a random factor and food type was a fixed factor tested over th e interaction term. Significance was assessed using Pillai’s trac e. To further investigate differences identified by the RM MANOVA, a two-wa y, mixed-model RM ANOVA was performed for each principal component separately us ing individual as a random factor and food type as a fixed factor tested over the inte raction term. Statistical significance was evaluated using Fisher’s LSD with an level of 0.05. To determ ine the order of the bestfit equation describing trends over the repeat ed measure time, single degree of freedom polynomial contrasts were utilized. Bonferroni corrections were not applied to any tests due to their tendency to increase the rate of type II errors (Cab in and Mitchell, 2000;

PAGE 67

56 Moran, 2003). The PCA, MANOVA, and RM ANOVA tests were performed using Systat 11 (SPSS Inc.). Morphological variables measured on liv ing specimens were log-transformed and regressed against the logarithm of total le ngth using Model II linear regressions to investigate changes in the rela tive dimensions of these variables over time. A slope of one for these regressions indicated isometric growth. Significant deviations from a slope of one were tested using Student’s t (Zar 1999) and indicated allometry for the given feature. Regressions were performed fo r each individual separately and then 95% confidence intervals were determined for all live individuals comb ined. For variables measured on both live and dead specimens, data taken from dead specimens was compared against the confidence intervals for the live individuals to determine if it could have been drawn from the same population. Comparing live and dead scaling data via this method allowed data colle cted exclusively on dead spec imens to be applied to live specimens within the appropriate size ra nge. Morphological variables measured exclusively on dead specimens were log-tran sformed and regressed against the logarithm of total length using Model II linear regressions The expected slope of these regressions depended on the dimensionality of the variab le being considered, with a slope of one expected for linear variables, a slope of tw o expected for planar variables (e.g. muscle cross section), and a slope of three expected for cubic variables (e.g. muscle mass, buccal volume).

PAGE 68

57 Results Individuals generally re sted on the bottom or swam slowly around the filming chamber until food was introduced. The shark would then initiate searching behavior, which was characterized by rapid movement along the bottom in a sweeping pattern that covered most of the filming chamber. F ood capture kinematics began with lower jaw depression or cranial elevation and progressed in a posteriorly directed fashion with the hyoid beginning to depress shortly after the onset of jaw opening and reaching its maximum excursion well after the time of maximum gape (Fig. 10). The temporal sequence of kinematic events was conservati ve across ontogeny and among food types. Capture success was generally high with al l dead food types, averaging 1.10.3 attempt per capture and changing little over ontogeny, however the numb er of attempts needed to capture live shrimp declined steadily from 2.50.8 during th e first time segment to 1.60.5 during the last. Figure 10: Composite photographic series of a representative food capture sequence for Chiloscyllium plagiosum The point on the food that is farthest from the shark’s mouth is indicated by a white dot. This was the point used in calculating the RSI. Times for this capture are shown in the upper right corner of each field. In chronological order the events show n are onset of lower jaw depression, maximum gape, food capture, maximum cranial elevation, maximum hyoid depression, and jaw closure on the food.

PAGE 69

58 Scaling of Kinematics When individuals were combined, the majo rity of kinematic variables scaled with positive coefficients over ontogeny indicating that slower, more extensive motions accompanied growth (Table 6). The only tw o variables that decreased over ontogeny were the time to onset of hyoid depression and RSI, indicating that as sharks grew they began depressing their hyoid earlier in the strike and their overall feeding modality became more suction-dominated. The distance moved by the predator and the velocity of the predator during the strike did not change with size, while the distance moved by the food increased slightly (Table 6). Taken t ogether, these relationshi ps indicate that the decrease in RSI over ontogeny was caused excl usively by greater motion of the food due to suction. When individuals were analyzed separately, differences were detected in the extent of the scaling coefficient for many variables but rarely the sign (Table 6), indicating that while substantial inter-indivi dual variability exists in the scaling of kinematics over ontogeny the general tre nds presented above are representative. Kinematic Trends The PCA reduced the set of kinematic vari ables into six principal components that accounted for 72.80% of the overall variability in the data set. Five kinematic variables did not load heavily on any of the PCs (Table 7). Three of these related to cranial elevation during the period from onset to maxi mum, indicating that the timing and extent of cranial elevation was char acterized by relatively little variation. Timing variables loaded on PCs 1, 3, and 4, with general tim ing variables loadi ng on PC1, variables describing depression of the cranium from maximum excursion loading on PC3, and

PAGE 70

59 variables describing the recovery of the hyoi d to its resting position loading on PC4. Variables associated with forward motion of the predator over the course of the strike loaded on PC2 while variables describing th e motion of the food loaded on PC5. The RSI, even though it is an index based on motion of both the predator and food, only loaded heavily on PC2. This indicates that va riability in aspects of predator motion was greater than for aspects of food motion. Las tly, variables describing the extent of buccal expansion, as indicated by maximum gape and maximum hyoid depression, loaded on PC6 (Table 7). Table 6: Regression parameters for kinematic variables versus total length (cm) for Chiloscyllium plagiosum Results shown are for all individuals combined (N=5). Where differences among individuals were detected using a modified Student’s t-test, the magnitude of these differences is indicated. Max=maximum; Depress=depression; Elev=elevation. Individuals Combined Individuals Separated Variable Slope y-int StErr P Range of Slopes Max Gape 0.90 -1.40 0.04 <0.001* 0.79* 1.11* Time to Max Gape 0.21 1.12 0.07 0.005* 0.02** 0.37** Max Cranial Elev Angle ---0.334 -0.93** 0.97** Time to Max Cranial Elev Angle 0.29 1.17 0.07 <0.001* 0.03* 0.48* Time to Onset of Cranial Elev ---0.345 -0.40 0.03 Time to Offset of Cranial Elev 0.75 0.79 0.08 <0.001* 0.43** 1.15** Duration of Cranial Elev 0.92 0.47 0.10 <0.001* 0.58* 1.50* Time to Onset of Hyoid Depress -0.32 1.64 0.15 0.032* -0.54 -0.31 Max Hyoid Depress 0.97 -1.77 0.08 <0.001* 0.54** 1.30** Tim to Max Hyoid Depress 0.37 1.10 0.06 <0.001* 0.13* 0.62* Time to Hyoid Retraction 0.26 1.65 0.05 <0.001* 0.08 0.47 Duration of Hyoid Depress 0.38 1.42 0.06 <0.001* 0.19* 0.55* Duration of Food Movement 0.30 0.96 0.09 <0.001* 0.03* 0.49* Total Strike Duration ---0.118 -0.48** 0.11** Distance Moved by Food 0.32 -0.58 0.08 <0.001* 0.13 0.41 Velocity of Food ---0.797 -0.12* 0.47* Distance Moved by Predator ---0.878 -0.27** 0.23** Velocity of Predator ---0.078 -0.53** -0.09** RSI -0.22 0.07 0.11 0.039* -0.51* -0.07* Strike Distance 0.87 -1.65 0.13 <0.001* 0.55** 1.55**

PAGE 71

60Table 7: Principal component loadings of kinematic variables associated with capture sequences in Chiloscyllium plagiosum Bold face values indicate variables dete rmined to load heavily on the respective component (loading scores>|0.5 ) (N=5). Together the six components explain 72.80% of the overall variability in the data set. For clarity, all lo adings<|0.25| are replaced by 0. Max=maximum; Depress=depression; Elev=elevation. Variable PC 1 PC 2 PC 3 PC 4 PC 5 PC 6 Duration of Prey Movement 0.844 0 0 0 0 0 Total Strike Duration 0.626 0.309 0 0 0 0 Time to Max Gape 0.623 0 0 0 0 0 Velocity of Food -0.528 0 0 0 -0.770 0 Time to Max Hyoid Depress 0.523 0 0 0.260 0 0 Distance Moved by Predator 0 -0.966 0 0 0 0 RSI 0 -0.914 0 0 0.382 0 Velocity of Predator -0.368 -0.870 0 0 0 0 Duration of Cranial Elev 0 0 0.949 0 0 0 Time to Offset of Cranial Elev 0 0 0.910 0 0 0 Duration of Hyoid Depress 0 0 0 0.971 0 0 Time to Hyoid Retraction 0 0 0 0.953 0 0 Distance Moved by Food 0 0 0 0 -0.949 0 Max Gape 0 0 0 0 0 0.754 Max Hyoid Depress -0.488 0 0 0 0 0.561 Time to Onset of Cranial Elev 0 0 0 0 0 0 Time to Onset of Hyoid Depress 0 0 0 0 0 0 Time to Max Cranial Elev Angle 0.356 0 0.491 0 0 0 Max Cranial Elev Angle -0.298 0 0.400 0 0 0.376 Strike Distance 0 -0.303 0 0 -0.306 0.465 Eigenvalue 4.72 2.65 2.27 1.89 1.63 1.08 Percent Variance Explained 19.99 14.00 11.56 10.55 10.15 6.56 The RM MANOVA conducted simultaneousl y on all six principal components indicated that differences existed over ontogeny (Pillai Trace df =3; F=5.99; P=0.001) and among individuals over ontogeny (df=12; F=3.61; P<0.001), but not among food types over ontogeny (df=12; F=1.35; P=0.190) or among food types for a given individual over ontogeny (df=48; F=1.05; P=0.338). Despite a hi gh degree of overlap, when each PC was examined with a separate RM ANOVA fo r differences among individuals they were detected regardless of ontogeny on all co mponents (Fig. 11) and over ontogeny on all components except PC2 (predator motion) (T able 8), indicating the presence of

PAGE 72

61 substantial inter-individual differences in kinematics. Differences among captures on various food types regardless of ontogeny ex isted only on PCs 1 (general timing) and 2 (predator motion), with strikes on live, small, and large prey separating on both PCs (Fig. 12). Interactions between the factors shark and food type were only found on PCs 1 (general timing) and 4 (hyoid timing). Di fferences over ontogeny among captures of various food types were only detected on PC s 2 (predator motion) (Fig. 13) and 5 (prey motion), indicating that modula tion in response to food attrib utes did not occur in the timing or extent of cranial motion over ontoge ny. Effect interactions among sharks within food types over ontogeny existed onl y on PCs 5 (prey motion) and 6 (buccal expansion) indicating complex interactions over time on th ese two components that are due to the combined influen ce of individual and food type. Scaling of Morphology Morphometric variables tended to scale w ith isometry or very slight allometry (Table 9). The greatest allometric scali ng coefficient was quantified for mouth width (1.05) and suggests that the widt h of the mouth at an age of 1 year is ~12% greater than would be predicted assuming isometry. Th e scaling coefficient exhibited by other morphological attributes typica lly produced an estimate varying by less than 3% from that predicted by isometry. The apparent ly high level of sta tistical significance associated with these deviations is attributab le to the low variability generated by taking measurements on a weekly basis. Measuremen ts taken on dead specimens fell within the 95% confidence intervals of those taken from live specimens and were dominated by isometric or very slightly a llometric patterns of growth.

PAGE 73

62 Figure 11: Comparison of principal component 1 (PC1) (general timing) and PC2 (predator motion) scores among individuals irrespective of time for Chiloscyllium plagiosum Together the variability explained by these two components is 33.99% of the over all variability in the kinematic data set. Polygons delimit the region of the plot occupied by bites fr om each individual. Note the high degree of intraindividual variability and the large degree of overlap am ong individuals, which is also a prevailing trend on the remaining PCs. On PC1 sharks 1 and 4 differ from the remaining sharks ( =0.05) and on PC2 sharks 4 and 5 differ ( =0.05). -3 -2 -1 0 1 2 3 -3-2-10123 PC 1-General TimingPC 2-Predator Motion Shark 1 Shark 2 Shark 3 Shark 4 Shark 5

PAGE 74

63 Table 8: Results of RM ANOVAs performed separately on each principal component from a PCA of prey capture kinematics in Chiloscyllium plagiosum (N=5). df=Degrees of Fr eedom; TS=Time Segment. For all between subjects comparisons the error degrees of freedom are 100, while for all within subjects comparisons the error degr ees of freedom are 300. PC1 General Timing PC2 Predator Motion df F P df F P Between Subjects: Shark 4 20.562 <0.001* 4 3.842 0.006* Food 4 6.203 0.004* 4 13.952 <0.001* Shark*Food 16 3.161 <0.001* 16 1.143 0.327 Within Subjects: Time Segment 3 8.421 <0.001* 3 5.409 0.003* TS*Shark 12 4.705 <0.001* 12 1.385 0.172 TS*Food 12 1.649 0.072 12 2.151 0.014* TS*Shark*Food 48 1.153 0.238 48 1.363 0.065 PC3 Cranial Timing PC4 Hyoid Timing df F P df F P Between Subjects: Shark 4 3.771 0.007* 4 4.974 0.001* Food 4 2.457 0.089 4 1.222 0.386 Shark*Food 16 1.024 0.438 16 2.173 0.010* Within Subjects: Time Segment 3 1.836 0.155 3 2.536 0.076 TS*Shark 12 3.184 <0.001* 12 1.966 0.027* TS*Food 12 1.362 0.193 12 0.663 0.786 TS*Shark*Food 48 1.359 0.073 48 1.884 0.147 PC5 Food Motion PC6 Buccal Excursion df F P df F P Between Subjects: Shark 4 10.431 <0.001* 4 10.740 <0.001* Food 4 2.530 0.082 4 2.881 0.061 Shark*Food 16 1.643 0.071 16 2.881 0.061 Within Subjects: Time Segment 3 2.462 0.070 3 0.299 0.887 TS*Shark 12 4.161 <0.001* 12 3.994 <0.001* TS*Food 12 3.803 <0.001* 12 1.522 0.115 TS*Shark*Food 48 1.581 0.012* 48 2.257 <0.001*

PAGE 75

64 Figure 12: Comparison of principal component 1 (PC1) (general timing) and PC2 (predator motion) scores for large, small, and live food irrespective of time for Chiloscyllium plagiosum Together the variability explained by these two components is 33.99% of the overall variability in the kinematic data set. Polygons delimit the region of the plot occupied by bites for each food type and indicate significant difference at the =0.05 level. Note the high degree of overlap among food types, which is also a prevailing trend on the remaining PCs. -3 -2 -1 0 1 2 3 -3-2-10123 PC 1-General TimingPC 2-Predator Motion Large Small Live

PAGE 76

65 -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC2 Score Large Clam, A -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC2 Score Large Krill, B -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC2 Score Small Clam, A -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC2 Score Small Krill, A -2.5 -1.5 -0.5 0.5 1.5 2.5 0100200300400 Time (days)PC2 Score Live Shrimp, B Figure 13: Trends in predator motion (PC2) over time by food type in Chiloscyllium plagiosum Model II regressions are based on all bites from all individuals but, for clarity, only daily averages are shown. Graphs labeled with differen t letters differ in the trend displayed over time at the =0.05 level. Over time bites on small (one half mouth width) food and large (mouth width) clam exhibit a slight trend toward greater predator motion, while bites on large (mouth width) krill and live shrimp diverge from this trend at ~250 days and are characterized by relatively less predator motion. Regression lines labeled with different letters differ at the =0.05 level.

PAGE 77

66Table 9: Parameters for log-log regressions against total length (cm) of external morphological variable data measured on live specimens of Chiloscyllium plagiosum Results shown are averages for individuals (N=5). The expected slope for is ometry is one in all cases. t0.05(1), 40=1.684. Variable Slope y-intercept r2 St Error P Mouth Width 1.05 -1.48 1.00 6.1E-4 <0.001* Mouth Length 1.02 -2.46 0.95 2.1E-3 <0.001* Pre-Oral Length 0.99 -1.35 0.99 7.8E-4 <0.001* Pre-Orbital Length 1.00 -1.13 0.99 9.3E-4 0.350 Pre-Branchial Length 1.01 -0.86 0.99 1.1E-3 <0.001* Pre-Pectoral Length 0.99 -0.78 0.99 1.0E-3 <0.001* Pectoral Fin Base to LJ 1.00 -0.95 0.99 8.0E-4 0.309 Head Length 1.01 -0.73 0.99 9.4E-4 <0.001* Branchial Length 1.00 -1.33 0.97 1.6E-3 <0.001* Head Depth at Hyoid 0.99 -1.14 0.99 9.8E-4 <0.001* Head Width 0.98 -0.94 0.99 9.3E-4 <0.001* The growth trajectories calculated for the masses and cross-sectional areas of muscles, as well as buccal volume, were domin ated by substantial allometry (Table 10). The cross-sectional ar ea of all of the jaw abducting a nd adducting muscles scaled with positive allometry. The scaling coefficient for the cross-sectional area of the coracohyoideus was especially remarkable (3.04), reflecting a difference of ~51% from the area predicted by isometry. Paradoxically, the weight of both th e coracohyoideus and the coracomandibularis scaled with negative al lometry (Table 10), which may be a factor of unquantified changes in ot her aspects of muscle size (e.g. length) that occur over ontogeny. The combined weight of the pala toquadrate and Meckel’s cartilage scaled with positive allometry but to a lesser extent than the cross-sectional area of the muscles acting on these elements. Th e volume of the buccal cavity when closed scaled with negative allometry while the volume of the buc cal cavity when open sc aled with slight positive allometry. This discrepancy lead to a substantial positive allometric scaling coefficient for the buccal reserve volume indicating that as C. plagiosum grows its relative buccal reserve volume substantially increases (Fig. 14).

PAGE 78

67Table 10: Parameters for log-log regressions against total length (cm) of morphological variable data taken from dead specimens only of Chiloscyllium plagiosum The expected slope for isometry is given in the column labeled SlEXP. t0.05(1), 8=1.860. Variable SlEXP Slope y-intercept r2 St Error P Buccal Volume Closed (ml) 3 2.75 -2.69 0.99 1.2E-2 <0.001* Buccal Volume Open (ml) 3 3. 04 -2.79 0.99 1. 2E-2 0.003* Buccal Reserve Volume (ml) 3 3.31 -3.46 0.99 1.2E-2 <0.001* Combined Jaw Weight (g) 3 3.08 -4.54 0.96 2.9E-2 0.018* Quadratomandibularis Area (mm2) 2 2.44 -4.17 0.98 1.5E-2 <0.001* Coracohyoideus Area (mm2) 2 3.04 -5.36 0.99 1.7E-2 <0.001* Coracomandibularis Area (mm2) 2 2.36 -4.52 0.99 1.3E-2 <0.001* Coracoarcualis Area (mm2) 2 2.39 -4.13 0.97 2.0E-2 <0.001* Quadratomandibularis Weight (g) 3 3.13 -5.21 0.98 2.0E-2 <0.001* Coracohyoideus Weight (g) 3 2.55 -4.51 0.99 9.0E-3 <0.001* Coracomandibularis Weight (g) 3 2.53 -4.41 0.97 2.0E-2 <0.001* Coracoarcualis Weight (g) 3 3.13 -5.16 0.97 2.4E-2 <0.001* Figure 14: Scaling of buccal volume relative to total length in Chiloscyllium plagiosum Measurements calculated from the weight of silicone casts obtained from dead specimens placed in resting and maximally expanded positions based on kinematic footage. Reserve volume is the difference between the open and closed position volume for a specimen. Inset shows lateral view of closed (top) and open (bottom) silicone casts from a representative individual. The expected slope for isometry is 3 in all cases. Model II regression equations are: Buccal Volume Closed y=2.75x–2.69; Buccal Volu me Open y=3.04x–2.79; Buccal Reserve Volu me y=3.31x–3.46.

PAGE 79

68 Discussion Changes in morphology and behavior asso ciated with the feeding apparatus of Chiloscyllium plagiosum over early ontogeny appear to enhance the capacity of this species to generate subambient pressures in the buccal cavity during feeding, facilitating inertial suction feeding. Though most morphol ogical aspects of the head scale with isometry, the musculature associated w ith jaw and hyoid abduction, and jaw adduction, hypertrophies with growth. Additionally, buccal reserve volume grows allometrically at a rate greater than the positive allometric increase in the time to maximum gape and maximum hyoid depression. Considered togeth er with a decrease in overall predator motion, the combined effect of these relations hips is a positively al lometric increase in the relative contribution of suction through ontogeny in C. plagiosum Ferry-Graham (1998b) described diffe rences in morphology and kinematics between hatchling and juvenile swellsharks Cephaloscyllium ventriosum and found isometric growth but differences in RSI that were attributed to in creased predator motion masking the suction component of the strike in hatchlings. A mathematical model of buccal volume changes during feeding in C. ventriosum suggested isometric growth of the buccal cavity, leading to an isometric pred iction for suction generation in this species (Ferry-Graham, 1998b). Isometric growth of the buccal cavity over ontogeny was also noted in the cottid fish Clinocottus analis but differential motion of the body relative to the premaxilla during strikes lead to an incr ease in the apparent degree of suction over ontogeny, as quantified by the RSI (Cook, 1996) To verify whether the apparent increase in suction over ontogeny in C. plagiosum is mediated by an allometric increase

PAGE 80

69 in suction generation or simply due to a decr ease in predator moti on with size, direct measurements of suction pressure are underway. Within the broader trend of directional development toward in creasing utilization of suction, limited modulation in resp onse to food type was observed in C. plagiosum Differences in general timing variables (PC1) detected among small, large, and live food bites regardless of ontogeny are in large part due to divergence in the velocity and duration of food movement. Because these va riables are determined by the interaction between the food and the water flow generated by the shark, which is partially dictated by physical attributes of the food, the apparent modulation should be in terpreted cautiously. If the force exerted by the entraining flow of water generated by th e shark is the same among bites on different food types, the respons e of different food types to this flow could vary in a consistent manner that woul d be statistically dete cted as modulation. Assessment of the response of various food type s to controlled suction forces is necessary to firmly assert that C. plagiosum is modulating kinematic timings in response to food type. When food presentation is via tongs restricting food mo tion, the nurse shark Ginglymostoma cirratum does not modulate its feeding kinematics in response to food type (Matott et al., 2005). The lack of differences in the pattern of response over ontogeny in C. plagiosum argues for stereotypy of f eeding strike kinematics but modulation of overall predator motion in response to food type (Table 8). Despite the conservative kinematic patte rns described above, differences were detected among individuals bot h irrespective of, and over th e course of, ontogeny. The occurrence of considerable variability in feeding kinematics among and between individual fish during discrete life stages is nearly universal (W ainwright and Lauder,

PAGE 81

70 1986; Sanderson, 1990; Cook, 1996; Ferry-Graha m, 1998a), even in species like G. cirratum that are obligate suction feeders and ex hibit considerable stereotypy (Motta, et al., 2002). The general lack of longitudinal studies that fo llow target individuals across ontogeny has led to a paucity of informati on regarding intra-specific variability in developmental trajectories w ith regard to feeding morphology and behavior (but see Chapter 2). As disparities among indivi duals can influence performance, niche utilization, and potentially survival (V an Valen, 1965; Bolnick, et al., 2003), the application of laboratory studies of feeding to natural settings is compromised if interindividual variability is not considered. The magnitude of developmental differences over ontogeny are likely to be amplified in natu ral settings due to greater variability in predator experience and heterogeneity of available resources (Morse, 1980), compounding the potential impact of intra-indi vidual effects on niche diversification. Intra-individual developmental differences could serve as a fe rtile source of variability, especially in species that exhibit conservative growth and narrow behavioral repertoires, permitting response to variations in the environment over phylogeny (Liem, 1980a). The pattern of morphological and behavi oral development described here for C. plagiosum is in stark contrast to that reported in Chapte r 2 for the leopard shark Triakis semifasciata The growth of the head of T. semifasciata is highly variable and strongly allometric, producing marked changes in head shape over ontogeny. This is accompanied by hypertrophy of the jaw and hyoid abducting musc ulature, but not to the extent seen here for C. plagiosum especially in the cas e of the coracohyoideu s. Additionally, BRV increases isometrically in T. semifasciata which is both a dietary and behavioral generalist (Talent, 1976; Compagno, 1984a; Kao, 2000). The feeding kinematics of T.

PAGE 82

71 semifasciata are characterized by extensive vari ability (Ferry-Graham, 1998a) and modulation of timing and modality variables in response to prey size and elusivity. Triakis semifasciata apparently utilizes a diverse, ramdominated feeding repertoire over early ontogeny to capture functionally and ta xonomically diverse pr ey in the wild. By contrast, C. plagiosum employs a narrow range of behavior characterized by limited variability and modulation over th e range of food types tested here. Dietary information for young C. plagiosum is lacking, but the ab ility to capture elusive shrimp is clearly demonstrated here. Capture of diverse pr ey items (Castro, 2000), apparently using a single, highly stereotyped feed ing behavior (Motta, et al., 2002), has been demonstrated in the orectolobid shark G. cirratum and is hypothesized to be prevalent in bottomdwelling, suction-feeding elasmobranchs (M otta and Wilga, 2001; Motta, 2004). The generation of suction forces adequate to entrain prey is contingent upon temporally coordinated buccal expansion and is severely limited by physical attributes of the aquatic medium (Muller and Osse, 1980; Wainwright et al., 2001b), suggesting that suction feeding species should exhibit little variation or modulation in strike kinematics. For both C. plagiosum and T. semifasciata an increase in capture success when feeding on live shrimp was noted, demonstrating that with experience young sh arks are able to capture more elusive food regard less of whether their behavioral repertoire consists of a single, specialized behavior or numerous, hi ghly variable behaviors. Thus, for young-ofthe-year sharks, at least two developmental st rategies exit that both have the potential to facilitate exploitation of a taxonomically diverse food base.

PAGE 83

72 Chapter 4: The comparative ontogeny of feed ing performance in tw o sharks: the leopard shark Triakis semifasciata and the whitespotted bambooshark Chiloscyllium plagiosum Abstract Development of the capacity and ability to e xploit prey is crucial to the survival of predatory organisms in the wild. Ontogeneti c trends in food capture performance were quantified using direct measures of suction pr essure and flow in front of the mouth for two elasmobranchs, the ram-f eeding generalist leopard shark Triakis semifasciata and the suction-feeding specialist whitespotted bambooshark Chiloscyllium plagiosum at four times during the first year of lif e. At any given total length, C. plagiosum tended to produce greater maximum subambient pressure and ingest more rounded, high-aspect ratio parcels of water than T. semifasciata Growth, and feeding kinematics that have been shown to scale with growth in these sp ecies, primarily accounted for variation in performance aspects over time. Maximum s ubambient pressure scaled with negative allometry in T. semifasciata and was mediated by an allometric increase in the time to reach maximum gape. Despite similar allo metric increases in the timing of buccal expansion, maximum subambient pressure in C. plagiosum scaled with isometry, presumably as a consequence of earlier onset of hyoid depression over ontogeny and an allometrically increasing buccal reserve volume. The length of the ingested parcel of water relative to mouth width and the aspect ratio of the parcel did not change with growth in either species, demonstrating phys ical constraints imposed by water on the use

PAGE 84

73 of suction in feeding. Though total lengt h was the primary factor responsible for developmental trends in both species, beha vioral changes irresp ective of size also contributed to performance. The contribution of these be havioral changes was minor relative to the effect of growth, and size-independent behavioral variability is hypothesized to contribute litt le to overall performance di fferences over the course of early ontogeny.

PAGE 85

74 Introduction The performance of the feeding apparatu s of aquatic vertebrates is influenced both by the structure and spatial organization of its elements and the temporal and spatial pattern of motion of those elements duri ng a feeding event (Reilly and Lauder, 1988; Ferry-Graham, et al., 2001b; Mo tta and Wilga, 2001; Carroll, et al., 2004). Changes in the mechanical properties of structural elements (Summers, et al., 1998b; Summers, 2000), allometry of musculoskeletal elements (Richard and Wainwright, 1995; Robinson and Motta, 2002; Herrel, et al., 2005), mechanical advantag e of muscular lever systems (Hernandez and Motta, 1997; Wainwright, et al., 2000; Adriaens, et al., 2001; Westneat, 2003), and morphology of the feeding apparatus (Carroll, et al., 2004; Dean and Motta, 2004) have all been shown to affect aspects of feeding performance. While structural changes within an individual take time to develop, even during metamorphosis when change occurs rapidly (Reilly and Lauder, 1990; Galis, 1993; Reilly, 1996), variation in the behavioral repertoire at discrete points during ontoge ny is nearly universally in aquatically feeding verteb rates (Wainwright and Laude r, 1986; Sanderson, 1990). Additionally, distinct modulat ion of food capture behavior by aquatic vertebrates is thoroughly documented (Deban, 1997; Ferry-Gra ham, et al., 2001b; Motta and Wilga, 2001) and has been shown to influence seve ral aspects of performance (Norton, 1991; Wainwright, et al., 2001a; Day, et al., 2005; Higham, et al., 2005). Inertial suction feeding is the dominant prey capture mechanism employed by aquatically feeding vertebrates (Laude r, 1985b) and several morphological and behavioral correlates for enhancing suction performance have been identified (Liem, 1993; Carroll, et al., 2004; Gibb and Ferry-Graham, 2005; Van Wassenbergh, et al.,

PAGE 86

75 2005). As suction feeding requires imparting fo rce to a food item via displacement of the water surrounding it, traits that increase the force available to be transferred are characteristic of suction feed ers. Morphological attributes that enhance suction include a smaller buccal aperture that is often late rally enclosed, larger jaw and hyoid abductor muscles, reduced oral dentition, a deep h ead profile, hypertrophied epaxial musculature and an expansible buccophar yngeal cavity (Muller and Osse 1984; Liem, 1993; Carroll, et al., 2004). Behavioral tra its that are correlated with increased suction performance include rapid, extensive hyoid depression, buccopharyngeal expansion, and cranial rotation (Svanback, et al., 2002; Carroll, et al., 2004; Gibb and Ferry-Graham, 2005). Over ontogeny gradual changes may occur in these morphological and behavioral traits, producing the potential for substantial devel opmental changes in performance (Motta and Kotrschal, 1992; Galis, et al., 1994; Koehl, 1996). While several aspects of the anatomy and food capture behavior of teleost fishes have been studied over ontogeny (Cough lin, 1991; Cook, 1996; Hernandez, 2000; Krebs and Turingan, 2003; Van Wassenbergh, et al., 2005), few studies have focused on the development of these aspect s in elasmobranchs (but s ee Ferry-Graham, 1997; 1998a; 1998b; Chapters 2 and 3). Adult elasmobr anchs are known to possess morphologically and functionally varied feeding mechanisms (Shirai and Nakaya, 1992; Wu, 1994; Motta, et al., 1997; Wilga and Motta, 1998a; Wilga and Motta, 2000) that are employed to exploit diverse prey (Cortes, 1999; Sims, 2000; Motta and Wilga, 2001). Ontogenetic dietary changes are well documented in elasm obranchs (Lowe, et al., 1996; Cortes, 1999; Kao, 2000; Ebert, 2002) and have been primarily attributed to changes in habitat and growth. It is likely, however that developmental change s in morphology and behavior

PAGE 87

76 affect performance in ways that could enab le either the exploitation of novel prey, facilitating ontogenetic dietary changes, or a change in how prey are captured. The use of disparate prey capture behaviors to occupy diverse feeding niches despite the anatomical simplicity of their feeding appa ratus (Motta and Wilg a, 2001; Motta, 2004), as well as the absence of metamorphi c changes in feeding morphology, makes elasmobranchs a worthwhile clade in wh ich to study the development of feeding morphology and behavior. The elasmobranch species selected for study were the leopard shark Triakis semifasciata (Triakidae) and the wh itespotted bambooshark Chiloscyllium plagiosum (Hemiscylliidae). Triakis semifasciata is a common demersal species found along the Pacific coast of North America and north ern Central America where it frequents estuaries, lagoons, and shallow bays (Com pagno, 1984a). This species is aplacental viviparous, giving birth to pups that are approximately 20-26 cm and grow between 2 and 4 cm yr-1 (Compagno, 1984a; Kusher, et al., 1992). Triakis semifasciata is an opportunistic generalist that feeds on a broad taxonomic and functional diversity of prey including benthic invertebrates and fishes throughout ontogeny (Talent, 1976; Kao, 2000). Despite possessing several morphological attributes typical of suction feeders, previous study has shown that T. semifasciata typically employs ram-dominated food capture (Ferry-Graham, 1998a). This species is, however, capable of modulating capture kinematics in response to food size and elusivity to utilize greater su ction (Chapter 2). The whitespotted bambooshark Chiloscyllium plagiosum is a common shallow water, reef-dwelling, epibenthic species of the Indo-West Pacific an d eastern coast of Southern Asia (Compagno, 1984b). Chiloscyllium plagiosum is oviparous, hatching at 12-20 cm

PAGE 88

77 TL (A. Cornish, pers. comm.; Tullis and Peterso n, 2000). This species is an opportunistic generalist that feeds pr imarily on benthic invertebrates a nd, with age, occasionally fish (A. Cornish, pers. comm.). Previous study has shown that C. plagiosum displays characteristic suction-feeding morphology a nd employs a more behaviorally conserved, suction-dominated food capture be havior (Wu, 1993; Chapter 3). The goals of this study were: 1) to quantify developmental changes in food capture performance for two elasmobranch species, a ram-feeding generalist and a suction-feeding specialist, fr om birth/hatching through the first year of life. The study was limited to the first year of life because juvenile mortality is high in many elasmobranchs but declines with age (Manire and Gruber, 1993; Heupel and Simpfendorfer, 2002), making this a critical de velopmental period.; 2) to determine the contribution of organism size and behavioral development to performance changes over ontogeny in both species; and 3) to integr ate knowledge of mo rphology, behavior, and performance to propose general developmenta l trends in the feeding behavior of youngof-the-year sharks that might impact indi vidual prey capture ca pacity and survival. Methods and Materials Experimental Animals Twelve neonatal T. semifasciata were obtained from Mote Marine Laboratory (MML), Sarasota, Florida and six hatchling C. plagiosum were obtained from SeaWorld, Orlando, Florida. Triakis semifasciata were approximately 1 month old upon arrival at MML, as determined by their size (Kusher, et al., 1992), and had b een previously fed commercial aquarium feed, while C. plagiosum were reared at S eaWorld and obtained

PAGE 89

78 prior to first feeding. Triakis semifasciata were maintained in a 2.4-meter diameter, 1400-liter semicircular co mmunal holding tank at 27 1C and 32 2 ppt salinity while C. plagiosum were maintained in a 340-li ter communal holding tank at 26 1C and 332 ppt salinity. During experiment al sessions, individuals of each species were isolated into a 90cm45cm30cm filming tank containing 55 liters of water from their respective holding tank. Animals were maintained on a diet cons isting of 3-4% of their body weight in various live and dead food types of various sizes ad libitum twice per week. This feeding frequency was maintained throughout the study but for feedings immediately preceding experimental sessions the ration was cut to 2% to encourage active feeding during filming (see below). During experimental se ssions, only pieces of Atlantic threadfin herring Opisthomena oglinum (for T. semifasciata ) or the clam Mercenaria mercenaria (for C. plagiosum ) that were scaled to the mouth wi dth of the animal were fed. These food types were selected from a subset of f ood types for which the behavioral response of these two species is known (C hapters 2 and 3) in order to maximize suction effort and performance (Carroll, et al., 2004). Though an experimental session often consisted of as many as ten food capture events per individual, only the first five were considered for analysis in order to avoid the effect s of satiation (Sass and Motta, 2002). Experimental Techniques The experimental period for this study was 52 weeks and was broken into four even segments for analysis, beginning w ith birth/hatching of the individuals. Experimental sessions only occurred dur ing the middle five weeks of each 13-week

PAGE 90

79 segment, providing a sample of feeding perf ormance at four disc rete times during the young-of-the-year period for both species. Feeding performance was measured during each time segment using two independent te chniques (suction pressure and particle motion), both of which were accompanied by high-speed recordings from which kinematic data were obtained. Recordings of capture events were made with a Redlake PCI 1000 high-speed digital camera (Redlake, San Diego, CA, USA) that was placed perpendicular to the aquarium to provide a lateral view. Recordings were made at 250 fps with a shutter speed of 1/1000 ms a nd illumination was provided by two, 500-Watt quartz-halogen lights. Animals were traine d to feed under illumination prior to the experiment and were allowed a 20-minute accl imation period prior to each feeding session. A rule beside the shark provided distance measure and only orthogonal views were retained for analysis. Kinematic da ta were measured from recordings using Redlake MotionScope PCI software versi on 2.21.1 and SigmaScan Pro version 4. The variables measured were selected for their functional relevance and prior employment in other studies of elasmobranch feeding (M otta, et al., 1997; Ferry-Graham, 1998a; Wilga and Motta, 1998b). From the onset of lowe r jaw depression (time 0 ms), the following kinematic variables were quantified: 1) strike distance, from the closest point on the food to the lower jaw of the shark (cm); 2) ma ximum gape (cm); 3) time to maximum gape (ms); 4) time to close, from time maximum gape was first reached to jaw closure (ms); 5) maximum cranial elevation (degrees); 6) time to maximum cranial elevation (ms); 7) time to onset of cranial elevation (ms); 8) time to offset of cranial elevation (ms); 9) duration of cranial elevation (ms); 10) time to onset of hyoid depression (ms); 11) maximum hyoid depression (cm); 12) time to maximum hyoi d depression (ms); 13) time to hyoid

PAGE 91

80 retraction (ms); 14) duration of hyoid depression (ms); 15) time gills start to open (ms); 16) total strike duration from onset lowe r jaw depression to jaw closure (ms); 17) duration of food movement duri ng the strike (ms); 18) distan ce the food moves during the strike (cm); 19) velocity of the food over the course of the strike (cm s-1); 20) distance the predator moves during the strike (cm); and 21) velocity of th e predator over the course of the strike (cm s-1). Variables (18) and (20) were us ed to calculate the ram-suction index (RSI) (Norton and Brainerd, 1993). The RSI is calculated as (DPREDATOR–DPREY)/ (DPREDATOR+DPREY) and indicates the relative contri bution of forward motion of the predator and motion of the prey to a given capture event. Both the distance moved by and velocity of the food were used as indicators of performance in subsequent analyses (see below). No measures were made of upper jaw protrusion excursion or timing because protrusion was not present in all cases. Simultaneous with the recordings descri bed above, food capture performance was measured by quantifying the subambient suction pressure at the position of the food. To obtain these data, a 5-French Millar MPC-500 Mikro-tip cat heter (Millar Instruments, Inc., Houston, TX, USA) was inserted through an aperture in a Plexig las false bottom that divided the filming tank vertically. Food items were then notched with a knife such that, when wrapped around the catheter, the tip of th e pressure transducer was exposed directly to the flow generated by the sh ark during capture. The catheter was placed at the level of the food rather than within the buccopharyngeal cavity of the shark because the suction pressure available to elicit food displacement was the performance vari able of interest. Additionally, the long-term e ffects of cannula implantation or permanent catheterization on elasmobranch growth are unknown and fre quently repeated implantation was not

PAGE 92

81 feasible on such small animals. The cat heter was connected to a Millar TCB-500 transducer control unit that was calibrated and zeroed at the beginning of each experimental session to record pressure diffe rential at depth. The output from the control unit was recorded with a Yokogawa DL 716 di gital oscilloscope and exported to a computer for subsequent analysis using SigmaScan Pro version 4 (SPSS Inc.). Recordings from the camera and oscilloscope were synchronized via a cyclically repeating electronic pulse. The performance variables measured from the synchronized recordings were: 1) the durati on of the subambient pressure pulse (pulse duration) (ms); 2) the area under the curve of the sum-ambient pressure pulse (pulse area) (kPa ms); 3) the maximum subambient pressure (kPa); 4) the time to reach maximum subambient pressure from onset of lower jaw depre ssion (ms); 5) the time to reach maximum subambient pressure from onset of hyoid depr ession (ms); and 6) the rate of pressure decrease from the onset of subambient pre ssure until maximum subambient pressure is reached (kPa ms-1). Twenty suction pulses with s ynchronized video were recorded from each time segment for each species, resulting in the analysis of a to tal of 160 captures. No more than three captures per individual du ring a given time segment were included in the analysis to avoid bias due to individual differences. In addition to measuring subambient pressure generation, performance was quantified during independent experimental sessions within each time segment by describing the pattern of particle motion into the buccopharyng eal cavity during food capture. To accomplish this, 4.1 g of Artemia spp. cysts were soaked for 1-3 hours in seawater until neutrally buoyant and then added to the filming tank, producing a final seeding density of ~3 g l-1 (Lauder and Clark, 1984). Artemia spp. cysts were selected in

PAGE 93

82 lieu of polystyrene or hollow-ground silver spheres (Muller and Osse, 1984; FerryGraham, et al., 2003; Day, et al., 2005; Hi gham, et al., 2005), because they are biodegradable and repeated exposure over ont ogeny is not detrimental to experimental animals. The filming tank was then enclos ed within a photo-opaque, non-reflective box that admitted light only through two, 3-mm wide slits on the top and one side of the box producing a sheet of quartz hal ogen illumination perpendicula r to the camera that ran through the midsagittal plane of the shark. Illuminating the filming tank in this way allowed discrimination of par ticles directly in-line with the shark and allowed the displacement of these particle s to be manually tracked duri ng food captures. Particles that entered the oral aperture during the stri ke were tracked backward field-by-field from the time they were ingested to their initi al resting position at th e onset of lower jaw depression. Connecting the initial positions of the distal-most particles in each direction allowed description of the parcel of water ingested by the shark during food capture from the lateral aspect (Fig. 15) (Day, et al., 2005; Higham, et al., 2005). A minimum of 20 particle trajectories were described per capture event and additional particles were tracked until subsequent traject ories did not alter the shape determined for the ingested parcel. The following attributes of the ingested parcel were then quantified as indicators of performance using SigmaScan Pro versi on 4: 1) maximum parcel length at any position (cm and relative to mouth width [MW] ); 2) maximum parcel height at any position (cm and MW); and 3) parcel area (cm2). Parcel height was then divided by parcel length to obtain an asp ect ration that was used to desc ribe the relative shape of the ingested parcel (Day, et al., 2005; Higham, et al., 2005). Ten flow patterns from independent capture events were recorded for each time segment for each species,

PAGE 94

83 resulting in the analysis of a total of 80 captures. No more than two captures per individual during a given time segment were incl uded in the analysis to avoid bias due to individual differences. Figure 15: Illustrative examples of the range of relative si ze and shape of the parcel of water ingested by Chiloscyllium plagiosum (top) and Triakis semifasciata (bottom) during capture ev ents. Images have been scaled such that head length is equal in all cases. The parcels are truncated ventrally by the Plexiglas bottom and dorsally by the rostrum of the shark. Frames to the left depict relatively rounded, high aspect ratio parcels while those to the right depict more elongated, low aspect ratio parcels relative to head length. Overall, high aspect ratio parcels (average=0.800.21) are typical of C. plagiosum (upper left) while low aspect ratio parcels (0. 570.11) are typical of T. semifasciata (lower right). Statistical Analyses To ascertain whether the use of the cath eter or particles influenced food capture behavior, kinematic data were compared agai nst data for the same species feeding on the same food type when neither method to meas ure performance was employed (Chapters 2 and 3). All kinematic data were log-tr ansformed and checked for normality and

PAGE 95

84 homogeneity of variance using the Kolm ogorov-Smirnov and Levene Median tests, respectively. Model II linear regressions we re performed against total length for each kinematic variable and the studentized residuals obtained (Q uinn and Keough, 2002). Size-corrected, studentized residuals for all kinematic variables were then used in a correlation matrix-based Principal Com ponents Analysis (PCA) (Quinn and Keough, 2002). An Equamax rotation was used because it produced higher loadings than any other rotation, enhancing data interpretation, and vari ables that loaded above an absolute value of 0.5 were identified as contributing heavily to the variability within the respective principal component (PC). Principal com ponents with an eigenvalue greater than 1.0 were retained for further analysis. The factor loading scores for each capture event on each principal component were then used in a series of one-way MANOVAs to identify differences between clusters of capture sequences in multivariate space. Capture sequences from each time segment using each technique to quantify performance were independently compared against sequences during the same time segment in which neither technique was employed (Chapters 2 and 3). Significan ce was assessed using Pillai’s trace because it is robust to multivar iate deviation from normality (Zar, 1999). Regressions were performed using SigmaSta t Pro version 3.1 (SPSS Inc.) while the PCA and MANOVA tests were performed using Systat 11 (SPSS Inc.). Once the effects of the presence of th e catheter and particle s had been assessed, log-transformed performance data for each variable were regressed against logtransformed total length data using Model II linear regressi ons to describe changes in performance over ontogeny. Model II linear re gressions were appropriate because error existed in measurements of both the depe ndent and independent variables (Quinn and

PAGE 96

85 Keough, 2002). Significance was assessed via ANO VA. In addition to testing for the presence of trends in performance vari ables over ontogeny, the nature of these relationships was assessed to determine whet her performance scaled isometrically or allometrically. The predicted isometric sca ling coefficient for each performance variable was determined by the dimensionality of the variable and existing kinematic and anatomical scaling data for these two species (Chapters 2 and 3). Li near variables, such as maximum parcel height and velocity of the food, were predicted to scale with a coefficient of one, whereas planar variables, such as parcel area and pulse area, were predicted to scale with a coefficient of two. Maximum parcel height and length relative to mouth width (i.e. with size accounted fo r), as well as parcel aspect ratio, were predicted to scale with a coefficient of zero. As strike duration does not scale with size in either species (Chapters 2 and 3), the duration of the pressu re pulse and the time of its onset relative to the onset of lower jaw depression were predicted to scale with coefficients of zero. The onset of hyoi d depression does not scale with size in T. semifasciata (Chapter 2), but scales with a coefficient of -0.3 in C. plagiosum (Chapter 3), thus the time of maximu m subambient pressure relative to the onset of hyoid depression was expected to scale with co efficients of zero and -0.3, respectively. Maximum subambient pressure, though influenced by behavioral deviation in the rate and duration of buccal expansion (Muller and Osse, 1984; Carroll, et al., 2004), has been shown to correlate strongly with volume ch ange in the buccal cavity during feeding (Svanback, et al., 2002; Carroll, et al., 2004). The difference in the volume of the buccal cavity (buccal reserve volume) has been show n to scale with a coefficient of 3.0 in T. semifasciata (Chapter 2) and 3.3 in C. plagiosum (Chapter 3), so maximum subambient

PAGE 97

86 pressure was predicted to scale similarly. Deviation from an isometric coefficient was tested for each variable using a modified Student’s t-test (Z ar, 1999). Differences between regressions at the species level for a given variable were also assessed using a modified Student’s t-test (Zar, 1999). To determine whether changes in perfo rmance variables o ccurred over ontogeny irrespective of size, studentiz ed residuals were obtained from the regressions of each variable against total length and these values were regressed agains t log-transformed age data (in days). Deviation from a coef ficient of zero was tested using ANOVA and indicated differences in that performance variable over ontogeny that could not be accounted for simply by an increase in TL. Fo r each performance variable that displayed a trend over ontogeny that c ould not be accounted for by TL alone, a backward stepwise multiple regression was used to elucidate the contribution of specific kinematic variables to this relationship (F-to-Enter=2.0, Fto-Remove=1.9, 20 steps) (Quinn and Keough, 2002). Only variables directly related to th e timing, duration, and extent of motion of cranial features of the shark were retained fo r this analysis (i.e. strike distance, duration of food movement, and RSI were not included) Because significant correlations among variables included in multiple regressions can lead to problems of multicollinearity that substantially weaken the c onclusions of the test (Qui nn and Keough, 2002), a Spearman rank order correlation analysis was performed to determine if kinematic variables were correlated. Several signifi cant correlations (P<0.05) were detected, so a correlation matrix-based PCA was performed to consolidate related variables into several uncorrelated PCs. Variables that loaded above an absolute value of 0.5 were identified as contributing heavily to the variability with in the respective PC. Principal components

PAGE 98

87 with an eigenvalue greater than 1.0 were retained for further analysis. The PCs were then used as independent variables in a second set of backward stepwise multiple regressions with the operational criteria pr ovided above for each performa nce variable that exhibited a trend over ontogeny that c ould not be accounted for by TL alone. All regressions, including the backward stepwise multiple re gressions, were performed with SigmaStat Pro version 3.1. Results Sharks of both species generally swam slowly around the filming tank or rested on the bottom until food was introduced. On aver age, strikes were typically faster in C. plagiosum (average total strike duration of 6917 ms) than in T. semifasciata (9616 ms), involved relatively less excursion of stru ctural elements, and were characterized by substantially less variability. For a detailed description of feeding kinematics in both species over ontogeny see Chapters 2 and 3. Differences were not detected in multidimensional space during any time segm ent between capture events with and without catheterized food, or between capture events with and without particles present (Table 11). At any given total length, the maximum suba mbient pressure generated by C. plagiosum tended to be greater than that generated by T. semifasciata though considerable overlap was apparent (Fig. 16). Additionally, at any given total length the aspect ratio of the parcel ingested tend ed to be higher (i.e. more rounded) in C. plagiosum but lower (i.e. more elliptical) in T. semifasciata (Fig. 17).

PAGE 99

88 Table 11: Comparison between capture events with food catheterized and without, and with particles present and absent, during each time segment fo r both species via MANOVA based on principal component loading scores. Values shown are for Pillai’ s trace. df=degrees of freedom. Note that effects were not detected during any time segment for either technique used to quantify performance. Triakis semifasciata Chiloscyllium plagiosum df F P df F P With catheter vs. without: Segment 1 5, 34 1.255 0.306 5, 39 0.752 0.590 Segment 2 5, 34 0.768 0.579 5, 39 2.077 0.096 Segment 3 5, 34 0.796 0.560 5, 39 0.756 0.587 Segment 4 5, 34 1.525 0.208 5, 39 2.118 0.084 With particles vs. without: Segment 1 6, 23 0.983 0.459 5, 29 0.299 0.909 Segment 2 6, 23 2.193 0.081 5, 29 1.468 0.231 Segment 3 6, 23 1.464 0.209 5, 29 0.504 0.771 Segment 4 6, 23 1.763 0.152 5, 29 1.049 0.408 0 5 10 15 20 25 2025303540455055 Total Length (cm)Max Sub-ambient Pressure (kPa) Triakis semifasciata Chiloscyllium plagiosum Figure 16: Values of maximum subambient pressures generated during food capture events in Chiloscyllium plagiosum and Triakis semifasciata relative to total length (TL). Note that at any given TL values tend to be higher (i.e. greater subambient pressure drop) in C. plagiosum than in T. semifasciata

PAGE 100

89 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 202530354045 Total Length (cm)Aspect Ratio Triakis semifasciata Chiloscyllium plagiosum Figure 17: Aspect ratio values for parcels of wate r ingested during food capture events in Chiloscyllium plagiosum and Triakis semifasciata relative to total length (TL). Note that at any given TL values tend to be higher (i.e. more rounded) in C. plagiosum than in T. semifasciata Scaling of Performance When compared to a slope of zero, most performance variables exhibited some trend relative to growth, with the value of all variables for which trends were present in T. semifasciata displaying positive coefficients (Table 12, Slope and P columns). The same was not true for C. plagiosum with pulse duration and the time to reach maximum subambient pressure relative to the on set of both lower jaw and hyoid depression decreasing with growth (Table 12). Though the rate of pres sure decrease did not change with growth in T. semifasciata it tended to increase substan tially (slope=4.42) (P<0.001) in C. plagiosum (Table 12). Although maximum parcel height relative to mouth width tended to increase slightly in T. semifasciata it exhibited no trend with growth in C.

PAGE 101

90 plagiosum Velocity of the food during the stri ke, maximum parcel length relative to mouth width, and parcel aspect ra tio notably did not change with growth in either species. Table 12: Regression parameters for performance variables versus total length and tests of isometry for both species. The observed slope (Slope) and expected slope (SlEXP) are given. Standard error and P values are from ANOVA tests for relationships between e ach variable and total length. The expected slope for isometry is given in the column labeled SlEXP. Where the expected slope was not zero, a modified Student’s t-test was used to assess isometry. Where the expected slope was zero (n/a) the results of the ANOVA were defaulted to. For variable s quantified with catheterized food, t0.05(1), 78=1.665. For variables quantified with particles present t0.05(1), 38=1.686. Variable Slope y-int r2 StErr P SlEXP t Triakis semifasciata Pulse Duration 1.08 -0.06 0.10 0.37 0.005* 0.0 n/a Pulse Area 3.93 -4.86 0.24 0.79 <0.001* 2.0 79.85* Maximum Subambient Pressure 2.45 -3.35 0.18 0.59 <0.001* 3.0 -30.26* Time of Maximum Pressure Relative: To Onset of Lower Jaw Depression ----0.481 0.0 n/a To Onset of Hyoid Depression ----0.812 0.0 n/a Rate of Pressure Decrease ----0.134 1.0 -2.70* Distance Food Moves 1.45 -2.15 0.25 0.41 0.001* 1.0 37.57* Velocity of the Food ----0.115 1.0 -3.77* Maximum Parcel Length (cm) 1.05 -1.21 0.34 0.24 <0.001* 1.0 7.67* Maximum Parcel Length (MW) ----0.292 0.0 n/a Maximum Parcel Height (cm) 1.40 -2.00 0.39 0.28 <0.001* 1.0 47.69* Maximum Parcel Height (MW) 0.60 -1.13 0.10 0.28 0.042* 0.0 n/a Parcel Area 2.74 -3.81 0.56 0.40 <0.001* 2.0 63.72* Parcel Aspect Ratio ----0.262 0.0 n/a Chiloscyllium plagiosum Pulse Duration -1.38 3.49 0.28 0.25 <0.001* 0.0 n/a Pulse Area 6.55 -8.21 0.62 0.58 <0.001* 2.0 154.91* Maximum Subambient Pressure 3.26 -4.31 0.59 0.31 <0.001* 3.3 -1.63 Time of Maximum Pressure Relative: To Onset of Lower Jaw Depression -1.11 3.12 0.26 0.21 <0.001* 0.0 n/a To Onset of Hyoid Depression -1.53 3.61 0.15 0.42 <0.001* -0.3 -57.90* Rate of Pressure Decrease 4.42 -7.20 0.58 0.42 <0.001* 1.0 160.16* Distance Food Moves 0.66 -1.03 0.13 0.28 0.026* 1.0 -21.82* Velocity of the Food ----0.876 1.0 -52.26* Maximum Parcel Length (cm) 0.73 -1.03 0.37 0.16 <0.001* 1.0 -31.35* Maximum Parcel Length (MW) ----0.065 0.0 n/a Maximum Parcel Height (cm) 1.12 -1.72 0.53 0.17 <0.001* 1.0 12.40* Maximum Parcel Height (MW) ----0.622 0.0 n/a Parcel Area 1.16 -1.80 0.37 0.25 <0.001* 2.0 -61.95* Parcel Aspect Ratio ----0.067 0.0 n/a

PAGE 102

91 Behavioral Effects on Performance Though growth accounted for most of the variability in performance over ontogeny, when the studentized residuals from the regressions of performance variables against total length were regresse d against time, a number of tr ends remained (Table 13). For T. semifasciata pulse duration, pulse area, di stance moved by the food, maximum parcel height, and parcel area exhibited weakly increasing trends over time (Table 13). Backward stepwise multiple regressions i ndicated that numerous kinematic variables partially accounted for these trends (Table 14). The relative extent of maximum gape was largely accountable for trends in all of these performance variable s except pulse area, which was accounted for primarily by total st rike duration (Table 14). In addition to maximum gape, trends in the remaining four performance variables were accounted for by a variety of kinematic variables, with no clear tendency for a ny single variable to dictate performance (Table 14). Two notable exceptions to this ge neralization were the time to hyoid retraction and the duration of hyoid depression, which exhibited strong relationships with distance the food moves dur ing a strike (Table 14). PCA consolidated the numerous initial kinematic variables into six PCs (Table 15) and eliminated problems of multicollinearity due to correlation among variables. Backward stepwise multiple regressions for T. semifasciata based on these PCs determined that trends in pulse duration and pulse area could be accounted for primarily by PC4 (timing variables describing cranial elevation, maximum gape, a nd initial hyoid displa cement), with PC6 (the duration and extent of cr anial elevation) also contribut ing to pulse duration (Table 16). Trends in the distance that the f ood moved and parcel area were accounted for primarily by PC5 (maximum gape and the time the gills open). The trend in maximum

PAGE 103

92 parcel height was also explained by PC 5, but PC1 (duration of hyoid and buccal displacement) also c ontributed (Table 16). Table 13: Regression parameters for performance vari ables against age in da ys for both species irrespective of size. For variables quantif ied in the presence of the catheter, t0.05(1), 19=1.729. For variables quantified with particles present, t0.05(1), 9=1.833. Variable Slope y-int r2 StErr P Triakis semifasciata Pulse Duration 0.13 -0.27 0.08 0.05 0.013* Pulse Area 0.26 -0.54 0.07 0.11 0.020* Maximum Sub-ambient Pressure ----0.080 Time of Maximum Pressure Relative to: Onset of Lower Jaw Depression ----0.286 Onset of Hyoid Depression ----0.638 Rate of Pressure Decrease ----0.650 Distance Food Moves 0.13 -0.28 0.13 0.05 0.022* Velocity of the Food ----0.119 Maximum Parcel Length (cm) ----0.300 Maximum Parcel Height (cm) 0.09 -0.20 0.14 0.04 0.016* Parcel Area 0.13 -0.27 0.13 0.05 0.023* Parcel Aspect Ratio ----0.166 Chiloscyllium plagiosum Pulse Duration 0.43 -0.92 0.00 0.13 0.001* Pulse Area ----0.808 Maximum Sub-ambient Pressure 0.17 -0.35 0.07 0.07 0.021* Time of Maximum Pressure Relative to: Onset of Lower Jaw Depression ----0.832 Onset of Hyoid Depression ----0.497 Rate of Pressure Decrease ----0.120 Distance Food Moves ----0.597 Velocity of the Food ----0.508 Maximum Parcel Length (cm) ----0.701 Maximum Parcel Height (cm) ----0.918 Parcel Area ----0.725 Parcel Aspect Ratio ----0.834

PAGE 104

93 Table 14: Kinematic variables contributing to performa nce differences over time irrespective of size in Triakis semifasciata Only variables included in the final step of the backward stepwise multip le regressions are shown (e.g. those that contribute significantly). The total perce nt variance explained is a measure of how well the multiple regression comprising the final kinematic variables describes the trend in the respective performance var iable. Max=maximum; Ret=retraction; Depress=depression; Pred=predator; Var=variance. Pulse Duration Pulse Area Distance Food Moves Max Parcel Height Parcel Area Variable Coeff StErr P Coeff StErr P Co eff StErr P Coeff StErr P Coeff StErr P Constant 0.00 0.02 -0.00 0.04 -0.00 0.02 -0.00 0.01 -0.00 0.01 -Max Gape 0.80 0.25 0.002 ---1.11 0.29 <0.001 0.59 0.20 0.006 1.11 0.25 <0.001 Time to Max Gape 0.58 0.26 0.031 -------0.58 0.21 0.009 ---Time to Close 0.47 0.13 <0.001 ------------Time to Onset Hyoid Depress ------0.46 0.21 0.035 ------Max Hyoid Depress ---------0.30 0.13 0.031 ---Time to Max Hyoid Depress -0.78 0.31 0.015 ------0.42 0.20 0.048 ---Time to Hyoid Retraction -------3.18 1.26 0.016 ------Duration of Hyoid Depress ------2.50 1.02 0.019 ------Time Gills Start to Open ---------0.09 0.03 0.011 0.15 0.04 <0.001 Total Strike Duration ---1.48 0.55 0.008 -------0.45 0.18 0.015 Distance Predator Moves ------------0.19 0.08 0.028 Total % Var Explained 31.7 14 .6 40.0 43.7 57.9

PAGE 105

94 Table 15: Principal component loadings after Equamax rotation of kinematic variables exclusively associated with motion of the sh ark during capture events in Triakis semifasciata Bold face values indicate variables determined to load heavily on the respective component (loading scores>|0.5 ) (N=12). Together the six components explain 83.42% of the overall variability in the data set. For clarity, all loadings < |0.25| are replaced by 0. Depress=depression; Max=maximu m; Elev=elevation; Var=variance. Variable PC1 PC2 PC3 PC4 PC5 PC6 Duration of Hyoid Depress 0.922 0 0 0 0 0 Time to Hyoid Retraction 0.887 0 0 0.281 0 0 Time to Close 0.716 0 0.406 0 0 0.261 Total Strike Duration 0.667 0 0 0.538 0 0.295 Max Hyoid Depress 0 -0.969 0 0 0 0 Time to Max Hyoid Depress 0 0.933 0 0 0 0 Distance Predator Moves 0 0 0.955 0 0 0 Velocity of the Predator -0.265 0 0.877 0 0 0 Onset of Cranial Elev 0 0 0 0.787 0 -0.321 Time to Max Cranial Elev 0 0.251 0 0.749 0 0.370 Time to Max Gape 0.444 0 0 0.738 0 0 Time to Onset of Hyoid Depress 0 0 0 0.706 0 0.302 Offset of Cranial Elev 0 0 0 0.592 0 0.619 Max Gape 0 0 0 0 0.800 0 Time Gills Start to Open 0 0 0 0 0.646 0 Duration of Cranial Elev 0 0 0 0 0 0.853 Max Cranial Elev 0 0.284 0 0 0 0.768 Eigenvalue 3.10 2.09 2.13 3.17 1.37 2.32 Percent Var Explained 18.25 12.27 12.55 18.64 8.08 13.64

PAGE 106

95 Table 16: Partial regression coefficients from the backward stepwise mu ltiple regressions of principal components (PC) that contribute to performance differences over time irrespective of size in Triakis semifasciata Only PCs included in the final step of the backward stepwise multiple regressions are shown (e.g. those that contribute significantly). The total percent va riance explained is a measure of how well the multiple regress ion comprising the final PCs describes the trend in the respective performance variable. Var=variance; Max=maximum. Pulse Duration Pulse Area Distance Food Moves Max Parcel Height Parcel Area Variable Coeff StErr P Coeff StErr P Co eff StErr P Coeff StErr P Coeff StErr P Constant 0.00 0.02 -0.00 0.04 -0.00 0.02 -0.00 0.01 -0.00 0.01 -PC1 ----------0.02 0.01 0.040 ---PC4 0.06 0.02 <0.001 0.08 0.04 0.048 ---------PC5 ------0.07 0.02 <0.001 0.04 0.01 <0.001 0.08 0.01 <0.001 PC6 -0.04 0.02 0.041 ------------Total % Var Explained 18.2 4.9 29.5 32.3 44.6

PAGE 107

96 For C. plagiosum only pulse duration and maxi mum subambient pressure displayed increasing tends over time irresp ective of growth (Table 13). Backward stepwise multiple regression revealed that th ese trends could be accounted for primarily by variables associated with the time taken to reach and extent of maximum cranial elevation, for pulse duration, or total strike duration, for maximum subambient pressure (Table 17). The clarity and significance of these relations hips was not enhanced by PCA so results of this analys is are not presented here. Table 17: Kinematic variables contributing to performa nce differences over time irrespective of size in Chiloscyllium plagiosum Only variables included in the final step of the backward stepwise multiple regressions are shown. The total percent variance explained is a measure of how well the multiple regression comprising the final kinematic variables describes the trend in the respective performance variable. Max=maximum; Var=variance. Pulse Duration Max Subambient Pressure Variable Coeff StErr P Coeff StErr P Constant -0.02 0.39 -1.59 0.42 -Max Cranial Elevation -0.17 0.08 0.045 ---Time of Max Cranial Elevation 0.60 0.20 0.005 ---Onset of Cranial Elevation -0.10 0.05 0.037 ---Total Strike Duration ----0.42 0.28 <0.001 Total % Var Explained 16.6 17.5 Discussion Trends exist in several measures of food capture performance for both T. semifasciata and C. plagiosum over ontogeny that can be ac counted for by size increase and, to a lesser degree, behavioral changes. Size effects are prevalen t in most studies of biomechanical performance (Losos, 1990; Elswor th, et al., 2003; Herrel, et al., 2005) and have led to the generation of numerous models intended to predict scaling coefficients of kinematic measures (Hill, 1950; O'Reilly, et al., 1993; Richard and Wainwright, 1995). Explanatory mechanisms for trends with si ze typically assume isometric growth and

PAGE 108

97 range from changes in muscle contraction rate to the disparity between the scaling coefficient of muscular force production (s lope=2) and the mass of skeletal elements (slope=3). The model used here to predic t performance was base d on knowledge of the scaling of kinematics in T. semifasciata and C. plagiosum which primarily conform to the model of Richard and Wainwright (1995) but deviate because of a llometric growth in several morphological and anatomical features (Chapters 2 and 3). The generation of greater subambient su ction pressure during feeding in teleost fishes is correlated with increased cross-sec tional area of muscles th at abduct and expand the buccopharyngeal cavity, faster buccophar yngeal and hyoid expansion, and increased cranial elevation (Muller and Osse, 1984; No rton and Brainerd, 1993; Ferry-Graham, et al., 2001b; Sanford and Wainwright, 2002; Carroll, et al., 2004). In both T. semifasciata and C. plagiosum the absolute subambient suction pre ssure increases with size, but to a greater extent in C. plagiosum In T. semifasciata the time to reach maximum hyoid depression and maximum cranial elevation do no t change with size but the time to open the mouth (time to maximum gape) increases (slope=0.35), despite being accompanied by positively allometric growth in the weight and cross-sectional area of jaw abducting muscles and isometric growth in buccal reserv e volume (Chapter 2). This suggests that in T. semifasciata maximum subambient pressure shoul d scale negatively allometrically as the rate of oral expansion decreases and resu lts show that this is indeed the case. In C. plagiosum the time to open the mouth, maximum hyoid depression, and maximum cranial elevation all increase (slopes between 0.2 and 0.4) despite positively allometric growth in abducting musculature (Chapter 3), suggesting that maxi mum subambient pressure should also scale with nega tive allometry. Maximum subambient pressure, however,

PAGE 109

98 scaled with isometry in C. plagiosum This discrepancy may be due to the relationship between the time of onset hyoid depression and size. In T. semifasciata the onset of hyoid depression exhibits no relationship with growth (Chapter 2), but in C. plagiosum the hyoid begins to be depressed relatively ea rlier in the strike w ith growth (slope=-0.32) (Chapter 3). Functionally this means that flow would be genera ted through a slowly expanding, smaller oral aperture earlier in th e strike, which should increase the velocity of flow into the buccopharyngeal cavity and, therefore, maximum subambient pressure (Muller and Osse, 1984; Svanbac k, et al., 2002). Indeed, in C. plagiosum the time of maximum subambient pressure relative to both the onset of lower jaw and hyoid depression scales with negative allometry, while the rate of pr essure decrease scales with positive allometry (slope=4.42) (Table 12), re sulting in isometric scaling of maximum subambient pressure. The timing of hyoid depression as the prim ary determinant of maximum subambient pressure has been hypot hesized for orectolobid sharks (Motta, 2004), and changes in the relative timing of hyoid depression appear to enhance pressure generation over ontogeny in C. plagiosum Both species used in this study have b een shown to utilize inertial suction during food capture (Chapters 2 and 3; FerryGraham, 1998a; Wu, 1 993; 1994), although the relative contribution of sucti on tends to be greater in C. plagiosum The viscosity and density of the aquatic medium dictate that suction is a relatively near-field phenomenon, as the effects of flow degrade as the cube of distance from the oral aperture (Muller and Osse, 1984; Svanback, et al., 2002; Ferry-Graha m, et al., 2003). The greatest parcel length of water sucked into the mouth by T. semifasciata was 4.45 cm (1.73 MW) and 2.15 cm (1.38 MW) for C. plagiosum Absolute maximum parcel length (cm) scaled with

PAGE 110

99 slight positive allometry in T. semifasciata but with negative allometry in C. plagiosum while maximum parcel length relative to mout h width did not exhibit any trend with size in either species. These va lues correspond well w ith theoretical and empirical measures of suction distance (Alexander, 1967; La uder and Clark, 1984; Muller and Osse, 1984; Wainwright, et al., 2001a) and demonstrate th e severe constraint imposed by the fluid medium on the effective range of suction feed ing. The ecological consequence of this relationship is that suction feeders must approach their prey closely before striking, potentially necessitating a stal king or ambushing predatory be havior (Wainwright, et al., 2001b; Motta, 2004). Both empirical (Muller and Osse, 1984; Higham, et al., 2005) and theoretical work (Weihs, 1980; Muller, et al., 1982) ha s shown that one mechanism of overcoming constraints on suction distance imposed by the aquatic medium is to increase predator velocity, generating a more focused, elongate d region of effective suction. Though the aspect ratio of the ingested pa rcel did not exhibit a trend with growth in either species, it tended to be higher (i.e. more rounded) in C. plagiosum than in T. semifasciata at any given size. In the current study the strike velocity of T. semifasciata averaged 19.106.75 cm s-1 while for C. plagiosum it was only 9.376.06 cm s-1. Additionally, the feeding modality of T. semifasciata is typically ram-domina ted (Ferry-Graham, 1998a; Chapter 2), while the feeding modality of C. plagiosum is strongly suction-dominated (Chapter 3). This corroborates the tendency for increased forward motion of the predator during the strike to influence the shape and le ngth of the parcel of water ingested. A second mechanism to overcome the limited effec tive distance of suction generation in the aquatic medium involves the use of the s ubstrate to truncate and focus flow (P.

PAGE 111

100 Wainwright, pers. comm.). Food was presen ted on the bottom to both species in the current study and flow was cl early ventrally truncated by the substrate (Figure 15). Additionally, the rostrum of both species appeared to limit the dorsal extent of the flow to some degree, an effect not noted in previous studies of species with terminal mouths (Lauder and Clark, 1984; Muller and Osse, 1984; Day, et al., 2005; Gibb and FerryGraham, 2005). The additive effects of truncating flow into the mouth by feeding near the substrate or other aquatic structures and having a s ubterminal mouth may enhance suction capacity by focusing suction force over a smaller area, overcoming limitations imposed by the aquatic medium. This coul d have consequences for prey capture performance, especially in species that live or hunt in close association with the substrate or within crevices. The small degree of variabil ity in suction performance that was not accounted for by growth was often accounted for by relativel y straightforward phys ical relationships described by kinematic variables. In T. semifasciata variation in pulse duration and area were accounted for by variables describing the timing of cranial elevation and the duration that the oral aperture was open. In C. plagiosum variation in pulse duration was accounted for only by variables describing the ti ming and extent of cranial elevation. The importance of cranial elev ation in determining suction feeding performance has been underscored for teleost fishes by Carroll et al (2004). Variability in the distance moved by the food, maximum parcel height, and parcel area in T. semifasciata were accounted for by the extent of maximum gape and the time the gills began to open. Subtle variation in maximum gape can drastically affect the ra te of flow at the or al aperture (FerryGraham and Lauder, 2001; Day, et al., 2005), aff ecting both the force available to impart

PAGE 112

101 motion to the food and the volume of fluid inge sted. This relationship, however, assumes that strike duration does not cha nge over ontogeny, which it does not in T. semifasciata (Chapter 2). With growth accounted for, maximum subambient pressure increased as strike duration decreased in C. plagiosum The low partial regr ession coefficients of kinematic variables after growth had been accounted for, however, suggest that sizeindependent behavioral variabil ity contributes relatively little to overall performance over the course of ontogeny in either species. Ecological and Evolu tionary Implications Substantial changes in suction feedi ng performance occur over ontogeny in both T. semifasciata and C. plagiosum that are primarily attributab le to growth and trends in kinematic variables with growth. In T. semifasciata these developmental trajectories generate an organism that is functiona lly adept at suction feeding, but less so than C. plagiosum Despite physical constraints imposed by the aquatic medium, the propensity to ram feed throughout ontogeny and produce gr eater absolute pressure with growth likely facilitate the known ont ogenetic dietary shift in T. semifasciata toward more functionally diverse and elusive prey items (Talent, 1976; Kao, 2000). Ram feeding has been shown in several species of bony fish to facilitate consumption of more elusive prey (Norton, 1991; Ferry-Graham and Lauder, 2001; Wainwright et al., 2001b) and is typified by the ingestion of a more oval parcel of wa ter (Higham, et al., 2005). As increasingly diverse benthic prey types ar e taken over ontogeny it is also possible that what suction T. semifasciata is able to generate is made more effective by feeding in close association with the substrate, which substantially modifies the focus and anterior extent of flow into

PAGE 113

102 the buccopharyngeal cavity. Shifts in ha bitat use over ontogeny my also influence dietary composition as leopard sharks ar e known to decreasingly utilize shallow, homogenous mud flats and increasingly uti lize more complex, sandyor rocky-bottomed habitats as they age (Barry a nd Cailliet, 1981; Compagno, 1984a). As C. plagiosum becomes functionally, anatomically, and behaviorally canalized to feed via inertial suction over ontogeny, it generates absolutely greater suction but maintains a high aspect ratio, more rounde d parcel of ingested water. These developmental aspects stringently constrain the distance from which C. plagiosum can effectively entrain prey, but pr ovide for a more rapid, brief, and forceful suction capture over ontogeny. Knowledge of the biological ro le for which this feeding mechanism is employed is crucial to making assertions a bout diet over ontogeny in this species. Orectolobiform sharks including C. plagiosum (C. Wilga, pers. obs.), G. cirratum (Compagno, 1984b; P. Motta, pers. obs .), and the epaulette shark Hemiscyllium ocellatum (Heupel and Bennett, 1998) are known to feed by thrusting their heads into soft sediment or crevices in rocky substrate. The capacity to capture prey via suction feeding may be augmented over ontogeny by the development of ambushing, stalking, or excavating behaviors that make available crevice-dw elling or burrowing prey by positioning them close to the mouth. If such changes in foraging behavior do occur over ontogeny, C. plagiosum is anticipated to increas ingly utilize benthic prey over ontogeny as opposed to exploiting prey in the water column. Know ledge of ontogenetic shifts in microhabitat utilization by C. plagiosum could augment this predicti on, but this information is unavailable on the scale needed to make accurate dietary predictions.

PAGE 114

103 Despite substantial differences in the de gree of anatomical complexity of the feeding apparatus of teleost fishes, aquati cally feeding tetrapods, and elasmobranchs, commonalities exist in the behavioral mech anism and effect of subambient buccal pressure generation that are di ctated by physical attr ibutes of the aquatic medium (Lauder and Shaffer, 1993; Wainwright, et al., 2001b; Dean, 2003; Motta, 2004; Gibb and FerryGraham, 2005). Regardless of the number of musculoskeletal components present or the temporal relationships governing the interact ions of these components during a strike, the generation of subambient pre ssure is enhanced by specifi c morphological and behavioral attributes (Wainwright, et al., 2001b; Svanback, et al., 2002; Carroll, et al., 2004; Gibb and Ferry-Graham, 2005; Higham, et al ., 2005). Over ontogeny and phylogeny this means that even large morphological cha nges may produce relatively small suction performance changes (Gibb and Ferry-Graham, 2005), dictating that expansion of the dietary niche of an individual or a species ma y be best accomplished by the learning of novel pre-strike hunting behaviors or the occupation of novel habitats once physical limits of performance imposed by the medium are reached.

PAGE 115

104 References Ackerman, L.T., 1971. Contributions to the biology of the leopard shark, Triakis semifasciata (Girard) in Elkhorn Slough, Monter ey Bay, California. Sacramento State College, Sacramento, pp. 54. Adriaens, D., Aerts, P., Verraes, W., 2001. Ontogenetic shift in mouth opening mechanisms in a catfish (Clariidae, Silurifomes): A response to increasing functional demands. Journal of Morphology 247, 197-216. Alexander, R.M., 1967. The functions and mech anisms of the protru sible upper jaws of some acanthopterygian fish. Zoological Journal of the Linnaean Society 151, 4364. Anderson, C.W., 1993. The modulation of feeding behavior in response to prey type in the frog Rana pipiens Journal of Experimental Biology 179, 1-12. Arendt, J.D., Wilson, D.S., 1997. Optimistic growth: Competition and an ontogenetic niche-shift select for rapid gr owth in pumpkinseed sunfish ( Lepomis gibbosus ). Evolution 51, 1946-1954. Arnold, S.J., 1983. Morphology, performance and fitness. American Zoologist 23, 347361. Barel, C.D.N., 1983. Towards a constructional morphology of cichlid fishes (Teleostei, Perciformes). Netherlands Journal of Zoology 33, 357-424. Barel, C.D., Anker, G.C., Witte, F., Hoogerhoud, R.J., Goldschmidt, T., 1989. Constructional constraint and its ecomorphological implications. Acta Morphologica Neerlands Scandinavia 27, 83-109. Barry, J.P., Cailliet, G.M., 1981. The uti lization of shallow marsh habitats by commercially important fishes in Elkhor n Slough, California. Cal-Neva Wildlife Transactions 1981, 38-47. Benton, M.J., 2004. Vertebrate Paleontology, 3rd Ed. Blackwell Publishers, 455 pp. Bock, W.J., 1980. The definition and recogni tion of biological adaptation. American Zoologist 20, 217-227. Bock, W.J., von Wahlert, G., 1965. Adaptation and the form-function complex. Evolution 19, 269-299. Bohn, T., Amundsen, P., 2001. The competitive edge of an invading specialist. Ecology 82, 2150-2163. Bolnick, D.I., Svanback, R., Fordyce, R.E., Yang, L.H., Davis, J.M., Hulsey, C.D., Forister, M.L., 2003. The ecology of individu als: Incidence and implications of individual specialization. The American Naturalist 161, 1-28. Brickle, P., Laptikhovsky, V., Pompert, J., Bishop, A., 2003. Ontoge netic changes in the feeding habits and dietary overlap betw een three abundant rajid species on the Falkland Islands shelf. Journal of the Marine Biological A ssociation of the UK 83, 1119-1125.

PAGE 116

105 Bryan, J.E., Larkin, P.A., 1972. Food specializ ation by individual trout. Journal of the Fisheries Research Board of Canada 29, 1615-1624. Cabin, R.J., Mitchell, R.J., 2000. To Bonferroni or not to Bonferroni: When and how are the questions. Bulletin of the Ecol ogical Society of America 81, 246-248. Carroll, A.M., Wainwright, P.C., Huske y, S.H., Collar, D.C., Turingan, R.G., 2004. Morphology predicts suction feeding performance in centr archid fishes. Journal of Experimental Biology 207, 3873-3881. Castro, J.I., 2000. The biol ogy of the nurse shark, Ginglymostoma cirratum off the Florida east coast and the Bahama Isla nds. Environmental Biology of Fishes 58, 1-22. Chu, C.T., 1989. Functional design and prey capture dynamics in an ecologically generalized surfperch (Embiotocida e). Journal of Zoology (London) 217, 417440. Compagno, L.J.V., 1984a. FAO Species Catalogu e. Sharks of the World. An Annotated and Illustrated Catalogue of Shark Species Known to Date. Part 2. Carcharhiniformes. FAO Fisheries Synopsis 125 Vol. 4. Compagno, L.J.V., 1984b. FAO Species Catalogue. Sharks of the World. An Annotated and Illustrated Catalogue of Shark Species Known to Date. Part 1. Hexanchiformes to Lamniformes, FAO Fisheries Synopsis 125. Cook, A., 1996. Ontogeny of feeding morphology and kinematics in juvenile fishes: a case study of the cottid fish Clinocottus analis Journal of Experimental Biology 199, 1961-1971. Cortes, E., 1999. Standardized diet compos itions and trophic levels of sharks. ICES Journal of Marine Science 56, 707-717. Cortes, E., Manire, C.A., Hueter, R.E., 1996. Diet, feeding habits, and diel feeding chronology of the bonnethead shark, Sphyrna tiburo in southwest Florida. Coughlin, D.J., 1991. The ontogeny of feeding beha vior of first feed ing Atlantic salmon ( Salmo salar ). Canadian Journal of Fisher ies and Aquatic Science 48, 1896-1904. Cutwa, M.M., Turingan, R.G., 2000. Intralocal ity variation in feeding biomechanics and prey use in Archosargus probatocephalus (Teleostei, Sparidae), with implications for the ecomorphology of fishes. Envir onmental Biology of Fishes 59, 191-198. Day, S.W., Higham, T.E., Ch eer, A.Y., Wainwright, P.C., 2005. Spatial and temporal patterns of water flow generated by suction-feeding bluegill sunfish Lepomis macrochirus resolved by Particle Image Velo cimetry. Journal of Experimental Biology 208, 2661-2671. Dean, M.N., 2003. Suction feeding in the pipid frog, Hymenochirus bettgeri : Kinematic and behavioral considerations. Copeia 2003, 879-886. Dean, M.N., Motta, P.J., 2004. Feeding behavior and kinematics of the lesser electric ray, Narcine brasiliensis (Elasmobranchii: Batoidea). Zoology 107, 171-189. Deban, S.M., 1997. Modulation of prey-capture behavior in the pl ethodontid salamander Ensatina eschscholtzii Journal of Experime ntal Biology 200, 1951-1964. Ebert, D.A., 2002. Ontogenetic changes in the diet of the sevengill shark ( Notorynchus cepedianus ). Marine and Freshwat er Research 53, 517-523. Ellis, J.R., Shackley, S.E., 1995. Ontogenetic changes and sexual dimorphism in the head, mouth, and teeth of the lesser spotted dogfish. Journal of Fish Biology 47, 155-164.

PAGE 117

106 Elsworth, P.G., Seebacher, F., Franklin, C. E., 2003. Sustained swimming performance in crocodiles ( Crocodylus porosus ): Effects of body size and temperature. Journal of Herpetology 37, 363-368. Ferry-Graham, L.A., 1998a. Effects of prey size and motility on prey-capture kinematics in leopard sharks Triakis semifasciata Journal of Experimental Biology 201, 2433-2444. Ferry-Graham, L.A., 1998b. Feeding kine matics of hatchling swellsharks, Cephaloscyllium ventriosum (Scyliorhinidae): the importance of predator size. Marine Biology 131, 703-718. Ferry-Graham, L.A., Lauder, G.V., 2001. Aquatic prey capture in ray-finned fishes: A century of progress and new direc tions. Journal of Morphology 248, 99-119. Ferry-Graham, L.A., Bolnick, D.I., Wa inwright, P.C., 2002. Using functional morphology to examine the ecology and evol ution of speciali zation. Integrative and Comparative Biology 42, 265-277. Ferry-Graham, L.A., Wainwright, P.C., Be llwood, D.R., 2001a. Prey capture in longjawed butterflyfishes (Chaetodontidae): th e functional basis of novel feeding habits. Journal of Experimental Marine Biology and Ecology 256, 167-184. Ferry-Graham, L.A., Wainwright, P.C., La uder, G.V., 2003. Quantification of flow during suction feeding in bluegill sunfishes. Zoology 106, 159-168. Ferry-Graham, L.A., Wainwright, P. C., Westneat, M.W., Bellwood, D.R., 2001b. Modulation of prey capture kinema tics in the cheeklined wrasse Oxycheilinus digrammus (Teleostei : Labridae). Journal of Experimental Zoology 290, 88-100. Fouts, W.R., Nelson, D.R., 1999. Prey capture by the Pacific angel shark, Squatina californica : visually mediated strikes and am bush-site characteristics. Copeia, 304-312. Friel, J.P., Wainwright, P.C., 1999. Evolution of complexity in motor patterns and jaw musculature of tetraodontiform fishes. Journal of Experimental Biology 202, 867880. Galis, F., 1990. Ecological and morphological aspects of changes in food uptake through the ontogeny of Haplochromis piceatus In: Hughes, R.N. (Ed.), Behavioural Mechanisms of Food Selection. Sp ringer-Verlag, Berlin, pp. 281-302. Galis, F., 1993. Interactions between the pha ryngeal jaw apparatus, feeding behaviour, and ontogeny in the cichlid fish, Haplochromis piceatus : A study of morphological constraints in evolutiona ry ecology. Journal of Experimental Zoology 267, 137-154. Galis, F., Terlouw, A., Osse, J.W.M., 1994. The relation between morphology and behavior during ontogenetic and evolutionary changes. Journal of Fish Biology 45, 13-26. Gibb, A.C., Ferry-Graham, L., 2005. Cranial m ovements during suction feeding in teleost fishes: Are they modified to enha nce suction production? Zoology 108, 141-153. Grossman, G.D., Coffin, R., Moyle, P. B., 1980. Feeding ecology of the Bay Goby (Pisces: Gobiidae). Effects of behavioral, ontogenetic, and temporal variation on diet. Journal of Experimental Marine Biology and Ecology 44, 47-59. Hernandez, L.P., 2000. Intraspecific scaling of feeding mechanics in an ontogenetic series of zebrafish, Danio rerio Journal of Experimental Biology 203, 3033-3043.

PAGE 118

107 Hernandez, L.P., Motta, P.J., 1997. Trophic cons equences of differential performance: Ontogeny of oral jaw-crushing performance in the sheepshead, Archosargus probatocephalus (Teleostei, Sparidae). Journal of Zoology (London) 243, 737756. Hernandez, L.P., Barresi, M.J.F, Devoto, S.H., 2002. Functional morphology and developmental biology of zebrafish: Reci procal illumination from an unlikely couple 1. Integrative and Co mparative Biology 42, 222-231. Herrel, A., Aerts, P., Fret, J., DeVree, F ., 1999. Morphology of the feeding system in agamid lizards: Ecological correlates. Anatomical Record 254, 496-507. Herrel, A., van Wassenbergh, S., Wouters, S., Adriaens, D., Aerts, P., 2005. A functional morphological approach to the scaling of the feedi ng system in the African catfish, Clarias gariepinus Journal of Experimental Biology 208, 2091-2102. Heupel, M.R., Bennett, M.B., 1998. Observations on the diet and feeding habits of the epaulette shark, Hemiscyllium ocellatum on Heron Island Reef, Great Barrier Reef, Australia. Marine and Freshwater Research 49, 753-756. Heupel, M.R., Simpfendorfer, C.A., 2002. Estim ation of mortality of juvenile blacktip sharks, Carcharhinus limbatus within a nursery are using telemetry data. Canadian Journal of Fisherie s and Aquatic Science 59, 624-632. Higham, T.E., Day, S.W., Wainwright, P. C., 2005. Sucking while swimming: evaluating the effects of ram speed on suction generation in bluegill sunfish Lepomis macrochirus using digital particle image velo cimetry. Journal of Experimental Biology 208, 2653-2660. Hill, A.V., 1950. The dimensions of animal s and their muscular dynamics. Science Progress (London) 38, 209-230. Huber, D.R., Motta, P.J., 2004. Comparative an alysis of methods for determining bite force in the spiny dogfish Squalus acanthias Journal of Experimental Zoology 301A, 26-37. Hunt Von Herbing, I., 2001. Development of fe eding structures in larval fish with different life histories: wi nter flounder and Atlantic c od. Journal of Fish Biology 59: 767-782. Huskey, S.H., Turingan, R.G., 2001. Variation in prey-resource utiliz ation and oral jaw gape between two populations of largemouth bass, Micropterus salmoides Environmental Biology of Fishes 61, 185-194. Kao, J.S., 2000. Diet, daily ration and gast ric evacuation of the leopard shark ( Triakis semifasciata ). California State University, Hayward. Kent, G.C., Carr, R.K., 2001. Comparat ive Anatomy of the Vertebrates, 9th Ed. McGraw-Hill, Inc., New York, 544 pp. Kiltie, R.A., 1982. Bite force as a basis for niche diversification between rain forest peccaries ( Tayassu tajacu and T. pecari ). Biotropica 14, 188-195. Koehl, M.A.R., 1996. When does morphology matter? Annual Review of Ecology and Systematics 27, 501-542. Krebs, J.R., Turingan, R.G., 2003. Intraspecific variation in gape-pre y size relationships and feeding success during ear ly ontogeny in red drum, Sciaenops ocellatus Environmental Biology of Fishes 66, 75-84.

PAGE 119

108 Kusher, D.I., Smith, S.E., Cailliet, G.M., 1992. Validated age and growth of the leopard shark, Triakis semifasciata with comments on reproduction. Environmental Biology of Fishes 35, 187-203. Lauder, G.V., 1980. Hydrodynamics of prey captu re by teleost fish. Biofluid Mechanics 2, 161-181. Lauder, G.V., 1981. Form and function: struct ural analysis in e volutionary morphology. Paleobiology 7, 430-442. Lauder, G.V., 1985a. Aquatic feeding in lower vertebrates. In: Hildebrand, M., Bramble, D.M., Liem, K.F., Wake, D.B. (Eds.) Functional Vertebrate Morphology. Belknap Press, Cambridge, pp. 210-229. Lauder, G.V., 1985b. Functional morphology of the feeding mechanism in lower vertebrates. Fortschrit te Der Zoologie 30, 179-188. Lauder, G.V., 1990. Functional morphology and systematics: Studying functional patterns in an historical context. An nual Review of Ecology and Systematics 21, 317-340. Lauder, G.V., Clark, B.D., 1984. Water flow patterns during prey capture by teleost fishes. Journal of Experimental Biology 113, 143-150. Lauder, G.V.J., Lanyon, L.E., 1980. Functional anatomy of feeding in the bluegill sunfish, Lepomis macrochirus : in vivo measurements of bone strain. Journal of Experimental Biology 84, 33-55. Lauder, G.V., Prendergast, T., 1992. Kinematics of aquatic prey capture in the snapping turtle Chelydra serpentina Journal of Experime ntal Biology. 164, 55-78. Lauder, G.V., Shaffer, H.B., 1993. Aquatic vertebrate skull design. In: Hanken, J., Hall, B.K. (Eds.), The skull. The University of Chicago Press, Chicago, pp. 113-149. Liem, K.F., 1978. Modulatory multiplicity in th e functional repertoire of the feeding mechanism in cichlid fishes. Journal of Morphology 158, 323-360. Liem, K.F., 1980a. Adaptive significance of intr aand interspecific differences in the feeding repertoires of cichlid fi shes. American Zoologist 20, 295-314. Liem, K.F., 1980b. Acquisition of energy by teleosts: adaptive mechanisms and evolutionary patterns. In: Ali, M.A. (E d.), Environmental Physiology of Fishes. Plenum Publishing Corp., New York, pp. 299-334. Liem, K.F., 1990. Aquatic versus terrestria l feeding modes: possible impacts on the trophic ecology of vertebrates. American Zoologist 30, 209-221. Liem, K.F., 1993. Ecomorphology of the teleost s kull. In: Hanken, J., Hall, B.K. (Eds.), The Skull. Functional and evolutionary mechanisms. The University of Chicago Press, Chicago, pp. 422-452. Liem, K.F., Summers, A.P., 2000. Integration of versatile functional design, population ecology, ontogeny and phylogeny. Netherlands Journal of Zoology 50, 245-259. Lima-Junior, S.E., Goitein, R., 2003. Ontogeneti c diet shifts of a Neotropical catfish, Pimelodus maculatus (Siluriformes, Pimelodidae): An ecomorphological approach. Environmental Biology of Fishes 68, 73-79. Losos, J.B., 1990. Ecomorphology, performance cap ability, and scaling of West Indian Anolis lizards: An evolutionary anal ysis. Ecological Monographs 60, 369-388. Lowe, C.G., Wetherbee, B.M., Crow, G.L., Te ster, A.L., 1996. Ontogenetic dietary shifts and feeding behaviors of the tiger shark, Galeocerdo cuvier in Hawaiian waters. Environmental Biology of Fishes 47, 203-211.

PAGE 120

109 Manire, C.A., Gruber, S.H., 1993. A preliminary report on the rate of natural mortality of juvenile lemon sharks, Negaprion brevirostris NOAA, pp. 65-71. Matott, M.P., Motta, P.J., Hueter, R.E., 2005. Modulation in feeding kinematics and motor pattern of the nurse shark Ginglymostoma cirratum Environmental Biology of Fishes 74, 163-174. McGowan, C., 1988. Differential development of the rostrum and mandible of the swordfish ( Xiphias gladius ) during ontogeny and its possible functional significance. Canadian Journal of Zoology 66, 496-503. Miller, D.J., Lea, R.N., 1972. Guide to the coas tal marine fishes of California. California Department of Fish and Game, Fishery Bulletin 157, 1-235. Moran, M.D., 2003. Arguments for rejecting th e sequential Bonferroni in ecological studies. Oikos 100, 403-405. Morse, D.H., 1980. Behavioral Mechanisms in Ecology. Harvard University Press, Cambridge, MA, 383 pp. Motta, P.J., 2004. Prey capture behavior and feeding mechanics of elasmobranchs. In: Carrier, J.C., Musick, J.A., Heithaus, M. R. (Eds.), Biology of Sharks and Their Relatives. CRC Press, Boca Raton, FL, pp. 596. Motta, P.J., Kotrschal, K.M., 1992. Correla tive, experimental, and comparative evolutionary approaches in ecomor phology. Netherlands Journal of Zoology 42, 400-415. Motta, P.J., Wilga, C.A.D., 1995. Anatomy of the feeding apparatus of the lemon shark, Negaprion brevirostris Journal of Morphology 226, 309-329. Motta, P.J., Wilga, C.D., 1999. Anatomy of th e feeding apparatus of the nurse shark, Ginglymostoma cirratum Journal of Morphology 241, 1-29. Motta, P.J., Wilga, C.D., 2001. Advances in th e study of feeding behaviors, mechanisms, and mechanics of sharks. Environmental Biology of Fishes 60, 131-156. Motta, P.J., Hueter, R.E., Tricas, T.C., 1991. An electromyographic analysis of the biting mechanism of the lemon shark, Negaprion brevirostris : functional and evolutionary implications. Journal of Morphology 201, 55-69. Motta, P.J., Clifton, K.B., Hernandez, P., Eggold, B.T., 1995a. Ecomorphological correlates in ten species of subtropical seagrass fish es: diet and microhabitat utilization. Environmental Biology of Fishes 44, 37-60. Motta, P.J., Hueter, R.E., Tricas, T.C., Su mmers, A.P., 2002. Kinematic analysis of suction feeding in the nurse shark, Ginglymostoma cirratum (Orectolobiformes, Ginglyostomatidae). Copeia 1, 24-38. Motta, P.J., Tricas, T.C., Hueter, R.E., Summers, A.P., 1997. Feeding mechanism and functional morphology of the jaws of the lemon shark Negaprion brevirostris (Chondrichthyes, Carcharhinidae). Jour nal of Experimental Biology 200, 27652780. Motta, P.J., Clifton, K.B., Hernandez, P ., Eggold, B.T., Giordano, S.D., Wilcox, R., 1995b. Feeding relationships among nine sp ecies of seagrass fishes of Tampa Bay, Florida. Bulletin of Marine Science 56, 185-200. Muller, M., Osse, J.W.M., 1984. Hydrodynamics of suction feeding in fish. Transactions of the Zoological Society of London 37, 51-135. Muller, M., Osse, J.W.M., Verhagen, J.H. G., 1982. A quantitative hydrodynamical model of suction feeding in fish. Jour nal of Theoretical Biology 95, 49-79.

PAGE 121

110 Nemeth, D.H., 1997. Modulation of attack behavior and its effect on feeding performance in a trophic generalist fish, Hexagrammos decagrammus Journal of Experimental Biology 200, 2155-2164. Norton, S.F., 1991. Capture success and diet of cottid fishes: the role of predator morphology and attack kinematics. Ecology 72, 1807-1819. Norton, S.F., Brainerd, E.L., 1993. Converg ence in the feeding mechanics of ecomorphologically similar species in the Centrarchidae and Cichlidae. Journal of Experimental Biology 176, 11-29. Norton, S.F., Luczkovich, J.J., Motta, P.J., 1995. The role of ecomorphological studies in the comparative biology of fishes. Envi ronmental Biology of Fishes 44, 287-304. Olson, M.H., 1996. Ontogenetic niche shifts in largemouth bass: variability and consequences for first-year growth. Ecology 77, 179-190. O'Reilly, J.C., Linstedt, S.L., Nishikawa, K. C., 1993. The scaling of feeding kinematics in toads (Anura: Bufonidae) American Zoologist 33, 147A. Osse, J.W.M., 1990. Form changes in fish la rvae in relation to changing demands of function. Netherlands Journal of Zoology 40, 362-385. Quinn, G.P., Keough, M.J., 2002. Experimental De sign and Data Analysis for Biologists. Cambridge University Press, Cambridge, United Kingdom, 537 pp. Reilly, S.M., 1995. The ontogeny of aquatic feeding behavior in Salamandra salamandra : Stereotypy and isometry in f eeding kinematics. Journal of Experimental Biology 198, 701-708. Reilly, S.M., 1996. The metamorphosis of feeding kinematics in Salamandra salamandra and the evolution of terrestrial feeding behavior. Journal of Experimental Biology 199, 1219-1227. Reilly, S.M., Lauder, G.V., 1988. Ontogeny of aquatic feeding pe rformance in the Eastern newt, Notophthalamus viridescens (Salamandridae). Copeia 1988, 87-91. Reilly, S.M., Lauder, G.V., 1989. Physiologi cal bases of feeding behavior in salamanders: do motor patterns vary with prey type? Journal of Experimental Biology 141, 343-358. Reilly, S.M., Lauder, G.V., 1990. Metamorphosis of cranial design in tiger salamanders ( Ambystoma tigrinum ): a morphometric an alysis of ontogenetic change. Journal of Morphology 204, 121-137. Reilly, S.M., Lauder, G.V., 1992. Morphology, be havior, and evolut ion: Comparative kinematics of aquatic feeding in salama nders. Brain, Behavior, and Evolution 40, 182-196. Richard, B.A., Wainwright, P.C., 1995. Scali ng the feeding mechanism of largemouth bass ( Micropterus salmoides ): kinematics of prey capture. Journal of Experimental Biology 198, 419-433. Robinson, M.P., Motta, P.J., 2002. Patterns of growth and the effects of scale on the feeding kinematics of the nurse shark ( Ginglymostoma cirratum ). Journal of Zoology 256, 449-462. Russo, R.A., 1975. Observations on the food habits of leopard sharks ( Triakis semifasciata ) and brown smoothounds ( Mustelus henlei ). California Fish and Game 61, 95-103. Sanderson, S.L., 1990. Versatility and specializ ation in labrid fishes: ecomorphological implications. Oecologia 84, 272-279.

PAGE 122

111 Sanderson, S.L., 1991. Functional stereotypy an d feeding performance correlated in a trophic specialist. Func tional Ecology 5, 795-803. Sanford, C.P., Wainwright, P.C., 2002. Use of sonomicrometry demonstrates the link between prey capture kinematics and sucti on pressure in largemouth bass. Journal of Experimental Biology 205, 3445-3457. Sass, G.G., Motta, P.J., 2002. The effects of satiation on prey capture kinematics in the largemouth bass, Micropterus salmoides Environmental Biology of Fishes 65, 441-454. Schwenk, K., 2000. Tetrapod feeding in the context of vertebrate morphology. In: Schwenk, K. (Ed.), Feeding: Form, function, and evolution in tetrapod vertebrates. Academic Press, San Diego, pp. 3-20. Shaffer, H.B., Lauder, G.V., 1985a. Aquatic prey capture in Ambystomatid salamanders: patterns of variation in muscle ac tivity. Journal of Morphology 183, 273-284. Shaffer, H.B., Lauder, G.V., 1985b. Patterns of variation in aquatic Ambystomatid salamanders: kinematics of the feed ing mechanism. Evolution 39, 83-92. Shirai, S., Nakaya, K., 1992. Functional mor phology of feeding apparatus of the cookiecutter shark, Isistius brasiliensis (Elasmobranchii, Dalatiinae). Zoological Science 9, 811-821. Sibbing, F.A., Nagelkerke, L.A.J., 2001. Res ource partitioning by Lake Tana barbs predicted from fish morphometrics and pr ey characteristics. Reviews in Fish Biology and Fisheries 10, 393-437. Sims, D.W., 2000. Filter-feeding and crui sing swimming speeds of basking sharks compared with optimal models: they filte r-feed slower than predicted for their size. Journal of Experimental Marine Biology and Ecology 249, 65-76. Skulason, S., Noakes, D.L.G., Snorrason, S.S., 1989. Ontogeny of trophic morphology in four sympatric morphs of Arctic charr Salvelinus alpinus in Thingvallavatn, Iceland. Biological Journal of the Linnean Society 38, 281-301. Stoner, A.W., Livingston, R.J., 1984. Ont ogenetic patterns in diet and feeding morphology in sympatric sparid fishes from seagrass meadows. Copeia 1984, 174-187. Summers, A.P., 2000. Stiffening the stingray skeleton: An investigation of durophagy in myliobatid stingrays (Chondrichthyes, Batoidea, Myliobatidae). Journal of Morphology 243, 113-126. Summers, A.P., Koob, T.J., Brainerd, E.L., 1998a. Stingray jaws strut their stuff. Nature 395, 450-451. Summers, A.P., Darouian, K.F., Richmond, A.M., Brainerd, E.L., 1998b. Kinematics of aquatic and terrestrial prey capture in Terrapene carolina with implications for the evolution of feeding in cryptodire turtles. Jour nal of Experimental Zoology 281, 280-287. Svanback, R., Wainwright, P.C., Ferry-Gra ham, L.A., 2002. Linking cranial kinematics, buccal pressure, and suction feeding performance in largemouth bass. Physiological and Biochemical Zoology 75, 532-543. Talent, L.G., 1976. Food habits of the leopard shark, Triakis semifasciata in Elkhorn Slough, Monterey Bay, California. Ca lifornia Fish and Game 62, 286-298.

PAGE 123

112 Van Damme, J., Aerts, P., 1997. Kinematic s and functional morphology of aquatic feeding in Australian snake-necked turt les (Pleurodira: Chelodina). Journal of Morphology 233, 113-125. Van Valen, L., 1965. Morphological variation an d width of ecological niche. American Naturalist 99, 377-390. Van Wassenbergh, S., Aerts, P., Herrel, A., 2005. Scaling of suction-feeding kinematics and dynamics in the African catfish, Clarias gariepinus Journal of Experimental Biology 208, 2103-2114. Verwaijen, D., Van Damme, R., Herrel, A., 200 2. Relationships between head size, bite force, prey handling efficiency and di et in two sympatric lacertid lizards. Functional Ecology 16, 842-850. Wainwright, P.C., 1994. Functi onal morphology as a tool in ecological research. In: Wainwright, P.C., Reilly, S.M. (Eds.) Ecological Morphology: Integrative Organismal Biology. The University of Chicago Press, Chicago, pp. 42-59. Wainwright, P.C., Lauder, G.V., 1986. Feeding bi ology of sunfishes: pa tterns of variation in the feeding mechanism. Zoological Journal of the Linnean Society 88, 217-228. Wainwright, P.C., Westneat M.W., Bellwood, D.R., 2000. Linki ng feeding behavior and jaw mechanics in fishes. In: Domenici, P ., Blake, R.W. (Eds.), Biomechanics in Animal Behavior. BIOS Scientific Publishers Ltd., Oxford, pp. 207-221. Wainwright, P.C., Ferry-Graham, L.A., Walt zek, T.B., Hulsey, C.D., Carroll, A.M., Svanback, R., 2001a. Evaluating suction feed ing performance in fishes. American Zoologist 41, 1617-1617. Wainwright, P.C., Ferry-Graha m, L.A., Waltzek, T.B., Carroll, A.M., Hulsey, C.D., Grubich, J.R., 2001b. Evaluating the use of ram and suction during prey capture by cichlid fishes. Journal of Experimental Biology 204, 3039-3051. Weihs, D., 1980. Hydrodynamics of suction feed ing of fish in motion. Journal of Fish Biology 16, 425-433. Werner, E.E., Gilliam, J.F., 1984. The ontogenetic niche and species in teractions in sizestructured populations. Annual Review of Ecology and Systematics 15, 393-425. Werner, E.E., Mittelbach, G.G., Hall, D.J., 1981. The role of foraging profitability and experience in habitat use by th e bluegill sunfish. Ecology 62, 116-125. Westneat, M.W., 1994. Transmission of force and velocity in the feeding mechanisms of labrid fishes (Teleostei, Pe rciformes). Zoomorphology 114, 103-118. Westneat, M., 2003. A biomechanical model for analysis of muscle force, power output and lower jaw motion in fishes. J ournal of Theoretical Biology, 269-281. Wilga, C.A.D., Motta, P.J., 1998a. Feeding mechanism of the Atlantic guitarfish Rhinobatos lentiginosus : Modulation of kinematic a nd motor activity. Journal of Experimental Biology 201, 3167-3184. Wilga, C.D., Motta, P.J., 1998b. Conservation an d variation in the feeding mechanism of the spiny dogfish Squalus acanthias Journal of Experi mental Biology 201, 13451358. Wilga, C.D., Motta, P.J., 2000. Durophagy in sharks: feeding mechanics of the hammerhead Sphyrna tiburo Journal of Experimental Biology 203, 2781-2796. Wilga, C.A.D., Wainwright, P.C., Motta, P.J., 2000. Evolution of jaw depression mechanics in aquatic vertebrates: in sights from Chondrichthyes. Biological Journal of the Linnean Society 71, 165-185.

PAGE 124

113 Wilga, C.D., Hueter, R.E., Wainwright, P. C., and Motta, P.J., 2001. Evolution of upper jaw protrusion mechanisms in elasm obranchs. American Zoologist 41, 12481257. Wu, E.H., 1993. The functional morphology of feeding mechanisms in orectolobiform sharks. Harvard University, Cambridge, Massachusetts, pp. 177. Wu, E.H., 1994. Kinematic analys is of jaw protrusion in orectolobiform sharks: a new mechanism for jaw protrusion in elas mobranchs. Journal of Morphology 222, 175-190. Zar, J.H., 1999. Biostatistical Analysis. Pren tice Hall, Upper Saddle River, New Jersey, 919 pp.

PAGE 125

About the Author David Lowry was born in St. Charles, Misso uri and moved to Boise, Idaho at age five. Encouraged by his parents’ love of nature, David read ra venously about diverse natural phenomena and spent time outdoors wh enever possible. The mysterious ocean intrigued David from a young age and his search for knowledge soon turned toward uncovering the secrets of the animals that dw elt therein. After graduating from Borah High School in 1996, David attended Hawai’i Pa cific University in Honolulu, Hawai’i where he obtained a Bachelors Degree in Ma rine Biology in 1999. It was during a winter visit home to Boise that David met his future wife Sonia. Immediately after graduation David and Sonia moved to Tampa, Florida to attend the University of South Florida, where David earned his Ph.D. in Biology in 2005. After completing his degree, David moved to Olympia, Washington to pursue a career in applied biological science.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001912520
003 fts
005 20071010090903.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 071010s2005 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001312
040
FHM
c FHM
035
(OCoLC)173959356
049
FHMM
090
QH307.2 (ONLINE)
1 100
Lowry, David C.
4 245
The early ontogeny of feeding in two shark species :
b developmental aspects of morphology, behavior, and performance
h [electronic resource] /
by David C. Lowry.
260
[Tampa, Fla] :
University of South Florida,
2005.
3 520
ABSTRACT: Early ontogeny is a time of rapid anatomical and behavioral development in most organisms. The degree of synchrony between form and function during this period, and the concomitant performance consequences, can strongly impact individual survival. Understanding the development of feeding during early ontogeny is important because nutrient acquisition universally influences organismal biology. A one-year, longitudinal feeding study was conducted for two elasmobranch species that were selected for their disparate morphology, behavior, and habitat: the whitespotted bambooshark Chiloscyllium plagiosum and the leopard shark Triakis semifasciata. To quantify changes in cranial morphology, external attributes of the feeding apparatus were measured weekly. Additionally, specimens were dissected to examine trends in the growth of select muscles and the volume of the buccal cavity. To quantify feeding behavior, individuals were observed weekly using high-speed digital cameras^ as they consumed various food types. Suction performance was evaluated using particle image velocimetry and direct measurements of suction pressure. The cranial morphology of C. plagiosum exhibited primarily isometric growth while the cranial morphology of T. semifasciata was dominated by allometric growth. Allometric increases were noted in the cross-sectional area of every muscle examined in both species, though the primary hyoid depressor, the coracohyoideus, hypertrophied to a greater degree in C. plagiosum. Although intra-individual differences throughout ontogeny complicated comparison, modulation in response to food attributes was clearly evident in T. semifasciata but broadly absent in C. plagiosum. Over ontogeny C. plagiosum generated allometrically greater suction while T. semifasciata generated relatively less. The shape of the parcel of water ingested during feeding did not change over ontogeny in either species. The capacity to perform diverse feeding behaviors thro ughout ontogeny is not constrained in T. semifasciata but tends to be stereotyped and accompanied by enhanced performance in C. plagiosum. A functionally generalized feeding apparatus and repertoire may benefit T. semifasciata by allowing the use of diverse feeding behaviors in variable environments, such as estuaries, over ontogeny. Morphological and behavioral conservation of the feeding apparatus throughout ontogeny, however, may allow C. plagiosum to exploit taxonomically varied crevice-dwelling reef organisms using a single specialized behavior.
502
Dissertation (Ph.D.)--University of South Florida, 2005.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 113 pages.
Includes vita.
590
Adviser: Philip Motta, Ph.D.
653
Ecomorphology.
Aquatic feeding.
Modulation.
Variability.
Developmental trajectory.
690
Dissertations, Academic
z USF
x Biology
Doctoral.
773
t USF Electronic Theses and Dissertations.
0 856
u http://digital.lib.usf.edu/?e14.1312