Sexual Dimorphism and Size-Related Changes in Body Shape In

Home , Sexual dimorphism, Surfperch

SEXUAL DIMORPHISM AND SIZE-RELATED CHANGES IN BODY SHAPE IN

TULE PERCH, A NATIVE CALIFORNIA LIVE-BEARING FISH

A Thesis

Presented to the faculty of the Department of Biological Sciences

California State University, Sacramento

Submitted in partial satisfaction of the requirements for the degree of

MASTER OF SCIENCE

Biological Sciences

(Ecology, Evolution, and Conservation)

Elizabeth Sarah Parvis

SPRING 2016

Elizabeth Sarah Parvis

SEXUAL DIMORPHISM AND SIZE-RELATED CHANGES IN BODY SHAPE IN

TULE PERCH, A NATIVE CALIFORNIA LIVE-BEARING FISH

A Thesis

Elizabeth Sarah Parvis

Approved by:

______, Committee Chair Ronald M. Coleman, Ph.D.

______, Second Reader Brett Holland, Ph.D.

______, Third Reader Jimmy Pitzer, Jr., Ph.D.

______Date

iii

Student: Elizabeth Sarah Parvis

I certify that this student has met the requirements for format contained in the University format manual, and that this thesis is suitable for shelving in the Library and credit is to be awarded for the thesis.

______, Graduate Coordinator ______Jamie Kneitel, Ph.D. Date

Department of Biological Sciences

Abstract

SEXUAL DIMORPHISM AND SIZE-RELATED CHANGES IN BODY SHAPE IN

TULE PERCH, A NATIVE CALIFORNIA LIVE-BEARING FISH

Elizabeth Sarah Parvis

Sexual dimorphism is prevalent in animal taxa and can result from a variety of factors including natural selection, sexual selection, and the sexes occupying dimorphic niches. Sexual selection typically acts on males via female choice or intra-sexual competition, while the dimorphic niche hypothesis usually applies to females due to reproductive constraints. Fishes exhibit sexual dimorphism as variation in size, body proportions, fins, and/or color.

The tule perch, Hysterocarpus traski, is a viviparous (live-bearing) and externally monomorphic, internally fertilizing fish. Despite appearing monomorphic, males and females are expected to differ in body shape because the different reproductive roles occupied by the sexes should influence patterns of selection and, ultimately, lead to differences in morphology. It is also expected that the body shape of a tule perch changes as the fish matures, especially the abdomen shape, as reproductive structures develop.

Dimorphic variation among tule perch was investigated using morphometric approaches. The objectives of this study were to determine if tule perch exhibit (1) v

sexual dimorphism in body shape, (2) size-related changes in body shape, and (3) sexual size dimorphism. Objectives 1 and 2 were accomplished by describing and testing mean body shape differences due to sex and size using a geometric morphometric approach.

The third objective was accomplished using a traditional morphometric approach.

Results of the Multivariate Analysis of Variance indicated that tule perch exhibited significant sexual dimorphism of body shape. Results of the Canonical Variates

Analysis demonstrated that the most effective discriminators between the sexes were in the mid-body and caudal peduncle regions. Specifically, females were narrower through the caudal peduncle and mid-body and had anterior anal fin insertion points that were more posteriorly-located than males.

Additionally, results of the multivariate regression indicated that tule perch exhibited size-related changes in body shape. The deformation grid obtained from the regression of shape to size illustrated that larger fish were deeper through the mid-body and blunter through the snout than smaller fish. The caudal peduncle was wider and more rounded and the pectoral fin was shifted ventrally in larger fish, and the eyes of larger

fish were relatively smaller and located higher on the body than those of smaller fish.

Finally, results of the two-tailed t-test indicated that tule perch did not exhibit sexual size dimorphism with regard to standard length, but did exhibit sexual size dimorphism with regard to weight.

______, Committee Chair Ronald M. Coleman, Ph.D.

______Date

vii

ACKNOWLEDGEMENTS

I would like to thank Dr. Ron Coleman for introducing me to tule perch and geometric morphometrics. I would like to acknowledge Dr. Brett Holland and Dr. Jimmy

Pitzer, Jr. for being a part of my advisory committee. I would like to thank Dr. Rob Titus for his support in obtaining a California Department of Wildlife (CDFW) Scientific

Collecting Permit. I would like to express extreme gratitude to CDFW staff Monty

Currier, Steve Baumgartner, and Lisa Corvington for their assistance in collecting and transporting tule perch. I would like to thank Dr. Samantha Hens (CSUS Department of

Anthropology) for her encouragement and support regarding geometric morphometric methodology. I would like to thank my brothers, Andrew and David Parvis, for their help with processing tule perch for data collection. Finally, I would like to thank my parents and Farhat Bajjaliya for their love, support, and patience throughout this process.

viii

TABLE OF CONTENTS Page

Acknowledgements ...... viii

List of Tables ...... x

List of Figures ...... xi

INTRODUCTION ...... 1

METHODS ...... 6

Specimen Collection ...... 6

Specimen Preparation ...... 7

Image Acquisition ...... 7

Landmark Identification and Digitization ...... 8

Shape Coordinates ...... 11

Statistical Analyses ...... 13

RESULTS ...... 14

Sexual Dimorphism of Body Shape ...... 14

Size-related Changes in Body Shape ...... 19

Sexual Size Dimorphism ...... 29

DISCUSSION ...... 31

Sexual Dimorphism of Body Shape ...... 31

Size-related Changes in Body Shape ...... 33

Sexual Size Dimorphism ...... 34

CONCLUSIONS...... 37

Literature Cited ...... 38

LIST OF TABLES

Table Page

1. Jack-knife test of group assignments ...... 16

2. Tule perch size data...... 30

LIST OF FIGURES

Figure Page

1. Landmarks used for geometric morphometric shape analysis of

tule perch...... 10

2. Generalized Procrustes Analysis (GPA) of 130 male (blue) and 139

female (pink) tule perch...... 12

3. Canonical Variates Analysis (CVA) of the sexes...... 15

4. Deformation grid of the first canonical variate (CV1)...... 17

5. Deformation grid of the first canonical variate (CV1) regressed...... 18

6. Deformation grid of regression of entire tule perch sample...... 21

7. Deformation grid of regression of female tule perch...... 22

8. Scree plot of percentage of variance explained versus principal

component number for female tule perch...... 23

9. Deformation grid of the first principal component (PC1) for female

tule perch...... 24

10. Deformation grid of regression of male tule perch...... 25

11. Scree plot of percentage of variance explained versus principal

component number for male tule perch...... 26

12. Deformation grid of the first principal component (PC1) for male

tule perch...... 27

xii

13. Deformation grid of the second principal component (PC2) for male

tule perch...... 28

xiii

INTRODUCTION

Sexual dimorphism is prevalent in animal taxa and can result from a variety of factors including natural selection, sexual selection, and the sexes occupying dimorphic niches (Blanckenhorn 2005; Kitano et al. 2007; Hassell et al. 2012). Sexual selection typically acts on males via female choice or intra-sexual competition, while the dimorphic niche hypothesis usually applies to females due to reproductive constraints

(Hedrick and Temeles 1989; Herler et al. 2010). Fishes exhibit sexual dimorphisms as variations in size, body proportions, fins, and/or color (Echeverria 1986).

The tule perch, Hysterocarpus traski, is the only freshwater member of the surfperch family Embiotocidae (Breder and Rosen 1966) and is an externally monomorphic, internally fertilizing fish. Additionally, the tule perch is endemic to

California (Moyle 2002) and is the only native, viviparous (live-bearing) freshwater fish species in the state (Baltz and Moyle 1981). Three subspecies of tule perch are recognized in three drainage systems: Clear Lake (H. t. lagunae), Russian River (H. t. pomo), and Sacramento-San Joaquin (H. t. traski) (Baltz and Moyle 1981; Baltz and

Loudenslager 1984). Hysterocarpus t. traski can be found as far north as Pittville on the

Pit River in Lassen County, and as far south as the Tulare basin in Tulare County (Bundy

1970).

In Sacramento-San Joaquin tule perch populations, males and females school separately, except during the mating season (Bundy 1970). Breeding occurs from May to late October (Bryant 1977). During this time, females mate with multiple males (Moyle

2002; Liu et al. 2013), but delay fertilization until January (Bundy 1970; Bryant 1977).

Between April and June, female tule perch give birth to one brood of fully-formed young

(Bundy 1970; Baltz and Moyle 1982). Tule perch are not born sexually mature, but reach maturity within a few weeks (Bryant 1977; cf. Micrometrus minimus, Warner and Harlan

1982).

Of the 23 surfperch species (Baltz 1984), tule perch are the only surfperch with permanent, non-sexual color dimorphism (Phelps 1989). Additionally, in this relatively monomorphic species, the sexes only differ outwardly in appearance in the region of the anal fin. Males have semilunar depressions in the body surface above the anal fin and a fleshy, glandular enlargement at its base (Bundy 1970; Bryant 1977). Sexual size dimorphism is not well documented in tule perch. It has been reported that males are significantly smaller than females (Phelps 1989), but some data suggests otherwise

(Bundy 1970, Parvis unpub. data).

Despite appearing monomorphic, males and females are expected to differ in body shape because the different reproductive roles occupied by the sexes should influence patterns of selection and, ultimately, lead to differences in morphology

(Casselman and Schulte-Hostedde 2004). It is also expected that the body shape of a tule perch changes as the fish matures, especially the abdomen shape as reproductive structures develop (Hassell et al. 2012). This variation can be investigated using a morphometric approach.

Morphometrics, the quantitative study of biological shape variation, plays an integral role in the study of organismal biology (Adams 1999; Webster and Sheets 2010).

Variation in body shape among and within species may indicate differences in functional

3 roles of parts, selective pressures, and/or growth processes (Zelditch et al. 2012).

Additionally, studying morphology can provide insight into the natural history, facilitate the identification of morphological-ecological relationships, and provide data that supports more informed inferences regarding the evolution of a particular species

(Echeverria 1985; Adams 1999).

A traditional morphometric approach is characterized by the application of univariate and multivariate statistical analyses to sets of variables such as body lengths, widths, and/or depths of organisms (Rohlf and Marcus 1993; Adams et al. 2004; Webster and Sheets 2010; e.g., Echeverria 1986; Lenarz and Echeverria 1991; Oliveira and

Almada 1995). While this approach is still occasionally used to examine body shape

(e.g., Casselman and Schulte‐Hostedde 2004; Coban et al. 2011), it has been supplanted by a newer, more powerful method: geometric morphometrics.

First introduced by Bookstein (1982), geometric morphometrics uses a configuration of landmarks to summarize shape (Webster and Sheets 2010). Inherently multidimensional, this robust approach is advantageous compared to traditional morphometrics because it supplies information regarding the spatial relationship among landmarks on an organism, taking the geometry of the entire organism into consideration, and provides visualization of shape and shape variation (Adams et al. 2004; Kassam et al.

2004; Webster and Sheets 2010). In the first ten years after Bookstein's (1982) original proposal of a geometric approach, considerable work was completed to develop and expand this field (see Strauss and Bookstein 1982; Bookstein 1986; Rohlf and Slice

1990; Rolhf and Marcus 1993). The geometric morphometric method continues to be

4 widely used to investigate questions regarding the effects of ecology and evolution on body shape (Zelditch et al. 2012).

Geometric morphometric approaches have been used to study a variety of taxa including fishes. Most geometric morphometric studies of fish have investigated stock identification (e.g., Cadrin and Silva 2005), phylogenetic relationships (e.g., Fink and

Zelditch 1995; Fink and Zelditch 1996), and the effects of predation (e.g., Hassell et al.

2012; Araujo et al. 2014) or food sources (e.g., Trapani 2003; Frederich et al. 2008) on body shape. Additionally, investigations of sexual dimorphism and ontogeny are popular in geometric morphometric studies with the majority of work focusing on egg-laying fishes (e.g., Douglas 1993; Zelditch and Fink 1995; Loy et al. 1998; Reis et al. 1998;

Hood and Heins 2000; Kassam et al. 2004; Kitano et al. 2007; Herler et al. 2010; Kitano et al. 2012) rather than live-bearing species (e.g., Woods 2007; Hassell et al. 2012;

Araujo et al. 2014).

Only one study has investigated sexual dimorphism of body shape in a viviparous fish using geometric morphometric methods. In this study, the shape and relative size of the abdomen in Brachyrhaphis rhabdophora was identified as the primary morphological difference between the sexes and was attributed to the different roles played by males and females in a live-bearing reproductive mode (Hassell et al. 2012). Additional work has been competed in live-bearing species using the traditional morphometric approach.

Sexual dimorphism was examined in four rockfish species that do not show obvious sexual dimorphism (Echeverria 1986). While females reached larger sizes than males in all four species, males had larger eyes and longer pectoral fins. This study was expanded

5 to include 30 additional species of rockfish (Lenarz and Echeverria 1991). The authors determined similar results regarding sexual size dimorphism: females were larger than males in 27 of the 34 species and males had greater pectoral fin length and eye size.

Additionally, it was discovered that the head, upper jaw, and longest dorsal spine of males were larger than those of females (Lenarz and Echeverria 1991).

Geometric morphometric studies of ontogeny in viviparous fishes have been conducted as well. In a study on B. rhabdophora, juveniles had relatively larger heads and more streamlined bodies than their adult counterparts, illustrating the selection for early development of sensory and feeding systems (Hassell et al. 2012). Additionally, it was determined that ontogenetic changes in Cymatogaster aggregate, the shiner surfperch, were strongest at the juvenile stage and were better predicted by habitat rather than geography (Woods 2007). This is the only investigation of ontogeny using geometric morphometrics in surfperch.

The objectives of the present study were to determine if tule perch exhibit (1) sexual dimorphism in body shape, (2) size-related changes in body shape, and (3) sexual size dimorphism. Objectives 1 and 2 were accomplished by describing and testing mean body shape differences due to sex and size using a geometric morphometric approach.

The third objective was accomplished using a traditional morphometric approach.

METHODS

Specimen Collection

Approximately 275 tule perch were collected from Baum Lake, Shasta County via an electrofishing boat on July 30, 2014 (California Department of Fish and Wildlife

SCP# 13031). This reservoir was selected due to the abundance of tule perch and an offer of collection assistance (M. Currier and S. Baumgartner) from the California Department of Fish and Wildlife. Additionally, with only 5.6 kilometers of shoreline and 36 surface hectares, Baum Lake is a relatively small waterbody, reducing the likelihood of significant ecophenotypic variation within the collected fish. Tule perch were collected over a three-hour period from the western and northeastern shores of the lake.

The tule perch were transported to Dr. Ron Coleman’s Evolutionary Ecology of

Fishes Laboratory at California State University, Sacramento in nine, 10-L styrofoam coolers equipped with bubblers. Once in the lab, the tule perch were housed in three 120-

L aquaria. Due to unknown reasons, the tule perch did not thrive in the lab and died over the next few weeks.

Dead fish were removed from the aquaria and stored in sealable plastic bags in the freezer. Fish were often preserved within minutes of death and never allowed to sit longer than 24 hours. Freezing was used as the preservation method because it causes less change in fish specimen length and weight than other preservation methods (DiStefano et al. 1994). Additionally, the freezing and thawing process does not cause significant shrinkage or swelling that would affect geometric morphometric analyses (Cadrin and

Silva 2005).

It is important to note that tule perch do not attain sexual maturity until they reach a minimum standard length: 58 and 55mm for males and females, respectively (Bryant

1977). Almost all of the fish utilized in this study were considered sexually mature, even if only newly so. However, because the standard length of the specimens ranged widely

(males: 60.4 – 131.4 mm; females: 42.0 – 138.1 mm), size-related changes were still expected to be evident.

Specimen Preparation

Tule perch were removed from the freezer and thawed in a plastic bag in a bucket of cool water for approximately 20 minutes. Each fish was assigned a three-digit identification number (e.g., 001, 002, etc.) for photograph tracking and reference purposes. The weight, standard length (with mouth closed), and sex of each newly- thawed fish were recorded for use in sexual size dimorphism analyses. Fish bent during the storage and/or the preservation process were not included in the analyses if they could not be manually straightened without distorting the lateral body shape.

Image Acquisition

A total of 270 fish were photographed with a Nikon D7100 digital camera, which records images at a resolution of 6000 X 4000 pixels. Each fish was placed in the center of a dissecting pad (32 X 23 cm) with the left side of its body facing up (Figure 1). The fins of each fish were spread out and pinned in place to provide a clearer view of insertion points into the body. For scale, a ruler was placed against the lower edge of the

8 dissecting pad. The identification number and sex were written on a small square of paper and placed within the viewing field for each fish that was photographed.

Photographs were obtained using a “free-hand” technique (i.e., without a tripod) to maximize efficiency and photograph quality. To prevent image distortion, the camera was adjusted so that the specimen was in the center of the viewing field. If the available ambient light was not sufficient, additional light was provided via a desk lamp, as the camera’s flash feature was not used. Multiple photographs of a specimen were taken to ensure the acquisition of a quality image. Once photographed, the fish was placed in a plastic freezer bag labeled with its respective identification number and photograph date, and returned to the freezer.

After each photography session, photographs were sorted through to select the best image for each specimen. Landmark clarity, lighting, and angle were considered when selecting a photograph. Once chosen, the photograph file was renamed with the specimen ID number and sex (e.g., 001F). The tpsUtil software was used to build a TPS file of all the selected images in preparation for data collection. All geometric morphometric software used for this study was publicly available at http://life.bio.sunysb.edu/morph/index.html.

Landmark Identification and Digitization

Landmarks were selected using criteria supplied by Zelditch et al. (2012). The

following 18 biologically homologous landmarks were identified: (1) anterior tip of snout, (2) most anterior point of eye outline, (3) most posterior point of eye outline, (4)

9 dorsal origin of the operculum, (5) posterior boundary of the supraoccipital bone, (6) anterior insertion of dorsal fin, (7) dorsal fin spine and ray boundary, (8) posterior insertion of dorsal fin, (9) dorsal insertion of caudal fin, (10) lateral line at caudal fin,

(11) ventral insertion of caudal fin, (12) posterior insertion of anal fin, (13) anterior insertion of anal fin, (14) posterior insertion of pelvic fin, (15) anterior insertion of pelvic fin, (16) posterior insertion of pectoral fin, (17) anterior insertion of pectoral fin, and (18) ventral origin of the operculum (Figure 1). Pins were inserted on each specimen at landmarks 4, 5, and 10 prior to taking photographs to ensure the accuracy of landmark location during digitization.

Landmarks were digitized using tpsDig2 software. This program was selected because the data format it produces can be read by all shape analysis software (Zelditch et al. 2012).

Figure 1. Landmarks used for geometric morphometric shape analysis of tule perch. (1) anterior tip of snout, (2) most anterior point of eye outline, (3) most posterior point of eye outline, (4) dorsal origin of the operculum, (5) posterior boundary of the supraoccipital bone, (6) anterior insertion of dorsal fin, (7) dorsal fin spine and ray boundary, (8) posterior insertion of dorsal fin, (9) dorsal insertion of caudal fin, (10) lateral line at caudal fin, (11) ventral insertion of caudal fin, (12) posterior insertion of anal fin, (13) anterior insertion of anal fin, (14) posterior insertion of pelvic fin, (15) anterior insertion of pelvic fin, (16) posterior insertion of pectoral fin, (17) anterior insertion of pectoral fin, and (18) ventral origin of the operculum.

Shape Coordinates

Shape coordinates were calculated from the landmark data provided by tpsDig2.

Using CoordGen8, a Generalized Procrustes Analysis (GPA) was performed to remove all non-shape variation due to position, orientation, and scale of the specimen (Zelditch et al. 2012; Figure 2). This process translates each specimen to a common position and rotates it such that the distance between the landmarks of each configuration is minimized within and between all configurations (Hood and Heins 2000; Hassell et al.

2012). Generalized Procrustes Analysis is the most widely used form of superimposition and was selected over other methods because it is the most consistent with the general theory of shape. Additionally, GPA reduces induced covariance among landmarks and produces smaller variance ellipses around most landmarks (Zelditch et al. 2012).

Figure 2. Generalized Procrustes Analysis (GPA) of 130 male (blue) and 139 female (pink) tule perch. Performed in CoordGen8 software, GPA mathematically removes non- shape variation due to size, scale, and position.

Statistical Analyses

To test for sexual dimorphism, a single factor permutation Multivariate Analysis of Variation (MANOVA) was performed using the software CVAGen8 (Zelditch et al.

2012). Additionally, a Canonical Variates Analysis (CVA) was performed in CVAGen8 to describe the differences in body shape between male and female tule perch with regard to variation within each sex. Differences between the sexes were visualized using deformation grids generated in CVAGen8.

To test for body shape change with respect to centroid size, a multivariate regression was performed using Regress8 (Zelditch et al. 2012). Centroid size is a measure of geometric scale that is the preferred size measurement in geometric morphometrics because it is uncorrelated with shape in the absence of allometry (Zelditch et al. 2012). Differences in body shape with respect to centroid size were visualized using deformation grids generated in Regress8. Additionally, a Principle Components Analysis

(PCA) was performed in PCAGen8 to explore the variation in body shape within the sexes. This variation was visualized using deformation grids generated in PCAGen8.

To test for sexual size dimorphism with regard to standard length, a two-sample t- test was performed. The assumption of normality was tested by plotting the data on a normal probability plot. The assumption of equal variances was tested using an F-test.

These procedures were repeated to test for sexual size dimorphism with regard to weight.

All analyses investigating sexual size dimorphism were performed using PAST Version

3.11.

RESULTS

Sexual Dimorphism of Body Shape

There was a significant difference in body shape between males and females

(F=59.17, df=268, p<0.001). The first canonical variate (CV1) was the only unique axis

(Wilk’s Λ=0.1782, X2=432.98, df=32, p<0.001; Figure 2) and accounted for 96.7% of the

variance. A jack-knife test of group assignments resulted in 258 correct assignments and

11 incorrect assignments (Table 1). Given the group sizes, the expected random rate of

correct assignments was 50.2% and the observed rate of correct assignments was 95.9%

correct.

The deformation grid associated with CV1 illustrated that the most effective

discriminators between the sexes were landmarks 8, 13, and 15 (Figure 4; see Figure 1

for landmark identification). Females were narrower through the caudal peduncle and

mid-body and had anterior anal fin insertion points that were more posteriorly-located

than males. Additionally, the deformation grid associated with the regression of CV1 illustrated females had steeper foreheads, as well as lower, more anteriorly-located eyes and more dorsally-located pectoral fins than males (Figure 5).

Figure 3. Canonical Variates Analysis (CVA) of the sexes. Results of the CVA, performed in CVAGen8, demonstrate that one dominant canonical variate axis described significant differences between male (blue circles) and female (pink squares) tule perch.

Table 1. Jack-knife test of group assignments. Results of the jack-knife analysis demonstrate that 95.9% of tule perch were assigned to the correct group (sex) based on body shape.

a posteriori assignments

Female Male

Female 135 4

Male 7 123 a priori

assignments

Figure 4. Deformation grid of the first canonical variate (CV1). Generated in CVAGen8, this deformation grid depicts the shape transformation associated with the CV1 axis, which represents the variables that maximally discriminate between the sexes. Circles on the grid represent the reference body shape for all male tule perch and the vectors demonstrate the direction and relative magnitude of local shape change to the reference body shape of all female tule perch.

Figure 5. Deformation grid of the first canonical variate (CV1) regressed. Generated in CVAGen8, this deformation grid depicts the shape transformation correlated with the CV scores, which represents all shape differences between the sexes rather than the maximal discriminators. Circles on the grid represent the reference body shape for all male tule perch and the vectors demonstrate the direction and relative magnitude of local shape change to the reference body shape of all female tule perch.

Size-related Changes in Body Shape

In the combined dataset (males and females), variation in shape was related to centroid size (F = 19.98, df1=32, df2=8576, p<0.01). Centroid size accounted for 6.9% of morphological variation. The deformation grid obtained from the regression of shape on size illustrated that larger fish were deeper through the mid-body and more blunt through the snout than smaller fish (Figure 6). Additionally, the caudal peduncle was wider and more rounded and the pectoral fin was shifted ventrally in larger fish. Finally, the eyes of larger fish were relatively smaller and located more dorsally on the body than those of smaller fish.

In females, variation in shape was related to centroid size (F=8.924, df1=32, df2=4384, p<0.01). Centroid size accounted for 6.1% of the morphological variation. The deformation grid obtained from the regression of shape on size illustrated that larger females were deeper through the mid-body, more blunt through the snout, slighter wider through the caudal area, and had steeper foreheads than smaller females (Figure 7).

Additionally, the anterior insertion of the anal fin was posteriorly shifted in larger females. The pectoral fin in larger females was shifted ventrally compared to its placement in smaller females. Finally, the eyes of larger females were relatively smaller and located more dorsally on the body than those of smaller females.

A PCA of female tule perch indicated that the first principle component accounted for 24.1% of the variance and was the only distinct axis (Figure 8). The deformation grid

20 associated with PC1 illustrated changes toward a deeper, rounder abdomen; shorter, more dorsally-angled tail; higher placed eye; and flatter, more asymmetric head (Figure 9).

In males, variation in shape was related to centroid size (F=13.41, df1=32, df2=4096, p<0.01). Centroid size accounted for 9.5% of the morphological variation. The deformation grid obtained from the regression of shape on size illustrated that larger males were deeper through the mid-body, more blunt through the snout, wider through the caudal region, and rounder through the posterior edge of the caudal peduncle than smaller males (Figure 10). The pectoral fin in larger males was shifted ventrally compared to its placement in smaller males. Finally, the eyes of larger males were relatively smaller and located higher on the body than those of smaller males.

A PCA of male tule perch indicated that the first and second principal components were the only distinct axes. The first principle component (PC1) accounted for 24.5% of the variance and the second principle component (PC2) accounted for

15.5% of the variance (Figure 11). The deformation grid associated with PC1 illustrated changes toward a deeper, rounder abdomen; shorter, wider, more rounded tail; higher placed eye; and a flatter, more asymmetric head (Figure 12). The deformation grid associated with PC2 illustrated changes toward a steeper, higher head; dorsal expansion through the mid-body; and a slightly wider, more dorsally-angled tail (Figure 13).

Figure 6. Deformation grid of regression of entire tule perch sample. Generated in Regress8, this deformation grid depicts the shape transformation associated with the regression of shape on centroid size for the combined (male and female) sample of tule perch. Circles on the grid represent the reference body shape for smaller tule perch and the vectors demonstrate the direction and relative magnitude of local shape change to the reference body shape of larger tule perch.

Figure 7. Deformation grid of regression of female tule perch. Generated in Regress8, this deformation grid depicts the shape transformation associated with the regression of shape on centroid size for female tule perch. Circles on the grid represent the reference body shape for smaller females and the vectors demonstrate the direction and relative magnitude of local shape change to the reference body shape of larger females.

Figure 8. Scree plot of percentage of variance explained versus principal component number for female tule perch. Generated in PCAGen8, this scree plot provides a

visualization of the finding that the first principal component is the only distinct axis.

Figure 9. Deformation grid of the first principal component (PC1) for female tule perch. Generated in PCAGen8, the deformation grid associated with PC1 illustrates changes toward a deeper, rounder abdomen; shorter, more dorsally-angled tail; higher placed eye; and flatter, more asymmetric head.

Figure 10. Deformation grid of regression of male tule perch. Generated in Regress8, this deformation grid depicts the shape transformation associated with the regression of shape on centroid size for male tule perch. Circles on the grid represent the reference body shape for smaller males and the vectors demonstrate the direction and relative magnitude of local shape change to the reference body shape of larger males.

Figure 11. Scree plot of percentage of variance explained versus principal component number for male tule perch. Generated in PCAGen8, this scree plot provides a visualization of the finding that the first two principal components are the only distinct

axes.

Figure 12. Deformation grid of the first principal component (PC1) for male tule perch. Generated in PCAGen8, the deformation grid associated with PC1 illustrates changes toward a deeper, rounder abdomen; shorter, wider, more rounded tail; higher placed eye; and a flatter, more asymmetric head.

Figure 13. Deformation grid of the second principal component (PC2) for male tule perch. Generated in PCAGen8, the deformation grid associated with PC2 illustrates changes toward a steeper, higher head; dorsal expansion through the mid-body; and a slightly wider, more dorsally-angled tail.

Sexual Size Dimorphism

There was no significant difference between males and females with regard to standard length (Table 2). The variances of male and female standard lengths did not differ significantly (F=1.14, df=270, p=0.46) and each standard length dataset was normally distributed (female correlation coefficient = 0.98; male correlation coefficient =

0.99).

Males were significantly heavier than females (t=-3.51, df=270, p<0.001; Table

2). The variances of male and female weights did not differ significantly (F=1.04, df=270, p=0.82). A log transformation was applied to both weight datasets prior to other analyses because they did not exhibit a normal distribution. After transformation, the datasets were normally distributed (female correlation coefficient = 0.99; male correlation coefficient = 0.98).

Table 2. Tule perch size data.

Average Standard Average Weight (g) Length (mm) Male (n=132) 87.1 23.0 Female (n=140) 84.5 18.0

DISCUSSION

Sexual Dimorphism of Body Shape

Tule perch collected in the present study exhibited sexual dimorphism with regard to body shape. The most effective morphological discriminators between the sexes occurred along the ventral edge of the body and in the caudal region.

The finding that females were narrower through the mid-body than males is not only inconsistent with expectations based on the respective reproductive roles of each sex, but also with previous research in both oviparous and viviparous species (e.g, Hood and Heins 2000; Cadrin and Silva 2005; Hassell et al. 2012; Unito-Ceniza et al. 2012).

One possible explanation for this discrepancy may be elucidated by the finding that females had a more posteriorly-placed anterior insertion of the anal fin than males. While males had a deeper ventral outline, the posterior shift of the anal fin, and by association the anus, in females may indicate the availability of additional volume in the body cavity for ovary development.

Another possible explanation may be due to the asynchronous enlargement of the gonads in the sexes. The tule perch used in this study were collected in the middle of breeding season. During this time, male gonad size is maximized, while female gonad size is minimized (Bryant 1977). Gonad size in females is highest in the spring, just before partuition (Bryant 1977). Further investigation comparing body shape of each sex during peak and quiescent gonad development periods would be necessary for clarification.

As previously mentioned, the anterior insertion of the anal fin was posteriorly- shifted in females. In addition to the implication of this finding regarding the placement of the anus in females, it also illustrates that males had a wider anal fin base. This finding is consistent with studies in oviparous fishes (e.g., Oryzias latipes, Koseki et al. 2000;

Puntius binotatus, Dorado et al. 2012; Lepisosteus osseus, McGrath and Hilton 2012;

Fundulus notatus, Welsh et al. 2013; Atractosteus spatula, McDonald et al. 2013), and has only been documented in one viviparous fish species, Sebastes umbrosus (Chen

1971).

As observed by Phelps (1989), tule perch are lek-type breeders. Males hold small display territories, which are defended by posturing, biting, and chasing away other males. It has been suggested that a wider anal fin base could increase the success in these activities as a larger fin base increases the surface area displayed to an opponent during antagonistic interactions (Oliveira and Almada 1995; McGrath and Hilton 2012).

An additional possibility is that a larger anal fin aids in swimming stability and maneuverability (Eklov and Jonsson 2007) and could facilitate optimal male-female placement during breeding to maximize fertilization success (Casselman & Schulte-

Hostedde 2004). Because tule perch are internal fertilizers, male positioning during mating is critical to fertilization and reproductive success.

The finding that females were narrower through the caudal peduncle is consistent with studies on oviparous fishes (Hood and Heins 2000; Cadrin and Silva 2005;

McDonald et al. 2013). Based on work completed in alligator gar (Atractosteus spatula), the authors suggested that a larger caudal peduncle may be beneficial for a competitive

33 male during spawning because it may (1) enhance the ability of a male to distribute milt over egg batches, and (2) afford an advantage in accessing eggs during spawning events

(McDonald et al. 2013). Although the former benefit may not apply to tule perch given their viviparity, the latter may translate as males compete for access to females during mating events.

Size-related Changes in Body Shape

In addition to exhibiting sexual dimorphism, tule perch in this study exhibited different body shapes with regard to size. Universal differences included deeper mid- bodies, blunter snouts, and wider caudal peduncles in larger fish. Furthermore, the eyes of smaller fish were much larger relative to their body size than those of larger fish.

These findings are consistent with previously documented patterns of ontogenetic change in a variety of fish species (e.g., Reis et al. 1998; Hood and Heins 2000; Cadrin and Silva

2005; Hassell et al. 2012). This pattern has been interpreted as a juvenile morphology that emphasizes sensory abilities and feeding systems manifested into a shape determined by the reproductive requirements of adults (Hassell et al. 2012).

Besides allowing room for reproductive structures, deeper bodies also provide a shape more suitable for swimming manueuverability (Webb 1984). The importance of this with regard to mating has been previously discussed in this study. Additionally, deeper bodies and blunter snouts may increase the chance of escape from predation by piscivorous fish. Generally, piscivores swallow their prey head first and are limited by prey body depth relative to their mouth width (Eklov and Jonsson 2007). Nilsson and

Bronmark (2000) found that piscivorous pike preferred more shallow-bodied fish because deep-bodied prey are more difficult, and require more handling time, to swallow.

Furthermore, Elkov and Jonsson (2007) found juvenile perch that were grown in environments with a pike predator presence had deeper bodies than juveniles in a pike- free environment.

Although age was not investigated in this study, Bryant (1977) demonstrated a well-fit, positive linear relationship between age group and mean standard length in tule perch. The work completed by Bryant (1977) and the consistency of the findings in this study with previously completed ontogenetic studies suggest that the documented shape differences may be ontogenetic rather than simply size-related.

Sexual Size Dimorphism

Tule perch in the present study did not exhibit sexual size dimorphism with regard to standard length. This finding conflicts with the only other investigation of sexual size dimorphism in tule perch. Phelps (1989) documented that male tule perch were significantly smaller than female tule perch and suggested earlier investment in reproduction by males than females was the cause.

However, both Phelps (1989) and Bryant (1977) reported that the sex ratios of their samples were heavily female-biased (75 and 25% and 63 and 37%, respectively).

Bryant (1977) also reported that the sex ratio of tule perch in the ovary, prior to partuition, was more evenly distributed (47 and 53% male and female, respectively) and suggested that females might be more easily trapped during sample collection than males.

The sex ratio of the sample used in this study was fairly even (49% male and 51% female). This study sampled tule perch using an electroshocking boat, which may be a more inclusive collection method, resulting in an unskewed sex ratio. It is also possible that the skewed sex ratio collected by Phelps (1989) may have affected the significant sexual size dimorphism of that study.

The finding in this study was similar to findings in studies on other internally fertilizing, viviparous fish (e.g., Bisazza 1997, Froeschke et al. 2007). For example, the goodeid fish Xenotoca eiseni only exhibit a slight sexual size dimorphism with the average male standard length equaling 91% that of females (Bisazza 1997). A study determined that male X. eiseni preferred similarly sized mates and that significantly more copulations were successful between similarly sized pairs than mismatched pairs (Bisazza

1997). Additionally, the author noted that in instances where the female was much larger than the male, the male struggled to place his anal fin in the correct position and was too far forward to copulate. It has been suggested that it is beneficial for males to be similarly sized to their potential mates given the difficulty of achieving internal fertilization underwater (Bisazza 1997).

Finally, tule perch exhibited sexual size dimorphism with regard to weight. This may be due to the aforementioned discrepancy in breeding status at the time of collection.

Bundy (1970) and Bryant (1977) found testicular enlargement to be the highest during the summer months. Furthermore, a study in another surfperch species, Micrometrus minimus, documented that testes actively produce sperm during the mating season, but are inactive during the remainder of the year (Warner and Harlan 1982). During breeding

36 season, gonads comprised an average of 9.4% of a male’s body weight. The difference in average weight between male and female tule perch in this study was 5.01 grams, which may be partially attributed to the enlarged gonads in males during breeding season.

CONCLUSIONS

This study suggests that tule perch may exhibit differences in body shape with regard to sex and size, despite their monomophic appearance. Additionally, tule perch in this study did not exhibit sexual size dimorphism with regard to standard length, but did exhibit sexual size dimorphim with regard to weight. This study contributed to the very limited knowledge of a California native fish species and augmented the collected works of geometric morphometric studies in viviparous fishes. Although additional investigation of sex-related body shape during peak and quiescent gonad development periods would be necessary for clarification, these findings provide opportunities for future research, which will increase the knowledgebase regarding the influence of ontogeny on the onset and development of dimorphic changes in fish, as well as the evolutionary biology leading to these differences.

Literature Cited

Adams, D.C. 1999. Methods for shape analysis of landmark data from articulated structures. Evolutionary Ecology Research 1: 959-9710.

Adams, D. C., Rohlf, F. J., and D.E. Slice. 2004. Geometric morphometrics: ten years of progress following the ‘revolution’. Italian Journal of Zoology 71: 5-16.

Araujo, M.S., Perez, S.I., Magazoni, M.J.C., and A.C. Petry. 2014. Body size and allometric shape variation in the molly Poecilia vivipara along a gradient of salinity and predation. BMC Evolutionary Biology 14:251.

Baltz, D.M. 1984. Life history variation among female surfperches (Perciformes: Embiotocidae). Environmental Biology of Fishes 10:159-171.

Baltz, D.M. and E.J. Loudenslager. 1984. Electrophoretic variation among subspecies of tule perch (Hysterocarpus traski). Copeia 1984:223-227.

Baltz, D.M. and P.B. Moyle. 1981. Morphometric analysis of tule perch (Hysterocarpus traski) populations in three isolated drainages. Copeia 1981:305-311.

Baltz, D.M. and P.B. Moyle. 1982. Life history characteristics of tule perch (Hysterocarpus traski) populations in contrasting environments. Environmental Biology of Fishes 7:229-242.

Bisazza, A. 1997. Sexual selection constrained by internal fertilization in the livebearing fish Xenotoca eiseni. Animal Behaviour, 54: 1347-1355.

Blanckenhorn, W. U. 2005. Behavioral causes and consequences of sexual size dimorphism. Ethology 111: 977-1016.

Bookstein, F. L. 1982. Foundations of morphometrics. Annual Review of Ecology and Systematics 13: 451-470.

Bookstein, F. L. 1986. Size and shape spaces for landmark data in two dimensions. Statistical Science 1: 181-222.

Breder Jr, C. M., and D.E. Rosen. 1966. Modes of reproduction in ﬁshes. Nat. Hist. Press, Garden City, NY, USA.

Bryant, G.L. 1977. Fecundity and growth of the tule perch, Hysterocarpus traski, in the lower Sacramento-San Joaquin Delta. California Fish and Game 63:140-156.

Bundy, D.S. 1970. Reproduction and growth of the tule perch Hysterocarpus traskii (Gibbons) with notes on its ecology. M.Sc. Thesis. University of the Pacific: USA.

Cadrin, S. X. and V.M. Silva. 2005. Morphometric variation of yellowtail flounder. ICES Journal of Marine Science: Journal du Conseil 62: 683-694.

Casselman, S. J., A.I. Schulte‐Hostedde. 2004. Reproductive roles predict sexual dimorphism in internal and external morphology of lake whitefish, Coregonus clupeaformis. Ecology of Freshwater Fish 13: 217-222.

Chen, L. C. 1971. Systematics, variation, distribution, and biology of rockfishes of the subgenus Sebastomus (Pisces, Scorpaenidae, Sebastes). Scripps Institution of Oceanography 18.

Coban, D., Yildirim, S., Kamaci, H. O., Suzer, C., Saka, S., and K. Firat. 2011. External morphology of European seabass (Dicentrarchus labrax) related to sexual dimorphism. Turkish Journal of Zoology 35: 255-263.

DiStefano, R. J., Roell, M. J., Wagner, B. A., and J.J. Decoske. 1994. Relative performances of four preservatives on fish and crayfish. Transactions of the American Fisheries Society, 123: 817-823.

Dorado E. L., Torres M. A. J., and C.G. Demayo. 2012. Sexual dimorphism in body shapes of the spotted barb fish, Puntius binotatus of Lake Buluan in Mindanao, Philippines. AACL Bioflux 5: 321-329.

Douglas, M.E. 1993. Analysis of Sexual Dimorphism in an endangered cyprinid fish (Gila cypha Miller) using video image technology. Copeia 1993: 334-343.

Echeverria, T. W. 1986. Sexual dimorphism in four species of rockfish genus Sebastes (Scorpaenidae). Environmental Biology of Fishes 15: 181-190.

Eklöv, P., and P. Jonsson. 2007. Pike predators induce morphological changes in young perch and roach. Journal of Fish Biology, 70: 155-164.

Fink, W. L., and M. L. Zelditch. 1995. Phylogenetic analysis of ontogenetic shape transformations: a reassessment of the piranha genus Pygocentrus (Teleostei). Systematic Biology 44: 343-360.

Fink, W. L. and M. L. Zelditch. 1996. Historical patterns of developmental integration in piranhas. American Zoologist 36: 61-69.

Frederich, B., Adriaens, D., and P. Vandewalle. 2008. Ontogenetic shape changes in Pomacentridae (Teleostei, Perciformes) and their relationships with feeding strategies: a geometric morphometric approach. Biological Journal of the Linnean Society 95: 92-105.

Froeschke, B., Allen, L.G. and D.J. Pondella II. 2007. Life history and courtship behavior of black perch, Embiotoca jacksoni (Teleostomi: Embiotocidae), from southern California. Pacific Science 61:521-531.

Hassell, E.M.A., Meyers, P.J., Billman, E.J., Rasmussen, J.E., and M.C. Belk. 2012. Ontogeny and sex alter the effect of predation on body shape in a livebearing fish: sexual dimorphism, parallelism, and costs of reproduction. Ecology and Evolution 2:1738-1746.

Hedrick, A. V., and E.J. Temeles. 1989. The evolution of sexual dimorphism in animals: hypotheses and tests. Trends in Ecology & Evolution 4: 136-138.

Herler, J., Kerschbaumer, M., Mitteroecker, P., Postl, L., and C. Sturmbauer. 2010. Sexual dimorphism and population divergence in the Lake Tanganyika cichlid fish genus Tropheus. Frontiers in Zoology 7:4.

Hood, C.S., and D.C. Heins. 2000. Ontogeny and allometry of body shape in the blacktail shiner, Cyprinella venusta. Copeia 2000: 270-275.

Kassam, D., Mizoiri, S., and K. Yamaoka. 2004. Interspecific variation of body shape and sexual dimorphism in three coexisting species of the genus Petrotilapia (Teleostei: Cichlidae) from Lake Malawi. Ichthyological Research 51: 195-201.

Kitano, J., Mori, S., and C.L. Peichel. 2007. Sexual dimorphism in the external morphology of the threespine stickleback (Gasterosteus aculeatus). Copeia 2007: 336-349.

Kitano, J., Mori, S., and C.L. Peichel. 2012. Reduction of sexual dimorphism in stream‐ resident forms of three‐spined stickleback Gasterosteus aculeatus. Journal of Fish Biology 80: 131-146.

Koseki, Y., Takata, K., and K. Maekawa. 2000. The role of the anal fin in fertilization success in male medaka, Oryzias latipes. Fisheries Science 66: 633-635.

Lenarz, W. H., and T.W. Echeverria. 1991. Sexual dimorphism in Sebastes. Environmental Biology of Fishes 30: 71-80.

Liu, J., Tatarenkov, A., O’Rear, T.A., Moyle, P.B. and J.C. Avise. 2013. Molecular evidence for multiple paternity in a population of the viviparous tule perch Hysterocarpus traski. Journal of Heredity 104: 217-222.

Loy, A., Mariani, L., Bertelletti, M., and L. Tunesi. 1998. Visualizing allometry: geometric morphometrics in the study of shape changes in the early stages of the two-banded sea bream, Diplodus vulgaris (Perciformes, Sparidae). Journal of Morphology 237: 137-146.

McDonald, D. L., Anderson, J. D., Hurley, C., Bumguardner, B. W., and C.R. Robertson. 2013. Sexual Dimorphism in Alligator Gar. North American Journal of Fisheries Management 33: 811-816.

McGrath, P. E., and E.J. Hilton. 2012. Sexual dimorphism in longnose gar Lepisosteus osseus. Journal of Fish Biology 80: 335-345.

Moyle, P.B. 2002. Inland fishes of California. Revised and Expanded Edition. University of California Press, Berkeley, CA, USA.

Nilsson, P. A., and C. Brönmark. 2000. Prey vulnerability to a gape‐size limited predator: behavioural and morphological impacts on northern pike piscivory. Oikos 88: 539-546.

Oliveira, R. F., and V.C. Almada. 1995. Sexual dimorphism and allometry of external morphology in Oreochromis mossambicus. Journal of Fish Biology 46: 1055- 1064.

Phelps, A. 1989. Behavioral ecology of tule perch: Life-history and reproductive behavior. Ph.D. Thesis. UC Davis: USA.

Reis, R. E., Zelditch, M. L., and W.L. Fink. 1998. Ontogenetic allometry of body shape in the Neotropical catfish Callichthys (Teleostei: Siluriformes). Copeia 1998: 177-182.

Rohlf, F. J., and Slice, D. 1990. Extensions of the Procrustes method for the optimal superimposition of landmarks. Systematic Biology 39: 40-59.

Rohlf, F.J. and L.F. Marcus. 1993. A revolution in morphometrics. Trends in Ecology & Evolution 8: 129-132.

Strauss, R. E. and F.L. Bookstein, F. L. 1982. The truss: body form reconstructions in morphometrics. Systematic Biology 31: 113-135.

Trapani, J. 2003. Geometric morphometric analysis of body-form variability in Cichlasoma minckleyi, the Cuatro Cienegas cichlid. Environmental Biology of Fishes 68: 357-369.

Unito-Ceniza, K. M., Torres, M. A. J., and C.G. Demayo. 2012. Describing body shape variation between sexes of an endemic eleotrid fish Hypseleotris agilis (Herre, 1927), from Lake Mainit, using landmark-based geometric morphometrics. International Journal of Biological, Ecological and Environmental Sciences 1: 201-204.

Warner, R. R., and R.K. Harlan. 1982. Sperm competition and sperm storage as determinants of sexual dimorphism in the dwarf surfperch, Micrometrus minimus. Evolution, 44-55.

Webb, P. W. 1984. Body form, locomotion and foraging in aquatic vertebrates. American Zoologist, 24: 107-120.

Webster, M. and H.D. Sheets. 2010. A practical introduction to landmark-based geometric morphometrics. The Paleontological Society Papers 16: 163-188.

Welsh, D. P., Zhou, M., Mussmann, S. M., Fields, L. G., Thomas, C. L., Pearish, S. P., Kilburn, S.L., Parker, J.L., Stein, L.R., Bartlett, J.A., Bertram, C.R., Bland, T.J., Laskowski, K.L., Mommer, B.C., Zhuang, X., and R.C. Fuller. 2013. The effects of age, sex, and habitat on body size and shape of the blackstripe topminnow, Fundulus notatus (Cyprinodontiformes: Fundulidae)(Rafinesque 1820). Biological Journal of the Linnean Society 108: 784-789.

Woods, P. J. 2007. Habitat‐dependent geographical variation in ontogenetic allometry of the shiner perch Cymatogaster aggregata Gibbons (Teleostei: Embiotocidae). Journal of Evolutionary Biology 20: 1783-1798.

Zelditch, M. L., and W.L. Fink. 1995. Allometry and developmental integration of body growth in a piranha, Pygocentrus nattereri (Teleostei: Ostariophysi). Journal of Morphology 223: 341-355.

Zelditch, M.L., Swiderski, D.L., Sheets, H.D. and W.L. Fink. 2012. Geometric Morphometrics for Biologists. 2nd Edition. Elsevier Academic Press, San Diego, CA, USA.