COMPARATIVE ANALYSIS OF FACTORS INFLUENCING MALE REPRODUCTIVE
SUCCESS IN SYMPATRIC FRESHWATER TURTLES

By

Jeanette M. McGuire

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment for the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Zoology
Ecology, Evolutionary Biology and Behavior

2011

ABSTRACT

COMPARATIVE ANALYSIS OF FACTORS INFLUENCING MALE REPRODUCTIVE
SUCCESS IN SYMPATRIC FRESHWATER TURTLES

By

Jeanette M. McGuire

Male reproductive success (RS) is primarily constrained by mate number. However,
when females vary in reproductive quality (i.e., reproductive frequency, clutch size, egg size),
female quality can contribute substantially to male reproductive success. I used a comparative
inter-specific approach, to document sources of variation in male reproductive success in three
species of freshwater turtles (Painted turtles (Chrysemys picta), Blanding's turtles (Emydoidea
blandingii) and Snapping turtles (Chelydra serpentina). Data were collected and analyzed using
extensive background information from a long-term study on the life history and demography of
all three species. Long-term studies (1953-2007) on the E.S. George Reserve in Michigan
allowed analyses to be preformed based on females of known age, size, and reproductive history.
In all three species, larger and older females were of higher reproductive quality than were
smaller or younger females, suggesting that males would benefit by mating with these females
compared to mating indiscriminately. Using multi-locus microsatellite genotypes and paternity
analysis, I examined offspring (Painted turtles N = 1065 from 171 clutches; Blanding's turtles N
= 772 from 98 clutches, Snapping turtles N = 1064 from 63 clutches) collected over four years

(Painted and Snapping turtles) and 8 years (Blanding's turtles) from nests of known females.
Incidence of multiple paternity was variable among years in all three species (Painted turtles,
6.1% - 30.0%; Blanding‟s turtles, 15.4% - 55.6%; Snapping turtles, 44.4% - 61.5%) and was
positively associated with female age in all three species. Sperm from the same male(s)
fertilized successive clutches (repeat paternity) among years in all three species. Reproductive
frequency contributed substantially to male reproductive success. Paternity analyses conducted
in Painted and Blanding's turtles revealed that the number of clutches sired represented a
significant component to male reproductive success. Mate number was not a substantial source
of variation of male reproductive success in Painted turtles, as few males mated with more than
one female. In Blanding's turtles mate number and clutch number were significantly associated
with male reproductive success. Female reproductive frequency increased with increasing age,.
Additionally, older females are more likely to have multiple males sire offspring within a clutch
suggesting that older females were comparatively more attractive to males or more receptive to
mating attempts by multiple males. Clutches of eggs from older females were more genetically
diverse (lower coancestry) that clutches of younger females. Therefore, older females contribute
disproportionally to population levels of genetic diversity. For species such as Blanding's turtles
that make extensive use of terrestrial habitats, we found that older females were more likely to
have offspring sired by males from other residence wetlands than do younger females. Results
highlight the need to protect corridors that facilitate movement and gene flow among permanent
wetlands. The presence of age-specific effects on female reproductive quality in all three species
highlight the importance of management strategies that reduce adult mortality rates.

Copyright by
Jeanette M. McGuire
2011

ACKNOWLEDGEMENTS

I am deeply indebted to both my advisor, Dr. Kim Scribner, and to my co-advisor, Dr. Justin
Congdon, for all of their hard work and dedication to helping me learn and grow as an academic
and also for their help and compassion when dealing with the trials of life. You both have taught
me more about science, turtles, and life than you know and I am forever grateful. I would like to
thank my committee, Drs. Kay Holekamp, Andrew McAdam, and Doug Schemske, for their
insight, comments, and guidance throughout the project.
I would like to thank the University of Michigan for maintaining the Edwin S. George Reserve
as a world class research facility and recognizing the value in long-term research. I'd like to
thank the Department of Zoology, Department of Fisheries and Wildlife, the Ecology,
Evolutionary Biology, and Behavior Program, and the Graduate School of Michigan State
University for their support, and especially the secretaries and office staff who never tired of my
endless questions and always answered with a smile.
I'd also like to thank all of those individuals that have helped to contribute to the long-term data
that greatly enhanced this project. In particular, I'd like to thank Owen Kinney, Roy Nagle, Todd
Quinter, and Anna Pock. I'd especially like to thank Josh Capps for not only his help in the field
but also for the endless hours of conversation on turtles, philosophy, and life.
I'd like to thank Nestor and Diana DeOcampo for their help, guidance, and friendship. I have
learned so much about teaching from both of you. You have been amazing to work with and I
am thankful for our friendship.
v

I'd also like to thank Fred Dyer for giving me my first research job and for showing me that
science could be a career. Those simple words changed my life.
Special thanks are due to the current and past members of the Scribner lab, especially Scot,
Kristi, Yen, Laura, Jeannette, Jen, Stacey, Hope and Jamie for your help, support, and friendship.
Last, but certainly not least, a very heartfelt thank you to my family and friends that have helped
me through each and every part of the last several years. You all helped me to so much through
everything and I can honestly say I could not have finished this degree without the love and
support of all of you. In particular, I'd like to thank Gary Dunn who has helped me with more
than I could have ever imagined. I dedicate my dissertation to you and to my family (blood and
extended) especially: Theresa, John, Jennifer, Erik, Jonathan, Shelley, Julia, Paul, Joey,
Francesca, Terry, Linda, Kris, Brett, Ellen, Kristi, and Jen. The words here cannot express how
much you have done for me. I love you all!

vi

TABLE OF CONTENTS:

LIST OF TABLES..........................................................................................................................ix
LIST OF FIGURES........................................................................................................................xi
INTRODUCTION...........................................................................................................................1
APPENDIX 0...................................................................................................................... 6
TABLE.................................................................................................................... 7
REFERENCES....................................................................................................................8
CHAPTER 1..................................................................................................................................10
VARIATION IN FEMALE REPRODUCTIVE QUALITY AND REPRODUCTIVE SUCCESS
OF MALE PAINTED TURTLES (CHRYSEMYS PICTA MARGINATA)
ABSTRACT......................................................................................................................11
INTRODUCTION.............................................................................................................12
METHODS........................................................................................................................15
RESULTS..........................................................................................................................21
DISCUSSION....................................................................................................................25
APPENDIX 1 ....................................................................................................................32
TABLES................................................................................................................33
FIGURES...............................................................................................................35
REFERENCES..................................................................................................................39
CHAPTER 2..................................................................................................................................46
FEMALE QUALITY AFFECTS MALE PAINTED TURTLE (CHRYSEMYS PICTA
MARGINATA) REPRODUCTIVE SUCCESS
ABSTRACT.......................................................................................................................47
INTRODUCTION.............................................................................................................48
METHODS........................................................................................................................49
vii

RESULTS..........................................................................................................................55
DISCUSSION....................................................................................................................61
APPENDIX 1 ................................................................................................................... 69
TABLES................................................................................................................70
FIGURES...............................................................................................................71
REFERENCES..................................................................................................................78
CHAPTER 3..................................................................................................................................85
MATING SYSTEM, MALE REPRODUCTIVE SUCCESS, AND POPULATION
CONNECTIVITY OF BLANDING'S TURTLES (EMYDOIDEA BLANDINGII)
ABSTRACT.......................................................................................................................86
INTRODUCTION.............................................................................................................87
METHODS........................................................................................................................92
RESULTS..........................................................................................................................98
DISCUSSION..................................................................................................................103
APPENDIX 1 ..................................................................................................................113
TABLES..............................................................................................................114
FIGURES.............................................................................................................115
REFERENCES................................................................................................................121
CHAPTER 4................................................................................................................................130
FACTORS INFLUENCING MATING SYSTEMS AND MALE REPRODUCTIVE SUCCESS
IN FRESHWATER TURTLES: A COMPARATIVE ANALYSIS
ABSTRACT....................................................................................................................131
INTRODUCTION...........................................................................................................132
METHODS......................................................................................................................142
RESULTS........................................................................................................................145
DISCUSSION..................................................................................................................149
APPENDIX 1 ..................................................................................................................158
TABLES..............................................................................................................159
FIGURES.............................................................................................................166
REFERENCES................................................................................................................169

viii

LIST OF TABLES

Table 0.1......................................................................................................................................... 7
A general description of traits and duration of nesting seasons of three freshwater turtle species
Table 1.1........................................................................................................................................33
Demonstration of the use of inferred paternal alleles at a polymorphic locus to distinguish among
four categories of repeat paternity (stored sperm use) that can have direct influence on male
reproductive success (RS) of painted turtles (Chrysemys picta marginata).
Table 1.2........................................................................................................................................34
Clutch size and egg widths (from x-radiographs) of 44 female painted turtles (Chrysemys picta
marginata) that did not produce any second clutches and 123 that did produce at least one second
clutches over sampling periods of 5-21 years (data = mean, standard deviation, (minimummaximum).
Table 2.1...................................................................................................................................... 70
Summary of the number of offspring, clutches, and mates per year and over four years (includes
residual paternity) for successful male Painted turtles (Chrysemys picta).
Table 3.1......................................................................................................................................114
Percents of matings (N = number of pairings) described during all years (2003-2006) between
males and females with respect to residence wetland (East, West, and non-resident).
Table 3.2................................................................................................................................ .....114
Percentages of among-year clutches produced by females where repeat paternity was observed
(N = number of comparisons) as a function of male and female marsh of residence (e.g. repeat
paternity occurring between east-resident females east-resident males).
Table 4.1......................................................................................................................................159
Summary of population characteristics, life-history trait values, morphology, physiology and
behavior for Painted turtles (CP; Chrysemys picta), Blanding's turtles (EB; Emydoidea
blandingii), and Snapping turtles (CS; Chelydra serpentina) cohabiting wetlands on the Edwin S.
George Reserve (ESGR). Summary data include mean (minimum-maximum), standard
deviation.
Table 4.2.....................................................................................................................................160
Predicted influence of traits on the incidence of multiple and repeat paternity assuming all other
traits remain constant. Where possible, rankings are provided that demonstrate the species (CP,
Chrysemys picta; EB, Emydoidea blandingii; CS, Chelydra serpentina), with the highest/lowest
ix

predicted multiple paternity based on the trait. Rationales are provided to describe the
underlying basis for each prediction.
Table 4.3......................................................................................................................................162
Primary relationship between measures of female qualities and age or size that have implications
for male reproductive success in Painted turtles (CP; Chrysemys picta), Blanding's turtles (EB;
Emydoidea blandingii), and Snapping turtles (CS; Chelydra serpentina).
Table 4.4......................................................................................................................................163
Comparison of among year reproductive intervals of females based on long-term data (>33
years) for Painted turtles (CP), Blanding's turtles (EB) and Snapping turtles (CS).
Table 4.5......................................................................................................................................164
Comparisons of the direction and magnitude of effects of factors influencing multiple paternity
in Painted turtles, Blanding's turtles, and Snapping turtles (2003-2006).
Table 4.6......................................................................................................................................165
Comparisons of factors influencing male reproductive success (total number of offspring) for
successful Painted turtle and Blanding's turtle males on the ESGR (2003-2006).

x

LIST OF FIGURES

Fig. 1.1...........................................................................................................................................35
a) The total number of reproductive painted turtle (Chrysemys picta marginata) females captured
each year at East Marsh (numbers with the last four years are the percent of clutches exhibiting
evidence of multiple paternity in years for which genetic data were available), and (b) the
proportion of second clutches produced at East Marsh each year from1983-2006.
Fig. 1.2...........................................................................................................................................37
a) The numbers of painted turtle (Chrysemys picta marginata) females with reproductive records
spanning 5–24 years) that produced first clutches only (circles), and those that produced second
clutches (triangles) each year (1983-2006), and (b) percent of females producing second clutches
(on the x-axis, 0.0 indicates females that did not produce any second clutches).
Fig. 2.1...........................................................................................................................................71
a) The number of reproductive Painted turtle (Chrysemys picta marginata) females per year
(1983-2006) with yearly occurrence of multiple paternity (2003-2006) and b) relationship
between body size and age of females.
Fig. 2.2...........................................................................................................................................72
a) Relationships between a) clutch size, and b) egg widths and body size of females, and c)
frequency of second clutch production and age groups of females.
Fig. 2.3...........................................................................................................................................73
a) Distribution of carapace lengths and (b) ages of reproductive females from East Marsh
captured in 1999 and reproductive females subsequently captured from 2000-2006 (14.6% and
51.8% of 1999 and 2000-2006 females respectively, were < 12 years old; the ages that females
reach reproductive maturity).
Fig. 2.4...........................................................................................................................................74
a-f) Histograms of body sizes of females sampled over different periods of the life history and
genetics studies.
Fig.
2.5...................................................................................................................................................75
a-f) Histograms of ages of females sampled over different periods of the life history and genetics
studies.
Fig. 2.6...........................................................................................................................................76
Percent of nests with evidence of the same male siring offspring with the same female (repeat
paternity) within- and among-years.
xi

Fig. 2.7..........................................................................................................................................77
Distribution by age and body size for all males sampled (a, b) and for males that mated
successfully in the population (c, d) 2003-2006.
Fig. 3.1........................................................................................................................................115
Description of an integrative approach combining multiple sources of long-term data used to
answer research questions [Q] pertaining to the relative importance of nesting ecology and
movements, characteristics of mated pairs, and male reproductive success, to gene flow and
spatial genetic structuring.
Fig. 3.2.........................................................................................................................................116
The (a.) numbers of resident, non-resident and total reproductive females captured each year
(1983-2006) on the E. S. George Reserve and ranks of the percent of nests with multiple
paternity from 1999–2006 (% multiple paternity for each rank was 1 = 15.4, 2 = 36.4, 3=42.9.
4=44.4. 5 = 50.0, 6 = 53.8, 7 = 55.6; average carapace length (+ 1SE) by year of (b.) resident and
(c.) non-resident reproductive females.
Fig. 3.3.........................................................................................................................................117
Distribution of Blanding's turtle (Emydoidea blandingii) nests on the 650 ha Edwin S. George
Reserve (boundary in dashed lines) from 1999-2006, residence wetland of females with the
number of nests per location are indicated within symbols (circles = east residents, squares =
west residents, triangles = non-residents of the E. S. George Reserve.
Fig. 3.4.........................................................................................................................................118
Frequency histograms of (a) ages and (b) body sizes of males that successfully mated with
females that nested on the E. S. George Reserve (ESGR) and the (c) ages and (d) body sizes of
all ESGR resident males. Ages and sizes for successful males were determined from the average
of all ages/sizes for the years in which a male was successful over the period of study (19992006). Ages for all resident males were adjusted to the age the male in 2003.
Fig. 3.5.........................................................................................................................................119
Percentage of nests with evidence of the same male siring offspring with the same female (repeat
paternity) as a function of inter-annual nesting interval.
Fig. 3.6.........................................................................................................................................120
Percentage of among-year comparisons where repeat paternity was observed from males from
the same (black) or different (gray) residence wetlands as the female.
Fig. 4.1.........................................................................................................................................166
Summaries of the a) number, b) body sizes and c) ages of reproductive female Painted turtles
(Chrysemys picta), Blanding's (Emydoidea blandingii) and Snapping turtles (Chelydra
serpentina) cohabiting wetlands on the Edwin S. George Reserve each year from 1977-2006.

xii

Fig. 4.2.........................................................................................................................................167
Conceptual diagram illustrating how population, life-history, morphology and behavior can
directly (solid arrows) and indirectly (dashed arrows) influence male reproductive success.
Fig. 4.3.........................................................................................................................................168
Percent of clutches showing evidence of repeat paternity (and sample sizes; N) within and among
years for Painted turtles, Blanding's turtles, and Snapping turtles.

xiii

INTRODUCTION

1

INTRODUCTION
Behavioral ecological research can be enhanced when life-history information is
available (Krebs and Davies 1997; Sih and Bell 2008; Bassar et al. 2010). Similarly, life-history
studies also benefit from behavioral ecology studies because information on behaviors can
identify sources of variation in body size and age specific reproductive success, which is a
central focus of life-history studies. Life history and behavioral ecology studies focus on
females, in part because of the comparative ease of observation, and because population levels of
recruitment are mediated by female reproductive success. Detailed information on male
reproductive success is rarely available, particularly for poikilothermic vertebrates, because
paternity cannot be inferred from direct observation alone, and mating behaviors of many
poikilothermic vertebrates takes place out of view (e.g., underwater). However, the importance
of including information on male reproductive success in life-history studies has recently been
demonstrated (Kokko and Mappes 2005, Kokko et al. 2008).
The study of male reproductive success of three species of long-lived turtles on the
University of Michigan‟s E. S. George Reserve (ESGR) provides two important aspects to
studies of mating systems. First, results from behavioral ecology studies are enhanced by
detailed age-specific reproductive information on individuals being studied. For many longlived organisms, accurate age-specific reproductive information is unavailable due to the inherit
difficulties and costs associated with conducting long-term studies. Secondly, more primitive
and less complex mating systems are the foundations for the evolution of complex systems. The
results from my research should provide a basis for the interpretation of comparatively more
complex mating systems that are complicated by factors such as social interactions, parental
care, and improved access to resources.
2

My dissertation research uses genetic techniques to document sources of variation in
male reproductive success in syntopic populations of three species of freshwater turtles (Midland
Painted turtles, Chrysemys picta marginata; Blanding‟s turtles, Emydoidea blandingii; and
Snapping turtles, Chelydra serpentina); see the summary of traits of the three species below
(Table 0.1). The study of male reproductive success was conducted within the context of a longterm study on life history and demography initiated by Owen Sexton (1953-1957), continued by
Henry Wilbur (1968-1972, Donald Tinkle (1975-1979), and Justin Congdon (1975-2007). Longterm data that quantify sources of variation in female reproductive quality (i.e., age- and sizespecific clutch size, clutch frequency, and egg size for all three species, and individual-based
frequencies of within-year second clutch production for Painted turtles) allows quantification of
the relative importance of mate number and female reproductive qualities to male reproductive
success.
In Chapter 1, entitled "Variation in Female Reproductive Quality and Reproductive
Success of Male Painted Turtles (Chrysemys picta marginata)" the primary objective was to
determine whether females primarily use stored sperm to fertilize second clutches within a year
as suggested by Gist and Congdon (1998). I documented how female quality can influence male
reproductive success by quantifying stored sperm use in intra-seasonally iteroparous females.
(Chapter 1 is currently in press in the Canadian Journal of Zoology (accepted August 2011).
In Chapter 2, entitled " Female quality affects male Painted turtle (Chrysemys picta
marginata) reproductive success" I quantified the relative importance of mate number and
quality to male reproductive success based on genetically determined parentage estimated from
offspring of nests collected over 4 years. I tested the following predictions: 1) mate number
contributes significantly to male reproductive success; 2) female quality contributes substantially
3

to male reproductive success; and 3) high-quality (e.g., older or larger) females will have a
higher incidence of multiple paternity.
In Chapter 3, entitled "Mating system, male reproductive success, and population
connectivity of Blanding's turtles (Emydoidea blandingii)" I characterized mating behaviors
influencing male reproductive success, and quantified the degree of spatial and genetic
structuring of threatened Blanding‟s turtles. I quantified components of variation in male
reproductive success in relation to: 1) number of mates, 2) mate quality (clutch size, clutch
frequency, egg size), 3) occurrence of multiple paternity 4) the number of eggs a male sired per
clutch, 5) repeat paternity (via stored sperm or remating), and 6) age and body size of both sexes
of adults. I also quantified the proportion of successful matings among individuals from
different resident wetlands on and off of the ESGR. Mating pair data were used to quantify
propensities for turtles to mate with individuals outside their residence wetlands to quantify gene
flow.
In Chapter 4, entitled " Factors influencing mating systems and male reproductive
success in freshwater turtles: A comparative analysis" I examined patterns of reproduction
among the three species. I tested hypotheses regarding factors associated with incidence of
multiple paternity and repeat paternity within the context of the species similarities and
differences in life history and demography. Factors associated with male reproductive success
were compared between the two species of Emydid turtles (Painted turtles and Blanding's
turtles). I did not have sufficient power for paternity analysis of Snapping turtles. The three
species are sympatric, and sampled over the same time periods to allow inter-annual variation to
be compared among the three species with respect to variation in female reproductive quality.

4

The research on three species of freshwater turtles provides opportunities to compare and
generalize findings among species to quantify how fluctuations in environmental and
demographic environments affect traits associated with reproductive success. The three turtle
species present a unique opportunity to contrast components of male reproductive success, test
for evidence of assortative mating, and aspects of the species mating system in general for
species that differ in adult sex ratio, population size, age- and size-specific variation in
reproductive quality, age at maturity and clutch size. Comparisons have here-to-fore not been
possible for long-lived poikilothermic vertebrates, and thus have the potential to have a major
impact on evolutionary and behavioral ecology fields.

5

APPENDIX 0

6

Table 0.1. A general description of traits and duration of nesting seasons of three freshwater turtle species.
Species

Mean
body
size of
females

Larger
sex

Adult sex ratio

(mm)

Min. age
at
maturity
of
females

Longevity
(years)

Clutch
size

Annual
clutch
frequency

Egg size/
Mean
body size
duration of
relationship nesting season

mean

(days)

(min-max)

Painted
turtles

136

female

2.3 M /F

7

~ 50

6

1.3 (0 – 3)

Strong

37

Blanding‟s

196

equal

Equal

14

>75

10

0.8 (0 – 1)

Weak

26

250

male

Equal

11

> 55

23

0.85 (0 – 1)

Weak

18

turtles
Snapping
turtles

7

REFERENCES

8

References
Bassar, R.D., López-Sepulcre, A., Walsh, M.R., Turcotte, M.M., Torres-Mejia, M., and Reznick,
D.N. 2010. Bridging the gap between ecology and evolution: integrating density
regulation and life-history evolution. Annals of the New York Academy of Sciences
1206(1): 17-34. doi:10.1111/j.1749-6632.2010.05706.x.
Gist, D., and Congdon, J.D. 1998. Oviductal sperm storage as a reproductive tactic of turtles.
Journal of Experimental Zoology 282: 526-534.
Kokko, H., and Jennions, M.D. 2008. Parental investment, sexual selection and sex ratios.
Journal of Evolutionary Biology 21(4): 919-948. doi:10.1111/j.1420-9101.2008.01540.x.
Kokko, H., and Mappes, J. 2005. Sexual selection when fertilization is not guaranteed. Evolution
59(9): 1876-1885. doi:10.1111/j.0014-3820.2005.tb01058.x.
Krebs, J.R., and Davies, N.B. 1997. Behavioural Ecology: An Evolutionary Approach, 4th ed.
Blackwell, Oxford.
Sih, A., and Bell, A.M. 2008. Chapter 5 Insights for Behavioral Ecology from Behavioral
Syndromes. In Advances in the Study of Behavior. Edited by Timothy J. Roper Marc
Naguib Katherine E. Wynne-Edwards Chris Barnard H. Jane Brockmann, and C. Mitani
John. Academic Press. pp. 227-281.

9

CHAPTER 1: VARIATION IN FEMALE REPRODUCTIVE QUALITY AND
REPRODUCTIVE SUCCESS OF MALE PAINTED TURTLES (CHRYSEMYS PICTA
MARGINATA)

10

Abstract
Although mate number is perceived to be the primary factor affecting male reproductive success
in polygynous systems, differences in female reproductive qualities may also influence variation
in male reproductive success. We combined 32 years of data on variation in reproductive
qualities (clutch size and clutch frequency) of female painted turtles (Chrysemys picta marginata
Agassiz, 1857) with genetic data on patterns of repeated paternity (i.e., stored sperm use) and
multiple paternity to examine the potential influence on male reproductive success. Over 24
years (1983-2006), the number of reproductive females each year averaged 84 (min-max = 62106) and on average 23% (min-max = 6%-40%) produced two clutches (intra-seasonally).
Among females with reproductive histories spanning 5-24 years (N = 167), 26% of individuals
produced only one clutch annually, whereas 74% produced two clutches within a season.
Among just intra-seasonally iteroparous females, second-clutch production varied from 7%-50%.
Repeated paternity was observed in 97.5% of 40 paired clutches and 44% of 9 among-year
comparisons of clutches from consecutive years. The frequent use of stored sperm to fertilize
sequential clutches within and potentially among years can substantially increase a male‟s
reproductive success, particularly if males can base mating decisions on phenotypic
characteristics correlated with female quality.

11

Introduction
Life-history studies have primarily focused on age-specific traits of females, but complete
understanding of the evolution of life histories will require knowledge of how interactions
between males and females influence the way each sex attempts to maximize fitness.
Inequalities in gametic investment by males and females (anisogamy) often results in different
tactics as they attempt to maximize reproductive success (Trivers 1972; Stockley 1997). For
example, the use of stored sperm by females can affect male reproductive success through sexual
conflict, including cryptic female choice, sperm competition, and post-copulatory sexual
selection (Stockley 1997; Jennions and Petrie 2000).
Male reproductive success (RS) is thought to be primarily constrained by mate number
(Bateman 1948), but male RS can also be influenced by female qualities such as reproductive
frequency, number of offspring, parental investment, and in some species the use of stored
sperm, (Darwin 1871; Trivers 1972; Kirkpatrick et al. 1990; Webster et al. 1995). Variation in
reproductive qualities of females can result from differences in: 1) resource availability among
years, 2) the ability to acquire and utilize resources among females (Stearns 1992), and 3)
characteristics of females such as body size and age (Congdon et al. 2003). Compared to males
that mate indiscriminately, individuals that have the ability to identify and mate with high-quality
females can sire more offspring within and among years.
The characteristics of turtles make them excellent models of less complex reproductive
systems (Avise 2001), particularly when genetic and long-term life-history data from the same
population are available (Clutton-Brock 1988; Linden and Møller 1989; Clutton-Brock and
Vincent 1991). In contrast to mating systems of mammals and birds, less complex systems of
poikilothermic vertebrates allows interactions between the sexes that influence reproductive
12

success to be examined in the absence of complex social systems and parental care (Trivers
1972; Clutton-brock 1991; Gross 1996; Avise 2001). Turtles were one of the first vertebrate
groups to evolve internal fertilization (Gist et al. 2000); they do not form pair bonds (Wilbur and
Morin 1988), and have little or no post-ovulatory parental care (Congdon and Gibbons 1990).
Females of many turtle species are intra-seasonally iteroparous (they produce more than one
clutch of eggs within a reproductive season), with females of some sea-turtle species laying up to
10 clutches in a year (Eckert 1987). Females often fertilize clutches of eggs with sperm from
more than one male (Galbraith et al. 1993; Pearse and Avise 2001b; Pearse et al. 2002; Lee
2008). Female turtles are able to store sperm for longer than a year (Ewing 1943; Gist and Jones
1989) and that allows temporal asynchrony between mating and fertilization (Birkhead and
Møller 1993; Shuster and Wade 2003).
Because of the short interval between sequential clutches within a reproductive season ,
Gist and Congdon (1998) hypothesized that a primary function of stored sperm is to fertilize
second clutches. Four studies (Fitzsimmons 1998; Kichler et al. 1999; Pearse et al. 2002;
Roques et al. 2006) found that stored sperm was used to fertilize sequential clutches within a
reproductive season, suggesting that that use of stored sperm is widespread among turtles.
Therefore, variation in reproductive frequency, within and among years, and a propensity for
using stored sperm to fertilize sequential clutches, can substantially influence male reproductive
success. If males discriminate among female phenotypes that are associated with an individual's
tendency to produce single and multiple clutches seasonally, among-individual variation in male
reproductive success could increase substantially.
The life-history and ecology of painted turtles (Chrysemys picta marginata Agassiz,
1857) have been studied on the University of Michigan‟s Edwin S. George Reserve (ESGR) near
13

Hell, Michigan for 43 of the past 55 years (Sexton 1959; Wilbur 1975; Tinkle et al. 1981;
Congdon et al. 2003). A summary of ESGR painted turtle characteristics follows. Males mature
at age 4-5 years and females from 6-12 years of age (Congdon et al. 2003), and maximum age of
males and females is about 40 and 50 years, respectively. Observations of mating activity of
painted turtles on the ESGR are infrequent because it occurs in highly vegetated wetlands in
early spring (Sexton 1959). Females allocate the majority of resources to first and second clutch
follicles during late summer to early fall in the year prior to egg laying (Congdon and Tinkle
1982). Although females provide no behavioral post-ovulatory parental care, they do provide
hatchlings with substantial yolk reserves necessary to sustain juveniles during the early posthatchling period (Congdon et al. 1983). Some females do not reproduce every year, but the
majority oviposit one or two clutches of eggs in terrestrial nests in late May through early July
(Tinkle et al. 1981). The shortest interval between first- and second-clutch nests is 10 days, but
soft-shelled eggs of second clutches can be detected by palpation six days after the first-clutch
nest (J. D. Congdon, unpublished data). Because the size and number of eggs in first and second
clutches are similar (Congdon et al. 2003; Harms et al. 2005), producing two clutches of eggs
essentially doubles the annual fecundity of females. Temporal and spatial separation of first and
second clutch nests reduces the probability that all eggs will be lost (i.e., not putting all eggs in
one basket).
Detailed information on the reproductive histories of females from the long-term studies
on the ESGR provides an opportunity to interpret the results from the genetic study. Data from
the long-term studies include: 1) numbers, body sizes, and ages of most adult males and females
in the population, 2) the number of females that reproduced annually, 3) annual reproductive
output of individual females (clutch size, clutch frequencies, egg size as egg widths from x14

radiographs, and relationships of reproductive variables to body size and age), and 4) the
identities of female parents and their offspring. Genotypes of adults and hatchlings allowed us to
document the frequency of first and second clutches of intra-seasonally paired clutches being
fertilized by the same male (or males) and thereby determine if females used stored sperm to
fertilize the second clutch; (Gist and Congdon 1998), whether the same male fertilized sequential
clutches between years, and the frequency of multiple paternity. Because both clutch size and
egg size increase with body size of females, and egg size and reproductive frequency increases
with age (Congdon et al. 2003), older and larger females represent individuals of higher
reproductive quality and should be sought after by males.
The primary goals of this study were to test the hypothesis that an important use of stored
sperm is to fertilize sequential clutches within a season and quantify age and body size specific
reproductive traits of females that may substantially influence male reproductive success.

Materials and Methods
Life history and Population Study
Research was conducted in accordance with the University of Michigan Animal Use and
Care Committee (UCUCA #8496). Painted Turtles (Chrysemys picta marginata) from East
Marsh on the ESGR, were intensively trapped from 1976 – 2007. All individuals were uniquely
marked by notching marginal carapace scutes at first capture in aquatic traps, on land, or at drift
fences completely surrounding East Marsh or located between wetlands and potential nesting
areas. At each capture, the date, individual identification, body size (carapace length, and
weight), reproductive condition (e.g., gravid or not gravid) were recorded. Hatchlings and
yearlings were assigned “age 1”, and at first capture, the ages of juveniles with distinct growth
15

rings were estimated by assuming one growth ring for each year of life (Gibbons 1976).
Thereafter ages of individuals were calculated from the interval between first and each
subsequent recapture.
Data on clutch size and reproductive frequency were collected over 26 consecutive years
(1983-2006), when East Marsh was completely enclosed by a 1.3 km fence during all nesting
seasons. The fence was monitored all days of each nesting season at approximately 20 minute
intervals from 0600 h until the end of painted turtle activity in the evening. Almost all females
leaving the marsh to nest (with first or second clutches) were captured each year, identified,
measured and weighed, and then X-radiographed to determine clutch size and widths of eggs
(Gibbons and Greene 1979; Hinton et al. 1997). We associated nests with females by observing
them in the act of nesting during extensive searches of nesting areas.
Over the 33 years spanning the entire study at East Marsh (1975-2007), 2,796 individuals
were marked, 13,917 recaptures were made, and females were detected with 4003 and 893 first
and second clutches, respectively. The total resident adult population at East Marsh between
2001 and 2006 was calculated from catchability data (Congdon and Gibbons 1996) based on all
captures and all methods (aquatic traps, on land, or at the fence) over six overlapping three year
periods (2000-2002, 2001-2003 2002-2004, 2003-2005, 2004-2006, 2005-2007). We used data
from extensive capture and reproductive records from East Marsh (mean = 9.5 years, min-max =
5–24 years) to assign 167 females to the categories of those that had produced „one clutch only‟
and those that produced „second clutches‟.

Paired clutches

16

Within-year pairs of clutches (N = 44 pairs; 88 nests) of eggs were obtained from nests of
35 marked females from 2003-2006. Nests were protected in situ in 2003-2004. During 2005
and 2006, eggs were transferred to a common protected natural nesting area within four hours of
nest completion. Transferred eggs were incubated in perforated plastic containers to exclude
burrowing mammal predation, and were buried approximately 6 cm below the surface of the
ground. Different methods of nest protection resulted in similar hatching successes (Chi-square;
P = 0.4867, df = 32), so samples were combined.

Genetic Samples
A sample of tail tissue (less than 2 mm) was taken from each hatchling and preserved in
95% ethanol until DNA extraction. Approximately 0.1cc of blood was collected from each
female parent and was stored in a blood storage buffer (100mM TrisHCL (pH 8.0), 100mM
EDTA, 10mM NaCl and 0.5% SDS). DNA was extracted from blood and tissue samples using a
Qiagen DNeasy kit (Qiagen), and quantified using fluorimetry.
Hatchlings from all clutches were genotyped at 7 polymorphic microsatellite loci: Cp2,
Cp10, Cp3 (Pearse et al. 2001), BTGA2, BTGA3, BTCA7 (Libants et al. 2004), and EB11
(Osentoski et al. 2002). DNA was amplified using the published conditions for polymerase
chain reaction (PCR), separated using gel electrophoresis on a 6% polyacrylamide gel, and
visualized using an FMBIOII scanner (Hitachi Inc.). Gels were hand-scored independently by
two experienced lab personnel and over 10% of PCR products from hatchlings were randomly
selected and reanalyzed electrophoretically and rescored independently to minimize scoring
errors. Estimates of genetic variability and multi-locus exclusion probability were quantified
using the program GERUD 2.0 (Jones 2005) and are based on the equations presented in (Dodds
17

et al. 1996). Analysis of stored sperm and multiple paternity were restricted to clutches with
three or more offspring, reducing the sample size to 40 pairs (80 nests) from 31 females. All 44
pairs were used for analyses of hatching success (described below).

Paternal genotype reconstruction
Paternal genotype reconstruction was performed on 40 pairs of nests (80 clutches) using
the program GERUD 2.0 (Jones 2005) based on one known (female) parent. We performed
three step-wise analyses involving paternal genotype reconstruction: 1) each nest independently,
2) offspring combined for paired clutches within-years, and 3) offspring combined for each
female among all years. Reconstructed paternal genotypes of males contributing to nests were
also compared visually (Fiumera et al. 2002; Jones et al. 2007) for consistency between first and
second clutches that were produced within-years and among years to detect use of stored sperm
(below). In cases of multiple paternity, GERUD 2.0 provides potential paternal reconstructions
along with estimates of the number of most likely male contributors based on paternal allele
counts. Because the female parent was known for all clutches of eggs, we could identify paternal
alleles and thus detect multiple paternity when three or more distinct paternal alleles were
present. Results from the visual comparison and step-wise GERUD 2.0 analyses were compared
for consistency to ensure the most accurate interpretation. Finally, we compared reconstructed
paternal genotypes of all contributing males, using the program GENECAP (Wilberg and Dreher
2004), for evidence that any male sired offspring with more than a single female within and
among years.

Detecting Repeat Paternity (the use of Stored Sperm)
18

When paternal genotype reconstructions (GERUD 2.0; Jones 2005) indicated that the
same male(s) sired sequential clutches, we identified the second nest as having “repeat
paternity”. Because of the short interval between first and second clutches within a season,
repeat paternity in intra-seasonally paired clutches was considered as evidence of the use of
stored sperm to fertilize the second clutch (Gist and Congdon 1998; Pearse et al. 2002). Because
there are ample opportunities to re-mate between years, we refer only to repeat paternity among
years. However, the male-biased adult sex ratio in the ESGR population (2 males per female)
reduces the probability of re-mating with the same male within or between years.
Previous research described the use of stored sperm based on inferences from
observations of paternal genotypes being “consistent” or “inconsistent” with respect to the first
clutch (Fitzsimmons 1998; Kichler et al. 1999; Pearse et al. 2002). Because of our focus on the
influence of female traits on male reproductive success, we expanded the category of
“consistent” to include patterns of repeat paternity that differentially influence male reproductive
success (Table 1.1). When hatchlings from first and second clutches of an individual female had
identical paternal alleles we categorized the condition as “complete” repeat paternity. Complete
repeat paternity was then subdivided into “complete type 1” (paternal alleles were consistent
with one male) or “complete type 2 (paternal alleles were consistent with more than one male).
When the composition of paternal alleles differed between the first and second clutches due to an
addition or deletion of paternal alleles, we categorized the condition as “incomplete” repeat
paternity. Incomplete repeat paternity was then divided into “inclusive incomplete” (when an
allele(s) from a new male was observed only in the second clutch), or “exclusive incomplete”
(when an allele(s) from a male siring offspring in the first clutch was not observed in the second
clutch). The terminology of inclusive and exclusive incomplete refers to the composition of
19

paternal genotypes present and not whether paternity resulted from re-matings or stored sperm
from previous years.

Statistical Analyses
Summary statistics of female attributes and life-history-traits values such as reproductive
frequency, age, body size, clutch size, and egg size (width) among females, and demographic
parameters such as the number of reproductive females for each year, and correlations between
traits were performed using a Spearman's Rank Correlation (SAS 7, 1998)(SAS 1998).
Relationships between female attributes and incidence of multiple paternity were
examined using a generalized linear mixed effect, logistic regression model (GLMM), with
“paternity” (a binary output with 0 = single paternity and 1= multiple paternity) as the dependent
variable (R 2.5 with the lme4 package for mixed models; R Development Core Team, 2007).
Since all females contributed at least one pair of nests, and paired clutches for some females
were sampled in more than one year, we accounted for structuring in the data by including both
female and year as random effects. Fixed effects included female age, female body size
(carapace length), clutch sequence (first or second), and clutch size.
When female turtles are isolated from males over several years, use of stored sperm to
fertilize subsequent clutches can result in decreased fertility and hatching success (Cuellar 1966;
Jun-Yi 1982), via sperm depletion or a decline in sperm viability (Goin et al. 1978; Gist and
Jones 1987; Palmer et al. 1998). We tested whether second clutches had a reduced probability of
hatching using a general linear model (GLM) with hatching percent as a dependent variable
(rather than a linear mixed effect model) because nests are independent events with respect to
hatching percent and the increased power allowed a more accurate test of fixed effects. We also
20

independently tested additional factors proposed to influence hatching success such as female
age (Blem et al. 1999), female body size (Ban et al. 2000), and whether a clutch was multiply or
singly-sired (Zeh and Zeh 1996; Byrne and Robert 2000; Pearse et al. 2002; Garner and Schmidt
2003; Roques et al. 2006).

Results
Frequency of second clutch production

Over the 24 years after East Marsh was fenced (1983-2006), the number of reproductive
females each year averaged 84 (SD = 12.5, min–max = 62-106; Fig. 1.1a) and varied by 46
individuals (43.4%, min-max = 58.5%-100% of the maximum number). The proportion of
reproductive females each year that produced second clutches averaged 23% (min–max = 6%40%; Fig. 1.1b). Variation in numbers of reproductive females among years was reflected in
changes in both first and second clutch producing individuals (Fig. 1.2a).
Among 167 females with extensive individual reproductive histories, individual
frequencies of second clutch production varied widely (min-max = 0%-50%; Fig. 1.2a). No
second clutches were produced by 44 females (26%), whereas 123 females (74%) produced
second clutches in some of the years sampled (Fig. 1.2a). Second clutch production averaged
25% (min-max = 7–50%; Fig. 1.2a) for the 123 females that produced second clutches.
Carapace length was positively correlated with age within all annual samples (1983-2006) of
reproductive females (all Spearman Rho (0.400-0.773), Ps < 0.006), and among just those
females that produced second clutches (all Spearman Rho (0.441-0.852), Ps < 0.004). In
addition, for just second-clutch producing females over all years combined, the frequency of
21

second clutch production was correlated with carapace length (Spearmans Rho = 0.235, P =
0.01) and age (Spearmans Rho = 0.206, P = 0.02). The annual percent of second clutches
produced in the population was positively correlated with the total number of reproductive
females documented each year from 1983-2006 (Spearman‟s Rho = 0.287, P < 0.001; Fig.
1.1a,b).

Measures of genetic diversity
The loci Cp2 and Cp10 (Pearse et al. 2001) showed evidence of null alleles and were
excluded from further analyses. For the remaining five loci, estimates of the measures of genetic
variability included the mean number of alleles per locus over five loci (17; min-max = 7-34),
mean expected heterozygosity (0.793; min-max = 0.682-0.888), and the expected exclusion
probability with one parent known (all loci = 0.994; min-max = 0.437-0.772).

Proportions of the total populations of adult males and females contributing to the paired
clutches
Over the four years of genetic sampling (2003-2006), the number of males and females in
the East Marsh population averaged 335 (min-max = 321-348) and 180 (min-max = 169–190),
respectively (adult sex ratio ~ 2:1). The proportions of high-quality females that were genotyped
(those that produced the paired clutches) and males that contributed to hatchlings in those
clutches were 6.8% and 9.6% of the estimated total population of each sex, respectively.
Within the sample of hatchlings from 40 pairs of nests, reconstructed male genotypes
indicated that no male sired offspring with more than one female that produced two clutches
within a reproductive season or among years. However, some males sired offspring with the
22

same female among consecutive years (over two years by two males and three years by one
male). In total, 32 males contributed to offspring from 24 females (operational sex ratio = 1.3:1
males to females contributing to paired clutches).

Within-year use of stored sperm
Among 40 paired clutches, 97.5% of the within-year second clutches exhibited evidence
that stored sperm was used to fertilize all or a portion of the second clutch of eggs. Among clutch
pairs, 77.5% exhibited “complete-type 1” use of stored sperm (i.e., identical paternal alleles
indicative of one male in both clutches), 7.5% exhibited “complete-type 2” use of stored sperm
(identical paternal alleles indicative of more than one male in both clutches), 7.5% displayed
inclusive incomplete use of stored sperm (i.e., some hatchlings had paternal alleles not found in
the first clutch), and 5% displayed exclusive incomplete (some hatchlings had paternal alleles not
found in the second clutch; Table 1.1). In all cases where overall composition of paternal alleles
differed between the first and second clutch, the second clutch showed evidence of incomplete
use of stored sperm.

Repeat Paternity Among-years
For seven females, paired clutches were sampled in two consecutive years (14 pairs) and
for one female in three consecutive years (3 pairs), allowing 9 among-year comparisons.
Overall, 44.4% (4 of 9 among-year comparisons, N = 8 females) of the clutches among years
showed evidence of repeat paternity (3 complete type 1, 0 complete type 2, 1 inclusive
incomplete, and 0 exclusive incomplete; Table 1.1). One female with clutches sampled over two
years, and one with clutches sampled over three consecutive years (3 among-year comparisons;
23

33.3%), had hatchlings with identical paternal alleles in all clutches. Both clutches of one female
in the second year (11.1%) had offspring fertilized by the same male as the previous year and an
additional male (evidence of inclusive incomplete use of stored sperm) and one among-year
comparison was inconclusive (Table 1.1). Four among-year comparisons (44%) had different
compositions of paternal alleles among years (no evidence of repeat paternity). For females
where relationships could be conclusively assigned, 50% of among-year comparisons (37.5% of
females) showed evidence of repeat paternity among years.

Multiple Paternity
Over 4 years, multiple paternity was evident in 16.3% of all clutches (N = 80) and varied
among years [22.2% (N = 18), 30.0% (N = 20), 8.8% (N = 34), and 0% (N = 8) in 2003, 2004,
2005 and 2006, respectively; Fig. 1.1a]. The proportion of clutches exhibiting multiple paternity
was highest in the two years with the fewest reproductive females and lowest in the two years
with the most reproductive females. The occurrence of multiple paternity did not differ
significantly between first (15%; N = 40) and second (17.5%, N = 40) clutches (GLMM, P =
0.735). However, in 5 out of the 40 pairs (12.5%), single and multiple paternity varied between
pairs of clutches, with two pairs having multiple paternity only in the first clutch, and three pairs
having multiple paternity only in the second clutch. Multiple paternity was positively, but not
significantly correlated with the number of eggs sampled from a clutch (GLMM, P = 0.08) and
female age (GLMM, P = 0.08), but not female body size (GLMM, P = 0.52), suggesting there
may be an important biological relationship between female age and probability of multiple
paternity.

24

Comparisons of clutch size and egg size in first and second clutches
Over the 24 years (1983-2006), x-radiographs were taken of 312 pairs of clutches of
eggs. Mean clutch size was significantly larger (7.2 eggs and 6.2 eggs for first and second
clutches respectively, Paired T = 4.39, P < 0.001), and egg widths (measured from Xradiographs) were slightly and significantly wider (18.1mm and 18.0 mm for first and second
clutches respectively; Paired T = 3.40, P < 0.008; Table 1.2) in first compared to second
clutches. However, min-max of clutch size were identical (3-13 eggs) and egg widths were
similar (15.4-20.8mm and 15.9-20.5mm) for first and second clutches respectively (Table 1.2).
Over the years in the study for which genetic data are available (2003-2006), the number of
offspring genotyped in first and second clutches was not significantly different (P = 0.256, df =
87). Hatching success was similar among years (95%, N = 18, 90%, N = 20, 85%, N = 40, and
89.5%, N = 10 for 2003, 2004, 2005, and 2006, respectively) and between first and second
clutches (90.7 and 86.6% respectively (GLM, P = 0.285, df = 87). Hatching success was not
associated with female age (GLM, P = 0.153, df = 80), body size (GLM, P = 0.71, df = 87), or
the incidence of multiple paternity (GLM, P = 0.421, df = 79).

Discussion
Annual male reproductive success can be affected by among-year variation in 1) total
number of reproductive females, 2) the proportion of reproductive females that produce second
clutches, and 3) the probability that a given mate will produce a second clutch (Fig. 1.1a,b). That
the proportion of second clutches is positively and significantly correlated with the number of
reproductive females suggests that resources available during the previous year may be the
source of the variation in clutch production (Stearns 1992; Kitaysky et al. 2000). Whereas it may
25

be difficult for males to predict how many females will be reproductive and what proportion will
produce two clutches in a given year, resource levels in the present year may allow males to
assess the quality of the next reproductive season (e.g., the proportion of second clutches or
number of reproductive females).
Among all females, and within the subset of intra-seasonally iteroparous females, age and
carapace length (body size) were positively correlated with frequency of second clutch
production. Males may be able use phenotypic traits associated with age (e.g., body size) to
identify not only females that lay two clutches in a year, but also identify within those females,
individuals with high propensity for producing intra-seasonally iteroparous clutches.
Additionally, other phenotypic cues may allow males to identify females that lay two clutches.
Red and yellow colors on the head, neck, and marginal scutes of adult painted turtles (and in
many other species of turtles) may vary in intensity and contrast resulting from variation in
carotenoid production, which has been suggested as a reliable indicator of individual quality in
other taxa (Kodric–Brown and Brown 1984; Badyaev and Hill 2000; Maan et al. 2006). If males
are able to evaluate indicators of female quality to target larger (older) females to increase the
probability of mating with an intra-seasonally iteroparous female, the increase in reproductive
success would be substantial (even more so if females use stored sperm to fertilize a substantial
proportion of second clutches within a year).
Gist and Congdon (1998) hypothesized that an important function of sperm storage in
turtles is to facilitate fertilization of sequential clutches within a season because of a short
receptive period between clutches. In sea turtles, ovulation of sequential clutches of eggs occurs
in less than 36 hr (Licht et al. 1979). Female turtles may not be receptive in the short time
between nesting and ovulation of the next clutch of eggs because they must recover
26

physiologically (e.g. from the accumulation of lactic acid during nesting; (Congdon and Gatten
1989; Jessop and Hamann 2004). Further, hormones involved in nesting and mating behavior
(e.g., testosterone, estradiol, and progesterone) must readjust to levels suitable for mating (Rostal
et al. 1998) because they also appear to provide information about current reproductive state.
For example, after gravid females were induced to lay eggs with oxytocin and released back into
wetlands, they moved from aquatic to terrestrial nesting areas and performed the entire sequence
of nesting behaviors without having eggs to lay (Tucker et al. 1995); J.D. Congdon, unpublished
data).
Comparisons of reconstructed paternal genotypes indicated that 97.5% of second clutches
of East Marsh females were fertilized with stored sperm [85% complete (combined type I and
type II) and 12.5% incomplete (combined exclusive and inclusive)]. The high frequency of
stored sperm use found in this study is similar to that in another population of painted turtles
(100%) in an open riverine environment (Pearse et al. 2002) and in other species of turtles
(Fitzsimmons 1998; Kichler et al. 1999; Roques et al. 2006). We found that sperm from the
male that fertilized the first clutch was used to fertilize all or part of the second clutch 77.5% of
the time, and two males shared paternity of both clutches 7.5% of the time. In 12.5% of paired
clutches, one male was not represented in either the first or the second clutch (incomplete
repeated paternity). Incomplete repeated paternity results in a male not siring any offspring in
one of the two clutches. Overall, male reproductive success has the potential to dramatically
increase without further investment in reproductive activities by mating with females that
frequently produce second clutches within a year.
The variation among females in production of second clutches could result in two male
tactics: 1) if there is no way to discriminate among females based on their reproductive potential
27

(i.e., clutch frequency), males should mate with as many females as possible (Bateman 1948), or
2) if there are phenotypic cues that allow males to identify females with a high propensity to
produce second clutches, males should increase efforts to mate with those females. Bateman
(1948) suggested that a primary mechanism for males to increase their reproductive success is
through mate number. However, tactic 2 (above) may increase a male‟s reproductive success
while maintaining or reducing the number of matings attempted with different females. Among
the ESGR males that successfully mated with an intra-seasonally iteroparous female, none were
found to have sired offspring of any other intra-seasonally iteroparous female during the four
years of this study. Our results do not support the assumption that mating with as many females
as possible is a primary way male painted turtles in this population increase reproductive
success.
Perhaps even more striking than the high incidence of repeat paternity within a year is
that the same males fertilized all, or a portion of, 44% of pairs of nests among years, and a single
male sired all offspring in 6 clutches that one female produced over three years. In European
pond turtles, a similar proportion (58%) of clutches fertilized by the same male among years was
observed, indicating that among-year use of stored sperm may also be important to male
reproductive success (Roques et al. 2006). Because females have more time (and presumably
more opportunities) to re-mate among years than within years, it is not possible to exclude remating as an explanation for the same male fertilizing paired clutches among years.
The use of stored sperm by ESGR females is the most parsimonious explanation, given a
population size of 515 adults and a sex ratio of 1.9 males per female. Our interpretation is that
re-mating is not the primary mechanism for fertilizing sequential clutches of a female intra- or
inter-seasonally. First, under the assumption of random mating, the probability of a given male
28

mating with a female = (1 male/335 total males), and for mating to occur twice with the same
2
-6
male would be (1/335) = 8.9 x 10 . Second, between reproductive seasons the majority of

females almost certainly encounter many potential mates. Third, the use of stored sperm reduces
the necessity for spatial and temporal synchrony of matings with female receptivity (Birkhead
and Møller 1993; Shuster and Wade 2003), which would allow for more time (and presumably
opportunity) to mate with more females. Finally, there is no evidence of pair bonding in turtles
(Galbraith 1993; Pearse and Avise 2001a; Roques et al. 2006).
If stored sperm is the mechanism that explains the repeat paternity among years, then the
question becomes: Why do females use stored sperm when males are plentiful and frequently
encountered? Regardless of whether repeat paternity is the result of stored sperm or re-mating,
the findings suggest that once a male is successful with a female, he has a high probability of
continuing to be successful with that female among years. Either way, the results are not
consistent with the general perception that the number of mates is the only substantial
determinant of male reproductive success.
Clutch sizes and hatching success were similar in first and second clutches. Female
turtles are able to store sperm over several years (Ewing 1943) and although declines in fertility
have been documented in turtles (Palmer et al. 1998), the decline in fertility was observed among
years, not within a season. The high probability that females use stored sperm to fertilize the
second clutch and that the hatching success of first and second clutches were similar, indicates
that mating with intra-seasonally iteroparous females has the potential to substantially increase
male reproductive success.

29

Two major influences on male reproductive success of painted turtles on the ESGR
appear to be the potential that the females have for producing second clutches and the propensity
for females to use stored sperm to fertilize sequential clutches within and among years. If males
have no way of discriminating among females, then the intensity of courtship by males and the
cost and benefits of courting or defending any female in the population should be equal.
However, if males can recognize high quality females (those that produce and use stored sperm
to fertilize multiple clutches within a year), the costs may increase since more males would be
attempting to mate with them and that could result in reducing a male‟s ability to mate with other
females. Even with increased costs, the benefits to males would apparently remain high for
those mating with females that are intra-seasonally iteroparous.
The large proportions of sequential clutches within a year that are fertilized with stored
sperm in different species (Fitzsimmons 1998; Kichler et al. 1999; Pearse et al. 2002; Roques et
al. 2006); this study) suggest that use of stored sperm within a year is widespread among turtles.
In many species of sea turtles, females lay 2-10 clutches of eggs in a single season (van Buskirk
and Crowder 1994). If repeat paternity is observed among clutches within a year (e.g.
Fitzsimmons 1998), then the benefits to male reproductive success could be great. However, the
inter-annual nesting intervals can extend 5 years or more (Limpus et al. 1994), a period that may
be too long for stored sperm use among-years. Additionally, with clutch sizes much larger
(many species over 100 eggs per clutch; Buskirk and Crowder 1994) than those observed in
painted turtles (average of approximately 7 eggs per clutch), the issue of sperm depletion within
and among years must also be considered in sea turtles. Females in other taxa (including birds,
reptiles, mammals, and insects) are also known to store sperm (Birkhead and Møller 1993;
Birkhead 1998), and several species are intra-seasonally iteroparous (Verhulst et al. 1997). If
30

frequent use of stored sperm to fertilize offspring in sequential reproductive bouts (within or
between years) occurs in these groups, it represents a potentially important component of male
reproductive success.

31

APPENDIX 1

32

Table 1.1. Demonstration of the use of inferred paternal alleles at a polymorphic locus to distinguish among four categories of repeat
paternity (stored sperm use) that can have direct influence on male reproductive success (RS) of painted turtles (Chrysemys picta
marginata).

Categories of

Paternal Alleles
First Second

Repeat Paternity

Clutch Clutch

Observed Incidence
Within Years
Among Years
N=9
N = 40 pairs
comparisons*
77.50%
33.30%
(N=31)
(N=3)

Inclusive Incomplete

A

7.50%

B
C
A
B

B
C
A
B
C

(N=3)

A

Complete: Type 2

A
B

A

Complete: Type 1

A
B

A

Implication of repeat paternity for Male RS

0.00%

7.50%
(N=3)

11.10%
(N=1)

5.00%

0.00%

Exclusive Incomplete

One male attains 100% paternity of two clutches

Two males attain partial paternity of both
clutches

Male 1 attains complete paternity of clutch 1 and
partial paternity of clutch 2
Male 2 attains partial paternity of clutch 2
Male 1 attains partial paternity of clutch 1 and
100% paternity of clutch 2
Male 2 attains partial paternity of clutch 1 and
0% paternity of clutch 2
Unable to determine without paternity analysis

B
B
(N=2)
C
A
B
2.50%
11.10%
Inconclusive
B
D
(N=1)
(N=1)
C
A
D
0.00%
44.40%
Males unable to capitalize on repeat paternity
No Evidence
B
E
(N=4)
C
F
*Nine among-year comparisons with one female sampled in three years (2 comparisons)
33

Table 1.2. Clutch size and egg widths (from x-radiographs) of 44 female painted turtles
(Chrysemys picta marginata) that did not produce any second clutches and 123 that did produce
at least one second clutches over sampling periods of 5-21 years (data = mean, standard
deviation, (minimum-maximum).
Female
Category

N
Clutches

Intra-seasonal
clutch sequence

Clutch size
6.7.0; 1.53
(2-11)

No second
clutches

263

First only

Second
clutches

312 Pairs

First

7.2, 1.46
(2-13)

Second

6.2 (3-13)

34

Egg width (mm)
17.6, 0.87
(15.1-19.5)
17.9, 0.94
(15.4-20.8)
18.0, 0.85
(15.8-20.5)

Fig. 1.1. a) The total number of reproductive painted turtle (Chrysemys picta marginata)
females captured each year at East Marsh (numbers with the last four years are the percent of
clutches exhibiting evidence of multiple paternity in years for which genetic data were
available), and b) the proportion of second clutches produced at East Marsh each year
from1983-2006.

35

Fig 1.1 (cont'd)

110

a. N = 374 females

0.0%

# females

100

8.8%

90
80
30.0%

70

22.2%

Proportion 2nd clutches

60
0.5

b. N = 4003 first and 893 second clutches

0.4
0.3
0.2
0.1
0.0
1980

1985

1990

1995

Year
36

2000

2005

Fig. 1.2. a) The numbers of painted turtle (Chrysemys picta marginata) females with
reproductive records spanning 5–24 years) that produced first clutches only (circles), and those
that produced second clutches (triangles) each year (1983-2006), and (b) percent of females
producing second clutches (on the x-axis, 0.0 indicates females that did not produce any second
clutches).

37

Fig 1.2 (cont'd)

38

REFERENCES

39

References
Avise, J.C. 2001. Introduction. J. of Hered. 92(2): 99. doi:10.1093/jhered/92.2.99.
Badyaev, A.V., and Hill, G.E. 2000. Evolution of sexual dichromatism: contribution
ofcarotenoid- versus melanin-based coloration. Biol. J. Linn. Soc. 69: 153–172.
Ban, S., Lee, H.-W., Shinada, A., and Toda, T. 2000. In situ egg production and hatching
successof the marine copepod Pseudocalanus newmani in Funka Bay and adjacent waters
off southwestern Hokkaido, Japan: associated to diatom bloom. J. Plankton Res. 22(5):
907-922. doi:10.1093/plankt/22.5.907.
Bateman, A.J. 1948. Intra-sexual selection in Drosophila. Heredity, 2: 349-368.
Birkhead, T. 1998. Sperm competition in birds. Reviews of Reproduction, 3(2): 123129.doi:10.1530/ror.0.0030123.
Birkhead, T.R., and Møller, A. 1993. Sexual selection and the temporal separation of
reproductive events: sperm storage data from reptiles, birds and mammals. Biol. J. Linn.
Soc. 50(4): 295-311.
Blem, C.R., Blem, l.L.B., and Barrientos, C.I. 1999. Relationships of Clutch Size and
HatchingSuccess to Age of Female Prothonotary Warblers. Wilson Bull. 111(4): 577581.
Byrne, P.G., and Robert, J.D. 2000. Does multiple paternity improve fitness of the frog Crinia
Georgiana? Evolution, 54(3): 968-973. doi:10.1111/j.0014-3820.2000.tb00096.x.
Clutton-Brock, T.H. 1988. Reproductive success studies of individual variation in contrasting
breeding systems. The University of Chicago Press, Chicago, IL.
Clutton-Brock, T.H. and Vincent, A.C. 1991. Sexual selection and the potential reproductive
rates of males and females. Nature (London) 351: 58-60.
Clutton-Brock, T.H. 1991. The Evolution of Parental Care. Princeton University Press,
Princeton, NJ.
Congdon, J.D., and Gatten, R.E., Jr. 1989. Movements and energetics of nesting Chrysemys
picta. Herpetologica, 45(1): 94-100.
Congdon, J.D., and Gibbons, J.W. 1990. The evolution of turtle life histories. In Life History and
Ecology of the Slider Turtle. Edited by J.W. Gibbons. Smithsonian Institution Press,
Washington, D.C. pp 45-54.
Congdon, J.D., and Gibbons, J.W. 1996. Structure and dynamics of a turtle community over two
40

decades. In Long-term Studies of Vertebrate Communities. Edited by M.L. Cody and
J.A. Smallwood. Academic Press, San Diego, Calif. pp 137-156.
Congdon, J.D., Tinkle, D.W., Breitenbach, G.L., and van Loben Sels, R.C. 1983. Nesting
Ecology and Hatching Success in the Turtle Emydoidea blandingi. Herpetologica, 39(4):
417-429.
Congdon, J.D., Nagle, R.D., Kinney, O.M., van Loben Sels, R.C., Quinter, T., and Tinkle, D.W.
2003. Testing hypotheses of aging in long-lived painted turtles (Chrysemys picta). Exp.
Geron. 38(7): 765-772.
Congdon, J.D., and Tinkle, D.W. 1982. Reproductive energetics of the painted turtle (Chrysemys
picta). Herpetologica, 38(1): 228-237.
Cuellar, O. 1966. Delayed fertilization in the lizard Uta stansburiana. Copeia 3: 549-552.
Darwin, C. 1871. The descent of man, and selection in relation to sex. Princeton University
Press, Princeton, NJ.
Dodds, K., Tate, M., McEwan, J., and Crawford, A. 1996. Exclusion probabilities for pedigree
testing farm animals. Theor. Appl. Genet 92: 966-975.
Eckert, K.L. 1987. Environmental unpredictability and leatherback Sea Turtle (Dermochelys
Coriacea) nest loss. Herpetologica, 43(3): 315-323.
Ewing, H. 1943. Continued fertility in female box turtles following mating. Copeia: 112–114.
Fitzsimmons, N.N. 1998. Single paternity of clutches and sperm storage in the promiscuous
green turtle (Chelonia mydas). Mol. Ecol. 7(5): 575-584.
Fiumera, A.C., Porter, B.A., Grossman, G.D., and Avise, J.C. 2002. Intensive genetic assessment
of the mating system and reproductive success in a semi-closed population of the mottled
sculpin, Cottus bairdi. Mol. Ecol. 11(11): 2367-2377.
Galbraith, D. 1993. Multiple paternity and sperm storage in turtles. Herpetol.l J. 3: 117–123.
Galbraith, D., White, B., Brooks, R., and Boag, P. 1993. Multiple paternity in clutches of
snapping turtles (Chelydra serpentina) detected using DNA fingerprints. Can. J. Zool. 71:
318–324.
Garner, T.W.J., and Schmidt, B.R. 2003. Relatedness, body size and paternity in the alpine newt,
Triturus alpestris. Proc. R. Soc. Lond. B. Biol. Sci. 270(1515): 619-624.
doi:10.1098/rspb.2002.2284.
Gibbons, J.W. 1976. Aging phenomena in reptiles pp 454-475. EAR, Inc., Maine.
41

Gibbons, J.W., and Greene, J.L. 1979. X-ray photography: a technique to determine reproductive
patterns of freshwater turtles. Herpetologica, 35: 86–89.
Gist,D. and Jones, J. 1989. Sperm Storage within the oviduct of turtles. J. Morphol 199: 379-384.
Gist, D., and Congdon, J.D. 1998. Oviductal sperm storage as a reproductive tactic of turtles. J.
Exp. Zool. 282: 526-534.
Gist, D.H., and Jones, J. 1987. Storage of sperm in the reptilian oviduct. Scanning Microscopy,
1: 1839-1849.
Gist, D.H., Turner, T.W., and Congdon, J.D. 2000. Chemical and thermal effects on the viability
and motility of spermatozoa from the turtle epididymis J. Repr. Fert. 119: 271–277.
Goin, C.J., Goin, O.B., and Zug, G.R. 1978. Introduction to herpetology. W.H. Freeman and Co.,
New York.
Gross, M.R. 1996. Alternative reproductive strategies and tactics: diversity within sexes. Trends
Ecol. Evol. 11(2): 92-98.
Harms, Heidi K., Paitz, Ryan T., Bowden, Rachel M., and Janzen, Fredric J. 2005. Age and
season impact resource allocation to eggs and nesting behavior in the painted turtle.
Physiol. Biochem. Zool. 78(6): 996-1004. doi:10.1086/432920.
Hinton, T.G., Fledderman, P.D., Lovich, J.E., Congdon, J.D., and Gibbons, J.W. 1997.
Radiographic determination of fecundity: is the technique safe for developing turtle
embryos? Chelonian Conserv. Biol. 2: 409–414.
Jennions, M.D., and Petrie, M. 2000. Why do females mate multiply? A review of the genetic
benefits. Biol. Rev. Camb. Philo. Soc. 75(1): 21-64.
Jessop, T.S., and Hamann, M. 2004. Hormonal and metabolic responses to nesting activities in
the green turtle, Chelonia mydas. J. Exp. Marine Biol. Ecol. 308(2): 253-267.
Jones, A.G. 2005. Gerud 2.0: a computer program for the reconstruction of parental genotypes
from half-sib progeny arrays with known or unknown parents. Mol. Ecol. Notes, 5(3):
708-711.
Jones, B., Grossman, G.D., Walsh, D.C.I., Porter, B.A., Avise, J.C., and Fiumera, A.C. 2007.
Estimating differential reproductive success from nests of related individuals, with
application to a study of the mottled Sculpin, Cottus bairdi. Genetics, 176(4): 2427-2439.
doi:10.1534/genetics.106.067066.

42

Jun-Yi, L. 1982. Sperm retention in the lizard Chamaeleo hoehnelii. Copeia, 2: 488-489.
Kichler, K., Holder, M.T., Davis, S.K., Marquez-M, R., and Owens, D.W. 1999. Detection of
multiple paternity in the Kemp's ridley sea turtle with limited sampling. Mol. Ecol. 8:
819-830.
Kirkpatrick, M., Price, T., and Arnold, S.J. 1990. The Darwin-Fisher theory of sexual selection
in monogamous birds. Evolution, 44(1): 180-193.
Kitaysky, A.S., Jr, G.L.H., Flint, E.N., Rubega, M.A., and Decker, M.B. 2000. Resource
allocation in breeding seabirds: responses to fluctuations in their food supply. Mar. Ecol.
Prog. Ser. 206: 283-296. doi:10.3354/meps206283.
Kodric–Brown, A., and Brown, J.H. 1984. Truth in advertising: The kinds of traits favored by
sexual selection. Am. Nat. 124: 305-322.
Lee, P.L.M. 2008. Molecular ecology of marine turtles: New approaches and future directions.
J. Exp. Mar. Biol. Ecol. 356(1-2): 25-42. doi:DOI: 10.1016/j.jembe.2007.12.021.
Libants, S., Kamarainen, A.M., Scribner, K.T., and Congdon, J.D. 2004. Isolation and crossspecies amplification of seven microsatellite loci from Emydoidea blandingii. Mol. Ecol.
Notes, 4(2): 300-302.
Licht, P., Wood, J., Owens, D.W., and Wood, F. 1979. Serum gonadotropins and steroids
associated with breeding activities in the green sea turtle Chelonia mydas : I. Captive
animals. Gen. Comp. Endocrinol. 39(3): 274-289.
Limpus, C.J., Eggler, P., and Miller, J.D. 1994. Long interval remigration in eastern Australian
Chelonia. NOAA Technical Memorandum NMFS-SEFSC-341 ed.
Linden, M., and Møller, A.P. 1989. Cost of reproduction and covariation of life history traits in
birds. Trends Ecol. Evol. 4(12): 367-371. doi: 10.1016/0169-5347(89)90101-8.
Maan, M.E., van der Spoel, M., Jimenez, P.Q., van Alphen, J.J.M., and Seehausen, O. 2006.
Fitness correlates of male coloration in a Lake Victoria cichlid fish. Behav. Ecol. 17(5):
691-699. doi:10.1093/beheco/ark020.
Osentoski, M.F., Mockford, S.M., Wright, J.M., Snyder, M., Herman, T.B., and Hughes, C.R.
2002. Isolation and characterization of microsatellite loci from Blanding's turtle,
Emydoidea blandingii. Mol. Ecol. Notes, 2(2): 147-149.
Palmer, K.S., Rostal, D.C., Grumbles, J.S., and Mulvey, M. 1998. Long-term sperm storage in
the desert tortoise (Gopherus agassizii). Copeia (3): 702-705.
43

Pearse, D., Janzen, F., and Avise, J. 2002. Multiple paternity, sperm storage, and reproductive
success of female and male painted turtles (Chrysemys picta) in nature. Behav. Ecol.
Sociobiol. 51(2): 164-171.
Pearse, D.E., and Avise, J.C. 2001 Turtle mating systems: behavior, sperm storage, and genetic
paternity. J. Hered. 92(2): 206-211. doi:10.1093/jhered/92.2.206.
Pearse, D.E., Janzen, F.J., and Avise, J.C. 2001. Genetic markers substantiate long-term storage
and utilization of sperm by female painted turtles. Heredity, 86: 378-384.
R. Core Development Team 2007. R: A language and environment for statistical computing. R
Foundation for Statistical Computing. Available from http://www.R-project.org.
[Accessed July 2009]
Roques, S., Díaz-Paniagua, C., Portheault, A., Pérez-Santigosa, N., and Hidalgo-Vila, J. 2006.
Sperm storage and low incidence of multiple paternity in the European pond turtle, Emys
orbicularis: A secure but costly strategy? Biol. Conserv. 129(2): 236-243.
Rostal, D.C., Owens, D.W., Grumbles, J.S., MacKenzie, D., and Amoss, M.S. 1998. Seasonal
Reproductive Cycle of the Kemp's Ridley's Sea Turtle (Lepidochelys kempi). Gen.
Comparat. Endocrinol. 109: 232-243.
SAS. 1998. SAS OnlineDoc(TM), Version 7-0. SAS Institute Inc., Cary, NC.
Sexton, O.J. 1959. Spatial and temporal movements of a population of the painted turtle,
Chrysemys picta marginata (Agassiz). Ecol. Monogr. 29(2): 113-140.
doi:http://www.jstor.org/stable/1942200.
Shuster, and Wade. 2003. Mating Systems and Strategies. Princeton University Press Princton,
New Jersey.
Stearns, S.C. 1992. The Evolution of Life Histories. Oxford University Press Inc., New York.
Stockley, P. 1997. Sexual conflict resulting from adaptations to sperm competition Trends Ecol.
Evol. 12(4): 154-159 doi:10.1016/S0169-5347(97)01000-8
Tinkle, D.W., Congdon, J.D., and Rosen, P.E. 1981. Nesting frequency and success: implications
for the demography of painted turtles. Ecology, 62: 1426-1432.
Trivers, R.L. 1972. Parental investment and sexual selection pp. 136-179. Aldine Transaction,
Chicago.

44

Tucker, J.K., Janzen, F.J., and Paukstis, G.L. 1995. Oxytocin induced nesting behavior in
females of the red-eared turtle, Trachemys scripta elegans, without oviductal eggs.
Herpetological Review, 26: 138.
van Buskirk, J., and Crowder, L. 1994. Life-history variation in marine turtles. Copeia 1: 66-81.
doi:Stable URL: http://www.jstor.org/stable/1446672.
Verhulst, S., Tinbergen, J.M., and Daan, S. 1997. Multiple breeding in the Great Tit. A trade-off
between successive reproductive attempts? Funct. Ecol. 11(6): 714-722.
Webster, M.S., Pruett-Jones, S., Westneat, D.F., and Arnold, S.J. 1995. Measuring the effects of
pairing success, extra-pair copulations and mate quality on the opportunity for sexual
selection. Evolution. 49: 1147-1157.
Wilberg, M.J., and Dreher, B.P. 2004. genecap: a program for analysis of multilocus genotype
data for non-invasive sampling and capture-recapture population estimation. Mol. Ecol.
Notes 4(4): 783-785.
Wilbur, H.M. 1975. The Evolutionary and mathematical demography of the turtle Chrysemys
picta. Ecology 56(1): 64-77.
Wilbur, H.M., and Morin, P.J. 1988. Life history evolution in turtles. In Biology of the Reptilia.
Vol. 16 Edited by C. Gans and R.B. Huey. Alan Liss, New York. pp. 387-439
Zeh, J.A., and Zeh, D.W. 1996. The evolution of polyandry I: intragenomic conflict and genetic
incompatibility. Proc. R. Soc. Lond. B. Biol. Sci. 263(1377): 1711-1717.

45

CHAPTER 2: FEMALE QUALITY AFFECTS MALE PAINTED TURTLE (CHRYSEMYS
PICTA MARGINATA) REPRODUCTIVE SUCCESS

46

ABSTRACT
Male reproductive success (RS) is often associated with mate number. However female
reproductive qualities (e.g., clutch size, egg size, and clutch frequency) and propensity to use
stored sperm to fertilize sequential clutches can also contribute to male RS. Life history data
on Painted turtles (Chrysemys picta marginata) collected from 1983-2007 were used to
quantify variation in reproductive quality of females. Genetic determinations of paternity
estimated from 155 nests from 59 females during 2003-2006 were used to quantify the
number of offspring sired by each male. Mate quality (primarily intra- and inter-annual
clutch frequency) was the major determinant of male RS rather than mate number. The
number of offspring sired was influenced by repeat paternity, resulting from rematings or
female use of stored sperm, that occurred in > 95% and 38% of clutches produced within or
among years, respectively. Older females produced clutches of eggs both within season and
among years more frequently than did young individuals. The incidence of multiple
paternity was also significantly higher in clutches of older females, particularly in years when
the total numbers of reproductive females were low. However, males sired offspring with
only one female within years (1.05 females) and over four years (1.09 females). Results
suggest that mating with more than a single female is difficult or costly; however, factors
such as prolonged courtship or mate guarding have not been reported in turtles. The skewed
adult sex ratio (2.3 males/female) may contribute to costs associated with courtship and
successful mating of males and females.

47

Introduction
A combination of ecological, demographic and genetic data allow the relative importance
of factors influencing male reproductive success (RS) to be indentified (Rowell and Servedio
2009; Salehialavi et al. 2011). Although male RS is often associated with mate number
(Bateman 1948; Clutton-Brock and Vincent 1991; Andersson 1994), RS of males may also be
influenced when females vary in quality (Andersson 1994; Owens and Thompson 1994;
Servedio and Lande 2006).
The relative costs and benefits of male mating tactics can be difficult to discern in species
with complex social systems. Therefore we examined the relative influence of mate number and
mate quality has on male RS in the relatively simple social system of Painted turtles (Chrysemys
picta). Painted turtles do not form pair bonds, have relatively simple courtship behaviors, and
provide no post-ovulatory parental care; however, females frequently use stored sperm to
fertilize second clutches (Pearse and Avise 2001; Pearse et al. 2001; Pearse et al. 2002; McGuire
et al. 2011). Reproductive qualities of female Painted turtles include clutch size, egg size, and
reproductive frequency (number of clutches produced within and among years) and all increase
with female body size and age in the study population (Congdon and Gibbons 1987; Congdon et
al. 2003). Approximately one third of females in study population produce only one clutch of
eggs per year, two thirds produce more than one clutch within a season (intra-annual iteroparity),
and some females do not reproduce each year (McGuire et al. 2011).
Although painted turtles live in resource rich wetlands, per capita resource availability
may be limited due to resource harvest and processing costs (Congdon 1989). If resources are
limited, production of second clutches in a given year may result in longer annual intervals
between reproductive bouts and that would reduce differences in reproductive output between
48

females that produce only one versus females that produce two clutches annually. Regardless of
potential constraints on second clutch production, variation in female qualities will influence the
RS of males, and if males can access female quality, selection should favor males that do so.
A study of the life history, demography, and reproductive ecology of Painted turtles on
the Edwin S. George Reserve (ESGR) near Hell, MI was conducted each year from 1975-2007.
We used detailed information on age- and size-specific variation among individuals of both
sexes, and long-term data on clutch size, egg size, and inter-clutch intervals of individual females
to document variation in female reproductive qualities.
Long term life history and population size data were combined with genetics data on
repeat paternity, multiple paternity, mate number to support interpretations and inferences about
the mating tactics and male RS. Genetic data establishing paternity of offspring from nests were
collected over 4 years to quantify how variation in the number and reproductive quality of
females influences male RS within and among years.
The goals of the study were to test the following predictions: 1) if resources are limiting,
the length of inter-clutch intervals will be longer for females producing second clutches than for
those producing a single clutch, 2) mate number will be a substantial component of male RS; 3)
female quality will be a substantial component of male RS; and 4) if males can assess
reproductive quality of females then individuals with high qualities (e.g., larger or older females
that have larger clutch sizes and shorter inter-clutch intervals) will have a higher incidence of
multiple paternity.

Materials and Methods
Field sample collection
49

Life history and genetic data were restricted to a single permanent wetland on the ESGR
(East Marsh) that was completely encircled by a 1.3 km drift fence from 1983-2006. The fence
was located between the wetland and nesting areas and was walked at approximately 20 min
intervals from 0600 h until the end of Painted turtle activity in the evening during all days of the
nesting season. Because nesting almost always takes more than an hour to complete, the
monitoring frequency on the fence was frequent enough to observe any females attempting to
construct a nest near the fence. Therefore, almost all females leaving the marsh to nest each year
(with first and second clutches) were captured and identified (Gibbons and Greene 1979; Hinton
et al. 1997).
During 15 of the years from 1976-2007, Painted turtles from East Marsh were intensively
trapped (mean aquatic captures per year 360, min–max = 219-578, SD = 105.06). During the
other years the study was restricted to the nesting season and trapping was less intense. At first
capture (in aquatic traps, dip nets, by hand, on the fence, or on land), all individuals were
uniquely marked by notching marginal carapace scutes. At each capture the date, location, sex,
and body size (carapace length) were recorded. Just prior to and during the nesting season, all
females all females on land were X-radiographed to determine clutch size and widths of eggs.
At first capture the ages of juveniles with distinct growth rings were estimated by assuming one
growth ring for each year of life (Gibbons 1976). Hatchlings and yearlings were assigned “age
1”. Ages of individuals were then calculated from the interval between the first capture (where
age was assigned) and each subsequent recapture.
Data on the number of reproductive females, their body sizes and ages, clutch sizes, egg
widths, were collected for each of 24 consecutive nesting seasons (1983-2006), the period that
50

East Marsh was enclosed within a drift fence. Data from extensive captures and reproductive
records of 167 East Marsh females (mean recapture intervals = 12.0, min-max = 5-33 years, SD
= 6.45) were used to assign individual probabilities of second-clutch production.
Offspring were associated with mothers by observing females in the act of nesting during
frequent searches of nesting areas. Post-nesting females returning to the fence without eggs were
recorded as non-gravid. Assignment of a reproductive event as a first or second clutch was
based on recapture histories when gravid females exited the wetland 10 or more days after their
first clutch and from differences in the number or position of eggs in sequential X-radiographs
within a nesting season. Because a few females did not return the East Marsh after nesting,
clutches of any individual caught in a nesting area > 15 days post first clutch (supported by Xradiograph data when available) were also assigned as a second clutch.
Hatchling sample collection

From 2003-2006 a total of 155 clutches of eggs and 1054 hatchlings obtained from 59
marked females were used for genetic analyses (see below). Nests were protected in situ in
2003-2004, and eggs were transferred to a common protected natural nesting area within four
hours of nest completion during 2005 and 2006. Transferred eggs were incubated in perforated
plastic containers to exclude burrowing mammal predators, and were buried approximately 6 cm
below the surface of the ground.

Genetic data collection
Approximately 1cc of blood was taken from adults and stored at room temperature in a
buffered solution (100mM TrisHCL (pH 8.0), 100 mM EDTA, 10 mM NaCl and 0.5% SDS)
51

until DNA was extracted. A < 2 mm tail tip sample was taken from all hatchlings and preserved
in 95% EtOH and blood samples were obtained from the caudle vein of adults and preserved in
blood storage buffer. Both sample types were stored at ambient temperature until they were
moved to refrigeration (4°C). DNA was extracted from blood from adults (N = 535) and tissue
from hatchlings (N = 1054) using a Qiagen DNeasy kit (Qiagen Inc, Valencia, CA) and
quantified using fluorimetry.
All samples were genotyped at 7 microsatellite loci (Cp2, Cp3, Cp10, (Pearse et al.
2001a), BTGA2, BTGA3, BTCA7, (Libants et al. 2004) and EB11 (Osentoski et al. 2002).
Amplification was performed using polymerase chain reaction (PCR) using the published
conditions and separated using gel electrophoresis on a 6% polyacrylamide gel. Products were
visualized using an FMBIOII scanner (Hitachi Inc., Tokyo, Japan). All gels were independently
scored by a minimum of two experienced laboratory personnel. To estimate error, and ensure
genotyping consistency among samples, 10% of adults were randomly selected and re-genotyped
for all loci. Estimates of expected heterozygosity, number of alleles, the exclusion probability
with one parent known, and tests for Hardy-Weinberg Equilibrium were performed using the
program CERVUS 3.1.0 (Kalinowski et al. 2007).
Paternity Assignment
Some commonly used parentage programs, such as CERVUS 3.0.1 (Kalinowski et al.
2007) or COLONY (Wang 2004) can over-estimate the number of contributing sires (Fiumera et
al. 2002; Jones et al. 2007), particularly when offspring are non-independent. Such is the case
when each offspring within a clutch shares the same known maternal parent. As a result, treating
each offspring as an independent event fails to incorporate prior information available in the
52

progeny arrays, and can lead to over-estimation of the number of contributing sires (Jones et al.
2007).
Because the mothers of Painted turtle offspring produced during the study were known
based on direct observation while constructing nests and offspring are grouped within a nest, we
used a two-tiered approach to assign a male to each offspring. We first assigned male(s) to a
clutch using the program NEST (Jones et al. 2007). We then assigned males to individual
offspring without the clutch grouping using program CERVUS 3.0.1 (Kalinowski et al. 2007).
We assigned a male to each offspring when the male(s) assigned by the NEST program also were
determined to have high probability of paternity based on a positive LOD score and with no sireoffspring genotype mismatches as determined by program CERVUS. Joint assignment of males
to offspring by both programs reduces over-estimation bias. We also reconstructed male
genotypes from offspring genotypes for each clutch using program GERUD 2.0 (Jones 2005).
Reconstructed male genotypes were compared to the genotypes of candidate ESGR males using
the program GENECAP (Wilberg and Dreher 2004) that allows reconstruction and comparisons
of multi-locus genotypes from males. GERUD 2.0 is highly accurate at assessing the number of
sires for a clutch when there are fewer than 6 sires (Sefc and Koblmuller 2008). The paternal
assignments from program NEST were compared to the assignments from GENECAP and
GERUD 2.0 along with visual assignments using the method described by (Fiumera et al. 2002)
to ensure consistency in paternal assignment. Repeat paternity was identified when the same
male was observed siring offspring with the same female in successive clutches within or among
years.
Statistical Analyses
53

Female reproductive quality - We tested whether female age was predictive of reproductive
frequency using mixed model analysis (GLMM). Reproductive intervals were estimated as the
number of years between successive reproductions (two clutches within a year were assigned an
interval of 0). Inter-nest intervals of > 4 yrs were removed from analyses because durations of 5
or more years were considered cases where pre-nesting movements occurred prior to monitoring
the fence or were due to injury of disease.
Because extensive reproductive histories were available for each of 167 females and all
females reproduced a number of times over the period 1983-2006, we included females as a
random effect with a random slope and intercept (Schielzeth and Forstmeier 2009). We assigned
female age, body size, and reproductive frequency as fixed effects. Of the 167 females with
extensive reproductive histories spanning 5-30 years, 44 produced a maximum of one clutch per
year and 123 produced a maximum of two clutches a year. We examined the variation among
females for relationships within season and among year reproductive frequencies. Specifically,
we tested whether the production of a second clutch (at year t) increased the time until the next
clutch was produced (during year t+1) using a binomial parameter (0,1) indicating whether the
preceding reproductive event involved was characterized by production of one or two clutches.
Male reproductive success - We quantified whether male attributes were significantly associated
with offspring numbers sired using linear mixed models (GLMM); restricting analysis to only
successful males. We included female as a random effect and tested the fixed effects of male age
and body size, total number of mates, total number of clutches sired per year, number of clutches
sired in all years, and multiple paternity of clutches sired on male RS (defined as the total
number of offspring sired).
54

We also tested whether body size and ages of males and females and clutch size of
females were associated with incidence of multiple paternity. To reduce structuring in the data
set, female and male parameters were run in separate models. For example, associations
between the occurrence of multiple paternity and female attributes were tested in models where
each clutch was represented once. Male attributes associated with multiple paternity were tested
in models where a clutch may be represented as many as two times (once for each male).
Evidence for assortative mating- We tested for evidence of assortative mating as a function of
ages and sizes of males and females of mating pairs using two independent tests. First, we used
a mixed model to determine whether male age and size was predicted by the mated female‟s age
and size. We included female as a random effect (intercept) in all models. Second, we tested
whether ages or sizes of mated pairs deviated from random pairs of males and females by
generating random mating pairs from a pool of all adult females and males sampled. Because
not all females reproduce in every year, and some females reproduce in more than one years, we
randomly sampled (with replacement) males and females to produce 155 mating pairs to have a
similar sample size as our genetic data. For each mating pair, we calculated ratios of female
age/size to male age/size (e.g., female age divided by male age for the mating pair) as described
by Rowe et al. (2007). Frequency distributions of ratios for the randomly generated mating pairs
were then compared with the actual mating pairs using a G-test. All statistical analyses were
implemented in the statistical software R (R Development Team 2006).

Results
Female Reproductive Qualities.

55

Over the 24 years that East Marsh was fenced (1983-2006), an average of 91.5 females
reproduced each year (Fig. 2.1a; min–max = 55-102; SD = 16.04), and on average, 26.0% (min –
max 13.0% – 45.0%, SD = 0.89) of nesting females produced second clutches each year.
Because female body sizes and ages are positively correlated (Fig. 2.1b) and both are positively
correlated with reproductive qualities (Congdon and Gibbons 1987; Congdon et al. 2003), we
examined whether body size and ages varied among years with different numbers of females.
The ranks of carapace lengths, ages, and number of reproductive females each year were not
correlated (Spearman Rho = 0.351, P = 0.09); Spearman Rho = 0.707, P = 0.74, respectively).
Clutch size and egg widths of first clutches averaged 6.9 (min-max = 3-12, SD = 1.48, N
= 1479) and 17.6 mm (min-max = 13.6-22.0, SD = 1.03, N = 1607), respectively and for second
clutches averaged 6.7 (min-max = 3-13, SD = 1.55, N = 364) and 17.8 mm (min-max = 15.617.9, SD = 1.03 N = 359), respectively. Clutch size and egg width were positively correlated
with body size of females (Fig. 2.2a,b). The frequency of second-clutch production increased
from 10% in primiparous females to 40% in the oldest females (Fig. 2.2c).
Among 167 females with extensive reproductive histories spanning 5-33 years, 44 (26%)
produced only one clutch in any year and 123 produced second clutches in 7-50% of years
monitored. The average frequency of second clutch production of all 167 females was 20%
(min-max = 0% - 50%, SD = 0.14). Among just those 123 females that produced second
clutches, second clutch production averaged 25% (min-max = 7–50%). Females with the highest
second clutch production rates had lower inter-year reproductive intervals (GLMM, t = -3.50, P
<0.001), and older individuals had shorter reproductive intervals than did younger females
(GLMM, t = -2.141, P < 0.05). Among-year reproductive intervals of females that only
56

produced first clutches were longer (mean = 1.39 years, SD = 0.73, min - max = 1-4, than the
among year intervals females that produced two clutches per year (mean = 1.13 years, SD = 0.41,
min-max 1-4; GLMM; t = -3.96, P < 0.001).
From 1999-2006 (the years that genetic samples were taken from adults), a total of 216
gravid females were captured at East marsh, and the number of females that reproduced each
year averaged 89 (min – max 70 – 106, SD = 13.13). In 1999, 100 reproductive females were
captured, and 116 different reproductive females were captured over the next 7 years. The range
of body sizes of the 116 females was similar to the females in the 1999 sample (Fig. 2.3a). A
mean of 16.6 (min-max = 11-28; SD = 5.80) new females were captured in each of the following
7 years and 42% (12% - 65%) of individuals were 6-11 years old (the ages of primiparous
females that generally have low reproductive qualities). Compared to the distribution of females
ages in 1999 (Fig. 2.3b), the distribution of ages of new individuals captured from 2000-2006
was biased toward young females (Fig. 2.3c). The proportions of new females < 12 years of age
captured during the years genetic samples were taken from hatchlings (2003-2006), were 45,
43%, 43%, and 65%, respectively.
Age and Size Distributions of Females
Body sizes (carapace length) of reproductive females with nests sampled to determine
paternity (N = 59 females) from 2003-2006 were not different from females captured from 19832
2006 (Fig. 2.4b; Chi-square, X = 2.85, df = 6, P = 0.827). Age distributions from 2003-2006

were normally distributed (Fig. 2.5a), but differed significantly from ages of reproductive
2
females captured from 1983-2006 (Fig. 2.5b; Chi-square, X = 48.81, df = 6, P < 0.001). The

57

size and age distributions of females in years with low numbers of reproductive females (2003
and 2004; Fig. 2.4c,d and Fig. 2.5c,d) were not different from the distributions of sizes and ages
of females in years with high numbers of reproductive females (2005 and 2006; Fig. 2.4e,f and
2
2
Fig. 2.5e,f; X = 3.29, df = 6, P = 0.772, age, X =7, df = 6, P = 0.321).

Genetic data and patterns of variation in multiple paternity
All loci were polymorphic and in Hardy-Weinberg Equilibrium. Estimates of genetic
variability included the mean number of alleles per locus over the five loci (mean = 17; min-max
across loci = 7 - 34), mean expected heterozygosity (0.793; min - max among loci = 0.6820.888), and the expected paternal exclusion probability with one parent known was 0.998.
Empirically estimated multi-locus error rate was less than 1%. No evidence of gametic
disequilibrium was detected (P > 0.05).
Of the 155 nests sampled to determine paternity, males were assigned to offspring in 108
nests (70%). Concordance in paternity assignment among programs (NEST, CERVUS, GERUD
with GENECAP) was 94%. Nests where offspring were assigned inconsistently among
programs were excluded from further analyses. Comparisons between paternity assignment
using the multiple-program approach to the "most-likely male" approach from CERVUS resulted
in 24.7 % of offspring assigned to different males, and in all cases would have resulted in an
overestimation of the number of contributing sires (as well as multiple paternity).
Multiple paternity was observed in 14.1% of clutches of eggs sampled and the
proportional occurrence varied among years (30.0, 12.5, 14.3 and 6.1% for 2003-2006,
respectively). The incidence of multiple paternity in clutches of females that produced second
58

clutches was similar to clutches of females that only reproduced once in a year (13.5% versus
2
13.8%; X = 0.048, df = 1, P = 0.83). Assuming that all males remain reproductive each year,

variation in the total number of reproductive females will cause the operational sex ratio in East
Marsh to change over the four years that multiple paternity was determined (Fig. 2.1a).
Therefore, we tested whether factors associated with incidence of multiple paternity differed in
years with high (2005, 2006) or low (2003, 2004) numbers of reproductive females. The
proportion of clutches from second-clutch producing females that were multiply sired was
different between the two years with high and two years with low numbers of reproductive
2
females (X = 292.55, df = 11, P <0.001). In years with low numbers of reproductive females,

offspring in one or both clutches of females producing second clutches were significantly more
likely to be sired by >1 male (higher incidence of multiple paternity) than clutches of females
producing single clutches (20.9 vs. 12.5%, respectively). In years when the number of
reproductive females was high, the incidence of multiple paternity was 9.2% and 14.3% in
females producing only single clutches and those producing second clutches, respectively, but
the differences were not significant.
Overall, clutch size and female body size (carapace length) were not significant
predictors of multiple paternity (GLMM, clutch size, Z = -0.316, P = 0.752, body size = Z = 0.228, P = 0.819). Female age was a significant predictor of multiple paternity over all years
(GLMM; age, Z = 2.013, P = 0.044), and in years with low numbers of reproductive females,
(GLMM; Z = 2.652, P = 0.008). Female age was not significant in years with high numbers
reproductive females (GLMM; Z = 0.305, P = 0.760). No other factors tested were significantly
associated with incidence of multiple paternity (GLMM; P > 0.10). The incidence of multiple
59

paternity was not different in the next clutch for females reproducing and those skipping
2
reproduction in a given year (X = 0.142, df = 1, P = 0.706).

Variation in Repeat paternity

The same male sired offspring with the same female between years (repeat paternity) in 38% of
nests, including inter-nest intervals of up to 4 years (Fig. 2.6). There was no significant
relationships between repeat paternity and the age or size of females (GLMM; age, Z = 0.043; P
= 0.966; size, Z = 0.792, P = 0.429) or males (GLMM; age, Z = 1.037; P = 0.300, size, Z =
1.032; P = 0.302). Repeat paternity was not predicted by multiple paternity (GLMM, P =
0.996).

Male reproductive success
The number of offspring sired varied among successful males each year (mean = 7.11,
min - max = 1 - 19 offspring, SD = 3.99; Table 2.1) and among the four years of study (mean =
8.83, min - max = 1 - 33 offspring, SD = 6.28). Male RS (number of offspring sired from 20032006) was not predicted by the number of mates (GLMM, t = 1.465; P = 0.144). The average
number of mates per year was 1.05 (SD = 0.22), and the average number of mates over the four
year period (2003-2006) was 1.09 (SD =0.30). No males sired offspring with more than one
female in any year except 2005 (mean = 1.15 mates).
The number of clutches sired was a significant predictor of male RS (GLMM, t = 10.114;
P < 0.001). The number of clutches sired averaged 1.4 per year (SD = 0.57) and 1.7 (SD = 0.87)
over the 4 years of study. Therefore increases in the number of offspring sired primarily resulted
from repeat paternity (use of stored sperm or remating with the same female). We also tested the
60

influence of the number of clutches on male RS resulting from only among-year variation in the
number of clutches by randomly removing either the first or second clutch of each pair. When
the total number of offspring and number of clutches sired were recalculated, the number of
clutches sired was still a significant predictor of male RS (GLMM, t = 11.708, P < 0.001).
Therefore, repeat paternity among years also has a substantial influence on male RS. Age and
2
size distributions of successful males did not differ from all sampled males (Fig. 2.7; age, X =
2
6.973, df = 8, P = 0.540; size, X = 19.226, df = 13, P = 0.116). Among successful males, the

total number of offspring sired was not associated with male body size or age (GLMM; size: t =
1.698; P = 0.095), GLMM; age: t = -0.936, P = 0.361), but was marginally negatively influenced
by multiple paternity (GLMM; t = -1.89; P = 0.06).

Evidence for Assortative mating
Frequency distributions of sampled mating pairs differed substantially from random
2
mating with respect to age (G = 20.17, df = 7, P = 0.005; X = 13.71, df = 7, P = 0.056) but not
2
body size (G = 1.39, df = 7, P = 0.986; X = 0.73, P = 0.994). The ages of females were

correlated with the ages of male mates (GLMM, t = 2.67, P = 0.01), whereas body sizes of
females and males was not (GLMM, t = 1.23, P = 0.20).

Discussion

61

Despite the almost stable population size of Painted turtles on the ESGR, the number of
gravid females captured varied annually 55-102 (mean = 91.5). From 1999-2006, the ages of the
majority of different reproductive females captured after 1999 were ages 6-12 years (the ages of
primiparous females) whereas the others were presumably those that had skipped reproduction.
During the last 3 years of the study (200-2006), we trapped East Marsh extensively and made 25
captures of 22 females that had not been captured on land during the nesting season (i.e., did not
reproduce that year); an average of 7% of females captured.
Approximately 26% of reproductive females were not observed to produce any second
clutches over intervals of 5-32 years whereas 74% produced clutches in 7-50% of all years.
Individuals of both groups skipped reproduction in some years. The minimal interval between
within season clutches is approximately 10 days and among year intervals exceed two years.
Over the 24 years that East Marsh was fenced (1983-2006), and during the period of genetics
study (1999-2006) the average number of females that reproduced each year were similar (means
= 91.5 and 86.6 females, respectively). The range of variation in numbers of reproductive
females and the sample of females that do not produce second clutches in any years was
surprising since females live in a resource rich wetland (Fiala and Congdon 1983) and females
use body lipid stores to support reproduction (Congdon and Tinkle 1982). However, there are
other short-lived (Deirochelys reticularia, Gibbons 1969) and long-lived (Chelydra serpentina,
Obbard 1983; Emydoidea blandingii, Congdon et al. 1993) adult female turtles that do not
reproduce every year. That some females do not reproduce each year, while some females can
produce two clutches in one year, represents a source of variation that can have direct benefits to
male RS.

62

Because clutch size and egg size of first and second clutches are similar, females that
produce second clutches allocate approximately two times the resources to reproduction in a year
than do females that produce one clutch. If resource limitation occurs in the relatively resource
rich environment of a wetland, we expected that it would be expressed as a longer inter-annual
reproductive interval (i.e., narrowing the difference in reproductive quality). However, the interannual intervals between reproductive events were similar in females that produce only one
clutch per year and those that produce second clutches (i.e., inter-annual reproductive intervals
do not influence differences in reproductive quality). In addition, among all females, older
females have shorter intervals between reproductive events and among just the females that
produce two clutches, older females produce paired clutches at higher frequencies (McGuire et
al. 2011).
When females vary substantially in reproductive quality, males should evolve to prefer
traits that are reliable indicators of the direct benefits (Gwynne 1981). Males demonstrate
preferences for reliable indictors of female quality in a variety of taxa (Amundsen 2000; CluttonBrock 2007; Zhang et al. 2010). Because body size and age of females are correlated in painted
turtles (Fig. 2.1b), body size is one reliable indicator that males may use to access female
reproductive quality (Congdon et al. 2003; McGuire et al. 2011); however, assessment of female
quality by male Painted turtles has not been documented.
Patterns of variation in multiple paternity
Over the four years of the genetic study, female age was positively associated with the
probability of a clutch being sired by more than one male (multiple paternity). The increase in

63

incidence of multiple paternity could result from an increase in the number of males attempting
to mate with high quality (larger and older) females.
If the majority of adult males in the population are reproductive each year, the adult sex
ratio will vary with the number of reproductive females. During 2003-2004 the numbers of
reproductive females was low (Fig. 2.1a), the occurrence of multiple paternity was high in
second clutch producing females and was positively associated with female age. When the
number of reproductive females was high (2005-2006), multiple paternity was lower and not
significantly associated with age than in 2003-2004. Both results suggest that when relatively
more adult males are available per female, more mating attempts per female occur.
The difference in association of the incidence of multiple paternity with age in years with
the high compared to low numbers of reproductive females may be due to a higher proportion of
young females (generally individuals with low-reproductive qualities) being added to the number
of reproductive females in a given year. A high number of young adults recruiting into the group
of reproductive females in years with low total number of females may increase the number of
males attempting to mate with high quality females.
Variation in repeat paternity
Reproductive frequency of females represents an important component of female quality
(Gibbons et al. 1982), whereas reproductive frequency of females and repeat paternity of a
female‟s clutches in combination are an important mechanism influencing male RS. Over 4
years sampled males mated with an average of a single female, therefore, increased male RS
occurs as a function of female reproductive frequency when males sire offspring in subsequent
64

clutches of females (i.e., repeat paternity). The majority of Painted turtle females apparently use
stored sperm from the same male to fertilize eggs in second clutches within a year (McGuire et
al. 2011) and in other populations (Pearse and Avise 2001; Pearse et al. 2001; Pearse et al. 2002).
Additionally, repeat paternity occurs in ~ 40% of clutches produced among years in ESGR
Painted turtles (Fig. 2.6) and also occurs in other species of turtles (Fitzsimmons 1998; Roques et
al. 2006). Collectively, the results indicate that repeat paternity (through the use of the same
male‟s stored sperm or by females or remating with the same male) contributes substantially to
male RS.
The high incidence of repeat paternity among years coupled with variation in female
reproductive frequencies and the importance of the number of clutches sired supports predictions
that mate quality is more important than mate number to variation in male RS in the ESGR
population (McGuire et al. 2011). Because both inter- and intra-annual reproductive frequency
increase with female age and size (Congdon et al. 2003; McGuire et al. 2011) body size is a
reliable indicator of female reproductive quality. However, the presence of age effects
associated with multiple paternity without significant body-size effects suggests that additional
cues in assessing female age must be present. If males can identify and successfully mate with
high quality females, they will have substantially higher RS than males that mate randomly and
the incidence of multiple paternity will increase among both older females and those that
frequently produce two clutches in a year regardless of age.
The frequency of repeat paternity among years and the relatively low estimate of multiple
paternity begs the question: why would some females sire offspring with the same males among
years if males are plentiful? Iteroparity provides an opportunity for females to change mates
65

between reproductive bouts, so we examined whether attributes of the males changed when
paternity differed among-years. The "trading up" hypothesis predicts that remating will occur if
a female subsequently encounters a higher-quality mate or mates (Uller and Olsson 2008). In
Painted turtles, changes in paternity among clutches produced in different years exhibited no
consistent pattern of increasing or decreasing body size or age with a "new" male compared to
the male who sired progeny in a previous clutch.

Male reproductive success
The total number of offspring sired by sampled males varied among males within year (119 offspring) and among males over the four years (1-32 offspring; Table 2.1). Neither body
size not age of the male was associated with variation in male reproductive success (Fig. 2.7).
However, we did find evidence of assortative mating by age. The number of clutches a female
produced coupled with repeat paternity was a substantial source of variation in male RS, whereas
the number of females a male mated with was not. If there are males that can identify and mate
with high quality females, they will on average leave more offspring compared to males that
mate indiscriminately (Altmann 1997; Shine et al. 2003; Zhang et al. 2010).
Although male RS can be influenced by the presence of multiple paternity the incidence
of occurrence in the present study averaged 14% over all years and was not a major determinant
of male RS. However, in other species of turtles and in reptiles in general, the incidence of
multiple paternity is much higher (Lee 2008; Uller and Olsson 2008) and therefore may be a
larger source of variation in male RS than it is in painted turtles.
66

Painted turtles are often described as having a polygamous mating system. However, we
found that over a four-year period, the vast majority of males successfully mated with a single
female. The number of mates for males was rarely greater than 1, whereas females clutches were
sired by have multiple males within and among years (mean number of mates 1.18 per year and
1.74 over 4 years). Our data suggest that at least in the ESGR painted turtle population, the
mating system is more accurately described as polyandrous.
Why don‟t the majority of males mate with more than one female within or among-years?
Several explanations are possible: 1) mating with multiple females may be costly particularly in
the ESGR population where the adult sex ratio is 2.3 males per female, 2) if male courtship
behaviors are prolonged or energetically expensive, or mate guarding is extensive, males may
not be able to obtain additional successful matings opportunities, and 3) males may attempt to
mate with several females, but female choice or post-copulatory sexual selection in the use of
stored sperm may reduce the number of successful matings per male (Eberhard 1996). Postcopulatory sexual selection is widespread among taxa and has been documented in insects, birds,
mammals, and reptiles (Birkhead and Pizzari 2002; Calsbeek and Bonneaud 2008). In the
absence of behavioral data on mating tactics, genetic methods that allow paternity to be inferred
precludes the ability to disentangle pre- and post-copulatory mechanisms leading to the outcome
of paternity. Thus, the patterns of repeat paternity, change of paternity, and male RS in painted
turtles could be due to pre- or post-copulatory mechanisms, and represents an area of research to
be further explored.
Our findings have implications for other iteroparous organisms, particularly those species
where females have the ability to store sperm and exhibit propensities for fertilizing subsequent
67

clutches with sperm of the same male. If the number of females is limited, then the first priority
should be securing any mate, but that tactic is not apparent in the ESGR population (i.e., almost
all successful males mate with only one female). Our results highlight an example where male
mate choice could evolve in the absence of male parental care, sex role reversal, or cooperative
breeding (Trivers 1972; Clutton-Brock and Vincent 1991; Clutton-brock 1991; Hauber and
Lacey 2005).

68

APPENDIX 2

69

Table 2.1. Summary of the number of offspring, clutches, and mates per year and over four years
(includes residual paternity) for successful male Painted turtles (Chrysemys picta).
Per Year
For Four Years
SD
Mean
Min-Max Mean
SD
Min-Max
3.99
Number of Offspring
7.18
1-19
8.64
5.81
1-32
0.57
Number of Clutches
1.41
1-3
1.68
0.87
1-5
0.23
Number of Mates*
1.05
1-2
1.09
0.30
1-2
*Only one year (2005) had an average number of mates that was greater than 1 (1.15
mates/male).

70

Fig. 2.1. a) The number of reproductive Painted turtle (Chrysemys picta marginata) females per
year (1983-2006) with yearly occurrence of multiple paternity (2003-2006) and b) relationship
between body size and age of females.

# of females

110

a.

100

14.3%

90

6.1%
12.5%
30.0%

80
70
60
50
1985

1990

1995

2000

2005

Year

Carapace length (mm)

180

b. CL = 125.86 + 0.69 (age)
r ²2= 0.33
R = 0.33

170
160
150
140
130
120
110
100
0

10

20

30

Age (years)

71

40

50

Egg width (mm)

Clutch size

Fig. 2.2. a) Relationships between a) clutch size, and b) egg widths and body size of females,
and c) frequency of second clutch production and age groups of females.

14
12
10
8
6
4
2
20

a. CS = -5.55 + 0.09 (CL)
rR2 = 0.26
² =0.26

b. EW = 7.27 + 0.08 (CL)
rR2 = 0.38
² = 0.38

18
16
14
12
110

120

130

140

150

160

% second clutches

Carapace length (mm)
0.5

c.

0.4
0.3
0.2
0.1
6-11

12-19

20-24

Age (years)

72

25-29

>29

Fig. 2.3. a) Distribution of carapace lengths and b) ages of reproductive females from East Marsh
captured in 1999 and reproductive females subsequently captured from 2000-2006 (14.6% and
51.8% of 1999 and 2000-2006 females respectively, were < 12 years old; the ages that females
reach reproductive maturity).

20

1999, N = 100
2000 - 2006
N = 116

a.

15
10

Number of females

5
0
115

120

125

130

135

140

145

150

Carapace length (mm)
10
8
6
4
2
0

b.

5
16
14
12
10
8
6
4
2
0

1999
N = 97

10

15

20

25

30

c.

5

35

40

2000 - 2006
N = 95

10

15

20

25

Age (years)

73

30

35

40

Number of females

Fig. 2.4. a-f) Histograms of body sizes of females sampled over different periods of the life
history and genetics studies.

16 a.
2003-2006 40 b.
1983-2006
14
12
30
10
8
20
6
4
10
2
0
0
120 130 140 150 160 120 130 140 150 160
16 c.
2003 16 d.
2004
14
14
12
12
10
10
8
8
6
6
4
4
2
2
0
0
120 130 140 150 160 120 130 140 150 160
16
16
2006
e.
2005 14 f.
14
12
12
10
10
8
8
6
6
4
4
2
2
0
0
120 130 140 150 160 120 130 140 150 160

Carapace length (mm)

74

Number of females

Fig. 2.5. a-f) Histograms of ages of females sampled over different periods of the life history and
genetics studies.

18
16 a.
14
12
10
8
6
4
2
0
10
16
14 c.
12
10
8
6
4
2
0
10
16
14
12
10
8
6
4
2
0

2003-2006 40 b. 1983-2006
30
20
10
0
20

30

16
14
12
10
8
6
4
2
0

2003

20

30

e.

10

40

20

30

40

10

10

40
2004

20

30

f.

Age (years)
75

30

d.

10

16
2005 14
12
10
8
6
4
2
0
40

20

40
2006

20

30

40

Fig. 2.6. Percent of nests with evidence of the same male siring offspring with the same female
(repeat paternity) within- and among-years.

120

% of nests

100

N = 40

80
60
N = 42

N = 13

N=5

1

2

3+

40
20
0

Within

Reproductive interval (years)

76

Fig. 2.7. Distribution by age and body size for all males sampled (a, b) and for males that mated
successfully in the population (c, d) 2003-2006.

80 b.

60

60

40

Number of males

80 a.

40

20

20

0

0
0

14

10 20 30 40 50

16
14
12
10
8
6
4
2
0

c.

12
10
8
6
4
2
0
0

80

10

20 30

40

50

Age (years)

100 120 140 160
d.

80

100 120 140 160

Carapace length (mm)
77

REFERENCES

78

References
Altmann, J. 1997. Mate choice and intrasexual reproductive competition: Contributions to
reproduction that go beyond acquiring more mates. In: Feminism and Evolutionary
Biology: boundaries, intersections, and frontiers. New York: Chapman Hall.
Amundsen, T. 2000. Why are female birds ornamented? Trends in Ecology & Evolution 15(4):
149-155.
Andersson, M. 1994. Sexual Selection. Princeton University Press, Princeton. .
Bateman, A.J. 1948. Intra-sexual selection in Drosophila. Heredity 2: 349-368.
Birkhead, T.R., and Pizzari, T. 2002. Postcopulatory sexual selection. Nature Reviews Genetics
3(4): 262-273.
Calsbeek, R., and Bonneaud, C. 2008. Postcopulatory fertilization bias as a form of cryptic
sexual selection. Evolution 62(5): 1137-1148. doi:10.1111/j.1558-5646.2008.00356.x.
Clutton-Brock. 2007. Sexual selection in males and females. Science 318: 1882-1885.
Clutton-Brock, and Vincent. 1991. Sexual selection and the potential reproductive rates of males
and females. Nature 351: 58-60.
Clutton-brock, T.H. 1991. The Evolution of Parental Care. Princeton University Press, Princeton,
NJ.
Congdon, J.D. 1989. Proximate and evolutionary constraints on energy relations of reptiles.
Physiological Zoology 62: 356-373.
Congdon, J.D., Dunham, A.E., and Sels, R.C.V.L. 1993. Delayed Sexual Maturity and
Demographics of Blanding's Turtles (Emydoidea blandingii): Implications for
Conservation and Management of Long-Lived Organisms. Conservation Biology 7(4):
826-833.
Congdon, J.D., and Gibbons, J.W. 1990. The evolution of turtle life histories. p. 45-54. In: Life
History and Ecology of the Slider Turtle. J.W. Gibbons (ed.), Smithsonian Institution
Press, Washington, D.C.
Congdon, J.D., and Gibbons, W. 1987. Morphological Constraint on egg size: a challenge to
optimal egg size theory? Proc. Natl. Acad. Sci. USA 84: 4145-4147.
Congdon, J.D., Nagle, R.D., Kinney, O.M., van Loben Sels, R.C., Quinter, T., and Tinkle, D.W.
2003. Testing hypotheses of aging in long-lived painted turtles (Chrysemys picta).
Experimental Gerontology 38(7): 765-772.
79

Congdon, J.D., and Tinkle, D.W. 1982. Reproductive Energetics of the Painted Turtle
(Chrysemys picta). Herpetologica 38(1): 228-237.
Coulson, T., Catchpole, E.A., Albon, S.D., Morgan, B.J.T., Pemberton, J.M., Clutton-Brock,
T.H., Crawley, M.J., and Grenfell, B.T. 2001. Age, Sex, Density, Winter Weather, and
Population
Crashes
in
Soay
Sheep.
Science
292(5521):
1528-1531.
doi:10.1126/science.292.5521.1528.
Crowley, p., Travers, S.E., Linton, M.C., Cohn, S.L., Sih, A., and Sargent, R.C. 1991. Mate
Density, Predation Risk, and the Seasonal Sequence of Mate Choices: A Dynamic Game.
The
American
Naturalist
137(4):
567-596.
doi:Stable
URL:
http://www.jstor.org/stable/2462382.
Darwin, C. 1871. The descent of man, and selection in relation to sex. Princeton University
Press, Princeton, NJ.
Dreiss, A.N., Cote, J., Richard, M., Federici, P., and Clobert, J. 2010. Age- and sex-specific
response to population density and sex ratio. Behav. Ecol. 21(2): 356-364.
doi:10.1093/beheco/arp198.
Eberhard, W. 1996. Princeton, Princeton University Press.
Emlen, S., and Oring, L. 1977. Ecology, sexual selection, and the evolution of mating systems.
Science 197(4300): 215-223. doi:10.1126/science.327542.
Fedorka, K.M., and Mousseau, T.A. 2002. Material and genetic benefits of female multiple
mating and polyandry. Animal Behaviour 64(3): 361-367. doi:10.1006/anbe.2002.3052.
Fiala, K.L., and Congdon, J.D. 1983. Energetic Consequences of Sexual Size Dimorphism in
Nestling Red-Winged Blackbirds. Ecology 64(4): 642-647.
Fitzsimmons, N.N. 1998. Single paternity of clutches and sperm storage in the promiscuous
green turtle (Chelonia mydas). Molecular Ecology 7(5): 575-584.
Fiumera, A.C., Porter, B.A., Grossman, G.D., and Avise, J.C. 2002. Intensive genetic assessment
of the mating system and reproductive success in a semi-closed population of the mottled
sculpin, Cottus bairdi. Molecular Ecology 11(11): 2367-2377.
Fromhage, L., McNamara, J.M., and Houston, A.I. 2008. A model for the evolutionary
maintenance of monogyny in spiders. Journal of Theoretical Biology 250(3): 524-531.
doi:10.1016/j.jtbi.2007.10.008.
Gibbons, J.W. 1969. Ecology and Population Dynamics of the Chicken Turtle, Deirochelys
reticularia. Copeia 1969(4): 669-676.
80

Gibbons, J.W. 1976. Aging phenomena in reptiles pp 454-475. EAR, Inc., Maine.
Gibbons, J.W., and Greene, J.L. 1979. X-ray photography: a technique to determine reproductive
patterns of freshwater turtles. Herpetologica 35: 86–89.
Gibbons, J. W., J. L. Greene, and K. K. Patterson 1982. "Variation in reproductive characteristics
of aquatic turtles." Copeia 1982: 776-784.
Gwynne, D.T. 1981. Sexual Difference Theory: Mormon Crickets Show Role Reversal in Mate
Choice. Science 213(4509): 779-780. doi:10.1126/science.213.4509.779.
Hauber, M.E., and Lacey, E.A. 2005. Bateman's Principle in Cooperatively Breeding
Vertebrates: The Effects of Non-breeding Alloparents on Variability in Female and Male
Reproductive Success. Integr. Comp. Biol. 45(5): 903-914. doi:10.1093/icb/45.5.903.
Hinton, T.G., Fledderman, P.D., Lovich, J.E., Congdon, J.D., and Gibbons, J.W. 1997.
Radiographic determination of fecundity: is the technique safe for developing turtle
embryos? Chelonian Conservation Biology 2: 409–414.
Jones, A., and Ardren, W. 2003. Methods of parentage analysis in natural populations. Molecular
Ecology 12: 2511-2523.
Jones, A.G. 2005. Gerud 2.0: a computer program for the reconstruction of parental genotypes
from half-sib progeny arrays with known or unknown parents. Molecular Ecology Notes
5(3): 708-711.
Jones, B., Grossman, G.D., Walsh, D.C.I., Porter, B.A., Avise, J.C., and Fiumera, A.C. 2007.
Estimating Differential Reproductive Success From Nests of Related Individuals, With
Application to a Study of the Mottled Sculpin, Cottus bairdi. Genetics 176(4): 24272439. doi:10.1534/genetics.106.067066.
Kalinowski, S., Taper, M., and Marshall, T. 2007. Revising how the computer program
CERVUS accommodates genotyping error increases success in paternity assignment.
Molecular Ecology 16: 1099-1006.
Klemme, I., Ylönen, H., and Eccard, J. 2007. Reproductive success of male bank voles
(Clethrionomys glareolus): the effect of operational sex ratio and body size. Behavioral
Ecology and Sociobiology 61(12): 1911-1918. doi:10.1007/s00265-007-0431-1.
Kokko, H., and Jennions, M.D. 2008. Parental investment, sexual selection and sex ratios.
Journal of Evolutionary Biology 21(4): 919-948. doi:10.1111/j.1420-9101.2008.01540.x.
Kokko, H., and Mappes, J. 2005. Sexual selection when fertilization is not guaranteed. Evolution
59(9): 1876-1885. doi:10.1111/j.0014-3820.2005.tb01058.x.
81

Lee, P.L.M. 2008. Molecular ecology of marine turtles: New approaches and future directions.
Journal of Experimental Marine Biology and Ecology 356(1-2): 25-42. doi:DOI:
10.1016/j.jembe.2007.12.021.
Libants, S., Kamarainen, A.M., Scribner, K.T., and Congdon, J.D. 2004. Isolation and crossspecies amplification of seven microsatellite loci from Emydoidea blandingii. Molecular
Ecology Notes 4(2): 300-302.
McGuire, J.M., Congdon, J.D., Scribner, K., and Capps, J. 2011. Variation in Female
Reproductive Quality and Reproductive Success of Male Painted Turtles (Chrysemys
picta marginata). Canadian Journal of Zoology 89 (11) 1136-1145.
McLoughlin, P.D., Boyce, M.S., Coulson, T., and Clutton-Brock, T. 2006. Lifetime reproductive
success and density-dependent, multi-variable resource selection. Proceedings of the
Royal
Society
B:
Biological
Sciences
273(1593):
1449-1454.
doi:10.1098/rspb.2006.3486.
Monnin, T., and Peeters, C. 1999. Dominance hierarchy and reproductive conflicts among
subordinates in a monogynous queenless ant. Behavioral Ecology 10(3): 323-332.
doi:10.1093/beheco/10.3.323.
Obbard, M.E. 1983. Population ecology of the common snapping turtle, Chelydra serpentina, in
North-central Ontario. University of Guelph, Guelph, Ontario.
Oliver, C., and Cordero, C. 2009. Multiple mating reduces male survivorship but not ejaculate
size in the polygamous insect Stenomacra marginella; (Heteroptera: Largidae).
Evolutionary Ecology 23(3): 417-424. doi:10.1007/s10682-007-9239-8.
Olsson, M., Madsen, T., and Shine, R. 1997. Is Sperm Really so Cheap? Costs of Reproduction
in Male Adders, Vipera berus. Proceedings of the Royal Society B: Biological Sciences
264(1380): 455-459.
Osentoski, M.F., Mockford, S.M., Wright, J.M., Snyder, M., Herman, T.B., and Hughes, C.R.
2002. Isolation and Characterization of microsatellite loci from Blanding's turtle,
Emydoidea blandingii. Molecular Ecology Notes 2(2): 147-149.
Owens, I.P.F., and Thompson, D.B.A. 1994. Sex Differences, Sex Ratios and Sex Roles.
Proceedings of the Royal Society B: Biological Sciences 258(1352): 93-99.
Pearse, D., Janzen, F., and Avise, J. 2002. Multiple paternity, sperm storage, and reproductive
success of female and male painted turtles (Chrysemys picta) in nature. Behavioral
Ecology and Sociobiology 51(2): 164-171.
Pearse, D.E., and Avise, J.C. 2001. Turtle Mating Systems: Behavior, Sperm Storage, and
Genetic Paternity. J Hered 92(2): 206-211. doi:10.1093/jhered/92.2.206.
82

Pearse, D.E., Janzen, F.J., and Avise, J.C. 2001. Genetic markers substantiate long-term storage
and utilization of sperm by female painted turtles. Heredity 86(3): 378-384.
Reynolds, J., and Gross, M.R. 1990. Costs and Benefits of Female Mate Choice: Is There a Lek
Paradox? The American Naturalist 136(2): 230-243.
Roques, S., Díaz-Paniagua, C., Portheault, A., Pérez-Santigosa, N., and Hidalgo-Vila, J. 2006.
Sperm storage and low incidence of multiple paternity in the European pond turtle, Emys
orbicularis: A secure but costly strategy? Biological Conservation 129(2): 236-243.
Rowe, S., Hutchings, J.A., and Skajaeraasen, J.E. 2007. Nonrandom mating in a broadcast
spawner: mate size influences reproductive success in Atlantic cod (Gadus morhua).
Canadian Journal of Fisheries and Aquatic Sciences 64(2): 219-226.
Rowell, J.T., and Servedio, M.R. 2009. Gentlemen prefer blondes: the evolution of mate
preference among strategically allocated males. The American Naturalist 173: 12-25.
Salehialavi, Y., Fritzsche, K., and Arnqvist, G. 2011. The cost of mating and mutual mate choice
in 2 role–reversed honey locust beetles. Behavioral Ecology 22(5): 1104-1113.
doi:10.1093/beheco/arr097.
Schielzeth, H., and Forstmeier, W. 2009. Conclusions beyond support: overconfident estimates
in mixed models. Behavioral Ecology 20(2): 416-420. doi:10.1093/beheco/arn145.
Schubert, M., Pillay, N., Ribble, D.O., and Schradin, C. 2009a. The Round-Eared Sengi and the
Evolution of Social Monogamy: Factors that Constrain Males to Live with a Single
Female. Ethology 115(10): 972-985. doi:10.1111/j.1439-0310.2009.01684.x.
Schubert, M., Schradin, C., Rödel, H., Pillay, N., and Ribble, D. 2009b. Male mate guarding in a
socially monogamous mammal, the round-eared sengi: on costs and trade-offs.
Behavioral Ecology and Sociobiology 64(2): 257-264. doi:10.1007/s00265-009-0842-2.
Sefc, K.M., and Koblmuller, S. 2008. Assessing Parent Numbers from Offspring Genotypes: The
Importance of Marker Polymorphism. J Hered: esn095. doi:10.1093/jhered/esn095.
Servedio, M.R., and Lande, R. 2006. Population genetic models of male and mutual mate choice
Evolution 60 674-685.
Shine, R., Phillips, B., Waye H., LeMaster, M., and Mason, R.T. 2003. The lexicon of love:
What cues cause size-assortative courtship by male garter snakes? Behavioral Ecology
and Sociobiology 53: 234-237.
Shuster, and Wade. 2003. Mating Systems and Strategies. Princeton University Press Princeton,
New Jersey.
83

Székely, T., Webb, J.N., Houston, A., and McNamara, J.M. 1996. An evolutionary approach to
offspring desertion in birds. Current Ornithology 13: 271-331.
Tinkle, D.W. 1979. Long-term field studies. Bioscience 29: 717.
Trivers, R.L. 1972. Parental investment and sexual selection pp. 136-179. Aldine Transaction,
Chicago.
Uller, t., and Olsson, M. 2008. Multiple paternity in reptiles: patterns and processes. Molecular
Ecology 17(11): 2566-2580.
Wang, J. 2004. Sibship reconstruction from genetic data with typing errors Genetics 166: 19631979.
Wang, J., and Santure, A.W. 2009. Parentage and Sibship Inference From Multilocus Genotype
Data Under Polygamy. Genetics 181(4): 1579-1594. doi:10.1534/genetics.108.100214.
Wilberg, M.J., and Dreher, B.P. 2004. genecap: a program for analysis of multilocus genotype
data for non-invasive sampling and capture-recapture population estimation. Molecular
Ecology Notes 4(4): 783-785.
Zhang, M., Li, J., Zhu, Y., Wang, X., and Wang, S. 2010. Male mate choice in Tibetan macaques
Macaca thibetana at Mt. Huangshan, China. Current Zoology 56(2): 213 - 221.

84

CHAPTER 3: MATING SYSTEM, MALE REPRODUCTIVE SUCCESS, AND
POPULATION CONNECTIVITY OF BLANDING'S TURTLES (EMYDOIDEA
BLANDINGII)

85

Abstract
Identification of factors that facilitate or impede gene flow among populations requires an
integrative approach, where multiple sources of information are simultaneously evaluated.
Long-term data on population demography, reproductive ecology, and core areas were combined
with genetic data to identifying mating pairs, and male reproductive success. We then examined
spatial and genetic structuring among threatened Blanding‟s turtles (Emydoidea blandingii)
inhabiting wetlands on and surrounding the Edwin S. George Reserve (ESGR) in southeastern
Michigan. Analyses were based on data from nests of 49 females and inferred male paternity of
77 nests collected over 7 years (1999-2006). Nests located in terrestrial areas were usually
nearer to residence than to non-residence wetlands; although nests were most often located in
closer proximity to temporary wetlands. For resident ESGR females, 34.1% of clutches were
sired by non-resident ESGR males, whereas 56% of clutches of non-resident females that nested
on the ESGR were sired by ESGR resident males. Males and females had multiple mates (mean
1.56 and 2.02, respectively). Over four years, the occurrence of multiple paternity was 47.6%
(min-max = 15.4%-55.6%). However, repeat paternity (same male siring offspring with the
same female among years) was common (69.6% of inter-annual clutch pairs from the same
female). Propensities of adults to mate with individuals from different wetlands and tendencies
for hatchlings to disperse to wetlands other than the mother‟s residence wetland were two
important but previously undocumented sources of gene flow among wetlands. Lack of
significant differences in allele frequencies among individuals from different residence wetlands
(Fst=0.002, P > 0.05) are concordant with observations of adult and juvenile movements and
genetic data identifying mating pairs. Landscape-level conservation efforts should include large
86

areas of terrestrial habitat around residence wetlands to facilitate adult and juvenile movements
and mating encounters that allow gene flow within and among populations..

Introduction
Quantification of the relative importance of factors that promote or limit inter-population
gene flow can be improved by integrating data from multiple sources (Steen et al., in press).
Direct observations of individual movements and habitat used to complete life functions (i.e.,
core habitat; Semlitsch and Bodie 2003) can be used to infer rates and direction of dispersal.
However, in the absence of data on spatial variation in gene frequency or genetically determined
parentage that identifies mating pairs, factors that affect gene flow cannot be directly quantified
(Lowe and Allendorf 2010). Genetic data that allow identification of mating pairs and male
reproductive success, when supported by information on movements of adults that affect mating
encounters and of juveniles dispersing from natal areas can provide more definitive
characterization of how aspects of species‟ reproductive ecology and habitat use throughout the
year facilitate or impede gene flow. Such data are often lacking or are available only for limited
periods of time. Thus researchers often lack the ability to generalize findings to other conditions
(e.g., among years), locations, or species.
Rates of immigration and emigration among populations are influenced by constraints
imposed by morphology, dispersal capabilities, risks, reproductive ecology, and the distribution
of available of resources such as food and mates (Clobert et al. 2001). Such data are particularly
important for semi-aquatic organisms such as turtles where core areas include permanent bodies
of water, ephemeral wetlands, and terrestrial habitats surrounding wetlands (Semlitsch 1998;
87

Calhoun et al. 2003; Roe et al. 2009; Congdon et al. 2011). Use of terrestrial areas and
permanent and temporary wetlands by turtles can increase encounters and matings between
individuals from different permanent wetlands. Therefore, use of landscapes between resident
wetlands can facilitate gene flow (Calhoun et al. 2003; Congdon et al. 2011). For example, some
species of turtles move among ephemeral wetlands and seasonally flooded areas to exploit
resources and mate (Bodie and Semlitch 2000; Kinney 1999), and some females make prenesting movements up to 20 km along wetland corridors and overland (Obbard and Brooks 1980;
Power 1989). Additionally, males and females move across upland areas to find mates
(Morreale et al. 1984; Gibbons et al. 1990), visit ephemeral wetlands to exploit seasonally
abundant resources (Bodie and Semlitsch 2000), and to locate overwintering sites (Buhlmann et
al. 2009). Movements also occur when aquatic habitats become uninhabitable due to droughts
(Cagle 1944; Gibbons et al. 1983; Morreale et al. 1984; Kennett and Georges 1990), which can
result in episodic but high rates of gene flow (Scribner et al. 1995). Therefore core areas of
turtles and other semi-aquatic vertebrates often extend well beyond a permanent wetland
(Semlitsch and Bodie 2003; Congdon et al.2011).
Because observations of matings in turtles are rare, and do not necessarily indicate the
copulating male sired offspring, genetic data are required to document the occurrence of multiple
matings and successful matings between individuals from different wetlands. The frequency of
such matings provide more informed predictions of levels of gene flow and population levels of
genetic diversity. Genetic diversity can be influenced by the number and proportion of
individuals that have more than one mate (Nunney 1993; Karl 2008) and variation in
reproductive success (Pusey et al. 1997; Miller et al. 2009; Frankham 1995).

88

Extensive demographic, ecological, and life history data are rarely available for longlived poikilothermic vertebrates (Avise 2001; Owens 2006), particularly those that are
considered threatened or endangered throughout most of their range (Herman et al. 2002;
Mockford et al. 2005; Mockford et al. 2007; Congdon et al. 2008). Blanding's turtle (Emydoidea
blandingii) is a notable exception because it is one of the most studied species in North America.
In many parts of their range, Blanding‟s turtles exist in small and isolated populations that
overlap with human development (Grgurovic and Sievert 2005). Individuals make extensive
movements (Ross and Anderson 1990; Kinney 1999; Hamernick 2001; Calhoun et al. 2003) and
home ranges and core areas are large (Hamernick 2001; Congdon et al. 2011). Females produce
a maximum of one clutch per year, nest survivorship is low and variable (Congdon et al. 1983;
Congdon et al. 2000), and hatchlings emerging from nests orient and disperse toward far dark
horizons (Butler and Graham 1995; Pappas et al. 2000). Females do not reach sexual maturity
until 14 to 20 years of age, and adult survivorship in some populations is > 0.97 (Congdon et al.
2001; Congdon et al. 2003). However, data on spatial genetic structuring and gene flow can
have substantial influence on the way the extensive life history and ecology data are interpreted.
The life history and ecology of Blanding's turtle has been studied on the University of
Michigan‟s Edwin S. George Reserve (ESGR) near Hell, MI for 43 of the past 57 years (Owen
Sexton, 1953-1957; Henry Wilbur, 1968-1972, Donald Tinkle 1975-1979, and Justin Congdon
(1975-2007). During the last 33 years of continuous study, 2,081 individuals (excluding
hatchlings) were marked, 5,800 recaptures were made, 1,050 x-radiographs of gravid females
were taken, and the locations of 614 nests and fate of 433 nests of known females were recorded.
Background data on nesting ecology and female reproductive success (Congdon et al. 1983,
Congdon et al. 1983, 2000), movements and use of core areas (Kinney 1999; Congdon et
89

al.2011), demography and life history (Congdon and van Loben Sels 1991; Congdon et al. 1993;
Congdon and van Loben Sels 1993; Congdon et al. 2001) are well documented.
The general goal of our research was to identify factors that contributed to gene flow
among Blanding‟s turtles that reside for decades in different wetlands (Congdon et al. 2011).
We used long-term data to generate questions and predictions about how life-history traits and
core-habitat use influence mating tactics and to test the predictions using genetic data. We
quantified components of variation in male reproductive success in relation to: 1) number of
mates, 2) mate quality (clutch size, clutch frequency, egg size), 3) occurrence of multiple
paternity 4) the number of eggs a male sired per clutch, 5) repeat paternity (via stored sperm or
remating), and 6) age and body size of both sexes of adults. We also quantified the proportion
successful matings among individuals from different residence wetlands on and off of the ESGR.

Traits of Blanding’s turtles and development of research questions
Female Blanding‟s turtles produce an average clutch of 9.9 eggs (min-max = 2-19, SE =
0.70), and inter-clutch intervals are greater than one year (i.e., some females do not reproduce
each year; Congdon et al. 2001). Reproductive characteristics of female Blanding‟s turtles (i.e.,
clutch size, clutch frequency and egg size) vary as a function of body size and/or age (Congdon
et al. 1983; 2000). Based on all captures on the ESGR, adult sex ratios among resident
individuals is approximately 1:1. Females can store sperm for several years, and clutches of
eggs are sometimes fertilized by more than one male (Refsnider 2009), traits that are similar to
those of other emydid turtles (Painted turtles (Chrysemys picta marginata): Pearse and Avise
2001; Pearse et al. 2001; Pearse et al. 2002; McGuire et al. 2011, European pond turtles (Emys
obicularis; Roques et al. 2006)
90

Because body sizes and ages of the majority of individuals on the ESGR are known, we
can assess whether body size, or some phenotypic traits that vary as a function of age, are
associated with differences in mating tactics and male reproductive success (Fig. 3.1, Question
1), and whether female reproductive qualities (clutch size and frequency) affect male
reproductive success (Fig. 3.1, Question 2).
We predict that repeat paternity reflecting either female use of stored sperm or repeated
matings will be frequent and thus will be an important component of variation in male
reproductive success (Fig. 3.1, Questions 3 and 4), and may have a substantial influence on
levels of gene flow among different wetlands. Variation in male reproductive success, multiple
paternity, and female use of stored sperm can influence cohort and population levels of
coancestry and inter-population variance in allele frequency (Scribner et al. 1993). Based on the
ability of Blanding‟s turtles to store sperm and results from other studies of turtles demonstrating
repeat paternity (Painted turtles: McGuire et al. 2011, Pearse et a al. 2001, Pearse et al 2002,
Pearse and Avise 2001, European pond turtles: Roques et al 2006, Green turtle (Chelonia
mydas), Fitzsimmons 1998),
Philopatry to residence wetlands and nests located relatively close to residence wetlands
increases the probability of offspring recruiting to the mother‟s residence wetland, and thereby
increases the probability of inbreeding (Fig. 3.1, Question 5). Adult Blanding's turtles of both
sexes have occupied one of the two permanent wetlands on the ESGR for 40 or more years and
very few adults changed residence (Congdon et al. 2011). Although copulations occur in
residence and ephemeral wetlands (Kinney 1999), the proportion of successful matings that
occur in each type of wetlands is unknown, as a result, genetic structuring between residence
wetlands may or may not be present (Fig. 3.1, Question 6).
91

If males and females consistently use the same ephemeral wetlands, repeated encounters
will increase the probability of repeat matings and repeat paternity between the same pair of
individuals in different years (Fig. 3.1, Question 3). Further, use of the same ephemeral wetlands
by resident and non-resident adults is a potential mechanism for gene flow through matings (Fig.
3.1, Question 7). Adult male and female Blanding‟s turtles make extended terrestrial excursions
to visit ephemeral wetlands where they can utilize seasonally abundant resources. Terrestrial
movements also increase the probability of encountering adults from different residence
wetlands (Baker and Gillingham 1983; Ross and Anderson 1990; Rowe and Moll 1991; Kinney
1999; Beaudry et al. 2009).
Nesting area selection and hatchling dispersal can contribute significantly to genetic
exchanges among wetlands, reducing coancestry and probabilities of inbreeding (Fig. 3.1,
Question 8). Some adult females on the ESGR make extensive terrestrial movements from
residence wetlands to nesting areas (up to 2 km, Congdon et al., 2011) that result in some
females leaving the ESGR to nest. As the distance between resident wetlands and nesting areas
increases, the probability of hatchlings dispersing into the mother‟s residence should decrease.

Methods
Field sampling of adults and determination of residence wetland.
Blanding's turtles on the ESGR were captured in aquatic traps, on land, and at drift fences
during 44 of the 54 years from (1953-2007). At first capture, all juveniles and adults were given
a unique identification mark by notching a sequence of marginal carapace scutes. At every
capture, the date, individual identification, body size (carapace length, and weight), and
92

reproductive condition (e.g., gravid or not gravid; females only) were recorded. To quantify
traits associated with female quality, females with eggs were X-radiographed to determine clutch
size and egg width (Gibbons and Greene 1979; Hinton et al. 1997). Hatchlings were given nest
specific cohort marks and both hatchlings and yearlings were assigned “age 1”. At first capture,
juvenile ages were estimated from counts of growth rings, assuming one growth ring for each
year of life (Gibbons 1976).
Age at maturity for females was determined from recapture histories and presence of
eggs documented with X-radiographs. Ages at maturity for males were not known, but a
minimum age at maturity was provisionally set at 21 years based on the asymptote of the
relationship between age and body size in the ESGR population (Congdon and van Loben Sels
1993). Adults were assigned as ESGR residents based on extensive aquatic captures over more
than four years in one of two permanent wetlands (East and West; 1.9 km apart) on the ESGR.
Hereafter, „residence‟ or „residence wetland‟ will be used interchangeably. Females were
assigned as non-residents based on extensive captures histories on land in ESGR nesting areas,
but with minimal or no aquatic captures in ESGR wetlands over 33 years (Congdon et al.,2011).

Hatchling sample collection
Hatchlings of 49 marked females were obtained from 77 nests (611 hatchlings) from
1999-2006. Nests were protected in situ using wire cages to exclude predators except during
2005 and 2006 when eggs were transferred to a protected natural nesting area within four hours
of nest completion and incubated in perforated plastic containers to exclude burrowing mammal
predation. Samples from all years were combined for analyses.
93

Collection of genetic data
From 1997-2007, approximately 1cc of blood was taken from the caudal vein of the tail
from all adults captured on the ESGR (males, N = 85 that were all residents of ESGR marshes;
females, N = 163 including 32 ESGR residents and 131 non-residents) to use for genetic
analysis. Based on annual catchability estimates of 85% (Congdon et al. 1983), nearly all
resident males and females were captured every three years, and the probability that individuals
were not captured over the duration of the genetic sampling (11 years) was very low. Blood
samples from adults were taken during 1997-2006 and a small tissue sample (< 2 mm) was taken
from the tail of hatchlings from protected nests from 1999-2006 for parentage analysis. All
blood samples were stored at room temperature in a blood storage buffer (100 mM TrisHCL (pH
8.0), 100 mM EDTA, 10 mM NaCl and 0.5% SDS), and tissue samples were preserved and
initially stored at room temperature in 95% EtOH and then refrigerated at (4°C). The DNA was
extracted from blood and tissue samples using a Qiagen DNeasy kit (Qiagen Inc., Valencia, CA)
and quantified using fluorimetry.
All individuals were genotyped at 8 microsatellite loci including BTGA5, BTCA11,
BTCA9, (Libants et al. 2004), EB11, EB19, EB17, EB09 (Osentoski et al. 2002), and GmuD70
(King and Julian 2004). Polymerase chain reaction (PCR) was performed to amplify the DNA
using published conditions and separated using gel electrophoresis on a 6% polyacrylamide gel.
Products were visualized using an FMBIOII scanner (Hitachi Inc., Kanagawa, Japan). All gels
were independently scored by two experienced laboratory personnel. To estimate error, and
ensure consistency of allele assignment among samples, 10% of adults were randomly selected
94

and re-genotyped. Ten percent of PCR products from hatchlings were also reanalyzed
electrophoretically and any discrepancies resulted in re-amplification and and validation.
Statistical analysis
Genetic data were used for parentage assignment and evidence of spatial genetic
structuring, which are both contingent on levels of genetic diversity and based on the assumption
of Hardy-Weinberg equilibrium. All loci were tested for deviations from Hardy-Weinberg
equilibrium, independence, presence of null alleles, measures of genetic diversity (average
number of alleles and expected heterozygosity). Summary statistics and the multi-locus paternal
exclusion probability were estimated using program CERVUS 3.0 (Kalinowski et al. 2007).
Paternity assignment
The use of multiple parentage programs helps to ensure a robust parentage assignment
(Jones and Ardren 2003; Jones et al. 2007; Lee 2008; Wang and Santure 2009). Aspects of a
species‟ biology, specific study objectives, and availability of other background information
influences the most appropriate programs to use (Jones and Ardren 2003; Lee 2008). In this
study, we observed females nesting so female parents were known for all offspring. Turtle
offspring are produced as a group (clutch) and therefore the use of statistical methods
implemented in programs that utilize clutch-level information (e.g. NEST; Jones et al. 2007) can
enhance the accuracy of paternity assignment and reduce over-estimation of the number of
contributing sires (Fiumera et al. 2002; Jones et al. 2007). In contrast to methods implemented
in programs such as CERVUS (Kalinowski et al. 2007) that use a likelihood approach and
considers each offspring as a random sample from the population, program NEST (Jones et al.
95

2007) uses a Bayesian framework that incorporates within-nest relatedness and prior information
of paternity for individuals within the clutch to facilitate paternal assignment. We performed a
two-tiered approach to assign a male to each offspring. We first used the program NEST to
assign male(s) to a clutch, and subsequently assigned each offspring to a male using program
CERVUS 3.0.1 (Kalinowski et al. 2007). Only in cases where the male(s) assigned by NEST
also had a positive log odds (LOD) score estimated in CERVUS was a male assigned as a sire.
When a sire was selected by one program but not confirmed with the second program, the sire
was assigned as "undetermined". In cases where neither NEST nor CERVUS was able to assign
a male(s), the nest was considered to be from an un-sampled (non-resident) male. Two or more
males were determined to have sired offspring in a clutch (multiple paternity) based on
assignment of two or more males or when a minimum of three paternal alleles were detected in
the clutch. Repeat paternity (assignment of the same male to offspring in two or more clutches
of a female in different years) was identified based on the same criteria. In cases where a sire
could not be identified, the offspring in all sampled nests of the female parent were combined
and re-analyzed using program NEST to assign offspring of a given female to one or multiple
un-sampled males. If NEST assigned the offspring to the same un-sampled male(s), the clutch
was also considered to be consistent with repeat paternity (sired by the same male).
Sources of variation in male reproductive success
We quantified sources of variation in male reproductive success using general linear
mixed models (GLMM) using the lme4 and languageR packages in the statistical software R
(Team 2007). Our dependent variable of male reproductive success was defined as the total
number of offspring sired over the course of the study. Fixed effects associated with
96

independent variables were quantified including male age and body size (for the years in which
offspring were sired), total number of mates, total number of clutches sired, and whether the
clutch was sired by multiple males (multiple paternity). Each fixed effect was tested
independently in a mixed model that included female as a random effect to account for the fact
that offspring from multiple nests were often available for the same female. Because some males
mated with more than one female, and were therefore represented multiple times in the dataset,
we also tested for factors associated with male RS using average values for male age, male body
size, number of mates (per year and all years), number of clutches (per year and all years), and
average clutch size.
Quantification of Gene Flow
We used parentage data to identify mating pairs and to quantify the occurrence of
pairings of individuals from different resident wetlands (East and West). For each female, we
compared female residence with that of her mate(s). For example, there were three categories of
matings possible for an east residence female. An east residence female could produce offspring
with an east-residence, a west-residence or non-resident male. More than one category may be
present for a female in cases where she mated with more than one male (multiple paternity). The
proportions of each category of matings were summarized by the residency of the female (e.g.,
the three potential categories for an east residence female should total 100%). Differences in the
proportions of pairing were compared using adjusted G-tests.
Females were observed in the act of nesting during extensive searches of nesting areas.
Nest locations were recorded by latitude and longitude for ESGR resident (east and west) and
non-resident females using the program PASSaGE v2 (Rosenberg and Anderson 2011) and
97

plotted using Google Earth. We used long-term capture-recapture data to identify recaptured
hatchlings of known mothers and to examine whether the probability of hatchling recruitment
into the mother's residence was associated with female nest location.

Evidence of Spatial Genetic Structure
We performed multiple analyses to characterize degree of spatial genetic structuring.
First, based on a priori assignment of individuals to a residence category (east, west or nonresident), we used program FSTAT v4.0.11 (Goudet 1995) to quantify variance in allele
frequency (Fst; Weir and Cockerham 1984) among individuals of each residence. Nominal alpha
levels for the hypothesis Fst=0 were adjusted using a Bonferroni correction. Secondly, we
estimated the number of genetic clusters using the program STRUCTURE v2.3.3 (Pritchard et al.
2000) based on all individuals (resident and non-resident males and females) sampled on the
ESGR with no a priori information regarding residence. We performed 10 independent runs for
6
5
each value of K = 1-4 using 3*10 MCMC iterations after a 2*10 step burn-in period. The

most likely number of clusters was determined by identifying values of K that produced the
highest log-likelihood value. If evidence for >1 genetic clusters was found, individuals were
assigned to a cluster based on the highest posterior probability of cluster assignment.

Results

98

Variation in Female Reproduction
From 1976-2006, the total number of ESGR east, west resident and non-resident females
that were captured on the ESGR were 34, 46, and 170 individuals, respectively. Although the
proportion of the total nesting females over all years that were non-residents was larger than
ESGR residents (G-test, G = 66.27, P < 0.05), annual captures of residents (mean = 12, min-max
= 2-21), west residents (mean = 11, min-max = 5-19), and non-residents (mean = 11, min-max =
5-22; Fig. 3.2a) were similar. The numbers of resident and non-resident reproductive females
captured each year were positively correlated (Fig. 3.2a; Pearson Rank Correlation; ρ = 0.49; P =
0.005).
We quantified two aspects of female quality (clutch size and egg size). Clutch size and
mean egg size (width) of clutches estimated from x-radiographs of all reproductive females
averaged 9.9 eggs (min – max = 2-19, SE = 0.70, N = 971) and 24.3 mm (min-max = 20.7-28.4;
SD = 0.04; N = 963 clutches), respectively. Clutch size increased with female body size (CS = 2
8.25+0.095 carapace length; r = 0.26, N = 280; (Congdon and van Loben Sels 1991), whereas

egg size and reproductive frequency increased with increasing female age (Congdon et al. 2000).
Mean carapace lengths of reproductive females each year (1976-2006; Fig. 3.2b,c) for east
residents (mean = 211.1 mm, min – max = 186-217, SE = 2.18), west residents (mean = 203.1
mm, min-max = 196-210, SE = 2.21), and non-residents (mean = 196.6 mm, min max = 183213, SE = 1.22). Resident and non-resident females that were part of the genetic study did not
differ in CL, age, clutch size, or egg widths (GLM, CL, t = 0.24, df = 48, P = 0.81; age, t =1.45,
df = 38, P=0.16; clutch size, t = -1.24, df = 33, P = 0.22; egg width, t = 0.42, df = 36, P = 0.68).

99

Genetic Diversity
All loci were in Hardy-Weinberg Equilibrium (HWE; P > 0.10). Loci were independent
(no evidence for gametic disequilibrium, P > 0.10) and polymorphic (average of 7.4 alleles per
locus, min - max = 4 - 15 alleles). Multi-locus expected heterozygosity was estimated to be 0.69
(min - max across loci 0.54 - 0.85). The multi-locus paternal non-exclusion probability with one
parent known was estimated to be 0.004. There was no evidence of null alleles at each locus (P
> 0.10). The multi-locus error rate was empirically estimated to be less than 1% across all loci.
Mating System and Reproductive Success
Females mated with multiple males within and among years and the within year average
number of sires per clutch was 1.5 (SD = 0.67, min-max = 1-3). Over the 8 years of the genetic
study, the average number of mates for a female was 2.02 males (SD = 1.05, min-max = 1-5) per
female. However, because some females were sampled more frequently than others, we also
adjusted the total number of males per female by the number of clutches sampled, after which
the number of sires per female averaged 1.36 (SD = 0.70, min-max = 1-3).
Occurrence of multiple paternity was 41.6% over all clutches but variable among years
(N = 77 nests; N = 49 females (27% of reproductive females); min-max = 15.4% - 55.6%; Fig.
3.2a). The incidence of multiple paternity was positively and significantly associated with
female age (Question 1: GLMM; Z = 2.037; P = 0.042), body size (carapace length; GLMM; Z =
2.19; P = 0.03), and clutch size (GLMM; Z = 1.998; P = 0.046). Incidence of multiple paternity
was similar for clutches of resident and non-resident ESGR females (GLM; Z = 1.238; P =
0.216). The total number of reproductive females each year was negatively correlated with the
yearly incidence of multiple paternity (Pearson's Rank Correlation; ρ = -0.857, P = 0.006).
100

Males that successfully mated with sampled females averaged 1.56 mates (SD = 0.91, N
= 32, min-max = 1-4) over the 8 years of the study. Due to the occurrence of repeat paternity,
the average number of clutches sired over the 8 years was higher (mean = 1.97, SD = 1.53, N =
32, min-max = 1-7). The average number of offspring sired over 8 years was 11.22 (SD = 8.62,
N = 32, min-max = 1-40).
Among males that successfully mated with sampled females that nested on the ESGR,
male reproductive success was significantly associated with the number of mates (GLMM; t =
6.042; P <0.001) and the number of clutches (GLMM; t = 13.091; P <0.001). Male reproductive
success was not significantly associated with male age (GLMM; t = 0.150; P = 0.88), size
(carapace length; GLMM; t = 0.618; P = 0.54), offspring sired per clutch (GLMM; t = 0.091; P =
0.928) or multiple paternity (GLMM; t = -1.846; P = 0.08). Because males mated with more
than one female, and were therefore represented multiple times in the dataset, we also tested for
factors associated with male RS using average values for all fixed effects in the model. There
were no differences between the use of the averaged and non-averaged fixed effects (male age,
size, number of mates per year and all years, number of clutches, and average clutch sizes). The
average number of mates per year, clutches per mate over all years, and offspring per mate over
all years were significant predictors of male reproductive success (GLMM; average mates per
year, t = 4.40, P <0.001; average clutches per mate, t = 4.57, P <0.001, average offspring per
mate t = 4.05, P <0.001). Male reproductive success was not influenced by average female body
size, average female age, average number of offspring sired per clutch, or multiple paternity
(Questions 2 and 4: GLMM; average age, t =0.44, P = 0.67; average CL, t =1.02, P = 0.32;
average offspring per clutch, t =1.16, P = 0.26, average multiple paternity; t = 0.04,P = 0.97).

101

The size distributions of successful males and all genetically sampled males were similar
2
(Fig. 3.4; X = 6.44, df = 18, P = 0.994). However, age distributions of successful and all
2
sampled males differed significantly (Fig. 3.4; X = 11.42. df = 5, P = 0.044). The youngest

successful male was 22 years of age.
Nests of 28 (57%) of 49 sampled females from 1999-2006 were sampled at least twice.
There was evidence of repeat paternity (Question 3) or different males in 7 and 8 clutches, of 28
females, respectively, whereas evidence from clutches of 13 females was inconclusive. Of 23
pair-wise comparisons of each female‟s nests among years, 69.6% showed evidence of repeat
paternity. Clutches of females reproducing in consecutive years had the highest incidence of
repeat paternity. Evidence of repeat paternity occurred over clutch intervals extending over the
four years of study (7 years; Fig. 3.5).
Gene flow estimated from paternity data
Resident females (N = 44) mated most often with males from the same residence wetland
(G- test = 9.06; P < 0.05; Table 3.1). The proportion of clutches of non-resident females (N =
46) sired by resident and non-resident males did not differ significantly (G-test = 1.57, P > 0.05).
However, non-resident females (N = 46) mated more often with west resident males than with
east resident males (G-test, G= 18.69, P < 0.05). No other comparisons were significant (Table
3.1). Resident and non-resident females had similar probabilities of having at least one offspring
in a clutch sired by a non-resident male (Questions 5 and 6: GLM; Z = -0.355; df = 30; P =
0.723). Regardless of residency, the probability of having at least one offspring in a clutch sired

102

by a non-resident male was positively associated with female age (Question 7: GLM; Z = 2.76;
df = 46; P = 0.006), but not body size (GLM; Z = 1.45; df = 50; p= 0.146).
The location of nests of east-resident females were generally nearer to their residence
than to the west-residence wetlands (Question 5), indicating wetlands may be spatially
genetically structured if offspring recruit primarily to the mother‟s residence. However, nonresident females nested throughout the reserve (Fig. 3.3).
Data from mother-offspring pairs (N=32) identified from the long-term mark-recapture
data indicated that dispersal of offspring from the mother‟s resident wetland may contribute to
gene flow among wetlands (Question 4). Half of the offspring (16) were from resident females (8
east, 8 west) and half (16) were from non-resident females; and 31 (96.9%) of the 32 offspring,
were from nests located in the central portion of the ESGR.

Population and Genetic Structuring
Genetic analysis of all individuals sampled on the ESGR (west resident, east resident and
non-residents) revealed that there is only a single genetic cluster (Question 9: STRUCTURE, K
= 1, log-likelihood = -5132.0). Estimates of variance in allele frequency among females sampled
from resident and non-resident wetlands were not significantly different from zero (Fst=0.002, P
> 0.05).

Discussion
Females and reproductive characteristics.
103

Reproductive females from residence wetlands located on and off the ESGR were
captured frequently while nesting on the ESGR. Although the average number of east-resident,
west-resident, and non-resident females captured each year from 1976-2006 were similar (Fig.
3.1a), the total number of all resident females (N = 80) was about one half of the number of nonresidents (N = 170) that could have originated from up to six permanent wetlands within 1.5 km
2
(1.77 km area) of the ESGR (Congdon et al., 2011).

Mating system and Reproductive Success
Mate number can influence the effective population size and population levels of genetic
diversity (Sugg and Chesser 1994). Additionally, when clutches of eggs have multiple sires
(multiple paternity), offspring within the clutch are a mixture of full- and half-siblings. Because
offspring from multiple-paternity nests are genetically more diverse (i.e., lower coancestry;
Scribner and Chesser 2001), the offspring from clutches sired by multiple males represent
potentially important sources of genetic variation. Thus, it is important to understand the
circumstances that contribute to the frequency of multiple matings by females. Both sexes of
Blanding‟s turtles on the ESGR mate with more than one individual and mate numbers of males
and females were similar (mean 1.56 and 2.02, respectively). The occurrence of multiple
paternity in clutches of eggs on the ESGR (41. 6%) was lower than that found in population in
Minnesota (56%; Refsnider 2009). Both estimates of multiple paternity in Blanding‟s turtles
were higher than have been observed in other freshwater species including Painted turtles
(Pearse et al. 2002, McGuire et al 2011), and the European pond turtle (Roques et al. 2006), and
are more similar to species with substantially higher clutch sizes such as Snapping turtles
104

(Chelydra serpentina, Galbraith et al. 1993) and sea turtles (Green turtles, Chelonia mydas,
Fitzsimmons 1998).

Multiple Paternity .-- The extensive movements of Blanding‟s turtles may also contribute to a
high incidence of multiple paternity (Ross and Anderson 1990; Hamernick 2001; Grourovic and
Sievert 2005). Two non-mutually exclusive movement-based mechanisms can affect multiple
paternity. First, encounters with different individuals are more likely to occur if adults visit
ephemeral wetlands widely distributed throughout the landscape compared to adults that remain
in a single residence wetland. Second, if the dispersal of individuals to ephemeral wetlands
decreases how often suitable mates are encountered, females may become less selective, and
mate indiscriminately with males when they are encountered (Kokko and Mappes 2005). In
contrast, an increase in the total number of males encountered may increase levels of female
choosiness and can decrease multiple matings by females (Kokko and Mappes 2005; Sztatecsny
et al. 2006; Rankin and Kokko 2007).
Overall, clutch size was positively associated with the frequency of multiple paternity.
However, it is unlikely that the relatively high occurrence of multiple paternity in Blanding‟s
turtles could be attributed solely to detection bias caused by clutch size variation among species.
Clutch size in Blanding's turtles is more similar to those in other Emydid turtles (average 9.9
eggs) and smaller than those of snapping turtles (average on the ESGR of 20.8 eggs) or green sea
turtles (>100 eggs; Fitzsimmons 1998).
The incidence of multiple paternity was variable among years (15.4% - 55.6%; Fig. 3.2)
and was negatively correlated with the number of reproductive Blanding‟s turtle females. A
similar relationship also occurred in Painted turtles, but only among females with the highest
105

reproductive qualities (McGuire et al.2011. A relationship between multiple paternity and
estimated breeding population sizes or density was found among species of sea turtles (Ireland et
al. 2003; Jensen et al. 2006). Based on the data provided in Jensen et al. (Table 3.1, 2006), the
mean occurrence of multiple paternity for species characterized by synchronized nesting was
higher than in species characterized by solitary nesting (75.0%; min-max = 58-92%) compared to
30.6% (min-max = 9-50%), respectively; a difference of 44%. The authors caution that multiple
paternity also varies from 9-62% among populations of a single species. Our results document
among-year variation in multiple paternity within a Blanding‟s turtle population, (min-max =
15.4- 55.6%; a difference of 40.2%). Among year variation in multiple paternity in Blanding‟s
turtles demonstrates that when levels of multiple paternity are estimated from different
populations sampled in different years, interpretations of differences may be confounded by
multiple sources of variation.
The age and body size of female Blanding‟s turtles on the ESGR were positive predictors
of incidence of multiple paternity (Question 1). Reproductive frequency and parental investment
(egg size) primarily increase as a function of female age, whereas clutch size (egg number)
primarily increases with body size (Congdon et al. 2001). If males are able to identify larger or
older females, they should increase efforts to mate with them; a result that could lead to a higher
occurrence of multiple paternity in high quality females. Additionally, if larger or older females
move more frequently or longer distances, they may have a higher probability of encountering
and mating with more males.
Male reproductive success.--Females appear to mate indiscriminately with respect to male size
and age. We found no evidence of male traits associated with male reproductive success. Male
attributes such as body size and age were not significant predictors of male reproductive success
106

(Fig. 3.4). Additionally, the youngest male to sire offspring was 22 years old, an age that is
consistent with estimates of age of maturity based on growth trajectories (Congdon et al. 2001).
Variation in reproductive success can accentuate the loss of genetic variation in the
population because some individuals contribute disproportionally to the population (and in some
situations some males may not contribute at all (Frankham 1995; Nunney 1996). Based on
reproductive data from females that nested on the ESGR, we documented sources of variation in
male reproductive success. The total number of offspring sired was variable among males (minmax = 1-40) and was significantly associated with the number of mates and the number of
clutches sired, but not with multiple paternity (Question 2). Because the number of mates was a
positive predictor of male reproductive success, pairings of males with multiple females can
increase levels of coancestry among offspring from different clutches (Chesser 1991).
Male reproductive success in many species has been shown to be primarily determined
by the number of mates (Bateman 1948; Clutton-Brock and Vincent 1991; Hauber and Lacey
2005). However, when substantial variation occurs in female reproductive qualities, female
quality can also contribute significantly to male reproductive success (Darwin 1871; Arnold and
Duvall 1994; Owens and Thompson 1994; McGuire et al. 2011). Our data on Blanding‟s turtles
indicates that male reproductive success was influenced by the number of female mates and the
number of clutches sired through remating or use of stored sperm by females (Question 4).
The number of years that viable sperm can be used by Blanding's turtle females has not
been documented; however in ESGR Blanding‟s turtles, repeat paternity was observed
throughout the duration of the study (maximum of 7 years between clutches). Additionally, the
number of offspring sired per clutch did not necessarily decline among year (e.g., 9/9 (100%)
and 10/10 (100%) despite a six-year interval; Question 3). Female turtles use stored sperm to
107

fertilize sequential clutches within, and potentially among years (Ewing 1943; Pearse and Avise
2001; Pearse et al. 2001; Pearse et al. 2002; Roques et al. 2006; McGuire et al. 2011). However,
use of stored sperm among years is associated with decreased fertility (Cuellar 1966; Jun-Yi
1982), that may result from sperm depletion or declining sperm viability (Goin et al. 1978; Gist
and Jones 1987; Palmer et al. 1998). In captivity, Eastern Box turtles (Terrapene carolina
carolina) females produce viable offspring for up to 4 years after removal from males (Ewing,
1943). Our observations of repeat paternity extending beyond 5 years without declines in the
proportion of offspring sired (e.g. 100% of offspring sired in clutches 6 years apart), suggests
that females may have remated with the same males in different years. Co-occupancy of habitats
(residence wetlands, overwintering sites, or ephemeral wetlands) could bring the same
individuals together repeatedly among-years. Regardless of mechanism, repeat paternity will
increase levels of relatedness (coancestry, Scribner et al. 1993) among hatchlings cohorts
produced in different years and potentially increase levels of spatial genetic structuring (Scribner
and Chesser 2001).

Gene Flow
Although resident females (east and west) mated predominantly with individuals from
their same residence wetland, some successful matings occurred between non-residents and
residents of the ESGR. Approximately one-third of east-resident and west-resident females
mated with a non-resident male (Table 3.1; Question 6). Non-resident females nesting on the
ESGR mated predominantly with other non-resident males and males from the west-residence.
Fewer successful matings occurred between non-resident females and east-resident males (Table
3.1). Matings were observed in permanent and ephemeral wetlands on the ESGR during most of
108

the activity season (Kinney 1999), and the genetics data from this study document matings
between individuals from different residences on the ESGR and between resident and nonresident individuals. Combined, the behavioral and genetics data support that successful matings
occur in ephemeral wetlands where encounters with individuals from different residences can
occur (Question 7). Males and females may utilize ephemeral areas to exploit seasonally
abundant resources (such as food) as well as to find mates (Kinney 1999; Pappas et al. 2000
Newton and Herman 2009; Congdon et al. 2011). Our data are consistent with the hypothesis
that in addition to resources, males also move long distances to encounter mates. Regardless of
wetland of residence, older females were more likely to have offspring sired by a non-resident
male. Older individuals may utilize larger core areas or they may have gained experience that
allows them to more efficiently exploit resources in ephemeral wetlands, and that in turn may
expose them to a greater number of potential mates. .
Females from east and west residence wetlands primarily nested in the eastern and
western areas of the ESGR, respectively (Fig. 3.3; Question 8). Therefore, the spatial
distribution of nests indicates a potential for spatial genetic structuring between east- and westresident individuals. Although the closest permanent wetland to nests may be the residence
wetland of their mother, ephemeral wetlands (that might lead to dispersal of hatchlings away
from residence wetlands) were much closer to many nests. For 32 ESGR resident motheroffspring pairs, 16 (50%) recaptured hatchlings came from non-resident females that that nested
on the ESGR Also, because 31 of the 32 parent-offspring pairs were from nests located in the
interior of the ESGR, nest-placement influenced the probability of juvenile recruitment into their
mother‟s residence wetland. Offspring of resident females that exit the ESGR to nest would also
be expected to have a low probability of dispersing into ESGR residence wetland. Patterns of
109

hatchling dispersal to residence wetlands other than the parents residences can homogenize allele
frequencies among residence wetlands (Question 8).
Population and genetic structuring
Results from the genetic analysis of population structuring (STRUCTURE; k=1)
demonstrated that individuals captured on the ESGR (residents and non-residents) represent a
single population and thus the population of Blanding's turtles extends beyond the boundaries of
the ESGR (Question 9). The lack of spatial genetic structuring is consistent with the occurrence
and frequency of matings between males and females from different residences and with the
probability that hatchling emerging from nests will disperse to wetlands other than their mother‟s
residence.

Conservation implications
Adult Blanding‟s turtles exhibit a propensity for long-distance terrestrial movements and
have the capacity to remain in terrestrial habitats over long periods (Power 1989; Ross and
Anderson 1990; Rowe and Moll 1991; Kinney 1999). Extensive movements made by adults,
high frequency of matings between resident and non-resident individuals, and high probability of
offspring to disperse from nests to permanent wetlands other than the parents residences
constitute two important but previously undocumented mechanisms that contribute to gene flow
among wetlands. Therefore, population boundaries extend well beyond permanent wetlands and
adjacent terrestrial areas suggested from the data collected during the long-term life history and
nesting ecology study on the ESGR.
Clutches of older females had a higher probability of being sired by multiple males and
by males that were non-residents. Therefore older females represent an important subset of the
110

population, not only because of their comparatively higher reproductive frequency and offspring
quality (Congdon et al. 2001), but also as an important means of distributing genes from multiple
males to offspring thereby reducing cohort levels of coancestry (Scribner and Chesser 2001). In
addition to the demographic value of older females to population dynamics and persistence
(Congdon et al. 1993), the loss of older females would also lead to increasing population levels
of coancestry and inbreeding that also influence population persistence. If there are age-based
differences in movements, greater exposure of older females to human-induced disturbance that
increase risk of mortality (such as roads) will substantially increase mortality thereby decreasing
the mean age in the population (Gibbs and Shriver 2002) and concomitantly decreasing rates of
gene flow among wetlands and increasing relatedness among offspring within and among year
cohorts.
Blanding's turtles are characterized by polygamous mating system, with both males and
females mating with multiple mates within and among years. The high percentage of multiple
paternity is consistent with previous work in other turtle species (and other taxa) demonstrating
that as male density declines, female choosiness may decrease, resulting in higher incidence of
multiple matings. Females apparently make temporary visits to ephemeral wetlands to exploit
seasonally abundant resources, whereas mate acquisition may be a more important factor than
resources for similar movements by males. Repeat paternity resulting from female use of stored
sperm or remating can accentuate loss of genetic diversity if multiple mates, and mates from
other residencies, are no longer encountered. Given the species‟ propensity for frequent and
extensive terrestrial movements, the integrity of the aquatic and terrestrial habitats are required to
allow long-distance movements of all ages among wetlands that are crucial to the persistence of
populations.
111

The results from this metapopulation case study highlight the importance of utilizing
multiple sources of data to ascertain mechanisms of inter-population gene flow. Interpretations
based on information on population size and core area gathered using a single method could have
resulted in inaccurate conclusions, that could in turn misdirect management decisions. For
example, data on fidelity to residence wetlands and the distribution of nests suggested that there
was population structuring within the ESGR. However the inclusion of data on paternity, male
reproductive success, mating pairs, and the degree of spatial genetic structuring led to a better
characterization of the population. Similarly, analyses of genetic data indentified mating pairs
and the spatial context in which encounters between males and females from different residence
wetlands would not have been possible without long-term life history and demography data.
The combination of multiple sources of data identified several factors that influence gene
flow. The higher reproductive output of older females coupled with increased multiple paternity
make them demographically and genetically valuable compared to young females. The
placement of nests of all females in areas that facilitate hatchling dispersal into wetlands other
than residences of mothers, and the propensity of adult males and females to utilize large areas of
terrestrial habitats surrounding ephemeral wetlands and mate with individuals from outside their
residencies were all important sources of gene flow. In combination this study and that of
(Congdon et al. 2011) demonstrate the importance of geographically expansive core areas of
Blanding's turtles to successful conservation and management efforts. Protection of the core
areas of Blanding‟s turtles will also likely aid in the conservation of the majority of other semiaquatic species.

112

APPENDIX 3

113

Table 3.1. Percents of matings (N = number of pairings) described during all years (2003-2006)
between males and females with respect to residence wetland (East, West, and non-resident).
Residence wetland of males
Residence

of
females

East

West

Non-resident

East (33)

57.6

9.1

33.3

West (11)

0.0

63.6

36.4

Non-resident (46)

wetland

8.7

47.8

43.5

Table 3.2. Percentages of among-year clutches produced by females where repeat paternity was
observed (N = number of comparisons) as a function of male and female marsh of residence (e.g.
repeat paternity occurring between east-resident females east-resident males).

Female
residence
wetland

East
West
Non-resident

Male residence wetland
East
West
Non-resident
64.3 (14) 25.0 (4)
50.0 (6)
NA
33.3 (3)
100.0 (1)
100.0 (1) 44.4 (9)
0.0 (3)

114

Fig. 3.1. Description of an integrative approach combining multiple sources of long-term data
used to answer research questions [Q] pertaining to the relative importance of nesting ecology
and movements, characteristics of mated pairs, and male reproductive success, to gene flow and
spatial genetic structuring.

[Question 1] Are offspring from clutches of older or larger females sired more often by
multiple males (multiple paternity) than clutches of offspring of younger or smaller
females?
[Question 2] Is female reproductive quality (number of clutches, number of offspring per
clutch) a substantial source of variation in male reproductive success (total number of
offspring sired).
[Question 3] Are clutches of a female fertilized by the same male(s) in different years (repeat
paternity)?
[Question 4] Does repeat paternity contribute significantly to male reproductive success?
[Question 5] Does long-term fidelity to residence wetlands and nest placement by females
contribute to spatial genetic structuring?
[Question 6] Does fidelity to a residence wetland promote mating with males from the same
marsh or influence repeat paternity?
[Question 7] To what extent does terrestrial movements of males and females facilitate matings
among individuals from different resident marshes?
[Question 8] Does nest location influence whether offspring recruit into the resident marsh of
the mother or disperse to another marsh?
[Question 9] Does repeat paternity and matings with individuals from different resident
marshes, coupled with variation in nest placement by females among years, an important
source of movements of male genes among marshes?
115

Fig. 3.2. The (a.) numbers of resident, non-resident and total reproductive females captured each
year (1983-2006) on the E. S. George Reserve and ranks of the percent of nests with multiple
paternity from 1999–2006 (% multiple paternity for each rank was 1 = 15.4, 2 = 36.4, 3=42.9.
4=44.4. 5 = 50.0, 6 = 53.8, 7 = 55.6; average carapace length (+ 1SE) by year of (b.) resident and
(c.) non-resident reproductive females.

116

Fig. 3.3. Distribution of Blanding's turtle (Emydoidea blandingii) nests on the 650 ha Edwin S.
George Reserve (boundary in dashed lines) from 1999-2006, residence wetland of females with
the number of nests per location are indicated within symbols (circles = east residents, squares =
west residents, triangles = non-residents of the E. S. George Reserve.

117

Fig. 3.4. Frequency histograms of (a) ages and (b) body sizes of males that successfully mated
with females that nested on the E. S. George Reserve (ESGR) and the (c) ages and (d) body sizes
of all ESGR resident males. Ages and sizes for successful males were determined from the
average of all ages/sizes for the years in which a male was successful over the period of study
(1999-2006). Ages for all resident males were adjusted to the age the male in 2003.

118

Fig. 3.5. Percentage of nests with evidence of the same male siring offspring with the same
female (repeat paternity) as a function of inter-annual nesting interval.

(1)
(7)
(2)
(3)

(3)

(5)
(2)

119

(2)

Fig. 3.6. Percentage of among-year comparisons where repeat paternity was observed from
males from the same (black) or different (gray) residence wetlands as the female.

(2) (2)

(5)

(7)
(6)

(9)
(3)
(7)

120

REFERENCES

121

References
Arnold, S.J., and Duvall, D. 1994. Animal Mating Systems: A synthesis based on selection
theory. The American Naturalist 143(2): 317-348.
Avise, J.C. 2001. Introduction. Journal of Heredity 92(2): 99. doi:10.1093/jhered/92.2.99.
Baker, R.E., and Gillingham, J.C. 1983. An analysis of courtship behavior in Blanding's turtle,
Emydoidea blandingii Herpetologica 39: 166-173.
Bateman, A.J. 1948. Intra-sexual selection in Drosophila. Heredity 2: 349-368.
Beaudry, F., Demaynadier, P.G., and Hunter, M. 2009. Seasonally Dynamic Habitat Use by
Spotted (Clemmys guttata) and Blanding's Turtles (Emydoidea blandingii) in Maine.
Journal of Herpetology 43(4): 636-645.
Bodie, J.R., and Semlitsch, R.D. 2000. Spatial and temporal use of floodplain habitats by lentic
and lotic species of aquatic turtles. Oecologia 122(1): 138-146. doi:10.1007/pl00008830.
Buhlmann, K.A., Congdon, J.D., Gibbons, J.W., Greene, J.L., and Mathis, A. 2009. Ecology of
Chicken Turtles (Deirochelys reticularia) in a Seasonal Wetland Ecosystem: Exploiting
Resourceand Refuge Environments. Herpetologica 65(1): 39-53. doi:10.1655/08-028r1.1.
Butler, B.O., and Graham, T. 1995. Early post-emergent behavior and habitat selection in
hatchling Blanding's turtles (Emydoidea blandingii), in Massachusetts. Chelonian
Conservation and Biology 1: 187-196.
Cagle, F.R. 1944. Home range, homing behavior, and migration in turtles. University of
Michigan Press, Ann Arbor.
Calhoun, A.K.J., Walls, T.E., Stockwell, S.S., and McCollough, M. 2003. Evaluating vernal
pools as a basis for conservation strategies: a Maine case study. Wetlands 23(70B): 81.
Chesser, R.K. 1991. Influence of Gene Flow and Breeding Tactics on Gene Diversity Within
Populations. Genetics 129(2): 573-583.
Clobert, J., Danchin, E., Dhondt, A.A., and Nichols, J.D. 2001. Dispersal. Oxford University
Press, Oxford.
Clutton-Brock, and Vincent. 1991. Sexual selection and the potential reproductive rates of males
and females. Nature 351: 58-60.
Congdon, J.D., Dunham, A.E., and van Loben Sels, R.C. 1993. Delayed Sexual Maturity and
Demographics of Blanding's Turtles (Emydoidea blandingii): Implications for
122

Conservation and Management of Long-Lived Organisms. Conservation Biology 7(4):
826-833.
Congdon, J.D., Graham, T., Herman, T., Lang, J., Pappas, M., and Brecke, B. 2008. Emydoidea
blandingii (Holbrook 1838)- Blanding's Turtle. Chelonian Research Monographs 5.
doi:doi: 10.3854/crm.5.015blandingiiv1.2008.
Congdon, J.D., Kinney, O.M., and Nagle, R.D. 2011. Spatial Ecology and Core Area Protection
of Blanding's Turtle (Emydoidea blandingii). Canadian Journal of Zoology 89:10971105.
Congdon, J.D., Nagle, R.A., Kinney, O.M., Osentoski, M., Avery, H., van Loben Sels, R.C., and
Tinkle, D.W. 2000. Nesting ecology and embyro mortality: implications for the
demography of Blanding's turtles (Emydoidea blandingii). Experimental Gerontology 36:
569-579.
Congdon, J.D., Nagle, R.D., Kinney, O.M., and van Loben Sels, R.C. 2001. Hypotheses of aging
in a long-lived vertebrate, Blanding's turtle (Emydoidea blandingii). Experimental
Gerontology 36(4-6): 813-827.
Congdon, J.D., Nagle, R.D., Kinney, O.M., van Loben Sels, R.C., Quinter, T., and Tinkle, D.W.
2003. Testing hypotheses of aging in long-lived painted turtles (Chrysemys picta).
Experimental Gerontology 38(7): 765-772.
Congdon, J.D., Tinkle, D.W., Breitenbach, G.L., and Sels, R.C.v.L. 1983. Nesting Ecology and
Hatching Success in the Turtle Emydoidea blandingi. Herpetologica 39(4): 417-429.
Congdon, J.D., and van Loben Sels, R.C. 1991. Growth and body size in Blanding's turtles
(Emydoidea blandingii): relationships to reproduction. Canadian Journal of Zoology 69:
239-245.
Congdon, J.D., and van Loben Sels, R.C. 1993. Relationships of reproductive traits and body
size with attainment of sexual maturity and age in Blandings's turtles. Journal of
Evolutionary Biology 6: 547-557.
Cuellar, O. 1966. Delayed fertilization in the lizard Uta stansburiana. Copeia 3: 549-552.
Darwin, C. 1871. The descent of man, and selection in relation to sex. Princeton University
Press, Princeton, NJ.
Ewing, H. 1943. Continued fertility in female box turtles following mating. Copeia: 112–114.
Fitzsimmons, N.N. 1998. Single paternity of clutches and sperm storage in the promiscuous
green turtle (Chelonia mydas). Molecular Ecology 7(5): 575-584.
123

Fiumera, A.C., Porter, B.A., Grossman, G.D., and Avise, J.C. 2002. Intensive genetic assessment
of the mating system and reproductive success in a semi-closed population of the mottled
sculpin, Cottus bairdi. Molecular Ecology 11(11): 2367-2377.
Frankham, R. 1995. Effective population size/adult population size ratios in wildlife: a review.
Genetical Research 66: 95-107.
Galbraith, D., White, B., Brooks, R., and Boag, P. 1993. Multiple paternity in clutches of
snapping turtles (Chelydra serpentina) detected using DNA fingerprints. Canadian
Journal of Zoology 71: 318–324.
Gibbons, J.W. 1976. Aging phenomena in reptiles pp 454-475. EAR, Inc., Maine.
Gibbons, J.W., and Greene, J.L. 1979. X-ray photography: a technique to determine reproductive
patterns of freshwater turtles. Herpetologica 35: 86–89.
Gibbons, J.W., Greene, J.L., and Congdon, J.D. 1983. Drought-related responses of aquatic turtle
populations Journal of Herpetology 17: 242-246.
Gibbons, J.W., Greene, J.L., and Congdon, J.D. 1990. Temporal and spatial movement patterns
of sliders and other turtles. In Life History and Ecology of the Slider Turtle. Edited by
J.W. Gibbons. Smithsonian Institute Press, Washington, DC. pp. 201-215.
Gibbs, and Shriver. 2002. Estimating the Effects of Road Mortality on Turtle Populations.
Conservation Biology 16(6): 1647-1652.
Gist, D.H., and Jones, J. 1987. Storage of sperm in the reptilian oviduct. Scanning Microscopy 1:
1839-1849.
Goin, C.J., Goin, O.B., and Zug, G.R. 1978. Introduction to herpetology. W.H. Freeman and Co.,
New York.
Goudet, J. 1995. FSTAT (Version 1.2): A Computer Program to Calculate F-Statistics. Journal of
Heredity 86(6): 485-486.
Graham, T., and Doyle, T.S. 1979. Dimorphism, courtship, eggs, and hatchlings of the
Blanding's turtle, Emydoidea blandingii (Reptilia, Testudines, Emydidae) in
Massachusetts. Journal of Herpetology 13: 125-127.
Grgurovic, M., and Sievert, P. 2005. Movement patterns of Blanding's turtles (Emydoidea
blandingii) in the suburban landscape of eastern Massachusetts. Urban Ecosystems 8:
203-213.

124

Griffith, S.C., Owens, I.P.F., and Thuman, K.A. 2002. Extra pair paternity in birds: a review of
interspecific variation and adaptive function. Molecular Ecology 11(11): 2195-2212.
doi:10.1046/j.1365-294X.2002.01613.x.
Hamernick, M.G. 2001. Home ranges and habitat selection of Blanding's turtle (Emydoidea
blandingii) at the Weaver Dunes, Minnesota. St. Mary's University of Minnesota,
Winona.
Hauber, M.E., and Lacey, E.A. 2005. Bateman's Principle in Cooperatively Breeding
Vertebrates: The Effects of Non-breeding Alloparents on Variability in Female and Male
Reproductive Success. Integr. Comp. Biol. 45(5): 903-914. doi:10.1093/icb/45.5.903.
Herman, T., Boates, S., Drysdale, C., Alcorn, E., Bleakney, S., Eaton, S., Elderkin, M., Gilhen,
J., Jones, C., Kierstead, J., McNeil, J., Mills, J., Mockford, S., Morrison, I., O'Grady, S.,
and Smith, D. 2002. National Recovery Plan for the Blanding's Turtle (Emydoidea
blandingii) Nova Scotia Population. . In Nova Scotia, Canada.
Hinton, T.G., Fledderman, P.D., Lovich, J.E., Congdon, J.D., and Gibbons, J.W. 1997.
Radiographic determination of fecundity: is the technique safe for developing turtle
embryos? Chelonian Conservation Biology 2: 409–414.
Ireland, J.S., Broderick, A.C., Glen, F., Godley, B.J., Hays, G.C., Lee, P.L.M., and Skibinski,
D.O.F. 2003. Multiple paternity assessed using microsatellite markers in green turtles
Chelonia mydas (Linnaeus, 1758) of Ascension Island, South Atlantic. Journal of
Experimental Marine Biology and Ecology 291: 149-160.
Jensen, M.P., Abreu-Grobois, F.A., Frydenberg, J., and Loeschcke, V. 2006. Microsatellites
provide insight into contrasting mating patterns in arribada vs. non-arribada olive ridley
sea turtle rookeries. Molecular Ecology 15(9): 2567-2575. doi:10.1111/j.1365294X.2006.02951.x.
Jones, A., and Ardren, W. 2003. Methods of parentage analysis in natural populations. Molecular
Ecology 12: 2511-2523.
Jones, B., Grossman, G.D., Walsh, D.C.I., Porter, B.A., Avise, J.C., and Fiumera, A.C. 2007.
Estimating Differential Reproductive Success From Nests of Related Individuals, With
Application to a Study of the Mottled Sculpin, Cottus bairdi. Genetics 176(4): 24272439. doi:10.1534/genetics.106.067066.
Jun-Yi, L. 1982. Sperm retention in the lizard Chamaeleo hoehnelii. Copeia 2: 488-489.
Kalinowski, S., Taper, M., and Marshall, T. 2007. Revising how the computer program
CERVUS accommodates genotyping error increases success in paternity assignment.
Molecular Ecology 16: 1099-1006.
125

Karl, S.A. 2008. The effect of multiple paternity on the genetically effective size of a population.
Molecular Ecology 17(18): 3973-3977. doi:10.1111/j.1365-294X.2008.03902.x.
Kennett, R.M., and Georges, A. 1990. Habitat Utilization and Its Relationship to Growth and
Reproduction of the Eastern Long-Necked Turtle, Chelodina longicollis (Testudinata:
Chelidae), from Australia. Herpetologica 46(1): 22-33.
King, T.L., and Julian, S.E. 2004. Conservation of microsatellite DNA flanking sequence across
13 Emydid genera assayed with novel bog turtle (Glyptemys muhlenbergii) loci.
Conservation Genetics 5(5): 719-725. doi:10.1007/s10592-004-1854-0.
Kinney, O.M. 1999. Movements and habitat use of Blanding's turtles in southeast Michigan:
implications for conservation and management. University of Georgia, Athens, GA.
Kokko, H., and Mappes, J. 2005. Sexual selection when fertilization is not guaranteed Evolution
59(9): 1876-1885. doi:10.1111/j.0014-3820.2005.tb01058.x.
Lee, P.L.M. 2008. Molecular ecology of marine turtles: New approaches and future directions.
Journal of Experimental Marine Biology and Ecology 356(1-2): 25-42. doi:DOI:
10.1016/j.jembe.2007.12.021.
Libants, S., Kamarainen, A.M., Scribner, K.T., and Congdon, J.D. 2004. Isolation and crossspecies amplification of seven microsatellite loci from Emydoidea blandingii. Molecular
Ecology Notes 4(2): 300-302.
Lowe, W.H., and Allendorf, F.W. 2010. What can genetics tell us about population connectivity?
Molecular Ecology 19(15): 3038-3051. doi:10.1111/j.1365-294X.2010.04688.x.
McGuire, J.M., Congdon, J.D., Scribner, K., and Capps, J. in press. Variation in Female
Reproductive Quality and Reproductive Success of Male Painted Turtles (Chrysemys
picta marginata). Canadian Journal of Zoology ##(#): ##-##.
Miller, K.A., Nelson, N.J., Smith, H.G., and Moore, J.A. 2009. How do reproductive skew and
founder group size affect genetic diversity in reintroduced populations? Molecular
Ecology 18(18): 3792-3802. doi:10.1111/j.1365-294X.2009.04315.x.
Mockford, S., Herman, T., Snyder, M., and Wright, J. 2007. Conservation genetics of Blanding‟s
turtle and its application in the identification of evolutionarily significant units.
Conservation Genetics 8(1): 209-219. doi:10.1007/s10592-006-9163-4.
Mockford, S.W., McEachern, L., Herman, T.B., Snyder, M., and Wright, J.M. 2005. Population
genetic structure of a disjunct population of Blanding's turtle (Emydoidea blandingii) in
Nova Scotia, Canada. Biological Conservation 123(3): 373-380. doi:DOI:
10.1016/j.biocon.2004.11.021.
126

Morreale, S.J., Gibbons, J.W., and Congdon, J.D. 1984. Significance of activity and movement
in the yellow-bellied slider (Pseudemys scripta). Canadian Journal of Zoology 62: 10381042.
Newton, E.J., and Herman, T.B. 2009. Habitat, movements, and behaviour of overwintering
Blanding's turtles (Emydoidea blandingii) in Nova Scotia. Canadian Journal of Zoology
87(4): 299-309.
Nunney, L. 1993. The Influence of Mating System and Overlapping Generations on Effective
Population Size. Evolution 47(5): 1329-1341. doi:Stable URL:
http://www.jstor.org/stable/2410151.
Nunney, L. 1996. The influence of variation in female fecundity on effective population size.
Biological Journal of the Linnean Society 59(4): 411-425. doi:10.1111/j.10958312.1996.tb01474.x.
Obbard, M.E., and Brooks, R. 1980. Nesting Migrations of the Snapping Turtle (Chelydra
serpentina). Herpetologica 36(2): 158-162. doi:http://www.jstor.org/stable/3891480.
Osentoski, M.F., Mockford, S.M., Wright, J.M., Snyder, M., Herman, T.B., and Hughes, C.R.
2002. Isolation and Characterization of microsatellite loci from Blanding's turtle,
Emydoidea blandingii. Molecular Ecology Notes 2(2): 147-149.
Owens, I.P.F. 2006. Where is behavioural ecology going? Trends in Ecology & Evolution 21(7):
356-361. doi:DOI: 10.1016/j.tree.2006.03.014.
Owens, I.P.F., and Thompson, D.B.A. 1994. Sex Differences, Sex Ratios and Sex Roles.
Proceedings of the Royal Society B: Biological Sciences 258(1352): 93-99.
Palmer, K.S., Rostal, D.C., Grumbles, J.S., and Mulvey, M. 1998. Long-Term Sperm Storage in
the Desert Tortoise (Gopherus agassizii). Copeia 1998(3): 702-705.
Pappas, M., Brecke, and Congdon, J.D. 2000. The Blanding‟s turtle of Weaver Dunes,
Minnesota. Chelonian Conservation and Biology 3: 557-568.
Pearse, D., Janzen, F., and Avise, J. 2002. Multiple paternity, sperm storage, and reproductive
success of female and male painted turtles (Chrysemys picta) in nature. Behavioral
Ecology and Sociobiology 51(2): 164-171.
Pearse, D.E., and Avise, J.C. 2001. Turtle Mating Systems: Behavior, Sperm Storage, and
Genetic Paternity. J Hered 92(2): 206-211. doi:10.1093/jhered/92.2.206.
Pearse, D.E., Janzen, F.J., and Avise, J.C. 2001. Genetic markers substantiate long-term storage
and utilization of sperm by female painted turtles. Heredity 86(3): 378-384.
127

Power, T.D. 1989. Seasonal movements and nesting ecology of a relict population of Blanding‟s
turtle in Nova Scotia. M.SC. Thesis,. Acadia University, Wolfville, Nova Scotia.
Pritchard, J.K., Stephens, M., and Donnelly, P. 2000. Inference of population structure using
multilocus genotype data. Genetics 155: 945-959.
Pusey, A., Williams, J., and Goodall, J. 1997. The Influence of Dominance Rank on the
Reproductive Success of Female Chimpanzees. Science 277(5327): 828-831.
doi:10.1126/science.277.5327.828.
Rankin, J.D., and Kokko, H. 2007. Do males matter? The role of males in population dynamics.
Oikos 116(2): 335-348. doi:10.1111/j.0030-1299.2007.15451.x.
Refsnider, J.M. 2009. High Frequency of Multiple Paternity in Blanding's Turtle (Emys
blandingii). Journal of Herpetology 43(1): 74-81. doi:10.1670/08-102r.1.
Roe, J.H., Brinton, A.C., and Georges, A. 2009. Temporal and spatial variation in landscape
connectivity for a freshwater turtle in a temporally dynamic wetland system. Ecological
Applications 19(5): 1288-1299.
Roques, S., Díaz-Paniagua, C., Portheault, A., Pérez-Santigosa, N., and Hidalgo-Vila, J. 2006.
Sperm storage and low incidence of multiple paternity in the European pond turtle, Emys
orbicularis: A secure but costly strategy? Biological Conservation 129(2): 236-243.
Rosenberg, M.S., and Anderson, C.D. 2011. PASSaGE: Pattern Analysis, Spatial Statistics and
Geographic Exegesis. Version 2 Methods in Ecology and Evolution 2 in press.
Ross, D.A., and Anderson, R.K. 1990. Habitat use, movements, and nesting of Emydoidea
blandingii in central Wisconsin. Journal of Herpetology 24: 6-12.
Rowe, J.W., and Moll, E.O. 1991. A radiotelemetric study of activity and movements of the
Blanding's turtle (Emydoidea blandingii) in northeastern Illinois. Herpetologica 25: 178185.
Scribner, K.T., S. Morreale, M.H., Smith, and J.W. Gibbons. 1995. Factors contributing to
temporal and age-specific genetic variation in the freshwater turtle Trachemys scripta.
Copeia. 1995:970-977.
Scribner, K.T., and Chesser, R.K. 2001. Group-structured models in analyses of he population
and behavioral ecology of poikilothermic vertebrates. Journal of Heredity 98: 180-189.
Scribner, K.T., Congdon, J.D., Chesser, R.K., and Smith, M.H. 1993. Annual Differences in
Female Reproductive Success Affect Spatial and Cohort-Specific Genotypic
Heterogeneity in Painted Turtles. Evolution 47(5): 1360-1373.
128

Semlitsch, R. 1998. Biological delineation of terrestrial buffer zones for pond-breeding
salamanders. Conservation Biology 12: 1113-1119.
Semlitsch, R., and Bodie, J.R. 2003. Biological Criteria for Buffer Zones around Wetlands and
Riparian Habitats for Amphibians and Reptiles. Conservation Biology 17(5): 1219-1228.
Sugg, D.W., and Chesser, R.K. 1994. Effective Population Sizes With Multiple Paternity.
Genetics 137(4): 1147-1155.
Sztatecsny, M., Jehle, R., Burke, T., and Hödl, W. 2006. Female polyandry under male
harassment: the case of the common toad (Bufo bufo). Journal of Zoology 270(3): 517522. doi:10.1111/j.1469-7998.2006.00120.x.
Team, R.C.D. 2007. R Core Development Team (2007) R: A language and environment for
statistical computing. R Foundation for Statistical Computing,Vienna, Austria. ISBN 3900051-07-0, URL http://www.R-project.org.
Wang, J., and Santure, A.W. 2009. Parentage and Sibship Inference From Multilocus Genotype
Data Under Polygamy. Genetics 181(4): 1579-1594. doi:10.1534/genetics.108.100214.
Weir, B.S., and Cockerham, C.C. 1984. Estimating F-Statistics for the Analysis of Population
Structure. Evolution 38(6): 1358-1370.

129

CHAPTER 4: COMPARATIVE ANALYSIS OF FACTORS INFLUENCING MATING
SYSTEMS AND MALE REPRODUCTIVE SUCCESS IN FRESHWATER TURTLES

130

Abstract
Interspecific comparative studies can help to identify factors associated with mating
tactics adopted by males and females, particularly when detailed life-history data are available.
We used 33 years of reproductive data combined with 4 years of genetic parentage data in three
species of freshwater turtles, Painted turtles (Chrysemys picta), Blanding's turtle (Emydoidea
blandingii), and Snapping turtles (Chelydra serpentina), to quantify how commonalities and
differences in species life history and demography influence multiple paternity, repeat paternity,
and male reproductive success. Incidence of multiple paternity was lower among Painted turtles
(14.1%) than among Blanding's turtles (46.8%) or Snapping turtles (51.7%). However,
incidence of multiple paternity varied among years in all three species. Female age was a
significant positive predictor of incidence of multiple paternity in Painted and Snapping turtles.
Female body size was a significant positive predictor of incidence of multiple paternity in
Blanding's and Snapping turtles. Clutch size was a significant positive predictor of incidence of
multiple paternity only in Blanding's turtles. The incidence of females using sperm from the
same male(s) to sire offspring among years was 38%, 69.56% , and 66.7% for Painted,
Blanding's and Snapping turtles, respectively. Paternity analysis conducted on Painted and
Blanding's turtles revealed that the number of clutches sired was a significant positive predictor
of male reproductive success in both species. Mate number was a significant positive predictor
of male reproductive success in Blanding's turtles but not in Painted turtles. Results demonstrate
that variation in reproductive quality among females, particularly frequency of clutch production,
significantly and positively affected male reproductive success in Painted and Blanding's turtles,
and may be a substantial source of variation in Snapping turtles.

131

Introduction
Long-term, individually-based reproductive ecology studies allow tests of long held but
largely emperically untested predictions in evolutionary ecology (Clutton-Brock et al. 1982;
Clutton-Brock 1988). Mating tactics and traits associated with increased reproductive success
are most appropriately viewed where the interplay between life history and behavioral ecology
allows the relative costs and benefits of such traits to become apparent (Sih and Bell 2008;
Bassar et al. 2010). Although life-history studies focus on age- and size-specific survivorship
and reproductive rates, traits that enhance competitive mating abilities often do so at the expense
of other fitness components, including longevity (Reznick et al. 2000). Thus, traits associated
with reproduction often exemplify the potential trade-offs between allocation of resources to
current versus future reproductive efforts.
Although the majority of life-history concepts are female based (Stearns 1992),
integration of data on males into life-history theory and population dynamics is important
(Rankin and Kokko 2007). For example, male density influences fertilization probability
(Rankin and Kokko 2007) and rates or intensity of female harassment (Lee and Hays 2004;
reviewed in Westneat and Fox 2010); both of which can result in females mating with multiple
males (Kokko and Mappes 2005; Sztatecsny et al. 2006). Male density and adult sex ratios can
also affect female mate choice, whereby females may become less choosy when males are less
frequently encountered (Kokko and Mappes 2005).
Males and females use different tactics to maximize reproductive success (Trivers 1972)
despite providing equal genetic contributions to offspring. Variation in reproductive success is
primarily a function of resource availability for females (Clutton-Brock et al. 1981; Stearns
1992). Male reproductive success is primarily determined by the number of females mated
132

(Bateman 1948; Andersson 1994; Rodriguez-Munoz et al. 2010), and by the traits of the females
mated (Arnold and Duvall 1994; Owens and Thompson 1994). However for both sexes,
variation in demographic and environmental conditions can influence mating behaviors and
reproductive success.
Among-year variation in factors associated with reproductive success can result from
variation in sex ratio and the number of reproductive individuals (Emlen and Oring 1977;
Crowley et al. 1991; Klemme et al. 2007), and within season variation in the ages and traits of
the subset of reproductive individuals (Coulson et al. 2001, Dreiss et al. 2010). Demographic
variation can alter the number and attributes of potential mates that are encountered and affect
the strength of mate preference (Dreiss et al. 2010; Milner et al. 2010). For example, the order in
which males are encountered by females and the attributes of those males can influence whether
or not individuals will reproduce, and levels of variation in reproductive success of males (e.g.
"trading up"; Pitcher et al. 2003; Uller and Olsson 2008).
Reproductive success is typically estimated based on short-term studies that sample a
subset of individuals rather than repeated measures of the same individuals over time (CluttonBrock 1988; Weatherhead and Boag 1997; but see also McElligott and Hayden 2000; fallow
bucks; Dama dama); Brown et al. 1997, mexican jay; Aphelcoma ultramarina). Documentation
of reproductive success over multiple reproductive episodes will capture more of the variation
experienced over an individual's lifetime and provide a more robust estimate of the factors that
influence variation among individuals in reproductive success (Clutton-Brock 1988; Linden and
Møller 1989).
Comparisons of mating tactics among species can reveal factors that have shaped life
history and behaviors (Tinbergen 1951; Christy and Salmon 1991), and can also provide a basis
133

for developing hypotheses regarding the evolutionary significance of traits (Doughty 1996).
However, because mating behaviors are often variable (Magellan and Magurran 2009), and can
be influenced by environmental and demographic variation that typically varies over space and
time (Emlen and Oring 1977), comparative studies can be enhanced by studying species that
cohabit the same area and over the same time period.
Turtles are good subjects for quantifying the relative effects of factors that influence male
reproductive success. Turtles provide no post-ovulatory parental care, do not form pair bonds,
and are not territorial. Therefore, data on turtle mating tactics may be interpreted in the absence
of complex social interactions. However, despite the benefits of studies of less complex systems,
less than 5% of behavioral ecology studies have been conducted on taxa other than birds, insects,
fish, and mammals (Owens 2006). Long-term studies of reptiles in general and of reproductive
success of long-lived, iteroparous species such as turtles specifically are rare (Avise 2001).
For many long-lived organisms, accurate age-specific reproductive information is
unavailable due to the inherent difficulties and costs associated with conducting long-term
studies (Tinkle 1979). A long-term study on life history and demography of freshwater turtles
was initiated by Owen Sexton (1953-1957), and continued by Henry Wilbur (1968-1972, Donald
Tinkle (1975-1979), and Justin Congdon (1975-2007) on the E.S. George Reserve (ESGR) in
southeastern Michigan. The long-term study was conducted on three species of freshwater turtle
(Painted turtles (Chrysemys picta), Blanding's turtles (Emydoidea blandingii) and Snapping
turtles (Chelydra serpentina)) that cohabit the same wetland complexes during the same
extended time period. Information on reproductive histories of individual females of known ages
and sizes, and demographic information such as adult sex ratios, are available for all three
species. Additionally, long-term data are available on variation in female reproductive quality
134

(i.e., age- and size-specific clutch size, clutch frequency, and egg size). Populations of all three
species on the ESGR have been essentially stable over the past 3 decades. Data from those
studies provide an opportunity to use information on demography, life history, and female
reproduction as a basis for conducting a shorter term genetic study of factors influencing the
mating system and male reproductive success
The goals of this study were to characterize and compare life-history, demographic, and
genetic information of turtles on the ESGR to test predictions of how variation among species in
life history and demography influence differences in 1) incidence of multiple paternity, 2)
incidence of repeat paternity and 3) male reproductive success. There are substantial differences
in population sizes and sex ratios, life-history trait values, and behavior among the three species
on the ESGR (Table 4.1). In the sections below, we summarize demographic, life-history , and
behavioral characteristics traits of the three species of turtles and provide information on
reproductive qualities of females (Table 4.1). We propose several factors that can directly or
indirectly influence the incidence of multiple or repeat paternity in Table 4.2 and suggest
predictions of higher or lower values of incidence of multiple or repeat paternity likely to be
exhibited by each species for trait values described. For example, based solely on differences in
sex ratio, we predict that multiple paternity will be higher in Painted turtles compared to
Blanding's or Snapping turtles because there are more than twice the number of adult males than
females in the ESGR Painted turtle population.
Predictions outlined in Table 4.2 for multiple and repeat paternity, assume that all other
traits are held constant. Traits that influence male reproductive success are not usually
independent (Fig. 4.2), and the patterns observed in multiple paternity, repeat paternity, and
reproductive success represent the outcomes of several processes (e.g. pre- and post-copulatory
135

mechanisms; Uller and Olsson 2008). We integrate the results from each individual prediction
(Table 4.2) and through comparisons within and among species, explain how factors interact to
influence male reproductive success (Fig. 4.2).

Sources of Variation in Female Quality
Variation in Clutch Size and Egg size
Within species variation in clutch size (number of eggs produced per reproductive bout)
can affect male reproductive success (number of offspring). In Painted turtles, average clutch
sizes are smallest (6.99, SD = 0.19) and are similar between the first (mean = 6.9, SD = 1.47) and
second (mean = 6.7, SD = 1.55) clutches produced within a season. Clutch sizes of Blanding's
turtles average 9.67 (SD = 0.40) and Snapping turtles clutch sizes average 20.80 (SD = 1.05).
Clutch size is positively related to females body size in all three species (Tables 4.1 and 4.3).
Although most female turtles do not provide behavioral or trophic parental care, they do
provide hatchlings with substantial yolk reserves necessary to sustain juveniles during early posthatchling periods (Congdon et al. 1983). Because egg widths are positively related to egg size,
yolk volume and hatchling size, variation in egg widths within a species represent another
measure of offspring (and thus female) quality. Egg widths average 17.3 mm (SD = 0.22) in
Painted turtles, 23.7 mm (SD = 0.16) in Blanding's turtle, and 27.6 mm (SD = 0.29) in Snapping
turtles. In Painted and Snapping turtles, egg widths from X-radiographs were positively
correlated with female body size and age (Tables 4.1 and 4.3), whereas in Blanding‟s turtles egg
widths were not strongly correlated with female body size or age.
The strong positive relationship between body size and egg size in a small-bodied species
such as Painted turtles is in part due to relaxation of constraints on egg width caused by the size
136

of the pelvic aperture (Tucker et al. 1978; Congdon and Gibbons 1987), but is not known to
occur in larger-bodied species, such as Snapping turtles. Larger Painted turtles produce larger
eggs, suggesting that increasing offspring quality outweighs the cost of producing fewer
offspring (Smith and Fretwell 1974). Because there are positive relationships between body size
and clutch size and egg size in Painted and Snapping turtles (Tables 4.1 and 4.3), variation in
female body size provides males with a reliable indicator of female quality.

Clutch Frequency
Average reproductive frequencies were 1.3, 0.80, and 0.85 for Painted, Blanding's, and
Snapping turtles, respectively (Painted turtles can produce more than one clutch per year; Table
4.1), and among year reproductive frequencies vary among females of all three species (Table
4.1).

Although Blanding's and Snapping turtles produce a maximum of one clutch per year and

Painted turtles can produce up to 2 clutches per year (Congdon et al., 2001, 2003), some females
of all three species fail to reproduce each year (Table 4.1).
Reproductive frequency is an important component of female reproductive output (i.e.,
number of offspring produced = clutch size * number of clutches) and the number of offspring a
male sires with a female is how male reproductive success is achieved. Therefore, production of
second clutches within a year by Painted turtle females coupled with a high frequency of repeat
paternity in those clutches represents a major way for males to increase reproductive success that
is not available to Blanding‟s and Snapping turtle males. Gist and Congdon (1998) hypothesized
that a primary function of stored sperm is to fertilize sequential clutches within a year due to the
short period (less than 6 days) between clutches when eggs in oviducts of females can be
fertilized before shelling of eggs begins. The Gist and Congdon hypothesis was supported in
137

ESGR painted turtles where repeat paternity was observed in >95% of second clutches (McGuire
et al. 2011) and was observed in > 85% of clutches in a Painted turtle population in Illinois
(Pearse and Avise 2001; Pearse et al. 2001; Pearse et al. 2002).
Female turtles store sperm (Ewing 1943; Gist and Jones 1989), and repeat paternity has
been documented within and among years in several turtle species (Fitzsimmons 1998; Kichler et
al. 1999; Pearse and Avise 2001; Pearse et al. 2001; Pearse et al. 2002; Roques et al. 2006; Lee
2008; McGuire et al. 2011). Therefore, we predict that repeat paternity among years will be
observed in all three species (Table 4.2) and the potential benefits to male reproductive success
will be substantial. Longer durations between reproductive intervals can increase the likelihood
of individuals encountering and mating with different individuals. Additionally, declines in
sperm viability or sperm depletion (Goin et al. 1978; Gist and Jones 1987; Palmer et al. 1998)
may decrease the likelihood of stored sperm use over longer time intervals. Increases in
encounters among years and multiple matings (Uller and Olsson 2008) could increase incidence
of multiple paternity (Table 4.2).
Female body size and age are positively associated with female quality (clutch size,
clutch frequency, and egg size; Congdon et al. 2001; Congdon et al. 2003). Males can obtain
reproductive benefits by mating with high-quality females. Therefore, the number of males
attempting to mate and the intensity of courtship may increase with larger and older females.
Therefore we predict that in all three species, incidence of multiple paternity will be higher in
clutches from older/larger females than in clutches from younger/smaller females (Table 4.2).
Increased encounters with males may decrease the incidence of repeat paternity in older/larger
females (Table 4.2).

138

Source of variation in number of mates
Courtship
Courtship behaviors can influence mate number by providing an opportunity for mate
choice (Halliday 1983; Wiley and Poston 1996). Male courtship in Painted turtles involves use
of the foreclaws to titillate the face of the female (Berry and Shine 1980; Harding 1997). Male
courtship in Blanding's turtles involves chasing, swaying, chin inflation, head chinning, head
weaving, shell thumping, gulping, and bubbling (Graham and Doyle 1979; Baker and
Gillingham 1983; Ernst et al. 1994). Snapping turtle mating includes face to face posturing with
head swinging (Legler 1955) and water gulping, bubbling, and biting of the head and neck of the
female (Taylor 1933; Ernst et al. 1994). Although courtship behaviors allow individuals to
evaluate potential mates, to date there are no studies that document mate choice in the Painted,
Blanding's or Snapping turtles.

Sexual Size Dimorphism
Sexual size dimorphism can affect courtship and mating behaviors (Berry and Shine
1980; Weatherhead et al. 1995; Fairbairn 1997; Zamudio 1998; Blanckenhorn 2005). Among
the species on the ESGR all three possible size combinations occur: 1) females are larger than
males in Painted turtles (Congdon et al. 2003), 2) male and females are similar in size in
Blanding's turtles (Graham and Doyle 1979; Congdon and van Loben Sels 1991; Rowe 1992;
Germano et al. 2000; Pappas et al. 2000), and 3) males are substantially larger than females in
Snapping turtles (Berry and Shine 1980; Congdon et al. 2008; Table 4.1). The larger size of
Snapping turtle males contributed to speculation that males force copulations (Berry and Shine
1980). If Snapping turtle males are able to forcefully copulate with females, the incidence of
139

multiple paternity may be higher than in other species (Table 4.2; Berry and Shine, 1980).
However, because only a subset of males would be large enough to force copulations, we predict
that more multiple matings may occur in smaller females and small males should not be able to
mate with large females (Table 4.2).

Population sizes and sex ratios
Population size is positively associated with multiple paternity among some sea turtle
populations (Ireland et al. 2003; Jensen et al. 2006). If freshwater turtles exhibit similar
tendencies, we predict that multiple paternity will be higher in Painted turtles (largest population)
than in Blanding's or Snapping turtles (both smaller populations; Table 4.2). Adult sex ratios are
male biased (2.3:1) in Painted turtles, and approximately 1:1 in resident Blanding‟s (Congdon et
al. 2011) and Snapping turtles (Congdon et al 1987). Based on differences in adult sex ratios
(Table 4.1), we predict that multiple paternity will be higher in Painted turtles than in Blanding‟s
and Snapping turtles (Table 4.2). However, because females would encounter more males (and
thus more opportunities to mate with different individuals), repeat paternity will be lower in
Painted turtles than Blanding's or Snapping turtles with respect to both population size and sex
ratio (Table 4.2).
If all males are reproductive each year, variation in the number of reproductive females
each year will influence the operational sex ratio (Fig 4.1). Mating attempts among the females
that are reproductive may increase in years when the number of reproductive females is low.
We predict that incidence of multiple paternity will increase and repeat paternity will decrease in
years where the number of reproductive females is low (Fig. 4.1, Table 4.2)

140

Morphological traits that influence terrestrial movement
The three species have different shell characteristics that influence the degree to which
individuals use terrestrial habitats, that can in turn influence mate encounter rates. The shells of
Blanding's turtles are hinged and provide protection from predators and reduce desiccation
(Greene 1988; Ernst and Barbour 1989; Martin et al. 2005) while they are in terrestrial habitats.
The large body size and aggressive behavior of Snapping turtles provides some protection from
some predators. However their reduced plastron and skin characteristics increases susceptibility
to desiccation while in terrestrial habitats (Finkler 2001). Painted turtles are the smallest of the 3
species and their body size combined with rudimentary kinesis of the plastron provides little
protection from predators and nesting females experience much higher predation rates than do
Blanding‟s or Snapping turtles.
Terrestrial movements of Painted and Snapping turtles are infrequent except for female
nesting migrations that are short in distance and duration compared to those of Blanding‟s
turtles. Male and female Blanding's turtles make frequent terrestrial movements to ephemeral
wetlands, and females make nesting migrations up to 1600m (Congdon et al. 2011). Painted and
Snapping turtle females make much shorter nesting migrations (an average of 65.9 and 34.5m
respectively). Unlike the hatchlings of Blanding's turtles, hatchlings of both Painted and
Snapping turtles have a high probability dispersing from nests directly into the residence wetland
of their parents. If matings occur in residence and ephemeral wetlands, then the number of
potential mates encountered in both habitats will probably be larger than just those encountered
within their residence wetland (Table 4.2). Additionally, if individuals are more widely
distributed across the landscape, mate density may decrease, which can affect mate choice and
mate number (Table 4.2).
141

Methods
Field Methods
The field methods used to collect size, age, and reproductive data for females of all three
species of turtles on the ESGR were essentially the same with two substantial exceptions. First,
sample collection of Painted turtles was restricted to East Marsh (and surrounding nesting areas)
because the population size of Painted turtles is substantially larger than those of Blanding's
turtles or Snapping turtles. Blanding's turtles and Snapping turtles were sampled from all areas
on the ESGR. Second, the use of a drift fence that completely surrounded East Marsh was not
equally effective at capturing females of all species. Whereas nearly all gravid Painted turtle
females were captured at the fence as they moved to nesting areas during daylight hours, some
Snapping turtle females nested in the early morning (between 0100-0500 h) when the drift fence
was not monitored consistently in most years. The use of the drift fence was least effective for
Blanding‟s turtles because some females move to ephemeral wetlands before the beginning of
the nesting season.
Reproductive frequency
To compare inter-annual reproductive frequency among the three species, we excluded
among-year intervals > 4 years because long intervals probably represent unusual durations
between reproductive events (e.g. those resulting from injuries, disease, temporary movements to
other wetlands, and nesting off of the ESGR). Because Painted turtles can produce two clutches
in a year, whereas Blanding's and Snapping turtles produce a maximum of one, we restricted
reproductive frequency to only among-year reproductive events. We calculated the mean
142

reproductive intervals of individuals to standardize variation in total inter-capture intervals
ranging from 5 to >30 years.
Genetic Methods
Snapping turtle DNA was extracted from blood (adults; N = 461) or tissue (hatchlings; N
= 89 nests, 1575 hatchlings) using the Qiagen DNeasy kit (Qiagen Inc, Valencia, CA) and
amplified 4 microsatellite loci Cp3, (Pearse et al. 2001), BTCA7, BTGA5 (Libants et al. 2004),
and D111 (Hackler et al. 2007). Painted turtle samples (N = 535 adults, 155 nests, 1054
hatchlings) were genotyped at 6 microsatellite loci (Cp2, Cp3, (Pearse et al. 2001), BTGA2,
BTGA3, BTCA7, (Libants et al. 2004) and EB11 (Osentoski et al. 2002)). Blanding's turtle
samples (247 adults, 77 nests, 611 hatchlings) were genotyped at 8 microsatellite loci including
BTGA5, BTCA11, BTCA9, (Libants et al. 2004), EB11, EB19, EB17, EB09 (Osentoski et al.
2002), and GmuD70 (King and Julian 2004).
For all three species, amplification of DNA was performed using polymerase chain
reaction (PCR) using published conditions and separated using gel electrophoresis on 6%
polyacrylamide gels. We used an FMBIOII scanner (Hitachi Inc., Tokyo, Japan) to visualize
PCR products. All genotypes were independently scored by a minimum of two experienced
laboratory personnel. To estimate error, and ensure genotyping consistency, 10% of adults were
randomly selected and re-genotyped for all loci. Estimates of expected heterozygosity, number
of alleles, the exclusion probability with one parent known, and tests for Hardy-Weinberg
Equilibrium were performed using the program CERVUS 3.1.0 (Kalinowski et al. 2007).
Paternity, multiple paternity, and repeat paternity
143

For Painted and Blanding's turtles, we utilized the methods for parentage, multiple
paternity, and repeated paternity as described in Chapters 2 and 3 that relied on consensus
between two parentage programs: NEST (Jones et al. 2007) and CERVUS (Kalinowski et al.
2007).
For Snapping turtles, we used a minimum allele count to determine whether more than
one male sired offspring within a clutch (multiple paternity). Because we observed females in
nesting activity, we could identify paternal alleles. When 3 or more paternal alleles at a single
locus could be identified, the clutch was assigned as multiple paternity.
Paternal alleles were used to reconstruct multi-locus paternal genotypes that were
compared among years to provide evidence of repeat paternity. When reconstructed paternal
genotypes were consistent among years at all loci, the clutch was identified as repeat paternity.
Estimates of repeat paternity were based on all available data for all three species (Painted turtles
2003-2006, Blanding's turtles 1999-2006, Snapping turtles 2002-2006).
We lacked sufficient power to determine paternity for Snapping turtles and therefore
analyses of male reproductive success were restricted to Painted turtles and Blanding's turtles.

Statistical Analyses
We quantified associations between female age, female body size (carapace length),
clutch size (number of eggs), with incidence of multiple paternity for Painted turtles, Blanding's
turtles, and Snapping turtles using generalized linear mixed models (GLMM) using multiple
paternity as a binomial dependent variable. Female was used as a random effect in all models
144

because nests from the same female were often sampled more than once in the course of the
study. Data for all three species were primarily based on samples collected from 2003-2006 to
facilitate comparisons among the species.
We tested for associations between female age and body size with incidence of multiple
paternity for Painted turtles and Snapping turtles using generalized linear mixed models
(GLMM) with repeat paternity as a binomial dependent variable. Low sample sizes precluded
our ability to use a mixed model in Blanding's turtles and so models were run as generalized
linear models (GLM) with repeat paternity as a binomial dependent variable.
To quantify associations between the total number of mates, the total number of clutches,
the average offspring per clutch, and whether or not a clutch had multiple sires (multiple
paternity) with male reproductive success (total number of offspring sired) using general linear
models (GLM). Mixed model analysis was not necessary because each male was only
represented once in the dataset. We used average values for the number of offspring per clutch
and multiple paternity determined for each male.

Results
Variation in female numbers and quality
The number of reproductive females of all three species varied among years (Fig. 4.1a).
However, the annual number of reproductive females was not correlated among the species
(Pearson; ρ <0.299, df = 23, P > 0.156).
The average carapace lengths of female Painted and Blanding's turtles increased slightly
over the course of the life history study; whereas the average carapace length of Snapping turtle
females decreased (Fig. 4.1b). Because all three species of turtles are long-lived (i.e., some
145

individuals remained in the study over many years) it was expected that the average ages of all
three species would increase over time (Fig. 4.1c).
Carapace length of females was significantly and positively related to age in all three
2
2
species, but the relationship was strongest in Snapping turtles (R = 0.59) and Painted turtles (R
2
= 0.33) compared to Blanding‟s turtle (R = 0.10; Table 4.1). The relationships between clutch

size and body size and clutch size and age were similar among species (Table 4.1). Egg widths
increased significantly with body size and age of females in Painted and snapping turtles, but
those relationships were weak in Blanding‟s turtles (Table 4.1).
Average among-year reproduction intervals were slightly shorter for Painted turtles than
for Blanding's or Snapping turtles (Painted turtles; mean = 1.5, min-max = 1-4, SD = 0.43;
Blanding's turtles, mean = 1.7, min-max = 1-4, SD = 0.55); Snapping turtles, mean = 1.7, minmax = 1-3, SD = 0.49; Table 4.4).

Analyses based on genetic data
Genetic diversity
For all three species, all loci were in Hardy-Weinberg Equilibrium (HWE; P > 0.10) and
loci were independent (P > 0.10). In Painted turtles, the mean number of alleles per locus over
the six loci was 19.2 (min-max across loci = 7-34), mean expected heterozygosity was 0.807
(min-max across loci = 0.689-0.885), and the multi-locus exclusion probability with one parent
known was 0.999. In Blanding's turtles, loci were polymorphic with an average of 7.4 alleles
(min-max = 4 - 15 alleles) and a multi-locus estimate of expected heterozygosity of 0.69 (minmax = 0.54 - 0.85). The multi-locus non-exclusion probability with one parent known was
146

0.996. In Snapping turtles, the mean number of alleles per locus was 17.8 (min-max = 4-27) and
the mean expected heterozygosity was 0.68 (min-max = 0.15 - 0.91). For Snapping turtles, the
multi-locus exclusion probability with one parent known was 0.984. Insufficient power
precluded paternity analysis. For all three species the empirically estimated multi-locus error
rate was less than 1% (mean over all loci).

Multiple paternity
Over four years (2003-2006), the incidence of multiple paternity averaged 14.1%, 46.8%,
51.7% for Painted turtles, Blanding‟s turtles, and Snapping turtles, respectively. Annual
proportions of multiple paternity within species also varied among years for Painted turtles (30.0,
12.5, 14.3 and 6.1%), Blanding‟s turtles (53.8, 55.6, 36.4, and 42.9%), and Snapping turtles
(61.5, 57.1, 44.4, and 50.0%) for 2003-2006, respectively.
The three species varied in the relationships between the number of reproductive females
and incidence of multiple paternity. In Painted turtles, the number of reproductive females was
inversely related to the incidence of multiple paternity among high-quality females (McGuire et
al. 2011). However the relationship between multiple paternity and the number of reproductive
female Painted turtles was less pronounced among all females (Chapter 2). In Blanding's turtles,
from 2003-2006 there was no substantial variation in the total number of reproductive females
(includes both residents and non-residents; Chapter 3). However, among all years of genetic
sampling for Blanding's turtles (1999-2006), the year with the lowest multiple paternity, also had
the highest number of reproductive females. In Snapping turtles, only one year (2004) had a low
number of reproductive females and multiple paternity was 57.1%, which was consistent with
the average amount of multiple paternity for the species.
147

The three species exhibited similarities in the attributes associated with incidence of
multiple paternity. In Painted turtles, female age was a significant predictor of multiple paternity
(Table 4.5; GLMM; age, Z = 2.01, P = 0.044) whereas female body size and clutch size were not
(body size, Z = -0.228, P = 0.819; clutch size Z = -0.316, P = 0.752). Age of Blanding's turtle
females was not a significant predictor of multiple paternity (Table 4.5; GLMM; Z = 0.977, P =
0.328), but body size was marginally significant (GLMM; Z = 1.896, P = 0.058) and clutch size
was a significant predictor of multiple paternity (GLMM; Z = 2.507, P = 0.012). However, when
the entire dataset is used for Blanding's turtles (1999-2003), female age, body size, and clutch
size were all significant (GLMM; age, Z = 2.04, P = 0.042; body size Z = 2.19, P = 0.029, clutch
size Z = 2.787, P = 0.005). In Snapping turtles, female age and body size (carapace length) were
both significant predictors of multiple paternity (Table 4.5; GLMM; age, Z = 2.67, P = 0.008;
body size, Z = 2.62, P = 0.009), whereas clutch size was not (GLMM; Z = 0.92, P = 0.357).

Repeat paternity
Paternal genotypes that were consistent with the same male(s) siring offspring with the
same female among years (repeat paternity) was observed in each species. The mean incidence
of repeat paternity was (38%, 69.56%, and 66.7% for Painted, Blanding's and Snapping turtles
respectively; Fig. 4.3). In two of the three species, the incidence of repeat paternity was higher
in clutches produced in consecutive years than for longer inter-clutch intervals (38.0%, 85.7%,
and 82.4% for Painted, Blanding's and Snapping turtles respectively; Fig. 4.3). In all three
species no female attributes (age, size) were associated with incidence of repeat paternity
(Painted turtle; GLMM; age, Z = 0.043; P = 0.966; size, Z = 0.792, P = 0.429; Blanding's turtle;

148

GLM (GLMM did not converge); age, Z = 0.958, P = 0.338; size, Z = 0.698, P = 0.485;
Snapping turtle; GLMM; age, Z = 1.464, P = 0.143, size Z = 1.44, P = 0.150).

Male reproductive success (total number of offspring)
In Painted turtles, male reproductive success was significantly associated with the
number of clutches sired (Table 4.6; GLM, t = 10.238; P < 0.001) and the average number of
offspring sired per clutch (GLM, t = 4.252, P < 0.001). Male reproductive success was not
associated with the number of mates (Table 4.6; GLM, t = 1.181; P = 0.242) or attributes of the
male (GLM; body size; t = 0.872; P = 0.387; age: t = -1.132, P = 0.272) and was marginally
associated with multiple paternity (GLM; t = -1.90; P = 0.06).

Male reproductive success in Blanding's turtles was influenced by the total number of
mates, the total number of clutches, and whether the clutch was sired by more than one male
(Table 4.6; GLM; total mates, t = 4.355, P <0.001; total clutches, t = 8.071, P <0.001; multiple
paternity, t = -2.348, P = 0.027). Male reproductive success was not influenced by the average
number of offspring sired per clutch (Table 4.6; GLM, t = 1.732, P = 0.096) or attributes of the
male (Table 4.6; GLM; body size, t = 0.794, P = 0.437; age, t = 0.084, P = 0.934).

Discussion
Variation in female reproduction
The number of reproductive females varied substantially among-years in all three species
(1983-2006; Fig. 4.1a). However, annually the numbers of reproductive females each year were
not correlated among species, suggesting that annual resource abundance was not the only factor
149

involved and that dietary differences among species may play a role in determining the number
of reproductive females each year. Although average body sizes and ages of Painted turtles and
Blanding's turtles increased over the duration of the life history study, the average body size of
female Snapping turtles declined from the mid-1990s through 2006 while ages generally
increased (Fig. 1b,c). The decline in body size apparently resulted from smaller primiparous
females recruiting into the reproductive population. Continued recruitment of small primiparous
females into the Snapping turtle reproductive population made up of larger females could
strengthen the relationship between body size and age. However, the relationship between body
2
size and age was essentially stable over the entire period (prior to 1991 R = 0.56 and from
2
1983-2006 (R = 0.59; Table 4.1). Therefore relationships between body size and age seem to

be unaffected by the recruitment of smaller reproductive females, but overall among female
variation in body size and reproductive quality would still increase.
Female reproduction is typically limited by the resource availability (Stearns 1992) and
reproductive intervals were variable among females within each species. Painted turtles on the
ESGR produce a maximum of two clutches per year while Blanding's and Snapping turtle
females produce a maximum of one clutch annually, however, in all three species some females
fail to reproduce every year (Congdon et al. 2001; Congdon et al. 2003). Even though Painted
turtle females produce second clutches in some years, inter-annual reproductive intervals were
similar to those of female Blanding's and Snapping turtles (Table 4.4). Producing a second
clutch of eggs represents a substantial reproductive investment that can delay the time until the
next reproductive event for Painted turtles. However, the females with a high propensity to
produce two clutches actually had shorter among-year reproductive intervals (Chapter 2) than did
150

female Painted turtles that produced only one clutch. Reproductive allocation to the single
clutch averaging 10 and 28 eggs is substantially different for Blanding‟s turtles and Snapping
turtles, respectively. (Table 4.1). However, the intervals between reproductive events were
similar (reproductive intervals = 1.66 for both Blanding's and Snapping turtles; Table 4.4).
Despite the production of second clutches by Painted turtles, and the substantial difference in the
size of the single cutch produced by Blanding‟s and Snapping turtles, similar reproductive
intervals of all three species appear to be possible due to the resource-rich environment of ESGR
wetlands (Fiala and Congdon 1983). However, among female variation in reproductive
frequencies within species remains a potentially important source of variation in male
reproductive success.
Multiple paternity
Within-species variation in the number of reproductive females influenced the incidence
of multiple paternity. Incidence of multiple paternity was inversely related to the number of
reproductive females among high-quality females in Painted turtles (McGuire et al. 2011), and
among all reproductive females (resident and non-residents) in Blanding‟s turtles (Chapter 3).
However, there was no relationship between the number of reproductive females and multiple
paternity in Snapping turtles, and the relationship was weak between multiple paternity and the
number of reproductive females in Painted turtles when females of all reproductive qualities
were considered. If the number of reproductive males remains constant, variation in the number
of reproductive females among years will change the operation sex ratio. For example, if the
number of reproductive females declines, the operational sex ratio becomes more male biased,
and more males can result in more attempts to mate with each female. In Blanding's turtles, the
151

number of reproductive females was inversely associated with the annual proportion of clutches
with multiple paternity (Chapter 3). In Painted turtles, the proportions of clutches exhibiting
multiple paternity was positively related to ages of females in years where the number of
reproductive females was low, but not when the number of reproductive females was high
(Chapter 2).
Several direct and indirect benefits may exist for males and females that have more than
one mate (Zeh and Zeh 1996). Aside from fertilization assurance, direct benefits are unlikely
because turtles provide no post-ovulatory parental care and no trophic resources are provided by
males prior to or following matings (Uller and Olsson 2008). Females may obtain indirect
benefits (e.g. "good genes") through pre-copulatory mate choice or post-copulatory sexual
selection (e.g. sperm competition or cryptic female choice) that can facilitate genetic
compatibility (Zeh and Zeh 1996; Uller and Olsson 2008). However, to date no empirical
evidence supports that female turtles benefit from polyandry (Lee and Hays 2004). An
alternative hypothesis for multiple matings is "convenience polyandry" (Thornhill and Alcock
1983), where females are expected to resist repeat matings unless the cost of resisting a male
exceeds the cost of mating.
In turtles, costs of matings for females could result from male harassment or through the
potential for forced copulation in situations where males are larger than the females. Harassment
of females has been shown to increase multiple matings in taxa other than turtles (water striders
(Aquarius remigis), Watson et al. 1998; common toad (Bufo bufo); Sztatecsny et al. 2006) and
has been proposed in sea-turtles (Green turtles; Chelonia mydas; Lee and Hays 2004). The malebiased adult sex ratio in the ESGR population of Painted turtles could result in female
152

harassment, and high levels of multiple paternity, particularly in years when low numbers of
reproductive females reproduce. However, Painted turtles had the lowest incidence of multiple
paternity of the three species, suggesting that if harassment increases multiple matings,
additional mechanisms (e.g. post-copulatory sexual selection) may also exist. Sexual size
dimorphism where males are larger occurs in Snapping turtles and forced copulation has been
suggested as a male mating tactic (Berry and Shine 1980). However, because size ranges of
males and females overlap, the size differences of only a subset of mating pairs would be
sufficiently large enough to make forced copulations probable. Thus if forced copulations is a
mechanism leading to females mating multiply, the highest incidence of multiple matings should
occur in smaller females. In contrast, we observed that multiple paternity occurred most
frequently in the largest females; a pattern suggesting that increased levels of multiple paternity
occur as a function of males seeking copulations with females of high reproductive quality.
Female body size represents one reliable trait that males may use to identify females of
higher reproductive quality because body size was positively correlated with clutch size in all
three species (Tinkle et al. 1981; Congdon et al. 1987; Congdon and van Loben Sels 1991).
However, to date, there is no behavioral evidence for male preference for larger females. Clutch
size was positively associated with incidence of multiple paternity in Blanding's turtles and
Snapping turtles, but was not associated with multiple paternity in Painted turtles (Table 4.5). In
Snapping turtles, body size was a significant predictor of multiple paternity but clutch size was
not, suggesting that there may be higher competition for larger, more fecund, females compared
to smaller females, and that the effect of body size on multiple paternity is not a function of
detection bias.

153

In the four years that genetic data were collected we did not observe an association
between female age and multiple paternity in Blanding's turtles (Table 4.5). However, when
samples from a previous study on the ESGR were included (Osentoski 2002), female age was
significantly and positively associated with incidence of multiple paternity (Chapter 3).
Therefore, female age was associated with incidence of multiple paternity of clutches of all three
species. Because female body size and age is positively related, males may use a reliable trait to
identify older and high quality females. Female reproductive frequency has been shown to
increase with age in all three species (Blanding's turtle, Congdon et al. 2001; Painted turtle,
Congdon et al. 2003; Snapping turtle J.D. Congdon unpublished data). The positive relationship
between the incidence of multiple paternity and female age in all three species suggests that
males can identify body size or some other phenotypic traits to identify high-quality older
females.
Males assess mate quality when investing resources toward pre-copulatory mating tactics
(Bonduriansky 2001; Craig et al. 2002; Zhang et al. 2010). In desert gobies (Chlamydogobius
eremius), males associated with larger (more fecund) females for longer periods than they did
with smaller females, and males increased signaling efforts when presented with a larger females
compared to smaller females (Wong and Svensson 2009). In Tibetan macaques (Macaca
thibetana), males spent more time associating with high-quality females and more males were
attracted to high-quality females than to lower-quality females (Zhang et al. 2010). Although
there is no evidence of social structure or pair bonding in turtles, nor direct behavioral evidence
of preferences, that multiple paternity increases among older and larger females suggests that
males may be able to identify and attempt to mate with high-quality females.

154

Male reproductive success
Multiple paternity was negatively associated with male reproductive success in
Blanding's turtles and was marginally (and negatively) significant in Painted turtles (Table 4.6).
Among successful males, multiple paternity will have an overall negative effect on male
reproductive success because multiple paternity results in sharing of paternity within a clutch.
The average number of offspring sired per clutch was a significant source of variation in male
reproductive success in Painted turtles, but was only marginally significant in Blanding's turtles
(Table 4.6). Male reproductive success was positively associated with the number of clutches a
male sired in Painted and Blanding's turtles (Table 4.6). However, the means by which males
were able to increase the number of clutches differed between Painted and Blanding's turtles.
The number of mates was a significant predictor in Blanding's turtles, but was not significant in
Painted turtles. In contrast, in Painted turtles the vast majority of males sired offspring with only
a single female. Males increased the number of clutches sired without increasing the number of
mates through repeat paternity with the same female within and among years. Blanding's turtle
males, however, increased the number of clutches sired through repeat paternity or by mating
with multiple females.
The time between production of clutches exhibiting repeat paternity was limited only by
the duration of the study. Repeat paternity extending three and more years was common in all
three species (Fig. 4.3) and could have resulted from remating with the same females, or from
the use of stored sperm by females. For Painted turtles, repeat paternity between clutches within
a year is probably the result of stored sperm use due to the short window of receptivity between
clutches (Gist and Congdon 1998) since repeat paternity was observed nearly 100% of the time
155

between clutches within a year (Fig. 4.3). Among years, opportunity for repeat matings
increases, and the likelihood of a female using stored sperm use decreases with the time elapsed
between nesting events. In addition, several studies have shown declines in fertility over time
when females are isolated from males, which could be the result of a decline in sperm viability or
sperm depletion (Goin et al. 1978; Gist and Jones 1987; Palmer et al. 1998). However, during
the present study, a decline in the number of fertilized eggs was not observed in clutches of all
three species. The same males sired all eggs in clutches with 4 and 7 years intervals in Painted
turtles and Blanding's turtles, respectively. No attributes of the female were associated with
repeat paternity in any species, therefore no subsets of females showed a greater propensity to
have the same male sire clutches among years. In combination, the relatively low numbers of
mates of both males and females, the frequency of occurrence of repeat paternity (especially for
Painted turtles), and the extended among-year inter-clutch intervals between repeat paternity
raise questions about the current view of turtle mating systems and behavioral biology that
warrants further explanation. Future work should focus on detailed behavioral observations and
experiments designed to disentangle remating from stored sperm mechanisms of repeat paternity
and attempt to answer the following questions: why aren't some females mating with different
males among years if males are plentiful; and, why aren't male Painted turtles successfully
mating with more than one female? Despite differences in the life history, demography, and
behavior of the three species (Table 4.1) we found commonalities in the attributes associated
with multiple paternity and the determinants of reproductive success. For all three species,
multiple paternity increased with female age and long-durations of repeated paternity were
observed. In Painted and Blanding's turtles, multiple paternity negatively influenced male
reproductive success. Thus males that mated with older females would share paternity more
156

often than males that mate with younger females. Female age was positively associated with
clutch frequency in all three species (Blanding's turtle, Congdon et al. 2001; Painted turtle,
Congdon et al. 2003; Snapping turtle J.D. Congdon unpublished data). In all three species,
increased clutch frequency among older females, combined with the high levels of repeat
paternity, suggests that, despite sharing paternity, the benefits of siring multiple clutches of older
females among years can substantially increase male reproductive success. The number of
clutches that a male sired was positively associated with reproductive success in both Painted
and Blanding's turtles. Although we could not test attributes associated with success in Snapping
turtles, because we observed high levels of repeat paternity among clutches, we predict the
number of clutches would also be a substantial determinant of male reproductive success in
Snapping turtles. Because body size is positively correlated with age in all three species, males
may be able to use size as a reliable indicator of female age. However, in Painted turtles, age but
not body size was a positive predictor of multiple paternity. Thus, males may also rely on
additional cues to identify older females.
Older females represent a valuable resource for populations both in terms of increased
reproductive output as well as for maintaining genetic diversity. Clutches sired by more than one
male can reduce levels of cohort-specific coancestry. Consistent age-specific patterns among the
three species with respect to clutch frequency and multiple paternity demonstrate the importance
of older females to populations. Conservation plans should include ways to maintain high
survivorships of older, high-quality females necessary to maintain stable populations and
maintain genetic diversity.

157

APPENDIX 4

158

Table 4.1. Summary of population characteristics, life-history trait values, morphology, physiology and behavior for Painted turtles
(CP; Chrysemys picta), Blanding's turtles (EB; Emydoidea blandingii), and Snapping turtles (CS; Chelydra serpentina) cohabiting
wetlands on the Edwin S. George Reserve (ESGR). Summary data include mean (minimum-maximum), standard deviation.
Category
Population
Life History

Traits
Adult population size (stable)
Adult sex ratios (M:F)
Age at maturity
Maximum longevity
Male carapace length (mm)
Female carapace length (mm)
Sexual size dimorphism
Relationship - Carapace length
to age
Maximum clutches per year
Reproductive frequency
Clutch size
Relationship to body size
Relationship to age
Egg width (mm)
Relationship to body size
Relationship to age

Behavior

Terrestrial movements

Painted turtle (CP)
550
2.3:1.0
9 (7-13)
~50
119.5 (93-146), 9.65
141.1 (112-169), 9.59
F>M
CL = 125.9 + 0.69 (age)
2
R = 0.33
2/yr
1.3
6.9 (2-13), 0.19
-4.54 + 0.08 (CL)
2
R = 0.23
6.13 + 0.05(age)
2
R = 0.06
17.33, SD = 0.22
18.31 + 0.07 (CL)
2
R = 0.33
16.35 + 0.07 (age)
2
R = 0.32
Primarily nesting
159

Blanding's turtle (EB)
190
1.0:1.0
17 (14-21)
>75
200.2 (161-226),15.24
198.5 (174-219),10.13
F=M
CL = 183.5 + 0.34 (age)
2
R = 0.10
1/yr
0.8
10.1 (2-19), 0.40
-8.58 + 0.08 (CL)
2
R = 0.22
20.33 + 0.02(age)
2
R = 0.04
23.72, SD = 0.16
20.33 + 0.02 (CL)
2
R = 0.04
24.07 + 0.008 (age)
2
R = 0.07
Extensive and frequent
(both sexes)

Snapping turtle (CS)
200
1.0:1.0
13 (11-16)
>55
273.7 (180-367),10.13
245.8 (152-300), 21.6
F<M
CL = 203.5 + 1.43 (age)
2
R = 0.59
1/yr
0.85
27.6 (12-44), 1.05
-14.56 + 0.16 (CL)
2
R = 0.27
21.14 + 0.147 (age)
2
R = 0.04
27.62, SD = 0.29
16.47 + 0.05 (CL)
2
R = 0.37
26.46 + 0.07(age)
2
R = 0.15
Primarily nesting

Table 4.2. Predicted influence of traits on the incidence of multiple and repeat paternity assuming all other traits remain constant.
Where possible, rankings are provided that demonstrate the species (CP, Chrysemys picta; EB, Emydoidea blandingii; CS, Chelydra
serpentina), with the highest/lowest predicted multiple paternity based on the trait. Rationales are provided to describe the underlying
basis for each prediction.

160

Table 4.2 (cont'd)
Multiple Paternity
Predictions

Repeat Paternity
Predictions

CP > (CS = EB)

CP < (CS = EB)

CP > CS > EB

CP < CS < EB

OSR
(fluctuating)

Increase or decrease
for a given year

Increase or decrease for a
given year

Life History
Female age
and size

Higher among older
or larger females

Lower among older or
larger females
1) Variable with individual
reproductive interval.
2) CP > (EB = CS)
3) CP only; high between
clutches within a year

Increased mating pressure on older or larger females may
result in females mating with more males
1) Increased mating pressure on females with high
reproductive frequency may result in more opportunities to
mate.
2) Longer intervals may decrease repeat paternity
3) Short receptivity window necessitates use of stored sperm

Clutch size

1) CS > EB > CP
2) Higher among
females that produce
more eggs per clutch

2) Lowest among females
that produce more eggs per
clutch

1) Clutch sizes increase ability to detect multiple paternity
2) Increased mating pressure on larger females may increase
mating opportunities

Sexual size
dimorphism

CS only:
1) Higher among
smaller females.

1) CS only:
Repeat paternity lowest
among smaller females
2) CS > (EB = CP)

1) Males larger than females; multiple paternity may
increase among smallest females
2) Proposed male-dominance system can result in females
re-encountering the same male

1) EB > (CS = CP)
2) EB < (CS = CP)

1) EB < (CS = CP)
2) EB > (CS = CP)

1) Extended terrestrial movements increases encounters with
non-resident individuals
2) Re-visiting the same habitats can increase encounters
between the same individuals

Trait
Population
Adult
population size
Adult sex ratio
(M:F)

Clutch interval

1) Higher among
females with higher
clutch frequencies

Rationale for predictions
Population size is larger in Painted turtles compared to
Snapping and Blanding's turtles
Male biased sex ratio increases female encounter with males
compared to even sex ratios
The number of reproductive females varies in all species
(Fig. 4. 1). If the number of reproductive males is constant,
the operational sex ratio(OSR) will change

Behavior
Terrestrial
movement

161

Table 4.3. Primary relationship between measures of female qualities and age or size that have implications for male reproductive
success in Painted turtles (CP; Chrysemys picta), Blanding's turtles (EB; Emydoidea blandingii), and Snapping turtles (CS; Chelydra
serpentina).
Relationships
Inter-annual reproductive frequency and age

CP
Positive

EB
Positive

CS
Positive

Intra-annual reproductive frequency and age
Egg width and age
Egg width and carapace length
Clutch size and carapace length

Positive
Positive
Positive
Positive

NA
Positive
No
Positive

NA
Positive
No
Positive

162

Table 4.4. Comparison of among year reproductive intervals of females based on long-term data (>33 years) for Painted turtles (CP),
Blanding's turtles (EB) and Snapping turtles (CS).
Species Reproductive Frequency
CP
Among-year interval
EB
Among-year interval
CS
Among-year interval

Mean
1.47
1.66
1.66

Min-Max
1-4
1-4
1-3

163

SD
0.44
0.55
0.49

Table 4.5. Comparisons of the direction and magnitude of effects of factors influencing multiple paternity in Painted turtles,
Blanding's turtles, and Snapping turtles (2003-2006).

Female age
Clutch size
Female body size

Painted turtles
Beta
Zpestimate value value
0.115
2.01
0.044
-0.046
-0.32 0.752
-0.011
-0.23 0.819

Blanding's turtles
Beta
Zpestimate
value value
0.032
0.98
0.328
0.476
2.51
0.012
0.074
1.90
0.058

164

Snapping turtles
Beta
Zpestimate
value value
0.087
2.67
0.008
0.033
0.78
0.433
0.051
2.62
0.009

Table 4.6. Comparisons of factors influencing male reproductive success (total number of offspring) for successful Painted turtle and
Blanding's turtle males on the ESGR (2003-2006).
Painted turtles

Blanding's turtles

Total number of mates

Beta
estimate
3.47

Total number of clutches

5.82

10.24

<0.001

5.40

8.07

<0.001

Average offspring sired per clutch

1.60

4.25

<0.001

0.91

1.73

0.096

Average multiple paternity

-4.05

-1.90

0.062

-4.58

-2.35

0.027

Male age (average)

-0.29

-1.13

0.272

0.05

0.08

0.934

Male size (average)

0.04

0.87

0.387

0.10

0.08

0.627

t-value

p-value

Beta estimate

t-value

p-value

1.18

0.242

7.95

4.36

<0.001

165

Fig. 4.1. Summaries of the a) number, b) body sizes and c) ages of reproductive female Painted turtles (Chrysemys picta), Blanding's
(Emydoidea blandingii) and Snapping turtles (Chelydra serpentina) cohabiting wetlands on the Edwin S. George Reserve each year
from 1977-2006.
Chrysemys picta

# females

100

Emydoidea blandingii
40

a.

30

70

40

25

80

30

Carapace length (mm)

50

20

20
15

140

10
270
b.
265
260
255
250
245
240
235

204 b.
202
200
198
196
194
192

b.
138
136
134
132

32

50 c.
45

c.

18

30
26

35

12

24

30

14

22

Resident fem
ales

25
1985 1990 1995 2000 2005

Years

c.

28

40

16

a.

50

60

Age (years)

60

a.

35

90

20

Chelydra serpentina

1985

1990

1995

Years
166

2000

2005

20
1980 1985 1990 1995 2000 2005

Years

Fig. 4.2: Conceptual diagram illustrating how population, life-history, morphology and behavior can directly (solid arrows) and
indirectly (dashed arrows) influence male reproductive success.

167

Fig. 4.3. Percent of clutches showing evidence of repeat paternity (and sample sizes; N) within
and among years for Painted turtles, Blanding's turtles, and Snapping turtles.

% of nests

120
100
80
60
40
20
0
100

Painted turtles
N = 40

N = 42

N=7

80

N = 13

N=5

Blanding's turtles
N=5

N = 11

60
40
20

NA

0
100

N = 17

Snapping turtles

80

N = 12

60

N = 10

40
20

NA

0

3+
2
Within
1
Reproductive interval (years)
168

REFERENCES

169

References
Andersson, M. 1994. Sexual Selection. Princeton University Press, Princeton. .
Arnold, S.J., and Duvall, D. 1994. Animal Mating Systems: A synthesis based on selection
theory. The American Naturalist 143(2): 317-348.
Avise, J.C. 2001. Introduction. Journal of Heredity 92(2): 99. doi:10.1093/jhered/92.2.99.
Baker, R.E., and Gillingham, J.C. 1983. An analysis of courtship behavior in Blanding's turtle,
Emydoidea blandingii Herpetologica 39: 166-173.
Bassar, R.D., López-Sepulcre, A., Walsh, M.R., Turcotte, M.M., Torres-Mejia, M., and Reznick,
D.N. 2010. Bridging the gap between ecology and evolution: integrating density
regulation and life-history evolution. Annals of the New York Academy of Sciences
1206(1): 17-34. doi:10.1111/j.1749-6632.2010.05706.x.
Bateman, A.J. 1948. Intra-sexual selection in Drosophila. Heredity 2: 349-368.
Berry, J.F., and Shine, R. 1980. Sexual size dimorphism and sexual selection in turtles (order
testudines). Oecologia 44(2): 185-191. doi:10.1007/bf00572678.
Blanckenhorn, W.U. 2005. Behavioral Causes and Consequences of Sexual Size Dimorphism.
Ethology 111(11): 977-1016. doi:10.1111/j.1439-0310.2005.01147.x.
Bonduriansky, R. 2001. The evolution of male mate choice in insects: a synthesis of ideas and
evidence. Biological Reviews 76(03): 305-339. doi:doi:10.1017/S1464793101005693.
Brown, J.L., Brown, E.R., Sedransk, J., and Shannon, R. 1997. Dominance, Age, and
Reproductive Success in a Complex Society: A Long-Term Study of the Mexican Jay.
The Auk 114(2): 279-286.
Christy, J., and Salmon, M. 1991. Comparative Studies of Reproductive Behavior in Mantis
Shrimps and Fiddler Crabs. American Zoologist 31(2): 329-337.
doi:10.1093/icb/31.2.329.
Clutton-Brock. 1988. Reproductive Success Studies of Individual Variation in Contrasting
Breeding Systems. The University of Chicago Press, Chicago, IL.
Clutton-Brock, and Vincent. 1991. Sexual selection and the potential reproductive rates of males
and females. Nature 351: 58-60.
Clutton-Brock, T.H., Albon, S.D., and Guinness, F.E. 1981. Parental investment in male and
female offspring in polygynous mammals. Nature 289(5797): 487-489.
170

Clutton-Brock, T.H., Guinness, F.E., and Albon, S.D. 1982. Red Deer: Behavior and ecology of
the two sexes. University of Chicago Press, Chicago.
Congdon, J.D., Breitenbach, G., van Loben Sels, R.C., and Tinkle, D. 1987. Reproduction and
nesting ecology of snapping turtles (Chelydra serpentina) in southeastern Michigan.
Herpetologica 43(1): 39-54.
Congdon, J.D., and Gibbons, W. 1987. Morphological Constraint on egg size: a challenge to
optimal egg size theory? Proc. Natl. Acad. Sci. USA 84: 4145-4147.
Congdon, J.D., Graham, T., Herman, T., Lang, J., Pappas, M., and Brecke, B. 2008. Emydoidea
blandingii (Holbrook 1838)- Blanding's Turtle. Chelonian Research Monographs 5.
doi:doi: 10.3854/crm.5.015blandingiiv1.2008.
Congdon, J.D., Kinney, O.M., and Nagle, R.D. 2011. Spatial Ecology and Core Area Protection
of Blanding's Turtle (Emydoidea blandingii). Canadian Journal of Zoology in press.
Congdon, J.D., Nagle, R.D., Kinney, O.M., and van Loben Sels, R.C. 2001. Hypotheses of aging
in a long-lived vertebrate, Blanding's turtle (Emydoidea blandingii). Experimental
Gerontology 36(4-6): 813-827.
Congdon, J.D., Nagle, R.D., Kinney, O.M., van Loben Sels, R.C., Quinter, T., and Tinkle, D.W.
2003. Testing hypotheses of aging in long-lived painted turtles (Chrysemys picta).
Experimental Gerontology 38(7): 765-772.
Congdon, J.D., Tinkle, D.W., Breitenbach, G.L., and Sels, R.C.v.L. 1983. Nesting Ecology and
Hatching Success in the Turtle Emydoidea blandingii. Herpetologica 39(4): 417-429.
Congdon, J.D., and van Loben Sels, R.C. 1991. Growth and body size in Blanding's turtles
(Emydoidea blandingii): relationships to reproduction. Canadian Journal of Zoology 69:
239-245.
Coulson, T., Catchpole, E.A., Albon, S.D., Morgan, B.J.T., Pemberton, J.M., Clutton-Brock,
T.H., Crawley, M.J., and Grenfell, B.T. 2001. Age, Sex, Density, Winter Weather, and
Population Crashes in Soay Sheep. Science 292(5521): 1528-1531.
doi:10.1126/science.292.5521.1528.
Craig, A.S., Herman, L.M., and Pack, A.A. 2002. Male mate choice and male-male competition
coexist in the humpaback whale (Megaptera nonvaeangliae). Canadian Journal of
Zoology 80(4): 745-755.
Crowley, p., Travers, S.E., Linton, M.C., Cohn, S.L., Sih, A., and Sargent, R.C. 1991. Mate
Density, Predation Risk, and the Seasonal Sequence of Mate Choices: A Dynamic Game.
The American Naturalist 137(4): 567-596. doi:Stable URL:
http://www.jstor.org/stable/2462382.
171

Doughty, P. 1996. Statistical Analysis of Natural Experiments in Evolutionary Biology:
Comments on Recent Criticisms of the Use of Comparative Methods to Study
Adaptation. The American Naturalist 148(5): 943-956.
Dreiss, A.N., Cote, J., Richard, M., Federici, P., and Clobert, J. 2010. Age- and sex-specific
response to population density and sex ratio. Behav. Ecol. 21(2): 356-364.
doi:10.1093/beheco/arp198.
Emlen, S., and Oring, L. 1977. Ecology, sexual selection, and the evolution of mating systems.
Science 197(4300): 215-223. doi:10.1126/science.327542.
Ernst, C., and Barbour, R. 1989. Turtles of the World. Smithsonian Institution Press,
Washington, D.C.
Ernst, C., Barbour, R., and Lovich, J. 1994. Turtles of the United States and Canada.
Smithsonian Institution Press, Washington, D.C.
Fairbairn, D.J. 1997. Allometry for sexual size dimorphism:Pattern and Process in the
Coevolution of Body Size in Males and Females. Annual Review of Ecology and
Systematics 28(1): 659-687. doi:doi:10.1146/annurev.ecolsys.28.1.659.
Fiala, K.L., and Congdon, J.D. 1983. Energetic Consequences of Sexual Size Dimorphism in
Nestling Red-Winged Blackbirds. Ecology 64(4): 642-647.
Finkler, M. 2001 Rates of water loss and estimates of survival time under varying humidity in
juvenile Snapping turtles (Chelydra serpentina). Copeia (2): 521-525
Germano, D.J., Bury, R.B., and Jennings, M. 2000. Growth and population structure of
Emydoidea blandingii from western Nebraska. Chelonian Conservation and Biology 3:
618-625.
Gist, D., and Congdon, J.D. 1998. Oviductal sperm storage as a reproductive tactic of turtles.
Journal of Experimental Zoology 282: 526-534.
Gist, D.H., and Jones, J. 1987. Storage of sperm in the reptilian oviduct. Scanning Microscopy 1:
1839-1849.
Goin, C.J., Goin, O.B., and Zug, G.R. 1978. Introduction to herpetology. W.H. Freeman and Co.,
New York.
Graham, T., and Doyle, T.S. 1979. Dimorphism, courtship, eggs, and hatchlings of the
Blanding's turtle, Emydoidea blandingii (Reptilia, Testudines, Emydidae) in
Massachusetts. Journal of Herpetology 13: 125-127.
172

Greene, H.W. 1988. Antipredator mechanisms in reptiles. Wiley, New York.
Hackler, J.C., Van Den Bussche, R.A., and Leslie, D.M. 2007. Characterization of microsatellite
DNA markers for the alligator snapping turtle, Macrochelys temminckii. Molecular
Ecology Notes 7(3): 474-476. doi:10.1111/j.1471-8286.2006.01624.x.
Halliday. 1983. The Study of Mate Choice. Cambridge University Press, Cambridge.
Harding, J. 1997. Amphibians and Reptiles of the Great Lakes Region. University of Michigan
Press.
Ireland, J.S., Broderick, A.C., Glen, F., Godley, B.J., Hays, G.C., Lee, P.L.M., and Skibinski,
D.O.F. 2003. Multiple paternity assessed using microsatellite markers in green turtles
Chelonia mydas (Linnaeus, 1758) of Ascension Island, South Atlantic. Journal of
Experimental Marine Biology and Ecology 291: 149-160.
Jensen, M.P., Abreu-Grobois, F.A., Frydenberg, J., and Loeschcke, V. 2006. Microsatellites
provide insight into contrasting mating patterns in arribada vs. non-arribada olive ridley
sea turtle rookeries. Molecular Ecology 15(9): 2567-2575. doi:10.1111/j.1365294X.2006.02951.x.
Kalinowski, S., Taper, M., and Marshall, T. 2007. Revising how the computer program
CERVUS accommodates genotyping error increases success in paternity assignment.
Molecular Ecology 16: 1099-1006.
King, T.L., and Julian, S.E. 2004. Conservation of microsatellite DNA flanking sequence across
13 Emydid genera assayed with novel bog turtle (Glyptemys muhlenbergii) loci.
Conservation Genetics 5(5): 719-725. doi:10.1007/s10592-004-1854-0.
Klemme, I., Ylönen, H., and Eccard, J. 2007. Reproductive success of male bank voles
(Clethrionomys glareolus): the effect of operational sex ratio and body size. Behavioral
Ecology and Sociobiology 61(12): 1911-1918. doi:10.1007/s00265-007-0431-1.
Kokko, H., and Jennions, M.D. 2008. Parental investment, sexual selection and sex ratios.
Journal of Evolutionary Biology 21(4): 919-948. doi:10.1111/j.1420-9101.2008.01540.x.
Kokko, H., and Mappes, J. 2005. Sexual selection when fertilization is not guaranteed. Evolution
59(9): 1876-1885. doi:10.1111/j.0014-3820.2005.tb01058.x.
Lee, P.L.M., and Hays, G.C. 2004. Polyandry in a marine turtle: Females make the best of a bad
job. Proceedings of the National Academy of Sciences of the United States of America
101(17): 6530-6535. doi:10.1073/pnas.0307982101.
Legler, J.M. 1955. Observations on the sexual behavior of captive turtles. Lloydia 18: 95-99.
173

Libants, S., Kamarainen, A.M., Scribner, K.T., and Congdon, J.D. 2004. Isolation and crossspecies amplification of seven microsatellite loci from Emydoidea blandingii. Molecular
Ecology Notes 4(2): 300-302.
Linden, M., and Møller, A.P. 1989. Cost of reproduction and covariation of life history traits in
birds. Trends in Ecology & Evolution 4(12): 367-371. doi:Doi: 10.1016/01695347(89)90101-8.
Magellan, K., and Magurran, A.E. 2009. The effect of social environment during ontogeny on
life history expression in the guppy Poecilia reticulata. Journal of Fish Biology 74(10):
2329-2337. doi:10.1111/j.1095-8649.2009.02245.x.
Martin, J., Marcos, I., and Lopez, P. 2005. When to come out from your own shell: risk-sensitive
hiding decisions in terrapins. Behavioral Ecology and Sociobiology 57(5): 405-411.
doi:10.1007/s00265-004-0887-1.
McElligott, A.G., and Hayden, T.J. 2000. Lifetime mating success, sexual selection and life
history of fallow bucks (Dama dama). Behavioral Ecology and Sociobiology 48(3): 203210. doi:10.1007/s002650000234.
McGuire, J.M., Congdon, J.D., Scribner, K., and Capps, J. 2011. Variation in Female
Reproductive Quality and Reproductive Success of Male Painted Turtles (Chrysemys
picta marginata). Canadian Journal of Zoology ##(#): ##-##.
McLoughlin, P.D., Boyce, M.S., Coulson, T., and Clutton-Brock, T. 2006. Lifetime reproductive
success and density-dependent, multi-variable resource selection. Proceedings of the
Royal Society B: Biological Sciences 273(1593): 1449-1454.
doi:10.1098/rspb.2006.3486.
Milner, R.N.C., Detto, T., Jennions, M.D., and Backwell, P.R.Y. 2010. Experimental evidence
for a seasonal shift in the strength of a female mating preference. Behav. Ecol. 21(2):
311-316. doi:10.1093/beheco/arp196.
Osentoski. 2002. Unpublished dissertation.
Osentoski, M.F., Mockford, S.M., Wright, J.M., Snyder, M., Herman, T.B., and Hughes, C.R.
2002. Isolation and Characterization of microsatellite loci from Blanding's turtle,
Emydoidea blandingii. Molecular Ecology Notes 2(2): 147-149.
Owens, I.P.F. 2006. Where is behavioural ecology going? Trends in Ecology & Evolution 21(7):
356-361. doi:DOI: 10.1016/j.tree.2006.03.014.
Owens, I.P.F., and Thompson, D.B.A. 1994. Sex Differences, Sex Ratios and Sex Roles.
Proceedings of the Royal Society B: Biological Sciences 258(1352): 93-99.
174

Palmer, K.S., Rostal, D.C., Grumbles, J.S., and Mulvey, M. 1998. Long-Term Sperm Storage in
the Desert Tortoise (Gopherus agassizii). Copeia 1998(3): 702-705.
Pappas, M., Brecke, and Congdon, J.D. 2000. The Blanding‟s turtle of Weaver Dunes,
Minnesota. Chelonian Conservation and Biology 3: 557-568.
Pearse, D., Janzen, F., and Avise, J. 2002. Multiple paternity, sperm storage, and reproductive
success of female and male painted turtles (Chrysemys picta) in nature. Behavioral
Ecology and Sociobiology 51(2): 164-171.
Pearse, D.E., and Avise, J.C. 2001. Turtle Mating Systems: Behavior, Sperm Storage, and
Genetic Paternity. J Hered 92(2): 206-211. doi:10.1093/jhered/92.2.206.
Pearse, D.E., Janzen, F.J., and Avise, J.C. 2001a. Genetic markers substantiate long-term storage
and utilization of sperm by female painted turtles. Heredity 86: 378-384.
Pearse, D.E., Janzen, F.J., and Avise, J.C. 2001b. Genetic markers substantiate long-term storage
and utilization of sperm by female painted turtles. Heredity 86(3): 378-384.
Pitcher, T.E., Neff, B.D., Rodd, F.H., and Rowe, L. 2003. Multiple mating and sequential mate
choice in guppies: females trade up. Proceedings of the Royal Society of London. Series
B: Biological Sciences 270(1524): 1623-1629. doi:10.1098/rspb.2002.2280.
Rankin, J.D., and Kokko, H. 2007. Do males matter? The role of males in population dynamics.
Oikos 116(2): 335-348. doi:10.1111/j.0030-1299.2007.15451.x.
Reznick, D., Nunney, L., and Tessier, A. 2000. Big houses, big cars, superfleas and the costs of
reproduction. Trends in Ecology &amp; Evolution 15(10): 421-425. doi:10.1016/s01695347(00)01941-8.
Rodriguez-Munoz, R., Bretman, A., Slate, J., Walling, C.A., and Tregenza, T. 2010. Natural and
Sexual Selection in a Wild Insect Population. Science 328(5983): 1269-1272.
doi:10.1126/science.1188102.
Rowe, J.W. 1992. Dietary Habits of the Blanding's Turtle (Emydoidea blandingii) in
Northeastern Illinois. Journal of Herpetology 26(1): 111-114.
Shuster, and Wade. 2003. Mating Systems and Strategies. Princeton University Press Princton,
New Jersey.
Sih, A., and Bell, A.M. 2008. Chapter 5 Insights for Behavioral Ecology from Behavioral
Syndromes. In Advances in the Study of Behavior. Edited by Timothy J. Roper Marc
Naguib Katherine E. Wynne-Edwards Chris Barnard H. Jane Brockmann, and C. Mitani
John. Academic Press. pp. 227-281.
175

Smith, C.C., and Fretwell, S.D. 1974. The Optimal Balance between Size and Number of
Offspring. The American Naturalist 108(962): 499-506.
Stearns, S.C. 1992. The Evolution of Life Histories. Oxford University Press.
Sztatecsny, M., Jehle, R., Burke, T., and Hödl, W. 2006. Female polyandry under male
harassment: the case of the common toad (Bufo bufo). Journal of Zoology 270(3): 517522. doi:10.1111/j.1469-7998.2006.00120.x.
Taylor, E.H. 1933. Observations on the courtship of turtles. University of Kansas Science
Bulletin 21: 269-271.
Thornhill, R., and Alcock, J. 1983. Th evolution of insect mating systems. Harvard University
Press, Cambridge Massachusetts.
Tinbergen, N. 1951. The Study of Instinct. Clarendon Press, Oxford.
Tinkle, D.W. 1979. Long-term field studies. Bioscience 29: 717.
Tinkle, D.W., Congdon, J.D., and Rosen, P.E. 1981. Nesting frequency and success: implications
for the demography of painted turtles. Ecology 62: 1426-1432.
Trivers, R.L. 1972. Parental investment and sexual selection pp. 136-179. Aldine Transaction,
Chicago.
Tucker, J.K., Funk, R.S., and Paukstis, G.L. 1978. The adaptive significance of egg morphology
in two turtles (Chrysemys picta and Terrapene carolina). . Bulletin of the Maryland
Herpetological Society 14(1): 10-22.
Uller, t., and Olsson, M. 2008. Multiple paternity in reptiles: patterns and processes. Molecular
Ecology 17(11): 2566-2580.
Watson, P.J., Arnqvist, G., and Stallman, R.R. 1998. Sexual Conflict and the Energetic Costs of
Mating and Mate Choice in Water Striders. The American Naturalist 151(1): 46-58.
Weatherhead, and Boag. 1997. Genetic estimates of annual and lifetime reproductive success in
red-winged blackbirds. Ecology 78: 884-896.
Weatherhead, P.J., Barry, F.E., Brown, G.P., and Forbes, M.R.L. 1995. Sex ratios, mating
behavior and sexual size dimorphism of the northern water snake, Nerodia sipedon.
Behavioral Ecology and Sociobiology 36(5): 301-311. doi:10.1007/bf00167791.
Westneat, D.F., and Fox, C.W. 2010. Evolutionary Behavioral Ecology. Oxford University
Press, Oxford.
176

Wiley, R.H., and Poston, J. 1996. Perspective: Indirect Mate Choice, Competition for Mates, and
Coevolution of the Sexes. Evolution 50(4): 1371-1381.
Wong, B., and Svensson, P. 2009. Strategic male signalling effort in a desert-dwelling fish.
Behavioral Ecology and Sociobiology 63(4): 543-549.
Zamudio, K.R. 1998. The Evolution of Female-Biased Sexual Size Dimorphism: A PopulationLevel Comparative Study in Horned Lizards (Phrynosoma). Evolution 52(6): 1821-1833.
Zeh, J.A., and Zeh, D.W. 1996. The Evolution of Polyandry I: Intragenomic Conflict and
Genetic Incompatibility. Proceedings: Biological Sciences 263(1377): 1711-1717.
Zhang, M., Li, J., Zhu, Y., Wang, X., and Wang, S. 2010. Male mate choice in Tibetan macaques
Macaca thibetana at Mt. Huangshan, China. Current Zoology 56(2): 213 - 221.

177