Sex Role Reversal and Sexual Selection in Seahorses and ...jan.ucc.nau.edu/~bio666-c/Lecture...

1

Sex Role Reversal and Sexual Selection in Seahorses and Pipefish

Stephen M. ShusterNorthern Arizona University

Sex Role Reversal•Males provide parental care.•Females seek multiple mates.

•Sexual dimorphism is

reversed.

2

Extreme Cases are Important for Understanding Sexual Selection

The sex-difference in the Variance in Relative Fitness is reversed,

VWfemales >> VWmales.

Female Male

3

Syngnathid Fish•Are among the

best known of sex role reversed

species.•Variable in degree

of reversal.•Phylogeny is well

understood.

The Goal:To use the degree of sex-role reversal in

Syngnathid fishes to compare Parental Investment Theory

with The Opportunity for Selection

as measures of the sex-difference in the intensity of sexual selection.

Syngnathid Genera of Interest

4

Nerophis(Strongly Role Reversed)

Males brood embryos externally.

Females initiate courtship.

Males are choosy; females are not.

Females inhibit the courtship displays of smaller females.

Females much larger than males, with blue skin folds.

Syngnathus(Moderately Role Reversed)Male brood pouch with sealed folds.

Both sexes active in courtship; females faster than males.

Both sexes choosy about size; M>F.

Females inhibit egg production in smaller females.

Females slightly larger than males, show pigment displays during courtship.

Hippocampus(Conventional Sex Roles)

Male brood pouch internal.

Males compete more to become pregnant than females compete to mate.

Both sexes are choosy about size.

Pairs are faithful; male aggression observed, but rarely.

Sexes are monomorphic.

5

SyngnathidPhylogeny

Conventional sex roles are NOT

ancestral.Mating system

evolution is NOT constrained by

phylogeny.

Two Theoretical Frameworks

Parental Investment Theory(PIT)

Bateman 1948; Williams 1966; Trivers 1972; Emlen & Oring 1977; Maynard Smith 1977; Clutton-Brock & Vincent 1991; Clutton-Brock &

Parker 1992; Ahnesjö et al. 2001.

The Opportunity for Selection(OS)

Crow 1958, 1962; Wade 1979; Wade & Arnold 1980; Arnold &Wade 1983; Wade 1995; Shuster & Wade 1991, 2003.

Parental Investment Theory•The sex difference in initial parental investment causes sexual selection.•Generates three

predictions for sex role reversed

species.

6

Predictions of PIT for Syngnathids

1. Male Investment > Female Investment

{Male – Female} Investment

Nerophis > Syngnathus > Hippocampus


2. Operational Sex Ratio = Strength of Sexual Selection

OSR (= Nfemales/Nmales)



3. Male Availability Limits Female Reproduction: A Sex-Difference in

Potential Reproductive Rate“The maximum number of independent offspring that

parents can produce per unit time (p. 60; Clutton-Brock & Vincent 1991).”

{Female – Male} PPR

7

Measuring PRR

•Allow members of each sex unlimited

opportunities to mate in controlled setting.

•PRR=Xoffspring/unit time.


3. A Sex-Difference in Potential Reproductive Rate

{Female – Male} PPR


Data: A Sex Difference in Initial Parental Investment

Nerophis1. Males suffer greater predation risk (Svennson 1988)

2. Males invest less in gametes than females (Berglund et al. 1986; 1989)

Syngnathus1. Males feed less than females (Svensson 1986).

2. Male and female investment in gametes is about equal (Berglund et al. 1986; 1989).

Hippocampus1. Females have higher energy expenditure than males

(Masonjones & Lewis 2000).

8

1. Male Investment > Female Investment{Male – Female} Investment

Prediction:++Nerophis > +Syngnathus > 0 Hippocampus

Observation:0 Syngnathus > - Nerophis > - - Hippocampus

PIT Prediction vs Observation

Data: The Operational Sex Ratio (OSR)

Nerophis1. Equal in nature (Berglund et al. 1989).

Syngnathus1. Equal in nature (Berglund et al. 1989).

2. Variable, but usually female biased (Vincent et al. 1993).3. Consistently female biased (Berglund & Rosenqvist 1993; Jones et al. 1999).

Hippocampus1. Not biased in mated males, male biased with unmated males

(Vincent 1994; Vincent and Sadler 1995).

2. Operational Sex Ratio = Strength of Sexual Selection, OSR (= Nfemales/Nmales)

Prediction:1++Nerophis > 1+ Syngnathus > 1 Hippocampus

Observation:1+? Syngnathus > 1 Nerophis > -1? Hippocampus


9

Data: A Sex Difference in PRRNerophis

1. Much greater for females, males mate once per cycle, females mate >1 (Berglund et al. 1989; McCoy et al. 2001).

Syngnathus1. Females produce more eggs than males can care for

(Berglund et al. 1988; 1989; 2.7X, Berglund & Rosenqvist; Jones et al. 2000).

Hippocampus1. Identical under natural conditions

(Vincent 1994; Vincent and Sadler 1995).

2. 17% greater for males under laboratory conditions (Masonjones & Lewis 2000).

3. Male Availability Limits Female Reproduction: A Sex-Difference in

Potential Reproductive Rate{Female – Male} PRR

Prediction:Nerophis > Syngnathus > Hippocampus

Observation:++Nerophis = ++Syngnathus > Hippocampus


Summary: PIT #1

• Observed Sex-Difference in

GameticInvestment not

in predicted order.

•Syngnathus has female investment equal to males.•Nerophis females invest more than males.

•Hippocampus females invest much more than males.

10

Summary: PIT #2

• Observed OSRs not

in predicted

order.

•Most sex-reversed, Nerophis, does not have predicted excess of females.

•Unexpected female excess observed in Syngnathus.•Expected male excess in Hippocampus…?

Summary: PIT #3• Observed Sex-Difference in

PRR consistent with predictions but not

consistent with observed dimorphism in Nerophis and

Syngnathus.

•Also, estimates of PRR require study-specific manipulation and

lack confidence limits.

The Opportunity for Selection•The sex difference in the

variance in mate numbers causes sexual selection.

•Provides a direct measure of the intensity of sexual selection.

•Allows phylogenetic comparisons and calculation of

confidence limits.•Provides an empirical estimate of the degree to which the sexes

will diverge.

11

What Is Measured?

I = VW/W2

•Compares the fitness of breeding parents relative to the pre-reproductive

population.•The variance in offspring numbers for

each sex is divided by the squared average in offspring numbers.

Ifemales - Imales = Imateswhen R = 1.

The Sex Difference in the Opportunity for Selection

Ifemales - Imales = (R-1) (Imales) + Imateswhen R ≠ 1.

The Opportunity for Selection

•In sex role reversed species,

VWfemales >> VWmales

•This relationship generates three

predictions.

12

Predictions of OS for Syngnathids

1. A Sex-Difference in the Variance in Offspring numbers

{VOfemales – VOmales}



2. A Sex-Difference in the Opportunity for Selection

{Ifemales – Imales}



3. Imates is Large Relative to Ifemales

{Imates / Ifemales}


13

We Need to Measure:

The variance in offspring numbers for females and for males.

Prediction 1. A Sex-Difference in the Variance in Offspring Numbers

{VOfemales – VOmales}

The Mean and Variance in Offspring Numbers

NOtotal = 25

O♀♀= 5; VO♀♀= 0; N=5.

O♂♂=5; VO♂♂= 0; N=5.

NOtotal = 25

O♀♀= 5; VO♀♀= 0; N=5.

O♂♂=5; VO♂♂= 0; N=5.

The Mean and Variance in Offspring

Numbers, by Sex

14

NOtotal = 25

O♀♀= 5; VO♀♀= 10; N=5.

O♂♂=5; VO♂♂= 0; N=5.

If One Female Mates Twice...

NOtotal = 25

O♀♀= 5; VO♀♀= 20; N=5.

O♂♂=5; VO♂♂= 0; N=5.

If One Female Mates Five

Times...

NOtotal = 25

O♀♀= 5; VO♀♀= 0; N=5.

O♂♂=5; VO♂♂= 20; N=5.

If One Male Mates Five

Times...

15

When some individuals are excluded from mating, the

variance in offspring numbers is increased.

This is the source of sexual selection.

Nerophis ophidion(Berglund et al. 1989)

Measured PRR using 15 females, 27 males

O♂♂=204±82, N=27; VO♂♂=6,724

H♀♀=1.8±0.7, N=15

If R = 1 as in nature,

then pS♀=15/27 = .555

And p0♀ = 1 - pS♀ = .445

The Distribution of Males with Females

pi♀ i i pi♀___________________________________________________________

p0♀ =12/27=.445 0 0

p1♀ = 5/27=.185 1 .185

p2♀ = 8/27=.296 2 .592

p3♀ = 2/27=.074 3 .222

∑ 1.000 1.000

Σ i pi♀ = M♀♀

16

VO♀♀total = VO♀♀within + VO♀♀among

= The average of the variances in offspring number within the classes of mating females

+The variance of the averages in offspring number

among the classes of mating females

The Total Variance in Female Offspring Number

VO♀♀within = Σ pi (iVO♂♂)

= (.445)[(0)(6,724)] + (.185)[(1)(6,724)] + (.296)[(2)(6,724)] + (.074)[(3)(6,724)]

= 0 + 1,234.9 + 3,980.6 + 1,492.7

= 6,717.3

The Variance in Offspring Number Within Females

VO♀♀among = Σ pi (MO♂♂-iO♂♂)2

= (.445)[204-(0)204]2 + (.185)[204-(1)204]2

+ (.296)[204-(2)204]2 + (.074)[204-(3)204]2

= 18,519.1 + 0 + 12,318.3 + 12,318.3

= 43,155.8

The Variance in Offspring Number Among Females

17


= 49,873.1And,

VOfemales – VOmales

= 49,873.1 - 6,724

= 43,149.1

The Sex Difference in the Variance in Offspring Numbers

1. A Sex-Difference in the Variance in Offspring numbers {VOfemales – VOmales}

Prediction:Nerophis > Syngnathus > Hippocampus

Observation:


43,149.1 > 6,275.3 > -18.7(Berglund et al. 1989; Vincent & Sadler 1995)

OS Prediction vs Observation

We Need to Measure:

The opportunity for selection on females and on males.

Prediction 2. A Sex-Difference in the Opportunity for Selection

{Ifemales – Imales}

18

Syngnathus typhle(Berglund et al. 1989)

Measured PRR using 25 females and 43 males

∑ i pi♂=P*♂♂= 3.33±0.64 matings/maleO♀♀/mating=23; VO♀♀/mating=324,

N=43

0

5

10

15

20

25

30

35

0 1 2 3 4 5 6

N Matings

N M

ales

VO♂♂total = VO♂♂within + VO♂♂among

= The average of the variances in offspring number within the classes of mating males


among the classes of mating males

The Total Variance in Male Offspring Numbers

VO♂♂within = Σ pi (iVO♀♀/mating)

= (.023)[(2)(324)] + (.698)[(3)(324)] + (.209)[(4)(324)] + (.070)[(5)(324)]

= 15.1 + 678.1 + 271.3 + 113.0

= 1,077.5

The Variance in Offspring Number Within Males

19

VO♂♂among = Σ pi (O♂♂ - iO♀♀/mating)2

= (.023)[77-(2)(23)]2 + (.698)[77-(3)(23)]2

+ (.209)[77-(4)(23)]2 + (.070)[77-(5)(23)]2

= 21.8 + 40.2 + 49.7 + 102.9

= 214.6

The Variance in Offspring Number Among Males


= 1,292.1 And,

Imales = VO ♂♂ / O♂♂2

= 1,292.1 / (76.6)2

= 0.22

The Opportunity for Selection on Males


Measured PRR using 25 females and 43 males,But R = 1 in nature.

∑ i pi♀=P♀♀= 1 mate/female (=3.33 matings/mate)

O♂♂=76.6, VO♂♂=1,292.4, N=43

0

5

10

15

20

0 0.5 1 1.5 2 2.5 3 3.5 4

N Filled Males

N F

emal

es

20





The Total Variance in Female Offspring Numbers

VO♀♀within = Σ pi (iVO♂♂ ) = (.442)[(0)(1,292)] + (.023)[(.5)(1,292)] + (.093)[(1)(1,292)] + (.209)[(1.5)(1,292)]+ (.070)[(2)(1,292)] + (.116)[(2.5)(1,292)]+ (.023)[(3)(1,292)] + (.023)[(3.5)(1,292)]

= 0 + 15.0 + 120.2 + 405.7 + 180.3+ 375.6 + 90.1 + 150.2

= 1,292.1



= (.442)[77-(0)(77)]2 + (.023)[77-(.5)(77)]2

+ (.093)[77-(1)(77)]2 + (.209)[77-(1.5)(77)]2

+ (.070)[77-(2)(77)]2 + (.116)[77-(2.5)(77)]2

+ (.023)[77-(3)(77)]2 + (.023)[77-(3.5)(77)]2

= 2,592.0 + 34.1 + 0 + 306.9 + 409.3+ 1,534.7 + 545.7 + 852.6

= 6,275.3


21


= 7,567.4 And,

Ifemales = VO♀♀ / O2♀♀

= 7,567.4 / (76.6)2

= 1.29

The Opportunity for Selection on Females



In Syngnathus,

1.29 – 0.22 = 1.07

2. A Sex-Difference in Opportunity for Selection {Ifemales – Imales}

Prediction:

Nerophis > Syngnathus > HippocampusObservation:

Syngnathus > Nerophis > Hippocampus

{1.29-.22}=1.07 > {1.20-.16}= 1.04 > {.64-.65}= -.01(Berglund et al. 1989; Vincent & Sadler 1995)


22

Nerophis ophidion(Berglund et al. 1989)

0

5

10

15

20

25

30

0 1 2 3

N Mates

N M

ales

0

2

4

6

8

10

12

14

0 1 2 3 4

N Mates

N F

emal

es

Only females have multiple mates; with sexual selection,

sexual dimorphism is pronounced.


0

5

10

15

20

0 0.5 1 1.5 2 2.5 3 3.5 4

N Filled Males

N F

emal

es

0

5

10

15

20

25

30

35

0 1 2 3 4 5 6

N Matings

N M

ales

Males and females have multiple mates; with sexual selection on both sexes, sexual dimorphism

is reduced.

We Need to Measure:

The sex difference in the opportunity for selection on females

and on males.

Prediction 3. Imates is Large Relative to Ifemales

{Imates / Ifemales}

23

Hippocampus fuscus(Vincent and Sadler 1995)

51 females, 47 males in nature

∑ i pi♂=P*♂♂=P♂♂= 0.89±0.52matings/male

O♀♀/mating=40.2; VO♀♀/mating=670.8, N=42

0

10

20

30

40

0 1 2 3

N Mates

N M

ales


= The average of the variances in offspring number within the classes of mating males


among the classes of mating males

The Total Variance in Male Offspring Numbers

VO♂♂within = Σ pi (iVO♀♀/mating)

= (.191)[(0)(671)] + (.723)[(1)(671)] + (.085)[(2)(671)]

= 0 + 485.3 + 114.2

= 599.4

The Variance in Offspring Number Within Males

24

VO♂♂among = Σ pi (O♂♂ - iO♀♀/mating)2

= (.191)[40-(0)(40)]2 + (.723)[40 -(1)(40)]2

+ (.085)[40 -(2)(40)]2

= 309.5 + 0 + 137.5

= 447.0

The Variance in Offspring Number Among Males


= 1,046.4 And,

Imales = VO ♂♂ / O♂♂2

= 1,046.4 / (40.2)2

= 0.65

The Opportunity for Selection on Males

Hippocampus fuscus(Vincent and Sadler 1995)

∑ i pi♀= P*♀♀= P♀♀= 0.82±0.51 matings/female

O♂♂/mating=40.2; VO♂♂/mating=270.8, N=42

0

10

20

30

40

0 1 2 3

N Mates

N F

emal

es

25





The Total Variance in Female Offspring Numbers

VO♀♀within = Σ pi (iVO♂♂ )

= (.235)[(0)(671)] + (.706)[(1)(671)] + (.059)[(2)(671)]

= 0 + 473.5 + 78.9

= 552.4



= (.235)[40-(0)(40)]2 + (.706)[40 -(1)(40)]2

+ (.059)[40 -(2)(40)]2

= 380.2 + 0 + 95.1

= 475.3


26


= 1,027.7 And,

Ifemales = VO♀♀ / O2♀♀

= 1,027.7 / (40.2)2

= 0.64

The Opportunity for Selection on Females



In Hippocampus,

0.64 – 0.65 = -0.01

RHfuscus = 51♀♀/47♂♂ = 1.085

However, R > 1

Thus,Ifemales - Imales = (R-1) (Imales) + Imates

AndImates(adj) = 0.067

27

That is, how much of selection on males is due to

sexual selection?

Imates / Ifemales = 0.067 / 0.636

= 0.103

How Much of Ifemales is Imates?

3. Imates is Large Relative to Ifemales {Imates / Ifemales}

Prediction:

Nerophis > Syngnathus > HippocampusObservation:


.87 > .84 > .10(Berglund et al. 1989; Vincent & Sadler 1995)


Summary: OS #1

• Observed Sex-Difference in Variance in

Offspring number consistent with

predictions.

•Relative magnitudes of {VOfemales – VOmales} are consistent with observed sexual dimorphism in all

genera.

28

Summary: OS #2

• Observed {Ifemales-Imales}

not in predicted order.

•BUT, magnitude of Imales is smaller in Nerophis than in Synganthus, in which females and males seek multiple

mates.

Summary: OS #3

• Observed Imates/Ifemalesconsistent with

predictions.

•Relative magnitudes of Imates/Ifemales consistent with observed sexual

dimorphism in all genera.

Conclusion I: OS = PIT?•NO!

•The OS method provides a direct

measure of selection intensity.

•The PIT approach relies on proxies for selection intensity.

29

Conclusion I: OS = PIT?•The sex difference in the opportunity for selection

is both necessary and sufficient to account for

differences in sexual dimorphism.

•The correlation between sex differences in gametic/parental

investment and sexual dimorphism is poor.

Conclusion II: OS > PITFor measuring the intensity of sexual selection within and among populations,the Opportunity for

Selection approach is:More quantitative.Easily interpreted.

Evolutionarily meaningful.

Sexual Selection:Its Not Just for

Males Anymore.

Sex Role Reversal and Sexual Selection in Seahorses and ...jan.ucc.nau.edu/~bio666-c/Lecture...

Documents

Transcript of Sex Role Reversal and Sexual Selection in Seahorses and ...jan.ucc.nau.edu/~bio666-c/Lecture...