Diversification of the livebearing Mexican : Pattern and process in macroevolution

Shane A. Webb, Ph.D. Department of Biology

Acknowledgments Mike Ritchie, Jeff Graves & Anne Magurran University of St. Andrews, Scotland

Diarmaid O’Foighil and LMS, UMMZ

Constantino Macias G. & Edgar Avila L., UNAM

ALA, Jim Langhammer, Juan Migual Artigas A., John Lyons

Department of Biology and School of Science & Health Professions, NGCSU Talk coverage Ritchie, M., R.M. Hamill, J.A. Graves, A.E. Magurran, S.A. Webb, and C. Macias-Garcia. 2007. Sex and differentiation; Population genetic divergence and sexual dimorphism in Mexican goodeid fish. J. Evol. Bio. 20(5): 2048-2055.

Ritchie, M.G., Webb, S.A., Graves, J.A., Magurran, A.E. and Macías Garcia, C. 2005. Patterns of speciation in Mexican Goodeid fish: Sexual conflict or adaptive radiation? J. Evol. Bio. 18(4): 922-929.

Webb, S.A., J.A. Graves, C. Macias-Garcia, A.E. Magurran, D. Ó Foighil, & M.G. Ritchie. 2004. Molecular phylogeny of the live-bearing Goodeidae (). Mol. Phyl. Evol. 30: 527-544.

Patterns

• Why are there many species in some places? Are there special places?

• The U.S. has 10-35 fish species per drainage system along the active margin, but these are members of ~12 different fish groups.

• In Mexico the Goodeidae contains ~40 species Process

Or are there special processes? Sexual selection

• Described by Darwin (Origin and Descent) – Male-male competition (intrasexual)  weapons – Female mate choice (intersexual)  ornaments

• Positive feedback (Fisher) • Handicap principle (Hamilton et al.) • Good genes / “bright” male (Hamilton & Zuk) • Sensory exploitation (Ryan) • Etc. Objectives I. Build a molecular phylogeny of the Goodeidae (historical biology based on inference) and compare results with previous work

II. Determine patterns of diversification within this natural group

III. Try to infer the processes important in diversification of this group Introduction to the Goodeidae • Composition and some natural history • Sexual selection and diversification Methods • Datasets and phylogenetic analyses • Calculation of the goodeid molecular clock • Quantifying sexual dimorphism Results and Discussion • Phylogeny • Patterns of divergence and the role of geologic events in diversification • A role for sexual selection? INTRODUCTION

Diversity and natural history Sexual selection and diversification Map

Profundulus Empetrichthyinae

Goodeinae baileyi Empetrichthyinae

Empetrichthys latos Image courtesy of Dr. Paul Loiselle

Pahrump Valley, NV, USA Photo by J. Deacon

MTB Michoacan, Mexico turneri (male) lateralis (male)

Image courtesy of C. Grimes Image courtesy of ALA

Skiffia lermae (male) furcidens (male) viviparity

Girardinichthys viviparus (female) matrotrophy

Lombardi & Wourms 1985 matrotrophy (embryogenesis)

40

35 ↑15,000%

30 other species Ameca 25 ↑ 2,700-39,000%

20

15 Dry weightDry (mg)

10

5 ↑1,000% 0 0 2 4 6 8 10 12 14 16 18 20 Standard length (mm) Lombardi & Wourms 1988 sexual dimorphism

♀

♂

Skiffia multipunctata costly ornaments Presa de Orandino, Michoacan predation by Thamnophis prediction Degree of sexual dimorphism should correlate with diversity Arnqvist et al. 2000; Gavrilets et al. 2001; Martin & Hosken 2003

i.e. Sexually dimorphic groups should be more speciose (possess higher speciation rates)

HO : Species richness is independent of sexual dimorphism. If no correlation, what is the pattern of speciation in the group?

Is most speciation allopatric? Not necessarily exclusive of sexual selection.

What is the rate of speciation? Is speciation clocklike (i.e. relatively constant through time)?

METHODS dataset 1 mtDNA sequence data

• Uniparentally inherited (maternal) • Evolves rapidly (doesn’t recombine) • Protein coding and non-coding regions • Circular dsDNA molecule (~16kb)

• COI (627 bp) and Control region (~400 bp) of 37 taxa DNA protocol • DNA extraction from tissues • PCR  Amplification of desired loci • Fragment isolation • Sequencing reactions • Autosequencing • Proofreading 2 strands aligned sequence data (mtCOI) Allodhub TATTTAGTATTTGGTGCCTGAGCCGGCATAGTTGGTACCGCCCTAA Allodzon ...... T...... G. Alloorob ...... C...... T. Allotcat ...... T...... T...... Allotdug ...... C..T...T... Amspl ...... T...... C...... Ataentow ...... T...... G.....T...... Chapenc ...... T...... C...... Charaud ...... T..T...... Charlat ...... T..T...... C...... Girmult ...... C...... Girviv ...... C...... Gooatr ...... C...... Hubtur ...... T...... T... Ilyfur ...... Skifbil ...... T...... C...... Skiffran ...... G...... C...... Skifmult ...... G...... C...... Xenocapt ...... T...... C...... Xenres ...C...... T...... GT Xenvar ...... T...... C...... Zoogquit ...... T.....G...... 123..3..3..3..3..3..3..3..3..3..3..3..3..3..3. phylogenetic analyses

• Outgroup – close relatives outside group – Profundulus labialis • Ingroup – 36 species of goodeines

• Parsimony analysis (PAUP 4) – Heuristic searches – 20 random stepwise addition replicates

dataset 2 sexual dimorphism • Standard length • Dorsal, caudal, and • Mid body depth anal-fin heights • Body depth at • D-, C- & A-fin areas midpoint of dorsal fin • D- & A-fin bases

Caveat: No measurement of color or behavior in our analyses Mahalanobis distance

• Multivariate • Discriminant function analysis

DM

females males quantifying dimorphism

more dimorphic DM

less

dimorphic DM

females males dataset 3 microsatellite data • 5-7 neutral loci were employed for each of four species that vary in sexual dimorphism – (X. melanosoma, C. lateralis, Z. quitzeoensis, G. atripinnis) • Genotypes were determined for 30 individuals (15males / 15females) of each species from four populations

• FST was calculated and adjusted for geographic distances among populations

Loci from Boto & Doadrio 2003 and Hamill et al. 2007 population 1 FST population 2

a b a a b a a b b b a = 1 a b b

a a b = 0.5 b a a b a b b a pop. subdivision accounts for a ~50% of the total variation b b (due to genetic drift?)

b a a a b a a a = 0 a b b b b b b a RESULTS & DISCUSSION molecular phylogeny

(mtCOI + CR)

dataset 1

Webb et al. 2004 extreme fin dimorphism

riverine clade

hypothesis of Hubbs and Turner 1939

Girardinichthyinae Goodeinae Ilyodon Characodontinae Xenotaenia Ataeniobiinae Alloophorus undescribed Ameca variata Xenoophorus "Xenotoca" Goodea Hubbsina Skiffia Ataeniobius Characodon hypothesis of Smith 1980 Allodontichthys Ilyodon Characodon group Xenotaenia Goodea group Alloophorus Chapalichthys Ameca Xenotoca variata Xenoophorus "Xenotoca" Zoogoneticus Goodea Allotoca Girardinichthys Hubbsina Skiffia Ataeniobius Characodon Empetrichthyinae Profundulus goodeine molecular clock (mtCOI)

• Calibrated with fossil and geological data: – Alloophorus, Ameca, Chapalichthys encaustus, Goodea atripinnis fossils are mid-late Pleistocene (250 kya; Smith et al. 1975) – Allopatric speciation of Girardinichthys (Sierra Madre de las Cruces ~5 Mya; Barbour 1973) – †Tapatia occidentalis (sister group of girardinichthyins) fossils dated as late Miocene (~6 Mya; Smith 1980, Smith & Miller 1986) • Corrected divergence (%)  time of event • Denominator is an underestimate with fossils, but shouldn’t affect overall relationship (slope) 0.18 0.9% per million years 0.16 y = 0.0091x + 0.0482 0.14 R2 = 0.401 †Tapatia 0.12

0.1 Goodea

atripinnis Nei distance Nei - 0.08 Alloophorus

0.06 Girardinichthys Tamura

0.04 Ameca

Chapalichthys 0.02 encaustus

0 0 1 2 3 4 5 6 time (mya) chronology of speciation events

Tamura-Nei sequence divergence (%)

Time (Mya) you are here from Webb et al. 2004 dataset 1 time to speciation

0.3

rank cc = 0.462 )

 p = 0.007 Internode interval 0.2 increasing through time

0.1 Speciation Waitingtime ( not clocklike.

0.0

0 10 20 30 40 Lineages from Ritchie et al. 2005 lineages through time plot dataset 1

runs test rejects linearity Recent p < 0.0001 extinction?

Early

radiation? affected by speciation and extinction

from Ritchie et al. 2005 datasets 1&2 sexual dimorphism and speciation 1.863 1.367, 1 1.785 1.931, 2 1.359, 3 2.273 2.689, 4 2.158 1.965, 5 2.257 2.157 1.902, 6 2.163 2.356 2.117, 7 2.619, 8 2.104 1.886, 9 2.228 2.044, 10 2.412 2.273 2.660, 11 2.537 2.760, 12 2.496, 13 2.182 2.149, 14 2.263 2.331 2.202, 15 2.407 2.581, 16 2.304 2.398 2.252, 17 2.509, 18 2.465 10 2.342 2.615, 19 2.333 2.502, 20 6 2.245, 21 2.357 2.246, 22 12:4 sig. 2.488, 23 2.407 2.468, 24 2.407, 25 p = 0.14 from Ritchie et al. 2005 Does not reject HO dataset 3 FST vs. geographic distance

0.75 FST dependent on dimorphism X. mel p = 0.028 C. lat.

0.50 G. atr.

Z. quit. Fst

0.25

Rejects HO 0.00 0 100 200 300 400 Distance (Km)

from Ritchie et al. 2007 Conclusions • The goodeid “clock” suggests the group is ~16.8 Myrs old, the Goodeinae is 14.9 Myrs old, and that approximately 3/4 of goodeine divergence events occurred during the Miocene (i.e. older).

• All goodeine genera but Xenotoca are monophyletic. Findings of previous authors not supported.

• The biogeographical history of the Goodeinae is difficult to reconstruct. There has been significant range change in older groups. Conclusions

• Comparative analysis failed to find a relationship between sexual dimorphism and rate of speciation, but intraspecific variation suggests a role for sexual selection in genetic differentiation.

• Speciation rate appears to decline over time (are goodeines an adaptive radiation?).

• Role of vicariance (atherinopsids and cyprinids), other extrinsic factors? Agave tequilana Sequential linear regression model

• Distance as a covariate • Mono- vs. dimorphism as a factor (dimorphism) • Species nested within dimorphism

• Fst dependent on dimorphism (F1,18 = 5.7; p=0.028)

• ANCOVA  least squares mean Fst for dimorphic species = 0.25; mono- = 0.16 vicariance plot dataset 1

1

0.9 y = 0.0319x + 0.0636 0.8 R2 = 0.1174 0.7 older groups 0.6 overlap 0.5 0.4 sister species 0.3 degree sympatry allopatric 0.2

0.1

0 0 2 4 6 8 10 12 14 T-N sequence divergence

Method of Barraclough et al. 1998