Tierärztliche Hochschule Hannover

Institut für Zoologie

Geographical variation in the granular

poison frog, Oophaga granulifera:

genetics, colouration, advertisement call

and morphology

DISSERTATION

zur Erlangung des Grades

eines Doktors der Naturwissenschaften

– Doctor rerum naturalium –

(Dr. rer. nat.)

vorgelegt von

Oscar Brusa

Cuggiono, Italien.

Hannover 2012

Wissenschaftliche Betreuung: Prof. Dr. Heike Pröhl

1. Gutachterin: Prof. Dr. Heike Pröhl

2. Gutachter: Dr. Sebastian Steinfartz

Tag der mündlichen Prüfung: 21 September 2012

Eidesstattliche Erklärung

Hiermit versichere ich, dass ich die vorliegende Arbeit selbständig und nur mit den

angegebenen Hilfsmitteln angefertigt habe. Alle Stellen, die dem Wortlaut oder Sinn

nach anderen Werken entnommen sind, wurden durch Angaben der Quellen als

Entlehnungen kenntlich gemacht.

Ich versichere, dass die vorliegende Dissertation nicht woanders eingereicht wurde.

Hannover, den 14/7/2012

Oscar Brusa

To:

All people who dream.

Table of contents

Zusammenfassung 9

Summary 11

1. Introduction 13

1.1 Biological species concept and speciation mechanisms 13

1.2 Geographic variation in animals 14

1.3 Aposematism 18

1.4 The remarkable case of polymorphic toxic species 19

1.5 The biogeographical peculiarities of the lower Central American 21

1.6 Evolutionary studies on amphibians in lower Central American 23

1.7 Neotropical poison-dart frogs 25

1.8 The granular poison-dart frog, Oophaga granulifera 28

1.9 Aims of the study 29

1.10 Study Area 31

2. Material and Methods 33

2.1 Field activity 33

2.2 Tissue sampling and DNA extraction 33

2.3 Phylogeography 34

2.3.1 Mitochondrial DNA sequencing 34

2.3.2 Mitochondrial DNA analyses 35

2.4 Population genetics 37

2.4.1 Microsatellite loci genotyping and analyses 37

2.5 Colour description and measurements 38

2.6 Bioacoustics 38

2.7 Morphology 39

3. Results 40

3.1 Colouration 40

3.2 Mitochondrial DNA variation and divergence 46

3.3 Mitochondrial DNA phylogeography and population structure 48

3.4. Microsatellite loci variation and population structure 51

3.5 Bioacoustics 53

3.6 Morphology 57

4. Discussion 65

4.1 General discussion 65

4.2 Why should a cryptic morph evolve in an aposematic species? 70

4.2.1 General background 70

4.2.2 Role of genetic drift 71

4.2.3 Role of predation 71

4.2.4 Role of toxicity 73

4.2.5 Role of sexual selection 75

5. Conclusion 76

6. References 77

Additional Information 110

1. Sequences accession numbers 110

2. Population structure ∆K results 113

Ackowledgments 114

9

Geographische Variation in dem granulierten Giftfrosch,

Oophaga granulifera: Genetik, Färbung, Paarungsrufe und

Morphologie.

Oscar Brusa

Zusammenfassung

Die vorliegende Arbeit untersucht die intraspezifische genetische Divergenz und phänotypische

Diversifizierung des neotropischen Giftfrosches, Oophaga granulifera, innerhalb seines

Verbreitungsgebietes. Dabei werden phylogeographische Muster, Populationsstrukturen und

phänotypische Merkmale, die potentiell an reproduktiver Isolation beteiligt sind, analysiert. Die

Analysen zur Phylogeographie and der Populationsstruktur wurden mit Hilfe der

mitochondrialen Gene von Cytochrom b und 16S sowie sieben polymorphen Mikrosatelliten-

Loci von neun bis elf Populationen durchgeführt. Für die Analyse der Phänotypen erfasste ich

den Farbphänotyp in 189 Fröschen und maß die Reflektionsspektren in fünf ausgewählten

Populationen. Zudem wurden in zehn Populationen männliche Paarungsrufe aufgenommen und

zehn morphologische Parameter gemessen. Die phylogenetische Rekonstruktion unterstützte die

Aufspaltung in eine südliche und eine nördliche mitochondriale Linie von O. granulifera. Dieses

Ergebnis wurde durch die Mikrosatelliten-Analyse bestätigt. Weiterhin zeigt die Art

hauptsächlich zwei dorsale Farbvarianten, rot im südlichen und grün im nördlichen

Verbreitungsgebiet. Die Verteilung von Farbmorphen ist asymmetrisch zur mitochondrialen

Variation. Die südliche Linie umfasste nur monomorphe rote Populationen, wohingegen die

nördliche Linie rote, grüne und polymorphe Populationen enthielt. Die polymorphen

10

Populationen kommen in einer Übergangszone zwischen den beiden wichtigsten Farbmorphen

vor. Hier traten neue graduell intermediäre Farbmorphen, von orange-selb bis grün-braun,

zusammen mit roten oder grünen Fröschen auf. Trotzdem folgte der Übergang von rot nach grün

keiner einfachen geographischen Kline. Das Vorkommen von graduell intermediären

Farbmorphen deutet auf signifikante Vermischung entlang der Übergangszone hin. Beide

genetischen Linien unterschieden sich signifikant in ihren Paarungsrufen und ihrer Morphologie.

In der nördlichen Linie unterschied sich die rote Farbmorphe von der grünen Farbmorphe in der

Pulsrate der Paarungsrufe und in der Körpergröße. Die Unterschiede in den Lauten und der

Morphologie beider Farbmorphen liefern Potential für „assortative mating“, wobei zueinander

passende Partner bevorzugt werden. Die mehrfachen Hinweise auf die tiefreichende

intraspezifische Divergenz deuten stark darauf hin, dass es sich bei den beiden identifizierten

Linien um verschiedene Arten handelt. Dem gegenüber lässt sich aufgrund der fehlenden

genetischen Diversifizierung zwischen den Farbmorphen innerhalb der nördlichen Linie und der

genetischen Zusammengehörigkeit der Übergangszone auf eine erst kürzlich aufgetretene

Diversifizierung der Körperfärbung schließen. Ausgehend von einer Süd-Nord

Ausbreitungsbewegung sind unterschiedliche lokale Selektionsdrücke der Schlüssel zur

Interpretation der beachtlichen Farbvariation in diesem aposematischen System.

11

Geographical variation in the granular poison frog, Oophaga

granulifera: genetics, colouration, advertisement call and

morphology.

Oscar Brusa

Summary

The present work investigates the intraspecific genetic divergence and phenotypic

diversification in the neo-tropical poison frog, Oophaga granulifera, across its distribution

range, analysing phylogeographic patterns, population structure and phenotypic traits

potentially involved in reproductive isolation. The phylogeographic and population structure

analyses were performed using the mitochondrial genes cytochrome b and 16S and seven

polymorphic microsatellite loci from nine to eleven populations. For phenotypic analysis, I

observed the colour phenotype in 189 frogs and measured the reflectance spectra in five

targeted populations. Moreover, male advertisement calls were recorded and ten

morphological variables measured in ten populations. The phylogenetic reconstruction

supported the divergence into southern and northern mitochondrial lineages in O. granulifera,

which was confirmed by the microsatellite analyses. The species presents two main dorsal

colour morphs, red in the south of the distribution and green in the north. The distribution of

colour phenotypes is asymmetric as compared to the mitochondrial variation. The southern

lineage included only monomorphic red populations while the northern lineage contained red,

green and polymorphic populations. The polymorphic populations occured in a transition area

between the main morphs in the northern lineage. Here novel gradual intermediate colour

12

variants, varying from orange-yellow to green-brown, occurred with the red or green

phenotypes. However the gradual change from red to green did not follow simple linear

geographic colour cline. The occurrence of these intermediate morphs suggests the presence

of gene flow between the red and green morphs as confirmed by the significant admixture

across the transition area. The two genetic lineages significantly differed in advertisement

calls and morphology. The red and green colour morphs in the northern lineage differed in

pulse rate of the advertisement call and body size revealing potential for assortative mating.

In conclusion, multiple lines of evidence show that O. granulifera presents deep intraspecific

divergence strongly suggesting that the two identified lineages may well represent different

species. On the contrary, the lack of genetic diversification among colour morphs within the

northern lineage and the genetic connectivity across the colour transition area suggest a recent

colour diversification. Originating from a south-north colonization movement, locally

diversified selection is likely to represent the key for interpreting the striking colour variation

in this aposematic system.

13

1. Introduction

1.1 Biological species concept and speciation mechanisms

The scientific definition of species in biology is a complex matter not free from controversy.

The topic has been the object of extensive investigation and reviewing (Ereshefsky, 1992;

Mallet, 1995; Dover, 1995, Mayden, 1997; Burton, 2004; Coyne & Orr, 2004; Hausdorf,

2011). If we refer to sexual organisms, two species are considered distinguished when they

are reproductively isolated (Mayr, 1947, 1963, 1970). The reproductive isolation can be pre-

zygotic (Bailey et al., 2004, Dopman et al., 2010) or post-zygotic (Sánchez-Guillén et al.,

2012). Pre-zygotic mechanisms avoid interbreeding among populations and the consequent

gene exchange through behavioural, physiological or morphological divergence (Schluter,

2001, 2009; Fitzpatrick et al., 2009). If the reproductive isolation persists over time, post-

zygotic isolation characterised by sterility or inviability of hybrids may arise following the

accumulation of genetic differences (Welch, 2004; Orr & Turelli, 2001). Genetic drift and

natural selection play an important role in speciation acting at pre- and post-zygotic levels

(Lande, 1976; Templeton, 1981; Schluter, 2001). Sexual selection is the other key factor able

to drive population differentiation, typically at the pre-zygotic stage (pre and post-mating)

(Noor, 1999; Panhuis et al., 2001; Kirkpatrick & Ravigné, 2002), through a differential

reproductive output among individuals determined by mate choice, mostly by females, that

determines a selective pressure on the opposite sex (Lande, 1981; Panhuis et al., 2001). The

speciation process is also influenced by the geographic distribution of the diverging

populations. Three biogeographic mechanisms of species formation have been identified:

allopatry (Feder et al., 2003; Kirkpatrick & Barton, 2006; Feder & Nosil, 2009), parapatry

(Gavrilets et al., 2000; Turelli et al., 2001; Doebeli & Dieckmann, 2004) and sympatry

14

(Schliewen et al., 2006; Barluenga et al., 2006; Mallet et al., 2009). If different populations

are separated by a large distance as compared to the species vagility or an insurmountable

physical barrier such as a river or a mountain range, that is allopatry by definition. Allopatric

speciation is a classic model and there is convergence among scientists on its importance in

speciation. Parapatric speciation occurs to populations occupying adjacent areas following

behavioural or ecological diversification. Sympatric speciation concerns populations found in

the same area when reproductive isolation arises by natural or sexual selection (White, 1968,

Endler, 1977; Feder et al., 2005; Barluenga et al, 2006). Sympatric models have been

supported by some studies (Schliewen et al., 2006; Barluenga et al., 2006; Mallet et al., 2009)

but the possibility that populations may diverge in the presence of potential gene flow is not

universally accepted (e.g. Fitzpatrick et al., 2008).

1.2 Geographic variation in animals

Intraspecific geographical variation in animals is a widespread phenomenon closely linked to

the speciation process (Mayr, 1942; Gould, 1972; Endler, 1977; Coyne & Orr, 2004). For

instance, populations or individuals occurring in different areas diverge in morphology

(Grant, 1999), colour (Zink & Remsen, 1986) or calls (Wilczynski & Ryan, 1999).

Geographical variation in body size and body shape has been reported in all vertebrate taxa

including, amphibians [e.g. newts (Arntzen & Wallis, 1999); anurans (Castellano & Giacoma,

1998; Schäuble, 2004)], birds [e.g. barnacle goose (Larsson & Forslund, 1991); Galapagos

finches (Grant & Grant, 1989)], mammals [e.g. bats, (Marchan-Rivadeneira et al., 2012);

rodents, (Colangelo et al., 2010)], reptiles [e.g. lizards, (Wiens et al., 2002; Malhotra &

Thorpe, 1997], fish (e.g. Barlow, 1961). Colour variation has been observed as well in several

species [e.g. fish (Kohda et al., 1996); amphibians (Hoffmann & Blouin, 2000); birds (Hill,

1993); reptiles (Rosenblum, 2004); mammals (Hoekstra et al., 2004)] and variation in

15

acoustic signals was found in amphibians (Gerhardt & Huber, 2002) and birds (Slabbekoorn

& Smith, 2002). In anurans, body size is relatively conserved within species (Cherry et al.,

1978; Green, 1993) but may vary along habitat, altitudinal and latitudinal gradients (Berven,

1982; Morrison & Hero, 2003; Rosso et al., 2004). Measuring variation in body size across

historically divergent lineages allows to infer whether population differences are explained by

geographical isolation alone, or whether localized selection contributes to fine-scale

population variation. Colouration and male advertisement calls are often heritable and subject

to selection (Endler, 1992; Ryan et al., 1996; Kingston et al., 2003; Hoekstra et al., 2004;

Hoffman et al., 2006). In polychromatic species individuals of the same sex and age display

colour variants that originate from the interplay between genetic and environmental

components (Roulin, 2004). Colour polymorphism can arise and be maintained by different

evolutionary processes. For instance, a reduction in gene flow among ecologically distinct

populations can lead to phenotypic divergence (Gray & McKinnon, 2006; Rueffler et al.,

2006). Similarly, male advertisement calls can diverge as a result of local adaptation (West-

Eberhard, 1983; Ryan & Wilczynski, 1991; Panhuis et al., 2001; Couldridge & van Staaden,

2004). All these traits have a partial (body size and colour) or exclusive (call) signalling

function which is important in intraspecific communication, being therefore potentially under

sexual selection (Arak, 1983; Forester & Czarnowsky, 1985; Ryan & Rand, 1990; Ryan &

Wilczynski, 1991; Ryan, 1999; Gabor et al., 2000). Body size can have an effect on mating

selection when larger males gain primary access to preferred females through greater

competitive ability (Berven, 1981; Elwood et al., 1987; Harari et al., 1999; Gabor et al., 2000;

Fisher & Cockburn, 2006) or when females selectively mate with larger males likely to have

superior fitness (Wedell & Sandberg, 1995; Harari et al., 1999; Gabor & Page, 2003). On the

other hand, males can also mate selectively with larger females because of fecundity

advantages (Verrell, 1985; Olsson, 1993; Harari et al., 1999; Dosen & Montgomerie, 2005;

Preston et al., 2005). Mate selection based on colour has received a great deal of attention for

16

the conspicuous traits that can arise in males through the selective process. For instance, the

blue structural colour of crown feathers (reflecting also in the ultraviolet range) has a direct

fitness effect for male blue tits because the females prefer to mate with those having a higher

reflectance (Andersson et al., 1998). Carotenoid-based colour traits of males are also selected

by females, for example in birds (Hill, 1990; MacDougall & Montgomerie, 2003) and fish

(Endler, 1983; Seehausen & van Alphen, 1998; Van Oppen et al., 1998; Barluenga et al.,

2006). Studies have shown that these colour traits are a cue for foraging ability, immune

system efficiency and overall body condition, characterising an honest signal. This depends

on the fact that carotenoid pigments cannot be synthesised endogenously and are a limited

resource in the environment. Therefore, a bright carotenoid-based colouration can only be

obtained diverting these pigments from their immune function according to a trade-off

mechanism (Blount et al., 2003; Baeta et al., 2008). Overall, the phenotypic variation

observed across the distribution of an animal species may depend on different factors that can

be summarised into three main components: genetic drift (Wright, 1948), natural (Hoekstra et

al., 2004) and sexual selection (Tilley et al., 1990; Ryan et al., 1996). These factors can act in

combination (Kirkpatrick & Ravigne, 2002). Genetic drift was introduced by Wright (1948)

and represents the stochastic genetic variation in a population. As it can be intuitively

understood, genetic drift is conceptually the opposite of selection, because the latter

modulates the expression of a certain trait according to its fitness outcome. Despite a

controversial evaluation of the actual importance of drift in the phenotypic evolution of

animal populations (Coyne, 1992; Gavrilets, 1997) it is well established that in some

circumstances, when significant numerical reductions of populations occur, such as in islands’

colonization or after a bottleneck, random fluctuations in allele frequencies are able to

determine evolutionary effects (Mayr, 1963; Carson & Templeton, 1984). It seems likely that

a diversification process within a species that eventually leads to speciation may often depend

on an interaction between selection and genetic drift (Mayr, 1963; Carson, 1975; Templeton,

17

1980). For instance, local genetic drift may determine intraspecific divergence in temporarily

isolated populations (Wright, 1948; Avise, 2000) but prezygotic isolation mechanisms, such

as mate choice, are necessary to impede the possible hybridisation between the diverging

populations if gene flow is re-established (Lande, 1982; Panhuis et al., 2001). The

intraspecific divergence in selected traits and its importance for speciation is well documented

(Butlin, 1987; Endler & Houde, 1995; Boughmann, 2002). Nonetheless, the vast majority of

studies have analysed these traits independently while comprehensive integrative analyses of

phenotypic geographical variation and genetic differentiation are lacking. The importance of

geographical variation in advertisement calls and in visual signals and its role in reproductive

isolation has been shown in anuran amphibians (Blair, 1958; Littlejohn et al., 1968; Gerhardt

& Huber, 2002; Hoskin et al., 2005), but very few studies investigated the relationships

among clines in visual signals, acoustic signals and genetic population structure (Pröhl et al.

2006; Amezquita et al., 2009). It is important to understand that the quantification of

geographical genetic and phenotypic diversity among populations represents a valuable

approach to get insights into the historical roles of drift and selection and evaluate the

potential for population diversification and speciation (Foster & Endler, 1999; Panhuis et al.,

2001; Gray & McKinnon, 2006; Hoffman et al., 2006). As mentioned, phenotypic

differentiation may lead to intra-specific divergence when populations are geographically

isolated or under diversifying selection (Dobzhansky, 1940; Thorpe et al., 1995; Hoekstra et

al., 2005; Pröhl et al., 2006). Therefore, it is necessary to analyse the geographical variation

considering historical and recent factors to understand the relationship between phenotypic

differentiation and genetic divergence, (Lougheed et al., 2006). Despite a possible

discordance in terms of direction and rate of evolution (Marroig & Cheverud, 2001; Wiens &

Penkrot, 2002), when analysed conjunctly genetic and phenotypic traits can reveal patterns of

covariation providing information on the relative importance of gene flow and local selection

in shaping spatial and temporal diversification (Lu & Bernatchez, 1999; McKay & Latta,

18

2002; Thorpe & Stenson, 2003; Gübitz et al., 2005; Nosil & Crespi, 2006). Given the

numerous pathways for signal divergence, it is not surprising that differentiation in calls is

often observed at the population level (Ryan et al., 1996; Foster & Endler, 1999) as well as

among species (Blair, 1958; Sanderson et al., 1992; Roberts & Wardell-Johnson, 1995;

Parsons & Shaw, 2001; Smith et al., 2003). At both scales, behavioural and morphological

characters can retain evolutionary and phylogenetic signals therefore providing a valuable

integrative information in phylogeographic studies (de Queiroz & Wimberger, 1993; Foster &

Endler, 1999; Wiens & Penkrot, 2002; Price, et al., 2007).

1.3 Aposematism

Aposematic animals have conspicuous warning colour patterns informing predators that the

individual is toxic, unpalatable or otherwise unprofitable. The occurrence of species using

toxic or distasteful substances as a defence against predators is a known phenomenon,

observed for example in insects (Frazer & Rothschild, 1960) molluscs (Faulkner, 1992), frogs

(Daly, 1987) and snakes (Smith, 1977). These chemical defences are often associated with

bright colour making the bearer of such signals easily recognizable by potential predators

(Wallace, 1867; Poulton, 1890; Ruxton et al. 2004). The association of these two traits

constitutes the base for the evolutionary theory known as aposematism. The related scenario

is that a potentially costly trait (bright colour) evolved to maximise the effect of another

potentially costly trait (toxicity). Distastefulness and toxicity other than with a bright

colouration may also be coupled with other distinctive colour patterns, odours, sounds and

behaviours originating an overall conspicuous strategy aimed at maximising the efficacy of

the unprofitability signal to predators (Poulton, 1890; Cott, 1940; Edmunds, 1974). Both

theory and empirical evidences suggest that an effective warning signal should be easy to

detect, learn, recall and associate with the defence (Benson, 1972; Gittleman & Harvey, 1980;

19

Roper & Redstone, 1987; Guilford, 1990; Alatalo & Mappes, 1996; Gamberale-Stille &

Tullberg, 1999; Lindstedt et al., 2011). A key point of the aposematic theory (Müller, 1879) is

that populations of aposematic species should converge into one conspicuous phenotype

because in this way the predator will easily learn to avoid that specific colour, making fewer

recognition errors (e.g., Guilford, 1986; Mallet & Barton, 1989; Joron & Mallet, 1998;

Gamberale-Stille, 2000; Kapan 2001; Beatty et al., 2004; Rowland et al., 2007; but see Rowe

et al., 2004; Ihalainen et al., 2007). Accordingly, selection is expected to favour the most

conspicuous and most common colour morph and decrease variation in warning coloration.

On the contrary, the presence of polymorphic populations is considered evolutionary not

optimal because it would imply a longer learning time by the predator with consequent

potential detrimental effect on the aposematic species individual fitness. Being the bright

colour associated with toxicity it represents a signal that the bearer cannot manipulate. This is

the reason why aposematic colouration has been explained within the theoretical framework

of honest signalling (Blount et al., 2009). The described scenario applies if we consider the

evolution of bright colour successive to the evolution of toxicity. On the other hand, one

cannot rule out the possibility that the bright colour evolved first, independently by toxicity.

In this case, the bright colour would have made the animal more detectable and therefore

more exposed to predation attempts, making the phenotype counter-selected via natural

selection. Subsequently, to compensate for the higher predation pressure caused by a

conspicuous appearance, the species evolved a system of chemical defence so that to originate

a deterrent effect to the predator and re-establish an evolutionary optimal strategy.

1.4 The remarkable case of polymorphic toxic species

One of the most striking examples of divergence in visual signals in animals is colour

polymorphism, defined as the co-occurrence of different colour morphs in a population

20

(Müller, 1879). Colour polymorphism belong to a broader phenomenon definable as colour

diversity or polychromatism, that is the occurrence of intraspecific colour variation, at the

interpopulation but not necessarily at the intrapopulation level. As previously mentioned

colour traits may depend on structural characteristics of integuments or pigments or a

combination of the two (Auber, 1957; Sieffermann & Hill, 2003; Endler, 1983). There are two

major classes of pigments responsible for colouration in animals: melanins and carotenoids

(Griffith et al., 2006). Melanins are synthesised endogenously and are responsible for both

black (across taxa) and rufous (mammals and birds) colour background and patterning in the

variants of eumelanin and pheomelanin respectively (Roulin, 2004). On the contrary, animals

are not able to synthesise carotenoids de-novo and therefore depend on the diet for displaying

a colouration based on these pigments. Animals can produce red, brown, orange and yellow

pigmentation rearranging enzymatically the many molecular variants of carotenoids (over 600

split into two classes, xantophylles and carotenes). Colour diversity depends mainly on

pigments variation and is a relatively common phenomenon in vertebrates, observed for both

melanins and, less frequently, carotenoids (Roulin, 2004; Hoffman & Blouin, 2000). It is also

worth to mention a less common class of endogenously synthesised pigments called pteridines

that can originate a colour range overlapping with those dependent on carotenoids. Pteridines

have been found for example in wings of butterflies (Rutowski et al., 2005; Bezzerides et al.,

2007), skin of fish (Grether et al., 2001), amphibians (Obika & Bagnara, 1964) and reptiles

(Steffen & McGraw, 2006) and irises of birds (Hill & McGraw, 2006). Pteridines can be

synthetized in the body therefore they are not considered components of honest signalling

systems. A biochemical analysis able to discriminate between carotenoids and pteridines is

therefore necessary for a correct inference concerning the evolutionary ecology of

pigmentation. Notably, colour polymorphism has been observed in association with toxicity

challenging the Müllerian prediction of phenotypic convergence in aposematic species. In

fact, aposematic animals are usually chromatically monomorphic (Edmunds, 1974;

21

Greenwood et al., 1981). Nonetheless, colour polymorphism exists in aposematic insects such

as lepidopterans [(Brakefield & Liebert, 1985; Stimson & Berman, 1990 including Heliconius

butterflies (Mallet, 1986)] ladybirds (Hodek & Honěk, 1996; Povolný, 1999) and true bugs

(Heteroptera; Štys, 2004). Among vertebrates dendrobatid frogs provide the most significant

example, showing a striking array of colour morphs, (Daly & Myers, 1967; Summers &

Clough, 2001; Summers et al., 2003; Darst & Cummings, 2006). This unexpected pattern is

known, for example, in the strawberry poison frog [Oophaga pumilio (Schmidt, 1857)] whose

colour diversity spans the visual spectrum including even cryptic variants (Summers et al.,

2003; Pröhl & Ostrowski, 2011). The presence of cryptic morphs is surprising because

predators are thought to learn to avoid unpalatable prey faster if they are conspicuous rather

than cryptic (Gittleman & Harvey, 1980; Roper & Redston, 1987; Alatalo & Mappes, 1996;

Lindström et al., 1999a) and if they are clearly contrasted against the average background

(Roper & Redston, 1987; Lindström, et al., 1999a; Gamberale-Stille, 2001; Prudic et al.,

2007). Several factors have been taken into consideration to explain the occurrence of colour

polymorphism in aposematic species. These include: genetic drift (Mallet & Singer, 1987;

Turner & Mallet, 1996; Mallet & Joron, 1999, Rudh et al., 2007), sexual selection (Summers

et al., 1999, Maan & Seehausen, 2011), variation in toxin availability (Wang, 2011) and

variation in predation regime (Mallet & Joron, 1999; Endler & Mappes, 2004; Nosil et al.,

2005; Ratcliffe & Nydam, 2008).

1.5 The biogeographical peculiarities of the lower Central American region

The numerous studies carried out on Central America anuran amphibians have provided

insights into the great level of variation typical of this biogeographic region (Ryan et al.,

1996; Crawford, et al. 2007; Robertson et al., 2009). Due to its complex geomorphologic and

climatic history, Central America presents high biological diversity at the inter- and

22

intraspecific levels with the latter resulting in the evolution of different genetic lineages in

several species (Coates & Obando, 1996; Campbell, 1999; Kohlmann et al., 2002; Savage,

2002). Particularly, the history of Costa Rica and Panama has been profoundly influenced by

largescale landscape changes (Campbell, 1999; Kohlmann et al., 2002; Savage, 2002). A

fundamental geomorphologic trait of the area is the Cordillera de Talamanca, a mountain

chain extending across lower Central American that originated approximately 3 million years

ago dividing the area in Pacific and Atlantic slopes. This structure markedly altered the

geomorphology and climate of the Caribbean, Central and Pacific areas (Holdridge, 1947;

Campbell & Frost, 1993; Coates & Obando, 1996), producing a sharp geoclimatic

differentiation. This diversity likely contributed to the exceptionally high species richness and

endemicity characteristic of this area (Holdridge, 1947; Savage, 2002). The Cordillera de

Talamanca plaid a role in the diversification patterns across many Central American taxa

[terrestrial snakes (Zamudio & Greene, 1997; García-París et al. 2000); beetles (Kohlmann

et al. 2007); frogs (Crawford, 2003; Robertson & Robertson, 2008). In Costa Rica, the

southwestern region presents high level of endemism following the uplift of the Cordillera

and the consequent isolation from northwestern CR and Caribbean lowland forests

(Holdridge, 1947; Savage, 2002). Another biologically highly diverse area is the Osa

Peninsula, that may have derived from an offshore island drifted into mainland Costa Rica

approximately 2 million years ago (Kohlmann et al., 2002). Genetic isolation of the Osa

Peninsula populations from mainland Costa Rica has been documented in frogs (Crawford,

2003; Crawford et al., 2007; Robertson, 2008) and snakes (Zamudio & Greene, 1997). These

studies suggest that the Osa Peninsula has retained a signature of isolation, despite

reconnection to the mainland. In Panama, the Caribbean region of Bocas del Toro is a

significant biogeographic area for animal and plant taxa, including frogs (Crawford, 2003;

Weigt et al., 2005; Crawford et al., 2007; Hagemann & Pröhl, 2007).

23

1.6 Evolutionary studies on amphibians in the lower Central American

region

Studies conducted on amphibians of Costa Rica and Panama returned a complex picture of the

interplay between genetic and phenotypic diversification. For instance, populations of a

species may be phenotypically highly diversified despite a lack of concordant genetic

divergence as identifiable using nuclear and mitochondrial markers (Hagemann & Pröhl,

2007; Hauswaldt et al., 2011). On the contrary, populations rather similar in terms of

phenotype may be divergent when analysed at the molecular level (Hauswaldt et al., 2011;

present study). Therefore, a compelling aim of modern evolutionary studies is to clarify the

patterns of association between the variation in traits such as body size, colour, call and

historical and recent population genetic structure. This together with information on

ecological variation would allow to get insights into the possible evolutionary origin of the

stunning diversity of the area. Despite the need to consider all the relevant traits to have a

global description of the possible interrelations between genetic diversity and phenotype few

studies have really adopted a comprehensive approach. In the case of Neotropical anuran

amphibians, recent investigations have focussed on reconstructing the phylogeographic

patterns (Wang & Shaffer, 2008; Brown et al., 2010; Vences et al., 2003), some incorporated

data on recent population structure (Wang & Summers, 2010; Hauswaldt et al., 2011) and few

related those genetic traits to biological important components such as advertisement calls and

morphology (e.g. Pröhl et al., 2006; Amezquita et al., 2009). A clear example of the lack of

such a comprehensive approach is represented by dendrobatid frogs, that present a remarkable

colour variation occurring in some species, for instance in the Panamanian populations of the

strawberry poison frog Oophaga pumilio (Maan & Cummings, 2012). Several studies have

dealt with dendrobatids behaviour (Pröhl, 1997; Summers et al., 1999; Pröhl & Hödl, 1999;

Pröhl, 2003, 2005; Pröhl et al., 2007; Richards-Zawacki & Cummings, 2011), aposematic

defense evolution [using theoretical (Ruxton et al., 2004, 2007; Lee et al., 2011), empirical

24

(Maan & Cummings, 2008; Darst et al., 2006; Noonan & Comeault, 2009) and phylogenetic

approaches (Santos et al., 2003; Brown et al., 2010; Wang, 2011)] and the phylogenetic

relations within the Oophaga (former Dendrobates) genus (Hagemann & Pröhl, 2007;

Hauswaldt et al., 2011). Nonetheless, only a few studies have considered a key trait, for its

importance in intraspecific divergence, that is the male advertisement call (e.g. Pröhl et al.,

2007), that has been shown to play a role in mate choice at the population level in O. pumilio

(Pröhl et al., 2003) and in other Neotropical amphibians (Pröhl et al., 2006). It is also

remarkable that other traits potentially important in sexual selection such as morphology have

been quite neglected to date. Moreover, significant levels of genetic divergence were found

among populations of Costa Rica and Panama leading to the definition of different lineages in

some dendrobatid species (Symula et al., 2003; Noonan & Wray, 2006; Hagemann & Pröhl,

2007; Hauswaldt et al., 2011) as well as other amphibians (Ryan et al., 1996; Crawford et al.,

2007; Robertson et al., 2009). The complex paleogeographic history of the Neotropics

accounts for this high level of diversity (Lougheed et al., 1999; Graham et al., 2004; Elmer et

al., 2007; Roberts et al., 2007; Santos et al., 2009). The bright colour of many dendrobatids is

commonly associated with toxicity and explained according to the aposematism theory

(Summers & Clough, 2001; Darst & Cummings, 2006). Nonetheless, colour diversity among

populations and dull colour morphs exist (Savage, 1968). An instructive case is the strawberry

poison frog, Oophaga pumilio (Schmidt, 1857), that is monomorphic in Costa Rica but shows

a high number of colour morphs in Panama, in the Bocas del Toro archipelago (Daly &

Myers, 1967; Summers et al., 2003). A phylogeographic study found a southern and a

northern lineage with the former including all the polymorphic populations together with the

southern Costa Rica monomorphic populations (Hagemann & Pröhl, 2007). The presence of

two main lineages in O. pumilio has been confirmed by the bioacoustic divergence in the male

advertisement call (Pröhl et al., 2007). The importance of advertisement calls in anuran

reproductive isolation has been shown in several studies (Littlejohn et al., 1968; Gerhardt &

25

Huber, 2002; Hoskin et al., 2005). Nonetheless, there is a general lack of studies combining

phylogeographic reconstruction with an analysis of advertisement call variation in poison

frogs. Investigating the correlation between genetic distance and divergence in bioacoustic

traits is useful to evaluate the patterns of reproductive isolation and speciation (e.g. Pröhl et

al., 2006). In fact, a low level of genetic differentiation between colour phenotypes has been

observed (Hagemann & Pröhl, 2007; Wang & Shaffer, 2008; Brown et al, 2010; Hauswaldt

et al., 2011; but see Wang & Summers, 2010). In this scenario sexual selection is likely to

have played a key role in the fixation of the colour peculiarities of these populations, with

females preferring males of their local insular colour morph (Summers et al., 1999; Reynolds

& Fitzpatrick, 2007; Maan & Cummings, 2008; Richard-Zawacki & Cummings, 2010).

Despite the obvious phenotypic divergence, there is no clear evidence of genetic

differentiation between colour morphs in this species (Pröhl et al., 2007; Hauswaldt et al.,

2011 but see Wang & Summers, 2010). Moreover, in dendrobatid species, studies

investigating the correlation between the geographical variation in colour and other traits

potentially involved in reproductive isolation are scant. Colour variation in poison frogs has

been related to several factors such as local variation in the predator communities, toxin

availability and assortative mating based on colour and call traits (Summers et al., 1997,

1999; Speed & Ruxton., 2004; Darts & Cummings, 2006; Saporito et al., 2007a; Speed &

Ruxton, 2007; Maan & Cummings, 2008, 2012). Within this framework, it is particularly

important to investigate the genetic and phenotypic variation across the species distribution to

have a clear picture of population-level diverging patterns.

1.7 Neotropical poison-dart frogs

During the last three decades the scientific knowledge on dendrobatid frogs, commonly

known as poison-dart frogs, has increased significantly. The number of described species

26

exponentially grew over the years and 247 are presently recognized. Dendrobatids are found

in dense primary and secondary forests, open areas, mountains and lowlands, and occupy a

variety of aquatic, terrestrial and arboreal habitats, in Central and South America. They are

distributed from Nicaragua to Bolivia and the Atlantic coast of Brazil and from the Pacific

coast of South America to the French Antilles (Savage, 2002). Approximately one-third of the

known species of dendrobatids have toxins in special dermal glands and are able to release

these toxins through the skin (Savage, 1968; Silverstone, 1975). The name poison-dart frogs

derives from the tradition of the native tribe of the Emberá, inhabiting the Chocó region of

western Colombia, to poison their hunting blowgun darts using the skin secretion of three

dendrobatid species, Phyllobates terribilis (Myers et al., 1978), P. aurotaenia (Boulenger,

1913), and P. bicolor (Duméril & Bibron, 1841). These three species contain in their skin two

of the most powerful toxins known, batrachotoxin and homobatrachotoxin (Myers et al.,

1978). Over 450 different lipophilic alkaloids have been isolated from dendrobatids and 24

major structural classes were defined (Daly et al., 1999). Many of these compounds found

application in medical research (Daly et al., 1997, 2000; Daly, 1998). It is known that

dendrobatids are not able to synthesise their toxins de-novo but they take them from the diet.

However, the details on the sequestration mechanism are still unknown (Daly et al., 1994a).

The dietary sources of alkaloids identified so far are formicine ants, a siphonotid millipede,

melyrid beetles and scheloribatid mites (Daly et al., 1987; Saporito et al., 2003; Dumbacher et

al., 2004; Saporito et al., 2004; Takada et al., 2005; Saporito et al., 2007a), while the

remaining alkaloids are still unknown elsewhere in nature. The vast majority of dendrobatids’

alkaloids are lypophilic molecules but one hydrophilic alkaloid, called tetrodotoxin, has been

detected in Colostethus inguinalis (Cope, 1868) (Daly et al., 1994b). In addition to toxicity,

Neotropical poison-arrow frogs are characterised by many other peculiar traits (Daly &

Myers, 1967; Weygoldt, 1987). Dendrobatids, differently from the vast majority of

amphibian species, are diurnal (except Aromobates nocturnus, critically endangered and

27

poorly known species found in north-west Venezuela), they present bright and dull colours,

conspicuous and cryptic behaviour, territoriality with aggressive defence observed in both

males and females and complex mating rituals (e.g., Silverstone, 1973; Wells, 1978, 1980a,

1980b, 1980c; Weygoldt, 1987; Zimmermann & Zimmermann, 1988; Summers, 1989;

Aichinger, 1991; Caldwell, 1997; Fandino et al., 1997; Juncá, 1998; Caldwell & de Oliveira,

1999; Summers et al., 1997; Summers et al., 1999; Siddiqi et al., 2004; Pröhl, 2005; Maan &

Cummings, 2008; Lüddecke, 2000; Bourne et al., 2001; Pröhl & Berke, 2001; Pröhl, 2003;

Narins et al., 2003, 2005; Summers & McKeon, 2004). They lay terrestrial eggs, either on the

ground or in water-filled leaf axils of phytotelmata plants (from the ancient greek phyto-,

meaning 'plant', and telma, meaning 'pond'). Many species are characterized by elaborate

reproductive behaviours, biparental care, eggs moistening, egg guarding, tadpoles

transportation on the dorsum of parent frogs and tadpole feeding with unfertilized oocytes by

the mother (genus Oophaga) and one or both sexes are involved in highly evolved parental

care (Weygoldt, 1980; Zimmermann & Zimmermann 1988; Pröhl, 2005). In many species,

the male advertises a reproductive territory by emitting a species-specific call and defending it

actively from conspecific male intruders (Pröhl, 2005). Despite a broad knowledge on the

different aspects of their biology, the phylogenetic relationships among dendrobatids at the

inter- and intraspecific level are not yet fully understood. Phylogenetic relationships have

been used to investigate the evolutionary origins of dendrobatid features (e.g., Summers et al.,

1999; Santos et al., 2003; Vences et al., 2003; Graham et al., 2004; Darst et al., 2005). The

growing awareness of the importance of well resolved phylogeny for dendrobatids has

recently brought to significant systematic revisions (e.g. Noonan & Gaucher, 2006;

Wollenberg et al., 2006; Grant et al., 2006; Santos et al., 2009).

28

1.8 The granular poison-dart frog, Oophaga granulifera

The granular poison frog, O. granulifera, shows a very limited distribution in south-central

pacific Costa Rica (Fig. 1). It inhabits the lowland forests of the pacific slope of Cordillera de

Talamanca in Costa Rica and possibly north Panama. It is endangered because of habitat

fragmentation and included in the IUCN Red List as a vulnerable species (IUCN Red List). In

the southern part of its distribution the species is characterised by a red dorsal patch and a pale

blue to grey colour on the remaining part of the body. The proportion of body area covered

with red increases from the south to the centre of the distribution. Three colour morphs were

known before the present study: a red colour morph in the south and central part of the

distribution, a green colour morph in the north and a yellow colour morph located in between

(Savage, 2002; Wang, 2011). The species was classified in the histrionica group (Myers et al.,

1984) together with other 8 species found in Costa Rica, Panama and Colombia [O. arborea

(Myers et al., 1984), O. histrionica (Berthold, 1845), O. lehmanni (Myers & Daly, 1976), O.

occultator (Myers & Daly, 1976), O. pumilio (Schmidt, 1857), O. speciosa (Schmidt, 1857),

O. sylvatica (Funkhouser, 1956) and O. vicentei (Jungfer et al., 1996). Currently the same

species are grouped in the Oophaga genus, a name that derives from the unusual feeding of

the tadpoles on unfertilized eggs provided by the mother (Grant et al., 2006). O. granulifera

(Taylor, 1958) shares several behavioural traits with the other Oophaga species, such as male

territorial behaviour, courtship behaviour and predominantly female parental care including

tadpole transportation and tadpole feeding (Goodmann, 1971; Crump, 1972; van

Wijngaarden, 1985; Bolaños, 1990; van Wijngarden & Bolaños, 1992; van Wijngarden &

Gool, 1994) other than tadpole morphological traits (Grant et al., 2006). However, it is rather

divergent from congeneric species in terms of bioacoustic properties of the male

advertisement call, particularly the length of the acoustic unit (call) within a call sequence

(Meyer, 1992, 1993). In addition, phylogenetic assessments concerning Oophaga spp. and

29

including a few samples of O. granulifera, found indications of a high genetic divergence of

the latter within the Oophaga genus (Clough & Summers, 2000; Hagemann & Pröhl, 2007).

Figure 1 Sampling localities of Oophaga granulifera. The populations from Wang (2011) are

marked with *. Population codes as in Table 1.

1.9 Aims of the study

In the work here presented, I investigated the phylogeography, the population structure and

the geographical variation in colour, advertisement call and morphology in the granular

poison frog, Oophaga granulifera (Taylor, 1958). I aimed at evaluating the relationships

between historical and recent genetic structure, colour diversity and geographical variation in

30

traits which are potentially involved in reproductive isolation such as male advertisement call

and morphology. In particular, I focused on populations located in a very narrow colour

transition area between the two main colour morphs red and green, where intermorph crossing

was expected and intermediate colour phenotypes were more likely to occur. This assumption

was strengthened by the description of an intermediate yellow phenotype (Savage, 2002;

Wang 2011). I aimed at clarifying which of two possible scenarios may describe O.

granulifera colour diversification: the presence of intermediate gradual phenotypes as a result

of gene flow among the main morphs or, alternatively, the occurrence of the three discrete

colour morphs in reproductive isolation. I analysed the colour variation across the transition

area using field observations and spectrometric measurements and estimate the

interpopulation gene flow using microsatellite loci. In addition, I studied the geographical

variation in male advertisement call and morphology as potential sexual signals involved in

mate choice, to investigate how they relate to genetic and colour variation. Recently, Wang

(2011) identified two monophyletic clades in the species, one of southern red populations and

one of yellow and green populations nested within a northern clade which also contains a red

population. The author considered nine populations: four southern red, one central red, two

northern intermediate yellow and two northern green (Fig. 1). I analysed the phylogeography

of O. granulifera including three newly found polymorphic populations occurring in the

colour transition area. I also considered seven other populations: one northern green, three

northern red located within 20 km from the transition area and three red from the Osa

peninsula in the south (Table 1; Fig. 1). This population sampling allowed a detailed

evaluation of the historical genetic divergence across the colour transition area. Moreover, I

analysed the male advertisement calls and the morphological characteristics of the different

populations as key factors in mating recognition and consequently in reproductive isolation, to

unravel the pattern of interactions between morphological differentiation and speciation.

Summarising, the present study applied an integrative approach based on phylogeographic

31

analysis, bioacoustic and morphological analyses, detailed colour phenotype observations and

spectral reflectance measurements. Here, I provide the most comprehensive reconstruction of

O. granulifera phylogeography to date, including populations containing previously

undescribed colour morphs. I clarified the existence of two deep genetic lineages within the

species supported also by the differentiation in the male advertisement calls and morphology.

1.10 Study area

The study was carried out in south-western Costa Rica sampling in the lowland forest along

the pacific coast. The study area extends from the Osa Peninsula in the south to the central

part of the Puntarenas province between the towns of Quepos and Parrita and is rather

diversified in terms of climate and vegetation cover. The Osa peninsula is the most pristine

area of Costa Rica, (hosting for example the Corcovado national park, a world known

biodiversity hotspot) characterized by an extensive forest coverage and a low urbanization.

The area between the towns of Palmar and Quepos is characterised by a high human impact,

deforestation and consequent habitat fragmentation. The surroundings of Quepos also host

extended oil palm plantations. On the study area, many rivers are present including the Rio

Térraba (Fig. 1), the largest of Costa Rica that divides the town of Palmar into two parts,

named Palmar Norte and Palmar Sur.

32

Table 1 Information for each study population: name, code, geographical location and

coordinates, number of individuals sequenced for 16S and cyt b and genotyped for

microsatellites, number of individuals captured for colour description, number of male

advertisement calls recorded and frogs measured for morphological analysis.

Population Code CR region GPS Cyt b 16S Msats Colour Call Morphology

Ventana VE Osa Peninsula N8°42.457–W83°29,589

2 – 7 16 9 –

Charcos CH Osa Peninsula N8°40.319–W83°30.320

3 3 8 19 8 19

Neotropica NE Osa Peninsula N8°42.082–W83°30.783

– – 8 24 15 15

Palmar PA South-West N8°58.112–W83°27.356

3 3 7 14 4 14

Firestone FS Central-West N9°16.668–W83°52.167

3 2 8 15 5 15

Dominical DO Central-West N9°15.065–W83°51.292

3 2 8 15 – –

Baru BA Central-West N9°16.416–W83°52.464

3 1 8 14 5 13

Matapalo MA Central-West N9°20.913–W83°52.167

4 1 10 20 3 20

Portalon PL Central-West N9°20.951–W83°56.277

3 2 15 19 4 19

Savegre SA Central-West N9°22.039–W83°58.277

1 1 8 20 6 19

Damitas DA North-West N9°32.124–W84°12.222

3 1 11 13 – 13

33

2. Material and Methods

2.1 Field activity

The data used in the present study were collected during three field seasons from 2008 to

2010 for a total of seven months. The work was carried out between April and July in the first

half of the rainy season when the frogs calling was intense. According to the frogs activity

field work was performed in most cases from 5:00 am in the morning to 1:00 pm when

temperature increased and the calling activity became weak. Data collection was also

performed during minor activity time of Oophaga granulifera in late afternoon until sunset. In

case of precedent very abundant rainfall or slight rain during the day, frogs remain active

throughout the day enabling continuous data collection.

2.2 Tissue sampling and DNA extraction

Tissues samples used for genetic analyses were obtained in the field by toe-clipping and

stored in 96% ethanol in 1.5 ml plastic tubes with screw cap. Toe-clipping is the removal of

one or more toes and is a common practice in amphibian studies (Hero, 1989; Waichman,

1992; Halliday, 1996). Other than for obtaining tissue for genetic analyses it is used to allow

individual recognition in behavioural studies removing several toes in specific combinations

and to obtain bone for estimating age using skeletochronology (Friedl & Klump, 1997;

Driscoll, 1998; McGuigan et al., 1998). Some studies revealed possible deleterious effects of

removing several toes in amphibians (Clarke, 1972; Humphries, 1979; Golay & Durrer, 1994;

Reaser & Dexter, 1996) while others did not (Lemckert, 1996; Williamson & Bull, 1996;

Parris & McCarthy, 2001). Nonetheless, it is recognized that the removal of one toe is likely

not to provoke any negative effect on survival (McCarthy & Parris, 2004). In the present

34

study only a portion of one toe (2-3 mm) was removed and no adverse effects were observed

after release. The sample was stored for some weeks in the 1.5 ml screw cap plastic tubes

filled with 96% ethanol. The preservation of tissue for DNA extraction in ethanol for a short

time is appropriate because ethanol inhibits the proteolytic enzymes responsible for tissue

degradation. Once in the laboratory, the samples were stored in the fridge for few days until

DNA extraction. The DNA was extracted in the genetic laboratory of the Institute of Zoology

at the TiHO (co-responsible Prof. Dr. Heike Pröhl) in Hannover and at the Department of

Zoology of Cambridge University (evolutionary genetics group, responsible Dr. Nicholas

Mundy) using a Qiagen DNeasy extraction kits and a Qiagen QIAquick purification kit

respectively (Qiagen, Hilden, Germany) following the manufacturer's instructions. The

general procedure involved digestion of the tissue in 1.5 ml tubes containing a proteinase K

solution and incubation at 37°C overnight. The proteinase K releases the DNA from the tissue

and inhibits the nucleases enzymes responsible for nucleic acid degradation. Then a binding

buffer was added and the samples were passed through filtering mini-columns embedded in

1.5 ml centrifuge tubes. After centrifugation, a washing buffer was added to the columns to

eliminate the impurities from the filter that binds the DNA and the flow-through discarded. As

a final step the DNA was eluted by adding an elution buffer in two successive aliquots each

followed by centrifugation. The eluate was stored at -20° C. Small known aliquots of purified

DNA mixed with a loading dye were run on 1% agarose gel stained with ethidium bromide to

quantify the DNA yield and prepare proper dilution for subsequent amplification steps.

2.3 Phylogeography

2.3.1 Mitochondrial DNA sequencing

Two fragments of mitochondrial DNA were considered for the phylogeographic analysis. I

amplified a 600 bp fragment of the cytochrome b gene (cyt b) from 28 individuals (10

35

populations) using primers MTAL (Hauswaldt et al., 2011) and cyt b-c (Bossuyt &

Milinkovitch, 2000). A portion (527 bp) of the 16S rRNA gene (16S) was amplified in a

subset of samples (N = 16; 9 population) using universal primers 16SA and 16SB (Palumbi et

al., 1991) (Table 1). PCRs were performed in 25 µl volume using Taq DNA-polymerase

(Eppendorf, Hamburg, Germany). The thermocycling conditions for cyt b comprised an initial

denaturation at 95°C for 2 min followed by 35 cycles of denaturation (95°C for 45s),

annealing (52°C for 45s), elongation (72°C for 1min), and a final elongation at 72°C for 10

min. The profile for 16S was identical except for an annealing temperature of 55°C. PCR

products were purified using Qiagen purification kits and filtering columns system described

above. The purified PCR product was then precipitated and prepared for sequencing

according to standard procedure. The fragments were sequenced in both directions in 10 µl

volume using BigDye 3.1. The sequences were run on an ABI sequencer 3500 (Applied

Biosystems, Darmstadt, Germany).

2.3.2 Mitochondrial DNA analyses

Sequences were aligned and edited with MEGA 4.1 (Tamura et al., 2007). Haplotypes were

extracted using the on–line tool “DNA to haplotype collapser and converter” available in

FABOX 1.40 (http://www.birc.au.dk/software/fabox/). I calculated the number of variable

sites as well as uncorrected pairwise distances between populations (p-distance) using MEGA

4.1. I performed Maximum Likelihood and Bayesian phylogenetic analyses with PAUP* 4.0

(Swofford, 2002) and MrBayes 3.1 (Ronquist & Huelsenbeck, 2003), respectively. All the cyt

b sequences of comparable size published for O. granulifera were included in the analyses

(Wang, 2011; see Additional Information for all accession numbers). The populations

considered in Wang (2011) were Corcovado (CO), Drake Bay (DB), Piedras Blancas (PB),

Rio Terribe (TE), Rio Baru (RB), Cerro Nara (CN), Rio Savegre (SV), Rio Damitas (DM) and

Fila Chonta (FC). The geographical areas of Costa Rica where the populations were found

36

and the correspondent colour phenotypes described in Wang (2011) were: south-west, red

(CO, DB, PB), central-south-west, red (TE), central-west, red (BR), central-north, yellow

(CN, SV), central-north, green (DM, FC) (Fig. 1). Hereafter, these populations will be cited in

the text using the abbreviations. I am not aware of any sampled population in common with

Wang (2011) (Fig. 1). As outgroups, I included eight 16S sequences from all but one of the

other species of the genus Oophaga (nomenclatural authorities) [O. arborea (Myers et al.,

1984), O. histrionica (Berthold, 1845), O. lehmanni (Myers & Daly, 1976), O. pumilio, O.

speciosa (Schmidt, 1857), O. sylvatica (Funkhouser, 1956) and O. vicentei (Jungfer et al.,

1996); two sequences for O. pumilio northern and southern lineages, one sequence for each

other species; O. occultator (Myers & Daly, 1976) was not included because there were no

sequences available], two 16S sequences of Dendrobates leucomelas and D. tinctorius and

two cyt b sequences of O. histrionica and O. sylvatica, (see Additional Information for all

accession numbers). For ML analyses, I selected the model of evolution that best fitted the

data using jMODELTEST 0.1 (Posada, 2008). Analyses were carried out using a heuristic

search with 10 random additions per replicate and TBR branch swapping. Node confidence

was assessed using 1000 bootstrap replicates. Bayesian posterior probabilities were calculated

using MrBayes 3.1 (MCMC sampling) applying the substitution model selected in

MrModeltest 3.7. Two markov chains were run from random trees for 4 × 106 generations and

monitored to ensure that the average standard deviation of split frequencies was <0.01. The

20% of the initial trees were discarded as burn-in. Finally, I reconstructed the cyt b haplotype

network using TCS 1.21 (Clement et al., 2000) and determined the minimum number of

mutational steps required to connect different clusters of haplotypes using the fix connection

limit option. I included the same published cyt b sequences (Wang, 2011) considered in the

phylogenetic analyses.

37

2.4 Population genetics

2.4.1 Microsatellite loci genotyping and analyses

Seven microsatellite loci [B9, C3, D4, E3, F1, G5, O1] previously isolated for O. pumilio

(Hauswaldt et al., 2009) were amplified via PCR from seven to 15 individuals belonging to

eleven populations. PCRs were performed in 25 µl volume using Taq DNA-polymerase

(Eppendorf, Hamburg, Germany). The thermocycling conditions comprised an initial

denaturation step at 95°C for 30 sec followed by 35 cycles of denaturation (95°C for 30 sec),

annealing [48°C (C3); 57°C (B9, D4, F1); 59°C (E3); 60°C (G5); 64°C (O1) for 45 sec] and

elongation (72°C for 45 sec) followed by a final elongation at 72°C for 10 min. Then

genotyping (electrophoresis) was performed on a ABI 3500 sequencer (GE Healthcare, Little

Chalfont, UK) using ladder GeneRuler 100bp Plus (Fermentas, Vilnius, Lithuania) and

fragments were scored with GeneMapper 4.1 software (Applied Biosystem, Foster City,

California, USA). I used GenAlEx 6.4 (Peakall & Smouse, 2006) to calculate observed and

expected heterozigosities (Ho and He) and the number of private alleles (AP). Average allelic

richness (AR), for at least seven individuals per population was calculated with FSTAT

(Goudet, 1995). I tested each locus and population for linkage disequilibrium and deviation

from Hardy-Weinberg equilibrium (HWE) with Genepop on the Web (Fisher’s exact tests,

Raymond & Rousset, 1995a,b; Rousset, 2008) using the Markov chain Monte Carlo (MCMC)

and default parameters (10,000 dememorizations, 1000 batches and 10,000 iterations per

batch). FST pair-wise values (Weir & Cockerham, 1984) and the molecular variance

(AMOVA; Excoffier et al., 1992) were computed with Arlequin 3.11 (Excoffier et al., 2005).

Significance was tested with 10,000 permutations. Finally, I performed a population

assignment test based on Bayesian posterior probabilities as implemented in Structure 2.3.3

(Pritchard et al., 2000). Runs were performed with 1×106 burn-in period and MCMC of

3×106, setting an admixture model (αinitial = 1.0; αmax = 10.0). I tested one to 13 possible

clusters (K), and 10 runs were conducted for each K. The uppermost hierarchical level of

38

population structure was determined according to Evanno et al. (2005), checking the greatest

rate of change in estimated likelihood between successive K-values.

2.5 Colour description and measurements

For colour description, 189 frogs were photographed with a digital camera Nikon Coolpix

4500 and colour phenotype characteristics recorded from 11 populations distributed across the

known distribution range of O. granulifera (Table 1; Fig. 1). Light reflectance measurements

were taken from 15 individuals each from populations Firestone, Dominical, Matapalo,

Portalon and Savegre using an Ocean Optics 2000+ (Ocean Optics Inc., Dunedin, FL, USA)

fiber optic spectrometer and a deuterium-tungsten DT-Mini-2-GS light source. Reflectance

spectra were obtained averaging five measurements each from the head, upper dorsum and

lower dorsum. Geographical coordinates and localities altitudes were documented with a

Garmin GPS 12 (Garmin Ltd., Schaffhausen, Switzerland).

2.6 Bioacoustics

Male advertisement calls were recorded using a Sennheiser directional microphone (MHK

416 P48; Sennheiser, Wedemark-Wennebostel, Germany) and a digital walkman Marantz

(PMD 671, Marantz Corporation, Kanagawa, Japan). The calls of 59 frogs from nine

populations were recorded (Table 1). Due to very low calling activity in the Dominical and

Damitas populations, the calls of frogs from these populations could not be recorded. In most

populations the frogs were found in close proximity to creeks, therefore many of the recorded

calls contained background noise, leading to low sample size in some populations. During

recording the distance to the frogs was no more than two metres to minimize environmental

noise interference. The calls were recorded mostly in the morning hours (5.00 to 13.00) when

the calling activity was more intense. After each recording, air temperature was measured to

39

allow analyses to be controlled for temperature effect on the calls. Pulse rate, call duration and

call rate were calculated using oscillograms. Dominant frequency was measured by using

spectrograms. For details on call parameters definition see Pröhl (2003). Bioacoustic analyses

were performed using Avisoft (Avisoft Bioacoustics, Berlin, Germany). I used one-way

ANOVA to test for differences in call parameters among populations and colour morphs.

2.7 Morphology

Morphological measurements were collected from 147 frogs belonging to nine populations

(Table 1). Body weight was measured using a digital scale to the nearest 0.01g. Body (snout-

vent), femor, tibia, humerus, radio-ulna and head lengths, head width, tympanum diameter

and between-nostril distance were measured with a digital calliper to the nearest 0.01mm. I

collected all measurements personally to avoid inter-operator variation. I used one-way

ANOVA to test for differences in morphological parameters among populations and colour

morphs.

40

3. Results

3.1 Colouration

I present the results on coloration first because they are important for interpreting the genetic

results. I found that the red colour phenotype gradually changed from the south (Charcos,

Neotropica, Ventana) to the north of the species distribution. I observed an increase in the size

of red patch over the body (Palmar, Firestone, Dominical, Baru) and then a change to an olive

green phenotype. In the transition area, populations were polymorphic and the two main

colour morphs (red and green) co-occurred with an array of intermediate phenotypes (Fig. 2,

3). The intermediate phenotypes were found together with the red morph in Matapalo and

with the green morph in Portalon and Savegre. The polymorphic populations occupy a

remarkably small area of approximately five Km while the whole sampled area was

approximately 110 km long. The southern (red) Charcos, Neotropica, and Ventana were all

located in the Osa peninsula and showed a similar colour pattern with a bright red dorsal

patch beginning at the head and covering 2/3 of the dorsum. The remaining body parts were

covered by a pale blue colour (Fig. 3a). In these three populations 57%, 33%, 55% of frogs

respectively had small red patches on the arms. The colour of the ventrum and limbs varied

from dark to pale blue with black reticulation often present on the dorsal surface and flanks,

but rarely on the ventrum. In Palmar, the frogs were almost completely red on the dorsum and

also usually had extended red patches on the arms and red spots on the proximal portion of the

legs (Fig. 3b). The red phenotype in Firestone was mainly dark red (75%) with less orange

individuals (25%), in Dominical the frogs were mainly orange (70%) while a minority was

red (30%) (Fig. 3c,d,e). In Baru I observed a uniform dark red colour. In Firestone and Baru,

54% and 33% of frogs had red spots on the ventral surface. In Dominical, in all frogs the

entire trunk was red or orange, including in some cases hands and feet and most of the

41

ventrum (33% of frogs). The rest of the ventrum and legs was azure (Firestone, Baru) to azure

grey (Dominical). In the populations Matapalo, Portalon and Savegre we found previously

undescribed colour phenotypes. In Matapalo, brightly red frogs (three out of 20) were found

to coexist with uniformly orange-green individuals (17 out of 20). The intermediate frogs in

this population showed different degrees of saturation, being either darker or lighter (Fig. 3

f,g). In Portalon, I observed highly variable intermediate colour phenotypes (50% of the

frogs) ranging from green with brown patches to red-brown and orange similar to those

identified in Matapalo. Also frogs with three concentric areas of different colour (brown,

yellow and green) were observed. The other half of the observed individuals displayed the

green phenotype (Fig. 3l). In Savegre, more uniform orange to gold-yellow intermediate

phenotypes (25%) were found. In this population, the green morph was most common (75%

of frogs) (Fig. 3i). In Damitas I found mostly green frogs, although four out of 13 individuals

showed a brown patch on the head smaller than the patches observed in Matapalo, Portalon

and Savegre. The ventrum and leg regions ranged from azure to green–azure (Fig. 2l). A

northern green population, not included in the present study, was located near the city of

Parrita outside the known distribution of O. granulifera. Moreover, the altitudal range

occupied by the species is wider than reported (20-100m asl; Savage, 2002; IUCN Red List of

Threatened Species), with populations (Baru, Matapalo, Portalon, Savegre) extending up to

400m asl. The reflectance spectra confirmed the colour intergradation among Firestone and

Dominical. The spectra from Matapalo, Portalon and Savegre showed the presence of a

gradual variation from green to orange (Fig. 4). The colour variation in O. granulifera can be

broadly categorised in the following chromatic areas: red (red), orange (intermediate), orange-

green (intermediate), green-yellow-brown (intermediate), green-brown (intermediate) and

green (green) (Table 2).

42

Table 2 Grouping of O. granulifera colour phenotypes into six categories. Symbols describe

the colour variation occurring in the populations as follow: Red (■), Orange (□), Orange-

green (◙), Green-yellow-brown (◘), Green-brown (○), Green (●).

Population Colour

Charcos ■

Neotropica ■

Ventana ■

Palmar ■

Firestone ■ □

Baru ■

Dominical ■ □

Matapalo ■ ◙

Portalon ◙ ◘ ○ ●

Savegre ◙ ◘ ○ ●

Damitas ○ ●

43

Figure 2 Sampling localities of Oophaga granulifera. The pie-graphs indicate the proportion

of red, intermediate and green phenotypes in each of the populations sampled for the present

study. The populations from Wang (2011) are marked with *. Population codes as in Table 1.

44

Figure 3 Dorsal colour variation in Oophaga granulifera. (a) Osa peninsula [■], (b) Palmar

[■], (c) Firestone [■], (d) Firestone [□] , (e) Dominical [□], (f,g) Matapalo [◙] , (h)

Portalon [◙] , (i) Savegre [○], (l) Damitas [●]. In [ ] colour symbols as in Table 2.

Photographs by Oscar Brusa

45

Figure 4 Average reflectance spectra of O. granulifera from populations (a) Firestone, (b)

Dominical, (c) Matapalo, (d) Portalon, (e) Savegre.

46

3.2 Mitochondrial DNA variation and divergence

There were 17 variable sites and five haplotypes identified in 16S (N = 16) and 44 variable

sites and twelve haplotypes identified in cyt b (N = 30). Both markers identified a southern

(Osa peninsula) and a northern (all remaining populations) genetic lineages. For 16S, the

average genetic distance between the population Charcos (Osa peninsula) and all the other

populations was 1.9% (Table 3). The distances for 16S among the other seven Oophaga spp.

included in the analyses varied between 0.8% and 3.2% with a mean of 1.9%. The average

distance between O. granulifera and the other Oophaga spp. was 5.5%, while the average

distance between Oophaga spp. and the species D. leucomelas and D. tinctorius was 9.2%

and 11.5% respectively (Table 3). Interpopulation genetic distances (p-distance) for cyt b

varied from 0.0 to 4.3% (Table 4). The average genetic difference between southern and

northern populations for the same marker was 4%, while the divergence among the northern

populations varied from 0.1% to 0.8% (Table 4). In the northern lineage, the average genetic

distance between Palmar and the other populations (0.8%) was higher than among the latter

(0.2%). From the AMOVA, most of the variation was among lineages (91.65%, Φct = 0.92, P

< 0.001), whereas variation within populations (3.75%, Φst = 0.97, P < 0.001) and within

southern and northern lineages (4.60%, Φsc = 0.55, P < 0.001) were much lower.

47

Table 3 16S sequence divergence (uncorrected p-distances) between the two lineages of Oophaga granulifera,

seven other Oophaga spp. and two Dendrobates spp. “O.gran. south” indicates the O. granulifera southern

lineage; “O.gran. n” indicates the O. granulifera northern lineage.

Table 4 Cyt b sequence divergence (uncorrected p-distances) among the investigated populations of Oophaga

granulifera. Populations abbreviated as in Table 1.

Species D.leu D.tin O.arb O.his O.leh O.pum O.spe O.syl O.vic O.g. S

D.leucomelas

D.tinctorius 0.062

O.arborea 0.085 0.106

O.histrionica 0.094 0.115 0.008

O.lehmanni 0.092 0.110 0.014 0.014

O.pumilio 0.096 0.119 0.020 0.028 0.030

O.speciosa 0.092 0.108 0.014 0.018 0.020 0.026

O.sylvatica 0.092 0.110 0.012 0.016 0.014 0.032 0.022

O.vicentei 0.096 0.110 0.016 0.018 0.020 0.024 0.018 0.024

O.gran. S 0.109 0.133 0.048 0.054 0.049 0.065 0.059 0.050 0.055

O.gran. N 0.106 0.124 0.049 0.053 0.053 0.070 0.064 0.051 0.059 0.019

Population CH VE PA FS DO BA MA PL SA

Charcos

Ventana 0.004

Palmar 0.036 0.035

Firestone 0.040 0.041 0.007

Dominical 0.041 0.042 0.008 0.001

Baru 0.042 0.043 0.009 0.004 0.005

Matapalo 0.040 0.041 0.007 0.001 0.002 0.004

Portalon 0.042 0.042 0.009 0.002 0.001 0.006 0.003

Savegre 0.041 0.042 0.008 0.001 0.000 0.005 0.002 0.001

Damitas 0.041 0.042 0.008 0.001 0.000 0.005 0.002 0.001 0.000

48

3.3 Mitochondrial DNA phylogeography and population structure

Maximum-likelihood (ML) and Bayesian analyses revealed a deep genetic separation of the

southern populations, Charcos and Ventana, from all the other populations for both

mitochondrial markers (Fig. 5). The phylogeny reconstructed with 16S sequences indicated a

clear differentiation between the two lineages of southern and northern O. granulifera with

100% node support for both ML and Bayesian trees. I also found a higher level of divergence

of the species compared with the other Oophaga spp. (Fig. 5a). Similarly, in the cyt b tree the

node separating the Osa peninsula from the other populations was highly supported by

bootstrap and posterior probability values (ML = 98, BI = 1.00, Fig. 5b). The cyt b tree also

showed moderate support for a derived monophyletic clade including Firestone, Dominical,

Baru, Matapalo, Portalon, Savegre and Damitas populations. Within this clade, the divergence

of two haplotypes from Baru and two from BR was supported by significant values of

bootstrap and posterior probability (ML = 86, BI = 0.99, Fig. 5b). Despite a moderately

supported divergence from the other northern populations, Palmar was included in the

northern lineage (Fig. 5b). The haplotype network (cyt b) confirmed the genetic separation of

the populations located south and north of the river Río Térraba (Fig. 6), with the two clusters

being separated by a minimum of 20 mutational steps. Eleven out of 16 haplotypes (seven in

the northern and four in the southern lineage) were private. Within the northern lineage,

haplotypes are shared among intermediate and monomorphic red or green populations except

for haplotypes H11 and H12 belonging to Portalon and Matapalo respectively (Fig. 6).

49

Figure 5 Phylogenies obtained analysing (a) 16S and (b) cyt b sequences. The ML bootstrap

values and Bayesian posterior probabilities are reported at nodes. For each O. granulifera

sequence we indicated: the species name, the population code (as in Table 1) and sample

number and in (b) the haplotype code (H1 to H16); “Wang” indicates the sequences from

Wang (2011) included in the analyses; font colours correspond to O. granulifera colour

morphs as follows: red = red morph, yellow = intermediate morphs, green = green morph. S =

southern lineage, N = northern lineage. * indicates sequences from Palmar, ** indicate

sequences from TE.

(a) (b)

50

Figure 6 Network of cyt b haplotypes of Oophaga granulifera. Haplotypes (H) are indicated

by circles whose areas correspond to the haplotype frequency; S and N indicate southern and

northern lineages; the dashed double arrow shows the minimum separation between lineages.

Correspondences between colours and populations are indicated in figure; * distinguishes the

populations sampled for this study (coded as in Table 1) from the populations sampled in

Wang (2011) included in the analyses (Fig. 1).

51

3.4. Microsatellite loci population structure

All microsatellite loci except one were polymorphic for the eleven O. granulifera populations.

Locus D4 was monomorphic in the three populations sampled south of the river Río Térraba

(Charcos, Ventana and Neotropica). No deviation in linkage disequilibrium was found (P >

0.05). In eight populations one to two loci deviated from HWE after Bonferroni correction

(Table 5). Within-population diversity was overall higher in southern populations than in

northern ones (Table 5). Allelic richness ranged from 5.11 (± 2.00) in Matapalo to 8.00 (±

3.27) in Ventana and was higher in the southern populations and Palmar (6.97 on average)

than in the northern populations (5.12). Private alleles were mainly found in the three

southern populations, Charcos, Ventana and Neotropica, as well as in Palmar sampled just

north of the river Terraba (9.5 on average), while they were scarce in other northern

populations (2.3 on average). Significant FST values (P < 0.05) were detected in 52 out of 55

pairwise comparisons among populations (Table 6). Mean FST value was 0.11 (± 0.05),

ranging from 0.01 (between Baru and Firestone) to 0.18 (between southern population

Charcos and the northernmost green population Damitas). Population of the northern mtDNA

lineage had higher average FST value (0.11 ± 0.05) than those belonging to the southern

lineage (0.05 ± 0.01). AMOVA results markedly differed from those with the mitochondrial

cyt b marker. Most of the variation was found within populations (84.53%, FST = 0.15, P <

0.01), whereas variation among lineages (9.63%, FCT = 0.10, P < 0.001) and among

population within lineages (5.84% FSC = 0.06, P < 0.001) were much lower. Bayesian

assignment analysis identified two well defined clusters (Fig. 7; see Additional Information

for ∆K results). The first one included the Osa Peninsula populations, while the second

grouped all the populations belonging to the northern mitochondrial lineage except Palmar.

Palmar presented an intermediate population structure having genotypes similar to both the

southern and the northern clusters (Fig. 7). The differentiation of Palmar from Charcos,

Neotropica, Ventana grouped in the southern cluster is confirmed by high pairwise Fst values

52

(Table 6). One individual from the northern population Dominical showed a genetic profile

close to the southern populations. Admixture was high within the clusters and low among

clusters (Fig. 7).

Table 5 Analyses of seven microsatellite loci of Oophaga granulifera; n = number of

individuals analysed; AR[14], mean allelic richness (± SD) estimated for a minimum of seven

individuals; Ho, observed heterozygosity (± SD); He, expected heterozygosity (± SD); HWE,

loci out of Hardy–Weinberg equilibrium after Bonferroni correction; AP, number of private

alleles.

Population n AR[14] Ho He HWE AP

Charcos 8 6.95 ± 3.07 0.71 ± 0.42 0.72 ± 0.32 D4, G5 6 Neotropica 8 6.62 ± 3.15 0.75 ± 0.37 0.71 ± 0.32 D4, G5 8 Ventana 7 8.00 ± 3.27 0.71 ± 0.35 0.73 ± 0.32 D4 15 Palmar 7 6.29 ± 2.75 0.55 ± 0.21 0.73 ± 0.14 O1 9 Firestone 8 4.92 ± 2.06 0.66 ± 0.21 0.66 ± 0.16 1 Dominical 8 5.12 ± 1.75 0.61 ± 0.15 0.67 ± 0.13 F1 2 Baru 8 4.83 ± 1.25 0.61 ± 0.25 0.66 ± 0.16 O1 – Matapalo 10 5.11 ± 1.99 0.70 ± 0.23 0.65 ± 0.19 3 Portalon 15 5.36 ± 2.15 0.69 ± 0.18 0.73 ± 0.14 O1 1 Savegre 8 5.40 ± 2.17 0.71 ± 0.20 0.68 ± 0.15 – Damitas 11 5.24 ± 2.76 0.64 ± 0.29 0.67 ± 0.24 C3 1

Table 6 Differentiation between Oophaga granulifera populations estimated by pairwise FST

values. (*) Significant FST values (P < 0.05). Population codes (Pop) as in Table 1.

Pop CH NE VE PA FS DO BA MA PL SA DA CH – NE 0.05* – VE 0.04* 0.06* – PA 0.14* 0.16* 0.13* – FS 0.17* 0.16* 0.16* 0.11* – DO 0.15* 0.16* 0.14* 0.14* 0.10* – BA 0.15* 0.13* 0.15* 0.10* 0.01 0.09* – MA 0.17* 0.17* 0.17* 0.13* 0.07* 0.11* 0.07* – PL 0.13* 0.12* 0.14* 0.11* 0.07* 0.04* 0.05* 0.04* – SA 0.14* 0.12* 0.12* 0.07* 0.05* 0.06* 0.03 0.06* 0.03* – DA 0.18* 0.16* 0.18* 0.11* 0.06* 0.08* 0.06* 0.06* 0.05* 0.02 –

53

Figure 7 Assignment probabilities of individuals of Oophaga granulifera from Costa Rica to

putative population clusters at K=2 using the program STRUCTURE. Populations codes (as

in Table 1, font colours as in Fig.5).

3.5 Bioacoustics

Call duration, call rate and dominant frequency were significantly different between southern

and northern lineages, whereas there was no difference in pulse rate (Table 7). In the southern

genetic lineage frogs had longer and slower calls and higher dominant frequency as compared

to the northern lineage (Table 8; Fig. 8). Variability among populations was highest for call

duration while it was lowest for dominant frequency (Table 9). The calls of Charcos,

Neotropica and Firestone showed the lowest variation. The calls of the males from Portalon

and Savegre were the most variable (Table 9). Within the northern lineage, pulse rate

significantly differed between the red and green colour morphs (F = 10.16, P < 0.001).

54

Table 7 ANOVA results testing for differences in call parameters between southern and

northern populations of Oophaga granulifera. Abbreviations for call parameters: PR = pulse

rate, CD = call duration, CR = call rate, DF = dominant frequency. * indicates significant

values.

Variable F P

PR (1s–1) 3.14 0.082

CD (s) 70.11 <0.0001*

CR (1s–1) 6.36 0.015*

DF (kHz) 6.36 <0.0001*

Table 8 Male advertisement call parameters in the Oophaga granulifera populations. N =

number of individuals recorded. Abbreviations for call parameters: PR = pulse rate, CD = call

duration, CR = call rate, DF = dominant frequency. All values given as mean ± SD.

Population N PR (1s–1) CD (s) CR (1s–1) DF( kHz)

Charcos 9 174.14±17.06 0.53±0.08 10.37±0.95 4.21±0.19

Neotropica 14 165.99±14.97 0.57±0.08 9.93±1.21 4.20±0.15

Ventana 7 203.59±36.38 0.44±0.10 11.70±1.83 4.32±0.16

Palmar 3 168.76±26.73 0.28±0.07 16.85±2.07 4.07±0.28

Firestone 4 172.49±18.74 0.33±0.03 14.60±1.40 4.15±0.08

Baru 4 181.51±22.85 0.31±0.1 16.33±2.12 4.16±0.18

Matapalo 3 209.13±49.36 0.36±0.09 15.75±2.36 3.97±0.11

Portalon 3 211.37±24.16 0.26±0.11 17.67±2.03 3.98±0.22

Savegre 5 221.98±33.71 0.32±0.13 14.27±1.80 4.27±0.06

55

Figure 8 Variation in male advertisement call parameters among southern and northern

populations of Oophaga granulifera: (A) call duration ; (b) call rate; (c) dominant frequency;

(d) pulse rate

(A)

(B)

56

(C)

(D)

57

Table 9 Coefficients of variation [(mean/standard deviation)*100] for the male advertisement

call parameters. µpop = mean across populations. Abbreviations for call parameters: PR =

pulse rate, CD = call duration, CR = call rate, DF = dominant frequency.

CVPR CVCD CVCR CVDF µpop

Charcos 9.8 14.5 9.1 4.6 9.6

Neotropica 9.0 13.5 12.2 3.7 9.4

Ventana 17.9 23.2 15.6 3.6 14.6

Palmar 15.8 24.3 12.3 6.9 15.1

Firestone 10.9 7.9 9.6 1.8 7.9

Baru 12.6 31.9 13.0 4.3 16.1

Matapalo 23.6 24.2 15.0 2.9 15.7

Portalon 11.4 41.5 11.5 5.5 19.7

Savegre 15.2 40.6 12.6 1.3 19.9

µpar

14.0

24.6

12.3

3.8

16.3

3.6 Morphology

The parameters weight, body length, femur length, tibia length, humerus length and radio-

ulna length significantly differed between southern and northern genetic lineages (Table 10;

Fig. 9). Frogs belonging to the red southern populations were lighter, smaller and with shorter

limbs than frogs from red, intermediate and green northern populations. Among the northern

populations, head length (F = 7.99; P < 0.0001), head width (F = 24.74; P < 0.0001),

tympanum diameter (F = 9.68; P < 0.0001) and between-nostril distance (F = 6.98; P <

0.0001) together with body weight (F = 15.57; P < 0.0001) and body length (F = 9.34; P <

58

0.0001) were significantly different between red and green colour morphs. In the northern

lineage, red frogs were heavier, longer and had an overall larger head compared to the green

ones.

Table 10 ANOVA comparisons between genetic lineages for ten morphological parameters. *

indicates significant values

Variable F P

weight F= 21.66 P < 0.0001*

body length F = 20.45 P < 0.0001*

femur length F = 28.61 P < 0.0001*

tibia length F = 66.75 P < 0.0001*

humerus length F= 7.15 P < 0.0001*

Radio-ulna length

F= 0.42 P < 0.0001*

head length F= 0.13 P = 0.71

head width F = 1.19 P = 0.28

tympanum diameter

F=0.253 P = 0.14

Between-nostril distance

F = 0.33 P = 0.56

59

Figure 9 Box plots for ten morphological variables comparing southern and northern genetic

lineages in O. granulifera. (A) body weight, (B) body (snout-vent) length, (C) tibia, (D)

femor, (E) humerus, (F) radio-ulna, (G) head-width, (H) head length, (I) tympanum diameter,

(L) between nostril-distance.

(A)

60

(B)

(C)

61

(D)

(E)

62

(F)

(G)

63

(H)

(I)

64

(L)

65

4. Discussion

4.1 General discussion

The fine-scale analysis of genetic and phenotypic variation conducted in the present study

revealed significant geographical divergence in genetics, colouration, advertisement calls and

body size, in Oophaga granulifera populations in Costa Rica. Specifically, the results here

presented show that O. granulifera possesses higher colour diversity than previously

described (Savage, 2002; Wang, 2011). Beside the red, yellow and green colour morphs, the

species shows an array of intermediate morphs and is polymorphic in the northern part of its

distribution. Most of this remarkable variation occurs in a small transition area between

populations monomorphic for red and green phenotypes. To the north of the river Río

Térraba, first the colour phenotype changes quantitatively following a well defined south-

north cline toward increase of the red body area. Then the colour phenotype differentiates into

an array of colours spanning the chromatic range between red and green. In contrast, all

populations occurring in the Osa peninsula in the south of the distribution are very consistent

in colour characteristics. I provide the first description of colour polymorphic populations of

O. granulifera. The three polymorphic populations consisted of yellow morphs together with

green or red and other novel intermediate phenotypes. Thus, in contrast to previous studies

(Savage, 2002; Wang, 2011), intermediate frogs were always found in polymorphic

populations. The distribution of colour morphs in O. granulifera is asymmetric compared to

the genetic structure resulting from the phylogenetic reconstruction. The southern lineage is

monomorphic red while the northern lineage contains red, green and polymorphic

populations. Within the northern lineage, a monophyletic clade includes all of the northern

populations except Palmar. In this clade, Firestone, Dominical, Portalon, Matapalo, Savegre

and Damitas are closely related while Baru appears to be more differentiated. In contrast to a

66

previous study (Wang, 2011), I did not find a monophyletic yellow-green clade, which

presumably relates to the more comprehensive sampling in the colour transition area

performed in the present study. A missing link between colour and genetic differentiation in

poison frogs has been previously observed in O. pumilio (Hagemann & Pröhl, 2007; Wang &

Shaffer, 2008; Brown et al., 2010; Hauswaldt et al, 2011; but see Wang & Summers, 2010).

A recent colour radiation might explain the lack of correspondence between genetic

divergence at mitochondrial loci and colour divergence. According to our phylogeny the

Palmar population belongs to the northern lineage while the TE population from Wang (2011)

belongs to the southern one. Interestingly, the sample location was north of the river Río

Térraba while Wang (2011) sampled south of it (Ian Wang, pers. comm.; in the cited study the

name of the river was misspelled as Rio Terribe). The River Río Térraba is the largest of

Costa Rica and has been estimated to have originated in the Quaternary period (Bergoeing, et

al., 1997). The historical presence of such an important physical barrier may well present the

major determinant of the observed inter-lineage divergence. Considering a mutation rate for

mitochondrial DNA of approximately 2% per million years (Avise, 2000), the genetic

distance I observed between lineages (4% for cyt b) is congruent with a role of the river Río

Térraba in the O. granulifera intra-specific divergence. The analyses of 16S data confirms the

divergence between the southern and northern lineages and show that the intraspecific genetic

distance in O. granulifera is comparable to the distance between different Oophaga species.

The data therefore show, using a much larger sampling than previous studies (nine

populations instead of one), that O. granulifera is highly divergent from all congeneric

species (Hagemann & Pröhl; Summers & Clough, 2000). The divergence into two main

lineages is furthermore supported by the haplotype network that defines two clusters

corresponding to the southern and northern lineages. Deep intraspecific genetic divergence

has been found in other dendrobatids. For instance, high divergence in cyt b haplotypes is

reported for Epipedobates femoralis and E. hahneli from Western Amazonia (12% and 7.1%,

67

Lougheed et al., 1999; Roberts et al., 2007), and in O. pumilio from Costa Rica and Panama

(7%, Hauswaldt et al., 2011). Nonetheless, the present study is the first providing concordant

lines of evidence for a deep inter-specific divergence, using historical and recent genetic

inference and phenotypic traits. Notably, the two O. granulifera lineages diverged both in

advertisement calls and in morphological traits and present highly reduced gene admixture.

The frogs of the Osa peninsula are smaller and have longer notes (calls), lower call rate and

higher dominant frequency. Thus, the two lineages have undergone divergent evolutionary

trajectories leading to a deep divergence in which reinforcement mechanisms, such as

assortative mating, were presumably involved in impeding hybridization. The opposite

geographical trend has been found in O. pumilio in Costa Rica, in which two southern and

northern genetic lineages roughly correspond to two bioacoustic groups, and the southern

frogs were bigger and had higher call rate (Pröhl et al., 2007; Hagemann & Pröhl, 2007;

Hauswaldt et al., 2011). This may be explained if stochastic genetic processes, linked for

example to demographic fluctuations during colonization movements, are important factors in

originating intraspecific divergence, secondarily maintained and increased by selection.

The microsatellite STRUCTURE analysis evidences the presence of two well distinguished

southern and northern genetic clusters corresponding to the mitochondrial lineages with the

exception of the population Palmar which has an intermediate genetic structure sharing

genotypes from both clusters (Fig. 7). This is likely to be dependent on the peculiar location

of this population which is in between the southern and the northern populations (Fig. 2). In

particular, the pattern may be explained considering the geographical proximity between the

populations occurring in the Río Térraba area (Palmar and TE) together with changes in the

water level and position of the river that may have occurred over time. Overall, the genetic

data show a deep intraspecific divergence between populations located south and north of the

river Río Térraba and indicate the populations of the river basin as historical link between the

two lineages. The differentiation I observed in both calls and morphology between red,

68

intermediate and green colour morphs points to a potential role of mate choice in Oophaga

granulifera colour divergence. Interestingly, in the colour transition area, the majority of the

intermediate phenotypes co-occur with the green phenotype (Portalon, Savegre).

Both genetic and colour geographical variations in O. granulifera do not follow a clear cline,

considering the admixture among populations within the northern cluster and the asymmetric

distribution of intermediate phenotypes in the colour transition area. The lack of a clear colour

cline and a genetically well defined hybrid area may be explained if some of the intermediate

colour phenotypes in O. granulifera are not hybrids between red and green morphs, but

evolutionary steps from the red to the green phenotype. This would explain the fact that we

did not observe any population where pure red and green morphs co-occur. The observed

pattern in colour variation may be the result of complex crossing between red, intermediates

and green morphs over time across the colour transition area.

Wang (2011) stated that the O. granulifera cryptic morph was derived from the bright red

morph. In my opinion Wang (2011) and our phylogeographic reconstruction are congruent

with both a direct evolution of the green morph from a red ancestor and an indirect evolution

via intermediate steps. The green morph might have evolved from the red one and all the

intermediate morphs found in the transition area would be the result of hybridization between

red and green morphs or, alternatively, the intermediate morphs evolved from the red and

subsequently diverged to the green morph. In the second case, the intermediate colours

observed in the transition area would be a mixture of pure phenotypes and hybrids. The

coexistence of bright and dull colour morphs we found in the O. granulifera northern lineage

was also observed in the O. pumilio southern lineage (Hagemann & Pröhl, 2007; Wang &

Shaffer, 2008; Hauswaldt et al., 2011). These colour phenotypes have different

conspicuousness against the natural background (Pröhl & Ostrowski, 2011; Wang, 2011),

supporting the hypothesis that aposematic and cryptic colour morphs can coexist in systems

primarily evolved as aposematic (Darst et al., 2006; Speed & Ruxton, 2007). Several

69

evolutionary factors have been considered to explain the puzzling colour divergence in poison

frogs: genetic drift, differential predation pressure, toxin availability and sexual selection

(Speed & Ruxton, 2007; Rudh et al., 2007; Summers et al., 1997, 1999; Saporito et al.,

2007a). Interestingly, I found that the polymorphic and green populations show reduced body

size. This is congruent with the evolution of a cryptic antipredator strategy based on lack of

detectability to the predator instead of conspicuous signalling (Hagman & Forsman, 2003). In

addition, the lower genetic variation in the northern populations as revealed by microsatellite

allele diversity points to a possible role of genetic drift in the evolution of colour divergence

in O. granulifera. The effect of stochastic variation on allele frequency was advanced to

explain the evolution of isolated populations (Wright, 1977, Nosil, 2007). Examples have

been reported in O. pumilio and Heliconius butterflies (Rudh et al., 2007; Mallet & Joron,

1999). The presence of highly variable gradual intermediate colour morphs in O. granulifera

and the high population admixture in the northern lineage suggest that a certain amount of

gene flow between colour morphs exists in the colour transition area. Consequently, the

colour divergence within the northern lineage implies the maintenance of selection. Female-

biased parental care as observed in the Oophaga group may have increased female choosiness

and created the conditions for sexual selection through assortative mating by visual and

acoustic cues (Summers et al., 1997). Since male vocalizations are key sexual signals in frogs,

the difference in the pulse rate of the male advertisement call we found between red and green

colour morphs suggest the potential occurrence of female mate choice. The divergence in

body size, may also suggest a role of sexual selection in intraspecific divergence in O.

granulifera. In fact, assortative mating mediated by body size is demonstrated in amphibians

(Berven, 1981; Gabor et al., 2000; Takahashi et al., 2010). Reduced body size may therefore

have originated by natural selection for increased crypsis and be maintained by differential

mating selection in the colour transition area. Empirical tests in the field using mate choice

70

experiments are required to evaluate the actual role of sexual selection in the evolution and

maintenance of colour diversity in O. granulifera.

4.2 Why should a cryptic morph evolve in an aposematic species?

4.2.1 General background

Colour pattern polymorphisms can arise and be maintained by selection favouring multiple

variants within populations (Hoffman & Blouin, 2000; Merilaita, 2006; Vercken et al., 2007;

Dijkstra et al., 2008). Nonetheless, an aposematic species is a peculiar case in which the

evolution of colour diversification is predicted to be highly constrained (Müller, 1879).

Particularly, the possible adaptive role of a cryptic colour variant in an aposematic system is

debated. In fact, this is in constrast with the aposematism theory which predicts the functional

association of toxicity and conspicuous colour to enhance the efficancy of the deterrent signal

to predators. In non toxic species, a cryptic colouration can facilitate predator avoidance

(Cott, 1940; Endler, 1992; Duellman, 2001) therefore being under natural selection

(Robertson, 2008) or can be involved in mate choice (Gomez & Thery, 2007). Laboratory

studies have tested predation rates as a function of colour pattern in polymorphic frogs and

found that predators attacked cryptic (background-matched) morphs less often (Tordoff, 1980;

Morey, 1990). However, for toxically defended anurans such as poison-arrow frogs, bold

colour patterns and overall conspicuousness are thought to protect individuals from predation

attempt and detrimental effects connected (Summers & Clough, 2001; Darst et al., 2006).

Disentangling the selective forces responsible for colour diversification in aposematic

dendrobatids is difficult because several different and potentially interacting processes could

underlie divergent phenotypic expression, including natural (Hairston, 1979; Endler, 1980)

and sexual selection (Endler, 1983; West-Eberhard, 1983; Panhuis et al., 2001; Masta &

71

Maddison, 2002; Maan et al., 2004). I discuss here those among these selective forces that are

considered the most significant and how these may apply to O. granulifera.

4.2.2 Role of genetic drift

Stochastic fluctuations in allelic frequencies have been proposed to play a role in the

evolution of animal populations (Wright, 1933). Recently the actual role of drift has been re-

evaluated and its potential evolutionary effects are not considered of general application.

Nonetheless, under particular conditions such as population reduction and isolation the effect

of drift may be significant and shaping the evolutionary trajectories taken by those

populations (Lande, 1977). This may be the case, for instance, for the colour diversity

observed in the Panamanian populations of O. pumilio, which are small and separated (Rudh

et al., 2007). In O. granulifera, population size reduction may have happened during the

colonization movements toward the northern part of the distribution. The data concerning

microsatellite allelic diversity here presented suggest a decrease in genetic diversity in the

northern intermediate and green populations, concordant with an effect of drift. Drift may

have, for example, determined fluctuations in the frequency of the red aposematic morph

creating the conditions for a colour diversification in combination with other factors. It is

interesting at this regard, considering that theoretical models predict an association between

the frequency of the aposematic morph in the population and the promptness with which

predators learn to avoid it (Speed, 1993; Turner & Speed, 1996; MacDougall & Dawkins,

1998; Servedio, 2000; Speed & Ruxton, 2007).

4.2.3 Role of predation

In aposematic toxic animals, it has been shown that conspicuousness, that is a highly

contrasted appearance to the background, can make individuals more vulnerable to attacks by

naïve (e.g., Lindström et al., 2001a,b; Riipi et al., 2001; Lindstedt et al., 2008) or specialized

72

predators (Yosef & Whitman, 1992). Moreover, the predators visual and learning abilities and

resistance to prey toxicity, may vary both within and among species determining locally

variable predation risk for a potential aposematic prey (Calvert, 1979; Fink & Brower, 1981;

Yosef & Whitman, 1992; Pinheiro, 1996; Mappes & Alatalo, 1997; Marples et al., 1998;

Exnerová et al., 2003, 2007). Another factor to take into consideration is the ecological

community to which the aposematic animals belong to. Also, the vulnerability to predation

depends on the amount of alternative prey available (Merilaita & Kaitala, 2002; Lindström et

al., 2004) and the hunger level of predators (Barnett et al., 2007; Sandre et al., 2010). Overall,

predation could select for intermediate signals in some circumstances (Endler & Mappes,

2004; Tullberg et al., 2005; Ruxton et al., 2009) or allow for signal variation in others (Rowe

et al., 2004; Darst et al., 2006; Ham et al., 2006; Ihalainen et al., 2007). Generally, predators

are thought to learn to avoid unpalatable prey faster if they are conspicuous rather than cryptic

(Gittleman & Harvey, 1980; Roper & Redston, 1987; Alatalo & Mappes, 1996; Lindström et

al., 1999b) and if they are clearly visible (highly contrasted) against their background (Roper

& Redston, 1987; Lindström et al., 1999b; Gamberale-Stille, 2001; Prudic et al., 2007). Many

typical warning colours (e.g. red, orange or yellow) may also be innately aversive to predators

in comparison with common cryptic colours, such as green and brown, which can increase the

benefits of being warning coloured (Smith, 1975; Schuler & Hesse, 1985). Despite the

advantages of clearly visible warning colours (Guilford, 1986; Gamberale-Stille, 2000;

Schlenoff, 1984; Marples et al., 1998; Thomas et al., 2003; Roper & Redston, 1987; Roper,

1990; and Sherratt & Beatty, 2003; Ham et al., 2006), intermediate warning signals may be

selected instead of conspicuous ones if the predation risk varies (Endler & Mappes, 2004). In

dendrobatids the importance of predation as a driving force in the evolution of colour

diversity has been recently evidenced using empirical tests to assess the actual association of

colour phenotype and predation pressure (Comeault & Noonan, 2011, Saporito et al., 2007b).

The efficacy of an aposematic defense is dependent on the frequency of the conspicuous

73

colour morph characterizing the system (Speed, 1993; Turner & Speed, 1996; MacDougall &

Dawkins, 1998; Servedio, 2000; Lindström et al., 2001b; Speed & Ruxton, 2007), because a

high frequency is necessary to elicit the avoidance learning by the predators increasing the

deterrent efficacy of the signal. In O. granulifera, as mentioned, a demographic reduction may

be related to a decrease in genetic variability in the northern green populations. This together

with a colonization movement south-north (Wang, 2011) point to possible demographic

fluctuations linked to the occupancy of new habitats. If, following genetic drift phenomena,

the frequency of the red morph decreased in the O. granulifera northern populations, an

increased predation pressure on the that conspicuous morph may have produce a positive

selection for a cryptic variant. Moreover, the less conspicuous behaviour observed in the

green morph (personal observation) is congruent with an overall cryptic strategy of the green

morphs in O. granulifera. Importantly, to allow a significant evaluation of the role of

predation in the evolution of colour diversity in dendrobatids, the actual level of toxicity of

cryptic and conspicuous morphs has to be taken into consideration (Maan & Cummings,

2012).

4.2.4 Role of toxicity

The presence of chemical defences is a fundamental trait of aposematic defences as the

predators learn to avoid a conspicuously colours animals due to the unpalatability and adverse

effect provoked by toxins (Muller, 1879). Therefore, the possible variation of the actual

toxicity in a population may influence the selection pressure at the local level and create

potential for the diversification in aposematic signals (Wang, 2011). It has been proposed that

once aposematism is established in a system, then colouration can become decoupled and

diversifies following local variation in toxicity. There is evidence that in O. pumilio, the dull

coloured less constrasted (cryptic) morphs are significantly less toxic than the bright coloured

morphs (Maan & Cummings, 2012). This finding is in contrast with Wang (2011) for O.

74

granulifera that showed an overall higher toxicity in green and intermediate morphs. These

two apparently contrasting findings may be explained considering a crucial point in

dendrobatid chemical defences that is the dependence on the diet as a source of toxins. In fact,

spatial and temporal variation in the prey communities from which dendrobatids derive their

alkaloids must be taken into consideration before depicting any related evolutionary scenario.

In the two mentioned studies these longterm observational data are missing therefore any

deduction has to be careful (see Saporito et al., 2007c). Another factor that has to be

thoroughly weighed is the methodology that has been followed up to date to evaluate the

toxicity of dendrobatids, which is an intradermal injection of frog skin extract in laboratory

mice. Both the species used and the way of assumption may not be optimal representation of

the real processes as they happen in nature. Nonetheless, the difference in toxicity was

quantitative (toxicity essays; both studies) and qualitative (number of identified alkaloids;

Wang, 2011) and is consistent across different cryptic morphs therefore it may represent an

important factor in the evolution of colour diversity in aposematic species. The possible

reason explaining the constrasting pattern observed in O. pumilio and O. granulifera may be

dependent on the different biogeographic conditions in which the two systems evolved.

Indeed, most of O. pumilio colour diversity is found in the Bocas del Toro archipelago in

Panama, formed by many different small islands of recent origin while O. granulifera is

found in mainland Costa Rica in a potentially continuous distribution. In O. granulifera, the

effect of an increased toxicity may have reduced the need for a bright colouration over time.

This is because, predators can learn to avoid a highly toxic not profitable prey even if dully

coloured. The related evolutionary mechanism is compatible with individual selection because

any predation attempt may be neutralized if powerful toxins cover the body surface, such as in

the case of dendrobatids, determining immediate repellent and noxious effects to the predator.

Evolving a dull colouration, the potential costs involved in the bright pigments production are

avoided. Several factors will have to be clarified to test the feasibility of this proposed

75

scenario. Other than the already mentioned possible fluctuations in toxicity and the

questionable ecological significance of the toxicity essay used, information on the pigments

involved in dendrobatids pigmentation are necessary to evaluate the actual cost of pigment

production.

4.2.5 Role of sexual selection

As discussed before, together with the role of natural selection mediated by predators, sexual

selection has been suggested to play an important role in the diversification of aposematic

signals in poison-dart frogs. Females show preference for brighter (Maan & Cummings, 2009)

and more colourful males (Maan & Cummings, 2008) as well for males of their own

population in O. pumilio (Summers et al., 1999; Reynolds & Fitzpatrick, 2007; Maan &

Cummings, 2008). I showed here, that aposematic red and cryptic green morphs have a

different pulse rate in the male advertisement call in the northern populations. A significant

differentiation in this key sexual signal indicates that the females may potentially select males

of their own colour, similarly with what has been observed in the congeneric species O.

pumilio.

76

5. Conclusion

The present work provides evidence for deep intraspecific divergence in the polymorphic

aposematic poison-dart frog Oophaga granulifera. I described the patterns of geographical

variation in historical and recent genetic structure and in key phenotypic traits such as

morphology and advertisement call. The data here presented all congruently show that the

southern and northern lineages identified are clearly diverging and may well represent

different species. The phylogeographic reconstruction and the recent population structure

indicate that the river Río Térraba is likely to be the geographic barrier between the lineages

that are, therefore, geographically, genetically and phenotypically separated.

As concerned to the recent evolution of colour diversity in O. granulifera, several factors may

potentially be involved in the origin and maintenance of this notable differentiation.

Alongside correlation data such as those here presented future research efforts will be

necessary to empirically evaluate the effective strength in mating preferences shaping the

observed variation in colouration. Moreover, many aspects of dendrobatid ecology, such as

actual predator community, prey community and related spatial and temporal variation in the

acquired toxicity, will have to be clarified to define a clearer evolutionary scenario explaining

the reason of the remarkable colour variation in this aposematic vertebrate.

77

6. References

Aichinger, M. (1991) Tadpole transport in relation to rainfall, fecundity and body size in five

species of poison-dart frogs from Amazonian Peru. Amphibia-Reptilia, 12, 49–55.

Alatalo, R.V. & Mappes, J. (1996) Tracking the evolution of warning signals. Nature, 382,

708–709.

Andersson, S., Örnborg, J. & Andersson, M. (1998) Ultraviolet sexual dimorphism and

assortative mating in blue tits. Proceedings of the Royal Society of London series B, 265,

445–450.

Arak, A. (1983) Sexual selection by male-male competition in natterjack toad choruses.

Nature, 306, 261–262.

Arntzen, J.W. & Wallis, G.P. (1999) Geographic variation and taxonomy of crested newts

(Triturus cristatus superspecies): morphological and mitochondrial DNA data.

Contributions to Zoology, 68, 181–203.

Auber, L. (1957) The distribution of structural colours and unusual pigments in the class

Aves. Ibis, 99, 463–476.

Baeta, R., Faivre, B., Motreuil, S., Gaillard, M. & Moreau, J. (2008) Carotenoid trade-off

between parasitic resistance and sexual display: an experimental study in the blackbird

(Turdus merula). Proceedings of the Royal Society series B, 275, 427–434.

Bailey, R.I., Thomas, C.D. & Butlin, R.K. (2004) Premating barriers to gene exchange and

their implications for the structure of a mosaic hybrid zone between Chorthippus brunneus

and C. jacobsi (Orthoptera: Acrididae). Journal of Evolutionary Biology, 17, 108–119.

Barlow, G.W. (1961) Causes and significance of morphological variation in fishes.

Systematic Zoology, 10, 105–117.

78

Barluenga, M., Stölting, K.N., Salzburger, W., Muschick, M. & Meyer, A. (2006) Sympatric

speciation in Nicaraguan crater lake cichlid fish. Nature, 439, 719–723.

Barnett, C.A., Bateson, M. & Rowe, C. (2007) State-dependent decision making: educated

predators strategically trade off the costs and benefits of consuming aposematic prey.

Behavioural Ecology, 18, 645–651.

Barton, N. (2004) Speciation: Why, how, where and when? Current Biology, 15, R603–R604.

Beatty, C.D., Beirinck, K. & Sherratt, T.N. (2004) The evolution of Müllerian mimicry in

multispecies communities. Nature, 431, 63–67.

Benson, W. (1972) Natural selection for Mullerian mimicry in Heliconius erato in Costa Rica.

Science, 176, 936–939.

Berven, K.A. (1981) Mate choice in the wood frog, Rana sylvatica. Evolution, 35, 707–722.

Berven, K.A. (1982) The genetic basis of altitudinal variation in the wood frog Rana

sylvatica. An experimental analysis of life history traits. Evolution, 36, 962–983.

Bezzerides, A.L., McGraw, K.J., Parker, R.S. & Husseini, J. (2007) Elytra color as a signal of

chemical defense in the Asian ladybird beetle Harmonia axyridis. Behavioral Ecology and

Sociobiology, 61, 1401–1408.

Blair, W.F. (1958) Mating call in the speciation of anuran amphibians. American Naturalist,

92, 27–51.

Blount, J.D., Metcalfe, N.B, Birkhead, T.R & Surai, P.F. (2003) Carotenoid modulation of

immune function and sexual attractiveness in zebra finches. Science, 300, 125–127.

Blount, J.D., Speed, M.P, Ruxton, G.D. & Stephens, P.A. (2009) Warning displays may

function as honest signals of toxicity. Proceedings of the Royal Society series B, 276, 871–

877.

Bolaños, F. (1990) Actividad de canto y territorialidad en Dendrobates granuliferus. Tesi de

maestria, Universidad de Costa Rica.

79

Brakefield, P.M. & Liebert, T.G. (1985) Studies of colour polymorphism in some marginal

populations of the aposematic jersey tiger moth Callimorpha quadripunctaria. Biological

Journal of the Linnean Society, 26, 225–241.

Bossuyt, F. & Milinkovitch, M.C. (2000) Convergent adaptive radiations in Madagascan and

Asian ranid frogs reveal covariation between larval and adult traits. Proceedings of the

National Academy of Sciences USA, 97, 6585–6590.

Bourne, G.R., Collins, A.C., Holder, A.M. & McCarthy, C.L. (2001) Vocal communication

and reproductive behavior of the frog Colostethus beebei in Guyana. Journal of

Herpetology, 35, 272–281.

Brown, J., Maan, M.E., Cummings, M.E. & Summers, K. (2010) Evidence for selection on

coloration in a Panamanian poison frog: a coalescent-based approach. Journal of

Biogeography, 37, 891–901.

Caldwell, J.P. (1997) Pair bonding in spotted poison frogs. Nature, 385, 211–214.

Caldwell, J.P., & de Oliveira, V.R.L. (1999) Determinants of biparental care in the spotted

poison frog, Dendrobates vanzolinii (Anura: Dendrobatidae). Copeia, 3, 565–575.

Calvert, W.H. (1979) Mortality of the monarch butterfly (Danaus plexippus L.): avian

predation at five overwintering sites in Mexico. Science, 204, 847–851.

Camp, D.C., Marshall, J.L., Austin, R.M. Jr. (2000) The evolution of adult body size in black-

bellied salamanders (Desmognathus quadramaculatus complex). Canadian Journal of

Zoology, 78, 1712–1722.

Campbell, J. (1999) Distribution patterns of amphibians in Middle America. In: Patterns of

Distribution of Amphibians (ed. Duellman). The John Hopkins University Press,

Baltimore, Maryland.

Campbell, J.A. & Frost, D.R. (1993) Anguid lizards of the genus Abronia: revisionary notes,

descriptions of 4 new species, a phylogenetic analysis, and key. Bulletin of the American

Museum of Natural History, 1–121.

80

Carson, H.L. (1975) The genetics of speciation at the diploid level. American Naturalist, 109,

73–92.

Carson, H.L. & Templeton, A.R.A. (1984) Genetic revolutions in relation to speciation

phenomena: the founding of new populations. Annual Review of Ecology and Systematics,

15, 97–131.

Castellano, S. & Giacoma, C. (1998) Morphological variation of the green toad, Bufo viridis,

in Italy: a test of causation. Journal of Herpetology, 32, 540–550.

Cherry, L.M., Case, S.M. & Wilson, A.C. (1978) Frog perspective on the morphological

difference between humans and chimpanzees. Science, 200, 209–211.

Clarke, R.D. (1972) The effect of toe clipping on survival in Fowler's toad (Bufo woodhousei

fowleri). Copeia, 1, 182–185.

Clement, M., Posada, D. & Crandall, K.A. (2000) TCS: a computer program to estimate gene

genealogies. Molecular Ecology, 9, 1657–1659.

Clough, M. & Summers, K. (2000) Phylogenetic systematics and biogeography of the poison

frogs: evidence from mitochondrial DNA sequences. Biological Journal of the Linnean

Society, 70, 515–540.

Coates, A.C. & Obando, J.A. (1996) The geological evaluation of the Central American

Isthmus. In: Evolution and Environment in Tropical America (eds. Jackson, B.C., Budd,

A.F. & Coates, A.G.). University of Chicago Press, Chicago. Illinois, 21–56.

Colangelo, P., Castiglia, R., Franchini, P. & Solano E. (2010) Pattern of shape variation in the

eastern African gerbils of the genus Gerbilliscus (Rodentia, Muridae): environmental

correlations and implication for taxonomy and systematics. Mammalian Biology, 75, 302–

310.

Comeault, A. A. & Noonan, B.P. (2011) Spatial variation in the fitness of divergent

aposematic phenotypes of the poison frog, Dendrobates tinctorius. Journal of Evolutionary

Biology, 24, 1374–1379.

81

Cott, H.B. (1940) Adaptive Coloration in Animals. Methuen, London, UK.

Couldridge, V.C.K & Van Staaden, M.J. (2004) Habitat-dependent transmission of male

advertisement calls in bladder grasshoppers (Orthoptera; Pneumoridae). Journal of

Experimental Biology, 207, 2777–2786.

Coyne, J. A. (1992) Genetics and speciation. Nature, 355, 511–515.

Crawford, A.J. (2003) Huge populations and old species of Costa Rican and Panamanian dirt

frogs inferred from mitochondrial and nuclear gene sequences. Molecular Ecology, 12,

2525–2540.

Crawford, A.J., Bermingham, E. & Polania, C. (2007) The role of tropical dry forest as a

long–term barrier to dispersal: a comparative phylogeographical analysis of dry forest

tolerant and intolerant frogs. Molecular Ecology, 16, 4789–4807.

Crump, M.L. (1972) Territoriality and mating behaviour in Dendrobates granuliferus (Anura:

Dendrobatidae). Herpetologica, 28, 195–198.

Daly, J.W. & Myers, C.W. (1967) Toxicity of Panamanian poison frogs (Dendrobates): some

biological and chemical aspects. Science, 156, 970–973.

Daly, J.W., Myers, C.W. & Whittaker, N. (1987) Further classification of skin alkaloids from

neotropical poison frogs (Dendrobatidae), with a general survey of toxic/noxious

substances in the Amphibia. Toxicon, 10, 1023–1095.

Daly, J.W., Garraffo, H.M., Spande, T.F., Jaramillo, C. & Rand, A.S. (1994a) Dietary source

for skin alkaloids of poison frogs (Dendrobatidae)? Journal of Chemical Ecology, 20, 943–

955.

Daly, J.W., Gusovsky, F., Myers, C.W., Yotsu-Yamashita, M. & T. Yasumoto (1994b) First

occurrence of tetrodotoxin in a dendrobatid frog (Colostethus inguinalis), with further

reports for the bufonid genus Atelopus. Toxicon, 32, 279–285.

Daly, J.W., Garraffo, H.M. & Myers, C.W. (1997) The origin of frog skin alkaloids: an

enigma. Pharmaceutical News, 4, 9–14.

82

Daly, J.W. (1998) Thirty years of discovering arthropod alkaloids in amphibian skin. Journal

of Natural Products, 61, 162–172.

Daly, J.W., Garraffo, H.M. & Spande, T.F. (1999) Alkaloids from amphibian skins. In

Alkaloids: chemical and biological perspectives (ed. Pelletier, S.W.). Pergamon, New

York, 1–161.

Daly, J.W., Garraffo, H.M. Spande, T.F. Decker, M.W. Sullivan, J.P. & Williams, M. (2000)

Alkaloids from frog skins: the discovery of epibatidine and the potential for developing

novel non-opioid analgesics. Natural Products Report, 17, 131–135.

Darst, C.R., Cummings, M.E. & Cannatella, D.C. (2006) A mechanism for diversity in

warning signals: conspicuousness versus toxicity in poison frogs. Proceedings of the

National Academy of Sciences of the USA, 103, 5852–5857.

Darwin C (1859) On the Origin of Species by Means of Natural Selection (John Murray,

London).

Dijkstra, P.D., Seehausen, O. & Groothuis, T.G.G. (2008) Intra-sexual competition among

females and the stabilization of a conspicuous colour polymorphism in a Lake Victoria

cichlid fish. Proceedings of the Royal Society of London Series B, 275, 519–526.

Dobzhansky T. (1940) Speciation as a stage in evolutionary divergence. American Naturalist

74, 312–321.

Doebeli, M. & Dieckmann, U. (2004) Adaptive dynamics of speciation: spatial structure. In:

Adaptive Speciation (eds. Dieckmann, U., Doebeli, M., Metz, J.A.J., Tautz, D.).

Cambridge University Press, Cambridge, UK, 160–167.

Dopman, E.B., Robbins, P.S. & Seaman, A. (2010) Components of reproductive isolation

between North American pheromone strains of the European corn borer. Evolution, 64,

881–902.

Dosen, L.D. & Montgomerie, R. (2005) Female size influences mate preferences of male

guppies. Ethology, 110, 245–255.

83

Dover, G. (1995) A species definition: a functional approach. Trends in Ecology &

Evolution, 12, 489–490.

Driscoll, D.A. (1998) Genetic structure, metapopulation processes and evolution influence the

conservation strategies for two endangered frog species. Biological Conservation, 83, 43–

54.

Duellman, W.E. (2001) The Hylid Frogs of Middle America. Society for the Study of

Amphibians and Reptiles, Ithaca, New York.

Dumbacher, J.P., Wako, A., Derrickson, S.R., Samuelson, A., Spande, T.F. & Daly., J.W.

(2004). Melyrid beetles (Choresine): a putative source for the batrachotoxin alkaloids

found in poison-dart frogs and toxic passerine birds. Proceedings of the National Academy

of Science USA, 101, 15857–15860.

Edmunds, M. (1974) Defence in Animals. A Survey of Anti-Predator Defences. Longman,

New York.

Elwood, R., Gibson, J., Neil, S. (1987) The amorous Gammarus: size assortative mating in G.

pulex. Animal Behaviour, 35, 1–6.

Endler, J.A. (1977) Geographic Variation, Speciation and the Origin of Species. Princeton,

University Press, Princeton, New Jersey.

Endler, J.A. (1980) Natural selection on color patterns in Poecilia reticulata. Evolution, 34,

76–91.

Endler, J.A. (1983) Natural and sexual selection on colour pattern in poeciliid fishes.

Environmenatl. Biology of Fish, 9, 173–190.

Endler, J.A. (1992) Signals, signal conditions, and the direction of evolution. American

Naturalist, 139, 125–153.

Endler, J. & Mappes, J. (2004) Predator mixes and the conspicuousness of aposematic signals.

American Naturalist, 163, 532–547.

84

Ereshefsky, M. (1992) The Units of Evolution: Essays on the Nature of Species, MIT Press,

Cambridge, Massachussetts.

Evanno, G., Regnaut, S. & Goudet, J. (2005) Detecting the number of clusters of individuals

using the software STRUCTURE: a simulation study. Molecular Ecology, 14, 2611–2620.

Excoffier, L., Smouse, P.E. & Quattro, J.M. (1992) Analysis of molecular variance inferred

from metric distances among DNA haplotypes: application to human mitochondrlal DNA

restriction data. Genetics, 131, 479–491.

Excoffier, L., Laval, G. & Schneider, S. (2005) Arlequin version 3.5: an integrated software

package for population genetics data analysis. Evolutionary Bioinformatics Online, 1, 47–

60.

Exnerová, A., Landová, E., Stys, P., Fuchs, R., Prokopová, M. & Cehlariková, P. (2003)

Reactions of passerine birds to aposematic and non-aposematic firebugs

(Pyrrhocorisapterus, Heteroptera). Biological Journal of Linnean Society, 78, 517–525.

Exnerová, A., Stys, P., Fuciková, E., Veselá, S., Svadová, K., Prokopová, M., Jarosık, V.,

Fuchs, R. & Landová, E. (2007) Avoidance of aposematic prey in European tits (Paridae):

learned or innate? Behavioural Ecology, 18, 148–156.

Fandino, M.C., H. Luddecke & A. Amezquita. (1997) Vocalisation and larval transportation

of male Colostethus subpunctatus (Anura: Dendrobatidae). Amphibia-Reptilia, 18, 39–48.

Faulkner, D. J. (1992) Chemical defenses of marine molluscs. In: Ecological Roles of Marine

Natural Products (ed. Paul, V. J.). Comstock Publishing Associates, Ithaca, New York,

119–163.

Feder, J.L., Berlocher, S.H., Roethele, J.B., Dambroski, H., Smith, J.J., Perry, W.L.,

Gavrilovic, V., Filchak, K.E., Rull, J. & Aluja, M. (2003) Allopatric origins for sympatric

host-plant shifts and race formation in Rhagoletis. Proceedings of the National Academy of

Science USA, 100, 10314–10319.

85

Feder, J.L., Xie, X., Rull, J., Velez, S., Forbes, A., Dambroski, H., Filchak, K. & Aluja., M.

(2005) Mayr, Dobzhansky, Bush and the complexities of sympatric speciation in

Rhagoletis. Proceedings of the National Academy of Science USA, 102, 6573–6580.

Feder, J. L. & Nosil, P. (2009) Chromosomal inversions and species differences: when are

genes affecting adaptive divergence and reproductive isolation expected to reside within

inversions? Evolution, 63, 3061–3075.

Fink, L.S. & Brower, L.P. (1981) Birds can overcome the cardenolide defense of monarch

butterflies in Mexico. Nature, 291, 67–70.

Fisher, D.O. & Cockburn, A. (2006) The large-male advantage in brown antechinuses: female

choice, male dominance, and delayed male death. Behavioral Ecology, 17, 164–171.

Fitzpatrick, B.M., Fordyce, J.A. & Gavrilets, S. (2008) What, if anything, is sympatric

speciation? Journal of Evolutionary Biology, 21, 1452–1459.

Fitzpatrick, B.M., Fordyce, J.A. & Gavrilets, S. (2009) Pattern, process and geographic modes

of speciation. Journal of Evolutionary Biology, 22, 2342–2347.

Forester, D.C. & Czarnowsky, R. (1985) Sexual selection in the spring peeper Hyla crucifer

(Amphibia, Anura): role of the advertisement call. Behaviour, 92, 112–128.

Foster, S.A & Endler, J.A. (1999) Geographic variation in behavior: perspectives on

evolutionary mechanisms. Oxford University Press, New York.

Frazer, J.F.D. & Rothschild, M. (1960), Defence mechanisms in warningly-coloured moths

and other insects Proceedings of the 11th International Congress of Entomology, Vienna,

1959, 249–256.

Friedl, T.W.P. & Klump, G.M. (1997) Some aspects of population biology in the European

treefrog, Hyla arborea. Herpetologica, 53, 321–330.

Gabor, C.R., Krenz, J.D. & Jaeger, R.G. (2000) Female choice, male interference, and sperm

precedence in the red-spotted newt. Behavioral Ecology, 11, 115–124.

86

Gabor, C. & Page, R. (2003) Female preference for large males in sailfin mollies, Poecilia

latipinna: the importance of predation pressure and reproductive status. Acta ethologica, 6,

7–12.

Gamberale-Stille, G. (2000) Decision time and prey gregariousness influence attack

probability in naïve and experienced predators. Animal Behavior, 60, 95–99.

Gamberale-Stille, G. (2001) Benefit by contrast: an experiment with live aposematic prey.

Behavioural Ecology, 12, 768–772.

García-París, M., Good, D.A., Parra-Olea., G. & Wake, D.B. (2000) Biodiversity of Costa

Rican salamanders: implications of high levels of genetic differentiation and

phylogeographic structure for species formation. Proceedings of the National Academy of

Sciences, USA, 97, 1640–1647.

Gavrilets, S. (1997) Evolution and speciation on holey adaptive landscapes. Trends in

Ecology and Evolution, 12, 307–312.

Gavrilets, S., Li, H. & Vose, M.D. (2000) Patterns of parapatric speciation. Evolution,

54, 1126–1134.

Gavrilovic, V., Filchak, K.E., Rull, J. & Aluja, M. (2003) Allopatric origins for sympatric

host–plant shifts and race formation inRhagoletis. Proceedings of the National Academy of

Science USA, 100, 10314–10319.

Gerhardt, H.C. & Huber, F. (2002) Acoustic Communication in Insects and Anurans.

University of Chicago Press, Chicago.

Gittleman, J.L. & Harvey, P.H. (1980) Why are distasteful prey not cryptic? Nature, 286,

149–150.

Golay, N. & Durrer, H. (1994) Inflammation due to toe-clipping in natterjack toads (Bufo

calamita). Amphibia-Reptilia, 15, 81–96.

87

Gomez, D. & Thery, M. (2007) Simultaneous crypsis and conspicuousness in color patterns:

comparative analysis of a neotropical rainforest bird community. American Naturalist, 169,

42–61.

Goodman, D.E. (1971) Territorial behaviour in a neotropical frog, Dendrobates granuliferus.

Copeia, 2, 365–370.

Goudet, J. (1995) FSTAT version 1.2: a computer program to calculate F-statistics. Journal of

Heredity, 86, 485–486.

Grant, B.R & Grant, P.R (1989) Evolutionary dynamics of a natural population: the large

cactus finch of the Galápagos University of Chicago Press, Chicago, Illinois.

Grant, T., Frost, D.R., Caldwell, J.P., Gagliardo, R., Haddad, C.L., Kok, P.J., Means, D.B.,

Noonan, B.P., Schargel, W.E. & Wheeler, W.C. (2006) Phylogenetics systematics of

dart–poison frogs and their relatives (Amphibia: Athesphatanura: Dendrobatidae).

Bulletin of the American Museum of Natural History, 299, 1–268.

Gray, S.M. & McKinnon, J.S. (2006) Linking color polymorphism maintenance and

speciation. Trends in Ecology and Evolution, 22, 71–79.

Green, D.M. (1993) Evolutionary relationships of Eastern Palearctic brown frogs, genus

Rana: paraphyly of the 24-chromosome species group and the significance of chromosome

number change. Zoological Journal of the Linnean Society, 109, 1–25.

Greenwood, J.J.D., Wood, E.M. & Batchelor, S. (1981) Apostatic selection of distasteful

prey. Heredity, 47, 27–34.

Grether, G.F., Hudon, J. & Endler, J.A. (2001) Carotenoid scarcity, synthetic pteridine

pigments and the evolution of sexual coloration in guppies (Poecilia reticulata).

Proceedings of the Royal Society series B, 268, 1245–1253.

Griffith, S.C., T.H. Parker & Valérie, A. Olson (2006) Melanin- versus carotenoid-based

sexual signals: is the difference really so black and red? Animal Behaviour, 71, 749–763.

88

Gübitz, T., Thorpe, R.S. & Malhotra, A. (2005) The dynamics of genetic and morphological

variation on volcanic islands. Proceedings of the Royal Society of London Series B, 272,

751–757.

Guilford, T. (1986) How do ‘‘warning colors’’ work? Conspicuousness may reduce

recognition errors in experienced predators. Animal Behaviour, 34, 286–288.

Guilford, T. (1990) Evolutionary pathways to aposematism. Acta Oecologica, 11, 835–842.

Hagemann, S. & Pröhl, H. (2007) Mitochondrial paraphyly in a polymorphic poison frog

species (Dendrobatidae; D. pumilio). Molecular Phylogenetics and Evolution, 45, 740–747.

Hagman, M. & Forsman, A. (2003) Correlated evolution of conspicuous coloration and body

size in poison frogs (Dendrobatidae). Evolution, 57, 2904–2910.

Hairston, N.G. (1979) Adaptive significance of color polymorphism in 2 Species of

Diaptomus (Copepoda). Limnological Oceanography, 24, 15–37.

Halliday, T. (1996) Amphibians Ecological Census Techniques, A Handbook (ed. Sutherland,

W.J.). Cambridge University Press, Cambridge, UK.

Ham, A.D., Ihalainen, E., Lindström, L. & Mappes, J. (2006) Does color matter? The

importance of color in avoidance learning, memorability and generalisation. Behavioural

Ecology and Sociobiology, 60, 482–491.

Harari, A.R., Handler, A.M. & Landolt, P.J. (1999) Size-assortative mating, male choice and

female choice in the curculionid beetle Diaprepes abbreviatus. Animal Behaviour, 58,

1191–1200.

Hartl, D.L. & Clark, A.G. (2007) Principles of Population Genetics, Sinauer, Sunderland,

Massachussetts.

Hausdorf, B. (2011) Progress toward a general species concept. Evolution, 65, 923–931.

Hauswaldt, J.S., Ludewig, A.K., Vences, M. & Pröhl, H. (2011) Widespread co-occurrence of

divergent mitochondrial haplotype lineages in a Central American species of poison frog

(Oophaga pumilio). Journal of Biogeography, 38, 711–726.

89

Hero, J.M. (1989) A simple code for toe clipping anurans. Herpetological Review, 20, 66–67.

Hill, G.E (1990) Females house finch prefer colourful males: sexual selection for a condition-

dependant trait. Animal Behaviour, 40, 563–572.

Hill, G.E. (1993) Geographic variation in the carotenoid plumage pigmentation of male house

finches (Carpodacus mexicanus). Biological Journal of the Linnean Society, 49, 63–86.

Hill, G.E. & McGraw, K.J. (2006) Bird Coloration. Vol II. Function and Evolution. Harvard

University Press, Cambridge, Massachussetts.

Hodek, I. & Honěk, A. (1996) Ecology of Coccinellidae. Kluwer Academic Publishers,

Dordrecht, The Netherlands.

Hoffman, E.A. & Blouin, M.S. (2000) A review of colour and pattern polymorphisms in

anurans. Biological Journal of the Linnean Society, 70, 633–665.

Hoffman, E.A., Schueler, F.W., Jones, A.G. & Blouin, M.S. (2006) An analysis of selection

on colour polymorphism in the northern leopard frog. Molecular Ecology, 15, 2627–2641.

Hoekstra, H.E., Drumm, K.E. & Nachman, M.W. (2004) Ecological genetics of adaptive

color polymorphism in pocket mice: geographic variation in selected and neutral genes.

Evolution, 58, 1329–1341.

Hoekstra, H.E., Krenz, J.G. & Nachman, M.W. (2005) Local adaptation in the rock pocket

mouse (Chaetodipus intermedius): natural selection and phylogenetic history of

populations. Nature Heredity, 94, 217–228.

Holdridge, L.R. (1947) Determination of world plant formations from simple climatic data.

Science, 105, 367–368.

Humphries, R.B. (1979) Dynamics of a breeding frog community. PhD Thesis. The

Australian National University, Canberra, Australia.

Ihalainen, E., Lindström, L. & Mappes, J. (2007) Investigating Müllerian mimicry: predator

learning and variation in prey defenses. Journal of Evolutionary Biology, 20, 780–791.

90

Joron, M. & Mallet, J.L.B. (1998) Diversity in mimicry: paradox or paradigm? Trends in

Ecology and Evolution, 13, 461–466.

Juncá, F.A., Altig, R. & Gascon, C. (1994) Breeding biology of Colostethus stepheni, a

dendrobatid with a nontrasnported nidicolous tadpole. Copeia, 747–750.

Kapan, D.D. (2001) Three-butterfly system provides a field test of müllerian mimicry. Nature,

409, 338–340.

Kingston, J.L., Rosenthal, G.G. & Ryan, M.J. (2003) The role of sexual selection in

maintaining a colour polymorphism in the pygmy swordtail, Xiphophorus pygmaeus.

Animal Behavior, 65, 735–743.

Kirkpatrick, M. & Ravigné, V. (2002) Speciation by natural and sexual selection: models and

experiments. American Naturalist, 159, 22–35.

Kirkpatrick, M. & Barton, N.H. (2006) Chromosome inversions, local adaptation and

speciation. Genetics, 173, 419–434.

Kohda, M., Yanagisawa, Y. Sato, T. Nakaya, K. Niimura, Y. Matsumoto, K. & Ochi, H.

(1996) Geographical colour variation in cichlid fishes at the southern end of Lake

Tanganyika. Environmental Biology of Fish, 45, 237–248.

Kohlmann, B., Wilkinson, J. & Lulla, K. (2002) Costa Rica desde el espacio/Costa Rica from

Space, 1st edn. Fundacíon Neotrópica, San José.

Kohlmann, B., Solis, A., Elle, O., Soto, X. & Russo, R. (2007) Biodiversity, conservation,

and hotspot atlas of Costa Rica: a dung beetle perspective (Coleoptera: Scarabaeidae:

Scarabaeinae). Zootaxa, 1457, 1–34.

Lande, R. (1976) Natural selection and random genetic drift in phenotypic evolution.

Evolution, 30, 314–334.

Lande, R. (1981) Models of speciation by sexual selection on polygenic traits. Proceedings of

the National Academy of Sciences, USA, 78, 3721–3725.

91

Larsson, K. & Forslund, P. (1991) Environmentally induced morphological variation in the

Barnacle Goose, Branta leucopsis. Journal of Evolutionary Biology, 4, 619–636.

Lee, T.J., Speed, M.P. & Stephens, P.A. (2011) Honest Signaling and the Uses of Prey

Coloration. The American Naturalist , 178, 1–9.

Lemckert, F. (1996) Effects of toe-clipping on the survival and behaviour of the Australian

frog Crinia signifera. Amphibia-Reptilia, 17, 287–290.

Lindstedt, C., Lindström, L. & Mappes, J. (2008) Hairiness and warning colors as

components of antipredator defense: additive or interactive benefits? Animal Behaviour,

75, 1703–1713.

Lindstedt, C., Eagerb, H., Ihalainena, E., Kahilainena, A., Stevens, M. & Mappes J. (2011)

Direction and strength of selection by predators for the color of the aposematic wood tiger

moth. Behavioral Ecology, 22, 580–587.

Lindström, L., Alatalo, R.V. & Mappes, J. (1999a) Inherited avoidance towards color,

gregariousness and conspicuousness: experiments with naive and experienced predators.

Behavioural Ecology, 10, 317–322.

Lindström, L., Alatalo, R., Mappes, J., Riipi, M. & Vertainen, L. (1999b) Can aposematic

signals evolve by gradual change? Nature, 397, 249–251.

Lindström, L., Alatalo, R., Lyytinen, A. & Mappes, J. (2001a) Predator experience on cryptic

prey affects the survival of conspicuous aposematic prey. Proceedings of the Royal Society

of London series B, 268, 357–361.

Lindström, L., Alatalo, R.V., Lyytinen, A. & Mappes, J. (2001b) Strong antiapostatic

selection against novel rare aposematic. Proceedings of the National Academy of Sciences,

98 9181–9184.

Lindström, L., Alatalo, R.V., Lyytinen, A. & Mappes, J. (2004) The effect of alternative prey

on the dynamics of Batesian and Mullerian mimicries. Evolution, 58, 1294–1302.

92

Lougheed, S.C., Austin, J.D., Bogart, J.P., Boag, P.T. & Chek, A.A. (2006) Multi-character

perspectives on the evolution of intraspecific differentiation in a neotropical hylid frog.

BMC Evolutionary Biology, 6, 23.

Lu, G. & Bernatchez, L. (1999) Correlated trophic specialization and genetic divergence in

sympatric lake whitefish ecotypes (Coregonus clupeaformis) support for the ecological

speciation hypothesis. Evolution, 53, 1491–1505.

Luddecke, H. (2000) Behavioral aspects of the reproductive biology of the Andean frog

Colostethus palmatus (Amphibia: Dendrobatidae). Revista de la Academia Colombiana de

Ciencias Exactas, Fisicas y Naturales, 23, 303–316.

Maan, M.E., Seehausen, O., Soderberg, L., Johnson, L., Ripmeester, E.A.P., Mrosso, H.D.J.,

Taylor, M.I., van Dooren, T.J.M. & van Alphen, J.J.M. (2004) Intraspecific sexual

selection on a speciation trait, male coloration, in the Lake Victoria cichlid Pundamilia

nyererei. Proceedings of the Royal Society of London Series B, 271, 2445–2452.

Maan, M.E. & Cummings, M.E. (2008) Female preferences for aposematic signal components

in a polymorphic poison frog. Evolution, 62, 2334–2345.

Maan, M.E. & Cummings, M.E. (2009) Sexual dimorphism and directional sexual selection

on aposematic signals in a poison frog. Proceedings of the National Academy of Sciences

of the USA, 106, 19072–19077.

Maan, M.E. & Seehausen, O. (2011) Ecology, sexual selection and speciation. Ecology

Letters, 14, 591–602.

MacDougall, A. & Dawkins, M. S. (1998) Predator discrimination error and the benefits of

Mullerian mimicry. Animal Behaviour, 55, 1281–1288.

MacDougall, A.K. & Montgomerie, R. (2003) Assortative mating by carotenoid-based

plumage colour: a quality indicator in American goldfinches, Carduelis tristis.

Naturwissenschaften 90, 464–467.

93

Malhotra, A. & Thorpe, R.S. (1997) Size and shape variation in a Lesser Antillean anole,

Anolis oculatus (Sauria: Iguanidae) in relation to habitat. Biological Journal of the Linnean

Society, 60, 53–72.

Mallet, J. (1986) Hybrid zones in Heliconius butterflies in Panama, and the stability and

movement of warning colour clines. Heredity, 56, 191–202.

Mallet, J. & Singer, M.C. (1987) Individual selection, kin selection and the shifting balance in

the evolution of warning colours: the evidence from butterflies. Biological Journal of the

Linnean Society, 32, 337–350.

Mallet, J.L.B. & Barton, N.H. (1989) Strong natural selection in a warning color Müllerian

mimicry. Evolutionary Ecology, 13, 777–806.

Mallet, J. & Joron, M. (1999) Evolution of diversity in warning color and mimicry:

polymorphism, shifting balance and speciation. Annual Review of Ecology and

Evolutionary Sytematics, 30, 201–233.

Mallet, J., Meyer, A., Nosil, P. & Feder, J.L. (2009) Space, sympatry and speciation. Journal

of Evolutionary Biology.

Marroig, G. & Cheverud, J.M. (2001) A comparison of phenotypic variation and covariation

patterns and the role of phylogeny, ecology, and ontogeny during cranial evolution of new

world monkeys. Evolution, 55, 2576–2600.

Mappes, J. & Alatalo, R.V. (1997) Effect of novelty and gregariousness in survival of

aposematic prey. Behavioural Ecology, 8, 174–177.

Masta, S.E. & Maddison, W.P. (2002) Sexual selection driving diversification in jumping

spiders. Proceedings of the National Academy of Science USA, 99, 4442–4447.

Marchan-Rivadanera, M.R., Larsen, P.A., Phillips, C.J., Strauss, R.E. & Baker, R.J. (2012),

On the association between environmental gradients and skull size variation in the great

fruit-eating bat, Artibeus lituratus (Chiroptera: Phyllostomidae). Biological Journal of the

Linnean Society, 105, 623–634.

94

Marples, N.M., Roper, T.J. & Harper, D.G.C. (1998) Response of wild birds to novel prey:

evidence of dietary conservatism. Oikos, 83, 161–165.

Mayden, R.L. (1997) A hierarchy of species concepts: the denouement in the saga of the

species problem. In: Species: the units of biodiversity (eds. Claridge, M.F., Dawah, H.A. &

Wilson, M.R.). Chapman and Hall, London, UK, 381–424.

Mayr, E, (1947) Ecological factors in speciation. Evolution, 1, 263–288.

Mayr, E. (1963) Animal species and evolution. Museum of Comparative Zoology, Harvard

University, Cambridge, Massachussetts.

Mayr, E. (1970) Populations, species and evolution. Harvard University Press, Cambridge,

Massachussetts.

McCarthy, M.A. & Parris, K.M. (2004) Clarifying the effect of toe clipping on frogs with

Bayesian statistics. Journal of Applied Ecology, 41, 780–786.

McGuigan, K., McDonald, K., Parris, K. & Moritz, C. (1998) Mitochondrial DNA diversity

and historical biogeography of a wet forest-restricted frog (Litoria pearsoniana) from mid-

east Australia. Molecular Ecology, 7, 175–186.

McKay, J.K. & Latta, R.G. (2002) Adaptive population divergence: markers, QTL and traits.

Trends in Ecology and Evolution, 17, 285–291.

Merilaita, S & Kaitala V. (2002) Community structure and the evolution of aposematic

coloration. Ecology Letters, 5, 495–501.

Merilaita, S. (2006) Frequency-dependent predation and maintenance of prey polymorphism.

Journal of Evolutionary Biology, 19, 2022–2030.

Meyer, E. (1992) Erfolgreiche Nachzucht von Dendrobates granuliferus Taylor, 1958.

Herpetofauna, 14, 11–21.

Meyer, E. (1993) Fortplanzung und Brutpflegeverhalten von Dendrobates granuliferus

Taylor, 1958. Veröffentlichungen Naturhistorisches Museum Schleusingen, 7/8, 113–142.

95

Morey, S.R. (1990) Microhabitat selection and predation in the pacific treefrog, Pseudacris

regilla. Journal of Herpetology, 24, 292–296.

Morrison, C & Hero, J. (2003) Geographic variation in life–history characteristics of

amphibians: a review. Journal of Animal Ecology, 72, 270–279.

Müller, F. (1879) Ituna and Thyridia: a remarkable case of mimicry in butterflies. Proceedings

of the Entomological Society of London, 20–29.

Myers, C.W., Daly, J.W. & Malkin, B. (1978) A dangerously toxic new frog (Phyllobates)

used by Emberá Indians of western Colombia, with discussion of blowgun fabrication and

dart poisoning. Bulletin of the American Museum of Natural History, 161, 307–366.

Myers, C.W., Daly, J.W. & Martinez, V. (1984) An arboreal poison frog (Dendrobates) from

western Panama. American Museum Novitates, 2783, 1–20.

Narins, P.M., Hö, W. & Grabul., D.S. (2003) Bimodal signal requisite for agonistic behaviour

in a dart-poison frog, Epipedobates femoralis. Proceedings of the National Academy of

Science USA, 100, 577–580.

Narins, P.M., Grabul, D.S., Soma, K.K., Gaucher, P. & Hödl, W. (2005) Cross-modal

integration in a dart-poison frog. Proceedings of the National Academy of Science USA,

102, 2425–2429.

Noonan, B.P. & Gaucher, P. (2006) Refugial isolation and secondary contact in the dyeing

poison frog Dendrobates tinctorius. Molecular Ecology, 15, 4425–4435.

Noonan, B.P., & Comeault, A.A. (2009) The role of predator selection on polymorphic

aposematic poison frogs. Biology Letters 5, 51–54.

Noor, M.A.F. (1999) Reinforcement and other consequences of sympatry. Heredity, 83, 503–

508.

Nosil, P., Vines, T.H. & Funk, D.J. (2005) Reproductive isolation caused by natural selection

against immigrants from divergent habitats. Evolution, 59, 705–719.

96

Nosil, P & Crespi, B.J. (2006) Experimental evidence that predation promotes divergence in

adaptive radiation. Proceedings of the National Academy of Sciences of the United States

of America, 103, 9090–9095.

Obika, M. & Bagnara, J.T. (1964) Pteridines as pigments in amphibians. Science, 143, 485–

487.

Ohmer, M.E., Robertson, J.M. & Zamudio, K.R. (2009) Discordance in body size, colour

pattern, and advertisement call across genetically distinct populations in a Neotropical

anuran (Dendropsophus ebraccatus). Biological Journal of the Linnean Society, 97, 298–

313.

Olsson, M. (1993) Male preference for larger females and assortative mating for body size in

the sand lizard (Lacerta agilis). Behavioral Ecology and Sociobiology, 32, 337–341.

van Oppen, M.J.H., Turner, G.F., Rico, C., Robinson, R.L., Deutsch, J.C., Genner, M.J. &

Hewitt, G.M. (1998) Assortative mating among rock-dwelling cichlid fishes supports high

estimates of species richness from Lake Malawi. Molecular Ecology, 7, 991–1001.

Orr, H.A. & Turelli, M. (2001) The evolution of postzygotic isolation: Accumulating

Dobzhansky-Muller incompatibilities. Evolution, 55, 1085–1094.

Palumbi, S.R., Martin, A., Romano, S., McMillan, W.O., Stice, L. & Grabowski, G. (1991) A

simple fool’s guide to PCR. Version 2.0. University of Hawaii, Honolulu, HI.

Panhuis, T.M., Butlin, R., Zuk, M. & Tregenza, T. (2001) Sexual selection and speciation,

Trends in Ecology and Evolution, 16, 364–371.

Parris, K.M. & McCarthy, M.A. (2001) Identifying effects of toe clipping on anuran return

rates: the importance of statistical power, Amphibia-Reptilia, 22, 275-289.

Parsons, Y.M. & Shaw, K.L. (2001) Species boundaries and genetic diversity among

Hawaiian crickets of the genus Laupala identified using amplified fragment length

polymorphism. Molecular Ecology, 10, 1765–1772.

97

Pinheiro, C.E.G. (1996) Palatability and escaping ability in Neotropical butterflies: tests with

wild kingbirds (Tyrannus melancholicus, Tyrannidae). Biological Journal of the Linnean

Society, 59, 351–365.

Peakall, R. & Smouse, P.E. (2006) GENALEX 6: genetic analysis in Excel. Population genetic

software for teaching and research. Molecular Ecology Notes, 6, 288–295.

Posada, D. (2008) jModelTest: phylogenetic model averaging. Molecular Biology and

Evolution, 25, 1253–1256.

Poulton, E.B. (1890) The Colour of Animals: their Meaning and Use. Kegan Paul, Trench,

Trubner, London, UK.

Povolný D. (1999) Mendelian dihybridism in central European polymorphic populations of

aposematic burnet-moth Zygaena ephialtes (Linnaeus, 1767) and its Müllerian-Batesian

mimicry (Lepidoptera, Zygaenidae). Acta Universitatis Agriculturae et Silviculturae

Mendelianae Brunensia, 47, 41–66.

Preston, B.T., Stevenson, I.R., Pemberton, J.M., Coltman, D.W. & Wilson, K. (2005) Male

mate choice influences female promiscuity in Soay sheep. Proceedings of the Royal

Society of London Series B, Biological Sciences, 272, 365–373.

Price, J.J., Friedman, N.R. & Omland, K.E. (2007) Song and plumage evolution the new

world orioles (Icterus) show similar lability and convergence in patterns. Evolution, 61,

850–863.

Pritchard, J.K., Stephens, M. & Donnelly, P. (2000) Inference of population structure using

multilocus genotype data. Genetics, 155, 945–959.

Pröhl, H. (1997) Territorial behavior of the strawberry poison–dart frog, Dendrobates

pumilio. Amphibia-Reptilia, 18, 437–442.

Pröhl, H. & Hödl, W. (1999) Parental investment, potential reproductive rates and mating

system in the strawberry poison-dart frog Dendrobates pumilio. Behavioural Ecology and

Sociobiology, 46, 215–220.

98

Pröhl, H. & Berke, O. (2001) Spatial distribution of male and female strawberry poison frogs

and their relation to female reproductive resources. Oecologia, 129, 534–542.

Pröhl, H. (2003) Variation in male calling behaviour and relation to male mating success in

the strawberry poison frog (Dendrobates pumilio). Ethology, 109, 273–290.

Pröhl, H. (2005) Territorial behavior in dendrobatid frogs. Journal of Herpetology, 39, 354–

365.

Pröhl, H., Koshy, R.A., Mueller, U., Rand, A.S. & Ryan, M.J. (2006) Geographic variation of

genetic and behavioral traits in northern and southern túngara frogs. Evolution 60, 1669–

1679.

Prudic, K., Skemp, A.K. & Papaj, D.R. (2007) Aposematic coloration, luminance contrast,

and the benefits of conspicuousness. Behavioural Ecology, 18, 41–46.

de Queiroz, A. & Wimberger, P.H. (1993) The usefulness of behavior for phylogeny

estimation: levels of homoplasy in behavioral and morphological characters. Evolution 47,

46–60.

Ratcliffe, J.M. & Nydam, M.L. (2008) Multimodal warning signals for a multiple predator

world. Nature 455, 96–99.

Raymond, M. & Rousset, F. (1995a) GENEPOP (version 1.2): a population genetics software

for exact test and ecumenicism. Journal of Heredity, 86, 248–249.

Raymond, M. & Rousset, F. (1995b) An exact test of population differentiation. Evolution,

49, 1280–1283.

Reaser, J.K. & Dexter, R.E. (1996) Rana pretiosa (spotted frog). Toe clipping effects.

Herpetological Review, 27, 195–196.

Reynolds, R.G. & Fitzpatrick, B.M. (2007) Assortative mating in poison-dart frogs based on

an ecologically important trait. Evolution, 61, 2253–2259.

99

Richards, C.L. & Knowles, L.L. (2007) Tests of phenotypic and genetic concordance and

their application to the conservation of Panamanian golden frogs (Anura, Bufonidae).

Molecular Ecology, 16, 3119–3133.

Richards-Zawacki, C.L. & Cummings, M.E. (2011) Intraspecific reproductive character

displacement in a polymorphic poison dart frog, Dendrobates pumilio. Evolution, 65, 259–

267.

Riipi, M., Alatalo, R., Lindström, L. & Mappes, J. (2001) Multiple benefits of gregariousness

cover detectability costs in aposematic aggregations. Nature, 413, 512–514.

Roberts, J.D. & Wardell-Johnson, G. (1995) Call differences between peripheral isolates of

the Geocrinia rosea complex (Anura: Myobatrachidae) in Southwestern Australia. Copeia,

4, 899–906.

Robertson, J.M. (2008) Genetic and phenotypic diversity patterns in two polymorphic,

neotropical anurans: biogeography, gene flow and selection. PhD thesis, Cornell

University, Ithaca, New York.

Robertson, J.M. & Robertson, A.D. (2008) Spatial and temporal patterns of phenotypic

variation in a Neotropical frog. Journal of Biogeography, 35, 830–843.

Robertson, J.M., Duryea, M.C. & Zamudio, K.R. (2009), Discordant patterns of evolutionary

differentiation in two Neotropical treefrogs. Molecular Ecology, 18, 1375–1395.

Ronquist, F. & Huelsenbeck, J.P. (2003) MrBayes 3: Bayesian phylogenetic inference under

mixed models. Bioinformatics, 19, 1572–1574.

Roper, T.J. & Redstone, S. (1987) Conspicuousness of distasteful prey affects the strength

and durability of one-trial avoidance learning. Animal Behaviour, 35, 739–747.

Roper, T.J. (1990) Responses of domestic chicks to artificially colored insect prey: effects of

previous experience and background color. Animal Behaviour, 39, 466–473.

Rosenblum, E.B., Hoekstra, H. E. & Nachman, M.W. (2004) Adaptive reptile color variation

and the evolution of the MCIR gene. Evolution, 58, 1794–1808.

100

Rosso, A., Castellano, S. & Giacoma, C. (2004) Ecogeographic analysis of morphological and

life-history variation in the Italian treefrog. Evolutionary Biology 18, 303–321.

Roulin, A. (2004) The evolution, maintenance and adaptive function of genetic colour

polymorphism in birds. Biological Reviews, 79, 815–848.

Rousset, F. (2008) GENEPOP’007: a complete re-implementation of the GENEPOP software for

Windows and Linux. Molecular Ecology Resources, 8, 103–106.

Rowe, C., Lindström, L. & Lyytinen, A. (2004) The importance of pattern similarity between

Müllerian mimics in predator avoidance learning. Proceedings of the Royal Society of

London series B, 271, 407–413.

Rowland, H.M., Ihalainen, E., Lindström, L., Mappes, J. & Speed, M.P. (2007) Co-mimics

have a mutualistic relationship despite unequal defense levels. Nature, 448, 64–66.

Rudh, A., Rogell, B. & Höglund, J. (2007) Non-gradual variation in colour morphs of the

strawberry poison frog Dendrobates pumilio: genetic and geographical isolation suggest a

role for selection in maintaining polymorphism. Molecular Ecology, 16, 4284–4294.

Rueffler, C., Van Dooren, T.J.M., Leimar, O. & Abrams, P.A. (2006) Disruptive selection and

then what? Trends in Ecology and Evolution, 21, 238–245.

Rutowski, R.L., Macedonia, J.M., Morehouse, N.I. & Taylor-Taft, L. (2005) Pterin pigments

amplify iridescent ultraviolet signal in males of the orange sulphur butterfly, Colias

eurytheme . Proceedings of the Royal Society series B, 272, 2329–2335.

Ruxton, G.D., Sherratt, T.N. & Speed, M.P. (2004) Avoiding attack. Evolutionary ecology of

crypsis, warning signals and mimicry. Oxford University Press, New York.

Ruxton, G.D., Speed, M. & Broom, M. (2007) The importance of initial protection of

conspicuous mutants for the coevolution of defense and aposematic signaling of the

defense: a modeling study. Evolution, 61, 2165–2174.

101

Ryan, M.J. & Rand, A.S. (1990) The sensory basis of sexual selection for complex calls in the

túngara frog, Physalaemus pustulosus (sexual selection for sensory exploitation).

Evolution, 44, 305–314.

Ryan, M.J. & Wilczynski, W. (1991) Evolution of intraspecific variation in the advertisement

call of a cricket frog (Acris crepitans, Hylidae). Biological Journal of the Linnean Society,

44, 249–271.

Ryan, M.J., Rand, A.S. & Weigt, L.A. (1996) Allozyme and advertisement call variation in

the túngara frog, Physalaemus pustulosus. Evolution, 50, 2435–2453.

Ryan, M.J. (1999) Sexual selection, receiver biases, and the evolution of sex differences.

Science, 28, 1999–2003.

Sanderson, N., Szymura, J.M. & Barton, N.H. (1992) Variation in mating call across the

hybrid zone between fire-bellied toads Bombina bombina and B. variegata. Evolution, 46,

595–607.

Sandre, S.L., Stevens, M. & Mappes, J. (2010) The effect of predator appetite, prey warning

coloration and luminance on predator foraging decisions. Behaviour, 147, 1121–1143.

Santos, J. C., Coloma, L. A. & Cannatella, D. C. (2003) Multiple, recurring origins of

aposematism and diet specialization in poison frogs. Proceedings of the National Academy

of Science USA, 100, 12792–12797.

Santos, J.C., Coloma, L.A., Summers, K., Caldwell, J.P., Ree, C. & Cannatella, R.D.

(2009) Amazonian amphibian diversity is primarily derived from late miocene

Andean lineages. PLoS Biology, 7, e56.

Saporito, R.A., Garraffo, H.M., Donnelly, M.A., Edwards, A.L., Longino, J.T. & J.W. Daly.

(2004) Formicine ants: an arthropod source for the pumiliotoxin alkaloids of dendrobatid

poison frogs. Proceedings of the National Academy of Science USA, 101, 8045–8050.

102

Saporito, R.A., Norton, R.A., Garraffo, H.M., Spande, T.F. & Daly, W.D. (2007a) Oribatid

mites as a major dietary source for alkaloids in poison frogs. Proceedings of the National

Academy of Science USA, 104, 8885–8890.

Saporito, R.A, Zuercher, R., Roberts, M., Gerow, K.G. &. Donnelly, M.A (2007b)

Experimental Evidence for Aposematism in the Dendrobatid Poison Frog Oophaga

pumilio.

Saporito, R.A., Donnelly, M.A., P. Jain, Garraffo, H.M., Spande, T. F. & Daly, J. W. (2007c)

Spatial and temporal patterns of alkaloid variation in the poison frog Oophaga pumilio in

Costa Rica and Panama over 30 years. Toxicon, 50, 757–778.

Savage, J.M. (1968) The dendrobatid frogs of Central America. Copeia, 4, 745–776.

Savage, J.M. (2002) The Amphibians and Reptiles of Costa Rica: A Herpetofauna Between

Two Continents, Between Two Seas. The University of Chicago Press, Chicago, Illinois.

Schäuble, C.S. (2004) Variation in body size and sexual dimorphism across geographical and

environmental space in the frogs Limnodynastes tasmaniensis and L. peronii. Biological

Journal of the Linnean Society, 82, 39–56.

Schlenoff, D.H. (1984) Novelty: a basis for generalisation in prey selection. Animal

Behaviour, 32, 919–921.

Schliewen, U., Kocher, T., McKaye, K.R., Seehausen, O. & Tautz, D. (2006) Evidence for

sympatric speciation? Nature, 444, E12–E13.

Schluter, D. (2009) Evidence for ecological speciation and its alternative. Science, 323, 737–

741.

Schuler, W. & Hesse, E. (1985) On the function of warning coloration, a black and yellow

pattern inhibits prey attack by naıve domestic chicks. Behavioural Ecology and

Sociobiology, 16, 245–255.

103

Seehausen, O. & van Alphen, J.J.M. (1998) The effect of male coloration on female mate

choice in closely related Lake Victoria cichlids (Haplochromis nyererei complex).

Behavioral Ecology and Sociobiology , 42, 1–8.

Servedio, M.R. (2000) The effects of predator learning, forgetting, and recognition errors on

the evolution of warning coloration. Evolution, 54, 751–763.

Sherratt, T.N. & Beatty, C.D. (2003) The evolution of warning signals as reliable indicators of

prey defense. American Naturalist, 162, 377–389.

Siddiqi, A., Cronin, T.W., Loew, E.R., Vorobyev, M. & Summers, K. (2004) Interspecific

and intraspecific views of color signals in the strawberry poison frog Dendrobates pumilio.

Journal of Experimental Biology, 207, 2471–2485.

Siefferman, L. & Hill, G. E. (2003) Structural and melanin coloration indicate parental effort

and reproductive success in male eastern bluebirds. Behavioral Ecology, 14, 855–861.

Silverstone, P.A. (1973) Observations on the behavior and ecology of a Colombian poison-

arrow frog, the kokoepá (Dendrobates histrionicus Berthold). Herpetologica, 29, 295–301.

Silverstone, P.A. (1975) A revision of the poison-arrow frogs of the genus Dendrobates.

Scientific Bulletin Of The Natural History Museum, Los Angeles County, 21, 1–55.

Slabbekoorn, H. & Smith, T.B. (2002) Bird song, ecology and speciation. Philosophical

Transactions of the Royal Society of London Series B, 357, 493–503.

Smith, S.M. (1977) Coral-snake pattern recognition and stimulus generalisation by naive great

kiskadees (Aves: Tyrannidae). Nature, 265, 535–536.

Smith, S.M. (1975) Innate recognition of coral snake pattern by a possible avian predator.

Science, 187, 759–760.

Smith, M.J., Roberts, J.D., Hammond, T.J. & Davis, R.A. (2003) Intraspecific variation in the

advertisement call of the sunset frog Spicospina flammocaerula (Anura: Myobatrachidae):

a frog with a limited geographic distribution. Journal of Herpetology, 37, 285–291.

104

Speed, M.P. (1993) Muellerian mimicry and the psychology of predation. Animal Behaviour,

45, 571–580.

Speed, M.P. & Ruxton, G.D. (2007) How bright and how nasty: explaining diversity in

warning signal strength. Evolution, 61, 623–635.

Stalker, H.D. & Carson, H.L. (1947) Morphological variation in natural populations of

Drosophila robusta Sturtevant. Evolution, 1, 237–248.

Steffen, J.E. & McGraw, K.J. (2007) Contributions of pterin and carotenoid pigments to

dewlap coloration in two anole species. Comparative Biochemistry and Physiology B, 146,

42–46.

Stimson, J. & Berman, M. (1990) Predator induced colour polymorphism in Danaus

plexippus L. Lepidoptera: Nymphalidae) in Hawaii. Heredity, 65, 401–406.

Štys, P. (2004) Phylogenetic constraints on occurence of warning coloration in the European

Heteroptera. In: Abstract Book ISBE 2004, 211. Jyväskyla, Finland: University of

Jyväskyla.

Summers, K. (1989) Sexual selection and intra-female competition in the green poison-dart

frog, Dendrobates auratus. Animal Behaviour, 37, 797–805.

Summers, K., Bermingham, E., Weigt, L., McCafferty, S., & Dahlstrom, L. (1997)

Phenotypic and genetic divergence in three species of dart-poison frogs with contrasting

parental behavior. Journal of Heredity, 88, 8–13.

Summers, K., Symula, R., Clough, M. & Cronin, T. (1999) Visual mate choice in poison

frogs. Proceedings of the Royal Society of London Series B, 266, 2141–2145.

Summers, K. & Clough, M. (2001) The evolution of coloration and toxicity in the poison frog

family (Dendrobatidae). Proceedings of the National Academy of Sciences of the United

States of America, 98, 6227–6232.

105

Summers, K. & McKeon, C.S. (2004) The evolutionary ecology of phytotelmata use in

Neotropical poison frogs. Miscellaneous Publications, Museum of Zoology, University of

Michigan, 193, 55–73.

Swofford, D.L. (2002) PAUP*: phylogenetic analysis using parsimony, version 4b10. Sinauer

Associates, Sunderland, MA.

Takada, W., Sakata, T., Shimano, S., Enami, Y., Mori, N., Nishida, R. & Kuwahara. Y.

(2005) Scheloribatid mites as the source of pumiliotoxins in dendrobatid frogs. Journal of

Chemical, Ecology, 31, 2403–2415.

Tamura, K. Dudley, J., Nei, M. & Kumar, S. (2007) MEGA4: Molecular Evolutionary

Genetics Analysis (MEGA) software version 4.0. Molecular Biology and Evolution, 24,

1596–1599.

Templeton, A.R. (1980) The theory of speciation via the founder principle. Genetics, 94,

1011–1038.

Templeton, A.R. (1981) Mechanisms of speciation: a population genetic approach. Annual

Review of Ecology and Systematics, 12, 23– 48.

Thomas, R.J., Marples, N.M., Cuthill, I.C., Takahashi, M. & Gibson, E.A (2003) Dietary

conservatism may facilitate the initial evolution of aposematism. Oikos, 101, 458–466.

Thorpe, R.S., Malhotra, A., Black, H., Daltry, J.C. & Wüster, W. (1995) Relating geographic

pattern to phylogenetic process. Proceedings of the Royal Society of London Series B, 349,

61–68.

Thorpe, R.S & Stenson, A.G. (2003) Phylogeny, paraphyly and ecological adaptation of the

colour and pattern in the Anolis roquet complex on Martinique. Molecular Ecology, 12,

117–132.

Tordoff, W. (1980) Selective predation of gray jays, Perisoreus canadensis, upon boreal

chorus frogs, Pseudacris triseriata. Evolution, 24, 1004–1008.

106

Tullberg, B.S., Merilaita, S. & Wiklund, C. (2005) Aposematism and crypsis combined as a

result of distance dependence: functional versatility of the colour pattern in the swallowtail

butterfly larva. Proceedings of the Royal Society series B, 272, 1315–1321.

Turelli, M., Barton, N.H. & Coyne, J.A. (2001) Theory and speciation. Trends in Ecology &

Evolution, 16, 330–343.

Turner, J.R.G. & Speed, M.P. (1996) Learning and memory in mimicry. Simulations of

laboratory experiments. Philosophical Transactions of the Royal Society of London Series

B, 351, 1157–1170.

Turner, J.R.G. & Mallet, J.L.B. (1996) Did forest islands drive the diversity of warningly

coloured butterflies? Biotic drift and the shifting balance. Phylosophical Transactions of

the Royal Society of London Series B, 351, 835–845.

Vences, M., Kosuch, J. Boistel, R. Haddad, C.F.B., La Marca, E., Lotters S. & Veith. M.

(2003) Convergent evolution of aposematic coloration in neotropical poison-dart frogs: a

molecular phylogenetic perspective. Organismal Diversity and Evolution, 3, 215–226.

Vercken, E., Massot, M., Sinervo, B. & Clobert, J. (2007) Colour variation and alternative

reproductive strategies in females of the common lizard Lacerta vivipara. Journal of

Evolutionary Biology, 20, 221–232.

Verrell, P.A. (1985) Male mate choice for large, fecund females in the red-spotted newt,

Notophthalmus viridescens: how is size assessed? Herpetologica, 41, 382–386.

Waichman, A.V. (1992) An alphanumeric code for toe clipping amphibians and reptiles.

Herpetological Review, 23, 19–21.

Wallace, A. R. (1867) Mimicry and other protective resemblances among animals.

Wang, I.J. & Shaffer, H.B. (2008) Rapid color evolution in an aposematic species: a

phylogenetic analysis of color variation in the strikingly polymorphic strawberry poison-

dart-frog. Evolution, 62, 2742–2759.

107

Wang, I.J. & Summers, K. (2010) Genetic structure is correlated with phenotypic divergence

rather than geographic isolation in the highly polymorphic strawberry poison-dart frog.

Molecular Ecology, 19, 447–458.

Wedell, N. & Sandberg, T. (1995) Female preference for large males in the bushcricket

Requena sp. 5 (Orthoptera: Tettigoniidae). Journal of Insect Behavior, 8, 513–522.

Weigt, L.A., Crawford, A.J., Rand, A.S. & Ryan, M.J. (2005) Biogeography of the tungara

frog, Physalaemus pustulosus: a molecular perspective. Molecular Ecology, 14, 3857–

3876.

Weir, B.S. & Cockerham, C.C. (1984) Estimating F-statistics for the analysis of population

structure. Evolution, 38, 1358–1370.

Welch, J.J. (2004) Accumulating Dobzhansky-Muller incompatibilities: Reconciling theory

and data. Evolution, 58, 1145–1156.

Wellborn, G.A. & Broughton, R.E. (2008) Diversification on an ecologically constrained

adaptive landscape. Molecular Ecology, 17, 2927–2936.

Wells, K.D. (1978) Courtship and parental behavior in a Panamanian poison-arrow frog

(Dendrobates auratus). Herpetologica, 34, 148–155.

Wells, K.D. (1980a) Behavioral ecology and social organization of a dendrobatid frog

(Colostethus inguinalis). Behavioral Ecology and Sociobiology, 6, 199–209.

Wells, K.D. (1980b) Evidence for growth of tadpoles during parental transport in Colostethus

inguinalis. Journal of Herpetology, 14, 428–430.

Wells, K.D. (1980c) Social behavior and communication of a dendrobatid frog (Colostethus

trinitatis). Herpetologica, 36, 189–199.

West-Eberhard, M.J. (1983) Sexual selection, social competition, and speciation. Quarterly

Review of Biology, 58, 155–183.

108

Weygoldt, P. (1980) Complex brood care and reproductive behavior in captive poison-arrow

frogs, Dendrobates pumilio, O. Schmidt. Behavioral Ecology and Sociobiology, 7, 329–

332.

Weygoldt, P. (1987) Evolution of parental care in dart poison frogs (Amphibia: Anura:

Dendrobatidae). Zeitschrift für Zoologische Systematik und Evolutionsforschung, 25, 51–

7.

White, M.J.D. (1968) Models of speciation. Science, 159, 1065–1070.

Wiens, J.J. & Penkrot, T.A. (2002) Delimiting species using DNA and morphological

variation and discordant species limits in spiny lizards (Sceloporus). Systematic Biology,

51, 69– 91.

van Wijngaarden, R. (1985) Some remarks on Dendrobates granuliferus. Lacerta, 43, 181–

183.

van Wijngaarden, R. & Bolaños, F. (1992) Parental care in Dendrobates granuliferus (Anura:

Dendrobatidae) with a description of the tadpole. Journal of Herpetology, 26, 102–105.

van Wijngaarden, R. & van Gool. S. (1994) Site fidelity and territoriality in the dendrobatid

frog Dendrobates granuliferus. Amphibia-Reptilia, 15, 171–181.

Williamson, I. & Bull, C.M. (1996) Population ecology of the Australian frog Crinia

signifera: adults and juveniles. Wildlife Research, 23, 249–266.

Wollenberg, K., Noonan, B.P. & Lötters, S. (2006) Polymorphism versus species richness-

systematics of large Dendrobates from the eastern Guiana Shield (Amphibia:

Dendrobatidae). Copeia. 4, 623–629.

Wright, S. (1931) Evolution in Mendelian populations. Genetics, 16, 97–159.

Yosef, R. & Whitman, D.W. (1992) Predator exaptations and defensive adaptations in

evolutionary balance: no defense is perfect. Evolutionary Ecology, 6, 527–536.

Zamudio, K.R. & Greene, H.W. (1997) Phylogeography of the bushmaster (Lachesis muta:

109

Viperidae): implications for Neotropical biogeography, systematics, and conservation.

Biological Journal of the Linnean Society, 62, 421–442.

Zimmermann, H., & E. Zimmermann (1988) Etho-Taxonomie und zoogeographische

artengruppenbildung bei pfeilgiftfröschen (Anura: Dendrobatidae). Salamandra, 24, 125–

160.

110

Additional Information

1. Sequences accession numbers

List of all sequences used in the study. Samples specifically sequenced for the present work

are marked with *. Table reports, sampled animal, sampling site (and code) and the haplotype

code (with relative accession number). For samples sharing the same haplotype identified in

the present study, the accession number is reported only once.

ID Sampling site [code] Cyt b (AccN°) 16S (AccN°)

O. granulifera 1-CH* Charcos [CH] HE775251 (H1) HE804764 (1)

O. granulifera 2-CH* Charcos [CH] HE775251 (H1) HE804764 (1)

O. granulifera 3-CH* Charcos [CH] HE775252 (H2) HE804765 (2)

O. granulifera 4-VE* Ventana [VE] HE775253 (H3)

O. granulifera 5-VE* Ventana [VE] HE775253 (H3)

O. granulifera 6-PA* Palmar [PA] HE775254 (H4) HE804766 (3)

O. granulifera 7-PA* Palmar [PA] HE775255 (H5) HE804766 (3)

O. granulifera 8-PA* Palmar [PA] HE775256 (H6) HE804766 (3)

O. granulifera 9-FS* Firestone [FS] HE775257 (H7) HE804767 (4)

O. granulifera 10-FS* Firestone [FS] HE775257 (H7)

O. granulifera 11-FS* Firestone [FS] HE775258 (H8) HE804767 (4)

O. granulifera 12-BA* Baru [BA] HE775258 (H8) HE804767 (4)

O. granulifera 13-BA* Baru [BA] HE775259 (H9)

O. granulifera 14-BA* Baru [BA] HE775260 (H10)

O. granulifera 15-DO* Dominical [DO] HE775257 (H7) HE804767 (4)

O. granulifera 16-DO* Dominical [DO] HE775257 (H7) HE804767 (4)

O. granulifera 17-DO* Dominical [DO] HE775257 (H7)

111

O. granulifera 18-MA* Matapalo [MA] HE775257 (H7)

O. granulifera 19-MA* Matapalo [MA] HE775257 (H7)

O. granulifera 20-MA* Matapalo [MA] HE775258 (H8) HE804767 (4)

O. granulifera 21-MA* Matapalo [MA] HE775262 (H12)

O. granulifera 22-PL* Portalon [PL] HE775257 (H7) HE804767 (4)

O. granulifera 23-PL* Portalon [PL] HE775257 (H7)

O. granulifera 24-PL* Portalon [PL] HE775261 (H11) HE804767 (4)

O. granulifera 25-SA* Savegre [SV] HE775257 (H7) HE804767 (4)

O. granulifera 26-DA* Damitas [DA] HE775257 (H7) HE804768 (5)

O. granulifera 27-DA* Damitas [DA] HE775257 (H7)

O. granulifera 28-DA* Damitas [DA] HE775257 (H7)

O. granulifera Wang0925-FC Fila Chonta [FC] HQ841077 (H7)

O. granulifera Wang0927-FC Fila Chonta [FC] HQ841078 (H16)

O. granulifera Wang0928-FC Fila Chonta [FC] HQ841079 (H15)

O. granulifera Wang0929-CN Cerro Nara [CN] HQ841080 (H15)

O. granulifera Wang0930-CN Cerro Nara [CN] HQ841081 (H7)

O. granulifera Wang0931-CN Cerro Nara [CN] HQ841082 (H15)

O. granulifera Wang0932-DM Rio Damitas [DM] HQ841083 (H15)

O. granulifera Wang0933-DM Rio Damitas [DM] HQ841084 (H7)

O. granulifera Wang0934-DM Rio Damitas [DM] HQ841085 (H15)

O. granulifera Wang0939-SV Rio Savegre [SV] HQ841086 (H7)

O. granulifera Wang0940-SV Rio Savegre [SV] HQ841087 (H7)

O. granulifera Wang0941-SV Rio Savegre [SV] HQ841088 (H7)

O. granulifera Wang1007-BR Rio Baru [BR] HQ841089 (H10)

O. granulifera Wang1009-BR Rio Baru [BR] HQ841090 (H8)

O. granulifera Wang1010-BR Rio Baru [BR] HQ841091 (H10)

O. granulifera Wang1012-PB Piedras Blancas [PB] HQ841092 (H14)

O. granulifera Wang1014-PB Piedras Blancas [PB] HQ841093 (H13)

112

O. granulifera Wang1015-PB Piedras Blancas [PB] HQ841094 (H14)

O. granulifera Wang1016-TE Río Térraba [TE] HQ841095 (H14)

O. granulifera Wang1017-TE Río Térraba [TE] HQ841096 (H14)

O. granulifera Wang1018-TE Río Térraba [TE] HQ841097 (H14)

O. granulifera Wang1019-DB Drake Bay [DB] HQ841098 (H14)

O. granulifera Wang1020-DB Drake Bay [DB] HQ841099 (H14)

O. granulifera Wang1021-DB Drake Bay [DB] HQ841100 (H14)

O. granulifera Wang1022-CO Corcovado [CO] HQ841101 (H14)

O. granulifera Wang1024-CO Corcovado [CO] HQ841102 (H14)

O. granulifera Wang1026-CO Corcovado [CO] HQ841103 (H14)

Oophaga histrionica _ HQ841104 DQ502032

Oophaga sylvatica _ HQ290567 AY364569

Oophaga arborea _ DQ502036

Oophaga lehmanni _ DQ502034

Oophaga pumilio-N Nicaragua DQ502235

Oophaga pumilio-S Panama EU342663

Oophaga speciosa _ DQ502037

Oophaga vicentei _ DQ502167

Dendrobates leucomelas

Dendrobates tinctorius

_ AF124119

DQ768797

113

2. Population structure ∆K results

114

Ackowledgments

Thanks to the Volkswagen foundation for the very generous resources provided.

Thanks to the MINAE for releasing research permits.

Thanks to Heike (Pröhl) for giving an opportunity to an unknown student from Italy.

Thanks to Nick (Mundy) for letting me enter in his lab at the University of Cambridge.

Thanks to Eberhard Meyer for providing information on some frog populations’ locations.

Thanks to Adriana Bellati for advising on molecular data analysis.

Thanks to Sabine Sippel for her collaboration in the lab.

Thanks to Sönke von den Berg for helping with graphics.

Thanks to Cristina who has been an important support during several years.

Thanks to my parents who are still doing an effort to understand where I want to go.

Thanks to all people who helped me out during the crazy time in the field.

Thanks to the tiny granular poison frogs for existing.

Thanks to the lethal snakes of Costa Rica for leaving me alive and able to conclude this work.

Thanks to Costa Rica cause I realized there is not just one way to live.

Thanks to Germany and England cause I understood cultural differences may be funny.

Thanks to Julia for showing me one does not need many things to be happy.