This item is only available as the following downloads:
DIEL ACTIVITY OF THE POISON FROG OOPHAGA PUMILIO IN RESPONSE TO WEATHER CONDITIONS IN BOCAS DEL TORO, PANAMA BY RALUCA CRUCEANU A Thesis Submitted to the Division of Natural Sciences New College of Florida in partial fulfillment of the requirements for the degree Bachelor of Arts Under the sponsorship of Dr. Alfred Beulig Sarasota, Florida May, 2012
Acknowledgements I would like to extend a warm thank you to Dr. Beu lig for keeping me positive with his witty humour and lively manner, to my fami ly for their support, Dr. Diana Weber for encouraging me to continue my research on these frogs, Peter Lahanas, my former roommate Jo, Carlos Ormond for making sure t hat my site remained untouched throughout my study, Sivens Glaude for the flowers every birthday, my committee members for the interest in my work and last but no t least, the frogs, which I will dearly miss.
Table of Contents Acknowledgements . . . .... ii Table of Contents . . . .. iii Abstract . iv Introduction . . . .... 1 Amphibians .. . . . .... 1 Poison frogs . . . . .... 2 Oophaga pumilio . . ... ... 3 Taxonomy . . .. .. 5 Habitat . . . . .... 5 Aposematism . .. ... 6 Genetics . . . 10 Skin toxins .. . . 10 Diet and Feeding . .. .... 11 Reproduction ... 11 Oviposition ... .... 12 Parental Care . .. . . .. 12 Feeding the young . . . . 14 Territoriality .. . .... .. 15 Vocalization . .... .. .. 21 Activity Levels . .. .. 25 Methods .. . 27 Results . . 28 Total Activity and Calling Activity . . .... 28 Temperature . 29 Humidity . . . 30 Light Intensity . .. 31 Rainfall . 32 Sound Waveforms and Spectrograms .. .. ... 37 Discussion .... 40 References . .. . 46
DIEL ACTIVITY OF THE POISON FROG OOPHAGA PUMILIO IN RESPONSE TO WEATHER CONDITIONS IN BOCAS DEL TORO, PANAMA Raluca Cruceanu New College of Florida, 2012 ABSTRACT Oophaga pumilio is the most polymorphic poison frog. The greenish -black spotted frog with yellow undersides can only be fou nd on Isla Coln and is entirely specific to the area. A novelty among anurans, fem ale Oophaga pumilio feed the larvae with unfertilized eggs. This species exhibits soph isticated parental care by both males and females. This study is the first in this line of research th at has shown a statistically significant difference between hourly time interval s for diel calling activity in the dry season, and a positive correlation between calling activity and light intensity. There was no significant difference between total activity an d temperature, total activity and humidity, total activity and light intensity, total activity and rainfall, calling activity and temperature, calling activity and humidity, calling activity and rainfall. Single advertisement calls, mating calls, territorial call s and courtship calls fell in the same frequency range, but had different waveforms. Dr. Alfred Beulig Division of Natural Sciences
Amphibians An amphibian can be either an animal that has spent part of its life in an aquatic environment then changed to an aquatic adult or an animal alternating between aquatic and terrestrial environments. Some amphibians are aquatic throughout their lives and others neither enter water nor have aquatic life st ages. Amphibians can best be defined as quadrupedal vertebrates with two occipital condyles on their skull and one sacral vertebra and glandular skin lacking epidermal surfaces (scal es, feathers and hair). Amphibians are classified intro three groups: anurans (tailless am phibians; frogs and toads), urodeles (tailed amphibians; salamanders) and caecilians (Gy mnophiona). Anurans have long hindlimbs and short stiff bodies that do not bend w hen they walk; urodeles have forelimbs and hindlimbs of equal size and move with lateral undulations; caecilians are limbless and employ snakelike locomotion (Pough et al., 2009). A common trait among amphibians is the production of eggs with a water-p ermeable jelly coat from which aquatic larvae hatch (Duellman and Trueb, 1986). E xtant amphibians (Lissamphibia) first appeared in the Jurassic, with clear differen tiations since mid-Cretaceous (144 65 million years ago) (Duellman and Trueb, 1986). The se vertebrates have provided extensive research in developmental biology and ver tebrate biology. Their transition from aquatic larvae to terrestrial adults shows a r emarkable mode of adaptation to changes in the environment. Reproductive strategie s of amphibians are closely related to environmental conditions. The diversity of these s trategies within species and within populations is reflected in their evolutionary and ecological differences. Rainfall is the primary factor in breeding activity of amphibians. In tropical environments, breeding seasons can be time-dependent or continuous. Caeci lians reproduce semi-anually,
urodeles annually or biennially, and anurans in the wet tropics have continuous reproduction. Location of mates may be visual, olf actory, auditory, tactile or a combination. Courtship is an integral part in mate selection. Auditory cues such as vocalization in anurans attract mates to breeding s ites. In the majority of amphibian populations, risk of predation on eggs and larvae i s high (Duellman and Trueb, 1986). Multiple small clutches reduce reproductive effort and enhance survivorship of early developmental stages. An anuran in a temperate zon e may reach sexual maturity in two years whereas an anuran in the tropical equatorial zone may reach maturity in one year. Anurans have provided a wealth of information in th e field of sexual selection due to their territorial defense mechanisms, acoustic comm unication, courtship, mating and reproductive strategies. Among the 6000 or so spec ies of amphibians known today, nearly 5400 are anurans (Pough et al., 2009).. Anurans can be terrestrial, aquatic, fossorial or a rboreal. They are particularly fascinating due to their various modes of reproduct ion and adaptability to a wide range of habitats. Poison frogs (Aromobatidae, Dendrobatidae) are high ly specialized anurans in the neotropics (South and Central America). Many are p olymorphic (many-shaped) regarding their color and pattern. Most are diurna l and live in forests, on the ground or along streams. Half exhibit brilliant colors while the other half exhibit cryptic color patterns. Aposematic frogs secrete toxins in their skin, mostly derived from their prey, to escape predators and deter enemies. Camouflage and anti-predation signals are especially puzzling as defense strategies. Some p oison-dart frogs deposit their eggs on
ground while one parent remains with the eggs until they hatch into tadpoles, and others carry the eggs with them until they are ready to ha tch. Representatives of the genus Oophaga (Dendrobatidae : Dendrobatinae), comprising nine species (Bauer, 1994), are among th e most variably colored vertebrates with the most developed parental care strategy. Th ey mate without an amplexus; the role of the male is to tend to the clutch, while that of the female is to carry the larvae to standing bodies of water. Tadpoles exhibit begging behavior when fed with nutritive eggs (Weygoldt, 1980; Brust, 1993; Pramuk and Hille r, 1999). Oophaga pumilio (Schmidt, 1857) is a diurnal, terrestrial poison fr og in the family Dendrobatidae, found in low-elevation tropical fore sts on the Atlantic side of Nicaragua, Costa Rica, and western Panama (Daly and Myers, 196 7; Savage, 1968; Myers and Daly, 1983). Their degree of aposematism is remarkable. Oophaga pumilio is one of the most variable poison frogs regarding coloration, color p attern, skin structure and skin alkaloid content. Pattern evolution among the frogs on the islands of the Bocas del Toro Archipelago in Panama produced a fascinating palett e of colors such as fire red with black spots, blue with black spots, orange, orange with black spots etc., each color pattern pertaining to its own island (Fig. 1). The selecti ve forces that have produced this remarkable variation remain unclear. This species shows extreme variation in color and pattern between populations that have been geograph ically isolated for < 10,000 years. The greenish-black spotted frog with yellow undersi des can only be found on Isla Coln and is entirely specific to the area. Oophaga pumilio is a model system for understanding the relationship between natural and sexual selection in producing color morphs, territoriality, mating and reproduction, de fense strategies and aposematism,
toxicity and diet in poison frogs (Saporito et al., 2004; Saporito, 2007; Saporito et al., 2010). Territoriality and reproductive behavior we re extensively studied since the 1960s. Concerning parental care, Oophaga pumilio are a novelty in anurans in that females feed the larvae with infertile eggs. This species has produced extensive controversy reg arding the influence of sexual and natural selection on color divergence and patte rning. Recent studies confirm that females are attracted to males with bright colorati on, and patterning has an effect on male-male interactions (Maan and Cummings, 2008; Cr others et al., 2011). Figure 1. Left: Distribution of O. pumilio (striped area) an d sample area (box), right: a detailed map of color morphs. Locality and abbreviations: Ba stimentos red frog beach BRF, Cayo Nancy CN, Isla Coln IC, Isla San Cristobal SC, Isla Popa IP, Pelican Cay PC, Cerro Brujo CB. Rudh et al., 2010
Table 1 Approximate area, distance and name of the closest body of land, sister population relationships of O. pumilio and a maximum estimate of effective p opulation size. Brown et al., 2010 Taxonomy Kingdom : Animalia Phylum : Chordata Class : Amphibia Order : Anura Family : Dendrobatidae Genus: Oophaga Species: pumilio Habitat Most Oophaga pumilio inhabit lowland rainforests. They can be found in banana plantations, cocoa plantations or in the leaf litte r of primary and secondary forests. Their abundance in the leaf litter is due to the presence of arthropods, their main food source.
Aposematism According to sexual selection theory, the evolution of exaggerated display behavior is driven by increased mating success, but limited by natural selection through predation. Aposematic species signal to visual pre dators that they are unpalatable prey by using bright colors (Poulton, 1890). Conspicuou s coloration advertises anti-predator defense across many taxa, including invertebrates, fish, amphibians, snakes, and birds. Many frogs of the family Dendrobatidae are aposemat ic, exhibiting extreme variation in coloration and patterning. Oophaga pumilio is considered the most polymorphic poison frog. The red/blue phenotype is encountered across Nicaragua to Costa Rica to the western border of Panama (Maan and Cummings, 2008) In the Bocas del Toro archipelago in western Panama the species exhibits intense variation in bo th hue and brightness. Fifteen different phenotypes span the entire range of the v isual spectrum on the mainland and across island populations (Siddiqi et al. 2004). The islands are relatively small, ranging from 2.4 km265 km2 (Brown et al., 2010) (Table 1), and have been geogr aphically isolated for < 10,000 years (Summers et al., 1997; Wang and Shaffer, 2008). Studies in Panama, especially in the archipelago have focused on aposematism, spectral analysis of color morphs and female choice experiments correlat ed with color patterns. Molecular phylogeographic studies indicate that the different populations of O. pumilio in the archipelago are genetically similar (Summer s et al., 1997; Hagemann and Prhl, 2007; Rudh et al., 2007; Wang and Shaffer, 2 008; Brown et al., 2010). Sexual selection alone cannot explain the divergenc e in coloration among populations so both natural and sexual selection me chanisms have to be considered.
Although geographic isolation could be a facilitato r in color divergence in the region, all islands were connected less than 10,000 years ago ( Summers et al. 1997; Anderson and Handley, 2002); this is a very short time for genet ic drift to result in the extreme phenotypic diversity we see today. Maan and Cummings (2008) studied O. pumilio female preferences for different components of aposematic coloration across several phenotypically distinct populations. Specific components of the aposematic signal (dorsa l color, ventral color, and spotting pattern) are affected differently by natural and se xual selection. Earlier studies (Summers et al. 1999; Reynolds and Fitzpatrick, 2007) tested the hypothesis that divergent sexual selection by female mate choice has contributed to the color diversity. It has recently been discovered that O. pumilio anurans have color vision: their eyes have one rod and three cone types that enable them to differentiate between the many conspecific color morphs (Siddiqi et al., 2004 ). Females rely heavily on this ability to select mates of the same color morph (Savage, 20 02; Siddiqi et al., 2004). Maan and Cummings (2008) found that male dorsal color is the most important determiner of female preferences. Interestingly enough, spotting or ventral color did not affect female choices. Crothers et al. (2011) examined how aposematic sign al variation influences male male interactions. Given evidence for female prefe rence for brighter coloration in some populations of this species, they tested whether ma lemale interactions could also be mediated by male brightness. Their results show ma le dorsal brightness influences the behaviors of male conspecifics, and that a males d orsal brightness predicts his own behavior such that bright males approach stimulus f rogs faster, direct more calls to bright
stimulus frogs, and exhibit lower advertising call pulse rates. Aposematic signals may thus be a product of intra-sexual selection (male-m ale competition). Many authors have hypothesized that color is the ta rget of divergent selection between O. pumilio populations in the Bocas del Toro archipelago (Dal y and Myers, 1967; Summers et al., 1997; Siddiqi et al., 2004; H agemann and Prhl, 2007; Prhl et al., 2007; Rudh et al., 2007; Wang and Shaffer, 2008). Summers et al. (1997) found relatively low mitochondrial DNA diversity among co lor morphs. Brown et al. (2010) used a combination of population genetics, phylogeo graphy and phenotypic analyses to test for divergent selection in coloration in O. pumilio They collected tissue samples from 15 distinct populations (Fig. 2) to develop a gene tree using the mitochondrial DNA (mtDNA) d-loop region and predicted the hypothetica l nuclear gene underlying coloration. They collected spectral reflectance an d body size measurements on individuals from four of the populations to perform a quantitative analysis of phenotypic divergence. Their results suggest divergence in co lor, not body size, is occurring at a faster rate than expected under neutral processes. Their study provides quantitative support for the claim that strong diversifying sele ction underlies color variation in the strawberry poison frog. They could not, however, i dentify the underlying mechanisms for the extreme divergence in O. pumilio coloration.
n Figure 2. (a) A map of the Bocas del Toro archipelago, Panama with photos of color morphs, (b) Gene tree for four populations of Oophaga pumil io Brown et al., 2010
Assuming birds are potential predators, Darst and C ummings (2006) used an avian visual model and live poison frogs to evaluat e conspicuousness as a combination of color and brightness contrast. They noted that no s tudy has empirically evaluated the relative importance of conspicuousness and unpalata bility in avoiding attack by predators but two earlier studies in the 1980s had already d emonstrated distastefulness of this anuran to the ant, Paraponera clavata (Fritz et al., 1981) and the ctenid spider, Cupiennius coccineus (Szelistowski, 1985); both studies express the prob ability that this could be due to the toxins in the skin of these fro gs. Genetics Oophaga pumilio have a 2n = 20 bimodal karyotype with six large pa irs (1 to 6) and four small pairs (7 to 10) of chromosomes (Boga rt, 1991). Skin toxins Poison frogs excrete toxins from granular glands in the skin as a chemical defense against predation and/or microorganisms (Daly and M yers, 1967; Myers and Daly, 1983). The poison consists of alkaloids, lipophilic substa nces that attack the peripheral nervous system. Nearly 500 different alkaloids have been id entified in Dendrobatoidea (Ltters et al., 2007). Pumiliotoxin (Fig. 3) is the principa l toxin in O. pumilio Lethal dosage (LD50) of pumiliotoxin for laboratory mice is 1.2 2.5m g/kg (Ltters et al., 2007). Figure 3. Pumiliotoxin chemical structure Ltters et al., 2007
Diet and Feeding Poison frogs obtain their poison via their prey. Oophaga pumilio specialize in minute prey. Prey must be small enough to be handle d with the tongue. Donnelly (1991) studied feeding patterns of strawberry poison frogs ; she dissected O. pumilio frogs and found that their diet consists of insects and mites : the stomachs of females contained 56% ants and 38% mites, those of males 51% ants and 42% mites, and those of juveniles 39% ants and 56% mites. It was recently discovered tha t Dendrobatids have an efficient system that accumulates alkaloids from dietary alka loid-containing arthropods (Saporito et al., 2004). Not all ants contain these pumiliot oxins; it has been suggested that ants might acquire them via their food or symbiotic bact eria (Smith and Jones, 2004). Reproduction The mating system of Oophaga pumilio is best described as sequential polygamy (Davies, 1991) comprising sequential and simultaneo us polygyny and polyandry. Females engage in polyandry to produce unfertilized eggs. They invest their energy into one small clutch. Once they mate with another male, they climb up in the canopy to deliver their eggs to the larvae. Constant availability of water and food allows fema les to produce clutches continuously. Donnelly (1989b) found unequal dist ribution of large eggs in the oviducts of females. Their presence was increased before th e rainy season, which suggests seasonal variation in reproduction. However, juve niles are present all year round.
Oviposition Oviposition takes place in a location preselected b y the male. The male has to ensure that the mating site is close to a water sou rce, so he chooses it based on water availability. Prhl and Berke (2001) found that th e sex ratio is strongly female biased in the habitat with more tadpole-rearing sites, which is reflective of the female reliance on these resources. In accordance with the data from D onnelly (1989b), female home ranges are large because they move among tadpole-rearing s ites and territories to select males (Prhl and Berke, 2001). The male leads the female with close-range courtshi p calls until they reach the oviposition site (usually a leaf) (Limerick, 1980). Upon arrival the male deposits sperm with wiping motions with his hind legs by rapidly e xtending his hind feet outwards and back (Limerick, 1980). He will continue calling un til the female joins him. Physical contact can consist of both trying to maneuver them selves into a position whereby the male is sitting on the females back so their cloac ae touch (pers. obs.) during this phase the male will deposit his sperm (Ltters et al., 20 07), or stroking the females back with a foreleg and then a hind leg while she deposits eggs (Limerick, 1980). He will leave the site while the female begins to lay her eggs. By r otating in small increments around her vertical axis, she will lay 3 5 eggs (Limerick, 1 980), 5 9 eggs (Weygoldt, 1980) or 3 7 eggs (Prhl and Hdl, 1999) and remain on them fo r 30 35 min (Limerick, 1980). Parental Care This species exhibits sophisticated parental care. After oviposition, the female leaves the site. The male will return to moisten t he eggs daily until the embryos develop (10 14 weeks). Eggs develop into tadpoles in 10 12 days (Weygoldt, 1980). The
female will return when the tadpoles begin to hatch and she will sit on top of the larvae until one wriggles up her back. She will take them in separate trips to a water pool in bromeliad (banana or heliconia) axils close to the ground or up in the canopy (bromeliads inhabited by a larva are not used for bathing) (Wey goldt, 1980) (Fig. 4). Figure 4. Leftbromeliad on a tree branch in the canopy, Rig htfemale on her way up to the larvae. Over the next 6 8 weeks she will inspect them on 7 11 occasions over 4 days, visiting each larva regularly and feed it unfertili zed eggs. In terrarium conditions, a female could lay one egg clutch per week to every t wo weeks to supply the larvae with food (Weygoldt, 1980). Their metamorphosis takes u p to 46 days (Prhl, 1995) or 43 52 days (Brust, 1993). If a male comes upon the clutch of eggs of another frog, it will destroy or consume them (Weygoldt, 1980), which is an interest ing show of intraspecific competition.
Feeding the young Most species of amphibians lay eggs which are depos ited either in water or on land. Eggs may hatch into aquatic larvae or miniat ures of the terrestrial adults. In some species females retain their eggs in the oviducts a nd give birth to metamorphosed young, while in others adults carry their eggs attached to the surface of their bodies, in pockets of the skin of gills or flanks, vocal sacs, etc. Viviparity is widespread among caecilians. At birt h, the young are thirty to sixty percent of their mothers body length. Fetuses ini tially feed on the yolk contained in the egg at the time of fertilization; once this source is exhausted, they emerge from the egg membranes and align themselves to scrape the walls of the oviducts with specialized embryonic teeth (Pough et al., 2002). Salamanders lay their eggs singly or in a mass of transparent gelatinous material; with the exception of aquatic species, eggs hatch into gilled aquatic larvae which transform into terrestr ial adults. A few species of salamanders in the genera Salamandra and Lyciasalamandra are viviparous; once they exhaust the yolk supply, embryos consume unfertiliz ed eggs and later scrape the reproductive tract of the female with specialized t eeth (Pough et al., 2002). A novelty among anurans, female Oophaga pumilio feed the larvae with unfertilized eggs. Brust (1993) and Pramuk and Hil ler (1999) demonstrated obligate oophagy in this species. In 1994, it was renamed Oophaga pumilio (formerly Dendrobates pumilio ) (Bauer, 1994). When ready to feed its larvae, the female inspects the water pool by staring into it. The feeding process starts with the recipient larva showing begging behavior: swimming slowly with a stiffened tail and vibrating movement s towards the female and touching her
vent or legs. She will then turn around and enter t he water backwards like a frog attempting to bathe (Weygoldt, 1980). Tadpoles fee d by biting a hole in the jelly envelope of the unfertilized egg and sucking its co ntents. A feeding session takes 3 5 minutes (Weygoldt, 1980). Females feed larvae once every one to two weeks in terrarium conditions (Weygoldt, 1980). Brust (199 3) monitored seven O. pumilio larvae on Isla Bastimentos, Panama. He observed that eac h received 1 5 nutritive eggs on 9 13 occasions, totaling 20 36 eggs; the larvae wer e fed at intervals of 1 9 days; tadpoles experimentally placed in leaf axils were not fed an d therefore died of starvation. Females do not attempt additional reproductions whereby the y fertilize eggs while caring for a brood. Territoriality Territoriality is widespread among all groups of ve rtebrates, and is defined as space-related dominance that ensures the territory holders access to critical resources such as food, breeding sites or mates (Davies and H alliday, 1978; Duellman and Trueb, 1986; Prhl, 2005; Wells, 2007). A territory is de fined as the calling area of one male, used exclusively and defended against other male in truders by use of territorial advertisement calls and contests (Kaufman, 1983). A territory is part of an individuals home range. Home ranges overlap whereas territorie s do not. Criteria for territoriality are: site tenancy, advertisement calls and aggressi ve defense/exclusion of competitors. Costs associated with territory defense are elevate d predation risk and time and energy expenditures. Benefits associated with territory d efense are better exposure to females and control over food resources, oviposition and ma ting sites (Duellman and Trueb, 1986).
Martof (1953) demonstrated the existence of territo riality in amphibians in breeding aggregations of the green anuran, Rana clamitans Herpetologists accepted the notion of territorial defense after studies on bull frogs Rana catesbiana (Emlen, 1968; Wiewandt, 1969) and dendrobatids Prostherapsis trinitatis (Test, 1954; Sexton, 1960) were published. Oophaga pumilio males show extensive territorial behavior; they de fend their territories over several months or years (Prhl and Berke, 2001). Male O. pumilio typically display territory ownership from perches, defined as any object (log, vine, tree branch, tree buttress, roots etc.) elevated 5 cm ab ove the forest floor (Graves et al., 2005). Bunnell (1973) was the first to offer a descriptive account of O. pumilio territorial behavior. She studied the Costa Rican bright red co lor morph with blue legs and fawn colored venter at the Organization for Tropical Stu dies at La Selva, Costa Rica. Her observational experiment shows that males remained in close proximity to their perch, were 2 3 m apart from each other (Fig. 5), and ma intained this territorial spacing by use of vocalization. Her manipulative experiment teste d responses to tape recorded calls. Data show antiphonal responses to playbacks: males stop calling when tape recording begins and start calling when tape recording ends. She described the most intensive calling time frame from 6 am 10 am, and deemed te mperature not significant to calling variables because it ranged only 21 23C. She in vestigated bout length (string calls separated by at least 1 second of silence), number of calls per bout and rate of calling within a bout.
Figure 5. Study area with points of capture mapped for each a nuran Bunnell, 1973 Territoriality in anurans is frequently associated with the defense of limited reproductive resources. Donnelly (1989a) investiga ted whether reproductive resources limit density in O. pumilio at La Selva, Costa Rica. She manipulated resource availability to pinpoint which resource is being de fended. Bromeliads are important for the egg stage of O. pumilio because other large bodies of standing water do no t occur in the habitat (1989a). Bromeliads are thus a valuabl e resource because eggs are deposited in the water-filled axils. She was the first to sh ow that competition among O. pumilio males is linked to limited bromeliad availability, because the addition of this resource led to an increase in the density of the territorial se x.
In resource defense polygyny (Emlen and Oring, 1977 ; Alcock, 1993) males defend resources to increase attractiveness to fema les. Females thus preferentially select males that defend high-quality resources needed f or offspring survival. Male O. pumilio were found not to defend tadpole rearing sites for females (Prhl and Berke, 2001). However, O. pumilio defend territories used for calling, courtship and oviposition (Prhl and Berke, 2001). Males compete strongly fo r sites with high female density, highly important in increasing their mating success Male reproductive success is thus limited by female availability (Prhl and Hdl, 199 9). Territories which include home ranges of females enhance male mating success (Prh l and Berke, 2001). Baugh and Forester (1994) performed several prior residence effect (Braddock, 1949) manipulative experiments. They allowed 32 ma le O. pumilio to establish territories in 32 aquaria. They were the first to establish an ethogram on the following agonistic behaviors: calling, push-ups, tracking, c harge and veer (no physical contact), charge and contact, grappling, pinning, chase escap e, and freeze position (referred to as statue behavior whereby the submissive anuran fre ezes in one position). They measured the index of aggression by evaluating ener gy expenditure: a behavior judged to be more energetically expensive received a higher v alue than a less costly one. They tested 1) responses to conspecifics, 2) resident re moval, 3) encounters involving size asymmetry and 4) encounters with a sympatric, confa milial species. They found that 1) residents were consistently dominating intruders du ring staged encounters; when reciprocal trials were carried out, residents were dominant over males to which they had previously been subordinate, 2) when territory mark ers (Sphagnum and potted plants) were removed, 7 out of 10 males still recognized an d defended their territory, 3) when the
n resident male was presented with a larger intruder, he still maintained territorial ownership, 4) the prior residence effect in O. pumilio was not altered when introduced to confamilial intruders ( Phyllobates lugubris ). Forester et al. (1993) reported results indicating that the visual stimulus of a male O. pumilio present in another males territory elicited more aggressive responses than did playbac ks of advertisement calls to the same subject. Their data demonstrates that male O. pumilio exhibit prior territorial residence effect (Braddock, 1949) in relation to the resident s knowledge of his territory, and resident males generally win against intruder males Dear enemy recognition has been demonstrated in a number of taxa including anurans (Davis, 1987), urodeles (Jaeger, 1981), liz ards (Fox and Baird, 1992; Lopez and Martin, 2002), and birds (Brooks and Falls, 1975; F alls and McNicholl, 1979). The dear enemy effect consists of territory holders exhibit ing lower levels of aggression toward their nearby neighbors in comparison to unfamiliar individuals. Bee (2003) conducted neighbor-stranger discrimination playback experimen ts of O. pumilio (red black-spotted color morph) on Isla Bastimentos, Panama. He noted males responded to playback stimuli but did not exhibit differential responses to neighbors and strangers. He found no evidence of dear enemy recognition and suggested that resident males may require visual cues/stimuli (Forester et al., 1993) to reco gnize individuals. He also noted the distance between territories to range 1.7 4 m (me an 2.7 m), in accordance with previous studies (Bunnell, 1973; McVey et al, 1981; Donnelly 1989b; Prhl and Hdl, 1999; Prhl, 2003). Gardner and Graves (2005) followed up on Baugh and Foresters (1994) and Bees (2003) research. They tested for responses o f resident males to familial and
unfamilial intruders (the dear enemy phenomenon) at La Suerte, Costa Rica. The dependent measures of territorial defense were: the amount of time spent oriented toward the intruder, physical contact with the intruder, a nd four calling variables: 1) number of call groups, 2) call note rate, 3) total time calli ng and 4) mean call group duration (Graves et al., 2005). They found that resident male O. pumilio do not exhibit dear enemy recognition of neighbors, and no significant diffe rences in resident calling characteristics when exposed to familiar and unfami liar male intruders. Staudt et al. (2010) found that there is a connecti on between foraging behaviour and territoriality in poison anurans. Oophaga pumilio defend their territory in response to food availability. They found males established te rritories in areas of high insect abundance whereas the area around their territories had a lower amount of insects and was not defended. Feeding sites may thus be the co re area of males territory in areas of high female density. The presence of alkaloid sour ces like formicine ants of the genera Brachymyrmex and Paratrechin provides them with two limited resources: mates and toxic diet. Male anurans expend a considerable amount of energy defending their territory. Oophaga pumilio have adapted to using acoustic communication signa ls to deter potential enemies thus avoiding engagement in costl y fights (Meuche et al., 2012). They often use vocalization as a social mediator to asse ss the fighting abilities of an opponent (Davies and Halliday, 1978). Meuche et al. (2012) found that male O. pumilio modify their calls in response to synthetic calls with different dominant frequency, and use dominant frequency to assess the fighting ability of an opponent during aggressi ve interactions.
Vocalization Vocalization is the primary mediator of social inte raction among anurans. Males produce loud, species-characteristic calls. These calls are distinctive and of widespread interest in studies of behavior and sexual selectio n. The most frequently heard vocalizations are advertisement calls (Fig. 6). Su ch calls can be categorized as territorial calls by males when rival males are present, and co urtship calls when females are present. Males have vocal slits (openings in the floor of th e mouth) which lead to inflatable sacs. These vocal sacs act as resonators to amplify the s ound. Air pumped out of the lungs over the vocal cords produces the fundamental frequ ency of a call (Fig. 7). Figure 6. Components of anuran communication system Duellman and Trueb, 1986
Figure 7. Schematic representation of an anuran showing struc tures involved in vocalization Duellman and Trueb, 1986 Acoustical features of advertisement calls are anal yzed with electronic equipment such as oscilloscopes or sonographs that produce os cillograms and audiospectrograms as representation of sounds. The oscillogram shows th e waveform of the soundhow the call amplitude (vertical axis) changes with respect to time (horizontal axis) (Fig. 8). The audiospectrogram shows the change in frequency (ver tical axis) with respect to time (horizontal axis); the darkness of the patterns ref lects the amplitude of the sound component (Fig. 9). The most significant components of vocalizations ar e the dominant frequency (pitch), duration and pulse rate (successive bursts of sounds). They are measured in cycles per second (Hz), time in seconds, and number per second.
Figure 8. a) Power spectrum of a single advertisement call gr oup from Oophaga pumilio, b) Oscillogram of the same call group, c) Oscillogram depicting four pulsed advertisement calls from the middle of the same call group. Horizontal bars depict time scales in b and c. Bee, 2003
Figure 9. Oscillogram (above), spectrogram (below and left) o f three advertisement calls from a call sequence. Oscillogram and sonogram show the pu lsed structure of the signal, the spectrogram shows the energy distribution with a peak near 4 kH z C.D.= call duration, C.P.= call period, I.I.=inter -call interval, P= pulses. Prhl, 2003 Male Oophaga pumilio exhibit aggressiveness and engage in physical comb at by grappling and wrestling the intruder. As a compone nt of intra-sexual competition, males advertise territorial ownership through vocalizatio ns (Prhl, 2003). Males calling during male contests call with a lower frequency and are t hus more successful in defending their territories (Prhl and Berke, 2001). Advertisement calls consist of a long series of ch irps (Duellman, 1966) or as described later by Zimmerman (1990) as` rapidly rep eated, brief buzzing sounds. Each burst of the buzzing sound is called a "call note," and a series of consecutive call notes with less than one second between call notes compri ses a call group (Bunnell, 1973). Zimmermann (1990) identified four call types for O. pumilio males: advertisement calls, courtship calls, aggressive calls and release calls all described as buzz calls with various modifications.
Calling characteristics such as call rate, pulse ra te and dominant frequency are correlated with mating success (Prhl, 2003). Bunn ell (1973) first suggested that females are attracted by auditory stimuli rather than visua l or chemical stimuli. Females prefer males that exhibit high rates of vocalization and o ccupy elevated display sites (Prhl and Hdl, 1999). Prhl (2003) suggests that males incre ase their call rate when detecting a female because higher call rates are attractive for females, whereas males that produce low calling frequencies are unsuccessful in attract ing mates. Activity Levels Diel activity of poison frogs is influenced by ligh t, temperature and relative humidity (Ltters et al., 2007). Dendrobates are a ctive and vocalize early in the morning (when air is full of moisture), retreat over noon w hen temperatures reach their maximum, reappear and resume their activity later in the aft ernoon (Ltters et al., 2007). Bunnell (1973) suggested that the most intensive pe riod of calling for O. pumilio is between 6 am and 10 am, but did not report data. She also deemed temperature not significant to calling variables because it ranged only 21C 23C. Graves (1999) and Graves et al (2005) studied daily activity patterns correlated with vocal display at La Suerte Biological Station in Costa Rica. Studies w ere done for nine consecutive days in 1997 (Graves 1999), and for seven weeks in 1998 (se ven males studied per week) (Graves et al 2005) to measure daily activity level s in relation to vocal display. Graves (1999) found that O. pumilio become active at dawn and peak in activity early in the morning (07:45 am to 09:15 am); activity gradually decreases to midday, after which a moderate level of activity is observed until dusk. Temperature was not significant to calling activity. Regression comparing 72-hr rainf all and total number of O. pumilio
observed in the plot was significant. Graves et al (2005) found that vocalization characteristics such as total time calling, mean ca ll group duration, number of call groups, call note rate, and proportion of days calling were greater between 06:30 am and 08:30 am than at other times through the day. There was a clear peak perch use by O. pumilio shortly after sunrise. Occupation of elevated disp lay sites was greater between 05:30 am and 09:30 am than at other times of day. Temperatu re was not significant to calling activity. Rainfall during the previous 24-hr perio d was positively correlated with all calling characteristics and perch use. Graves (1999) and Graves et al. (2005) determined t he total number of frogs engaged in various activities throughout the day bu t did not specifically pinpoint how many frogs vocalized and when their calling peak wa s. Knowing when the calling activity is at its highest is crucial in studies de pendent on calling variables. In dim light, frogs are hidden in the leaf litter rather than cal ling (personal observation) so it is important to know how weather conditions affect the se anurans. Graves (1999) assumed that the proportion of animals found on elevated pe rches would provide an indication of the extent of territorial and courtship vocalizatio n, relative to foraging activity. In my study I 1) quantitatively determined the number of vocalizing male frogs and 2) investigated differences in hourly total activity a nd hourly calling activity and the influence of weather conditions such as temperature humidity, light intensity and rainfall on total activity and calling activity during a day
Methods I conducted this study in January 2012 at the Insti tute for Tropical Ecology and Conservation Biological Station (092490N; 8219 83W) in Boca del Drago, North Isla Coln, Province Bocas del Toro, Republic of Pa nama. A study plot was established in a primary forest us ing a random numbers table, a compass, and flagging tape. Vocalizing frogs were marked with a flag a day prior to the onset of the field experiment. Data was collected during 10 observation periods ea ch day, spaced at 1-hr intervals. The first observation period began at 7 :30 am, and the last began at 4:30 pm. At the onset of an observation period, temperature, humidity and light level measurements were taken (for example in the 7:30 am 8:30 am observation period, measurements were taken at 7:30am, for 8:30 am 9: 30 am period, measurements were taken at 8:30 am, etc.). The forest floor was scanned thoroughly and consist ently by walking along 3 transects equidistant from the longer edges of the plot (measuring 35 ft by 44 ft) for 45 minutes and resting for 15 minutes. All frogs seen on the ground, on perches and vocalizing within the plot were counted in every ob servation period. Frogs calling outside the plot were not counted. Temperature was recorded to the nearest 0.1C with a digital thermometer. Humidity was recorded with Sling Psychrometer. Lig ht Intensity (Foot Candles) was recorded with a Field Scout Light Meter C#3415FQF. Amount of rainfall (meters) within 24 hrs was recorded daily with a standard rain gaug e located in an open pasture at 6:30am. Vocalizations were recorded during observa tion periods using a Sony PCM-
M10 recorder and a Sennheiser ME66/K6 unidirectiona l microphone. These devices were placed approximatively 1 m from the calling ma les. Recording times were chosen independent of time of day. Advertisement calls were digitized (16-bit resolutio n; sampling rate= 22.05 kHz) and bandpass filtered bet ween 3.5 and 7 kHz to reduce background noise due to wind and other signaling an imals, using Raven Pro 1.4 soundediting software running on a Dell Inspiron 1525 po rtable computer. A one-way ANOVA ( p < 0.5) statistical test was used for total activity and calling activity and linear regression was used for weather conditions (temperature, humidity, light intensity, rainfall). Results Data reflects overall daily activity in hourly incr ements of a) foraging frogs, frogs engaged in courtship, and calling frogs and b) only the vocalizing frogs, in response to weather conditions: previous 24-hr rainfall, hourly temperature, hourly humidity, and hourly light intensity. Total Activity and Calling Activity The total number of O. pumilio frogs observed at a single observation time ranged from 0-16. The number of calling males observed at a single observation time ranged from 0-10 and varied across time periods. There we re no significant differences in the total activity of O. pumilio among observation time s ( p = 0.5398, df = 9, F = 0.8912) (Fig. 1). However, activity peaks for calling activity w ere significantly different ( p = 0.0011, df = 9, F = 3.791) from one observation period to the other w ith peak calling 10:30 am 11:30 am (Fig. 2).
n Figure 1. Mean SD number of O. pumilio observed at each obs ervation period. Figure 2. Mean SD number of calling O. pumilio observed at each observation period. Temperature Temperature ranged between 23.9 C to 31.4 C durin g the observation times in this study. Mean ambient temperature for all obse rvation times was 26.3C, with means for each of the six days of the study ranging from 24.7 C to 27.7 C. Regression comparing temperature to total number of O. pumilio observed was not significant ( F = 0.3731, p = 0.5437, df = 1, r2 = 0.006503) (Fig. 3). Regression comparing mean nrn r r n n r r r r r r r 8 7 6 5 4 3 2 1 0 rn r r n n r r r r r r r 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
temperature to calling number of O. pumilio observed was not quite significant ( F = 3.065, p = 0.0853, df = 1, r2 = 0.05019) but the trend was for decreased calling as temperature increased (Fig. 4). Figure 3. Temperature and number of O. pumilio observed at ea ch observation period. Figure 4. Temperature and number of calling O. pumilio observ ed at each observation period Humidity Mean ambient humidity for all observation times was 90.1%, with means for the six days of the study ranging from 85.8% to 94.2% F Regression comparing humidity to total number of O. pumilio observed was not significant ( F = 1.149, p = 0.2883, df =1, r2 = 0.01976) (Fig. 5). Regression comparing mean hu midity to calling number of
O. pumilio observed was not significant as well ( F = 0.3676, p = 0.5467, df = 1, r2 = 0.006298) (Fig. 6). Figure 5. Humidity and number of O. pumilio observed at each observation period. Figure 6. Humidity and number of calling O. pumilio observed at each observation period. Light Intensity Illumination ranged between 20 ft cd to 360 ft cd d uring the observation times in this study. Mean intensity for all observation ti mes was 72.58 ft cd, with means for each of the six days of the study ranging from 18.33 ft cd to 136.66 ft cd. Regression comparing light intensity to total number of O. pumilio observed was not significant ( F = 0.3731, p = 0.5437, df = 1, r2 = 0.006503) (Fig. 7). Regression comparing mean l ight
intensity to calling number of O. pumilio was significant ( F = 6.401, p = 0.0141, df = 1, r2 = 0.09939) with an increase in calling at higher l ight intensity (Fig. 8). Figure 7. Light intensity and number of O. pumilio observed a t each observation period. Figure 8. Light intensity and number of calling O. pumilio ob served at each observation period. Rainfall Regression comparing rainfall to total number of fr ogs was not significant ( F = 0.2442, p = 0.6471, df = 1, r2= 0.05754) (Fig 9). Regression comparing rainfall to the number of calling frogs was not significant as well ( F = 0.1471, p = 0.7208, df = 1, r2= 0.03547) (Fig 10).
Rainfall (cm) 11 10 9 8 7 6 5 4 3 2 1 Total # frogs 13 12 11 10 9 8 7 6 Figure 9. Rainfall (24 hr) and total number of frogs Rainfall (cm) 11 10 9 8 7 6 5 4 3 2 1 # Calling Frogs 5 4 3 2 Figure 10. Rainfall (24 hr) and number of calling frogs.
Figure 11. Diagram of the study plot; territory sizes are appr oximate, territory size measurements were not taken. Numbers refer to frog resident # in the territory. Fourteen perches were monitored for perch occupati on while calling, presence in territory, and absence from territory (Fig. 11). T he frogs that advertised the most were frog #5 and frog #14 (Table 2). Both frog #5 and f rog #14 held contested perches (Table 3). Intruders ranged between 10 26 males in a da y; intruders called within the plot 6.7 38.3 percent of the time (Table 4). n
Table 2 Distribution of activities of resident frogs throug hout the day Frog # % perch occupation while advertising % presence in territory % absence 1 20 8.3 71.7 2 0 0 100 3 0 0 100 4 3.3 18.3 78.4 5 53 1.7 45.3 6 1.6 3.3 95.1 7 0 0 100 8 8.3 1.7 90 9 8.3 0 91.7 10 21.6 6.7 71.7 11 28.3 1.7 70 12 20 3.3 76.7 13 6.6 1.7 91.7 14 40 16.7 43.3 Note: frogs # 2, 3, 7 were absent throughout the study period, they were most likely intruders. Table 3 Most contested perches Perch # # of combats 5 2 14 2 13 1 Table 4 Calling frequency of intruders on territories and p ercent of intruders calling Day 1 2 3 4 5 6 Number of unflagged males calling 10 26 21 10 23 13 % calling 6.7 43.3 35 16.7 38.3 21.7
Males engaged in fights exhibited lower advertising call pulse rates (Table 5), in accordance with the findings of Crothers et al. (20 11). Table 5 Dominant Frequencies, Amplitudes and Pulse Rates in analyzed sequences Activity Frequency range (kHz) Amplitude for call group (kU) Mean Pulse Rate (number bursts/sec) Male advertising alone 3.5 6 5 0.5 Male engaged in fight when female is present 3.7 6 3; 5; 7 1.3 Male courting a female 3.3 5.7 3 1.4 Male engaged in mating with female 3.5 6 S1 S2 S3 S4 S1 S2 S3 S4 1018 8; 15 13 155 115 10101010 15 16 13 0.67 0.13 0.28 0.23 Note: kU are kilo units of amplitude in Raven sou nd analysis software, S1, S2, S3, S4 are all separate four sequences of the same mating callse e Figs. 15, 16, 17, 18.
Sound Waveforms and Spectrograms Figure 12. Waveform and spectrogram of male advertising by him self; units: amplitude kU, frequency kHz, time sec
Figure 13. Waveform and spectrogram of male fighting another i n the presence of a female; units: amplitude kU, frequency kHz, time sec.
n Figure 14. Waveform and spectrogram of male courting female; u nits: amplitude kU; frequency kHz; time sec. One mating was observed (male #5). The mating tim e until oviposition occurred was 96 min (Fig. 14, 15, 16, 17). Figure 15. Sequence 1. Waveform and spectrogram of a mating ca ll; units: amplitude kU; frequency kHz; time sec.
Figure 16. Sequence 2. Waveform and spectrogram of a mating ca ll; units: amplitude kU; frequency kHz; time sec. Figure 17. Sequence 3. Waveform and spectrogram of a mating ca ll; units: amplitude kU, frequency kHz, time sec.
Figure 18. Sequence 4. Waveform and spectrogram of a mating ca ll; units: amplitude kU, frequency kHz, time sec. Discussion I used techniques similar to those employed in the studies conducted by Graves (1999) and Graves et al. (2005). An important diff erence is that the current study was conducted in the dry season in January, while Grave s study (1999) was conducted June to July and Graves et al.s study (2005) July to Au gust during the wet season when reproductive activity is at its highest. Graves (1999) and Graves et al. (2005) found that a ctivity (foraging, courting, calling, etc.) rises early in the morning (6 am 9 am), decreases slightly to mid-day, and spikes again before sunset. Bunnell (1973) found t he most intensive calling activity period early morning 6 am 10 am. In my study, ov erall activity was low early in the morning, somewhat higher from 10:30 am until noon, and somewhat constant late
afternoon, with a spike an hour before sunset. Cal ling activity however, was at its lowest early in the morning and late afternoon, and at its highest 9:30 am 12:30 pm. Graves (1999) and Graves et al. (2005) concluded th at rainfall is significantly correlated with activity. In my study, the amount of rainfall does not accurately reflect that received in the forest. The rain gauge was lo cated in an open pasture 20 minutes away from the study site. The study site was 500 m in from the entrance in the forest. I suggest placing the gauge near the forest entrance. It sometimes rained during the 10 hr observation periods. Calling activity right after the rain seems to have been higher (pers. obs.), probably due to the surfacing of frogs previ ously hidden in the leaf litter. As expected, temperature and humidity were not sign ificant. However, regression comparing mean light intensity to calling number of O. pumilio was very significant. The lower the light intensity, the more frogs stay hidd en in the leaf litter; the higher the light intensity, the greater the chances of frog visibili ty thus the greater the number of calling males. Contests with intruders over perches usually occurr ed when a female was present. Bunnell (1973) offered an observational account of two fights in the field: as soon as an intruder was within 1 1.5 m of the resident male s territory, the resident male would approach the intruder until it retreated. In this study all observed fights were won by the resident males, thus prior residence had a strong e ffect on the outcome. It is important to note that perch #14 was in the middle of the plot, highly visible from any one point, while perches #5 and #13 were on opposite branches of a l arge tree buttress at the edge of the plot. Even though male #5 was not clearly visible (the branch was covered in leaves), he advertised 53 percent and was absent 45.3 percent o f the time (Table 2) which shows that
he invested a great amount of energy calling. His efforts were successful and he was observed mating once. Perch #14 and perch #8 were right next to each other; they were two slanted branches which touched the ground in th e same pointwhenever a female was present they would both have to descend their p erch at the same spot (Fig. 19). Due to their level of aggression and one observed fight frogs #14 and #8 did not exhibit dear enemy recognition. Figure 19. Left arrow: perch # 14, Center arrow: intersection of perches, Right arrow: perch # 8. Bee (2003) r !""# "r$ %&"'()* !" $+) '" ( ", !-)" $.!/"frequencies I analyzed when a male advertised alone when the male was engaged in a fight when the female was present, when a male was courting a female, and when a male
was engaged in mating, fell in a similar range (3.5 6 kHz) to their data (Table 5). The oscillograms (waveforms) of the call groups differe d in amplitude (kU) which shows that the calls were different even though the frequencie s were in the same range. Cicadas broadcast at 4.91 kHz (Wong et al., 2009), a frequency similar to that of O. pumilio (4 kHz in Prhl, 2003 and 4.77 kHz in Wong et al., 2009) Whenever cicadas called, they produced heteroacoustic disturbances: frog calling activity was zero except when frogs were actively engaged in courtship. It remains to be seen whether cessation of calls during cicada advertisements has any effec t on male mating success such that it disturbs courting vocalizations will a choosy fem ale leave his territory if he stops vocalizing? Studies over longer time periods and across the geo graphic range of the species will be necessary to address the extent of variatio n in Oophaga pumilio activity patterns. In addition, female choice can be examined testing for a hot shot, hot spot hypothesis. If a males site is highly contested by intruders, his site must be a hot spot; if many females visit his site he must be a hot shot. Ma les defend territories that provide suitable calling sites, space for courtship and ovi position, and prevent interference by competitors (Prhl and Hdl, 1999). Females spend more time on parental care than males (Prhl and Hdl, 1999). If the female invests energy in producing infertile eg gs to feed her small clutch in separate trips, mate choice must be highly selective. Survi val rate of eggs is lowof the 11 clutches Limerick (1980) had observed, only two had eggs that hatched, so females much choose males who successfully defend their territor y to ensure the survival of their offspring.
Ways to improve the current experiment would be to 1) measure the distance between territories to compare it to previous studi es (Bunnell, 1973; McVey et al, 1981; Donnelly 1989b; Prhl and Hdl, 1999; Bee, 2003; Pr hl, 2003), to 2) count the number of females engaged in either courtship or mating wi th a resident male, and the number of females present when males fight to test for the h ot shot, hot spot hypothesis, and to 3) record advertisement calls such that recordings con tain more than 10 call groups to be able to compare them to Prhls (2003) and Bees (2 003) calling parameters. Finally, it is worth mentioning, that this study is the first in this line of research that has shown a statistically significant difference be tween hourly time intervals for diel calling activity in the dry season, and a positive correlat ion between calling activity and light intensity.
References Alcock, J. 1993. Animal Behavior. 5th ed. Sinauer Associates, Sunderland, MA. Anderson, R. P. and Handley, C. O. 2002. Dwarfism i n insular sloths: biogeography, selection, and evolutionary rate. Evo lution 56:10451058. Bauer, L. 1994. New names in the family Dendrobatid ae (Anura, Amphibia). RIPA. Fall, 1-6. Baugh, J. R. and Forester, D. C. 1994. Prior reside nce effect in the Dart-Poison Frog, Dendrobates pumilio Behaviour 131(3/4): 207-224. Bee, M. A. 2003. A test of the Dear Enemy Effect in the Strawberry DartPoison Frog ( Dendrobates pumilio ). Behavioral Ecology and Sociobiology 54(6): 601-610. Bogart J. P. 1991 The influence of life history on karyotypic evolution in frogs. In Amphibian cytogenetics and evolution (ed. D. M. Gre en and S. K. Sessions), 233257. Academic Press, San Diego, USA. Braddock, J. C. 1949. The effect of prior residence upon dominance in the fish Platypoecilus maculatus Physiol. Zool. 22: 161-169. Brooks, R. J. and Falls, J. B. 1975. Individual rec ognition by song in whitethroated sparrows. I. Discrimination of songs of ne ighbors and strangers. Canadian Journal of Zoology 53: 879-888. Brown, J. L., Maan, M. E., Cummings, M. E., and Sum mers, K. 2010. Evidence for selection on coloration in a Panamanian poison frog: a coalescentbased approach. Journal of Biogeography 37: 89190 1. Brust, D.G. 1993. Maternal brood care by Dendrobate s pumilio: a frog that feeds its young. Journal of Herpetology 27: 9698. Bunnell, P. 1973. Vocalizations in the territorial behavior of the frog Dendrobates pumilio Copeia 1973(2): 277-284. Crothers, L., Gering, E., and Cummings, M. 2011. Ap osematic signal variation predicts male-male interactions in a polymorphic po ison frog. Evolution 65:599605. Daly, J. W. and Myers, C. W. 1967. Toxicity in Pana manian poison frogs (Dendrobates): some biological and chemical aspects Science 156: 970973.
Darst, C. R. and Cummings, M. 2006. Predator learni ng favors mimicry of a lesstoxic model in poison frogs. Nature 440:208-211. Davis, M. S. 1987. Acoustically mediated neighbor r ecognition in the North American bullfrog, Rana catesbeiana Behavioral Ecology and Sociobiology 21:185-190. Davies, N. B and Halliday, T. R. 1978. Deep croaks and fighting assessment in toads Bufo bufo Nature 274:683685. Davies, N. B. 1978. Ecological questions about terr itorial behaviour. Pages 317350 in J. R. Krebs and N. B. Davies, editors. Behav ioural ecology: an evolutionary approach. Blackwell Scientific, Oxford England. Davies N. B. 1991. Mating systems. In: Krebs JR, Da vies NB (eds) Behavioral ecology: An evolutionary approach. Blackwell, Oxfor d, 263-294. Donnelly, M. A. 1989a. Demographic effects of repro ductive resource supplementation in a Territorial Frog, Dendrobates pumilio Ecological Monographs 59(3): 207-221. Donnelly, M. A. 1989b. Effects of reproductive reso urce supplementation on space-use patterns in Dendrobates pumilio. Oecologica 81: 212-218. Donnelly, M. A. 1991. Feeding patterns of the Straw berry Poison Frog, Dendrobates pumilio (Anura: Dendrobatidae). Copeia 1991(3): 723-730. Duellman, W. E. 1966. Aggressive behavior in Dendro batid frogs. Herpetologica 22: 217-221. Duellman, W. E and Trueb, L. 1986. Biology of Amphi bians. McGraw-Hill, New York. Emlen, S. T. 1968. Territoriality in the bullfrog Rana catesbiana Copeia 1968:240-243. Emlen, S. T., and Oring, L. W. 1977. Ecology, sexua l selection, and the evolution of mating systems. Science 197:215223. Falls, J. B. and McNicholl, M. K. 1979. Neighbor-st ranger discrimination by song in male blue grouse. Canadian Journal of Zoology 57 : 457-462.
Forester, D. C, Cover, J., and Wisnieski, A. 1993. The influence of time of residency on the tenacity of territorial defense by the Dart-Poison Frog Dendrobates pumilio Herpetologica 49(1): 94-99. Fox, S. F. and Baird, T. A. 1992. The dear enemy ph enomenon in the collared lizard, Crotaphytus collaris with a cautionary note on experimental methodology. Animal Behaviour 44:780-782. Fritz, G., Rand, A. S., and dePamphilis, C. W. 1981 The aposematically colored frog, Dendrobates pumilio is distasteful to the large, predatory ant, Paraponera clavata Biotropica 13: 158159. Gardner, E. A. and Graves, B. M. 2005. Responses of Resident Male Dendrobates pumilio to Territory Intruders. Journal of Herpetology 39 (2): 248-253. Grant, T., Frost, D. R., Caldwell, J. P, Gagliardo, R., Haddad, C. F. B., Kok, P. J. R., Means, D. B., Noonan, B. P., Schargel, W. E., a ndWheeler, W. C. 2006. Phylogenetic systematics of dart-poison frogs and their relatives (Amphibia: Athesphatanura: Dendrobatidae). Bulletin of the American Museum of Natural History 299: 1-262. Graves, B. M. 1999. Diel activity patterns of the s ympatric Poison Dart Frogs, Dendrobates auratus and Dendrobates pumilio in Costa Rica. Journal of Herpetology 33(3): 375-381. Graves, B. M., Stanley K. A., Gardner, E. A. 2005. Correlates of Vocal Display in a Costa Rican Population of Strawberry Poison-Dart Frogs, Dendrobates pumilio Journal of Herpetology, 39(1):101-107. Hagemann, S. and Prhl, H. 2007. Mitochondrial par aphyly in a polymorphic poison frog species (Dendrobatidae; D. pumilio). Mo lecular Phylogenetics and Evolution, 45:740 747. Jaeger, R. G. 1981. Dear enemy recognition and the costs of aggression between salamanders. American Naturalist 117:962-974. Kaufman, J. H. 1983. On the definitions and functio ns of dominance and territoriality. Biol. Revue 58: 120. Limerick, S. 1980. Courtship Behavior and Ovipositi on of the Poison-Arrow Frog Dendrobates pumilio Herpetologica 36(1): 69-71. Lopez, P. and Martin, J. 2002. Chemical rival recog nition decreases aggression levels in male Iberian Wall Lizards, Podarcis hispanica Behavioral Ecology and Sociobiology 51: 461-465.
n Ltters, S., Jungfer, K. H., Henkel, F. W., and Schmidt, W. 2007. Poison frogs: Biology, species and captive husbandry. Edition Chi maira, Frankfurt am Main, Germany. Maan, M. E. and Cummings, M. E. 2008. Female Prefer ences for aposematic signal components in a polymorphic poison frog. Evo lution 62(9): 23342345. Martof, B. S. 1953. Territoriality in the green fro g Rana clamitans Ecology 34:165-174. McVey, M. E, Robert, Z. G, Perry, D. and MacDougal, J. 1981. Territoriality and homing behavior in the poison-dart frog ( Dendrobates pumilio ). Copeia 1981(1): 1-8. Meuche, I., Lisenmair, E. K., and Prhl, H. 2012. Intrasexual competition, territoriality and acoustic communication in male s trawberry poison frogs ( Oophaga pumilio ). Behav. Ecol. Sociobiol. 66:613621. Noonan, B. P., Schargel, W. E., and Wheeler, W. C. 2006. Phylogenetic systematics of dart-poison frogs and their relative s (Amphibia : Athesphatanura : Dendrobatidae). Bull. Am. Mus. Nat Hist. 6262. Pough, F. H., Janis, C. M, Heiser, J. B. 2002. Vert ebrate Life, 8th Ed. Prentice Hall Inc., Upper Saddle River, NJ. Poulton E. B. 1890. The colours of animals: their m eaning and use, especially considered in the case of insects. Kegan Paul, Lond on. Pramuk, J. B., and Hiler, B. I. 1999. An investigat ion of obligate oophagy of Dendrobates pumilio tadpoles. Herpetological Review, 30: 219221. Prhl, H. 1995. Territorialund Paarungsverhalten von Dendrobates pumilio. Diplomarbeit (Master's thesis). Tierrztliche Hochs chule Hannover, Germany. Prhl, H. 1997. Territorial behavior of the strawbe rry poison-dart frog, Dendrobates pumilio (Anura:Dendrobatidae). Rev. Biol. Trop. 45:16691676. Prhl, H. and Hdl, W. 1999. Parental investment, p otential reproductive rates, and mating system in the strawberry dart-poison fro g, Dendrobates pumilio Behavioral Ecology and Sociobiology 46(4): 215-22 0.
Prhl, H and Berke, O. 2001. Spatial distribution o f male and female Strawberry Poison Frogs and their relation to female reproduct ive resources. Oecologia 129:534542. Prhl, H. 2003. Variation in Male Calling Behaviour and Relation to Male Mating Success in the Strawberry Poison Frog ( Dendrobates pumilio ). Ethology 109(4): 273290. Prhl, H. 2005. Clutch loss affects the operational sex ratio in the strawberry poison frog Dendrobates pumilio. Behavioral Ecology and Sociobiology 58: 310-315. Prohl, H., Hagemann, S., Karsch, J., and Hobel, G. 2007. Geographic variation in male sexual signals in Strawberry Poison Frogs ( Dendrobates pumilio ). Ethology 113(9):825837. Reynolds, R. G, and Fitzpatrick, B. M. 2007. Assort ative mating in poison-dart frogs based on an ecologically important trait. Evo lution 61: 2253-2259. Rudh, A., Rogell, B. and Hglund, J. 2007. Non-gr adual 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, 16:42844294. Saporito, R. A., Garraffo, H. M., Donnelly, M. A., Edwards, A. L., Longino, J. T., and Daly, J. W. 2004. Formicine ants: an arthropod source for pumiliotoxin alkaloids of dendrobatid poison frogs. Proc. Natl. Acad. Sci. USA, 101:8045-8050. Saporito, R. 2007. The relationship between alkaloi d-based chemical defenses and diet in dendrobatid poison frogs. PhD dissertation, Florida International University, Miami, FL. Saporito, R. A, Donnelly, M. A., Madden, A. A., Gar raffo, H. M., and Spande, T. F. 2010. Sex-related differences in alkaloid chemic al defenses of the Dendrobatid Frog Oophaga pumilio from Cayo Nancy, Bocas del Toro, Panama. Journal of Natural Products 73(3): 317-321. Savage, J. M. 1968. The dendrobatid frogs of Centra l America. Copeia 1968(4): 745-776. Savage, J. M. 2002. The Amphibians and Reptiles of Costa Rica: A Herpetofauna between two Continents, between two Seas. Universit y of Chicago Press, Chicago.
Schmidt, O. 1857. Diagnosen neuer Frosche des zoolo gischen Cabinets zu Krakau. Sitz. Math-Naturwiss. Akad. Wissen. Wien 24 (1): 10-15. Sexton, O. 1960. Some aspects of the behavior and o f the territory of a dendrobatid frog, Prostherapis trinitatus Ecology, 41(1):107-115. Siddiqi, A., Cronin, T. W., Loew, E. R., Vorobyev, M., and Summers, K. 2004. Interspecific and intraspecific views of color sign als in the strawberry poison frog Dendrobates pumilio J. Exp. Biol. 207:24712485. Smith, S. Q. and Jones, T. H. 2004. Comment. Tracking the cry ptic pumiliotoxins. Proc. Natl. Acad. Sci. USA, 101:7841-7842. Smith, S. Q. and Jones, T. H. 2004. Comment. Tracki ng the cryptic pumiliotoxins. Proc. Natl. Acad. Sci. USA 101:78417842. Staudt, K., Ospina, S. M, Mebs, D., and Prhl, H. 2 010. Foraging behaviour and territoriality of the strawberry poison frog ( Oophaga pumilio ) in dependence of the presence of ants. Amphibia-Reptil ia 31(2):217-227. Summers, K., Bermingham, E., Weigt, L., and McCaffe rty, S. 1997. Phenotypic and genetic divergence in three species of dart-poi son frogs with contrasting parental behavior. J. Hered. 88:813. Summers, K., Symula, R., Clough, M., and Cronin, T. 1999. Visual mate choice in poison frogs. Proc. R. Soc. Lond. B Biol. Sci. 2 66:21412145. Summers, K., Cronin, T. W., and Kennedy, T. 2003. V ariation in spectral reflectance among populations of Dendrobates pumilio the strawberry poison frog, in the Bocas del Toro Archipelago, Pan ama. J. Biogeogr. 30:3553. Szelistowski, WA. 1985. Unpalatability of the poiso n arrow frog Dendrobates pumilio to the ctenid spider Cupiennius coccineus. Biotrop ica 17(4):345346. Test, F. H. 1954. Social aggressiveness in an amphi bian. Science, 120 (3108):140-141. Wagner W. E. 1989. Graded aggressive signals in Bl anchard's cricket frog: vocal responses to opponent proximity and size. Anim Beha v 38:10251038.
Wang, I. J. and Shaffer, H. B. 2008 Coloration is a highly labile trait in an aposematic species: A phylogenetic analysis of colo r evolution in the strikingly polymorphic strawberry poison-dart frog. Evolution 62:2742-2759. Wells, K. D. 1978. Courtship and parental behavior in a Panamanian poisonarrow frog ( Dendrobates auratus ). Herpetologica 34:148-155. Wells, K. D. 2007. The ecology and behavior of amph ibians. The University of Chicago Press, Chicago. Weygoldt, P. 1980. Complex brood care and reproduct ive behaviour in captive poison-arrow frogs, Dendrobates pumilio O. Schmidt. Behavioral Ecology and Sociobiology 7:329-332. Wiewandt, T. A. 1964. Vocalization, aggressive beha vior, and territoriality in the bullfrog, Rana catesbiana Copeia 1964:276-285. Wong, S., Parada, H., and Naris, P. M. 2009. Hetero specific Acoustic Interference: Effects on Calling in Oophaga pumilio. Biotropica 41(1):7480. Zimmermann, E. 1990. Behavioral signals and reprodu ction modes in the neotropical frog family Dendrobatidae. In: Hanke, W Ed. Biology and Physiology of Amphibians. Fischer, Stuttgart, 61-73