The True Poison-Dart Frog: The Golden Poison Frog
Phyllobates terribilis
By: Sean K. Stewart

Introduction to Poison-Dart Frogs

The American tropics from Nicaragua in Central America, south to about 20 degrees latitude in South America, are home to a fascinating family of frogs that have captured the attention of naturalists, scientists, photographers and hobbyists the world over. The family, Dendrobatidae, comprises at least 170 species of frogs that occupy a wide range of habitats. Some species live on the ground in lowland or montane rain forest; other forest species live in trees. Species can be found along streams, away from water or even in open, dry country where they find sufficient humidity on shaded ground under vegetation. Included in this family are the frogs commonly referred to as “poison-dart” frogs of the genera Dendrobates, Phyllobates, Epipedobates, Minyobates, Colostethus and Aromobates (1).

The belief that all the brightly colored Dendrobatid frogs are used for poisoning “arrows” is a modern literary myth (2). In fact, of the 170 known species of Dendrobatids, only three extraordinarily toxic species of Phyllobates from rain forest on the Pacific drainage of western Colombia are known to be used for poisoning blowgun darts (not arrows) for hunting and protection from jaguars, the Panthera onca (not warfare). These frogs secrete a poison that is stronger than the toxic mixture of “jungle plants” called curare. Curare is known to relax muscles into a state of inactivity and/or cause debilitating hallucinations. The curare prepared by the Macushi Indians of southwestern Guyana retained its toxicity for 170 years (3)! Just as curare is a widely used poison for arrows and blowgun darts east of the Andes, dendrobatid toxins serve as dart poisons in western Colombia.

A Historical Perspective of the Genus Phyllobates

In 1823-24, British naval Captain Charles Stuart Cochrane documented his travels through Colombia in a journal. As he crossed the western Andes on foot he encountered a frog “about three inches long, yellow on the back, with very large black eyes” called the rana de veneno by the Spanish. Cochrane describes the use of these frogs as follows:

“Those who use this poison catch the frogs in the woods, and confine them in a hollow cane, where they regularly feed them until they want the poison, when they take one of the unfortunate reptiles, and pass a pointed piece of wood down his throat, and out one of his legs. This torture makes the poor frog perspire very much, especially on the back, which becomes covered with a white froth: this is the most powerful poison that he yields, and in this they dip or roll the points of their arrows, which will preserve their destructive power for a year. Afterwards, below this white substance, appears a yellow oil, which is carefully scraped off, and retains its deadly influence for four to six months, according to the goodness (as they say) of the frog. By this means, from one frog sufficient poison is obtained for about fifty arrows (4).”

These “arrows” were about eight inches long, sharpened to a point and cut like a corkscrew at the top end. This spiral screw was rolled in the poison. The “arrow” was then blown out of a twelve-foot long reed shaft and could travel for one hundred yards with great accuracy.

“A tiger when hit, runs ten or a dozen yards, staggers, becomes sick, and dies in four or five minutes. A bird is killed as with a bullet; and the arrow and wounded part of the flesh being cut out, the reminder is eaten without danger.”

These frogs Cochrane described a century ago are thought to be Phyllobates bicolor and Phyllobates aurotaenia. Both frogs are found along the San Juan River and their secretions have been used by the northern Noanama Choco’ Indians to poison their blowgun darts. The southern Embera’ Choco’ Indians, however, poison their darts simply by wiping them across the back of a living specimen of the most toxic dart-frog known, the Phyllobates terribilis. All three of these toxic species of dendrobatids are confined to a relatively small area in western Colombia. Therefore, the use of dendrobatid frog secretions to poison blowgun darts is not as widespread as reported (5). There are other species of Phyllobates that exist in Central America, outside the range where dart poisons are known to have been employed, and are much less toxic than their Colombian relatives.

(Genus Phyllobates Distribution Illistration)

Phyllobates terribilis
Natural habitat

Phyllobates terribilis is found in lowland rainforest (100-200 meters elevation) of pacific coastal Colombia. It occurs in rough, hilly landscapes at the western foot of a northerly inclined spur of the Cordillera Occidental. The humid forest is believed to receive at least 5 meters of rainfall per year and the forest is broken as a result of stream dissection. The forest tends to be open on gravel slopes that are usually wet due to seepage; slopes tend to be steep and hillside soils are often covered in gravel. Leaf litter is reported to be sparse. The main ground vegetation is composed of saplings and treelets, small palms, herbaceous plants and ferns (6).

The Phyllobates terribilis are collected from the upper Rio Saija drainage in the vicinity of Quebrada Guangui’ and at La Brea. The morphology and toxicity of these frogs are uniform except for a microgeographic color variation which ranges from golden-yellow to deep orange to a metallic silver-green. The metallic silver green morph is generally found in the locality of La Brea in forest on the riverfront. Learning the true extent of their range has been hindered by the risks of surveying land owned by drug cartels of Colombia.

                                (Photos courtesy of Sean Stewart, Ron Gagliardo and Derek Rader)


Phyllobates terribilis is a diurnal species that is terrestrial in habit. I have not observed captive terribilis climbing up the sides of their enclosure or climbing on the enclosure’s vegetation frequently. They are bold frogs that seem unconcerned with hiding.

Phyllobates terribilis attains a snout-vent length averaging 47mm both in the wild and in captivity, making it one of the largest poison dart frogs known. In captivity females tend to be slightly larger than males (47mm and 45mm respectively). This dimorphism is probably seen in the wild, where females can reach 50mm, snout to vent. The large size and potent toxicity of this frog is manifested in its captive behavior. Compared with other species of dendrobatids, the terribilis is a bold frog that is not intimidated by human on-lookers. When feeding my terribilis, I often shake crickets into their enclosure from above to watch them jump at the cricket container and strike crickets out of mid air!


The Phyllobates terribilis are brightly colored frogs for a good reason. The brilliant color is an advertisement to warn potential predators of their toxicity. These frogs release their toxins from microscopic skin glands at times of stress. When a predator has seized a terribilis in its mouth, the predator will at least experience a foul taste and probably numbness and burning as well.

Phyllobates terribilis are the most toxic species of frog; twentyfold more toxic than other poison dart frogs (7). During the evolution of poison-dart frogs piperidine alkaloid synthesis, found in all groups of toxic dendrobatids, was suppressed in favor of batrachotoxins which are extraordinary toxic alkaloids (8). One wild terribilis specimen may contain up to 1900 micrograms of toxins! Myers and Daly, indirectly judged the human lethal dose of this toxin to be in the range of 2-200 micrograms. The poison would be lethal to humans if it entered an open wound.

The non-protein toxin increases the permeability of the outer membranes of nerve and muscle cells. Specifically, the alkaloids prevent sodium channels from closing, thus preventing nerves from electrically transmitting impulses and therefore, muscles remain in an inactive state of contraction. Ultimately, the heart experiences arrhythmias which can lead to fibrillation and heart failure. Amazingly, in Phyllobates terribilis a sodium channel regulatory site in nerve and muscle cells seems to be altered slightly to confer self-protection against their own toxins. However, the terribilis remain sensitive to similar acting plant toxins to which they have never been exposed. This specific adaptation at the molecular level permits the “terrible frog” to host toxins for protection from almost all predators while not adversely affecting themselves. This is just one of many illustrations of how nature provides astonishing examples of specialized adaptations that allow living things to outrival each other by utilizing resources. But even the terribilis has natural predators. A small frog-eating snake, the Leimadophis epinephelus, shares the same range as all the species of Phyllobates and can eat juvenile terribilis with no ill effects. The exact protective mechanism this snake employs is unknown. Detoxification starting with the snake’s salivary secretions has been proposed.

Interestingly, captive hatched and reared terribilis do not express any toxicity making this species safe to hobbyists. In their natural environment, poison-dart frogs secrete toxins on their skin by metabolizing specific organic substrate(s) provided by their natural diet of ants and other insects. However, captive hatched and raised frogs do not produce toxins because their captive diet lacks the necessary ingredient(s) to produce these toxins.

Phyllobates terribilis in captivity

In October of 1997, I received 8 captive-born terribilis from a German breeder. They were approximately 1.5 cm in length and still in there juvenile coloration. At this time the juvenile terribilis were black along the dorsum and venter with gold dorsolateral striping. The general appearance of all 8 specimens is of a black frog with paired stripes of gold. They were of the silver-green color variety. They were reared together as a group in a 2x2x2 foot enclosure and were not aggressive toward one another. I raised them at 75-85?F, with a 12-hour photoperiod and daily mistings to maintain humidity above 70% relative humidity.

It has been reported that the age of first reproduction is approximately 13 months. In my experience, Phyllobates terribilis began to reproduce at approximately 18 months of age. This sexual maturation may correlate to body size rather than age. Males can be sexually mature at 37 mm, where females mature at about 40-41 mm. Males use two trill calls to engage females in courtship behavior: a “long, melodious trill” lasting 6-7 seconds along with an abbreviated 2-3 second version. The trill call has been described as uniform trains of notes uttered at a rate of 13 per second. The frequency is about 1800 Hz, a little lower than the frequency of calls in related species (P. bicolor, aurotaenia, lugubris and vittatus).

Cephalic amplexus in which the dorsal surfaces of the male’s hands are pressed against the female’s throat has been observed by some. I have observed multiple males calling, which is my first clue of courtship behavior. The interested frogs move around together before they enter a breeding hut. No amplex positioning has been observed. Rather, it seems that the female lays the eggs in the presence of a male and he fertilizes them soon after.

Clutch size range from 8-18 with the average clutch size being 13-14 eggs. I have found clutches in excess of 30 eggs, but assume this is from multiple females laying. I have, on many occasions, observed multiple males and females crowding amongst each other in a small breeding hut. There were no recognizable behavioral patterns in this group of adult frogs but a breeding frenzy was observed.

The eggs are small (2.5 mm excluding the jelly) and like to be well hydrated. I found that clutches that are not submersed in water and are kept drier do better than clutches that are wet. I have had only about 50% success with hatching eggs; losing many eggs within the first week. They look as if they rupture and turn a milky-beige color. In their natural habitat males carry the tadpoles to streams. Tadpoles take 9-12 days to develop and hatch-out very small (8-11mm) compared to Dendrobates species of similar adult size (i.e. D. tinctorious and azureus). At the time the larvae hatches, branching gills are apparent. Gills reduce in size soon after and the tadpole grows a few millimeters before it begins to feed. Hatchling tadpoles are grayish-brown in color and the body pigmentation becomes paler as development continues. Another pigmentation change is seen corresponding with limb development. The grayish tadpole darkens to almost black as metamorphosis begins. Tadpoles begin to feed about 6 days after hatching. At 75?F, terribilis tadpoles metamorphose in approximately 60 days, when they are approximately 15-20 mm. At metamorphosis, the tadpole has acquired bronze flecking which is concentrated dorsolaterally forming paired stripes; the body appears black at the time the forelimbs appear. The dorsolateral stripes and upper limbs become bright gold as the frog matures and the flecking in the lower limbs becomes golden bronze.

I rear my tadpoles communally in rubbermaid tubs with pothos used for cover. The tadpoles are not cannibalistic. I feed the tadpoles daily, with either Aquarian vitamin and mineral flake or a 3:2:1 mixture of spirulina, chlorella and sera micron.

I raise the froglets on pin-head crickets and fruit flies. They are aggressive feeders and grow quickly. I prefer daily feedings of crickets gut-loaded with a high-calcium diet. I recommend raising no more than 15 juveniles in a 10-gallon aquarium for the first 3 months. Then, 2-4 juveniles per 10-gallon enclosure. There have been rumors that terribilis froglets were overly sensitive to fecal build-up in crowded terrariums, but I have not noticed any hypersensitivity when kept as above. The dorsolateral stripes remain well defined in juveniles until they reach a snout-vent length of 15 mm. At this size lighter pigment appears dorsally between the stripes which also widen ventrally. Ventral surfaces remain essentially black. By 20 mm snout-vent length, the juvenile terribilis is almost uniformly colored on its back and sides and the ventral surface is also picking up pigment. It seems the ventral color change lags a few weeks behind. This pale green color becomes more brilliant as the frog increases in size. Some adults grow into a bluish-green frog, whereas others lose most of their greenish hue and become completely silver!

I maintain breeding colonies as a group of six. I feed them 3/8-inch crickets at least every other day. Ideally, I try to feed a little everyday. The crickets are gut-loaded with a high-calcium cricket diet that I further supplement with Rep-Cal’s phosphorous-free calcium and vitamin D3, Rep Cal’s Herptivite multi-vitamins and Quintrex redaxanthin.

Sexing adult terribilis is not an exact science. Females tend to be larger and I have not heard a female call. I do not rule out the possibility that female terribilis vocalize, because of what I have experienced with female P. bicolor. I have observed a female bicolor become vocal (a short, low-pitched blow sound) when stimulated by the male’s call. If the females vocalize it would be expected to be a much different call in frequency, rhythm and duration to that of the males. Males can often be best identified by the presence of black discoloration under the jaw as a result of vocalization (vocal folds).

I have found keeping Phyllobates terribilis in my collection to be a fascinating experience. They are breathtaking in size and beauty and some of their behaviors are unique among poison-dart frogs. They reproduce well in captivity and the tadpoles metamorphose quickly compared to the Dendrobates, but egg husbandry does remain a challenge. On the other hand, taking care of these hardy tadpoles communally makes rearing easy. They are bold, diurnal frogs that allow the keeper unlimited observation into the world of the true poison-dart frog.

Literature Cited

1. Canela, J. and Vazquez, J. 1998. Multicoloured Poison Frogs, Dendrobates. Reptilia 2 (March): pp. 19.
2. Myers, C.W. and Daly, J.W. 1993. Tropical Poison Frogs. Science 262: pp. 1193.
3. Plotkin, M.J. 1993. Tales of a Shaman’s Apprentice. Penguin Books. New York, New York. pp. 131.
4. Cochrane, Capt. Charles Stuart 1825. Journal of a residence and travels in Colombia during the years 1823-1824. In two volumes. London, Henry Colburn, vol. 1, xvi + pp. 524.; vol. 2, xvii + pp. 517.
5. Myers, C.W. and Daly, J.W. 1983. Dart-Poison Frogs. Scienctific America 248 (2): pp. 120-131.
6. Myers, C.W., Daly, J.W. and Malkin B. 1978.A Dangerously Toxic New Frog (Phyllobates) Used By Embera’ Indians of Western Colombia, With Discussion of Blowgun Fabrication And Dart Poisoning. Bulletin of the American Museum of Natural History. 161: article 2, pp. 309-365.
7. Myers, C.W. and Daly, J.W. 1983. Dart-Poison Frogs. Scienctific America 248 (2): pp. 120-131.
8. Myers, C.W., Daly, J.W. and Malkin B. 1978.A Dangerously Toxic New Frog (Phyllobates) Used By Embera’ Indians of Western Colombia, With Discussion of Blowgun Fabrication And Dart Poisoning. Bulletin of the American Museum of Natural History. 161: article 2, pp. 309-365.