Locally, the Asian toad (Bufo melanostictus), the introduced American bullfrog (Rana catesbeiana) and a few other mid-size ranids such as the copper-cheeked frog and common greenback are probably the most frequently encountered anurans or tail-less amphibians. There is, however, one other frog that exists even in fairly urbanised area where ample damp shelters exist. The chubby banded bullfrog (Kaloula pulchra) is a rotund little thing with a rude belch of a croak. The species epithet means 'beautiful' in Latin but the animal's attractive mottling probably serves to camouflage the creature in its leaf litter home rather than betray its pretty hide to frog-eating fowl.
It's not clear why this cute frog acquired the moniker bullfrog, which is more appropriate for its robust American cousin that has taken over a number of local habitats and wrecked havoc on both human and animal life in other ecosystems. As with its compatriot the red-eared slider, the American bullfrog will (even if local breeding records are elusive) alas be a permanent fixture of local waterways, so long as the population is replenished by escapees from farms and intentional releases by the thoughtlessly devout. And soon, it's not implausible that we start to see obese green tree frogs clinging to roadside shrubs as carefree petowners tired of their exotic purchase.
The banded bullfrog is one of the largest members in the family Microhylidae. Also known as narrow-mouthed frogs for the slim snouts sported by many species, microhylids are the most diverse anuran family by genera, with about 70 genera across more than 420 species (the family encompasses just 8% of all frog species but 20% of total frog genera). Their current distribution is thought to suggest a Gondwanan origin and an early divergence from more derived anuran families. The highest number of species are found in the relatively isolated biogeographical spheres of Madagascar and islandic Southeast Asia, and the eastern fringe of the latter region harbours some of the most curious representatives of the family.
Guilds of egg guarders
The microhylids of Australasia (grouped in the subfamilies Asterophryinae and Genyophryninae) are distinguished by direct development, whereby eggs hatch into miniature froglets rather than textbook tadpoles. But this developmental foreshortening is just one facet of the strange reproductive adaptations that microhylids have evolved on the world's largest tropical island, New Guinea. As David Bickford, an evolutionary ecologist, discovered after four years in New Guinea's mountains, endemic microhylids have radiated into a number of ecological guilds and a few have taken up parental care strategies that belie the minuscule proportions of these quite tiny frogs.
Revealing the gist of his findings last Wednesday, Bickford noted that the study of parental care in amphibians offers tremendous scope for integrating different facets of biology including ecology, behaviour, evolution and reproductive biology. Much work in this field has been done on mammals and birds (it seems that for a long time, people had a hard time accepting that ectotherms, which have been around far longer than endothermic tetrapods, could display more than a rudimentary form of parental care), but in fact, amphibians demonstrate possibly not only the most diverse reproductive strategies amongst terrestrial vertebrates, but also the most variety in terms of modes of parental care (Crump M.L., Parental care among the amphibia. Adv. Stud. Behav. 1996; 25:109–144, see also this excellent dissertation chapter).
The stereotypical amphibian practices external fertilisation and is oviparous (egg-laying), but internal fertilisation, vivipary and ovovivipary are also known. Some require an aquatic component in their life-cycle; others have bypassed this need to varying degrees. There are an assortment of nest builders, egg guarders, species that transport their eggs and tadpoles, 'marsupial' frogs with pouches on their backs where the eggs are brooded, the production of trophic eggs to feed tadpoles, and most curiously, a gastric-brooding species which carries the eggs in the stomach until the froglets emerged from the mouth. Unfortunately, the ability of Rheobatrachus silus to regulate its production of digestive acids can no longer be studied as the frog and its closest relative are both extinct.
From his fieldwork in New Guinea, Bickford has observed two hitherto unknown modes of amphibian parental care: froglet attendance after the eggs hatch and froglet transport by the male who ferries his brood to locations away from the spawning site.
Besides their reproductive diversity, amphibians also pose an easier bounty for field biologists, who can in undisturbed habitats encounter the animals in staggering densities. And in places like New Guinea, where new species of birds, butterflies, plants as well as mammals thought extinct can still be found in remote forests yet to be logged, one can "walk out on any given night... and see a few hundred frogs, if you know where to look." Bickford adds that many of the species he worked with for his doctoral thesis were undescribed and one widely-distributed frog once thought to be conspecific may in fact be a complex of up to a dozen species.
In search of highland hoppers
To unravel why New Guinean microhylids evolved their modes of parental care and the forces that drove this radiation, Bickford based himself at the Crater Mountain Biological Research Station in Chimbu Province in the central highlands of Papua Niugini, where he recruited an army of local Pawaia tribesmen in his hunt for hoppers. He was able to group the frogs into six ecomorphs with distinct ecologies, diets, behaviour and morphologies, a phenotypic diversity that he notes is more often seen between families than within a single family.
Direct development from egg to froglet is a recurrent phenomenon in a number amphibian groups, and one could say that a logical progression of this trend was the emergence and radiation of amniotes. With frogs, Bickford suggests three driving forces that would favour direct development: predation, intra-specific competition and environmental factors. Freshwater habitats are rife with predators happy to make a meal out of spawn and every tadpole stage. Where free-standing water is at a premium, tadpoles compete against each other for food and space.
In the New Guinea highlands at elevations of 800-1,350 m, the sheer abundance of water in the form of year-round rainfall (Bickford's study area gets nearly 8 m of rain annually) coupled with steep topography causes torrential problems for tadpoles which risk getting washed away to unsuitable sites. The few species that do thrive as tadpoles there have evolved suctorial mouthparts to help them stick around. In this milieu, the microhylids that form the dominant frog fauna of New Guinea appear to have radiated from a phylogenetic trajectory that has made them completely independent of an aquatic environment.
Different selective pressures exist on terrestrial ecosystems. Lacking a waterproof capsule, amphibian eggs laid on land must be placed at sites that can keep them hydrated. Alternatively, the parent soaks the eggs in a regular bath of bladder water to prevent dessication. There are also land-based creatures that have a taste for tetrapod spawn. And as any visitor to the rainforest floor soon realises, every square inch of earth is patrolled by foraging ants for which frog eggs are a luscious treat. Parental care of some form, defined as any investment by the parent that takes place after the formation of the zygote that results in increased offspring survivorship, would thus confer a selective advantage to frogs that are also in the process of freeing themselves from the grip of a waterbound stage in their lifecycle.
Bickford's study subjects, as noted earlier, come in six ecological guilds. There are arboreal microhylids (mostly Oreophryne spp.) which never descend from the canopy and resemble the true tree frogs from the families Hylidae and Rhacophoridae. Scansorial microhylids (e.g. Cophixalus and Austrochaperina) are frogs that forage on the forest floor at night and retreat to the trees by day. Terrestrial ecomorphs like Liophryne schlaginhaufeni and the monotypic Sphenophryne cornuta spend their lives in the leaf litter. Fossorial frogs are adapted for burrowing headfirst, while cunicular microhylids dwell in tunnels where they await prey. And one group, the riparians, has returned to the water for food (though not to breed) and bear feet with webbing, a rather rare feature for this family of frogs.
Methods and findings
To study the frogs, Bickford used standard ecological methods: visual encounter surveys at night (600 transects) and 3,500 5x5 metre leaf litter plots. In situ behaviour was recorded by marking the sites where brooding frogs were found and checking the status of the clutch every day. To determine the form of parental care, the observer had to stay with a clutch for 24/7 for 3-4 days, "until we were pretty confident that we had seen everything that the frogs are going to do." This focal method was done for every species and for five individuals of each species. Hydration measurements were also taken of individual eggs to ascertain if the adults had a role in maintaining the moisture level of the clutches.
Experimental methods included introducing ants to observe parental response, physical disturbance (i.e. poking the frog on the nose with the blunt end of a pencil) as well as removal of the adult to find out what happens when there is no parental care. The latter required careful control so that each treated clutch and its paired control clutch were comparable: they had to be within one development stage of each other; differed in clutch size by no more than 15%, found in similar geographic locations and discovered within 3 days of each other.
The team found 243 clutches from 16 species in 10 genera. Just over three-quarters of the clutches were of 4 species: Hylophorbus rufescens, a terrestrial frog; the undescribed Oreophryne sp. 'Rattler', an arboreal species that lays its eggs on the underside of leaves; Liophryne schlaginhaufeni; and the fossorial Copiula sp., which distinguished itself in its readiness to commit oophagy (eating one's own eggs) when confronted by researchers. "It's almost immediate oophagy and a direct response to our disturbing them," recalls Bickford. "You would tip over a leaf or take off a piece of log and boo! he eats the eggs right in front of you and then tries to get away." More than half the clutches found of this species were devoured by the parent.
Male frogs attended 235 of the 243 clutches, while females guarded 6 clutches. The sex of the remaining two parents were unknown as they got away posthaste. With over 150 species in ± 20 genera, New Guinean microhylids are possibly the only large group of tetrapods when the male parental care predominates.
Zooming to the trait of egg attendance, Bickford found that this was practiced by all microhylid ecoguilds in his study site except for arboreal species, which demonstrate at most intermittent or nocturnal attendance. He postulates that this might be due to the absence of predation pressure in the treetops. Active egg guarding was likewise seen in all guilds except the arboreal frogs, which failed to perform any defensive acts like inflating their bodies, flicking away ants or actually eating the would-be predators. This suggests that under natural conditions, ant predation is not a selection factor for the treetop dwellers.
The parent removal experiments, which involved the ubiquitous Hylophorbus rufescens, confirmed that parental care is vital to egg survival, with all treated clutches perishing from ant predation. The same 100% mortality took place in the tree, but for different reasons. Hydration measurements on 19 clutches from the arboreal Oreophryne sp. yielded significant differences in capsular diameter between eggs attended by adults and those where the parent was removed. Sans parent, the eggs shrivel and die.
Two forms of oophagy were observed. One was defensive: the consumption of eggs infected by fungi or parasitised by drosophilid flies. The other was a reflex action to predator (or human) disturbance. All the species studied performed the first form but the arboreal frogs were the only ones that failed to eat their eggs before making a getaway.
Picture by David Bickford.
Froglet attendance, whereby the adult stays with the young after they hatch until the froglets disperse, was seen in most of the frogs except for one scansorial species and all the arboreal microhylids. And going one step further, Bickford's team found to their surprise that in three species, Liophryne schlaginhaufeni, Sphenophryne cornuta, and an undescribed Cophixalus species, the froglets actually hopped on to the father's back and got a free ride to new homes. Bickford suspects this behaviour to be more widespread than thought on the island and offers potential selective benefits stemming from lower chances of mass predation, reduced competition from greater dispersal and enhanced geneflow amongst populations with fewer opportunities for inbreeding between siblings. No other frog family is known to exhibit male froglet transport; the behaviour has only been reported for females of a single cave-dwelling species in Jamaica.
Bickford's phylogenetic analysis of New Guinea microhylids coupled maximum parsimony with two different probabilistic methods: maximum likelihood and Bayesian. The tree that he erected, while not the last word, indicates that the arboreal guild lies close to the base and probably split off early in the group's history. The scansorial frogs presented a fuzzy picture, as "these guys did not come up as a strong group" with two independent branches emerging. The monophyly of the riparian frogs appear certain (and they share a clade with one group of fossorial frogs), but both the fossorial and terrestrial guilds seem to be polyphyletic. One particular clade contains three different ecomorphs: terrestrial, cunicular and fossorial.
Based on his results, Bickford believes that the two currently recognised New Guinea microhylid subfamilies, Genyophryninae and Asterophryinae, are in need of serious revision and should probably be combined into a single subfamily, Asterophryinae. The genera Austrochaperina, Callulops, Cophixalus, and Oreophryne should also be regarded as polyphyletic.
The field studies and phylogenetic results suggest that for New Guinea microhylids, egg attendance is a plesiomorphic trait, while egg guarding is apomorphic, a derived behaviour that independently emerged as the tree branched off. Bickford also found no strong signal from parametric bootstrap analysis to refute the hypothesis of froglet attendance and froglet transport as a monophyletic event. Parental care is thus a primitive behaviour in this group of frogs that started with egg attendance and evolved in concert with direct development. The subsequent diversification into six ecological guilds could be due to the family's early (and near unchallenged) colonisation of New Guinea and success in filling unoccupied forest microhabitats.
Further exploration is called for on the two new modes of froglet parental care. The question also remains as to why males call the shots in this family. The island's microhylids also exhibit other bizarre adaptations. One group lacks tongues and uniquely, eat mushrooms. "There's something crazy going on there," quips Bickford. Other arenas for study include the biogeography of the genus Oreophryne, which is distributed with high endemicism from New Guinea to the Philippines. Bickford is also involved in an effort to uncover the existence of cryptic species amongst common Southeast Asian frogs, including some commercial important species that are likely to be in decline from overharvesting. Like the eggs of oophagic parents, these icons of ponds and padi fields are vanishing in leaps and bounds, no thanks to our appetite for the tender meat of these sweet chickens.