To metamorphose, or stay as a child: that is the question…for newts, but not frogs

Roger Downie, Froglife and University of Glasgow

The standard life cycle in urodele and anuran amphibians (newts, salamanders; frogs and toads) has adults mating in freshwater, laying eggs there, which develop into  larvae (carnivorous in urodeles; usually herbivorous in anurans) which grow until, at a certain size, they metamorphose into juveniles, losing their larval characteristics, and moving on to land; there they grow further until ready to mature into adults.

There are many variants on that norm, for example, eggs that develop on land directly into juveniles without going through larval or metamorphic stages. In a previous ‘croak’ (Downie, 2026) I discussed another variant: neoteny and paedomorphosis where, in animals with complex life histories, having distinctly different larval and adult forms, maturation occurs in the larval form, cutting out the metamorphosis where the larval form is normally transformed into the adult. Denoel et al. (2005) report that paedomorphic development is the norm in 57 species of urodele amphibians (newts and salamanders). It is ‘obligate’ (occurs all the time) in four complete families and also in a few species from other families. This means that the adults are aquatic, retain external gills and gill slits and also tail fins. A well-known example is the Mexican Axolotl.

However, there is another intriguing variant in urodeles where, in a species that mostly shows the standard life cycle, some individuals are paedomorphic. We term this ‘facultative’ paedomorphosis, meaning that it is not the norm, but occurs in response to some sort of decision or stimulus.

Does this occur in our UK species? It has been reported in both smooth and palmate newts though not commonly, and not in great-crested newts (Beebee & Griffiths, 2000). Paterson (2017) found a single example of a paedomorphic palmate newt in a SUDS pond in East Kilbride. Banks (1985) reported the fairly common occurrence of paedomorphic smooth newts from a site near Sunderland in the 1960s and 1970s. Allain & Smith (2017) found paedomorphic smooth newts in a Cambridgeshire pond and later, Allain & Phillips (2023) reported a high proportion of paedomorphic smooth newts in a disused residential swimming pool in Norfolk. It is possible that paedomorphosis occurs more commonly in our British species, but has simply not been reported.

The most sustained set of reports on facultative paedomorphosis in Europe derives from a population of palmate newts in the Larzac area of southern France, studied over many years by Denoel and colleagues. The region is a limestone plateau 500-800 metres above sea level. Denoel (2007) surveyed 174 ponds during the newt breeding seasons of 2002-2006. He found paedomorphic individuals in 46 ponds, with numbers ranging from 1-277. Twenty of the ponds each had more than 50 paedomorphs, so this is not an occasional occurrence at this site.

Denoel et al.(2005) discuss paedomorphosis in palmate newts as an example of phenotypic plasticity: i.e. where the final form of an animal varies according to a decision made at some point during its development. In the case of paedomorphosis, the decision involves the costs and benefits of metamorphosis. Metamorphosis takes time and has energy costs: it also involves a major change in way of life and habitat from the known (the birth pool) to an uncertain new one in the surrounding terrestrial environment.

Mathiron et al. (2017) carried out a laboratory experiment on a sample of paedomorphic newts from Larzac in the context of two field findings: a) paedomorphic males are less frequent than females; b) summer drought is associated with the disappearance of paedomorphs. In the laboratory, they simulated drought conditions by reducing water levels and also by raising temperature. They found that paedomorphs responded to the reduced water levels by metamorphosing and to increased temperature by metamorphosing earlier. The responses were sex-biased: males were more likely than females to metamorphose at any particular water level or temperature, according with the field observations. The authors discussed their results in the context of climate change, bringing in hotter, drier summers, as have occurred in the last couple of years across Europe, and altering the balance between the benefits of paedomorphosis and metamorphosis in the Larzac population.

Denoel et al. (2026) measured the costs of metamorphosis to palmate newts collected in Larzac. They induced metamorphosis by reducing water levels in the laboratory and compared weight trajectories in individuals which metamorphosed with those that retained the paedomorphic phenotype. At the low water level 33 from 40 metamorphosed; at high water, only 6 of 40 metamorphosed. Metamorphosis had costs in body weight loss and in reduced food consumption: these costs were greater for females than for males.

I have presented merely a snapshot of the work done on the Larzac newts, which Denoel has advocated should become a population protected for their intraspecific biodiversity.

Both obligate and facultative paedomorphosis provide urodeles with flexible ways to alter their life histories, so it initially seems surprising that paedomorphosis is absent from frogs and toads: why should this be? Wassersug (1975) tackled this question. In urodeles, larvae and adults are not so different from each other. The larvae have four walking legs, like the adults, and both life history stages are carnivorous (though their food capture techniques are somewhat different). Both have slender elongate bodies with tails, so metamorphosis in urodeles does not involve such a drastic set of changes. In evolutionary terms, this makes it relatively easy to eliminate metamorphosis and mature in the larval form.

However, this is not the case for anurans, where the larval and adult forms are strikingly distinct. The tadpole larva has no visible limbs till near the time of metamorphosis; it swims by means of a long wide tail; food is mostly microalgae and biofilms sucked into the wide mouth and filtered from the water; the larval teeth are unrelated to the adult teeth. All these features and more are re-modelled during metamorphosis which, because of the major changes involved, takes significant time and energy. The major differences between a tadpole and an adult anuran make it hard to see how a tadpole could become reproductively active.

In most anurans, metamorphosis occurs  when the tadpole is still relatively small compared to the adult, and a substantial period of post-metamorphic growth is needed before maturation: this again would make paedomorphosis difficult. The nearest an anuran comes to a paedomorphic-like life history occurs in the remarkable ‘paradoxical frogs’ of South America and Trinidad, where I was fortunate to be able to study them (Downie et al., 2009 a,b).  There are 11 species in the genus Pseudis, all fully aquatic as adults with powerful swimming hindlimbs and webbed feet. They are evolutionarily related to the treefrogs of the Family Hylidae, although they do not look much like them. The stunning feature of these frogs is their tadpoles, which grow to a huge size, metamorphosing into smaller adults, quite the opposite of the usual anuran life history.   In the Pseudis paradoxa tadpoles we measured, the maximum length, including tail, at metamorphosis was 15-23 cm (i.e. up to 9 inches long), with a snout to hindlimb length of 4-5cm. The females were larger in mass and length than the males at this stage. Adult males and females collected in the field were a little larger than the metamorphs, but not much, so these tadpoles are at almost adult size when they metamorphose. Examination of the reproductive organs of  P. paradoxa late stage tadpoles shows that sperm development is well advanced and that in the ovaries, pre-yolk formation oocytes are fully developed. We do not know when these animals first reproduce after metamorphosis, but examination of their anatomy suggests that males are capable of reproducing almost immediately, and females soon after feeding and resultant yolk formation.

It is usually thought that there are serious limits on tadpole size: the nutritional content of their food is low; their swimming ability is poor compared to fish, so they are highly vulnerable to predators like herons; they often inhabit temporary water bodies which may dry up before a tadpole can reach a substantial size. It is not really clear how Pseudis  has been able to overcome these limits: their food is aquatic vegetation; they live in permanent swamps, avoiding the desiccation problem, but where predators are abundant. They tend to inhabit the tall swamp fringing vegetation, so perhaps that protects them from predators. A mystery for further research…

Click here for references