Offspring transportation in poison frogs: tadpoles approach adult frogs to escape competition and cannibalism

15 05 2017

Poison frog tadpoles seek parental transportation to escape their cannibalistic siblings

Lisa M. Schulte and Michael Mayer. (2017) Journal of Zoology, DOI: 10.1111/jzo.12472

Parental care can be found in many different animal taxa and it is highly beneficial for the survival of the offspring. This is also the case for Neotropical poison frogs: most species guard and moisten their terrestrial clutches and then transport their tadpoles to appropriate water bodies. The poison frog Ranitomeya variabilis, however, is a little bit different: it already deposits its eggs inside of phytotelmata (small water bodies within plants). Once the tadpoles are ready to hatch, the male typically returns to transport each of them separately into other phytotelmata. The separation is important because the tadpoles of this species not only compete for the same resources within the small pools, but are also cannibalistic among each other. However, on various occasions we found that males did not return for their offspring and let their tadpoles hatch all together into the same phytotelm.

Poison frog and tadpole

A male Ranitomeya variabilis transporting a tadpole. Photo by Lisa M. Schulte

This made us wonder if such abandoned tadpoles, facing competition and cannibalism, actively seek parental care in form of transportation whenever they get the chance – i.e. as soon as a frog enters their pool. In order to test this hypothesis, we conducted experiments where tadpoles of the same clutch that shared a pool with their siblings were presented with con- and heterospecific frogs, and both moving and non-moving frog models.

Experimental set-up

Our experimental setup demonstrating tadpoles approaching a conspecific frog. First published in Schulte & Mayer (2017)

Our results revealed that abandoned tadpoles actively approached the frogs and sometimes even climbed onto their backs – although the frogs had not entered the pools with the purpose of tadpole transportation. Furthermore, this behaviour was not restricted to the biological parents, and even unfamiliar species that normally occur in rivers and never enter phytotelmata were approached by the tadpoles. This leads us to the assumption that the benefits for the tadpoles to escape sibling competition and cannibalism are so high that they even accept the risk of undirected transportation by heterospecific frogs. The plastic models, however, did not trigger any behaviour, possibly because tadpoles do not recognize frogs by visual or tactile cues only, but might be dependent on chemical or multiple stimuli.

The latter is also the case in related egg-feeding species where tadpoles beg for food. We therefore suggest that our observations might be comparable to begging behaviours – with the difference that in our study system the tadpoles seem to be begging for transportation instead of food. Another approach to interpret our results is a potential parent-offspring conflict, where the tadpoles require more care than the adult frogs are willing to give. Both interpretation approaches open up a wide range of new questions, highlighting that our study system should be of great relevance for anybody interested in animal parental care.

Lisa M. Schulte

Recluse spiders and their highly modified, self-sufficient spinning apparatus

20 04 2017

Recluse spiders produce flattened silk rapidly using a highly modified, self-sufficient spinning apparatus

I. L. F. MagalhaesA. M. RaveloC. L. SciosciaA. V. PerettiP. Michalik and M. J. Ramírez. (2017) Journal of Zoology, DOI: 10.1111/jzo.12462


Silk is one of the most fascinating biological materials – it can be as tough as steel, and twice as elastic. This remarkable substance is exclusively produced by arthropods – such as silkworms, which are the larval stages of the moth Bombyx mori.  Among all arthropods, a particular group has achieved the highest complexity of the types of silk they produce and the uses they make of it – spiders. There are some 46,000 described species of spiders, and, without a single known exception, all of them are capable of producing silk. Spiders have up to four (but usually three) pairs of specialized organs at the end of their abdomens called “spinnerets”, which are mobile appendages associated with the glands that produce silk. Silk is secreted by these glands in the liquid state, and the very act of pulling it through the apertures of the silk ducts – called “spigots” – makes it change its conformation to solid. Each type of silk gland produces a different type of silk, and a single spider species may have as much as seven types of glands.


Figure 1. Scanning electronic microscopy of silk strands spun by the African recluse spider Loxosceles simillima.

Among spiders, a particular genus produces a unique kind of silk found nowhere else in the animal kingdom. The recluse spiders of the genus Loxosceles, infamous for containing some species whose venom can cause severe symptoms in humans, are also peculiar regarding their silk. Their anterior spinnerets have a modified spigot with a slit-like aperture – in comparison, regular spigots have a circular aperture. As a consequence, the silk threads they produce are extremely flattened (Fig. 1). Interestingly, they have only two of these modified spigots. However, their entire webs are composed of these flattened strands (Fig. 2). How is it possible for them to build such large webs from a single pair of spigots?


Figure 2. A female recluse spider Loxosceles amazonica sitting on her web. Photo courtesy of A. Anker and P.H. Martins

While observing females of Loxosceles laeta, a common South American species, we noticed they have a very peculiar spinning behavior. Their spinnerets seemed to move very rapidly, with silk accumulating at a fast rate; then the spider moved, stretching the silk strands (Video 1). Interestingly, they never used the legs or the substrate to pull the silk strands. As I mentioned above, silk changes from liquid to solid when it is pulled – thus, if neither the legs nor the substrate were being used to pull the strands, how was the spider spinning?

To investigate this, we immobilized some females by placing them in glass tubes and filmed their spinnerets while they produced silk. At first, we were amazed to see how fast they moved – about 8 to 13 times per second. When we reduced the speed of the videos, we were stunned – the spider was apparently using the posterior spinnerets to pull the silk strands produced in the anterior spinnerets! (Video 2) This was a much unexpected observation, because so far all arthropods use their legs or the substrate to pull silk strands. If Loxosceles was indeed using the spinnerets themselves to pull silk strands, this would be the first record of a self-sufficient spinning apparatus – that is, spinnerets that can complete all the work of spinning without external help.

Of course, we wanted to confirm this result. Therefore, we proceeded to manipulate specimens. We selected three females and interfered with their posterior spinnerets  – either by gluing them together so they could not move (but letting the anterior spinnerets free, so silk could still come out of them), or by removing the hairs from the posterior spinnerets, which we suspected might be important for pulling silk strands. If the posterior spinnerets were truly involved in silk spinning, then the manipulated specimens should be unable to produce webs normally, even if the anterior spinnerets had not been interfered with. And that is what we observed – even after several weeks, these manipulated spiders could not spin normal webs, but only rather thin meshes of threads, or no web at all (Fig. 3).


Figure 3. Comparison of the webs of manipulated recluse spiders (A, B) and an untreated specimen (C). The specimen in A has only one pair of posterior spinnerets glued together and could build a web, although much thinner than the normal webs (C). The specimen with both pairs of posterior spinnerets glued together (B) could spin no web at all. First published as Figure 3 in Magalhaes et al. (2017)

Loxosceles have highly modified spinnerets (Fig. 4). After comparing their morphology with that of closely related spider families, we could observe that these modifications are all related to their unique spinning behavior. For example, they have a unique type of hairs in the posterior spinnerets that allow them to pull the silk strands from the anterior spinnerets. The anterior spinnerets are unusually long, and separated from the posterior spinnerets by a notable distance – which we hypothesize might be related to pulling a longer strand of silk in each beat of the spinnerets. Finally, they also have a very different muscular system associated to spinnerets (Video 3), with long and thin muscles which allow them to move the spinnerets in a fast way.


Scanning electronic microscopy of the spinnerets of the recluse spider Loxosceles laeta in lateral (A) and ventral (B) views. At the tip of the lateral spinnerets (ALS), there is a modified major ampullata gland spigot (MAP; C). The modified silk strands (E, F) are pulled from this spigot by the posterior spinnerets (PMS, PLS) with the help of the curved setae (CS) and toothed setae (TS; D). First published as Figure 1 in Magalhaes et al. (2017)

We have found that Loxosceles not only has unique silk strands – they also have a unique spinning behavior, which is possible because of the morphological modifications in their spinnerets and spigots. More importantly, this allowed us to discover the first known arthropod capable of spinning using only its spinnerets, without help of the legs or the substrate – once more highlighting the beautiful diversity of the spinning work of spiders.

Ivan Magalhaes


Video 1: A female recluse spider spinning its web.

Video 2: The spinnerets of a recluse spider while at work, reduced to 20% of the original speed.

Video 3: A 3D reconstruction of the muscles associated to the spinnerets of a male recluse spider obtained using X-ray microscopy.


Morpho-dynamics and early development of interlimb coordination in baboons

29 03 2017

Intrinsic limb morpho-dynamics and the early development of interlimb coordination of walking in a quadrupedal primate

F. DruelleG. Berillon and P. Aerts; Journal of Zoology, Volume 301, Issue 3, pages 235–247, March 2017

The development toward locomotor autonomy is a long journey for primate (altricial) species. Although the process is multidimensional, the mechanical properties of the changing body are likely to influence the direction and the timing of this journey. Indeed, according to the biomechanical theory, the animal body can be seen as a machine the shape and size of which impact movement via their resistance to linear and angular acceleration. In this context, the mechanical properties, particularly at the level of the legs, are likely to facilitate and/or constraint the learning processes of locomotion.


Olive baboons (Papio anubis) in captivity. Photo by Francois Druelle

In order to explore the interactions between the intrinsic body mechanics and the manner in which primates improve their walking gaits, we followed a group of six young baboons living with their mothers and within their social group in captivity. The baboon is an interesting model for our aims because it is described as a committed quadruped and the locomotor development lasts around a year and a half. Therefore, a follow-up of these animals was possible in the context of a doctoral study. We were able to assess the interlimb coordination pattern during quadrupedal walking of the young baboons via the quantification of the temporal aspects of their gaits during their daily activities. We assessed the mechanical properties of their bodies via external measurements and the use of a geometrical model. The natural pendular period of a limb, i.e. the period of oscillation at which exchange between potential and kinetic energy would be maximal and metabolic energy would be minimal, is put forward for representing the intrinsic limb morpho-dynamics.


Figure 1 from the Druelle et al. (2017) article, demonstrating left forelimb touchdown (a and b) and lift-off (c and d). Photos by Francois Druelle

Our results are of great interest because they demonstrate the presence of an early optimization of the limbs in baboons. Indeed, during development, fore and hind limbs are strongly convergent at the level of their natural pendular period. Interestingly, this affects the coordination pattern but this effect does not interact with age, therefore revealing a parallel factor of the neuromuscular maturation. These limbs’ mechanical properties are likely to be adaptive in facilitating the learning processes toward quadrupedal walking in baboons. Indeed, an early acquisition of locomotor skills in young baboons should favor their survivorship. Female baboons usually resume sexual cycling 12 months after they gave birth involving that at this time their infant must be able to locomote efficiently on the ground in order to keep following their group. Our finding is therefore in line with the specific needs related to the baboons’ socio-ecological niche. These results also question the presence of other body optimizations in young primates according to their species-specific ecological niches.

François Druelle

New Journal of Zoology Podcast

10 03 2017

A new episode of the Journal of Zoology podcast is now available and you can listen to it here.

JZoologySpiderPodcastIn this episode, Travis DeVault talks to us about their experiment testing whether bird collisions with vehicles are affected by experience, we will learn from Fanny Ruhland about the brain of tarantulas and whether they show behavioural or morphological left-right asymmetry, and Daniel Rocha tells us about using baited camera traps to study carnivores and their prey in the Amazon.

You may subscribe to iTunes to receive the latest Journal of Zoology podcasts.

Author Spotlight – Alien versus predators: effective induced defenses in an invasive frog

28 02 2017

Alien versus predators: effective induced defenses of an invasive frog in response to native predators

E. Pujol-Buxó, C. García-Guerrero and G. A. Llorente

Have you ever asked yourself, as a biologist, when should a species be considered “invasive”? You may even have discussed it with your colleagues. And when is a species “allochtonous”? Seems easy to define from a geographic point of view, but… what does an evolutionary point of view have to say about it? When does a species behave as allochtonous? Will an introduced population of a species ecologically act as allochtonous if it is introduced very near its natural range? Going further… what can we expect when species are introduced within the same ecoregion that comprises its native populations? This is what I asked myself – and even more – before planning the study “Alien vs. Predators: Effective induced defenses of an invasive frog in response to native predators” published in Journal of Zoology. If we choose my favorite study system, the introduced populations of the Mediterranean painted frog (Discoglossus pictus), in a practical sense the question is more simple: we have had Discoglossid frogs, Libellulid dragonflies and Notonectid backswimmers all around the Mediterranean basin for millions of years… so, focusing on their predator-prey relationships, will they ever function as unknown species if we are translocating these species inside this ecoregion?  The fact is that although Discoglossus pictus has been introduced from another continent (Africa), its invasive (European) range is included in the same Mediterranean ecoregion. This means that both areas share several – most, in fact – genera and species of aquatic predators. So I thought I could plan some experiments with this array of species to shed some light onto this complex set of questions.


A tadpole of the Mediterranean painted frog (Discoglossus pictus). Photo by Eudald Pujol-Buxo

The first step was to see if in a natural environment – in order to control the presence of predators we had to mimic the “natural environment” with a complex mesocosm set – inducible defenses of the introduced frog in the presence of native predators would be clearly detectable. In other words, we wanted to know if the phenotypic plasticity of the introduced frog could be something similar to a native-native relationship with the predators. However, even when putting a lot of effort and motivation into a project, there can also be unwanted incidents: the project was in near termination when an unwanted and uncontrolled extra set of predators entered the mesocosms… ‘but how?’ you could ask. If you are reaching the article for the first time through this blog post, you will now have the privilege of laughing at what went wrong. How did some predators sneak into some mesocosms, forcing us to discard them? The story is that, for the clearing and maintenance of the experimental area, our university uses a small sheep herd. Somehow the sheep decided – and managed – to find their way into the enclosed area of active experimentation, and they also discovered that by destroying some nets they could access the mesocosms’ water and drink it.  It must have been nice for them during the hot spring, but for us the result was close to a total disaster after three months of hard work – and not counting in previous planning. Luckily enough, the herd was somehow “scientifically grateful” and left untouched enough mesocosms of each treatment to make us believe that it was worth keeping up the experiment using this subset. Finally, we saw that continuing the experiment had been the right choice when the results confirmed our suspicions: the frogs clearly reacted to the native predators. And even more: the results were so clear that the lack of some mesocoms would not be a problem at all. Nice! Now, that motivated me enough to move to the second round of experiments in lab to answer the next question.


A mesocosm for the experiment. Photo by Eudald Pujol-Buxo

Once the results of the previous experiment were clear, the next question we wanted to answer was: are these inducible morphologies and behaviour really effective in reducing predation risk of the introduced frogs? Or it is just some spurious outcome? For this we had to plan some predation trials, and – believe me – it is really a fun work to do. We reared induced and non-induced tadpoles, predators, and prepared some water tanks with plastic plants and natural stones: after some weeks, we were ready for enjoying the work in this new experiment. I have to admit I was eager to see in front of my eyes – and I mean, the action, not the plot of data – how the native predators tried to feast on the introduced tadpoles and how the induced individuals managed to avoid it. Would this be the outcome? Or would the results in this second experiment contradict the first, showing us that what we thought as “inducible defenses” were in fact a non-adaptive change in morphology and behavior? More importantly for our motivation, this time nothing could go wrong: sheep are not allowed in the lab. The results were clear, easy to interpret, and did not contradict the first experiment. And yes, making preliminary tests with Notonecta backswimmers allowed us to enjoy seeing them in action. Backswimmers are indeed spectacular predators when they become hungry: we could not keep them together with the Libellulid darter nymphs because they ate some of them as well. So, morphological results were similar, and induced tadpoles really survived better the predation trials than non-induced. Summing up, we detected marked reactions of antipredator phenotypic plasticity in Discoglossus pictus tadpoles and these reactions seemed clearly effective in reducing their mortality and injury rates against the predators.

More experiments would be needed to confirm this, but we believe that the introduced frog benefits from a previous experience of these – or very similar – predator species. In conclusion, even though native and invasive ranges of Discoglossus pictus are in different continents, the similarity of predator communities of both areas may make this fact unimportant in terms of predator-prey interactions during its larval phase. Going back where we started, the introduced frog is not behaving like an introduced species in this case. We hope that the readers find the study interesting and inspiring for further work on the topic.

Eudald Pujol-Buxó

Author Spotlight: How soil features are shaping the bite force and skull morphology in subterranean rodents

17 01 2017

The role of soil features in shaping the bite force and related skull and mandible morphology in the subterranean rodents of genus Ctenomys (Hystricognathi: Ctenomyidae)

L.R. BorgesR. MaestriB.B. KubiakD. GalianoR. Fornel and T.R.O. Freitas

Tuco-tucos (genus Ctenomys) are subterranean rodents widespread in the southern cone of South America. They are members of the caviomorph lineage (e.g. the guinea pig and their relatives), which arrived in South America ~50 million years ago via transoceanic dispersal from Africa. Today, more than 60 species of Ctenomys are described.


Ctenomys minutus, a species of tuco-tuco. Photo by Daniel Galiano


Another species of tuco-tuco, Ctenomys ibicuiensis. Photo by Daniel Galiano

To live underground, the tuco-tucos must be able to excavate the soil. They use both claws and teeth to break up the soil, cut roots, tubers and other plant material, therefore opening their way to live underground. Adaptations to excavate with the incisors are particularly well demonstrated in this group. As soil features change, it is possible that distinct adaptations arise – the more compact the soil, the harder it is for digging. The question we asked ourselves was: does species living in more compact (harder) soils have a stronger bite force than species living in less compact (soft) soils?

In our paper published in Journal of Zoology we tried to answer this question. We used Freeman and Lemen’s bite force index (a useful formula to estimate bite force of rodents when direct measurements are not available) to estimate bite force values for 24 Ctenomys species, and we used a bulk density variable for soil compaction at each species’ distribution. Morphometric geometric techniques were applied on 1,122 specimens of the same species to investigate skull and mandible features correlated with bite force.


Figure 1 from the article by Borges et al.: Landmarks used for measuring the shape of the skull and the mandible of the tuco-tucos.

We found that the species with strong bite force values do tend to occur in highly compact soils, while species with low bite force values tend to occur in less compacted soils. However, it turns out that species with low bite force values are also found in highly compacted soils. This makes us believe that different species developed different strategies to manage the excavation process. While some species probably rely mostly on their teeth (those species with high bite forces values occupying harder soils), others may rely on other distinct strategies to excavate, using their claws for instance, making these species able to have weaker bite force even though inhabiting highly compacted soils. We also discovered that a wider skull and a robust mandible are associated with the strongest bites, while an elongated skull and mandible are correlated with the weakest bite forces, in what seems to be a recurrent pattern for rodents and other mammal species.

Our next step is to figure out how appendages (limbs, the shoulder blade) contribute to excavation in species of tuco-tucos.

Renan Maestri



HIDDEN GEM: Contributions to a Knowledge of the Hemipterous Fauna of St. Helena, and Speculations on its Origin

12 12 2016

The Zoological Society of London has been publishing scientific papers in zoology since 1830, and our backfiles contain a wealth of ‘hidden gems‘ written by early explorers and zoologists. This article by F. Buchanan White, M.D., F.L.S., was published in 1878 in the Proceedings of the Zoological Society of London, a predecessor of Journal of Zoology. It is a fascinating early article on biogeography that reads almost like a detective story, speculating on the origin of the native fauna and flora of St Helena, a remote island in the middle of the South Atlantic Ocean. As the article describes, St Helena is situated in extreme isolation, nearly 1200 miles from the African continent and 1800 miles from South America, and has no indigenous terrestrial mammals nor any land or freshwater amphibians, reptiles or fish. However, as summarised by Buchanan White, naturalists such as T.V. Wollaston and J.C Meliss had previously reported having found native terrestrial invertebrates on the island, such as land molluscs, various Coleoptera and Hemiptera, as well as spiders and scorpions, many of which were endemic, or ‘peculiar’, to the island. Furthermore, Meliss as well as J.D. Hooker had studied the flora of St Helena and had found 77 species of plants that appeared to be ‘absolutely peculiar’.


P.Z.S. 1878, Plate XXXI: ‘Hemiptera of St Helena’

These findings inevitably induced the question ‘Whence and by what means came this very peculiar fauna and flora?’, presenting a real puzzle for the naturalists of the day, and in this beautifully written article Buchanan White summarizes and discusses the main proposed theories, or ‘speculations’, about the geographical origins of the indigenous fauna and flora of St Helena, and the mechanisms by which they have ended up on this remote island in the middle of the ocean. The author then proposes his own theory, involving the glacial period and possible former islands acting as ‘stepping stones’ for the fauna to spread to the island, referring to ideas presented earlier by other naturalists such as Wallace and Darwin. Buchanan White concludes the paper with species descriptions of Hemiptera from St Helena, collected by Wollaston during his 6 months of exploration of the island. By the time of the writing of this article, Wollaston had sadly passed away, and before his death he had asked Buchanan White to describe all the new species in his collection ‘in a single paper and not piecemeal’, resulting in this remarkable paper from our archives which you can access and read for free.

Elina Rantanen