Imprimir Republish


Eight-legged puppets

Wasp larvae manipulate the behavior of spiders

JOBER SOBCZAK / UFSCARDuel: Hymenoepimecis bicolor wasp attacks Nephila clavipes spiderJOBER SOBCZAK / UFSCAR

When they were searching for spiderwebs with two to three hundred spiders of a species that live in spider colonies, biologists Marcelo Gonzaga and Jober Sobczak found something even more amazing: a wasp parked on top of a spiderweb, with no spider in sight. An ordinary person walking by probably would have paid no attention to this situation, but Gonzaga quickly set up his camera. His instinct was right: very soon, a fly was caught in the web, the spider came out of the leaf that was sheltering it and, before it had time to enjoy the day’s meal, the spider was attacked. The wasp grabbed the spider and inserted an ovipositor into the mouth of the spiderweb’s owner, releasing a paralyzing substance. The effect of the substance was long enough to allow the wasp to lay an egg on the upper part of the victim?s abdomen. The researcher kept his finger on the camera’s shutter and documented everything. “I was very lucky – I never thought I would find anything like this,” says Gonzaga, a professor at the Federal University of Uberlândia (UFU), in the State of Minas Gerais and member of the National Science and Technology Institute of Hymenoptera of the Southeast Region of Brazil.

The lucky encounter happened in 2001, in the Japi mountain range, a Mata Atlantica, rainforest, reserve that surrounds the town of Jundiaí, State of São Paulo. This led to the first description of the parasitoid use of a spider (Araneus omnicolor) by a wasp (Hymenoepimecis veranae) in the country. A similar case had been reported seven years before by William Eberhard, of the University of Costa Rica. Ever since then, Gonzaga has maintained contact with his colleague from Central America and described the parasitoid behavior – in which the host is always killed by the parasite – of six other wasp species, two of which had been unknown to science until then.

The procedure described by Gonzaga and Sobczak – a doctorate student whose advisor is Angélica Penteado Dias, of the Federal University of São Carlos (UFSCar), and whose co-advisor is Gonzaga – is a complex one, involving the host’s manipulation by the wasp. After paralyzing the spider (the poison affects the central nervous system through the subesophagic ganglion), the wasp inspects the victim’s abdomen and, if necessary, kills and removes the competing larva. Then, the wasp deposits an egg which releases a larva that adheres to the spider’s body, makes holes in the spider’s abdomen and feeds on the spider’s hemolymph, a blood-like fluid.

060-063_Vespa_180Two weeks later, when the larva spins the cocoon inside which it will become an adult wasp, the spider suddenly changes the structure of the web. It no longer produces a spiral of viscous threads that captures its prey and organizes a simpler and more resistant structure – in some cases, the spider creates a silk barrier to protect the cocoon. The larva then attaches itself to the web by means of structures in the form of hooks that spring out of its carapace the last time the larva shed its skin. Then, the larva kills the spider and spins its cocoon.

In 40% of the 85 spiderwebs examined by the researcher, the results of which were described in a paper published in 2007 in Naturwissenschaften journal, the two biologists found two mature spider females hosting the parasites. They also found a cocoon attached to the web. This is a high proportion, because wasps do not attack just any spider. Gonzaga and Sobczak did not find any male spiders hosting the parasites, but it is too soon to say that they are not victims because they are smaller or were not found by the researchers at the time of the year – March and April – during which the research study was conducted. Gonzaga explains that the big female spiders can fight and change the situation by eating the wasp. “We have never found big female spiders with parasites,” he says. The small spiders are not a very good target because they are not big enough to support the development of a wasp larva. In addition, the immature inhabitants of the spiderweb still have to go through changes (of the skeleton, which is external) until they reach adulthood, and, in this case, the larva would be attached to a dry, empty shell with nothing nutritious to feed on. Wasps have to choose their victims very well in order to survive.

Finding more parasite wasps and describing the interaction between them and their host was the first step of the research study. So far, the species identified in Brazil have been found in the Mata Atlantica, rain forest, in the States of São Paulo and Espírito Santo. However, this does not mean that the parasitoid wasps that prey on spiders are exclusive to this ecosystem; The geographic restriction is established by the region where Gonzaga conducted his research studies – he worked in the Mata Atlantica, rainforest, while attending a post-doctorate course at the State University of Campinas (Unicamp), and during the time he was working as a researcher at UFSCar with a grant from the FAPESP’s Young Researchers program.

Gonzaga has been in Uberlândia since 2008. He is getting ready to explore the cerrado  (Brazilian savanna) region at Estação Ecológica do Panga, a research center of the UFU. The area lies 30 kilometers away from the university. “I have already seen that the interaction between the Nephila and the bicolor Hymenoepimecis, which I had observed in the Japi mountain range, also exists in the cerrado  region,” the biologist reports. He also discovered a species of parasitoid wasp that feeds on spiders that live in colonies. He still has to describe his discovery. The description of this parasite-like process includes a comparison of the findings of Gonzaga and Sobczak in Brazil with the findings of Eberhard in Costa Rica. “So far, the behavioral patterns have been repeated and the changes in the spiderwebs are similar.”

Another phase of the study, which is being conducted concurrently, entails trying to understand what triggers the signal for the spider to spin the web again. Brazilian biologists have already discarded the hypothesis that the energy sucked up by the larvae changes the spider’s behavior, as they had described in an experiment published in 2010 in the Ethology, Ecology and Evolution journal. They went to the field to place bamboo and screen cages around the branches and bushes that supported 15 webs, in order to prevent insects from being entangled in the webs. Then they controlled the spiders´ food intake for 21 days. A group of parasite-free spiders was fed one fly a day, while another parasite-free spider group was not fed. A third group, of parasite-hosting spiders was fed one fly a day. The treatment weakened the spiders, but they did not spin webs of the kind that support the cocoons.

Spinning the threads – modified web under the influence of the H. bicolor larva, on the right

MARCELO GONZAGA / UFUSpinning the threads – modified web under the influence of the H. bicolor larva, on the rightMARCELO GONZAGA / UFU

In collaboration with a group from Unicamp, Sobczak is attempting to discover possible chemical alterations in the larvae at their final stage of development. “We don’t know if the larva injects an unknown compound into the host or induces the spider to produce a similar substance to the one that it releases prior to shedding, as the modified spiderweb resembles the web that spiders spin at such times,” Gonzaga explains.

Sobczak has spent a lot of time in the Japi mountain range, where he has been doing fieldwork to evaluate the importance of the modification of the webs for the survival of the larvae. According to Gonzaga, an ordinary spiderweb unravels in a few days, unless it is constantly maintained by the spider. A modified web, on the other hand, lasts for over a month, which is more than enough time for the larva, which spends approximately one week in the cocoon, to grow into an adult wasp. The spiderweb that shelters the cocoon is more robust because it is made of reinforced threads that attach it to branches and leaves and because it does not have the adhesive threads that trap insects. “Prey that flounders in the web could knock down the cocoon,” believes the biologist from UFU.

In the attempt to prove that the modified spiderweb interferes in the survival of the wasps, Sobczak has been placing cocoons in ordinary webs. As soon as the spider modifies the web and the larva is ready to kill it and cocoon itself, he transfers the spider and the larva into an empty spiderweb. The larva then spins its cocoon in a spiderweb that will continue to trap insects and that will not be maintained by the spider. The initial results are pointing in the expected direction: the cocoons are not successful in ordinary spiderwebs.

Gonzaga has a lot of work ahead. The manipulation of hosts by parasites has many fascinating examples, as is the case of the moth caterpillar, which acts as the nanny of the wasp cocoons that feed on it during the larva stage. This example was studied by a Dutch group from the University of Amsterdam, in collaboration with researchers from the Federal University of Viçosa (UFV), State of Minas Gerais, led by Eraldo Lima and Angelo Pallini.

Thyrinteina leucocerae caterpillars and Glyptapanteles wasps were collected on the campus of the UFV and maintained in the laboratory. In this case, approximately 80 eggs are deposited inside the host’s body. The larvae leave from holes they bore in the caterpillar, while the caterpillar remains next to the cocoon until the wasp is fully developed, according to an article published in 2008 in the PLoS One journal. When predators attack the cocoon, the guardian makes vigorous movements with its head and drives away half of the attackers. However, it dies a short while after the adult wasp emerges from the cocoons. It is possible that the larvae that remain alive inside the caterpillar pull the puppet strings, but nobody knows how they do that.

Even more spectacular manipulations by parasitoid wasps have already been described in other countries, but it is hard to imagine that they were seen in very special places. In view of the astonishing biological diversity found in Brazilian ecosystems, it is a question of time – and a lot of hard work – until more examples are seen that could resemble the most hair-raising terror films.

See more pictures.

The project
Maternal care in spiders of the Theridiidae family (nº 2006/59810-8); Type Young Researchers; Coordinator Marcelo de Oliveira Gonzaga – UFSCar; Investment R$ 52,450.78 (FAPESP)

Scientific article
GONZAGA, M. O. et al. Modification of Nephila clavipes (Araneae Nephilidae) webs induced by the parasitoids Hymenoepimecis bicolor and H. robertsae (Hymenoptera Ichneumonidae). Ethology Ecology & Evolution. v. 22, p. 151-65. 2010.