Some wasp species use spiders as food resources, overcoming several anti-predator barriers that are exerted by spiders. Tromatobia ichneumonid wasps are spider egg predators that usually attack Araneidae species, although there are few records of predation on Clubionidae, Philodromidae, Linyphiidae, Tetragnathidae, and Theridiidae spiders. Here, we describe the interaction between Tromatobia sp. and Chrysso compressa, a subsocial theridiid spider that exhibits extended maternal care, in the Atlantic Forest of southeastern Brazil. We observed that the larva of Tromatobia sp. develop inside the egg sacs of C. compressa, preying on the entire egg mass and building cocoons that change the color and morphology of the egg sacs. Despite these structural modifications, we registered an adult female of C. compressa guarding and caring for the cocoons (attacked egg sac) of the predators as if they were offspring (non-attacked egg sac). To the best of our knowledge, this study represents the first record of Tromatobia preying on Chrysso eggs.

egg sac, maternal care, Pimplinae, Serra do Japi

Wasps are important natural enemies of spiders and adopt several foraging strategies to subdue their prey (Rayor 1996). For example, some wasp species act exclusively as koinobiont ectoparasitoids (polysphinctine wasps sensu Gauld & Dubois, 2006) or predators (spider-hunting wasps – Crabronidae, Pompilidae and Sphecidae; Mayr et al. 2020) of juvenile, sub-adult, or adult spiders. Some other wasp species are specialized in attacking and parasitizing individual eggs (e.g., Baeus Austin, 1985) or consuming part of the egg mass of spiders (e.g., Tromatobia Gauld et al., 2002). Consequently, spiders have evolved anti-predator mechanisms that prevent wasp attacks.

Spiders can minimize predation risk by avoiding detection, recognition, and access to predators by using several defensive cues and behaviors (Pekár 2013; Gawryszewski 2017). The anti-predator strategies of spiders include crypticity, mimicry, construction of three-dimensional webs (Blackledge et al. 2003) and refuges, presence of thick silk barriers in egg sacs, extended parental care, and, in some extreme cases, group living and sociality (Pekár 2013; Gawryszewski 2017). Understanding how wasps overcome these barriers and succeed in capturing spiders is an interdisciplinary research topic that combines studies on ecology, evolutionary behavior, physiology, and phylogeny.

Darwin wasps of the genus Tromatobia Foster, 1869 are specialized in attacking aerial-web building spiders (Fitton et al. 1987, 1988; Finch 2005; Yu et al. 2016; Broad et al. 2018), completing their larval life cycle inside spider egg sacs, and consuming part or the entire egg mass (Austin 1984, 1985; Villanueva-Bonilla et al. 2016). There is high interspecific variation in the number of eggs, from one to 14, that Tromatobia females can lay inside spider egg sacs (Nielsen 1923). This variation also occurs within the same species, as Tromatobia blancoi Gauld, 1991 can produce six-nine eggs per egg sac of the spider Araneus thaddeus (Hentz, 1847) (Jiménez 1987). Although rare, most reports of Tromatobia-spider interactions involve Araneidae species of the genera Araneus, Araniella, Argiope, Cyclosa, and Zygiella (e.g., Nielsen 1923; Jiménez 1987; Fitton et al. 1988; Quicke 1988; Oehlke and Sacher, 1991; Cortés et al. 2000; Sobczak 2012), indicating fine specialization for this family. However, there are a few records of Tromatobia attacking Clubionidae, Philodromidae, Linyphiidae, Tetragnathidae, and Theridiidae species (e.g., Austin 1985; Fitton et al. 1988; Oehlke and Sacher 1991; He et al. 1992). In fact, knowledge about the biology, ecology, and behavior of interactions with these minor hosts is scarce and requires further field observations and experimental studies.

Herein, we report the interaction between the egg predator wasp Tromatobia sp. (Ichneumonidae) and the cobweb spider Chrysso compressa (Keyserling, 1884) (Theridiidae) in the Brazilian Atlantic Forest, with notes on the behavior of the C. compressa guarding the predator’s cocoon.

We collected 22 egg sacs of C. compressa (N = 6 in January, N = 10 in February, N = 4 in March, N = 1 in April, and N = 1 in May), of which five (22.7%) were attacked by Tromatobia sp. Healthy egg sacs harboured an average of 46 ± 21 eggs (N = 17), were light in color, round-shaped (diameter = 10.1 ± 3.7 mm, N = 17), and were usually located under the bodyguard of the adult female (Fig. 2A). The morphology of egg sacs that were attacked by Tromatobia sp. was found to be altered after the development of wasp larva; they were elongated and oval (average total length = 9.1 ± 1.2 mm and average total width = 2.5 ± 0.9 mm, N = 3), with the shape of cocoons adhered to each other longitudinally, on their longer sides. At this stage, the wasp cocoons, grey or brownish in color (Fig. 2B) became exposed and interspersed with a few silk remnants of the original egg sac (Fig. 2C). The attacked egg sacs that were collected in the present study contained two (N = 3) or three (N = 1) cocoons of Tromatobia sp. Adult wasps emerged through holes with approximately 2 mm in diameter located in the apical portion of the cocoon (Fig. 2D). In total, we obtained nine adult individuals of Tromatobia sp. (Fig. 2E), five females (average body length = 7.3 ± 0.9 mm) and four males (body length = 6.6 ± 0.8 mm).

Figure 2. 

A adult female of C. compressa taking care of the egg sac B adult female of C. compressa taking care of cocoons of the egg predator wasp Tromatobia sp. C cocoon of the egg predator wasp Tromatobia sp. D cocoon hole through which adult wasps of Tromatobia sp emerge. E lateral view of an adult female of Tromatobia sp., and F lateral view of an adult wasp that did not emerge from the cocoon. Photos: Brenda Santiago and Diego Pádua.

We observed that the maternal care provided by C. compressa included the protection of egg sacs under the female body, between the forelegs and held by the pedipalps. We also registered the same behavior of females protecting the cocoons of Tromatobia, even in the case when predators had already fed on the entire egg mass (Fig. 2B). In addition, we observed that the spider we offered an alien egg sac adopted and attached it to the native egg sac. The spider bodyguarded both egg sacs containing wasp cocoons until an adult wasp emerged from the native egg sac. At this time, we removed the female spider and left the alien egg sac alone and undisturbed for 68 days; however, adult wasps did not emerge from this egg sac. Then we opened it carefully and found a cocoon with a dead male adult wasp almost fully-developed inside (Fig. 2F).

To the best of our knowledge, this is the first report on an interaction between Tromatobia and Chrysso. Previous studies have indicated a strong affinity of Tromatobia for araneid orb-web spiders, with a few unusual records of attacks on other families including Theridiidae that construct three-dimensional webs. However, some records are rather poorly documented and potentially unreliable. Thus, our report is probably the best-documented record of Tromatobia parasitising non-araneid spiders. We showed that Tromatobia wasps could overcome anti-predator barriers and affect more than 20% of C. compressa egg sacs. Unlike the case of egg parasitoid species, in which each wasp attacks an individual spider egg, Tromatobia sp. consumes the entire egg mass and affects the spiderling population to a greater extent. We recorded two to three cocoons per egg sac, similar to the observation made by Sobczak et al. (2012) for the sympatric orb-web spider Araneus omnicolor Keyserling 1893. However, other studies have reported instances where up to 14 individuals were observed per egg sac in Tromatobia species (e.g., Nielsen 1923; Jiménez 1987). Hence, is still unknown whether there is an optimal number of eggs and whether and how larvae compete for food resources.

Behavioral manipulations of hosts induced by parasitoid species have been well-studied in the last few years (Weinersmith 2019). In cases involving spiders and wasps, parasitoid species can alter host behavior by inducing the construction of modified webs that are used to protect the cocoon, thereby maximizing survival (Eberhard 2000, 2010; Gonzaga et al. 2017). However, Tromatobia sp. does not change the behavior of C. compressa but takes advantage of bodyguarding, a pre-existing behavior of maternal care. Although there is an evident morphological change in the egg sac, the wasp may use concealment mechanisms, such as chemical or tactile camouflage (Kaminski et al. 2020), to deceive adult spiders. We observed that the wasp of the alien egg sac that was adopted by C. compressa in the laboratory did not emerge from the cocoon after we removed the female spider. This anecdotal report suggests that the presence of adult females may be important for wasps. Hence, the effect of bodyguarding behavior on wasp survival should be further investigated in future studies.

Egg protection is crucial for the survival of spider progenies given the high diversity of selective pressures exerted by multiple predators and parasitoids (Bristowe 1971; Li et al. 1999; Yip and Rayor 2014). The behavior of many spider species that keep their egg sacs suspended, for example, is a strategy to avoid attack by wandering generalist predators, such as ants (Turnbull, 1973), but using this strategy can increase the conspicuousness of egg sacs to visually oriented predators (e.g., birds and wasps, Hieber 1992). Instead, egg sacs of C. compressa are usually under female guard within the refuge, and attacks may occur during foraging when females move to the web to catch prey. However, it is still unknown how Tromatobia sp. adults overcome the barriers of C. compressa in detecting, reaching, and successfully attacking egg sacs. In this way, further investigation should focus on identifying specific behaviors performed by the wasps and cues provided by spiders that facilitate egg predation.

This study was financially supported by the Instituto Nacional de Ciência e Tecnologia dos Hymenoptera Parasitoides da Região Sudeste Brasileira (HYMPAR/Sudeste – CNPq/FAPESP/CAPES) and CAPES, Finance Code 001 (Grant Number 88887.513500/2020-00 to BKSS and grant n° 88887.372005/2019-00 to DGP). We thank the Serra do Japi Foundation and the Postgraduate Program in Animal Biology of the University of Campinas for their structural and logistic support for this research, and the Invertebrate Collection of Instituto Nacional de Pesquisas da Amazônia (INPA) for the use of imaging equipment. We thank Editage (www.editage.com) for the English language editing. We thank the reviewers Dr Keizo Takasuka, Dr Tamara Spasojevic, and the editor Dr Gavin Broad for the corrections and suggestions in the text. We also thank Dr Broad for sending the photographs of the holotype of Tromatobia lineiger.