Capture efficiency of an ant-eating spider, Zodariellum asiaticum (Araneae: Zodariidae), from Kazakhstan.
Abstract: Zodariellum asiaticum (Tyschchenko 1970) is an ant-eating spider from Central Asia. Using five syntopically occurring ant species, namely Cataglyphis aenescens, Formica cunicularia (both Formicinae), Messor aralocaspius, Tetramorium caespitum (both Myrmicinae), and Tapinoma erraticum (Dolichoderinae) in a laboratory study of preycapture behavior, I evaluated capture frequency, attack latency, number of attacks, and paralysis latency. Although spiders captured all five ant species, capture efficiency varied when spiders were tested with the different ant species, being highest when the spiders were tested with F. cunicularia. I concluded that small juvenile Z. asiaticum probably adapt to feed primarily on species of small dolichoderine and myrmicine ants and that large juvenile and the adult Z. asiaticum adapt to feed primarily on large formicine ants.

Keywords: Stenophagy, specialization, myrmecophagy, prey-capture behavior
Article Type: Report
Subject: Predation (Biology) (Research)
Spiders (Behavior)
Author: Pekar, Stano
Pub Date: 09/01/2009
Publication: Name: Journal of Arachnology Publisher: American Arachnological Society Audience: Academic Format: Magazine/Journal Subject: Biological sciences; Zoology and wildlife conservation Copyright: COPYRIGHT 2009 American Arachnological Society ISSN: 0161-8202
Issue: Date: Sept, 2009 Source Volume: 37 Source Issue: 3
Topic: Event Code: 310 Science & research
Geographic: Geographic Scope: Kazakhstan Geographic Code: 9KAZA Kazakhstan
Accession Number: 253537152
Full Text: Spiders are well known for being euryphagous predators, (i.e., consuming a wide variety of prey), but many have an aversion to ants. This makes any group of spiders that routinely eat ants ("myrmecophagy") especially interesting. Myrmecophagic spiders are of further significance because they often seem to adopt ant-specific prey-capture behavior, and they may actively choose ants in preference to other prey (e.g., Huseynov et al. 2008). There is particular interest in spider species that might be exclusively myrmecophagic.

Although myrmecophagy may be a rare phenomenon in spiders as a whole, examples are found in an assortment of spider families, including the Gnaphosidae, Oecobiidae, Salticidae, Theridiidae, Thomisidae, and Zodariidae (Glatz 1967; Heller 1976; Porter & Eastmond 1982; Jocque 1991; Castanho & Oliveira 1997; Jackson et al. 1998). Species that appear to be exclusively myrmecophagic include a theridiid Dipoena and a thomisid Aphantochilus (Umeda et al. 1996; Castanho & Oliveira 1997). However, regardless of whether a species is exclusively or partially myrmecophagic, arachnologists have tended to envision myrmecophagic spiders as preying on ants in general rather than as having become adapted to particular kinds of ants. Yet there is evidence that at least some of the myrmecophagic spiders show a preference for certain genera or even species of ants within the family Formicidae as a whole. Researchers need to gather more information about myrmecophagic spiders so that we can determine how important the targeting of particular ant taxa is for these predators. "Targeting" includes a variety of adaptations by which a spider might specialize on particular kinds of ants, including adaptation related to morphology, physiology, and behavior.

The myrmecophagic spiders I investigated are from the family Zodariidae, one of the most diversified families of spiders (Platnick 2009). The family Zodariidae is known for including a number of myrmecophagic species (Jocquea 1991), but the natural history of the great majority of these species is still poorly known. Available evidence suggests the species in four of the genera in the subfamily Zodariinae, namely Diores, Trygetus, Zodariellum, and Zodarion, are exclusively myrmecophagous (Marikovsky & Tyschchenko 1970; Pekar et al. 2005; Haddad & Dippenaar-Schoeman 2006). However, evidence that the predator distinguishes prey below the level of family (Formicidae) has come from only one of these genera, namely Zodarion (Pekaa r 2005; Pekaar et al. 2008).

Here I consider a species from the genus Zodariellum. The species in this genus are morphologically uniform (Marusik & Koponen 2001), meaning that distinctive interspecific differences are evident only in the details of sexual organ shape, not in structures such as the spider's chelicerae that function directly in predation. This suggests that, for finding evidence of adaptation to specific types of ants, we should investigate behavioral and predatory-related physiological traits.

In this study, I focus on some traits related primarily to behavior. Ant-eating spiders typically capture ants using a 'bite-and-release' tactic (e.g., Jackson & van Olphen 1992; Cushing & Santangelo 2002). One probable advantage of this mode of attack is that it enables the predator to avoid being injured or killed when ants counter-attack. Using this mode of attack, species from the genus Zodarion can subdue a number of different kinds of ants, but apparently the spider's efficiency in capturing different kinds of ants varies considerably. Evidence that efficiency varies comes from data on paralysis latency (i.e., the time elapsing between when the ant is attacked and when it becomes immobile) and the frequency of attacks before the prey is eaten (e.g., Pekaar 2005). These are the parts of the predatory sequence to which I paid particular attention in this study of Zodariellum.

Worldwide, there are 22 species in the genus Zodariellum, about 10 of which appear to be endemic to Central Asia (Platnick 2009). There are published anecdotal prey records for two species, Z. asiaticum Tyschchenko 1970 and Z. sahariense Denis 1959, feeding on ants (Pierre 1959; Marikovsky & Tyschchenko 1970). Marikovsky reported that the ant on which Z. asiaticum preys is primarily Formica cunicularia Latreille, but predation was also observed on Tetramorium caespitum (Linnaeus), Messor aralocaspius Ruzsky, and Cataglyphis aenescens (Nylander) (Marikovsky & Tyschchenko 1970; Marikovsky 1979).

Zodariellum asiaticum, the species I investigated, occurs in southeastern Kazakhstan. The specimens I used (eight female and seven subadult individuals; body lengths 3.5^.5 mm) were collected in April on the sandy slopes of a semi-desert habitat along the Illi River, near the city of Kapchagay (43[degrees]56'93.4N, 77[degrees]03'56.7E). Spiders were identified using Marikovsky & Tyschchenko (1970) and Marusik & Koponen (2001) and kept in glass tubes (diameter 10 mm, length 60 mm) with moistened substrate (plaster of Paris). All spider specimens are deposited in the collection of arachnids of the Department of Botany and Zoology, Masaryk University, Brno.


The ant fauna occurring syntopically with Z. asiaticum was surveyed at the same sites (ants identified using Marikovsky 1979). Five ant species were used in the experiments: Cataglyphis aenescens (body length 4.5-8 mm) and Formica cunicularia (4-6 mm) (both Formicinae), Messor aralocaspius (4.5-8 mm) and Tetramorium caespitum (3-3.5 mm) (both Myrmicinae), and Tapinoma erraticum (Latreille) (3.5-4 mm) (Dolichoderinae). We collected ants used as prey in the field a few hours before we used them in the experiment.

I chose one of the ant species (Messor) to serve as the standard for initial feeding (i.e., after being collected, the spiders were fed with a single individual of Messor the next day). Three days later, each spider was offered successively, in random order, a single ant from each of the five species. I tested each spider with each ant species only once.

There was a 2-day interval between successive trials. A single trial consisted of releasing an ant into a dish occupied by a spider (diameter 40 mm; filter paper glued to the bottom; thin layer of fluon on the sides). Each spider had been in the Petri dish for one day before the trial began. Spiders usually attacked within 30 s. If the spider did not attack the ant within 10 min, I terminated the trial. These aborted trials were classified as rejection of prey. For each trial, I recorded attack latency (i.e., the time between when the spider oriented itself toward the ant and the first attack), number of successive attacks, and paralysis latency (i.e., the time between the first attack and the prey becoming completely immobilized).

Data were analyzed using Linear Mixed-Effects Models (LME) from the NLME-package within the R-environment (R Development Core Team 2007). I chose this method because observations were not independent (i.e., there was repeated use of the spider individuals) and LME is designed for taking into account repeated measurements (Pinheiro & Bates 2000). Both latencies appeared to come from asymmetrical (skewed to the right) distributions. Therefore, I applied logarithmic transformation, after which the data distribution approximated the normal distribution. I expected that the size of ant prey might affect the number of attacks and the paralysis latency. Therefore, I used prey size as a covariate when analyzing the data. Frequency of capture was compared using the Cochran Q test.

Predatory sequences in encounters between Z. asiaticum and ants were similar to predatory sequences in encounters between species of Zodarion and ants (e.g., Pekar 2004). Z. asiaticum approached ants quickly from behind and attacked, usually by contacting the ant's dorsal thorax or leg, releasing the ant, and then continuing to attack several more times or waiting until the ant became paralyzed.

Spiders attacked all five ant species, the most frequently attacked species being F. cunicularia and the least frequently attacked being T. erraticum (Fig. 1A). However, differences in frequency of attacking ant species were not statistically significant (Cochran Q test, [Q.sub.4] = 4.8, P = 0.31). Differences in latency of making the first attack were also non-significant across ant species (LME, [F.sub.4,42] = 1.9, P = 0.13, Fig. 1B) and for number of attacks (LME, [F.sub.4,42] = 0.7, P = 0.6, Fig. 1C). When number of attacks were considered in relation to ant size, differences in the data were not significantly different (i.e., number of attacks is independent of ant size) (LME, [F.sub.1,42] = 0.07, P = 0.79). However, there were significant differences between ant species for paralysis latency (LME, [F.sub.4,42] = 10.8, P < 0.0001, Fig. 1D), latency for Messor being about three times as long as latencies for Cataglyphis, Formica and Tapinoma. Latency for Tetramorium was similar to latency for Messor (Fig. 1D). For formicine ants (Cataglyphis and Formica), there were no significant differences in paralysis latency when considered whether latency depended on ant size (LME, [F.sub.1,23] = 0.14, P = 0.72, Fig. 2A). For myrmicine ants (Messor and Tetramorium), on the other hand, latency for large ants was significantly longer than for small ants (LME, [F.sub.1,24] = 13.4, P = 0.001, Fig. 2B).


I found that Z. asiaticum captured each of the five ant species used in this study, suggesting that this spider may have some general adaptations that enable it to be effective at capturing ants in general. Yet there are some critical differences indicating that this predator distinguishes ants below the level of family and has acquired adaptations by which it can be particularly effective as a predator of certain kinds of ants. Preference is a motivational trait that drives a predator's prey-choice behavior (Huseynov et al. 2008), and it seems likely that my data on capture frequency and attack latency are especially closely linked to the preferences of Z. asiaticum. I might predict that a predator will be more likely to attack and quicker to attack prey it prefers. However, my hypothesis is that paralysis latency and the numbers of attacks made by Z. asiaticum on different ant species are determined to a large extent by biochemical specificity of the spider's venom. If these predictions and hypotheses are correct, then I have no evidence from this study of preference for particular kinds of ants, nor does the number of attacks made before the ant is immobilized provide evidence of venom specificity. However, significant differences among ants were found for paralysis latency. The data for paralysis latency do not simply corroborate the venom-specificity hypothesis, but they add strength to this hypothesis and imply that further investigation would be of interest.

More specifically, paralysis data support a hypothesis that Z. asiaticum has adapted in special ways as a predator that targets formicine ants and F. cunicularia in particular. Stated more precisely, Z. asiaticum specializes on the formicine ants. In this context, "specialized" refers to having special characteristics (in this instance, formicine-specific venom) that make a predator especially effective at preying on a particular kind of prey.

It is of interest that no size-dependent relationship was evident for formicine ants. This is what we might expect if the venom of Z. asiaticum is specific to these ants. In an earlier study (Pekaa r et al. 2008), we demonstrated similar results for Zodarion germanicum (C.L. Koch 1837), as this species performed best, in terms of growth, development and survival, on a diet consisting of only formicine ants. For this species as for Z. asiaticum, there was no significant variation in paralysis efficiency dependent on ant size when the ants were formicines. It is also noteworthy that Marikovsky reported a paralysis latency of about 4 min when Z. asiaticum attacked F. cunicularia in the field (Marikovsky & Tyschchenko 1970), as this corresponds closely to the latencies I found in the laboratory.

Without further experiments, the connection between paralysis latency and venom specificity remains only a hypothesis. Paralysis latency may be influenced by variables that could not be controlled in this study. For example, the volume of venom injected may influence paralysis latency, but venom volume was not determined in this study. That venom volume needs to be considered is illustrated by recent findings from a study on an unrelated spider of the genus Cupiennius. This spider was shown to inject larger volumes of venom when it detected that the prey was especially dangerous (Wigger et al. 2002).

For the genus Zodarion, two distinct groups appear to be identifiable in the context of prey specificity. First, there are species that may normally exploit a single ant species. More specifically, there appear to be species that exploit solely ants from the genus Messor. Being polymorphic, ants of this genus may be feasible prey for exploitation by Zodarion throughout the spider's life cycle, as there would always be available, regardless of the developmental stage to which an individual of Zodarion might belong, a suitable size morph of this ant species (Pekaa r, unpublished). The other group consists of Zodarion species whose adults tend to prey most often on monomorphic ant species. I hypothesize that these Zodarion species, specifically the smaller juveniles, prey primarily on ant species that are smaller than the species on which the adults prey. This implies that the spiders adapt to the ants they target as they progress through their life cycles. Often the stage-specific switch may be from smaller to larger ant species belonging to the same ant subfamily (Pekaa r et al. 2008).

Field data suggest that Z. asiaticum frequently exploits ant species from different subfamilies. For Z. asiaticum, there may be a developmental adaptation in the ants primarily targeted, with small juveniles feeding on small dolichoderine ants such as Tapinoma and on small myrmicine ants such as Tetramorium, and with large juveniles and adults feeding primarily on large formicines such as F. cunicularia. On one occasion, I found a juvenile of Z. asiaticum feeding on Tapinoma in the field, which helps to support this hypothesis. If this hypothesis is corroborated, then it will be of interest to investigate the venom specificity of the early juvenile stages of Z. asiaticum to determine whether small juveniles have venom that is specific not to formicines, but instead to dolichoderines or myrmicines or both.


I would like to thank M. Martisova for help with the spider collection in the field and A. Gromov for help with the logistics in Kazakhstan. Further, I am very grateful to R. Jackson for useful comments and extensive text alterations. The study was supported by project no. MSM0021622416 provided by the Ministry of Education, Youth and Sports of the Czech Republic.

Manuscript received 2 February 2009, revised 7 August 2009.


Castanho, L.M. & P.S. Oliveira. 1997. Biology and behaviour of the neotropical ant-mimicking spider Aphantochilus rogersi (Araneae: Aphantochilidae): nesting, maternal care and ontogeny of ant hunting techniques. Journal of Zoology 242:643-650.

Cushing, P.E. & R.G. Santangelo. 2002. Notes on the natural history and hunting behavior of an ant-eating zodariid spider (Arachnida, Araneae) in Colorado. Journal of Arachnology 30:618-621.

Glatz, L. 1967. Zur Biologie und Morphologie von Oecobius annulipes Lucas (Araneae, Oecobiidae). Zeitschrift fur Morphologie der Tiere 61:185-214.

Haddad, C.R. & A.S. Dippenaar-Schoeman. 2006. Epigeic spiders (Araneae) in pistachio orchards in South Africa. African Plant Protection 12:12-22.

Heller, G. 1976. Zum Beutefangverhalten der ameisenfressenden Spinne Callilepis nocturna (Arachnida: Araneae: Drassodidae). Entomologica Germanica 3:100-103.

Huseynov, E.F., F.R. Cross & R.R. Jackson. 2005. Natural diet and prey-choice behaviour of Aelurillus muganicus (Araneae: Sal ticidae), a myrmecophagic jumping spider from Azerbaijan. Journal of Zoology 267:159-165.

Huseynov, E.F., R.R. Jackson & F.R. Cross. 2008. The meaning of predatory specialization as illustrated by Aelurillus m-nigrum,an ant-eating jumping spider (Araneae: Salticidae) from Azerbaijan. Behavioral Processes 77:389-399.

Jackson, R.R., D. Li, A. Barrion & G.B. Edwards. 1998. Prey-capture techniques and prey preferences of nine species of ant-eating jumping spiders (Araneae: Salticidae) from the Philippines. New Zealand Journal of Zoology 25:249-272.

Jackson, R.R. & A. van Olphen. 1992. Prey-capture techniques and prey preferences of Chrysilla, Natta and Siler, ant-eating jumping spiders (Araneae, Salticidae) from Kenya and Sri Lanka. Journal of Zoology 227:163-170.

Jocquea, R. 1991. A generic revision of the spider family Zodariidae (Araneae). Bulletin of the American Museum of Natural History 201:1-160.

Marikovsky, P.I. & V.P. Tyschchenko. 1970. Pauk-myrmekofil (Zodarion asiaticum Tysch. sp. nova) i nekotorye tcherty jevo biologii [Myrmecophilous spider (Zodarion asiaticum Tysch. sp. nova) and some aspects of its biology]. Trudy Alma-Atinskogo Gosudarstvenogo Zapovednika 9:196-201.

Marikovsky, P.I. 1979. Muravy pustynh Semiretchia [Ants of Semiretchie Deserts]. Nauka, Akademia nauk Kazachskoj SSR, Alma-Ata.

Marusik, Y.M. & S. Koponen. 2001. Spiders of the family Zodariidae from Mongolia (Arachnida: Araneae). Reichenbachia 34(4):39-48.

Norgaard, E. 1956. Environment and behaviour of Theridion saxatile. Oikos 7:159-192.

Pekaa r, S. 2004. Predatory behavior of two European ant-eating spiders (Araneae, Zodariidae). Journal of Arachnology 32:31-41.

Pekaa r, S. 2005. Predatory characteristics of ant-eating Zodarion spiders (Araneae: Zodariidae): potential biological control agents. Biological Control 34:196-203.

Pekaar, S., J. Kraal & Y.D. Lubin. 2005. Natural history and karyotype of some ant-eating zodariid spiders (Araneae, Zodariidae) from Israel. Journal of Arachnology 33:50-62.

Pekaar, S., S. Toft, M. Hrusskovaa & D. Mayntz. 2008. Dietary and prey-capture adaptations by which Zodarion germanicum, an anteating spider (Araneae: Zodariidae), specialises on the Formicinae. Naturwissenschaften 95:233-239.

Pierre, F. 1959. Le mimetisme chez les araignees myrmecomorphes. Annee Biologie 35:191-201.

Pinheiro, J.C. & D.M. Bates. 2000. Mixed-Effects Models in S and S PLUS. Springer, New York.

Platnick, N.I. 2009. The World Spider Catalog, Version 9.5. American Museum of Natural History, New York. Online at http://research. 1.html.

Porter, S.D. & D.A. Eastmond. 1982. Euryopis coki (Theridiidae), a spider that preys on Pogonomyrmex ants. Journal of Arachnology 10:275-277.

R Development Core Team. 2007. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna. Online at

Umeda, Y., A. Shinkai & T. Miyashita. 1996. Prey composition of three Dipoena species (Araneae: Theridiidae) specializing on ants. Acta Arachnologica 45:95-99.

Wigger, E., L. Kuhn-Nentwig & W. Nentwig. 2002. The venom optimisation hypothesis: a spider injects large venom quantities only into difficult prey types. Toxicon 40:749-752.

Stano Pekar: Department of Botany and Zoology, Faculty of Sciences, Masaryk University, Kotlafska 2, 611 37 Brno, Czech Republic. E-mail:
Gale Copyright: Copyright 2009 Gale, Cengage Learning. All rights reserved.