Printer Friendly

The effect of encounter experience on hunting behavior in the spider wasp, Pepsis cerberus Lucas (Hymenoptera: Pompilidae).

Abstract. -- Laboratory experiments were conducted to assess the effects of encounter experience with a host on the hunting behavior of naive females of the spider wasp, Pepsis cerberus (Hymenoptera: Pompilidae). The time required to complete the overall hunting sequence decreased significantly from a mean of 189.4 min for the first encounter with a host spider, Aphonopelma steindachneri, to 153.1 min after the eighth encounter. The performance of certain behavioral components of hunting improved with experience (initial approach to the host, and antennation / paralyzation) whereas others did not (grooming of antennae, burial and oviposition, and closure of the nest entrance). Results show that certain components of otherwise inflexible instinctive behaviors can be modified by experience.


The hunting behavior of tarantula hawk wasps of the genus Pepsis (Hymenoptera: Pompilidae) has been a topic of interest in both the popular and scientific literature since the early descriptions of Petrunkevitch (1926) and Passmore (1936). Pepsis wasps are one the most conspicuous representatives of the arthropod fauna of North American deserts (Hurd 1952; Punzo 1994a). Pompilid wasps are characterized by the following behavioral traits (Evans & Eberhard 1970): (1) utilization of spiders as a food source (host) for larvae; (2) each wasp larva is provided with a single paralyzed host; (3) paralyzed hosts are dragged backwards by the female wasp over the ground surface and placed in the nest (burrow) where a single egg is deposited on the abdomen of the spider; (4) the abdomen of the wasp is used to compress the soil in the closure of the nest entrance. Although adult male and female Pepsis wasps are nectivorous, (Field 1992), the females selectively hunt large theraphosid (tarantula) spiders as hosts for their developing larvae (Williams 1956; Punzo 1990).

In the earlier literature the hunting behavior of Pepsis wasps has often been cited as a classical example of a stereotypical instinctive (innate) behavior comprised of an inflexible sequence of behavioral acts (Lucas 1919; Petrunkevitch 1926; Cazier & Mortenson 1964) referred to as fixed action patterns (FAPs) (Tinbergen 1951). In contrast, more recent analyses have shown that experience can modify certain components of innate behavioral acts in response to local environmental demands (Punzo 1996; 2000), giving rise to the concept of modal action patterns (MAPs, Barlow 1968).

Some degree of plasticity in hunting behavior has been observed for some species of Pepsis wasps. For example, females of P. formosa may or may not rub the lateral area of their abdomens with the mesothoracic legs during encounters with host spiders (Punzo & Garman 1989). In addition, following paralyzation of the host, females frequently (but not always) exhibit lapping behavior whereby they drink hemolymph oozing from the site where the stinger was inserted (wound site) or liquid material from the mouth of the host. Thus, there is flexibility in the sense that lapping behavior may or may not be exhibited, and if it does occur, the site where it is initiated may vary. There is some evidence that the occurrence of lapping behavior is influenced by the amount of time the female wasp spent in flight searching for a suitable host (Punzo 2000).

The hunting behavior for a few species of Pepsis wasps has been described in detail, including Pepsis marginata (Petrunkevitch 1926; 1952), Pepsis thisbe (Williams 1956; Punzo 1994b), and P. formosa (Punzo & Garman 1989). To summarize, female wasps fly over variable distances and appear to visually locate spiders wandering over the ground surface. In other cases, females will alight on the ground using rapid walking movements to locate occupied spider burrows. If a spider has been encountered on the surface, the wasp taps the spider's body with her antennae), a behavior known as antennation, presumably utilizing olfactory cues to assess the suitability of the host. If suitable (correct species), the wasp then moves away from the host and passes the antennae through her mandibles, a behavioral component known as grooming. After a variable period of time, the wasp returns to the spider, resumes antennation, and then rushes under the host and uses her mandibles to seize one of the spider's posterior legs. Once this occurs, the spider is usually flipped on its back and the wasp searches for a suitable site to insert the stinger. Following insertion of the stinger and evenomation, the spider exhibits rapid paralysis and a curling of its legs over its abdomen. The wasp then moves away from the paralyzed host, resumes grooming, and may or may not exhibit lapping behavior.

If the spider is within its burrow, the wasp enters and forces the spider onto the ground surface before continuing the sequence of behavioral acts described above. Because there are no reports of a Pepsis female attacking and stinging a tarantula within its burrow (Punzo 2000), it has been suggested that the confines of the burrow may not provide adequate space for the maneuverability required to grasp the spider and insert the stinger (Baerg 1958; Petrunkevitch 1952; Punzo & Garman 1989).

Following paralyzation, the wasp will excavate her own burrow (nest) if a spider burrow is not available; otherwise, she will utilize the spider's burrow as her nest. The spider is dragged into the burrow and a single egg is deposited on its abdomen. The wasp then uses her legs to cover the burrow entrance and her abdomen to compress the soil. Thus, the hunting sequence of Pepsis wasps can be divided into the following behavioral components (acts): (1) initial approach and antennation (APA), moving away and grooming (MG), antennation and paralyzation (ANP), burial and oviposition (BO), and closure of the nest (CL) (Evans & Eberhard 1970; Punzo 1991).

This study attempts to determine what effects previous encounter experience with the host may have on the hunting behavior of Pepsis cerberus, with particular emphasis on the amount of time required to complete the overall hunting sequence and the individual behavioral components of hunting.


Host spiders, as well as eggs and developing larvae of P. cerberus, were collected from burrows of parasitized spiders (Aphonopelma steindachneri) located in Big Bend National Park (Brewster County, Texas, USA), during spring and summer of 2003, and transported to the laboratory. This park lies within the northern region of the Chihuahuan Desert. To ensure genetic diversity among wasps used in these experiments, parasitized spiders were collected from burrows separated by a minimum distance of 7.5 km, a distance in excess of the home ranges reported for these wasps (Punzo 2000).

Following pupation, emergent adult male and female wasps were housed separately in 1-L glass cylindrical containers and maintained at 22 [+ or -] 0.5[degrees]C, 65-70% relative humidity. They were fed on a diet consisting of honey and a 2% glucose solution. Thus, all female wasps used in these experiments (15 to 18 days of age; mean body weight: 3.06 g [+ or -] 0.09 SE) were naive (had no previous encounter experience with the host spider). In addition, all spiders were adult females (mean body weight: 8.25 g [+ or -] 0.14 SE) that had been obtained from egg sacs deposited by spiders in captivity, and had no previous experience with wasps.

Ten encounters with a host spider were staged in a test chamber for each of 10 female wasps. A different female tarantula was used for each encounter, and a 24-hr intertrial interval was used between individual encounters. The test chamber was constructed of clear plexiglass, and was 50 cm (length) by 40 cm (width) by 40 cm (height). The floor was covered to a depth of 15 cm with a substrate consisting of a mixture of sand and adobe soil originally obtained from spider collection sites in the field. This depth was sufficient to allow the wasp to excavate a burrow (nest). The top of the chamber contained a hole (3 cm in diam) located at its center that was used to introduce for encounter trials. All observations were made through a one-way mirror to minimize disturbance to the animals, and were recorded using a Panasonic 815D video recorder.

Before the initiation of encounter trials, all of the wasps used in these experiments were allowed to explore the test chamber for 2 hr a day for five days in order to allow wasps to become familiar with the confines of the chamber. Preliminary observations showed that this significantly reduced the amount of time and energy used by wasps in trying to escape. Before each trial, a female spider was removed from its cage and placed in an inverted plastic container for transport to the test chamber. To initiate a trial, the spider was placed at the center of the chamber floor, and the plastic container removed. Following a 5-min period, a female wasp was introduced into the chamber using the hole located at its top. The time required for the completion of the following behavioral components of the hunting sequence was recorded: APP (initial; approach and antennation); MG (moving away and grooming); ANP (antennation and paralyzation); BO (burial and oviposition); and CL (closure of the nest entrance). Following data collection, wasp and paralyzed spiders were preserved in 80% ethanol and placed in the Invertebrate Collection at the University of Tampa.

All statistics used in data analysis followed procedures as outlined by Sokal & Rohlf (1955). To analyze the effects of encounter experience on the time required to complete the behavioral components of hunting an analysis of variance (ANOVA) was used since a Bartlett's Test showed homogeneity of variances and G-Tests indicated that error variances were normally distributed. Scheffe F tests were used to determine significance between various means for individual behavioral components.


Results show that the amount of time required by female wasps to complete the overall hunting sequence decreased as a function of the number of encounters with a host spider (Fig. 1), from a mean of 189.4 min during the first encounter, to 153.1 min for the tenth encounter (F = 27.21, df = 9,45, P < 0.01). This reduction in time was significant after eight encounters. This improvement in performance can be considered a change in behavior (behavioral plasticity) that resulted from experience, thereby satisfying a common definition for learning (Bitterman 1975). A similar effect of experience on enhanced efficacy of hunting in pepsine wasps has been reported for P. formosa (cf. Punzo & Garman 1989; Punzo 1991) and P. thisbe (Punzo 1994b). In terms of energy budgets (cost-benefit analysis), additional time spent in acquiring needed resources is rendered unavailable for other activities such as mating and reproduction. In addition, the more time an animal spends engaged in foraging activities, the more it will be exposed to potential predators. Thus, the ability to reduce the amount of time allocated toward search, pursuit, capture, and handling of prey should increase the overall fitness of predators (Krebs 1978; Punzo 2002). In this respect, selection should favor the evolution of adaptive behavioral programs that allow for a certain degree of plasticity (Punzo 1996).


With respect to specific behavioral acts, there was a significant reduction in the amount of time required to complete APP and ANP components (F = 32.67, P < 0.01). In contrast, female wasps showed no improvement in performance for the MG, BO and CL components (Table 1). Thus, APP and ANP exhibit a degree of plasticity as a result of experience, whereas MG, BO and CL are not similarly affected. It is plausible that certain biomechanical and temporal constraints associated with digging, oviposition and nest closure activities negate any possibility for their improvement as the result of experience. It is also interesting to note that the specific acts that did show improvement (APP and ANP) are precisely those components that place the wasp in the greatest danger. Theraphosid spiders are formidable opponents and are capable of seizing and killing animals as large or larger than themselves including other spiders, scorpions, and solifugids, as well as small reptiles and mammals (Baerg 1958; Punzo & Henderson 1999). Thus, any capacity to reduce the time required to immobilize a spider should increase survivorship in these wasps.

In conclusion, the results of these experiments indicate that the hunting behavior of females of P. cerberus improves as a function of increasing encounters with a host. The significant decrease in the time allocated to the initial approach toward the host, as well as paralyzation, should result in a decrease in the amount of energy required to complete the overall hunting sequence. This should result in an increase in the energy available for other activities such as oogenesis, territorial defense, predator avoidance, and reproduction. This represents an optimization of foraging behavior and would increase the number of hosts that an individual wasp could locate within a given period of time.


I thank J. Bottrell and L. Ludwig for assistance in collecting wasps and spiders in the field, R. J. Edwards and anonymous reviewers for critical comments on an earlier draft of the manuscript, and K. Dove for consultation on statistical analyses. I also thank R. Skiles, Park Biologist, Big Bend National Park (BBNP), Texas, for logistical support, and a research permit issued by the National Park Service and BBNP (#BIBE-27). A Faculty Development Grant to FP from the University of Tampa made much of this work possible.


Baerg, W. J. 1958. The Tarantula. Univ. Kansas Press, Lawrence, Kansas, 88 pp.

Barlow, G. W. 1968. Ethological units of behavior. Pages 37-55, in D. Ingle (ed.), The central nervous System and fish behavior. Univ. Chicago Press, Chicago, 488 pp.

Bitterman, M. E. 1975. The comparative analysis of learning. Science, 188:699-709.

Cazier, M. A. & M. Mortenson. 1964. Bionomical observations on tarantula hawks and their prey (Hymenoptera: Pompilidae). Ann. Ent. Soc. Amer., 57:533-541.

Evans, H. E. & M. Eberhard. 1970. The Wasps. Univ. Michigan Press, Ann Arbor, 134 pp.

Field, J. 1992. Guild structure in solitary hunting spider wasps (Hymenoptera: Pompilidae) compared with null predictions. Ecol. Entomol., 17:198-208.

Hurd, P. D. 1952. Revision of the Nearctic species of the genus Pepsis (Hymenoptera: Pompilidae). Bull. Am. Mus. Nat. Hist., 98:261-334.

Krebs, J. R. 1978. Optimal foraging decision rules for predators. Pp. 23-63, in J. R.

Krebs & N. B. Davies (eds.), Behavioural Ecology: An Evolutionary Approach. Blackwell Publs., Oxford, 494 pp.

Lucas, R. 1919. Pompiliden-Studien. Arch. Naturgesch., 83A:1-180.

Passmore, L. 1936. Tarantula and tarantula hawk. Nature Mag., 27:31-40.

Petrunkevitch, A. 1926. Tarantula versus tarantula hawk: a study in instinct. J. Exp. Zool., 45:367-394.

Petrunkevitch, A. 1952. The spider and the wasp. Sci. Amer., 187:20-24.

Punzo, F. 1990. The hemolymph composition and neurochemistry of the spider wasp, Pepsis formosa (Say) (Hymenoptera: Pompilidae). Comp. Biochem. Physiol., 96A:341-345.

Punzo, F. 1991. Nuerochemical events associated with learning and hunting behavior in The spider wasp, Pepsis formosa (Hymenoptera: Pompilidae). Florida Sci., 54:51-61.

Punzo, F. 1994a. The biology of the spider wasp Pepsis thisbe (Hymenoptera: Pompilidae) from Trans Pecos Texas. I. Adult morphometrics, larval development, and the ontogeny of larval feeding. Psyche, 101:229-241.

Punzo, F. 1994b. The biology of the spider wasp, Pepsis thisbe (Hymenoptera: Pompilidae) from Trans Pecos Texas. II. Temporal patterns of activity and hunting behavior with special reference to the effects of experience. Psyche 101:243-256.

Punzo, F. 1996. Localization of brain function and neurochemical events associated with learning in insects. Rec. Trends Comp. Biochem. Physiol., 2:9-16

Punzo, F. 2000. Desert Arthropods: Life History Variations. Springer, Heidelberg, Germany, 301 pp.

Punzo, F. 2002. Reversal learning and complex maze learning in the spider Aphonopelma hentzi (Girard) (Araneae, Theraphosidae). Bull. Br. Arachnol. Soc. 12:153-158.

Punzo, F. & B. Garman. 1989. Effects of encounter experience on the hunting behavior of the spider wasp, Pepsis formosa (Say) (Hymenoptera: Pompilidae). Southwest. Nat., 34:513-518.

Punzo, F. & L. Henderson. 1999. Aspects of the natural history and behavioural ecology of the tarantula spider, Aphonopelma hentzi (Chamberlin) (Orthognatha, Theraphosidae). Bull. Br. Arachnol. Soc., 11:121-128.

Sokal, R. & F. J. Rohlf. 1995. Biometry. 2nd edition. W. H. Freeman, New York, 907 pp.

Tinbergen, N. 1951. The Study of Instinct. Oxford Univ Press, New York, 228 pp.

Williams, F. X. 1956. Life history studies of Pepsis and Hemipepsis wasps in California (Hymenoptera: Pompilidae). Ann. Ent. Soc. Amer., 49:447-466.

FP at:

Fred Punzo

Department of Biology, University of Tampa

Tampa, Florida 33606
Table 1. The effect of encounter experience on the time (min) required
for females of Pepsis cerberus to complete various behavioral components
of the overall hunting sequence. Data expressed as means ([+ or -] SE)
for 10 wasps, APA (approach and antennation); MG (moving away and
grooming); ANP (antennation and paralyzation); BO (burial and
oviposition); CL (closure of the nest entrance). Values in columns
followed by different letters are significantly different (P < 0.01).

 Behavioral Components

 1 27.8a (3.7) 8.5 (1.2) 7.1a (0.7) 25.7 (4.9) 108.8 (12.4)
 2 31.4a (4.2) 11.3 (2.5) 6.6a (0.4) 29.4 (5.2) 116.3 (9.9)
 3 23.5a (4.7) 9.5 (1.9) 7.7a (0.6) 34.5 (6.7) 124.3 (13.1)
 4 21.4a (3.6) 13.4 (3.2) 5.7a (0.5) 26.7 (3.7) 104.2 (7.8)
 5 10.3b (1.6) 10.6 (2.1) 6.6a (0.7) 31.2 (5.5) 97.5 (6.5)
 6 12.3b (2.4) 8.8 (1.1) 4.3b (0.5) 28.8 (3.4) 119.5 (10.2)
 7 10.2b (1.1) 12.2 (2.5) 3.4b (0.6) 32.5 (5.7) 106.6 (11.2)
 8 6.5c (0.4) 9.8 (1.8) 3.1b (0.4) 33.4 (4.1) 114.7 (9.9)
 9 7.2c (0.6) 10.5 (0.8) 2.6b (0.3) 27.8 (3.8) 99.7 (8.9)
10 6.7c (0.3) 9.3 (0.6) 2.8b (0.2) 30.4 (4.1) 118.9 (10.6)
COPYRIGHT 2005 Texas Academy of Science
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2005 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Punzo, Fred
Publication:The Texas Journal of Science
Geographic Code:1U7TX
Date:May 1, 2005
Previous Article:A systematic checklist of the land snails of Louisiana.
Next Article:A comparison of brood stock origin for eastern wild turkey restoration in eastern Texas.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters