Host-Handling Behavior: An Innate Component of Foraging Behavior in the Parasitoid Wasp Ampulex compressa Tamar Keasar*, Noa Sheffer , Gustavo Glusman & Frederic Libersat  * Department of Life Sciences, Achva College, Mobile Post Shikmim, Israel  Department of Life Sciences, Ben Gurion University, Beer Sheva, Israel Introduction Learning, defined as any adaptive experience-induced change in behavior (Thorpe 1956), is well documen- ted in insects (Papaj 1993). Some of the best-known cases of insect learning concern foraging behavior. Insects learn to navigate to their food sources (Collett 1996); to associate chemical, visual or mechanical cues with the presence of attractive or repulsive foods (Hammer & Menzel 1995; Dukas & Bernays 2000); to adjust their diet selection, and time in feeding patches, to prey availability (Papaj & Prokopy 1989); and to handle their prey more efficiently as they gain experience (Chittka & Thomson 1997). Much of the research on insect learning deals with pollinating insects, in particular with honeybees and bumblebees. A wide array of choice tests has focused on associative learning in bees, i.e. learning to respond to a sensory cue that signals a food reward (Seeley 1995). A second direction of study concerns the operant learning capabilities of bees, namely learning to perform a behavioral sequence quickly and accurately to obtain a food reward. Pollinators improve in speed and accuracy of flower handling with repeated trials in field and laboratory tests, and retain their flower-handling skills overnight (Laverty 1994; Keasar et al. 1996). Flowers that are morpho- logically complex require a longer learning period than structurally simple flowers. Individuals that for- age on two flower types at the same time require a longer learning period (Chittka & Thomson 1997; Gegear & Laverty 1998). Parasitoids resemble pollinators in that they need to find, evaluate, and handle their prey (hosts), although most host tissue is consumed by the parasi- toid’s offspring, rather than by the foraging individ- ual (Quicke 1997; O’Neill 2001). Associative learning is clearly involved in the host-searching be- havior of parasitoids. Adult and larval parasitoids learn to respond to kairomones or visual cues associ- ated with the substrate of their prey (Vet & Groene- wold 1990; Kerguelen & Carde ´ 1998; Steidle 1998; Dukas & Duan 2000; Gandolfi et al. 2003). General- ist eucoilid parasitoids were shown to respond to these cues more than related specialist species (Pool- man Simons et al. 1992). Associative learning is also Correspondence Tamar Keasar, Department of Life Sciences, Achva College, Mobile Post Shikmim 79800, Israel. E-mail: tkeasar@bgumail.bgu.ac.il Received: July 26, 2005 Initial Acceptance: September 22, 2005 Final Acceptance: November 10, 2005 (L. Sundstro ¨ m) doi: 10.1111/j.1439-0310.2006.01214.x Abstract The reproductive success of parasitoids depends on their ability to locate, select, and handle hosts quickly and efficiently. Learning has been shown to play a role in host finding and host choice, but the role of learning in the handling of hosts has received little attention. We tested whether the speed and accuracy of host handling improve with experi- ence in the wasp Ampulex compressa. This parasitoid performs an elaborate behavioral sequence when parasitizing its host, the cockroach Periplaneta americana. We provide a quantitative description of the behavioral sequence, and show that: (a) the duration of the whole sequence, (b) the number of completed stages, and (c) the precision of an important stage in the handling sequence, host antennal cutting, are similar in inexperienced and experienced individuals. We discuss features of A. compressa’s life history that may select for innate host handling. Ethology Ethology 112 (2006) 699–706 ª 2006 The Authors Journal compilation ª 2006 Blackwell Verlag, Berlin 699