 1999 John Wiley & Sons, Inc CCC 0012-1630/99/010021-07 Steve Ray Ben Ferneyhough School of Biological & Molecular Sciences Oxford Brookes University Gipsy Lane, Oxford, OX3 0BP, UK Behavioral Development and Olfactory Learning in the Honeybee (Apis mellifera) Received 19 February 1998; accepted 28 May 1998 ABSTRACT: Honeybees (Apis mellifera) can be readily conditioned in the laboratory to spe- cific odors paired with a subsequent sucrose reward. A series of experiments are reported which demonstrate that the ability of bees to acquire and retain this learning is affected by the stage of behavioral development (caste). Results show rapid acquisition of the proboscis ex- tension reflex (PER) conditioning in adult forager bees, however much slower acquisition and poorer retention of the same learning paradigm in younger adults, i.e., nurse bees and guard bees. Further, if nurse bees are made to forage precocially by manipulation of the hive pop- ulation, these bees show excellent acquisition and retention of PER conditioning comparable to normal adult forager bees. Results are discussed in terms of olfactory learning requirements of bees performing caste-specific behaviors and the maturation of the bee nervous system.  1999 John Wiley & Sons, Inc. Dev Psychobiol 34: 21–27, 1999 Keywords: olfactory learning; development; proboscis extension reflex; caste; Apis mellifera INTRODUCTION The honeybee is renowned for its propensity to learn and remember a variety of tasks in all sensory modal- ities when tested both free flying in the field (Gould, 1986) and restrained in the laboratory (Menzel & Bit- terman, 1983). Perhaps the most widely researched learning paradigm in the literature is the proboscis ex- tension reflex (PER) conditioning (Bitterman, Menzel, Fietz, & Schafer, 1983; Menzel, 1985; 1990; Menzel, Erber, & Masuhr, 1974). PER conditioning has been established in a variety of laboratories investigating both the parametric characteristics of this learning (Hammer & Menzel, 1995) and the physiological basis of learning (Bicker & Menzel, 1989). The condition- Correspondence to: S. Ray ing consists of pairing a specific odor presentation to a restrained bee with a subsequent sugar reward. Most adult forager bees will require only one trial to acquire this learning and show excellent long-term retention. However, some variability in PER conditioning is in- timated in the literature (Menzel et al., 1974; von Frisch, 1967); unfortunately, this is substantiated by little systematic research. Several laboratories report seasonal variation in PER conditioning (Menzel, 1990; Menzel et al., 1974, p. 178; Ray & Ferneyhough, 1997a; von Frisch, 1967, p. 244) despite the fact the bees are trained and tested in the laboratory with little variation in conditions. One variable which may account for some seasonal variability and, of general interest to all bee learning research, is the well-defined and variable roles each animal performs within the colony at various stages in its development (Page & Robinson, 1991). Similarly, the number of bees performing specific colony related