Introduction

Many of the chapters in this volume have highlighted the way in which different sensory systems work together to influence perception and behavior as well as the activity of neurons in different regions of the brain. In order for these cross-modal interactions to be of value, however, the various sensory signals arising from a common source must be coordinated within the brain. This requirement applies both to amodal features, which include the timing, location, and intensity of the stimuli, and to modality-specific properties, such as voice pitch and facial features.

This capacity to combine information across different sensory channels to form a coherent multisensory representation of the world has its origin in the way in which the sensory systems interact during development. Behavioral and neurophysiological studies carried out on a variety of species have highlighted the importance of sensory experience in this process (see Gutfreund & Knudsen, Chap. 38; Wallace, Chap. 39; and Lickliter & Bahrick, Chap. 40, this volume). In particular, experience plays a critical role in matching the neural representations of spatial information provided by the different sensory systems. As in the adult brain, most developmental studies of multisensory processing at the neuronal level have focused on the superior colliculus (SC; see Stein, Jiang, & Stanford, Chap. 15, this volume).

The choice of species in which to investigate the development of multisensory spatial integration is governed by a number of scientific as well as economic and clinical factors. Like many other groups interested in the early development and plasticity of sensory systems (e.g., Chapman, Stryker, & Bonhoeffer, 1996; Dantzker & Callaway, 1998; Henkel & Brunso-Bechtold, 1995; Juliano, Palmer, Sonty, Noctor, & Hill, 1996; Moore, Hutchings, King, & Kowalchuk, 1989; Weliky & Katz, 1999; see also Sur, Chap. 42, this volume), we use the ferret, Mustela putorius furo, for our work on the SC. The ferret is a carnivore whose central nervous system is organized much like that of the more extensively studied cat. Both are altricial species, although the ferret is particularly immature at birth. Indeed, evoked-potential measurements show that these animals are unable to hear until near the end of the first postnatal month (Moore & Hine, 1992). Eye opening typically occurs a few days after this, although responses to complex visual stimuli can be recorded in the thalamus and cortex through the closed eyelids as early as 21 days after birth (Krug, Akerman, & Thompson, 2001). Despite their relative immaturity, young ferrets are robust and highly suitable for both in vivo and in vitro electrophysiological recording experiments. Moreover, mature ferrets can be readily trained to make behavioral responses to remotely presented stimuli (e.g., Kelly, Kavanagh, & Dalton, 1986; King, Parsons, & Moore, 2000; Moore, Hine, Jiang, Matsuda, Parsons, & King, 1999; von Melchner, Pallas, & Sur, 2000). This allows the functional significance of experience-driven changes in neuronal response properties to be assessed and provides a valuable link to psychophysical studies in humans.

In this chapter, we examine the development of sensory map registration in the mammalian SC, with an emphasis on the steps that lead to the emergence of spatially tuned auditory responses and their alignment with visual receptive fields (RFs). Many of these studies have been carried out in ferrets, with experimental approaches ranging from measurement of orienting behavior to the response characteristics of individual neurons (see King, 1999; King & Schnupp, 2000; King et al., 2000, for recent reviews).