key: cord-0034308-2n4axtcn authors: Brunjes, Peter C. title: Lessons from lesions: the effects of olfactory bulbectomy date: 1992-12-03 journal: Chem Senses DOI: 10.1093/chemse/17.6.729 sha: 0219019d8b2748b0aa802de596b4c40973c95805 doc_id: 34308 cord_uid: 2n4axtcn Olfactory bulb removal has been used to examine a wide-ranging number of topics. The present review outlines the categories of studies employing the technique, discusses some problems with the methodology and with previous interpretations of observed results, and suggests some potential avenues of investigation. Ablation techniques have played a central role in our present conception of neural organization and function. Attempts to understand brain circuitry in the first two-thirds of this century were dominated by reports examining patterns of lesion-induced degeneration; indeed, much of classical neuroanatomy was built on the technique. Similarly, attempts to examine neural function have also relied on the lesion approach: the role of many regions has been inferred after their removal or isolation from the remainder of the brain. These studies, in turn, have been used to interpret the effects of human brain damage, or, in some cases, to suggest treatment regimens (e.g. Valenstein, 1986) . Because of their prevalence, it is important to examine lesion techniques on a periodic basis. One reason is to ensure that researchers continue to appreciate the underlying assumptions upon which the methods are based. Brain lesions result in much more than a simple destruction of a discrete area: a cascade of changes emerges after the procedure. Edema and debris initially present at the lesion site can potentially affect relatively large areas of the brain. Thereafter many secondary consequences occur, resulting from disrupted vasculature, anterograde, retrograde and transneuronal degeneration, and changes in synaptic number and function as the result of factors such as sprouting, denervation supersensitivity and other forms of synaptic rebalancing (see section IIIB). These dynamic and myriad changes are often ignored by researchers examining single functional results at a particular time, despite the fact that there have been many previous reviews of the pitfalls associated with the methodology (e.g. Isaacson, 1976; Lynch, 1976; Schoenfeld and Hamilton, 1977; Steward, 1989) . Perhaps the easiest approach to examining the consequences of lesions would be to study their effects on brain areas receiving input from primary sensory neurons. These regions are intensely studied due to the fact that information flow is easily manipulated and relatively unidirectional. As a result, their circuitry, development and chemistry are relatively well understood. A reasonable assumption is that a lesion in these zones should simply deprive animals of that sensory modality. For example, removal of a cochlear nucleus should result in unilateral deafness or the removal of the olfactory bulbs should render animals anosmic. However, there are a wealth of other consequences of these brain lesions. Studying the range of these changes should lead to understanding €) Oxford University Press 729 of both the function of the brain regions in question and to an enhanced ability to interpret the consequences of lesion experiments in general. A second reason to review results from lesion studies periodically is to attempt to integrate the information in order to produce a more unified view of both the function of a particular area and the consequences of the lesions. Lesions of a specific brain region are done in order to test a number of different theoretical issues or to examine function from several perspectives. Often the studies are done without regard to each other, as many investigators do not look outside their own area of interest to find alternative interpretations. Below is a brief overview of the kinds of uses that have been found for olfactory bulbectomy, and a brief description of the changes encountered as a result. An enormous number of investigations have employed the surgical technique, and, as a result, a comprehensive list would be very difficult to produce. The present paper is designed simply to examine the numerous ways in which bulbectomy has been employed in order to demonstrate the extraordinary variety of changes which are a consequence (Section II) and to attempt to delineate both problems with interpreting the results of the technique and areas demanding further investigation (Section III). Several surgical approaches have been used to remove the olfactory bulbs. By far the most common has been to remove a portion of the overlying frontal bone and then to aspirate the bulb, sometimes after a caudal transection at the level of the rostral pole of the neocortex. Details of actual procedures are often sketchy (or even omitted) in many of the published reports, and histological verifications of the extent of the lesions, and/or correlations of lesion size with the amount of subsequent alterations in behavior, are quite rare. These details and analyses are far from trivial, as the topology of the bulb makes excision of the entire structure quite difficult, especially without damaging surrounding tissue (see below). Several excellent and extensive reviews of the anatomy, physiology and neurochemistry of the olfactory system have been published (e.g. Shepherd, 1972; Macrides et al., 1985; Switzer et al., 1985; Scott and Harrison, 1987; Halasz, 1990) . The olfactory bulbs, bilateral extensions of the rostral telencephalon which constitute ~4% of the volume of the rat brain (Cain, 1974a) , house two separate structures: the main and accessory olfactory bulbs. The main bulb receives input from the olfactory receptor cells which reside in the caudal third of the nasal cavity in the olfactory mucosa. The accessory olfactory bulb receives its afferent supply from the vomeronasal organs, tubeshaped structures found on the ventral midline of the anterior nasal cavity (Wysoscki, 1979; Halpern, 1987; Meredith, 1991) . Both regions have separate and parallel outputs. The main bulb projects via the lateral olfactory tract to the anterior olfactory nucleus, primary olfactory cortex, olfactory tubercle, entorhinal cortex, nucleus of the lateral olfactory tract, and anterior and posterolateral divisions of the cortical nucleus of the amygdala. The accessory olfactory bulb projects to separate regions of the amygdala: the medial nucleus and posteromedial division of the cortical nucleus (Switzer et al., 1985 and references cited therein; Price, 1987) . Therefore, bulb removal results in at least partial denervation of all of these regions. Fibers of the nervus terminalis also originate in the olfactory cavity and course along the medial surface of the bulb, entering the brain with the anterior cerebral artery and synapsing in the septa! region (Schwanzel-Fukuda and Silverman, 1980; Wirsig and Leonard, 1986a,b, 1987; Meredith, 1991) . Therefore, total bulbectomy would eliminate this afferent fiber system which has been demonstrated to contribute GnRH fibers to the accessory olfactory bulb and ventral forebrain. Interestingly, a superficial and incomplete bulb lesion might leave ventral portions of the main olfactory bulb intact while destroying both the vomeronasal-accessory olfactory bulb system and the nervus terminalis. The bulb also receives a massive input from higher brain structures. Indeed, it has been estimated that there are more centripetal bulb inputs than sensory afferents. These inputs include reciprocal fibers from many of the targets innervated by the bulb, a large cholinergic and GABAergic input from the region of the horizontal limb of the diagonal Anterior is at top, medial to right. Caudal border is in true coronal plane. Note that while border between the olfactory bulb and the anterior olfactory nucleus is quite clear, it does not lie in a corona] plane. The media] side of the bulb extends much further posteriorly than does the lateral side. (B) Coronal section (medial to right, dorsal to top) taken approximately 3/4 of a millimeter caudal to the frontal pole of the neocortex. Note that at this level a considerable amount of olfactory bulb is still present, along with a substantial portion of the anterior olfactory nucleus. Taken together, the two panels demonstrate that transection at the frontal pole will leave a significant portion of bulb intact, while damaging the rostral tip of the anterior olfactory nucleus. Key: 1 = main olfactory bulb, 2 = anterior olfactory nucleus, 3 =• = subependymal zone, 4 = accessory olfactory bulb, 5 = frontal pole of neocortex. Scale bar in (A) = 500 jun. the effects of bulbectomy on the subsequent re-innervation of the rostral forebrain. Results indicate that after bulb removal olfactory axons can make synaptic contacts with the remaining forebrain, including the olfactory tubercle, anterior olfactory nucleus, and the frontal neocortex. Furthermore, the axons do, indeed, appear to alter the cytoarchitecture of the target regions, producing cytology similar to that seen in normal olfactory glomeruli (Graziadei et al., 1978 (Graziadei et al., , 1979 Graziadei and Karlan, 1980; Graziadei and Samamen, 1980; Graziadei and Monti-Graziadei, 1986; Zigova et al., 1988; Monti-Graziadei and Graziadei, 1992) and inducing alterations in neurotransmitter expression (Guthrie and Leon, 1989) . However, the connections do not appear to be capable of transmitting usable olfactory information (Butler et al., 1984 ; but see Monti- Graziadei and Graziadei, 1992) . A2. Physical changes: higher structures (a) Connectivity. As classical tract-tracing techniques relied on degeneration argyrophilia, bulbectomy has been repeatedly used to examine olfactory system connections in a variety of species (Girgis, 1970; fish: Davis et al., 1981; reptiles: Gamble, 1956; Halpern, 1976; Ulinski and Peterson, 1981; Lohman et al., 1988; mammals: Powell et al., 1965; White, 1965; Heimer, 1968; Scalia, 1968; Lohman and Metink, 1969; Scalia and Winans, 1975; Skeen and Hall, 1977; Turner et al., 1978; Turner and Mishkin, 1978; Shammah-Lagnado and Negrao, 1981; Carlsen et al., 1982; Wouterlood and Nederlof, 1983; Wouterlood et al., 1985) as well as to study the development of the olfactory system (Leonard, 1975; Singh, 1977b) . The process of degeneration. The primary output pathway of the bulb, the lateral olfactory tract, forms a uniform bundle on the ventrolateral surface of the forebrain, and its terminations within the primary olfactory cortex have been well specified. As a result, olfactory bulbectomy has been used to examine the time course and morphology of lesion-induced axonal degeneration (e.g. Westrum, 1975; Caviness et al., 1977; Heimer and Kalil, 1978; Price, 1981, 1986a,b; Leonard, 1981; Westrum and Bakay, 1986) . (c) Anatomical reorganization. Lesion-induced axonal sprouting has been demonstrated to occur within several of the projections from the olfactory bulb. For example, after partial transection of the lateral olfactory tract remaining fibers sprout around the area of injury to re-innervate caudal portions of the primary olfactory cortex. These reorganized projections have been demonstrated by performing a subsequent bulbectomy and then staining for degeneration products (e.g. Devor, 1976; Small and Leonard, 1983 ). Westrum and his colleagues (e.g. Westrum, 1975 Westrum, , 1988 Westrum and Bakay, 1986; Westenbroek et al., 1988) as well as others (e.g. Ichikawa, 1987) have performed extensive electron microscopic investigations of the sequence of denervation and re-innervation of the primary olfactory cortex after bulbectomy. Sprouting of dopaminergic fibers into the olfactory tubercle has also been demonstrated to occur after bulbectomy (e.g. Gilad and Reis, 1979) . Similar sorts of reorganization have also been reported for the goldfish, which exhibits an impressive ability to compensate after central lesions (e.g. Zippel et al., 1988; Stewart, 1992) . (d) Neurochemical reorganization. Bulb lesions cause widespread changes in brain neurochemistry (reviewed by Hirsch, 1980; Van Riezen and Leonard, 1990) which have been exploited in order to examine a number of topics. Overall, the findings from the studies vary considerably, perhaps due to differences in levels of resolution of the techniques employed, and, importantly, due to differences in post-lesion survival times. As described in Section IIB10 below, bulb removal has been extensively employed as a model for studying the effects of antidepressant drugs. Many of the antidepressants affect monoamine systems and, as a result, most studies examine these transmitters. However, as outlined below, bulbectomy affects many neurochemical systems. (i) Norepinephrine: several investigators have reported that bulbectomy causes generalized reductions in neocortical norepinephrine (Pohorecky et al., 1969a,b; Pohorecky and Chalmers, 1971; Eichelman et al., 1972; Oishi and Ueki, 1978; Cairncross et al., 1979a) . However, Edwards et al. (1977) compared lesions of varying size and determined that the reductions were only seen after considerable damage to retrobulbar areas. Therefore, the effect seen earlier was probably due to damage to the noradrenergic fibers coursing anterior to the genu of the corpus callosum. Edwards et al.'s (1977) paper presents an important caveat to much of the work done examining changes in central transmitter content and activity (and, indeed, the bulbectomy literature as a whole), by suggesting that lesions limited to the olfactory bulb may have a relatively small impact. Noradrenergic changes have also been described in several other brain regions. Both Yoshimura and Ueki (1981) and Tonnaer et al. (1980) reported that bulbectomy results in increased amounts of norepinephrine in the hypothalamus. Iwasaki et al. (1986) , however, found similar changes only in those bulbectomized rats exhibiting increased mouse-killing behavior (see Section IIB5). Elevated noradrenergic turnover or transmitter levels have been found in the amygdala (Tonnaer et al., 1980; Jancsar and Leonard, 1984; Iwasaki et al., 1986) , although conflicting reports also exist (e.g. Broekkamp et al., 1986) . The affinity of a-adrenoreceptors was increased in the amygdala and hippocampus, suggesting a supersensitivity, but affinities were not altered in the cerebral cortex (Tiong and Richardson, 1990) . Higher number, but no changes in the affinity, of a 2 -adrenoreceptors, have also been reported in the homogenized forebrain (Hong et al., 1987) . Increased norepinephrine content has been reported in the primary olfactory cortex (Harvey et al., 1975) and some authors report a small increase in brainstem norepinephrine content although the finding has been inconsistent (e.g. Hirsch, 1980) . (ii) Dopamine: several studies have examined the consequences of bulbectomy on dopaminergic systems. Iwasaki et al. (1986) found decreased levels of dopamine in the lateral hypothalamus of muricidal bulbectomized rats. Tyler et al. (1979) reported bulbectomy-induced decreases in tyrosine hydroxylase activity in the striatum. As mentioned, bulbectomy-induced sprouting of dopaminergic fibers has been demonstrated in the olfactory tubercle. After bulb removal an increase in tyrosine hydroxylase activity and immunoreactivity, high affinity uptake of dopamine, activity of dopamine-sensitive adenylate cyclase, D, and D 2 receptor density, and Na + ,K + ATPase activity occurs in the olfactory tubercle (e.g. Gilad and Reis, 1979; Lingham and Gottesfeld, 1986; Swann and Gottesfeld, 1987) . (iii) Serotonin: the fact that both raphe lesions and bulbectomy induce aggressiveness in rats, along with the observation that many antidepressant drugs affect the serotonergic systems, has led to quite a bit of interest in bulbectomy-induced changes in brain serotonin. However, conflicting or incomplete evidence exists. For example, Neckers et al. (1975) reported decreased whole brain tryptophan hydroxylase activity six days after bulbectomy in mice. Harvey et al. (1975) and Garris et al. (1984) reported increased serotonin in the ventral forebrain. Decreased turnover of serotonin in the amygdala has also been reported (Jancsar and Leonard, 1984; Van Riezen and Leonard, 1990) . (iv) Acetylcholine: while Yoshimura (1981; see also Harvey et al., 1975) found no differences in choline acetyltransferase levels in seven brain areas between bulbectomized and control rats, Broekkamp et al. (1986) reported an increase 15 days after bulbectomy in the olfactory tubercle. Yoshimura et al. (1974) reported decreased acetylcholine content in the cerebral cortex after bulb removal. (v) Other transmitters and neurochemicals: changes in excitatory amino acid neurotransmitters in the primary olfactory cortex have been reported after bulbectomy. Many of these studies have been done in order to determine what neurotransmitters are used by bulb relay neurons. Concentrations of amino acids in the primary olfactory cortex are compared before and after bulbectomy and any decreases attributed to the lack of bulb input (e.g. Collins and Probett, 1981; Scholfield et al., 1983; Collins, 1984; Sandberg et al., 1984; and references therein) . Bulbectomy has also been reported to result in increased glycine in the olfactory cortex (Harvey et al., 1975) . Bulb removal causes a large reduction in histochemical staining for cytochrome oxidase, an enzyme involved in oxidative phosphorylation and, therefore, metabolism in the primary olfactory cortex (Onoda and Imamura, 1984) . The finding is consistent with many studies of anatomical or functional deafferentation (e.g. Korol and Brunjes, 1990) . Unilateral bulbectomy causes an increase in the number of opioid binding sites in the contralateral bulb (Hirsch and Margolis, 1980) , and changes in DNA and RNA concentrations in the contralateral hemisphere (Shafa, 1979; Shafa et al., 1980) . Alterations in amygdala GAB A (Jancsar and Leonard, 1981) and glutamic acid decarboxylase activity in the ventral tegmental area (Tyler et al., 1979) have been reported, along with alterations in benzodiazepine (Hirsch, 1981) and imipramine (Al-Khatib et al., 1988) receptor binding. Changes in the physiology of the primary olfactory cortex have also been examined (Becker and Freeman, 1968) . It is apparent from the above that bulb removal induces a broad number of changes in the central nervous system, from the rewiring of synaptic assemblies to the rebalancing of neurochemical systems. The resultant changes in behavior and physiological functioning, outlined below, are certainly as diverse. 1. Anosmia. The fact that olfactory bulbectomy results in an inability to detect odors has been known for a long time (e.g. Swann, 1934; Brooks, 1937; Brown and Ghiselli, 1938; Allen, 1941) , and the fact has been a primary motivation for using the technique. Bilateral bulbectomy has been demonstrated to result in deficits in olfactory detection and discrimination in rats (e.g. Long and Tapp, 1970; Slotnick and Schoonover, 1984; Amemori and Bures, 1988; Whishaw and Tomie, 1989) , as well as a number of other species (e.g. salamander: Mason and Stevens, 1981; sheep: Baldwin and Meese, 1977; Bell et al., 1979; Cohen-Tannoudji et al., 1986; rabbit: Klosovskii and Kosmarskaya, 1963; kangaroo rats: Webster and Webster, 1975; pigs: Parrot et al., 1985) . However, unilateral bulbectomy apparently does not affect olfactory sensitivity (Slotnick and Schoonover, 1984) . As mentioned above, the nasal cavity contains a number of chemoreceptive detectors, including the vomeronasal organ and endings of the trigeminal nerve. Bulbectomy or partial lesions are often used to distinguish which of these neural systems underlies particular behaviors (e.g. Ladewig et al., 1980; Wysocki et al., 1982; Hart and Leedy, 1985; Meredith, 1991) or to test the possibility of olfaction mediating a particular behavior (e.g. territorial marking, Thiessen et al., 1970) . points must be addressed (e.g. Murphy, 1976; and Section IIIC) . First, due to the convoluted nasal turbinates, lavage techniques seldom destroy all receptors. This observation may explain why some studies report that olfactory receptor damage is effective in reducing sexual behavior (e.g. Powers and Winans, 1973) while others claim it does not (see above). Secondly, bulbectomy destroys both the main and accessory olfactory systems, while nasal lavage may not affect the vomeronasal organ (but see Kelche and Aron, 1984) . Thirdly, some techniques of producing peripheral anosmia can be confounded by systemic poisoning (Sieck and Baumbach, 1974) . The large number of studies examining modifications in sexual behavior after bulbectomy has prompted study of the central consequences of the procedure. For example, bulbectomy results in increased estrogen receptor binding in the amygdala of female rats (McGinnis et al., 1985) and reduced androgen binding in the amygdala and hypothalamus of male rats (Lumia et al., 1987) . Changes such as these lend credence to the view that the behavioral consequences of bulbectomy may result from more complicated alterations than simple anosmia. Potential secondary changes, including widespread alterations in hormonal secretion and in photoperiodicity, are reviewed in the next section. 3. Hormones and photoperiods. Bulbectomy causes widespread and quite complicated changes in the neuroendocrine axis beyond the alterations in steroid hormone binding patterns mentioned above. For example, bulb removal sensitizes female rats to the anti-gonadal effects of melatonin administration (Reiter et al., 1980) and reverses the testicular regression induced by melatonin injections in male rats (Pieper et al., 1986b) . Changes in gonadotropin and prolactin secretion also occur in bulbectomized animals (Pieper and Gala, 1979; Galaer al., 1984 Galaer al., , 1985 Pieper et al., 1984 Pieper et al., , 1990a Clancy et al., 1986) and bulbectomy prevents the negative feedback of testosterone on the secretion of both LH and FSH by the pituitary in hamsters (Pieper et al., 1987) . Interestingly, in some respects peripherally-induced anosmia appears to have similar effects as bulbectomy. Furthermore, selective lesions of the vomeronasal-AOB system are apparently without effect, suggesting that the vomeronasal-accessory bulb system is not involved in the hormonal interactions (Mediavilla et al., 1985; Sanchez-Barcelo etal., 1985; Pieper et al., 1989) . Combining bulb removal with blindness results in striking changes in the neuroendocrine axis. Reiter et al. (1970) reported that neither procedure, when performed alone, affected the reproductive organ weights of adult female rats (however, see Whitten, 1956) . However, when both operations were combined, ovaries were found to be smaller than in controls (although number of eggs ovulated may not differ between groups; Peppier et al., 1973) . Pinealectomy reversed the effects of the combined sensory deprivation, suggesting that changes in melatonin levels were responsible for the alterations in gonadal size. Further research (Donofrio and Reiter, 1972; Leadem and Blask, 1972a,b; Shiino et al., 1972; Blask and Reiter, 1975; Ronnekliev and McCann, 1975) has indicated concomitant decreases in the size of the pars distalis of the pituitary, reductions in overall numbers of adenohypophyseal cells, smaller prolactin secreting cells and alterations in circulating growth hormone. Seasonal breeders (e.g. hamsters) exhibit a period of low reproductive competence characterized by reductions in gonadal size and function and inhibited gonadotropin secretion. These reductions appear to be controlled by melatonin produced by the pineal gland. As mentioned, the effects of melatonin (and thus short photoperiod) can be prevented by olfactory bulbectomy. Therefore, bulbectomized hamsters maintained on short photoperiod do not exhibit anestrus (Pieper et al., 1986) or testicular regression (Pieper et al., 1984) . Non-seasonal breeders, such as rats and mice (and pigs, Booth and Baldwin, 1983 ), normally are not influenced by day-length. However, olfactory bulbectomy has been demonstrated to potentiate photoperiodicity. Bulbectomized rats maintained on short daily periods of light have smaller testes and seminal vesicles and lower plasma testosterone than controls or animals on longer photoperiods, and these effects can be reduced by pinealectomy (e.g. Nelson and Zucker, 1981; Nelson et al., 1985; Nelson, 1990; Pieper et al., 1990a,b; and references therein). The hormonal underpinnings of these changes have been extensively studied (e.g. Pieper et al., 1990a,b) . Taken together, the results described above indicate that bulb lesions have robust and quite complicated effects on the pituitary-gonadal axis. Bulbectomy has also been demonstrated to interact with other hormonal systems. For example, lower basal levels and enhanced stress-induced increases in corticosterone (Loyber et al., 1976; Cairncross et al., 1977 Cairncross et al., , 1979b Broekkamp et al., 1986) heavier adrenal gland (Eichelman et al., 1972) , and an enhanced insulin response to glucose overloads (Perassi et al., 1972) have been reported after bulb removal. These changes, combined with the central neurochemical alterations outlined above (Section IIA2d), suggest that bulbectomy results in a complicated cascade of physiological adaptations. Bulbectomy has been reported to lengthen the active period of the circadian cycle in rats, mice and hamsters (Possidente et al., 1990; Pieper and Lobocki, 1991) , to delay the onset of entrained activity by over an hour and a half (Bittman et al., 1989; Possidente et al., 1990) , to increase activity levels during the dark phase of the circadian cycle (Giardina and Radek, 1991) , and to result in elevated cAMP levels in the suprachiasmatic nucleus (Vagell et al., 1991) . Pieper and Lobocki (1991) found circadian effects in castrated hamsters both with and without replacement hormones, and concluded, therefore, that the finding was not the result of changes in the pituitary-gonadal axis, but a direct effect on circadian oscillators in the suprachiasmatic nucleus. Aggression. It has long been known that bilateral bulbectomy increases the irritability and aggressiveness of rats (e.g. Watson, 1907; Vergnes and Karli, 1963; Didiergeorges and Karli, 1966; Didiergeorges et al., 1966; Douglas et al., 1969; Bernstein and Moyer, 1970; Sieck and Gordon, 1972) . However, understanding the effects of bulb removal on aggressive behavior is complicated by species and sex differences (e.g. Hilger and Rowe, 1975) , as well as the fact that many forms of aggression exist and they are not all equally affected by the surgery. For example, removal of the bulbs inhibits intermale aggressive behavior in mice (Ropartz, 1968; Rowe and Edwards, 1971; Bandler and Chi, 1972; Denenberg et al., 1973; Svare and Gandelman, 1974; DaVanzo et al., 1983; Whalen and Johnson, 1988) , guinea-pigs (Beauchamp et al., 1977) , hamsters (Hilger and Rowe, 1975; Murphy, 1976) and gerbils (Christenson et al., 1973; Rieder and Lumia, 1973; Hull et al., 1974) , presumably since the resulting anosmia makes the recognition of strangers difficult. Such an effect was not noted in sheep (Parrot and Baldwin, 1984) . Similarly, bulbectomy reduces pain-induced aggression: gentle electric shocks applied to mice housed with a novel partner elicited aggressive behavior more reliably in control animals than in bulbectomized mice (Fortuna and Gandelman, 1972) . However, bulbectomy potentiates interspecies aggression in rats. Normally, only a small number (10-30%) of male rats will kill mice when the animals are confined together. However, after bulbectomy 50-100% of the subjects exhibit muricide (Vergnes and Karli, 1963; Didiergeorges and Karli, 1966; Didiergeorges et al., 1966; Bandler and Chi, 1972; Ueki et al., 1972a,b; Thome et al., 1973 Thome et al., , 1974 Mast et al., 1974; Yoshimura « al., 1974; Hull and Homan, 1975; Oishi and Ueki, 1978; Yamamoto and Ueki, 1978; Yoshimura and Ueki, 1981; Moutzoukis et al., 1985; Thome and Rowles, 1988) . A considerable effort has been made to examine the neurochemicaJ changes which might underlie this behavior, with most studies concentrating on the major centripetal inputs into the olfactory bulb and, therefore, on potential retrograde changes in higher brain structures resulting from bulbectomy. Examinations of changes in the noradrenergic (Oishi and Ueki, 1978; Yamamoto et al., 1982; Hong et al., 1987) , cholinergic (Yoshimura and Ueki, 1974; Yoshimura et al., 1974; Yoshimura, 1981) and serotonergic (Neckers et al., 1975; Yamamoto and Ueki, 1977, 1978; Vergnes, 1978; Yamamoto et al., 1982 Yamamoto et al., , 1985 Garris et al., 1984; Hong et al., 1987; Al-Khatib et al., 1988) systems have been reported (see section IIA2D). Lesions of the amygdala reduced muricide, suggesting that disrupted relationships between it and the bulb might be responsible for the observed effects (Eclancher et al., 1975; Shibata et al., 1982) . It has also been reported that androgen injections can ameliorate the effects of bulb removal (e.g. Lumia et al., 1975 Lumia et al., , 1976 Lumia et al., , 1977 . Enhanced aggression after bulbectomy has been demonstrated in several other situations. Bulb removal has been reported to result in hyper-reactivity to heat and shock stimulation (Brown and Remley, 1971) , enhanced emotionality (Richman et al., 1972; Ueki et al., 1972a,b; Sieck, 1973; Nurimoto et al., 1974) , and increased aggression induced by food deprivation (Fortuna, 1977) . However, bulb lesions also result in a decrease in fear and timidity and, therefore, in enhanced exploration of novel environments (e.g. Mollenauer et al., 1974) . As in many of the other topics reviewed here, a considerable effort has been made to dissociate the effects of anosmia from the consequences of central nervous system damage. Several studies have compared the effects of peripherally-induced anosmia to bulbectomy and reported that only after the latter are rats more aggressive (e.g. Alberts and Friedman, 1972; Spectorand Hull, 1972; Sieck, 1973; Cain, 1974b; Cain and Paxinos, 1974; Tiffany et al., 1979) . These studies have been used to strengthen the position that there are 'non-sensory' effects of the surgery. However, two other lines of inquiry must also be considered. First, some investigators (e.g. Murphy, 1976 , territorial aggression in hamsters) have found little difference between the two techniques. Secondly, studies designed to separate the consequences of bulb removal from damage to retrobulbar regions suggest that damage to areas caudal to the bulb is more more effective in inducing enhanced aggressiveness than lesions restricted to the bulb (e.g. Sieck, 1973; Cain, 1974a; Albert et al., 1981) . 6. Activity. Many investigators report that bulb removal increased exploratory activity in an enclosed arena ('open field', Klein and Brown, 1969; Sieck, 1972; Sieck and Gordon, 1972; Sieck and Baumbach, 1973; Sieck et al., 1973 Sieck et al., , 1974 Hilger and Rowe, 1975; Burge and Edwards, 1976; Hull et al., 1979; Misslin and Ropartz, 1981; Gomita et al., 1984; Broekkamp et al., 1986; Thome and Rowles, 1988) and decreases activity in running wheels (Marks et al., 1971; Borer et al., 1974) . Once again, peripherallyinduced anosmia does not seem to affect activity (Sieck and Baumbach, 1974) , and considerable effort has been spent trying to determine the underlying changes in neurochemistry responsible (e.g. Leonard and O'Connor, 1987) . Increased activity has been interpreted as: (i) resulting from anosmia, so that animals are forced to seek more stimulation in order to gauge the nature of the novel environment; (ii) a result of a decrease in the rate of habituation in general (e.g. Cheal and Domesick, 1979) ; or (iii) as a consequence of alterations in the function of the iimbic system' areas innervated by the bulb (see Section HID). Learning. It should come as no surprise that anosmic animals do poorly on tasks in which olfactory cues are salient (Herrick, 1933) . However, there are many studies suggesting that more complicated changes result from bulbectomy. Examinations of alterations in learning abilities in lesioned animals are complicated once again by the many tasks employed and species examined. For example, bulbectomized pigeons are slower to learn to peck a key for food reward than controls, but once the habit is acquired, they are no slower at learning a visual discrimination task. Bulbectomized birds were reported to be 'hesitant and inactive during early stages of training' (Wenzel and Salzman, 1968 ; see also Wenzel et al., 1969) . Bulbectomized rats trained in a visual discrimination task exhibit deficits in acquisition as well as in the final level of performance (Phillips, 1970) . Bulbectomized rats do not differ from chance in a spontaneous alternation situation, suggesting that they do not remember which arms they have previously frequented, or cannot discriminate them (Douglas et al., 1969; Klein and Brown, 1969) . They also exhibit inferior performance on passive avoidance tasks (Marks et al, 1971; Sieck, 1972; Archer et al., 1984a,b; Moutzoukis