Does the Lateral Bundle of the Medial Olfactory Tract Mediate Reproductive Behavior in Male Crucian Carp?

Transcription

1 Chem. Senses 28: , 2003 Does the Lateral Bundle of the Medial Olfactory Tract Mediate Reproductive Behavior in Male Crucian Carp? Finn-Arne Weltzien, Erik Höglund, El Hassan Hamdani and Kjell B. Døving Division of General Physiology, Department of Biology, University of Oslo, Oslo, Norway Correspondence to be sent to: Kjell B. Døving, Division of General Physiology, Department of Biology, PO Box 1051, University of Oslo, 0316 Oslo, Norway. Abstract The olfactory tract in crucian carp (Carassius carassius) is divided into three distinct bundles: the lateral tract (LOT) and the lateral (lmot) and medial (mmot) bundles of the medial tract. The LOT has been shown to mediate information associated with feeding behavior, whereas the mmot mediates information associated with alarm response. The role of the medial olfactory tract (lmot and mmot) in reproductive behavior is still under debate. In the present experiment, male reproductive behavior towards prostaglandin-injected females was investigated before and after cutting off the different olfactory tract bundles, to determine which of the tract bundles is essential for mediating reproductive behavior in male crucian carp. The fish were maintained in physiological saline before and after surgery to preserve the remaining tract bundles. Operations were performed symmetrically on both sides and post-operative inspections revealed the functionality of the intact tracts. Sham-operated males and males with only the lmot intact showed typical reproductive behavior, with following of the female and inspections of the female anal papilla. However, males in which the lmot was cut, leaving both the mmot and the LOT intact, showed reduced reproductive behavior. Our results suggest that the lmot mediates reproductive behavior in male crucian carp. Key words: fish, olfaction, smell, nerve transection, spawning, pheromones Introduction The fish olfactory system mediates different types of behaviors that are essential for various life processes, such as feeding, alarm reaction and reproduction (Døving, 1986). In gadids, siluroids and cyprinids, the paired olfactory tract is clearly divided into three distinct bundles (Sheldon, 1912). These are designated as the lateral olfactory tract (LOT), the lateral bundle of the medial olfactory tract (lmot) and the medial bundle of the medial olfactory tract (mmot). The different tract bundles contain fibers that originate from different portions of the olfactory bulbs (Dubois-Dauphin et al., 1980; Satou et al., 1983; Satou, 1990). Also, each bundle connects to different brain areas, although some overlap between bundles seem to exist (Finger, 1975; Rooney et al., 1992). Moreover, the different bundles contain fibers with different electrophysiological and histological properties (Døving and Gemne, 1965). These specific morphological and electrophysiological properties strongly suggest that the individual bundles also have specific functional properties and that activity in each individual bundle mediates information about a specific behavior. Døving and Selset (Døving and Selset, 1980) first provided evidence that a particular tract bundle mediates a particular behavioral response. Electrical stimulation of discrete tract bundles in free-swimming Atlantic cod (Gadus morhua) showed that the LOT mediates feeding behavior, the lmot mediates spawning behavior and the mmot mediates alarm behavior. Further evidence for a functional separation of the olfactory tract bundles has been reported in several behavioral and physiological studies. Goldfish (Carassius auratus) lost their ability to discriminate between different amino acids after transection of the LOT (von Rekowski and Zippel, 1993) and the LOT mediates feeding responses in crucian carp, Carassius carassius (Hamdani et al., 2001). Alarm behavior was absent in crucian carp after transection of the mmot (Hamdani et al., 2000). The MOTs (bundle not specified) are necessary for mediation of reproductive behavior in male goldfish (Stacey and Kyle, 1983; Kyle et al., 1987). Moreover, known sex pheromones selectively elicited electrical activity in the MOTs of male goldfish (Sorensen et al., 1991), whereas goldfish sperm release was induced by electrical stimulation of the MOTs (Demski and Dulka, 1984). However, the particular bundle of the MOT responsible for mediating reproductive behavior has never been differentiated in an experimental situation involving male and female conspecifics. Goldfish, crucian carp and common carp (Cyprinus carpio) are the most extensively studied fish species with regard to the role of olfactory signals in reproduction. When Oxford University Press All rights reserved.

2 294 F.-A. Weltzien et al. approaching ovulation, female goldfish release a mixture of steroid-hormone-derived sex pheromones through the urine, eliciting reproductive behavior in spermiating males (Stacey and Sorensen, 1986; Scott and Sorensen, 1994; Sorensen et al., 1995, 1996). 17α,20β-Dihydroxy-4-pregnen-3-one (DHP) is released preovulatory, while prostaglandin F 2α (PGF 2α )is released after ovulation, both pheromones exerting primer and releaser effects. The males respond to DHP through elevated plasma concentrations of luteinizing hormone, leading to spermiation and increased sperm volume, the primer effect (Dulka et al., 1987), but DHP also stimulates male reproductive behavior, the releaser effect (DeFraipont and Sorensen, 1993; Poling et al., 2001). PGF 2α stimulates male reproductive behavior, including active chasing and nudging of the female (Sorensen et al., 1988, 1996; Bjerselius and Olsen, 1993; Stacey et al., 1994), but also stimulates male release of luteinizing hormone (Sorensen et al., 1989), although through different mechanisms than those mediating the DHP effect (Zheng and Stacey, 1996, 1997). In behavioral studies, injections in immature females of PGF 2α are often used. This will induce release of sex pheromones and reproductive behavior in males and females within minutes (Partridge et al., 1976; Stacey, 1981). The aim of the present study was to investigate which of the olfactory tract bundles is responsible for mediating reproductive behavior in male crucian carp. This was investigated by observing the behavior of spermiating males that had undergone different types of olfactory tract transections towards PGF 2α -injected females. In view of previous results (Døving and Selset, 1980; Hamdani et al., 2000, 2001), we decided to focus our investigation on the role of lmot in crucian carp reproductive behavior. Materials and methods Experimental animals Crucian carp (both juveniles and sexually maturing individuals) were caught in a small lake outside of Oslo, Norway (60 N), in June The fish were transported to the Department of Biology and kept in 700 l fiberglass stock tanks without vegetation. The temperature of the running water was constant between 7 and 9 C, while the photoperiod was maintained at 12 h/12 h light/dark. The fish in the stock tanks were fed once a day on Tetrapond pond sticks (Tetra GmbH, Germany). The fish were sorted on the basis of sex and kept in separate tanks from the time when spermiating males were first observed (no sexual dimorphism was observed). The fish used in experimental trials had the following body measurements: male (spermiating) body wt was 24.6 ± 1.3 g (mean ± SEM; n = 23) and male gonadosomatic index GSI = [100 (gonad wt/body wt)] was 2.8 ± 0.2. The female (vitellogenic, used as spawning partners) body wt was 42.4 ± 2.2 g (n = 16), while the female GSI was 13.8 ± 0.5. Experimental design The reproductive behavior trials were conducted in six 40 l aquaria containing still, aerated physiological saline (in g/l: NaCl, 8.53; KCl, 0.22; MgSO 4 7H 2 O, 0.25; CaCl 2 6H 2 O, 0.28) at 21ºC. All aquaria contained artificial gravel and floating vegetation made of green yarn attached with rubber bands to floating corks. This artificial spawning substrate covered ~30% of the water surface. All behavioral trials involved the use of one male (with running milt, tested only once) and one PGF 2α -injected vitellogenic female. On day 1, males were acclimatized to the physiological saline by transferring them from the stock tank to individual experimental aquaria. At the time of transfer, the water temperature in the experimental aquaria was identical to the stock tank temperature. The following day (day 2), when the physiological saline had reached room temperature (21ºC), one PGF 2α -injected female was introduced to each individual male and the behavior was observed and videorecorded for 10 min. If the male showed more than four short followings (Fs) or short followings directed against the female anal papilla (FsG; see below), the male was considered sexually active and used in the trials. Males that failed the test were excluded from the experiment. The video-recorded behavior performed by the males that passed the test was further analyzed and the frequency of reproduction-related behaviors (see below) was used as a preoperative index of reproductive behavior (pre-operation). These sexually active males thereafter underwent bilateral transection of one or more of the olfactory tracts according to the procedure given below. Earlier studies have shown that the LOT mediates feeding responses (Hamdani et al., 2001), while the mmot mediates alarm behavior (Hamdani et al., 2000) in crucian carp. In view of this and also other results (Døving and Selset, 1980), which indicated that reproductive behavior is mediated through the lmot, we decided to investigate the reproductive behavior in three groups of crucian carp: (i) sham-operated males; (ii) males with the LOT and the mmot transected, named lmot intact ; and (iii) males with the lmot transected, but the LOT and the mmot intact, named lmot cut. The operated males were returned to their aquaria directly after surgery and allowed to recover overnight. The following morning, a new PGF 2α -injected female was introduced to each male and the behavior was video-recorded for 10 min and further analyzed. The frequency of reproductionrelated behavior was used as a post-operative index of reproductive behavior (post-operation). Females were injected i.m. on both sides of the dorsal muscles ~15 min before the start of the trial with 200 µl PGF 2α (~0.5 µg PGF 2α /g body wt; dissolved in 0.9% NaCl. This induces spawning behavior for at least 1 h post-injection (Stacey, 1981; Bjerselius and Olsen, 1993). Following PGF 2α -injection, each female was used for three consecutive trials with three different males

3 Reproductive Behavior in Crucian Carp In nine males, the LOTs and the mmots were left intact, while the lmots were transected (lmot cut). After surgery, the cranial cavity was filled with 2% agar dissolved in physiological saline to reduce infection. Thereafter, the males were returned to their observation aquaria. At the end of the experiment, each fish was visually inspected for possible regeneration of the tract bundles and for proper blood flow in the vessels along the tracts. Figure 1 Schematic outline of the olfactory system and the forebrain in crucian carp (Carassius carassius). Cyprinids (and siluroids and gadids) have a short olfactory nerve leading from the sensory epithelium to the olfactory bulb, and long olfactory tracts between the olfactory bulb and telencephalon. Three individual tract bundles can be seen by the naked eye: the lateral olfactory tract, LOT; the lateral bundle of the medial olfactory tract, lmot; and the medial bundle of the medial olfactory tract, mmot. from different treatment groups. The order of males was randomized. After the trials, the males were decapitated for determination of body wt, sex and GSI. Experimental trials were conducted from the middle of June to early September, between and h. All experiments were conducted in accordance with the Norwegian Animal Welfare Act as approved by the Norwegian Animal Research Authority. Surgical procedure After the pre-operation behavior trial, sexually active males were transferred from the observation aquaria, anesthetized with benzocaine (45 mg/l; placed in a stand with running water with anesthetic through the mouth and over the gills and operated on under a stereomicroscope. The skin was removed from a 0.7 cm 2 area on the skull just posterior from the eyes and the underlying dorsal cranium was removed. Cranial fluid was removed using filter paper, while the mesenchymal tissue in the brain case and the meninges around the olfactory tracts were removed with fine forceps. The olfactory tracts were clearly visible as three distinct bundles running from the olfactory bulb to the brain (Figure 1). The bundles were gently separated with a needle and specific bilateral bundle transections were made using fine scissors, taking care not to disrupt the blood vessels running parallel with the tract bundles. Approximately 2 mm of the transected bundles was removed to prevent regeneration (von Rekowski and Zippel, 1993; Zippel et al., 1993). The males were divided into three treatment groups, as follows. 1. Eight males were sham-operated, i.e. the brain case was opened and the meninges and mesenchymal tissue were removed to expose the olfactory tracts (sham). 2. In six males, the lmots were left intact, while the LOTs and the mmots were transected (lmot intact). Behavioral analysis All behavioral experiments were conducted with one spermiating male and one PGF 2α -injected female at a time. The video-recorded male female interactions were analyzed blind for the following behaviors, modified from earlier studies (Partridge et al., 1976; Bjerselius et al., 2001): 1. Short following (Fs): the male follows the female with its head in close vicinity of the tail of the female for a period of <5 s. The frequency of the behavior was recorded. 2. Short following gut (FsG): the male follows the female with its head in close vicinity of the female anal papilla for a period of <5 s. The frequency of the behavior was recorded. 3. Pushing (Pu): the male pushes the female with the side of his head. A pushing male makes a horizontal movement with its head so that the operculum hits the side or the head of the female. The frequency of the behavior was recorded. 4. Biting/nipping (BN): the male bites or nips the female. The frequency of the behavior was recorded. 5. Spawning: the male and the female swim together towards the vegetation. They turn to the side and swim in parallel while they release gametes. We did not observe behaviors in the present experiments that could be related to spawning. 6. Picking (Pi): the number of bites directed toward the gravel substrate. This behavior was included because it has been used to describe reduced spontaneous feeding activity in sexually aroused males. To verify that the observed behavior variants were male female specific, we also observed male male behavior in the observation aquarium (n = 5). The male male trials were conducted as for the male female trials, with one male being acclimated in its aquarium for 24 h and the second male added on day 2. Data analysis Data were analyzed using Statistica version 5.1 (Statsoft Inc., Tulsa, OK) and presented as mean values ± SEM. Possible effects of olfactory tract bundle transections on male reproductive behavior were tested using a repeatedmeasurement analysis of variance (ANOVA), followed by the Tukey HSD post hoc test for unequal sample sizes (Spjotvoll and Stoline test).

4 296 F.-A. Weltzien et al. Table 1 Average (±SEM) scores of reproductive behavior patterns in the three experimental groups of male crucian carp pre- and post- operation. The male behavior towards a PGF 2α -injected female was observed for a period of 10 min Behavior Sham (n = 8) lmot intact (n = 6) lmot cut (n = 9) ANOVA Pre-operation Post-operation Pre-operation Post-operation Pre-operation Postoperation Short following (Fs) Short following gut (FsG) 4.13 ± ± ± ± ± ± 0.71 F(2,20) = 0.76; P = ± ± ± ± ± ± 0.17* F(2,20) = 21.34; P = Pushing 0.63 ± ± ± ± ± ± 0.24 F(2,20) = 0.05; P = 0.95 Biting/nipping 1.38 ± ± ± ± ± ± 0.46 F(2,20) = 1.67; P = 0.21 Picking 4.50 ± ± ± ± ± ± 0.18 F(2,20) = 1.64; P = 0.22 *Significant difference (P < 0.05) from the scores within the same behavior. Results The reproductive behavior of spermiating male crucian carp towards PGF 2α -injected females was observed before and after bilateral transection of the different olfactory tract bundles (Table 1). The following behavior patterns were observed: Fs, the male made repeated short followings after the female; FsG, the male made repeated short followings directed towards the female anal papilla; Pu, the male pushed the female with its head; and BN, the male bit or nipped on the female body. Occasional picking at the gravel or at the floating vegetation was also observed. There was no significant difference between the treatment groups (sham; lmot intact; lmot cut) in the observed behaviors before surgery (Table 1). However, bilateral transection of thelmotsignificantlyaffectedfsg[f(2,20) = 21; P = ] and resulted in a significant decrease in FsG in post-operated compared to pre-operated males (P = ; Figure 2). Furthermore, the fish with transected lmot had significantly lower scores of FsG compared to the other treatment groups: sham pre-operation (P = ), sham post-operation (P = ); lmot intact pre-operation (P = ), lmot intact post-operation (P = ). We did not observe significant pre- or post-operative effects between the treatment groups in the other observed behavior types (Table 1). In the male male behavior trials, two different scenarios were evident. Either the two males remained largely inactive throughout the 10 min observation period, or they displayed aggressive behavior. Two different types of aggressive behavior were evident in the male male situation: highspeed chasing around the aquarium and also one male blocking the swimming path of the other. None of the Figure 2 Average (±SEM) scores of short following gut (FsG) in the three experimental groups of male crucian carp, pre- and post-operation. The male behavior towards a PGF2α-injected female was observed for a period of 10 min. Asterisk indicates significant difference compared to the lateral bundle of the medial olfactory tract (lmot) pre-operation, or compared to the other groups pre- and post-operation (P < ). male female-typical behavior patterns were observed between two males. Discussion Extensive experimental evidence indicates that ovulated females in most species of fish, including the crucian carp (Bjerselius and Olsén, 1993), release pheromone(s) that stimulates reproductive behavior in male conspecifics. As these pheromones fail to elicit a full response in males in which olfactory function has been impaired, reproductive behavior most likely is mediated by the olfactory system (Liley, 1982). In the present work, we observed significant differences in reproductive behavior between males with

5 Reproductive Behavior in Crucian Carp 297 the lmot intact and those with the lmot cut. Although previous studies have concluded that reproductive behavior is mediated by the MOT (Stacey and Kyle, 1983), the present work gives the first direct evidence that only the lateral bundle of the MOT is necessary for full reproductive behavior in male crucian carp. Whether this feature is valid for all teleosts remains an open question. Several earlier investigations have indicated a role for the MOT in teleost reproductive events. For example, only axons in the MOT project to the ventromedial telencephalon and the preoptic area in goldfish (Oka et al., 1982; von Bartheld et al., 1984) and Atlantic cod (Rooney et al., 1992). Electrical stimulation in these brain areas elicits sperm release in anesthetically immobilized goldfish (Demski and Hornby, 1982). Lesions in the same areas impair courtship behavior in male goldfish (Kyle and Peter, 1982; Koyama et al., 1984) (Table 2). Also, electrical stimulation of the olfactory tracts after transection of one or more of the tract bundles revealed that the MOT induces sperm release in goldfish (Demski and Dulka, 1984). Known sex pheromones specifically elicited electrical activity in the MOT of male goldfish (Sorensen et al., 1991). A further specification of the roles of the lateral and medial bundles of the MOT in reproduction-related activity was impossible or not attempted in the studies noted above. To our knowledge, only Døving and Selset (Døving and Selset, 1980) have successfully differentiated between the different bundles of the MOT, as they induced different types of stereotypic behavior in individual free-swimming Atlantic cod by electrical stimulation of single tract bundles. Electrical stimulation of the lmot induced spawning-like behavior, although the responses to conspecifics were not tested in these fish. Stacey and Kyle (Stacey and Kyle, 1983) investigated the effect of olfactory tract transection on male goldfish reproductive behavior. Although they investigated single tract bundles, they concluded that transection of the entire MOT Table 2 teleosts reduced reproductive behavior in male goldfish (Stacey and Kyle, 1983; Kyle et al., 1987). However, a closer look at their results reveals some interesting points: transection of the lmot reduced the duration of spawning behavior. However, the same effect was observed after transection of the mmot. Moreover, transection of the mmot and the LOT reduced the duration of spawning behavior, an effect not observed when the lmot and the LOT were transected. These results led the authors (Stacey and Kyle, 1983) to suggest that both the lmot and the mmot were involved in mediation of spawning behavior, hence their general conclusion that the entire MOT is necessary for full reproductive activity. The present study demonstrates that bilateral transection of the male lmot decreases the frequency of short following directed to the female anal papillae (FsG) in crucian carp. Moreover, transection of the mmot and LOT did not affect any of the observed reproductive behavior patterns, thus contrasting the earlier results (Stacey and Kyle, 1983). However, Stacey and Kyle (Stacey and Kyle, 1983) measured the total duration (and not the frequency) of spawning behavior and they did not differentiate between Fs and FsG as in the present study. Such distinction in analysis might explain the differences in the results between the studies. In studies on anosmic goldfish (Partridge et al., 1976), the authors did not differentiate between Fs and FsG. They (Partridge et al., 1976) showed that the anosmic goldfish decreased the time spent following an ovulating female, but increased the frequency of the same behavior, as compared to control fish. Other sexual behaviors observed in that study, including Pu and spawning behavior, showed the opposite relation between frequency and time spent, further indicating that following the female is not solely a sexual behavior induced by pheromones. We did not observe actual spawning behavior in the present study, although this is frequently seen when experiments on goldfish are made under similar conditions (Bjerselius et al., 2001). Also, the frequency of Pu was relatively low A summary of the literature on reproduction-related behavioral or physiological reactions associated with the medial olfactory tract (MOT) in Associated with Response Species Reference MOT Mediates reproductive behavior Carassius auratus (Stacey and Kyle, 1983) MOT Elicits sperm release Carassius auratus (Demski and Dulka, 1984) MOT Tract activated by known sex pheromones Carassius auratus (Sorensen et al., 1991) MOT Axons project to the preoptic area Carassius auratus (Oka et al., 1982; von Bartheld et al., 1984) Gadus morhua (Rooney et al., 1992) MOT Stimulation of preoptic area elicits sperm release Carassius auratus (Demski and Hornby, 1982) MOT Lesions in preoptic area impair courtship behavior Carassius auratus (Kyle and Peter, 1982; Koyama et al., 1984) lmot Mediates spawning-like behavior Gadus morhua (Døving and Selset, 1980) lmot Mediates reproductive behavior Carassius carassius Present paper

6 298 F.-A. Weltzien et al. (0.6/10 min) in the present study compared to experiments on goldfish: 5 10/10 min (Partridge et al., 1976; Bjerselius et al., 2001). We can think of two possible explanations. First, the wild-captured crucian carp do not seem to acclimatize as well to laboratory conditions (they seem to have a higher activation level, i.e. they get easily scared) compared to the highly domesticated goldfish, even after several months in captivity. This may influence their reproductive expenditure, leading to reduced sexual performance. Secondly, it is possible that our experimental set-up did not allow the males to perform their whole repertoire of sexual behavior. However, because we essentially copied the set-up from earlier work (Bjerselius et al., 2001), we find this explanation rather unlikely. None the less, the present study showed that spermiating males with the lmot cut had a clear decrease in FsG compared to individuals with the LOT and the mmot cut. In both Atlantic cod (Døving and Selset, 1980) and crucian carp (Hamdani et al., 2000), the mmot mediates alarm reaction. Spawning-like behavior was induced by electrical stimulation of the lmot in cod (Døving and Selset, 1980) and our results are in accordance with that finding. Thus, the findings presented here and in previous studies suggest that only the lmot and not the entire MOT is necessary for mediating full reproductive behavior in crucian carp. In the present study, not all behavior patterns that are assumed to be reproduction-related showed reduced frequency in males with the lmot cut. They all responded to the female with Fs, Pu, and BN and with frequencies not significantly different between treatment groups or between pre- and post-operation trials. Although the olfactory system is required to mediate full reproductive response, these behavior patterns (Fs, Pu and BN) may also be mediated by other sensory systems (e.g. the visual and/or tactile), as they were present with similar frequencies even in males in which the entire olfactory tract system had been cut (n =2;data not shown). This is consistent with earlier results (Partridge et al., 1976), who found that male goldfish that had been made anosmic by occlusion of the nares still followed ovulated females about three times as much as they followed non-ovulated females. Also, while the female blue gourami (Trichogaster trichopterus) evidently uses the olfactory system to locate the source of male sex pheromone (Lee and Ingersoll, 1979), anosmia has little effect on the spawning success of females that are already in visual contact with the males (Pollack et al., 1978). However, our results are in contrast to those of Bjerselius et al. (Bjerselius et al., 2001), who based their analyses of reproductive behavior in male goldfish on a sexual behavior index that was the sum of the number of Fs, Pu and spawning attempts within a set time period. Even after adjusting for the number of spawning attempts observed by Bjerselius et al. (Bjerselius et al., 2001), we found no significant differences between treatment groups or between pre- and post-operated crucian carp males using this index. The behavior trials described in Bjerselius et al. (Bjerselius et al., 2001) were similar to those described in the present study and included one male and one PGF 2α -injected female. It is thus possible that the behavioral patterns on which their index is based are not mediated solely through the olfactory system, but also through the visual and/or the tactile systems. Alternatively, although the general repertoire of behavioral patterns is similar in goldfish and crucian carp, differences could exist between the species concerning the expression of the different reproductive behavior patterns. In conclusion, this study demonstrates that the lmot mediates odor-induced reproductive behavior in male crucian carp. Previous studies on crucian carp have shown that the mmot mediates alarm reaction (Hamdani et al., 2000) and the LOT mediates feeding behavior (Hamdani et al., 2001). These ablation experiments indicate that each of the distinct olfactory tract bundles transmits a unique message to the brain. In Atlantic cod, electrical stimulation of the olfactory tract bundles induces different behavior patterns associated with feeding (LOT), reproduction (lmot) and alarm (mmot) (Døving and Selset, 1980). Thus, in two species of teleosts from two different families, there is accordance between the morphological entities and the behavioral messages that the olfactory tract bundles mediate. This agreement between morphology and function supports the idea of labeled lines discussed by Erickson (Erickson, 1963) and also poses the question of the universality of these findings. Future studies of labeled lines will be aided greatly by new techniques that allow tracing of the connections that specific sensory neurons make in the brain (Zou et al., 2001). One might ask if the association between the morphological entities and the behavioral messages that the olfactory tract bundles mediate is a general phenomenon, valid for all teleosts, or if it is a coincidence for these two species only. Furthermore, there are other behavioral patterns associated with the olfactory system that have not been associated with a particular bundle of the olfactory tract, for example homing ability (Døving and Stabell, 2003). The question of which bundle of the olfactory tract mediates homing behavior must be raised in forthcoming studies. Acknowledgements This work was supported by the Norwegian Research Council. References Bjerselius, R. and Olsén, K.H. (1993) A study of the olfactory sensitivity of crucian carp (Carassius carassius) and goldfish (Carassius auratus) to 17α,20β-dihydroxy-4-pregnen-3-one and prostaglandin F2α. Chem. Senses, 18, Bjerselius, R., Lundstedt-Enkel, K., Olsén, H., Mayer, I. and Dimberg, K. (2001) Male goldfish reproductive behavior and physiology are severely affected by exogenous exposure to 17β-estradiol. Aquat. Toxicol., 53, Demski, L.S. and Dulka, J.G. (1984) Functional anatomical studies on

7 Reproductive Behavior in Crucian Carp 299 sperm release evoked by electrical stimulation of the olfactory tract in goldfish. Brain Res., 291, Demski, L.S. and Hornby, P.J. (1982) Hormonal-control of fish reproductive-behavior brain gonadal steroid interactions. Can. J. Fish. Aquat. Sci., 39, DeFraipont, M. and Sorensen, P.W. (1993) Exposure to the pheromone 17α,20β-dihydroxy-4-pregnen-3-one enhances the behavioral spawning success, sperm production, and sperm motility of male goldfish. Anim. Behav., 46, Døving, K.B. (1986) Functional properties of the fish olfactory system. Prog. Sensory Physiol., 6, Døving, K.B. and Gemne, G. (1965) Electrophysiological and histological properties of the olfactory tract of the burbot (Lota lota L.). J. Neurophysiol., 28, Døving, K.B. and Selset, R. (1980) Behavior patterns in cod released by electrical stimulation of olfactory tract bundlets. Science, 207, Døving, K.B. and Stabell, O.B. (2003) Trails in open waters: sensory cues in salmon migration. In Collin, S.P. and Marshall, J. (eds), Sensory Processing in the Aquatic Environment. Springer, New York, pp Dubois-Dauphin, M., Døving, K.B. and Holley, A. (1980) Topographical relation between the olfactory-bulb and the olfactory tract in tench (Tinca tinca L.). Chem. Senses, 5, Dulka, J.G., Stacey, N.E., Sorensen, P.W. and Van der Kraak, G.J. (1987) A steroid sex-pheromone synchronizes male female spawning readiness in goldfish. Nature, 325, Erickson, R.P. (1963) Sensory neural patterns and gustation. In Zottermans, Y. (ed.), Olfaction and Taste. Pergamon Press, Oxford, pp Finger, T.E. (1975) The distribution of the olfactory tracts in the bullhead catfish, Ictalurus nebulosus. J. Comp. Neurol., 161, Hamdani, E.H., Stabell, O.B., Alexander, G. and Døving, K.B. (2000) Alarm reaction in the crucian carp is mediated by the medial bundle of the medial olfactory tract. Chem. Senses, 25, Hamdani, E.H., Kasumyan, A. and Døving, K.B. (2001) Is feeding behaviour in crucian carp mediated by the lateral olfactory tract? Chem. Senses, 26, Koyama Y., Satou, M., Oka, Y. and Ueda, K. (1984) Involvement of the telencephalic hemisperes and the preoptic area in the sexual behavior of the male goldfish, Carassius auratus: a brain-lesion study. Behav. Neural Biol., 40, Kyle, A.L. and Peter, R.E. (1982) Effects of forebrain lesions on spawning behavior in the male goldfish. Physiol. Behav., 28, Kyle, A.L., Sorensen, P.W., Stacey, N.E. and Dulka, J.G. (1987) Medial olfactory tract pathways controlling sexual behavior in teleosts. In Demski, L.S. and Schwanzel-Fukuda, M. (eds), The Terminal Nerve (Nervus Terminalis): Structure, Function and Evolution. Annals of the New York Academy of Sciences, New York, pp Lee, C.T. and Ingersoll, D.W. (1979) Social chemosignals in five Belontiidae (Pisces) species. J. Comp. Physiol., 93, Liley, N.R. (1982) Chemical communication in fish. Can. J. Fish. Aquat. Sci., 39, Oka, Y., Ichikawa, M. and Ueda, K. (1982) Synaptic organization of the olfactory bulb and central projection of the olfactory tract. In Hara, T.J. (ed.), Chemoreception in Fishes. Elsevier, Amsterdam, pp Partridge, B.L., Liley, N.R. and Stacey, N.E. (1976) The role of pheromones in the sexual behaviour of the goldfish. Anim. Behav., 24, Poling, K.R., Fraser, E.J. and Sorensen, P.W. (2001) The three steroidal components of the goldfish preovulatory pheromone signal evoke different behaviors in males. Comp. Biochem. Physiol., 129B, Pollack, E.I., Becker, L.R. and Haynes, K. (1978) Sensory control of mating in the blue gourami, Trichogaster trichopterus (Pisces, Belontiidae). Behav. Biol., 22, Rooney, D., Døving, K.B., Ravaille-Veron, M. and Szabo, T. (1992) The central connections of the olfactory bulbs in cod, Gadus morhua L. J. Hirnforsch., 33, Satou, M. (1990) Synaptic organization, local neuronal circuitry, and functional segregation of the teleost olfactory bulb. Prog. Neurobiol., 34, Satou, M., Fujita, I., Ichikawa, M., Yamaguchi, K. and Ueda, K. (1983) Field potential and intracellular potential studies of the olfactory bulb in the carp: evidence for a functional separation of the olfactory bulb into lateral and medial subdivisions. J. Comp. Physiol., 152, Scott, A.P. and Sorensen, P.W. (1994) Time course of release of pheromonally active gonadal steroids and their conjugates by ovulatory goldfish. Gen. Comp. Endocrinol., 96, Sheldon, R.E. (1912) The olfactory tracts and centers in teleosts. J. Comp. Neurol., 22, Sorensen, P.W., Hara, T.J., Stacey, N.E. and Goetz, F.W. (1988) F prostaglandins function as potent olfactory stimulants that comprise the postovulatory female sex pheromone in goldfish. Biol. Reprod., 39, Sorensen, P.W., Chamberlain, K.J. and Stacey, N.E. (1989) Differing behavioral and endocrinological effects of two female sex pheromones on male goldfish. Horm. Behav., 23, Sorensen, P.W., Hara, T.J. and Stacey, N.E. (1991) Sex pheromones selectively stimulate the medial olfactory tracts of male goldfish. Brain Res., 558, Sorensen, P.W., Scott, A.P., Stacey, N.E. and Bowdin, L. (1995) Sulfated 17,20β-dihydroxy-4-pregnen-3-one functions as a potent and specific olfactory stimulant with pheromonal actions in the goldfish. Gen. Comp. Endocrinol., 100, Sorensen, P.W., Brash, A.R., Goetz, F.W., Kellner, R., Bowdin, L. and Vreize, L. (1996) Origins and functions of F prostaglandins as hormones and pheromones in the goldfish. In Goetz, F.W. and Thomas, P. (eds), Proceedings of the Fifth International Symposium on Reproductive Physiology of Fish, Austin, TX, pp Stacey, N.E. (1981) Hormonal regulation of female reproductive behavior in fish. Am. Zool., 21, Stacey, N.E. and Kyle, A.L. (1983) Effects of olfactory tract lesions on sexual and feeding behavior in the goldfish. Physiol. Behav., 30, Stacey, N.E. and Sorensen, P.W. (1986) 17α,20β-Dihydroxy-4-pregnen-3- one: a steroidal primer pheromone which increases milt volume in goldfish. Can. J. Zool., 64, Stacey, N.E., Cardwell, J.R., Liley, N.R., Scott, A.P. and Sorensen, P.W. (1994) Hormones as sex pheromones in fish. In Davey, K.G., Peter, R.E. and Tobe, S. (eds), Perspectives in Comparative Endocrinology. National Research Council of Canada, Ottawa, Canada, pp von Bartheld, C.S., Meyer, D.L., Fiebig, Z. and Ebbesson, S.O.E. (1984) Central connections of the olfactory bulb in the goldfish, Carassius auratus. Cell Tissue Res., 238, von Rekowski, C. and Zippel, H.P. (1993) In goldfish the qualitative discriminative ability for odors rapidly returns after bilateral nerve

8 300 F.-A. Weltzien et al. axotomy and lateral olfactory tract transection. Brain Res., 618, Zheng, W. and Stacey, N.E. (1996) Two mechanisms for increasing milt volume in male goldfish. J. Exp. Zool., 276, Zheng, W. and Stacey, N.E. (1997) A steroidal pheromone and spawning stimuli act via different neuroendocrine mechanisms to increase gonadotropin and milt volume in male goldfish (Carassius auratus). Gen. Comp. Endocrinol., 105, Zippel, H.P., Hofmann, M., Meyer, D.L. and Zeman, S. (1993) Functional and morphological regeneration of olfactory tracts and subtracts in goldfish. J. Comp. Physiol., 172A, Zou, Z., Horowitz, L.F., Montmayeur, J.P., Snapper, S. and Buck, L.B. (2001) Genetic tracing reveals a stereotyped sensory map in the olfactory cortex. Nature, 414, Accepted March 20, 2003