- Email: [email protected]
Social status and scent-marking behaviour in Lemur catta
Anim. Behav., 1990, 40, 774-788
Short Communications Social Status and Scent-marking Behaviour in Lemur catta
Among prosimian primates, ringtailed lemurs, Lemur eatta, stand out for the complexity of their olfactory communication system (Epple 1986). They also form large social groups and exhibit marked dominance relations (Jolly 1966), which are known to affect the frequency of scent-marking behaviour in many mammals (Rails 1971). The use of scents in the establishment of dominance relations is suggested by the central role of tailwaving displays during 'stink fights' between adult males (Jolly 1966), but detailed~tudies of olfactory and agonistic behaviour within intact groups are lacking. I investigated the relationship between scent deposition and social status in a group of 31 L. catta at the Duke University Primate Center between September 1987 and April 1988. The group inhabited a 3.4-ha natural habitat enclosure and consisted of 19 adults (born before 1986), seven subadults (born in 1986) and five infants. Each adult and subadult member served as focal animal once a week. Focal samples were 10 min in length and were conducted on a randomized schedule. A total of 21 such group samples resulted in 91 h of observation. I recorded all occurrences of scent marking of focal animals. Both sexes marked anogenitally. Antebrachial marking, tail-anointing and tail-waving displays were performed only by males. The two latter behaviour patterns were scored together as tail-marking. I determined the dominance hierarchy on the basis of decided agonistic interactions (Kappeler, in press). Since females are dominant over males, this was done for each sex separately. Because dominance in L. catta is dependent on sex and age, I also examined differences in marking behaviour between age and sex classes. I compared total marking frequencies because all animals were observed for the same amount of time. The dominance hierarchy among adult females was significantly linear (Landau's index=0-91; Martin & Bateson 1986), but the male hierachy was characterized by many (N=31) reversals among middle-ranking individuals, and therefore was not perfectly linear (Landau's index = 0.58). However, for the purpose of the present study, it seemed justified to describe the social status of males in this manner because of clear-cut differences between high- and low- ranking animals. There was a significant positive correlation 0003-3472/90/100774 + 15 $03.00/0
between male relative rank and the frequency of all three kinds of scent marking (anogenital: r s = 0.74, P<0-01; antebrachial: rs=0-81, P<0-001; tailmarking: r s = 0-48, P < 0"05), but their was no correlation (r s =0"02, Ms) between a female's rank and her frequency of anogenital marking (Table I). The anogenital marking frequencies of adult and subadult females did not differ significantly (adult: X_+ SE= 4"2 + 0"7 (individual/h); subadult: X_+ SE= 5-3+2-6; Wx=7, NS; Wilcoxon-Mann-Whitney test), but adult females performed significantly more anogenital marking than adult males (females: X_+sE=4.2+0.7; males: .~+SE=0"7+ 0'2; z = 3'24, P=0'0014). However, when all kinds of male marking were considered, adult male and female marking frequencies were not significantly different (males: -~+ SE 7.2 + 1.9; females: X + SE= 4-2 + 0'7; z = 0"96, NS). Similarly, the total marking frequencies of subadult males and females were also not significantly different (males: -~+ SE= 4"3+ 1"0; females: X + S E = 5 ' 3 + 2 ' 6 ; Wx= 14, ys). There was also no difference in their anogenital marking frequencies alone (males: _~+SE=0'4+ 0-3; females: _~+SE=5'3+2'6; Wx=16-5, NS). Adult males and subadult males did not differ significantly in their anogenital (adult: .~+SE= 0-7 + 0"2; subadult: ,~+SE=0-4+0-3; z=0"45, NS) or antebrachial (adult: X _ S E = 4 ' 9 + l ' 3 ; subadult: .~+SE=3'6+0'7; z=0"28, NS) marking frequencies, but adult males engaged significantly more often in tail-marking (adult: X _ S E = I ' 5 + 0 ' 4 ; subadult: .~+ SE= 0'2 + 0' 1; z = 2'94, P = 0"003). One suggested function of increased scent marking of dominant mammals is to suppress reproductive activity in subordinate group members (Epple 1986). In group-living prosimians, such indirect reproductive competition among females in absent (Kappeler 1989). The lack of a positive correlation between female rank and marking frequency is congruent with this observation. Scents of female lemurs appear to contain information about their reproductive state (Kappeler 1988), but the function of scent marks deposited outside the brief annual mating season has not been investigated. Epple (1986) suggested that the ubiquity of scents deposited by dominant animals can serve as a permanent threat signal and help to control the behaviour of subordinates, resulting in a stable dominance hierarchy and the avoidance of physical combat. The high frequency of agonistic interactions among male L. eatta and the relative instability of their hierarchy argue against the hypothesis. At least two alternative explanations of =
9 1990 The Association for the Study of Animal Behaviour 774
Short Communications
775
Table I. The identity, sex age, and dominance rank of all focal animals, along with their
frequencies of anogenital, antebrachial and tail-marking in 21 10-min focal samples Frequency Identity
Sex
Age
Rank
Anogenital
Antebrachial
Tail-marking
Lys Bau Ane Cor Bar The Adm Art Phy Hie Hes Dio Iro Rom Tha Phi Pra Ana Leo Nic Cer Mic Cal Dat Isa Mil
Female Female Female Female Female Female Female Female Female Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male
Adult Adult Adult Adult Adult Adult Subadult Subadult Subadult Adult Adult Adult Adult Adult Adult Adult Adult Adult Adult Adult Adult Adult Subadult Subadult Subadult Subadult
1 2 3 4 5 6 >6 >6 >6 1 2 3 4 5 6 7 8 9 10 11 12 13 > 13 > 13 > 13 > 13
l0 12 25 15 19 8 22 1 33 12 2 6 5 0 1 3 1 4 0 0 0 0 5 1 0 0
---
---
--
--
---66 24 29 22 9 11 5 17 17 9 5 8 5 16 13 5 17
--22 2 6 7 5 5 1 3 5 4 5 5 0 0 2 0 2
The ranks of subadults of both sexes were not determined, because they did not engage in enough agonistic interactions among themselves. the positive correlation between male social status a n d m a r k i n g b e h a v i o u r seem possible: (1) increased m a r k i n g frequencies o f h i g h - r a n k i n g males m a y simply be a n o n - f u n c t i o n a l correlate of a general increase in social activity associated with high rank, a n d (2) male scent m a r k s m a y serve as a m e c h a n i s m in indirect intra-sexual competition. These h y p o t h eses are n o t mutually exclusive, but indirect evidence favours the latter one. First, scents o f male lemurs m a y be primarily deposited for the i n f o r m a t i o n o f other males. Female L. catta responded to only 1-2% ( N = 4 1 5 ) of male m a r k s with active investigation a n d / o r o v e r m a r k i n g within 30 s, whereas o t h e r males r e s p o n d e d to m o r e t h a n 6 % o f t h e m (Kappeler, u n p u b l i s h e d data). A similar sex difference h a s been d o c u m e n t e d in L. coronatus ( K a p p e l e r 1988). Second, a physiological m e c h a n i s m o f m a l e male reproductive competition has been described in Mierocebus murinus. O d o u r s o f d o m i n a n t m o u s e lemur males are k n o w n to c o n t a i n volatile c o m p o nents t h a t elicit a decrease o f p l a s m a testosterone
levels a n d sexual activity in other males via hyperprolactinaemia (Schilling et al. 1984). Moreover, scent m a r k s have individual characteristics (Mertl 1975), a n d m a y serve as ' h o n e s t ' signals because o f their direct testosterone dependence. They may thus provide an efficient m e c h a n i s m for indirect reproductive competition. A l t h o u g h this is not a n obligatory conclusion from these observations, it certainly is a plausible working hypothesis. Investigations of qualitative differences in scents of males o f high and low rank, a n d d o c u m e n t e d changes of scent quality a c c o m p a n y i n g significant r a n k changes o f individuals are needed to illuminate this hypothesis in more detail. I t h a n k Peter Klopfer, Stephen Nowicki, Carel van Schaik a n d K a r e n Petras for c o m m e n t s a n d discussion. This is D u k e University P r i m a t e Center publication No. 485. PETER M. KAPPELER
Department o f Zoology, Duke University, Durham, NC 27706, U.S.A.
Animal Behaviour, 40, 4
776 References
Epple, G. 1986. Communication by chemical signals. In: Comparative Primate Biology 2A (Ed. by G. Mitchell & J. Erwin), pp. 531 580. New York: A. R. Liss. Jolly, A. 1966. Lemur Behavior. Chicago: University of Chicago Press. Kappeler, P. M. 1988. A preliminary study of olfactory behavior of captive Lemur coronatus during the breeding season. Int. J. Primatol., 9, 135-146. Kappeler, P. M. 1989. Agonistic and grooming behavior of captive crowned lemurs (Lemur coronatus) during the breeding season. Hum. Evol., 4, 207-215. Kappeler, P. M. In press. Female dominance in Lemur catta: more than femalefeedingpriority? Foliaprimatol. Martin, P. & Bateson, P. 1986. Measuring Behaviour. Cambridge: Cambridge University Press. Mertl, A. S. 1975. Discrimination of individuals by scent in a primate. Behav. Biol., 14, 505-509. Rails, K. 1971. Mammalian scent marking. Seienee~ 171, 443-449. Schilling, A., Perret, M. & Predine, J. 1984. Sexual inhibition in a prosimian primate: a pheromone-like effect. J. Endocrinol., 102, 143-151. (Received 10 November 1989; initial acceptance 8 February 1990;final acceptance 14 May 1990; MS. number." As-680)
Brood Parasitism and Egg Robbing among Three Freshwater Fish Brood parasitism, where eggs are laid in the nest of another species and the young reared by the host parent, is well documented among birds. It can be destructive, where all the host brood are killed (e.g. cuckoos) or mild, where the parasites and host brood are reared together (e.g. cowbirds) (Lack 1968; Payne 1977). Both types of brood parasitism also occur among fish (Kramer & Smith 1960; Hunter & Hasler 1965; McKaye & Oliver 1980; Steele & Pearson 1981; Sato 1986) where examples of the mild form may be mutually adaptive in that the host protects the parasite's eggs which in turn decrease the predation pressure on the host's eggs (McKaye 1981). We present the first quantitative study of brood parasitism in fish documented in Asia, consisting of a system where the disadvantage to the host species results from egg predation by a third non-parasitic species mingling with the parasite. The host species, the freshwater perch Siniperca kawamebari (Serranidae), occurs in the southern part of Honshu, Kyushu and Shikoku Islands of Japan, and at the southern end of the Korean Peninsula. During the breeding season males establish territories of less than 1 m in diameter and aggressively repel intruders of its own and other
species. Within his territory each male cleans the surface of several reed stems for use as spawning sites. We here term all such spawning sites within a territory a 'nest'. Females visit the male's territory only to oviposit, and the male thereafter aggressively defends the eggs and fry against all intruders (Imai & Nakahara 1957). From early May to mid-August 1989 we observed 116 nests of 105 males every other day in a 150 m section of a branch of the Ibo River in Hyogo Prefecture, Japan. At the start of our observations 15 nests owned by 15 different males and containing in all 5323 eggs were already present. The number of eggs (Xq'-SD diameter=2.5_+ 0-7 mm, N = 30), usually laid in two or, rarely, three lines on several stems (X_+SD= 2'3 _+1'6 stems per nest, N = 116 nests) in each nest, varied greatly between clutches (_Y+SD= 79 _+26 eggs per clutch, N = 676 clutches), as did the total number of eggs contained in each nest (.Y_+SD= 576 _+424 eggs, N = 116 nests). Up to seven clutches were deposited on the same stem ( . ~ + s n = 1.8_+ 1.2 clutches per stem, N=449 stems), suggesting oviposition by several females within the same nest. Oviposition continued until the beginning of August. The parasitic species, the minnow, Pungtungia herzi (Cyprinidae), began to deposit their eggs in the nests of S. kawamebari from mid-May. The spherical eggs of P. herzi were considerably smaller (X_+SD=2.1__0.6mm, N=30) than the host's eggs, and were very adhesive and were strongly attached to the nest stem, unlike those of S. kawamebari which were easily dislodged. Deposition of eggs increased explosively from the beginning of June and continued up to the end of July. Pungtungia herzi spawned in a large group. The host male could drive one or two intruders from his territory, but as the group size of the parasites increased, he was overwhelmed by their numbers. Of 116 nests of S. kawamebari examined, 76 (65.5%) were parasitized by P. herzi, indicating that brood parasitism was frequent in this population. Eggs of both species were usually adjacent and rarely overlapped. Although we searched intensively for eggs of P. herzi deposited outside a nest of S. kawamebari within an area of 4000 m 2 in the same study area, we found only one P. herzi with its eggs under a large rock, but were unable to clarify whether or not the fish actually guarded those eggs. About 1100 eggs were present under the rock, whereas we found 2-1298 eggs of P. herzi (X'+SD= 192___294 eggs, N = 76) per nest of S. kawamebari, indicating that the parasitic mode of reproduction is important for this species. To analyse the timing of egg laying ofP. herzi, we divided the developmental phase of S. kawamebari
Short Communications Social Status and Scent-marking Behaviour in Lemur catta
Among prosimian primates, ringtailed lemurs, Lemur eatta, stand out for the complexity of their olfactory communication system (Epple 1986). They also form large social groups and exhibit marked dominance relations (Jolly 1966), which are known to affect the frequency of scent-marking behaviour in many mammals (Rails 1971). The use of scents in the establishment of dominance relations is suggested by the central role of tailwaving displays during 'stink fights' between adult males (Jolly 1966), but detailed~tudies of olfactory and agonistic behaviour within intact groups are lacking. I investigated the relationship between scent deposition and social status in a group of 31 L. catta at the Duke University Primate Center between September 1987 and April 1988. The group inhabited a 3.4-ha natural habitat enclosure and consisted of 19 adults (born before 1986), seven subadults (born in 1986) and five infants. Each adult and subadult member served as focal animal once a week. Focal samples were 10 min in length and were conducted on a randomized schedule. A total of 21 such group samples resulted in 91 h of observation. I recorded all occurrences of scent marking of focal animals. Both sexes marked anogenitally. Antebrachial marking, tail-anointing and tail-waving displays were performed only by males. The two latter behaviour patterns were scored together as tail-marking. I determined the dominance hierarchy on the basis of decided agonistic interactions (Kappeler, in press). Since females are dominant over males, this was done for each sex separately. Because dominance in L. catta is dependent on sex and age, I also examined differences in marking behaviour between age and sex classes. I compared total marking frequencies because all animals were observed for the same amount of time. The dominance hierarchy among adult females was significantly linear (Landau's index=0-91; Martin & Bateson 1986), but the male hierachy was characterized by many (N=31) reversals among middle-ranking individuals, and therefore was not perfectly linear (Landau's index = 0.58). However, for the purpose of the present study, it seemed justified to describe the social status of males in this manner because of clear-cut differences between high- and low- ranking animals. There was a significant positive correlation 0003-3472/90/100774 + 15 $03.00/0
between male relative rank and the frequency of all three kinds of scent marking (anogenital: r s = 0.74, P<0-01; antebrachial: rs=0-81, P<0-001; tailmarking: r s = 0-48, P < 0"05), but their was no correlation (r s =0"02, Ms) between a female's rank and her frequency of anogenital marking (Table I). The anogenital marking frequencies of adult and subadult females did not differ significantly (adult: X_+ SE= 4"2 + 0"7 (individual/h); subadult: X_+ SE= 5-3+2-6; Wx=7, NS; Wilcoxon-Mann-Whitney test), but adult females performed significantly more anogenital marking than adult males (females: X_+sE=4.2+0.7; males: .~+SE=0"7+ 0'2; z = 3'24, P=0'0014). However, when all kinds of male marking were considered, adult male and female marking frequencies were not significantly different (males: -~+ SE 7.2 + 1.9; females: X + SE= 4-2 + 0'7; z = 0"96, NS). Similarly, the total marking frequencies of subadult males and females were also not significantly different (males: -~+ SE= 4"3+ 1"0; females: X + S E = 5 ' 3 + 2 ' 6 ; Wx= 14, ys). There was also no difference in their anogenital marking frequencies alone (males: _~+SE=0'4+ 0-3; females: _~+SE=5'3+2'6; Wx=16-5, NS). Adult males and subadult males did not differ significantly in their anogenital (adult: .~+SE= 0-7 + 0"2; subadult: ,~+SE=0-4+0-3; z=0"45, NS) or antebrachial (adult: X _ S E = 4 ' 9 + l ' 3 ; subadult: .~+SE=3'6+0'7; z=0"28, NS) marking frequencies, but adult males engaged significantly more often in tail-marking (adult: X _ S E = I ' 5 + 0 ' 4 ; subadult: .~+ SE= 0'2 + 0' 1; z = 2'94, P = 0"003). One suggested function of increased scent marking of dominant mammals is to suppress reproductive activity in subordinate group members (Epple 1986). In group-living prosimians, such indirect reproductive competition among females in absent (Kappeler 1989). The lack of a positive correlation between female rank and marking frequency is congruent with this observation. Scents of female lemurs appear to contain information about their reproductive state (Kappeler 1988), but the function of scent marks deposited outside the brief annual mating season has not been investigated. Epple (1986) suggested that the ubiquity of scents deposited by dominant animals can serve as a permanent threat signal and help to control the behaviour of subordinates, resulting in a stable dominance hierarchy and the avoidance of physical combat. The high frequency of agonistic interactions among male L. eatta and the relative instability of their hierarchy argue against the hypothesis. At least two alternative explanations of =
9 1990 The Association for the Study of Animal Behaviour 774
Short Communications
775
Table I. The identity, sex age, and dominance rank of all focal animals, along with their
frequencies of anogenital, antebrachial and tail-marking in 21 10-min focal samples Frequency Identity
Sex
Age
Rank
Anogenital
Antebrachial
Tail-marking
Lys Bau Ane Cor Bar The Adm Art Phy Hie Hes Dio Iro Rom Tha Phi Pra Ana Leo Nic Cer Mic Cal Dat Isa Mil
Female Female Female Female Female Female Female Female Female Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male
Adult Adult Adult Adult Adult Adult Subadult Subadult Subadult Adult Adult Adult Adult Adult Adult Adult Adult Adult Adult Adult Adult Adult Subadult Subadult Subadult Subadult
1 2 3 4 5 6 >6 >6 >6 1 2 3 4 5 6 7 8 9 10 11 12 13 > 13 > 13 > 13 > 13
l0 12 25 15 19 8 22 1 33 12 2 6 5 0 1 3 1 4 0 0 0 0 5 1 0 0
---
---
--
--
---66 24 29 22 9 11 5 17 17 9 5 8 5 16 13 5 17
--22 2 6 7 5 5 1 3 5 4 5 5 0 0 2 0 2
The ranks of subadults of both sexes were not determined, because they did not engage in enough agonistic interactions among themselves. the positive correlation between male social status a n d m a r k i n g b e h a v i o u r seem possible: (1) increased m a r k i n g frequencies o f h i g h - r a n k i n g males m a y simply be a n o n - f u n c t i o n a l correlate of a general increase in social activity associated with high rank, a n d (2) male scent m a r k s m a y serve as a m e c h a n i s m in indirect intra-sexual competition. These h y p o t h eses are n o t mutually exclusive, but indirect evidence favours the latter one. First, scents o f male lemurs m a y be primarily deposited for the i n f o r m a t i o n o f other males. Female L. catta responded to only 1-2% ( N = 4 1 5 ) of male m a r k s with active investigation a n d / o r o v e r m a r k i n g within 30 s, whereas o t h e r males r e s p o n d e d to m o r e t h a n 6 % o f t h e m (Kappeler, u n p u b l i s h e d data). A similar sex difference h a s been d o c u m e n t e d in L. coronatus ( K a p p e l e r 1988). Second, a physiological m e c h a n i s m o f m a l e male reproductive competition has been described in Mierocebus murinus. O d o u r s o f d o m i n a n t m o u s e lemur males are k n o w n to c o n t a i n volatile c o m p o nents t h a t elicit a decrease o f p l a s m a testosterone
levels a n d sexual activity in other males via hyperprolactinaemia (Schilling et al. 1984). Moreover, scent m a r k s have individual characteristics (Mertl 1975), a n d m a y serve as ' h o n e s t ' signals because o f their direct testosterone dependence. They may thus provide an efficient m e c h a n i s m for indirect reproductive competition. A l t h o u g h this is not a n obligatory conclusion from these observations, it certainly is a plausible working hypothesis. Investigations of qualitative differences in scents of males o f high and low rank, a n d d o c u m e n t e d changes of scent quality a c c o m p a n y i n g significant r a n k changes o f individuals are needed to illuminate this hypothesis in more detail. I t h a n k Peter Klopfer, Stephen Nowicki, Carel van Schaik a n d K a r e n Petras for c o m m e n t s a n d discussion. This is D u k e University P r i m a t e Center publication No. 485. PETER M. KAPPELER
Department o f Zoology, Duke University, Durham, NC 27706, U.S.A.
Animal Behaviour, 40, 4
776 References
Epple, G. 1986. Communication by chemical signals. In: Comparative Primate Biology 2A (Ed. by G. Mitchell & J. Erwin), pp. 531 580. New York: A. R. Liss. Jolly, A. 1966. Lemur Behavior. Chicago: University of Chicago Press. Kappeler, P. M. 1988. A preliminary study of olfactory behavior of captive Lemur coronatus during the breeding season. Int. J. Primatol., 9, 135-146. Kappeler, P. M. 1989. Agonistic and grooming behavior of captive crowned lemurs (Lemur coronatus) during the breeding season. Hum. Evol., 4, 207-215. Kappeler, P. M. In press. Female dominance in Lemur catta: more than femalefeedingpriority? Foliaprimatol. Martin, P. & Bateson, P. 1986. Measuring Behaviour. Cambridge: Cambridge University Press. Mertl, A. S. 1975. Discrimination of individuals by scent in a primate. Behav. Biol., 14, 505-509. Rails, K. 1971. Mammalian scent marking. Seienee~ 171, 443-449. Schilling, A., Perret, M. & Predine, J. 1984. Sexual inhibition in a prosimian primate: a pheromone-like effect. J. Endocrinol., 102, 143-151. (Received 10 November 1989; initial acceptance 8 February 1990;final acceptance 14 May 1990; MS. number." As-680)
Brood Parasitism and Egg Robbing among Three Freshwater Fish Brood parasitism, where eggs are laid in the nest of another species and the young reared by the host parent, is well documented among birds. It can be destructive, where all the host brood are killed (e.g. cuckoos) or mild, where the parasites and host brood are reared together (e.g. cowbirds) (Lack 1968; Payne 1977). Both types of brood parasitism also occur among fish (Kramer & Smith 1960; Hunter & Hasler 1965; McKaye & Oliver 1980; Steele & Pearson 1981; Sato 1986) where examples of the mild form may be mutually adaptive in that the host protects the parasite's eggs which in turn decrease the predation pressure on the host's eggs (McKaye 1981). We present the first quantitative study of brood parasitism in fish documented in Asia, consisting of a system where the disadvantage to the host species results from egg predation by a third non-parasitic species mingling with the parasite. The host species, the freshwater perch Siniperca kawamebari (Serranidae), occurs in the southern part of Honshu, Kyushu and Shikoku Islands of Japan, and at the southern end of the Korean Peninsula. During the breeding season males establish territories of less than 1 m in diameter and aggressively repel intruders of its own and other
species. Within his territory each male cleans the surface of several reed stems for use as spawning sites. We here term all such spawning sites within a territory a 'nest'. Females visit the male's territory only to oviposit, and the male thereafter aggressively defends the eggs and fry against all intruders (Imai & Nakahara 1957). From early May to mid-August 1989 we observed 116 nests of 105 males every other day in a 150 m section of a branch of the Ibo River in Hyogo Prefecture, Japan. At the start of our observations 15 nests owned by 15 different males and containing in all 5323 eggs were already present. The number of eggs (Xq'-SD diameter=2.5_+ 0-7 mm, N = 30), usually laid in two or, rarely, three lines on several stems (X_+SD= 2'3 _+1'6 stems per nest, N = 116 nests) in each nest, varied greatly between clutches (_Y+SD= 79 _+26 eggs per clutch, N = 676 clutches), as did the total number of eggs contained in each nest (.Y_+SD= 576 _+424 eggs, N = 116 nests). Up to seven clutches were deposited on the same stem ( . ~ + s n = 1.8_+ 1.2 clutches per stem, N=449 stems), suggesting oviposition by several females within the same nest. Oviposition continued until the beginning of August. The parasitic species, the minnow, Pungtungia herzi (Cyprinidae), began to deposit their eggs in the nests of S. kawamebari from mid-May. The spherical eggs of P. herzi were considerably smaller (X_+SD=2.1__0.6mm, N=30) than the host's eggs, and were very adhesive and were strongly attached to the nest stem, unlike those of S. kawamebari which were easily dislodged. Deposition of eggs increased explosively from the beginning of June and continued up to the end of July. Pungtungia herzi spawned in a large group. The host male could drive one or two intruders from his territory, but as the group size of the parasites increased, he was overwhelmed by their numbers. Of 116 nests of S. kawamebari examined, 76 (65.5%) were parasitized by P. herzi, indicating that brood parasitism was frequent in this population. Eggs of both species were usually adjacent and rarely overlapped. Although we searched intensively for eggs of P. herzi deposited outside a nest of S. kawamebari within an area of 4000 m 2 in the same study area, we found only one P. herzi with its eggs under a large rock, but were unable to clarify whether or not the fish actually guarded those eggs. About 1100 eggs were present under the rock, whereas we found 2-1298 eggs of P. herzi (X'+SD= 192___294 eggs, N = 76) per nest of S. kawamebari, indicating that the parasitic mode of reproduction is important for this species. To analyse the timing of egg laying ofP. herzi, we divided the developmental phase of S. kawamebari