Cod courtship song: a song at the expense of dance?

Color profile: Disabled Composite Default screen

542

Cod courtship song: a song at the expense of dance? Frode Engen and Ivar Folstad

Abstract: Sexually selected characters may reveal information about individual quality during mate choice. Fin display and sound emitted with the aid of specific drumming muscles are characters described as being of importance in the reproductive behaviour of cod (Gadus morhua L.). We examined whether the mass of drumming muscles or fin size was sexually dimorphic, and whether these characters could provide information about male cod that was potentially of benefit to mate-seeking females. The mass of drumming muscles, but not fin size, was sexually dimorphic, with males having larger muscles than females. Neither the mass of drumming muscles nor fin size apparently revealed information about traits that may be associated with parasite resistance in males (i.e., parasite intensities and leukocyte densities). However, variation in fertilization potential (i.e., spermatocrit level) among males was related to both mass of drumming muscles and fin size. Thus, by evaluating sound and fin size, mate-seeking females may obtain information about fertilization ability among males. This may be of particular importance for females in a species whose eggs commonly remain unfertilized. Furthermore, males with large drumming muscles and small fins had low spermatocrit levels. This may reflect reductions in sperm density resulting from frequent ejaculations by attractive males. A costly allocation of resources for the development of drumming muscles at the expense of fin muscles used for propulsion is presented as a tentative explanation as to why females should pay attention to these particular traits during courtship. Increased investment in “song” may thus appear at the expense of “dance.” Résumé : Les caractères soumis à la sélection sexuelle peuvent être révélateurs de la qualité d’un individu lors du choix d’un partenaire. L’agitation des nageoires et les sons, émis grâce à l’action de muscles tambourineurs spécifiques, sont des caractères reconnus d’importance dans le comportement reproducteur de la Morue franche (Gadus morhua L.). Nous avons cherché à établir si la masse des muscles tambourineurs ou la taille des nageoires sont des caractères sexuels dimorphiques et si ces caractères peuvent témoigner de la supériorité de certains mâles aux femelles en quête d’un partenaire. La masse des muscles tambourineurs, mais pas la taille des nageoires, est un caractère dimorphique et ces muscles sont plus gros chez les mâles que chez les femelles. Ni la masse des muscles, ni la taille des nageoires ne semblent transmettre d’information au sujet de traits de caractère qui pourraient être associés à la résistance aux parasites chez les mâles (i.e., la gravité des infections ou la densité des leucocytes). Cependant, la variation du potentiel reproducteur (i.e., la valeur du spermatocrite) des mâles est reliée à la fois à la masse des muscles tambourineurs et à la taille des nageoires. En jaugeant les sons et la taille des nageoires des mâles, les femelles en quête d’un partenaire peuvent juger de leur capacité de fécondation. Cela peut avoir une importance particulière chez les femelles d’une espèce dont les oeufs restent souvent non fécondés. De plus, les mâles qui ont de gros muscles tambourineurs et de petites nageoires ont un spermatocrite faible, peut-être par réduction de la densité des spermatozoïdes à la suite d’éjaculations fréquentes chez les mâles attirants. L’allocation coûteuse des ressources pour produire des muscles tambourineurs forts au détriment des muscles qui servent à la propulsion des nageoires pourrait expliquer pourquoi les femelles devraient accorder attention à ces caractéristiques durant la période de cour. Il semble donc que cet investissement plus grand dans le « chant » se fasse au détriment de la « danse ». [Traduit par la Rédaction]

Engen and Folstad

Introduction Females generally produce fewer and energetically richer gametes than males. Owing to the greater investment by females than males in each zygote, females are believed to experience the greatest loss during an unsuccessful reproductive event (Andersson 1994), and they are also expected to be the more choosy sex in selecting a mate (CluttonReceived June 30, 1998. Accepted December 4, 1998. F. Engen and I. Folstad.1 Department of Ecology, Institute for Biology, University of Tromsø, 9037 Tromsø, Norway. 1

Author to whom all correspondence should be addressed (e-mail: [email protected]).

Can. J. Zool. 77: 542–550 (1999)

J:\cjz\cjz77\cjz-04\Z99-010.vp Thursday, September 09, 1999 4:42:27 PM

550

Brock and Vincent 1991). To discriminate among males during mate choice, females may obtain information by evaluating sexually selected characters that act as signals in communication between the sexes. Signals, like most other characters, are affected by trade-offs between costs and benefits. However, signals differ from other characters in that their benefits, i.e., positive changes in fitness, are bestowed by the recipient of the signal (Hasson 1997). The reliability of signals may depend on their costs (Zahavi 1975, 1977), their design, or conventions (see Hasson 1997). Both direct and indirect benefits may be derived from female mate choice. Direct benefits can be material benefits such as food, territory quality, male parental care, or highquality ejaculates that assure fertilization of eggs (Andersson 1994). Information about the fertilization potential of males © 1999 NRC Canada


Engen and Folstad

543

Fig. 1. Courtship behaviour of cod, showing the sequence of fin display and sound production from the moment a female enters a male’s territory (1) to the actual act of spawning (13). This drawing of the dance is modified from Brawn (1961a).

may, for example, be revealed through sexually selected traits that can indicate gonadal development and sperm numbers (Trivers 1972; Williams 1978; Williams 1992; Mjelstad 1991; Matthews et al. 1997, but see Gibson and Jewell 1982). Thus, by utilizing information derived from sexually selected traits, a female may assess a male’s ability to fertilize her eggs. Indirect benefits from female choice may be gained if females can obtain information about heritable qualities of mates. One such heritable quality may be parasite resistance, and females choosing the most parasite-resistant mates may benefit indirectly through producing parasite-resistant offspring (Hamilton and Zuk 1982). A mechanistic explanation as to why sexually selected characters signal parasite resistance has been suggested by Folstad and Karter (1992). Specific hormones stimulate the development of characters used in sexual selection, but also reduce individual immunocompetence. This dual effect of some hormones is the basis of “the immunocompetence-handicap hypothesis,” which suggests that hormone profiles, and consequently both immunosuppression and the expression of male secondary sexual characters, are self-regulated in response to parasite burden (Folstad and Karter 1992). Information that directly or indirectly benefits females may be provided by signaling between the sexes during courtship. Such signaling behaviour has been reported for several fish species, and it has been described from observations of cod (Gadus morhua L.) in aquaria by Templeman (1958) and Brawn (1961b). Brawn described an intense “flaunting display” in which the courting male swims alongside and in front of the female, making many circles and sharp turns, following a tortuous course with all the median fins fully erect (see Fig. 1). The display becomes most intense when the male pauses in front of the female and presents a lateral view of himself. According to Brawn (1961b), the “effec-

tiveness” of this display is increased when the male lowers the first or second dorsal fin together with the first ventral fin and then suddenly erects them again. This remarkable use of fins in courtship behaviour suggests that they may be of particular importance in courtship communication. Indeed, fin size has been reported to reveal information about individual quality (Brønseth and Folstad 1997), and female preference for males with large fins has been reported for several fish species (Bischoff et al. 1985; Basolo 1990; Macias Garcia et al. 1994). Furthermore, male cod also make grunting sounds which, according to Brawn (1961b), increase the excitement of the female during courtship. The grunting sound is produced by three pairs of muscles surrounding three lobes external to the swim-bladder wall (Brawn 1961a). These drumming muscles make repeated slight contractions that vibrate the swim-bladder wall and generate sounds (Jones and Marshall 1953). Only males produce sound during courtship and the loudness of the sound is positively associated with fish size (Brawn 1961c). Additionally, considerable metabolic demands are placed on the sonic muscles during sound production by vocalizing fish (Brantley 1992). Courtship sounds are reported to be of importance in fish reproduction (Tavolga 1958; Gerald 1971; Winn 1972; Stout 1975), and it has also been suggested that females can differentiate sounds from different males, using this “as one means of assessing the desirability of potential mates” (Myrberg et al. 1986). In pelagic spawners like cod, which release their gametes freely into the water, it is seldom that all eggs are fertilized (Bauer and Bauer 1981; Nakatsuru and Kramer 1982; Petersen et al. 1992; Levitan 1993). According to McKenzie (1940) a “great number of eggs” are not fertilized during natural reproduction in cod. Moreover, spermatocrit (i.e., the percentage of the total volume of milt composed by sperm cells) © 1999 NRC Canada



544

and the proportion of sperm cells exhibiting progressive forward motion together account for up to 14% of the total variability in fertilization success in cod (Trippel and Neilson 1992). Spermatocrit is commonly considered to be a reliable indicator of male fertilization potential (Daye and Glebe 1984; Munkittrick and Moccia 1987; Trippel and Neilson 1992), consequently the spermatocrit level should be important for fertilization in cod. Information about spermatocrit levels may thus be important in mate choice, representing a direct benefit to reproductively active female cod. In cod, females receive nothing but gametes when they reproduce: there is no parental care and males do not occupy a territory with resources available for the female. If sound emission and fin display function as signals during courtship in cod, these characters should, if they are not exclusively Fisherian traits (Fisher 1930), be related to individual quality. Therefore, we investigated whether the mass of the drumming muscles and the length of the median fins are sexually dimorphic in cod, and whether these characters could reveal information about immunological parameters, parasite intensities, or fertilization potential during mate choice.

Materials and methods The spawning period of the arcto-Norwegian tribe of cod lasts from February to May, with the main spawning occurring in March–April (Sars 1879; Rollefsen 1932; Wiborg 1957; Ellertsen et al. 1981), and peak spawning is thought to take place on 31 March (Pedersen 1984). The 38 male and 18 female cod included in this study were reproductively active and were caught on a spawning ground in Balsfjorden, northern Norway, during 5 h (6 a.m. to 11 a.m.) on 10 March 1994. The location of the spawning ground was described by Johnsen (1981). To prevent fins from being damaged and to keep the handling times of individual fish approximately equal we caught all fish by means of fishing lines. Immediately after a fish was caught, a blood sample was collected from the ductus cuvieri using a heparinized vacutainer in order to estimate the density of immunologically important blood cells. Hematocrit (the percentage of the total volume of blood composed of erythrocytes) was measured by centrifuging blood in capillary tubes for 3 min 20 s at 11 500 rpm with a Compur Mini Centrifuge (M1100). A blood smear was made, fixed, and later stained for identification of blood cells according to Rowley (1990). Lymphocytes, granulocytes, and erythrocytes were counted in three randomly selected areas of these blood smears under a microscope. Lymphocyte:erythrocyte and granulocyte:erythrocyte ratios were calculated using the mean values of these three counts. The repeatability of the mean values of three counts of the lymphocyte:erythrocyte and granulocyte:erythrocyte ratios were r2 = 0.86, p < 0.0001, and r2 = 0.84, p < 0.0002, respectively (n = 10 in both cases). The densities of circulating lymphocytes and granulocytes were estimated by multiplying these ratios by the hematocrit value (cf. Skarstein and Folstad 1996). After the blood sample was collected, the genital area of the fish was dried with an absorbent paper towel to prevent water from spilling into the milt sample, which was collected by stripping. Spermatocrit was estimated by centrifuging homogenized milt samples according to the procedure described above for hematocrit. This method gives a reliable estimate of the actual number of sperm in a milt sample (Bouck and Jacobsen 1976; Obraztsov 1985; Aas et al. 1991). The fish were then individually frozen in polyethylene bags. In the laboratory, the ungutted fish was weighed to the nearest gram and body length was measured to the nearest millimetre. The third fin ray of the two most anterior dorsal fins and the two ven-

Can. J. Zool. Vol. 77, 1999 tral fins, counting from the anterior end of the fins, was measured to the nearest millimetre. Our fin measurements are highly intercorrelated (see below). Fins were selected according to the ease with which they could be reliably measured and their presumed conspicuousness in an individual in a lateral posture (see Fig. 1). The three pairs of drumming muscles covering the second, third, and fourth lobes of the swim-bladder wall are attached on the pleural rib and the swim-bladder wall (Brawn 1961c). They are easily separated from the surrounding tissue and the combined dry mass of these three pairs of muscles was measured to the nearest 0.01 g (an accuracy representing 0.3 and 0.5% of the mean mass of drumming muscles of males and females, respectively). The age of each cod was determined by counting the hyaline zones of the otoliths. During postmortem examinations we tried to estimate the intensities of all macroparasites in each host; six groups of macroparasites were identified in our host sample: (i) the copepod Lernaeocera branchialis L. infecting the gills; (ii) metacercariae of the digenean Cryptocotyle lingua infecting the skin; (iii) the adult digeneans Hemiurus levinseni and Derogenes varicus infecting the stomach; (iv) the acanthocephalan Echinorhynchus gadi infecting the gastrointestinal tract; and nematodes of the family Anisakidae, which were divided into two groups that infect different organ systems: (v) those located on the liver and (vi) those infecting both the pyloric caeca and stomach. Because of low prevalence and intensity of D. varicus, the two adult digenean species were treated as one group. To estimate the intensity of C. lingua metacercariae, the skin (excluding that of the fins) was removed between the neck and operculum at the anterior end of the fish and the tail fin at the posterior end. The skin was then digested in a fluid containing 1% pepsin and 2% HCl at 38°C with continual agitation at 100 rpm for 30 h. The solution was then filtered through 50-µm plankton mesh before the metacercariae were counted using a stereomicroscope (cf. Lysne et al. 1994). All parasites were identified according to Odhner (1905), Möller and Anders (1986), and Hemmingsen and MacKenzie (1993). An index to the overall parasite burden of the fish was constructed (cf. Markusson and Folstad 1997). Based on the distribution of the original parasite intensities, infections with each of the six parasite groups (see above) were divided into quartiles for each sex separately. The fish were then given ordinal scores ranging from 1 (least infected) to 4, depending on the specific parasite intensity. Fish with no parasites belonging to one group were not included in the calculation of the quartiles and were consequently given a score of 0 for that parasite group. The six parasite scores for each fish were then summed, and this sum constitutes the individual fish’s parasite index value. A principal component analysis (PCA) was performed on the four measurements of fin size for each sex separately. This method reduces the original four measurements to one variable describing the length of the four median fins of each fish. For males and females, these new variables described respectively 91% (eigenvalue 3.6) and 87% (eigenvalue 3.5) of the variation in the original fin measurements. When investigating sexual dimorphism in fin size, one variable resulting from one PCA of the four fin measurements for both sexes was used (in contrast to elsewhere, where fin measurements for each sex were used separately). The variable used when examining sexual dimorphism in fin size described 90% (eigenvalue 3.6) of the variation in the original fin-size measurements. Between 87 and 98% of the variation in the original variables included in any PCA were explained by the new variable.

Statistical analyses Statistical analyses were done using Statsoft Statistica (Macintosh version 3.0b). Associations were tested using regression analyses, and differences between groups were assessed using unpaired t tests. Frequency distributions were tested using Kolmogorov–Smirnov one-sample tests. Only data on dry mass of drumming muscles © 1999 NRC Canada



Engen and Folstad

545

Table 1. Descriptive statistics for the parameters examined in the study of reproductively active cod. Males (n = 38) Age (years) Body mass (g) Hematocrit (%)a Granulocytes a,b Lymphocytesa,b Spermatocrit (%)c First dorsal fin (mm)d Second dorsal fin (mm)d First ventral fin (mm)d Second ventral fin (mm)d Mass of drumming muscles (g)e Lernaeocera branchialis Cryptocotyle lingua Echinorhynchus gadi Intestinal digeneans f Anisakidae Ig Anisakidae IIh

Females (n = 18)

Mean

SD

Mean

6.5 1397.9 24.7 0.0012 0.0084 75.9 66.5 55.1 55.7 49.0 2.99 0.4 593.2 39.0 55.5 84.1 222.3

0.9 734.1 4.2 0.0013 0.0053 15.2 9.5 8.1 8.0 9.1 2.92 0.6 1850.2 95.1 74.4 260.4 487.1

6.7 1545.9 27.2 0.0014 0.0060 67.8 56.0 58.4 51.0 2.04 0.4 1142.7 16.5 22.4 126.3 315.6

SD 0.8 515.1 3.5 0.0016 0.0018 8.9 7.5 8.1 7.7 1.07 0.8 2405.4 51.6 24.3 183.7 323.6

a

n = 16 for females. Density estimate. c n = 33. d Length of fin ray (see the text). e Dry mass of the sound-producing muscles. f Combined intensities of H. levinseni and D. varicus. g Infections found on the liver. h Infections located on the stomach and pylorus caeca. b

Table 2. Results from regression analyses between mass of drumming muscles (DM) and the investigated parameters for reproductively active male and female cod. Males (n = 38)

Females (n = 18)

DM (mass)a

DM

DM (mass)a

DM

Age

r = 0.05 ns

r = 0.44 p = 0.006

r = 0.52 ns

r = 0.32 ns

Lymphocytes

r = –0.03 ns

r = 0.22 ns

r = –0.40 ns

r = –0.21 ns

Granulocytesb

r = –0.05 ns

r = –0.27 ns

r = –0.50 ns

r = –0.42 ns

Parasite index

r = 0.08 ns

r = 0.37 p = 0.022

r = 0.28 ns

r = 0.34 ns

Note: “ns” denotes “not significant.” a Mass of drumming muscles controlled for fish body mass. b n = 16 for females.

needed transformation (logarithmic) in order to meet the assumption of normality. When controlling for confounding variables, residual values were used (Kleinbaum et al. 1988). All statistical tests applied in this study were two-tailed, with a 5% significance level. No adjustment for multiple comparisons was done when testing the predictions, because according to Rothman (1990), this leads to fewer errors of interpretation when the data under evaluation are not random numbers but actual observations from nature.

Results Descriptive statistics for the parameters used in this study are given for males and females separately in Table 1. The

results of univariate regression analyses between mass of drumming muscles or fin size and other variables are given in Tables 2 and 3, respectively. Confounding variables In males, mass of drumming muscles and fin size were both positively related to fish body mass (r = 0.75, p < 0.001, and r = 0.90, p < 0.001, respectively; n = 38) and age (Tables 2 and 3). However, after body mass was controlled for, mass of drumming muscles and fin size were no longer correlated with age. In females, mass of drumming muscles and fin size were significantly related to body mass (r = © 1999 NRC Canada



546

Can. J. Zool. Vol. 77, 1999 Table 3. Results from regression analyses between fin size (FS) and the investigated parameters for reproductively active male and female cod. Males (n = 38)

Females (n = 18)

FS (mass)a

FS

FS (mass)a

FS

Age

r < 0.01 ns

r = 0.47 p = 0.003

r = 0.16 ns

r = 0.45 ns

Lymphocytes

r = –0.11 ns

r = 0.15 ns

r = –0.39 ns

r = –0.30 ns

Granulocytesb

r = –0.24 ns

r = –0.16 ns

r = –0.30 ns

r = –0.30 ns

Parasite index

r < 0.01 ns

r = 0.38 p = 0.019

r = 0.13 ns

r = 0.24 ns

Note: “ns” denotes “not significant.” a Fin size controlled for fish body mass. b n = 16 for females.

0.65, p = 0.004, and r = 0.89, p < 0.001, respectively; n = 18) but not to age (Tables 2 and 3). Consequently, among both males and females, mass of drumming muscles and fin size were analyzed both before and after controlling for fish body mass.

Fig. 2. Box-and-whisker plot of the residual masses of drumming muscles after controlling for fish mass in reproductively active male (n = 38) and female cod (n = 18), showing that males have relatively heavier drumming muscles than females (t = 3.35, df = 54, p = 0.001). Mean, standard error, and standard deviation are shown.

Mass of drumming muscles and fin size Males and females with large fins also had a large mass of drumming muscles (r = 0.68, p < 0.0001 (n = 38), and r = 0.68, p = 0.003 (n = 18), respectively). However, when both mass of drumming muscles and fin size were controlled for body mass, these relationships disappeared (r = 0.01, p = 0.93 (n = 38), and r = 0.28, p = 0.26 (n = 18), for males and females, respectively). Sexual size dimorphism The sexes did not differ in either mass of drumming muscles or fin size when body mass was not controlled for (t = 1.26, df = 54, p = 0.21, and t = –0.7, df = 54, p = 0.45, respectively). Males had smaller fins than females for a given body mass, but the difference was not significant (t = –0.17, df = 54, p = 0.87). However, males had significantly heavier drumming muscles after individual body mass was partialled out (Fig. 2). Parasite index The parasite index for males increased with body mass (r = 0.42, p = 0.009; n = 38), and it was also positively related to mass of drumming muscles when fish body mass was not controlled for (Table 2). Furthermore, fin size increased significantly with parasite index, but not after body mass was partialled out (Table 3). In females, no significant relationship was found between parasite index and fish body mass (r = 0.20, p = 0.44; n = 18). Additionally, the parasite index did not correlate with mass of drumming muscles or fin size in females before or after body mass was partialled out. Immunology Among males, granulocyte density was not related to fish body mass (r = –0.25, p = 0.13; n = 38). Furthermore,

granulocyte density did not correlate with mass of drumming muscles (Table 2) or fin size (Table 3), either before or after fish body mass was controlled for. No relationship was found between granulocyte density and female body mass (r = –0.08, p = 0.78; n = 16). Additionally, granulocyte density was not related to mass of drumming muscles or fin size in females either before or after body mass was controlled for. Lymphocyte density in males did not correlate with body mass (r = 0.21, p = 0.20; n = 38) or with mass of drumming muscles or fin size. Lymphocyte density was not related to body mass in females (r = 0.17, p = 0.55; n = 18), nor did it correlate with their mass of drumming muscles or fin size. Spermatocrit There was no relationship between spermatocrit level and body mass (r = 0.12, p = 0.51; n = 33), nor was spermatocrit level related to mass of drumming muscles before fish mass was controlled for (r = –0.17, p = 0.34; n = 33). However, for a given body mass, individuals with light drumming muscles had significantly higher spermatocrit levels than individuals with heavy drumming muscles (Fig. 3). Sperma© 1999 NRC Canada



Engen and Folstad Fig. 3. Correlation between spermatocrit (percentage of sperm cells in milt) and residual mass of drumming muscle after controlling for fish body mass in reproductively active male cod (r = –0.38, p = 0.031, n = 33).

tocrit level and fin size showed no significant relationship before body mass was controlled for (r = 0.31, p = 0.09; n = 33), but were positively correlated after body mass was controlled for (Fig. 4). Consequently, both fin size and mass of drumming muscles could explain a significant amount of the variation in spermatocrit levels. Therefore, we conducted multiple regression analyses post hoc to estimate the combined explanatory value of fin size and mass of drumming muscles on spermatocrit level. When individual body mass was not controlled for, mass of drumming muscles and fin size together explained 27% of the variation in spermatocrit levels (F[2,30] = 6.89, p = 0.004). On the other hand, when mass of drumming muscles and fin size were controlled for fish body mass, 27% of the variation in spermatocrit levels was still explained (F[2,30] = 6.84, p = 0.004). Furthermore, for a given fin size, mass of drumming muscles was negatively correlated with spermatocrit level (r = –0.41, p = 0.016; n = 33), and when fin size was controlled for mass of drumming muscles, spermatocrit level increased with fin size (r = 0.55, p = 0.0009; n = 33). Consequently, by evaluating these characters in relation to each other, with a measurement accuracy comparable to ours, spermatocrit levels can be estimated with a certainty of 17–30%.

Discussion Based on the description of the reproductive behaviour of cod, sound and fin display of males seem to be involved in transferring information to females. The mass of the drumming muscles of cod was sexually size dimorphic, with males having larger muscles than females, whereas fin sizes did not differ between the sexes during the spawning period. Moreover, both mass of drumming muscles and fin size seem to have the potential to reveal information about the fertilization potential of males. It seems less likely, however, that these characters serve as a source of information about individual parasite resistance. Sexually dimorphic drumming muscles have been documented for other sound-producing fish species such as Porichthys notatus (Brantley and Bass 1994; Bass 1997) and Opsanus tau (Gray and Winn 1961). In haddock (Melanogrammus aeglefinus), which belongs to the cod family (Gadidae), the sexually dimorphic drumming muscles of mature

547 Fig. 4. Correlation between spermatocrit (percentage of sperm cells in milt) and residual fin size for a given fish body mass in reproductively active male cod (r = 0.41, p = 0.017, n = 33).

males increase to nearly twice their normal size during the spawning period (Templeman and Hodder 1958; Hawkins et al. 1967). This developmental pattern may reflect the important function of the drumming muscles, and thus of sound production, in the reproductive behaviour of haddock (Templeman and Hodder 1958). Sexually selected traits may signal immunocompetence (Folstad and Karter 1992), and increased expression of secondary sexual characters has previously been reported to be negatively related to leukocyte densities (Saino and Møller 1994; Skarstein and Folstad 1996). In this study, the level of circulating lymphocytes and granulocytes was not correlated with the sexually dimorphic mass of drumming muscles or fin size. Therefore, it seems less likely that these morphological characters reveal information about individual qualities related to immunological activity in male cod. Moreover, an intraspecific negative relationship between parasite burden and the development of secondary sexual characters had been suggested earlier (Hamilton and Zuk 1982; reviewed by Hamilton and Poulin 1997). In this investigation, parasite index was positively related to both mass of drumming muscles and fin size in males. These characters thus seem to reveal the parasite burden of an individual. However, a positive relationship between parasite index and fish body mass was also found, but when body mass was controlled for, the associations between parasite index and the morphological characters examined disappeared. Assuming a positive relationship between individual food consumption and body mass, the positive relationship between parasite burden and fish body mass might have been expected, as four of the six parasite groups included in this study are long-lived and food-transmitted. Consequently, even if mass of drumming muscles and fin size can provide information about parasite burden, the associations might result from a increase in parasite intensity and character size with fish mass. Therefore, the size of these characters need not be influenced by the need to transmit information about heritable parasite resistance as an indirect benefit to mateseeking females. The negative association between spermatocrit level and mass of drumming muscles may be explained by variation in ejaculation frequency related to differences in attractiveness between males in at least three ways. First, cod are batchspawners (Kjesbu 1989), and batch-spawners exhibit a decrease in spermatocrit level with increasing number of © 1999 NRC Canada



548

batches spawned (Billard et al. 1971; Billard 1983; Büyükhatipoglu and Holtz 1984; Munkittrick and Moccia 1987). If males with the largest drumming muscles are the most attractive, the association between spermatocrit level and mass of drumming muscles in cod can be explained. Second, in males with the highest overall mating success, the amount of sperm contributed during each mating may be reduced (Warner et al. 1995) to the point that sperm release may be stopped when the ratio of eggs fertilized to sperm released is at a maximum (Shapiro and Giraldeau 1996). This could also result in low spermatocrit levels among attractive males. Third, in several species with high mating frequencies, males often experience low sperm output (Dewsbury 1982; Pitnick and Markow 1994). This may occur if the time required to replenish the sperm supply after the previous spawning is too short, something that might be more common among attractive than unattractive males. Consequently, if males with large drumming muscles are more attractive than those with small drumming muscles, a high batch number, optimization of sperm release among matings, and a short time between spawnings may all have led to a negative relationship between mass of drumming muscles and spermatocrit level in the cod examined. Males with large drumming muscles may thus be attractive, but fin size was positively related to spermatocrit level. Why should males with small fins be attractive? In species breeding during darkness, vocal signals may be of greater importance in mate attraction than visual ones (Ryan 1990). In courting male cod, fin size can, however, be revealed either visually or during the courtship dance in which water movements initiated by fin movements may be perceived by the female through her lateral-line organ. Fin size may thus be important in short-range mate attraction. The drumming sound of male cod may, on the other hand, be more effective in long-range mate attraction. In order to be attractive, males should therefore need to invest in sound-producing muscles, and a trade-off may exist in the allocation of resources to either drumming muscles or to muscles important for fin movements. The allocation of resources to the drumming muscles may thus occur at the expense of resources for the growth of muscles important for propulsion. Since deviation from optimal fin size may influence swimming capabilities or maneuverability, females may, by evaluating sound, fin display, and dance, obtain information about the total amount of a male’s resources available for development of muscle tissue and also his relative investment in signals (i.e., drumming muscle tissue). That is, an individual’s investment in drumming muscles occurs at the expense of swimming capability, which may be revealed during the intimately executed courtship dance. This trade off may ensure costs and consequently reliability to drumming sound. The positive relationship between fin size and spermatocrit level could thus be explained, since males that invest in muscles important for fin movement rather than in muscles important for sound production should experience fewer spawnings and thus have higher spermatocrit levels. This corresponds to the findings of the present study. Males with small fins for a given drumming muscle mass have lower spermatocrit levels than males with large fins, and males with heavy drumming muscles for a given fin size have low spermatocrit levels.

Can. J. Zool. Vol. 77, 1999

This tentative conclusion, that there is a trade-off in allocation of resources for muscle growth, may explain why it is not enough for females to evaluate drumming sound alone in order to estimate a male’s investment in current reproduction. Rather, by evaluating fin size and dance as well, females may also obtain information about the amount of residual resources (i.e., the amount of resources used for sound-producing muscles compared with the amount used for muscles important for propulsion). Thus, sound production, fin display, and dance, which initially may seem to be independent behavioural traits, could have a common denominator in allocation of resources for muscle development. Additionally, knowledge about this allocation may reveal information about a male’s fertilization potential, which, for female cod at risk of not getting all their eggs fertilized, may be of great importance and represent information about direct benefits to be gained in their search for a mate.

Acknowledgments We thank Ståle Liljedal, Eystein Markusson, Inger Martinussen, Anders Pape Møller, Jarle Tryti Nordeide, and Frode Skarstein for valuable comments on the manuscript and two gentlemen, Huse and Rønning, for excellent assistance during fieldwork.

References Aas, G.H., Refstie, T., and Gjerde, B. 1991. Evaluation of milt quality of Atlantic salmon. Aquaculture, 95: 125–132. Andersson, M. 1994. Sexual selection. Princeton University Press, Princeton, N.J. Basolo, A.L. 1990. Female preference for male sword length in the green swordtail, Xiphophorus helleri (Pisces: Poeciliidae). Anim. Behav. 40: 332–338. Bass, A.H. 1997. Comparative neurobiology of vocal behaviour in teleost fishes. Mar. Freshwater Behav. Physiol. 29: 47–63. Bauer, J.A., and Bauer, S.E. 1981. Reproductive biology of pigmy angelfishes of the genus Centropyge (Pomacanthidae). Bull. Mar. Sci. 31: 495–513. Billard, R. 1983. A quantitative analysis of spermatogenesis in the trout, Salmo truttafario. Cell Tissue Res. 230: 495–502. Billard, R., Breton, B., and Jalabert, B. 1971. La production spermatogénétique chez la truite. Ann. Biol. Anim. Biochim. Biophys. 11: 199–212. Bischoff, R.J., Gould, J.L., and Rubenstein, D.I. 1985. Tail size and female choice in the guppy (Poecilia reticulata). Behav. Ecol. Sociobiol. 17: 253–255. Bouck, G.R., and Jacobsen, J. 1976. Estimation of salmonid sperm concentration by microhematocrit technique. Trans. Am. Fish. Soc. 105: 534–535. Brantley, R.K., Jr. 1992. Ontogeny of inter- and intrasexual dimorphism in a vocalizing fish: behavioral, morphological and endocrine correlates. Ph.D., Cornell University. Ithaca, N.Y. Brantley, R.K., and Bass, A.H. 1994. Alternative male spawning tactics and acoustic-signals in the plainfin midshipman fish Porichthys notatus Girard (Teleostei, Batrachoididae). Ethology, 96: 213–232. Brawn, V.M. 1961a. Aggressive behaviour in the cod (Gadus callarias L.). Behaviour, 18: 107–147. Brawn, V.M. 1961b. Reproductive behaviour of the cod (Gadus callarias L.). Behaviour, 18: 176–197. © 1999 NRC Canada



Engen and Folstad Brawn, V.M. 1961c. Sound production by the cod (Gadus callarias L.). Behaviour, 18: 239–255. Brønseth, T., and Folstad, I. 1997. The effect of parasites on courtship dance in threespine sticklebacks: more than meets the eye? Can. J. Zool. 75: 589–594. Büyükhatipoglu, S., and Holtz, W. 1984. Sperm output in rainbow trout (Salmo gairdneri) — effect of age, timing and frequency of stripping and presence of females. Aquaculture, 37: 63–71. Clutton-Brock, T.H., and Vincent, A.C.J. 1991. Sexual selection and the potential reproductive rates of males and females. Nature (Lond.), 351: 58–60. Daye, P.G., and Glebe, B.D. 1984. Fertilization success and sperm motility of Atlantic salmon (Salmo salar L.) in acidified water. Aquaculture, 43: 307–312. Dewsbury, D.A. 1982. Ejaculate cost and male choice. Am. Nat. 119: 601–610. Ellertsen, B., Solemdal, P., Strømme, T., Sundby, S., Tilseth, S., and Westgård, T. 1981. Spawning period, transport and dispersal of eggs from the spawning area of Arcto-Norwegian cod (Gadus morhua L.). Rapp. P.-V. Reun. Cons. Int. Explor. Mer, 178: 260–267. Fisher, R.A. 1930. The genetical theory of natural selection. Clarendon Press, Oxford. Folstad, I., and Karter, A.J. 1992. Parasites, bright males and the immunocompetence handicap. Am. Nat. 139: 603–622. Gerald, J.W. 1971. Sound production during courtship in six species of sunfish (Centrarchidae). Evolution, 25: 75–87. Gibson, R.M., and Jewell, P.A. 1982. Semen quality, female choice and multiple mating in domestic sheep: a test of Trivers’ sexual competence hypothesis. Behaviour, 80: 9–31. Gray, G.-A., and Winn, H.E. 1961. Reproductive ecology and sound production of the toadfish, Opsanus tau. Ecology, 42: 274–282. Hamilton, W.D., and Zuk, M. 1982. Heritable true fitness and bright males: a role for parasites? Science (Washington, D.C.), 218: 384–387. Hamilton, W.J., and Poulin, R. 1997. The Hamilton and Zuk hypothesis revisited: a meta-analytical approach. Behaviour, 134: 299–320. Hasson, O. 1997. Towards a general theory of biological signaling. J. Theor. Biol. 185: 139–156. Hawkins, A.D., Chapman, K.J., and Symonds, D.J. 1967. Spawning of haddock in captivity. Nature (Lond.), 215: 923–925. Hemmingsen, W., and MacKenzie, K. 1993. A checklist of the protozoan and metazoan parasites reported from the Atlantic cod, Gadus morhua L. Bull. Eur. Assoc. Fish Pathol. 13: 134–137. Johnsen, T. 1981. Otolittsoner og vekstprosesser hos torsk (Gadus morhua L.) i Balsfjorden. Cand. Scient. Thesis, University of Tromsø, Tromsø, Norway. Jones, F.R.H., and Marshall, N.B. 1953. The structure and function of the teleostan swimbladder. Biol. Rev. Camb. Philos. Soc. 28: 16–83. Kjesbu, O.S. 1989. The spawning activity of cod, Gadus morhua L. J. Fish Biol. 34: 195–206. Kleinbaum, D.G., Kupper, L.L., and Muller, K.E. 1988. Applied regression analysis and other multivariate methods. 2nd ed. Duxbury Press, Belmont, Calif. Levitan, D.R. 1993. The importance of sperm limitation to the evolution of egg size in marine invertebrates. Am. Nat. 141: 517–536. Lysne, D.A., Hemmingsen, W., and Skorping, A. 1994. The distribution of Cryptocotyle spp. metacercariae in the skin of caged Atlantic cod (Gadus morhua). J. Fish Biol. 45: 352–355. Macias Garcia, C., Jiminez, G., and Contreras, B. 1994. Correlational evidence of asexually-selected handicap. Behav. Ecol. Sociobiol. 35: 253–259. Markusson, E., and Folstad, I. 1997. Reindeer antlers: visual indicators of individual quality? Oecologica, 110: 501–507.

549 Matthews, I.M., Evans, J.P., and Magurran, A.E. 1997. Male display rate reveals ejaculate characteristics in the Trinidadian guppy Poecilia reticulata. Proc. R. Soc. Lond. B Biol. Sci. 264 (1382): 695–700. McKenzie, R.A. 1940. Nova Scotian autumn cod spawning. J. Fish. Res. Board Can. 5: 105–120. Mjelstad, H. 1991.Displaying intensity and sperm quality in the capercaillie Terao urogallus. Fauna Norv. Ser. C, 14: 93–94. Munkittrick, K.R., and Moccia, R.D. 1987. Seasonal changes in the quality of rainbow trout (Salmo gairdneri) semen: effect of a delay in stripping on spermatocrit, motility, volume and seminal plasma constituents. Aquaculture, 64: 147–156. Myrberg, A.A., Jr., Mohler, M., and Catala, J.D. 1986. Sound production by males of acoral reef fish (Pomacentrus partitus): its significance to females. Anim. Behav. 34: 913–923. Möller, H., and Anders, K. 1986. Diseases and parasites of marine fishes. Verlag Möller, Kiel. Nakatsuru, K., and Kramer, D.L. 1982. Is sperm cheap? Limited male fertility and female choice in lemon tetra (Pisces, Characidae). Science (Washington, D.C.), 216: 753–755. Obraztsov, A.N. 1985. Estimation of sperm concentration in rainbow trout (Salmo gairdneri Rich.). In Genetic and ecological implications in fish culture. [In Russian with English summary.] pp. 111–116. Odhner, T. 1905. Die Trematoden des arktischen Gebietes. Erlangung der Doktorwürde, University of Uppsala, Sweden. Pedersen, T. 1984. Variation of peak spawning of arcto-Norwegian cod (Gadus morhua L.) during the time period 1929–1982 based on indices estimated from fishery statistics. In The propagation of cod Gadus morhua. L. Edited by E. Dahl, D.S. Danielssen, E. Moksness, and P. Solemdal, Institute of Marine Research, Flødevigen Biological Station, Arendal, Norway. Flødevigen rapporter No. 1. pp. 301–316. Petersen, C.W., Warner, R.R., Cohen, S., Hess, H.C., and Sewell, A.T. 1992. Variable pelagic fertilization success: implications for mate choice and spatial patterns of mating. Ecology, 73: 391–401. Pitnick, S., and Markow, T.A. 1994. Male gametic strategies: sperm size, testis size, and the allocation of ejaculate among successive mates by the sperm-limited fly Drosophila pachea and its relatives. Am. Nat. 143: 785–819. Rollefsen, G. 1932. Litt om skreiens gyting. Årsberetn. Norg. Fisk. 1932(2): 95–97. Rothman, K.J. 1990. No adjustments are needed for multiple comparisons. Epidemiology, 1: 43–46. Rowley, A.F. 1990. Collection, separation and identification of fish leukocytes. In Techniques in fish immunology. Edited by J.S. Stolen, T.C. Fletcher, D.P. Anderson, B.S. Roberson, and W.B. van Muiswinkel. SOS Publications, Fair Haven, N.J. pp. 113–135. Ryan, M.J. 1990. Sexual selection, sensory systems and sensory exploitation. Oxf. Surv. Evol. Biol. 7: 157–195. Saino, N., and Møller, A.P. 1994. Secondary sexual characters, parasites and testosterone in the barn swallow, Hirundo rustica. Anim. Behav. 48: 1325–1333. Sars, G.O. 1879. Indberetninger til Departementet for det indre. Departementet for det indre, Christiania, Norway. Shapiro, D.Y., and Giraldeau, L.-A. 1996. Mating tactics in external fertilizers when sperm is limited. Behav. Ecol. 7: 19–23. Skarstein, F., and Folstad, I. 1996. Sexual dichromatism and the immunocompetence handicap: an observational approach using Arctic charr. Oikos, 76: 359–367. Stout, J.F. 1975. Sound communication during the reproductive behavior of Notropis analostanus (Pisces: Cyprinidae). Am. Midl. Nat. 94: 296–325. Tavolga, W.N. 1958. The significance of underwater sounds pro© 1999 NRC Canada



550 duced by males of the gobiid fish, Bathygobius soporator. Physiol. Zool. 31: 259–271. Templeman, W. 1958. How cod spawn — Nielsen’s observations. J. Fish Res. Board Can. 68: 15–16. Templeman, W., and Hodder, V.M. 1958. Variation in fish length, sex, stage of sexual maturity, and season in the appearance and volume of the drumming muscles of the swim-bladder in the Haddock, Melanogrammus aeglefinus (L.). J. Fish Res. Board Can. 15: 355–390. Trippel, E.A., and Neilson, J.D. 1992. Fertility and sperm quality of virgin and repeat-spawning Atlantic cod (Gadus morhua) and associated hatching success. Can. J. Fish. Aquat. Sci. 49: 2118–2127. Trivers, R.L. 1972. Parental investment and sexual selection. In Sexual selection and the descent of man. Edited by B. Campbell. Heinemann, London. pp. 136–197. Warner, R.R., Shapiro, D.Y., Marcanato, A., and Petersen, C.W. 1995. Sexual conflict: males with the highest mating success

Can. J. Zool. Vol. 77, 1999 convey the lowest fertilization benefits to females. Proc. R. Soc. Lond. B Biol. Sci. 262: 135–139. Wiborg, K.F. 1957. Factors influencing the size of the year classes in the Arcto-Norwegian tribe of cod. Fisk Dir. Skr. Ser. Havunders. 11: 1–24. Williams, G.C. 1992. Natural selection: domains, levels and challanges. Oxford University Press, Oxford. Williams, M.B. 1978. Sexual selection, adaptation and ornamental traits: the advantage of seeming fitter. J. Theor. Biol. 72: 377–383. Winn, H.E. 1972. Acoustic discrimination by the toadfish with comments on signal systems. In Behaviour of marine animals: current perspectives in research. Vol. 2. Vertebrates. Edited by H.E. Winn and B.L. Olla. Plenum Press, New York. pp. 361–385. Zahavi, A. 1975. Mate selection — a selection for a handicap. J. Theor. Biol. 53: 205–214. Zahavi, A. 1977. The cost of honesty. J. Theor. Biol. 67: 603–605.

© 1999 NRC Canada
