Behavioral Ecology Advance Access originally published online on March 5, 2007
Behavioral Ecology 2007 18(3):535-540; doi:10.1093/beheco/arm007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
No terminal investment in pipefish males: only young males exhibit risk-prone courtship behavior
a Department of Biology, Norwegian University of Science and Technology, N-7491 Trondheim, Norway b Department of Ecology and Evolution/Animal Ecology, Uppsala University, Norbyvägen 18d, S-752 36 Uppsala, Sweden
Address correspondence to A.M. Billing. E-mail: anna.billing{at}bio.ntnu.no.
Received 15 May 2006; revised 7 January 2007; accepted 25 January 2007.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Animals are expected to trade-off current and future reproduction in order to maximize lifetime reproductive success. Old individuals may accept higher risks during courtship and mate choice as their residual reproductive value (RRV) diminishes (the terminal investment hypothesis). Alternatively, young individuals may be forced to take higher risks during courtship to compensate for their lower competitiveness and/or attractiveness (the compensation hypothesis). In this study, we used the sex-role reversed pipefish Syngnathus typhle to test how mate choice and courtship behavior of males with different RRV were affected by an increase in predation risk. Males of different ages were given the opportunity to court and choose between 2 partners. In half of the trials, a predator was present in a separate aquarium. We found no support for the terminal investment hypothesis: no difference in response to the increased predation risk by males of different ages was evident. In agreement with the compensation hypothesis, young males invested more in courtship behavior compared with older males. In addition, in the absence of a predator, we found that a high female activity was important for male mate choice decisions. During increased predation risk, this relationship was, however, reversed and males preferred less active, and thus less conspicuous, partners. This suggests that both female activity and size are important factors for male mating decisions in this species and that these decisions mainly are affected by predation risk and advantages in mate acquisition.
Key words: courtship behavior, male mate choice, predation risk, reproductive trade-off, Syngnathus typhle, terminal investment.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
According to life-history theory, animals should adjust their behavior in order to maximize their lifetime reproductive success and thereby their fitness (Williams 1966
). Because of the multitude of factors affecting reproductive success, there are always trade-offs between different behavioral choices. For animals with more than one reproductive event during their lifetime, one such trade-off is between current and future reproduction (Williams 1966; Magnhagen and Vestergaard 1991). A high investment in current reproduction might reduce the capacity or possibility to invest in future reproductive events, that is, the residual reproductive value (RRV) is reduced. This is especially true for individuals experiencing an increased predation risk during reproduction (Lima and Dill 1990; Magnhagen 1991). Because reproductive activities, such as mate search, mate choice, courtship, mating/copulation, or parental care, typically involve conspicuous behaviors that increase the risk of predation, there is a clear conflict between maximizing reproductive output and minimizing the risk of predation (Lima and Dill 1990; Forsgren and Magnhagen 1993; Clark 1994).
Theoretical models predict that the strength of mate preferences should decrease with an increasing cost of mate choice (Crowley et al. 1991; Kokko and Monaghan 2001). If an individual experiences an increase in predation risk during mate choice or courtship behavior, one would expect a reduction in choosiness and courtship activity. In fishes, this has been shown for a number of species, for example, guppy (Poecilia reticulata; Godin and Briggs 1996), goby (Pomatoschistus minutus; Forsgren 1992; Forsgren and Magnhagen 1993), pipefish (Syngnathus typhle; Berglund 1993; Fuller and Berglund 1996), and stickleback (Gasterosteus aculeatus; Candolin 1997).
The cost of predation in terms of lifetime reproductive success can, however, differ between individuals of different ages. With increasing age, the probability of future reproductive events, and thereby the RRV of an individual, is reduced. Old individuals are therefore expected to make a terminal investment in reproduction and to accept large risks, be choosy and court intensely regardless of level of predation risk in order to get a partner of high quality (Magnhagen and Vestergaard 1991; Clutton-Brock and Parker 1992; Pärt et al. 1992; Clark 1994; Candolin 1998). On the other hand, young individuals breeding for the first time have a high RRV and predation during reproduction would therefore result in loss of their entire lifetime reproductive success. Young individuals are therefore expected to respond more to an increase in predation threat by a greater reduction in choosiness, courtship, and general activity compared with older individuals. Furthermore, for species with two or more reproductive events per breeding season, the effect of age can be expected to be even more pronounced in the later part of the breeding season as the RRV for old individuals then is even further reduced. Young individuals, on the other hand, will still have quite a high RRV as they have a high chance of reproducing again the coming season. These life-history theory arguments are called the "terminal investment hypothesis" (Clutton-Brock 1984; Gustafsson and Pärt 1990).
A contrasting set of predictions stems from sexual selection theory. If small or young individuals are less competitive and/or attractive than large or old individuals, these individuals may have to compensate for their shortcomings by taking larger risks in mate choice and courtship, in order to secure high-quality matings, or any matings at all. We call this the "compensation hypothesis." Here we test which of these 2 hypotheses best explains male mating decisions in the sex-role reversed pipefish Syngnathus typhle.
S. typhle is found along most European coasts, mainly in shallow eelgrass, Zostera marina, meadows (Berglund et al. 1986a). S. typhle is a polygamous species with predominantly male mate choice and female–female competition, but also females discriminate between partners if given the opportunity (Berglund et al. 1986a, 2005; Sandvik et al. 2000). Fecundity increases with increasing body size for both sexes, and both males and females prefer large over small partners (Berglund et al. 1986a, 1986b, 2005; Sandvik et al. 2000). Males have also been shown to prefer dominant over subdominant females (Berglund and Rosenqvist 2001a). However, large females are also more dominant, and dominance and attractiveness are therefore correlated in female pipefish (Berglund and Rosenqvist 2001a, 2001b). S. typhle has exclusive male parental care, and during brooding, the male supplies the eggs with oxygen and an osmoregulated environment (Berglund et al. 1986a, 1986b).
The colors and elongated body shape of S. typhle make it very cryptic in the eelgrass, and it relies heavily on crypsis as a defense against predators (Vincent et al. 1994). However, during courtship and copulation, the crypsis of the pipefish is greatly reduced (Vincent et al. 1994). Copulation is preceded by the male and female engaging in a prolonged nuptial dance that includes body shakes and ornament display (Vincent et al. 1994; Berglund et al. 1997; Bernet et al. 1998). Male choosiness and courtship intensity has earlier been shown to decrease with an increased level of predation risk in this species (Berglund 1993; Fuller and Berglund 1996). However, if this effect differs between individuals with different RRV has not yet been studied.
To investigate whether individuals with different RRV respond differently to an increase in predation risk, we conducted aquaria experiments where we examined male mate choice and male courtship behavior under different predation pressure, between males of different ages and at 2 different periods (early and late) during the breeding season.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The study was carried out at Kristineberg Marine Research Station and Klubban Biological Station in Sweden (58°15'N, 11°28'E) from May to July 2002. The fish were caught during 2 periods (mid-May before the onset of reproduction and mid-June toward the end of the first male brooding period) by trawling in shallow eelgrass meadows in the Gullmar Fjord. After capture, the fish were brought to the laboratory and kept in 225-l barrels with continuously renewed sea water. The number of fish in each barrel was around 70, and sexes were kept separate. The pipefish were fed ad libitum with frozen mysids and both frozen and live brine shrimps, Artemia sp. The barrels were cleaned daily and equipped with plastic eelgrass for shelter. The conditions in the barrels followed the natural light regime, temperature, and salinity in the waters around the research station during the study period.
The experiments were carried out in 2 separate periods during the natural breeding season of pipefish, one early (25 May–6 June) during the first male brooding period and one late (23 June–7 July) during the second male brooding period. To avoid effects of prolonged captivity, new fish (both males and females) were caught before the start of the second trial period. To ensure that all males used in the second period had already carried one brood, we caught pregnant males and allowed them to give birth in the lab.
Like most fish species, pipefish has indeterminate growth, and the bimodal size distribution in the local population allows us to assume that the size of the fish reflects their age (Berglund et al. 1989; Berglund 1991). We thus divided the males into 2 different size categories to represent different age classes. Small males (128.5 ± 8.9 mm) were considered 1 year old, and large males (194.9 ± 16.1 mm) were considered to be older. We did not use medium size males, and the sizes of the fish in the 2 age categories were not overlapping.
The mate choice experiments took place in 120-l mate choice aquaria (40 x 50 cm and 60 cm high, Figure 1). The aquaria were divided lengthwise into 3 compartments, 2 small in the back and one large in the front of the aquarium, using transparent dividers. Two females of different sizes were placed in the rear compartments, and a male from one of the two age categories (order randomly assigned) was placed in the larger front compartment, from which he had an equal possibility to inspect both female compartments. Focal males were only used once, but 8 of the 238 stimulus females had to be used twice. These females, however, were never paired with the same female more than once, and there was always a period of at least 2 days between the first and the second time a particular female was used. In order to provide the pipefish with shelter, the aquaria were equipped with live eelgrass, planted in rinsed seashore sand on the bottom. During the experimental period, the aquaria were also supplied with continuously renewed sea water entering the aquaria from the rear part of the 2 female compartments with the outlet located in the male compartment. This created a current from the females to the male, allowing the male to smell any olfactory cues emitted from the females during the mate choice trial.
|
In half of the trials (randomly assigned), a locally abundant predator, Atlantic cod, Gadus morhua, was present in a separate aquarium placed behind the female compartments of the mate choice aquarium (Figure 1). This setup has been used successfully in earlier mate choice experiments (Berglund 1993
). In this earlier study, the male responded clearly to the predator. Furthermore, the placement of the tank ensures that the male can observe both females and predator at the same time. The predator can also see all individuals, and the closeness of the females to the predator ensures the predacious activity of the predator to remain high. The cod was picked at random from a pool of 20 cods available for the experiment. No cod was used more than once per day, and each cod was used approximately 6 times in total (determined by measuring the cod used after each trial). All fish could see each other during the whole trial, and all trials lasted for 4 h. The experiments were video filmed using Hitachi Hi-8 WM-H80E video cameras connected to Panasonic AG-6730 time-lapse super-VHS video recorders equipped with AG-IA670 time code generators/computer interfaces, using the 24-h time-lapse mode (which uses one-eighth of normal speed). Intense light (one 100-W spot and one 20-W luminescent lamp placed 0.5 m above each aquarium), in addition to natural light from windows, was used during filming. The behavior of all individuals was later analyzed from the videotapes using the Noldus Observer 4.1 software.
For the male, time spent in front of either of the 2 stimulus female compartments was recorded, and a male was considered to have chosen a female if he spent more than 50% of his time in front of that female. Earlier experiments have confirmed that this predicts subsequent mating decisions (Berglund and Rosenqvist 1993). For both males and females, the total duration of swimming, resting, and dancing was recorded. Dancing was defined as a male and a female synchronously swimming up and down facing each other. The length of the dance varies considerably from short sequences of 10–20 s up to 10 min or more. In addition, number of male body shakes (a predominantly male courtship behavior performed in front of females) was recorded. The shake behavior consists of a short shake of the body and can occur both before, during, or after dancing. It can also be performed without dancing at all. The recorded activity of the females was incorporated as covariates in the statistical analyses to control for effects of female activity on male behavior.
We also recorded the activity of the cod. There was a low variation in cod activity with all being active throughout most of the trials (cod activity [mean ± standard deviation {SD}): 97.5 ± 6.5% of the total trial time). The cod activity did not differ significantly between trials with males of different ages (Kruskal–Wallis test: 
2 = 0.4189, degrees of freedom [df] = 1, P = 0.518; mean percentage of total time ± SD—young males: 96.7 ± 8.3; old males: 98.3 ± 4.3) or between the breeding periods (Kruskal–Wallis test: 2 = 3.1677, df = 1, P = 0.075; mean percentage of total time ± SD—early: 99.0 ± 1.5; late: 96.0 ± 8.9).
We used 3 measurements of male preference: "total preference" (time spent in front of the large female minus the time in front of the small female, i.e., positive values indicating a choice of the large female and vice versa), "dancing preference" (the difference in time dancing with the large female minus time dancing with the small female), and "shake preference" (number of shakes in front of the large female minus number of shakes in front of the small female). Total preference was tested both for all males and when only including males engaging in courtship activities (dance or shake).
To assess of male courtship behavior, we tested 2 measures of dancing behavior: "dancing willingness" (whether a male engaged in dancing behavior or not) and "dancing duration" (total duration a male danced regardless of which female the dance was directed to) and 2 measures of shake behavior: "shake willingness" (whether a male shaked or not) and "shake frequency" (total number of shakes a male performed during the trial regardless of which female it was directed to).
In order to obtain a basic view of the preference of the males used in the experiment, we first tested the preference measures against zero (i.e., no preference). The preference measures (total preference, dancing preference, and shake preference) were then tested against full factorial analysis of covariance (ANCOVA) or general linear models (GLMs) with male age, time of season, and predator presence as factors and relative female activity (large female activity divided by total female activity) as a covariate. The courtship measures (dance and court willingness, dancing duration, and shake frequency) were tested against full factorial ANCOVA or GLMs with male age, time of season, and predator presence as factors and total female activity (the sum of the activity of the large and the small female) as a covariate. To remove nonsignificant interaction terms from the models, we used stepwise backward model selection in the statistical software S-PLUS 6.2 following the recommendations from Crawley (2002) for model selections of ANCOVA and GLMs. The model selection was set to always include the main effects of the factors included in the model. Due to a low sample size in the models only including courting males, these models did not include 3- and 4-way interactions. The normality of the models was checked from diagnostic plots of the residuals of the models.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
When testing total preference including all males, we found no discrimination between females of different sizes (Wilcoxon signed-rank test: Z = 0.069, P = 0.945, n = 123). However, when only including males that engaged in courtship behavior, there was a significant preference for the large female (one-sample t-test: t = 2.077, df = 46, P = 0.043). In addition, males courted large females to a greater extent than small females—dance preference: Wilcoxon signed-rank test: Z = 2.738, P = 0.007, only dancing males included, n = 42 and shake preference: Wilcoxon signed-rank test: Z = 2.364, P = 0.018, only shaking males included, n = 20.
When all males were included in the analysis, male age and time of the season had no significant effects on total preference (Table 1). However, there was a significant interaction effect of predator presence and the relative female activity on total male preference (Table 1). To further investigate the nature of this interaction effect, we performed post hoc tests separating replicates with and without a predator present. In the absence of a predator, there was a positive relationship between the time a male spent close to a female and relative activity of that female (linear model: F1,60 = 7.850, r2 = 0.12, P = 0.007; Figure 2). However, when a predator was present, this relationship became negative (linear model: F1,59 = 4.463, r2 = 0.07, P = 0.039; Figure 2).
|
|
When only courting males were included, there were no significant effects of time of the season or predator presence on total preference (Table 1). However, there was a significant interaction effect of male age and relative female activity on courting male preference (Table 1). Again we performed post hoc tests to separate the effect of relative female activity on young and old males. Young males spent more time close to the most active female regardless of her size (linear model: F1,21 = 6.257, r2 = 0.23, P = 0.021; Figure 3), whereas old males did not discriminate between active and nonactive females (linear model: F1,22 = 0.532, r2 = 0.02, P = 0.473; Figure 3).
|
Male dancing preference was not affected by age, time of the season, or predator presence (Table 1). There was a tendency for males to dance with large rather than small females when the large female was more active than the small (Table 1). Furthermore, male shake preference was not affected by any of the factors investigated, even if there was a tendency for males to shake more with large than with small females when the large female was more active than the small (Table 1).
Male willingness to engage in dancing behavior and the time males spent dancing once they decided to do so was not significantly affected by male age, time of season, predator presence, or total female activity (Table 2). Young males tended to dance for a longer period of time than large males (Table 2). Male shake willingness was not significantly affected by any of the factors (Table 2). There was a strong tendency for more young males to shake than old males (young males: 14 of 61 and old males: 6 of 62) and for more males to shake in the absence of the predator than when a predator was present (predator absent: 14 of 62 and predator present: 6 of 61). The only factor significantly affecting male shake frequency was male age (Table 2). Young males shaked significantly more than large males (Table 2). Time of season, predator presence, and total female activity did not affect male shake frequency (Table 2).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
As seen before, males engaging in courtship behavior showed a preference for larger females. This was true both for relative time spent in front of the large female and the relative amount of courtship directed to her. This is in accordance with the results from earlier mate choice studies on this species (Berglund 1991
). However, we did not find a preference for the large female when including all males. This might have been due to a lower sexual interest from noncourting males or from countering effects of the factors included in the experiment (i.e., predator presence, male age, and time of season).
We found no predator-induced reduction of preference for large females for any of our preference measures (total male preference, dancing preference, and shake preference) in spite of this being shown earlier in this population (Berglund 1993; Fuller and Berglund 1996). Moreover, we did not find any differences between males of different ages or at different periods of the breeding season in response to predator presence, that is, individuals did not respond differently to the change in predation risk depending on their RRV, and these results therefore does not support the terminal investment hypothesis.
The lack of a reduction in choosiness as a response to predator risk found here may be explained if the pipefish did not perceive the cod as a potent enough threat. However, we did find a response to the presence of a predator in terms of an interaction effect between relative activity of the females and predator presence on total male preference. In the absence of a predator, the male preferred to spend time close to the most active female, whereas the reverse was true when a predator was present (Figure 1). An active female is more likely to attract attention from the male. An active and displaying female should also be easier to assess and evaluate in a mate choice situation compared with an inactive female. There should thus be an advantage in choosing an active female. However, a high activity might also draw the attention of predators, and choosing a partner that is conspicuous under predation risk can be fatal as it also increases the risk for the choosing individual of being detected (Lima and Dill 1990; Su and Li 2006). This has been shown in guppies, where females in populations under high predation risk chose duller males compared with females in populations under low predation risk (Dill et al. 1999) It has, however, not been shown before in S. typhle.
When analyzing total preference only including courting males, we found an interaction effect of male age and relative female activity. There was a positive relationship between the time young males spent close to the large female and the relative activity of the large female. This indicates that young males that engaged in courtship behavior tended to spend most time close to the most active female, both with and without a predator present. There was no such relationship for the old males. As argued earlier, choosing a conspicuous partner under elevated predation risk could be seen as a risk-prone behavior. Furthermore, we found that young males did shake significantly more often than old males. Shaking is a behavior that considerably reduces crypsis, at least to human observers, and probably makes the fish more easily discovered by predators. Thus, young males behaved more conspicuously than old males and thereby accepted higher risks. In addition, we found a tendency for young males to dance longer than old males and for a higher proportion of young males to be willing to engage in shakes compared with old males. Although these results were not significant, the tendencies all point in the same direction, further supporting our finding that young males behaved more recklessly.
One explanation of this result is provided by the compensation hypothesis: young males, with a lower competitive ability compared with other males and/or lower attractiveness to females, must court more and take larger risks in order to obtain matings. In nature, mate availability is usually not a limiting factor for pipefish males even if there is some evidence suggesting that males may have difficulties finding partners under male-biased sex ratio or at a low frequency also under even sex ratio (Jones et al. 2005). In S. typhle, the male potential reproductive rate is much lower than that of the female (Berglund et al. 1989; Berglund and Rosenqvist 1990). In addition, the adult sex ratio in this population has been shown to be even (Berglund et al. 1986a, 1986b). Combined, these factors cause a female-biased operational sex ratio (Berglund and Rosenqvist 1993), so there is most likely a surplus of females ready to mate at any given time and even a young male should have no problems in finding a partner (Jones et al. 2005).
Because females in this species also show a preference for large males, a young male may be unsuccessful in attracting large females. A higher shake frequency in young males could therefore indicate that these actually need to take larger risks compared with older males in order to get a large, high-quality partner. Young males may thus compensate for their lower attractiveness and competitive ability by an increased shake frequency. Older males, on the other hand, may not need to court as intensively as younger males in order to achieve matings with large females.
Another putative explanation to the higher shake frequency in young males could be derived from a low winter survival for these individuals. A young male with a low chance of surviving would be expected to invest heavily in the current reproductive event and not to save resources to next year. However, the size distribution in this population suggest a relatively high winter survival of 1-year olds, that is, a high winter mortality is unlikely to explain our findings (Berglund 1991).
We conclude that young males, contrary to predictions from the terminal investment hypothesis, take higher risks by courting more than older males and by seeking active partners regardless of predation risk. Young males seem to compensate for their lower attractiveness and/or competitive ability by increasing their level of courtship, and the mating advantages stemming from such risk-prone behavior probably outweighs the costs in terms of a decrease in the RRV. The relative activity of the females was, in addition to female size differences, important for male mate choice in this species. This preference was reversed under predation risk, and males avoided conspicuous females at times of high predation risk.
|
ACKNOWLEDGEMENTS |
|---|
Many thanks to Ronny Höglund, Dominique Mazzi, Camilla Nilsson, Sarah Robinson-Wolrath, and Tanja Viio for field assistance and Åsa Borg, Frode Fossøy, Ingebrigt Uglem, Åslaug Viken and the editor, and 2 anonymous referees for valuable comments on earlier drafts. Funding was obtained from the Swedish Research Council (A.B.) and the Norwegian Research Council (G.R.) and the work was carried out at Kristineberg Marine Research Station and Klubban Biological Station.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Berglund A. Egg competition in a sex-role reversed pipefish: subdominant females trade reproduction for growth. Evolution (1991) 45:770–774.[CrossRef][ISI]
Berglund A. Risky sex: male pipefish mate at random in the presence of a predator. Anim Behav (1993) 46:169–175.[CrossRef][ISI]
Berglund A, Rosenqvist G. Male limitation of female reproductive success in a pipefish: effects of body-size differences. Behav Ecol Sociobiol (1990) 27:129–133.[CrossRef][ISI]
Berglund A, Rosenqvist G. Selective males and ardent females in pipefishes. Behav Ecol Sociobiol (1993) 32:331–336.[ISI]
Berglund A, Rosenqvist G. Male pipefish prefer dominant over attractive females. Behav Ecol (2001a) 12:402–406.
Berglund A, Rosenqvist G. Male pipefish prefer ornamented females. Anim Behav (2001b) 61:345–350.[CrossRef][ISI]
Berglund A, Rosenqvist G, Bernet P. Ornamentation predicts reproductive success in female pipefish. Behav Ecol Sociobiol (1997) 40:145–150.[CrossRef][ISI]
Berglund A, Rosenqvist G, Svensson I. Mate choice, fecundity and sexual dimorphism in two pipefish species (Syngnathidae). Behav Ecol Sociobiol (1986a) 19:301–307.[CrossRef][ISI]
Berglund A, Rosenqvist G, Svensson I. Reversed sex roles and parental energy investment in zygotes of two pipefish (Syngnathidae) species. Mar Ecol Prog Ser (1986b) 29:209–215.
Berglund A, Rosenqvist G, Svensson I. Reproductive success of females limited by males in two pipefish species. Am Nat (1989) 133:506–516.[CrossRef][ISI]
Berglund A, Widemo MS, Rosenqvist G. Sex-role reversal revisited: choosy females and ornamented, competitive males in a pipefish. Behav Ecol (2005) 16:649–655.
Bernet P, Rosenqvist G, Berglund A. Female-female competition affects female ornamentation in the sex-role reversed pipefish Syngnathus typhle. Behaviour (1998) 135:535–550.
Candolin U. Predation risk affects courtship and attractiveness of competing threespine stickleback males. Behav Ecol Sociobiol (1997) 41:81–87.[CrossRef][ISI]
Candolin U. Reproduction under predation risk and the trade-off between current and future reproduction in the threespine stickleback. Proc R Soc Lond B Biol Sci (1998) 265:1171–1175.
Clark CW. Antipredator behavior and the asset-protection principle. Behav Ecol (1994) 5:159–170.
Clutton-Brock TH. Reproductive effort and terminal investment in iteroparous animals. Am Nat (1984) 123:212–229.[CrossRef][ISI]
Clutton-Brock TH, Parker GA. Potential reproductive rates and the operation of sexual selection. Q Rev Biol (1992) 67:437–456.[CrossRef]
Crawley MJ. Statistical computing: an introduction to data analysis using S-Plus. (2002) Chichester (UK): John Wiley & Sons Ltd.
Crowley PH, Travers SE, Linton MC, Cohn SL, Sih A, Sargent RC. Mate density, predation risk, and the seasonal sequence of mate choices: a dynamic game. Am Nat (1991) 137:567–596.[CrossRef][ISI]
Dill LM, Hedrick AV, Fraser A. Male mating strategies under predation risk: do females call the shots? Behav Ecol (1999) 10:452–461.
Forsgren E. Predation risk affects mate choice in a gobiid fish. Am Nat (1992) 140:1041–1049.[CrossRef][ISI]
Forsgren E, Magnhagen C. Conflicting demands in sand gobies: predators influence reproductive behaviour. Behaviour (1993) 126:125–135.
Fuller R, Berglund A. Behavioral responses of a sex-role reversed pipefish to a gradient of perceived predation risk. Behav Ecol (1996) 7:69–75.
Godin J-GJ, Briggs SE. Female mate choice under predation risk in the guppy. Anim Behav (1996) 51:117–130.[CrossRef][ISI]
Gustafsson L, Pärt T. Acceleration of senescence in the collared flycatcher (Ficedula albicollis) by reproductive costs. Nature (1990) 347:279–281.[CrossRef]
Jones AG, Rosenqvist G, Berglund A, Avise JC. The measurement of sexual selection using Bateman's principles: an experimental test in the sex-role-reversed pipefish Syngnathus typhle. Integr Comp Biol (2005) 45:874–884.
Kokko H, Monaghan P. Predicting the direction of sexual selection. Ecol Lett (2001) 4:159–165.[CrossRef][ISI]
Lima SL, Dill LM. Behavioral decisions made under the risk of predation: a review and prospectus. Can J Zool (1990) 68:619–640.
Magnhagen C. Predation risk as a cost of reproduction. Trends Ecol Evol (1991) 6:183–186.[CrossRef]
Magnhagen C, Vestergaard K. Risk taking in relation to reproductive investments and future reproductive opportunities: field experiments on nest-guarding common gobies, Pomatoschistus microps. Behav Ecol (1991) 2:351–359.
Pärt T, Gustafsson L, Moreno J. "Terminal investment" and a sexual conflict in the collared flycatcher (Ficedula albicollis). Am Nat (1992) 140:868–882.[CrossRef][ISI]
Sandvik M, Rosenqvist G, Berglund A. Male and female mate choice affects offspring quality in a sex-role-reversed pipefish. Proc R Soc Lond B Biol Sci (2000) 267:2151–2155.[Medline]
Su KFY, Li D. Female-biased predation risk and its differential effect on the male and female courtship behaviour of jumping spiders. Anim Behav (2006) 71:531–537.[CrossRef][ISI]
Vincent A, Ahnesjö I, Berglund A. Operational sex ratios and behavioural sex differences in a pipefish population. Behav Ecol Sociobiol (1994) 34:435–442.[CrossRef][ISI]
Williams GC. Natural selection, the costs of reproduction, and a refinement of Lack's principle. Am Nat (1966) 100:687–690.[CrossRef][ISI]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||