Introduction

Mating is often costly for males due to energy expenditure and sperm depletion (Simmons et al. 2017). When males invest more resources in current reproduction, resources available for future reproductive events decline (Höglund and Sheldon 1998). Thus, mating events throughout the lifetime of an iteroparous species cannot be considered independent, and each mating event is expected to affect subsequent mating behavior (Kokko 1997).

The results of previous studies examining the effects of mating experience on male mating behavior are inconsistent, and vary with the species and experimental methods used. For example, mating experience increased the subsequent male mating success in the European corn borer Ostrinia nubilalis, although mating with the experienced males reduced fitness of females (Milonas et al. 2011). On the other hand, males that mated early in their lives decreased their calling effort in field crickets Teleogryllus commodus (Chiswell et al. 2014). Similarly, male–male fighting efforts may either increase or decrease after mating. Recently mated males with low sperm volume outcompeted those that have not recently mated and possess much larger volume of sperm in the blue crab Callinectes sapidus (Kendall and Wolcott 1999). Judge et al (2010) showed that male field crickets Gryllus pennsylvanicus that encountered females but did not transfer sperm were more likely to win fights with a virgin opponent than males that transferred sperm. The cause of these inconsistencies can be attributed in part to the following four factors: (1) costs of mating behavior (Dewsbury 1982); (2) perception of future mating opportunities (Ortigosa and Rowe 2003; van Son and Thiel 2006); (3) self-perception of attractiveness as a consequence of mating experience (Fawcett and Bleay 2009; Chiswell et al. 2014); and (4) quality of mating partners.

Each of these four factors is expected to have different effects on subsequent male reproductive efforts. For example, the cost of mating will decrease investment in subsequent reproductive effort (Bonduriansky 2001). In contrast, if higher mate encounter rates or self-attractiveness is perceived by a male through prior mating experience, he would invest more in reproductive effort given the higher fitness returns on the investment (Kemp 2006). If males encounter higher quality (e.g., larger fecundity) females, they are expected to invest more resources in their current courtship [e.g., guppy Poecilia reticulata, Dosen and Montgomerie (2004); fiddler crab Uca mjoebergi, Reading and Backwell (2007)] and/or copulation effort (Aumont and Shuker 2018). Accordingly, investment in subsequent mating is expected to decline due to energy and/or sperm depletion.

Additionally, courtship and male–male fighting efforts are the two important components of male reproductive effort (Andersson 1994). Previous studies on the effect of mating experience on reproductive effort focused on single components of the effort, such as either courtship or fighting. However, a trade-off is expected between these two components. The two components are not only positively correlated (i.e., stronger males are considered more attractive by females, e.g., Pizzari and Birkhead 2000), they can also show negative or no correlation (e.g., Shackleton et al. 2005; McGhee et al. 2007; Hedrick and Bunting 2014; Okada et al. 2014). Therefore, these two components of reproductive effort (courtship and fighting efforts) were measured in the field cricket Gryllus bimaculatus De Geer (Orthoptera: Gryllidae) in this study. Virgin male G. bimaculatus did not show a clear preference for the size of females, but mated males invested more effort in courting larger females (Bateman and Fleming 2006). Because of the flexible behavioral plasticity due to mating experience, G. bimaculatus is suitable for examining the relationship between mating experience and subsequent reproductive effort.

The aim of this study was to examine how mating experience influences the two components of male reproductive effort and to distinguish between the relative importance of factors, such as encounter with females, copulation success, and female quality in subsequent mating bouts. Male G. bimaculatus emit calling song to attract females and fight to defend their territories (Alexander 1961; Simmons 1986, 1988a, b). Calling is energetically expensive (Hoback and Wagner 1997), and greater calling activity attracts more females in some field crickets (e.g., Hunt et al. 2004). Therefore, the calling activity can be considered as a courtship effort. The fighting efforts were recorded as the intensity of fighting and fighting outcomes (i.e., win or lose).

I measured the calling activity of males before and after they were allowed to encounter females, and then they were allowed to contest virgin males that had not encountered females. If the cost of mating is a relatively important factor, males that copulated with females would invest less in subsequent reproductive efforts, while if self-perception of attractiveness is a more important factor, copulation success would promote subsequent reproductive efforts. If the perception of mating opportunities is a more important factor, encounters with females would promote subsequent reproductive efforts, irrespective of copulation success. Furthermore, males that mate with larger females (i.e., higher quality females) would decrease their subsequent reproductive efforts.

Materials and methods

Insects

All individuals used in this experiment were second-generation progeny of G. bimaculatus purchased from a local pet store. Approximately 300 adult crickets were reared in plastic breeding containers (58 × 38.5 × 31.5 cm) at room temperature (24–27 °C) under a 14 h:10 h (light:dark) photoperiod (light cycle: 01:00–15:00). The crickets were provided newspapers for shelter, moistened sand in a 500 mL plastic cup as a source of water, and an excess of insect food pellets (Oriental Koubo, Tokyo, Japan). The sand was sprinkled with water every 2–3 days. New adults were collected from the breeding tank within a day of emergence by checking the container daily. Since adult G. bimaculatus males need 4 days to become sexually active (Simmons 1986), this procedure ensured that all experimental individuals were virgins. Adults were individually housed in a 500 mL container (diameter 10 cm) supplied with a cotton-plugged water vial, an excess of food, and cardboard shelters. A sample size of 39 males was used for experiments.

Measurement of courtship effort and mating trials

The virgin males used in the experiments were aged 7–9 days post the final molt. The duration of male calling was measured for 2 h on the day prior to the mating trial, starting at 5 min after the lights had been switched off for the day (i.e., 15:05). Each male was housed in a 500 mL container (diameter 10 cm) supplied with a water vial, food, and a cardboard shelter, and maintained at 24–27 °C. The calling activity was visually checked using the one–zero sampling method. Calling or non-calling was recorded at 1 min intervals, and the total duration of calling was calculated by summing the number of calling events for 2 h. The raising of forewings was recorded as the calling activity. Because I did not check all the sound production, calling activity might be overestimating. However, the males did not exhibit locomotion or foraging with their forewings lifted (T. Kuriwada, personal observation), the raising of the forewings will be able to be considered as a courtship state. The focal crickets were not acoustically isolated from other individuals during monitoring. Male G. bimaculatus are distributed within hearing distance of other males in the field (Simmons 1988a). In field crickets, there are males that do not emit calling song when there are more competitive males around (e.g., Cade 1981). The experimental design can also be used to measure the effect of mating experience on alternative strategies or tactics, such as smaller males being less likely to engage in calling.

A virgin female aged 10–18 days post the final molt was housed with a male for 20 min in a plastic container (18.5 × 10.5 × 15 cm) covered with a damp filter paper, with each female being used only once. All observations started at 13:30. Courtship latency was measured as the time from the start of cohabitation to the initiation of a courtship song by males, which was taken to be indicative of a male’s motivation to mate (i.e., a shorter courtship latency reflects a high motivation to copulate). Males with a higher motivation will invest more reproductive effort, indicating that motivation would affect courtship and fighting efforts. Multiple pairs (ca. 9–10 pairs) were observed simultaneously (i.e., these crickets were visually but not acoustically isolated from the other pairs). The duration of copulation was defined as the time between the attachment of the male and female genitalia and the subsequent disentanglement of the pair. The length of time engaged in copulation was considered a measure of a male’s investment in the female. If the pair did not initiate copulation after 20 min, it was recorded that the male encountered a female but did not copulate. The body weight of the females was measured (to the nearest 0.1 mg) immediately after observations using an electronic balance (Mettler AE50 balance; Mettler-Toledo, Greifensee, Switzerland). The male crickets were returned to their original containers after the mating trial.

After the mating trials, calling effort was measured at 15:05 in the above way. Although there is a post-copulatory refractory period of approximately 5–6 min in G. bimaculatus under the presence of females, the male refractory period was 1 h under the absent of females (Ootsubo and Sakai 1992; Hall et al. 2000). Because the males were isolated from the females after the mating trial, the males were expected to resume their investment in subsequent reproductive efforts approximately 1 h after mating. Accordingly, calling effort was measured at least 1 h after mating trial in the present study. Changes in calling effort were calculated by subtracting the duration of calling recorded prior to mating from that recorded after mating.

Measurement of male–male fighting behavior

At 4.5 h after the mating treatment, each focal male was randomly paired with a rival male that had not encountered females. The rival males were used only once for the experiment. Males were marked on the pronotum with a dot of yellow paint (Magic Opaque Color; Teranishi, Osaka, Japan) to enable identification during the trial. Marking was randomly assigned to either a focal or rival male. Encounters between pairs of males were examined only once. Each pair of males was introduced into a plastic arena (18.5 × 10.5 × 15 cm) containing a clean filter paper and maintained at 25–26 °C. Transfer of crickets was conducted as gently as possible to minimize handling disturbance, and it was observed that males generally started fighting immediately after being housed together. The winners of these encounters were easily identified, as they tremulated, stridulated, and often chased the loser (Alexander 1961; Hofmann and Schildberger 2001). The level of aggression (contest intensity) was recorded using a categorical scale of aggression (modified from Hofmann and Schildberger 2001), where 0 = mutual avoidance, 1 = immediate dominance, 2 = antenna fencing and aggressive song, 3 = bilateral maxillae/mandible spreading, and 4 = grappling (including maxillae/mandible engagement). Body weight of the focal and opponent males was measured (to the nearest 0.1 mg) using an electronic balance immediately after observation. Since the outcome of aggressive encounters between rival males is dependent not only on reproductive effort but also on other factors such as the winner–loser effect, the outcome per se may not be directly indicative of reproductive effort. However, in this regard, possible experimental conditions such as fighting experience and social and nutritional status were controlled to reflect the reproductive effort associated with the outcome of fighting.

Statistical analyses

All statistical analyses were performed using R version 3.4.2 (R Core Team 2017). A generalized linear model (GLM) with binomial error distribution and logit link function was used (“glm” function) to examine the body weight of both sexes, male age, and the influence of courtship latency on copulation success (i.e., whether copulation occurred or not). A GLM with a gamma error structure and log link function was used to determine whether the duration of copulation was affected by the body weight of both sexes, male age, and courtship latency. The duration of calling before and after a mating trial was compared using Welch’s paired t test to determine whether mating experience increased male courtship efforts. In the following analyses, I assumed that males assessed the quality of females based on the weight difference between the females they encountered and themselves, and added the weight difference between males and females (i.e., female weight minus male weight) to the model instead of the weight of the females themselves. A linear mixed model (LMM) using the lmer function in the lme4 package (version 1.1.14) (Bates et al. 2015) was applied to examine the effects of explanatory variables (see Table 2). Since the courtship efforts of some males were measured simultaneously (i.e., the males were not acoustically isolated), the males that were measured on the same day were treated as a single group; therefore, the day of the experiment was treated as a random factor in the analysis. The interaction terms (i.e., copulation success × body weight) were also included in the analysis. An ordinal multinomial logistic regression was performed using the polr function in the MASS package (Venables and Ripley 2002) to examine the effects of mating experience on the intensity of male–male contests. A GLM with a binomial error structure and a logit link function was used to examine the effect of mating experience on fighting success (i.e., whether focal males won or not). The explanatory variables in the two models for analysis of contest outcomes were the differences in body weight between the fighting males (relative body size difference between males: focal male weight minus rival male weight), difference in body weight between sexes, male age, courtship latency, and copulation success. The change in calling effort was also added as an explanatory variable to examine whether courtship effort affects the outcome of fighting. The interaction terms between copulation success and the body weights were also included in the models. The likelihood ratio test was used to examine the statistical significance of each coefficient in the models.

Results

The mating-related behaviors of 39 male crickets were examined in this study. The interaction terms did not significantly affect the difference in courtship effort (male weight × copulation success P = 0.885; weight difference between sexes × copulation success P = 0.763) or fighting outcome (intensity of fighting: weight difference between males × copulation success P = 0.863; weight difference between sexes × copulation success P = 0.477; fighting success: weight difference between males × copulation success, P = 0.187; weight difference between sexes × copulation success, P = 0.412). Therefore, the interaction terms were excluded from subsequent analyses.

Nineteen of the 39 males copulated with females within 20 min of pairing in the mating trials. Males that copulated with females exhibited shorter courtship latency (mean ± standard deviation: 122 ± 120 s, N = 19) than those that did not copulate (339 ± 400 s, N = 20, Table 1), and heavier females were more likely to copulate within 20 min (Table 1; Fig. 1). The mean duration of copulation (± standard deviation) was 50 ± 22 s, and it was found that the time engaged in copulation was not significantly affected by courtship latency (β = 0.0564 ± 0.0490 [SE], χ2 = 1.714, P = 0.191), male weight (β = 0.124 ± 0.0931, χ2 = 2.263, P = 0.133), female weight (β = − 0.0417 ± 0.0384, χ2 = 1.537, P = 0.215), or male age (β = 8.700 ± 8.272, χ2 = 1.445, P = 0.229].

Table 1 Results of a generalized linear model used to determine the effects of explanatory variables on copulation success
Full size table
Fig. 1
figure 1

Body weight of females that did or did not copulate with males during mating trials. The boxes represent the lower (25%) and upper (75%) quartiles, the solid line is the median, and the whiskers indicate 1.5 times the interquartile range

Full size image

Calling duration was found to increase following mating (mean ± SD of pre-mating: 12 ± 12 min, post-mating: 22 ± 19 min, t64.74 = 3.004, P = 0.004). Male age, male weight, copulation success, and courtship latency had no significant effects on the extent of changes in calling duration (Table 2). However, the males increased their calling duration when he encountered a relatively large female (Fig. 2; Table 2).

Table 2 Results of a linear mixed model used to determine the effects of explanatory variables on changes in calling effort pre- and post-mating
Full size table
Fig. 2
figure 2

Relationships between changes in the duration of calling before and after mating trials (i.e., the duration of calling after a mating trial minus that before the trial) and difference in body weight between males and mating partners. The line was fitted based on the results of a linear mixed model. Gray circles and lines indicate pairs that did not mate within 20 min, and black circles and lines indicate pairs that did mate within 20 min

Full size image

Two males were excluded from the analysis of fighting behavior owing to a lack of aggressive behavior during their encounters. Copulation success, male age, differences in body weight between males, change in calling effort before and after mating, and courtship latency did not affect success in male–male contests (Table 3). Notably, males that had encountered relatively heavier females in mating trials were more likely to lose a subsequent aggressive contest with a rival male (Fig. 3; Table 3). Nevertheless, copulation success per se did not significantly affect fighting success (Table 3). None of the explanatory variables assessed had an influence on the intensity of fighting (Table 4). None of the experimental males lost their limbs during fights, irrespective of the intensity of fighting.

Table 3 Results of a generalized linear model used to determine the effects of explanatory variables on fighting success (i.e., win = 1, loss = 0)
Full size table
Fig. 3
figure 3

Effect of the weight difference between a previously encountered female on the fighting outcome (i.e., win or loss) of an experimental male in a subsequent male–male aggressive contest. Boxplot description is given in the caption for Fig. 1

Full size image
Table 4 Results of ordinal multinomial logistic regression used to determine the effects of explanatory variables on intensity of fighting behavior
Full size table

Discussion

The present study indicates that encounters with females tend to enhance subsequent male courtship efforts irrespective of copulation success. The results demonstrate that the perception of future mating opportunities is relatively more important in determining male courtship effort than either of the other two factors assessed (i.e., cost of mating and self-perception of attractiveness). Furthermore, encountering relatively heavier females (i.e., higher quality females) increased a male’s subsequent courtship efforts. The results suggest that encountering high-quality females will increase the expectation of future encounters with such females, and this may have the effect of promoting investment in subsequent courtship efforts.

Copulation success did not affect male fighting effort. The result suggests that the perception of future mating opportunity and self-perception of attractiveness do not have significant effects on the fighting effort. On the other hand, males encountered with heavier females tended to lose subsequent male–male contests. There was a positive correlation between copulation success and female weight. If high copulation success is associated with a high intensity of male courtship, such males would incur high energy expenditure, and may be at a disadvantage in subsequent male–male contest due to depletion of energy. Although sperm expenditure was not examined in G. bimaculatus in the present study, a previous report showed that male house crickets Acheta domesticus transfer larger amounts of sperm when they encounter larger females (Gage and Barnard 1996). In addition, courtship latency can be considered as indicator of male motivation to mate and a higher level of motivation may lead to increased male courtship effort because males with a shorter courtship latency achieved higher copulation success. However, the courtship latency, copulation duration, and the difference in calling effort before and after encountering females had no significant effects on fighting success. Therefore, there was no significant relationship between male investment and fighting success. It is important to note that copulation duration may not be a good indicator of male investment because there is no evidence to indicate that longer copulation is related to greater sperm transfer.

The calls of male field crickets serve as both an agonistic signal towards rival males and a courtship signal towards females (Alexander 1961; Simmons 1988a; Zuk and Simmons 1997). Moreover, it has been demonstrated that less competitive males are drawn to the songs of their more competitive counterparts (i.e., satellite behavior; Cade 1981). In the present study, crickets were not acoustically isolated from other individuals during mating trials or during the measurement of male calling effort. Consequently, the calling effort of focal males may have been affected by the calling activities of nearby audible males. Since male body weight did not affect change in calling duration, competitive interaction through acoustic signals is not likely to be strong in the present conditions. Acoustic conditions should be manipulated in future experiments to assess this possibility. For example, if the experiment with different ambient calling levels (silent, medium, and high) is conducted, the mating effects on alternative behaviors would be detected more clearly.

Previous studies have compared the behavior of individuals in mating and non-mating treatment groups to examine the effect of mating experience on mating behavior (Judge et al. 2010; Pérez-Staples et al. 2010). However, both female and male conditions varied within the mating treatment groups. The present study indicates that encountering and the quality of females had notable influences on subsequent reproductive efforts. Mating and fighting strategies and tactics of males will flexibly modified depending on previous mating partners.