INTRODUCTION

Many insects are polyandrous but why females mate multiply still remains controversial. Several hypotheses have been proposed to explain the benefits of female remating: (1) genetic benefits—females replace sperm of previous mate with sperm from a genetically superior mate, encourage competition between sperm to ensure fertilization by sperm of high quality, ensure genetic diversity, or avoid genetic incompatibility (Madsen et al., 1992; Zeh and Zeh, 1996; Newcomer et al., 1999; Tregenza and Wedell, 2002; Jennions et al., 2004); (2) sperm replenishment-–females may not obtain sufficient sperm to fertilize the full load of their eggs during their life time from only one mating, or sperm they obtain may degrade during long-time storage (Thornhill and Alcock, 1983; Ridley, 1988; Sheldon, 1994); (3) material benefits-–females remate to acquire male derivate nutrients for higher fecundity (Gwynne, 1984; Wiklund et al., 1993; Wang and Millar, 1997; Wilson et al., 1999; Jiménez-Pérez et al., 2003); and (4) convenience-–females remate to minimize the costs of harassment from males (Thornhill and Alcock, 1983; Parker, 1984; Rowe, 1992). However, these hypotheses need not be mutually exclusive (Simmons, 2001).

Nysius huttoni is a polyandrous seed bug from New Zealand, both sexes of which can mate frequently with the same mates or different mates (Eyles, 1965; QW, unpublished data). Mean mating duration is about 2 h and permanently paired bugs in the laboratory can mate an average of 68 times in their lifetime (Wang and Shi, 2004). However, these bugs are not gregarious in the reproduction season in the field (QW, unpublished data). Therefore, the very high number of matings recorded in the laboratory is less likely to occur in the field. Here we carried out a series of mating treatments and monitored female reproductive outputs to determine what benefits N. huttoni females might obtain from remating. If female remating is for genetic benefits, then females that mate with different partners are expected to produce more viable offspring than those that remate with the same partners (Tregenza and Wedell, 1998; Jennions and Petrie, 2000). Otherwise, females might remate for other purposes, such as to gain more nutrients and/or sperm.

In the present study, we attempted to determine whether diverse partners or mating frequencies played the major role in maximizing female reproductive success in N. huttoni.

MATERIALS AND METHODS

Insects

Insects were collected from twin cress (Coronopus didymus (L.) Sm.) on the campus of Massey University, Palmerston North, New Zealand during January (143 adults and 376 nymphs) and mass-reared in a transparent plastic cylinder (25 cm high by 21 cm diameter) with a 15 cm gauze covered hole in the lid for ventilation, at 20±1°C, 75±10% RH, with a photoperiod of 16:8 h (light: dark) (lighting provided with high frequency broad-spectrum biolux tubes, Osram, Germany). Two 6.5 cm medical cotton wool wicks were provided as oviposition sites. N. huttoni needs an average of 80 d to complete its lifecycle under these conditions (He et al., 2003). We reared these bugs for 80 d before experiments. Food provided for the colony and following experiments consisted of sunflower seeds and stems of twin cress and was replaced every day.

All experiments were carried out under the above conditions. For the experiments we sexed the 5th instar nymphs according to Eyles (1963a) and reared them individually in glass vials (2.5 cm diameter × 8.0 cm high) with a 1.5 cm diameter mesh-covered hole in lids. The date of emergence was recorded for each insect. Because both sexes became sexually mature about 10 d after emergence (He et al., 2003), 13-d-old adults were used for experiments.

Mating and Reproductive Output

To determine whether the number of times females mated and whether remating with the same and different males affected their reproductive outputs, we set up five female treatments: (1) mated once (M-1), (2) mated twice with the same males (SM-2), (3) mated twice with different males (DM-2), (4) mated five times with the same males (SM-5), and (5) permanently paired with the same males (SM-P). For each of treatments (1) to (3), 45 females were used, and for each of the remaining treatments, 22 females were employed. Each female was reared and mated in a glass vial as mentioned above. First mating was allowed when both sexes were 13-d-old. For treatments (1) to (4), males were removed after mating was completed but for treatment (5) females were paired with the same males until death. In the SM-2 treatment, the once-mated sexes were reared separately for 24 h, after which time females were allowed to mate again with the same males, and then females were reared individually until death. For the DM-2 treatment, once-mated females were maintained for 24 h, after which time they were allowed to mate with 14-d-old males that mated once 24 h ago with other females, and then females were reared individually until death. In the SM-5 treatment, females were allowed to mate with the same males five times at 24-h intervals, and then reared individually until death.

A 2.5 cm medical cotton wood bud was provided as an oviposition substrate for each female 24 h after the first mating, and then replaced daily. Eggs laid on each bud were counted and incubated in a Petri dish (1.2 by 8.8 cm) lined with filter paper. Egg hatch was checked and recorded daily. This allowed us to obtain data of overall and daily reproductive outputs for each female.

Statistical Analysis

Mean number of eggs laid (fecundity) and mean hatch rate (fertility rate) in different treatments were analyzed using an analysis of variance (ANOVA). Means were separated using a Tukey's studentized range test. Mean daily fertility rate was analyzed using an analysis of regression (AOR). The slopes of regression lines were then analyzed using an analysis of covariance (ANCOVA). Means were separated using a least squares means test. Because over 90% of females died 35 d after the first oviposition, daily fertility rate data for the first 34 d of the oviposition period were included in the analysis. Overall and daily fertility rate data were arcsin transformed and normalized prior to ANOVA or ANCOVA. Rejection level was set at P > 0.05. All analyses were made using SAS (SAS Institute 1996).

RESULTS

Results show that the total number of eggs laid increased slightly with the number of times N. huttoni females had mated (Fig. 1). However, there was no significant difference between treatments (ANOVA: F=2.25; df=4, 174; P=0.066).

Fig. 1.
figure 1

Mean fecundity of females with different mating status (M-1, females mated once; SM-2, females mated twice to the same males; DM-2, females mated twice to different males; SM-5, females mated five times to the same males; SM-P, females and males paired for life). Bars are SE. Bars with the same letter are not significantly different (ANOVA, P > 0.05).

Full size image
Fig. 2.
figure 2

Mean fertility rate of females with different mating status (M-1, females mated once; SM-2, females mated twice to the same males; DM-2, females mated twice to different males; SM-5, females mated five times to the same males; SM-P, females and males paired for life). Bars are SE. Bars with the same letter are not significantly different (ANOVA, P > 0.05).

Full size image

The proportion of full egg load fertilized in the female lifetime (fertility rate) highly depended on the number of matings in this species (ANOVA: F=61.01; df=4, 174; P < 0.0001) (Fig. 2). Once mated (M-1) females had significantly lower fertility rate than females twice-mated to the same males (SM-2) (P=0.034) or to the different males (DM-2) (P < 0.001). Although the fertility rate in DM-2 was slightly higher than in SM-2, the difference was not significant between these two treatments (P=0.095). Females mated five times to the same males (SM-5) and paired with the same males for life (SM-P) had significantly higher fertility rate than SM-2 and DM-2 females (P < 0.0001). No significant difference in fertility rate was found between SM-5 and SM-P treatments (P=0.393).

In all treatments, daily fertility rates significantly declined over time (Fig. 3A–E). However, the declining rates (slopes) were significantly different between treatments (ANCOVA: F=128.68; df=3, 164; P < 0.0001). Daily fertility rate in M-1 females declined significantly faster than in SM-2 and DM-2 females (P < 0.0001) (Fig. 3A–C). There was no significant difference in regression slopes between SM-2 and DM-2 treatments (P > 0.05) (Fig. 3B and C). Furthermore, in twice-mated treatments daily fertility rates declined significantly faster than in SM-5 and SM-P treatments (P < 0.0001) but the latter treatments were not significantly different (P > 0.05) (Fig. 3B–E).

Fig. 3.
figure 3

Daily fertility rate decline over time.

Full size image

DISCUSSION

Matings are costly, such as the energy costs of sexual behavior and the risks of predation, disease transmission and injury inflicted by the male (Daly, 1978). Therefore, the benefits of multiple matings must outweigh the costs to justify rematings. A number of studies have shown a positive relationship between the number of matings and fecundity in females, suggesting that females acquire nutrients from the male by remating (Kon et al., 1993; Pardo et al., 1995; Wang and Millar, 1997; Wilson et al., 1999; Jiménez-Pérez et al., 2003). However, the present study does not support the material benefit hypothesis because neither the number of matings nor the number of partners females had significantly affected female fecundity in N. huttoni (Fig. 1). Similar results were also reported for other insects (Kraan and Straten, 1988; Svärd and Wiklund, 1988; Ono et al., 1995; Rodríguez, 1998; Kawagoe et al., 2001). In addition, males of this seed bug do not provide food for females to obtain mating. It is suggested that males of these insects do not invest in nutritional substances as a reward for females during mating.

Cases of polyandry for genetic benefits are common in the recent literature (Zeh and Zeh, 1996; Newcomer et al., 1999; Tregenza and Wedell, 2002; Jennions et al., 2004; Neff and Pitcher, 2005). Jennions and Petrie (2000) even suggested that multiple matings by females for purely non-genetic benefits is unlikely as rematings invariably lead to the possibility of genetic benefits. In a previous study, Yang and Wang (2004) found that male genital traits rather than non-sexual traits such as body size and weight were highly selected by N. huttoni females in premating sexual selection. This preference of males with certain genital traits by females implies that to some extent females may obtain genetic benefits such as good genes from remating in this bug. Among various kinds of genetic benefits from female multiple matings with more than one male, offspring viability or hatching success rate is the most obvious (Madsen et al., 1992; Olsson et al., 1994; Tregenza and Wedell, 1998, 2002; Jennions and Petrie, 2000). However, the present study demonstrates that N. huttoni females remated to the same males or to the different males produced similar numbers of viable offspring. This result suggests that females do not obtain genetic benefit from remating in terms of offspring viability. Because how these viable offspring perform in their later life remains unknown, we cannot exclude the possibility that N. huttoni females may obtain other genetic benefits from multiple matings, such as genetic diversity and future reproductive advantage of these offspring.

A previous study showed that one mating could provide enough sperm to fertilize eggs throughout a female's lifespan in N. huttoni although noted that the fertility declined over time (Eyles, 1963b). Our results clearly demonstrate that one mating by females was not sufficient to fertilize the full egg load in N. huttoni. Furthermore, with the increase of the number of matings, whether this be with the same males or different males, the overall fertility rate significantly increased and the daily fertility rate declined significantly slower. It is thus strongly suggested that the replenishment of sperm is a major benefit females obtain from multiple matings in this bug. In his review of the relationship between mating frequency and fertility, Ridley (1988) found that the females of over half of 48 species for which there were data ran out of sperm if not allowed to remate. Thornhill and Alcock (1983), Hunter et al. (1993) and Sheldon (1994) also provided the value of evidence for sperm replenishment by female multiple matings. Our studies indicate that around 10% of eggs laid by N. huttoni females mating multiply failed to hatch. This cannot be attributed to the depletion of sperm. It is thus possible that a proportion of eggs and/or sperm are naturally sterile or unviable in this bug.

Finally, the results indicate that five matings were sufficient for N. huttoni females to maximize their reproductive success (Figs. 2 and 3D and E). Additional matings appear to be superfluous in terms of fertilization. However, the females of this bug could mate as many as 68 times if a male and a female were paired for their lifetime in the laboratory (Wang and Shi, 2004). In a longicorn beetle Nadezhdiella cantori, females could mate on average 46 times per day (Wang and Zeng, 2004) and in some other animal species, females could mate hundreds of times per each clutch produce(see review by Hunter et al., 1993). Such a large number of matings are obviously not necessary for the fertilization of the eggs. Parker (1984) proposed the convenience hypothesis to explain this phenomenon, which suggests that females may mate multiply simply because the costs of repeated harassment from males outweigh the costs of additional matings. N. huttoni males are almost always promiscuous and can mate every day (QW, unpublished data). This property may result in superfluous matings for ‘convenience’ by females of this bug in the laboratory. However, this phenomenon may be less likely to occur in the field as mentioned earlier in the paper.