Rapid female multiple mating in red flour beetles (Tribolium castaneum)

Color profile: Disabled Composite Default screen

888

Rapid female multiple mating in red flour beetles (Tribolium castaneum) Aditi Pai and Guiyun Yan

Abstract: Many female insects mate with multiple males within a single fertile period despite costs such as expenditure of energy and time and contraction of sexually transmitted diseases. In the red flour beetle, Tribolium castaneum, females remate with different males within minutes of the first copulation. If rapid multiple mating is adaptive then multiply mated females should have higher fitness than singly mated females. In this study, we determined the remating frequency of female beetles, characterized female mating behavior, and examined the fitness consequences of female multiple mating. We found that female T. castaneum mated, on average, with 4–6 nonvirgin males within a 1-h observation period. The number of males present in a mating arena did not significantly affect copulation frequency or the intermating interval. However, number of males present significantly affected the length of a single copula as a result of disturbance by rival males when more males were present. Female mating with multiple males in 24 h did not significantly improve egg production, F1-adult production, egg-to-adult viability, fertility retention, and female survivorship. Thus, multiple mating did not enhance long-term female fitness. Polyandrous mating behavior may have evolved through other mechanisms such as fertility assurance and increased offspring genetic diversity or fitness. Résumé : De nombreux insectes femelles s’accouplent avec plusieurs mâles durant une même période de fertilité, en dépit des coûts en dépenses d’énergie et de temps et du risque de contracter des maladies transmises sexuellement. Chez le tribolium rouge de la farine, Tribolium castaneum, les femelles s’accouplent de nouveau avec d’autres mâles dans les minutes qui suivent un premier accouplement. Si les accouplements multiples en succession rapide sont des comportements adaptatifs, les femelles qui s’accouplent à répétition devraient avoir un fitness plus élevé que les femelles qui ne s’accouplent qu’une seule fois. Dans notre étude, nous avons déterminé la fréquence des accouplements multiples chez ces coléoptères femelles, caractérisé le comportement reproducteur des femelles et évalué les conséquences des accouplements multiples des femelles sur le fitness. Les triboliums rouges de la farine femelles s’accouplent en moyenne avec 4–6 mâles non vierges durant une période d’observation de 1 h. Le nombre de mâles présents dans l’aire de reproduction n’affecte pas significativement la fréquence des accouplements, ni la longueur des intervalles entre les accouplements. Cependant, le nombre de mâles présents influence significativement la durée d’un accouplement, à cause des perturbations causées par les mâles rivaux lorsqu’ils sont nombreux. Les accouplements à répétition de femelles avec plusieurs mâles au cours d’une période de 24 h n’améliorent pas significativement la production d’oeufs, ni la production d’adultes de F1, ni la viabilité du stade oeuf au stade adulte, ni le maintien de la fertilité, ni la survie des femelles. Ainsi, les accouplements multiples n’augmentent pas le fitness à long terme des femelles. Le comportement d’accouplement polyandre peut avoir évolué par d’autres mécanismes, tels que l’assurance de la fertilité et l’augmentation de la diversité génétique et du fitness de la descendance. [Traduit par la Rédaction]

Introduction

896 Pai and Yan

Many female insects mate with multiple males within a single fertile period despite potential costs associated with mating such as expenditure of energy and time and risk of contracting sexually transmitted diseases (Drummond 1984; Parker 1984). Two major hypotheses have been proposed to explain the evolution of female multiple mating (Arnqvist and Nilsson 2000; Jennions and Petrie 2000). The first hypothesis is that females benefit directly if mating with multiReceived 11 December 2002. Accepted 9 April 2003. Published on the NRC Research Press Web site at http://cjz.nrc.ca on 30 May 2003. A. Pai1 and G. Yan. Department of Biological Sciences, State University of New York at Buffalo, Buffalo, NY 14260, U.S.A. 1

Corresponding author (e-mail: [email protected]).

Can. J. Zool. 81: 888–896 (2003)

J:\cjz\cjz8105\Z03-070a.vp Tuesday, May 27, 2003 3:41:38 PM

ple males provides (i) sufficient sperm to fertilize all the eggs (Ehrlich and Ehrlich 1978), (ii) nutrients or beneficial chemicals (Boggs and Gilbert 1979; Ridley 1988), (iii) paternal care for the offspring (Nakamura 1998), and (iv) protection from sexual harassment by other males (Watson et al. 1998). There is strong evidence for direct benefits of multiple mating in insect species. For example, female black swallowtail butterflies (Papilio polyxenes asterius) replenish sperm supply from additional matings (Lederhouse 1981), and in several butterfly species females gain nutrients from sperm (Boggs and Gilbert 1979; Oberhauser 1989, 1997; Boggs 1990). The second hypothesis is that female fitness increases indirectly through enhanced fitness of their offspring via good genes, attractiveness genes, compatible genes, or higher genetic diversity (Eberhard 1996; Zeh and Zeh 1996; Jennions and Petrie 2000). For example, good genes improve offspring quality in guppies (Poecilia reticulata; Evans and Magurran 2000), whereas in bumblebees (Bombus terrestris) enhanced genetic diversity improves off-

doi: 10.1139/Z03-070

© 2003 NRC Canada


Pai and Yan

spring quality (Baer and Schmid-Hempel 1999). Madsen et al. (1992) showed that polyandry in Swedish adders (Vipera berus) leads to more viable F1 offspring, and increase in F1-offspring viability is possibly caused by increased sperm competition, which allows “genetically superior” males to fertilize female’s eggs. In Drosophila melanogaster, success of sperm in fertilizing eggs depends on female genotype, thus providing evidence for the importance of genetic compatibility between sperm and eggs (Clark et al. 1999). Both hypotheses for polyandry predict that multiply mated females have a higher fitness than singly mated females. Polyandrous species have been used as model systems to study several aspects of sexual selection such as precopulatory mate choice, postcopulatory paternity-biasing processes (sperm competition and (or) cryptic female choice), or consequences of sexual conflict. One such model system is the red flour beetle, Tribolium castaneum, in which females exhibit a high degree of polyandry. Females engage in continuous mating and remate with different males even within minutes of the first copulation. This species has been widely used to study many aspects of sexual selection such as precopulatory mate choice (Boake 1985, 1986; Lewis and Austad 1994; Lewis and Iannini 1995; Arnaud and Haubruge 1999), postcopulatory processes (Wool and Bergerson 1979; Bloch Qazi et al. 1996, 1998; Edvardsson and Arnqvist 2000), and consequences of polyandry (Bernasconi and Keller 2001; Pai and Yan 2002). However, the extent of female multiple mating in natural populations is unknown and needs to be estimated because the extent of polyandry is an important aspect of sexual selection. Previous studies have demonstrated that mating with multiple males benefits the females by improving the fitness of offspring (Bernasconi and Keller 2001; Pai and Yan 2002) and by increasing progeny production (Lewis and Austad 1994; but see Pai and Yan 2002). However, hypotheses for benefits from multiple mating are not mutually exclusive and other fitness consequences of polyandry are also possible (Jennions and Petrie 2000) and need to be explored to understand maintenance of polyandry in this species. Accordingly, the aims of the present study are to characterize female multiple-mating behavior and to compare long-term consequences of rapid multiple mating. We characterized female mating behaviors when exposed to various numbers of males. We then compared mating frequency and copulation duration in the presence and absence of male–male interactions by exposing females to single and multiple males. Finally, we examined differences between singly and multiply mated females in their fitness components, including long-term fecundity (number of eggs laid), progeny production (number of F1 adults produced), egg-to-adult viability, female fertility retention (the duration that females laid fertile eggs after removal of the males), and female survivorship.

Materials and methods Beetle rearing All experiments used the cSM strain of T. castaneum (for the origin of cSM, see Wade 1977). Beetles were sexed as pupae to ensure adult virginity. Beetles were raised in 8dram shell vials containing -5 g standard medium (95% by weight, fine-sifted, whole-wheat flour and 5% dried, powdered, brewer’s yeast). Unless otherwise stated, experimental

889

vials were maintained in a dark incubator regulated at 29°C and 70% RH. To facilitate identification and separation of male from female beetles, beetle body color was used as a marker. The black body color is determined by a codominant allele, and there is no evidence that the body color affects mate recognition. Red cSM females and black cSM males were used throughout the study. Unless otherwise stated, beetles used to initiate the experiments were virgins and were 2–5 weeks post emergence. Female multiple-mating behavior We determined the magnitude of polyandry by determining the number of males that a female mated within 1 h. Both sexes are known to mate repeatedly (Wojcik 1969). Typically, the male initiates courtship by approaching a female and mounts her from the back, and copulation may last up to 0.5 h (Lewis and Austad 1994; Bloch Qazi et al. 1996). Our protocol for observing beetle mating behaviors followed the method of Arnaud and Haubruge (1999). Briefly, the experiment was conducted using a 5× magnifying glass in a mating arena, a 35 mm diameter plastic petri dish lined with filter paper, and a thin layer of flour under dim light. Because virgin females may be more motivated to copulate than nonvirgin females (Eberhard 1996), and subsequently the magnitude of multiple matings would be overestimated, we used nonvirgin males and females as a precaution. All females and males used in this experiment were previously exposed to beetles of the opposite sex for the same duration (30 min), and therefore had similar reproductive history. A male was introduced into a mating arena with a female. As soon as the pair completed copulation and separated, the male was removed and a fresh male was introduced into the mating arena and the process was repeated for 1 h. In this way, we recorded the exact number of partners females mated within 1 h. This experiment was repeated using the same females but different males 48 h after the first observation. A total of 15 females were used for the two observations. Mating-behavior characterizations Beetle courtship was observed in a mating arena as described above. We allowed beetles to habituate for 30 min before observation and conducted the observation with a 5× magnifying glass in dim light under room temperature and humidity. Individual females were exposed to 1, 2, 4, 8, or 16 males for 30 min. All behavioral observations consistently used red cSM females and black cSM males so that males could be readily distinguished from females. Beetles used in this experiment were virgins. The sex of the individuals that initiated or terminated copulation, frequency and duration of copulation, and presence or absence of disturbance were recorded. Here, we define disturbance as the phenomenon of rival males climbing on top of a copulating pair or butting into the pair. Unlike the previous experiment, the exact number of partners that a female mated with was not determined because individual males could not be distinguished. Each treatment was replicated 10 times (N = 10 × 5 = 50). Measurement of fitness components Fitness assays were conducted for females mated with various numbers of males simultaneously. This experimental © 2003 NRC Canada



890

design has the advantage of permitting precopulatory female mate choice and male–male interactions. The simultaneous multiple-mating experiment enabled virgin females to mate with 1, 2, 4, 8, or 16 virgin males simultaneously for 24 h, thus females could choose from a variety of males. Male intrasexual interactions, which may be very important in natural populations, were allowed. To maintain a similar density of beetles, and hence encounter rate for all treatments, the amount of flour medium in the vial increased with the number of males. In particular, when a female was mated with 2 males, 3 g of flour medium was added, and the amount of flour medium was then increased by 3 g for each successive treatment. Thus, 4-male treatment had 6 g of flour medium and 16-male treatment had 12 g of flour medium. Male beetles were removed from the vial after 24 h, and females were transferred to vials with -5 g fresh flour medium every 6 days until females stopped producing viable eggs for 2 consecutive weeks. Fecundity (the number of eggs) was recorded for as long as females continued to produce eggs and number of deaths was recorded for 192 days after mating, but progeny production was recorded only from the first month after mating. Ten replicates were conducted for each treatment (N = 10 × 5 = 50). Two independent mating experiments were conducted (referred to as SIM1 and SIM2 throughout the text). All beetle cultures were maintained in a dark incubator regulated at 29°C and 70% RH, except in the first month of SIM1 when the humidity could not be regulated. For the females in SIM1, 5 fitness components including fecundity, progeny production, egg-to-adult viability, fertility retention, and female survivorship were examined. Unlike in SIM1, we did not measure fecundity and egg-to-adult viability in SIM2. Our protocol for fitness-component measurement was similar to that described by Yan and Stevens (1995). Briefly, fecundity was determined by counting the number of eggs every 6 days until 2 weeks after a female stopped producing eggs. Progeny production was estimated as the number of F1 pupae and adult offspring produced after eggs were allowed to grow for 5 weeks. Viability of eggs was calculated as the proportion of eggs that survived to adulthood from the progeny-production data. Fertility retention was defined as the number of days since initial mating that a female produced at least one viable offspring. Number of deaths in each treatment was recorded every 6 days. The total number of replicates for each mating treatment was 20 (10 replicates × 2 trials), except that fecundity and viability analysis had 10 replicates (10 replicates × 1 trial). Data analysis To examine how the number of available mates affected female mating frequency, we used an analysis of variance (ANOVA) with the number of males as a factor. Intermating interval was compared using a nonparametric Kruskal–Wallis test because these data did not meet the normality criteria. To compare copulation duration among treatments we used a nested ANOVA with the number of males present and replicate as factors. Replicate was nested within treatment (i.e., number of males present). To further test whether disturbance affected copulation duration, we used a nested ANOVA model with the number of males present and disturbance as factors. Disturbance was nested within the treatment (i.e., number of

Can. J. Zool. Vol. 81, 2003 Fig. 1. Distribution of copulation frequency of nonvirgin female red flour beetles, Tribolium castaneum, with nonvirgin males within a 1-h observation period.

males present). The effect of the number of mates on fecundity and egg-to-adult viability was examined using ANOVA with the number of mates as a factor. To compare F1-adultprogeny production and fertility retention an ANOVA model with the number of mates and trial as factors was used. When necessary, data were transformed to fit the normality criteria. To compare survival of females among different mating treatments we used the Kaplan-Meier survival analysis. All statistical analyses were conducted with JMP (SAS Institute Inc. 1995). Four females (~2%) that did not produce any offspring in the first 4 weeks of the study were discarded from all analyses.

Results Characterization of female mating behaviors Nonvirgin females mated, on average, with 3.8 and 6 males on the first and third day, respectively, and this difference was significant (Student’s t test, t[28] = 3.62, P = 0.004; Fig. 1). In experiments with virgin beetles, males initiated contacts in most cases (81%), but females were usually the first to turn away from copulas (60% cases). Ninety-four percent of females mated at least once and up to a maximum of 11 times within the 30-min observation period. The number of males present did not affect copulation frequency (ANOVA, F[4,42] = 1.46, P = 0.22; Fig. 2A) or the intermating interval (Kruskal–Wallis test, χ42 = 2.61, P = 0.62; Fig. 2B). However, the duration of a single copulation differed significantly among treatments (nested ANOVA, F[4,42] = 2.74, P = 0.03; Fig. 2C). Copulations in a singlemale environment were significantly longer than copulations in environment with 2 or more males (ANOVA orthogonal contrast, t[2] = 2.15, P = 0.03). Presence or absence of disturbance by rival males significantly affected the duration of copulation (nested ANOVA, F[8,163] = 6.12, P < 0.001; Fig. 2D). In the absence of disturbance, copulations were significantly shorter for all treatments except the treatment © 2003 NRC Canada


Fig. 2. Copulation frequency and duration, and interval between copulations when a female T. castaneum had single and multiple mates. Copulation frequency within the 30min observation period (A), average interval between two copulations (B), copulation duration (C), copulation duration in the presence and absence of disturbance (D), and incidence of disturbance (E). Means and standard errors are shown.


Pai and Yan


891

© 2003 NRC Canada


892

Can. J. Zool. Vol. 81, 2003

Fig. 3. Temporal pattern of offspring production in multiple-mating experiments. Average monthly egg production (A) and average monthly F1-adult production (B).

with 4 males (2-male treatment, t[30] = 4.05, P < 0.001; 8male treatment, t[34] = 2.85, P = 0.007; 16-male treatment, t[46] = 2.17, P = 0.03). In general, disturbance increased with the number of males present (Pearson’s test, χ42 = 16.5, P = 0.009; Fig. 2E). Fecundity, F1-adult production, and egg viability Female fecundity and F1-adult production declined significantly over time (fecundity in SIM1: ANOVA, F[4,245] = 65.9, P < 0.001 (Fig. 3A); F1-adult production in SIM2: F[4,225] = 81.4, P < 0.001 (Fig. 3B)). The first month after mating showed the highest egg and F1-adult production (Fig. 3). Number of mates significantly affected long-term fecundity (Table 1). In SIM1, females with 1 or 2 partners produced significantly more eggs than females with 4 or more partners (ANOVA orthogonal contrast, t[1] = 3.26, P = 0.002). However, F1-adult production and egg-to-adult viability were not affected by the number of mates (Table 1). Overall, 36–65% of eggs hatched and survived to adulthood (Table 2). It is not appropriate to compare F1-adult production between SIM1 and SIM2, because this trait was examined only from the first month onwards in SIM1 but was examined for the entire length of the experiment in SIM2. Fertility retention and female survivorship To test whether females that mate multiply retain fertility for a longer time than females that mate singly, fertility retention was compared among females with different numbers of mates by simply comparing the duration for which the female produced viable offspring after initial mating. Because all females in this experiment received sperm on the same day, any difference in the ability to produce viable offspring is due to variation in the number of mates. Comparison of number of days that females remained fertile showed no significant difference among the 5 treatments or between the 2 trials (Tables 1 and 2). Female beetle survivorship did not vary among the treatments (Kaplan-Meier survival analysis, χ42 = 3.56, P = 0.46) or among the two trials (t[94] = 0.36, P = 0.71).

Discussion This study has demonstrated that female T. castaneum are

highly polyandrous. We found that nonvirgin females mated, on average, with 4–6 males within a 1-h observation period. For virgin females, the average number of copulations within 30 min was 3.8 and intermating interval was about 4 min. Virgin females exposed to single and multiple males did not differ in copulation frequency or intermating intervals. Female T. castaneum (both virgin and nonvirgin) mate readily and do not show any refractoriness to remating after first copulation. However, we detected no increase in the 5 fitness correlates (fecundity, F1-adult-offspring production, F1 egg-to-adult viability, fertility retention, and life span) of female T. castaneum, owing to multiple mating. The extent of multiple mating in a single fertile period is highly variable among insects. For example, a queen bee (Apis dorsata) may copulate with up to 35 drones (Moritz et al. 1995, 1996; Oldroyd et al. 1996), whereas a female tiger swallowtail butterfly (Papilio glaucus) mates with a maximum of 4 males in a 2-month period (Lederhouse 1995). Similarly, the interval between the first copulation and the subsequent copulation varies substantially among species. For example, female predatory mites Parasitus fimetorum remate immediately after the first copulation (Yasui 1997a), whereas female bean bugs (Reptortus clavatus) do not remate for 15–20 days after initial mating (Sakurai 1996). The interval between successive copulations may affect the degree of sperm transfer within female tracts, age of competing sperm, and hence the outcome of sperm competition (Yasui 1997b; Simmons and Siva-Jothy 1998; Danielsson and Askenmo 1999; Drnevich et al. 2000). Therefore, based on the interval between successive copulations we distinguish between rapid female multiple mating (e.g., mating within minutes or hours) and periodic multiple mating (e.g., mating within weeks). We make this distinction because the fitness consequences of multiple mating are likely to be different for rapid and periodic multiple mating. This study only determined occurrence and consequences of rapid female multiple mating. The consequences of periodic multiple mating, wherein females had access to males throughout their lifespan, are examined in a separate study (A. Pai and G. Yan, unpublished data). The results from observing mating behavior of females confirm that T. castaneum is a good model system to study polyandry and the resulting postcopulatory processes because © 2003 NRC Canada



Pai and Yan

893 Table 1. Means and standard errors (in parentheses) of fitness components of female red flour beetles, Tribolium castaneum, when mated with multiple males. Trial 1

2

Number of mates 1 2 4 8 16 1 2 4 8 16

Fecundity 300.5 383.2 187.0 235.4 213.1

(54.9) (47.9) (37.3) (41.2) (33.6)

F1-adult-progeny production 96.4 (44.3) 157.8 (42.4) 60.3 (23.5) 90.0 (27.8) 31.3 (9.4) 516.0 (45.8) 403.6 (58.0) 379.9 (82.4) 343.5 (38.2) 435.8 (61.1)

Egg-to-adult viability

Fertility retention (days)

Life-span (days)

0.48 0.65 0.55 0.62 0.36

83.3 88.4 82.5 79.1 70.7 88.2 75.0 75.7 79.6 79.3

163.9 (18.7) 140.2 (17.9) 96.5 (19.1) 124.7 (19.7) 132.7 (17.8) 129 (5) 126 (8) 124.6 (6.3) 134 (0) 128.4 (5.5)

(0.09) (0.08) (0.09) (0.10) (0.07)

(8.6) (5.2) (10.5) (5.4) (5.2) (9.5) (8.8) (9.3) (9.9) (11.9)

Note: F1-adult-progeny production was examined from the first month for trial 1 and the entire length of the experiment for trial 2.

Table 2. Results of analysis of variance on the effect of number of mates on fitness components of female T. castaneum. Fitness component

Source of variation

Fecundity

Number Error Number Trial Number Error Number Error Number Trial Number Error

F1-adult progeny

Egg-to-adult viability Fertility retention

of mates of mates of mates × trial of mates of mates of mates × trial

df 4 45 4 1 4 68 4 41 4 1 4 84

SS 248 785.3 858 121.4 157.43 1 585.1 114.4 1 391.13 0.52 2.81 0.19 0.05 0.22 10.19

F 3.26

P 0.01

1.92 77.48 1.39

0.11