The Mating Systems of Three Species of Minute Pirate Bug, Orius Sauteri, O

Home , Coelopa frigida, Mating system, Traumatic insemination

DOI: 10.1111/eea.12740 The mating systems of three species of minute pirate bug, Orius sauteri, O. minutus,andO. strigicollis Toru Arakawa, Kiyoko Taniai & Taro Maeda* Insect Interaction Research Unit, National Institute of Agrobiological Sciences, NARO, 1-2 Ohwashi, Tsukuba, Ibaraki 305-8634, Japan Accepted: 13 August 2018

Key words: polygamous, monandrous, natural enemies, behaviour, morphology, genitalia, SEM, biocontrol agents, biological control, polygamy, Heteroptera, Anthocoridae

Abstract We compared the mating systems and genital morphology of three Orius species dominant in Japan: Orius sauteri (Poppius), Orius minutus (L.), and Orius strigicollis (Poppius) (Heteroptera: Anthocori- dae). In all three species, the males were polygamous and could inseminate at least three females at a rate of one female per day. Compared to the other two species, the O. minutus males had a lower rate of success in three consecutive mating orders. The mated females laid the same number of eggs irrespective of mating sequence. Orius minutus and O. strigicollis females were found to be monandrous, and they re-mated with another male only when the 1st mating failed. In contrast, O. sauteri females were polyandrous and accepted a 2nd male even when the 1st mating was successful. Multiple mating with a single male did not increase their fecundity, and it decreased the hatching success of eggs. Mat- ing with multiple males did not affect the fecundity or hatching success. Our scanning electron microscopy observation suggested that the genitalia of O. minutus and O. strigicollis were more similar in shape. The similarities in mating system and morphology between O. minutus and O. strigicollis support a close phylogenetic relationship.

polyandry (Fisher et al., 2013), in which ‘sexy sperm’ will Introduction be inherited (Yasui, 1998). Under different conditions, Most studies of mating systems of insect natural enemies polyandry could be chosen. If it is costly to reject sexual have been conducted to understand either the evolution of harassment by males, females accept males’ attempts; this the mating systems themselves or the evolution of sex allo- is called convenience polyandry (Arnqvist & Nilsson, cation decisions (Hardy et al., 2005). The evolutionary 2000; Rowe & Arnqvist, 2002). consequences of polyandry are of interest, as it is not yet Minute pirate bugs (which make up the genus Orius) established why females would mate with multiple males are effective predators of important agricultural pests (Pizzari & Wedell, 2013). Mating involves various costs, including thrips, spider mites, aphids, and whiteflies (Lat- such as time and energy, predation risk, physical injury, tin, 1999). Many studies of Orius species have focused on and parasite/pathogen infection (Arnqvist & Nilsson, their foraging and reproductive ability because of their 2000). If the benefit of re-mating exceeds these costs, importance in biological control. In contrast, there have females should favour polyandry (Arnqvist & Nilsson, been only a few studies on the mating system of Orius. 2000; Arnqvist & Andres, 2006). Non-genetic benefits of Leon-Beck & Coll (2009) revealed that Orius laevigatus re-mating, among other benefits, are adequate sperm (Fieber) females are monandrous and males of the species (Elzinga et al., 2011), direct benefit from ejaculate compo- are polygamous. In contrast, Kobayashi & Osakabe (2009) nents (Paukku & Kotiaho, 2005; Boulton & Shuker, 2013), demonstrated that Orius sauteri (Poppius), Orius strigicol- nuptial gifts (Gwynne, 2008), and fertilization success. lis (Poppius), and Orius nagaii Yasunaga females displayed Genetic benefits are mating with a ‘good’ male and getting multiple mating with several males. If the genus Orius ‘good’ genes (Tregenza & Wedell, 1998, 2000; Kuijper includes both polyandrous and monandrous species, it is et al., 2012). Sperm competition could occur in true worthwhile to compare the mating systems of closely related species in order to help clarify the evolution of *Correspondence: E-mail: [email protected] polyandry.

Horton & Lewis (2011) suggested that Orius species O. minutus demonstrate traumatic insemination in which males use external genitalia to pierce and penetrate the female abdominal wall and ejaculate into the abdominal cavity during copulation. Although females’ defensive mechanisms have evolved, the mating costs in species with trau- O. strigicollis matic insemination are slightly higher compared to those of other species (Reinhardt et al., 2014). Indeed, physical damage can reduce longevity and reproductive success (Stutt & Siva-Jothy, 2001; Morrow & Arnqvist, 2003; O. sauteri Backhouse et al., 2012; Tatarnic & Cassis, 2013; Reinhardt et al., 2014). In discussions of the mating system of Orius spp., sexual conflict caused by traumatic insemination should be considered. On mainland Japan, O. sauteri, Orius minutus (L.), O. O. nagaii strigicollis,andO. nagaii are commonly distributed in a Figure 1 The phylogenetic relationships of four Orius species that wide area across Japan (Yasunaga, 1997a,b,c). These spe- are commonly distributed in a wide area across Japan. cies often occur sympatrically on various flowering plants (Ohno & Takemoto, 1997; Hinomoto et al., 2009) and have similar prey, including thrips, aphids, and hatchlings females in the 3rd experiment to determine whether the and larvae of various insects. Three of these species – O. females are monandrous or polyandrous. Notably, even if sauteri, O. minutus,andO. strigicollis – display good females show multiple mating with multiple males, it does potential as biological control agents (Ohno & Takemoto, not always mean they are polyandrous. If subsequent cop- 1997). A previous investigation of the distribution of these ulation occurs only when the previous copulation failed, three indigenous species in Japan revealed that two or all this is considered a form of monandry and is referred to as three species often coexisted on the same plants (Hino- ‘pseudopolyandry’ (Fisher et al., 2013). To distinguish moto et al., 2009). In addition to their similar habitat, the pseudo- from true polyandry, we checked whether the 1st three species’ prey preference, body size, and appearance mating was successful. We conducted the 4th experiment are very similar. to determine whether mating has non-genetic benefits. As The authors of a phylogenetic study suggested that O. multiple mating with a single male does not relate to sauteri, O. minutus,andO. strigicollis are closely related, genetic benefits, we allowed females to mate multiple times and that the relationship between O. minutus and O. strigi- with single males. In the 5th experiment, we compared the collis is closer than that between either species and O. sau- fecundity and hatching success of females mated with a teri (Muraji et al., 2000a,b; Figure 1). Morphological single male or three males. We discussed the genetic bene- observations of these three species (Yasunaga, 1997b) fits of polyandry based on the results of experiments 4 and show good agreement with the phylogenetic trees (Muraji 5. For the 6th experiment, we conducted a detailed scan- et al., 2000a). We were intrigued by the question whether ning electron microscopy (SEM) comparison of the male the mating systems of these Orius species would also coin- genitalia and mating systems of O. sauteri, O. minutus,and cide with their phylogenetic and morphological traits. It is O. strigicollis. known that genital structures vary widely in the family Anthocoridae, and we speculated that differences in genital Materials and methods structures among O. sauteri, O. minutus,andO. strigicollis might be involved in the differences in their mating beha- Insects viours and mating systems. Orius sauteri and O. minutus were collected from red clo- We conducted six experiments with the three species. In ver and chestnut, respectively, in Ibaraki, Japan. Orius the 1st experiment, we observed the species’ mating beha- strigicollis was obtained from a commercial source (Arysta viour and recorded the duration of copula and the number Life Science, Tokyo, Japan). The insects were maintained of eggs to observe what constituted effective copulation for in a plastic case (24 9 17 9 5cm)usingEphestia kueh- fertilization. In the 2nd experiment, we focused on the niella Zeller eggs (Ga-Ran; Agrisect, Inashiki, Ibaraki, mating capability of the males; we counted how many Japan) as a diet source and Sedum mexicanum Britt. twigs females a male could inseminate and how many eggs were as oviposition substrates under 25 Æ 1 °C, 60–70% r.h., deposited. We investigated mating frequency of the and a L13:D11 photocycle. A mesh sheet (15 9 10 cm) Mating system of three minute pirate bugs 143 was put on the bottom of the plastic case as a shelter for eggs deposited was assessed. Unlike most Heteroptera, the immature bugs. A plastic tube (1 cm diameter, 7.5 cm Anthocoridae (to which the Orius species belong) have a long) filled with water and plugged with cotton wool was unique fertilization system, in which fertilization takes provided as a moisture source. To obtain similarly aged place before the chorion is deposited (Horton, 2008). adult virgin bugs for the experiments, 4th or 5th instars Oocyte maturation does not occur in unmated females. were picked up from the plastic case and isolated individu- Unfertilized eggs generally degenerate and are resorbed ally in ventilated vials (1 cm diameter, 7.5 cm long) with before they reach maturation, and thus, unmated females E. kuehniella eggs and S. mexicanum twigs until they deposit few or no eggs (Horton, 2008). In preliminary reached adulthood. Ephestia kuehniella eggs were glued to experiments, we had confirmed that the virgin females of a piece of adhesive paper (3 9 6mm). three Orius species never oviposited. Fertilization success of many insects may be estimated as the proportion of Experiment 1: Duration of copula required for fertilization. eggs laid that hatched or the presence of sperm in the To elucidate the effective mating time, we investigated the spermatheca(e). In this study, on the basis of the unique mating duration and the insemination of the females. A fertilization system of Anthocoridae, we estimated the virgin female and a virgin male were placed together in a insemination success by counting the eggs. clear plastic tube (0.5 ml), and their mating behaviour was We performed a generalized linear model (GLM) analy- observed under fluorescent lighting. All of the adults used sis to identify significant factors affecting the number of in the bioassays were 3–5 days old. In preliminary eggs, using JMP v.8.0.2 software (SAS Institute, Cary, NC, experiments, we had confirmed by light microscopy that USA). The model assumed normal distribution (fam- copulation occurred under these conditions. On the basis ily = Gaussian, link = identity). of those observations, we defined the start of copula as the point when a mounted male curled his abdomen and the Experiment 3: Receptivity of the females. To investigate female stopped walking. The end of copula was recorded whether females are monandrous or polyandrous, we when the male’s abdomen parted from the female’s provided individual females with two successive mating abdomen. In cases in which the mating was not initiated opportunities. All of the adults were 3–5daysold.First, within 15 min, the observation was stopped. Males were the receptivity of virgin females was tested in a clear plastic removed soon after the mating, and thus the females tube (0.5 ml) as described above. When a male did not mated only once with a single male. The mated females attempt to mate, it was excluded from the data. The mated were kept individually in plastic tubes (1 cm diameter, females were kept individually soon after the mating in a 7.5 cm long) with the food source and the oviposition plastic tube (1 cm diameter, 7.5 cm long) with the food substrate described above. We confirmed the successful source and the oviposition substrate for 5 days, and all insemination by counting the eggs laid in the 7 days after eggs were checked. The female was then provided access to oviposition started or by counting the matured eggs in the a 2nd male under the same conditions as those used for ovary via dissection. No virgin females of these species had the 1st male. The females that had not laid any eggs were any matured eggs. dissected soon after the 2nd mating assay to confirm that they had no eggs. In cases in which a female did not accept Experiment 2: Insemination capacity of the males. A male the mating attempts of a male for 15 min, the female was was brought together with three individual females classified as refusing to mate. consecutively, and we checked the occurrence of successful copulation by direct observation. All the adults used in the Experiment 4: Multiple mating with a single male. To bioassays were 3–20 days old. By using adults of various evaluate the effects of multiple mating with a single male ages, we also investigated the effect of age on insemination on the fecundity and hatching success, we kept O. sauteri capacity. An adult virgin male bug was put into a plastic females with a single male for 3 or 10 days. As we used a tube (1 cm diameter, 7.5 cm long) containing diet eggs single male for each female, the results did not include and an S. mexicanum twig together with an adult virgin genetic benefit caused by polyandry. The total number of female (1st female), and the pair was left for 24 h under eggs deposited by these females and their hatching success conditions of 25 Æ 1 °C, 60–70% r.h., and a L13:D11 was examined. In this approach, the frequency of the photocycle. The female was then replaced with a new mating was not determined, but in our preliminary virgin specimen twice, that is, with the 2nd and 3rd observations we confirmed that the males attempted to females. The 1st, 2nd, and 3rd females were individually mate with the same females again and again in a plastic introduced into new plastic tubes with diet eggs and an tube. Therefore, the results of this experiment may include oviposition substrate twig. Seven days later, the number of the effect of the non-genetic benefit by multiple mating 144 Arakawa et al. and the cost of sexual harassment. The females of O. then fixed with 2.5% glutaraldehyde at room temperature minutus and O. strigicollis did not accept additional for 2 h. After being washed again with PBS 39,the mating attempts, and thus this experiment was not samples were dehydrated in a series of EtOH. The EtOH conducted for these two species. was exchanged with tert-butyl alcohol, and the samples were frozen in a refrigerator. We freeze-dried the samples Experiment 5: Multiple mating with three males. To using a freeze-drying machine (JFD-380P; JEOL, Tokyo, evaluate the effects of polyandry on the fecundity and Japan) and stuck them on an aluminium plate. The hatching success of O. sauteri, we allowed females to mate samples were sputter-coated with platinum using an SC- either once or 39 with different males. This experiment 701 Quick Coater (Sanyu Denshi, Tokyo, Japan). The design includes effects of both the number of males observations were conducted using a JSM-6301F SEM (monandry vs. polyandry) and the mating frequency (JEOL), and the analysis was conducted with digital image (once vs. 39 matings). To clarify the precise effects of analysis software (SEM Afore; JEOL). polyandry, further experiment is necessary as conducted by Tregenza & Wedell (1998) or Dunn et al. (2005). In Results this study, we will discuss the effects of polyandry by considering mating frequency in experiment 4. For Experiment 1: Duration of copula required for egg fertilization example, if no effects of mating frequency will be found in We observed that each Orius male 1st mounted a female, experiment 4, the results of experiment 5 could be then curled its abdomen underneath hers from the right discussed in terms of polyandry. The mating was side (Figure 2A and B). The mounted female stopped conducted with a virgin male in a plastic tube (0.5 ml) to walking, and then the male inserted his genitalia. Some confirm the mating. After the 1st mating, all females were females escaped from males by walking or struggling. Even kept individually in plastic tubes (1 cm diameter, 7.5 cm after being mounted by males, females sometimes rejected long) containing diet eggs and an S. mexicanum twig. The the copulation by raising their ovipositor (Figure 2D). diet and the twig were replaced every 3–4 days, and the When females accepted the copulation, they did not raise eggs laid were counted. The 2nd and 3rd matings were their ovipositors (Figure 2C). conducted by a similar procedure at 7-day intervals (on The duration of copula of O. sauteri ranged from 22 to days 7 and 14 after the 1st mating) with virgin males. The 1 162 s (Figure 3A), and 38.6% (27/70) of the females fecundity of these females and the hatching success were deposited eggs. We found no relationship between the examined. duration of copula and fertilization success (logistic regression analysis: likelihood ratio v2 = 1.53, P>0.05; Experiment 6: Scanning electron microscopy of male n = 64). The shortest duration of copula resulting in genitalia. Male genitalia were detached in 70% EtOH and oviposition (hereafter the minimum effective time) was washed with phosphate-buffered saline (PBS) 39,and 219 s.

Ovipositor

Figure 2 (A,B) Mating position of three Orius species. (C) The ovipositor usually ﬁts to the end of the abdomen, but (D) it was raised when the female rejected the BD mating attempt. Mating system of three minute pirate bugs 145

A O. sauteri oviposited. The minimum effective time was 273 s. The 35 n = 64 longer the copula duration, the higher the fertilization suc- v2 = = = 30 cess ( 5.58, P 0.018; n 130). The linear regression analysis showed that there was no 25

days correlation between the duration of copula and the number 20 of eggs for any of the three species: O. sauteri, R2 = 0.054, 15 2 F1,24 = 1.383, P = 0.25; O. minutus, R = 0.021, F1,20 = 2 10 0.428, P = 0.52; O. strigicollis,R = 0.001, F1,57 = 0.039,

No. eggs/7 5 P = 0.84. 0 Experiment 2: Insemination capacity of males 0 200 400 600 800 1000 1200 For O. sauteri and O. strigicollis, approximately 60–70% of B O. minutus the males successively mated with three females, and 45 n = 78 approximately 25% of the males mated twice (pie charts in 40 Figure 4A,C). Mating frequency in O. minutus was signiﬁ- v2 = = < ys 35 cantly different ( 28.92, d.f. 6, P 0.001; pie chart in 30 Figure 4B). The general linear model including mating 25 sequence, female age, and male age indicated that the 20 number of eggs did not differ, irrespective of the mating 15 sequence in any of the species (GLM: P>0.05; Figure 4, bar 10

No. eggs/7 da diagrams). In addition, the age of the males and females 5 did not affect the number of eggs (GLM: P>0.05). 0 0 100 200 300 400 500 600 700 800 Experiment 3: Receptivity of the females Mating success was calculated as the percentage of females C O. strigicollis whose mating time was longer than the minimum effective 25 n = 130 time (Figure 3, dotted line). The mating success of virgin females of O. sauteri was >60% (Figure 5A). For the 2nd 20 mating attempt, almost 50% of the females mated again 15 (36/77), but the mating success was signiﬁcantly different between non-inseminated and inseminated females. Non- 10 inseminated females had the same mating success as virgin eggs in ovary 5 females. In contrast, inseminated females had lower mat-

No. ing success (Bonferroni multiple comparison with Fis- 0 cher’s exact probability test: P<0.05). 0 500 1000 1500 2000 2500 Virgin females of O. minutus displayed high mating suc- Mating time (s) cess (Figure 5B). The 2nd mating success was approximately 40% (21/51 females). All mated females in the 2nd Figure 3 Relationship between the duration (s) of copula and the mating attempt were inseminated, and their mating suc- number of eggs laid by (A) Orius sauteri,(B)Orius minutus,and cess did not differ from that of the virgin females (Fischer’s (C) Orius strigicollis. The dashed line indicates the minimum exact probability test: P>0.05). No inseminated females effective time, deﬁned as the shortest copulation duration of the accepted additive mating attempts. The response of inseminated females. O. strigicollis females was the same as that of O. minutus females (Figure 5C). The mating success was high for both Orius minutus copulas lasted 16–693 s, and 28.6% (20/ virgin and non-inseminated females. On the other hand, 70) of females laid eggs (Figure 3B). Our logistic regres- inseminated females never re-mated. The total 2nd mating sion analysis showed that a longer duration of copula success was 19.6% (10/51 females). resulted in an increase in the number of females inseminated (v2 = 9.79, P = 0.002; n = 78). The minimum Experiment 4: Multiple mating with a single male effective time for O. minutus was 124 s. When females were kept with a male for 3 days, 10 of 13 For O. strigicollis, copula duration ranged from 23 to females oviposited. With the 10-day treatment, six of seven 2 239 s (Figure 3C), and 61.4% (35/57) of the females females oviposited. The total number of eggs did not differ 146 Arakawa et al.

A O. sauteri 2.8% 35 34 25 13

s 40 25.0% 30 NS 69.4% 20

n = 36 10

No. eggs/7 day 0 1st 2nd 3rd virgin B O. minutus 49 34 12 23 40 22.6% NS 28.3% 30 20 41.5%41.5% n=53 eggs/7 days 10

No. 0 1st 2nd 3rd virgin C O. strigicollis 38 36 25 35 4.9% 40 NS

30 Figure 4 Mating frequency of males (pie 61.0% 26.8% + 61.0% chart) and mean ( SE) number of eggs

gs/7 days 20 (bar graphs) deposited by the 1st, 2nd, and n = 41 10 3rd females mated with a male and that of virgin females for (A) Orius sauteri,(B) 0 1 2 3 No. eg 0 Orius minutus, and (C) Orius strigicollis. No. inseminated females 1st 2nd 3rd virgin Numbers capping the bars indicate the Mating order numbers of females tested. between the 3- and 10-day treatments (t-test: t = 0.776, unclear. The flagellum slides into the female’s abdominal d.f. = 14, P>0.05; Figure 6). The hatching success of the intersegment to inject spermatozoa. The ventral side of the 10-day treatment was lower than that of the 3-day treat- cones was smooth and flat in all three species (Figure 8). The ment (t = 2.664, d.f. = 13, P<0.05; Figure 6). However, flagellum of O. sauteri was relatively straight and short, and the number of hatched larvae did not differ between the spiralled strongly near its base (Figure 8A), whereas the flag- 3- and 10-day treatments (t = 0.842, d.f. = 13, P>0.05). ella of O. minutus and O. strigicollis were more elongated, bent, and weakly spiralled (Figure 8B and C). In O. sauteri Experiment 5: Multiple mating with three males the cone of the genitalia was relatively thick and narrow, like When females mated once, 23 of 51 females oviposited. In a cashew nut (Figure 8A), but in O. minutus and O. strigicollis contrast, eight of nine females oviposited when they mated the cone was flat and wide like a spatula (Figure 8B and C). with three males. The total number of eggs for the females’ The dendicule was small in O. sauteri, but large in O. lifetimes did not differ between single mating and multiple minutus (Figure 8A and B). The dendicule of O. strigicol- mating with different males (t = À0.00557, d.f. = 29, lis (Figure 8C) was smaller than that of O. minutus,and P>0.05; Figure 7). Hatching success was not affected by its structure permitted the easiest differentiation between the mating frequency by multiple males (t = 0.376, the two species. A coeloconic sensillum-like structure was d.f. = 12, P>0.05; Figure 7). seen on the thick surface of the cone of O. sauteri (Figure 8A), and a similar sensillum-like structure was Experiment 6: Scanning electron microscopy of the male genitalia observed at the side of the flat cone in the other two spe- The paramere of the male genitalia consists of two major cies (Figure 8B and C). We observed three circles near parts,theconeandtheflagellum.Thefunctionoftheconeis the hinge of the cone, and these circles were similarly in Mating system of three minute pirate bugs 147

10 6 9 6 A O. sauteri 112 31 46 100 40 0.7

(%) 35 80 0.6 30 s 60 0.5 25 40 0.4 20

ecundity 0.3

20 hing succes F 15 c

Mating success 0 10 0.2 Inseminated Non-inseminated Hat 5 0.1 1st mating 2nd mating 0 0 3 days 10 days 3 days 10 days B O. minutus Figure 6 + 76 18 33 Mean ( SE) fecundity (number of eggs deposited) and ) 100 hatching success of Orius sauteri with multiple mating opportunities with single males for 3 or 10 days. Numbers at the 80 top of each bar indicate the numbers of females tested. 60

40 23 8 8 6 ng success (% ng success 20 60

Mati 0.7 0 50 Inseminated Non-inseminated 0.6 40 1st mating 2nd mating cess 050.5 30 0.4 C O. strigicollis

Fecundity 0.3 73 36 15 20 100 atching suc 020.2 H 10 80 0.1

60 0 0 1 male 3 males 1 male 3 males success (%) 40

20 Figure 7 Mean (+ SE) fecundity (number of eggs deposited) and hatching success of Orius sauteri with mating opportunities with Mating 0 Inseminated Non-inseminated single virgin males once or 39 with different virgin males. Numbers at the top of each bar indicate the numbers of females 1st mating 2nd mating tested.

Figure 5 Mating receptivity of (A) Orius sauteri,(B)Orius minutus, and (C) Orius strigicollis at 1st and 2nd mating attempts by different males. The mated females were subjected to a 2nd inseminate at least three females, and there were no signiﬁ- mating experiment, then checked to determine whether they cant differences in the number of eggs laid among the 1st, were fertilized. Numbers at the top of each panel indicate the 2nd, and 3rd females. These results suggest that the males numbers of females tested. of these species could produce more offspring by mating with multiple females. Leon-Beck & Coll (2009) demon- line in all three species (Figure 8). This is the 1st report strated that O. laevigatus males also inseminated at least of such a series of circles in the paramere. The function of three females but the 3rd-mated O. laevigatus females the circles is not yet known. deposited signiﬁcantly fewer eggs than those mated with virgin or once-mated males. This suggests that sperm depletion would occur in O. laevigatus. In this study, Discussion although the 3rd matings by the O. minutus males were The results of our experiments demonstrated that the less successful than those of the other two species, we did males of the three Orius species are polygamous. They can not observe such a decline in the insemination ability of 148 Arakawa et al.

A O. sauteri (a) Ventral side Flagellum

(e) Sensillum-like (d) Cone structure Cone (b) Dorsal side (c) Three circles

Flagellum Dendicule (a) Ventral side B O. minutus (e) Sensillum-like Flagellum structure

(d) Cone (c) Three circles Cone (b) Dorsal side

Dendicule

Flagellum (a) Ventral side C O. strigicollis Flagellum Figure 8 SEM observation of the paramere of (A) Orius sauteri,(B)Orius minutus, and (C) Orius strigicollis.Overview(a) (e) Sensillum-like (d) Cone from inside (ventral side), and (b) from Cone structure outside (dorsal side), (c) three circles, (d) (c) Three circles (b) Dorsal side cone, and (e) sensillum-like structures on the cone. Flagellum, cone, and dendicule are indicated with a white arrow in (a) and (b). Three circles near the hinge of the cone are indicated with white arrowheads l Flagellum in (c). Scale bars: 10 m, except in (e), Dendicule where it represents 1 lm. males. As oviposition represents a successful deposition of O. strigicollis. The failure of mating could be caused by the chorion and successful preceding fertilization of Orius non-genetic or genetic factors in females and/or males; this (Horton, 2008), the insemination ability of these three spe- topic remains to be investigated. In our study, to exclude cies seems to continue for at least three copulations. at least pseudo-copulation (in which the copula duration In our investigation, a fraction of both the O. minutus was too short to transfer sperm), we determined the mini- and O. strigicollis females mated twice, but the 2nd mating mum effective time. After the exclusion of pseudo-copula- occurred only when the 1st mating failed. Many females tion, we found that females likely accepted a 2nd mating could not be inseminated by single mating; the percentage only to compensate for a failed insemination. We thus of females inseminated by single mating was slightly less conclude that O. minutus and O. strigicollis are functionally than 30% for O. minutus and slightly more than 60% for monandrous. Such functional monandry has been referred Mating system of three minute pirate bugs 149 to as ‘pseudopolyandry’ (Fisher et al., 2013). We also sperm and/or sperm-associated ejaculates for fertilization found that longer mating times resulted in a higher fertil- during each mating. In addition, multiple mating with sin- ization success but did not have any effect on the number gle males did not increase fecundity. These results sug- of eggs for both O. minutus and O. strigicollis. The reason gested that the O. sauteri females did not obtain a direct why O. strigicollis had a higher fertilization success com- benefit from mating. Leon-Beck & Coll (2009) reported pared to O. minutus may be related to its longer mating that sexual harassment by males decreased the fecundity of time. O. lavigatus. In our experiments, living together with a Monandry is assumed to arise via either male- or male tended to decrease the hatching success, and thus female-driven processes (Arnqvist & Nilsson, 2000; Arn- males’ harassment might have negative effects on hatching qvist & Andres, 2006; Hosken et al., 2009). The functional success, although the mechanisms underlying this are monandry of O. minutus and O. strigicollis appears to be unclear. As the number of hatched larva did not differ controlled by females, because females were observed to between our 3- and 10-day treatments, the females would reject unwanted mating attempts by struggling and raising not get a direct benefit of multiple mating with a single their ovipositors. Female-controlled monandry may male, or alternatively, a direct benefit might be cancelled evolve when female fitness is maximized by mating only by males’ harassment. once (Arnqvist & Nilsson, 2000; Jones, 2001; Wedell et al., We also observed that mating with multiple males did 2002; Hosken et al., 2009; King & Bressac, 2010; Jiao et al., not increase fecundity. Explicitly, this result includes the 2011). In other words, when the cost of mating is high, effects of the number of matings as well as the effects of monandry would be favoured. The costs of time and polyandry. Further experiments should be conducted to energy, predation risk, and parasite/pathogen infection are determine the precise effects of polyandry. In this study, unknown, and physical injury might occur during insemi- however, the experiment concerning the multiple mating nation. The Anthocoridae and Cimicidae both exhibit with a single male did not show any effects of the number traumatic insemination (Horton & Lewis, 2011). Physical of matings on fecundity. Considering the results of the damage and infection can result in short longevity and less multiple mating with a single male experiment and the reproductive success (Stutt & Siva-Jothy, 2001; Morrow & experiment regarding the polyandry, both the mating fre- Arnqvist, 2003; Backhouse et al., 2012; Tatarnic & Cassis, quency and polyandry are likely to have no effect on fecun- 2013; Reinhardt et al., 2014). Our SEM examination indi- dity. A meta-analysis showed that polyandry elevates egg cated that the genitalia of these Orius species could damage hatching success (Slatyer et al., 2012), but in our study the females during copulation. However, our observations hatching success did not differ. In addition to fecundity suggested that females of all three Orius species examined and hatching success, maternal longevity also could be use an extragenital copulatory tube for copulation and that affected by multiple mating including polyandry (e.g., the males do not seem to scar the females (manuscript in Kawagoe et al., 2001; McNamara et al., 2008). Arnqvist & preparation). It will be interesting to determine the cost of Nilsson (2000) showed that the negative effect of re-mat- extragenital insemination without wounding of the ing on female longevity might be compensated for by posi- female. tive effects (increased egg production rate and fertility). To Like O. minutus and O. strigicollis, the females of O. sau- reveal direct and indirect benefits of polyandry, further teri accepted the 2nd mating when they had not been studies on maternal longevity and the fitness of offspring inseminated. In addition, some O. sauteri females mated following sperm competition might be informative. again even after they had started oviposition. This means Our SEM observations of the male genital paramere that O. sauteri is truly polyandrous. In many insect species, suggested a closer relationship between O. minutus and O. females mate multiple times because of direct benefits strigicollis than between either of these species and O. sau- (such as an increase in fecundity; Arnqvist & Nilsson, teri. There was a particularly strong similarity between O. 2000) or indirect genetic benefits (Tregenza & Wedell, minutus and O. strigicollis with respect to the shape of the 1998; Jennions & Petrie, 2000; Slatyer et al., 2012). The cone and the position of the coeloconic sensillum-like increase in fecundity occurs by gaining material benefits structure on the cone. The sensillum-like structure was 1st from males such as sperm (Kraus et al., 2004; Wang & reported in Orius insidiosus (Say) and Orius pumilio Davis, 2006), sperm-associated accessory substances (Arn- (Champion) by Shapiro et al. (2010). Both of these species qvist & Nilsson, 2000; Gillott, 2003), and nuptial gifts had a thick narrow cone, and the sensillum-like structures (Vahed, 1998; Arnqvist & Nilsson, 2000). In the light of were found near the tip. The sensillum-like structure of O. our present finding that there was no significant relation- sauteri was on the thick surface, but the position was at the ship between the mating duration and the number of eggs, middle, where the cone curved strongly. In contrast, the it is likely that the males transferred the same amount of sensillum-like structures were found at the middle side of 150 Arakawa et al. the flat cones in O. minutus and O. strigicollis. Our obser- Dunn DW, Sumner JP & Goulson D (2005) The benefits of vations of the mating position in frozen samples showed multiple mating to female seaweed flies, Coelopa frigida that the cone-side possessing the sensillum-like structure (Diptera: Coelpidae). Behavioral Ecology and Sociobiology – faced the female abdomen (data not shown). It is possible 58: 128 135. that the sensillum plays a role in recognizing the female Elzinga JA, Chevasco V, Grapputo A & Mappes J (2011) Influence of male mating history on female reproductive success among either mechanically or chemically, and it may assist in the monandrous Naryciinae (Lepidoptera: Psychidae). Ecological orientation of the paramere for successful penetration of Entomology 36: 170–180. the female’s abdominal integument. It would be of interest Fisher DN, Doff RJ & Price TAR (2013) True polyandry and to examine ultra-thin sections of the sensillum-like struc- pseudopolyandry: why does a monandrous fly remate? BMC ture to determine whether nerves cells exist there. Evolutionary Biology 13: 157. We also identified three circles that formed a line near Gillott C (2003) Male accessory gland secretions: modulators of the hinge of the cone in all three Orius species. Since the female reproductive physiology and behavior. Annual Review phylogenetic relationships of the Anthocoridae and related of Entomology 48: 163–184. families have not yet been fully elucidated (Horton, 2008; Gwynne DT (2008) Sexual conflict over nuptial gifts in insects. – Jung et al., 2010), these novel structures might prove use- Annual Review of Entomology 53: 83 101. ful features for distinguishing relatedness. The genital Hardy ICW, Ode PJ & Siva-Jothy MT (2005) Mating systems. Insects as Natural Enemies: A Practical Perspective morphology of these three Orius species coincides with (ed. by MA Jervis), pp. 261–298. Springer, Dordrecht, The their phylogenetic relationships as determined by mito- Netherlands. chondrial cytochrome oxidase subunit I genes (Muraji Hinomoto N, Nagamori S, Kakimoto K, Shimizu T, Higaki T et al., 2000a). The degree of relatedness also reflected their et al. (2009) Molecular identification and evaluation of Orius mating systems; O. sauteri is polyandrous, whereas the two species (Heteroptera: Anthocoridae) as biological control closely related species O. minutus and O. strigicollis are agents. Japan Agricultural Research Quarterly 43: 281–288. functionally monandrous. Studies of additional species Horton DR (2008) Minute pirate bugs (Hemiptera: Anthocori- might improve our understanding of how much antho- dae). Encyclopedia of Entomology (ed. by J Capinera), pp. corid mating systems vary as a function of phylogeny and 2402–2412. Springer, Dordrecht, The Netherlands. morphology. In addition, we need to investigate the mat- Horton DR & Lewis TM (2011) Variation in male and female ing behaviour and mating frequency in natural conditions, genitalia among ten species of North American Anthocoris (Hemiptera: Heteroptera: Anthocoridae). Annals of the Ento- because small-scale bioassays in the laboratory might pro- mological Society of America 104: 1260–1278. vide spurious results (Collet et al., 2018). Hosken DJ, Stockley P, Tregenza T & Wedell N (2009) Mono- gamy and the battle of the sexes. Annual Review of Entomology Acknowledgements 54: 361–378. Jennions MD & Petrie M (2000) Why do females mate multiply? We thank Mr. Takashi A. Inoue of NARO for technical A review of the genetic benefits. Biological Reviews 75: 21–64. advice on the SEM operation. Jiao X, Guo L, Chen Z, Wu J, Chen J et al. (2011) Experimental evidence for female-driven monandry in the wolf spider, Par- – References dosa astrigera. Behavioral Ecology and Sociobiology 65: 2117 2123. Arnqvist G & Andres JA (2006) The effects of experimentally Jones TM (2001) A potential cost of monandry in the lekking induced polyandry on female reproduction in a monandrous sandfly Lutzomyia longipalpis. Journal of Insect Behavior 14: mating system. Ethology 112: 748–756. 385–399. Arnqvist G & Nilsson T (2000) The evolution of polyandry: mul- Jung S, Kim H, Yamada K & Lee S (2010) Molecular phylogeny tiple mating and female fitness in insects. Animal Behaviour and evolutionary habitat transition of the flower bugs (Hetero- 60: 145–164. ptera: Anthocoridae). Molecular Phylogenetics and Evolution Backhouse A, Sait SM & Cameron TC (2012) Multiple mating in 57: 1173–1183. the traumatically inseminating warehouse pirate bug, Xylocoris Kawagoe T, Suzuki N & Matsumoto K (2001) Multiple mating flavipes: effects on fecundity and longevity. Biology Letters 8: reduces longevity of females of the windmill butterfly Atropha- 706–709. neura alcinous. Ecological Entomology 26: 258–262. Boulton RA & Shuker DM (2013) Polyandry. Current Biology 23: King BH & Bressac C (2010) No fitness consequence of experi- R1080–R1081. mentally induced polyandry in a monandrous wasp. Behaviour Collet M, Amat I, Sauzet S, Auguste A, Fauvergue X et al. 147: 85–102. (2018) Insects and incest: sib-mating tolerance in natural Kobayashi T & Osakabe M (2009) Pre-winter copulation populations of a parasitoid wasp. bioRxiv (https://doi.org/10. enhances overwintering success of Orius females (Heteroptera: 1101/169268). Anthocoridae). Applied Entomology and Zoology 44: 47–52. Mating system of three minute pirate bugs 151

Kraus FB, Neumann P, van Praagh J & Moritz RFA (2004) Sperm Rowe L & Arnqvist G (2002) Sexually antagonistic coevolution in limitation and the evolution of extreme polyandry in honey- a mating system: combining experimental and comparative bees (Apis mellifera L.). Behavioral Ecology and Sociobiology approaches to address evolutionary processes. Evolution 56: 55: 494–501. 754–767. Kuijper B, Pen I & Weissing FJ (2012) A guide to sexual selection Shapiro J, Shirk P, Kelley K, Lewis T & Horton D (2010) Identity theory. Annual Review of Ecology, Evolution, and Systematics of two sympatric species of Orius (Hemiptera: Heteroptera: 43: 287–311. Anthocoridae). Journal of Insect Science 10: 1–17. Lattin JD (1999) Bionomics of the Anthocoridae. Annual Review Slatyer RA, Mautz BS, Backwell PRY & Jennions MD (2012) Esti- of Entomology 44: 207–231. mating genetic benefits of polyandry from experimental stud- Leon-Beck M & Coll M (2009) The mating system of the flower ies: a meta-analysis. Biological Reviews 87: 1–33. bug Orius laevigatus. Biological Control 50: 199–203. Stutt AD & Siva-Jothy MT (2001) Traumatic insemination McNamara KB, Elgar MA & Jones TM (2008) A longevity cost of and sexual conflict in the bed bug Cimex lectularius. Pro- re-mating but no benefits of polyandry in the almond moth, ceedings of the National Academy of Sciences of the USA Cadra cautella. Behavioral Ecology and Sociobiology 62: 1433– 98: 5683–5687. 1440. Tatarnic NJ & Cassis G (2013) Surviving in sympatry: paragenital Morrow EH & Arnqvist G (2003) Costly traumatic insemination divergence and sexual mimicry between a pair of traumatically and a female counter-adaptation in bed bugs. Proceedings of inseminating plant bugs. American Naturalist 182: 542–551. the Royal Society of London B 270: 2377–2381. Tregenza T & Wedell N (1998) Benefits of multiple mates in the Muraji M, Kawasaki K & Shimizu T (2000a) Nucleotide sequence cricket Gryllus bimaculatus.Evolution52:1726–1730. variation and phylogenetic utility of the mitochondrial COI Tregenza T & Wedell N (2000) Genetic compatibility, mate fragment in anthocorid bugs (Hemiptera: Anthocoridae). choice and patterns of parentage: invited review. Molecular Applied Entomology and Zoology 35: 301–307. Ecology 9: 1013–1027. Muraji M, Kawasaki K & Shimizu T (2000b) Phylogenetic utility Vahed K (1998) The function of nuptial feeding in insects: a of nucleotide sequences of mitochondrial 16S ribosomal RNA review of empirical studies. Biological Reviews 73: 43–78. and cytochrome B genes in anthocorid bugs (Heteroptera: Wang Q & Davis LK (2006) Females remate for sperm replenish- Anthocordiae). Applied Entomology and Zoology 35: 293–300. ment in a seed bug: evidence from offspring viability. Journal Ohno K & Takemoto H (1997) Species composition and seasonal of Insect Behavior 19: 337–346. occurrence of Orius spp. (Heteroptera: Anthocoridae), preda- Wedell N, Wiklund C & Cook PA (2002) Monandry and polyan- cious natural enemies of Thrips palmi (Thysanoptera: Thripi- dry as alternative lifestyles in a butterfly. Behavioral Ecology dae), in eggplant fields and surrounding habitats. Applied 13: 450–455. Entomology and Zoology 32: 27–35. Yasui Y (1998) The ‘genetic benefits’ of female multiple mating Paukku S & Kotiaho JS (2005) Cost of reproduction in Calloso- reconsidered. Trends in Ecological Evolution 13: 246–250. bruchus maculatus: effects of mating on male longevity and the Yasunaga T (1997a) The flower bug genus Orius wolff (Hetero- effect of male mating status on female longevity. Journal of ptera: Anthocoridae) from Japan and Taiwan, Part I. Applied Insect Physiology 51: 1220–1226. Entomology and Zoology 32: 355–364. Pizzari T & Wedell N (2013) The polyandry revolution. Philo- Yasunaga T (1997b) The flower bug genus Orius wolff (Hetero- sophical Transactions of the Royal Society B 368: 20120041. ptera: Anthocoridae) from Japan and Taiwan, Part II. Applied Reinhardt K, Anthes N & Lange R (2014) Copulatory wounding Entomology and Zoology 32: 379–386. and traumatic insemination. The Genetics and Biology of Sexual Yasunaga T (1997c) The flower bug genus Orius wolff (Hetero- Conflict(ed.byWRRice&SGavrilets),pp.115–139. Cold ptera: Anthocoridae) from Japan and Taiwan, Part III. Applied Spring Harbor Laboratory Press, Cold Spring Harbor, MA, USA. Entomology and Zoology 32: 387–394.