Introduction
A trade-off is a negative relationship between two life history traits in which increases in one are associated with decreases in the other as a result of a compromise allocation of the limited resources that are available to it (Begon et al., Reference Begon, Townsend and Harper2006). A trade-off can be revealed through experimental manipulation in which individuals are randomly assigned to different treatments. This procedure is better than simple field observation of negative phenotypic correlation because individuals may differ, for instance, in the quantity of resources they have available. Trade-offs are fundamental in predicting the optimal life history in any given environment, and traits commonly considered are reproduction and adult longevity (Zera & Harshman, Reference Zera and Harshman2001).
Reproduction is a life history trait that may occur in one terminal burst (semelparity) or as a series of repeated events (iteroparity), whose costs are eggs production and mating (De Loof, Reference De Loof2011). In insects, eggs production is coordinated with the availability of food, duration of the photoperiod and mating, among other factors (Gullan & Cranston, Reference Gullan and Cranston2010). Concerning mating, it is indispensable for the transfer of sperm in species with sexual reproduction. However, apart from the investment in energy and time that it requires, mating can impose risks to females due to physical injuries, pathogen infection, and predation. Nevertheless, in the majority of insects, females permit multiple mating with the same or with multiple males (i.e., polyandry) even when males provide no other benefits to females such as food or parental care (Arnqvist & Nilsson, Reference Arnqvist and Nilsson2000). Suggested positive effects of mating on female fitness are: sperm supply can maintain egg fertility, accessory substances can increase egg production rate, the mating act itself can stimulate egg production (Bradshaw & McMahon, Reference Bradshaw, McMahon, Jørgensen and Fath2008) and extend female lifespan (Burpee & Sakaluk, Reference Burpee and Sakaluk1993). For males, each copulation offers the opportunity to produce new offspring, so that their reproductive success is closely related to the number of females that they have the possibility of inseminating. Copulations can last a long time, and long-lasting copulations increase males' fertilization success relative to their rivals, in response to high sperm competition intensity under high population density and male-biased sex ratio (Sillén-Tullberg, Reference Sillén-Tullberg1981).
Stink bugs are iteroparous and they exhibit multiple copulations (MC) throughout their lifespan (Panizzi, Reference Panizzi, Schaefer and Panizzi2000). The green belly stink bug Dichelops furcatus (F.) (Hemiptera: Penatatomidae) is a Neotropical phytophagous stink bug common in Argentina and southern Brazil (Pereira et al., Reference Pereira, Salvadori, Lau, Marsaro and Panizzi2013). It is a pest on wheat and corn: when feeding injects toxins in the stem of the seedlings harming the vigor of the plant, and the pest status was attributed to the massive adoption by growers of the notillage cultivation system (Panizzi, Reference Panizzi, Schaefer and Panizzi2000). The adults overwinter in the fallow of maize and soybean crops, and in mid spring the post-hibernating adults disperse to wheat and later to young maize. Despite being a pest, very little is known about the population dynamics of D. furcatus and the characteristics of the process that results in offspring production.
Our objectives were: (1) to estimate demographic attributes of D. furcatus optimal reproduction (i.e., allowing females to copulate freely), (2) to analyze the effect of only one mating vs. multiple matings, and (3) to evaluate the trade-off between fecundity and adult female longevity.
Methods
Colonies of stink bugs were established from wild individuals collected from the Experiment Station ‘Julio Hirschhorn’ of the Faculty of Agricultural Sciences and Forestry of the National University of La Plata (34°59′24.32″S, 58°0′18.51″W; 27 ms. n. m.). Adults of D. furcatus were maintained in wired cages (20 × 20 × 30 cm), fed with Phaseolus vulgaris (L.) (Fabales: Fabaceae) pods, and kept at 24 ± 1°C, 70 ± 10% relative humidity, and a photoperiod of 14:10 h (L:D). Once the stink bugs reached the last (fifth) nymphal instar, they were kept separately until they molted into the adult stage. Then, adults were assigned to one of the following three treatments: (1) MC: each female was allowed to mate repeatedly with a single male until female's death (n = 20). Every time a male is dead before female's death, the male was replaced. (2) Single copulation (SC): each female was kept together with one male only until the first copulation was observed. We performed 18 replicates (n = 18), registering also the duration of the copulation process. (3) Virgin females (VF): each female was kept singly without copulation until its death (n = 20). The individuals were maintained in the same conditions as mentioned for insect culture. Every replicate was checked daily, and egg masses were collected and kept separately in a Petri dish until the emergence of nymphs. We recorded the date in which each oviposition was made, the number of egg masses oviposited by each individual female, the number of eggs per egg mass (egg mass size), and the number of viable (hatched) and inviable (unhatched) eggs, as well as females' mortality. For each female, we recorded the entire reproductive period (i.e., the time elapsed between the first and the last oviposition made), the time interval between two successive ovipositions, as well as the duration of the pre-reproductive (time until start of oviposition) and post-reproductive (time from last oviposition to death) periods. We estimated the mean longevity of females in the jth group, L (Gj), based on the duration of the adult life of each individual female, l (n, Gj), as L (Gj) = [∑ n=1, N l (n, Gj)]/N (Gj), where N (Gj) represents the total number of females in the jth group.
The mean fecundity of females in the jth group, Fec(Gj), was estimated as Fec(Gj) = [∑ n=1, N fec(n, Gj)]/N (Gj), where fec(n, Gj) is the fecundity of each individual female.
We estimated the total egg fertility of females in the jth group, F (Gj), dividing the total number of eggs from which nymphs emerged by the total number of eggs laid. For each female, we divided the reproductive period into two equal halves: first half (i.e., from the first oviposition to the mid time of the reproductive period) and second half (i.e., from the mid time of the reproductive period to the last oviposition). We compared fecundity and fertility between the first and the second halves.
Number of lived days equivalent to one deposited egg:
For answering this question, we assumed that the resources available to D. furctaus adult females would be allocated into reproduction and survival. We also assumed that the resource requirements to produce and oviposit a viable (fertilized) or an inviable (not fertilized) egg are the same, as well as that the requirements of a female to survive on a daily basis are independent of her age.
Taking into account the mean values of longevity and fecundity of MC females (L (MC) and Fec(MC)) and of VF females (L (VF) and Fec(VF)), we considered that, from the total amount of resources required for the development of an adult of D. furcatus, the amount invested by an MC female in its possible greater number of laid eggs is proportional to the amount invested by a VF female in its possible greater longevity. That is to say, k·[Fec(MC)-Fec(VF)] = [L (VF)-L (MC)], where k is the coefficient of proportionality. Therefore, k = (L (VF) − L (MC))/(Fec(MC) − Fec(VF)) represents the number of lived days equivalent to one deposited egg, and the inverse of the coefficient 1/k = (Fec(MC) − Fec(VF))/(L (VF) − L (MC)) represents the number of deposited eggs that are equivalent to one lived day. If, as a result of the trade-off between longevity and reproduction, Fec(MC) is expected to be greater than Fec(VF) and L (VF) is expected to be greater than L (MC), then the coefficient k takes values in the interval (0, infinite). On the contrary, if L (VF) ~ L (MC) and Fec(MC) ~ Fec(VF), k gets an undetermined value.
If the total amount of resources required for a D. furcatus adult female were independent of the treatment to which it was subjected, we expressed it for the n-th female of the SC group in units of days, TRd (n, SC), adding to the observed longevity, l (n, SC), the number of days equivalent to its fecundity, k·fec(n, SC), symbolically: TRd (n, SC) = l (n, SC) + k·fec(n, SC). In a similar way, we expressed the total amount of resources required for the n-th female of the SC group in units of eggs, TRe (n, SC), adding to the observed fecundity, fec(n, SC), the number of eggs equivalent to its longevity, (1/k)·l (n, SC), symbolically: TRe (n, SC) = fec(n, SC) + (1/k)·l (n, SC). We would expect that the corresponding mean values, TRd (SC) = [∑ n=1, N TRd (n, SC)]/N (SC), and TRe (SC) = [∑ n=1, N TRe (n, SC)]/N (SC), where N (SC) represents the total number of females in the SC group, did not differ significantly from the mean values calculated from MC (values from the VF group equals values from the MC group): TRd (MC) = L (MC) + k·Fec(MC), and TRe (MC) = Fec(MC) + (1/k)·L (MC).
Statistical analysis
Differences in longevity, duration of the reproductive period, fecundity, fertility, and time interval between two successive ovipositions were tested by one-way analysis of variance and differences among treatments were analyzed by means of the Tukey's HSD test. In cases where the assumptions of parametric tests were not reached, even after data transformation, we performed the non-parametric Kruskal–Wallis test and differences between pairs of treatments were analyzed by means of the Mann–Whiney U test.
We performed a correlation by the Pearson product-moment correlation test, between longevity and fecundity, for each one of the treatments. Differences between TRd and TRd (SC) and between TRe and TRe (SC) were evaluated by a t test. All statistical analyses were carried out using the free software R (R Core Team, 2014) and P < 0.05 values were considered significant (Zar, Reference Zar1996).
Results
For MC females, the more and the less long-lived individuals lived 205 and 73 days, respectively (the difference between them was almost 4 and a half months). Maximum and minimum fecundities were 415 and 83 eggs/female, and the lowest fertility was five times less than the highest. Longevity and fecundity of all MC females were positively correlated: r = 0.50; t (df=19) = 2.458 (P = 0.024).
For SC females, maximum and minimum longevity were 280 and 34 days, respectively (the difference between them was more than 8 months). Fecundity also fluctuated widely: maximum and minimum fecundity were 445 and two eggs/female, respectively. The lowest fertility was more than 200 times less than the highest. Longevity and fecundity of all SC females were not significantly correlated: r = 0.261; t (df=16) = 1.082; P = 0.295.
For VF females, maximum and minimum longevity were 396 and 20 days, respectively. Regarding fecundity, four of the females did not reproduce, and another four females laid only one egg mass. Without considering the females that did not lay eggs, maximum and minimum fecundity were 154 eggs/female (this female lived 265 days) and three eggs/female (oviposited by a female that lived 258 days). The lowest fertility was more than 50 times less than the highest. As for SC, longevity and fecundity of all VF females were not correlated: r = 0.215, t (df=19) = 0.933; P = 0.363.
Longevity was different among treatments (H (2, n=58) = 13.438; P = 0.001), being lower on females of the group MC (table 1). When we differentiated the duration of the pre- and post-reproductive and reproductive periods, we also found differences among treatments. The pre-reproductive period was similar for MC and SC females, and it was shorter than the period recorded for VF (H (2, n=54) = 23.392; P < 0.001). Contrarily, there were no differences in the duration of the reproductive period among treatments (H (2, n=54) = 3.357; P = 0.197). However, the post-reproductive period of MC females was shorter than the one of SC and VF (H (2, n=54) = 10.989; P = 0.004) (table 1).
Table 1. Effect of mating frequency on demographic attributes of Dichelops furcatus females multicopulated (MC), copulated once (SC), or virgin (VF) (means ± SD).
Different letters indicate significant differences among treatments.
We also found differences in the fecundity of females among treatments (H (2, n=58) = 26.114; P < 0.001): females of the groups MC and SC were more fecund than females of the group VF (table 1). The size of egg masses differed among treatments (H (2, n=2) = 7.556; P = 0.023): VF = 9.716 eggs/egg mass (SD = 3.182, n = 16), SC = 14.329 eggs/egg mass (SD = 4.355, n = 18), and MC = 12.543 eggs/egg mass (SD = 0.980, n = 209), and the differences between SC and MC were not significant (0.073). The mean egg mass size of SC females was somewhat greater than MC (although not significantly) and each SC female oviposited a lower number of masses than MC (although not significantly). Overall fecundity of SC females did not differ significantly from MC (F = 2.715; df = 41; P = 0.107) (table 1). Regarding eggs' fertility, VF only oviposited unfertile eggs, and when VF was excluded from the analysis, we found differences in the fertility of eggs oviposited by SC or MC females (F = 8.773; df = 35; P = 0.005).
In general, MC and VF females represented contrasting results in terms of longevity and fecundity, and SC females showed intermediate values (fig. 1).
Fig. 1. Mean longevity and fecundity of adult multicopulated (MC), copulated once (SC), or virgin (VF) females of Dichelops furcatus. Bars indicate confidence intervals (95%). Different letters indicate significant differences in longevity (lowcase letters) or fecundity (capital letters) among treatments.
Finally, the time interval between two successive ovipositions was significantly different among treatments (H (2, n=47) = 28.305; P < 0.001), being the shortest for MC, and SC shorter than VF.
Analyzing the effect of only single copulation vs. multiple copulations
During adulthood, MC females spent a pre-reproductive period of almost a fortnight, which represented 13.48% of the total longevity. The reproductive period represented 62.81% and the post-reproductive period represented the remaining 23.71% of the total longevity (table 1). When we divided the reproductive period into two equal parts, females laid 59.61% of all the eggs oviposited in their life in the first half, with a fertility of 60%. The time interval between two successive ovipositions was 4.30 days on average (SD = 1.65). We found no significant differences on the number of oviposited eggs, the fertility, and the duration of time intervals between successive ovipositions, between the first and the second half of the reproductive period (fecundity: F = 3.081; df = 38; P = 0.087, fertility: F = 0.134; df = 35; P = 0.716, time interval between successive ovipositions: U (n1=20, n2=20) = 131.00; P = 0.062).
During adulthood, SC females spent a pre-reproductive period of 13.11 days on average, which represented 7.405% of the total longevity. The reproductive period represented 47.631% and the post-reproductive period represented the remaining 44.965% of the total longevity (table 1). When we divided the reproductive period into two equal parts, SC females laid 57.53% of all the eggs oviposited in their life in the first half, with a fertility of 56%. The time interval between two successive ovipositions was 6.76 days on average (SD = 5.04). In the second half of the reproductive period, females oviposited a significantly lower number of eggs (52.18 ± 40.20) (U = 58; P = 0.014), with a lower fertility (U = 37; P < 0.001) (fig. 2) and with longer time intervals between successive ovipositions (U = 54.5; P = 0.014). Each copulation recorded in the SC group lasted between 5 and 20.9 h and the mean value was 18.27 h (SD = 4.93; n = 18).
Fig. 2. Fertility of individual egg masses oviposited by multicopulated (MC) and copulated once (SC) females of D. furcatus.
Females of the VF group showed a significantly higher (almost double) longevity than MC females (table 1). During adulthood, VF females spent a pre-reproductive period of 44.94 days on average, which represented 20.52% of the total longevity. The reproductive period represented 50.06% and the post-reproductive period represented the remaining 29.42% of the total longevity (table 1). Due to the scarce oviposition events recorded for females of the group VF, we did not divide the reproductive period into two halves for further analysis.
Number of lived days equivalent to one deposited egg
Mean longevity of VF females was 1.96 times greater than the longevity of MC females. On the contrary, mean fecundity of MC females was 4.71 times greater than fecundity of VF females. The coefficient k = 0.69 indicates that the amount of resources that would be needed to produce one egg is only enough to live 0.69 days. Consequently, 1/k = 1.45 indicates that the amount of resources that would be needed to live 1 day allows to produce almost one and a half egg.
For the MC group, the calculated total amount of resources utilized by an adult female expressed in ‘days equivalent’ and in ‘eggs equivalent’ were: 242.65 and 350.86, respectively, and the corresponding estimated mean values for SC group were: TRd (SC) = 277.86 ‘days equivalent’ (SD = 161.89, n = 18), and TRe (SC) = 460.01 ‘eggs equivalent’ (SD = 268.39, n = 18) (fig. 3) which did not differ from the former values: t (df=17) = −0.923 (0.2 > P > 0.1) for ‘days equivalent’, and t (df=17) = −1.725 (0.1 > P > 0.05) for ‘eggs equivalent’.
Fig. 3. Total amount of resources invested by each female of D. furcatus in producing eggs (columns on the left side) or in surviving (columns on the right side), expressed in ‘equivalent eggs’ and ‘equivalent days’, respectively. Bars represent 95% confident intervals.
Discussion
Life-history evolution can be constrained by trade-offs among various traits that contribute to fitness, and fitness can be defined as a measure of the proportion of individuals that are propagated into the following generations. Traits commonly considered are reproduction and adult longevity and reduction of life span as a result of mating has been reported for many species of arthropods of diverse orders and families (de Loof, Reference De Loof2011 and references therein).
Reproductive costs can imply ecological costs, as increased vulnerability to parasitism of mating pairs, and physiological costs when resources are diverted into reproduction and away from other essential processes (Reguera et al., Reference Reguera, Pomiankowski, Fowler and Chapman2004).
Besides, mating itself is an energy-demanding process which can represent a major reproductive cost and may directly and undirectly affect female fitness in a number of ways (Arnqvist & Nilsson, Reference Arnqvist and Nilsson2000). In effect, females may suffer male harassment during courting, physical damage by male genital barbs and spines (Crudgington & Siva-Jothy, Reference Crudgington and Siva-Jothy2000), and reduction of life span and foraging time (de Loof, Reference De Loof2011 and references therein). Nevertheless, mating can function as a necessary stimulus for the initiation of vitellogenesis or ovary development (Gillott, Reference Gillott2003), and can affect female's fecundity through three different mechanisms: the act of mating per se, the presence of sperm, and the transfer of accessory substances (Arnqvist & Nilsson, Reference Arnqvist and Nilsson2000). Anyway, multiple mating is widespread in insect species (Arnqvist & Nilsson, Reference Arnqvist and Nilsson2000). The adaptive significance of multiple mating has been extensively debated and the general idea is that, to be adaptive, the costs of multiple mating must be offset by benefits that enhance female's fitness. Two types of benefits are commonly used to explain the adaptive value of multiple mating: non-genetic benefits (Arnqvist & Nilsson, Reference Arnqvist and Nilsson2000) and genetic benefits (Jennions & Petrie, Reference Jennions and Petrie2000). The former, also denominated direct or first generation benefits, derive from the quality of the sperm of certain males that may increase female's fecundity, longevity, or mating rate. In this sense, Okada et al. (Reference Okada, Archer, Katsuki, Suzaki, Sharma, House and Hosken2015) demonstrated that irrespective of whether females mated with one or four males, mating multiply improved female fecundity and lifetime reproductive success in horned flour beetles (Gnatocerus cornutus, Coleoptera: Tenebrionidae). Additionally, if males transfer insufficient sperm, females may mate multiply to ensure all eggs are fertilized, hence obtaining fecundity benefits (García-González, Reference García-González2004). The latter also denominated indirect or second generation benefits result in an increase in the genetic diversity of the offspring, avoiding genetic incompatibilities derived from incestuous matings and inbreeding (Jennions & Petrie, Reference Jennions and Petrie2000). In this way, if future environmental conditions change, the chances that at least part of the offspring will be able to cope with the changes are increased.
Arnqvist & Nilsson (Reference Arnqvist and Nilsson2000) performed a meta-analysis to address the effects of multiple mating on the fitness of female insects, and concluded that a single mating does not maximize female fitness in general, and that beneficial effects of remating (increased fecundity and fertility) outweigh its detrimental effects on female longevity. Particularly among the Heteroptera, the effects of mating on egg development and oviposition appears to be variable (Fortes et al., Reference Fortes, Salvador and Cônsoli2011 and references therein). It was shown that N. viridula does not need mating-derived stimuli for ovary maturation, as already observed for other stink bug species (Masner, Reference Masner1966; Wightman, Reference Wightman1973; Davey et al., Reference Davey, Maimetes and Ruegg1986; Adams, Reference Adams2000) in which VF never laid eggs. However, multiple mating increased reproduction in N. viridula and Edessa meditabunda (Hemiptera: Pentatomidae) and the cost of increasing fecundity was the reduction of longevity (Mitchell & Mau, Reference Mitchell and Mau1969; Silva et al., Reference Silva, Laumann, Ferreira, Blassioli Moraes, Borges and Čokl2012).
In our study, we found that mating is not required for D. furcatus to oviposit, as VF were able to lay eggs. The same situation was denoted for Halyomorpha halys (Hemiptera: Pentatomidae) (Chu et al., Reference Chu, Zhou, Li and Liu1997; Cambridge, Reference Cambridge2016). The effects of mating on egg development and oviposition are variable, and even when VF lay eggs, alterations in the ovarian developmental period, oocyte size and composition, and/or number of eggs laid may occur (Fortes et al., Reference Fortes, Salvador and Cônsoli2011).
MC were costly in terms of reduced longevity for D. furcatus compared with once mated and VF. However, although multicopulated females lived a shorter period, only the pre- and post-reproductive periods were shortened and the time interval between consecutive ovipositions was reduced, agreeing with the observations made for E. meditabunda (Silva et al., Reference Silva, Laumann, Ferreira, Blassioli Moraes, Borges and Čokl2012). Although the total number of eggs laid by a D. furcatus female was not affected by the number of copulations, D. furcatus females copulated only once concentrated ovipositions in the first half of the reproductive period while multicopulated females oviposited a similar number of eggs in the first and second halves of the reproductive period. Congruent results in terms of the moment of the reproductive period in which ovipositions were made were found for N. viridula: females mated once or twice produced more egg clusters and concentrated egg-laying activity in the early part of adult life, whereas those mating more often laid eggs throughout the life span, with fewer egg clusters (Fortes & Cônsoli, Reference Fortes and Cônsoli2011). These authors also found that mating did not stimulate oocyte development in N. viridula, but the lack of mating activity appeared to have stimulated oocyte resorption in 17-day-old females. Oocyte resorption may explain the lower fecundity of virgin D. furcatus, when compared to multi or once-copulated females.
Coefficients k and 1/k indicate that the amount of resources needed to produce one egg is only enough to live 0.69 days, or in other words, that the amount of resources needed to live 1 day allows producing more than a single egg. This was another way to analyze the amount of resources invested by MC females compared with VF females: while the reduction in the number of days lived was 104.95 days, the increase in fecundity was 151.75 eggs: more than one egg per day lost.
Concerning egg fertility, we found a significant increase in D. furcatus females with a higher number of copulations. Similarly, Kawada & Kitamura (Reference Kawada and Kitamura1983) found that an increase in the number of copulations moderated the decline of fertility of Halyomorpha mista (Hemiptera: Pentatomidae). In a congruent way, D. furcatus females copulated only once concentrated viable eggs in the first half of the reproductive period, while in the second half all eggs were inviable, indicating that females mated only once received insufficient sperm to ensure all eggs to be fertilized, or that sperm quality deteriorated over time leading to a decrease in fertilization capacity. The relatively long time that each copula lasted suggests that it is a strategy of the males tending to ensure that their sperm fecundates all or the largest number of oocytes. In polyandric females, the competition between males continues after the copulation in the form of competition between ejaculates of different males, a process that is known as ‘spermatic competition’. Males may try to avoid or minimize sperm competition by defending the female from other males' copulation attempts. This defense can take different forms, one of which is the prolonged copulations with the female that physically prevent copulation by other males. This behavior was rather common among Pentatomidae, as with N. viridula (McLain, Reference McLain1992).
MC reduced the longevity of D. furcatus females, and although fecundity was not affected, fertility was incremented in multicopulated females. These findings provide information about the reproduction of this phytophagous insect, relevant to improve our understanding of the population dynamics of this pest.
Acknowledgements
This study was supported by the Program of Incentive for Professors-Researchers of the National Ministry of Culture and Education of Argentina (Programa de Incentivos a Docentes-Investigadores del Ministerio de Cultura y Educación de la Nación), grants 11/N777 and 11/N825.
