Evidence for adaptive self-superparasitism in the dryinid parasitoid Haplogonatopus atratus when conspecifics are present

Self-superparasitism – ovipositing in or on a host that was previously parasitized by the same individual – had generally been considered maladaptive for solitary parasitoids. Recently, however, some researchers have suggested that self-superparasitized hosts produce higher fitness returns per host than singly parasitized hosts when these hosts are parasitized by conspecifics. The present study is the first to verify this possibility in the parasitoid Haplogonatopus atratus . Fitness returns from the first and second ovipositions under triple parasitism were compared with those from the first oviposition under double parasitism. Fitness returns were assessed in terms of survival rate, adult size, and developmental time. No differences were found between the two groups in adult size and developmental time. When the interval between the second and third ovipositions was 1 h, the total survival rate of the first and second comers under triple parasitism was higher than the survival rate of the first comer under double parasitism, and as high as the total survival rate of two first comers under double parasitism. This suggests that the value of self-parasitized hosts is as high as that of unparasitized hosts. When the second-to-third interval of triple parasitism was 24 h, self-superparasitism produced no fitness returns. Moreover, fitness returns from ovipositing into self-superparasitized hosts were similar to those from ovipositing into singly parasitized hosts.     

Superparasitism strategy in a semisolitary parasitoid with imperfect self/non-self recognition, Echthrodelphax fairchildii

To assess the adaptiveness of self‐ and conspecific superparasitism in the parasitoid Echthrodelphax fairchildii Perkins (Hymenoptera: Dryinidae), we measured the rate of superparasitism avoidance and fitness returns from superparasitism for different intervals between the first and second ovipositions. We also tested for any preference in oviposition side and whether any such preference was adaptive. The rate of superparasitism avoidance in both self‐ and conspecific superparasitism was about 10% for oviposition intervals of 2–8 h, and higher for intervals of 1 and ≥24 h (but did not exceed 35%). When conspecifically superparasitizing (but not self‐superparasitizing), females exhibited a slight preference for the side without the first‐comer. Under conspecific superparasitism, the survival rate of second comers was independent of the oviposition side and interval, but slightly lower than that of immature parasitoids for single parasitism. The adult size of second‐comers on the side harboring the first‐comer was larger than that of second‐comers on the side without it, when the oviposition interval was <24 or 96 h. These results explained the overall low rate of avoidance of conspecific superparasitism, but not the variation in avoidance rate or the preference for side without the first‐comer when conspecifically superparasitizing. Assuming that fitness returns are influenced more by offspring quantity than by their quality, self‐superparasitism produced positive fitness returns only when the oviposition interval was <24 h and the side without the first‐comer was selected. This suggests that the observed behavior was not adaptive. Possible reasons for the discrepancies between observed and optimal behaviors, including an imperfect ability for self/non‐self recognition, are discussed.     

Seemingly maladaptive refraining from infanticidal probing at the third parasitism attack by the semi-solitary parasitoid Echthrodelphax fairchildii （Hymenoptera： Dryinidae）

We assessed the adaptiveness of refraining from infanticidal probing at the third attack under triple parasitism in the parasitoid Echthrodelphaxfairchildii Perkins, by comparing fitness returns from the third attack with and without probing. Fitness returns were assessed in terms of the survival rate, mean developmental time, and mean adult head width of the progenies. Not probing was maladaptive under triple parasitism with oviposition intervals of 1/24 and 24/24 hours （where the numbers before and after the slash refer to the first-to-second and second-to-third oviposition intervals, respectively）, although no significant difference was detected in fitness returns for oviposition intervals of 24/1 hours. We suggest that the cost of probing （especially the decrease in the chance of future ovipositions） is a reason for the seemingly maladaptive absence of probing.     

Adaptive significance of facultative infanticide in the semi-solitary parasitoid Echthrodelphax fairchildii

Abstract.  1. This is the first work to show how imperfect kin recognition influences the behavioural strategy of parasitoids.
2. The female of Echthrodelphax fairchildii lays an egg on the right or left side of the host thorax. When superparasitising on the side without the first offspring, it often probed the opposite side with its sting and killed the first offspring more than 50% of the time. The frequency of probing increased with increasing inter-oviposition interval, and was higher in conspecific superparasitism than in self superparasitism.
3. The adaptiveness of probing was assessed by the difference in fitness returns between superparasitism with probing and superparasitism without probing. Fitness returns from self‐superparasitism and those from conspecific superparasitism were assessed by the fitness performance of both first and second comers, and that of second comers respectively.
4. In terms of the survival rate of immatures (one fitness component), probing under conspecific superparasitism guaranteed higher fitness returns than not probing for longer oviposition intervals, but under self‐superparasitism not probing guaranteed higher fitness returns for shorter intervals because superparasitism for such intervals often resulted in two-adult emergence.
5. In terms of head width (another fitness component), probing produced higher fitness returns for shorter oviposition intervals for both self and conspecific superparasitism.
6. Assuming that offspring quantity contributes more to fitness returns than does offspring quality, the observed frequencies of probing were considered adaptive for parasitoids whose abilities to distinguish between self and conspecifically parasitised hosts and to estimate inter-oviposition intervals are imperfect.     

Imperfect preference for singly parasitized hosts over doubly parasitized hosts in the semisolitary parasitoid Echthrodelphax fairchildii: implications for profitable self-superparasitism

Self‐superparasitism can be profitable (i.e., a fitness gain) when conspecific female adult parasitoids prefer singly parasitized hosts over doubly parasitized hosts. This preference is expected to evolve when the value (i.e., the fitness gain from oviposition) of doubly parasitized hosts is lower than that of singly parasitized hosts. We examined whether such a preference is found in the small brown planthopper, Laodelphax striatellus (Fallén) (Homoptera: Delphacidae), and its semisolitary infanticidal parasitoid Echthrodelphax fairchildii Perkins (Hymenoptera: Dryinidae). We compared the preferences and host values between each of four pairs of double and triple parasitism, each of which had the same time interval between the first and last oviposition bouts. Ovipositions on doubly and singly parasitized hosts occurred with similar frequencies in each of the four pairs, even though the doubly parasitized hosts were of lower value than the singly parasitized hosts. However, the value of doubly parasitized hosts with the first and second parasitoid offspring on the same side of the host was higher than that of hosts with the two offspring on different sides, and the value of the former did not differ significantly from that of singly parasitized hosts. The preferences between singly and doubly parasitized hosts with the two offspring on the same or different sides were as expected from differences in their values, except for one pair of double and triple parasitisms. This exception is considered attributable to an imperfect ability to evaluate hosts.     

Temporary host paralysis and avoidance of self-superparasitism in the solitary endoparasitoid Meteorus pulchricornis

We suggest a new mechanism for the avoidance of self-superparasitism by a solitary endoparasitoid of free-living hosts. Meteorus pulchricornis (Wesmael) (Hymenoptera: Braconidae) is a solitary larval endoparasitoid of a wide range of Lepidoptera. First, we tested the tendency to avoid multiple oviposition on Spodoptera litura (Fabricius) (Lepidoptera: Noctuidae) larvae, examining the distribution pattern of eggs deposited in 20 hosts in 24 h. The standardized Morisita index of dispersion (Ip) for the number of eggs in a host larva was almost always below zero, and the 95% confidence limits of the means were lower than zero for both experienced and inexperienced females, indicating that parasitoids tend to avoid multiple oviposition, although not completely. Second, we observed changes in host movements (head shaking and forward movement) after oviposition and the occurrence of additional ovipositions in pre-oviposited hosts, because a recent study indicated the necessity of visual cues provided by host movements for oviposition by this species. The incidence of host movements decreased markedly after oviposition and remained at a low level for approximately 1 h, during which time additional ovipositions were rarely observed. Temporary host paralysis after oviposition would be a new mechanism for the avoidance of self-superparasitism in a solitary endoparasitoid using visual cues from moving hosts for oviposition.     

Transient host paralysis as a means of reducing self-superparasitism in koinobiont endoparasitoids

The term ‘idiobiont’ refers to those parasitoid species that permanently paralyse their hosts during parasitism, causing the cessation of host growth and development. This is in contrast to koinobiont parasitoids, which allow their hosts to continue developing after being parasitized. While no koinobiont species induce permanent paralysis in their hosts, a minority of koinobionts induce a temporary paralysis that does not interfere with overall host growth and development. We characterized transient paralysis induction in two koinobiont aphid parasitoids in the genus Binodoxys (Hymenoptera: Aphidiinae). Both Binodoxys species induced transient paralysis in Aphis glycines, with paralysis time ranging between 4.5 and 8 min (depending upon parasitoid species and host instar). In a separate experiment, B. communis was capable of inducing transient paralysis in nine aphid species. We addressed two hypotheses potentially explaining the adaptive value of temporary host paralysis in experiments using A. nerii, which is readily accepted but engages in strong defensive behaviour. The first hypothesis is that paralysis increases oviposition success by interfering with host defences and the second is that it aids in the avoidance of self-superparasitism. Paralysed aphids were more likely to be rejected by B. communis than were aphids that had never been stung or that had recovered from paralysis. This result supports the avoidance-of-self-superparasitism hypothesis and is inconsistent with the hypothesis that transient paralysis increases oviposition success of B. communis.     

Infanticide in a Dryinid Parasitoid, Haplogonatopus atratus

When a parasitized host is superparasitized, the second comer—an immature parasitoid originating from the second oviposition—is usually handicapped relative to the first comer in the competition for the host. In Haplogonatopus atratus, however, the survival rate of second comers is frequently higher than that of first comers. Infanticide is assumed to be responsible for this phenomenon. The present report reveals that, when the female wasp of H. atratus superparasitizes, she examines several membranes between the host's abdominal terga with her stinger before laying a supernumerary egg and that, when she locates the first comer, she pierces it with her stinger to kill it. Moreover, once she catches a host she performs infanticide irrespective of whether the host is self- or conspecifically parasitized.     

Oviposition decisions in an endoparasitoid under self-superparasitism conditions

Superparasitism is a widespread phenomenon. Having accepted superparasitism, mated female parasitoids must decide on the sex of each egg they subsequently lay into the same host. Theory predicts that this decision is either based on host quality, when more male eggs are laid in hosts that are already parasitized because they are perceived to be of poorer quality; or more eggs are laid of the sex that is most likely to be a strong larval competitor, i.e. generally females.Anastatus disparis is a facultative endoparasitic egg parasitoid. We used ‘artificial’ hosts to explore outcomes of decision making by A. disparis during superparasitism under a manipulated absence of larval competition. When only one egg was laid it was always female. As the number of eggs laid increased, so more of them were male. This supports the theory that oviposition decisions are based on host quality; more male eggs were laid in hosts that were already parasitized and thus of poorer quality.In a second experiment, eggs were exposed to parasitoids for different periods of time. Half the eggs were dissected to determine the number of parasitoid eggs that had been laid. The remaining eggs were incubated and the number and sex of offspring that ultimately emerged, following larval competition, were recorded. Under superparasitism conditions fierce larval competition ensued; only one offspring survived and they were predominantly female.In conclusion, oviposition decisions by female A. disparis accepting self-superparasitism were made based on host quality.     

Mass-rearing biology of Fopius vandenboschi (Hym., Braconidae)

Abstract:   Basic biological information concerning the reproductive activities of Fopius ( Biosteres ) vandenboschi (Fullaway) (Hym., Braconidae, Opiinae) were determined to facilitate insectary mass production. Mean (±SE) progeny production per generation (yield) of 200 ♀♀ F. vandenboschi was 6835.8 ± 433.2 and 3755.8 ± 112.4 parasitoids, with 24-h exposure to second instar Bactrocera dorsalis (Hendel), and Ceratitis capitata (Wiedemann), respectively. Ceratitis capitata as a factitious host was more sensitive to parasitism, producing a significantly higher percentage of uneclosed host puparia frequently containing parasitoid cadavers (22.5 ± 0.8%), than B. dorsalis (3.4 ± 0.2%). The overall mean production of female offspring was ≥50%♀♀ in both host species. Unlike several other opiine parasitoids, young (5-day-old) females of F. vandenboschi produced low proportion of daughters (26–37%♀♀) and significantly shifted production to female-biased offspring (53–71%♀♀) at older age intervals (16–30 days). Based on the parasitoid yield data with a 24-h exposure period to second instar host larvae of B. dorsalis , about 150 cage replicates (of 200 ♀♀ each) may be sufficient to mass produce about 1 million parasitoids of F. vandenboschi during the reproductive period of 5–30 days, with an average of 56% female offspring. A short oviposition exposure period of 6 h was not optimal for mass production of F. vandenboschi . Total yield at 6 h exposure was one-fifth the production at 24-h exposure, using B. dorsalis hosts.     

Effect of expected offspring survival probability on host selection in a solitary parasitoid

Optimal Foraging Theory predicts that parasitoid females should optimize their host selection to maximize their lifetime fitness gain and parasitize the most profitable hosts. In particular, in solitary parasitoids, females should avoid superparasitism, at least when sufficient unparasitized hosts are available. However, when unparasitized hosts are scarce, they should prefer, among already parasitized hosts, those that provide the best survival probability to their progeny, which depends on the age and the developmental stage of the first parasitoid. To test this hypothesis in a solitary ectoparasitoid, Pachycrepoideus vindemmiae Rondani (Hymenoptera: Pteromalidae), we first assessed the survival probability of a second parasitoid according to the time elapsed since initial parasitism. We then analyzed the female selection behavior in patches containing a mixture of hosts parasitized over various time intervals. Our results showed that the older the opponent larva was, the lower the survival probability of the second parasitoid was. However, when the first individual had reached the prepupal stage, both individuals could complete their development. At this stage, the survival probability of the second parasitoid was surprisingly high but such individuals were reduced in size. Our study also showed that host acceptance by females was strongly correlated with the survival probability of their progeny when the first parasitoid was from 0 to less than 10 days‐old. When the first parasitoid had reached the prepupal stage, females usually rejected these hosts, although the survival probability of the offspring was quite high. This discordance between female host selection behavior and progeny survival probability is discussed.     

Egg-laying tactic in Phyllomorpha laciniata in the presence of parasitoids

Female insects are expected to choose oviposition sites that have the best conditions for offspring development and survival. Natural enemies, such as predators and parasitoids, may have a strong influence on the selection of oviposition substrates by phytophagous insects. The golden egg bug, Phyllomorpha laciniata (Villers) (Heteroptera: Coreidae) has an unusual reproductive strategy. Females mainly use conspecifics, both males and other females, as egg-laying substrates, but occasionally they oviposit on plants as well. Survival of the eggs is higher when eggs are carried by conspecifics than when they are laid on plants, due to predation and parasitism. We investigated egg-laying behavior in the forced presence of the egg parasitoid Gryon bolivari (Giard) (Hymenoptera: Scelionidae). Specifically, we studied whether females provide egg protection by avoiding oviposition under the risk of egg parasitization. We expected a lower oviposition rate under parasitoid presence, and the eggs, if any, to be placed preferably on conspecifics and not on plants, thus ensuring higher survival of the progeny. The results show that P. laciniata 's egg-laying rate was lower when they were enclosed with parasitoids than when parasitoids were absent, especially when plants were the only substrate to oviposit on. Moreover, females showed strong preference for laying eggs on conspecifics rather than on plants. Egg-laying in P. laciniata appears to be not only influenced by the availability of conspecifics, but also by the presence of egg parasitoids. This indicates that females may be able to detect G. bolivari and avoid oviposition when parasitoids are present. We discuss the possibility of conspecifics as enemy-free space.     

Larval competition between Aphidius ervi and Praon volucre (Hymenoptera: Braconidae: Aphidiinae) in Macrosiphum euphorbiae (Hemiptera: Aphididae)

Interspecific competition between parasitoid larvae may influence the size, structure, and stability of the population, leading to a reduction in total parasitism and thus restricting the pest control. Aphidius ervi (Haliday) and Praon volucre (Haliday) are endoparasitoids that possess a wide host range and present considerable potential for the biological control of the aphid Macrosiphum euphorbiae (Thomas). The larval competition between A. ervi and P. volucre, and the possible intrinsic competitive superiority of one of the parasitoids in M. euphorbiae, have been studied. In single parasitism experiments, mated parasitoid females (n=10) were maintained individually in contact with M. euphorbiae hosts (n=30) inside petri dishes containing lettuce leaf discs and maintained in environmental chamber at 22 ± 1°C, 70 ± 10% RH, and 12-h photophase. The multiple parasitism experiments consisted of exposing single parasitized aphids (n=120) to the second parasitoid species. Two oviposition events were performed with a 4-h interval between them, namely the following: sequence A (oviposition by A. ervi, followed by P. volucre) and sequence B (oviposition by P. volucre, followed by A. ervi). Oviposition sequence A generated 24 A. ervi and 55 P. volucre adults, whereas oviposition sequence B generated 23 and 49 adults. P. volucre is an intrinsically superior competitor compared with A. ervi, and the use of the two species simultaneously may result in competitive exclusion and influence the stability of the parasitoid population.     

Hyperparasitism in a Generalist Ectoparasitic Pupal Parasitoid,Pachycrepoideus vindemmiae (Hymenoptera: Pteromalidae), on Its Own Conspecifics: When the Lack of Resource Lead to Cannibalism

Hyperparasitism is a normal behavior of parasitoids, which often happens among species. Conspecific hyperparasitism, such as some kinds of heteronomous hyperparasitic behaviors, has been only reported in some species belonging to Aphelinidae. In this article, the conspecific hyperparasitism of Pachycrepoideus vindemmiae (Pteromalidae) is reported, with Drosophila puparia as hosts. Hosts were exposed to P. vindemmiae females twice to parasitism with nine, twelve, and fifteen day intervals between the two exposures. None of the infested hosts emerged more than one offspring, and emergence of parasitoid offspring occurred in two obvious events, synchronously with the exposure time intervals, which suggested that offspring emerging during the first and second events would come from the primary and secondary parasitoids, respectively, and the inference with the developmental duration of offspring also indicated this. With two P. vindemmiae strains that could be identified by a simple sequence repeat marker, the above speculation of the origin of those offspring emerging during the two events was confirmed. Dissection of hosts exposed twice revealed a cannibalism behavior of larvae from the secondary foundresses on the primary conspecific pupae. Our results suggested a conspecific hyperparasitism behavior of the secondary parasitoids on the primary conspecifics. Measures showed a reduced body size for the adults from the conspecific hyperparasitism. Foundresses from the conspecific hyperparasitism had less fitness variables than those from primary parasitism, with shorter longevity, less life time fecundity, lower values of infestation degree, and lower success rate of parasitism. However, when the parasitoids from the conspecific hyperparasitism met healthy Drosophila puparia, their offspring would recover to normal size. Frequency of the conspecific hyperparasitism behavior enhanced with the decreasing of proportion of healthy hosts in the oviposition patch. The conspecific hyperparasitism of P. vindemmiae on the primary conspecifics would be helpful to last the population when healthy hosts are absent in the oviposition patch.     

Impact of superparasitism time intervals on progeny survival and fitness of an egg parasitoid

1.Superparasitism influences the fitness of female parasitoids and their progeny, and increasing time interval between oviposition bouts generally reduces survival probabilities of the second clutch. However, the timing of superparasitism may, under certain conditions, favour the second clutch. 2. This study investigated the effects of superparasitism time intervals on survival and fitness of both clutches, allowing the egg parasitoid Trichogramma euproctidis to parasitise previously parasitised Ephestia kuehniella host eggs at different time intervals. 3. In short intervals (0–1 h), a significant advantage was found for the second clutch over the first clutch (93.1% survival). In contrast, the second clutch was outcompeted by the first clutch in 17–19 h and 27–29 h intervals. Females deposited their eggs into larvae (intraspecific hyperparasitism) in a 39–41 h interval with a probability of survival of 57.1%. Females mostly refused to hyperparasitise pupae (～80% rejection at 124–126 h), and when they did, their progeny never survived. 4. Hyperparasitism significantly increased parasitoid mortality in both clutches from less than 20% (superparasitism only) to over 35%. 5. Except on newly laid eggs (0–1 h), superparasiting females were frequently observed attempting to stab immatures of all stages (from 36.4% to 89.4% of all ovipositions depending on treatment), but infanticide only appeared to succeed on larvae (39–41 h). 6. When the second clutch survived, emerging parasitoids were smaller than control individuals, probably due to resource depletion. 7. These results suggest that T. euproctidis females can detect that a host has been previously parasitised but they cannot perceive superparasitism time intervals.     

Reproductive biology and small-scale rearing of cherry bark tortrix and its candidate biological control agent

Abstract:  The larval parasitoid, Campoplex dubitator (Hym.,: Ichneumonidae), is under consideration as an agent for classical biological control of cherry bark tortrix (CBT), Enarmonia formosana (Lep.,: Tortricidae), in North America. A comprehensive risk analysis of the candidate agent will require prosperous cultures of both the pest and its parasitoid. We present a rearing method for small-scale production of both species using a bean-based artificial diet, with additional information on the reproductive biology of C. dubitator . Based on estimated survivorship probabilities, a CBT egg had a 70% chance of developing completely to the adult stage under this system. The success of parasitism, however, was very dependent on the instar of the CBT host larva at the time of oviposition. All parasitised first instar larvae died shortly after the attack, rendering them unsuitable for oviposition, while 50% of parasitised second instar larvae died prematurely. In contrast, early mortality was 15–30% for larvae parasitised in the third to fifth instars. Regardless of the instar at oviposition, approximately 90% of the surviving hosts yielded parasitoids, showing a high acceptance by C. dubitator of second to fifth instars for oviposition.     

Cannibalism of parasitoid‐attacked conspecifics in a non‐carnivorous caterpillar

Cannibalism, the killing and consumption of conspecifics, can even occur in insect species typically considered to be non‐carnivorous. Of particular interest is the cannibalism of parasitoid‐attacked conspecifics, which could reduce parasitism levels in subsequent generations for that conspecific population. This study reports on the occurrence and some of the consequences of cannibalism in parasitoid‐attacked obliquebanded leafroller, Choristoneura rosaceana (Harris) (Lepidoptera: Tortricidae). We show that larvae of C. rosaceana, which is considered to be an herbivorous caterpillar species, did not prey upon live conspecifics, but readily consumed conspecifics attacked by Habrobracon gelechiae Ashmead (Hymenoptera: Braconidae). Further examination found that C. rosaceana larvae feeding on parasitoid‐attacked conspecifics, since their fourth instar, suffered a higher mortality and reduction in body size than those fed on plant material only. The cannibalism of attacked conspecifics did not appear to offer any nutrient benefits for the cannibal. To our best knowledge, this is the first empirical example of the occurrence and some of the consequences of cannibalism by a non‐carnivorous insect on its parasitoid‐attacked conspecifics. We discuss the adaptive significance of such cannibalism on parasitoid‐attacked conspecifics with respect to a trans‐generational fitness gain for the population through the killing of the parasitoids, thereby reducing parasitism in subsequent generations.     

Bottom-up cascading effects in a tritrophic system: interactions between plant quality and host-parasitoid immune responses

Abstract.  1. Little is known about underlying mechanisms by which plants indirectly affect parasitism success in hymenopteran endoparasitoids. The hypothesis that host-plant effects can challenge the innate immune system of an insect host was experimentally tested in this study using a model tritrophic, crucifer – lepidopteran [ Plutella xylostella (L.)] – parasitoid [ Cotesia plutellae (Kurdjumov)], system.
2. The effects of host-plant suitability on herbivore performance and parasitism were examined. The bottom-up effect of plant suitability on host-parasitoid immune responses was then evaluated using measures of cellular and humoral effectors.
3. Host-plant quality showed a significant effect on the encapsulation response of P. xylostella to first instar but not to second instar parasitoid larvae. Encapsulation was never sufficient to prevent parasitoid emergence.
4. Poor host-plant suitability suppressed phenoloxidase activity in the absence of the parasitoid. The suppressive effect of C. plutellae on phenoloxidase activity was much greater and no plant effects were detectable after insects had been parasitized.
5. Despite strong plant effects on parasitism, those on immune effectors of the host were transitory or overwhelmed by the effect of the parasitoid.
6. These results demonstrated that plant-mediated variation in parasitism success by C. plutellae were not as a result of plant nutritional status or other attributes affecting the immune function of P. xylostella , nor to host-plant effects on superparasitism.
7. In these experiments, P. xylostella was a fully permissive host to C. plutellae and host-plant-mediated effects on the innate immune response appeared to play no part in parasitoid survival within hosts.     

斑痣悬茧蜂寄生、发育和生殖特征与寄主幼虫日龄的关系

【目的】“圆屋顶形”假说认为,对单寄生性姬蜂和茧蜂适合度而言,中间龄期幼虫寄主的品质高于更早和更晚龄期幼虫。该假说得到许多研究支持,但这些研究常以寄主幼虫脱皮划分虫龄,很少观测生殖特征,从而难以确切和全面描述适合度随寄主生长发育变化而变化的关系。本研究旨在检验“圆屋顶形”假说。【方法】本研究以斜纹夜蛾Spodoptera litura不同日龄幼虫为寄主,观测斑痣悬茧蜂Meteorus pulchricornis寄生和发育特征,并测定成蜂生殖力。【结果】线性回归分析表明,雌蜂对中间日龄寄主幼虫的寄生率大于对两端日龄寄主幼虫的寄生率;蜂卵至成虫的存活、成虫体型大小及其生殖力（产卵量）等适合度相关特征均表现出中间日龄寄主幼虫处理大于两端日龄幼虫处理。【结论】研究结果支持“圆屋顶形”假说。     

Offspring production and self-superparasitism in the solitary ectoparasitoid Spalangia cameroni (Hymenoptera: Pteromalidae) in relation to host abundance

Parasitoid fitness strongly depends on the availability and quality of hosts, which provide all resources required for larval development. Several factors, such as host size and previous parasitation, may affect host quality. Because self-superparasitism induces competition among a female's offspring, it should only occur if there is an imperfect recognition of self-parasitized hosts or if there is a fitness advantage to self-superparasitism. Against this background, we investigated self-superparasitism and offspring production in Spalangia cameroni (Hymenoptera: Pteromalidae) in relation to the abundance of a novel host, Ceratitis capitata (Diptera: Tephritidae). Individual pairs of parasitoids were provided with either two (low host abundance) or ten (high host abundance) pupae per day. Under high host abundance, lifetime fecundity (number of eggs laid), offspring number, number of pupae parasitized and hosts killed were greater than under low host abundance, whereas the number of eggs per host was lower; and the proportion of hosts that did not produce offspring tended to be lower. The latter suggests the occurrence of ovicide, when hosts are scarce due to an at least imperfect recognition of previously self-parasitized hosts. Offspring production per parasitized pupa was higher when hosts were scarce and levels of self-superparasitism high, suggesting the existence of beneficial effects of self-superparasitism.