Summary of Reviewers' Comments









Richard Stouthamer1, and E. Fred Legner2



1Department of Entomology, Wageningen Agricultural University

P.O. Box 8031, 6700EH Wageningen, The Netherlands


2 Department of Entomology, University of California, Riverside,

CA 92521, USA (



          Studies with new field isolates of a Peruvian strain in 1995 by Richard Stouthamer et al. (unpublished) have shown a greater involvement of larval cannibalism and much complexity in these species' reproduction.  Indeed survival mechanisms in parasitoids include many behaviors; among which increased cannibalism by more aggressive larvae may be triggered during times of host scarcity.  The pteromalid parasitoid Muscidifurax raptorellus Kogan & Legner is polymorphic for its oviposition behavior, fly pupae parasitized by the solitary form always give rise to one offspring, while fly pupae parasitized by the gregarious form give rise to more than one offspring in approximately 60% of the hosts. Legner has studied the inheritance of gregarious emergence, i.e. more than one wasp emerges per host, in great detail.  He found this trait to be polygenically controlled by 2‑19 genes.  Females carrying a higher percentage of their genome from the gregarious form have a higher percentage of their offspring emerging gregariously from a host.  More surprisingly he also found that the father's genetic background influenced the rate at which his progeny emerged gregariously.  Males from a gregarious line, when mated with females from a solitary line, caused the females to start having offspring more gregariously.  The opposite effect takes place when males from the solitary line are used.  Legner's hypothesis was that the males transferred some behavior modifying substance with their sperm to the female, which resulted in a higher or lower level of gregarious oviposition.  Here an alternative hypothesis is tested in which the level of gregarious emergence from hosts is determined by not only the number of eggs a female oviposits in a host, but also by the level of larval aggression.  The father's influence on the level of gregarious emergence in this hypothesis only takes place through his contribution to the genetic makeup of the larvae, with the larvae with a larger fraction of their genome from the solitary line showing higher levels of larval aggression.  ADD:  Under our experimental conditions the identity of the male did not influence the level of gregarious oviposition by a female, but it did have a significant influence on the number of larvae that survived.  The maternal behavior modification hypothesis did not apply significantly in this case where the cultures originated from new field acquisitions, the duration of the experiment was shorter and the parasitization environment was different than in the earlier investigations.


Keywords: oviposition behavior, larval aggression, clutch size




Species of the pteromalid genus Muscidifurax are parasitoids of synanthropic flies.  The genus is indigenous to the Americas where it shows great diversity (Kogan & Legner 1970).   Typically most species superparasitize with 2-3 eggs but only one wasp emerges per host.  However, in a South American species Muscidifurax raptorellus Kogan & Legner, two populations are known that differ in the number of wasps emerging per housefly pupa (Legner 1987a,b).  The form originating from Chile produces offspring gregariously, i.e. from approximately 60% of the hosts more than one wasp emerges while up to 17 wasps have been known to emerge from a single housefly pupa (Legner 1987b).  However, the form originating from Peru rarely produces more than one offspring per host.  The genetics of level of gregarious oviposition has been studied in detail by Legner (1987, 1988a,b; 1989a,b,c; 1991a,b; 1993).  The level of gregarious emergence of the offspring of a virgin female was found to be determined by the proportion of the genome originating from the gregarious form (Legner 1987b).  The higher this proportion the higher the level of gregariousness.  This trait appears to be polygenically determined and 2‑19 genes are involved in coding for this character (Legner 1991a).  More surprisingly, the level of gregariousness in offspring emergence was not only a function of the mother’s genetic makeup but the father's genetic background also influenced the level of gregariousness in which his offspring emerged.  Considerable amounts of data indicated that the male actually influences the females oviposition behavior in such a way that if the male originates from a gregarious line his mate lays her eggs more gregariously, and female lays fewer eggs per host when she mates with males originating from the solitary line (Legner 1987b, 1988b, 1993).  High levels of variation in these experiments suggested that other mechanisms might interplay in the paternal influencing of maternal behavior.  A number of possible behavior modifying substances, that could be transferred from the male to the female through the seminal fluid, have been suggested (Legner 1987b), these ­include hormones and bacteria.  Because males of the solitary line reduce the gregariousness of the offspring, and males of the gregarious line increase it one would have to assume the presence of two different substances: one that increases the level of gregarious oviposition and another that decreases this level.


Another hypothesis to explain level of gregarious emergence suggests that it is determined by two factors: a) the number of eggs a female lays per host and b) the level of aggressive interactions between the larvae emerging from the eggs.  The influence of the father in this hypothesis is primarily through his genetic contribution to his offspring (factor b), while under Legner's (1987b) original hypothesis the number of eggs a female lays is influenced by factor a.  These hypotheses were discriminated by mating females to males either of the gregarious form or of the solitary form and by determining  in half of the hosts the number of eggs that were oviposited and in the other half the number of wasps that emerged.  The mothers in these tests were  F1 hybrid females between the gregarious and solitary line.  These females were chosen because they are known to show both kinds of responses; i.e. when mated to a male of the gregarious line the females produce more offspring per host than when allowed to oviposit as virgins (Legner 1987b), and when such hybrid females are mated to solitary males they produce fewer offspring per host than virgin females.


Materials and Methods


Culture origin:  New field collections of Muscidifurax raptorellus from Peru and Chile were obtained in 1995.  FROM WHAT SPECIFIC AREA AND TIME OF YEAR IN EACH COUNTRY WERE THE CULTURES OBTAINED? A Peruvian/Chilean hybrid and a Chilean strain used in these experiments were maintained on pupae of the housefly (Musca domestica L.).  The hybrid was formed by mating large numbers of virgin Chilean females to Peruvian males followed by propagation of the culture for 6 generations, after which the hybrid continued to reproduce solitarily in mass culture.  ADD REFERENCE TO Kogan & Legner (1970).


Experimental setup: All experiments were done at 25EC and 24 hrs light.  Hybrid females between the solitary and gregarious lines were secured by mating solitary females with gregarious males.  Daughters of this cross were used in the experiment, where they were randomly assigned to one of three treatment groups.  The treatment groups consisted of mating the females to males of either (a) gregarious or (b) solitary strain or (c) leaving them unmated.  To each group respectively 16, 13 and 21 females were assigned.  After 24hrs the males were removed and each female was given 20 hosts daily for oviposition  The hosts of each day were randomly assigned to two groups of each 10 hosts.  The hosts of the first group were opened within 48 hrs after parasitization had taken place to determine the number of eggs laid per host.  The hosts of the second group were individually placed in a gelatin capsule (size 000) for wasp emergence.  In total the parasitization of these wasps was followed for 5 days, thus per mother 50 hosts were opened to determine the number of eggs laid per host and 50 hosts were kept for emergence.  WHAT ARE THE DIMENSIONS (or volume) OF THE HOUSEFLY PUPARIA?  -- WHAT IS THE PARASITIZATION ENVIRONMENT? (e.g., In screened polystyrene vials (46 cm3), with a basal are of 7 cm2, as in earlier experiments)


The experiment was done in such a way that the mother's group ­identity, i.e. whether she was mated to a solitary male or a gregarious male, was not known to the person who did the egg and wasp counts.  This avoided prior knowledge from influencing the results.


Analysis of variance was performed on the data derived from the two samples of fifty hosts that were either opened for egg counts or were left for wasp emergence for each mother: the percentage of the hosts that had been opened containing more than one egg, the percentage of hosts from which more than one wasp emerged, the difference between these two percentages, the total number of eggs in the egg sample, the total number of wasps in the wasp sample and the difference between the total number of eggs and the total number of wasps, the total number of parasitized hosts, i.e. those containing one or more eggs, the total number of parasitized hosts resulting in adult offspring and the difference between these two measurements per female.  No transformation was done for the statistical analysis of the data.  To determine the significance of the differences between means a Duncan's Multiple range test was used with alpha =0.05.




          Superparasitism (= insertion of more than one parasitoid egg per host) occurs in both the Peruvian and Chilean strain, and subsequent cannibalism by hatched larvae always follows.  The Peruvian strain deposits a lower number of eggs per host than the Chilean strain but cannibalism intensity is higher.  Therefore, it is difficult to determine the exact number of eggs initially deposited by both species.  The number of adult parasitoids that survive is always less in the Peruvian strain and usually averages about one.  On the contrary, more adult survivors usually occur in the Chilean strain, averaging about seven at a host density of 20 per 24 hrs.  A standardization of host density, size, age and duration of exposure to parasitization is essential in experiments as they influence the number of eggs deposited and the rate of cannibalism.





The results shown in table 1 indicate that the percentage of hosts in which more than one egg was found did not differ significantly between females mated with a male from the gregarious line or the solitary line.  There was a significant difference however between females mated to solitary males versus virgin females in the proportion of the eggs that were placed gregariously on a host.  The percentage of wasps that emerged gregariously did however differ significantly between all groups with that percentage being the highest for the offspring of the gregarious male, intermediate for the offspring of the virgin females and the lowest for the offspring of the solitary males. The difference between the percentages of hosts with gregarious eggs vs gregarious wasps also differed between the three groups.  The group mated to the gregarious males showed the smallest difference, the virgin females were intermediate while the offspring of the solitary males showed the largest reduction in the percentage gregariousness.


There were no significant differences between the groups in the total number of eggs per 5 day sample (table 2), however the number of wasps that emerged from the five day sample did differ significantly: the gregarious group had a significantly higher number of adult progeny than the solitary or virgin group. These two groups however did not differ from each other in this characteristic.  The difference between the number of eggs per five-day sample and the number of wasps per sample differed among all groups with the gregarious group showing the smallest reduction in numbers and the solitary the highest.


Finally there appeared to be no difference in the number of hosts that contained eggs between the groups nor in the number of hosts giving rise to wasps or in the difference between the numbers of hosts containing eggs versus those giving rise to wasps (table 3).




Indeed survival mechanisms in parasitoids include many behaviors; among which increased cannibalism by more aggressive larvae may be triggered during times of host scarcity.  In the current experiments the number of eggs laid per host by a hybrid the type of male to which she was mated did not appreciably influence female.  However, the genetic background of the male did influence the percentage of hosts from which more than one wasp emerged.  Consequently, the hypothesis that the male influences the female’s egg laying behavior has to be rather in favor of the hypothesis that larvae with a large proportion of their genome stemming from the solitary line somehow reduce the number that will eventually emerge from a host.  The number of eggs laid per mother did not differ between the groups, and neither did the number of hosts that has been parasitized.  Therefore, the reduction in the number of hosts from which wasps emerge gregariously was caused by a reduction during the larval stage.  Clearly, the offspring of the solitary males had the highest mortally, followed by that of the virgin females and the least mortality took place in the offspring of the gregarious males.  Several authors have reported that aggressive interactions take place between the larvae of the species M. raptor Girault & Sanders.  Wylie (1971) found that larvae of M. raptor will spend approximately the first ten hours after hatching moving about on the host while occasionally feeding on the pupa.  If a host has several eggs on it, the larva that hatches first attacks and kills most or all of the eggs.  If two larvae happen to survive the first instar, combat will take place in the second instar.  Podoler and Mendel (1977) report similar observations for M. raptor.  Legner (1987b) also reports that <2% of the hosts parasitized by the solitary form of M. raptorellus receive more than one egg, but rarely more than one wasp emerges.  This could also indicate that larval aggression occurs in this species and indeed aggression was observed between the larvae of the solitary form of this species.  Apparently the larval combat generally leads to only one wasp emerging per host, (table 3).


The level of aggressiveness between the larvae can be influenced to a maximal extent when males of the different lines are crossed with hybrid females between these lines.  The fertilized eggs result in female larvae that have on average either 75% of their genome from the solitary line ((S & X G %)& X S %)& or 25% of the solitary line ((S& X G%)& X G%)&.  In the first case the larvae apparently engage in combat which results in a reduction in the percentage of hosts that give rise to more than one offspring (fig. 1c).  While in the second case the larvae show less aggressive interactions, and per host, more of the eggs survive to give rise to adult wasps (fig. 1a).  The offspring of the virgin females is intermediate because all of their offspring will have on average 50% of their genome stemming from the solitary line (fig 1c.).  There could also be an influence of the father on the survival of male and female larvae in these crosses.  In the case of solitary fathers the fertilized eggs, i.e. daughters have on average 75% of the genome of the solitary line and can be expected to fight more than their male siblings, which have 50% of their genome from the solitary line.  In the case of offspring of the gregarious males, their daughters have only 25% of their genome coming from the solitary line while the males have 50% of their genome coming from the solitary line.  Consequently, in those cases where one male and one female are placed in a host in case of the "gregarious" group it would be expected that the one male one female combination would be more prevalent than in case of the "solitary" group, while at the same time the one male group should be larger in the gregarious line than in the solitary line. Although indeed these trends are visible in the data (table 4) they are rather subtle, probably mainly because the fraction of the hosts receiving one male and female egg is limited in these crosses.


The gregarious emergence is the result of two, possibly linked, effects: the number of eggs a female oviposits per host, this effect is determined purely by the female’s genome and secondly the level of aggressive interactions between the larvae, this level is influenced by the genetic background of both parents.  Females of the solitary line lay fewer eggs per host than the females of the gregarious line.  It is suspected that larvae of the gregarious line show little aggressive behavior towards each other, while the larvae of the solitary line are aggressive. The gregarious strain also superasitizes:  that is, more eggs are laid in a host than will survive to adulthood.  The cause of the mortality in the ones that do not survive could be due to genetic inferiority or to some degree of aggressiveness. These two groups of genes could interact in such a way that they would enhance each other.  The higher the percentage of the genome stemming from the gregarious line the higher the number of hosts with more than one egg and the lower the level of larval aggression, in contrast the lower the percentage of the gregarious genome the lower the level of gregarious oviposition and the higher the level of larval aggression.  Therefore, one would expect a slight s shaped curve to describe the relationship between the percentage of the genome from the gregarious line in a virgin mother and the percentage of the hosts giving rise to more than one offspring.  When this relationship (Legner 1987b, 1991a) is inspected indeed such a deviation, albeit slight, is found.  The findings of Legner (1987b, 1991a) on the number of genes influencing the gregarious and solitary oviposition should be reexamined in the light of these two groups genes being involved.  The finding that within one species both solitary, fighting forms and gregarious, non‑fighting forms appears to be unique to M. raptorellus.  This will allow the testing of some of the theory developed by Godfray (1987) on the evolution of solitary and gregarious oviposition.


There is also a heterosis in the F-1 hybrids that is not explained in this discussion.  The heterosis is expressed in several ways:  higher numbers of offspring and hosts killed, etc.  Also, backcrossing data suggest that at least eight loci were actively segregating for gregariousness behavior.






Support for P. Strippentow from a European Union Tempus grant is gratefully acknowledged.  Conversations with Molly Hunter and Marcel Visser greatly helped to clarify our thinking about this phenomenon.  Leo Koopman, André Gidding and Frans van Aggelen kindly supplied host for these experiments.





Godfray, H. C. J.  1987.  The evolution of clutch size in parasitic wasps.  Am. Nat. 129: 221‑233


               ADD:   Kogan, M. & E. F. Legner.  1970.  A biosystematic revision of the genus Muscidifurax (Hymenoptera: Pteromalidae) with descriptions of 

                            four new species.  Canad. Entomol. 102(10):  1268-1290.


Legner, E. F.  1987a.  Further insights into extranuclear influences on behavior elicited by males in the genus Muscidifurax.  Proc. 52 annual conf.

     Cal. Mosquito and Vector Control Association: 127‑130.


Legner, E. F.  1987b.  Inheritance of gregarious and solitary development in Muscidifurax raptorellus.  Can. Ent. 119: 791‑808.


Legner, E. F.  1988a.  Hybridization in principal parasitoids of synanthropic diptera: The genus Muscidifurax. Hilgardia 56(4): 36pp.


Legner, E. F.  1988b.  Muscidifurax raptorellus females exhibit postmating oviposition behavior typical of the male genome.  Ann. Entomol. Soc.

     Am. 81: 522‑527


Legner, E. F.  1989a.  Wary genes and accretive inheritance in Hymenoptera.  Ann. Entomol. Soc. Amer. 82: 245‑249.


Legner, E. F.  1989b.  Paternal influences in males of Muscidifurax raptorellus.  Entomophaga 34: 307‑320


Legner, E. F.  1989c.  Phenotypic expression of polygenes in Muscidifurax raptorellus, a synanthropic fly parasitoid. Entomophaga 34: 523‑530.


Legner.  E. F.  1991a.  Estimations of number of active loci, dominance and heritability in polygenic inheritance of gregarious behavior in

     Muscidifurax raptorellus.  Entomophaga 36: 1‑18


Legner, E. F.  1991b.  Recombinant males in the parasitic wasp Muscidifurax raptorellus. Entomophaga 36: 173‑181


Legner, E. F.  1993.  Theory for quantitative inheritance of behavior in a protelean parasitoid, Muscidifurax raptorellus (Hymenoptera:  

     Pteromalidae).  Eur. J. Entomol. 90:  11-21.


Podoler, H. & Z. Mendel.  1977.  Analysis of solitariness in a parasite-host system (Muscidifurax raptor – Ceratitis capitata).  Ecol. Entomol 2: 



Wylie, H. G.  1971.  Observations on intraspecific larval competition in three hymenopterous parasites of fly puparia. Can. Ent. 103: 137‑142





Table 1.  Mean and standard error of the percentage of hosts that contained more than one egg (egg/hst), from which more than one wasp emerged (wasp/hst), and the difference between these values for hybrid females that had mated either with a gregarious male, had remained virgin or had mated with a solitary male.


n                >egg/hst                                      >wasps/hst                                  difference


gregarious male                           16              38.4 (2.39)                                   28.7 (2.13)                                   9.8  (2.93)


virgin                                             21              33.8 (2.08)                                   12.3 (1.86)                                   21.6  (2.56)


                  solitary male                                13              43.1 (2.65)                                   5.3 (2.36)                                                   37.8  (3.25)




Table 2. Mean and standard error of the total number of offspring (eggs or wasps) and the difference between these numbers, for hybrid females that had mated either with a gregarious male, had remained virgin or had mated with a solitary male.



n                >egg/hst                                      >wasps/hst                                  difference


gregarious male                           16              60.6 (2.25)                                   54.8 (1.97)                                  5.8 (1.95)


virgin                                         22                  57.0 (1.97)                                   43.2 (1.72)                                   13.8 (1.71)


                  solitary male                                13              58.8 (2.51)                                   38.7 (2.19)                                   20.1 (2.17)




Table 3.  Mean and standard error of the number of parasitized hosts indicated by the presence of one or more eggs, by the emergence of one or more wasps and the difference between these numbers, for hybrid females that had mated either with a gregarious male, had remained virgin or had mated with a solitary male.


n                 hosts with egg                            host with wasps                         difference


gregarious male                           16              41.6 (1.24)                                   40.3 (1.40)                                   1.3 (0.93)


virgin                                             21              40.9 (1.08)                                   38.9 (1.22)                                   2.1 (0.81)


solitary male                                13              39.2 (1.37)                                   37.3 (1.55)                                   1.9 (1.03)




Table 4. Total egg and wasps distribution per host for females that had either mated with males from a gregarious line, had remained virgin or had mated with males from a solitary line.  Egg allocation per host, represented as the percentage of all hosts in egg sample containing either 1, 2, 3 or 4 eggs.  Wasp distribution per hosts represented as the percentage of hosts containing a male (M) or a female (F) or a combination of the two.



No.                              Sample dissected for egg counts             


mothers     ---------------------------------------------------------


        % of hosts with 1-4 eggs                  


No.                               1                2                3                4



gregarious male                           16              677                             62.5          32.2          5.0             0.3


                                virgin                                            22                  855                      65.1              30.4           3.7             0.7


solitary male                                13              507                      56.8                 36.9          5.3             1.0




No.                              Sample left for wasp counts

of                                 -----------------------------------------------------------------

mothers                       % of hosts from which males and/or females emerged


No.             1M            1F              2M            2F              1M1F        1M2F        3M            3F



gregarious male                           16              661           10.8          59.6          0               23.1          3.5            0.3            0               2.6


                              virgin                                             22 803            87.5          0           12.1              0                 0                0         0.4                 0


solitary male                                13              483           9.3            86.1         0               3.5             1.0              0              0                 0