Behavioral Ecology Vol. 10 No. 2: 122-127
© 1999 International Society for Behavioral Ecology
Don't count your eggs before they're parasitized
contest resolution and the trade-offsduring patch defense in a parasitoid wasp
a Department of Crop Protection, University of Adelaide, Waite Campus, South Australia b Department of Applied Mathematics, University of Adelaide, South Australia
Address correspondence to S. A. Field, Department of Entomology, Hebrew University of Jerusalem, PO Box 12, Rehovot, 76100, Israel. E-mail: sfield{at}agri.huji.ac.il
Received 7 August 1997; revised 18 April 1998; accepted 29 May 1998.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Although aggressive conflicts over hosts occur among females in many species of insect parasitoids, few studies have examined the mechanisms by which these contests are resolved. In Trissolcus basalis, a parasitoid of pentatomid bug egg masses, females co-exploiting an egg mass (patch) encounter one another repeatedly and fight for possession of the patch. We investigated the resolution of pairwise contests by experimentally varying the release time of females and the size of the patch. Logistic regression showed that the female arriving first on the patch was more likely to win both the first agonistic encounter and to retain overall possession of the patch. This advantage to the first female suggests a resource-correlated asymmetry in favor of the first female, due to her having invested more offspring in the patch. Although escalations were more common when the asymmetry in arrival times was small, the majority of encounters within all contests were nevertheless resolved without escalated fighting, with the resident attacking and the intruder backing down. Thus contest resolution basically followed a "bourgeois" rule. Residents tolerated intruders more frequently when the patch size was larger and the second female was released later, illustrating the trade-off faced by residents between defending the patch and continuing to exploit it.
Key words: agonistic behavior, asymmetries, bourgeois strategy, egg parasitoid, Trissolcus basalis.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
The occurrence of ritualized fighting and conventional settlement of animal conflicts was a problem largely solved by the development of the theory of evolutionarily stable strategies (ESSs; Maynard Smith and Price, 1973
). Initial studies showed that such behavior between evenly
matched opponents could evolve by frequency-dependent selection acting at the
level of the individual (Maynard Smith,
1974; Maynard Smith and Price,
1973). The inclusion in models of asymmetries between contestants,
such as differences in resource value, fighting ability, and even
"arbitrary asymmetries," revealed further conditions that favor
the settlement of contests without injurious fighting (e.g.,
Hammerstein, 1981;
Leimar and Enquist, 1984;
Maynard Smith and Parker,
1976). Empirical work has confirmed that asymmetries are often
used to settle contests in nature (reviewed in
Enquist and Leimar, 1987;
Leimar and Enquist, 1984),
although under certain ecological conditions injurious and even fatal fighting
can evolve (Enquist and Leimar,
1990; Grafen,
1987).
Animal contests are often seen as "single encounter" events, in
which two individuals contest a resource such as a mate or food item, and
possession of the resource is settled decisively at the first meeting. The
resource is then utilized by the winner and is of little or no further value
to the loser. However, in insect parasitoids contesting a patch of hosts, the
resources under issue cannot be consumed immediately by the winner of a fight
but remain in situ and can be usurped by competitors through superparasitism
(laying of an egg in a host previously parasitized one or more times; see
van Alphen and Visser, 1990).
Thus the resource may remain of considerable value to the loser of a fight for
many hours or even days (Field et al.,
1997; Sirot, 1996;
Visser et al., 1992). This
makes contests in parasitoids, especially those in which hosts are clumped
into patches of defendable size, likely to consist of a series of repeated
interactions between the same individuals.
There are many empirical reports of aggressive contests between female
parasitoids (reviewed in Griffiths and
Godfray, 1988; Godfray,
1994; Waage,
1982), but analyses of the mechanisms of contest resolution are
rare. In one well-studied example, the gregarious bethylid wasp Goniozus
nephantidis (Muesebeck), both larger size and holding ownership status
increased the chances of winning contests over paralyzed hosts
(Hardy and Blackburn, 1991;
Petersen and Hardy, 1996). The
effect of ownership status has been interpreted as an asymmetry related to
resource value. It was suspected that owners were maturing eggs while they
guarded a paralyzed host, increasing the relative value of the host to the
owner (Petersen and Hardy,
1996). Less substantial evidence for a size advantage was also
presented by Lawrence (1981),
who noted that larger females of the braconid Biosteres longicaudatus
(Ashmead) tended to aggressively displace smaller ones as they foraged on
high-density host patches. In the ichneumonid Venturia canescens
(Gravenhorst) (Hughes et al.,
1994), the braconid Coeloides filiformis (Ratz.), and
other parasitoids of scolytid bark beetles, if a female is disturbed by a
conspecific while examining a host, she usually retains her position
(Mills, 1991). These studies
suggest some kind of asymmetry or convention favoring the resident
individual.
This paper examines the mechanisms of contest resolution in the scelionid
parasitoid wasp Trissolcus basalis, which parasitizes the egg masses
of pentatomid bugs (Hemiptera: Pentatomidae). The host eggs are laid in rafts
that constitute defendable patches (see
Waage, 1982;
Wilson, 1961). Females often
co-exploit patches in the field, and superparasitism by conspecifics is common
(Field et al., 1998b). Females
lay one egg per oviposition and can discriminate unparasitized hosts from
hosts parasitized by themselves (self-discrimination)
(Wilson, 1961), but they
cannot discriminate between hosts parasitized by themselves and conspecifics
(conspecific discrimination) (Field and
Keller, 1999). As the offspring of a superparasitizing female have
a high probability of survival if the time interval between ovipositions is
short, patch defense has clear advantages for the first female to oviposit
(see Field et al., 1997).
In T. basalis, previous authors have noted that in pairwise
contests, one female eventually achieves dominance on the egg mass and
excludes the other (Cumber,
1964; Wilson,
1961). However, the mechanisms by which such dominance is achieved
have not been examined, nor has a satisfactory evolutionary explanation for
the phenomenon been suggested. The present investigation addressed these
problems by studying the behavioral rules used by females during contests and
by considering the resultant emergence of resident/intruder roles in an
evolutionary context.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
T. basalis were initially reared from egg masses of the horehound bug, Agonoscelis rutila (Hemiptera: Pentatomidae), collected near Adelaide, South Australia. They were subsequently reared on A. rutila as described by Field et al. (1997
). Females aged between 2
and 8 days were used in experiments, and in contests both females were of the
same age. They were isolated from conspecifics 24 h after emergence.
Behavioral observations
We observed agonistic behavior in the laboratory by releasing a pair of
females into a small petri dish containing an A. rutila egg mass.
Observations were illuminated by a fiber-optic light source and videotaped
using a camera mounted overhead. Continuous time records of behavior were made
using a TRS-80 Model 100 portable computer programmed with event-recording
software (The Observer, Noldus Information Technology, Wageningen, The
Netherlands). Individuals were distinguished by a colored mark of fluorescent
dust applied to the dorsal mesosoma.
To explore the mechanisms of contest resolution, we varied the patch size and the release time of females to generate asymmetries in resource value and residency status. The four experimental treatments were 12A (12 hosts per patch, females released simultaneously, n = 12); 24A (24 hosts, females released simultaneously, n = 12); 24B (24 hosts, second female released after the first had parasitized approximately half the hosts in the patch, n = 12); and 24C (24 hosts, second female released after the first had parasitized all the hosts in the patch, n = 13). Females sometimes spent some time searching in the arena before reaching the egg mass, and even when released simultaneously, one female often arrived on the egg mass slightly (up to 5 min) before the other.
Three types of encounters between females were distinguished. The first type was nonagonistic encounters, in which females either physically contacted one another or came into extremely close proximity but exhibited no aggression. The second and third types were two levels of agonistic encounters: backdowns and escalated fights. Backdowns consisted of one individual attacking and the other retreating without retaliation. Escalated fights consisted of one individual attacking and the other retaliating and initiating a bout of mutual aggression. These bouts of aggression varied from a brief struggle to a vigorous bout of wrestling and biting and ranged in duration between 3 and 30 s. Injuries were rare. In the four observed instances of injuries, one female had the distal segment of an antenna bitten off. This did not appear to impair the female's ability to examine and parasitize hosts.
The overall sequence of interactions in contests is summarized in Figure 1. The numbers refer to overall frequencies of behavioral options taken by females, which are explained in detail in the Results section. The sequence begins at step 1, with encounters between females as they search on the patch. Females may continue exploiting the patch without aggression, or one of them may attack the other, leading to step 2. The other may now choose not to retaliate, in which case it temporarily retreats from the patch, then later returns and attempts to continue searching and ovipositing. Meanwhile, the attacker remains on the patch and alternates between ovipositing and patrolling. If, instead, the other retaliates, a fight occurs (step 3), and both females actively contest possession of the patch. As with step 2, the loser of the fight retreats and later returns, while the winner remains and continues ovipositing and patrolling.
|
In our analysis, we included data from all repetitions of this basic
sequence prior to one of the females beginning "stationary"
behavior. This behavior indicates the start of a new and distinct behavioral
phase in the contest (Field et al.,
1998a), and at this point we considered contests to be resolved
into final resident/intruder roles. Thus, the labels "resident"
and "intruder" in Figure
1 are provisional, and refer to the role the female was currently
in at the time of the encounter (i.e., resident if she had won the most recent
encounter and intruder if she had lost).
Analysis of contest resolution
As contest outcome is a binary variable (i.e., win/lose), we used logistic
regression (see Hardy and Field,
1998) to analyze the effect of relative body size and relative
release time on the probability of becoming resident on the patch. For each
contest, one of the females was randomly chosen as the focal female, and her
size, relative release time, and success or failure used for analysis. Due to
missing values in the size data, we analyzed size and release time in separate
univariate logistic regression models. To determine how influential the result
of the first agonistic encounter was in determining the final outcome, models
with size and release time as factors were also run looking only at the result
of the first agonistic encounter. In each model, overall fit was assessed
using the deviance statistic (D) and significance using the change in
deviance (G). After verifying that patch size did not interact with
release time in either model, we pooled data from patch size 12 (treatment
12A) with those from patch size 24 (treatments 24A, 24B, and 24C).
Because T. basalis are extremely small wasps and accurate
morphometric measurements are difficult to obtain, we used weight as a measure
of size. To determine weight, females were stored at -60°C before being
freeze-dried overnight and then weighed to an accuracy of 10-6 g.
Relative size (RS) was calculated as the difference between the weight of the
focal female (Wff) and that of the other female
(Wof) as a proportion of the weight of the focal female:
 |
We also analyzed behavioral decisions between the first agonistic encounter and the establishment of final residency. We tested for differences among treatments in the number of escalated fights per contest using a nonparametric one-way ANOVA.
Finally, we tested whether the degree of adherence to the
"bourgeois" rule for contest resolution (i.e., attack if currently
resident; back down if intruder; Maynard
Smith, 1982), varied across different relative release times. This
amounted to testing the effect of the absolute difference in release time on
the probability that no escalated fights occurred during a contest. If no
escalations occurred, then both females behaved in a bourgeois fashion, as the
contest was resolved by one female repeatedly attacking and the other always
backing down. Again, this was done using logistic regression.
As contests consisted of repeated encounters, we could also calculate the frequency of bourgeois resolution of separate encounters within a contest. For the current resident, the frequency of bourgeois resolution was the sum of the frequency with which it attacked the intruder upon encounter (as opposed to tolerating its presence) and the frequency with which it retaliated if the intruder attacked. For the intruder, bourgeois resolution was the frequency with which it failed to attack the resident, either by tolerating its presence when foraging or by retreating from the patch when attacked. For example, the overall frequencies (all treatments pooled) can be obtained from Figure 1: (912 + 27)/1089 = 0.86 for the resident and (142 + 893)/1089 = 0.95 for the intruder. We verified that frequencies for resident and intruder were homogeneous within treatments, then pooled data and tested for differences between treatments using pairwise chi-square tests with significance level set at.05 and adjusted using the sequential Bonferoni procedure.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Logistic regression analysis of the first encounter and the overall outcome of each contest showed that outcomes were not determined by relative size (first encounter: G = 1.99, df = 1, p =.16; overall outcome: G = 0.80, df = 1, p =.37), but were strongly influenced by the relative release time of the females (first encounter: model significance G = 14.84, df = 1, p <.001; model fit D = 33.49, df = 34, p =.49, see Figure 2a; overall outcome: model significance G = 31.32, df = 1, p <.001; model fit D = 16.69, df = 34, p =.99, see Figure 2b). A further model confirmed that the outcome of the first encounter was a strong predictor of the probability of winning overall (G = 15.67, df = 1, p <.001), with the female that won the first encounter going on to achieve residency status in 78% (38 of 49) contests.
|
The overall frequencies of behavioral choices made by resident and intruder after the first agonistic encounter show in detail how resident-intruder roles emerged during contests (Fig. 1). Upon encounter (step 1), the current intruder rarely attacked the current resident. Most often the resident attacked the intruder and the intruder backed down, or, less frequently, both females tolerated one another's presence and continued exploiting the patch. If the resident attacked the intruder (step 2), the intruder almost invariably backed down, but if the intruder attacked the resident (step 2), the resident usually retaliated, leading to an escalated fight. In this eventuality (step 3), the resident usually retained its position, regardless of whether it initiated the fight or not.
The overall average number of agonistic encounters per contest (backdowns plus fights) was 19.2 ± 3.51 (mean ± 95% CI), and of these 1.1 ± 0.29 were escalated fights. There were fewer escalated fights per contest in treatment 24C than in any other treatment (H = 18.0, df = 3, p <.001, Dunn's multiple comparison, p <.05). All contests in treatments 12A and 24A contained at least one escalated fight, but escalated fighting was absent in two contests in treatment 24B and 12 contests in 24C.
Logistic regression showed that the degree of adherence to the bourgeois strategy varied with relative release time: when the second female was released later, it was less likely to initiate an escalated fight at some time during a contest (model significance G = 25.43, df = 1, p <.001; model fit D = 29.12, df = 34, p =.71; Figure 3. The frequency of bourgeois behavior within contests also varied across treatments. As the patch size increased and the difference in release time became greater, the trend was toward greater frequency of bourgeois behavior by both resident and intruder (Figure 4).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
In this study, we investigated contest resolution in T. basalis by experimentally varying the arrival times of females on a host patch and thus creating asymmetries in resource value and ownership status between them. When a substantial difference in arrival time (>30 min) existed, the first female invariably became the final resident. Furthermore, there was no evidence of an effect of body size on contest outcome, suggesting that the advantage was wholly due to some asymmetry derived from the difference in arrival time. There are two main possibilities: a resource-value asymmetry (the first female to arrive valued the patch more highly because it had oviposited more and therefore had more of its reproductive interests at stake); or a resource-uncorrelated ownership asymmetry (the second female recognized the first as having arrived previously and respected residency). We now consider these possibilities in light of the empirical evidence from our study and in light of the predictions from theoretical models of animal contests.
Our data support the general theoretical prediction that decision rules
will evolve such that highly asymmetric contests are settled without injurious
fighting (e.g., Maynard Smith and Parker,
1976). Although contests always contained at least one escalated
fight when the asymmetry was small, as the asymmetry became larger, bourgeois
settlement of contests became more common. Thus it appears that females were
to some degree able to recognize the asymmetry and adjust their strategy
accordingly. Females contested more vigorously if there was little or no
asymmetry, but intruders soon backed down if a moderate asymmetry was present
and usually refrained from serious fighting altogether when the asymmetry was
extreme (Figure 3).
Females would not have had precise information about these asymmetries, but
must have in some way estimated them from their experiences on the patch. As
the first female usually detected the entry of the second female within a few
minutes, the first female would immediately have had a good estimate of the
degree of asymmetry in both residency status and resource value. However, the
second female could only make an indirect estimate of how long the first
female had been present, by foraging on the patch and gauging the level of
patch depletion. Field and Calbert
(1998) showed that females'
decisions concerning when to begin fighting on a patch are influenced by their
encounter rates with unparasitized hosts and with other females. A female
entering a patch and encountering unparasitized hosts infrequently and another
female often, could reliably estimate that a large asymmetry existed and
thereafter behave accordingly.
Our data do not enable us to unequivocally distinguish between the resource-value asymmetry and an uncorrelated ownership asymmetry as the source of the advantage to the first female to arrive. Given that the resource-value asymmetry existed and could be perceived by females, it is perhaps more parsimonious to favor it as an explanation. However, looking at the repeated encounters between females within contests, the overwhelming majority of these were resolved by the female that had lost the most recent encounter (the current intruder) backing down without an escalated fight (Figure 1). The current intruder rarely attacked the resident, retaliated if it was attacked by the resident, or won an escalated fight if it occurred (Figure 1). The resident, on the other hand, almost invariably attacked the intruder upon encounter and usually won an escalated fight in the rare event that the intruder retaliated (Figure 1). Thus to a large extent, the final outcome of the contest could be predicted by the result of the first agonistic encounter. This is true even in treatments 12A and 24A, where the asymmetry was minimal. From this we see that once a female started winning encounters, she continued to do so, regardless of whether the asymmetries were small or large.
Thus bourgeois settlement was the rule not only for overall contests, but also for separate encounters within contests. For the resident, the frequency of within-contest bourgeois varied between 0.7 to 0.97, with an overall frequency of 0.86, and for the intruder, it ranged from 0.93 to 0.99, with an overall frequency of 0.95 (Figure 4). Although the frequency of bourgeois behavior by the intruder was relatively constant regardless of patch size or release time, that of the resident was markedly lower at the smaller patch size and also when the difference in release times was smaller.
Although the deviations from within-contest bourgeois by the intruder are
sufficiently low to be accounted for in terms of errors in role perception,
those by the resident require further explanation. The frequency of bourgeois
behavior by residents was lowest at the smaller patch size. This may have been
because they had invested less offspring on average than at the larger patch
size and thus had less incentive to defend because their potential losses to
superparasitism were smaller. It has been suggested previously that there may
be an upper threshold patch size, above which defense becomes unfeasible
(Waage, 1982). However, it is
also possible that as patches become smaller, defense tends to become
uneconomical, and the rate of gain of offspring could be maximized by paying
less attention to intruders.
Similarly, to understand the fact that within-contest bourgeois behavior by the resident was more frequent when the difference in release time was greater, it is important to note that resident females exploiting a patch face a trade-off between exploiting the remaining unparasitized hosts and protecting the offspring already invested in the patch. This may be neatly summed up in a variation of an old adage: "Don't count your eggs before they're parasitized"—in other words, there is little point defending a totally unparasitized patch. Locating and ovipositing in a suitable host usually occupies at least several minutes, so the resident must tolerate some degree of intrusion simply to make its own investment in the patch. This consideration was most important when both females were released simultaneously, in which case many hosts remained unparasitized when aggression began. However, when the intruder was released later in the contest, the resident had already parasitized most of the patch, so it could concentrate mostly on defense, and the frequency of withincontest bourgeois consequently increased almost to unity.
Tolerating some intrusion while exploiting a patch might be especially
relevant in the field, where it is possible that more than a single intruder
may arrive on a patch (Field et al.,
1998a), making perfect defense unrealistic. The frequency of
tolerance by residents may reflect some average expectation of intrusion.
There is less reason to expect the intruder to abandon the bourgeois approach
and attack the resident; because the payoff to superparasitism is initially
high and remains so for several hours
(Field et al., 1997), it may do
equally well by waiting for the resident to leave (see
Field et al., 1998a). This
explains why the frequency of deviation by the intruder remains in the range
explicable by errors. Alternatively, occasional attacks by the intruder might
not be errors at all, but instead represent adaptation to field conditions,
under which there would always be some uncertainty as to whether the other
female was the same one encountered before or a different one that might be
more easily overthrown.
To summarize, contests between pairs of female T. basalis competing for a host patch were resolved on the basis of which female arrived first on the patch and had therefore oviposited more. When a large arrival-time asymmetry existed between females, contest resolution followed the bourgeois rule: "attack if resident, backdown if intruder." Even if the asymmetry was small, the majority of repeated encounters within contests were settled in a bourgeois fashion. It appears that, when large asymmetries existed, females accurately perceived them, and the major deviations from bourgeois behavior within contests occurred when residents tolerated intruders due to the economic necessity of continuing to exploit unparasitized hosts in the patch.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
We thank the two anonymous referees for constructive criticisms and Ian Hardy, Phil Taylor, Mike Keller, and Andy Austin for comments on earlier versions. S.A.F. was supported by an Australian Postgraduate Award in Australia and a Golda-Meir Postdoctoral Fellowship in Israel.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Cumber RA, 1964. The egg parasite complex (Scelionidae: Hymenoptera) of shield bugs (Pentatomidae, Acanthosomidae: Heteroptera) in New Zealand. NZ J Sci 7:536-554.
Enquist M, Leimar O, 1987. Evolution of fighting behaviour: the effect of variation in resource value. J Theor Biol 127:187-205.
Enquist M, Leimar O, 1990. The evolution of fatal fighting. Anim Behav 39:1-9.
Field SA, Calbert G, 1998. Patch defence in the egg parasitoid Trissolcus basalis: when to begin fighting?Behaviour 135:629-642.
Field SA, Calbert G, Keller MA, 1998a. Patch defence in the egg parasitoid Trissolcus basalis: the time-structure of pairwise contests, and the `waiting game'. Ethology 104:821-840.
Field, SA, Keller, MA, 1999. Short-term host discrimination by the parasitoid wasp Trissolcus basalis (Hymenoptera: Scelionidae). Aust J Zool (in press).
Field SA, Keller MA, Austin AD, 1998b. Field ecology and behaviour of the egg parasitoid Trissolcus basalis (Wollaston) (Hymenoptera: Scelionidae). Trans R Soc S Aust 122:65-71.
Field, SA, Keller MA, Calbert G, 1997. The pay-off from superparasitism in the egg parasitoid Trissolcus basalis, in relation to patch defence. Ecol Entomol 22:142-149.
Godfray HCJ, 1994. Parasitoids: behavioral and evolutionary ecology. Princeton, New Jersey: Princeton University Press.
Grafen A, 1987. The logic of divisively asymmetric contests: respect for ownership and the desperado effect. Anim Behav 35:462-467.
Griffiths NT, Godfray HCJ, 1988. Local mate competition, sex ratio and clutch size in bethylid wasps. Behav Ecol Sociobiol 22:211-217.
Hammerstein P, 1981. The role of asymmetries in animal contests. Anim Behav 29:193-205.
Hardy ICW, Blackburn TM, 1991. Brood guarding in a bethylid wasp. Ecol Entomol 16:55-62.
Hardy ICW, Field SA, 1998. Logistic analysis of animal contests. Anim Behav 56:787-792.[Web of Science][Medline]
Hughes JP, Harvey IF, Hubbard SF, 1994. Host-searching behavior of Venturia canescens (Grav.) (Hymenoptera: Ichneumonidae): interference—the effect of mature egg load and prior behavior. J Insect Behav 7:433-454.
Lawrence PO, 1981. Interference competition and optimal host selection in the parasitic wasp, Biosteres longicaudatus. Ann Entomol Soc Am 74:540-544.
Leimar O, Enquist M, 1984. Effects of asymmetries in owner-intruder conflicts. J Theor Biol 111:475-491.
Maynard Smith J, 1974. The theory of games and the evolution of animal conflicts. J Theor Biol 47:209-221.[Web of Science][Medline]
Maynard Smith J, 1982. Evolution and the theory of games. Cambridge: Cambridge University Press.
Maynard Smith J, Parker GA, 1976. The logic of asymmetric contests. Anim Behav 24:159-175.
Maynard Smith J, Price GR, 1973. The logic of animal conflict. Nature 246:15-18.
Mills NJ, 1991. Searching strategies and attack rates of parasitoids of the ash bark beetle (Leperisinus varius) and its relevance to biological control. Ecol Entomol 16:461-470.
Petersen G, Hardy ICW, 1996. The importance of being larger: parasitoid intruder-owner contests and their implications for clutch size. Anim Behav 51:1363-1373.
Sirot E, 1996. The pay-off from superparasitism in the solitary parasitoid Venturia canescens. Ecol Entomol 21:305-307.
van Alphen JJM, Visser ME, 1990. Superparasitism as an adaptive strategy for insect parasitoids. Annu Rev Entomol 35:59-79.[Medline]
Visser ME, Luyckx B, Nell HW, Boskamp GJF, 1992. Adaptive superparasitism and patch time allocation in solitary parasitoids: marking of parasitized hosts in relation to the pay-off from superparasitism.Ecol Entomol 17:76-82.
Waage JK, 1982. Sib-mating and sex ratio strategies in scelionid wasps. Ecol Entomol 7:103-112.
Wilson F, 1961. Adult reproductive behaviour in Asolcus basalis (Hymenoptera: Scelionidae). Aust J Zool 9:739-751.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||