Sergey Ya. Reznik, Natalia D. Voinovich & TaisiaYa. Umarova. Effects of experience on host selection in Trichogramma principium (Hymenoptera, Trichogrammatidae).

Zoological Institute, Russian Academy of Sciences, Universitetskaya nab., 1, St. Petersburg, 199034, Russia

The oviposition behavior of insect parasitoids is usually considered as a sequence of stimulus-response reactions during which host acception is determined by a combination of positive and negative stimuli received by the parasitoid from the host. It is well known that learning may influence the process of host perception (Vinson, 1985; Alphen & Vet, 1986). Moreover, responsiveness of insects to specific stimuli is modulated by endogenous factors: lesser positive stimuli may cause a response if motivation to oviposit is high, while reduced motivation could raise the threshold of acceptance (Barton Browne, 1993).

The present paper concerns parasitization behavior in Trichogramma, an egg parasitoid that is not only widely used for biocontrol (Smith, 1996), but also represents a convenient object for laboratory investigations on insect physiology and behavior. It was repeatedly reported that some host species were readily accepted for oviposition by certain Trichogramma females, while the other females of the same strain delayed parasitization or even permanently refused to parasitize the same host (see Reznik et al., 1998 for the references).

For some years, our specific approach was to study the reaction of Trichogramma females to a poor quality host (near the lower threshold of acceptance). This method reveals minute behavioral variations, because all reactions are usually more variable in the threshold neighborhood. We have shown that Trichogramma females that refused to parasitize the Angoumois grain moth, Sitotroga cerealella Oliv. eggs carried a high number of mature ovarian eggs, they moved actively, they even might incidentally contact the host, but the usual sequence of behavioral reactions resulting in parasitization was interrupted at the stage of arrestment and host recognition, suggesting that refusal to oviposit may be considered as egg retention. As for Trichogramma females that accepted S. cerealella eggs, they appeared to maintain a relatively stable 'parasitization state' when sequentially offered two equal portions of the host eggs (Reznik et al., 1998). The aim of the present study was to estimate stability of the parasitization state in the case when not equal but different hosts (namely, different host species and/or stages of egg development) were sequentially provided.

Materials and methods

We used a laboratory strain of Trichogramma principium Sug. et Sor., collected in Kazakhstan from Noctuidae eggs and cultivated for more than 50 generations on the eggs of the grain moth. Two host species: S. cerealella and more preferred laboratory host, Mediterranian flour moth, Ephestia kuehniella Zeller were used. In certain variants, S. cerealella eggs that had developed 5 days at a temperature of 20 °C were used [it was demonstrated earlier that 5-day old S. cerealella eggs are still suitable for Trichogramma development, but less attractive for ovipositing females (see Reznik & Umarova, 1990 for the references)].

All experiments were conducted at 20 °C under a light regime of L18:D6. Trichogramma females were placed individually into small test tubes. Honey was streaked on the glass to feed the wasps. After completion of Trichogramma larvae development, darkened (parasitized) host eggs were recorded. In all experiments, females sequentially received two portions of host eggs and oviposition during the first and second period of the experiment was recorded for each female separately (see Reznik et al., 1998 for more detailed methods). The main characteristics used to describe the result of the experiment was the percentage of females accepting a given host species and age combination (i.e., the percentage of females laying at least one egg).

Our previous works suggested that individual Trichogramma females tend to behave in a similar manner during the first and second periods of the experiment (i.e. to parasitize or not to parasitize). To estimate the degree of this stability, we used two special indices:

C is the number of 'continuing' females, which oviposited during both the first and second periods,

F is the number of 'finishing' females, which oviposited only during the first period,

S is the number of 'starting' females, which oviposited only during the second period,

The mean percentage of ovipositing females, indices of stability and induction of parasitization state were compared among treatments by the Kruskal-Wallis distribution free test.

Results

In the first experiment, females were sequentially offered two portions of the grain moth eggs, either 1-day or 5-day old. In total, 2200 T. principium females were studied in 11 replicates of this experiment.

As expected, T. principium that was offered 1-day old eggs oviposited much more often compared to females that were offered 5-day old eggs (Table 1). Moreover, the wasps oviposited more often in 5-day old eggs when previously offered with 1-day old eggs compared to females which were sequentially offered two portions of 5-day old eggs. The stability of parasitization state (SP) was only slightly (not significantly) higher in females provided with 1-day old eggs, while the induction of parasitization (IP) depended significantly on the age of the second portion of grain moth eggs (Table 1).

Table 1. The effect on parasitization when T. principium females were sequentially offered two different ages of S. cerealella eggs (the first experiment) *)

Host species / age		Ovipositing females, %		Parasitization state: stability and induction
1^st period	2^nd period	1^st period	2^nd period	SP	IP
Sitotroga, 1-day old	Sitotroga, 1-day old	37.8±5.8 ^a	56.3±5.8 ^a	89.2±2.0 ^a	34.1±7.1 ^a
Sitotroga, 1-day old	Sitotroga, 5-day old	45.2±5.7 ^a	39.7±5.0 ^a	79.4±5.8 ^a	6.8±1.5 ^b
Sitotroga, 5-day old	Sitotroga, 1-day old	16.1±2.5 ^b	47.4±5.6 ^a	82.8±5.7 ^a	40.0±5.8 ^a
Sitotroga, 5-day old	Sitotroga, 5-day old	13.1±3.8 ^b	18.9±3.1 ^b	78.0±9.1 ^a	10.5±1.5 ^b

*) Means±SEM are shown. Means followed by the same superscript letter in the same column are not significantly different at 0.05 level (Kruskal-Wallis distribution free test).

In the second experiment, females were sequentially offered eggs of different host species. Each treatment of this experiment was conducted in two variants. In the first variant, each host was used for two days immediately followed by an alternative host for two consecutive days (same as in the first and third experiments). In the second variant, two 48 h periods of contact with host eggs were separated by a 48 h host deprivation interval inserted between the first and second periods. During host deprivation, females were offered honey solution, but no hosts were provided. In total, data on 713 females were obtained in 8 replicates of this experiment.

The effect of host deprivation interval was analyzed by the Kruskal-Wallis test in two treatments separately. At Ephestia->Sitotroga sequence, 48 h long host deprivation interval inserted between two host exposures had no effect on the induction of parasitization and on the percentage of females ovipositing during the second period while stability of parasitization was slightly but significantly (p=0.011) lower in the variant with host deprivation. At Sitotroga->Ephestia sequence, all these characteristics were unaffected by the host deprivation interval.

Results of the second experiment based on the pooled data of two variants are presented in Table 2. At Ephestia->Sitotroga sequence, ca 2/3 of T. principium females started parasitization during 48 h of contact with E. kuehniella eggs and practically all of these wasps continued oviposition during the second period, when S. cerealella eggs were provided. As for females which did not oviposit during the first period, only 1/4 of them started to parasitize S. cerealella. At Sitotroga->Ephestia sequence, less than a half of females started ovipositing during the first period (again, almost all continued to oviposit during the second period), but the induction of parasitization was much more often than that at Ephestia->Sitotroga sequence. As a result, the percentage of females ovipositing during the second period at Sitotroga->Ephestia sequence was nearly equal to that at Ephestia->Sitotroga sequence.

Table 2. The effect on parasitization when T. principium females were sequentially offered 1-day old eggs of two different host species (the second experiment) *)

In the third experiment, S. cerealella eggs that had developed 5 days were used with the aim of increasing the difference in the rate of acceptance between Sitotroga and Ephestia. In total, data on 520 females were obtained in 6 replicates of this experiment.

As expected, the decrease in 'quality' of 5-day old S. cerealella eggs reduced percentage of females ovipositing during the first exposure by one half (compare Tables 2 and 3). However, percentage of females ovipositing during the second exposure at Ephestia->Sitotroga sequence was almost the same. Obviously, the reason is that the parasitization stability index SP approached 100% at both host sequences, i.e., wasps that started parasitization almost always continued to parasitize independently of the new host species and age. As a result, females twice more often accepted 5-day old eggs of S. cerealella when previously offered E. kuehniella eggs compared to those which were sequentially offered two portions of S. cerealella eggs (Table 3).

Table 3. The effect on parasitization of S. cerealella eggs when T. principium females were preliminary offered eggs of two different host species (the third experiment) *)

Discussion

Our previous experiments showed that Trichogramma females tend to maintain parasitization when sequentially offered two equal sets of fresh eggs of the grain moth (Reznik & Umarova, 1998). In the present study, parasitization behavior was stable even when preferred and usually rejected hosts eggs were offered sequentially. Moreover, it was stable even with the 48 h host deprivation interval inserted between two host exposures. The mechanism of this stability is not clear yet. It is known that learning may influence host acceptance. However, (with the exception of imprinting) learning usually causes gradual, not qualitative changes. Moreover, oviposition on a given species usually does not influence the level of affinity for other hosts (Kaiser et al., 1989). Besides, the learned behavior is typically not persistent (see reviews: Vinson, 1985; Papaj & Lewis, 1992). It is known that the oviposition behavior is also subject to endocrine control (Raabe, 1986; Barton Browne, 1993). Possibly, the stability of the parasitization state is based on endocrine mechanisms: presence of oviposition-stimulating neurohormone which was released during the first period of the experiment may influence the female behavior during the second period. In this case, the egg retention might be considered not as a simple behavioral reaction, but as a sort of special physiological state when not oogenesis, but oviposition is subject to hormonal regulation. In such a situation, the influence of oviposition experience on further behavior of Trichogramma females may be divided into gradual reversible behavioral changes caused by learning and stable qualitative switches caused by endocrine mechanisms. However, this is nothing more than a hypotheses, which is difficult to verify by usual endocrinological methods because of a microscopic size of the wasps.

Be it as it may, our experiments have demonstrated that initial acceptance of a preferred host may lower the threshold for subsequent acceptance of usually rejected host. Thus, the probability of host acceptance significantly depends not only on the current host, but also on the previously offered hosts. Obviously, under natural conditions, parasitoids may sequentially contact patches of host eggs of different species and/or ages. Therefore, the results of the present study should be taken into account in further investigations on ecology and behavior of insect parasitiods.

References

Alphen, J.J.M. van & L.E.M. Vet. 1986. An evolutionary approach to host finding and selection. In: Insect parasitoids. (J.K. Waage and D.J. Greathead. Eds.) pp. 21-63. London, Academic Press.

Barton Browne, L. 1993. Physiologically induced changes in resource-oriented behavior. Annual Rev. Ent. 38: 1-25.

Kaiser, L., Pham-Delegue, M.N. & C. Masson. 1989. Behavioural study of plasticity in host preferences of Trichogramma maidis (Hymenoptera: Trichogrammatidae). Physiol. Ent. 14: 53-60.

Papaj, D.R. & A.C. Lewis. 1992. Insect learning: ecological and evolutionary perspectives. London, Chapman and Hall. 398 pp.

Raabe, M. 1986. Insect reproduction: regulation of successive steps. Adv. Insect Physiol. 19: 29-154.

Reznik, S.Ya. & T.Ya. Umarova. 1990. The influence of host's age on the selectivity of parasitism and fecundity of Trichogramma. Entomophaga 35: 31-37.

Reznik, S.Ya., Umarova, T.Ya. & N.D. Voinovich. 1998. Egg retention in the presence of a host in Trichogramma females. J. appl. Ent. 122: 555-559.

Smith, S.M. 1996. Biological control with Trichogramma: advances, successes, and potential of their use. Annual Rev. Ent. 41: 375-406.

Vinson, S.B. 1985. The behavior of parasitoids. In: Comprehensive insect physiology, biochemistry and pharmacology. (G.A. Kerkut and L.I. Gilbert. Eds.). Vol. 9. pp. 417-469. Oxford, Pergamon Press.