Acceptance and suitability of Acyrthosiphonpisum and Sitobion avenae as hosts of the aphid parasitoid Aphidius ervi ( Hymenoptera : Braconidae )

Aphidius ervi Haliday is a parasitoid of natural populations of both Acyrthosiphon pisum (Harris) on alfalfa and Sitobion avenae (Fabricius) on wheat in Chile. In this study the performance of the aphid parasitoid on both host species was evaluated. Regardless of origin, both females and males of A. ervi were significantly smaller when reared on S. avenae than on A. pisum. Males and females of A. ervi reared on S. avenae on wheat also took significantly longer to develop than when reared on A. pisum. There was no significant difference in the sex ratio of the parasitoid when reared on these two host aphids. Survival of parasitoids from the A. pisum alfalfa system was significantly lower when reared on S. avenae relative to those reared on A. pisum, but no significant difference in survival on both hosts was detected in parasitoids from the S. avenae wheat system. Behaviour of parasitoid females during oviposition, such as frequency of encounters, number of attacks and stabbings of host aphids, were significantly higher in the A. pisum A. ervi interaction than in the S. avenae A. ervi interaction, regardless of the origin of the parasitoid. Aphid defensive reactions such as kicking or production of cornicle secretion, were significantly higher in the A. pisum A. ervi than the S. avenae A. ervi interaction. The results are discussed in relation to the effect of host quality and host aphid body size on parasitoid develop­ ment.

Aphidius ervi is widely distributed in Eurasia, parasi tizing mainly Macrosiphinae aphids such as Acyrthosi phon pisum (Harris) on legumes and, to a lesser degree, Macrosiphum euphorbiae (Thomas) and Aulacorthum solani (Kaltenbach) on other host-plants (Takada & Tada, 2000).In Japan, its distribution is restricted to A. pisum and Acyrthosiphon kondoi Shinji (Takada & Tada, 2000).Aphidius ervi has become one o f the most widely used agents for the control o f A. pisum , and also o f A. kondoi on lucerne in several countries around the world (Powell & Wright, 1988).Although Sitobion avenae (Fabricius) on cereals is a suitable host o f A. ervi, this parasitoid is of only minor importance as an aphid biocontrol agent o f aphids on cereals in Europe (Cameron et al., 1984).This situation is expected to improve if cereals are grown near perennial legume fields (Starý, 1978).In Chile, the intro duction o f this parasitoid from cereal hosts in France occurred nearly 25 years ago (Zúftiga, 1990), and it now parasitizes here both A. pisum on alfalfa and S. avenae on wheat (Starýetal., 1993;Starý, 1993).
In Chile, where A. ervi parasitizes both aphid species, this parasitoid shows behavioural differences in its response to aphid host-plant volatiles, which seem to result from conditioning and adult experience, and not from genetic differentiation o f A. ervi subpopulations on alfalfa and wheat (Daza-Bustamante et al., 2002;Rodriguez et al., 2002).Thus, A. ervi responds more strongly to the volatile olfactory cues emanating from the aphid-host-plant complex in which they were reared, regardless o f the origin o f the parasitoid (wheat or alfalfa) (Daza-Bustamante et al., 2002).We hypothesise that there will be differences in host acceptance and suitability o f the aphid hosts from alfalfa and wheat, in particular that A. ervi from A. pisum on alfalfa and S. avenae on wheat reared on A. pisum and S. avenae, will differ in their acceptance and subsequent performance on these host aphids.The present work evaluates parasitoid host acceptance, aphid defensive behaviour, and parasitoid performance on A. pisum on alfalfa and S. avenae on wheat.

Stock cultures
Aphidius ervi was obtained from parasitised aphids (mummies) collected on alfalfa, Medicago sativa L., or wheat, Triticum aestivum L., at INIA-La Platina fields in Santiago, Chile.The parasitoids that emerged were reared on the same aphid host -plant combination as they were collected from.The colonies were maintained at 20 ± 1°C and 14L : 10D photope riod, and were supplemented twice (at two-monthly intervals) with specimens brought from the same aphid host -plant combi nation in the field.They were allowed to intercross ad libitum with individuals in the existing colony.Individuals used in this study were raised in the laboratory for at least 6 generations.

Host acceptance by parasitoids and aphid defensive behav iour
Batches of 50-80 parasitoids were collected as mummies from stock cultures and placed singly in small glass vials.Newly-emerged parasitoids were sexed, fed diluted honey and allowed to mate overnight.Pilot experiments indicated that this time period was sufficient to insure mating.Nymphs of both host aphid species from the stock cultures were synchronised by allowing adults to reproduce and using those nymphs produced within a 6-h period.Synchronised second-instar nymphs were placed singly in Petri dishes (3.5 cm diameter, 1 cm height) con taining a leaf of the aphid's food plant.Aphids were allowed to settle for 10 min, after which a single naïve mated female parasitoid was introduced into each dish.All experiments were per formed between 11:00 and 17:00 h.Aphids and parasitoids were used only once.The behaviour of parasitoids and aphids was observed under a stereoscopic microscope and data recorded using the software "The Observer".The observation period lasted for 5 minutes.The following parasitoid and aphid behav ioural events were recorded: (i) Encounter -a female parasitoid approaches an aphid and taps the aphid's body with its antennae, (ii) Attack -a female parasitoid bends its abdomen forward, and either reaches the aphid body or doesn't, (iii) Stab -a female parasitoid reaches an aphid's body with its ovipositor, and (iv) Kick -an aphid kicks backwards with its hind legs after being contacted by a foraging parasitoid.Experi ments were repeated until a set of 20 replicates of each treat ment was completed in which parasitoids or aphids showed at least one of the behavioral events described above.Shortly after the experiments the aphids and parasitoids were transferred to an oven, dried for 2 days at 50°C, and individually weighted on an electronic microbalance (sensitivity 0.0001 mg).

Parasitoid performance
Synchronised second-instar nymphs of either A. pisum or S. avenae were placed singly in Petri dishes (3.5 cm diameter, 1 cm height).After a 10-min settling period, a mated naïve female parasitoid, which had been treated in the same way as in the above experiment, was introduced into the Petri dish.To avoid superparasitism, wasps were allowed to stab a host only once with its ovipositor.Potentially parasitised aphids were clipcaged on alfalfa or wheat plants in a controlled environment chamber at 22 ± 1°C, 70 ± 10% r.h. and a 14L : 10D photope riod.Parasitised aphids were observed daily (between 10:00 and 10:30 h) until adult parasitoid emergence, and total develop mental time, pupal survival, secondary sex ratio and adult body size (measured as adult dry mass) were recorded.A mummy was considered to be formed when an aphid developed a brownish and opaque colour.After emergence, parasitoids were killed and their dry mass determined as described above.All aphids were parasitised since they all reached the mummy stage.Twenty replicates were performed of each treatment for the evaluation of survival and secondary sex ratio.For the evalua tion of development time and adult body size, experiments were repeated until twelve females and twelve males developed.All parasitoids and aphids were used only once in the experiments.

Statistical analyses
Since the assumption of homogeneity of variances was not valid for the data obtained when recording aphid acceptance by the parasitoid, it was analysed using the non-parametric Kruskal-Wallis test (Siegel & Castellan, 1988).Data on the developmental time of the parasitoid on different host aphids once V(x+0.5)transformed did not invalidate the assumption of homogeneity of variances; hence, a three-way ANOVA was used with parasitoid origin, aphid host and parasitoid sex as the main effects.No transformation of the data on the dry weight of parasitoid adults resulted in homogeneity of variances.Thus, the non-parametric Scheirer-Ray-Hare extension of the Kruskal-Wallis test for factorial ANOVA was used in this case (Sokal & Rohlf, 1998).Finally, the G-test was used to compare survival and secondary sex ratio of the parasitoids (Sokal & Rohlf, 1998).

Aphid -parasitoid interaction
Female parasitoids from the S. avenae -wheat system were smaller than females from the A. pisum -alfalfa system (Table 1).Frequency o f encounters, attacks and stabs by the parasitoid were higher on A. pisum than on S. avenae, when the origin o f the parasitoid was the A. pisum -alfalfa system, but not when it was the S. avenaewheat system (Table 1).Body size o f A. ervi from S. avenae on wheat was smaller than from A. pisum on alfalfa (Table 1).Incidence o f aphid defense against para sitoid attack, i.e. kicking, was significantly higher in A. pisum than in S. avenae, regardless o f parasitoid origin (Table 1).Body size o f A. pisum was larger than that o f S. avenae (Table 1).

Parasitoid performance
Regardless o f sex, total parasitoid development time was significantly higher on S. avenae than on A. pisum when parasitoid origin was the S. avenae -wheat system, but not when it was the A. pisum -alfalfa system (Table 2,  Regardless o f parasitoid origin, body size o f parasitoids was significantly smaller when the host aphid was S. aValues given are means with standard errors in parenthesis.Values in each column followed by the same letter are not significantly different according to ANOVA with V(x + 0.5) transformed data and a = 0.05.Body size analysis was performed with the Scheirer-Ray-Hare extension ofthe Kruskal-Wallis test.

Table 1 .
Behavioural events in host recognition and acceptance of Acyrthosiphon pisum and Sitobion avenae by the parasitoid Aphidius ervia Mean values or number of events are shown with standard errors in parenthesis.Values in each column followed by the same letter are not significantly different according to Kruskal-Wallis test with a = 0.05.

032; remaining inter actions non-significant). There were marginally signifi cant differences in parasitoid body size in relation to sex, mainly because o f the smaller males obtained from
S. avenae, when their origin was the S. avenae -

wheat sys tem. Furthermore, regardless o f parasitoid origin there were no significant differences in sex ratio between para- sitoids reared on
S. avenae and on A. pisum (0.

90 and 1.15, respectively, G-test, p > 0.05). Survival to the adult stage was significantly lower in parasitoids from the
A. pisum -alfalfa system reared on S. avenae than those reared on A. pisum (

75 and 100%, respectively, G-test, p < 0.05). No significant differences in survival were detected in parasitoids from the
DISCUSSIONA.ervi from alfalfa encountered, attacked and stabbed less frequently S. avenae than A. pisum .

On the other hand, A. ervi from wheat did not show significant differ ences in its response to the two aphid hosts. Kicking was more frequently performed by
A. pisum than by S. avenae, regardless o f the origin o f the parasitoid (i.e. from A. pisum on alfalfa or S. avenae on

wheat), in spite o f the smaller body size o f parasitoids reared on
S. avenae -wheat than on A. pisum -

alfalfa. Differences in encoun ter, attack and stabbing may be related to aphid body size: A. pisum is larger than S. avenae and thus more likely to be encountered, attacked and stabbed by A. ervi. As expected from the increased rate o f encounters, andTable 2 .
Performance of Aphidius ervi parasitoids (females and males) on the host aphids Acyrthosiphon pisum and Sitobion ave naea