In the experiments using inexperienced specimens, in the 1-h observation tests G. pennsylvanicum females showed a high host acceptance rate (75% of marked eggs) and a high host suitability rate (75%) on the target species L. occidentalis (table 2). Compared to the target species, the number of drummed eggs was significantly lower for G. insidiator and R. erythropus (χ2 = 10.67, d.f. = 1, P = 0.0011; χ2 = 14.73, d.f. = 1, P = 0.0001, respectively), while no eggs of N. viridula were drummed (χ2 = 18.91, d.f. = 1, P < 0.0001). The number of drilled eggs of L. occidentalis was significantly higher than those of C. marginatus, G. acuteangulatus, G. insidiator and C. lateralis (χ2 = 12.52, d.f. = 1, P = 0.0005; χ2 = 19.65, d.f. = 1, P = 0.0001; χ2 = 22.73, d.f. = 1, P = 0.0001; χ2 = 14.44, d.f. = 1, P = 0.0001, respectively), while eggs of G. juniperi (χ2 = 26.19, d.f. = 1, P < 0.0001), R. erythropus (χ2 = 26.19, d.f. = 1, P < 0.0001) and S. cingulatus (χ2 = 16.31, d.f. = 1, P = 0.0001) were not drilled. Despite the fact that some eggs of C. marginatus, G. acuteangulatus, G. insidiator and C. lateralis (for all these species: χ2 = 20.91, d.f. = 1, P < 0.0001) were drilled by the females of G. pennsylvanicum, they did not accept the eggs as hosts, as evidenced by the lack of marking behaviour, and no egg-parasitoid offspring were produced from those eggs. Moreover, drilling did not affect hatching of the heteropteran eggs as first instars hatched successfully from 100% of these tested eggs.
In the 4 and 48-h tests, G. pennsylvanicum host suitability of the target species L. occidentalis was high, respectively 80% and 100%. No non-target species tested were suitable hosts, producing no egg-parasitoid offspring, with the exception of a single egg of G. juniperi in the 48-h test (5% of the total eggs tested), from which one female G. pennsylvanicum successfully emerged. The difference between the target and non-target species was statistically significant (χ2 = 32.48, d.f. = 1, P = 0.0001). In two other cases, and only in the 48-h tests, one egg of C. marginatus and one of C. lateralis were accepted by G. pennsylvanicum, but they were not suitable as host because no adult G. pennsylvanicum emerged successfully. In the case of the parasitized egg of C. marginatus, the female egg-parasitoid died during emergence: the exit hole chewed by the female did not coincide with the pseudoperculum, but was oriented at the opposite side of the egg, and the specimen could not create a sufficiently large hole to escape. In the case of parasitization of the egg of C. lateralis, after egg dissection, a dead female G. pennsylvanicum was detected in the egg, and no chew marks were observed on the chorion.