Host specificity from the collected spiders
Of 9,314 spider specimens collected in the field, the potential spider hosts, i.e. the webbuilding spiders, were represented by five families amounting to 85.7% of all collected spiders. These were Theridiidae (92%, N = 7,353), Araneidae (6.2%), Dictynidae (0.5%), Tetragnathidae (0.6%) and Linyphiidae (0.6%). The parasitoid, however, attacked only Theridiidae (100%, N = 105).
Using all data from both the samples from the parasitism rate study (N = 105) and the rearing of wasps (N = 138), it was shown that parasitized theridiid spiders belonged mainly to Theridion varians (43%, N = 104), Neottiura bimaculata (24.4%), Phylloneta impressa (6.6%), T. pinastri (4.5%) and unidentified juveniles of the “Theridion” group (21.5%). The frequency of attacked theridiid hosts did not correspond to their availability both in 2007 (Goodness of fit, χ24 = 1674, p < 0.0001) and in 2008 (Goodness of fit, χ24 = 1137, p < 0.0001), ().
Comparison of relative frequency of available and parasitized “Theridion” spider hosts in the field in 2007 (A) and 2008 (B). High quality figures are available online.
The vast majority of the parasitized spiders (89.5%) were at the juvenile stage (from N = 105). Adults, which comprised 10.5% of all parasitized hosts, were parasitized merely in August 2007, and these were exclusively females of T. varians. The mean size of the prosoma of the parasitized host was 0.71 mm (N = 105, SE = 0.013) and changed somewhat during the season ().
Comparison of the size (prosoma length) of available juvenile “Theridion” hosts during season. Points are means; uncertainty bars represent 95% confidence intervals. High quality figures are available online.
The natural sex ratio of subadult and adult “Theridion” spiders was biased in favor of females, as only 25% (from N = 204) of spiders were males. The sex ratio of parasitized “Theridion” spiders was significantly skewed in favor of females, as only 5% (from N = 75) of spiders were males (Proportion test, χ24 = 12.2, p = 0.0005).
Ovopositioning results in the laboratory
Under laboratory conditions, Z. percontatoria attacked exclusively spiders of the family Theridiidae and ignored linyphiid, araneid or dictynid spiders (GLM, F3,32 = 19.1, p = 0.004). When six taxa of the family Theridiidae (N = 30) were offered, parasitism differed significantly among them (GEE, χ25 = 87.2, p < 0.0001). N. bimaculata, T. varians and “Theridion” were parasitized with similarly high frequency, P. impressa was parasitized with a low frequency, and D. melanogaster and Enoplognatha sp. were not parasitized at all ().
Relative frequencies of six taxa parasitized by wasps in laboratory experiments. High quality figures are available online.
In 2007 the rate of parasitism by Z. percontatoria wasps on theridiid spiders was 1.74% (from N = 4,814). In 2008 the rate halved to 0.83% (from N = 2,539). Comparison of the host spectrum between the two seasons, 2007 and 2008, revealed a shift in host preference. While in 2007 the highest portion of parasitized hosts was represented by T. varians, in 2008 it was represented by N. bimaculata (). In 2007 (April – August) T. varians was the most abundant species with a mean abundance of 40.1 specimens (N = 7, SD = 23.5). This species was also the most parasitized (49.5%) spider host (). In 2008, the abundance of T. varians decreased to 12.9 (N = 8, SD = 13.8), and its parasitized portion decreased to 14.6%. Neottiura bimaculata had a lower abundance in 2007, specifically 4.3 (N = 7, SD = 4.4) spiders and a parasitized portion 16.3%. In 2008, it was the most abundant species with 20.9 specimens (N = 8, SD = 6.5) and a higher portion of parasitism (63.4%) (). The abundance of P. impressa was similar in both years, i.e. 7 specimens per day (N = 7, SD = 9.8) in 2007 and 5.4 specimens (N = 8, SD = 4.84) in 2008. Its portion of parasitism increased from 5.4 to 12.2% between years.
The body size of parasitized hosts changed significantly according to the seasons (ANOVA, F2,102 = 10.4, P < 0.0001): it was smallest in autumn and largest in summer (). The frequency of occurrence of the three main hosts, T. varians, N. bimaculata and P. impressa, at a size preferred by the parasitoid, also changed with the seasons. In spring, all three host species were juvenile and thus occurred at high frequencies at the preferred body size (). In spring (of both years) the highest parasitized portion was in T. varians (43.5%, N = 10), followed by N. bimaculata (26%), the “Theridion” group (15.5%), P. impressa (13%) and T. pinastri (2%). In summer, spiders with a suitable body size were rare, as the majority of them had reached adulthood. The prosoma size of adult T. varians, however, falls within the preferred host range and adult females were thus accepted as host (). This was not the case in P. impressa, whose adult prosoma length was far larger than 1.1 mm. Neither the first instar spiderlings (prosoma size < 0.5 mm) of T. varians or N. bimaculata that hatched from egg sacs were accepted as hosts (field data). In summer, the highest parasitized portion was observed on T. varians (57%, N = 26), followed by indeterminable spiders at early instars (prosoma > 0.6 mm) of the “Theridion” group (26%), N. bimaculata (8.7%), and P. impressa (6%) and T. pinastri (2.3%). In autumn, spiderlings (prosoma size > 0.6 mm) of all three species were accepted by the wasp (). The parasitized portion was highest on N. bimaculata (37%, N = 13), followed by T. varians (28.6%), indeterminable spiders at early instars of “Theridion” group (25.7%), P. impressa (5.7%) and T. pinastri (3%).
Figure 4. Relative frequencies of prosoma sizes of three major available host species (bars) and parasitized individuals (line) in spring (A), in summer (B) and in autumn (C). Bars were generated from spider relative abundance in the particular prosoma size (for (more ...)