INTRODUCTION
Beauveria pseudobassiana Rehner & Humber (Hypocreales: Cordycipitaceae) is a natural antagonist of oak lace bug (Corythucha arcuata Say 1832, Heteroptera: Tingidae), a sap-feeding insect native to North America that is causing foliar damage of oak trees (Quercus spp.), first observed in Croatia in 2013 (Hrašovec et al. 2013). Previous studies have shown that this entomopathogenic fungus is consistently present in C. arcuata populations and that it is the most common entomopathogen on this species (Kovač et al. 2020, 2021; Kovač 2021). While studying C. arcuata and its natural infections with entomopathogenic fungi on different locations in Croatia, we subsequently observed dark brown ascomata emerging from the insects infected with B. pseudobassiana. We inferred hyperparasitism, a phenomenon in which one fungus (the mycoparasite) parasitizes another fungus (the host that is also a parasite). A hyperparasitic interaction consists of at least three trophic levels: a primary host, which is parasitized by the primary parasite, which serves as secondary host to the secondary parasite or hyperparasite (Bermúdez-Cova 2023). The aim of this study was to report new tritrophic association with the hyperparasitic fungus on C. arcuata infested with B. pseudobassiana in natural populations in Croatia.
MATERIALS AND METHODS
During the previous research of C. arcuata, carried out in 2019, 2021 and 2022 in different localities in Croa-tia, many overwintering adults infected with entomo-pathogen B. pseudobassiana were sampled (Kovač 2021, Kovač et al. 2021). Methodology used for sampling and processing of the specimens is described in Kovač et al. (2021). Most of the collected samples were kept in a refrigerator at 4°C and used for the present research. Information on the locations and time of collection can be found in Table 1. Altogether 271 individuals infected with B. pseudobassiana were gathered and inspected for the presence of hyperparasitic fungus S. parasitica. Adults were first checked under a dissecting microscope (Olympus, model SZX7, Tokyo, Japan) at magnification 40X to visually discern the fungal structures, and the fruiting bodies (perithecia) with emerged spores were cut with a sterile needle and microscopically analysed with a phase-contrast microscope (Olympus, model BX53, Tokyo, Japan). The fungus was identified based on the morphology of the fruiting bodies, as well as the size and shape of the ascospores, same as described in Georgieva et al. (2020). The perithecium had globose bases, yellow-brown in colour, with necks of 80–95 µm × 40-42 µm in size. Ascospores were cylindrical, one-celled, 7.5-8 µm long, and 2.5–3 µm wide. The fungus was also isolated on the Potato Dextrose Agar (PDA) medium by taking the whole perithecia with a sterile needle and gently tapping the apices of perithecium necks with spores on a medium. The cultures were incubated for 2–3 weeks at ± 25°C, and purified when needed. After the development of mycelium they were microscopically examined.
RESULTS
After examination of 271 C. arcuata individuals infected with entomopathogenic B. pseudobassiana 34 of them exhibited mycoparasitic growth on B. pseudobassiana (Figure 1a–f). The mycoparasite was identified as Syspastospora parasitica (Tulasne) Cannon & Hawksworth (Ascomycota: Hypocreales). The incidence of hyper-parasitism varied depending on the locality and collection period (Table 1).

Figure 1 (a) Corythucha arcuata adult infested with B. pseudobassiana and hyperparasitised by S. parasitica; (b) mature S. parasitica perithecia growing on B. pseudobassiana; (c) ascospores moving to the tip of the perithecium; (d) S. parasitica perithecium up close; (e) mature ascospores; (f) dissected perithecium with released ascospores.
Table 1C. arcuata infected by Beauveria pseudobassiana and hyperparasitised by Syspastospora parasitica, and information on sampling periods and sites.
*MU = Management unit, FO = Forest office, dep = department
DISCUSSION AND CONCLUSIONS
In this research a new tritrophic association is reported for the first time: a mycoparasitic fungus Syspastospora parasitica (syn. Melanospora parasitica (Tul.)) attacking B. pseudobassiana that had infected adult C. arcuata. This hyperparasitic fungus has been reported on various entomopathogenic fungi including B. pseudobassiana (Georgieva et al. 2020), but never in tritrophic association with C. arcuata. According to previous studies, B. pseudobassiana was established as a dominating species affecting C. arcuata in natural populations of Croatia (Kovač 2021, Kovač et al. 2021), so this new association is an interesting addition to that finding.
The fungus was first described as Sphaeronema parasitica Tul. on Isaria crassa by Tulasne (1857). So far, Syspas-tospora parasitica was recorded as a secondary parasite of the entomopathogenic fungi B. pseudobassiana and B. varroae on pine processionary moth Thaumetopoea pity-ocampa (Denis and Schiffermüller 1775) (Lepidoptera: Notodontidae) larvae and pupae (Georgieva et al. 2020), as well as on other Beauveria species such as B. tenella on Melolontha spp. (Coleoptera: Scarabaeidae) (Müller-Kögler 1961), and B. bassiana on ash weevil Stereonychus fraxini (Coleoptera: Curculionidae) (Markova 1991), Colorado potato beetle Leptinotarsa decemlineata (Say) (Posada et al. 2004) Cydia pomonella (L.) (Lepidoptera: Tortricidae), and on pupae of pine sawfly Neodiprion sertifer (Geoffroy) (Hymenoptera: Diprionidae) (Rosnev and Tsankov 1978). It has also been reported on other entomopathogenic fungi infecting insects such as Paeci-lomyces tenuipes (Peck) Samson attacking Bombyx mori L. (Lee and Nam 2000), Spicaria farinosa (= Paecilomyces farinosus) (Fr.) Vuill. infecting Boarmia bistortata Goeze (Lepidoptera: Geometridae), and Spicaria fumosorosea (= P. fumosoroseus) (Wize) Vassiljevsky infecting T. pity-ocampa (Müller-Kögler 1961).
Although fungal hyperparasitism is widespread in nature, it is still poorly studied, and not ecologically and evolutionarily well understood (Parratt and Laine 2016, Haelewaters et al. 2018, Bermúdez-Cova 2023). However, since entomopathogenic fungi are becoming an increasingly important factor in reducing the population of many insect pests, or at least in keeping their abun-dance at an acceptable level, in the context of biological control the interest in hyperparasitic fungi is growing.
