The global decline of insects (Hallmann et al. 2017; Sánchez-Bayo and Wyckhuys 2019; Seibold et al. 2019; Wagner et al. 2021) poses a major threat to humankind, as insects play an essential role for our ecosystems through providing functions such as pollination (Klein et al. 2007), natural balance of populations (DeBach and Rosen 1991), and nutrient cycling (Yang and Gratton 2014). Therefore, long-term assessments of insect population trends are necessary to implement and adapt conservation measures to preserve insects in the future. In recent years, several insect monitoring studies were conducted (e.g. Seibold et al. 2019; Karlsson et al. 2020; Lehmann et al. 2021; Staab et al. 2023), which are all facing various constraints. First, they are limited through funding sources (often granted only for a few years) and trained personnel for insect identification. Second, many long-term studies focus only on specific target groups, such as butterflies (Thomas 2005), wild bees (Turley et al. 2022), beetles, true bugs, grasshoppers and spiders (Seibold et al. 2019). Such groups are relatively well studied, easy to identify and historically used as indicators for diversity and landscape changes. In addition, calibrating estimates of abundance for long term studies can be difficult even for target groups as it highly depends on the trapping type used for the assessment (Battles et al. 2024).

An alternative to focusing on specific insect taxa is to use biomass as a proxy for abundance of insects. This has been done in the landmark study by Hallmann et al. (2017), in which biomass of flying insects was used to monitor insect population trends, showing a decline of 76% in insect biomass over 27 years on the study sites, but also in several other studies (Lister and Garcia 2018; Uhler et al. 2021; Welti et al. 2021; Müller et al. 2023). The benefits of using biomass is that the weighing setup is cheap, it can be done in a standardized way, and there is no need for time-consuming counting of specimens or taxonomic identification. However, the use of biomass as a proxy for diversity is also highly controversial (Hallmann et al. 2021b; Redlich et al. 2021). Intensive land use can lead to the mass occurrence of few insect species, compromising the relationship between biomass and diversity. In addition, the assessment of insect diversity using metabarcoding results still has methodological weaknesses that lead to bias in the data patterns (Zizka et al. 2022). Consequently, it depends on many factors to derive statements on species diversity from biomass information.

Malaise traps are used widely to assess biomass of flying insects in a standardized way. The two insect orders that are most abundant in such Malaise trap samples are Diptera and Hymenoptera (Geiger et al. 2016; Srivathsan et al. 2023). Although numerous, especially small-sized parasitoid Hymenoptera (microhymenoptera) represent only a small fraction of the total biomass, where they are mostly considered ‘by-catch’ and remain unprocessed. This is problematic, because Hymenoptera are assumed to be the most diverse animal order in the world (Forbes et al. 2018), and play an important role in many terrestrial ecosystems. Ecologically, parasitoid Hymenoptera are important as top-down regulators of their specific hosts, often herbivorous insects (LaSalle 1993; Heraty 2017) and are essential for the resilience of ecosystems (LaSalle and Gauld 1991; Grissell 1999). In ecological studies, they can serve as indicators for the presence of their hosts (Anderson et al. 2011). Due to their high position in the food web, parasitoids are also particularly prone to extinction (Hassell 2000; Wagner et al. 2021; Klaus et al. 2023), although only few studies focused on their potential decline (but see Gatter et al. 2020; Janzen and Hallwachs 2021).

One reason for the neglect of parasitoid Hymenoptera in biodiversity studies is that they harbour many so-called ‘dark taxa’, which are extremely species-rich groups of small-sized insects that are hard to identify and that harbour mostly undescribed taxa (Shaw and Hochberg 2001; Hausmann et al. 2020a; Hartop et al. 2022). The taxonomic work that would be necessary to describe and identify dark taxa often lacks funding sources as taxonomy is poorly recognized in science and the public (but see Behm 2018). As a consequence, the scarcity of experts contributes to the taxonomic impediment (Engel et al. 2021). Altogether, this leads to a huge knowledge gap not only on the diversity of parasitoid Hymenoptera in terrestrial ecosystems, but also on their potential threat through agricultural activities (Sánchez-Bayo and Wyckhuys 2019), like mowing or pesticide application (Brühl et al. 2021; Schöfer et al. 2023).

To fill this gap, we studied microhymenoptera in Malaise trap samples of an ongoing insect monitoring program in south-western Germany. As Hallmann et al. (2017) highlighted the relevance of insect biomass for monitoring insect abundance trends, we studied the relationship of insect total biomass, and abundance and diversity of microhymenoptera to see, if a decrease in insect total biomass of Malaise trap samples, can be also indicatory of a decrease of microhymenoptera. Moreover, we provide a phenology for different families of microhymenoptera of semi-arid grassland of south-western Germany. These data can be used for the timing of potentially harmful agricultural activities, e.g. mowing (Haas-Renninger et al. 2023b) or pesticide application, to avoid high activity periods of microhymenoptera.

To the very best of our knowledge, this is the first study on the diversity and the phenology of different families of microhymenoptera in meadow ecosystems using Malaise traps. Because it is currently not possible to study the abundance and the diversity of taxonomic groups in parallel at the same time using molecular methods, several thousands of microhymenoptera had to be sorted manually on family level, which took 18 person-months in the present study. Therefore, we restricted our analysis to data from one year and three sites. However, because all three sites are very similar, we consider it likely that these data are representative for insect communities in semi-arid grasslands of south-western Germany.

A current discussion revolves around the relationship between biomass of insects and biodiversity, and the potential of certain taxa to serve as indicators for biomass or species richness (Redlich et al. 2021; Uhler et al. 2021). This is important for the interpretation of recent data on insect decline and its ecological consequences, based on biomass data from malaise trap catches (Hallmann et al. 2017; Uhler et al. 2021). Recently, Hallmann et al. (2021a) found a significant relationship between hoverfly abundance and species richness with the total biomass of insects in malaise trap catches from six locations in North Rhine-Westphalia, Germany. These locations were also studied in the Krefeld study (Hallmann et al. 2017), at two time points, 1989 and 2014. From this, Hallmann et al. (2021a) concluded that the biomass decline reported by Hallmann et al. (2017) is in fact associated with a decline in abundance and diversity of hoverflies, and possibly other taxa as well. A positive correlation with hoverflies was confirmed by Redlich et al. (2021), and Uhler et al. (2021), using insect biomass data from malaise traps at 179 locations in South Germany from different habitats. Interestingly, the correlation of total insect biomass with species richness differed depending on temperature and habitat but was significantly positive in all habitats for Diptera, Lepidoptera and Orthoptera, but not for other groups on all studied habitats. From this, it was concluded that hoverflies might be indicators for large-bodied insect taxa such as Orthoptera and Lepidoptera, while taxa with small body size, but high diversity have less impact on biomass (Redlich et al. 2021). In contrast to this idea, our data on microhymenoptera from low-intensity managed meadows revealed not only a significant correlation between total insect biomass and diversity, but also with abundance. We assume that this is caused by the dependence of microhymenoptera on the abundance of their hosts, which at least partially belong to the Lepidoptera and Diptera. This finding indicates that the decline of insect biomass reported by Hallman et al. (2017) may also affect microhymenoptera.

There was a positive correlation between temperature, precipitation and abundance of microhymenoptera, which has been observed also in previous studies for insects (Pilotto et al. 2020; Uhler et al. 2021; Welti et al. 2021; Müller et al. 2023). This finding is reasonable considering the fact that insects are ectothermic organisms, which means that with increasing temperature, their metabolism increases, which can lead to higher flight activity (Klowden 2013). In his study on the influence of different climatic factors on parasitoid Hymenoptera, Ulrich (2000) identified precipitation as the most important climatic factor. In contrast, Juillet (1964) concluded that temperature and wind velocity are the main indicators of activity of Ichneumonoidea, and that precipitation is not an informative indicator, because of its occasional occurrence. Still, it remains unclear if microhymenoptera abundance and activity are affected by the weather directly or via their host densities.

At all sites, wasps were collected in larger numbers only after end of May in spring, and after September in autumn and had between one and four peaks in activity over the year, mostly between June and July, in July, and between July and August. It is unclear, if these peaks represent several generations of wasps throughout the year or are caused by different species within each family with non-matching peaks. As mowing showed to be harmful for microhymenoptera (Haas-Renninger et al. 2023b), our phenology data indicates two periods in the year which might be suitable for activities such as mowing, up to end of May in spring, and after September in autumn. However, our data do not provide evidence for a decline of microhymenoptera due to mowing. In fact, from the five mowing events and one grazing event observed during our study, only two were associated with a subsequent decline in abundance, while an increase was observed in four cases (Fig. 3). The large losses of up to 64% in individuals caused by mowing (Haas-Renninger et al. 2023b) may not affect the population sizes in the field. It can be suggested that mowing had no impact because microhymenoptera have a short lifespan and their population sizes strongly depend on newly emerged specimens. It might be possible that our study sites were quickly re-populated after mowing by wasps, which emerged from unmown habitats from surrounding areas (Fig. 2).

In our study, the most abundant microhymenopteran families were egg parasitoids of the families Mymaridae, Scelionidae and Trichogrammatidae, and Platygastridae. Boness (1953), who used sweep netting, ground eclectors and vegetation beating found that Chalcidoidea, the superfamily of Mymaridae and Trichogrammatidae, are the most abundant parasitoid wasps in meadows. This is consistent with Haas-Renninger et al. (2023b), which found that Chalcidoidea in general and especially Mymaridae dominated. However, Haas-Renninger et al. (2023b) found Ceraphronidae to be the family with the second largest abundance, which were found in much lower number in this study. This might be due to the different trapping methods, as Haas-Renninger et al. (2023b) used insect suction samplers. The egg-parasitoids in the families Scelionidae, Mymaridae and Trichogrammatidae show two main activities, the first between June and July and the second between August and September. This pattern was also observed by Ulrich (2005) for egg-parasitoid using emergence traps. Other endoparasitoids, such as some Braconidae, Encyrtidae and Ceraphronidae, that use various hosts, show a high activity already in mid-May, and in the case of Platygastridae even earlier in April. In the case of Braconidae, of which around one quarter of specimens are expected in the micro fraction (Haas-Renninger et al. 2023a), the observed phenology might be due to the activity of small-sized Aphidiinae, that are parasitoids of aphids. The high abundance of Encyrtidae between June and July at Apfelberg might be due to the mass-emergence of a few species. This is for example known from the genus Copidosoma, which are polyembyronic parasitoids of Lepidoptera and produce a high number of offspring (Smith et al. 2017). For the families that we found in the samples but did not discuss here, their biology comprises too many different host orders or life stages of hosts (Figitidae, Aphelinidae, Eulophidae, Eupelmidae, Eurytomidae, Pteromalidae), or they occurred in numbers that were too low to see a clear phenology (Cynipidae, Ormyridae, Signiphoridae, Tetracampidae, Mymarommatidae), partly due to the fractioning process (Ichneumonidae, Torymidae, Proctotrupidae, Aculeata) (Haas-Renninger et al. 2023a).

In our study, we used size-fractioned samples to separate small-sized specimens from larger ones and focused solely on the micro fraction, as we were interested in microhymenoptera. The efficiency of the fractionator to separate Hymenoptera families has already been tested (Haas-Renninger et al. 2023a) and has shown to be very useful in terms of specimen handling and expenditure of time considering the high number of sorted Malaise trap samples in this study. Our focus on the micro fraction explains the low numbers of Ichneumonoidea and Aculeata, which are generally very numerous in whole Malaise trap samples (Srivathsan et al. 2023). Another constraint of our study is the focus on the family level. Different genera and species within a family can have different phenologies depending on the host stages they are associated with (Gaasch et al. 1998). Nevertheless, we are confident that our results can show general activity patterns of microhymenopteran families, which might be interesting not only for hymenopterists focusing on a specific family, but also for conservation management. Therefore, this might be the first step towards assessing parasitoid Hymenoptera in a long-term monitoring approach and to consider them as useful indicators for biodiversity (LaSalle and Gauld 1993; Fraser et al. 2008; Anderson et al. 2011; Stevens et al. 2013). Species of higher trophic levels such as parasitoids are at particularly high risk of extinction due to habitat loss and landscape homogenisation (Hassell 2000; Wagner et al. 2021; Klaus et al. 2023). Another critical factor is climate change which was suggested by Janzen and Hallwachs (2021) as the main driver for a decline of parasitoid Hymenoptera species richness in a protected rain forest over 34 years.

We selected our study sites dependent on the conservation value which is based on the number and conservation status of indicator species, in our case wild bees, in a specific habitat (Görn and Fischer 2011). The microhymenoptera abundance at the protected area Apfelberg showed more similar patterns to the non-protected area Weissach than to the protected area Steiner Mittelberg, where abundance was nearly twice as high. This might be due to the location of the Malaise trap at Steiner Mittelberg, where the trap was placed in a low-intensity managed meadow, between an organically cultivated field and a forest edge (Fig. 1B). Therefore, many specimens might have been trapped while migrating from meadow to forest and vice versa, or while migrating through the corridor between field and forest, especially as the Malaise trap was oriented orthogonally to this corridor. Edges of forests are known as insect flyways (Townes 1972) and trap locations directly at such border lines can massively influence the trap result (Hausmann et al. 2020b), which is especially the case for Weissach (Fig. 1C). By contrast, the Malaise traps at Apfelberg (Fig. 1A) was placed more centrally in the meadow. As we could not see strong differences in the phenology of the different microhymenoptera families as well as the family composition between the sites, we believe that our site selection based on conservation values of wild bees together with the hierarchical cluster analysis was reasonable for the purpose of our study.

Our data set forms the baseline for microhymenoptera occurrence in low-intensity managed meadows in south-western Germany. Our pre-sorted microhymenoptera specimens are a valuable resource for taxonomical approaches, such as Large-Scale Integrative Taxonomy (LIT) (Hartop et al. 2022), which aims to enable fast and reliable species delimitation based on preliminary species hypotheses acquired through inexpensive data, such as pictures and barcodes. Parts of the microhymenoptera specimens sampled in our study have already been taxonomically treated in the project GBOL III: Dark Taxa (https://gbol.bolgermany.de/gbol3/de/gbol-dark-taxa/) resulting in new microhymenopteran species descriptions (Moser et al. 2023). With our study, we want to raise awareness for the importance of standardized long-term monitoring projects to observe population trends of insects. Our study also highlights the value of the ‘black gold’ of mass samples, containing an unknown diversity of microhymenoptera waiting to be discovered.