High species richness of tachinid parasitoids (Diptera: Calyptratae) sampled with a Malaise trap in Baihua Mountain Reserve, Beijing, China

Tachinidae are one of the most speciose families of Diptera and the largest group of non-hymenopteran parasitoids. Little is known about their diversity, distribution patterns, and seasonal variation in most ecosystems. This study reports on tachinid flies collected by a Malaise trap over 73 weeks in Baihua Mountain Reserve, northern China, and investigates the patterns of local species richness and its temporal distribution. The most species-rich season was summer, but the majority of specimens were recovered in spring. A total of 755 tachinid specimens were collected, consisting of 144 species in 85 genera, comprising 26.5% of the species and 49.7% of the genera recorded from northern China. A total species richness of 243 was estimated, indicating that only a portion of the community of tachinid flies was collected at this location and suggesting that the diversity of tachinids might be underestimated across Beijing and northern China. This work is a first step in assessing patterns of tachinid diversity in China using quantitative sampling and establishes a baseline for comprehending the temporal and spatial diversity of these ecologically significant parasitoids.

www.nature.com/scientificreports/ Tachinidae is an amazingly diverse Diptera family that contains exclusively parasitoid species and is distributed worldwide (ca. 8,500 described spp.) 26,27 . Tachinids are the largest group of non-hymenopteran parasitoids, and as parasitoids of herbivorous insects 28 , and thus are of great ecological importance in both natural and managed ecosystems. They primarily seek hosts among immature stages of herbivorous Lepidoptera, Hemiptera and Coleoptera [28][29][30] , and their diversity and population dynamics are suggested to reflect that of their hosts 30 . Understanding the temporal and spatial variation of tachinid diversity could indicate the health of local communities and may provide a basis for conservation work 31 . In addition, monitoring the temporal abundance of tachinids provides basic biological information on their voltinism, timing of development, and association with potential hosts, which is essential in assessing their effectiveness as potential means of biological control for pests 32 .
So far, only a few studies have quantified the species richness of local tachinids over different seasons in the Palaearctic Region [31][32][33][34] . In China, among the previous biodiversity surveys focused on tachinid species [35][36][37][38][39][40][41] , only a few documented or assessed the species richness and composition of tachinid communities in a certain geographic location. In this study, we collected tachinids using Malaise traps with continuous sampling over three seasons in the Baihua Mountain Reserve in Beijing, Northeastern China. Baihua Mountain Reserve forms an important ecological barrier for Beijing, and is known for its high biodiversity, environmental heterogeneity, and varied climate. This study aims to: (i) produce a species list as a baseline database of the Tachinidae in Baihua Mountain Reserve; (ii) estimate the total species richness of the tachinid community in the study area; (iii) examine how tachinid abundance and species richness vary over seasons; and (iv) assess voltinism and phenological patterns for several species of the tachinid community.

Materials and methods
Study area. The sampling was conducted in Baihua Mountain Reserve, Northeastern China. The study site is located at the forest-farmland edge where grass and shrubland successions occurred after the artificial coniferous forests were cut down (N39°50′11.04″, E115°34′41.52″, 1224 m; Fig. 1).    45 All the tachinid specimens were identified to the species level with integration of the recent literature 27,43,44,46-53 . Analysis. We used EstimateS 9.1.0 software package for Windows 54 to estimate the total species richness based on the classic Chao 1 Richness Estimator (Chao-1), Incidence Coverage-based Estimator (ICE) and Abundance Coverage-based Estimator (ACE). The rarefaction curve and its 95% confidence intervals (95% CL) were calculated and plotted using 1,000 permutations. Each sample corresponded to the pooled catches from two weeks, thereby representing specimens collected the first and second half of each month respectively. To explore the patterns of species composition over time, we used a Non-metric Multidimensional Scaling (NMDS) ordination plot with the 'vegan' 55 package implemented in R, following construction of Bray-Curtis dissimilarity metrics. Differences in tachinid communities were further quantified by a Multi Response Permutation Procedure (MRPP) analysis 56 using the 'vegan' package in R. MRPP was performed with 999 permutations on Bray-Curtis distance. The seasonal preference of different tachinid species was estimated by the indicator species analysis 57 using PC-ORD (version 5.0) for Windows 58 . Only tachinid species with an observed indicator value (IV) > 50 and statistical significance P < 0.05 (in a Monte Carlo test (MCT) with 499 permutations) were considered to have a significant preference for a certain season.
Species richness. The species accumulation curve based on all samples does not near its asymptote (Fig. 4), indicating that only a portion of the species were collected. The total species richness at the sampling site, estimated using the sample-based Chao-1 estimator, is 243 species (95% CI 196-333; Fig. 5; Table 1). The results were similar using the ACE estimator (240 species) and the ICE estimator (257 species). The total species richness of the tachinid community in the area, estimated using the sample-based Chao-1 estimator, for the period from June 2017 to October 2017 was 161 species (95% CI 111-278; Fig. 5, Table 1), and from 6 July 2018 to 6 July 2019 was estimated to be 152 species (95% CI 120-220; Fig. 5, Table 1).
Temporal distribution. Data analyses were based on all specimens captured between 2017 and 2019 (Supplementary Information File 1). Summer is the most species-rich season, with 257 species captured in the summer of 2017 (Table 2). Spring, conversely, is the season with the most specimens, with 257 specimens collected in the spring of 2019 (Table 2). Two Exoristinae species comprised almost 50% of all specimens collected in the spring of 2019; other dominant species were Phorocera normalis (31.1%) and Phryno vetula (16%).
The NMDS ordination for tachinid communities was plotted in Fig. 6. A clear division of the tachinid fauna were observed during the three seasons (Fig. 6). The stress obtained for this dimensionality was 0.1325, and the MRPP of the seasonal patterns observed in the NMDS plot was statistically significant (P = 0.001, A = 0.08723),      Supplementary Information File 1). In addition, we found that many species only appear in Table 2. The number of tachinid species and specimens over summer 2017 to spring 2019. The data for the summer of 2018 and 2019 is insufficient, so there is no analysis carried out.

Discussion
This study demonstrates the species richness and temporal distribution of Tachinidae sampled using a Malaise trap in Baihua Mountain Reserve over a sampling period of 73 weeks. A total of 755 tachinid individuals were collected and sorted into 144 species in 85 genera, including several species and genera newly recorded in China and one new species (Supplementary Information File 1 and Supplementary Information File 2). Our results indicate the effectiveness of Malaise traps in comprehensive insect community monitoring, and demonstrated that the Chinese tachinid fauna needs further study as more species are likely to be discovered. The high species richness found in our study suggests that the species richness of tachinids in Beijing and northern China might be vastly underestimated. Initially, the number of species captured in this study is surprisingly high in comparison to the current local species records. The captured species number accounts for 55.8% and 81.7% of the published tachinid species and genera number in Beijing, 26.5% and 49.7% in northern China, and 11.5% and 30.4% of the tachinid species and genera number in China 59 . Such high percentages resulted from our sampling in a generally well-collected region, suggesting a largely undiscovered diversity of the tachinid species in this area. Another line of evidence is that the Chao-1 estimator indicated that the total tachinid species richness (Fig. 5, Table 1) in this single site is close to the current number of tachinid species recorded from across Beijing 59 . In fact, up to 77.8% of the species in our collection are represented by only one or two specimens (Fig. 3). The heavily skewed species abundance distribution curves (Figs. 3 and 4) suggest that many species in the community have not yet been sampled, and that the number of tachinid species will grow as the collecting time increases (Fig. 4).
The high number of species of Tachinidae in our study site might be correlated with the wide variety of climates and habitats located in Baihua Mountain Reserve 11,60 . This reserve contains various microclimates and microhabitats, including coniferous forest, open woodland, broad-leafed forest, grassland, and streams, with elevations ranging from 500 to 2043 m 61 . Our sampling site was located at the forest and farmland edge, with an abundance of wild open grassland that is important in maintaining a high level of insect biodiversity 62 . Although this edge habitat could encourage greater species richness and exotic species as it attracts insects from both open and forest-associated communities [63][64][65][66] , this location only represents two of the common habitats of this reserve. Therefore, the estimated 243 species from one single trapping site are very likely to be a subset of species occurring in the Baihua Mountain Reserve. The hyper-diversity of tachinid communities reinforces the needs for expanding monitoring efforts to uncover the diversity of not only tachinids, but also many other taxa in the region.
This study demonstrated the high sampling efficiency of Malaise traps. We recorded a relatively higher number of species in comparison to previous studies conducted in locations of similar latitudes over similar sampling periods 32,34,67,68 (Table 3). For example, Stireman 34 documented 79 species, and estimated a total of 122 species. www.nature.com/scientificreports/ Inclán and Stireman 32 recorded a community of 117 species and estimated the total species to be 190. Despite the differences in the habitats (natural versus disrupted) between the previous studies 32, 34 and the current study, one explanation for this discrepancy may be the different sampling methods (pan traps versus Malaise traps). The pan trap is designed to catch insects which rely on visual orientation and active near the ground, while Malaise traps passively intercept insects flying at various heights [69][70][71] . Thus, the sampling efficiency of Malaise taps are higher than pan traps for Tachinidae biodiversity surveys. The diversity and population dynamics of tachinids can reflect the hosts and their hostplants due to their trophic level as parasitoids of herbivorous insects 30 . Based on the most comprehensive study to date 52 , we found host information for only 50 of the 144 species recorded in this study ( Supplementary Information File 4), comprising 810 host species belonging to six orders, 59 families and 450 genera ( Supplementary Information  File 4). Surprisingly, of the 50 species with known hosts, only 28 species had their host species also recorded in Baihua Mountain Reserve, suggesting the lack of host species studies of this remarkably diverse group of insects in Beijing and in north China. Understanding the relationship between these parasitoids and their hosts can be used as a baseline for the management of reserves, and as indicators of the diversity and population dynamics of their hosts 30 .

Data availability
All data generated or analysed during this study are included in this published article (and its Supplementary  Information files).