Species composition and seasonal dynamics of aphid parasitoids and hyperparasitoids in wheat fields in northern China

Parasitoids are important natural enemies of aphids in wheat fields of northern China, and interest in them has increased in recent years. However, little is known regarding parasitoids of wheat aphids, which has hindered the study and understanding of aphid-parasitoid interactions. In the present study, three primary parasitoids and 15 hyperparasitoids were collected in wheat fields during a 2-year survey in northern China (2014, 2015) and a 2-year investigation at Langfang, Hebei Province (2015, 2016). Among them, Aphidius uzbekistanicus Luzhetski was found most frequently among the primary parasitoids, while Pachyneuron aphidis (Bouché) dominated the hyperparasitoid community. Investigation of the dynamics of wheat aphids and parasitoids revealed that the primary parasitoids appeared early in the growing period and that the hyperparasitoids appeared later. Analysis of the seasonal dynamics revealed that growth of the parasitoid population followed that of the aphid population and that the parasitism rates were highest in the late growing period.

To develop such a molecular detection technique for Chinese cereal aphid parasitoids, we first compiled information on wheat aphid parasitoids to familiarize ourselves with these species. Most of the literature on wheat aphid parasitoids in China is focused on primary parasitoids, particularly species of Aphidiinae 1,2,26-32 . However, aphid 33 and primary parasitoid 33,34 populations can be influenced by hyperparasitoids, of which reports are mostly lacking. Most records are from the 1990s, with five hyperparasitoid species in this system being recorded in northern China and four in central China. These include Aphidencyrtus (Syrphophagus) aphidivorus (Mayr), Dendrocerus carpenteri (Curtis), Pachyneuron aphidis (Bouché), and other species that could only be classified to genus 1,35,36 . Recently, Zhao et al. (2011) and Hu et al. (2016) recorded nine hyperparasitoid species in northwestern China and provided detailed data on the abundance for each species; however, some also could not be identified to the species level 37,38 .
To understand the biological control effect these aphid parasitoids have, it is important to understand their relative efficiency, which would allow conservation efforts to target the most important species. How species from different trophic levels interact with each other and how they affect the food web as a whole, such as the effect of hyperparasitoids on the biological control efficiency, are still unclear. Moreover, comprehensive information about the aphid-parasitoid species is missing and the dominant species of aphid hosts, primary parasitoids, and hyperparasitoids at different growing periods is still unknown.
Here, we investigated (1) the parasitoid species composition in 18 wheat fields in 2014 and 30 fields in 2015 in northern China (Figs 1 and 2) the population dynamics of wheat aphids and their parasitoids in Langfang, Hebei Province from 2015-2016. This work will support future research on the species characterization and community composition of wheat aphid parasitoids in China.

Results
Parasitoid species composition in commercial wheat fields in northern China. In 2014, 460 primary parasitoid individuals from two species and 1,047 hyperparasitoid individuals belonging to 12 species emerged from the mummies collected from 18 wheat fields. In 2015, 1,276 primary parasitoid individuals belonging to three species and 2,048 hyperparasitoid individuals belonging to 14 species emerged from the mummies collected within 30 wheat fields.
Two primary parasitoid species, Aphidius uzbekistanicus Luzhetski and Aphidius gifuensis (Ashmead), were found in 2014, and a third primary parasitoid species, Aphidius ervi Haliday, was found in 2015. From 2014 to 2015, the relative catch of A. pusilla decreased slightly from 0.67% to 0.59%, that of S. eliavae decreased from 3.15% to 1.03% and that of D. laticeps decreased from 29.13% to 6.98%. In contrast, the relative catches of the remaining hyperparasitoid species increased from 2014 to 2015. All had relative catches under 10% except for P. aphidis, D. laticeps and S. taeniatus (Fig. 2).

Discussion
Our study, spanning a 2-year period, represents the full investigation of the diversity and community composition of wheat aphid parasitoids in northern China. Among the 13 published studies that record wheat aphid parasitoids in detail, only three were conducted in northern China 1,2,27 . The primary parasitoids recorded by these studies consisted of six species: Aphidius avenae Haliday, A. gifuensis, A. ervi, Ephedrus persicae Froggatt, E. plagiator (Nees), and Praon sp. 25 ; one additional species, Lipolexis gracilis (Förster), was present the following next year by the same authors as before 1 . A study conducted in 2007 found primary parasitoid species of the genera Aphidius, Diaeretiella, Ephedrus, and Lysiphlebus 2 . Compared with these previous works, the present study found fewer primary parasitoid species, all in the genus Aphidius: Aphidius uzbekistanicus, A. gifuensis and A. ervi. Only one previous work mentioned A. uzbekistanicus 29 ; and it appears not to an uncommon primary aphid parasitoid in wheat fields in China. In contrast, it is a consistent and important member of cereal aphid parasitoid communities in the Palaearctic 39,40 . However, in the present study, A. uzbekistanicus replaced A. avenae as the most abundant of primary parasitoid species in China 37,38 . Interestingly, no A. avenae were found in this study, a finding that brings into question the results of previous studies. Either A. avenae was misidentified in previous work or A. uzbekistanicus has replaced A. avenae in northern Chinese cereal crops for competitive reasons. One new species, A. ervi was found in 2015. The primary parasitoid collection in 2015 consisted of more than twice as many samples that in 2014. This expanded collection has likely increased the odds of discovering new species.
Hyperparasitoids have received less attention than primary parasitoids in past studies. Only five species have been reported for northern China, belonging to four families: Figitidae, Encyrtidae, Megaspilidae, and Pteromalidae. Of these, P. aphidis was the only species identified to species level; the remaining taxa were identified only to the genus level 1 . The relatively comprehensive species descriptions were generated based on collections in northwestern China which was different from our background 37,38,41 . We recollected most of these hyperparasitoid species in our study and identified nine species, belonging to four families: one species each from Encyrtidae and Megaspilidae, three species from Figitidae, and four species from Pteromalidae. Aphidencyrtus (Syrphophagus) aphidivorus. Among them, D. carpenteri (Megaspilidae), A. suspensus, A. vulgaris and P. aphidis (Pteromalidae) also appeared in our survey. In previous studies, Alloxysta spp. could not be classified to species 37,38 . In contrast, we identified five Alloxysta species: A. consobrina, A. fracticornis, A. japonicus, A. pusilla and A. victrix. In addition, three Syrphophagus species, i.e., S. eliavae, S. taeniatus, and Syrphophagus sp., along with D. laticeps and P. villosa were identified in the study, for a total of 15 hyperparasitoid species. Wheat aphid population densities were low during this study early on and peaked midway through the growing period. Parasitism rates increased throughout almost the entire growing season and followed the trends in growth and decline of the aphid populations, a typical seasonal dynamic of aphid-parasitoids 10 . In general, parasitoids seemed to have suppressed aphid populations during the late growing period. The primary parasitoids appeared during the early growing season, and hyperparasitoids appeared in the late growing season (Figs 3  and 4). In the early growing period, the parasitism rates were quite low and grew slowly such that the primary parasitoids appeared unable to prevent the increase in aphid numbers until late May, when the hyperparasitoids emerged 13 . This pattern was also apparent from the aphids' and parasites' seasonal dynamics during the middle growing period, when the parasitism rates were too low to stop the aphid population's growth. Hyperparasitoids may disrupt the biological control mediated by primary parasitoids and influence the latter's aphid control efficiency 33,42,43 . In some cases, hyperparasitoids can promote aphid suppression by stabilizing insect-parasitoid dynamics 42,44 , an interesting relationship which needs further study. Considering the delay in the control effect of parasitoids, efforts to protect parasitoids should be made in the early spring (using habitat strips and non-crop planting patterns) to ensure successful colonization. The use of pesticides should be reduced or avoided during the early growing season to protect parasitoid populations.
The aphid-parasitoid system model is a simple, common pest-natural enemy system model 8,10,11,45,46 . However, in our study system, the relationships between specific species of aphids and parasitoids remain unclear. In some studies, parasitoids have been reared with prior knowledge of their target aphid species 9,47 , but in many cases this species-specific relationship is unknown. To this end, barcoding techniques or PCR detection can provide more robust species identification, including the identification of cryptic species [47][48][49][50] . However, regardless of the methodological approach, a comprehensive understanding of the parasitoids species and composition will facilitate research on aphid-parasitoid relationships 23 .

Materials and Methods
Wheat aphid parasitoid species composition. Study area. The study was carried out in Beijing City and Tianjin City, Hebei Province, northern China. This agricultural landscape consists of various habitats, including crops, fallow land, grassland and forest. Wheat is the dominant crop during winter and spring. A total of 18 fields and 30 fields were investigated in 2014 and 2015, respectively (Fig. 1). All fields were randomly selected, and each field was > 1 ha. The distance between fields ranged between 1.5 km and 95.7 km. Sampling was conducted between May 18 and June 5 in 2014, and from May 16 to May 31 in 2015. There were three sampling dates per field each year, with a 7-10-day interval between sampling dates.
Parasitoid sampling method. A minimum of 150 aphid mummies were collected at five randomly selected sampling points per wheat field at each sampling date. Sampling sites were separated by at least 5 m, and were at least 10 m away from the field edge. All collected mummies were stored individually in 1.5 mL reaction tubes and brought to the laboratory for parasitoid rearing at 25 ± 1 °C, a photoperiod of 16:8 h (L:D), and 65-75% relative humidity. The lids of the reaction tubes were opened and the opening plugged with absorbent cotton to allow ventilation and prevent the escape of emerged parasitoids. Samples were inspected daily for parasitoids emergence, and emerged parasitoids were stored at 4 °C in 1.5-mL centrifuge tubes filled with 75% ethanol for identification. The number of emerged parasitoids from each sampling field was recorded (Table S1).
Parasitoid identification. Most of the reared aphid parasitoids were identified based on morphological characteristics. In addition, part of the mitochondrial cytochrome c oxidase subunit one (COI) and the 16 S gene DNA were sequenced from a subsample of the emerged parasitoid specimens to confirm the species identification. DNA was extracted from the parasitoids using a non-destructive DNA extraction (see below). A detailed description of the identification methods and the number of sequences per parasitoid species is provided in Tables S2-3. Emerged parasitoids were morphologically identified under a fluorescent stereomicroscope (SZX16, Olympus, Japan) using specific identifying characteristics. Photographs of the key identification characters were taken using a polarizing microscope (DM2500, Leica, Germany) and a digital camera (EOS 505D, Canon, Japan). Examples of such photographs are provided for different parasitoid species in the supplementary material ( Figure S1).
The DNA of three specimens of each parasitoid species was extracted using a non-destructive extraction method. Air-dried parasitoids were submerged in 180 μL Buffer ATL and 20 μL Proteinase K (QIAGEN DNeasy kit, QIAGEN, Germany) at 56 °C for 2 hours and then removed before performing the subsequent steps. The primers LepF1 (5′-3′ ATT CAA CCA ATC ATA AAG ATA TTG G) and LepR1 (5′-3′ TAA ACT TCT GGA TGT CCA AAA AAT CA) 51 were used to amplify an approximately 700-bp region of the COI gene for each parasitoid. The primers LR-N13398 (bee version) (5′-3′ CAC CTG TTT ATC AAA AAC AT) 52 and LR-J12888 (supplementary honeybee) (5′-3′ TCG ATT TGA ACT CAR ATC ATG TA) 53 were used to amplify an approximately 450-bp region of the 16S gene. Amplification of an approximately 1000-bp sequence of the 18S gene was performed using the universal primers 18SL0001 (5′-3′ TAC CTG GTT GAT CCT GCC AGT) and 18SR1100 (5′-3′CGA CGA TCC AAG AAT TTC AC) 54 . PCR followed the universal primer amplification protocols described in Yang et al. 55 . PCR products were visualized using a UV transilluminator (Universal Hoodα, Bio-Rad, USA) after electrophoresis at 180 V in a 1% agarose gel for approximately 12 min. Those PCR products which showed a single band of the expected fragment size were sent for sequencing in both forward and reverse directions (Sangon Biotech, Beijing, China). The sequences were then edited and checked using BLAST (Basic Local Alignment Search Tool, NCBI National Center for Biotechnology Information, https://blast.ncbi.nlm.nih.gov/Blast.cgi).
Seasonal dynamics of wheat aphids and parasitoids. Study area. In 2015 and 2016, the seasonal dynamics of aphids and parasitoids was recorded in three winter wheat (var. Zhongmai 175, Institute of Crop Sciences, CAAS, Beijing, China) plots located at the Langfang Experimental Station (GPS coordinates 116.6°E, 39.5°N), Institute of Plant Protection (IPP) of the Chinese Academy of Agricultural Sciences (CAAS) at Langfang. The three plots were 27 × 27 m and separated by more than 5 m from each other. No insecticides or herbicides were applied, and standard agronomic practices for northern China were used during the trial, including fertilization (375 kg/ha urea, 225 kg/ha phosphorus diamine, 150 kg/ha potassium sulfate), regular tillage, and weeding.
Population sampling. Mummy parasitoid samples were collected from the three plots, including five 1 × 1 m 2 sampling points per plot. The densities of aphids and mummies were recorded every five days between April 30 and June 15 in 2015 and between April 24 and June 9 in 2016. On each sampling date, the numbers of aphids and aphid mummies were counted on 20 randomly selected wheat plants during the collection of mummy samples to calculate aphid and mummy densities 8,9 . The storage of the mummy samples, the rearing of parasitoids and their identification were as described above.
Data analysis. Parasitism rates were calculated by dividing the number of aphid mummies by the sum of mummies and living aphids combined. A Chi-squared test (proc freq) was used to assess the differences among parasitoid taxa, including the two primary parasitoids and various hyperparasitoids on wheat collected from different growing periods in 2015 and 2016 (Table S4). Analyses were performed by using SAS 9.3 software (SAS Institute Inc., Cary, USA).