Metabarcoding Analysis of Pollen Species Foraged by Osmia excavata Alfken (Hymenoptera: Megachilidae) in China

To meet the pollination need of economic crops, Osmia excavata has been successfully used to improve the pollination efficiency of Rosaceae and Brassicaceae plants. As a widely used pollinator of economic crops, a systematic study of flower-visiting species and diversities of O. excavata stocked in China was not found. To investigate the foraging pollen species and diversities of O. excavata, beebread from 20 experimental plots in China was collected by the trap-nesting method and analyzed by DNA metabarcoding technology. A total of 26 pollen plants in 14 genera and nine families were identified. A further analysis showed that the richness and abundance of the wild flowering plants in orchards and farmlands were lower than those in the nearby semi-natural habitats. The favorite pollen comes from economic crops apple and rape and wild flowering plants Juncus interior, Rosa gymnocarpa, and Rosa laevigata. Through a diversity index analysis, it was found that the Anhui region has the highest pollen plant diversity, while the Liaoning region has the lowest. Our results can provide a basis for flower-visiting species and diversities of O. excavata.


INTRODUCTION
Pollinators are of great importance in ecosystems and contribute about $136 billion annually to agricultural systems in China (Ouyang et al., 2019). Due to the intensification and expansion of agriculture, the diversity of pollinators has been declining rapidly for decades, which has led to the loss, fragmentation, and degradation of pollinator habitats and a severe decline in global pollination (Foley et al., 2011;Féon et al., 2013;Belsky and Joshi, 2019). In the past 50 years, the dependence of global agriculture on pollinators has increased by more than 300%, while the abundance of pollinators worldwide has continued to decline (Potts et al., 2016). Previous studies have shown that planting of pollen plants in pollinator habitats can significantly increase the number and species of pollinators (Jönsson et al., 2015;M'Gonigle et al., 2016;Warzecha et al., 2018;Williams and Lonsdorf, 2018;Nichols et al., 2019). Therefore, it is necessary to clarify how to choose the best plants for different pollinators.
The internal transcribed spacer II (ITS2) barcode based on metabarcoding method can identify pollen plant species from mixed pollen and shows great potential in the analysis of beebread ingredients (Pornon et al., 2016;De-Vere et al., 2017;Lucas et al., 2018). Compared with traditional palynology, the molecular method not only has a higher sensitivity for the identification of mixed pollen but also effectively reduces time and cost (Lang et al., 2018;Tremblay et al., 2019). The ITS2 region has proved to be a core barcode for pollen identification (Richardson et al., 2015;Gous et al., 2019;Tremblay et al., 2019). The barcode combination of rbcL, trnH-psbA, and ITS2 successfully identified the main pollen plants, Vitex negundo and Vitex quinata, in the beebread of Megachile strupigera by He et al. (2020). Therefore, DNA metabarcode not only helps to identify species and determine the relative abundance of pollen plants in beebread but is also used to construct pollination networks, monitor plant pathogens, assess species diversity, and predict biodiversity patterns .
To meet the pollination demand of economic crops, pollinators with commercial potential have been found and developed (Ryder et al., 2020). Bumblebees have been successfully used as pollinators for eggplant, tomato, peach, and pear (Wu et al., 2006). Megachile rotundata was used for the pollination of alfalfa (Xu et al., 2009). In orchards, the bee of genus Osmia was considered as efficient pollinators of some Rosaceae plants, such as apricot, plum, cherry, peach, pear, and apple (Parker, 1981;Torchio et al., 1987;Wei et al., 1991;Xu et al., 1994). Importantly, Osmia bee can be artificially stocked due to its advantages such as easy management, low cost, and high pollination efficiency (Wei and Zhao, 1995;Wei et al., 2000b;Dai, 2004 (Wei et al., 2002). For O. excavata, it has many advantages, such as low-temperature tolerance, long daily activity time, high flower-visiting efficiency, centralized pollination range, and easy management (Wei and Yuan, 1997). Therefore, it was considered as an excellent and large-scale stocking pollinator. The pollination of O. excavata can create more than $27.8, $9.4, and $3.0 billion for cherries, apples, and rapes planting, respectively, each year (Liu et al., , 2019a. Pollination can significantly improve the fruit set rate and fruit quality, and it has broad application prospects in the future of fruit planting industry (He and Zhou, 2009).
However, the short flowering period of fruit trees limits the nutritional requirements of O. excavata during the foraging period (Yan et al., 2018). Because of this limitation, survival and large-scale stocking of O. excavata face great challenges. A previous study showed that the diversity of pollen plants is the main factor affecting the diversity of wild bee species (Kovács-Hostyánszki et al., 2017). Therefore, the effective solution is to grow other pollen plants in the orchard. This will not only improve the biodiversity of plantations but also provide adequate food sources for bees (Boyle et al., 2020). So, it is necessary to understand the foraging preferences of bees. In this study, beebread of O. excavata was collected by trap-nesting from 20 experimental plots in China. The species and abundance of plants were determined by ITS2 barcode based on metabarcoding method, and the dominant plants that meet the nutrition needs of bees were further analyzed.

Sample Collection and DNA Extraction
Osmia excavata can build a nest in trap nests made by reed tubes and make beebread to breed its offspring (Wang et al., 2001;Huang et al., 2013). According to the main activity period (March-April) of O. excavata, trap nests were placed in similar habitats (e.g., flower composition, competitors, and predators, etc.) in each experimental plot. Trap nests of O. excavata were collected from 20 regions of China (March-May 2020), covering three types of land: orchards (monoculture or mixedspecies plantation), farmlands, and semi-natural habitats (SNHs; agroforestry ecotone). Detailed geographic information of each sampling plot and surrounding main plant types are listed in Table 1. In the laboratory, mixed pollens within five groups of beebread from the trap nests of each site were stored at -80 • C for later use. Mixed pollens were extracted using the E.Z.N.A. R Soil DNA Kit according to the manufacturer's protocol (Omega Bio-Tek, Norcross, GA, United States). The concentration and purity of purified DNA were determined by a NanoDrop spectrophotometer (Thermo Fisher Scientific, Inc., United States).

Data Analysis
Raw sequencing data were filtered to delete the tail-base of reads (Q < 20) using Trimmomatic v0.38 (Bolger et al., 2014). The chimeric sequences were removed and effective sequences clustered into operational taxonomic units (OTUs; ≥97% similarity) by Usearch v10 (Edgar, 2013). Based on the Unite database v8.2, 1 the OTU representative sequences were analyzed taxonomically to obtain the community composition of each sample using the RDP classifier Bayes algorithm with default parameters (Wang et al., 2007). Alpha diversity indexes (Shannon, Simpson, and Chao1) were calculated using Mothur v1.30 (Schloss et al., 2009). The Fisher's least significant difference (LSD) in SPSS v25.0 (SPSS, Inc., Chicago, IL, United States) was used to analyze the differences of diversities of beebread between different land types and experimental plots. Species accumulation curves (SACs), the abundance, and composition of beebread were prepared and visualized by the "ggplot2" package (Hadley, 2016) and the "pheatmap" package (Raivo, 2019).

Sequencing and Assembly
A total of 4,563,930 clean ITS2 sequences from all experimental plots were obtained with an average length of 335 ± 9 bp and clustered into 64 OTUs (Supplementary Table 1). A statistical analysis showed that the SACs tend to be flat, indicating that the sample sizes reach the conditions that species are fully discovered. The species richness in the environment should not increase significantly with the increase of sample sizes (Supplementary Figure 1).

Impacts of Different Types of Land on the Composition of Beebread
The composition and abundance of pollen in beebread from different types of land use were compared ( Table 1). The results showed that economic crop pollens mainly came from the single-or mixed-fruit plantations and intensive farmland, and wild plant pollens mainly came from the agroforestry ecotone (Figure 2A). Pollen plants in beebread from semi-natural habitat (SNH; agroforestry ecotone) samples have the highest species abundance (4.7 ± 0.95, n = 10), then followed by that from mixed plantations (3.33 ± 0.76, n = 45), farmlands (3.05 ± 0.94, n = 20), and single-fruit plantations (2.08 ± 0.56, n = 25; Figure 2B). A further analysis found that the species abundance of pollen plants in beebread from farmlands and mixed-fruit plantations did not differ significantly (p = 0.36), but there were significant differences between other experimental plots (p < 0.05; Supplementary Table 3).

DISCUSSION
Wild bees are important pollinators for economic crops and wild plants. They rely on different landscapes to obtain flowering resources and nesting habitats (Kratschmer et al., 2019). In this study, 20 experimental plots in China were set up to collect the beebread of O. excavata. A total of 26 pollen plant species were identified, belonging to 14 genera of nine families and including 17 economic crops and nine wild plant species (Robinia pseudoacacia, Deutzia discolor, J. interior, P. rhoeas, P. sinensis, Rosa canina, R. gymnocarpa, R. laevigata, and R. primula). Based on the composition and relative abundance of pollens (Figures 1, 2), O. excavata is an effective pollinator for economic crops (B. juncea, Brassica sp., M. pumila, Malus sp., P. brachypoda, and P. spinosa).
Previous studies indicated that the larvae of O. excavata released at the early flowering stages of economic crops will grow into effective pollinators, especially Rosaceae and Brassicaceae plants (Yang et al., 1997;Wei et al., 2000a;Hawkins et al., 2015;Gresty et al., 2018). However, the flowering period (7-15 days; Supplementary Table 5) (Lech et al., 2008;Liu et al., 2008Liu et al., , 2010Wang et al., 2012;Lu et al., 2013;Lin et al., 2019;Zhang et al., 2019Zhang et al., , 2020Li et al., 2020) of most economic crops is short and cannot cover the entire adult stage of O. excavata (40-55 days; Wei et al., 2002;Men et al., 2018). Through the observation of professional beekeepers and our field work on the stocking of O. excavata, when O. excavata had no diverse food sources except economic crops, this gap seriously affects the bees to reserve beebread for their offspring, which causes a low recovery rate after artificial stocking. The number of reduced populations needs to be supplemented every year, excluding the situation of unnatural death, human activities, and population migration.
Although when the beebread from H5/H6 experimental plots were collected, the economic crops (pear and rape trees) in the plantation were in the late flowering stage, which led to the lack of pollen resources, O. excavata could still obtain pollens from wild plants (J. interior, R. gymnocarpa, R. laevigata, and R. primula; Figure 1). This means that in order to reduce the impact of the lack of pollen resources of economic crops, O. excavata can use the pollen of wild plants as their main source of nutrition for meeting survival needs. Plantations (orchards and farmlands) were dominated by a single type of economic crops and lack of flowering resources of wild plants (Figure 2A). A further analysis revealed that more wild plants in beebread were from SHNs (Figure 2A). Papanikolaou et al. (2017) found that increasing  Piko et al. (2021) concluded that the mixed flower field model can effectively provide attractive food resources, thereby greatly increasing the abundance and diversity of bumblebee. Bihaly et al. (2020) also emphasized that planting a variety of flowering herbaceous plants on a small scale could significantly increase the number of bee nests in the orchard. In response to concerns about the food source of the O. excavata, combing the previous results, a conjecture is proposed: a mixed flower field model suitable for O. excavata (MFF-OE model). We can plant a variety of favorite wild flowering plants around the plantation to meet

DATA AVAILABILITY STATEMENT
The sequence data that support the findings of this study are openly available in GenBank of NCBI at https: //www.ncbi.nlm.nih.gov/. The associated BioProject, SRA, and BioSample numbers are PRJNA742556, SRR15011441, and SAMN19967841, respectively.

ETHICS STATEMENT
All experiments procedures for this study complied with the current animal ethics guidelines and did not involve any protected animals.

AUTHOR CONTRIBUTIONS
AL and DH were responsible for the conceptualization and supervision of the study. HL, FD, and YH were responsible for the methodology and writing -original draft preparation. HL, FD, YL, KZ, and HZ were responsible for the software. ZZ, CZ, AL, and DH performed the validation. HL and FD conducted the formal analysis and visualization. FD, YL, KZ, HZ, and DH conducted the investigation. FD and DH were responsible for the resources. HL, FD, YL, KZ, HZ, and DH were responsible for data curation. YH, ZZ, CZ, AL, and DH were responsible for writing -review and editing. DH was responsible for project administration. HL, ZZ, AL, and DH were responsible for funding acquisition. All authors contributed to the article and approved the submitted version.