A Chromosome-Scale and Annotated Reference Genome Assembly of Plecia longiforceps Duda, 1934 (Diptera: Bibionidae)

Abstract

Urbanization is a leading factor effecting global biodiversity, driving rapid evolutionary processes in the local biota. Species that adapt and proliferate in city environments can become pests, with human activities facilitating their dispersal and excessive outbreaks. Here we present the first genome data of Plecia longiforceps, a lovebug pest in Eastern Asia with intensive aggregations recently occurring in the Seoul Metropolitan Area of Korea. PacBio HiFi and ONT Pore-C sequencing data were used to construct a highly continuous assembly with a total size of 707 Mb and 8 major pseudochromosomes, its integrity supported by the N50 length of 98.1 Mb and 96.8% BUSCO completeness. Structural and functional annotation using transcriptome data and ab initio predictions revealed a high proportion (69.3%) of repeat sequences, and synteny analysis with Bibio marci showed high levels of genomic collinearity. The genome will serve as an essential resource for both population genomics and molecular research on lovebug dispersal and outbreaks, and also implement studies on the eco-evolutionary processes of insects in urbanizing habitats.

Graphical Abstract

Graphical abstract.

Significance.

The lovebug species Plecia longiforceps is a recently highlighted urban pest exhibiting active patterns of outbreaks and range expansion in eastern Asia. We present a high-quality, chromosome-scale, and annotated reference genome for P. longiforceps, providing fundamental data for population and evolutionary studies to investigate the process of the species’ dispersal and adaptation to novel environments. Our results also effectively demonstrate the integration of HiFi and Pore-C sequencing technologies for de novo genome assembly in arthropods.

Introduction

Urbanization, with its rapidly increasing speed and scale, is one of the most prominent factors of environmental change and its subsequent effects on biodiversity (Simkin et al. 2022). Cities bring degradation and fragmentation of natural habitats, but also introduce novel landscapes and microclimates, all of which have substantial impacts on the species composition and interactions within the biotic community (Gibb and Hochuli 2002; Faeth et al. 2005). Furthermore, the complex transportation network of humans and goods in urban areas is often accompanied by the importation of foreign organisms, and together with the aforementioned alterations of environmental resistance, facilitates the establishment and proliferation of nonnative species (Byers 2002; McKinney 2006). These evolutionary challenges posed on the local biota drive rapid processes of adaptation (Alberti et al. 2017), as selection toward advantageous traits in human environments, such as pesticide resistance, tolerance to heat island effects, and better dispersal abilities, is frequently observed in urban communities (e.g. Piano et al. 2016; Zhu et al. 2016).

Some species that successfully adapt to human environments can become pests, defined on several criteria including their ability to threaten public health, damage artifacts, and occur in bothering numbers (Robinson 1996). Many species of true flies (Diptera) fall into the last category, as they seek benefit in survival and mating by living in large groups as eggs, larvae, or adults (Krivosheina et al. 2019). Urban environments and human activities can contribute to more intensive aggregations, as drainage systems, home gardens, and city parks provide suitable habitats for larvae, while artificial heat and volatiles attract adults (Busvine 1980; Krivosheina et al. 2019). Outbreaks also occur in the case of invasive species, with biological and environmental components determining the time and intensity of emerging populations (Crooks and Soulé 1999). Therefore, it is important to assess adaptive and evolutionary factors for effective management of outbreak pests in urban surroundings.

Due to their large size and weak, slow flying habits, mating aggregations of March flies (Bibionidae) can be especially conspicuous and thus disturbing to the public. Seasonal emergences of the St. Mark's Fly (Bibio marci) and Fever Fly (Dilophus febrilis) are regarded as nuisance pests in Europe (Busvine 1980). The Common Lovebug (Plecia nearctica) has a well-documented history of dispersal and explosive swarming along the US Gulf Coast, associated with anthropogenic factors including import through human transportation, habitation in highway subgrades, and attraction to common pollutants (Hetrick 1970; Buschman 1976; Callahan et al. 1985). However, compared with research on the ecology and pest status of March flies, little work has been done on the genomic or evolutionary level.

Plecia longiforceps is the latest member of Bibionidae to draw attention as urban pests, showing outbreak patterns closely resembling that of P. nearctica. First described from southeast China (Duda 1934), the insect spread along the Okinawa islands (Fitzgerald and Nakamura 2015; Tone and Osada 2020), and is recently causing major concern in the Seoul Metropolitan Area of Korea as mating swarms reach residential and commercial areas, with changes in seasonality suggesting adaptation to higher latitudes (Kim et al. 2022). As public demands grow for explanations to this emerging source of disturbance, research is required on the evolutionary background of successful expansions in East Asia and proliferation in urban habitats. In addition to field work and physiological experiments, genomic data will be critical in tracking the process of their dispersal, comparing urban and rural populations, and finding adaptive signals in city-associated traits. To address the need of molecular information, we constructed a chromosome-level and annotated genome assembly of P. longiforceps, based on high-fidelity (HiFi) long-read sequencing, Pore-C multiway contact sequencing, and RNA-sequencing data. Providing the first high-quality reference genome for the genus Plecia, we aim to facilitate genomic and evolutionary studies to understand the dispersal and adaptation processes of lovebugs in novel habitats.

Results and Discussion

Genome Assembly

The estimated genome size based on the k-mer distribution of HiFi reads was 702 Mb with 1.1% heterozygosity and 44.1% repeat content (supplementary fig. S1, Supplementary Material online). From the HiFi assembly, 1.0% (32 contigs) was identified as bacterial and nematode sequences and removed (supplementary fig. S2, Supplementary Material online). The final assembly using Pore-C contact information (Fig. 1a and b, Table 1) consisted of 491 scaffolds with a total length of 707 Mb, in accordance to the k-mer based estimate; also, more than 92.4% of the genome (653 Mb) was anchored to the eight largest chromosomal pseudomolecules. Genome statistics showed the N50 scaffold length of 98.1 Mb, 33.2% GC content, and 96.8% BUSCO completeness (94.1% single-copy, 2.6% duplicated, 1.2% fragmented, and 3.2% missing) (Table 1), demonstrating the high continuity and completeness of the de novo genome assembly.

Fig. 1.

Chromosome-scale genome assembly of P. longiforceps. a) Pore-C contact map of chromosomal interactions within the P. longiforceps genome assembly, showing eight major pseudomolecules. Outer boxes (blue) indicate final chromosome-level scaffolds, and inner boxes (green) are contigs after error correction in YaHS. b) Snail plot representing assembly statistics, including genome size, largest scaffold length, N50 and N90 lengths, GC composition, and BUSCO evaluation metrics. c) Genomic synteny between the eight major pseudochromosomes of P. longiforceps (pl1 to pl8) and B. marci chromosomes (bm1 to bm5, bmX). Repeat and gene densities of each pseudochromosome are depicted above, calculated in overlapping 100 kb windows.

Table 1.

Summary of genome assembly and annotation statistics for P. longiforceps

	Item	Value
Assembly statistics	Genome size (bp)	706,927,914
	Number of scaffolds	491
	Number of contigs	801
	Longest scaffold length (bp)	192,788,022
	Scaffold N50 (bp)	98,122,942
	Scaffold N90 (bp)	19,750,146
	GC content (%)	33.21
	BUSCO scores (insecta_odb10)	C: 96.7% [S: 94.1%, D: 2.6%], F: 1.2%, M: 2.1%, n: 1367
Annotation statistics	Repeat elements (length)	1,668,214 (489,828,923 bp)
	Total gene models	12,648
	Protein-coding genes	12,198
	Annotated proteins	10,699
	BUSCO scores (insecta_odb10)	C: 83.1% [S: 81.1%, D: 2.0%], F: 3.1%, M: 13.8%, n: 1367

Structural and Functional Annotation

The combined reference transcriptome consisted of 210,898 sequences with 94.9% BUSCO completeness (1,297 complete BUSCOs), and individual assemblies also had 88.6% to 92.8% completeness (supplementary fig. S3, Supplementary Material online). Of the P. longiforceps genome, 69.3% (490 Mb) was masked as repetitive, including retroelements (26.9%), DNA transposons (8.2%), and simple repeats (4.0%). However, the largest portion was unclassified (208 Mb, 29.4%), implying the lack of genomic information on related taxa. In MAKER, 12,198 genes with an average length of 7,584.42 bp were predicted, encoding proteins with an average of 467.46 amino acids, comprising 13.1% of the genome; 450 tRNAs were also predicted. The final annotated gene set showed 83.1% BUSCO completeness (Fig. 1c, Table 1, supplementary table S1, Supplementary Material online).

Functions of 10,699 gene models (87.7%) were assigned based on at least one hit from the Swiss-Prot, NR, Pfam, CDD, and eggNOG databases; 8,250 and 4,283 were assigned GO terms and KEGG pathways, respectively (Table 1, supplementary table S1, Supplementary Material online). Notably, Pfam searches assigned 128 genes to the cytochrome P450 (CYP) family (PF00067), which is relatively a high number, among insects (Dermauw et al. 2020). CYPs are well known for their role in insecticide resistance, mediating the detoxification of commonly used pesticides including pyrethroids and neonicotinoids (Manjon et al. 2018), and large CYPomes have been reported in common pests such as mosquitoes (Yang and Liu 2011) and house flies (Scott et al. 2014). The increased number of CYPs in the P. longiforceps genome may imply tolerance to insecticides, especially considering that intense swarms were observed in areas under heavy spraying due to a previous outbreak of stick insects (Jung et al. 2020). For other detoxification genes, glutathione S-transferases (PF00043 and PF02798; 35 genes) and carboxylesterases (PF00135; 57 genes) showed higher numbers of annotations compared with model species (Koirala et al. 2022; Cruse et al. 2023), while ABC transporters (PF00005, 50 genes) were moderate (Wu et al. 2019), and UDP-glucosyltransferases (PF00201, 4 genes) were exceptionally scarce (Ahn et al. 2012). Another gene group of interest, heat shock proteins (HSPs) are key factors in tolerance to stresses such as heat, cold, and desiccation, and thus indicators of adaptive competence in novel environments (King and MacRae 2015). HSPs found in this study include 28 Hsp40 (PF00226), 10 Hsp60 (PF00118), 9 Hsp70 (PF00012), and 3 Hsp90 (PF00183) annotations. Although these putative numbers of detoxification- and heat resistance-related genes will require further curation to identify full-length, partial or fragmented genes, and pseudogenes, the annotated regions provide a basis for investigations on the adaptation of P. longiforceps to urban conditions.

Genome Synteny

Synteny analysis between the P. longiforceps (pl1 to pl8) and B. marci (bm1 to bm5, bmX) genomes (Fig. 1c) identified 433 collinearity blocks containing 7,878 genes, showing high levels of collinearity between chromosomal units, specifically pl1 with bm3 and bm5, pl2 with bm1, pl3 with bm4, and pl4 and pl6 with bm2. Also, a large portion of pl8 was syntenic to the X chromosome of B. marci, suggesting its relation with sex determination. Based on this preliminary data, research on the karyotype and sex chromosome system of P. longiforceps will be required to elucidate the processes of chromosomal rearrangement in the evolution of March flies.

Conclusion

Prior to this study, no genomic data had been published for the genus Plecia. In the family Bibionidae, two chromosome-scale assemblies (B. marci [GCA_910594885], D. febrilis [GCA_958336335]) were available, with only the former annotated. As the first genomic data for Plecia, our data will facilitate comprehensive studies on the evolutionary background of their range expansion and outbreak phenomena. The assembly can serve as a high-quality reference genome to identify genomic markers such as microsatellites and single-nucleotide polymorphisms, which have been widely used in population genomics of invasive pests (e.g. Schmidt et al. 2021; Kang et al. 2023). Furthermore, through comparative studies with P. nearctica, and also with relatively rural species such as P. adiastola, we can identify advantageous traits in inhabiting city areas, and factors effecting explosive increases in population. Overall, we hope this data provides a fundamental basis for understanding the adaptation and outbreaks of lovebugs in urban ecosystems.

Materials and Methods

Sample Collection and Sequencing

Final-instar larvae of P. longiforceps were collected from forest soil at Neulsolgil Park, Incheon, Korea (37°23′20.2″N 126°43′14.3″E) in April 2023, and were reared to obtain pupa and adults. Identification was confirmed based on morphological characteristics of male terminalia (Fitzgerald and Nakamura 2015; Kim et al. 2022). Samples were flash-frozen in liquid nitrogen and stored in −80 °C.

HiFi and Pore-C experiments were conducted at the National Instrumentation Center for Environmental Management, Korea. High molecular weight genomic DNA was extracted from a single male adult specimen using the Wizard Genomic DNA Purification Kit (Promega, USA). Extract quality was assessed by 1% agarose gel electrophoresis and quantified with the Quantus Fluorometer (Promega, USA). The HiFi library was prepared using the SMRTbell prep kit 3.0 (Pacific Biosciences (PacBio), USA) and sequenced with two runs on a PacBio Sequel IIe platform. Reads with a quality score of Q20 or higher were obtained using the Circular Consensus Sequencing protocol in SMRT Link ver. 12.0.0.177059.

Twenty male adult specimens were used for Pore-C DNA extractions following the standard procedure (Deshpande et al. 2022) with slight adaptations (supplementary methods, Supplementary Material online). Pore-C libraries were constructed using the Ligation Sequencing Kit V10 (Oxford Nanopore Technologies [ONT], UK) and sequenced on an ONT PromethION platform. Reads with a phred quality score of 7 or higher were selected using the Guppy 6.5.7 software provided by ONT.

RNA-sequencing procedures were conducted at Macrogen, Korea. Total RNA was extracted from four male adults, and one each of female adult, pupa, and final-instar larva, using the RNeasy Mini Kit (Qiagen, Germany). Extractions from male adults were combined into a single sample. RNA integrity was checked using TapeStation (Agilent, USA). RNA-seq libraries were prepared using the TruSeq Stranded mRNA Library Prep Kit (Illumina, USA) and sequenced on an Illumina NovaSeq6000 platform.

Genome Assembly

Genome profiles were estimated from raw HiFi data using GenomeScope 2.0 (Ranallo-Benavidez et al. 2020) based on k-mer (k = 21) frequency. HiFi reads were assembled using Hifiasm-0.19.5 (Cheng et al. 2021). Contaminants were identified and removed from the draft assembly using BLAST + 2.14.1 (Camacho et al. 2009). Pore-C data was processed using the Workflow Pore-C pipeline (https://github.com/epi2me-labs/wf-pore-c) developed by ONT. Results were input to YaHS (Zhou et al. 2023) for chromosome-level scaffolding, and chromosomal contact information was visualized in Juicebox 2.15 (Durand et al. 2016). Genome completeness was assessed with BUSCO 5.4.7 (Manni et al. 2021) using the insecta_odb10 dataset.

Transcriptome Assembly

The four RNA-seq datasets were preprocessed to remove Illumina adapters using Trimmomatic 0.39 (Bolger et al. 2014), foreign contaminants using Kraken2-2.1.3 (Wood et al. 2019), and rRNA sequences using SortMeRNA 4.3.6 (Kopylova et al. 2012). The datasets were concatenated into a single pair for de novo transcriptome assembly using Trinity v2.15.1 (Grabherr et al. 2011). MMseqs2 v.15.6f452 (Steinegger and Söding 2017) was used to extract representative sequences to produce a reference transcriptome for P. longiforceps. Each life stage/sex transcriptome was also assembled separately using Trinity. Completeness of each assembly was assessed with BUSCO using the insecta_odb10 dataset.

Gene Annotation and Synteny Analysis

The MAKER v3.01.04 pipeline (Campbell et al. 2014) was used for gene annotation. A taxon-specific repeat model was generated using RepeatModeler 2.0.5 (Flynn et al. 2020) and combined with FamDB models for repeat identification in RepeatMasker 4.1.5 (Smit et al. 2013). The repeat model, reference transcriptome, and Diptera protein data from Swiss-Prot were used for the initial round of gene prediction. SNAP v.2006-07-28 (Korf 2004) and AUGUSTUS 3.5.0 (Stanke et al. 2008) were trained for ab initio gene prediction in the next two rounds of MAKER to produce the final transcript and protein sequence datasets. Completeness of the annotated protein set was assessed with BUSCO using the insecta_odb10 dataset.

MMseqs2 was used to search the protein models against Swiss-Prot, NCBI NR proteins, Pfam, and CDD databases, and eggNOG-mapper v2.1.12 (Cantalapiedra et al. 2021) was used to search against the EggNOG database and assign GO terms and KEGG pathways. For genome synteny analysis with the published B. marci genome (Sivell et al. 2021), MMseqs2 was used to search for matching protein sequences between P. longiforceps and B. marci, and collinearity blocks were generated using MCScanX (Wang et al. 2012). Detailed procedures on genome assembly and annotation are available in supplementary methods, Supplementary Material online.

Supplementary Material

Supplementary material is available at Genome Biology and Evolution online.

Funding

This work was supported by the National Institute of Biological Resources funded by the Ministry of Environment of the Republic of Korea [NIBR202405104; NIBR202410201]; the National Research Foundation funded by the Ministry of Science and ICT of the Republic of Korea [2021R1C1C1003452; 2021R1A5A1033157], and the Creative-Pioneering Researcher Program through Seoul National University.

Data Availability

Raw sequencing data, genome assembly, and transcriptome assemblies are deposited at NCBI under the BioProject PRJNA1126849. Predicted protein and transcript models are available at Figshare (DOI: 10.6084/m9.figshare.26112745).