All procedures involving animals in this study were conducted in strict accordance with the Animal Ethics Procedures and Guidelines of the People’s Republic of China and complied with internationally accepted principles for the ethical use of animals in research. The capture, handling, examination, and specimen collection of rodents were formally reviewed and approved by the Animals’ Ethics Committees of Dali Medical College (1990–2000) and Dali University (from 2001 onward), which were responsible for ethical oversight during different phases of this long-term investigation.

Ethical approval was granted under the following approval codes and dates: DLYXY1990-0109 (approved on 9 January 1990), DLXY2001-1116 (approved on 16 November 2001), and DLDXLL2020-1104 (approved on 4 November 2020). These approvals collectively covered all field investigations conducted between 2000 and 2024 across the five provincial regions of southwest China.

Rodent trapping and handling were performed by trained personnel using standardized and humane protocols designed to minimize animal stress and suffering. Captured animals were handled individually, examined promptly, and processed following approved biosafety and animal welfare procedures. All instruments were disinfected between examinations, and appropriate personal protective equipment was used by field and laboratory staff to ensure both animal welfare and investigator safety.

As this study represents a retrospective analysis of data and specimens obtained from long-term routine surveillance and ecological investigations, no experimental manipulation or invasive procedures beyond standard parasitological examination were performed. Representative voucher specimens of rodent hosts and ectoparasitic insects were deposited in the specimen repository of the Institute of Pathogens and Vectors, Dali University, ensuring traceability, reproducibility, and future reference.

This study was a retrospective analysis based on long-term historical field investigations and taxonomic identification conducted at 116 survey sites across five provincial regions of southwest China (Yunnan, Sichuan, Guizhou, Chongqing, and Xizang) between 2000 and 2024 (21°08′–33°41′ N, 97°21′–110°11′ E). Only survey sites where both fleas and sucking lice were investigated were included, while sites with only one insect group or ectoparasitic mites were excluded.

Simultaneous investigations of fleas and sucking lice were conducted at 10 survey sites in eastern Xizang Autonomous Region. Most areas of Xizang were not covered due to high altitude, harsh climatic conditions, limited accessibility, and logistical constraints. Field surveys were conducted mainly from March to May and from September to November to avoid the rainy season and extreme winter conditions.

At each survey site, cage traps (18 × 12 × 9 cm; Guixi Mousetrap Apparatus Factory, Jiangxi, China) were placed in indoor habitats (human dwellings, barns, and stables) and outdoor habitats (farmlands, shrublands, and woodlands). Trap bait was adjusted by habitat type, with corn or peanuts used outdoors and steamed bread or fried dough used indoors.

In open habitats, traps were arranged in straight lines, spaced 5 m apart within lines and 20 m between lines, with 25 traps per line. Indoors, traps were placed at approximately 15 m² intervals along walls. An average of 200 traps were set daily for 15 consecutive days at each site, resulting in approximately 3,000 trap placements per site to ensure methodological consistency and comparability [38, 41].

Captured rodents were collected the following morning, individually placed in cloth bags, and transported to the laboratory. Fleas and sucking lice were collected by combing and manual inspection of each rodent over a white tray. Ectoparasites from each host were preserved separately in 70%–75% ethanol. Rodents were identified morphologically using standard external measurements, and host sex and age were determined based on body size, pelage characteristics, genital development, and anogenital distance [42–44]. All instruments were disinfected with 75% ethanol (Wuhan Jisi Instruments & Equipment, China) between examinations, and field personnel used appropriate personal protective equipment [2].

Collected fleas and sucking lice were digested in 5% or 10% sodium hydroxide (Shanghai Macklin Biochemical Technology, China) or potassium hydroxide (Shanghai Aladdin Biochemical Technology, China), dehydrated through a graded ethanol series (30%–100%), clarified in ethanol–xylene (Shanghai Aladdin Biochemical Technology) (1:1, v/v), then in xylene, and mounted on glass slides (Jiangsu Qipinsheng Medical Supplies Co., Ltd, China) with Canadian balsam (Shanghai Macklin Biochemical Technology) or equivalent mounting media [16, 45]. After oven drying, specimens were examined under an Olympus CX31 trinocular microscope (Olympus Corporation, Tokyo, Japan). Species identification was performed using standard taxonomic keys and the literature, and all identifications were independently verified by multiple taxonomists [16, 46, 47]. Damaged or unidentifiable specimens were excluded from statistical analyses.

Infestation was assessed using the constituent ratio (C_r), prevalence (P_M), mean abundance (MA), and mean intensity (MI). P_M is the percentage of hosts infested with ectoparasitic insects. MA is the average number of insects per examined host, and MI is the average number of insects per infested host.

The indices were calculated as follows [38, 39]:

C_r = (N_i / N) × 100%

P_M = (H_i / H) × 100%

MA = N_i / H

MI = N_i / H_i

where N_i is the number of individuals of insect species i, N is the total number of insects collected, H is the total number of hosts examined, and H_i is the number of infested hosts.

Differences in P_M were analyzed using the chi-square (χ²) test. MA and MI were compared using nonparametric tests (Mann–Whitney U test or Kruskal–Wallis H test), with Bonferroni correction for multiple comparisons. Statistical significance was set at p < 0.05. All analyses were performed using SPSS 26.0 software (IBM Corporation, NY, USA).

Community characteristics were evaluated using species richness (S), Shannon–Wiener diversity index (H’), Pielou evenness index (E), Simpson dominance index (D), and Berg–Parker dominance index (d), as follows [48]:

H’ = −Σ (N_i / N) ln (N_i / N)

E = H’ / ln S

D = Σ (N_i / N)²

d = N_max / N

where N_i is the number of individuals of species i, N is the total number of individuals, N_max is the number of individuals of the most abundant species, and ln denotes the natural logarithm.

The association coefficient (V) was used to evaluate the relationship between flea and sucking louse infestations in R. tanezumi [49]:

V = (ad − bc) / √[(a + b)(a + c)(b + d)(c + d)]

where a represents hosts infested with both fleas and lice, b represents hosts infested with lice only, c represents hosts infested with fleas only, and d represents hosts free of both groups. Values of V range from −1 to +1. Statistical significance was assessed using the chi-square test.

Spearman’s rank correlation coefficient (r) was used to assess interspecific relationships among flea and louse species [50]:

r = Σ (X_i − X̅)(Y_i − Y̅) / √[Σ (X_i − X̅)² Σ (Y_i − Y̅)²]

where X_i and Y_i are observed values of variables X and Y, X̅ and Y̅ are their respective means, and n is the number of samples. Heat maps were generated using Origin 2024 software (OriginLab Corporation, USA) to visualize correlation patterns.

The nutritional status of R. tanezumi was evaluated using the relative fatness index (K), calculated as [39, 51–53]:

K = 100W / L³

where W is body weight (g) and L is body length (cm). Individual K values were calculated for each rat, and the mean K value (mean ± standard deviation) was used as a threshold to classify hosts into low-fatness and high-fatness groups for comparative analyses.

A total of 3,069 R. tanezumi were captured from 64 of the 116 survey sites (Figure 1). Among these 64 sites, ectoparasitic insect infestation was detected at 51 sites, including 32 sites where rats were co-infested with two insect groups (fleas and sucking lice), 12 sites with flea-only infestations, and 7 sites with louse-only infestations, for a total of 51 infested sites (32 + 12 + 7 = 51). Of the 3,069 R. tanezumi examined, 1,240 were infested with ectoparasitic insects, and a total of 12,539 insect specimens were collected. The 1,240 infested rats comprised 131 individuals co-infested with fleas and lice, 486 infested only with fleas, and 623 infested only with lice (131 + 486 + 623 = 1,240). All ectoparasitic insects were taxonomically identified as belonging to 7 families, 25 genera, and 34 species within two orders, Siphonaptera (fleas) and Phthiraptera (sucking lice). The overall infestation indices for ectoparasitic insects (fleas + lice) on R. tanezumi were P_M = 40.40%, MA = 4.09 insects per examined host, and MI = 10.11 insects per infested host. Among the 34 insect species identified, 10 flea species are known or potential vectors of zoonotic diseases, including plague, murine typhus, flea-borne spotted fever, and bartonellosis. These vector flea species were Xenopsylla cheopis, Leptopsylla segnis, Monopsyllus anisus, Pulex irritans, Leptopsylla serinus, Stenoponia apertus, Ctenocephalides felis, Paradoxopsyllus custodis, Frontopsylla spadix, and Neopsylla specialis. Among them, X. cheopis is the most important vector of plague and murine typhus in southwest China. Details of these vector flea species and the zoonotic diseases they transmit are presented in Table 1. [8, 10–12, 14, 16, 54–62].

Of the 12,539 ectoparasitic insects collected from R. tanezumi, 2,442 fleas were identified as 30 species, 23 genera, and five families within the Order Siphonaptera, whereas 10,097 sucking lice were identified as four species, two genera, and two families within the Order Phthiraptera. Flea species accounted for 88.24% of all insect species but only 19.48% of individuals, whereas sucking lice represented 11.76% of species and 80.52% of individuals. Compared with sucking lice (four species with 10,097 individuals), fleas exhibited markedly higher species richness (30 spp.) but lower individual abundance (2,442 individuals). The overall infestation indices for fleas were P_M = 20.10%, MA = 0.80, and MI = 3.96, whereas those for sucking lice were P_M = 24.57%, MA = 3.29, and MI = 13.39. The infestation indices for sucking lice were significantly higher than those for fleas (p < 0.05).

A total of 617 R. tanezumi rats were infested with fleas, including 486 rats infested only with fleas and 131 rats co-infested with fleas and sucking lice. The hierarchical taxonomic structure of fleas and the proportional distribution of flea individuals across taxonomic levels were visualized using a Sankey diagram (Figure 2), illustrating the constituent ratios of 2,442 flea specimens across one order, five families, 23 genera, and 30 species. At the family and genus levels, Pulicidae (C_r = 65.68%) and Xenopsylla (C_r = 65.48%) were dominant, followed by Leptopsyllidae (C_r = 25.80%) and Leptopsylla (C_r = 19.78%). At the species level, X. cheopis and L. segnis were the dominant flea species, jointly accounting for 85.26% of total flea individuals (Table 2, Figure 2). Among the 30 flea species identified, the infestation indices of X. cheopis (P_M = 13.16%, MA = 0.52, MI = 3.96) were significantly higher than those of other flea species (p < 0.05) (Table 2).

A total of 754 R. tanezumi rats were infested with sucking lice, including 623 rats infested only with lice and 131 rats co-infested with lice and fleas. Among the two louse families and genera identified, Hoplopleuridae and Hoplopleura accounted for 65.01% of the total lice, whereas Polyplacidae and Polyplax accounted for 34.99%. Of the four louse species recorded, Hoplopleura pacifica was dominant, representing 65.01% (6,564/10,097) of total lice, followed by Polyplax spinulosa at 34.62% (3,496/10,097). The infestation indices of H. pacifica (P_M = 16.68%, MA = 2.14, MI = 12.82) were significantly higher than those of P. spinulosa (P_M = 14.34%, MA = 1.14, MI = 7.95) (p < 0.05). Moreover, the infestation indices (P_M, MA, and MI) of H. pacifica were significantly higher than those of the two dominant flea species (p < 0.05) (Table 2). Representative micrographs of two dominant flea species (X. cheopis and L. segnis) and two dominant louse species (H. pacifica and P. spinulosa) are shown in Figure 3.

Of the 3,069 R. tanezumi rats examined, 3,042 had sex records (1,517 males and 1,525 females), and 3,061 had age records (2,197 adults and 864 juveniles). Infestation differences by sex and age were visualized using radar charts (Figure 4). Male rats exhibited higher constituent ratio, prevalence, and MA of sucking lice (C_r = 67.11%, P_M = 26.83%, MA = 4.45) than female rats (C_r = 32.89%, P_M = 22.62%, MA = 2.17) (p < 0.05). Adult hosts showed higher C_r (87.92%) and infestation indices (P_M = 26.81%, MA = 4.04, MI = 15.07) than juvenile hosts (P_M = 19.10%, MA = 1.41, MI = 7.39) (p < 0.001). In contrast, differences in flea infestation indices between sexes and age groups were not statistically significant (p > 0.05) (Figure 4).

Based on relative fatness (K), rats were categorized into low-fatness (K = 2.29 ± 0.33 g/cm³) and high-fatness (K = 3.52 ± 0.87 g/cm³) groups. Flea prevalence was higher in the low-fatness group (P_M = 17.88%) than in the high-fatness group (P_M = 13.02%), whereas MA was lower (MA = 0.51 vs. 0.57) (p < 0.05). For sucking lice, both P_M and MA were significantly higher in the low-fatness group (P_M = 27.87%, MA = 5.16) than in the high-fatness group (P_M = 18.93%, MA = 2.31) (p < 0.001). MI did not differ significantly between nutritional groups for either insect group (p > 0.05) (Table 3).

Infestation indices of fleas and sucking lice varied across environmental gradients (Figure 5). Along longitude gradients, both insect groups had the highest C_r at 97–100°E, whereas the highest P_M and MA occurred at ≥109°E (p < 0.05). Along latitude gradients, fleas showed the highest C_r, P_M, and MA at 24–26°N (p < 0.001). Sucking lice exhibited the highest C_r and P_M at 21–23°N (p < 0.001) and the highest MA at 27–29°N (p < 0.001). Along altitude gradients, fleas showed the highest C_r and MI at ≤1000 m (p < 0.05) but the highest P_M and MA at 2001–3000 m (p < 0.05). Sucking lice had the highest P_M, MA, and MI at >3000 m (p < 0.05) (Figure 5). Habitat-based analyses indicated that fleas had significantly higher P_M and MA in indoor habitats, whereas sucking lice showed higher C_r, P_M, and MA in outdoor habitats (p < 0.001). Fleas exhibited higher C_r, P_M, and MA in flatland landscapes (p < 0.001), whereas sucking lice showed higher infestation indices in mountainous landscapes (p < 0.05) (Table 4).

The flea community on R. tanezumi was characterized by high species richness (30 species) but low individual abundance (2,442 individuals), whereas the louse community comprised only four species with high abundance (10,097 individuals). Consequently, flea communities showed higher species richness and Shannon–Wiener diversity index, whereas louse communities exhibited higher Pielou evenness and Simpson dominance indices (Table 5).

Within flea and louse populations, females had higher constituent ratios than males. Adult lice accounted for a substantially higher proportion of the louse population than juvenile lice (Figure 6). In contrast, only adult fleas were observed because juvenile flea stages are not ectoparasitic.

The association coefficient (V) was used to assess the relationship between flea and sucking louse infestations, based on the contingency data presented in Table 6. The calculated value of V was −0.0389, close to zero and statistically significant (p < 0.05), indicating an essentially independent distribution of fleas and sucking lice on R. tanezumi (Table 6).

Interspecific relationships among the seven dominant insect species (five flea species and two louse species) were assessed using Spearman’s rank correlation coefficient and visualized as a heatmap (Figure 7). Positive correlations (0 < r < 1) are shown in red, and negative correlations (−1 < r < 0) are shown in blue, with significance levels indicated as ***p < 0.001, **p < 0.01, and *p < 0.05. The analysis revealed both positive and negative associations among species. For example, M. anisus and P. custodis showed a strong positive correlation (r = 0.90, p < 0.001), whereas H. pacifica and P. spinulosa exhibited a significant negative correlation (r = −0.53, p < 0.001) (Figure 7).

Southwest China, as defined in this study, comprises five provincial regions (Yunnan, Guizhou, Sichuan, Chongqing, and Xizang) and covers a vast area of approximately 2,341,467 km², with a population of about 205 million. This region is a natural focus for numerous zoonotic diseases, including plague, murine typhus, bartonellosis, leptospirosis, HFRS, and scrub typhus. Historically, plague and murine typhus, two major flea-borne diseases, were prevalent in Yunnan, Sichuan, and Xizang, with successive human cases reported over extended periods [63–68]. More recent studies have identified Yunnan and Sichuan as important endemic areas for bartonellosis, with rodents, cats, and companion animals as the primary sources of infection [28, 31]. Notably, 1,157 human cases of murine typhus were reported in Xishuangbanna Prefecture, Yunnan Province, in 2011 alone, corresponding to an incidence of 102.10 per 100,000 population [66]. The oriental house rat (R. tanezumi) is not only a common agricultural pest but also a significant medical and veterinary pest because of its close association with a wide range of zoonotic diseases [32, 33]. The pathogens responsible for these diseases can be readily transmitted among rat populations and from rats to humans or domestic animals via the blood-feeding activity of ectoparasitic insects, particularly fleas [8, 10, 13]. Consequently, investigating the infestation and distribution of ectoparasitic insects on R. tanezumi in southwest China is of considerable medical and veterinary importance.

Previous studies in southwest China have primarily focused on ectoparasitic mites, including chiggers and gamasid mites, on R. tanezumi [2, 3, 34, 40], while ectoparasitic insects have not been comprehensively analyzed due to limited taxonomic coverage. Earlier parasitological investigations in this region typically concentrated on mites or, in some cases, on fleas as a single insect group [54, 69]. In contrast, the present study is the first to simultaneously investigate fleas and sucking lice parasitizing R. tanezumi across five provinces in southwest China. The original data were derived from a multi-provincial field investigation conducted between 2000 and 2024 (Figure 1). The extensive spatial coverage, encompassing 116 survey sites, provides a large-scale ecological dataset on rodent-associated ectoparasites across this vast territory. In total, 3,069 R. tanezumi individuals were captured at 64 of the 116 survey sites, and these sites were distributed across all five provincial regions (Figure 1), further confirming that R. tanezumi is a widely distributed and dominant rodent species in southwest China [2, 3]. In this study, 34 species and 12,539 individuals of ectoparasitic insects (fleas and sucking lice) were identified from R. tanezumi, with overall infestation indices of P_M = 40.40%, MA = 4.09 insects per examined host, and MI = 10.11 insects per infested host. These results clearly demonstrate that R. tanezumi is highly susceptible to infestation by both fleas and sucking lice.

Among the 34 ectoparasitic insect species identified, 30 were flea species and only four were louse species, indicating substantially higher species diversity in fleas than in sucking lice. Globally, nearly 3,000 flea species have been described, with more than 600 species recorded in China [70–72]. A considerable proportion of flea species are recognized vectors of zoonotic pathogens, including those causing plague, murine typhus, flea-borne spotted fever, and bartonellosis [7–16]. In the present study, 10 of the 30 flea species identified on R. tanezumi were confirmed as vector species capable of transmitting plague, murine typhus, and other zoonotic diseases [54–62]. These species included X. cheopis, L. segnis, M. anisus, P. irritans, L. serinus, S. apertus, C. felis, P. custodis, F. spadix, and N. specialis (Table 1). In southwest China, plague occurs in both domestic rodent and wild rodent foci. X. cheopis is the principal vector in domestic rodent plague foci and also the major vector of murine typhus in these areas and surrounding regions [16]. In contrast, F. spadix and N. specialis are the dominant vectors in wild rodent plague foci [60, 61]. Moreover, X. cheopis, P. irritans, C. felis, and L. segnis also function as intermediate hosts for Hymenolepis nana, Hymenolepis diminuta, and Dipylidium caninum [8, 10–12, 14, 16, 55–57]. To date, no previous study has reported such a comprehensive assemblage of ten vector flea species on a single rodent host in southwest China. This finding provides new evidence of multi-vector coexistence and highlights potential zoonotic transmission hotspots associated with R. tanezumi populations.

Compared with sucking lice (four species comprising 10,097 individuals), fleas exhibited high species richness (30 species) but low individual abundance (2,442 individuals) on R. tanezumi. Overall infestation indices for sucking lice (P_M = 24.57%, MA = 3.29, MI = 13.39) were significantly higher than those for fleas (P_M = 20.10%, MA = 0.80, MI = 3.96) (p < 0.05). Despite the heavier infestation burden of sucking lice, the flea community showed substantially higher species richness (S) and Shannon–Wiener diversity index (H’) than the louse community (Table 5). This contrast reflects fundamental differences in community structure, with the louse community characterized by low species diversity and high dominance. Such patterns are closely related to life-history traits and host specificity. Sucking lice are obligate ectoparasites of eutherian mammals that permanently reside on their hosts throughout all life stages, resulting in high host specificity developed through long-term coevolution [73–75]. In contrast, fleas possess a four-stage life cycle, with only adult stages parasitic on warm-blooded animals, and many flea species exhibit low host specificity [14, 76–82]. In the present study, as many as 30 flea species were found on R. tanezumi, whereas only four louse species were identified, further illustrating these ecological differences. The lower host specificity of fleas facilitates pathogen transmission among diverse wild and domestic hosts, including humans [83, 84], whereas the high host specificity of sucking lice limits their inter-host transmission potential [85, 86].

Significant sex- and age-related biases were observed in sucking louse infestation. Male R. tanezumi exhibited significantly higher constituent ratios and infestation indices (P_M and MA) than females (p < 0.05), and adult rats showed higher C_r and infestation indices (P_M, MA, and MI) than juveniles (p < 0.001). In contrast, flea infestation indices did not differ significantly among host sexes or age groups (p > 0.05) (Figure 4). These patterns suggest that host physiological status has a stronger influence on obligate ectoparasites such as sucking lice than on fleas. Similar sex- and age-related biases in parasitic infections have been reported previously [87, 88] and may be attributed to differences in energy expenditure, immune function, body surface area, and exposure risk between males and females or between adults and juveniles [89, 90].

Host nutritional status also influenced ectoparasite infestation. Relative fatness (K), a widely used index of host nutritional condition [39, 51, 52], showed that rats with poor nutrition (low-fatness) had significantly higher P_M and MA of sucking lice than those with better nutritional status (p < 0.001) (Table 3). This finding provides novel evidence that poor nutritional condition increases susceptibility to louse infestation in R. tanezumi. In contrast, the effect of host nutrition on flea infestation appeared inconsistent. Although flea P_M was higher in the low-fatness group (p < 0.05), MA was lower in the high-fatness group (p < 0.05), suggesting that the relationship between flea infestation and host nutrition may be complex and warrants further investigation.

Beyond host-related factors, ectoparasite infestation was strongly influenced by environmental conditions. Infestation indices for fleas and sucking lice varied along gradients of longitude, latitude, altitude, habitat type, and geographical landscape (Table 4, Figure 5), reflecting pronounced environmental heterogeneity within the same host species. These findings are consistent with previous studies showing that ectoparasite infestation can vary substantially under different environmental conditions, even within the same host species [41,91]. Integrating multiple environmental gradients in the present study identifies niche preferences and habitat-dependent infestation patterns of fleas and sucking lice on R. tanezumi that have not been previously reported at this spatial scale.

The association coefficient (V) was used to assess the relationship between flea and sucking louse infestations in R. tanezumi. The calculated V value was close to zero (V = −0.0389, p < 0.05) (Table 6), indicating that infestations by fleas and sucking lice are essentially independent. This suggests that the presence of one insect group does not significantly influence infestation by the other, and that neither competitive exclusion nor facilitation occurs between these two ectoparasite groups on R. tanezumi [49, 92]. To our knowledge, this study is among the first to quantitatively demonstrate the independence of flea and louse distributions on a single rodent host using association coefficient analysis.

Heatmap-based interspecific correlation analysis using Spearman’s rank correlation coefficient revealed varying degrees of positive and negative associations among flea and louse species (Figure 7). Positive correlations indicate a tendency for co-occurrence on the same host, whereas negative correlations suggest potential interspecific competition during host selection [93, 94]. Within ectoparasite populations, female fleas and female sucking lice had higher constituent ratios than males (Figure 6), likely reflecting longer female lifespan in many insect species [95, 96]. The markedly higher proportion of adult lice relative to juveniles remains unexplained and highlights the need for further investigation into louse population dynamics.

Although this retrospective study provides comprehensive insights into the infestation status, distribution patterns, and ecological characteristics of ectoparasitic insects on R. tanezumi in southwest China, several limitations should be acknowledged. Survey sites were not evenly distributed across all provincial regions, with only 10 sites in eastern Xizang, leaving large areas unsampled (Figure 1). The long study period (2000–2024) also introduces potential temporal variation related to climate and vegetation changes, which were not analyzed due to a lack of long-term environmental data. Additionally, this study focused on overall infestation patterns rather than seasonal dynamics. Detailed investigations of seasonal variation would require fixed-site, monthly sampling over at least one full year, representing a fundamentally different study design.