*Article Trichinella* **spp. in Wild Boars (***Sus scrofa***), Brown Bears (***Ursus arctos***), Eurasian Lynxes (***Lynx lynx***) and Badgers (***Meles meles***) in Estonia, 2007–2014**

**Age Kärssin 1,2,\*, Liidia Häkkinen 1, Annika Vilem 1,2, Pikka Jokelainen 2,3,4 and Brian Lassen <sup>5</sup>**


**Simple Summary:** Trichinellosis is an important foodborne zoonosis. In Estonia, *Trichinella* infections are endemic in wild animals. This paper summarizes findings of *Trichinella*-parasites during an 8-year period in Estonia in selected host species: wild boars, brown bears, Eurasian lynxes, and badgers. The results highlight that testing wildlife hunted for human consumption for *Trichinella* is important, and that there is room for improvement in the proportion of hunted animals tested.

**Abstract:** In this study, we summarize *Trichinella* findings from four wild, free-ranging host species from Estonia during 2007–2014. *Trichinella* spp. larvae were detected in 281 (0.9%, 95% confidence interval (CI) 0.8–1.0) of 30,566 wild boars (*Sus scrofa*), 63 (14.7%, 95% CI 11.6–18.3) of 429 brown bears (*Ursus arctos*), 59 (65.56%, 95% CI 55.3–74.8) of 90 Eurasian lynxes (*Lynx lynx*), and three (60.0%, 95% CI 18.2–92.7) of five badgers (*Meles meles*). All four European *Trichinella* species were detected: *T. britovi* in 0.7% of the wild boars, 7.2% of the brown bears, 45.6% of the lynxes, and 40.0% of the badgers; *T. nativa* in 0.1% of the wild boars, 5.8% of the brown bears, and 20.0% of the lynxes; *T. pseudospiralis* in 0.02% the wild boars; and *T. spiralis* in 0.03% of the wild boars and 4.4% of the lynxes. The results include the first description from Estonia of *T. britovi* in brown bear and badgers, *T. pseudospiralis* in wild boars, and *T. spiralis* in wild boars and lynxes. The results indicate high infection pressure in the sylvatic cycles across the years—illustrating continuous risk of spillover to domestic cycles and of transmission to humans.

**Keywords:** foodborne; game meat; *Trichinella*; wildlife; zoonosis

#### **1. Introduction**

*Trichinella* spp. are zoonotic parasitic nematodes that can be transmitted to humans by consumption of undercooked or raw meat of infected animals. A multicriteria-based approach placed *Trichinella spiralis* as the third and *Trichinella* spp. other than *T. spiralis* as the fifth on a prioritization ranking list of foodborne parasites in Europe, and the fourth and the third, respectively, in Eastern Europe [1].

Meat of game animals, especially meat of wild boars (*Sus scrofa*), is considered one of the main sources of *Trichinella* infections for humans in Europe [2], and it is acknowledged as the main source in Estonia [3]. Cases of human trichinellosis have been reported from Estonia [4], and the proportion testing positive for antibodies against *Trichinella* spp. was 3.1% in the general adult human population and 4.9% among hunters [5].

**Citation:** Kärssin, A.; Häkkinen, L.; Vilem, A.; Jokelainen, P.; Lassen, B. *Trichinella* spp. in Wild Boars (*Sus scrofa*), Brown Bears (*Ursus arctos*), Eurasian Lynxes (*Lynx lynx*) and Badgers (*Meles meles*) in Estonia, 2007–2014. *Animals* **2021**, *11*, 183. https://doi.org/10.3390/ani110 10183

Received: 7 December 2020 Accepted: 11 January 2021 Published: 14 January 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

*Trichinella* spp. are endemic in wildlife in Estonia [6]. A high proportion, 42.1%, of investigated wild boars that were hunted in 2012–2013 tested positive for antibodies against *Trichinella* [7], and the biomass of *Trichinella* has increased in the main reservoir hosts raccoon dogs (*Nyctereutes procyonoides*) and red foxes (*Vulpes vulpes*) [6,8]. To add to the knowledge on epidemiology of these zoonotic parasites, the aim of this study was to summarize *Trichinella* findings during 2007–2014 in selected sylvatic hosts that are hunted in Estonia: wild boars, brown bears (*Ursus arctos*), Eurasian lynxes (*Lynx lynx*), and badgers (*Meles meles*).

#### **2. Materials and Methods**

#### *2.1. Ethics*

No animals were killed for the purpose of this study. No data of the hunters were handled in this study.

#### *2.2. Setting*

Estonia is a Baltic country located in northeastern Europe. Altogether 158,670 wild boars, 348 brown bears, 797 lynxes and 1527 badgers were hunted in Estonia in 2007– 2014 [9]. Meat of all these host species included in this study is eaten in the country.

#### *2.3. Samples and Data*

The muscle samples included in this study were collected from wild boars, brown bears, lynxes, and badgers, primarily from the predilection muscle groups (diaphragm, muscles of foreleg, or tongue), from across Estonia in 2007–2014 by hunters and by meat inspectors in game meat processing plants. The samples were sent to the Estonian Veterinary and Food Laboratory for *Trichinella* testing as part of meat inspection, either for primary or confirmatory testing.

Data on sex and age category of the animal, the county where the animal was hunted, and the year when the animal was hunted were extracted from the submission forms that accompanied the samples. The age category of wild boars, brown bears and lynxes was 'juvenile' for animals the hunters estimated to be up to 2 years of age and 'adult' for animals the hunters estimated to be over two years of age. The counties were categorized into eastern vs. western counties and southern vs. northern counties (Table 1).


*Animals* **2021**, *11*, 183

**Table 1.** Estonia, 2007–2014, by sex, age category, region, and year. Univariable

Prevalence

 of

*Trichinella*

infection in wild boars (*Sus scrofa*), brown bears (*Ursus arctos*), Eurasian lynxes (*Lynx lynx*) and badgers (*Meles meles*) hunted in

 odds to test positive in comparison

 to the reference sex (male), age category

(juvenile), east-west


**Table 1.** *Cont.*


**Table 1.** *Cont.*


**Table 1.** *Cont.* *Trichinella pseudospiralis;* Ts: *Trichinella spiralis*; Tb+Tn: mixed infection with *Trichinella britovi* and *Trichinella nativa*; Tb+Ts: mixed infection with *Trichinella britovi* and *Trichinella spiralis;* Tspp:

*Trichinella* species, no species-level result.

#### *2.4. Artificial Digestion*

The artificial digestion of the samples was carried out at the Estonian Veterinary and Food Laboratory, which is the national reference laboratory for parasites with its three regional laboratories. The laboratories are accredited by ISO 17025 and authorized as official laboratories for *Trichinella* digestion analyses according to EU 2075/2005 Annex I Chapter 1 [10]. One of the regional laboratories used Stomacher method according to EU 2075/2005 Annex I Chapter II [10] until 2009.

The testing included both primary and confirmatory testing. Other laboratories performing *Trichinella* testing send positive samples to the national reference laboratory for confirmation and species identification.

A minimum of 10 g muscle tissue was used from each animal in a pooled sample and 50 g for an individual sample, with the exception that in 2007, one regional laboratory used 5 g muscle tissue as the minimum in pooled samples for up to 20 animals. If a pooled sample was positive, the pool was divided into smaller pools and individual samples were tested to identify the infected animals.

If larvae were found, they were identified morphologically, counted, washed with tap water, and stored in ethanol, according to the procedures recommended by the European Union Reference Laboratory for Parasites [11]. The number of larvae per gram of muscle tissue (lpg) was calculated for each positive animal.

#### *2.5. Molecular Analysis*

Larvae collected in 2007–2010 were identified to the species level at the European Union Reference Laboratory [11], and larvae collected since 2011 at the Estonian Veterinary and Food Laboratory. The same multiplex PCR [12] was used for all the analyses. The method does not include sequencing.

#### *2.6. Statistical Analyses*

Only results from primary testing were used for prevalence estimates. Animals were excluded from statistical analyses if their individual infection status could not be determined, due to testing as part of a pooled sample followed by unsuccessful identification of the infected individuals.

We compared the prevalence estimates by sex, by age group, by eastern vs western counties and by southern vs northern counties, using two-by-two tables. In addition, we report univariable odds ratios for testing positive for *Trichinella*, using these same dichotomous variables, as well as counties as dummy variables and years as dummy variables.

For the statistical analyses, we used Microsoft Excel, OpenEpi and R [13,14]. We report 95% confidence intervals (95% CI, Mid-P Exact) for proportions. Associations were considered statistically significant if two-tailed *p* < 0.05.

#### **3. Results**

The proportion of animals included in this study from the officially reported hunting bag of the study period was 19.3% for wild boars, 123.6% for brown bears, 12.0% for lynxes and 0.3% for badgers [9]. A total of 44 wild boars and two lynxes were excluded from statistical analyses, because their individual infection status could not be determined. The final sample was 31,090 animals tested as primary testing (Table S1), and 20 positive animals (15 wild boars, one brown bear and four lynxes) that had been tested for confirmatory purposes. Data on larval burden were missing for one wild boar and two lynxes. For *Trichinella* spp. species identification, altogether 426 larval samples were tested; *Trichinella* species was not determined in 70 (16.4%) of the larval samples.

Of the altogether 31,090 animals tested as primary testing, 406 (1.3%, 95% CI 1.2–1.4) were positive for *Trichinella* spp. larvae. Altogether 281 (0.9%, 95% CI 0.8–1.0) of the 30,566 wild boars, 63 (14.7%, 95% CI 11.6–18.3) of the 429 brown bears, 59 (65.6%, 95% CI 55.3–74.8) of the 90 lynxes, and three (60.0%, 95% CI 18.2–92.7) of the five badgers were

*Trichinella* positive (Table 1). In wild boars and lynxes, a higher prevalence was observed in adults than in juveniles (*p* = 0.003 and *p* = 0.045, respectively) (Table 1). In wild boars, the prevalence was higher in the western counties than in the eastern counties (*p* < 0.001), and in the northern counties than in the southern counties (*p* = 0.026) (Table 1). The prevalence in lynxes was higher in the eastern counties than in the western counties (*p* = 0.020), and in the northern counties than in the southern counties (*p* = 0.045). The prevalence varied by year from 0.4% to 1.6% in wild boars, from 7.9% to 36.7% in brown bears, from 41.7% to 90.9% in lynxes, and from 0.0% to 100.0% in badgers (Figure 1, Table 1). In wild boars, the prevalence was higher in 2011 (*p* = 0.001), 2013 (*p* < 0.001), and 2014 (*p* < 0.001) than it was in 2007 (Table 1).

The larval burden appeared generally higher in wild boars than in brown bears and lynxes (Figure 1, Table 1). Nine wild boars had more than 100 lpg.

**Figure 1.** Percentage of *Trichinella* spp. positive animals; Mean (solid line) and median (dashed line) number of larvae per gram tissue in tested wild boars (*Sus scrofa*, **A**), brown bears (*Ursus arctos*, **B**), Eurasian lynxes (*Lynx lynx*, **C**) and badgers (*Meles meles*, **D**), 2007–2014, Estonia.

Mono-species *Trichinella* infection was found in 97.5% (95% CI 95.0–99.0) of the wild boars, 94.3% (95% CI 85.4–98.5) of the brown bears, 69.4% (95% CI 55.5–81.0) of the lynxes, and all badgers that were positive and had the *Trichinella* species identified. The *Trichinella* species that were detected are shown by county and by year in Figure 2, Table 1 and Table S2. The isolates of 2007–2010 were deposited in International *Trichinella* Reference Centre [11].

*Trichinella britovi* was the most common *Trichinella* species found in all the investigated host species. It was found in animals from all counties (Figure 2, Table S2), in 0.7% (95% CI 0.6–0.8) of wild boars, 7.2 % (95% CI 5.1–10.0) of brown bears, 45.6% (95% CI 35.5–55.9) of lynxes, and 40.0% (95% CI 7.4–81.8) of badgers (Table 2). *Trichinella britovi* infections were found in 31 brown bears: in five hunted in Ida-Virumaa in 2007, 2010 and 2011; in four hunted in Harjumaa in 2010, 2011, 2012, and 2013; in four hunted in Järvamaa in 2008 and 2013; in four hunted in Jõgevamaa in 2011 and 2013; in two hunted in Läänemaa in 2009 and 2010; in four hunted in Lääne-Virumaa in 2009, 2010 and 2013; in one hunted in Põlvamaa in 2012; in two hunted in Pärnumaa in 2008 and 2012; and in three hunted in Tartumaa in 2009, 2013 and 2014; and in two badgers, which had been hunted in Lääne-Virumaa and Viljandimaa in 2013—these are the first confirmed findings of this parasite species in these host species in Estonia (this study; [15]).

**Figure 2.** *Trichinella* species distribution in the tested animals that were positive and *Trichinella* species identification was successful, 2007–2014 (**A**–**H**, respectively), Estonia. *T. britovi*—ring, *T. nativa*—black dot, *T. pseudospiralis*—star; *T. spiralis*—cross, *T. britovi* and *T. nativa*—grey dot, *T. britovi* and *T. spiralis*—ring with cross.


speciesidentifiedinwildboars(*Susscrofa*),brownbears(*Ursusarctos*),Eurasianlynxes(*Lynxlynx*)andbadgers(*Melesmeles*)hunted

infection) or in mixed infection; d No data on larval burden for one wild boar and two lynxes; 95% CI: 95% confidence interval, Mid-P Exact; n pos: number of *Trichinella* positive animals;

lpg: number of *Trichinella* larvae per gram of muscle tissue.

The second most common *Trichinella* species was *T. nativa*, was found in all the investigated host species except badgers. Infected animals originated from 11 of the 15 counties; no findings were detected on the islands Hiiumaa and Saaremaa, and the southeastern counties Võrumaa and Valgamaa (Fig. 2, Table S2). *Trichinella nativa* was found in 0.1% (95% CI 0.0–0.1) of wild boars, 5.8 % (95% CI 3.9–8.4) of brown bears, and 20.0% (95% CI 12.7–29.2) of lynxes (Table 2).

*Trichinella pseudospiralis* was found in 2009 for the first time in wild boars in Estonia (Table 1; [11,16]). During the study period, this species was found in 0.02% of wild boars (Table 2); the prevalence was highest in Saaremaa, 0.2% (95% CI 0.1–0.5; Table S2).

The first *Trichinella spiralis* finding in a game animal in Estonia was identified in a lynx hunted in 2008 (shipped and tested in 2009), and further findings were detected in wild boars and lynxes hunted in 2009. The species was found in 0.03% (95% CI 0.0– 0.05) of wild boars and 4.4% (95% CI 1.4–10.4) of lynxes (Table 2). It was detected in nine counties: Harjumaa, Järvamaa, Läänemaa, Lääne-Virumaa, Põlvamaa, Pärnumaa, Raplamaa, Saaremaa, and Viljandimaa (Table S2).

#### **4. Discussion**

The high number of observations in this study add substantially to the knowledge on epidemiology of *Trichinella* spp. in Estonia and highlight that wildlife, including game animals, have a key role in it. *Trichinella* spp. are important zoonotic parasites in the country and the region [17], and it is crucial that the One Health approaches addressing them cover not only domestic animals and humans, but also wildlife.

It should be emphasized that hunted animals are always a convenience sample: hunting periods affect the age of the animals included in the sample, and the representativeness of a hunter-harvested sample is challenging to evaluate. Moreover, it should be noted that e.g., animals injured in traffic accidents or hunted illegally are not included in the official hunting statistics. This could explain the higher number of brown bears in our sample than in the hunting bags.

The sampling was done by hunters and veterinary inspectors, who were advised to sample from the predilection muscle groups, if these were available [10]. The sampling was not supervised by the authors, and possible variation in sample material may have affected the results to the direction of underestimation of the prevalence and in particular of the larval burden. The predilection muscle groups vary by host species [10], and for detailed comparisons, sampling the exact same muscle within each host species would be optimal.

The background information on the animals was provided by the hunters, and the authors had no means to validate these data. Misclassification of some animals to wrong age category or sex remain possible, and no data were provided for many animals (Table 1; Table S1).

The methodology we used is harmonized at international level and thus yields comparable results. The prevalence estimates reported in this study are generally in line with results from previous studies focusing on these host species in Estonia, which estimated the prevalence to be 0.3–1.0% in wild boars, 29.4% in brown bears, and 47.4–50% in lynxes [8,18]. The proportion of badgers that tested positive in this study was significantly higher (60.0%, three of five, *p* = 0.006) than an earlier estimate for badgers hunted in 1965–2000 (6.7%, six of 89) [19].

While the results of this study are not directly comparable with those from other countries due to different sampling and study designs, it is clear that *Trichinella* parasites thrive in the region. The prevalence in wild boars in this study was lower than that observed in Latvia [20], but higher than that in Poland [21]. The prevalence in brown bears and lynxes was higher than that in Finland [22], while the prevalence in lynxes was lower than that in Latvia [8,23]. The prevalence in badgers is considered low in several countries [24], however the proportion of positives in this study was substantial, in line with what has been observed in Latvia, and higher than that in Finland [22,23].

The results of this study confirm that *T. britovi* has been winning host-terrain, while *T. nativa* is well-established in whole mainland Estonia. It is noteworthy that *T. pseudospiralis* was found in animals from the southwestern part of the country. Several studies have described an increase of *T. pseudospiralis* findings in wild boar samples in Europe [25]. One possible vector of *T. pseudospiralis* are predatory birds, including migratory birds [25]. In Estonia, the findings of *T. pseudospiralis* have been made in animals killed near the sea or wetlands areas, which are good nesting sites for birds. Further research focusing on the potential host species living in these specific environments could provide insight into the role of birds in the epidemiology of *T. pseudospiralis*. Another noteworthy finding was *T. spiralis* from a lynx killed in 2009 in Järvamaa county, approximately 30 km from where a human trichinellosis outbreak was documented ten years earlier, and where *T. spiralis* was found in a domestic pig during the investigation [18]. *Trichinella spiralis* could be infecting wildlife in Estonia similarly as described by Oksanen and others [22], as spillover from the domestic cycle. That we did not find freeze-sensitive *T. spiralis* in the main reservoir animal host species in our previous study [6] might be explained by two consecutive colder years before 2011/2012 [26].

In contrast to our previous epidemiological study in the reservoir hosts raccoon dogs and red foxes, where no obvious geographical differences in *Trichinella* prevalence were seen [6], geographical differences in the prevalence were observed in wild boars and lynxes in this study. Interestingly, we previously demonstrated a higher seroprevalence in wild boars in the southwestern part of the country [7], and the results of this study confirm a higher infection prevalence in western and southern counties. The geographical variation may be due to several factors, including climate, temperature, and snow cover [16,20].

The results of this study exemplify that wild boars can serve as an indicator for *Trichinella* spp. monitoring, being annually hunted in high numbers and routinely tested for *Trichinella*. Wild boars have been popular game in Estonia after their population rapidly increased since the second half of the 1990s, supported by relatively mild winters and supplementary feeding [27–30]. Importantly, the results of this study reflect the situation before the African swine fever (ASF) outbreak in Estonia, which started in September 2014, and will thus serve as baseline data for future studies that could evaluate how the ASF-related changes affected the wild boar population and the parasites these animals can host.

The results of this study highlight that testing wildlife hunted for human consumption for *Trichinella* remains important, and that there is room for improvement in the proportion of hunted animals tested. Wildlife are important for epidemiology of *Trichinella* spp. in Estonia, and hunting wild game for human consumption provides a potential transmission route to humans.

#### **5. Conclusions**

In Estonia, *Trichinella* infections are common in wildlife, including in game animals hunted for human consumption. High infection pressure was evident in sylvatic cycles, and the risk for spillover to and from domestic cycles and transmission to humans remain relevant.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/2076-261 5/11/1/183/s1: Table S1: Primary testing included in the study: number of wild boars (*Sus scrofa*), brown bears (*Ursus arctos*), Eurasian lynxes (*Lynx lynx*) and badgers (*Meles meles*) tested for *Trichinella* in Estonia, 2007–2014, by sex, age category, county, and by year. Table S2: Prevalence of *Trichinella* infection in wild boars (*Sus scrofa*), brown bears (*Ursus arctos*), Eurasian lynxes (*Lynx lynx*) and badgers (*Meles meles*) hunted in Estonia, 2007–2014, by county. Univariable odds to test positive in comparison to the reference county (Harjumaa), and larval burden data and the *Trichinella* species identified are summarized.

**Author Contributions:** Conceptualization, A.K., B.L. and P.J.; methodology, A.K., B.L. and P.J.; investigation, A.K., L.H. and A.V.; data curation, A.K., B.L. and P.J.; writing—original draft preparation, A.K., B.L. and P.J.; writing—review and editing, B.L and P.J.; supervision, B.L. and P.J. All authors

contributed to the writing process. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data is contained within the article or supplementary material.

**Acknowledgments:** We thank the hunters and veterinary inspectors for collecting the samples and providing the background information about the animals. We thank Ingrid Ott and Moonika Musting from Tartu VFL, Ülla Rajamets and Kristel Lilles from Tallinn VFL, Eda Laas and Merle Reiman from Rakvere VFL, Mai Truutsi from Saaremaa VFL for performing laboratory analyses and collecting data, and Katrin Lõhmus for helping with the laboratory database. We thank European Union Reference Laboratory for Parasites for the species identification of *Trichinella* larvae collected in 2007–2010, and Edoardo Pozio and Gianluca Marucci for the excellent training in the multiplex PCR method.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Distribution and Host Selection of Tropical Rat Mite,** *Ornithonyssus bacoti***, in Yunnan Province of Southwest China**

**Peng-Wu Yin 1,2, Xian-Guo Guo 1,2,\*, Dao-Chao Jin 1, Rong Fan 2, Cheng-Fu Zhao 2, Zhi-Wei Zhang 2, Xiao-Bin Huang <sup>2</sup> and Ke-Yu Mao <sup>2</sup>**


**Simple Summary:** The tropical rat mite (*Ornithonyssus bacoti*) is a transmission vector of rickettsia pox and a potential vector of hemorrhagic fever with renal syndrome (HFRS). This article reports the distribution and host selection of *O. bacoti* in Yunnan Province of Southwest China. The original data came from the investigations in 39 counties of Yunnan. The prevalence (*PM*), mean abundance (*MA*) and mean intensity (*MI*) were calculated to reflect the infestations of the dominant rat hosts with *O. bacoti* mites. The patchiness index and Taylor's power law were used to measure the spatial distribution of the mites. A total of 4121 *O. bacoti* mites were identified from 15 species of small mammal hosts in 27 of the 39 investigated counties, and 99.20% of them (4088/4121) were found on rodents. The majority of total *O. bacoti* mites was found in the flatland landscape (91.28%) and indoor habitat (73.48%). Moreover, 51.78% and 40.09% of *O. bacoti* mites were identified from *Rattus tanezumi* and *R. norvegicus*, the two synanthropic rat species. The mites had some host-specificity, with a preference to two dominant hosts (*R. tanezumi* and *R. norvegicus*), and they were of aggregated distribution on *R. tanezumi*.

**Abstract:** (1) Background: As a species of gamasid mite, the tropical rat mite (*Ornithonyssus bacoti*) is a common ectoparasite on rodents and some other small mammals. Besides stinging humans to cause dermatitis, *O. bacoti* can be a vector of rickettsia pox and a potential vector of hemorrhagic fever with renal syndrome (HFRS). (2) Objective: The present study was conducted to understand the host selection of *O. bacoti* on different animal hosts and the distribution in different environmental gradients in Yunnan Province of Southwest China. (3) Methods: The original data came from the investigations in 39 counties of Yunnan, between 1990 and 2015. The animal hosts, rodents and some other small mammals were mainly trapped with mouse traps. The *O. bacoti* mites on the body surface of animal hosts were collected and identified in a conventional way. The constituent ratio (*Cr*), prevalence (*PM*), mean abundance (*MA*) and mean intensity (*MI*) were used to reflect infestations of animal hosts with *O. bacoti* mites. The patchiness index and Taylor's power law were used to measure the spatial distribution pattern of *O. bacoti* mites on their hosts. (4) Results: A total of 4121 tropical rat mites (*O. bacoti*) were identified from 15 species and 14,739 individuals of hosts, and 99.20% of them were found on rodents. More than half of *O. bacoti* mites (51.78%) were identified from the Asian house rat (*Rattus tanezumi*), and 40.09% of the mites from the Norway rat (*R. norvegicus*) (*p* < 0.05). The infestations of *R. tanezumi* (*PM* = 7.61%, *MA* = 0.40 and *MI* = 5.31) and *R. norvegicus* (*PM* = 10.98, *MA* = 1.14 and *MI* = 10.39) with *O. bacoti* mites were significantly higher than those of other host species (*p* < 0.05). The infestations of two dominant rat hosts (*R. tanezumi* and *R. norvegicus*) with *O. bacoti* mites varied in different environmental gradients (latitudes, longitudes, altitudes, landscapes and habitats) and on different sexes and ages of the hosts. The prevalence of juvenile *R. norvegicus* rats with *O. bacoti* mites (*PM* = 12.90%) was significantly higher than that of adult rats (*PM* = 9.62%) (*p* < 0.05). The prevalence (*PM* = 38.46%) and mean abundance (*MA* = 2.28 mites/host) of *R. tanezumi*

**Citation:** Yin, P.-W.; Guo, X.-G.; Jin, D.-C.; Fan, R.; Zhao, C.-F.; Zhang, Z.-W.; Huang, X.-B.; Mao, K.-Y. Distribution and Host Selection of Tropical Rat Mite, *Ornithonyssus bacoti*, in Yunnan Province of Southwest China. *Animals* **2021**, *11*, 110. https://doi.org/10.3390/ ani11010110

Received: 20 October 2020 Accepted: 30 December 2020 Published: 7 January 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

rats with *O. bacoti* mites in the high latitude were higher than those in the low latitudes (*p* < 0.05). The majority of the total collected 4121 *O. bacoti* mites was found in the flatland landscape (91.28%) and indoor habitat (73.48%) (*p* < 0.05). The *PM* (10.66%) and *MA* (0.49 mites/host) of *R. tanezumi* rats with *O. bacoti* mites were significantly higher in the indoor habitat than in the outdoor habitat (*p* < 0.05). The tropical rat mites showed an aggregated distribution pattern on their first dominant host, *R. tanezumi*. **Conclusion:** The tropical rat mite (*O. bacoti*) is a widely distributed species of gamasid mite in Yunnan Province, Southwest China, and its dominant hosts are two synanthropic species of rats, *R. tanezumi* and *R. norvegicus*. It is mainly distributed in the flatland landscape and indoor habitat. It has some host-specificity, with a preference to rodents, especially *R. tanezumi* and *R. norvegicus*. The *O. bacoti* mites are of aggregated distribution on *R. tanezumi* rats.

**Keywords:** Acari; gamasid mite; *Ornithonyssus bacoti*; distribution; host selection; Yunnan; China

#### **1. Introduction**

As a worldwide species of gamasid mite, the tropical rat mite (*Ornithonyssus bacoti*) is widely distributed in nearly all parts of the world, except the Arctic and Antarctic regions [1,2]. It is a common ectoparasite on rodents (rats, mice and voles) and some other small mammals (e.g., insectivores and tree shrews), frequently occurring on the body surface and in the nests of its hosts. *Ornithonyssus bacoti* is the most important species of gamasid mites associated with medicine, and nearly all its stages (larvae, protonymphs, deutonymphs and adults) in the life cycle can invade and sting animal hosts for blood meal [3–5]. The dermatitis caused by the stinging of *O. bacoti* mites is frequently reported throughout the world [5–16]. Besides directly stinging humans, *O. bacoti* is associated with the transmission of some zoonoses. Together with another species of gamasid mite (*Allodermanyssus sanguineus*), the mite *O. bacoti* is the confirmed vector of rickettsia pox, caused by the pathogen *Rickettsia akari* [1,17,18]. Rickettsia pox is a zoonosis associated with rodents, and it is mainly prevalent in the United States, Canada, Russia, Ukraine, India and Egypt, etc. [17,19–21]. Besides being the intermediate host of the animal parasite *Litomosoides carinii*, the mite *O. bacoti* has been proved to be the potential vector of hemorrhagic fever with renal syndrome (HFRS) caused by hantavirus [22–25].

To date, there have been a lot of studies on the tropical rat mite, *O. bacoti*. Early on in the 1940s, some scholars began to feed *O. bacoti* in the laboratory and designed some special devices suitable for *O. bacoti* and some other sucking mites [26,27]. Many previous publications on *O. bacoti* and some other species of gamasid mites were about the mites' ultrastructure [1,8,28–30], chromosome karyotype, gene sequencing, phylogeny and control [1,29,31–34], but only few studies were about the mite ecology. To date, there have been no systematically ecological studies on *O. bacoti* in Yunnan Province of Southwest China. Between 1990 and 2015, our research group made a long-term field investigation and accumulated abundant original data on gamasid mites in Yunnan. To take advantage of the previous investigation, the present paper analyzed the distribution and host selection of *O. bacoti* in Yunnan for the first time, which is an attempt to get more knowledge about *O. bacoti* and to enrich the ecological information of the mite.

#### **2. Materials and Methods**

#### *2.1. Collection and Identification of Ornithonyssus bacoti and Its Animal Hosts*

The original data came from a long-term field investigation in 39 counties of Yunnan Province in Southwest China, between 1990 and 2015, and the investigated 39 counties were shown in Figure 1, in Results. A stratified sampling investigation was made in different geographical localities, latitudes, longitudes, altitude, landscapes and habitats. To capture animal hosts, mouse traps were placed in the indoor and outdoor habitats of each investigation site, in the evening, and then checked the following morning. The indoor habitats covered houses, barns, stables, etc., and the outdoor habitats covered paddy fields, cornfields, bush habitats, woodlands, etc. [35]. Each trapped host was euthanized through anesthesia with ether (cotton balls soaked with ether), within a closed container. Under the anesthesia, the gamasid mites on the body surface of each host were all collected and then preserved in Eppendorf tubes with 70% of ethanol. After the mite collection, each animal host was identified into species according to its morphological features [4,36,37]. Through the dehydration and clarification, the collected gamasid mites were mounted onto glass slides with Hoyer's medium and they were then identified into species under microscopes [38,39]. Based on the identification of gamasid mites, the tropical rat mite (*O. bacoti*) was chosen as the target of the present study. In the animal euthanasia, most rodent pests in agriculture and forestry were anaesthetized to death because the local government encourages people to kill and eradicate them. Some non-pest small mammals (e.g., weasels, moles and some squirrels) were anaesthetized only for two to five minutes, according to their body size, and they were finally released to the wild when they woke up. The capturing of rodents and other small mammals was officially permitted by the local authority of wildlife service in Yunnan Province, China. The use of animals (including animal euthanasia) for research was officially approved by the Animals' Ethics Committee of Dali University, and the permitted number was DLDXLL2020-1104. The specimens of voucher mites and representative animal hosts are deposited in the specimen repository of the Institute of Pathogens and Vectors, Dali University, Yunnan, China.

#### *2.2. Infestation Statistics*

In a conventional way, the constituent ratio (*Cr*), prevalence *(PM*), mean abundance (*MA*) and mean intensity (*MI*) were calculated to reflect the infestations of dominant hosts with tropical rat mites (*O. bacoti*) [4,40–42]. In the present study, *Cr* represents the percentage of *O. bacoti* mites, *PM* the percentage of infested hosts by the mites, *MA* the number of the mites on each examined host and *MI* the number of the mites on each infested host.

#### *2.3. Analysis on Host Selection and Distribution*

The infestations of dominant animal hosts with *O. bacoti* mites were compared on different sexes and ages of the hosts to reflect the host selection of the mites. The infestations were compared in different latitudes and longitudes, to reflect the mite's horizontal distribution, and compared in different altitudes, to reflect the mite's vertical distribution. The latitudes and longitudes were divided into three gradients, and the altitudes were divided into four gradients. The three latitude gradients include <24◦ N, 24◦ N–26◦ N and >26◦ N, and the four longitude gradients are <100◦ E, 100◦ E–102◦ E, 102◦ E–104◦ E and >104◦ E. The four altitude gradients are <1000 m, 1000–2000 m, 2001–3000 m and >3000 m.

#### *2.4. Analysis of Spatial Distribution Pattern*

The patchiness index (*m\*/m*) and Taylor's power were used to measure the spatial distribution pattern of tropical rat mites (*O. bacoti*) among different individuals of its dominant hosts [43,44]. The formulae of patchiness index and Taylor's power are listed as follows.

$$m\*/m = \frac{m + \left(\frac{\sigma^2}{m} - 1\right)}{m}; \lg \sigma^2 = \lg a + b \lg m \tag{1}$$

In the above formulae, *m\*/m* = patchiness index, *m* = mean of *O. bacoti* mites on its dominant hosts and *σ<sup>2</sup>* = variance of the mites; lg *a* = intercept on the Y-axis, and *b* = the slope or regression coefficient. When *m\*/m* > 1, *a* > 1 and *b* > 1 or *a* > 1, *b* = 1, the spatial distribution pattern is determined as the aggregated distribution; when *m\*/m* = 1, *a* = 1 and *b* = 1, the random distribution; when *m\*/m* < 1, *a* < 1, *b* < 1, the uniform (or even) distribution [45,46].

#### *2.5. Analysis on Interspecific Association*

Based on a contingency table (see Table 10 in Results), the association coefficient (*V*) was used to measure the interspecific association between the tropical rat mite (*O. bacoti*) and some other related species of gamasid mites on the dominant animal hosts. In the contingency table, *O. bacoti* was defined as "species X", while the other related mite species was defined as "species Y". The association coefficient (*V*) is as follows.

$$V = \frac{ad - bc}{\sqrt{(a+b)(c+d)(a+c)(b+d)}}\tag{2}$$

In the above formula, *V* = the association coefficient between species X and Y; *a* = the host individuals on which species X and Y simultaneously occur; *b* = the host individuals on which species Y occurs, but species X does not occur; *c* = the host individuals on which species X occurs, but species Y does not occur; and *d* = the host individuals on which both species X and Y do not occur. When *V* > 0 and *p* < 0.05, the interspecific relationship between species X and Y is determined as the positive association; when *V* < 0 and *p* < 0.05, we have the negative association; P is the significance probability in Chi-square test (*χ2*).

#### *2.6. Significance Test*

Chi-square test (*χ2*) was used to test the significance of *Cr*, *PM* and *V*. Nonparametric test was used to test the significance of *MA* and *MI*. All the statistical analyses were performed with R software version 3.5.3.

#### **3. Results**

#### *3.1. Abundance of Ornithonyssus bacoti*

A total of 139,111 gamasid mites were collected from 74 species and 17,638 individuals of animal hosts, rodents and some other small mammals, in 39 counties of Yunnan. Of 139,111 collected gamasid mites, 137,210 of them were identified as 156 species and 39 genera in 13 families, and the remaining 1901 mites remained unidentified because of blemished, obscure and damaged structures or suspected new species. A total of 4121 tropical rat mites (*O. bacoti*) were identified from 15 species and 14,739 individuals of hosts, and they accounted for only 2.96% (4121/139,111) of the total mites. The identified 4121 *O. bacoti* mites were distributed in 27 counties (Figure 1).

**Figure 1.** The 39 investigated counties and 27 counties with tropical rat mites (*Ornithonyssus bacoti*) collected (marked with "\*") in Yunnan Province, Southwest China (1990–2015).

#### *3.2. Host Selection of Ornithonyssus bacoti*

The identified 4121 *O. bacoti* mites came from 15 species, 8 genera and 4 families of animal hosts in 3 orders, Rodentia, Soricomorpha and Scandetia. On the order level of animal hosts, 99.20% of *O. bacoti* mites (4088/4121) were found on the order Rodentia, which was significantly higher than that on Soricomorpha and Scandetia (*p* < 0.05). The percentages of *O. bacoti* mites found on the family Muridae (4088/4121 = 99.20%) and the genus *Rattus* (3953/4121 = 95.92%) were the highest of four host families and eight genera (*p* < 0.05). On the species level of hosts, the majority of *O. bacoti* mites was identified from two dominant rat hosts, with 51.78% of the mites on the Asian house rat (*Rattus tanezumi*) and 40.09% of the mites on the Norway rat (*R. norvegicus*) (*p* < 0.05). Of the 15 species of hosts, the infestations of the dominant rat hosts (*R. tanezumi* and *R. norvegicus*) with *O. bacoti* mites were significantly higher than those of other 13 species of hosts (*p* < 0.05) (Table 1). Positive linear correlations existed among *PM*, *MA* and *MI*, with *r* = 0.685 between *MA* and *MI*, *r* = 0.646 between *MA* and *PM* and *r* = 0.332 between *MI* and *PM* (*p* < 0.05) (Figure 2).

**Table 1.** Infestations of two dominant rat hosts (*Rattus tanezumi* and *R. norvegicus*) with tropical rat mites (*Ornithonyssus bacoti*) in Yunnan Province, Southwest China (1990–2015).


\*Annotation: The other 13 species of animal hosts are *Rattus nitidus*, *Mus musculus*, *R. brunneusculus (R. sladeni)*, *Suncus murinus*, *Crocidura attenuata*, *Apodemus draco*, *Niviventer andersoni*, *Tupaia belangeri*, *A. chevrieri*, *N. confucianus*, *Eothenomys miletus*, *M. caroli and A. peninsulae*.

**Figure 2.** The linear regression between *PM*, *MA* and *MI* of 15 host species with *O. bacoti* mites in Yunnan Province, Southwest China (1990–2015).

Different sexes and ages of two dominant rat hosts (*R. tanezumi* and *R. norvegicus*) showed some differences in the infestations with *O. bacoti* mites. The infestations of male rats (*R. tanezumi* and *R. norvegicus*) with *O. bacoti* mites were slightly higher than those of female rats, but the differences were of no statistical significance (*p* > 0.05) (Tables 2 and 3). The prevalence of juvenile *R. norvegicus* rats with the mites (*PM* = 12.90%) was significantly higher than that of adult rats (*PM* = 9.62%) (*p* < 0.05). The mean abundance (*MA* mites/host) and mean intensity (*MI* mites/host) of juvenile *R. norvegicus* rats with the mites were slightly higher than those of adult rats, but the differences were of no statistical significance (*p* > 0.05). The infestations of juvenile *R. tanezumi* rats with the mites (*PM*, *MA* and *MI*) were also slightly higher than those of adult rats, but the differences were of no statistical significance (*p* > 0.05) (Tables 2 and 3).

**Table 2.** Infestations of different sexes and ages of *R. tanezumi* rats with *O. bacoti* mites in Yunnan Province, Southwest China (1990–2015).


\* Annotation: The animal hosts without records of sexes and ages were not included in the above table.


**Table 3.** Infestations of different sexes and ages of *R. norvegicus* rats with *O. bacoti* mites in Yunnan Province, Southwest China (1990–2015).

\* Annotation: The animal hosts without records of sexes and ages were not included in the above table.

#### *3.3. Horizontal Distribution of Ornithonyssus bacoti*

The infestations of two dominant rat hosts (*R. tanezumi* and *R. norvegicus*) with *O. bacoti* mites showed some differences in different latitudes and longitudes (horizontal distribution). The *PM* (38.46%) and *MA* (2.28 mites/host) of *R. tanezumi* with *O. bacoti* mites, together with *MA* (2.04 mites/host) of *R. norvegicus* with the mites, were higher in the high latitude (>26◦ N) than in other latitudes (*p* < 0.05) (Tables 4 and 5). The *MA* (0.63 mites/host) of *R. tanezumi* rats with the mites was higher in the longitude <100◦ E than in other three longitudes (*p* < 0.05), and the *PM* (16.81%) and *MA* (2.10 mites/host) of *R. norvegicus* with the mites were higher in the longitude 100◦ E–102◦ E than in other three longitudes (*p* < 0.05) (Tables 4 and 5).

**Table 4.** Infestations of *R. tanezumi* rats with *O. bacoti* mites in different latitudes and longitudes of Yunnan Province, Southwest China (1990–2015).


\* Annotation: The animal hosts without records of latitudes and longitudes were not included in the above table.

**Table 5.** Infestations of *R. norvegicus* rats with *O. bacoti* mites in different latitudes and longitudes of Yunnan Province, Southwest China (1990–2015).


\* Annotation: The animal hosts without records of latitudes and longitudes were not included in the above table.

#### *3.4. Vertical Distribution of Ornithonyssus bacoti*

The infestations of two dominant rat hosts (*R. tanezumi* and *R. norvegicus*) with *O. bacoti* mites showed some differences in different altitudes (vertical distribution). The *PM* (27.27%) and *MA* (0.82 mites/host) of *R. tanezumi* rats with *O. bacoti* mites were highest above 3000 m, but *MI* (7.62 mites/host) was highest below 1000 m (*p* < 0.05). The *PM* (13.40%), *MA* (0.77 mites/host) and *MI* (5.74 mites/host) of *R. norvegicus* rats with the mites were highest at 1000–2000 m (*p* < 0.05) (Table 6).

**Table 6.** Infestations of *R. tanezumi* and *R. norvegicus* rats with *O. bacoti* mites in different altitudes of Yunnan Province, Southwest China (1990–2015).


\* Annotation: The animal hosts without records of altitudes were not included in the above table.

#### *3.5. Landscape and Habitat Distribution of Ornithonyssus bacoti*

The majority of total collected 4121 *O. bacoti* mites was found in the flatland landscape (1894/2075 = 91.28%) and indoor habitat (3028/4121 = 73.48%) (*p* < 0.05). The infestations of two dominant rat hosts (*R. tanezumi* and *R. norvegicus*) with *O. bacoti* mites showed some differences in two kinds of landscapes (mountainous and flatland landscapes), but

the differences were of no statistical significance (*p* > 0.05) (Table 7). The infestations of *R. tanezumi* and *R. norvegicus* with the mites also showed some differences in two kinds of habitats. The *PM* (10.66%) and *MA* (0.49 mites/host) of *R. tanezumi* rats with the mites were significantly higher in the indoor habitat than in the outdoor habitat (*p* < 0.05). The *PM* (11.50%), *MA* (1.30 mites/host) and *MI* (11.31 mites/host) of *R. norvegicus* rats with the mites were higher in the indoor habitat than in the outdoor habitat, but the differences were of no statistical significance (*p* > 0.05) (Table 8).

**Table 7.** Infestations of *R. tanezumi* and *R. norvegicus* rats with *O. bacoti* mites in different landscapes of Yunnan Province, Southwest China (1990–2015).


\* Annotation: The animal hosts without records of landscapes were not included in the above table.

**Table 8.** Infestations of *R. tanezumi* and *R. norvegicus* rats with *O. bacoti* mites in different habitats of Yunnan Province, Southwest China (1990–2015).


\* Annotation: The animal hosts without records of habitats were not included in the above table.

#### *3.6. Spatial Distribution Pattern and Interspecific Association*

A total of 5285 Asian house rats (*R. tanezumi*), the first dominant host of the tropical rat mites (*O. bacoti*), were captured in 35 of 39 investigated counties, but there were only 24 counties where *R. tanezumi* harbored *O. bacoti* mites. To establish a linear regression equation based on Taylor's power law, the 24 counties were recombined as four "sample units", according to their adjacent locations, and then the mean (*m*) and variance (*σ2*) of *O. bacoti* mites on *R. tanezumi* rats in each sample unit were calculated (Table 9). According to the calculated *m* and *σ2*, a linear regression equation was established as lg *σ<sup>2</sup>* = lg 39.30 + 1.42 lg *m*, where both *a* (39.30) and *b* (1.42) were beyond 1 (*a* > 1, *b* > 1), the border value for determining the aggregated distribution. The calculated patchiness

index (*m\*/m*) in each sample unit was also higher than 1 (*m\*/m* > 1), the border value for determining the aggregated distribution (Table 9). On the body surface of *R. tanezumi* rats, there were a lot of *L. nuttalli* mites (the other species of gamasid mites) that co-occurred with *O. bacoti* mites, and therefore the interspecific association between *O. bacoti* and *L. nuttalli* was studied. The result showed that there was a slight negative association between *O. bacoti* and *L. nuttalli* (*V* = −0.0794, *V* < 0, *p* < 0.05) (Table 10).

**Table 9.** The mean (*m*), variance (*σ2*) and patchiness index (*m\*/m*) of *O. bacoti* mites on *R. tanezumi* rats in each recombined sample unit of Yunnan, Southwest China (1990–2015).


Annotation: Each sample unit represents the following counties: 1 = Njian + Dali + Binchuan +Yangbi + Xiangyun + Weishan + Jianchuan + Heqing; 2 = Lushui + Fugong + Weixi + Gongshan + Lijiang; 3 = Gengma + Lianghe + Longchuan + Yingjiang + Longyang + Ruili + Yongde + Cangyuan; 4 = Luliang + Fuyuan + Maguan + Suijiang + Menghai + Yuanjiang + Simao + Mengzi + Jinghong + Ninger + Jinping + Hekou + Qiubei + Wenshan.

> **Table 10.** The contingency table for measuring the interspecific association between *O. bacoti* mites and *L. nuttalli* mites on the body surface of *R. tanezumi rats* in Yunnan, Southwest China (1990–2015).


#### **4. Discussion**

In laboratories, the tropical rat mite (*O. bacoti*) is often found on the experimental rats and mice, and it does a great harm to experimental animals [6,47,48]. Therefore, it is important to make a systematic study on *O. bacoti*. Some previous ecological studies of gamasid mites mainly focused on some local species surveys, faunal studies and community investigations [4,36,49]. Although some local investigations on the fauna and community of gamasid mites included the constituent ratio of *O. bacoti*, there were few independent and systematic studies of *O. bacoti* [16,38,50]. The present study systematically analyzed the distribution and host selection of *O. bacoti* in Yunnan Province of Southwest China for the first time. The original data came from a long-term investigation between 1990 and 2015, and the investigated 39 counties covered the different localities of Yunnan Province, Southwest China. The tropical rat mite (*O. bacoti*) was found in 27 of 39 investigated counties (Figure 1), and it suggests that *O. bacoti* is a widely distributed species of gamasid mites in Yunnan.

The present study showed that 99.20% of tropical rat mites (*O. bacoti*) were found on rodents (the order Rodentia), even though three orders of hosts (Rodentia, Soricomorpha and Scandetia) harbored the mites. Although *O. bacoti* mites occurred on different categories of hosts (15 species, 8 genera, 4 families and 3 orders), most of them were identified from two dominant rat species, the Asian house rat (*R. tanezumi*) and the Norway rat or brown rat (*R. norvegicus*). The infestations of *R. tanezumi* and *R. norvegicus* with *O. bacoti* mites were significantly higher than those of other 13 host species. The results suggest that *O. bacoti* has some host-specificity and it has a preference to *R. tanezumi* and *R. norvegicus* in Yunnan. The higher prevalence (*PM*) of juvenile *R. norvegicus* rats with *O. bacoti* mites than that of adult rats (Table 2) may imply the preference of the mites to juvenile hosts. Rodents are closely

related to human beings, and they are the infection source and reservoir hosts of many zoonotic diseases [18,51]. The rodent-preference of *O. bacoti* would increase the potential risk of the mite's attacking humans and spreading some zoonoses. The Asian house rat (*R. tanezumi*) and the Norway rat (*R. norvegicus*) are two major species of rodents associated with human settlements in Yunnan Province and some other places of China [52,53]. *Rattus tanezumi* (often called *R. flavipectus* in China) is widely distributed in the vast areas south of the Yangtze River, in Southern China. It is a very common rodent species in residential areas (especially the indoor habitats) in Central and Southern Yunnan [54,55]. *Rattus norvegicus* is widely distributed in the whole China, and it is also a very common rodent species in residential areas (especially the indoor habitats) in Yunnan, often co-occurring in the same areas with *R. tanezumi* [35,56–58]. The previous studies revealed that the main hosts of *O. bacoti* included some synanthropic rats and mice with humans (especially *R. norvegicus*) and experimental rats and mice [3,7,18]. The most commonly used laboratory rat is descended from *R. norvegicus*, which retains many biological characteristics of its ancestor *R. norvegicus* [59,60]. The frequent occurrence of *O. bacoti* on *R. tanezumi* and *R. norvegicus* would highly increase the risk of the mites' attacking humans and spreading some zoonoses.

*Rattus tanezumi* and *R. norvegicus* were two dominant rat hosts of *O. bacoti* mites, and therefore the present paper analyzed the infestations of the two host species with the mites in different horizontal gradients (latitudes and longitudes) and vertical gradients (altitudes). The results showed that the infestations of the rats (*R. tanezumi* and *R. norvegicus*) with *O. bacoti* mites showed some differences in different horizontal and vertical gradients. Some infestation indices (*PM* and *MA*) were higher in the high latitude (>26◦ N) and low longitudes (<100◦ E and 100◦ E–102◦ E) than in other latitudes and longitudes (Tables 4 and 5). The *PM* and *MA* of *R. tanezumi* rats with the mites were highest above 3000 m, but *MI* was highest below 1000 m. The *PM*, *MA* and *MI* of *R. norvegicus* rats with the mites were highest at 1000–2000 m (Table 6). The results indicated an unstable fluctuation in different vertical gradients. The climates in Yunnan province greatly vary from region to region because of complex topography and altitude gradients. Even within the same latitude or longitude gradient zone, the climate at a mountainous site with higher altitude may be very different from that at a flatland site with lower altitude [61–63]. The different infestations of the rats (*R. tanezumi* and *R. norvegicus*) with *O. bacoti* mites in different horizontal and vertical gradients may be related to different climates (temperature, humidity and rainfall, etc.) in different geographical localities. However, it is difficult to explain the unstable fluctuation of the mite infestations in different horizontal and vertical gradients, and more research studies still remain to be conducted.

Located in the southwest of China, Yunnan is a mountainous province where mountainous landscapes with higher altitude and lower temperature account for 84% of the whole territory, and flatland landscapes with lower altitude and higher temperature are often embedded in mountainous landscapes [63,64]. Although the flatland landscape only takes a small portion of the whole territory, the majority of *O. bacoti* mites (91.28%) was found in the flatland landscape, and this suggests that *O. bacoti* is mainly distributed in the flatland landscape. The infestations of the rats (*R. tanezumi* and *R. norvegicus*) with the mites showed some differences in mountainous and flatland landscapes, but the differences were of no statistical significance (Table 7). In habitat distribution, 73.48% of *O. bacoti* mites were collected in the indoor habitat. The *PM* and *MA* of *R. tanezumi* rats with the mites were significantly higher in the indoor habitat than in the outdoor habitat (Table 8), indicating the preference of the mites for the indoor habitat. The previous study showed that the optimum temperature for the development of *O. bacoti* was about 25 ◦C ± 5 ◦C, and higher than 30 ◦C or lower than 20 ◦C was not suitable for the mites' development [65]. The outdoor habitat in the present study involved a series of different sub-habitats, or microhabitats, such as cultivated farmlands (e.g., paddy fields and cornfields) and uncultivated bush areas and woodlands; the micro-climates in the outdoor habitat are often unstable. In comparison with the outdoor habitat, the indoor habitat is a relatively closed

environment with a relatively stable and warm temperature and low humidity [65,66]. The stable and warm micro-climate with relatively low humidity in the indoor habitat may be more suitable to the growth and reproduction of *O. bacoti* mites. The frequent occurrence of *O. bacoti* in the indoor habitat would highly increase the risk of the mites' invading and stinging humans. When rats and mice are not available for *O. bacoti* mites to suck the blood of, the mites in the indoor habitat may quickly move onto humans for the blood meal and then expand their range of activity [1,67,68].

The measurement of spatial distribution pattern of a certain population is one of important issues in arthropod ecology [69,70]. There are usually three types of spatial distribution patterns: uniform (or even) distribution, random distribution and aggregated distribution [52,71–73]. There are a variety of statistical methods to measure the spatial distribution pattern of a certain population, and the patchiness index and Taylor's power law are two of them [43,44,74]. According to the statistics of the patchiness index and Taylor's power law, tropical rat mites (*O. bacoti*) were determined to be of aggregated distribution on *R. tanezumi*, the first dominant host. The aggregated distribution indicates that the mite infestation is not even among different hosts. Some hosts may harbor a large number of mites, forming a clump of mites on their body surface, while some other hosts may have no or very few mites on their body surface. The aggregated distribution pattern of *O. bacoti* in the present study is consistent with that of some other ectoparasites, such as chigger mites; this is a common phenomenon in many parasites [38,52,71]. The aggregated distribution may be beneficial to the survival, mating and defense of the parasites [4,38,72,73].

The analysis of the interspecific relationship between any two different species is also an important issue in animal ecology [75,76]. The association coefficient (*V*) used in the present study is a simple way to measure the interspecific relationship between any two species [77–80]. The negative value of the association coefficient (*V* = −0.0794) may imply that there is a slight negative association between *O. bacoti* and *L. nuttalli*, but the value of "*V* = −0.0794" was very close to "0", and more research is still needed.

#### **5. Conclusions**

The tropical rat mite (*O. bacoti*) is a widely distributed species of gamasid mite in Yunnan Province, Southwest China, and its dominant hosts are two synanthropic species of rats, *R. tanezumi* and *R. norvegicus*. It is mainly distributed in the flatland landscape and indoor habitat. It has some host-specificity, with a preference to rodents, especially *R. tanezumi* and *R. norvegicus*. The *O. bacoti* mites are of aggregated distribution on *R. tanezumi* rats.

**Author Contributions:** Conceptualization, funding acquisition, project administration, resources, supervision, validation, and writing—review & editing, X.-G.G.; supervision, D.-C.J.; data curation, formal analysis, software, visualization, and writing—original draft, P.-W.Y.; investigation, R.F., C.-F.Z., Z.-W.Z. and K.-Y.M.; methodology, X.-B.H. All authors have read and agreed to the published version of the manuscript.

**Funding:** The project was supported by the National Natural Science Foundation of China (Nos. 81960380 and 81672055), to Xian-Guo Guo, and the Innovation Team of Vector Biology, Dali University (No. ZKLX2019104).

**Data Availability Statement:** The experimental data used to support the findings of this study are available from the corresponding author request.

**Acknowledgments:** Up until now, more than 60 people have joined this study, including field investigation, collection of gamasid mites, specimen making and identification of the mites. Here we would like to express our sincere thanks to the following people, who have made special contributions to the field investigation and laboratory work: Qiao-Hua Wang, Yong Zhang, Cong-Hua Gao, Nan Zhao, Jian-Chang He, Guo-Li Li, Xue-Song He, Yun-Ji Zou, De-Cai Ouyang and Shuang-Lin Wang, some colleagues and college students.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


*Article*
