Distribution of Deer Keds (Diptera: Hippoboscidae) in Free-Living Cervids of the Tuscan-Emilian Apennines, Central Italy, and Establishment of the Allochthonous Ectoparasite Lipoptena fortisetosa

Simple Summary In recent years, the increased presence of wildlife in habitats close to urban settlements has raised concerns about the risk of pathogen transmission from wild animals to humans due to the spread of different parasites. For this reason, a survey aimed at describing the dispersal and parasitism level of two cervid ectoparasites was carried out in the northern Apennines, in central Italy. The presence of two hippoboscids, the autochthonous Lipoptena cervi and allochthonous L. fortisetosa, native to Eastern Asia and recently recorded in Italy, were assessed on their main host species (red deer, fallow deer, and roe deer), considering host sex and age. The alien species L. fortisetosa was found to be widespread in the study area, most likely competing with L. cervi. Moreover, red deer seemed to be the favored host of both flies, with differences in sex and age class preferences. This study demonstrated the importance of regularly monitoring the populations of these parasites, especially the invasive species, due to the risks to human health, as these insects are potential vectors of pathogens. Abstract Lipoptena fortisetosa and L. cervi are hematophagous ectoparasites belonging to the Hippoboscidae family and preferentially living on cervids. In recent years, they have received specific attention due to the great increase in the abundance of their host species, and to their medical and veterinary importance as possible vectors of pathogens harmful to humans and animals. The aim of this study was to investigate the parasitism level of both of these flies on their main hosts in Italy, which are red deer, fallow deer, and roe deer, and to highlight a possible preference for a species, sex, or age class among the hosts. Deer keds were collected by examining 326 cervids hunted in the Tuscan-Emilian Apennines. Outcomes showed that L. fortisetosa has greatly spread throughout the study area, where it competes with the autochthonous L. cervi. Moreover, red deer was the favored host species of both ectoparasites, while different preferences for host sex and age classes were observed in the two hippoboscids. The regular monitoring of deer ked populations, especially the allochthonous L. fortisetosa, which is continuously spreading in Europe, is recommended to expand the knowledge on these parasitic species that are potentially dangerous to public health.


Introduction
In recent years, a substantial expansion range of free-living ungulate species has occurred in many European countries [1], including Italy [2,3] and particularly the Tuscany region in the northern Apennines [4], leading to a consequent increase in the abundance Each skin sample was accompanied by a form containing detailed information such as animal species, sex, and age class. Age classes were the following: • Fawn (<1 year for all cervid species); • Subadult (between 1 and 4 years for fallow deer; 1 and 2 years for roe deer; 1 and 3 years for female red deer and 1 and 5 years for male red deer); • Adults (>4 years for fallow deer; >2 years for roe deer; >3 years for female red deer and >5 years for male red deer).
Fur samples were thawed and then visually examined for deer keds. All the parasites were manually removed with forceps and morphologically observed under a stereomicroscope (Leica/Wild MZ16, equipped with an L2 illuminator; Leica Microsystems, Wetzlar, Germany) for taxonomic identification using keys and previously described characters [6,9,39,40].
Subsequently, all the insects were separated by species and sex, counted, and stored in 70% ethanol or frozen at −20 °C, pending further analyses.
All cervid handling procedures followed the regional, national, and institutional guidelines.
Data were transcribed and reported into different software programs: QGIS 3. 16  Each skin sample was accompanied by a form containing detailed information such as animal species, sex, and age class. Age classes were the following: • Fawn (<1 year for all cervid species); • Subadult (between 1 and 4 years for fallow deer; 1 and 2 years for roe deer; 1 and 3 years for female red deer and 1 and 5 years for male red deer); • Adults (>4 years for fallow deer; >2 years for roe deer; >3 years for female red deer and >5 years for male red deer).
Fur samples were thawed and then visually examined for deer keds. All the parasites were manually removed with forceps and morphologically observed under a stereomicroscope (Leica/Wild MZ16, equipped with an L2 illuminator; Leica Microsystems, Wetzlar, Germany) for taxonomic identification using keys and previously described characters [6,9,39,40].
Subsequently, all the insects were separated by species and sex, counted, and stored in 70% ethanol or frozen at −20 • C, pending further analyses.
All cervid handling procedures followed the regional, national, and institutional guidelines. From here onward, any mention of parasites per host animal is always referred to as the sum of the keds collected from the two skin samples described above.

Parasitological Index
The infestation of deer keds on the three different host species was described using parasitological indices according to Margolis et al. [41]. All the descriptors were stratified by host and ectoparasite species. The distribution of Lipoptena spp. was evaluated by the parasitological index of density (average number of parasites per unit area (cm 2 ) of Animals 2021, 11, 2794 5 of 15 the host body), prevalence (percentage of infested deer); abundance (average number of parasites per host), mean intensity (average number of parasites per infested host), and minimum and maximum intensity (Imin-Imax). Moreover, the variance-to-mean ratio (variance of infestation divided by mean abundance) was calculated as the aggregation index, considering the distribution as overdispersed (or aggregated) if the value was >1, as demonstrated by Barbour and Pugliese [42].

Statistical Analyses
Data were statistically analyzed using R software [43]. Preliminary analysis highlighted aggregated parasite distribution; therefore, generalized linear models (in particular negative binomial regression) with the abundance of each parasite species as the dependent variable were built using the MASS package [44][45][46].
First, the chi-square test was used to compare parasite prevalence among the three host species.
Differences in parasite abundance among the three host species were evaluated using a univariable model to determine the primary host species; therefore, multivariable models were used to evaluate the influence of host-related variables on parasite abundance in the formerly determined host species.

Results
Out of the 326 examined cervids, 287 harbored deer keds (88.0%). The morphological analyses of the 23,074 collected flies revealed the presence of two hippoboscid species, identified as L. fortisetosa and L. cervi. Of the total insects, 18,441 were L. fortisetosa, and 4633 were L. cervi; even though the total highest number of insects was L. fortisetosa, some host animals were more infested with L. cervi. Of the total examined hosts, 127 cervids carried both the Lipoptena species, while 26 out of 107 tested roe deer, 45 out of 181 tested red deer, and two out of 38 tested fallow deer did not harbor any parasites. The data on the overall deer ked infestations in the three host species are given in Table 1. The number of males and females of the two ectoparasites, together with the values of the parasitological parameters for each of the three host species, are reported in Table 2.
The highest number of deer keds on a host subject was found in red deer, while the number of parasites obtained from roe deer and fallow deer was much lower. A maximum of 1,844 L. fortisetosa were collected from a host specimen, while a maximum of 398 flies of L. cervi were picked off a single red deer.
The chi-square test highlighted significant differences among the prevalence of both parasite species in the three hosts (p = 0.000 and p = 0.000, respectively for L. fortisetosa and L. cervi). In particular, fallow deer showed the highest prevalence for L. fortisetosa (94.8%), while it displayed the lowest prevalence value for L. cervi (21.5%). Although fallow deer was infested more often with L. fortisetosa than the other two hosts, the abundance of this parasite was highest for red deer, as confirmed by the univariable negative binomial regression (p = 0.000). Additionally, L. cervi abundance was higher in red deer than in the other two host species (p = 0.000). Notably, only 17 L. cervi were found on the 38 analyzed fallow deer (Table 2). The aggregation index was >1 for all the cervid species for both keds, meaning that the parasites were aggregated over the host populations, as illustrated in Figure 2. Further details on the epidemiological parameters stratified by the sex and age of the hosts are provided in Table 3. The multivariable negative binomial regression, taking into consideration sex and age, was constructed for red deer only since it was the primary host species for both parasites. The results are reported in Tables 4 and 5 for L. cervi and L. fortisetosa, respectively. Lipoptena cervi was significantly less abundant in females than in males and in fawns than in subadults, while L. fortisetosa was significantly less abundant in adults than in subadults. The interactions between sex and age classes were significant for both parasite species, as evident in Figure 3. The plots of the residuals in Figure 4 show a good residual pattern, with similar residual distributions across the levels of the predicted values.

Discussion
This study documents the presence of both native L. cervi and allochthonous L. fortisetosa in the Tuscan-Emilian Apennines (central Italy). These hippoboscids have already been documented in Italy, but the literature is still limited to local areas [6]. Cervus elaphus, C. capreolus, and D. dama were hunted and sampled for ectoparasites, revealing a considerable distribution of these flies. All three host species were infested with both parasites, showing the adaptability of these parasites to the examined cervids. Although L. cervi seems to have a greater worldwide distribution than that of L. fortisetosa [47], our survey proves that locally allochthonous species may be were largely more abundant than autochthonous species, demonstrating that the introduced L. fortisetosa is numerous in the study area and strongly competes with native hippoboscids which not only live in the same geographic territories but also share the same host species. Our study confirms the coexistence of L. cervi and L. fortisetosa in the same area, as evidenced in other European regions, such as northeastern Poland and Lithuania [19,48]. Moreover, L. fortisetosa was found to share the same host with other dipteran ectoparasite species in Japan, where it

Discussion
This study documents the presence of both native L. cervi and allochthonous L. fortisetosa in the Tuscan-Emilian Apennines (central Italy). These hippoboscids have already been documented in Italy, but the literature is still limited to local areas [6]. Cervus elaphus, C. capreolus, and D. dama were hunted and sampled for ectoparasites, revealing a considerable distribution of these flies. All three host species were infested with both parasites, showing the adaptability of these parasites to the examined cervids. Although L. cervi seems to have a greater worldwide distribution than that of L. fortisetosa [47], our survey proves that locally allochthonous species may be were largely more abundant than autochthonous species, demonstrating that the introduced L. fortisetosa is numerous in the study area and strongly competes with native hippoboscids which not only live in the same geographic territories but also share the same host species. Our study confirms the coexistence of L. cervi and L. fortisetosa in the same area, as evidenced in other European regions, such as northeastern Poland and Lithuania [19,48]. Moreover, L. fortisetosa was found to share the same host with other dipteran ectoparasite species in Japan, where it was sampled on Japanese deer with Lipoptena sikae [49].
Although many cervid species have been reported as suitable definitive hosts for L. cervi, red deer and moose seem to be the favored species in Europe, while the Japanese deer, C. nippon, is considered the main and original host for L. fortisetosa [15]. On red deer, L. cervi can reach a very high frequency of infestation, ranging between 78% and 100% [19,50], while it is less abundant on fallow deer [10]. A heavy infestation of L. cervi in four hunted roe deer was recorded in Romania, with the average number of flies exceeding 2500 parasites per host [51]. Nevertheless, in other countries, a lower infestation prevalence was noted for L. cervi on the same host species, varying from 36% to 64% [52][53][54]. In Lithuania, L. cervi was less abundant on roe deer specimens than on in the other two examined host species, red deer and moose [19]. Additionally, L. fortisetosa showed a preference for attacking red deer over roe deer since the prevalence of infestation was 49% [55] or 100% [19], and 23% [54] or 90% [19], respectively. To the best of our knowledge, no studies balancing L. fortisetosa infestation on red deer, roe deer, and fallow deer have been carried out, but our data are consistent in stating that both parasites prefer red deer hosts over the other two cervids.
Interestingly, although some host species showed a high overall fly infestation prevalence (i.e., 94.7% L. fortisetosa on D. dama), no one species reached the 100% prevalence recorded on moose by several authors [19,38,56]. Moreover, in our survey, the overall density of Lipoptena spp. was higher on red deer than on the other two host species, at 0.12/cm 2 for L. fortisetosa and 0.03/cm 2 for L. cervi. Yet, a much greater number of deer keds was counted on moose, on which these parasites reached as many as 17,500 specimens on a single bull [56]. Since moose generally harbors a large number of hippoboscids, we could deduce that this species is more suitable for these parasites. Kadulski [52] found that among moose, red deer, roe deer, fallow deer, and Sika deer, the prevalence and intensity of infestation were directly proportional to the size of the host. Our findings are consistent with this conclusion since roe deer are smaller than red deer and fallow deer. Visual stimuli are considered important during the host location behavior of hippoboscids [57], and these parasites probably tend to attack larger species that are more easily detectable because of their size. In addition, several hematophagous ectoparasites use chemical cues (CO 2 or odors) during the host-finding process [58]. Lourenço and Palmeirim [59] found that two Nycteribiidae species (Hippoboscoidea superfamily) mainly used carbon dioxide for long-distance host locations. This cue is emitted by all vertebrates, and animals larger in size tend to release it in larger amounts [60]. The antennae of L. cervi and L. fortisetosa are equipped with a well-developed sensory pattern on the external surface of the pedicel, suggesting that these different types of sensilla are likely able to perceive the chemical cues emitted by the hosts, supporting the hypothesis that more than visual signals alone are responsible for identifying host locations [6,61]. We can speculate that the large amount of CO 2 released by the hosts may contribute to explaining how the roe deer are attacked less than the other two species. Haarløv [10] suggested that red deer species occur in habitats that are more suitable for the development of pupae that fall on substrates that are more suitable for their survival. In our case, red deer and fallow deer coexist in the study area and share the same territories, making it difficult to hypothesize that the local habitat strictly affects host choice. Most likely, instead, preference for red deer could be due to the physical features of this host species, such as the structure of its coat. In fact, the host fur represents the environment in which hippoboscids live, so it should have the conditions they need to survive. Red deer have long and robust guard hairs with a dense layer of underhairs at the base; however, roe deer and fallow deer have shorter hairs forming a softer but thicker covering that may obstructed parasites from reaching the skin, making trophic activity more difficult [10,61].
In this paper, the objective of verifying the possible differences between host age classes and sex was determined only for red deer since this species was the favorite host for both L. cervi and L. fortisetosa.
Kadulski [52] observed that the intensity of infestation increases with the size of a host. Our results show significant differences in the choice of host age classes by both deer keds. Other authors highlighted that fawns are attacked less than subadults or adults [38,56]. This preference could be due to fawn behavior since they follow dams during their first year of life. Given that fawns are together with the adult females, the flies are more likely to choose the larger subject since it is more visible and releases a greater amount of CO 2 . Regardless, body size alone cannot explain the host choice displayed by these hippoboscids; in fact, other aspects, such as the behavior and ecology of the host species, interact to affect this selection. Madslien et al. [38] hypothesized that L. cervi prefers parasitizing subadults over adult moose since the former moves more, increasing the chance of encountering a deer ked. Another explanation for this preference could be the resistance that some host species seem to develop toward hippoboscid attacks, as suggested for reindeer [62] and moose [38]. However, deer do not show similar resistance in the case of severe infestations [56].
According to our results, L. cervi is significantly more abundant in males. Additionally, in this case, the larger size and the more intense motility of host males may explain this preference, but it is also possible to hypothesize that odor secretions emitted by red deer can affect host choice as well. Deer have specialized regions with glands, whose activity may change between sexes, producing secretions to mark their territory. As demonstrated for white-tailed deer (Odocoileus virginianus), this glandular activity is higher in males, especially in dominant subjects [63]. Additionally, in C. elaphus, there are quantitative and qualitative differences between males and females in terms of their released compounds [64]. As reported by Johnson and Leask [65] for C. capreolus, glandular activity and active testosterone metabolism can increase just prior to and during the mating season. In Italy, the breeding season of red deer occurs from late summer to early fall, overlapping with the host seeking period of L. cervi, possibly affecting the preference of the parasite for males. Further studies are needed to confirm the influence of sex differences in the odor secretions on L. cervi host selection.
Although L. cervi and L. fortisetosa are restricted to a limited group of species, they are able to adapt to new hosts and do not appear to strictly follow a parasitization scheme. In fact, we found both hippoboscids on all three examined deer species and on all host age classes and sexes. Apparently, these flies cannot be too selective in terms of sex and age classes since they are obligate ectoparasites and need to find a host shortly after emergence. The host species, however, seems to be an important prerequisite for Lipoptena spp.; in fact, red deer are favored by both flies.
Host density is one of the most important factors that needs to be considered when studying the distribution of hippoboscid ectoparasites, even if it does not explain all of the variation in the expansion of these flies. For instance, Meier et al. [35], suggested that a local increase in host density may allow for the rapid establishment of allochthonous ectoparasites. Additionally, L. cervi occurred in Finland in 1960 when the moose density experienced a large population growth [66]. The study of the relationship between parasites and hosts is fundamental, especially when it concerns allochthonous species which are able to adapt to new hosts, competing with native species/fauna. Hosts can represent the easiest transfer option for ectoparasites so that they can be disseminated in new territories during host movements and introductions. Just as the expansion of L. cervi in the northeastern United States is considered to be due to the anthropogenic introduction of European deer [9], it is likely that L. fortisetosa spread to Europe due to the relocation of its original host, Cervus nippon. However, the possible hybridization between sika and red deer, or the translocation of C. elaphus-related subspecies to Europe cannot be ignored. Currently, C. nippon is recorded in 20 European countries, while L. fortisetosa is present in 13 European countries [67]. In Italy, a great increase in cervid abundance has been recorded in recent years, and the presence of C. nippon has been recently documented [3,5]. This situation confirms the risk related to the increase in the abundance of native ectoparasites, together with the spread of alien parasitic species further favored by global warming.

Conclusions
The results of the present study show the great expansion of the allochthonous parasite L. fortisetosa, recently detected in Italy. This fly, originally restricted to the main host C. nippon, has a strong adaptability to other host species, such as red deer, fallow deer, and roe deer. Moreover, it seems to strongly compete with the autochthonous hippoboscid L. cervi, being more numerous in the study area. The favored host of both flies was red deer, even if all three examined host species harbored parasites. Different preferences for sex and age classes of the hosts were observed in the two hippoboscids. Although some explanations were hypothesized for these outcomes, at present, it is difficult to provide a specific explanation, since each choice occurred due to the interactions of many factors. Thus, further investigations are ongoing. Another aspect worthy of attention is related to the possible health risk implicated in the expansion of allochthonous species as potential vectors of harmful pathogens. Therefore, hippoboscid populations should be continuously monitored to promptly identify possible substantial expansion or adaptation to other host species, which can lead to further spread with negative consequences from both ecological and health perspectives. Regular monitoring of deer keds should also be carried out to improve the knowledge of these parasites and establish specific management strategies to limit hippoboscid expansion.  Institutional Review Board Statement: No animal experimentation was carried out in this study. Skin samples were taken from dead animals, which had been hunted in accordance with national and regional laws on conserving, managing and hunting wildlife in Italy.

Data Availability Statement:
No data were deposited in an official repository.