Biotic and abiotic factors influencing the prevalence, intensity and distribution of Eucoleus aerophilus and Crenosoma vulpis in red foxes, Vulpes vulpes from Romania

To understand the importance of host’ habitat choice in determining parasite burden, we studied the distribution of two helminth parasites of the red fox (Vulpes vulpes) in south-eastern Europe (Romania): Crenosoma vulpis and Eucoleus aerophilus, both widely distributed respiratory nematodes parasitic in various carnivores. Even though the life cycle and biology of the two nematodes follow a different pattern, both parasites appear to be co-distributed and often co-infect foxes with variable prevalences across their range. Between July 2016 and August 2018, 550 red foxes, V. vulpes were collected by hunters in different localities from 22 counties of Romania and examined by necropsy. All parasites found in the trachea and bronchial system were collected and preserved in 70% ethanol. We characterised red fox/parasite habitats using seven predictors (fragmentation, altitude, presence/absence of water surface, per cent cover of arable land/grassland/urbanized areas/forest cover/wetlands). Prevalence, abundance, intensity, and sex ratio were calculated and statistically analysed using the R software. Out of the 550 examined foxes, 76.2% were infected with lungworms. The overall prevalence was 32.0% for C. vulpis and 72.5% for E. aerophilus. The mean intensity of infection was 13.70 for C. vulpis 6.15 for E. aerophilus. For both nematodes, the prevalence was significantly higher in males than in females, and there was no influence of hosts’ age. No statistical differences were found for intensity and mean intensity in the case of infection with C. vulpis and E. aerophilus between age and sex categories. The abundance of C. vulpis showed a strong positive relationship with the presence of wetlands and habitat fragmentation. We found a significant correlation between the abundance of E. aerophilus and altitude, with foxes from higher elevations showing higher prevalences.


Background
Parasitic helminths are common in most carnivore species and may cause significant pathology in their hosts. As most species are transmitted via food ingestion (either from an intermediate host or accidentally), host diet and habitat use may play key roles in the acquisition of infections (Samuel et al., 2001). For generalist carnivores, the diet is determined primarily by the habitat choice and may show wide seasonal fluctuations in temperate regions (Rosalino et al., 2011). To understand the importance of host' habitat choice in determining parasite burden, we studied the distribution of Crenosoma vulpis and Eucoleus aerophilus (=Capillaria aerophila), two widely distributed respiratory nematodes of the red fox (Vulpes vulpes), in south-eastern Europe, Romania.
Eucoleus aerophilus is an oviparous capillariid, infecting mainly the tracheal and bronchial mucosa of canids, felids and mustelids worldwide (Anderson, 2000). The eggs embryonate and develop in the environment, to become infective and probably need to be eaten by an earthworm to contaminate a carnivore finally (Anderson, 2000). Even though E. aerophilus is a relatively common parasite of carnivores, its life cycle is still poorly understood, and the obligate or facultative role of earthworms is still uncertain (Traversa et al., 2011).
Even though the life cycle and biology of the two nematodes follow a different pattern, both parasites appear to be co-distributed and often co-infect foxes. The prevalences of C. vulpis and E. aerophilus show wide variability and distribution seems to be patchy across their range (Nevárez et al., 2005;Davidson et al., 2006;Hodžić et al., 2016;Schug et al., 2018). Only a few studies have evaluated the influence of environmental factors on the spatial distribution of these respiratory nematodes (Tolnai et al., 2015;Maksimov et al., 2017;Čabanová et al., 2018). However, our understanding on the drivers of distribution and environmental risk factors associated with lungworm infection in carnivores is still limited, despite their very common occurrence.
Our study aimed to analyse the potential environmental and hostrelated (biotic) factors which can influence the epidemiological features (presence, prevalence, intensity) of C. vulpis and E. aerophilus infection in red foxes.

Parasite data
Between July 2016 and August 2018, 550 red foxes (315 males, 235 females; 180 young, 360 adults) were collected by hunters in different localities from 22 counties of Romania, through the country veterinary authority, as part of the rabies surveillance program (supplementary material) (Fig. 1). Romania is a climatically and geographically heterogeneous country, with the topography almost evenly divided among mountains (31%), plains (33&), and hills (36%) and five types of climate (alpine, cool continental, wet temperate continental, wet warm continental and warm oceanic) (http://www.meteoromania.ro/clima/ clima-romaniei/). All foxes originated from areas with wet continental climates.
For safety reasons, only foxes negative for rabies have been used for the present work. For each fox, the location (county, locality, precise location if available), age (young, less than one-year-old; adult, more than one-year-old, according to Harris, 1978) and sex were recorded. During the necropsy, the respiratory tract was longitudinally opened and carefully examined for nematodes under a stereo zoom microscope. All parasites found in the trachea and bronchial system were collected and preserved in 70% ethanol. All parasites were identified to species level and sexed using keys and morphological descriptions (Anderson et al., 2009;Latrofa et al., 2015).

Environmental predictors
The collection locations were geo-referenced and environmental predictors were collected for each point, using a 2 × 2 km cell grind containing the geo-referenced coordinates of the collection site (Sándor et al., 2017). These cells (area = 400 ha) are similar to the average redfox home range size in semi-natural and natural habitats in Europe (mean 413.42; min = 12.95, max = 1990.00; SD 393.1192 ha, n = 84 studies, (Lucherini and Lovari, 1996;Bartoń and Zalewski, 2007;Holmala and Kauhala, 2008;Šálek et al., 2015).
We characterised red fox/parasite habitats using seven predictors (fragmentation, altitude, presence/absence of water surface, per cent cover of arable land/grassland/urbanized areas/forest cover). Fragmentation was an integer number, assigned for each cell using the number of land-cover units crossed by the NW-SE diagonal at the cell level. The CORINE LandCover (European Environment Agency, http:// www.eea.europa.eu/) dataset was used as source for land-use data (see Table 1 for the associated CORINE LandCover categories). There was no statistical difference (x 2 = 0.2257956, df = 3 p < 0.97) between land-use composition inside the sampled 2 × 2 km plots and the overall land-use composition of the region, thus we consider that our results may be generalized for the whole region. Altitude for each cell was extracted from WorldClim database (www.worldclim.org), the data being downloaded from WorldClim database with the highest resolution provided, of 30 arc-seconds (one pixel equals ca. 0.6 km 2 ).

Statistical procedures
Prevalence, abundance, intensity, and sex ratio were calculated and statistically analysed using the R software (Version 3.2.3). Sample prevalence data were analysed using Fisher's exact test. Relationship between parasite prevalence and environmental predictors (land-use and altitude) was tested using Spearman Rank Correlation. Differences were considered significant when p < 0.05.

Results
Three species of pulmonary nematodes have been identified: Angiostrongylus vasorum, Crenosoma vulpis and Eucoleus aerophilus. The data regarding the prevalence and intensity of A. vasorum were presented elsewhere (Deak et al., 2017). Out of the 550 examined foxes, 419 (76.2%) were infected with C. vulpis or E. aerophilus ( Table 2). The overall prevalence was 32.0% for C. vulpis and 72.5% for E. aerophilus. The intensity of infection varied between 1 and 265 nematodes per fox for C. vulpis (mean intensity 13.70) ( Table 3) and between 1 and 51 nematodes per fox for E. aerophilus (mean intensity 6.15) ( Table 4). For both C. vulpis and E. aerophilus, the prevalence was significantly higher in males than in females, and there was no influence of the age (Tables  5 and 6). Male foxes also had significantly more co-infections than females (Table 7). No statistical differences were found for intensity and mean intensity in the case of infection with C. vulpis and E. aerophilus between age and sex categories (Tables 2 and 3).
Some of the foxes were received without exact collecting locations (n = 196, 35.64%), and these were excluded from the habitat-related analyses. The abundance of C. vulpis showed a strong positive relationship with the presence of wetlands and was so significant that all the other habitat predictors showed a negative correlation (Table 8). Both, prevalence (z = 4.440, p < 0.01) and intensity (z = 4.668, p < 0.01) were linked to the presence of wetlands. Habitat fragmentation (number of different land-use patches inside the 400 ha plot) showed a positive correlation with C. vulpis abundance. However, fragmentation was uncorrelated to the number or surface area of wetlands (z = 0.0023, p > 0.1).
We found a significant correlation between the abundance of E. aerophilus and altitude (z = 1.977, p = 0.0480), with foxes from higher elevations showing higher prevalences (independent of host sex or age). There was no relationship, however between altitude and the intensity of parasitism.

Discussion
The data on the influence of host-related factors (sex, age) on the prevalence, intensity and abundance of C. vulpis and E. aerophilus show wide variations. Several studies have shown that males foxes are more commonly infected with C. vulpis than females (Goble and Cook, 1942;Willingham et al., 1996), while others did not find any differences between sexes (Jeffery et al., 2004). Similarly, E. aerophilus was found G. Deak, et al. IJP: Parasites and Wildlife 12 (2020) 121-125 more commonly in males foxes than in females (Morgan et al., 2008). Our study confirms this characteristic also in Romania, for both parasites. The higher infection rate of males has been attributed to several reasons such as increased susceptibility to infection due to higher testosterone levels, increased exposure to parasites due to more risky behaviour, different diet due to their increased body size (Schmid-Hempel, 2011) or longer distances travelled predominantly by subadult males (Walton et al., 2018). It is uncertain why E. aerophilus had higher mean intensity in female foxes compared to males. It is speculated that variation in habitat use, foraging time, and/or diet may influence the acquisition of helminth species (Bush, 1990;Fedynich et al., 2005). Consecutively, different exposure probabilities to infective parasitic stages will result in different host-sex related intensities. Only a few studies have investigated the possible influence of environmental factors on the distribution and infection rate of C. vulpis and E. aerophilus. Moreover, results are not consistent between these studies, although all were geographically located in Central Europe (Tolnai et al., 2015;Maksimov et al., 2017;Čabanová et al., 2018). The only environmental factor which seems to influence the abundance of C. vulpis in our study was the presence of wetlands. Tolnai et al. (2015) showed that the infection rate with C. vulpis in foxes from Hungary was positively correlated only with annual precipitation. Maksimov et al. (2017) investigated the risk factors associated with shedding of C. vulpis larvae in the faeces of dogs in Germany and concluded that larval shedding is higher in the winter and higher in dogs less than one-yearold. The higher proportion of moorlands was the most important among the environmental factors associated with a higher risk for C. vulpis larval shedding in domestic dogs in Germany. Agricultural fields and water bodies were associated with a lower risk (Maksimov et al., 2017), which is in contrast to our results. Čabanová et al. (2018) concluded that the distribution of C. vulpis in foxes from Slovakia is not significantly influenced by any of the 16 environmental factors used as predictor variables of the regression modelling. Environmental factors do not necessarily affect the distribution and infection rate of C. vulpis, which is a heteroxenous nematode. They could act on the snail populations, intermediate hosts, thus indirectly impacting the nematode' distribution and infection rate. Generally, relative air and soil humidity,  Transitional woodland-shrub Forest 333 Sparsely vegetated areas Grassland G. Deak, et al. IJP: Parasites and Wildlife 12 (2020) 121-125 atmospheric and soil temperature, as well as litter depth, influence the distribution of land snails (Nunes and Santos, 2012). Wetlands and moorlands, characterized by a high level of humidity, meet optimal conditions for increased snail populations' sizes. Except for the altitude, no influence of other environmental factors was found in the case of prevalence, intensity and abundance of E. aerophilus in our study. The current knowledge on the biology and transmission dynamics of E. aerophilus to foxes does not allow to produce any strong hypothesis on why foxes at higher altitudes had higher prevalences. Several factors such as parasite infectivity and survival in the environment or different dietary preferences and behavior of foxes could be a possible cause. However, the infection rate with  Significance levels: n.s. = not significant. Significance levels: p < '*' 0.05, n.s. = not significant. Significance levels: p < '**' 0.01. Significance levels: p < '**' 0.01, n.s. = not significant. Significance levels: p < '**' 0.01, n.s. = not significant. E. aerophilus was positively correlated with annual precipitation and negatively correlated with the mean annual temperature in a survey from Hungary (Tolnai et al., 2015). As shown in Fig. 1, C. vulpis was mostly absent from the lowlands (south east and the western border with Hungary). In spite that our results suggest that the altitude, annual precipitation, and temperature seems to influence prevalence, intensity and abundance of E. aerophilus, others appreciated that drivers influencing the occurrence of this parasite are still unknown (Traversa and Di Cesare, 2014). Moreover, the eggs of E. aerophilus develop in the environment and the role of earthworms in the life cycle is questionable. More recent data indicate that the life cycle of E. aerophilus is direct, with a facultative, not mandatory, intervention of earthworms (Bowman, 2002;Taylor et al., 2007).
Our study is the first to analyse and demonstrate the influence of the habitat fragmentation, which was positively correlated with C. vulpis abundance. In conclusion, the present study shows the common occurrence of C. vulpis and E. aerophilus in foxes from Romania and demonstrates the influence of only a few environmental factors on their abundance and no influence of most other environmental factors. This might be related to the ubiquity of foxes, their variable diet and the complex biology of these nematodes.