A total of 150 red fox carcasses were collected across Bolzano province (Italy) by wildlife technicians in the period 2020–2021. Animals were either found dead, within existing passive surveillance programs, or culled during hunting season and population control campaign approved by local wildlife management authorities. Carcasses were transported to the Bolzano Laboratory of Istituto Zooprofilattico delle Venezie and frozen at −80°C for 72 h before examination to inactivate eggs of Echinococcus spp., thus preventing infection risk for personnel (19).

At necropsy, a fecal sample was collected from the rectum of each fox using sterile gloves; the small intestine was then tied at both ends, removed and transferred to the Department of Animal Medicine Production and Health, University of Padova. Classical copromicroscopic examination was carried out on separate fecal aliquots by means of two techniques: (i) a classical flotation technique (FT) using a Zinc chloride solution (specific gravity 1.350), followed by observation at optic microscope for identification of eggs at the lowest taxonomic level possible; (ii) a filtration/sieving technique followed by multiplex PCR on positive samples, for identification of taeniid eggs [named CMPCR, as per (20)]. Briefly, to this aim, 2 grams of feces were subjected to flotation; after a centrifugation step, supernatant was passed through 40 and 20 μm mesh sieves and retained eggs were collected. DNA was then extracted, and three couples of primers were initially used to amplify the ND1 gene for Echinococcus multilocularis and 12S rRNA for both E. granulosus and Taenia spp. Positive samples were then amplified using a PCR assay targeting a fragment of the cytochrome oxidase gene and sequenced to obtain species identification as per Citterio et al. (20).

Furthermore, isolation of helminths from the small intestine was performed through scraping, filtration and counting technique (SFCT), which was used as reference test. Briefly, the small intestine was cut into short segments (30 cm) which were longitudinally opened and successively rinsed in tap water to collect all the content in a beaker; the intestinal wall was then scraped, collecting all the washed material in the same beaker. Rinses were then filtered using 1,000 and 212 μm sieves. Non-filtered material was entirely collected and observed under the stereomicroscope (Olympus-SZX12) for the isolation and counting of the helminths. Besides, aliquots corresponding to the 25% of the total volume of washes was examined under the stereomicroscope from the second sieve. Individual parasites were identified morphologically at the lowest taxonomic level possible, using identification keys found in literature (21–30). Briefly, nematodes referred to as members of the superfamily Ascaroidea were identified as Toxocara sp. based on the presence of a glandular ventriculus at the end of the esophagus, and the presence of a digitiform-shaped caudal extremity and caudal alae in males, all absent in the genus Toxascaris. Moreover, elongated, narrow cervical alae were used as a mark to distinguish the species T. canis from Toxocara cati, which holds broader and shorter cervical alae. A minimum length 5.6 and 6.1 cm was set as the cut-off to define mature male and female specimens, respectively (21–23). Bursate nematodes were referred to the family Ancylostomatidae. Buccal capsule morphology was observed to distinguish Uncinaria stenocephala from Ancylostoma spp., considering the presence of two chitinous plates at the ventral border in the former, and one to three developed pairs of ventral teeth, jointly with two dorsal ones, in the latter (22). Moreover, smaller spicules are present in males of U. stenocephala (0.64–0.76 μm) compared to A. caninum (0.8–0.95 μm) (22). Size, presence of a well-developed bursa and spicula or eggs in uterus were used to confirm sex and sexual maturity (22, 24). Small nematodes of the genus Pterygodermatites were distinguished from the aforementioned genera for the presence of two rows of well-developed, peculiar, cuticular spines along the sub-ventral body surface. The presence of eggs in the uteri or unequal spicules was used for the sex determination (25, 26). Presence or absence of rostellum was used to distinguish among Taeniidae and Mesocestoides spp., respectively. Morphology and measurements of the hooks was used to identify Taenia at species level (27, 28). Lack of a rostellum and presence of four large suckers, together with presence of a central, parauterine organ, were the main characteristics for the identification of Mesocestoides sp. (29, 30).

Identification of cestodes at species level was molecularly confirmed by PCR and sequencing. DNA extraction was performed on 7 Taeniidae (n = 2 for Taenia crassiceps, n = 4 for Taenia polyacantha and n = 1 for Hydatigera [Taenia] taeniaeformis) and 1 from Mesocestoides sp. by using the NucleoSpin^® Tissue Kit (Macherey-Nagel, Germany), according to the manufacturer's instructions. The mt-CO1 region was amplified using primers JB3 and JB4.5 as already reported by Bowles et al. (31). PCR amplicons were purified and sequenced from both ends at Macrogen (Macrogen Europe, Amsterdam, The Netherlands). Alignment was performed with Clustal W integrated into MEGA v6.0, and sequences were compared with the non-redundant database available in the GenBank^® database using the software BLAST (32).

An agreement table between each of the two copromicroscopic tests (FT and CMPCR) and the scraping one (SFCT) was displayed; the concordance was then calculated for each of the two tests as the number of samples with same result out of the total number of samples examined (% concordance).

The sensitivity (Se) of copromicroscopic tests and 95% confidence interval (95%CI) was assessed for ascarids and Ancylostomatidae (only for FT), and for taeniids (both for FT and CMPCR), using SFCT as the reference standard. Analysis of the sensitivity of the two tests was performed through EpiTools (https://epitools.ausvet.com.au), using the Clopper-Pearson (exact) method for 95% CI definition.

The intensity of infection [i.e., the number of individuals of a particular parasite species in a single infected host, sensu Bush et al. (33)], was displayed and descriptively analyzed to assess its influence on the outcome of the copromicroscopic tests.

To unveil the relation between the presence of eggs in the feces and the adults in the intestines, we performed a literature search to complement our findings with data generated by similar studies previously conducted. Prevalence values in other red fox populations were retrieved from literature, focusing on studies conducted exclusively in Europe and using scraping (SCT or SFCT) and copromicroscopic (qualitative flotation or quantitative techniques) approaches on the same animals. For this aim, the database Google Scholar was searched using the terms “copromicroscopic” OR “eggs” AND “red fox” as keywords. Data from single publications investigating different, geographically separated, red fox populations, were considered independently as separated datasets. The level of concordance between the prevalence values obtained with the scraping and the copromicroscopic approaches was assessed through the Bland Altman plot, which graphically describes the agreement between two methods (34). Finally, the pooled prevalence and its 95% CIs were estimated for each parasite taxon, based on the inverse variance method and logit transformation, to account for different sampling sizes among studies (35).

Overall, SFCT allowed for the detection of parasites from 114 out of 150 foxes (76.0%). Among these, simple infections were present in few cases (n = 35; 30.7%). Three species of nematodes (T. canis, U. stenocephala and Pterygodermatites sp.), and five species of Cestodes (E. multilocularis, T. crassiceps, T. polyacantha, H. taeniaeformis, Mesocestoides sp.) were detected and identified, and selected specimens identification was confirmed by molecular analyses (identity >99.0% with T. crassiceps [acc.n. OP738082]; 99.7% with T. polyacantha [acc.n. OP738083]; 98.4% with H. taeniaeformis [acc.n. OP738085]; 99.5% with Mesocestoides litteratus [acc.n. OP738084]).

Flotation allowed to detect mixed or simple infections in 79 samples (52.7%). The eggs were referable to Toxocara sp., Ancylostomatidae and Taeniidae, as well as to Trichuris sp. and Capillaria sp., whose adult forms were not detectable by our technique, applied exclusively to the small intestine. No eggs of Pterygodermatites sp. or Mesocestoides sp. were found with copromicroscopic examination.

As for nematodes, the overall concordance between SFCT and FT was 62.7% for ascarids, and 70.7% for members of the Ancylostomatidae. Most non-concordant values were due to negative results at the FT when adult specimens were found in the intestine with SFCT. In few cases, adult parasites were not detected in the intestine, when ascarids and/or Ancylostomatidae eggs were observed at FT (Table 1). The sensitivity (Se) of FT was below 40% for both groups of nematodes. Nevertheless, excluding single-sex infections and infections solely by immature worms, Se increased to 46.1% for ascarids and to 42.8% for Ancylostomatidae (Table 1).

The number of parasites counted in positive animals at SFCT (intensity) ranged from 1 to 100 for ascarids and from 1 to 32 for hookworms. The intensity of infection did not seem to strongly influence the outcome of the copromicroscopic test, since animals positive at FT were fairly distributed among the different intensities. The highest intensity resulting in a negative copromicroscopic test were 26 ind. for ascarids, and 12 ind. for Ancylostomatidae.

For taeniids, scraping and FT had a low level of concordance (51.3%) that slightly increased for CMPCR (58.7%; Table 2). In the case of cestodes, all positive samples at copromicroscopic tests were confirmed by isolation of cestodes from the intestine, consequently low concordance was only due to false negative results at copromicroscopy (Table 2).

Among cestodes, the intensity of infection varied greatly between Taenia spp., Mesocestoides spp. and E. multilocularis, ranging from 1 to 22 for T. crassiceps (n = 33), from 1 to 82 for T. polyachanta (n = 31), from 1 to 26 for Mesocestoides sp. (n = 32), and from 4 to 19,800 for E. multilocularis (n = 24). A single animal was found infected with a single specimen of H. taeniaeformis. Similar to nematodes, the intensity of infection did not seem to strongly influence the outcome of the copromicroscopic test, since animals with highest intensity of adults in the intestine, respectively 22 for T. crassiceps, 82 for T. polyachanta and 19,800 for E. multilocularis, resulted negative at the FT test. As already mentioned, no Mesocestoides sp. eggs were detected at the FT.

At CMPCR, eggs referable to taeniids were detected in 19 samples, but only in 13 samples it was possible to identify the species or the genus through Multiplex PCR and sequencing, i.e., T. crassiceps (n = 7, 1 co-infection), E. multilocularis (n = 5, 3 co-infections), T. polyacantha (n = 2, 1 co-infection) and Taenia sp. (n = 2, 1 co-infection). The specific identifications at CMPCR were mostly in agreement with the results of adults' isolation from the intestine (SFCT), as showed in Table 3, although, in some cases of co-infections, one parasite species may remain undetected.

Based on the literature search, we found five studies investigating helminthic infection in red fox populations using, at the same time, scraping and copromicroscopic techniques (18, 36–39). However, three studies (36, 38, 39) were excluded from the analysis because the prevalence values at necropsy and at copromicroscopy were calculated on different subsets of animals, and the prevalence values in the group of animals analyzed by both methods were not reported. In one study, the investigated red fox population was split in three different datasets, due to their geographical separation. Therefore, a total of five datasets, including ours, were used in the meta-analysis. The prevalence values reported in each study for each considered group of parasites are listed in Table 4, along with the estimated pooled prevalence, and their concordance is displayed in Figure 1, using a Bland-Altman graph.

The prevalence value was constantly lower when estimated by copromicroscopy, rather than scraping, with mean difference ranging from about 15% for ascarids, to around 40% in the case of Ancylostomatidae and taeniids (Figure 1). Mesocestoides was the most underestimated genus at copromicroscopy, since the mean difference in prevalence values reaches about 70%. Moreover, the difference in the estimates between the two approaches increased with increasing prevalence values, although this was evident for cestodes, and less clear for nematodes. These differences were similarly highlighted by the calculation of the pooled prevalence, which was always higher at necroscopy (Table 4).