Trichinella spp. biomass has increased in raccoon dogs (Nyctereutes procyonoides) and red foxes (Vulpes vulpes) in Estonia

Background Raccoon dogs and red foxes are well-adapted hosts for Trichinella spp. The aims of this study were to estimate Trichinella infection prevalence and biomass and to investigate which Trichinella species circulated in these indicator hosts in Estonia. Methods From material collected for evaluating the effectiveness of oral vaccination program for rabies eradication in wildlife, samples from 113 raccoon dogs and 87 red foxes were included in this study. From each animal, 20 g of masseter muscle tissue was tested for the presence of Trichinella larvae using an artificial digestion method. The Trichinella larvae were identified to species level by multiplex polymerase chain reaction method. Results The majority of tested animals were infected with Trichinella spp. The parasite species identified were T. nativa and T. britovi. The apparent infection prevalence was 57.5% in raccoon dogs and 69.0% in red foxes, which were higher than previous estimates. In addition, the larval burden had also increased in both hosts. We estimated that in 2011–2012, the Trichinella spp. biomass was more than 15 times higher in raccoon dogs and almost two times higher in red foxes than in 1992–2000 (based on mean larval burden), and almost 20 times higher in raccoon dogs and almost five times higher in red foxes than in 2000–2002 (based on median larval burden). Conclusions Raccoon dogs and red foxes are relevant reservoirs for Trichinella spp. in Estonia. The biomass of Trichinella circulating in sylvatic cycles was substantial and had increased: there is substantial infection pressure in the sylvatic cycle.


Background
Trichinella spp. are zoonotic parasitic nematodes transmitted by carnivorism. Sylvatic Trichinella infections are endemic in Estonia, a EU country located in north-eastern Europe that is bordered by Latvia in the south and Russia in the east [1][2][3][4][5]. For example, while anti-Trichinella antibodies were not detected in the domestic pigs investigated in our recent study, a substantial proportion of hunted wild boars were Trichinella seropositive [5]. Assessment of the sylvatic component and awareness about it are important because there is a risk of spill-over to domestic animals and humans [5].
The raccoon dog (Nyctereutes procyonoides) is a suitable indicator species and well-adapted reservoir host for all four Trichinella species circulating in Europe, and the red fox (Vulpes vulpes) particularly for T. spiralis and T. britovi [4,[6][7][8][9][10]. The invasive raccoon dog [11], and the native red fox are common and numerous sylvatic carnivores in Estonia [12]. A total of 12,577 raccoon dogs and 7144 red foxes were hunted in Estonia during the hunting season 2011-2012 [13].
The most recent epidemiological data on Trichinella infections in raccoon dogs and red foxes in Estonia were based on material collected in 2000-2002 [4]. The apparent prevalence of Trichinella spp. was 42.0% in raccoon dogs and 40.6% in red foxes, which did not differ significantly from estimates from 1992 to 2000 [14]. In both earlier studies, T. nativa and T. britovi were identified in the target hosts. Our study aimed to update the Trichinella infection prevalence estimates in raccoon dogs and red foxes in Estonia and to identify the Trichinella species causing the infections. We compared the findings with the two previous estimates and with those reported from other European countries and estimated how the biomass of Trichinella has changed in Estonia.

Methods
For the evaluation of the effectiveness of the oral vaccination program for rabies eradication in wildlife [15], head samples from 1214 raccoon dogs and 625 red foxes were collected from whole Estonian territory (average density 4.3 animals per 100 km 2 ) from August 2011 to March 2012. The animals sampled were apparently healthy hunted animals, rabies indicator animals killed due to abnormal behavior near human settlements, and animals killed in traffic or found dead. We could investigate muscle samples from 200 of these heads (113 raccoon dogs and 87 red foxes), which was evaluated to be a sufficient sample size to estimate the infection prevalence with 80% confidence level.
To obtain a geographically representative sample for this study, the number of samples from each county was adjusted according to the surface area of the county, and a random sample was drawn from the samples available from there. Data on the estimated age (less than 1 year old = juvenile, at least 1 year old = adult) and gender of each animal had been collected on the submission forms. Age group was unknown for 31 animals and gender for 69 animals.
The samples were kept refrigerated until analysis, but few samples were or could have been frozen (n = 3 from raccoon dogs and n = 2 from foxes arrived frozen). From each animal, 20 g of masseter muscle tissue was analyzed for the presence of Trichinella spp. larvae using the European Union reference method, i.e. magnetic stirrer method, for artificial digestion [16]. The mean time between sampling and digestion was ten days (range: 1-92 days).
Larvae from each positive sample were evaluated morphologically and then counted, rinsed with water, collected, and stored in ethanol at room temperature until identification to species level. The species of Trichinella were identified using a previously described multiplex polymerase chain reaction method [17].
The sample size assessment and preliminary statistical calculations were done with OpenEpi software [18]. The confidence intervals (CI) of the prevalence estimates were calculated using Mid-P exact. Comparisons with the prevalence estimates, by host species and Trichinella species, from other European countries and previous Estonian studies were done using two by two tables. Two-tailed P-values (Mid-P exact) < 0.05 were considered statistically significant.
Logistic regression models were built with STATA 13.0 (Stata Corporation, College Station, Texas, USA) software for three outcomes: testing positive for Trichinella spp., testing positive for T. nativa, and testing positive for T. britovi. The variables we evaluated were 'host species' (raccoon dog or red fox), 'age' (juvenile or adult), 'gender' (female or male), 'county' (the 15 counties included as dummy variables i.e. allocated numbers that do not indicate any particular order), and 'cause of death' (whether the animal had been hunted, killed due to abnormal behavior, killed in traffic, or found dead). The variables with P-value ≤ 0.25 in univariable analysis were included in a multivariable model, followed by a stepwise backward elimination of those with P ≥ 0.05 that did not act as confounders.
Trichinella spp. biomass was quantitatively estimated for 1000 host animals and for the hunting bag, using estimate of weight of the host, estimate of proportion of muscle of the host weight (based on information available for small mammals of similar size), point estimate of Trichinella spp. prevalence, and mean or median larvae per gram of muscle tissue.
The presence of two sylvatic species, T. nativa and T. britovi, was confirmed (Table 1). Trichinella nativa was detected as the only species present or in mixed infections in 31.9% of raccoon dogs and 31.0% of red foxes, and T. britovi was detected as the only species present or in mixed infections in 24.8% of raccoon dogs and 35.6% of red foxes. Of those animals that hosted Trichinella spp. larvae that were determined to the species level, T. nativa was detected as the only species present or in mixed infection in 70.6% of raccoon dogs and 54.0% of red foxes. Of those animals that hosted Trichinella spp. larvae that were determined to the species level, T. britovi was detected as the only species or in mixed infection in 54.9% of raccoon dogs and 62.0% of red foxes. The prevalence of mixed infections had Table 1 Trichinella species identified in raccoon dogs (Nyctereutes With this particular Trichinella species, either as the only species or in mixed infection increased in red foxes from the estimate of the previous Estonian study (Table 3) [4]. The apparent Trichinella spp. infection prevalences estimated from the samples from 2011 to 2012 were higher than those from 2000 to 2002 in both raccoon dogs and red foxes (Tables 2 and 3) [4]. Moreover, the prevalence estimated from samples from 2011 to 2012 was higher than the one from 1992 to 2000 in red foxes; however, the estimate from 2011 to 2012 did not differ significantly from the estimate from 1992 to 2000 in raccoon dogs (Tables 2 and 3) [14].
The median larval burden had increased in both raccoon dogs and red foxes from those reported in the previous study: from 7.2 lpg to 98.0 lpg in raccoon dogs and from 3.0 lpg to 8.2 lpg in red foxes [4]. Furthermore, the proportion of animals with low larval burden (< 1 lpg) had decreased from 18.1% to 7.7% in raccoon dogs, and from 23.7% to 11.7% in red foxes, further indicating that the circulating parasite biomass of Trichinella larvae had increased [4]. The Trichinella spp. biomass was estimated to have increased 18.6-fold in raccoon dogs and 4.6-fold in red foxes (based on median larval burden) ( Table 4).
Trichinella nativa was not detected in samples from the large islands Saaremaa and Hiiumaa, nor the most southeastern county Võrumaa, while T. britovi was found in samples collected from all counties (Fig. 1).
None of the variables were significant factors for testing positive for Trichinella spp. in either of the hosts nor in both hosts together. The final model for testing positive for T. nativa had two variables, 'age' and 'county' , and the area under the receiver operating characteristic (ROC) curve was 0.72. The odds of testing positive for T. nativa were 3.6 times (P = 0.009, 95% CI: 1.4-9.3) higher in adults than in juveniles, and higher in counties Põlvamaa and Pärnumaa when compared with Harjumaa where the capital is located (P = 0.009, OR = 15.

Discussion
We summarized the results of European studies on Trichinella spp. infection prevalence in raccoon dogs ( Table 2) and red foxes (Table 3). Lower prevalences than our estimate from Estonia have been observed in both hosts in Finland, Latvia, Lithuania, Poland and Germany (Tables 2 and 3) [19][20][21][22][23][24][25][26]. Moreover, in red foxes, the Trichinella spp. infection prevalence was higher in Estonia than what has been reported in Austria, Belgium, Denmark, France, Great Britain, Hungary, Italy, Ireland, Netherlands, Norway, Northern Ireland, Portugal, Romania, Serbia, Slovakia, Spain and Switzerland (Table 3) . However, as different sampling schemes, sample sizes, sample material, and detection methods were used, these studies are not all directly comparable with our study.
In Europe, according to the International Trichinella Reference Centre [57], the northern species T. nativa has been found in raccoon dogs in Estonia, Finland, Latvia, Russia, and Sweden; and in red foxes in Estonia, Finland, Germany, Latvia, Norway, Poland, Sweden and Ukraine. The published studies on T. nativa in raccoon dogs and red foxes report lower prevalences (single and mixed infections included) in Latvia, Lithuania and Norway than our estimate from Estonia [4,22,42]. In Poland and Germany, T. nativa has been found in red foxes (Table 3) [20,21,[24][25][26]. When comparing the result of our study with that from the previous Estonian study, the T. nativa infection prevalence had increased in both raccoon dogs and red foxes (Tables 2 and 3) [4].
In Europe, T. britovi has been found in raccoon dogs in Estonia, Finland, Germany, Latvia and Lithuania (Table 2) [4,57]. It is the most common Trichinella species in red foxes in Europe [10]. The prevalence of T. britovi we observed in raccoon dogs in single and mixed infections was similar to that reported from Lithuania, lower than that from Latvia, and higher than those from western Poland and Germany (Table 2) [4,21,22,24]. The prevalence of T. britovi we observed in red foxes, including both single and mixed infections, was higher than those reported from Austria, France, Hungary, Norway, Poland, Portugal, Romania, Serbia, Slovakia and Switzerland (Table 3) [21-24, 32, 33, 45-48, 50, 54-56]. A similar to our prevalence estimate for T. britovi was detected in red foxes in Latvia [22]. When comparing the result of our study with that from the previous Estonian study, the T. britovi infection prevalence had increased in red foxes (Table 3) [4]. Moreover, mixed infections were more common in our study than what was observed in raccoon dogs and red foxes in the neighboring country Latvia and in red foxes in Lithuania (Tables 2 and 3) [4,22].
Despite the fact that we used 20 g of tissue for the digestion, our study likely underestimated the actual infection prevalence and larval burdens, because the available material was not optimal for finding Trichinella larvae [9,58,59]. In experimentally infected raccoon dogs, the T. nativa larval density in masseter muscle was about half of that in foreleg muscles [58].
The storage conditions and transport time could also affect the results [29,40,60].
The high Trichinella infection prevalence in raccoon dogs and red foxes, as well as the overall circulation of the parasites in the sylvatic cycle, may be supported by human behavior. For example, the local hunters use carcasses of hunted raccoon dogs as baits [61], which might help the transmission. According to winter tracking index and hunters' estimations, after the rabies vaccination program started in 2005 [62], the red fox population size first increased, with a peak in 2009-2010, and then decreased [12]. The raccoon dog population size has increased since the second half of last century [62] and has relatively stabilized after 2011-2012 [12]. These changes are also reflected in the increased hunting bag sizes [13] and may  have relevance beyond simply higher numbers. There was an association between the abundance index of raccoon dogs and the proportion of Trichinella-infected raccoon dogs and red foxes in Finland [19]. Estonia is located in the transition zone of maritime and continental climate [63]. The coldest months with mean air temperature below zero are December to February [64]. According to data covering these three months from six weather stations located in Harjumaa, Lääne-Virumaa, Pärnumaa, Saaremaa, Tartumaa and Võrumaa, the mean number of days with snow cover was 12.7% (from 4% in Pärnumaa to 26% in Saaremaa) higher in 2002-2011 than in 1992-2001 (data received on request from Estonian Environment Agency). The snow cover could reduce the destructive effect of freezing-thawing cycles on carcasses of infected animals and thus facilitate survival of Trichinella larvae [65,66].
Raccoon dogs and red foxes act as reservoir hosts for Trichinella spp. in the sylvatic cycle, where the infection can spread to game animals, such as wild boars, that are hunted for human consumption. The Trichinella seroprevalence in wild boars is high in Estonia [5], and the odds of testing Trichinella-seropositive were higher if the wild boar was hunted in certain counties, including Pärnumaa and Saaremaa, when compared with Harjumaa. In this study, a similar comparison was made, with Harjumaa as the reference county. Raccoon dogs and red foxes had higher odds to test T. nativa positive in Pärnumaa, whereas the odds to test T. britovi positive were higher in Pärnumaa and Saaremaa. Moreover, the highest larval burden was detected in a young raccoon dog from Pärnumaa. This raccoon dog had a mixed infection. These two counties could thus be interesting for further studies.
We estimated that in 2011-2012, the Trichinella spp. biomass was more than 15 times higher in raccoon dogs and almost two times higher in red foxes than in 1992-2000 (based on mean larval burden), and almost 20 times higher in raccoon dogs and almost five times higher in red foxes than in 2000-2002 (based on median larval burden) ( Table 4). Using the increased hunting bag in the calculation as an indication of increased population size or as an indication of biomass removed from the circulation by hunting, the role of these hosts as reservoirs was clearly illustrated ( Table 4). The widespread distribution of Trichinella infections in Estonian wildlife underlines that there is a high infection pressure within the eastern European sylvatic cycles. Moreover, the results of this study indicate that there is an increase in the infection pressure. Trichinella spp. thrive in Estonia, and there is a continuous risk of spill-over to domestic animals and humans.

Conclusions
In Estonia, the proportion of both raccoon dogs and red foxes that hosted Trichinella were higher than ten years earlier. In addition, the larval burdens had also increased in these hosts, and an increased biomass of Trichinella larvae was circulating in sylvatic cycles. Trichinella nativa and T. britovi were found in both host species. There is a substantial and increasing Trichinella infection pressure to the food chains and humans.