*Article* **Long-Term Determinants of the Seroprevalence of** *Toxoplasma gondii* **in a Wild Ungulate Community**

**Patricia Barroso 1,\*, Ignacio García-Bocanegra 2, Pelayo Acevedo 1, Pablo Palencia 1, Francisco Carro 3, Saúl Jiménez-Ruiz 1,2, Sonia Almería 4, Jitender P. Dubey 5, David Cano-Terriza <sup>2</sup> and Joaquín Vicente 1,6**


Received: 17 November 2020; Accepted: 8 December 2020; Published: 9 December 2020

**Simple Summary:** *Toxoplasma gondii* is a zoonotic intracellular parasite which infects a wide range of warm-blooded animals. Long-term studies provide the necessary perspective required to understand those processes which took place over many years in order to address epidemiology and ecology in complex host communities. This study is focused on evaluating what the main long-term determinants of the seroprevalence of *T. gondii* are in the wild ungulate community from Doñana National Park (southwestern Spain). With this purpose, we assayed sera from 1573 wild ungulates (wild boar, red deer, and fallow deer), collected for 13 years (from 2005 to 2018). We found high seroprevalence values of *T. gondii* (% ± CI 95%; wild boar 39 ± 3.3; red deer 30.7 ± 4.4; and fallow deer 29.7 ± 4.2. Several factors operating in the medium and long-term (individual, environmental, population and stochastic) explained the risk of *T. gondii* in wild boar and deer, some of them operating at the community level.

**Abstract:** *Toxoplasma gondii* is an obligate intracellular protozoan which infects warm-blooded vertebrates, including humans, worldwide. In the present study, the epidemiology of *T. gondii* was studied in the wild ungulate host community (wild boar, red deer, and fallow deer) of Doñana National Park (DNP, south-western Spain) for 13 years (2005–2018). We assessed several variables which potentially operate in the medium and long-term (environmental features, population, and stochastic factors). Overall, the wild ungulate host community of DNP had high seroprevalence values of *T. gondii* (STG; % ± confidence interval (CI) 95%; wild boar (*Sus scrofa*) 39 ± 3.3, *n* = 698; red deer (*Cervus elaphus*) 30.7 ± 4.4, *n* = 423; fallow deer (*Dama dama*) 29.7 ± 4.2, *n* = 452). The complex interplay of hosts and ecological/epidemiological niches, together with the optimal climatic conditions for the survival of oocysts that converge in this area may favor the spread of the parasite in its host community. The temporal evolution of STG oscillated considerably, mostly in deer species. The relationships shown by statistical models indicated that several factors determined species patterns. Concomitance of effects among species, indicated that relevant drivers of risk operated at the community level. Our focus, addressing factors operating at broad temporal scale, allows showing their impacts on the

epidemiology of *T. gondii* and its trends. This approach is key to understanding the epidemiology and ecology to *T. gondii* infection in wild host communities in a context where the decline in seroprevalence leads to loss of immunity in humans.

**Keywords:** parasite; long-term study; protozoan; shared infections; zoonoses; wildlife-livestock interface

#### **1. Introduction**

*Toxoplasma gondii* is a zoonotic obligate intracellular protozoan which infects warm-blooded vertebrates [1]. It has an indirect life cycle where wild and domestic felids are the definitive hosts, excreting oocyst in feces. Humans, as well as many mammal and bird species, serve as intermediate hosts of *T. gondii* and can become infected by vertical transmission, the fecal-oral route, through the ingestion of water or food contaminated with sporulated *T. gondii* oocysts, or through the consumption of tissues from animals infected with encysted bradyzoites [1,2].

*T. gondii* has been detected in wildlife and livestock worldwide [1]. Previous Spanish studies revealed a widespread distribution of this parasite in both wild and domestic ungulates, showing significant differences in the presence of *T. gondii* among geographic areas [3–7]. In Mediterranean ecosystems in southern Spain, antibodies against *T. gondii* have been detected in wild ungulates including wild boar (*Sus scrofa*), red deer (*Cervus elaphus*), fallow deer (*Dama dama*), roe deer (*Capreolus capreolus*), Barbary sheep (*Ammotragus lervia*), mouflon (*Ovis aries musimon*) and Iberian ibex (*Capra pyrenaica*). In these studies, seroprevalences of 40.2%, 15.6% and 10.5% were reached in wild boar, fallow deer and red deer, respectively [3,8], being lower or calculated from a few samples in the other species. Regarding livestock, serosurveys in this region revealed rate levels ranging between 18.6–83.3% and 16.2–24.3% in domestic ungulates [5,8,9] and pigs, respectively [7,10,11]. Higher seroprevalences from this area have been reported in wild carnivores, especially in the Iberian lynx (*Lynx pardinus*), reaching rates of 81.5% [12].

Host-pathogen dynamics are subjected to several processes which operate over broad temporal scales; however, little attention has been paid to *T. gondii*, and particularly, in intermediate host communities at the wildlife-livestock interface [4,8]. Wide temporal data series are essential to address epidemiology and ecology in complex host communities with the necessary perspective required to understand processes taking place over many years [13–16]. In Doñana National Park (DNP, South West Spain), the wild ungulate community (including wild boar, red deer and fallow deer) occurs sympatrically with free-ranging cattle and horses, and one of the most important meta-population of the endangered Iberian lynx [17]. Studies on *T. gondii* in DNP has been exclusively conducted in felid populations with conservational purposes, showing a widespread infection in the area and reporting seroprevalence rates up to 60% [12,18,19].

The multiple transmission routes and capacity of *T. gondii* to find niches into the hosts studied provided an excellent scenario to improve our understanding of the transmission dynamics of this pathogen. While *T. gondii* has normally been considered an excellent model to study host-pathogen interactions, we also showed that it may also be used to address the study of population, community and environmental factors. The present long-term study illustrates the interplay of factors, particularly factors operating at broad temporal scale that may contribute to the spread and maintenance of a pathogen over host communities. In this context, we present data on serosurveillance of *T. gondii* in wild ungulates (wild boar, red deer, and fallow deer) from DNP for a 13-years, with the specific aims of: (I) evaluating the factors (individual, populational and environmental) modulating the seroprevalence of *T. gondii* (STG), and (II) assessing the factors operating in the long-term (population and stochastic) in order to explain the temporal trend of STG in the intermediate host ungulate community from 2005 to 2018.

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

#### *2.1. Study Area*

This study was conducted in DNP (54,000 ha), one of the most relevant biodiversity reserves in Europe, located on the Atlantic coast of southwestern Spain (37◦09 N, 6◦30 W). Human access to the park is restricted and agriculture and hunting are prohibited inside the park; cattle and horse breeding are allowed and are, mainly focused on autochthonous and traditional breeds [20].

The habitat consists of a greater proportion of sand dune habitat and marshland, combined with pine forest and Mediterranean scrubland (see Figure 1 and [21] for a more detailed description). Between the scrublands and the edge of the marshland, there is a narrow north-south longitudinal strip of humid ecotone of high ecological richness.

**Figure 1.** Map of the study area, Doñana National Park. The sampling areas (cattle management units: Coto del Rey (CR), Sotos (SO), Doñana Biological Reserve (RBD), Puntal (PU) and Marismillas (MA)) are delimited and the ecotone and small human settlements are displayed by a dark band and blue "X", respectively. Red deer (squares), fallow deer (circles), and wild boar (triangles) sampled are shown. Black and white symbols mean animals positive and negative for antibodies against *Toxoplasma gondii*, respectively.

DNP has a dry sub-humid Mediterranean climate with strong seasonality, especially in terms of water availability to animals and vegetation. The average annual temperature is 17 ◦C, and the mean annual precipitation is 550 mm, with high intra and interannual fluctuation (170 to 1000 mm), which determine the dynamics of the marshlands [22]. During the wet seasons (winter and spring) the marshlands may flood, so ungulates concentrate and browse in the remaining uncovered scrublands. In late summer and autumn, the hardest season for ungulates due to the shortage of resources and the seasonal drought, an aggregation of wild and domestic ungulates on the ecotone and around water points occurs [21].

The territory of DNP included in this study is divided into five cattle management areas from north to south: Coto del Rey (CR), Sotos (SO), Doñana Biological Reserve (RBD), Puntal (PU) and Marismillas (MA). Free-ranging livestock is distributed through the entire park, except in the northernmost area (CR). In this area, despite the existence of a low number of horses since the last year, cattle husbandry is prohibited since 2002 as a conservation measure for the endangered Iberian lynx. A meta-population of 94 individuals of Iberian lynx currently inhabit DNP and the surrounding areas [17]. The remaining community of carnivores is comprised by red fox (*Vulpes vulpes*), Eurasian badger (*Meles meles*), Eurasian otter (*Lutra lutra*), polecat (*Mustela putorius*), European genet (*Genetta genetta*), Egyptian mongoose (*Herpestes ichneumon*), and occasionally, wild cat (*Felis silvestris silvestris*), whose presence is very scarce, probably due to the presence of a larger predator such as the Iberian lynx [23]. Furthermore, domestic carnivores including stray cats (*Felis silvestris catus*) and dogs (*Canis lupus familiaris*) are also occasionally present throughout DNP, although a population control plan of stray dogs and cats has been carried out in DNP since 2007 [24].

Finally, population control (by culling) of the wild ungulate population is practiced exclusively by park rangers as part of the park management scheme, and it is also used to carry out a health-monitoring program [14].

#### *2.2. Animal Sampling*

From October–January of 2005 to 2018 (sampling seasons 2005–2006 to 2017–2018), 423 red deer, 452 fallow deer and 698 wild boar were randomly (in terms of sex, age and health status) sampled in the population control context performed by park rangers and necropsied as part of the DNP health-monitoring program (approved by the Research Commission of DNP in accordance with management rules established by the Autonomous Government of Andalusia). Table S1 displays the sample size by species, sampling site and season, as well as the seroprevalences found. For each individual, the geographical location of the sighting was recorded through a portable GPS (Garmin Ltd., Olathe, KS, USA).

The sampling was performed according to European (EC Directive 86/609/EEC; [25]) and Spanish laws (RD 223/1988; [26]), current guidelines for the ethical use of animals in research [27], the Animal Experiment Committee of Castilla-La Mancha University and the Spanish Ethics Committee (PR-2015-03-08). Necropsies and sample collection were undertaken in the field by qualified veterinarians. During the examination, blood samples were collected into sterile plastic tubes (Vacutainer®, Becton-Dickinson, NJ, USA) from the heart, thoracic cavity, or preferably by endocranial venous sinuses puncture [28].

#### *2.3. Serological Testing*

Sera were obtained after centrifugation at 40× *g* for 5 min and stored at −20 ◦C until assayed for antibodies. Antibodies to *T. gondii* were tested using the modified agglutination test (MAT) as previously described [29]. This technique has been employed broadly for the diagnosis of antibodies against *T. gondii* in both domestic and wildlife species [1]. Two recent large studies in wild pigs and white-tailed deer in the USA added evidence for the validity of serological analysis by MAT in those species since viable *T. gondii* was isolated from a large number of seropositive animals and the rate of isolating viable parasites was positively associated with MAT titers in those studies [30–33]. Each serum sample was tested at 1:25 and 1:50 dilutions, including positive and negative controls in each test. Sera with a titer of 1:25 or higher were considered positive and those with doubtful or positive results were re-tested [12,34,35].

#### *2.4. Data Collection*

#### 2.4.1. Individual Factors

The sex and age of the animals were determined, classifying them into three age classes on the basis of dentition eruption patterns [36]: calves (<1-year-old), juveniles (1–2 years) and adults (≥3 years) for deer species, and piglets (<6 months), juveniles (0.5–2 years) and adults (>2 years) for wild boar.

Considering the well-known debilitating effect of tuberculosis (TB) progression on immune response [37], we assessed the potential effect of TB severity on the seroprevalence against *T. gondii*. For this purpose, the presence of concomitant tuberculosis-like lesions (TBL) was used as a proxy to infection by the *Mycobacterium tuberculosis* Complex (MTC), since it provides a relatively accurate diagnosis without the need for expensive laboratory confirmation [14,15]. The presence of TBL was

recorded by macroscopic inspection of the head, thoracic and mesenteric lymph nodes as well as abdominal and thoracic organs in the laboratory (see [38] for a detailed methodology). When TBL are identified in at least two of the three anatomical compartments examined (head, thorax, and abdomen) we considered the TBL as generalized, indicative of a more severe and evolved infection [38]. According to the generalized TBL status, wild ungulates were grouped in two classes: those without TBL or showing localized TBL in a single anatomical compartment, and a second class including animals with presence of generalized TBL.

#### 2.4.2. Environmental Factors

As for environmental information, several variables were included in our analysis to assess their effect on STG because of their importance to ungulate behavior, distribution, and transmission of pathogens in DNP and South Spain [15,21,39]. A grid of one hectare of surface was built, generating territorial units in which the proportional cover of dense scrub, low-clear shrubland, herbaceous grassland, woodland, bare land and watercourse vegetation were calculated for each territorial unit (see [40]). This grid was merged with the geographical location of the animals through a point sampling tool with QGIS version 2.12.1 [41]. Landcover data was obtained from Andalusia Environmental Information [42].

Given the effects reportedly associated with urban areas [43], the coast [44] and surface water on the infection risk of *T. gondii* [45], the effect of the nearest location of animals to these areas was assessed. For that purpose, we calculated the straight-line distance (m) from the exact location of each animal sampled to the nearest: urban area (DURB), small human settlements (DHS), coast line (DCOAST), water point (DWAT) and marsh-shrub ecotone (DE) (see [14,21,46]).

#### 2.4.3. Population Factors

To estimate the population density of wild ungulates we applied distance sampling methodology [47]. Every year during September, and two hours before sunset, we sampled twice 7 line transects of 10–15 km each one, distributed throughout the study area. Additionally, for wild boar we repeated the transect one hour after sunset in order to increase the sample size. We carried out the surveys during September because it is the month of maximum detectability for these species [48], and to obtain density results just before the health-monitoring program. Moreover, during September the marshland was dry, and it allowed us to sample all the habitats in DNP. The surveys were carried out from a vehicle (average speed was 10 km/h), and the perpendicular distance between animals and transect was recorded with a telemeter (Garmin Ltd., Olathe, KS, USA). The analysis were carried out using Distance Sampling 6.2 software [49] by considering stratification. We defined three strata according to its abundance and visibility: shrubland, marshland and ecotone. The data of all the years (2005–2018) were considered to estimate a detection function for each stratum, and we considered the data of each strata, sampling season and livestock management area to estimate the encounter rate and mean group size. Data were right-truncated when the probability of detection was lower than 0.15 [47]. Half-normal, uniform and hazard rate models for the detection function were fitted against the data using cosine, hermite polynomial, and simple polynomial adjustment terms, fitted sequentially. The selection of the best model was based on the Akaike's Information Criterion (AIC) [50].

The abundance of the diverse community of carnivores from DNP was monitored by means of track surveys along prefixed transects on sandy substrate according monitoring team program (ESPN-EBD-CSIC). Tracks left on moist sand over a 24 h period were tracked in transects of 1.5 m width and 2 km length, from dawn to midday and were expressed as Kilometric Abundance Index (KAI) of footprints. The surveys include 12 different transects distributed across the DNP which were repeated during three consecutive days, being cleaned daily.

As for livestock, we calculated the cattle and horse stocks per square kilometer for each sampling site and season.

#### 2.4.4. Stochastic Factors

Meteorological information (average rainfall and temperature) was collected from the meteorology station located at RBD for each sampling season [51]. In Mediterranean environments, rainfall and temperature have potential relevance to the dynamics of ungulate populations, as well as effects on the susceptibility or exposure to pathogens [15]. Specifically for *T. gondii*, both factors are key for the survival of oocyst in the environment [52]. Therefore, they were considered here for their potential effect in *T. gondii* epidemiology.

#### *2.5. Risk Factor Analysis*

Initially, collinearity between environmental and population variables was explored [53]. Given the high collinearity observed between environmental variables and with the purpose of simplifying the environmental information, a principal component analysis (PCA) was performed, obtaining two uncorrelated environmental factors: closed habitats, in which dense scrub and woodlands predominates, and watercourse habitats in which watercourse vegetation predominates.

Generalized linear mixed models (GzLMMs; binomial family) were used to assess the effect of the range of explanatory variables on the individual serological status against *T. gondii* (negative/positive). The statistical differences in STG among sampling areas (CR, SO, RBD, PU and MA) were evaluated in a first exploratory approach, the purpose of which was showing spatial differences in the serological status against *T. gondii*. A GzLMM for each species (red deer, fallow deer, and wild boar) was designed. In these models, serological status against *T. gondii* was the response variable; the sex, age class, and the sampling area were the explanatory variables. The sampling season and month were fitted in the model as random-effect factors.

Concerning the final model, it included sampling area and season as random-effect factors, since the main aim of this study was to generalize the effect of the variables included on the serological status against *T. gondii*regardless of the sampling area. Models were also performed separately for each species (red deer, fallow deer, and wild boar). The explanatory variables included individual, environmental, population and stochastic factors. Individual factors encompassed sex, age class, and tuberculosis status. Environmental factors comprised DWAT, DE, DCOAST, DHS, DURB, closed habitats, and watercourse habitats. Regarding populational factors, the population densities of wild (fallow deer, red deer, and wild boar) and domestic ungulates (cattle and horses), as well as the abundance (KAI of footprints) of wild carnivores (all together genet, Eurasian badger, red fox, and Egyptian mongoose, and separately, the abundance of Iberian lynx) were included. Finally, the stochastic factors were the previous seasons' rainfall and temperature. The two-way interactions between individual-stochastic factors separately (sex-age, and previous season's rainfall-temperature) and all together (sex-rainfall and age-rainfall), as well as between population-individual factors (density-age), and population-stochastic factors (previous season´s rainfall-density) were also included in the models. For the GzLMMs, a binomial error and a logit link function were used. Stepwise selection processes for the final models were performed on the basis of the AIC [50] (Table S2). Furthermore, the assumptions of binomial GzLMMs were met in all the best models selected [53]. The predicted probabilities of serological response to *T. gondii* obtained from these models were used to represent the results. Finally, cross-correlations and autocorrelations between STG and its predicted response probability between the different species were carried out to explore similarities of temporal patterns [54].

The statistical analyses were performed using R-studio software version 4.0.2 [55]. All models were performed using the R package glmer [56]. Significant *p*-values were set at 0.05.

#### **3. Results**

#### *3.1. General*

The STG (MAT ≥ 1:25; % ± confidence interval (CI) 95%) in wild boar was 39.0 ± 3.3 (*n* = 698), followed by red deer 30.7 ± 4.4 (*n* = 423), and fallow deer 29.7 ± 4.2 (*n* = 452). Among the

seropositive animals, titers of 1:25 were detected in 34.5% wild boar, 27.7% red deer, and 55.2% fallow deer, whereas titers ≥ 1:50 were found in 72.3% red deer, 65.5% wild boar, and 44.8% fallow deer. We observed increasing age trends in STG in all wild ungulate species, except for wild boar females (Figure 2a) since, interestingly, piglets already showed high STG. With respect to gender, males tended to present higher STG than females in deer species (32.1–28.8% and 32–26.4% for red deer and fallow deer, respectively), whereas the opposite was observed in wild boar (STG = 26.2% for males, STG = 29.7% for females; Figure 2a; see statistical comparisons below).

**Figure 2.** (**a**) Seroprevalence (±CI 95%) of *Toxoplasma gondii* depending on age class and sex in red deer, fallow deer and wild boar (**b**) Seroprevalence (±CI 95%) of *T. gondii* obtained from selected generalized linear mixed models (GzLMMs) for the species studied depending on the sampling area, from north to south areas (see Figure 1 for a map of the areas with their full names).

Contrasted STG were apparent among areas, which was consistent across species. In this sense, seroprevalence decreased from north to south, more markedly in red deer (Figure 2b). The temporal evolution of STG, and trends in the estimated density/abundance of each different species are summarized in Figure 3a,b, respectively. In this regard, the STG exhibited strong annual fluctuations, mostly in deer species (Figure 3a). Actually, it is noteworthy the significant decrease of STG in these species since the season 2013–2014. No autocorrelations or cross-correlations were observed.

**Figure 3.** Temporal trend of the (**a**) seroprevalence of *Toxoplasma gondii* (±CI 95%), and (**b**) population density of red deer, fallow deer, wild boar, and cattle (individuals/km2), and Kilometric Abundance Index of Iberian lynx.

#### *3.2. Factors Determining the Seroprevalence of T. gondii*

There were statistically significant differences in the STG between sampling areas for all the wild ungulates species (red deer, *F* = 13.4, df = 410, *p* ≤ 0.01; fallow deer, *F* = 4.5, df = 436, *p* ≤ 0.01; and wild boar *F* = 10.3, df = 682, *p* ≤ 0.01), confirming the north to south spatial decreasing gradient (Figure 2b).

The results of the GzLMMs on the status against *T. gondii*, incorporating broader environmental and populational information are shown in Table 1. The conditional R<sup>2</sup> obtained from these models were 0.37, 0.53 and 0.25, for red deer, fallow deer, and wild boar, respectively.





negative, "\*" represents interactions among explanatory variables.

#### 3.2.1. Individual Factors

The sex and age classes were statistically significant factors in the models on red deer and wild boar. However, no sex or age-related differences were found in fallow deer. Regarding red deer, females had lower STG than males, and irrespective of sex, the pattern increased with the age. Concerning wild boar, different sex-related age patterns were shown, increasing for males but not for females (the sex by age interaction was marginally significant).

Regarding TB status, the prevalence of TBL (% ± CI 95%) for wild boar, red deer and fallow deer were 77.4 ± 3.1, 42.5 ± 4.7, and 16.4 ± 3.7, respectively. Wild boar had the highest prevalence of generalized TBL (% ± CI 95%; 27.73 ± 3.5), followed by red deer (17.7 ± 4.2) and fallow deer (8.11 ± 2.9). The STG was higher in red deer and fallow deer presenting generalized TBL (Figure 4) compared to generalized TBL-free individuals (TBL-free plus not generalized TBL positive). As for the wild boar, a complementary model was performed with the purpose of exploring the effect of the presence of TBL (positive or negative), since no effect of the presence of generalized TBL was observed. In this model, wild boar presenting TBL showed higher STG than negative individuals (*F* = 8.96, df = 695, *p* = 0.05).

**Figure 4.** Seroprevalence (±CI 95%) of *Toxoplasma gondii* depending on the tuberculosis status in red deer and fallow deer (interpreted as positive animals with generalized presence of tuberculosis-like lesions (TBL), and in wild boar (interpreted as positive animals with presence of TBL).

#### 3.2.2. Environmental Factors

The further to the coastline, the higher the STG was (see e.g., Figure 5a for red deer) in all the species. Moreover, the closer to small human settlements, the higher the STG for wild boar was (Figure 5b). The increased availability of closed habitat significantly associated with lower STG in fallow deer (Figure 5c).

**Figure 5.** Seroprevalence (±CI 95%, represented by the shaded band) of *Toxoplasma gondii* obtained from selected generalized linear mixed models (GzLMMs) in (**a**) red deer depending on the distance to the coast line (m), (**b**) wild boar depending on the distance to the nearest human settlement (m), (**c**) fallow deer depending on the cover level of closed habitats, measured according to the principal component analysis (PCA) scores from axis 1, (**d**) red deer depending on the Kilometric Abundance Index of carnivores species (KAI), (**e**) fallow deer depending on the Kilometric Abundance Index of Iberian lynx (KAI), and (**f**) wild boar depending on the density of fallow deer (individuals/km2).

#### 3.2.3. Population Factors

The abundance of carnivores significantly and positively associated with the exposure to *T. gondii* in red deer (Figure 5d), and similarly, the Iberian lynx abundance positively associated with the seropositivity to this parasite in fallow deer (Figure 5e). The fallow deer density negatively associated with the STG in wild boar (Figure 5f) and fallow deer, but positively in the case of red deer. As for red deer, a negative association was found between STG and density. In contrast, wild boar showed higher STG at higher densities.

#### 3.2.4. Stochastic Factors

To represent our results, and considering the mean values obtained, we established the following categories of rainfall and temperature for displaying results: low rainfall (≤521.10 mm), high rainfall (>521.10 mm), low temperature (≤17.5 ◦C) and high temperature (>17.5 ◦C). Lower annual temperature was associated with higher STG in red deer. Furthermore, higher annual rainfall was associated with higher seropositivity to *T. gondii* in red deer. Regarding fallow deer, the interaction between rainfall and temperature was significant: overall, there was a trend to higher STG in cold years, and this pattern was more marked in dry years (see Figure 6a). Rainy years were statistically associated with higher STG in male red deer, but not in females (significant rainfall by sex interaction, Figure 6b). Concerning the wild boar, the rainfall was positively associated with the STG in juveniles, but this effect was not shown in other age classes (significant annual rainfall by age interaction, Figure 6c).

**Figure 6.** Seroprevalence (±CI 95%) of *Toxoplasma gondii* obtained from selected generalized linear mixed models (GzLMMs) in (**a**) fallow deer depending on the interaction between annual rainfall (mm) and temperature (◦C), (**b**) red deer depending on the interaction between annual rainfall (mm) and sex, and (**c**) wild boar depending on the interaction between annual rainfall (mm) and age.

#### **4. Discussion**

#### *4.1. General Patterns of the Seroprevalence of T. gondii*

The STG reported in the present study oscillated considerably (from 29.7 to 39%) between the three species tested and sharing the same environment. This may be caused by differences in the susceptibility, the feeding behavior, or the habitat use of those species determining the exposure [4,52]. The seroprevalence detected in wild boar (39%) concurs with studies conducted in Europe [3,57–59]. However, most studies from European countries, STG in wild boar ranged from 6 to 25% [6,60,61]. In this regard, trophic relationships by predation and/or scavenging of a wide range of warm-blooded animals of the DNP may operate.

Concerning deer, the STG obtained (30.7% and 29.7%, for red deer and fallow deer, respectively) are also in accordance with those reported in the literature over Europe in general [62,63] and Spain in particular [8,34], ranging from 10.5 to 48%. Specifically for fallow deer, STG (29.7%) were among the highest reported in European studies that were mainly focused on Spain [8,34,63]. It was only exceeded by the rates obtained by Calero-Bernal et al. [64] in south-central Spain (reaching the 48%). The higher rate of movement between areas reported for fallow deer in DNP may imply higher exposure to *T. gondii*, explaining the high STG observed in this species [39].

Overall, the wild ungulate host community of DNP showed higher STG compared with those reported in the literature of the European and Iberian contexts [34,61,62]. Mechanisms determining seroprevalence in different host species of the studied community are related to the life cycle of *T. gondii*, which involves both an environmental and a trophic transmission route (i.e., trophic relationships among potential hosts of the community; [65]). Terrestrial herbivores should have the lowest *T. gondii* exposure, only through the ingestion of oocyst-contaminated vegetation, soil and/or drinking water. In DNP, the environmental presence of oocyst excreted by felids may be playing a major role (see below). The high biodiversity inhabiting DNP, which provides a wide range of hosts and ecological/epidemiological niches, and the optimal climatic conditions for the survival of the oocysts may favor the spread of the parasite in the DNP host community.

The specific role of the different factors in a long-term perspective is also detailed further in the discussion. Interestingly, while the STG exhibited strong annual fluctuations, mostly in deer species, it was more stable in wild boar. In populations from The Netherlands, seroprevalence in wild boar similarly established at around 35% [66]. Whereas authors stated that the actual mechanisms behind the stabilization requires further investigation, an epidemiological SIS-model that included a reversion to susceptible after infection (with loss of antibodies that may have been preceded by a loss of tissue cysts), fitted the data much better.

The north to south spatial gradient observed is similar to that exhibited by the prevalence of other shared pathogens tested in the wild ungulate community of DNP in previous studies [14,67]. This pattern may relate to spatial variation in the contamination of the environment by *T. gondii* oocyst. The main large human settlements around DNP are concentrated in the northern part of the park, with a subsequent higher presence of peri-domestic cat populations [68], which may contaminate the environment with oocysts. Iberian lynx populations also show a north to south decreasing pattern in DNP [69], contributing to a lesser extent to this contamination. *T. gondii* oocysts were found in feces of 17% of cats sharing a habitat with Iberian lynx [19]. In this regard, feral cats are the more likely reservoir host of parasites affecting the Iberian lynx and wildlife species in general, especially in areas where feral cats are abundant and widespread such as DNP surroundings [70,71].

#### *4.2. Individual Factors*

In wild boar, overall, females had significantly higher STG than males. This result is in accordance with previous studies in this species [3,57–59]. However, the age pattern observed was opposite to that of males. Several authors have reported that no statistically significant effect of age on STG in wild boar was observed [3,72,73], whereas only one study found a significantly higher prevalence in adult wild boar [59]. Nevertheless, we must consider sex by age pattern to understand the differences. The increased exposure to *T. gondii* through life, together with the high persistence of antibodies against *T. gondii*, could explain the age pattern found in the STG in males. Even so, in females, the decline in the STG rather than indicating a decrease of exposure to the parasite, may be indicative of a subtle equilibrium of chronic infection and reduced specific humoral response that is not detected. Ecological and evolutive aspects determining differences in exposure may be behind this pattern. However, further research is required. Finally, piglets exhibiting high seroprevalences could be explained by maternal-derived antibodies, whose titers depend on those shown by sows, according to previous studies [10].

As for the red deer, the STG was significantly higher in adult individuals, which has been previously reported in many studies on *T. gondii* [8,57,58,73]. An apparent similar trend, but not significant, was observed in fallow deer (Figure 2a). The increased exposure to *T. gondii* along life together with the high persistence of antibodies against *T. gondii* could explain the age pattern found in deer species.

Concerning concomitant TB infection, overall, the positive TB status of the animals significantly associated with STG in all the species studied. For deer species, the generalized presence of TBL was relevant, as well as the presence of TBL for wild boar. There are several studies on TB-*T. gondii* co-infection in humans [74,75] but not in animals. The relationship between TB and STG observed may be mediated by exposure over time (age-related) and environment. In the latter case, the conditions favoring the persistence of MTC and *T. gondii* oocyst in the environment are similar (see Sections 4.3 and 4.5).

#### *4.3. Environmental Factors*

The distance to the coastline significantly positively associated with STG for all wild ungulates. Despite the usually reported contamination of seawater with *T. gondii* [44], less favorable conditions to the survival of oocysts may occur in the surroundings of the coastline, according to previous studies developed in northern Spain [4]. These conditions are mainly the high temperatures reached in the sandy soils which favor the desiccation of the oocysts. The availability of closed habitat (more covered by vegetation) negatively associated with STG in fallow deer (see Figure 5c). This species typically uses and occupies meadows in the park, and individuals sampled in more densely covered areas may have experienced lower exposure to *T. gondii*. By contrast, individuals of the other ungulate species combine the use of both types of habitats which may determine the absence of this effect.

The closer to small human settlements the higher the STG was for wild boar (see Figure 5b). This result is in accordance with previous studies where human settlements have become areas of epidemiological relevance for *T. gondii* infection, mainly mediated by the presence of peri-domestic species [4,45]. There are several human dwellings inside the park, around which peri-domestic cats could settle and consequently contaminate the surroundings of these areas with oocysts. Moreover, wild boar could become infected through the consumption of food scraps from garbage located in these small settlements.

#### *4.4. Population Factors*

The abundances of Iberian lynx and other carnivores were statistically positively associated with STG in deer species. The presence and abundance of felids have been considered a relevant risk factor associated with *T. gondii* in livestock and wildlife worldwide [1,7–10]. It has been reported that Iberian lynx could prey on fallow deer and less frequently on juveniles of red deer during seasons of rabbit´s scarceness in DNP, especially in winter and autumn, reaching 5–10% of the biomass in the Iberian lynx diet [76]. This leads to increased environmental contamination with *T. gondii* in areas with presence of wild ungulates. Furthermore, *T. gondii* infected red deer carcasses could pose a potential risk of *T. gondii* infection for carnivores species via scavenging and may therefore play a role as an amplifier of infection in the community [63]. This allows *T. gondii* to finish its life cycle, perpetuating its maintenance in the DNP host community. Little is known about the abundance of small mammals, as well as their STG in DNP. Further studies should focus on investigating their role in the maintenance and spread of this parasite in DNP.

STG significantly increases with density in wild boar. Several authors have described *T. gondii* as a density-dependent parasite for swine [3,45]. Density, together with ecological and behavioral factors typical of this species, could determine increased exposure by wild boar. Scavenging (including cannibalism) may be increased in high density situations when resources are scarce. In the dry season, both the availability of carcasses and exacerbated cannibalism behavior [77], but also ingestion of rodents and birds occur [78]. The negative relationship found both in red deer and fallow deer between density and the risk to test positive can be explained by high recruitment of susceptible individuals (non-infected offspring) associated with high density years. Moreover, unlike for wild boar, this negative association indicates that no density-dependent effect in *T. gondii* infection occurs in deer species. These species become infected only through the ingestion of water or food contaminated with sporulated *T. gondii* oocysts, so no direct transmission route exists as in the case of wild boar, which possesses scavenging habits. However, the positive relationship showed between STG in red deer and fallow deer density may be mediated by an increased susceptibility and/or exposure at high densities of ungulates due to the competition for scarce resources, but the exact mechanism deserves further research.

#### *4.5. Stochastic Factors*

Temperature was a significant factor for red deer, displaying the lower STG during the following seasons to the warmest ones. Furthermore, rainfall significantly interacted with temperature to explain STG in fallow deer, so that the effect of the previous season´s temperature on STG was more marked when the previous season was dry. Previous driest and warmest seasons markedly associated with a lower STG (see Figure 6a). Drought together with warm temperatures leads to higher rates of evaporation and the subsequent desiccation, limiting the survival and sporulation of the oocysts in the environment. As consequence, the exposure of herbivores to infective *T. gondii* oocyst decreases [34,35,52].

Additionally, the interaction between rainfall and individual factors was significantly associated with the STG in red deer and wild boar. Concerning red deer, higher previous annual rainfall was related to higher STG in stags, but not in females. This effect of rainy years on the prevalence of pathogens exhibited by males with respect to females have been shown in previous studies on TB in DNP, and may relate to increased exposure and/or susceptibility mediated by sexual behavior and life history traits [14]. An immunosuppressant effect of the intense rut that typically occurs in rainy years has been suggested. Intense rut implies greatest investments by red deer stags in terms of reproductive effort (testosterone metabolite levels and sexual signals), and the conflict between the immune response and the reproductive effort this species is well known [79,80]. *T. gondii* tissue cysts in many organs, including viscera, are believed to persist for the lifetime of the host. In addition, deer rutting typically occurs in the ecotone, which provides excellent wet conditions for oocysts to persist. Despite the same risk derived from the reproductive efforts during the rut exist for fallow deer, no sex-dependent effect of rainfall was observed in STG in this species, which is not surprising since rutting in fallow deer takes place later in Autumn, when rainfall conditions normally are less determinant.

As for wild boar, the positive effect of the rainfall on STG was more markedly in juveniles than adults and piglets. The early dispersal behavior of young males from the natal area may leads to higher exposure to *T. gondii* [81]. This, together with the increased survival of oocysts would give rise to higher infection rates in this age group [35,52].

#### **5. Conclusions**

This study provides evidence that factors behind the risk of *T. gondii* infection in wild boar, red deer and fallow deer are related to environmental and trophic transmission routes, so as to

individual, population and species characteristics. We provided evidence for most of these relationships (e.g., climate or population mediated) and trends. Concomitant pattern among species, indicated that overall, drivers of risk also operated at the community level. However, this research raised several questions that deserve further research. Approximately one-third of the human world's population is chronically infected while seroprevalence tends to decrease since the early 1960s in many countries [82]. As this decline in seroprevalence leads to loss of immunity, it becomes more relevant for the identification of the epidemiological role of wild host and the understanding of the epidemiology and ecology of *T. gondii* infection in wild host communities and at their interfaces with livestock and human. Thus, game meat, in particular venison, consumption should not be neglected as a public health risk for humans, with the subsequent impact to the public health [31]. For these purposes, addressing host population, community and environmental factors at broad temporal scale is key.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2076-2615/10/12/2349/s1, Table S1: Sample size and seroprevalence of *Toxoplasma gondii* by species, season and sampling site in wild ungulates; Table S2: Summary of the stepwise model selection procedure, based on the AIC, used to explain the serological status against *Toxoplasma gondii*.

**Author Contributions:** Conceptualization, J.V., I.G.-B. and P.A.; Methodology, J.V., I.G.-B., P.A., and F.C.; Validation, J.V., I.G.-B., P.A., and F.C.; Formal Analysis, J.V., P.B., F.C. and P.P.; Investigation, P.B., P.P., F.C., S.J.-R., and D.C.-T., Resources, J.V. and P.A.; Data Curation, P.B., F.C. and P.P.; Writing—Original Draft Preparation, P.B. and J.V.; Writing—Review and Editing, P.B., J.V., P.A., P.P., F.C., S.J.-R., S.A., J.P.D. and D.C.-T.; Visualization, P.B.; Supervision, J.V., I.G.-B. and P.A.; Project Administration, J.V. and I.G.-B.; Funding Acquisition, J.V. and P.A. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Ministerio de Economía y Competitividad (MINECO; AEI/FEDER, UE; AGL2016-76358-R). S.J. is co-supported by the UCLM and the European Social Fund (2018/12504). P.P. received support from the MINECO (FPU/16/00039).

**Acknowledgments:** The authors would like to thank the EBD-CSIC monitoring team and their colleagues at the IREC (UCLM, Spain), Doñana National Park and Doñana ICTS-RBD, which provided logistic and technical support and the data essential for carrying out this work, and to all those who participated in the fieldwork and data collection of the wild ungulates, with special thanks to the park ranger Jose Antonio Muriel, and María de los Ángeles Risalde for her kind help with this study.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**


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

© 2020 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 (http://creativecommons.org/licenses/by/4.0/).

## *Communication* **Intestinal Helminths in Wild Rodents from Native Forest and Exotic Pine Plantations (***Pinus radiata***) in Central Chile**

**Maira Riquelme 1, Rodrigo Salgado 1, Javier A. Simonetti 2, Carlos Landaeta-Aqueveque 3, Fernando Fredes <sup>4</sup> and André V. Rubio 1,\***


**Simple Summary:** Land-use changes are one of the most important drivers of zoonotic disease risk in humans, including helminths of wildlife origin. In this paper, we investigated the presence and prevalence of intestinal helminths in wild rodents, comparing this parasitism between a native forest and exotic Monterey pine plantations (adult and young plantations) in central Chile. By analyzing 1091 fecal samples of a variety of rodent species sampled over two years, we recorded several helminth families and genera, some of them potentially zoonotic. We did not find differences in the prevalence of helminths between habitat types, but other factors (rodent species and season of the year) were relevant to explain changes in helminth prevalence. Given that Monterey pine plantations are one of the most important forestry plantations worldwide, and due to the detection of potentially zoonotic helminths, more research should be conducted in this study area and elsewhere in order to better understand the effect of pine plantations on parasites and pathogens in rodents and other wildlife hosts.

**Abstract:** Native forests have been replaced by forestry plantations worldwide, impacting biodiversity. However, the effect of this anthropogenic land-use change on parasitism is poorly understood. One of the most important land-use change in Chile is the replacement of native forests by Monterey pine (*Pinus radiata*) plantations. In this study, we analyzed the parasitism (presence and prevalence) of intestinal helminths from fecal samples of wild rodents in three habitat types: native forests and adult and young pine plantations in central Chile. Small mammals were sampled seasonally for two years, and a total of 1091 fecal samples from seven small mammal species were analyzed using coprological analysis. We found several helminth families and genera, some of them potentially zoonotic. In addition, new rodent–parasite associations were reported for the first time. The overall helminth prevalence was 16.95%, and an effect of habitat type on prevalence was not observed. Other factors were more relevant for prevalence such rodent species for *Hymenolepis* sp. and season for *Physaloptera* sp. Our findings indicate that pine plantations do not increase helminth prevalence in rodents compared to native forests.

**Keywords:** Chile; helminths; land conversion; Rodentia; small mammals; zoonosis; wildlife

#### **1. Introduction**

Anthropogenic land-use change can impact biodiversity and human health, being a major driver of biodiversity loss and zoonotic disease emergence [1] Land-use change can modify host–parasite interactions through a variety of mechanisms that involve changes in

**Citation:** Riquelme, M.; Salgado, R.; Simonetti, J.A.; Landaeta-Aqueveque, C.; Fredes, F.; Rubio, A.V. Intestinal Helminths in Wild Rodents from Native Forest and Exotic Pine Plantations (*Pinus radiata*) in Central Chile. *Animals* **2021**, *11*, 384. https://doi.org/10.3390/ani11020384

Academic Editors: Rafael Calero-Bernal and Ignacio Garcia-Bocanegra Received: 23 January 2021 Accepted: 31 January 2021 Published: 3 February 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/).

abundance, behavior and immune response of hosts, as well as the composition and structure of host community [2]. Additionally, land-use change can modify abiotic conditions, which, in turn, may influence the transmission and life cycle of parasites such as helminths, which have several life stages with close contact with the environment [3].

Rodents are one of the most important reservoir hosts of zoonotic pathogens [4], and reservoir species are commonly found in high abundance in anthropogenic-modified habitats (e.g., agricultural lands, pasture lands) [5]. Wild rodents can act as definitive, intermediate and paratenic hosts of several endoparasites (helminths and protozoa) with zoonotic potential, such as *Capillaria hepatica* Bancroft 1893, *Cryptosporidium* spp., *Giardia* spp., *Toxoplasma gondii* Nicolle and Manceaux 1908, *Schistosoma* spp., etc. [6,7]. In addition, helminths are the most prevalent group of macroparasites in rodents [8]. Therefore, the study of the effect of anthropogenic land-use change on gastrointestinal helminths in rodents is needed to better understand host–parasite interactions in wildlife with public health implications.

The forest industry constitutes an important economic activity in tropical and temperate regions of developing countries, replacing large areas of native forests and grasslands by plantations of fast-growing exotic trees [9]. Despite the relevance of plantations worldwide, the study of the effect of monoculture plantations on wildlife parasites has been scarce [2], but some studies have found that plantations may increase the abundance of some ectoparasitic mites in rodents [10] as well as parasite richness in several wild mammals [11]. An important plantation in the global forest industry is Monterey pine, *Pinus radiata* (D. Don, 1836), accounting for 32% of productive plantations worldwide [12]. In Chile, Monterey pine plantations are one of the most important land uses in south-central regions [13], covering approximately 1.9 million ha and accounting for 68% of forestry plantations of the country (http://www.corma.cl). This forest plantation modifies biodiversity, including the composition and structure of rodents [14–16], as well as the prevalence and load of mites, *Ornithonyssus* sp. (Mesostigmata), in rodent hosts [10].

Several gastrointestinal helminths have been detected in native and exotic rodents in Chile [17–21]. However, to our knowledge, the effect of land use on intestinal helminths of rodents inhabiting monoculture plantations has not been investigated. Therefore, the aim of this study was to describe the presence and compare the prevalence of gastrointestinal helminths in feces from wild rodents inhabiting forestry plantations and native forests in a highly disturbed landscape from central Chile.

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

#### *2.1. Ethical Statement*

All procedures for trapping and handling rodents followed the guidelines of the American Society of Mammalogists for the use of wild mammals in research [22] and followed standardized safety guidelines recommended by the Centre for Disease Control and Prevention [23]. Sampling procedures were authorized by the Servicio Agrícola y Ganadero (SAG; Chilean Fish and Wildlife Service) (License No. 6831/2015) and the Ethics Committee of the Faculty of Science, University of Chile.

#### *2.2. Study Area and Sampling Sites*

The study was conducted in Tregualemu (35◦58 S, 072◦44 W), located in the coastal range of the Maule Region in central Chile. The landscape contains interspersed remnants of native forest and extensive stands of Monterey pine of different ages [24]. The native forest of the study area includes the Queules National Reserve and contiguous forests that make up an area of 600 ha. The forest is composed of *Nothofagus glauca* (R. Phil.) Krasser 1896 and *N. obliqua* (Mirb.) Oerst, 1871, as dominant species and the evergreen *Cryptocarya alba* (Molina) Looser 1950 and *Peumus boldus* Molina, 1782. Monterey pine plantations in this area are managed by maintaining a developed understory or multiple vegetation strata within monocultures [25]. Understory vegetation in mature pine plantation contains *Aristotelia chilensis* Stunz 1914, *P. boldus* with exotic species such as *Genista monspessulana* (L.) L.A.S.Johnson 1962 and *Rubus ulmifolius* Schott, 1818 [26].

Small mammals were sampled in three dominant habitat types: (1) native forest, (2) adult pine plantations (>15 years old) and (3) young pine plantations (3–4 years old). A total of 12 sampling sites were selected: three sites each in native forest and adult pine plantations and six in young pine plantations. Each site was separated by at least 400 m from each other (mean distance between sites = 2025 m).

#### *2.3. Small Mammals and Fecal Sampling*

Small mammals were sampled once each season (summer: January; autumn: April– May; winter: July–August; spring: November) over 2 years (2016–2017). Each site was sampled in all periods, except for one adult pine plantation and one young pine plantation site that were not sampled in autumn 2017 due to logistical constraints. At each site, live Sherman traps were placed separated by 10 m, forming a 7 × 10 grid (70 traps), for four consecutive nights. All traps were baited with rolled oats and vanilla essence and checked daily at dawn. After capture, animals were identified to species [27], measured, weighed, sexed and marked with uniquely numbered ear tags (National Band and Tag Co., Newport, KY). Fecal samples were collected directly from the animal or from the trap and then preserved in 70% ethanol in 2-mL microtubes. After handling, animals were released in the same place they were captured.

#### *2.4. Parasitological Analysis*

The presence of helminth eggs and larvae in feces was analyzed using the routine coprological method of modified Telemann [28,29]. Briefly, each stool sample was placed in a 15-mL Falcon tube with 10 mL of 70% ethanol and 2 mL of diethyl ether and centrifuged at 2000 rpm for 10 min at room temperature. After centrifugation, the supernatant was removed and 100 μL of the sample was placed on a glass microscope slide, and then, a 24 × 24-mm cover slip was placed on the surface of the sample. Finally, slides were scanned under 10× and 40× objective lenses of a light microscope. Identification of helminths was based on published taxonomic keys [28] and using references of eggs collected from adult helminths by Landaeta-Aqueveque et al. [30]. Laboratory analyses were performed at the Parasitology Laboratory, Faculty of Veterinary Science, University of Chile, and at the Parasitology Laboratory, Faculty of Veterinary Science, University of Concepcion.

#### *2.5. Data Analysis*

Prevalence (the proportion of positive animals divided by the total number tested) was calculated and 95% confidence intervals (CI) were determined using the Clopper–Pearson method [31] for each rodent species. These analyses were performed using the software Quantitative Parasitology v.3.0 [31]. Generalized linear models (GLMs) with binomial distribution and logit function were used to identify variables that may explain prevalence. Models were built using all parasites combined and also using the most frequent helminths separately. The explanatory variables analyzed were habitat type, season and rodent species. For each multivariate model, we calculated Akaike information criterion adjusted for sample size (AICc), differences in AICc and Akaike weights (wr). The best models were based on the lowest AICc values [32]. Selected models were validated by the Hosmer– Lemeshow goodness-of-fit test. We performed these analyses with R software (R Core Team 2017), including packages "MuMIn" [33] and "ResourceSelection" [34].

#### **3. Results**

#### *3.1. Small Mammal Sampling*

A total of 1962 individuals were captured. Small mammals belonged to seven rodent species and a marsupial, the elegant fat-tailed mouse opossum *Thylamys elegans* (Waterhouse, 1839). The most common species captured were *Abrothrix olivacea* Waterhouse, 1837 (56.5%), *Oligoryzomys longicaudatus* Bennett, 1832 (20.3%) and *Abrothrix longipilis* Waterhouse, 1837 (17.5%), followed by *Phyllotis darwini* Waterhouse, 1837 (2.7%) and the introduced *Rattus rattus* Linnaeus, 1758 (2.3%). Other small mammals captured were

*Irenomys tarsalis* Philippi, 1900, *Octodon bridgesi* Waterhouse 1844 and *T. elegans* (0.7% all three species pooled).

#### *3.2. Parasitological Analysis*

In total, 1091 fecal samples were collected and 185 samples (16.95%) were positive to any helminth egg or larvae and were found in the majority of small mammal species analyzed (Tables 1 and 2). The helminths identified belonged to three phyla: Acanthocephala, Nematoda and Platyhelminthes (Table 1; Figure 1). Another helminth egg was found in several samples from both *Abrothrix* species but could not be identified (Table 1; Figure 1). A larva was also detected in some samples (Table 1) but could not be identified through microscope. The most frequent helminth eggs were *Hymenolepis* sp. (6.1%) and *Physaloptera* sp. (3.5%). Prevalence of helminths in samples of the most abundant rodent species was *A. olivacea* (17.7%), *A. longipilis* (22.2%), *O. longicaudatus* (6.7%) and *P. darwini* (20%) (Table 2). Most helminths were found in both native forest and pine plantations, except Strongylida, which was found only in the native forest, and *Moniliformis* sp., which was only recorded in *Abrothrix* species in both adult and young pine plantations (Table 3).

**Table 1.** Presence of helminth eggs and larvae from fecal samples of small mammals. N = sample size. Number in parentheses is the number of positive samples.


**Table 2.** Helminths in fecal samples of small mammals at three habitat types. Sample size (N), number of positive individuals (+), prevalence (P) and 95% confidence intervals (CI) are reported.


**Figure 1.** Helminth eggs recorded from fecal samples of rodents. All images were obtained at 40×. (**A**) *Capillaria* sp.; (**B**) *Hymenolepis* sp.; (**C**) Strongylida; (**D**) *Syphacia* sp.; (**E**) *Moniliformis* sp.; (**F**) *Physaloptera* sp.; (**G**) unidentified egg; (**H**) unidentified larva.

**Table 3.** Number of infected host individuals reported by helminth species, host species (most abundant species) and by habitat type. NF = native forest; AP = adult pine plantation; YP = Young pine plantation; *n* = number sample size. Numbers in parentheses indicate prevalence.


#### *3.3. Data Analysis*

GLM analyses were conducted including the results of the three most abundant rodent species (*A. olivacea*, *A. longipilis* and *O. longicaudatus*). *Phyllotis darwini* was not included because this rodent was only present in young pine plantations [15]. The best GLM model selected according to AICc values for all helminths included season and host species as explanatory variables (Table S1). This model indicates that prevalence of helminths was higher in spring compared to autumn and summer, and *O. longicaudatus* had lower prevalence compared to both *Abrothrix* species (Table 4). Habitat type was included in the third best GLM model (Table S1), showing that helminth prevalence was not significantly different between native forest and adult pine plantations (estimate = −0.19, standard error (SE) = 0.31, *p* = 0.54) or young pine plantations (estimate = −0.16, SE = 0.30, *p* = 0.59). Regarding GLM analyses using the most common helminth species, the best model for *Hymenolepis* sp. included only host species as an explanatory variable (Table S2). *Oligoryzomys longicaudatus* had lower prevalence of *Hymenolepis* sp. compared to both *Abrothrix* species (Table 5). Habitat type was included in the second best GLM model (Table S2), indicating that *Hymenolepis* sp. prevalence was not significantly different between native forest and adult pine plantations (estimate = 0.25, SE = 0.55, *p* = 0.65) or young pine plantations (estimate = 0.46, SE = 0.55, *p* = 0.39). For *Physaloptera* sp., the best model included the variables season and host species (Table S3), indicating that this parasite had lower prevalence in autumn compared to all other seasons (Table 6). Similar to previous results, habitat type was included in the second best GLM model (Table S3), showing that *Physaloptera* sp. prevalence was not significantly different between native forest and adult pine plantations (estimate = −1.02, SE = 0.78, *p* = 0.19) or young pine plantations (estimate = 0.52, SE = 0.56, *p* = 0.35).

**Table 4.** Results of the best generalized linear model (GLM) that predicted the probability of positivity for helminths in rodents. Spring and *O. longicaudatus* were used as the reference categories. Hosmer–Lemeshow test: χ<sup>2</sup> = 10.96, *p* = 0.20 (*p* > 0.05 is interpreted as fit).


\* *p* values < 0.05.

**Table 5.** Results of the best GLM that predicted the probability of positivity for *Hymenolepis* sp. in rodents. *O. longicaudatus* was used as the reference category. Hosmer–Lemeshow test: χ<sup>2</sup> = 5.16, *p* = 0.74 (*p* > 0.05 is interpreted as fit).


\* *p* values < 0.05.

**Table 6.** Results of the best GLM that predicted the probability of positivity for *Physaloptera* sp. in rodents. Autumn and *O. longicaudatus* were used as the reference categories. Hosmer-Lemeshow test: χ<sup>2</sup> = 2.43, *p* = 0.97 (*p* > 0.05 is interpreted as fit).


\* *p* values < 0.05.

#### **4. Discussion**

In this study, we recorded helminth families or genera that are associated with rodents in Chile and elsewhere [17,35,36]. Furthermore, some of the helminth genera recorded may belong to zoonotic species, such as *Hymenolepis diminuta* Rudolphi 1819, *H. nana* Siebold 1852, *Syphacia obvelata* Rudolphi 1802 and *S. muris* Yamaguti 1941, which have been reported in Chile [30,37]. The following host–parasite associations are registered for the first time: *P. darwini–Physaloptera* sp., *A. longipilis–Moniliformis* sp., *A. longipilis–Physaloptera* sp., *A. longipilis–Capillaria* sp. and *O. longicaudatus–Capillaria* sp. We used traditional routine coprological analyses to detect helminths in feces, which have been successfully employed to address the effect of land use on parasites elsewhere [11,35,38]. However, further studies are needed to identify the helminth species present in the study area which should focus on identification of adult helminths [30] and/or species identification through molecular tools, which has been scarcely used for helminth egg identification from fecal samples of rodents [39].

The findings of some helminths only in native forest (Strongylida in *O. longicaudatus*) or only in pine plantations (*Moniliformis* in *Abrothrix* spp.) might be a consequence of the low prevalence of these parasites (0.37–1.3%, respectively). Our results indicate that helminth prevalence in wild rodents varies among seasons and host species but is not affected by habitat type. Therefore, the replacement of native forest by pine plantations in the study area would not be a driver for increased transmission of helminths with zoonotic potential in rodents. A previous investigation in the study area has shown a similar outcome, in which the prevalence of mites (*Androlaelaps* sp.) parasitizing *A. olivacea* was similar between the native forest and pine plantations [10]. On the other hand, the prevalence of other mite species (*Ornithonyssus* sp.) on the same host was increased in young pine plantations [10]. These contrasting results show that the effect of Monterey pine plantations on parasitism is parasite-dependent, which might be consequence of differences in the ecology, life cycle and other attributes between parasite species and their relation to the environment and their hosts. In fact, the effects of land-use change on helminth parasitism depend on the life history traits of each host–parasite association, as shown in several studies across the world [3,35,40,41].

*Hymenolepis* sp. was the most common helminth found, with higher prevalence in both *Abrothrix* species compared to *O. longicaudatus*. Most species of the genus *Hymenolepis* have an indirect cycle, including free living and ecto-parasite arthropods as intermediate hosts [42]. Therefore, differences in prevalence among rodents could be related to food habits, where the diets of *A. olivacea* and *A. longipilis* include a higher composition of insects (up to 25–32%) [27] compared to *O. longicaudatus,* for which invertebrates represent a minor component of the diet (5−10%) [27,43]. On the other hand, *Physaloptera* sp., the second most prevalent helminth, presented a seasonal variation, with lower prevalence in autumn compared to other seasons. As some *Physaloptera* species have complex cycles with intermediate hosts such as Coleoptera, Blattodea and Orthoptera, seasonal variation could be related to environmental conditions acting on variations in abundance of intermediate hosts and/or dietary changes of hosts, as suggested by Cawthorn and Anderson [44]. In fact, some invertebrates such as ground beetles (Coleoptera: Carabidae) in the study area have seasonal variations, increasing their abundance and richness in summer [45].

Several studies have investigated the impact of Monterrey pine plantation on biodiversity worldwide [25,46]. However, the effect on parasites and pathogens has only been recently addressed [10,15,47]. In this context, more studies are needed for a better understanding on the effects of pine plantations on parasites and pathogens of wildlife, including an ecological approach using multiple temporal and spatial scales [48]. This information will be useful for addressing the nexus between wildlife and zoonosis risks into land use planning, within the One Health framework [49,50].

#### **5. Conclusions**

In this study, we reported several intestinal helminths in fecal samples from wild rodents in the landscape of central Chile, but we did not find differences in helminth prevalence between rodents inhabiting native forests and Monterey pine plantations (adult and young plantations). As Monterey pine plantations are one of the most important forestry plantations worldwide, more research should be conducted in order to better understand the effect of pine plantations on other micro- and macroparasites of public health concern.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/2076-2 615/11/2/384/s1, Table S1: Comparison of models used for helminths; Table S2: Comparison of models used for *Hymenolepis* sp.; Table S3; Comparison of models used for *Physaloptera* sp.

**Author Contributions:** Conceptualization, J.A.S., F.F., and A.V.R.; methodology, J.A.S., F.F., and A.V.R.; formal analysis, M.R., R.S., C.L.-A., and A.V.R.; investigation, M.R., R.S., and A.V.R.; resources, J.A.S., C.L.-A., F.F., and A.V.R.; writing—original draft preparation, M.R., R.S., and A.V.R.; writing review and editing, M.R., R.S., J.A.S., C.L.-A., F.F., and A.V.R.; supervision, J.A.S., F.F., and A.V.R.; project administration, A.V.R.; funding acquisition, A.V.R., F.F., and J.A.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by ANID/FONDECYT No. 3160037 and additional support by FONDECYT No. 1140657 and ANID/PAI, Convocatoria Nacional de Subvención a la Instalación en la Academia, 2018, No. PAI77180009.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of University of Chile (Number 0151027 FCS-UCH, October 27, 2015), and by the Servicio Agrícola y Ganadero (SAG; Chilean Fish and Wildlife Service) (License No. 6831/2015).

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The study did not report any data.

**Acknowledgments:** We thank M. Barceló, J. Veloso, M. Silva, A. Arzabe, H. Mendoza, C. Muñoz, L. Moreno and other volunteers for their assistance during field sampling. We are grateful to Carlos Reyes (CONAF) and Ronny Zuñiga for logistical support in the field and Claudio Veloso for the loan of traps. We also thank Patricio Toro for his invaluable assistance in the laboratory. We thank Forestal Masisa S.A. and Corporación Nacional Forestal for allowing us to work on their properties.

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

#### **References**


## *Communication* **Molecular Characterization of New Haplotype of Genus** *Sarcocystis* **in Seabirds from Magdalena Island, Southern Chile**

**Igor C. L. Acosta 1,\*, Solange M. Gennari 1,2, Horwald A. B. Llano 1, Sebastián Muñoz-Leal 1,3 and Rodrigo M. Soares <sup>1</sup>**


**Simple Summary:** Sarcocystidae is a family of apicomplexan protozoa highly prevalent in vertebrates. The definitive hosts of sarcocystids eliminate oocysts or sporocysts that infect intermediate hosts. After infection, mature tissue cysts (sarcocysts) develop in intermediate hosts, mostly in muscle and neurological tissues. *Sarcocysts* are infectious for definitive hosts, which acquire them through carnivorous or scavenging habits. Intermediate hosts and definitive hosts are the natural hosts of sarcocystids in which infections are usually mildly or not symptomatic. In 2017, muscular and neurological tissues of 22 birds from Magdalena Islands, southern coast of Chile, were screened for the presence of DNA of sarcocystids. DNA of organisms of the genus *Sarcocystis* was identified in two Chilean skuas (*Stercorarius chilensis*). The genetic makeup of the parasite detected in skuas was unprecedented and probably represent a new species in the genus. It is well known that *Sarcocystis* may cause severe infections in aberrant hosts, which are susceptible animals that do not behave as natural hosts for the parasite and have low resistance to the infection, thus more studies are needed to characterize this parasitosis in skuas and other hosts to understand the epidemiology of the infection and its impact on the health of marine fauna.

**Abstract:** Evidence of sarcocystid infection was investigated in samples of 16 penguins (*Spheniscus. magellanicus),* four Dominican gulls (*Larus dominicanus*) and two Chilean skuas (*Stercorarius chilensis*) found in Madalenas Islands, Chile, in 2017. Samples of skeletal muscle, cardiac muscle and brain from all birds were screened by a pan-sarcocystid nested-PCR targeting a short fragment of the gene encoding the small ribosomal unit (nPCR-18Sa). The only two positive samples by nPCR-18Sa, both from skuas, were tested by a nested-PCR directed to the internal transcribed spacer 1 (nPCR-ITS1), also a pan-sarcocystidae nested-PCR, and to a nested-PCR directed to the B1 gene (nPCR-B1), for the exclusive detection of *Toxoplasma gondii*. The two nPCR-18Sa-positive samples were nPCR-ITS1-positive and nPCR-B1-negative. The nPCR-ITS1 nucleotide sequences from the two skuas, which were identical to each other, were revealed closely related to homologous sequences of *Sarcocystis halieti*, species found in seabirds of northern hemisphere. Larger fragments of genes encoding 18S and partial sequences of genes coding for cytochrome oxidase subunit 1 were also analyzed, corroborating ITS1 data. The haplotypes found in the skuas are unprecedent and closely related to species that use birds as the definitive host. Further studies need to be carried out to detect, identify and isolate this parasite to understand the epidemiology of the infection and its impact on the health of marine fauna.

**Keywords:** wild birds; coccidian; molecular; apicomplexa; marine

**Citation:** Acosta, I.C.L.; Gennari, S.M.; Llano, H.A.B.; Muñoz-Leal, S.; Soares, R.M. Molecular Characterization of New Haplotype of Genus *Sarcocystis* in Seabirds from Magdalena Island, Southern Chile. *Animals* **2021**, *11*, 245. https://doi.org/10.3390/ani11020245

Academic Editors: Rafael Calero-Bernal and Ignacio Garcia-Bocanegra Received: 16 November 2020 Accepted: 15 January 2021 Published: 20 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/).

#### **1. Introduction**

The phylum Apicomplexa is composed of obligate intracellular parasites that are characterized by having a specialized structure called the apical complex, which is used to invade vertebrate host cells [1] Within this phylum, the Sarcocystidae family comprises more than 196 species of coccidia that form cysts in tissues of intermediate hosts. Although taxonomic controversies still exist, this family has been divided into three subfamilies: Sarcocystinae, represented by the genera *Frenkelia* and *Sarcocystis*; Cystoisosporinae, containing the genus *Cystoisospora*; and Toxoplasmatinae, a subfamily with a few species grouped in the genera *Besnoitia*, *Hammondia*, *Neospora* and *Toxoplasma* [2–5].

*Toxoplasma gondii* is a coccidian parasite with worldwide distribution. It infects virtually all warm-blooded animals, including humans, but only cats (domestic and wild) act as definitive hosts. Toxoplasmosis has been reported in many avian species; however, little information is available in relation to populations of *Spheniscus magellanicus*, *Stercorarius chilensis* and *Larus dominicanus* [6]. Recently, *T. gondii* antibodies were detected in 57 (43.18%) out of 132 serum samples collected from free-living Magellanic penguins (*Spheniscus magellanicus)* on Magdalena Island, Chile, with titers that ranged from 20 to 320 [7].

The genus *Sarcocystis* has an obligate two-host life cycle. Asexual stages develop in intermediate hosts, usually omnivores, through forming cysts in the musculature and central nervous system. Infection of intermediate hosts occurs through their ingestion of food or water contaminated with sporocysts. Sexual stages only develop in the definitive host, which is typically a carnivore or an omnivore, and infection in this case occurs through ingestion of meat contaminated with cysts [8]. Sarcocystids of the genus *Sarcocystis* may cause severe infections in aberrant hosts, which are susceptible animals that do not behave as natural hosts of the parasite and have low resistance to the infection. Thus, *Sarcocystis* potentially pose risk to human and animal health, depending on the susceptible host behaving as aberrant host or not [8].

A few studies have documented the presence of *Sarcocystis* spp. in wild animals in Chile. Presence of cysts of this parasite has been confirmed in muscle tissues of pudu deer (*Pudu puda*), guanacos (*Lama guanicoe*) and sea lions (*Otaria byronia)* [9–11]. However, *Sarcocystis* has not yet been described in Chilean wild birds.

The Chilean skua *Stercorarius chilensis* is a large predatory seabird that inhabits shore ecosystems along the southern cone of South America from central Peru to northern Argentina, with occasional occurrence on the coasts of Ecuador, Brazil, Uruguay and Antarctica [12]. Skuas belong to the order Charadriiformes and are considered to be opportunist feeders, preying on a wide diversity of animals such as small seabirds, fish, scraps and carrion [13,14]. Populations of skuas may be small, but they do not approach the thresholds for vulnerable classification following a population-size criterion (<10,000 mature individuals) [15].

Considering other coastal birds' species, the Kelp Gull (*Larus dominicanus*) is an opportunistic feeder like numerous Laridae and consumes a wide variety of fishes, invertebrates and fisheries waste [16,17]. A high diversity in the use of habitat types has been recorded throughout its distributional range in the Southern Hemisphere, including Argentina, Brazil, Chile, Peru and Uruguay, and the breeding population has been estimated at least 160,000 pairs [17]. In contrast, the Magellanic penguin (*S. magellanicus*) has approximately 1.1 to 1.6 million breeding pairs that nest along the eastern and western coasts of South America, in Argentina Chile and the Malvinas/Falkland Islands [18]. *Spheniscus magellanicus* has a primarily piscivorous diet with the presence of some cephalopods and crustaceans [19].

To date, more than 25 species of *Sarcocystis* have been found to use birds as intermediate hosts [8,20]. *Sarcocystis falcatula*, *Sarcocystis calchasi* and the recently described unnamed species *Sarcocystis* sp. Chicken-2016-DF-BR, which can possibly be interpreted as *Sarcocystis wenzeli* [21] are species that may be pathogenic for intermediate hosts [22–24].

Focusing on *T. gondi* and the genus *Sarcocystis*, the aim of this study was to screen for the evidence of new species or species genotypes of Sarcocystidae in seabird carcasses from southern Chile, a region with scarce data on the occurrence of this group of parasites. Molecular evidence of a unique haplotype of genus *Sarcocystis* was found in two Chilean skuas (*S. chilensis*).

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

#### *2.1. Ethical Considerations*

Sample collections on Magdalena Island were conducted under license no. 039/2016 issued by the National Forestry Corporation (Corporación Nacional Forestal; CONAF), and permit no. 2799 issued by the National Fisheries Service (Servicio Nacional de Pesca; SERNAPESCA), Chile. This study was approved by the Ethics Committee on Animal Use (CEUA-no. 9701041113) of the School of Veterinary Medicine, University of São Paulo (FMVZ-USP).

#### *2.2. Collection of Samples*

In January 2017, fragments from the pectoral muscle, heart and brain, comprising approximately 5–10 g each, were collected from fresh seabird carcasses on Magdalena Island. This island is located in the Strait of Magellan, near the city of Punta Arenas, in southern Chile (52◦55 10.0" S; 70◦34 37.7" W), and constitutes a natural reserve named "Monumento Natural Los Pinguinos". Necropsies were performed in situ and samples were stored in sterile microtubes at −20 ◦C until the time of analysis. Samples were collected from 22 birds: 16 penguins (*S. magellanicus*), four Dominican gulls (*L. dominicanus*) and two Chilean skuas (*S. chilensis*), totaling 66 samples (22 from pectoral muscles, 22 from hearts and 22 from brains).

#### *2.3. Molecular Identification*

Tissue samples of 25–50 mg were subjected to DNA extraction using the DNeasy Blood and Tissue kit (Qiagen, Hilden, Germany) following the manufacturer's recommendations, with the exception of final elution of the product into 50 μL of elution buffer from Qiagen Kit. As internal control for the evaluation of the successfulness of the DNA extraction, DNA samples were tested by conventional PCR targeting mitochondrial hypervariable region in the penguins derived samples and by PCR directed to mitochondrial 16S rRNA gene in the tissues from the other seabirds [25–27].

Initial screening targeting the Sarcocystidae family was performed using a pansarcocystid nested PCR based on primers [28] directed to a short fragment of 18S rDNA gene (nPCR-18Sa). The nPCR-18Sa positive samples were further investigated for the presence of DNA of *T. gondii* to amplify partial fragments of gene B1 (nPCR-B1) using the primers described by [29]. The nPCR-18Sa positive samples were also tested by a second pan-sarcocystid nested PCR directed to internal transcribed spacer 1 (nPCR-ITS1) [30,31] The nPCR-ITS1 were used in order to obtain genetic sequences capable of differentiating the species of the Sarcocystid screened with nPCR-18Sa. The nPCR-ITS1 positive samples were further tested with a third pan-sarcocystid nested PCR, now targeting a larger fragment of 18S rDNA gene (nPCR-18Sb) using primers described by [32], as well with a *Sarcocystis* specific nested PCR [30] directed to cytochrome oxidase subunit I (nPCR-CO1). The primers are depicted in Table 1.

The first round of nPCR-18Sa were performed with 3.0 μL of extracted DNA, 1.8 μL of 10× PCR Buffer (KCL 50 mM; Tris HCl 10 mM; pH 9.0) (Life Technologies Corporation, Carlsbad, CA 92008 USA), 0.7 μL of MgCL2 (50 mM), 1.4 μL of mixed dNTPs (10 mM), 0.1 μL of each primer (25 μM), 0.14 μL of PlatinumTM Taq DNA Polymerase (5 U/μL) (Life Technologies Corporation, Carlsbad, CA 92008 USA) and ultrapure autoclaved water to a volume of 18 μL per reaction. The PCR thermal cycling consisted of an initial incubation at 94 ◦C for 30 sec, followed by 30 cycles (94 ◦C for 25 sec, 55 ◦C for 1 min, 72 ◦C for 1.5 min) and a final extension at 72 ◦C for 10 min. For the second rounds: 1 μL of template derived from the first reactions, 2.5 μL of 10× PCR Buffer (KCL 50 mM; Tris HCl 10 mM; pH 9.0) (Life Technologies Corporation, Carlsbad, CA 92008 USA), 2.5 μL of MgCL2 (50 mM), 4.0 μL of mixed dNTPs (10 mM), 1.25 μL of each primer (10μM), 0.15 μL of PlatinumTM Taq DNA Polymerase (5 U/μL) (Life Technologies Corporation, Carlsbad, CA 92008 USA)and

ultrapure autoclaved water to a volume of 25 μL per reaction. The PCR thermal cycling consisted of an initial incubation at 94 ◦C for 4 min, followed by 30 cycles (94 ◦C for 30 s, 55 ◦C for 1 min, 72 ◦C for 2.0 min) and a final extension at 72 ◦C for 10 min. For the second rounds the same quantities of the reagent mixture with primers at 50 μM, using 2 μL of the product of the PCR diluted in ultra-pure water (1:2). The nPCR thermal cycling consisted of an initial incubation at 94 ◦C for 4 min, followed by 35 cycles (94 ◦C for 30 s, 55 ◦C for 1 min, 72 ◦C for 2.0 min) and a final extension at 72 ◦C for 10 min.


**Table 1.** Primers for the detection of Sarcocystidae in tissues of seabirds from Magdalena Island, Chile.

<sup>a</sup> Primers used in the first round of amplification (1); primers used in the second round of amplification (2); primers used in both first and second round of amplification (1 + 2).

> The first round of nPCR-B1 were performed with 1.0 μL of extracted DNA, 2.5 μL of 10× PCR Buffer (KCL 50 mM; Tris HCl 10 mM; pH 9.0), (Life Technologies Corporation, Carlsbad, CA 92008 USA), 0.75 μL of MgCL2 (50 mM), 4.0 μL of mixed dNTPs (10 mM), 1.25 μL of each primer (10 μM), 0.15 μL of PlatinumTM Taq DNA Polymerase (5 U/μL) (Life Technologies Corporation, Carlsbad, CA 92008 USA)and ultrapure autoclaved water to a volume of 25 μL per reaction. The PCR thermal cycling consisted of an initial incubation at 94 ◦C for 3 min, followed by 25 cycles (94 ◦C for 25 s, 55 ◦C for 1 min, 72 ◦C for 1.5 min) and a final extension at 72 ◦C for 10 min. For the second rounds: 1 μL of template derived from the first reactions, 2.5 μL of 10× PCR Buffer (KCL 50 mM; Tris HCl 10 mM; pH 9.0) (Life Technologies Corporation, Carlsbad, CA 92008 USA), 2.5 μL of MgCL2 (50 mM), 4.0 μL of mixed dNTPs (10 mM), 1.25 μL of each primer (10 μM), 0.15 μL of PlatinumTM Taq DNA Polymerase (Life Technologies Corporation, Carlsbad, CA 92008 USA) and ultrapure autoclaved water to a volume of 25 μL per reaction. The nPCR thermal cycling consisted of an initial incubation at 94 ◦C for 3 min, followed by 35 cycles (94 ◦C for 25 s, 55 ◦C for 1 min, 72 ◦C for 1.5 min) and a final extension at 72 ◦C for 10 min.

> The first round of nPCR-18Sb, nPCR-ITS1 and nPCR-CO1 were performed with 4 μL of extracted DNA, 2.5 μL of 10× PCR Buffer (KCL 50 mM; Tris HCl 10 mM; pH 9.0) (Life Technologies Corporation, Carlsbad, CA 92008 USA), 1.0 μL of MgCL2 (50 mM), 0.5 μL of mixed dNTPs (10 mM), 1.0 μL of each primer (10 μM), 0.2 μL of PlatinumTM Taq DNA Polymerase (Life Technologies Corporation, Carlsbad, CA 92008 USA) (5 U/μL) (Termofischer Scientific) and ultrapure autoclaved water to a volume of 25 μL per reaction. The PCR thermal cycling consisted of an initial incubation at 94 ◦C for 3 min, followed by 35 cycles (94 ◦C for 30 s, 56 ◦C for 30 s, 72 ◦C for 50 s) and a final extension at 72 ◦C for 5 min. For the second rounds: 2 μL of template derived from the first reactions, 2.5 μL of 10× PCR Buffer (KCL 50 mM; Tris HCl 10 mM; pH 9.0) (Life Technologies Corporation, Carlsbad,

CA 92008 USA), 1.0 μL of MgCL2 (50 mM), 0.5 μL of mixed dNTPs (10 mM), 2.5 μL of each primer (10 μM), 0.2 μL of PlatinumTM Taq DNA ) (Life Technologies Corporation, Carlsbad, CA 92008 USA) (5 U/μL) and ultrapure autoclaved water to a volume of 25 μL per reaction. The thermal cycling was the same used in the first round.

DNA of *Sarcocystis neurona*, *Neospora caninum* and *Hammondia hammondi* was used as positive controls and ultrapure DNAse-free water as the negative control for all reactions.

PCR products were resolved on 2.0% agarose gels and viewed through UV transillumination. Amplicons of the expected sizes were treated with ExoSAP-IT (Affimetryx/Thermo Fisher Scientific, Santa Clara, CA, USA), prepared for sequencing using the Big Dye Terminator cycle sequencing kit (Applied Biosystems, Foster City, CA, USA) and sequenced in an ABI automated sequencer (ABI 3500 Genetic Analyzer, Applied Biosystems). Sequencing was performed using the same primers as in the nPCR consensus. Sequence edition and contig assemblies were done by using the software Codoncode aligner, Codoncode Corporation. Final sequences were compared with homologous available in GenBank, using the BLASTn algorithm (Table S1) (https://blast.ncbi.nlm.nih.gov/Blast.cgi).

For the phylogenies, sequences were aligned using the Clustal W program, as implemented in the BioEdit Sequence Alignment Editor [34]. The phylogenetic tree based on ITS1 was inferred using MEGA X [35], through the maximum likelihood method and T92 model of evolutionary distances. Branch supports were tested through 1000 bootstrap replications.

#### **3. Results**

#### *3.1. Molecular Identification*

Sixty-six tissue samples from 22 seabirds were screened by nPCR-18Sa and only two samples of pectoral muscle from two Chilean skuas were positive. None of these two samples were positive for the *T. gondii*-specific nested-PCR (nPCR-B1). The two nPCR-18Sa-positive samples were also positive by nPCR-ITS1, nPCR-18Sb and nPCR-CO1 (Figure 1, left panel). After sequencing nPCR-ITS1, nPCR-18Sb and nPCR-CO1 amplicons and removal of primer-derived sequences, 861, 783 and 547 base pairs were obtained, respectively. Fragments of the sequences obtained are shown in Figure 1, right panel. The homologous sequences of the two samples were identical to each other, thus only one set were submitted to the GenBank, under the accession numbers MW160469, MW161469, MW157378. Through BLAST searches, ITS1, CO1 and 18S genetic sequences were compared with sequences producing the most significant alignments, with query coverage ≥ 99% and percentage similarities ≥ 99.00% in the cases of CO1 and 18S. All ITS1 sequences with query cover ≥ 96% were used for analyses of the genetic sequence of the skuas.

**Figure 1.** Left panel: Agarose gel electrophoresis of nPCR-ITS1 (**1**), nPCR-CO1 (**2**), and nPCR-18Sb (**3**) amplicons from *Sarcocystis* sp. ex *Stercorarius chilensis,* nPCR-ITS1 amplicons (**4**) from *Sarcocystis neurona* and Ladder Scada 100 bp, Sinapse, Inc. (**L**). Right panel: segments of the electropherogram obtained after sequencing nPCR-18Sb (**top**), nPCR-CO1 (**middle**), and nPCR-ITS1 (**bottom**) amplicons from *Sarcocystis* sp. ex *Stercorarius chilensis.*

> The ITS1 fragment from the skuas showed 96.14–96.28% identity to sequences of *Sarcocystis halieti* from herring gulls (*Larus argentatus*) (MN450340–MN450356), great cormorants (*Phalacrocorax carbo*) (MH130209, JQ733513) and white-tailed sea-eagles (*Haliaeetus albicilla*) (MF946589–MF946596). *Sarcocystis* sp. from Cooper's hawk (*Accipiter cooperii*)

(KY348755), *Sarcocystis columbae* from common woodpigeons (*Columba palumbus*) (GU253885, HM125052) and herring gulls (*Larus argentatus*) (MN450338–MN450339) and *Sarcocystis corvusi* from Eurasian jackdaws (*Corvus monedula*) (JN256119) showed less than 94% sequence identity with homologous sequences from the Chilean skuas at ITS1 locus.

In contrast to ITS1, much less genetic variability was observed within the CO1 and 18S coding genes. The haplotype obtained from the skuas was named *Sarcocystis* sp. ex *Stercorarius chilensis*.

The CO1 sequences from *Sarcocystis* sp. ex *Stercorarius chilensis* were 100% identical with homologous sequences of *S. corvusi* (MH138314), *S. columbae* (MH138312) and *S. halieti* (MH138308-09; MF946583). *Sarcocystis fulicae* (MH138316), *Sarcocystis wobeseri* (MH138315), *Sarcocystis cornixi* (MH138313) and *Sarcocystis* sp. ex *Accipiter cooperii* (KY348756) differed by only one nucleotide substitution from *Sarcocystis* sp. ex *Stercorarius chilensis* at this locus. Regarding the 18S rRNA gene, the maximum percentage identity was found between *Sarcocystis* sp. ex *Stercorarius chilensis* and *S. halieti* (99.74%).

#### *3.2. Phylogeny*

The ITS1-based phylogeny demonstrated that the species that shares the most recent ancestral commonality with *Sarcocystis* sp. ex *Stercorarius chilensis* was *S. halieti* (Figure 2).

These sequences were separated with high support from a major clade comprising the species *Sarcocystis* sp. ex *Columba livia* (FJ232948), *Sarcocystis calchasi* (KC733715–KC733718) and *Sarcocystis wobeseri* (MN450365–MN450373, HM159421, JN256121), which exploit Anseriformes, Charadriiformes, Columbiformes, Psittaciformes and other birds as intermediate hosts.

Altogether, the *Sarcocystis* species that were most similar to *Sarcocystis* sp. ex *Stercorarius chilensis* used birds as intermediate hosts.

**Figure 2.** Phylogenetic tree of *Sarcocystis* species based on ITS1 sequences. The tree was constructed using the maximum likelihood method and Tamura 3 parameters nucleotide substitution model. The final alignment contained 49 sequences and 814 aligned nucleotide positions. All positions containing gaps and missing data were eliminated (complete deletion option). Numbers in branches represent bootstrap values after 1000 replicates. The red dot identifies the sequence of *Sarcocystis* sp. ex *Stercorarius chilensis*.

#### **4. Discussion**

*Toxoplasma gondii* has high genotypic diversity and several new genotypes have been described in wildlife [36–38], which has aided to understand the shape of molecular evolution and the epidemiology of the infection [38]. Similarly, several species descriptions have been made for the genus *Sarcocystis,* most of them with the aid of molecular methods [39–43]. This study presents the results of a molecular screening of Sarcocystidae focusing on animal species and geographical areas where these parasites have rarely or not

yet been identified. DNA of organisms of the genus *Sarcocystis* was identified in two Chilean skuas, whereas DNA of *T. gondii* were not found in any sample. The *Sarcocystis* haplotype detected in skuas was named *Sarcocystis* sp. ex *Stercorarius chilensis.*

Although all the samples tested negative for the presence of *T. gondii*, antibodies against this parasite were detected previously in 57 (43.18%) of the 132 serum samples from free-living Magellanic penguins from the same region, with titers that ranged from 20 to 320 [7]. Herein, infected animals were not encountered possibly because the sampling was insufficient to find at least one positive animal, as the prevalence of the infection in seabirds in the sampled area are not known and the sampling might have not been representative of the population surveyed. In addition, the mass of tissue that was tested might have been insufficient because of the very sparse and focal distribution of *T. gondii* cysts in the tissues of the HI, thus, digestion of samples previously to the DNA extraction and subsequent DNA detection would be more appropriate than direct DNA extraction, as used in this study [44].

Oocysts of *T. gondii* can sporulate and survive in seawater for months [45,46]. Marine mammals in different groups (cetaceans, pinnipeds and sirenians) and seabirds might become infected through consumption of water containing the oocysts. Thus, *T. gondii* oocysts from felidae feces might enter the marine environment and contaminate both the water and several invertebrate species, which could act as vectors of infection for mammals and seabirds [47]. Mice can be experimentally infected when fed with *T. gondii*-contaminated oysters (*Crassostrea virginica*) [45] proving that *T. gondii* was able to survive for several months in these mollusks [48]. Anchovies and Pacific sardines can be experimentally contaminated with *T. gondii* oocysts, which indicates that migratory filter feeders may serve as biotic vectors for this parasite [49]. Another study proved that freshwater crustaceans were able to bioaccumulate *T. gondii* oocysts. It should be noted that crustaceans are part of penguins' and many seabirds' food chain [50,51]. Thus, although the birds screened here were found not infected by *T. gondii*, marine fauna are at risk of acquiring the infection, by ingesting oocysts carried by transport hosts (oysters, fish and other) or through predation of intermediate hosts in the marine or in the coastal environment.

Based on molecular data, *Sarcocystis* sp. ex *Stercorarius chilensis* is an undescribed *Sarcocystis* species, closely related to *S. halieti*. The molecular identification based on ITS1, CO1 and 18S rRNA gene sequences showed a closed relationship between *Sarcocystis* sp. from Chilean skuas and other *Sarcocystis* spp. that use birds as intermediate hosts and predatory birds as definitive hosts. As expected, the most variable locus was ITS1, and phylogenies based on 18S rRNA and CO1 genes showed insufficient discrimination power to differentiate between species within the genus [39,41].

The most similar sequences to ITS1 of *Sarcocystis* sp. ex *Stercorarius chilensis* are those from *Sarcocystis* spp. that use hawks as definitive hosts. *Sarcocystis* sp. ex *Stercorarius chilensis* grouped together with *S. halieti*, a species that uses white-tailed sea-eagles (*Haliaeetus albicilla*) and Eurasian sparrowhawks (*Accipiter nisus*) as definitive hosts [39]. Other taxa found through ITS1-based BLAST searches encompass *Sarcocystis* spp. that also use hawks as definitive hosts (*Accipiter cooperii*, *Accipiter nisus*), except for *S. corvusi*, for which this information remains unknown [52,53]. Accipiter hawks (*Accipiter gentilis*, *Accipiter nisus*) are definitive hosts for *S. calchasi* [54–56].

Several studies have expanded the knowledge on the host specificity of *Sarcocystis*, as unequivocal identification of the parasite can be achieved after identifying sarcocysts and oocysts to species level using molecular methods. *Sarcocystis halieti* and *Sarcocystis lari* were found to have formed oocysts in the intestine of white-tailed sea eagle (*Haliaeetus albicilla*), showing for the first time the potential role of sea eagle as definitive host of those species of *Sarcocystis* [53]. Likewise, european seabirds were found to harbor several species of *Sarcocystis* after DNA of *Sarcocystis lari*, *S. wobeseri*, *S. columbae* and *S. halieti* were detected in sarcocysts infecting muscle of herring gulls (*Larus argentatus*), great black-backed gulls (*Larus marinus*) and great cormorants (*Phalacrocorax carbo*) in Lithuania [40–42].

The four morphologically indistinguishable *Sarcocystis* species, *Sarcocystis lari*, *S. wobeseri*, *S. columbae* and *S. halieti,* could only be differentiated in *L. argentatus* by means of ITS1 sequence analysis [42]. Likewise, only ITS1 clearly discriminated *Sarcocystis* sp. ex *Stercorarius chilensis* from *S. halieti*, which reinforces that molecular characterization using this marker is of paramount importance to distinguish closed related species within the genus.

It is well known that a single animal can host more than one *Sarcocystis* species [40]. Here, sarcocysts were not individually excised and subjected to molecular examination, notwithstanding, the possibility of mixed infected samples of skuas was discarded because single peaks and clean sequence throughout the chromatograms were obtained for each sequence. Thus, a haplotype could be confidently assigned to the samples.

Although screening *Sarcocystis* by using molecular methods without morphological characterization of parasitic structures is obviously not enough to name a new species, this procedure may provide subsides to future studies on the epidemiology of the infection and its impact on the health of marine fauna. To our knowledge, *Sarcocystis* in south American seabirds were identified only once [43], which suggests that a wide field of research on diversity of sarcocystidae can be explored on this continent.

#### **5. Conclusions**

Although few animals have been screened in this study and morphological characterization of the parasites was not carried out, evidence of an unprecedented haplotype of *Sarcocystis* was found in skuas from Chile, which demonstrate that molecular screening of *Sarcocystis* can be a valuable tool to prospect for new species, contributing to knowledge on the epidemiology of sarcocystosis and life cycle of *Sarcocystis*. Sporocysts shed with feces, sarcocysts in tissues or rapid dividing structures in acute sarcocystosis (schyzonts and merozoites) can be more easily and accurately identified as data on *Sarcocystis* genetic sequences increases. Nevertheless, a complete study encompassing aspects of life cycle and morphological data is necessary to fully describe *Sarcocystis* sp. ex *Stercorarius chilensis* and additional studies are needed to better understand the epidemiology of the infection and its impact on the health of marine fauna.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/2076-2 615/11/2/245/s1, Table S1. Results from BLAST search using sequences of ITS1, COX1 and 18S of *Sarcocystis* sp. ex *Stercorarius chilensis*.

**Author Contributions:** Conceptualization, I.C.L.A., S.M.G. and R.M.S. writing—original draft preparation, I.C.L.A. and R.M.S.; review and editing, I.C.L.A., S.M.G., H.A.B.L., S.M.-L. and R.M.S.; visualization, I.C.L.A., S.M.G., H.A.B.L., S.M.-L. and R.M.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo), Brazil and by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Brazil. R.M.S. and S.M.G. received a fellowship from CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico).

**Institutional Review Board Statement:** This study was approved by the Ethics Committee on Animal Use (CEUA-no. 9701041113 - 2014) of the School of Veterinary Medicine, University of São Paulo (FMVZ-USP). Sample collections on Magdalena Island were conducted under license no. 039/2016 issued by the National Forestry Corporation (Corporación Nacional Forestal; CONAF), and permit no. 2799 issued by the National Fisheries Service (Servicio Nacional de Pesca; SERNAPESCA), Chile.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** To Roberto Fernández and Ricardo Cid (CONAF) for their valuable assistance during the sample collections on Magdalena Island.

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

#### **References**

