Introduction

Parasites may have significant influence on individuals and populations by changing the behaviour of individual hosts (Lefèvre et al. 2009), regulating host population sizes through direct effects on birth rate and mortality (Møller 2005), mediating competitive interactions among hosts (Thomas et al. 2005) or acting as ecosystem engineers (Thomas et al. 1999).

Studies done world-wide have shown that variations in gastrointestinal parasitic infections are mainly connected with rainfall fluctuations (Chattopadhyay and Bandyopadhyay 2013; Akturazzaman et al. 2013). Adequate moisture and optimum temperature favour the growth and survival of the infective stages of nematodes, leading to higher contamination of pastures or food (Regassa et al. 2006; Shirale et al. 2008). It has also been shown that seasonal changes in the impact of parasites on hosts may influence immune function, for example, when connected with the breeding season (Møller et al. 2003). In addition, host condition, which is often correlated with host population density, is likely to influence parasite virulence (Donnelly et al. 2012). In more seasonal environments, it has been suggested that if density-dependent virulence is widespread, wildlife disease is likely to be more virulent (Donnelly et al. 2012).

The significant impact of parasites on the population dynamics of wildlife has emerged as a critical issue in the conservation of threatened species. Mounting evidence indicates that infectious agents can considerably influence local populations by causing temporary or permanent decline of their numbers (Thompson et al. 2010). Therefore, understanding the role of infectious agents in shaping wildlife dynamics requires accurate data on the diversity and abundance of potential pathogens in natural ecosystems, especially at the local scale (Smith et al. 2009).

The European bison (Bison bonasus L., 1758) is the largest terrestrial mammal in Europe. After extinction in the wild in 1919, it was restored from captive survivors and reintroduced to over 30 locations in eastern Europe (Pucek et al. 2004). The world population of the species has been increasing (3310 free-living bison in 2013) (Raczyński 2014). However, free-ranging populations are small (only 9 populations have over 100 individuals) and isolated by geographical distance. The bison population in the Białowieża Primeval Forest (BPF) is the largest in the world (965 individuals in total in the Polish and Belarusian parts of the Forest), and it constitutes the core of the global population of the species (29 % of free-ranging bison). Apart from lack of connectivity between isolated populations and their small size, the other threats to bison are low genetic variation due to a severe genetic bottleneck (Tokarska et al. 2011), as well as diseases and parasites (Pucek et al. 2004). Although bison are traditionally managed in forest habitats, increasing evidence shows that bison are a refugee species adapted to open and mixed habitats (Kerley et al. 2012; Bocherens et al. 2015). Most free-ranging populations of bison inhabiting forests are supplied with food in the winter which mitigates migrations and reduces farm crop depredation. This leads to unnatural winter aggregations, strong decline of individual home range size, and an increase of local winter densities of bison (Krasińska et al. 2000; Radwan et al. 2010). Within several years of the restoration, bison assimilated several new species of parasites. This included the blood-sucking nematode of abomasa Ashworthius sidemi (Trichostrongylidae) which is specific to Asian deer species (Dróżdż et al. 2000, 2003). The invasion of a new, dangerous parasite in the endangered population of bison in the BPF is alarming in light of the observed decline in their physical condition and the increasingly female-biased calf sex ratio. This is as predicted by Trivers-Willard hypothesis and may be associated with an increase in parasitic load especially the invasive A. sidemi (Hayward et al. 2011). In recent years, the presence of A. sidemi was confirmed genetically in cattle (Moskwa et al. 2015). This indicates the possible transmission of the parasite from wildlife to livestock.

The European bison population is widespread in the forest and forest surroundings at an area covering around 800 km2 (Kowalczyk et al. 2013). Males live alone (62 % of males) or in small bull groups, while females with calves and subadults roam in groups numbering on average 11–15 individuals (Krasińska and Krasiński 1995, 2013; Krasińska et al. 2000). Higher aggregations of bison occur during the rutting season (August–October). These behavioural traits are responsible for higher infection rates among females and yearlings. During the winter (December–March), most bison concentrate around feeding sites where they are supplementary fed with differing intensity. They stay in very limited ranges, sometimes for over 4 months, in herds up to 70 individuals (Hofman-Kamińska and Kowalczyk 2012; Kowalczyk et al. 2011). This is the second factor favouring inter-individual transmission of parasites in the bison population.

In this study, we aimed to investigate seasonal pattern and sexual differences in intestinal helminth egg excretion in wild bison. We predicted that both the prevalence and number of eggs per gram (EPG) of faeces would be higher in winter and among females. Our hypotheses were due to both the observed differences in social behaviour of bison cows and bulls and the increased contact rate between individuals at feeding sites in winter.

Material and methods

Study area

The study was conducted in the Polish part (ca. 600 km2) of the Białowieża Primeval Forest (BPF; 52° 29′–52° 37′ N, 23° 31′–24° 21′ E) located along the Polish-Belarusian border.

The climate of the BPF is transitional between Atlantic and continental types with clearly marked cold and warm seasons. Meteorological data have been collected daily since 2000. The mean annual temperature is +8.0 °C. The mean temperature of the coldest month—January—is −3.5 °C. The warmest month is July with a mean temperature of +19.9 °C. The vegetative season lasts on average 215 days (range 198–238) and was calculated by Tylkowski (2013). Snow cover has persisted from 41 to 120 days per year with a maximum recorded depth of 55 cm. Mean annual precipitation is 650 mm.

Faeces collection and analyses

This study was based on analysis of faecal parasitic load and was conducted during a 12-month period from April 2008 to March 2009 in the Polish part of the Białowieża Primeval Forest. Each month, usually in the morning, fresh faecal samples were collected separately from males and females (12 samples of each sex). Faeces were collected from radio-collared individuals, and herds with collared bison (15 to 25 individuals are tracked annually) or occasionally from non-collared individuals, always after visual sex identification. Our sampling strategy included both temporal and spatial stratification to avoid pseudo replication. Usually, the same herds or individuals were not sampled more than once within 1 month to secure as random sampling as possible. This was possible due to the spatial and social organisation of the bison population (Krasińska and Krasiński 1995; Krasińska et al. 2000) and the distribution of individuals and herds (in total 451–456 bison in 2008–2009) (Raczyński 2009, 2010) across the 800 km2 area. The samples were put into 30-ml labelled tubes. All samples were stored for up to 1 month at −20 °C and investigated immediately after defrosting. Detected eggs of parasites maintained their morphological figures typical for the genus. Up to date, we have not observed any damages in eggs structure that were previously preserved by freezing the faeces for few weeks (Demiaszkiewicz, pers. commun.). In total, 288 faecal samples were collected (144 for males and 144 for females). The same materials were used to study the annual pattern of shedding Eimeria spp. oocysts in European bison (Pyziel et al. 2011).

Three grams of each sample were examined for helminth eggs by direct flotation in saturated sucrose solution (specific gravity 1.27) with the use of a centrifuge (Taylor et al. 2007). Samples were investigated using an Olympus BX50 light microscope at ×100–400 magnification. The eggs were determined to the family or genus according to their dimensions and morphological figures (Shorb 1939; Taylor et al. 2007).

We used two measures to describe parasitic load: (1) the prevalence denoted the percentage of faeces with parasite eggs and (2) EPG—number of eggs per gram of faecal sample. Prevalence in this study do not reflect the proportion of bison infected, but the proportion of bison shedding parasite eggs, and we used this term to avoid longer explanation each time we talk about the proportion of faeces with bison parasites. Parasite egg number in faeces does not directly show infection intensity, which is possible to get by carrying out dissection analysis (Dróżdż et al. 2000). However, in the case of protected and endangered species such as European bison, this is the only method that allows the collection of samples for parasitological studies throughout the year and to track seasonal changes in parasite load. The number of excreted eggs stated in coprological studies can depend on the parasite. This is due to differences in the parasite’s biology, life cycle, or prepatent period. Therefore, we have not compared one parasite to another but followed the seasonal changes within each parasite separately.

Statistical analysis

To model which factors affect the prevalence and number of parasite eggs in bison faeces, we described the following set of parameters for all sampled faeces: the presence and the number of parasite eggs (separately for each parasite group), sex of the bison, and the season when samples were collected (winter: December–March; snow-free period: April–November). In order to analyse factors affecting the prevalence of each parasite group, we fitted generalised linear models (GLM) for binomial data (Zuur et al. 2009). We set the probability of the bison to be infected as a dependent binominal variable: “1” was attributed to parasite presence and “0” to parasite absence. Number of eggs per gram of faeces was calculated for the parasites in which the highest numbers of positive samples were found—Trichostrongylidae and Aonchotheca sp.—and we fitted negative binomial generalised linear models (negative binomial GLM) for count data which dealt with observed overdispersion of models (Zuur et al. 2009). For each parasite group, we ran separate models with the main effects of the explanatory variables (season and bison sex). Final models comprised only significant factors (P < 0.05).

We checked the normality and homoscedasticity in the distribution of the final model residuals by inspecting the quantile-quantile distribution plot and model residuals against plots of fitted values (estimated responses). All statistical analyses were completed in R (version 3.1.2; R Development Core Team 2014).

Results

We found eggs of four types of gastrointestinal nematodes from the genus Aonchotheca, Nematodirus, Trichuris, and Trichostrongylidae family and tapeworm eggs from the genus Moniezia, in bison faeces in the BPF. Trichostrongylidae eggs were expelled from bison at the highest percentage (27.8 %) but in low numbers (Table 1). The prevalence of other parasites did not exceed 12 %. Parasite species differed in EPGs, which was the highest in Trichuris sp. and Moniezia spp. (Table 1). For most of the period without snow cover, no eggs of Aonchotheca sp., Nematodirus spp. and Moniezia spp. were found (Figs. 1 and 2). Trichuris sp. eggs showed no pattern of seasonality and they were expelled at a regular intervals (Fig. 1).

Table 1 The mean number of parasite eggs in one gram of faeces (EPG) and their prevalence values in European bison males (n = 144) and females (n = 144) in the Białowieża Primeval Forest
Full size table
Fig. 1
figure 1

The annual cycle of parasite egg prevalence in European bison faeces (n = 288) in the Białowieża Primeval Forest

Full size image
Fig. 2
figure 2

Percentage of European bison excreting parasite eggs during winter (December–March) and the snow-free period (April–November) in the Białowieża Primeval Forest. Trichostrongylidae (Tricho), Aonchotheca sp. (Aonch), Trichuris sp. (Trichu), Nematodirus spp. (Nemat) and Moniezia spp. (Mon). ***P < 0.001; **0.001 < P < 0.01; *0.01 < P < 0.05

Full size image

Binomial generalised linear models (GLMs) were used to describe the effects of season and bison sex on the probability of bison to be infected with Trichostrongylidae (Tricho), Aonchotheca sp. (Aonch), Trichuris sp. (Trichu), Nematodirus spp. (Nemat) and Moniezia spp. (Monie) (Table 2). Averaged negative binomial generalised linear models (GLMs) were used to describe the effects of season and bison sex on the EPGs of Trichostrongylidae (Tricho) and Aonchotheca sp. (Aonch) infections in bison (Table 3). The models revealed that with the exception of Trichuris sp., whose eggs were more often detected in female faeces (P = 0.01), there were no significant differences in the prevalence of infections between bison males and females (Tables 1 and 2). EPG was significantly higher in females for Aonchotheca sp. eggs (P = 0.04) (Table 3).

Table 2 Parameter estimates for the binomial generalised linear models (GLMs), describing effects of season and bison sex on the probability of bison to be infested with Trichostrongylidae (Tricho), Aonchotheca sp. (Aonch), Trichuris sp. (Trichu), Nematodirus spp. (Nemat) and Moniezia spp. (Mon) in the BPF in 2008–2009
Full size table
Table 3 Parameter estimates for the negative binomial generalised linear models (GLMs), describing the effects of season and bison sex on the number of Trichostrongylidae (Tricho) and Aonchotheca sp. (Aonch) eggs per gram of faeces (EPG) in bison population in the BPF in 2008–2009
Full size table

The prevalence and the number of eggs excreted by European bison showed seasonal variation (Figs. 1, 2 and 3). A comparison of parasite prevalence between winter (December–March) and the snow-free period (April–November) revealed that in winter, a significantly higher share of individuals were infected (Trichostrongylidae: P < 0.001, Aonchotheca sp.: P < 0.001, Moniezia spp.: P = 0.008 and Nematodirus spp.: P = 0.02) (Table 2, Fig. 2). In winter, the prevalence of these parasites was 3–14 fold higher than observed in the snow-free period. Parasite prevalence showed seasonal variation, with the highest values recorded in the winter months: Trichostrongylidae, March (91.7 %); Aonchotheca sp., January (58.3 %); Moniezia spp., January (20.8 %); and Nematodirus spp., January (12.5 %) (Fig. 1). In addition, in winter, EPG of Trichostrongylidae was higher than in the snow-free period (P = 0.004) (Table 3, Fig. 3).

Fig. 3
figure 3

Number of Trichostrongylidae and Aonchotheca sp. eggs per gram of faeces (EPG) during winter (December–March) and the snow-free period (April–November) in the Białowieża Primeval Forest. Trichostrongylidae (Tricho) and Aonchotheca sp. (Aonch). **0.001 < P < 0.01

Full size image

Discussion

The parasitic fauna of European bison has been well described. A total of 88 species of parasites have been discovered in 100 years of investigations (Karbowiak et al. 2014a, b). There is an increasing trend in parasite species richness as well as the prevalence and intensity of infections. This could be caused by possible transmission from other ruminants or the increased contact rate between bison in winter due to their high density arising from supplementary feeding in fixed locations (Karbowiak et al. 2014b; Radwan et al. 2010; Pyziel et al. 2011). It is known that European bison are infested with parasite larvae through contact with contaminated water and pastures (Treboganova 2010). Constant and long-term use of the same area during winter causes an accumulation of invasive material and increased risk of helminth invasions (Treboganova 2010).

In our study, the eggs of four types of gastrointestinal nematodes and one tapeworm were recorded. The prepatent period—the time between the infection of the host and the appearance of the first eggs in faeces—lasts from 2–4 weeks in Trichostrongylidae and Nematodirus spp. (Skrjabin et al. 1954) to 7–8 weeks in Trichuris sp. (Skrjabin et al. 1957) and 9–12 weeks in Anchieta sp. (Moravec et al. 1987). After this time, female nematodes lay eggs which are released into the environment with the host faeces. The development of the third stage invasive larvae takes from 5 to 15 days in Trichostrongylidae, 15 to 30 days in Nematodirus spp., 23 to 25 days in Trichuris sp. and up to 54 to 62 days in Aonchotheca sp. (Skrjabin et al. 1954; 1957; Moravec et al. 1987). The life cycle of Moniezia spp. of tapeworms is complex; they utilise Oribatida sp. mites as intermediate hosts and have a prepatent period that lasts 5–7 weeks (Połeć and Moskwa 1994). A relatively short prepatent period suggests that the aggregation of bison in winter may cause effective transmission of parasites and increased parasitic load within this period.

The prevalence and number of eggs described in our study were relatively low, with the exception of eggs of the Trichostrongylidae family. This group includes parasites such as A. sidemi or Ostertagiinae, which have previously been reported in large numbers in European bison (Dróżdż et al. 2002, 2003; Demiaszkiewicz and Lachowicz. 2007; Demiaszkiewicz and Pyziel 2010; Demiaszkiewicz et al. 2012). They also have the shortest prepatent period and time required for the development of infective larvae (Skrjabin et al. 1954); this favours the possibility of re-infection. Similar (24.3 %) overall prevalence of Eimeria spp. invasion studied on the basis of the same faecal samples was found, but with much higher oocyst count per gram of faeces (OPG) (Pyziel et al. 2011). However, no symptoms of clinical coccidiosis were observed.

A. sidemi is a pathogenic, blood-sucking nematode of the abomasum which infested Polish populations of European bison in 1999 (Dróżdż et al. 2002). This parasite, which originated from Asian cervids (mainly the sika deer Cervus nippon), was dissected from culled bison in the BPF for the first time in 2000 (Demiaszkiewicz et al. 2008). From 2004 to 2011, all the culled bison were found to be infected with this parasite (Demiaszkiewicz et al. 2008; Demiaszkiewicz and Pyziel 2010), but since then, the prevalence of infections has been decreasing (Kołodziej-Sobocińska pers. commun).

No difference in parasitic infection was found between bison sexes, with the exception of Trichuris sp. prevalence and number of Aonchotheca sp. eggs excreted, which was higher in females. Similar pattern was observed for Eimeria spp. Oocysts were more prevalent in bison cows (35 %) than in bison bulls (14 %) (Pyziel et al. 2011). Recent studies have revealed that male-biased parasitism is not universal and that there are many other factors which influence parasite infection such as age of individuals, sexual size dimorphism, level of hormones, individual host’s variability (i.e. behavioural, physiological) or immunocompetence (Kiffner et al. 2013). In the BPF, higher parasite infection prevalence and the intensity of some parasites in females may result from (1) the bigger herds formed by cows, leading to a higher rate of parasite transmission between individuals, compared to males which roam solitarily or in small groups (Radwan et al. 2010; Pyziel et al. 2011); and (2) immunosuppression caused by pregnancy, which may facilitate the observed pattern (Lloyd 1983; Krishnan et al. 1996). According to previous studies, levels of EPGs and OPGs (oocysts per gram) as well as diversity of parasite composition were higher among the youngest cattle and bison living in high densities (ex. captive individuals) (Höglund et al. 2002; Pyziel et al. 2014). A number of factors not tested in the present study could influence the course of parasite invasion (Lloyd 1983; Wakelin 1985; Stear et al. 1999; Bush et al. 2001). In some studied parasites, the relatively low infection prevalence and intensity, based on the detection of eggs in faeces, could be the reason for not finding differences between the sexes.

We found a seasonal pattern of parasite egg shedding by bison in the BPF. In winter, we registered a higher prevalence of excreted eggs from Trichostrongylidae family, Aonchotheca sp., Moniezia spp. and Nematodirus spp. Trichostrongylidae eggs were also more numerous in this period. A similar pattern was observed in elk (Cervus elaphus) in Wyoming, USA (Hines et al. 2007). In winter (November–April), supplementary fed elk had significantly more gastrointestinal nematode eggs than non-fed individuals in January and February but, interestingly, significantly less in April. This was explained by improved nutrition on the feed grounds (Hines et al. 2007). To explain the seasonality of infectious diseases, with maximum values during winter, several factors were proposed such as host social behaviour, high contact rates, harsh weather conditions, poorer nutrition and investment in reproduction causing a weakened immune system (London and Yorke 1973; Finkenstädt et al. 1998; Newton-Fisher et al. 2000; Dowell 2001; Hosseini et al. 2004; Altizer et al. 2006). A decline of 10 % in the body condition of bison in winter has been recorded (Hayward et al. 2015). Study by Pyziel et al. (2011) revealed the highest prevalence of Eimeria spp. oocysts in bison faeces in early spring and significant influence of winter aggregations of bison on coccidian oocysts shedding.

In addition, some free-living stages of species from Trichostrongylidae family (e.g. Trichostrongylus colubriformis, Teladorsagia circumcincta in sheep) are known to be resistant to low temperatures; therefore, they may infect hosts during the winter season (O’Connor et al. 2006). For many parasites, transmission depends on host encounters with parasite stages present in the environment (Altizer et al. 2006). It is thought that both eggs and larvae are present in bison faeces year round (Demiaszkiewicz pers. commun.); however, such studies in wild ruminants have not been conducted.

Seasonal aggregation caused by the increased use of water sources, fruiting trees during the dry season, or dens and supplementary feeding sites in winter may have epidemiological consequences for animals (Gremilion-Smith and Woolf 1988; Altizer et al. 2006; Radwan et al. 2010; Navarro-Gonzalez et al. 2013; Kołodziej-Sobocińska et al. 2014). European bison behaviour varies between summer and winter. They spend summer in smaller herds roaming on larger ranges, while in winter, most individuals aggregate around supplementary feeding sites and use limited area (Krasińska and Krasiński 1995; Krasińska et al. 2000; Radwan et al. 2010). Therefore, the high population density at feeding sites in winter may facilitate parasite transmission due to the repeated defecation of bison in fixed locations. The relationship between winter densities of bison at feeding sites and blood-sucking nematode A. sidemi prevalence in the BPF was showed by Radwan et al. (2010). Recent data from dissected individuals revealed much higher A. sidemi infection intensity for supplementary fed bison, which form large herds (Kołodziej-Sobocińska pers. commun.).

Previous studies proved that the significant reduction of parasite egg shedding in the warm, snow-free period could be caused by several factors including lower bison densities, higher food availability and overall improved condition of the bison, the “self-cure” phenomenon or a stronger immune response (von Szokolay and Rehbinder 1984; Møller et al. 2003; Altizer et al. 2006; Hines et al. 2007). In fallow deer, a “spring-rise” of Trichostrongylidae, Trichuris sp. and Aonchotheca sp. was observed during the late winter-early spring period following a “self-cure” response (von Szokolay and Rehbinder 1984). Studies of the “self-cure” phenomenon in sheep infected with Haemonchus contortus, under field conditions in East Africa, revealed that the onset of “self-cure”, as judged by a dramatic fall in faecal egg counts, was found to be simultaneous in sheep grazing on both infected and uninfected pastures (Allonby et al. 1973). Furthermore, the results of autopsies carried out before and after self-cure showed that a remarkable and equal loss of adult worm burdens had also occurred. These results indicate that “self-cure” of H. contortus infections under natural conditions occurs in the absence of reinfection and is apparently non-immunological in origin. Since the phenomenon is always associated with a period of significant rainfall, it has been suggested that new growth of pasture may be a significant etiological factor (Allonby et al. 1973). The limited rate of reinfection after bison spread out in spring may explain the pattern observed in the BPF.

It has been established that gastrointestinal nematode invasions are the most widely spread infections among wild ruminants. The health status of the bison population in the BPF has been monitored since its restoration to the wild in the 1950s (Dróżdż 1961). Previous studies were aimed mainly at the monitoring of the prevalence and intensity of parasite infections (Demiaszkiewicz et al. 2012; Karbowiak et al. 2014a, b) but rarely surveyed seasonal and sexual patterns or factors which may have influenced parasitic invasions (Radwan et al. 2010; Pyziel et al. 2011).

Our study revealed a significantly higher percentage of bison excreted Trichostrongylidae, Aonchotheca sp., Moniezia spp. and Nematodirus spp. eggs in winter and an influence of bison sex on parasite egg shedding in Trichuris sp. and Aonchotheca sp. This increases our understanding and deepens our knowledge about factors involved in the spread of parasites and their seasonal variation in wild populations of ungulates. This knowledge may benefit the conservation management of endangered species such as the European bison to reduce the impact of different threats and to design proper management activities.