Footrot is a worldwide economically important, debilitating disease caused by Dichelobacter nodosus. In sheep (Ovis aries), it is characterized by lesions of varying severity, depending on the strain, whereas Alpine ibex (Capra ibex) seem to develop severe lesions, whatever the strain. Healthy carriers occur in livestock but are rare in wild ruminants. Using a triangulation approach (retrospective questionnaire survey, necropsy database screening, and pathogen prevalence estimation in selected ibex colonies with and without footrot), we aimed at evaluating the importance of footrot in the ibex population, identifying potential risk factors for disease occurrence in this species, and defining the epidemiological role of ibex. Our study revealed that footrot occurs throughout the entire ibex territory (34% of the Swiss ibex colonies affected) but only as a sporadic disease (mostly one case per disease event), although the situation differed among footrot-positive colonies because half of them had experienced outbreak recurrences. Risk factor analysis for the occurrence of footrot in ibex colonies suggested an absence of an effect of meteorologic conditions, region, contacts with sheep or cattle (known to be very common healthy carriers of D. nodosus) and existing local disease control program. We found a significant effect only of contacts with sheep having footrot. Pathogen prevalence was very low in all investigated colonies. In conclusion, our results support previous data suggesting that ibex are susceptible spillover hosts, likely infected mainly by sympatric sheep displaying clinical signs.
Footrot is a debilitating, contagious, and economically important bacterial foot disease affecting domestic and wild ruminants (Rogdo et al. 2012; Wimmershoff et al. 2015), although it mainly concerns domestic sheep (Ovis aries). It is primarily caused by Dichelobacter nodosus (Kennan et al. 2011), but extrinsic factors, such as interdigital skin damage caused by trauma or humid weather conditions, can favor bacterial invasion (Zuba 2012; Mauldin and Peters-Kennedy 2016). Secondary infections with Fusobacterium necrophorum may exacerbate the disease (Swinton et al. 1998; Witcomb et al. 2014; Frosth et al. 2015). Fusobacterium necrophorum is highly host dependent and does not seem to be ubiquitous in soil but is shed into the environment through feces (Clifton et. al 2019). Dichelobacter nodosus can survive in the environment up to 14 d (Egerton et al. 1989; Stewart and Claxton 1993), and bacterial survival time may extend up to 6 wk when hoof trimmings from sheep are left behind (Winter 2009; Cederlöf et al. 2013).
In sheep, mild and severe disease forms are associated with the bacterial genes, AprB2 for benign strains, or AprV2 for virulent strains (Stäuble 2014). In Alpine ibex (Capra ibex), severe lesions are associated with both strain types (Moore-Jones et al. 2020). In wild ruminants, footrot occurs in ibex and mouflon (Ovis orientalis) in the European Alps (Moore-Jones et al. 2020). Because treatments are not practicable in wildlife, severe lesions can lead to death (Volmer et al. 2008; Wimmershoff et al. 2015). Considering that the ibex is protected in Switzerland and that its successful reintroduction (Meile et al. 2003) resulted in genetic bottlenecks potentially reducing population fitness (Biebach 2009; Grossen et al. 2018), footrot may be relevant to species conservation. Although it was shown that D. nodosus infections are likely not problematic in wildlife on a national level (Moore-Jones et al. 2020), intercolony differences have not yet been investigated.
The Swiss ibex population is subdivided into colonies with different genetic backgrounds (Biebach and Keller 2009) located throughout the Swiss Alps in remote rocky mountainous areas (up to 3,500 m elevation). Considering ibex habitat and body mass (80–100 kg for males), carcasses are rarely submitted for postmortem examination, and health data are limited. Furthermore, genetic and environmental variations among ibex colonies may cause differences in pathogen maintenance and in the occurrence and extent of footrot outbreaks. For ibex, the more frequent use of wet, north-facing pastures at middle elevations (the Prealps) during summer months, together with the more frequent encounters with sheep on such pastures, may increase the likelihood of infection with D. nodosus and footrot development (Delétraz 2002; Belloy et al. 2007).
Healthy carriers of D. nodosus occur worldwide in domestic species, such as sheep, cattle (Bos taurus), goats (Capra hircus), llamas (Lama glama), and alpacas (Vicugna pacos; Green and George 2008; Knappe-Poindecker et al. 2014; Locher et al. 2015; Ardüser et al. 2020). Interspecies pathogen transmission may occur as encounters between wild and domestic ruminants are common on summer grazing pastures. We previously showed that healthy carriers of D. nodosus are frequent in cattle (benign strains) and sheep (benign and virulent strains) in Switzerland, whereas prevalence is very low in wild ruminants. These data suggest that sheep and cattle are maintenance hosts for D. nodosus in the environment and act as the main source of infection, whereas wild ruminants seem to be only occasional spillover hosts (Moore-Jones et al. 2020). However, because current estimates for D. nodosus prevalence only exist on a country level, information on intercolony differences are lacking.
The aims of this study were to assess the occurrence and apparent importance of footrot in Swiss ibex colonies, identify factors possibly influencing disease occurrence, and better define the epidemiological role of ibex. Our results should allow for assessing the need for targeted footrot management measures in this protected wild ungulate.
MATERIALS AND METHODS
The study area was the Swiss Alpine region (26,835 km2 or 65% of the Swiss territory), which is home to 46 ibex colonies (Fig. 1). A colony located in the Jura Mountains was also included in the study. Altogether, the Swiss ibex population is estimated to be about 18,000 animals subdivided into 47 colonies distributed in 16 cantons (Swiss political subunits), four bioregions (Jura, Prealps, Alps, and South), and seven climate regions (southern Alps, northwestern Alps, central Alps, southwestern Alps, Engadin, northcentral Grisons, and west Jura). Definitions of bioregions and climate regions are from the Federal Office of Meteorology and Climatology (2019). Estimated colony sizes ranged from 27 (the Jura colony and one Alpine colony) to 1,965 animals (Federal Office of Environment 2018).
We used a triangulation approach, that is, the corroboration of information was obtained by multiple methods in which each approach has different key sources of potential bias that are unrelated to each other (Lawlor et al. 2016; Kutz and Tomaselli 2019). First, we collected data on past and current known footrot outbreaks through a retrospective questionnaire survey among game wardens. Second, we screened the digital database (1999–2019) of the Centre for Fish and Wildlife Health (FIWI; University of Bern, Bern, Switzerland), which conducts the national general surveillance program for wildlife health in Switzerland (Ryser-Degiorgis and Segner 2015), for footrot cases. Diagnosis of footrot at the FIWI was based on macroscopic lesions or PCR. Third, we collected interdigital swabs for PCR analysis (Stäuble et al. 2014) in four ibex colonies (two with recurrent outbreaks and two without reported outbreaks) to estimate and compare D. nodosus prevalence on a colony level.
We conducted the survey (April–July 2018) among 76 professional game wardens that had at least one ibex colony in their surveillance district, using an online open source questionnaire software (LimeSurvey 2007). We improved initial versions of the questionnaire after suggestions of game wardens and a wildlife ecologist specialized in ibex biology. With the support of the hunting authorities of all ibex cantons, we emailed a link to the final questionnaire to be filled out online by the concerned game wardens.
The questionnaire was composed of three main parts, including 15–274 closed questions, depending on the number of colonies within the surveillance district of the respondent and on the number of recorded outbreaks. Space was provided for comments. If necessary, we subsequently contacted respondents to clarify uncertainties or to obtain more-detailed information.
Part I of the questionnaire gathered personal information (name, canton, years in service) and the number of ibex colony or colonies in the surveillance districts. Part II focused on the mentioned ibex colony or colonies: colony identification (name, number) and number of observed footrot outbreaks (animals were considered positive for footrot if they presented with footrot-like lesions, as previously described [Belloy 2007; Wimmershoff 2015]). An outbreak was defined as the occurrence of one or more cases of footrot in the same colony, with a maximum interval of 6 mo between any two cases; known interactions (direct contacts between two individuals, simultaneous or consecutive use of the same pasture) were also noted between ibex and livestock (sheep, cattle, goats, and South American camelids) or other wild ungulates (Alpine chamois [Rupicapra rupicapra]; red deer [Cervus elaphus] and roe deer [Capreolus capreolus]); along with occurrence of salt licks (mainly for domestic ungulates) and time period (predefined as 2000 or earlier, 2001–09, and 2010–17) of known outbreaks in sheep. Part III collected information on every outbreak that occurred in each ibex colony, including the duration (first–last case), the number of affected animals, and whether the diagnosis had been confirmed by laboratory analysis (yes versus no, benign versus virulent strains of D. nodosus, and name of laboratory).
We exported the data from the online survey to a Microsoft Excel (2016 version, Redmond, Washington, USA) spreadsheet. Questionnaires from respondents that appeared to be duplicates from the same Internet protocol address were excluded from the analysis. For statistics, we used the R software (version 3.5.1, R Foundation for Statistical Computing, Vienna, Austria). Maps were generated using QGIS 2.18.16 Las Palmas (Free Software Foundation Inc., Boston, Massachusetts, USA).
Risk factor analysis
As a first step, we performed a risk factor analysis with the reported occurrence of footrot outbreaks (yes/no) per year on a colony level as a dependent variable. Independent variables included 6-yr-based continuous and six categorical variables (Table 1). For the categorical variables, we assumed the answer of the respondents was true for the entire observation period for the corresponding colony, except for the reported occurrence of sheep with footrot, which was attributed to predefined periods (before 2000, 2001–09, and 2010–17). We performed a univariable analysis using a generalized mixed model, with colonies included as random effects, applying the glmer function of the lme4 packages for R. Because only one colony represented the bioregion Jura, that weather variable was not considered for further analysis. Because the values of the variable yearly rainfall sum and sheep with footrot did not converge in the chosen univariable glmer model, we used a log model with the glm function of the glm2 package for R (Marschner et al. 2018) without taking into account colonies as a random effect. For the variable presence versus absence of sheep in the ibex colony, we used the Fisher's exact test because all colonies had contact with sheep. We then tested all factors associated with P<0.2, within a multivariable mixed-effect model, performing a manual backward elimination procedure with a cutoff level at P<0.05.
In a second step, the same variables (Table 1) were analyzed as potential risk factors for the frequency of footrot outbreaks among colonies that reported at least one outbreak during the entire observation period. We defined outbreak frequency as rare versus recurring, based on disease incidence (number of outbreaks in the colony/number of years under observation by the respondent) using the incidence median (0.08) as a cutoff. Observation time ranged from 2 to 23 yr (mean, 20.4 yr). Rare occurrence was defined as an incidence of <0.08 (only one outbreak observed in a colony over the entire observation period). Recurring outbreaks were defined as an incidence of >0.08 (two or more outbreaks). We used a generalized linear model with the glm function of the glm2 package (Marschner et al. 2018) for univariable analysis. Because none of the variables were associated with P<0.2, we did not perform a multivariable analysis. For the variable presence versus absence of salt licks, we used the Fisher's exact test because there were salt licks in all colonies.
Prevalence estimation within colonies
Taking advantage of an on-going footrot-prevalence study in free-ranging wild ruminants in Switzerland conducted on a country level (Moore-Jones et al. 2020), we collected and analyzed interdigital swab samples in four colonies to estimate D. nodosus prevalence within, and compare it among, colonies, hypothesizing that pathogen prevalence may be higher in colonies with disease cases. Two large colonies in the canton of Grisons with, and two colonies without, reports of footrot outbreaks were selected based on information collected from the questionnaire survey. Another characteristic of these four colonies was that they had a hunting bag large enough to ensure the required sample size. The two colonies with recurrent outbreaks (two or more outbreaks) were defined as cases, and colonies without outbreaks were considered controls. We used the free online tool WinEpi (De Blas 2010) for sample size calculation. Input parameters for sample size calculations included population size, design prevalence of 5%, desired precision (0.05), and confidence level (95%). We used the following estimated population sizes for the sample size calculation (Federal Office of Environment 2018): case colony 1 (Julier, colony size=1,215), case colony 2 (Safien-Rheinwald-Mesocco, colony size=1,701), control colony 1 (Albris, colony size=1,395), and control colony 2 (Flüela-Rätikon, colony size=1487). We aimed to sample 69–70 animals per colony (i.e., 139 cases and 140 controls). The material obtained consisted of 142 samples from case colonies and 144 samples from control colonies (Table 2).
Questionnaire part I
All 76 game wardens responded to the questionnaire, and all 47 ibex colonies were represented (100% response rate). The mean time of service for all respondents was 15.1 yr (range, 4.5 mo to 39 yr). In total, 48 outbreaks were reported from 34% of the colonies (16/ 47), with up to 14 outbreaks per colony (Table 3 and Fig. 2) and 62% of the ibex cantons (10/ 16) reported outbreaks.
Questionnaire part II
Interspecies interactions in both footrot-positive (one or more outbreak) and footrot-negative (no outbreak) colonies most frequently involved chamois, followed by sheep, cattle, and goats (Fig. 2). The presence of salt licks was reported in 74% of the colonies (35/ 47) with no information for one colony. Outbreaks of footrot in sheep within the territory of ibex colonies were reported in 43% of the colonies (20/47).
Questionnaire part III
Outbreaks were recorded from 1996 to 1997 and again from 2003 to 2018, whereas no outbreaks were noticed during the 5-yr interval between those two periods. The number of affected animals per outbreak ranged from one to >20 animals per colony. A total of 29% (14/48) documented outbreaks were reported from the same colony. However, those 14 outbreaks involved only one to five animals each (Table 4).
Of the 29 outbreaks for which a duration was given, footrot cases were mostly reported to be first observed in summer (38%, 11/29) and in spring (45%, 13/29). The involvement of D. nodosus was confirmed by PCR in 31% of the outbreaks (15/48). According to the respondents, most cases (8/15, 53%) were associated with virulent strains, a few (2/15, 13%) with benign strains, and the type of strain was not specified for the remaining cases (5/15, 33%).
No case of footrot was documented at the FIWI from 1999 to 2009 (i.e., in the periods 1999–2000 and 2001–09), whereas there were 10 cases confirmed from 2010 to 2017, and one case in 2018. Analysis by PCR confirmed the involvement of D. nodosus in 68% (11/16) of the suspect ibex. Three animals (including one with suggestive lesions) were too autolytic for PCR, and two had unspecific lesions. All 10 cases sent to the FIWI corresponded to outbreaks reported in the questionnaire survey. These animals were adult males with a mean age of 8.6 yr (range, 5–12 yr) originating from six colonies in four cantons, of which three are located in the bioregion Alps and one in the Prealps. For four animals, strain identification was performed by PCR (Stäuble et al. 2014), revealing the involvement of benign strains in two outbreaks (three animals investigated) and of virulent strains in one outbreak (one animal investigated).
Risk factor analysis
In the univariable analysis, the variable sheep with footrot, observed in the colony area was significantly associated with the occurrence of footrot outbreaks in ibex (P=0.001). For the variables colony size (P=0.062) and temperature in spring and summer (P=0.088), there was only a tendency of association with footrot outbreaks (Table 5). None of the other variables showed an association with the occurrence of outbreaks (P≥0.265). In the multivariable analysis, the variable sheep with footrot, remained statistically significant (P=0.045). For the comparison between colonies with a rare (one reported outbreak) versus recurrent outbreaks (two or more), none of the tested variables were associated with a P<0.2 value in the univariable analysis.
Considering climatic differences between periods with observed outbreaks (1996–97 and 2003–18) and the period without observed outbreaks (1998–2002), all three rainfall variables (yearly, spring-summer, and autumn-winter) had a protective effect on the occurrence of footrot. There was no effect of the three temperature variables (yearly, spring-summer, and autumn-winter).
Prevalence estimation on a colony level
Of the 142 samples from case colonies, only two animals (1%), one from each colony, were positive for benign and virulent strains, respectively. Both animals were adult male ibex with severe footrot lesions. No healthy carriers were detected, neither in the case nor in the control colonies, and prevalence did not differ between them (Table 2).
Footrot occurrence varied among colonies. Prevalence of D. nodosus was very low in the four investigated colonies, independently of disease occurrence. Importantly, in these selected colonies, D. nodosus (benign or virulent) was only detected in ibex with foot lesions. This is in agreement with previous studies reporting the association between D. nodosus of any strain and footrot lesions in ibex (Wimmershoff et al. 2015; Moore-Jones et al. 2020) and supports previous data suggesting that ibex do not maintain D. nodosus (Moore-Jones et al. 2020).
The 100% response rate of the questionnaire survey allowed us to collect extensive information on the occurrence of past and present footrot outbreaks in Swiss ibex colonies. The answers were provided by cantonal game wardens (i.e., trained professionals) whose duties included field observations of wildlife and reporting of disease. Importantly, they are familiar with visual identification of footrot because of a study conducted in the 1990s (Belloy et al. 2007), meaning that they did not send all cases for examination but that the information they provided should be reliable. Nevertheless, the number of outbreaks (per colony and in total) may have been underestimated because not all diseased and dead animals can be detected in wildlife (Ryser-Degiorgis 2013) because of potential landscape inaccessibility and weather hazards. Furthermore, the answers were based on memorized events, which may have caused underreporting or misreporting. The problem of potentially missing outbreaks was partly overcome in data analysis by use of the incidence of outbreaks rather than their total number per colony. Working with retrospective questionnaire surveys entails a risk of data inaccuracy, but participatory approaches are the only option to collect information that is not systematically reported. The associated data weaknesses can, at least partly, be overcome thanks to triangulation, (Lawlor et al. 2016; Kutz and Tomaselli 2019). In Switzerland, case selection for pathologic examination of wildlife is done by field partners participating in the national health surveillance program. Therefore, they rarely report cases of diseases with typical external lesions (such as footrot) to laboratories because they can largely identify the disease themselves (Ryser-Degiorgis and Segner 2015; Pewsner et al. 2017; Pisano et al. 2019). Previous studies have already demonstrated that questionnaire surveys represent a useful tool in wildlife research that aims to better understand disease dynamics in wildlife populations (Casaubon et al. 2012; Kutz and Tomaselli 2019; Tomaselli et al. 2019; Pisano et al. 2019).
Our study revealed that footrot occurs throughout the entire ibex territory (affecting 34% of Swiss ibex colonies) but that it is a sporadic disease in this species. This suggests that footrot is currently of negligible importance on a national population level. However, a few colonies have experienced outbreak recurrence, hinting at potential intercolony differences.
In contrast to suggestions made in former studies (Wimmershoff et al. 2015), humid weather conditions did not emerge as a risk factor for the occurrence of footrot outbreaks and, therefore, could not explain intercolony differences. Even when comparing climatic differences between periods with and without observed outbreaks, meteorologic data did not provide a biologic explanation for occurrence or nonoccurrence of footrot. Colony size and on-going footrot control programs in domestic animals did not make a difference either.
We identified only the occurrence of sheep with footrot pasturing on territories of ibex colonies as a risk factor for footrot occurrence in ibex, but this has to be taken with caution because footrot data in sheep were not an accurate variable. The most frequently reported interspecies interactions, whether in colonies with or without reported footrot outbreaks, involved sheep and cattle, which was in line with previous studies (Ryser-Degiorgis et al. 2002; Casaubon et al. 2012; Moore-Jones et al. 2020), but their occurrence did not come out as a risk factor for footrot in ibex, which is surprising considering how widespread the pathogen is in the livestock population (Greber and Steiner 2013; Zingg et al. 2017; Ardüser et al. 2020). Other factors, such as preexisting foot lesions, reduced fitness, co-infections with other microorganisms, or genetic background of the colony, may have a role in footrot occurrence in ibex. An in-depth analysis is difficult to perform, considering the sporadic character of the disease.
Cattle and sheep are maintenance hosts for both benign (sheep and cattle) and virulent (sheep) strains of D. nodosus (Ardüser et al. 2020), whereas infections are rare in ibex (Moore-Jones et al. 2020). It is, therefore, likely that domestic livestock is the source of infection for ibex. Molecular studies are required to draw solid conclusions, but the hypothesis of another source of infection in wildlife habitat is highly questionable. Measures to reduce virulent D. nodosus prevalence in domestic sheep may make a difference in the occurrence of footrot in Swiss ibex colonies because our data suggest that sheep with footrot sharing pastures with ibex influence footrot occurrence in the latter. Because D. nodosus primarily persists on diseased sheep feet, they may be the main source of infection for naïve sheep (Clifton et al. 2019) and possibly for sympatric ibex. However, it is likely that the control program in sheep will not reduce outbreaks linked to the benign strains of D. nodosus. The sporadic occurrence of the disease in ibex does not justify costly measures to avoid contacts between livestock and wildlife.
Although the presence of salt licks did not make a difference in footrot occurrence, such aggregation points promote interspecies contacts and concentrate pathogens in the environment, favoring interspecific transmission (Schmitt et al. 1997; Palmer and Whipple 2006; Richomme et al. 2006).
All ibex with footrot signs that were necropsied at the FIWI, as well as the two positive individuals mentioned in this study, were adult males as previously found by others (Delétraz 2002; Belloy et al. 2007; Wimmershoff et al. 2015). Behavioral differences between sexes, such as the segregation of groups of males in spring and summer on pastures located at lower elevations, may lead to more frequent encounters between adult male ibex and livestock, compared with females and their young, who tend to stay in higher, rockier habitats (Bon et al. 2001; Delétraz 2002; Belloy et al. 2007). The frequency and intensity of interspecies interactions likely determine the occurrence of footrot in ibex, reinforcing the idea that efforts toward the elimination of aggregation points may be a worthwhile approach to prevent pathogen spillover.
We are grateful to all game wardens, hunters, and cantonal hunting inspectors who contributed to the sample collection and to Stefanie Gobeli Brawand for coordinating and supervising sample analysis. This study was supported by a grant of the Fondation Bruno Galli-Valerio c/o Service de la consommation et des affaires vétérinaires.