One Health 16 (2023) 100479 Contents lists available at ScienceDirect One Health journal homepage: www.elsevier.com/locate/onehlt Using integrated wildlife monitoring to prevent future pandemics through one health approach P. Barroso a, D. Relimpio a, J.A. Zearra a, J.J. Cerón b, P. Palencia a, c, B. Cardoso a, e, E. Ferreras a, M. Escobar d, G. Cáceres f, J.R. López-Olvera d, *, C. Gortázar a a SaBio Instituto de Investigación en Recursos Cinegéticos (IREC) CSIC-UCLM-JCCM, 13071 Ciudad Real, Spain b Interdisciplinary Laboratory of Clinical Pathology, Interlab-UMU, University of Murcia, 30003 Murcia, Spain c Department of Veterinary Sciences, University of Turin, 10095 Turin, Italy d Wildlife Ecology & Health (WE&H) research group and Servei d’Ecopatologia de Fauna Salvatge (SEFaS), Departament de Medicina i Cirurgia Animals, Universitat Autònoma de Barcelona (UAB), Bellaterra, 08193 Barcelona, Spain e CIBIO/InBio, Centro de Investigação em Biodiversidade e Recursos Genéticos, Universidade do Porto, Porto, Portugal f Epidemiology Department, Ministry of Agriculture, Fisheries and Food, Madrid, Spain A R T I C L E I N F O A B S T R A C T Keywords: In the One Health context, Integrated Wildlife Monitoring (IWM) merges wildlife health monitoring (WHM) and Anthropogenic imbalances host community monitoring to early detect emerging infections, record changes in disease dynamics, and assess Disease risk the impact of interventions in complex multi-host and multi-pathogen networks. This study reports the Host community deployment and results obtained from a nationwide IWM pilot test in eleven sites representing the habitat di­ Integrated wildlife monitoring One health versity of mainland Spain. In each study site, camera-trap networks and sampling of indicator species for anti­ Wildlife health monitoring body and biomarker analysis were used to generate information. The results allowed identifying differences in biodiversity and host community characteristics among the study sites, with a range of 8 to 19 relevant host species per point. The Eurasian wild boar (Sus scrofa) was the most connected and central species of the host communities, becoming a key target indicator species for IWM. A negative relationship between biodiversity and disease risk was detected, with a lower number and prevalence of circulating pathogens in the sites with more species in the community and larger network size. However, this overall trend was modified by specific host- community and environmental factors, such as the relative index of wild boar - red deer interactions or the proximity to urban habitats, suggesting that human-driven imbalances may favour pathogen circulation. The effort of incorporating wildlife population monitoring into the currently applied WHM programs to achieve effective IWM was also evaluated, allowing to identify population monitoring as the most time-consuming component, which should be improved in the future. This first nationwide application of IWM allowed to detect drivers and hotspots for disease transmission risk among wildlife, domestic animals, and humans, as well as identifying key target indicator species for monitoring. Moreover, anthropogenic effects such as artificially high wildlife densities and urbanisation were identified as risk factors for disease prevalence and interspecific transmission. 1. Introduction Monitoring (WHM) is essential to detect changes in disease occurrence and pathogen emergence, although WHM systems still need substantial Shared infections at the wild-domestic interface are caused by improvement [4,5]. transmissible pathogens maintained by at least one host species from While currently most WHM programs rely mainly on general and either compartment [1], and may be relevant to animal health, public targeted disease surveillance [6], monitoring wildlife population is health, wildlife management, and biodiversity conservation, with eco­ required to effectively achieve Integrated Wildlife Monitoring (IWM) nomic, sanitary and ecological impacts [2]. Wildlife is key in the from a One Health perspective [4]. The collection of long-term and epidemiology of shared diseases [3], and therefore Wildlife Health large-scale information through IWM provides knowledge into the * Corresponding author. E-mail address: jordi.lopez.olvera@uab.cat (J.R. López-Olvera). https://doi.org/10.1016/j.onehlt.2022.100479 Received 19 August 2022; Received in revised form 21 December 2022; Accepted 22 December 2022 Available online 26 December 2022 2352-7714/© 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). P. Barroso et al. One Health 16 (2023) 100479 Fig. 1. Map of the 11 pilot monitoring sites with a division of mainland Spain into five large bioregions (1–5) according to the Spanish Wildlife Disease Surveillance Scheme. Silhouettes of livestock species present at each point are shown. The presence of fences and protected areas are indicated with icons. epidemiology of pathogens and their establishment as endemic in host 2.2. Camera trap survey and calculation of abundance and cross-species populations, which is essential to control shared infections [7]. Thus, a interactions balanced IWM system including wildlife abundance monitoring is key to detect emergent pathogens and changes in pathogen dynamics, to crit­ A grid of 20 camera traps (Browning Strike Force HD ProX, Browning ically assess wildlife disease hazards and the impact of interventions, Arms Company®, Morgan, Utah, USA) was deployed for two months in and to better understand complex multi-host and multi-pathogen net­ each study site following previous studies [15]. works [8]. The cost and difficulties of covering each disease and host have 2.2.1. Random Encounter Model method promoted the search for indicator species [9] and non-specific indicators The Random Encounter Model (REM) has been successfully applied of population health, including acute-phase proteins (APP), indicators of to estimate animal density from camera trap data for different mammal redox status, and markers of immune system status [10–12]. species inhabiting an area, by modelling the rate of random encounters This study aims to report the implementation of the first nationwide between animals and camera traps without the need for individual pilot trial of balanced IWM in Europe, using a network of 11 pilot recognition [16,17]. Density (individuals/km2) was estimated as: monitoring sites in Spain. The insights obtained in this pilot IWM trial y π will help to further develop a comprehensive IWM and serve as a D= t v⋅r⋅(2 + α) reference for implementing IWM systems in other regions. where y is the number of encounters of the target species, t is the survey 2. Material and methods effort, v is the average daily distance travelled by an individual (day range) [18], and r and α are the radius and angle of the camera trap 2.1. Study sites and target species detection zone, respectively. All analyses were carried out using the ‘camtools’, ‘trappingmotion’, and ‘emdbook’ R packages [19,20]. The study sites selected represented the main habitats, climates and REM was applied to wild boar, red deer (Cervus elaphus) and red fox wildlife management systems of mainland Spain [13] (Fig. 1). The (Vulpes vulpes) as the most frequently detected species. Relative abun­ Eurasian wild boar (Sus scrofa) was selected as the target species due to dance indexes, namely trapping rate and relative occupancy index, were its abundance, distribution and epidemiological relevance for infections calculated for the remaining, less frequently detected species [21] shared between wild and domestic animals and humans [14]. (Table S1). 2 P. Barroso et al. One Health 16 (2023) 100479 Fig. 2. Social networks of the indirect interactions between (a) all the species, (b) all the potentially implicated hosts in the transmission of the Mycobacterium tuberculosis complex (MTC; red deer, roe deer, fallow deer, red fox, wild boar, Egyptian mongoose, Eurasian badger, cattle, sheep and goat), (c) the most epide­ miologically relevant hosts for the MTC in Spain (red deer, fallow deer, wild boar, badger, cattle, sheep and goat), obtained by camera trap records in eleven pilot monitoring sites in Spain. Nodes represent different species and edges represent indirect interactions recorded by camera traps between two nodes. The size of the nodes represents the degree and the weight of the edges represents the number of interactions between nodes, evidencing the centrality of the wild boar and the weight of wild boar-red deer interactions. Grey silhouettes: domestic; black: wild. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 2.2.2. Cross-species interactions the set of 11 study sites) for all the hosts potentially participating in the Direct interactions were defined as the simultaneous presence of transmission of the Mycobacterium tuberculosis complex (MTC), as well as individuals in a photograph. Indirect interactions were defined as the retaining only the most epidemiologically relevant hosts for MTC in photo-capture of individuals by the same camera trap during a time Spain. window of 24 h, based on the environmental survival of the target The metrics of the nodes and the network were estimated for all the pathogens [22,23]. The relative direct and indirect intra and inter- networks using the “igraph” R package (Table S2) [24,25]. The statistical specific interaction indexes for the epidemiologically relevant pairs of differences in network and node metrics among species and sites were species were calculated (Table S1). assessed using a Kruskal-Wallis test. 2.3. Social network characterization 2.4. Health monitoring Twelve static social networks were constructed, one for each study A total of 398 hunter-harvested wild boar from the 11 study sites site and one for all of them together, using the “igraph” R package [24]. (range: 7 to 60, mean ± standard error (SE): 36.2 ± 0.2) were sampled Additional networks were constructed (both for each study site and for between 2020 and 2021. Sex and age were recorded, classifying the wild 3 P. Barroso et al. One Health 16 (2023) 100479 boars as juveniles (younger than 12 months), yearlings (between 12 and The average prevalence of antibodies (% ± confidence intervals (CI) 95%) against Aujeszky’s disease virus (ADV), porcine circovirus type 2 (PCV2), hepatitis E virus (HEV), Mycobacterium tuberculosis complex (MTC), Brucella suis and Toxoplasma gondii, and mean concentration (± standard error) of acute phase proteins (APPs; including C-reactive protein (CRP), haptoglobin (Hp) and paraxonase 1 (Pon1)), a biomarker of redox balance 6.4–26.2 24 months) or adults (older than two years), according to tooth eruption. 95% CI (serum thiols -Thiol- and cupric reducing antioxidant capacity -Cuprac-), and a marker of the immune system (adenosine deaminase -ADA-) in eleven pilot monitoring sites in Spain. Sample size (N) is shown. 2.4.1. Specific health indicators Blood samples were collected from the endocranial venous sinus Toxoplasma gondii [26]. Sera were obtained by centrifugation at 400g for 5 min and kept at Pon1 (IU/mL) − 20 ◦ C until assayed for antibodies against six different pathogens (0.0–46.2) (Table S3). 0.7 ± 0.2 (0.3–2.2) 16.3 2.4.2. Non-specific health indicators Serum concentrations of two positive APPs (Haptoglobin -Hp- and C- reactive protein -CRP-), a negative APP (paraoxonase -Pon1-), two bio­ 10.5–34.9 95% CI markers of redox balance (serum thiols -Thiol- and cupric reducing antioxidant capacity -Cuprac-), and a marker of immune response (adenosine deaminase -ADA-) were determined as previously described [11,27,28], using an automated analyser (Olympus AU 600, Beckman (22.9–215.9) 84.1 ± 20.4 Brucella suis (0.0–57.1) Coulter). All the assays were validated in the indicator species used in (IU/L) the study, providing inter and intraassay imprecision values lower than ADA 22.7 15%, and being linear after serial sample dilutions with saline. 2.5. Land use information 2.0–16.2 95% CI The CORINE land cover categories were compressed into seven broad land-use classes: woodland, scrubland, herbaceous grassland, 857.6 ± 856.8 (0.4–9426.0) bare land, arable land, urban land and water habitats [29]. The per­ (0.0–36.7) (mmol/L) centage cover of each land-use class was calculated for each sampling Cuprac MTC site. Additionally, the straight-line distances in metres from the centroid 9.1 of each sampling site to the nearest road, urban area, water point, and river were calculated using qGIS 3.4 [30]. 17.1–26.0 95% CI 2.6. Statistical approach and methods To correct for intersite variability, the concentration and activity (5.0–32.7) (mmol/L) 0.3 ± 0.1 values of the unspecific health indicators were transformed to a per­ (0.0–1.4) centage of the maximum value for each study site (0–100%). The Thiol HEV 20.1 resulting percentages were grouped into four categories, namely biomarker of inflammation (positive APPs: Hp and CRP), biomarker of stress (negative APP: Pon1), biomarker of immune system activation 24.4–54.8 95% CI (ADA) and biomarker of redox balance (Cuprac and Thiols). The rela­ tionship of the four groups with the presence and prevalence of anti­ bodies against the pathogens studied was evaluated using chi-squared and a Kruskal-Wallis test, respectively. We evaluated spatial differences (3.2–78.3) 0.3 ± 0.1 (0.0–0.8) in these biomarkers through a Kruskal-Wallis test with Dunn’s multiple- PCV2 (g/L) 39.6 comparison post hoc analyses [31]. Hp Bivariate analyses were performed to assess the association between populational, environmental and health variables by using Pearson’s 21.0–43.6 correlation and Spearman’s rank tests according to the distribution of 95% CI the continuous variables and sample size, and Wilcoxon’s sum rank test for the categorical variables (Table S4). The significance of the differ­ ences between variables was evaluated using Fisher’s test. 27.8 ± 8.3 (0.0–58.3) (3.3–71.1) Finally, generalized linear mixed models (GLMMs) were fitted to (μg/mL) assess the relationship between unspecific biomarkers of health, envi­ ADV 32.3 CRP ronmental and population-related variables, and seropositivity to pathogens, selecting as response variables first the biomarker of inflammation (Hp and CRP), and then the seropositivity to MTC (bino­ 398 398 N N minal response). As a previous step, collinearity among the explanatory variables was explored [32], and principal component analysis (PCA) Average concentration (Range) was performed to obtain two uncorrelated factors; agricultural and Average prevalences (Range) grassland-water habitats (see Table S5). The variables included in the GLMMs are shown in Table S6. The GLMMs were fitted with a gaussian error distribution and the identity link function The models with an increase in Akaike’s Information Criterion lower than two (ΔAIC <2) were considered suitable to explain the variability observed in the response variable [33]. Once the best model was selected, normality and Table 1 the absence of residual pattern in data variation were checked. The GLMMs were fitted in the library “lme4” [34]. Significance was set at 4 P. Barroso et al. One Health 16 (2023) 100479 Fig. 3. Markers of health status in wild boar from 11 pilot monitoring sites in Spain. Outliers are shown as blue triangles. Percentage of maximum serum con­ centration (%) of (a) biomarker of inflammation in relation to individual seropositivity to Mycobacterium tuberculosis complex (MTC), (b) biomarker of stress in relation to the prevalence of antibodies against porcine circovirus type 2 (PCV2), (c) biomarker of immune system activation in relation to individual seropositivity to PCV2, (d) biomarker of immune system activation in relation to the prevalence of antibodies against B. suis, (e) biomarker of redox balance in relation to individual seropositivity to T. gondii, and (f) biomarker of redox balance in relation to the prevalence of antibodies against hepatitis E virus (HEV). Integrated wildlife monitoring results on 11 pilot monitoring sites in Spain. Mean prevalence of antibodies against (g) Mycobacterium tuberculosis complex (MTC) in relation to the density of red deer (individuals/km2) obtained by Random Encounter Model, (h) MTC in relation to the relative interaction index of red deer (Rd)-wild boar (Wb), and (i) Aujeszky’s disease virus (ADV; blue), Hepatitis E virus (HEV; yellow) and MTC (grey) in relation to the latitude, at study site level. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 0.05. All analyses were conducted using R software 4.0.2 [35]. the wild boar was the key and most connected species of the system, both globally and among the relevant species in MTC epidemiology, as 3. Results shown by its presence in all study sites (Tables S7 and S8), its connec­ tions with other species both overall and among the species relevant in 3.1. Population monitoring MTC epidemiology (Figs. 2, S1 and S2) and its higher centrality metric values (Fig. S3). Also, the interaction between wild boar and red deer The estimated population density values for wild boar (ind/km2; was the most intense one, both overall and when only considering mean ± SD (range): 13.39 ± 0.14 (0.52–57.49), red deer (18.99 ± 0.19 species relevant to MTC epidemiology (Figs. 2, S1 and S2). (0–41.21), and red fox (1.15 ± 0.01 (0.31–1.85)) are shown in Table S7. The wild boar was detected in all 11 study sites, red fox in ten sites and red deer in eight (Table S7). All the species detected in each site (mean 3.2. Health monitoring ± SE: 14 ± 0.03; range: 8–19) are shown in Table S8. Table S9 shows the relative interaction index of each pair of species The prevalences of antibodies against Aujeszky’s disease virus (percentage of camera traps per day) and the percentage of camera traps (ADV), porcine circovirus type 2 (PCV2), hepatitis E virus (HEV), MTC, (%) in which a given interaction was recorded. Brucella suis and Toxoplasma gondii detected in wild boar sera are shown The global and study site host networks are shown in Figs. 2 and S1, in Table S11 and Table 1. respectively (doi:https://doi.org/10.5281/zenodo.6861521). Fig. S2 The serum concentration of APPs, Thiols, Cuprac and ADA are shown displays the networks for all the epidemiologically relevant species in Table S12, Table 1 and Fig. S4. The percentage of positive APPs (Hp implicated in the transmission and maintenance of the MTC in Spain by and CRP, biomarker of inflammation) was significantly higher in wild study site. The network metrics are compiled in Table S10. Fig. S3 shows boar positive to MTC antibodies (W: 175.5, p = 0.02). Regarding the node centrality values for each species. negative APPs, the activity of Pon1 positively correlated with the Although red deer occasionally reached higher densities (Table S7), prevalence of antibodies against PCV2 (R: 0.62, p = 0.03). ADA (biomarker of immune system activation) was positively associated with 5 P. Barroso et al. One Health 16 (2023) 100479 detected was negatively correlated with the diversity of host species (Fig. 4a) and positively correlated with the relative interaction index of wild boar-red deer interactions (Fig. 4b). The network size of each study site was negatively correlated with wild boar seroprevalence against MTC (R: − 0.68, p: 0.02) and ADV (R: − 0.63, p: 0.04; Fig. S5a). The eigenvector centrality of wild boar was positively correlated with the prevalence of antibodies against PCV2 in this species (R: 0.77, p < 0.01; Fig. S5b) and with its population density (R: 0.64, p: 0.04; Fig. S5c). The best GLMM selected (Table S15) explained the seropositivity to MTC (as specific marker of health status) as a function of serum con­ centration of the biomarker of inflammation (CRP and Hp), wild boar density, and population management in terms of fencing (wi = 0. 63, R2 = 31.30%, Table S16a). Wild boar with higher concentrations of CRP and Hp, as well as those from high-density fenced populations had a higher probability of being seropositive to MTC (Table S16a). The variability in Hp and CRP as unspecific markers of health status was explained by the seropositivity to MTC (specific marker of disease), wild boar density, population management in terms of fencing, and proportion of urban lands (wi = 0.66, R2 = 14.40%, Table S16b). Consequently, wild boar seropositive to MTC had higher concentrations of CRP and Hp and were sampled in dense non-urban fenced populations (Table S16b). Finally, Fig. 5 shows the estimated distribution of the costs associ­ ated to the implementation and development of the IWM experimental pilot trial. 4. Discussion By combining WHM with wildlife population assessment using a One Health approach [4], this pilot IWM yielded novel insights on the effect of host community composition, population abundance and manage­ Fig. 4. Integrated wildlife monitoring results on 11 pilot monitoring sites in ment on the epidemiology of shared infections [1]. This opens the door Spain. Number of different pathogens for which antibodies were found by for the establishment of community-based wildlife health indicators and ELISA in relation to (a) the number of different species detected by camera traps the identification of hotspots for disease spread and transmission [9]. per study site (R = -0.61, p ≤ 0.05), and (b) the relative interaction index of red Higher biodiversity and larger network size reduced the number and deer (Rd) - wild boar (Wb) at study site level (R = 0.49, p ≤ 0.05). Each dot is surrounded by the local network obtained by camera trapping in such pilot prevalence of circulating pathogens (Figs. 4a and S5a) as proposed by point, with the number of different species recorded (nodes) and the number of the “dilution effect” hypothesis [36] but opposite to previous studies interactions (edges). The link between red deer (brown node) and wild boar [37]. The “dilution effect” hypothesis postulates that pathogens are less (yellow node) is highlighted in grey or black, according to the number of in­ likely to encounter their competent hosts, i.e., individuals in which a teractions, in graph -b-. (For interpretation of the references to colour in this given pathogen can replicate and spread to a new susceptible individual, figure legend, the reader is referred to the web version of this article.) in species-rich communities. Thus, biodiversity would lead to lower host exposure to pathogens and reduced transmission rates [36,38]. How­ PCV2 (W: 423.0, p < 0.01) and the prevalence of antibodies against ever, this overall trend was modulated by specific host-community and B. suis (R: 0.75, p < 0.01). Finally, the percentage of the biomarker of the environmental factors, such as the density and interaction of wild boar redox balance (Cuprac and Thiols) correlated with the positivity to and red deer, two key reservoir species for MTC (Fig. 3g and h) [39,40]. T. gondii (W: 437.0, p = 0.05) and the prevalence of antibodies against More importantly, the interaction of these two species also increased the HEV (R: 0.69, p = 0.03) (Fig. 3a-3f). overall number of pathogens detected (Fig. 4b), suggesting that popu­ lation management leading to imbalances may favour pathogen circu­ lation [37]. Similarly, the higher density and spatial aggregation of wild 3.3. Integrated monitoring boar and other ungulates in the intensively managed fenced hunting estates of south-western Spain favours disease transmission at the The correlations between population and environmental variables, wildlife-livestock interface [22,41], increasing MTC prevalence and specific (antibody prevalences) and unspecific markers of disease, (Table 16a) and explaining the negative correlation of ADV and MTC are shown in Tables S13 and S14, respectively. The prevalence of anti­ seroprevalence with latitude (Fig. 3i) [41–43]. The higher intraspecific bodies against T. gondii in wild boar was positively correlated with the wild boar contact rates in these overabundant populations would also trapping rate of free-ranging domestic cats and the percentage of urban explain the negative relationship between wild boar network centrality habitat (Table S13). The prevalence of antibodies against MTC in wild and the seroprevalence of the pig pathogen PCV2 (Fig. S5b) [44], further boar was positively correlated with red deer density and the relative supporting the dilution effect of biodiversity [36]. Apart from wildlife index of wild boar-red deer interactions (Fig. 3g and h). Latitude was management, environmental anthropization - i.e., a set of processes negatively correlated with the prevalence of antibodies in wild boar transforming or adapting landscapes by human actions [45]- also fav­ against ADV and MTC and positively for HEV (Fig. 3i). The prevalence of oured pathogen circulation [46,47], as shown by the higher T. gondii antibodies against ADV in wild boar was positively correlated with the prevalence in wild boar from periurban habitats with stray cats, distance to roads or urban areas, and the prevalence against B. suis was agreeing with previous reports [48]. negatively correlated with the relative interaction index of wild boar-red The distribution, abundance and centrality of wild boar (Tables S7 deer interactions (Table S13). Finally, the number of different pathogens and S8, Figs. S3 and S5c) and its connections with the other species, 6 P. Barroso et al. One Health 16 (2023) 100479 Fig. 5. Sampling effort of carrying out integrated wildlife monitoring (IWM) in each pilot monitoring site in Spain (n = 11). Effort (full working days (8 h) per person) required for population (69.2% of effort; including field work, image processing, calculation of the alternative abundance indicators, and application of Random Encounter Model (REM) method from camera trap data to estimate wildlife abundance) and health monitoring (30.7% of effort; sample collec­ tion, sera extraction, and laboratory analyses for specific and unspecific health indicators) per study site. particularly with those relevant in MTC epidemiology (Table S8, Fig. 2, validation is needed. However, the proteins and methods selected S1 and S2), make this species key for the introduction, spread and could vary for APPs in some groups such as ruminants [55]. maintenance of diseases in the network [49,57]. Moreover, the wild This study also allowed quantifying the effort required by each of the boar is a spreader and reservoir of pathogens shared between wildlife components of IWM, namely health and population monitoring, which and livestock [39], presenting high excretion rates and therefore a high must be balanced and integrated. The additional efforts to effectively risk of infection transmission [40]. Thus, the wild boar is a suitable implement IWM depend on the WHM schemes currently applied, which target species for targeted WHM [4] and to design and implement IWM differ significantly among European countries [6]. Due to the traditional at the wildlife-livestock interface in the European context. approach of WHM as a combination of general and targeted surveillance Overall, the unspecific markers of disease responded to population, [6], population monitoring is likely the main required improvement for environmental and disease factors (Table S14) and responded to the all WHM schemes [4]. However, this key information is precisely the same determinants as the specific disease marker assessed (seroposi­ most costly and time-consuming component of IWM (Fig. 5). tivity to MTC) (Tables 2, S15a and S15b). Higher antibody prevalences Although the REM methodology used in this pilot study already were positively associated with biomarkers of inflammation (MTC and optimized the labor cost when characterizing the host community [16], Hp and CRP), stress (PCV2 and Pon 1), activation of the immune system further decreasing the effort required to reach the essential reliable (PCV2 and B. suis and ADA), and redox balance (T. gondii and Cuprac population monitoring would increase the feasibility of IWM and facil­ and Thiols) (Table S14) agreeing with previous reports [11,12,50]. The itate the generalization of its application to wider geographical contexts increase in inflammation biomarkers (Hp and CRP) is often correlated [15]. Apart from reducing the workload of population monitoring, ef­ with the extent of lesions [51], and these analytes have lower sensitivity forts to improve IWM efficiency by developing, validating, and estab­ to viral diseases when compared to bacterial infections [52]. Some vi­ lishing non-specific markers of disease and/or wildlife population health ruses may even attenuate inflammatory responses to ease infection [53], status deserve further investigation. This would be easier and more cost- which could explain the negative relationship of the seroprevalence effective than targeting several pathogens. Also, non-invasive environ­ against HEV and Hp (Table S14). Apart from diseases, the unspecific mental DNA sampling could help to broaden the spectrum of hosts and markers were also affected by environmental variables such as urban pathogens to survey, decreasing sampling effort [56]. Furthermore, cover (Pon 1), cropland (Hp and Pon 1), and the number of different planning a central national wild animal serum repository would be hosts in the network (Cuprac and Thiols) (Table S14). This, together desirable to enable researchers and agencies for traceback of pathogens with the great variability among study sites of these unspecific markers in the future. Finally, the number and diversity of monitoring sites needs (Table S12), shows that considering interactions between environ­ to be broadened to reach a comprehensive and representative coverage mental, population and epidemiological variables and establishing of peninsular Spain while considering the challenging administrative baseline values for each population is required in order to use them as structure of the country and the different agents involved, as previously reliable indicators of health for IWM [10,54]. Although unspecific recommended [6]. markers have the potential to become reliable indicators for IWM, their non-specificity is simultaneously their main advantage and drawback, 5. Conclusion since it can be cheaper and easier to perform than analyzing all the potential pathogens individually but requires defining the conditions The implementation of a nationwide pilot IWM scheme showed the influencing these indicators for each population. In general, methods added value of combining general and targeted health surveillance with employed to measure markers of redox balance and immune response population monitoring and host community characterization. Thereby, potentially work for any mammal species, although species-specific IWM steps up from wildlife disease surveillance linking host abundance 7 P. Barroso et al. One Health 16 (2023) 100479 and community complexity with health surveillance and disease National and Regional Veterinary Services and Wildlife Services. epidemiology. However, reaching a balance among the three pillars of IWM and designing a scheme that is technically sound, economically References sustainable, and suitable for each country-specific situation requires continuous effort. [1] C. Gortázar, J.F. Ruiz-Fons, U. Höfle, Infections shared with wildlife: an updated perspective, Eur. J. Wildl. Res. 62 (2016) 511–525, https://doi.org/10.1007/ Supplementary data to this article can be found online at https://doi. s10344-016-1033-x. org/10.1016/j.onehlt.2022.100479. [2] C. Gortázar, P. Acevedo, F. Ruiz-Fons, J. Vicente, Disease risks and overabundance of game species, Eur. J. Wildl. Res. 52 (2006) 81–87, https://doi.org/10.1007/ s10344-005-0022-2. Funding [3] K.E. Jones, N.G. Patel, M.A. Levy, A. Storeygard, D. Balk, J.L. Gittleman, P. Daszak, Global trends in emerging infectious diseases, Nature. 451 (2008) 990–993, This study benefitted from a contract to IREC-UCLM from Ministry of https://doi.org/10.1038/nature06536. [4] B. Cardoso, I. García-Bocanegra, P. Acevedo, G. Cáceres, P.C. Alves, C. Gortázar, Agriculture, Fisheries and Food, Madrid, Spain, and Research Grant Stepping up from wildlife disease surveillance to integrated wildlife monitoring in PID2020-115046GB-I00 Ecología, salud pública y gestión del jabalí Europe, Res. Vet. Sci. 144 (2021) 149–156, https://doi.org/10.1016/j. urbano, funded by the Spanish Ministerio de Ciencia e Innovación. P.P. rvsc.2021.11.003. [5] C. Machalaba, M. Uhart, M.-P. Ryser-Degiorgis, W.B. Karesh, Gaps in health received support from the MINECO-UCLM through an FPU grant security related to wildlife and environment affecting pandemic prevention and (FPU16/00039) and from UNITO Dipartimento di Scienze Veterinarie preparedness, 2007–2020, Bull. World Health Organ. 99 (2021) 342–350, https:// (8B/2022/VET). B.C. is supported by the Fundação para a Ciência e doi.org/10.2471/BLT.20.272690. Tecnologia, FCT (grant number 2020.04872.BD). M.E. received support [6] B. Lawson, A. Neimanis, A. Lavazza, J.R. López-Olvera, P. Tavernier, C. Billinis, J. P. Duff, D.T. Mladenov, J.M. Rijks, S. Savić, G. Wibbelt, M.-P. Ryser-Degiorgis, from the MINECO through an FPU grant (FPU19/04651). T. Kuiken, How to start up a National Wildlife Health Surveillance Programme, Animals. 11 (2021) 2543, https://doi.org/10.3390/ani11092543. CRediT authorship contribution statement [7] P. Barroso, P. Acevedo, J. Vicente, The importance of long-term studies on wildlife diseases and their interfaces with humans and domestic animals: a review, Transbound. Emerg. Dis. 68 (2020) 1895–1909, https://doi.org/10.1111/ P. Barroso: Data curation, Formal analysis, Investigation, Method­ tbed.13916. ology, Software, Writing – original draft. D. Relimpio: Data curation, [8] J. Vicente, M. Apollonio, J.A. Blanco-Aguiar, T. Borowik, F. Brivio, J. Casaer, S. Croft, G. Ericsson, E. Ferroglio, D. Gavier-Widen, C. Gortázar, P.A. Jansen, Investigation, Methodology, Validation, Visualization, Writing – review O. Keuling, R. Kowalczyk, K. Petrovic, R. Plhal, T. Podgórski, M. Sange, & editing. J.A. Zearra: Data curation, Investigation, Methodology, M. Scandura, K. Schmidt, G.C. Smith, R. Soriguer, H.-H. Thulke, S. Zanet, Visualization, Writing – review & editing. J.J. Cerón: Formal analysis, P. Acevedo, Science-based wildlife disease response, Science. 364 (2019) 943–944, https://doi.org/10.1126/science.aax4310. Investigation, Methodology, Resources, Validation, Writing – review & [9] C. Gortázar, P. Barroso, R. Nova, G. Cáceres, The role of wildlife in the editing. P. Palencia: Conceptualization, Formal analysis, Methodology, epidemiology and control of foot-and-mouth-disease and similar transboundary Software, Visualization, Writing – review & editing. B. Cardoso: Data (FAST) animal diseases: a review, Transbound. Emerg. Dis. 69 (2021) 2462–2473, https://doi.org/10.1111/tbed.14235. curation, Investigation, Methodology, Visualization, Writing – review & [10] F. Ceciliani, J.J. Ceron, P.D. Eckersall, H. Sauerwein, Acute phase proteins in editing. E. Ferreras: Investigation, Methodology, Visualization, Writing ruminants, J. Proteome 75 (2012) 4207–4231, https://doi.org/10.1016/j. – review & editing. M. Escobar: Data curation, Investigation, Method­ jprot.2012.04.004. [11] M.D. Contreras-Aguilar, A. Tvarijonaviciute, I. Monkeviciene, M. Martín-Cuervo, L. ology, Visualization, Writing – review & editing. G. Cáceres: Concep­ G. González-Arostegui, L. Franco-Martínez, J.J. Cerón, F. Tecles, D. Escribano, tualization, Resources, Visualization, Writing – review & editing. J.R. Characterization of total adenosine deaminase activity (ADA) and its isoenzymes in López-Olvera: Conceptualization, Investigation, Methodology, Re­ saliva and serum in health and inflammatory conditions in four different species: sources, Supervision, Validation, Visualization, Writing – review & an analytical and clinical validation pilot study, BMC Vet. Res. 16 (2020) 384, https://doi.org/10.1186/s12917-020-02574-2. editing. C. Gortázar: Conceptualization, Funding acquisition, Method­ [12] J. Vicente, J. Martinez-Guijosa, A. Tvarijonaviciute, I.G. Fernandez-de Mera, ology, Project administration, Resources, Supervision, Validation, C. Gortazar, J.J. Ceron, S. Martinez-Subiela, Serum haptoglobin response in red Visualization, Writing – original draft, Writing – review & editing. deer naturally infected with tuberculosis, Comp. Immunol. Microbiol. Infect. Dis. 64 (2019) 25–30, https://doi.org/10.1016/j.cimid.2019.01.021. [13] MAPAMA, Plan de Actuación sobre Tuberculosis en Especies Silvestres (PATUBES), Declaration of Competing Interest Madrid, Spain. https://www.mapa.gob.es/es/ganaderia/temas/sanidad-animal-h igiene-ganadera/patubes2017_3_tcm30-378321.pdf, 2017. [14] F. Ruiz-Fons, A review of the current status of relevant zoonotic pathogens in wild The authors have no potential conflicts of interest to declare. swine (Sus scrofa) populations: changes modulating the risk of transmission to humans, Transbound. Emerg. Dis. 64 (2017) 68–88, https://doi.org/10.1111/ Data availability tbed.12369. [15] P. Palencia, J.M. Rowcliffe, J. Vicente, P. Acevedo, Assessing the camera trap methodologies used to estimate density of unmarked populations, J. Appl. Ecol. 58 Data are available at the Zenodo repository (doi:https://doi. (2021) 1583–1592, https://doi.org/10.1111/1365-2664.13913. org/10.5281/zenodo.6855764), including the metadata from camera [16] P. Palencia, P. Barroso, J. Vicente, T.R. Hofmeester, J. Ferreres, P. Acevedo, Random encounter model is a reliable method for estimating population density of trapping pictures (i.e. date, time and species), data on interspecific in­ multiple species using camera traps, remote Sens, Ecol. Conserv. 8 (2022) teractions and covariables to REM calculations in Excel files. 670–682, https://doi.org/10.1002/rse2.269. [17] J.M. Rowcliffe, J. Field, S.T. Turvey, C. Carbone, Estimating animal density using Acknowledgement camera traps without the need for individual recognition, J. Appl. Ecol. 45 (2008) 1228–1236, https://doi.org/10.1111/j.1365-2664.2008.01473.x. [18] P. Palencia, J. Fernández-López, J. Vicente, P. Acevedo, Innovations in movement We are grateful to the managers and technicians of the study sites, and behavioural ecology from camera traps: day range as model parameter, along with the hunters and vets who helped us access the hunted wild Methods Ecol. Evol. 12 (2021) 1201–1212, https://doi.org/10.1111/2041- 210X.13609. boar for sampling. The authors are grateful to Ajuntament de Barcelona, [19] P. Palencia, Trappingmotion: Integrate camera-trapping in movement and who supported field sampling through the contracts 13/051, 15/0174, behavioural studies, R package version 0.1.1 (2020). https://github.com/ 16/0243 and 16/0243-00-PR/01. We also thank for their collaboration: PabloPalencia/trappingmotion (accessed January 13, 2022). [20] J.M. Rowcliffe, Camtools. https://github.com/MarcusRowcliffe/camtools, 2022. Y.F., A.G. and C.R. and JCCM staff in the province of Ciudad Real, J.A.M. [21] T.G. O’Brien, Abundance, density and relative abundance: A conceptual in Doñana National Park, ARRECAL Huesca and M-C.A./D⋅F-L. in the framework, in: Camera Traps Anim. Ecol., Springer, Japan, Tokyo, 2011, province of Huesca, to M.M. in the area of Galicia, to F.G., J.M. and the pp. 71–96, https://doi.org/10.1007/978-4-431-99495-4_6. [22] E. Kukielka, J.A. Barasona, C.E. Cowie, J.A. Drewe, C. Gortazar, I. Cotarelo, park rangers of Redes Natural Park in Asturias, to B.C. in the area of J. Vicente, Spatial and temporal interactions between livestock and wildlife in Sierra de Gredos Regional Park, and finally to F.E. and C.M. in Murcia. south Central Spain assessed by camera traps, Prev. Vet. Med. 112 (2013) 213–221, We would also like to thank our colleagues from IREC for their help https://doi.org/10.1016/j.prevetmed.2013.08.008. during fieldwork and data processing, as well as to all collaborating 8 P. Barroso et al. One Health 16 (2023) 100479 [23] J. Niedballa, A. Wilting, R. Sollmann, H. Hofer, A. Courtiol, Assessing analytical [43] J. Vicente, F. Ruiz-Fons, D. Vidal, U. Höfle, P. Acevedo, D. Villanúa, I. methods for detecting spatiotemporal interactions between species from camera G. Fernández-de-Mera, M.P. Martín, C. Gortázar, Serosurvey of Aujeszky’s disease trapping data, remote Sens, Ecol. Conserv. 5 (2019) 272–285, https://doi.org/ virus infection in European wild boar in Spain, Vet. Rec. 156 (2005) 408–412, 10.1002/rse2.107. https://doi.org/10.1136/vr.156.13.408. [24] G. Csardi, T. Nepusz, The igraph software package for complex network research, [44] J. Vicente, J. Segalés, U. Höfle, M. Balasch, J. Plana-Durán, M. Domingo, Inter. J. Compl. Syst. 1695 (2006) 1–9. C. Gortázar, Epidemiological study on porcine circovirus type 2 (PCV2) infection in [25] B. Martínez-López, A.M. Perez, J.M. Sánchez-Vizcaíno, Social network analysis. the European wild boar (Sus scrofa), Vet. Res. 35 (2004) 243–253, https://doi.org/ Review of general concepts and use in preventive veterinary medicine, 10.1051/vetres:2004008. Transbound, Emerg. Dis. 56 (2009) 109–120, https://doi.org/10.1111/j.1865- [45] J. Martínez-Fernández, P. Ruiz-Benito, M.A. Zavala, Recent land cover changes in 1682.2009.01073.x. Spain across biogeographical regions and protection levels: implications for [26] A. Arenas-Montes, I. García-Bocanegra, J. Paniagua, J.J. Franco, F. Miró, conservation policies, Land Use Policy 44 (2015) 62–75, https://doi.org/10.1016/ M. Fernández-Morente, A. Carbonero, A. Arenas, Blood sampling by puncture in j.landusepol.2014.11.021. the cavernous sinus from hunted wild boar, Eur. J. Wildl. Res. 59 (2013) 299–303, [46] X. Fernández-Aguilar, M. Gottschalk, V. Aragon, J. Càmara, C. Ardanuy, https://doi.org/10.1007/s10344-013-0701-3. R. Velarde, N. Galofré-Milà, R. Castillo-Contreras, J.R. López-Olvera, [27] P.C. Jocelyn, Spectrophotometric assay of thiols, Methods Enzymol. 143 (1987) G. Mentaberre, A. Colom-Cadena, S. Lavín, O. Cabezón, Urban wild boars and risk 44–67, https://doi.org/10.1016/0076-6879(87)43013-9. for zoonotic Streptococcus suis, Spain, Emerg. Infect. Dis. 24 (2018) 1083–1086, [28] A. Tvarijonaviciute, F. Tecles, M. Caldin, S. Tasca, J. Cerón, Validation of https://doi.org/10.3201/eid2406.171271. spectrophotometric assays for serum paraoxonase type-1 measurement in dogs, [47] H. Wang, R. Castillo-Contreras, F. Saguti, J.R. López-Olvera, M. Karlsson, Am. J. Vet. Res. 73 (2012) 34–41, https://doi.org/10.2460/ajvr.73.1.34. G. Mentaberre, M. Lindh, J. Serra-Cobo, H. Norder, Genetically similar hepatitis E [29] CNIG, Corine Land Cover (España), https://land.copernicus.eu/pan-european/cor virus strains infect both humans and wild boars in the Barcelona area, Spain, and ine-land-cover, 2018 (accessed October 21, 2021). Sweden, Transbound. Emerg. Dis. 66 (2019) 978–985, https://doi.org/10.1111/ [30] QGIS Development Team, QGIS Geographic Information System. Open Source tbed.13115. Geospatial Foundation Project. https://qgis.org/es/site/, 2019 (accessed October [48] V. Lizana, C. Gortázar, A. Muniesa, Ó. Cabezón, A. Martí-Marco, J. López-Ramon, 1, 2019). J. Cardells, Human and environmental factors driving Toxoplasma gondii [31] O.J. Dunn, Multiple comparisons using rank sums, Technometrics. 6 (1964) prevalence in wild boar (Sus scrofa), Res. Vet. Sci. 141 (2021) 56–62, https://doi. 241–252. org/10.1016/j.rvsc.2021.10.007. [32] A.F. Zuur, E.N. Ieno, N.J. Walker, A.A. Saveliev, G.M. Smith, GLM and GAM for [49] R. Triguero-Ocaña, B. Martínez-López, J. Vicente, J.A. Barasona, J. Martínez- count data, in: Environ. Forum, 2009, pp. 209–243, https://doi.org/10.1007/978- Guijosa, P. Acevedo, Dynamic network of interactions in the wildlife-livestock 0-387-87458-6_9. interface in mediterranean Spain: an epidemiological point of view, Pathogens. 9 [33] K.P. Burnham, D.R. Anderson, Model Selection and Multimodel Inference: A (2020) 1–16, https://doi.org/10.3390/pathogens9020120. Practical Information-Theoretic Approach, Springer, New York, 2002. [50] C.F. Pierce, V.R. Brown, S.C. Olsen, P. Boggiatto, K. Pedersen, R.S. Miller, S. [34] D. Bates, M. Maechler, B. Bolker, S. Walker, R.H.B. Chistensen, H. Singman, B. Dai, E. Speidel, T.J. Smyser, Loci associated with antibody response in feral swine (Sus F. Sheipl, G. Grothendieck, O. Green, J. Fox, Linear Mixed-Effects Models Using scrofa) infected with Brucella suis, Front. Vet. Sci. 7 (2020), 554674, https://doi. “Eigen” and S4, 2019. org/10.3389/fvets.2020.554674. [35] R Core Team, R: A Language and Environment for Statistical Computing. https:// [51] A. Ráez-Bravo, J.E. Granados, J.J. Cerón, F.J. Cano-Manuel, P. Fandos, J.M. Pérez, www.r-project.org/, 2019. J. Espinosa, R.C. Soriguer, J.R. López-Olvera, Acute phase proteins increase with [36] F. Keesing, R.S. Ostfeld, Dilution effects in disease ecology, Ecol. Lett. 24 (2021) sarcoptic mange status and severity in Iberian ibex (Capra pyrenaica, Schinz 1838), 2490–2505, https://doi.org/10.1111/ele.13875. Parasitol. Res. 114 (2015) 4005–4010, https://doi.org/10.1007/s00436-015- [37] J.A. Barasona, C. Gortázar, J. de la Fuente, J. Vicente, Host richness increases 4628-3. tuberculosis disease risk in game-managed areas, Microorganisms. 7 (2019) 182, [52] P.L. McCarthy, A.L. Frank, R.C. Ablow, S.J. Masters, T.F. Dolan, Value of the C- https://doi.org/10.3390/microorganisms7060182. reactive protein test in the differentiation of bacterial and viral pneumonia, [38] D.J. Civitello, J. Cohen, H. Fatima, N.T. Halstead, J. Liriano, T.A. McMahon, C. J. Pediatr. 92 (1978) 454–456, https://doi.org/10.1016/S0022-3476(78)80448-X. N. Ortega, E.L. Sauer, T. Sehgal, S. Young, J.R. Rohr, Biodiversity inhibits parasites: [53] V. Chandra, A. Kar-Roy, S. Kumari, S. Mayor, S. Jameel, The hepatitis E virus ORF3 broad evidence for the dilution effect, Proc. Natl. Acad. Sci. 112 (2015) protein modulates epidermal growth factor receptor trafficking, STAT3 8667–8671, https://doi.org/10.1073/pnas.1506279112. translocation, and the acute-phase response, J. Virol. 82 (2008) 7100–7110, [39] V. Naranjo, C. Gortazar, J. Vicente, J. de la Fuente, Evidence of the role of https://doi.org/10.1128/JVI.00403-08. European wild boar as a reservoir of mycobacterium tuberculosis complex, Vet. [54] A. Parra, A. García, N.F. Inglis, A. Tato, J.M. Alonso, M. Hermoso de Mendoza, Microbiol. 127 (2008) 1–9, https://doi.org/10.1016/j.vetmic.2007.10.002. J. Hermoso de Mendoza, J. Larrasa, An epidemiological evaluation of [40] N. Santos, V. Almeida, C. Gortázar, M. Correia-Neves, Patterns of Mycobacterium Mycobacterium bovis infections in wild game animals of the Spanish Mediterranean tuberculosis-complex excretion and characterization of super-shedders in naturally- ecosystem, Res. Vet. Sci. 80 (2006) 140–146, https://doi.org/10.1016/j. infected wild boar and red deer, Vet. Res. 46 (2015) 129, https://doi.org/10.1186/ rvsc.2005.05.010. s13567-015-0270-4. [55] B. Wells, G.T. Innocent, P.D. Eckersall, E. McCulloch, A.J. Nisbet, S.T. Burgess, Two [41] P. Acevedo, J. Vicente, U. Höfle, J. Cassinello, F. Ruiz-Fons, C. Gortázar, Estimation major ruminant acute phase proteins, haptoglobin and serum amyloid a, as serum of European wild boar relative abundance and aggregation: a novel method in biomarkers during active sheep scab infestation, Vet. Res. 44 (2013) 103, https:// epidemiological risk assessment, Epidemiol. Infect. 135 (2007) 519–527, https:// doi.org/10.1186/1297-9716-44-103. doi.org/10.1017/S0950268806007059. [56] J.A. Barasona, J. Vicente, I. Díez-Delgado, J. Aznar, C. Gortázar, M.J. Torres, [42] M. Boadella, P. Acevedo, J. Vicente, G. Mentaberre, A. Balseiro, M.C. Arnal, Environmental presence of Mycobacterium tuberculosis complex in aggregation D. Martínez, I. García-Bocanegra, C. Casal, J. Álvarez, Á. Oleaga, S. Lavín, points at the wildlife/livestock Interface, Transbound. Emerg. Dis. 64 (2017) M. Muñoz, J.L. Sáez-Llorente, J. de la Fuente, C. Gortázar, Spatio-temporal trends 1148–1158, https://doi.org/10.1111/tbed.12480. of Iberian wild boar contact with Mycobacterium tuberculosis complex detected by [57] B. Martínez-López, A.M. Pérez, J.M. Sánchez-Vizcaíno, Social network analysis. ELISA, Ecohealth. 8 (2011) 478–484, https://doi.org/10.1007/s10393-011-0713- Review of general concepts and use in preventive veterinary medicine, y. Transbound. Emerg. Dis. 56 (2009) 109–120, https://doi.org/10.1111/j.1865- 1682.2009.01073.x. 9