Study Area

South American coatis were captured in two areas within the city of Campo Grande, located in the state of Mato Grosso do Sul, Brazil, in the central-western region of the country: (i) Parque Estadual do Prosa (PEP), a conservation unit in the core of Campo Grande with 135 hectares, and (ii) private area of the Air Force Village (AFV), a complex of 197 hectares divided into a military operational zone and a residential area inhabited by at least 730 humans with their domestic animals (Barreto et al. 2024). The predominant vegetation in Campo Grande is Cerrado, a tropical savanna that is characterized by hot, semi-humid summers and distinct seasonal variations, with a dry period from April to October and a rainy season from November to March (Barreto et al. 2021) (Figure 1).

Trapping Procedures

South American coatis were captured every three weeks for ten consecutive days from March 2018 to April 2019 using 80 metal traps (1 m × 0.40 m × 0.50 m), with 40 traps per area. The traps were strategically placed to ensure human accessibility and shade, covering most of the PEP and AFV areas. Captured animals were anesthetized with a combination of tiletamine hydrochloride and zolazepam hydrochloride (Telazol, Zoetis® New Jersey, USA) at approximately 6 mg/kg, intramuscularly. After chemical restraint, animals were marked with numbered color-coded ear tags, and a microchip was implanted subcutaneously between the shoulder blades for future identification. Each animal was measured, and its age was estimated according to Olifiers et al. (2010). Blood was collected from the femoral vein using tubes containing EDTA (ethylenediamine tetraacetic acid), then transferred to RNAse/DNAse-free cryotubes and stored at -80°C until molecular analysis. A thorough inspection for ectoparasites was conducted by a single person on the entire body of the individual for three minutes. Ticks were removed with forceps, stored in 100% alcohol, and identified to the genus (larvae) or species (nymphs and adults) using a stereomicroscope following taxonomic literature (Dantas-Torres et al., 2019; Martins et al. 2010).  All specimens were deposited in the Acari Collection of Instituto Butantan, São Paulo, Brazil.

All experimental procedures were approved by the Instituto Chico Mendes de Biodiversidade (ICMBio) under SISBIO license 49662-8, the Ethics Committee on Animal Use of the School of Agricultural and Veterinary Sciences, UNESP (CEUA FCAV/UNESP 06731/19), the Ethics Committee on Animal Use of Universidade Católica Dom Bosco (CEUA UCDB 001/2018), and the Air Force Cooperation Agreement (Nº01/GAP-CG/2018).

Parasite Sampling

DNA was extracted from 200 µL of blood using the Illustra Blood Mini Kit (GE Healthcare®, Chicago, IL, USA), following the manufacturer's instructions. To assess the quality of the extracted DNA and avoid false-negative results, the DNA samples were tested using PCR targeting the mammalian gapdh gene (Birkenheuer et al. 2003). Ultra-pure sterile water (Life Technologies®, Carlsbad, CA, USA) was used as a negative control in the PCR assays. PCR screening for Anaplasma, Ehrlichia, Hepatozoon, hemotropic Mycoplasma, and Neorickettsia was conducted by targeting specific genes for each genus, as described by Perles (2022a; b; 2023a; b).

Network structure

To examine the relationships among the detected hemoparasites, we used a modularity network metrics (M) (Sano et al. 2024). To assess the significance of network metrics, Monte Carlo procedures were employed, comparing the observed metrics with 1,000 random matrices generated from null models. We generated them based on the original weighted matrix using the algorithm proposed by Pinheiro et al. (2019). Significance was determined when the network structure deviated significantly from the null model at p < 0.05. Subsequently, to evaluate the relative importance of each node in the network structure, we computed the individual role for each South American coati and each parasite species. This involved determining their network functional role, classifying nodes based on their position and significance within the network. The classifications included (i) ultraperipheral vertices (all interactions within their module), (ii) peripheral vertices (most interactions within their module), (iii) non-hub connector vertices (many interactions to other modules), (iv) non-hub kinless vertices (interactions evenly distributed among all modules), (v) provincial hubs (most interactions within their module), (vi) connector hubs (many interactions to most other modules), and (vii) kinless hubs (interactions homogeneously distributed among all modules) (Mello et al. 2013; Queiroz et al. 2021). For the network construction we used the 'igraph' package (Csardi and Nepusz 2006), and the host-parasite interaction incidence matrix was handled with the 'Bipartite' package (Dormann et al. 2008; Dormann 2011).

Relationship between biotic and abiotic variables in infections

A variety of specific statistical models were assessed to investigate the influence of covariates (explanatory variables) on each parasite species' role within South American coatis. The examined variables included: (i) weight; (ii) age; (iii) sex; (iv) quantities of immature ticks collected from each individual (Perles et al. 2022c); and (v) collection area. Additionally, a full model (all variables) and null model, devoid of any variables, were established. The species role of each coati was treated as a categorical variable, representing distinct categories. To appropriately model this response variable, the analysis was conducted utilizing generalized linear models (GLM) with the multinomial family, which allowed us to account for the multiple, unordered categories effectively. All candidate models were compared in a model selection approach based on Akaike information criterion corrected for small samples (AICc) (Akaike 1974), considering as plausible all models with ΔAICc ≤ 2, using the package ‘AICcmodavg’ version 2.3–1 (Mazerolle 2020). All data were analyzed using R 4.2.1.