Vector-borne zoonotic bacteria in Tunisia: from fragmented data to an integrated One Health surveillance framework

1 Introduction

Vector-borne and vector-associated bacterial zoonoses represent an important but still insufficiently integrated component of infectious disease surveillance in Tunisia. The country occupies a strategic biogeographical position at the interface between the Mediterranean basin, North Africa and the Sahara, with diverse ecological zones ranging from humid northern forests to semi-arid livestock areas and Saharan camel-rearing systems. This environmental heterogeneity supports a wide diversity of arthropod vectors, including ticks and fleas, as well as multiple domestic and wildlife hosts that may contribute to bacterial circulation (Sarih et al., 2005; Bouattour et al., 2010; Balti et al., 2021; Baede et al., 2024; Selmi et al., 2024a,b).

During the last three decades, Tunisian studies have documented several bacterial groups of zoonotic or suspected zoonotic relevance, including Rickettsia, Bartonella, Anaplasma, Ehrlichia, Borrelia, and Coxiella (Letaief et al., 1995; Ghorbel et al., 1994; Znazen et al., 2005; Ben Arab et al., 2008; M’ghirbi et al., 2012; Ben Said et al., 2016; Barkallah et al., 2018; Gharbi et al., 2024). More recently, molecular and microbiome-based studies have expanded this landscape by revealing additional vector-associated bacteria, including tick endosymbionts, opportunistic bacterial genera and emerging or poorly characterized taxa such as Francisella-like endosymbionts, Midichloria mitochondrii and Rickettsiella spp. (Benyedem et al., 2022; Baede et al., 2024; Kratou et al., 2025a,b). These advances have substantially improved knowledge of bacterial diversity in Tunisian vectors and hosts.

However, the available evidence remains fragmented. Human data are often limited to specific clinical syndromes or hospital-based studies, such as Mediterranean spotted fever, blood culture-negative endocarditis, Q fever or suspected Lyme borreliosis (Znazen et al., 2005; Ben Arab et al., 2008; Hammami et al., 2013; Mahfoudhi et al., 2015; Lamloumi et al., 2024). Veterinary and entomological studies, in contrast, have generated extensive data from dogs, cats, ruminants, horses, dromedaries, wildlife, ticks and fleas, but these findings are rarely connected to simultaneous human surveillance (Belkhiria et al., 2017; Zouari et al., 2017; Selmi et al., 2018; Selmi et al., 2020a,b,c; Selmi et al., 2021a,b; ElHamdi et al., 2022). As a result, Tunisia has accumulated valuable but compartmentalized evidence, with limited integration between human medicine, veterinary medicine, entomology, wildlife ecology, environmental exposure, and public health decision-making.

This fragmentation is particularly important because the epidemiological meaning of bacterial detection differs widely between compartments. A bacterium detected in a human clinical sample has different public health significance from the same bacterium detected in a tick, a wildlife host, a microbiome dataset, or seropositive animal. Molecular detection in vectors does not necessarily prove vector competence. Animal positivity does not automatically demonstrate reservoir competence, and microbiome signals should not be interpreted as evidence of zoonotic threats. However, these findings should be interpreted according to several clinical and epidemiological supports. This distinction is essential for agents such as Anaplasma spp., which are strongly documented in animals and ticks but not yet confirmed in humans in Tunisia. Besides, tick-associated endosymbionts, such as Francisella-like organisms and Midichloria, showed a primarily ecological rather than clinical relevance (Belkahia et al., 2015; Ben Said et al., 2017a,b; Benyedem et al., 2022; M’ghirbi et al., 2022; Kratou et al., 2025b).

A One Health approach is highly relevant for assessing vector-borne bacterial zoonoses in Tunisia where multiple interfaces coexist, including dog-tick-human, cat/dog-flea-human, livestock-tick-farmer, camel-Hyalomma-human, Ixodes-human-wildlife, and tick-microbiome systems.

The present review aims to provide a critical synthesis of vector-borne and vector-associated zoonotic bacteria reported in Tunisia, with emphasis on current evidence, interpretation limits and One Health perspectives. Rather than simply listing reported pathogens, the review evaluates what Tunisian data actually demonstrate, what remains uncertain, and how existing findings could be translated into integrated surveillance and public health actions.

2 Methodological approach and one health analytical framework

This review was designed as a structured critical narrative synthesis of vector-borne and vector-associated bacterial zoonoses reported in Tunisia. Because the available studies differed substantially in bacterial targets, host species, arthropod vectors, ecological settings, sample types, diagnostic methods and study designs, no meta-analysis or pooled prevalence estimate was attempted. Instead, the review used an evidence-mapping approach adapted to a One Health risk-interpretation framework, with emphasis on transparency, diagnostic certainty, compartmental convergence and cautious interpretation of heterogeneous evidence (Page et al., 2021; Tricco et al., 2018; FAO, WHO, and WOAH, 2019, 2020).

Literature searches were conducted in PubMed/MEDLINE, Scopus, Web of Science Core Collection, and Google Scholar, and were complemented by searches in major publisher platforms, including ScienceDirect, SpringerLink, Wiley Online Library, Taylor & Francis Online, MDPI and Frontiers. Additional records were identified through backward and forward citation checking, local Tunisian or regional journals, conference abstracts, theses and author-held bibliographic files when they contained usable primary or contextual evidence. The overall study selection process is summarized in Figure 1. The final search was completed on 1 June 2026. No lower date limit was imposed, and all retrievable records published or available up to that date were considered. English and French records were included; records in other languages were considered only when sufficient bibliographic and methodological information could be reliably extracted. The complete search structure, sources and eligibility criteria are provided in Supplementary Table 1.

Search strings combined country terms with bacterial, disease, vector and host terms. The country terms were Tunisia OR Tunisian. The main genus-level and disease terms were Rickettsia, rickettsiosis, Mediterranean spotted fever, Bartonella, cat-scratch disease, endocarditis, Anaplasma, anaplasmosis, Ehrlichia, ehrlichiosis, Borrelia, Lyme, relapsing fever, Coxiella, Q fever, Francisella, Midichloria, Rickettsiella, Mycoplasma, tick microbiome and endosymbiont. These terms were combined with vector and host terms including tick, ticks, flea, fleas, ectoparasite, arthropod, Ixodes, Rhipicephalus, Hyalomma, Ornithodoros, Ctenocephalides, camel, dromedary, dog, cat, livestock, small ruminant, cattle, horse, wildlife, rodent, hedgehog, and fox. Equivalent French terms were used when relevant.

Records were eligible when they reported original or interpretable evidence on zoonotic, suspected zoonotic, veterinary, vector-associated, wildlife-associated, environmental or microbiome-related bacteria in Tunisia. Eligible evidence included human clinical studies, case reports, animal surveys, vector or ectoparasite screening studies, wildlife studies, environmental or food-chain investigations, microbiome studies, conference abstracts and theses when sufficient methodological details were available. Review articles were not used as primary evidence for evidence grading, but were retained when they provided historical, contextual or methodological information. Records were excluded from the Tunisian evidence synthesis when they were not conducted in Tunisia, addressed viral, parasitic, fungal or non-target bacterial diseases only, provided no usable diagnostic or methodological information, consisted only of inaccessible webpage records, or duplicated a study already extracted. Duplicate records, accepted manuscripts and repeated uploads were merged, and the most complete version was retained.

The evidence was organized into seven bacterial categories: Rickettsia spp., Bartonella spp., Anaplasma spp., Ehrlichia spp., Borrelia spp., Coxiella burnetii/Coxiella-like organisms and other vector-associated bacteria. Eligible records were screened, assessed for relevance, and categorized according to bacterial group and One Health compartment. The overall study selection and screening process are summarized in Figure 1, while the detailed distribution of included records among bacterial categories is provided in Supplementary Table 2. Because some studies reported evidence for more than one bacterial genus, genus-level categories represent overlapping contributions and should not be interpreted as independent study totals.

Data extracted from eligible studies included the bacterial taxon or genotype, study type, host or patient group, vector species when applicable, sample type, geographic location, ecological setting, diagnostic approach, molecular targets or serological methods, sequencing or phylogenetic analyses when available, detection rates or seroprevalence estimates, and relevance for One Health surveillance. Findings were interpreted according to the compartment in which they were generated: humans, dogs and cats, livestock, dromedary camels, wildlife, vectors or ectoparasites, environmental or food-chain sources, and microbiome/endosymbiont datasets.

Evidence strength was assessed using a structured qualitative decision framework rather than study counts alone. The classification considered study independence, reproducibility, diagnostic robustness, taxonomic resolution, clinical or epidemiological relevance, One Health compartment coverage, ecological plausibility, geographical representativeness, sampling quality and the quality of negative evidence. Strong evidence required convergent support from robust diagnostic confirmation and/or repeated, clinically relevant or cross-compartmental findings. Moderate evidence indicated credible but incomplete evidence, usually with reliable detection but limited clinical, geographical or compartmental support. Limited evidence indicated isolated, indirect, weakly resolved or microbiome-only signals. Evidence was considered absent or not documented when no relevant Tunisian evidence was identified for a given compartment, or when only well-defined negative studies were available. The decision framework is provided in Supplementary Table 3.

Diagnostic evidence was weighted according to biological and epidemiological certainty. Culture, validated PCR/qPCR, sequencing, multilocus typing and phylogenetic characterization were weighted more strongly than serology, especially when they enabled species- or genotype-level identification. Serological evidence was interpreted cautiously because it may indicate previous exposure rather than active infection, shedding or species-level attribution, and because cross-reactivity may occur in several bacterial groups. In humans, serology was considered more informative when associated with compatible clinical syndromes, confirmatory testing, paired serology, or repeated hospital-based observations. In animals, seropositivity was interpreted mainly as exposure or sentinel evidence rather than proof of active infection, reservoir competence or zoonotic transmission. Microbiome detections and 16S rRNA-based signals were interpreted as ecological or endosymbiotic evidence unless supported by species-level confirmation and independent clinical or epidemiological data. The diagnostic weighting and interpretive rules are summarized in Supplementary Table 4.

3 Current evidence by bacterial group

Available Tunisian studies reveal a heterogeneous evidence base across bacterial groups, with marked differences according to human, animal, wildlife and vector-associated available data. These differences are summarized visually in Figures 2, 3, which map the relative strength of evidence across One Health compartments for the main bacterial groups considered in this review. Taxon-level evidence for Rickettsia, Bartonella, Anaplasma and Ehrlichia spp. is summarized in Table 1, while corresponding critical gaps and surveillance priorities are presented in Table 2. Evidence for Borrelia spp., Coxiella spp. and other vector-associated bacteria is summarized in Table 3, with related gaps and priorities presented in Table 4. For each genus or bacterial category, the evidence is therefore interpreted according to the type of host investigated, the arthropod vectors involved, the diagnostic methods used, and the strength of the One Health link. Particular caution is needed when interpreting molecular detection in vectors, as the presence of bacterial DNA does not necessarily imply active infection, vector competence or confirmed zoonotic transmission.

TABLE 1. One Health evidence matrix for Rickettsia, Bartonella, Anaplasma and Ehrlichia spp. reported in Tunisia.

TABLE 2. Critical evidence gaps, surveillance implications and research priorities for Rickettsia, Bartonella, Anaplasma and Ehrlichia spp. in Tunisia.

TABLE 3. One Health evidence matrix for Borrelia spp., Coxiella spp. and other vector-associated bacteria reported in Tunisia.

TABLE 4. Critical evidence gaps, surveillance implications and research priorities for Borrelia spp., Coxiella burnetii and other vector-associated bacteria in Tunisia.

3.1 Rickettsia spp.

Among vector-borne bacterial pathogens reported in Tunisia, Rickettsia spp. represent one of the most structured and clinically relevant evidence bases. Historically, Tunisian rickettsiology was dominated by Mediterranean spotted fever (MSF), a disease primarily associated with Rickettsia conorii. The classical transmission cycle involves the brown dog tick Rhipicephalus sanguineus sensu lato, which has historically been considered the main vector in the Mediterranean region (Letaief et al., 1995; Znazen et al., 2013; Lamloumi et al., 2024). However, the progressive use of serology, molecular diagnosis, sequencing and multispacer typing has expanded this classical view toward a broader spectrum of Rickettsia species detected in humans, domestic animals, livestock, wildlife, and arthropods (Znazen et al., 2006; Znazen et al., 2013; Khrouf et al., 2014; Khrouf et al., 2016; Belkahia et al., 2021; Balti et al., 2021; Kratou et al., 2023, 2025a).

The most established human rickettsial disease in Tunisia remains MSF. Recent hospital-based data confirm that suspected MSF continues to be diagnosed in Tunisian patients, particularly in clinical cases combining fever, rash and inoculation eschar (Lamloumi et al., 2024). R. conorii therefore remains the species of greatest clinical and public health importance. Severe presentations should also be considered, as illustrated by the report of Guillain-Barré syndrome associated with R. conorii infection, indicating that rickettsioses should not always be regarded as benign febrile illnesses in endemic areas (Taher et al., 2025). At the same time, Tunisian data indicate that human rickettsioses are not limited to R. conorii. Serological and molecular investigations have reported evidence compatible with infections caused by or related to R. massiliae, R. helvetica, R. monacensis, R. aeschlimannii, R. felis, and R. typhi (Znazen et al., 2006, 2013; Sfar et al., 2008; Khrouf et al., 2016). However, these species should not be considered epidemiologically equivalent, as they belong to distinct ecological systems (Znazen et al., 2006, 2013; Khrouf et al., 2016; Belkahia et al., 2021). While R. conorii is primarily associated with Mediterranean spotted fever and the dog-associated tick Rh. sanguineus s.l. (Letaief et al., 1995; Znazen et al., 2013; Lamloumi et al., 2024), R. massiliae is mainly linked to Rhipicephalus ticks (Khrouf et al., 2014; Khrouf et al., 2016; Belkahia et al., 2021), R. aeschlimannii to Hyalomma ticks, particularly in camel and livestock-associated environments (Demoncheaux et al., 2012; Selmi et al., 2020b; Kratou et al., 2023), and R. helvetica and R. monacensis to Ixodes ticks (Sfar et al., 2008; Belkahia et al., 2021; Kratou et al., 2025a). In addition, R. felis is mainly associated with flea-borne cycles involving Ctenocephalides felis (Znazen et al., 2006; Khrouf et al., 2014; Zouari et al., 2017), whereas R. typhi belongs to flea-borne murine typhus cycles (Letaief et al., 1995). Therefore, the detection of different rickettsial species in Tunisia reflects multiple ecological niches rather than a single homogeneous transmission cycle. Historical seroepidemiological data also showed exposure to R. conorii and R. typhi among blood donors in central Tunisia, supporting the long-standing endemicity of both MSF and murine typhus in the country (Letaief et al., 1995). Nevertheless, the true clinical burden of non-conorii rickettsioses remains insufficiently defined, mainly because routine diagnosis often remains oriented toward classical MSF and because species-level confirmation is not systematically performed.

Ticks are the best-documented arthropods associated with Rickettsia spp. in Tunisia. Rh. sanguineus s.l. remains the key tick species in the domestic and peri-domestic cycle of R. conorii, particularly through its close association with dogs and human dwellings (Znazen et al., 2013; Gharbi et al., 2024). This tick complex has also been associated with R. massiliae and other spotted fever group rickettsiae, supporting its importance beyond the classical MSF framework (Khrouf et al., 2014; Khrouf et al., 2016). Other Rhipicephalus species, including Rh. turanicus, have been found carrying rickettsial DNA in ticks collected from livestock and forest-associated environments, suggesting that rural and pastoral systems may contribute to the circulation of several spotted fever group rickettsiae (Belkahia et al., 2021; Kratou et al., 2025a).

The genus Hyalomma is particularly relevant in arid and livestock-associated settings. Among Tunisian Hyalomma species, Hy. marginatum is of particular epidemiological importance because it is recognized as a main vector of Crimean-Congo hemorrhagic fever virus (CCHFV) in the Mediterranean region (Wasfi et al., 2016). Adults of H. marginatum are relatively large ticks and exhibit an active host-seeking strategy, using environmental cues to locate mammalian hosts (Estrada-Peña et al., 2017; Apanaskevich and Horak, 2008). R. aeschlimannii was detected in Hy. dromedarii collected from dromedaries in southern Tunisia, while broader screening of cattle- and camel-associated Hyalomma ticks revealed additional spotted fever group rickettsiae, including R. massiliae and R. sibirica subsp. mongolitimonae (Demoncheaux et al., 2012; Selmi et al., 2020b; Kratou et al., 2023). These findings are important for pastoral, livestock-rearing areas, but they should be interpreted cautiously. The detection of bacterial DNA in livestock may indicate exposure to infected arthropod vectors, but does not by itself demonstrate that livestock serve as competent reservoirs for human infection (Demoncheaux et al., 2012; Selmi et al., 2020b; Kratou et al., 2023).

In northern humid and forested regions, Ixodes ricinus and Ixodes inopinatus contribute to a distinct ecological component of the Tunisian rickettsial landscape. The detection of R. monacensis and R. helvetica in Ixodes ticks highlights forest-associated transmission systems that differ from the classical dog-Rhipicephalus-Mediterranean spotted fever model (Sfar et al., 2008; Kratou et al., 2025a). Although these findings expand the diversity of known Rickettsia species in Tunisia, their local clinical relevance remains insufficiently characterized. Domestic arthropods also contribute to the diversity of rickettsial circulation. C. felis has been associated with R. felis and other rickettsial DNA detections, supporting consideration of flea-associated cycles in domestic environments involving cats, dogs and humans (Znazen et al., 2006, 2013; Khrouf et al., 2014). More recently, R. felis DNA was detected in Cimex lectularius, suggesting a possible association with domestic arthropods; however, this observation requires cautious interpretation because molecular detection alone does not demonstrate biological vector competence or human transmission (Selmi et al., 2024b).

Domestic animals, particularly dogs, represent valuable sentinel compartments for assessing exposure to rickettsial agents. The close relationship between dogs and humans, together with their association with Rh. sanguineus s.l., supports their relevance for monitoring domestic and peri-domestic transmission contexts. High canine seroprevalence against R. conorii in northern Tunisia supports their use as indicators of exposure risk (Gharbi et al., 2024). However, seropositivity reflects previous exposure rather than necessarily active infection, and molecular confirmation in animal blood remains limited. Similarly, livestock-associated tick studies have reported R. massiliae, R. aeschlimannii, R. sibirica subsp. mongolitimonae and R. monacensis in ticks collected from cattle, sheep and goats (Belkahia et al., 2021; Kratou et al., 2023).

Wildlife data have expanded the ecological interpretation of rickettsial circulation in Tunisia. Wild rodents were found to harbor DNA of several vector-borne bacteria, including rickettsial taxa, suggesting that rodent-associated systems may participate in sylvatic or peri-domestic cycles (Selmi et al., 2021a). The North African hedgehog, Atelerix algirus, has also been investigated as a host for infected ticks and fleas, with detection of rickettsial taxa among other vector-borne agents (Balti et al., 2021). More recently, R. monacensis was reported in a red fox, extending the wildlife signal to wild carnivores and reinforcing the need to consider forest and peri-domestic interfaces (Selmi et al., 2025). These observations are valuable, but the role of rodents, hedgehogs and red foxes as confirmed reservoirs remains insufficiently demonstrated because several studies rely on limited sample sizes, opportunistic sampling or restricted geographical coverage (Balti et al., 2021; Selmi et al., 2021a; Selmi et al., 2025).

The Tunisian evidence on Rickettsia spp. also reflects clear diagnostic progress. Earlier studies relied mainly on serology, including immunofluorescence assay and Western blot, while later work incorporated qPCR, reverse line blot, sequencing and multispacer typing (Znazen et al., 2006, 2013; Khrouf et al., 2016). These molecular tools have enabled more precise identification of species and genotypes, moving Tunisian rickettsiology beyond a single-pathogen MSF framework. The use of skin swabs has also been proposed as a useful non-invasive tool for early molecular diagnosis of rickettsial infections, especially when eschar or skin lesions are available (Khrouf et al., 2016). However, diagnostic heterogeneity remains a limitation. Differences between serological, molecular and typing methods complicate comparisons across studies and highlight the need for standardized laboratory protocols and reporting criteria.

Despite being one of the best-documented bacterial groups in Tunisia, Rickettsia spp. still present important knowledge gaps that limit the interpretation of their epidemiological significance. A major limitation is that much of the available vector evidence relies on molecular detection of Rickettsia DNA in ticks, fleas or other arthropods. Although these findings are essential for identifying potential transmission interfaces, they do not by themselves demonstrate biological vector competence, efficient transmission or direct risk for human infection (Belkahia et al., 2021; Kratou et al., 2023; Selmi et al., 2024b). In addition, available data remain geographically heterogeneous, with studies concentrated mainly in northern forested areas, selected cattle farms, small-ruminant systems, southern camel-associated settings and a limited number of clinical centers, while several governorates and ecological compartments remain insufficiently investigated (Demoncheaux et al., 2012; Belkahia et al., 2021; Selmi et al., 2020b; Kratou et al., 2025a). Wildlife-associated evidence is particularly limited and remains exploratory; systematic investigations of rodents, hedgehogs, wild carnivores and their associated ectoparasites are required before their potential epidemiological roles can be accurately defined (Balti et al., 2021; Selmi et al., 2021a; Selmi et al., 2025). Therefore, although the evidence base for Rickettsia spp. is relatively strong compared with several other bacterial groups, particularly for R. conorii, it remains uneven across species, regions and One Health compartments. Future surveillance should prioritize the diagnostic approaches, sentinel monitoring, livestock- and associated tick surveillance, vector competence studies and wildlife investigations within a coordinated One Health framework. The taxon-level evidence for Rickettsia spp. is summarized in Table 1, while the major knowledge gaps and surveillance priorities are presented in Table 2.

3.2 Bartonella spp.

In Tunisia, the Bartonella evidence base links severe human disease, particularly blood culture-negative endocarditis, with domestic and peri-domestic ectoparasites, stray dogs, cats, dromedary camels and synanthropic rodents (Znazen et al., 2005; Hammami et al., 2013; Belkhiria et al., 2017; Zouari et al., 2017; Selmi et al., 2020a; Selmi et al., 2021a). The available data suggest that bartonellosis should not be viewed as a sporadic clinical curiosity, but as a neglected zoonotic bacterial disease requiring coordinated medical, veterinary and entomological attention.

Hospital-based cohorts from Sfax identified Bartonella spp. as etiological agents in approximately 8.9–9.8% of infective endocarditis cases, with B. quintana being the dominant species and B. henselae reported less frequently (Znazen et al., 2005; Hammami et al., 2013). B. henselae is also the main causative agent of cat-scratch disease, typically characterized by a local inoculation skin lesion followed by regional lymphadenopathy (Lin and Saccoccio, 2023), whereas B. quintana, mainly transmitted by body lice (Pediculus humanus corporis), is an important cause of blood culture-negative endocarditis and trench fever (Puspitasari et al., 2025). These observations highlight the clinical relevance of Bartonella infections in Tunisia, particularly in relation to endocarditis. Tunisian patients with Bartonella endocarditis showed severe valvular damage and substantial mortality, emphasizing the importance of early diagnosis and appropriate management (Hammami et al., 2013; Letaief et al., 2007; Znazen et al., 2005). In addition, a case of cat-scratch disease associated with Guillain-Barré syndrome suggests that B. henselae may be involved in atypical neurological presentations, although isolated case reports should not be interpreted as indicators of population-level frequency (El Ouni et al., 2015).

Beyond human disease, molecular screening of fleas collected from domestic animals revealed zoonotic Bartonella species, including B. henselae, B. clarridgeiae and B. elizabethae, in C. felis, Ctenocephalides canis and P. irritans (Zouari et al., 2017). These findings support the importance of domestic and peri-domestic flea-associated systems involving cats, dogs, livestock and humans in Tunisian Bartonella ecology. However, the available evidence should be interpreted cautiously, as detection of Bartonella DNA in fleas does not demonstrate local biological vector competence or confirm transmission to humans or animals (Zouari et al., 2017). Similarly, although human evidence is relatively strong for Bartonella-associated endocarditis, it remains geographically and clinically restricted, with most available data originating from retrospective hospital-based studies in the Sfax region (Znazen et al., 2005; Hammami et al., 2013). Therefore, further integrated investigations are required to better define the epidemiological significance of Bartonella circulation in Tunisia.

Available evidence in Tunisia suggests that the role of ticks in Bartonella circulation remains unclear. Molecular screening of Hyalomma ticks collected from dromedary camels and rodent-associated ectoparasites did not detect Bartonella DNA in the investigated samples (Selmi et al., 2020a; Selmi et al., 2021a). However, these negative findings should be interpreted within the specific ecological and methodological context of each study and do not exclude the possibility of tick-associated Bartonella circulation under other conditions.

Stray dogs showed substantial serological evidence of exposure to several Bartonella species. Molecular detection of Candidatus Bartonella merieuxii in Tunisian dogs, a species initially described in Iraq, suggests that dogs may represent potential sentinels of domestic and peri-domestic Bartonella diversity (Belkhiria et al., 2017). However, their role as true reservoirs remains unconfirmed, as available cross-sectional PCR-based studies do not demonstrate long-term bacterial maintenance, transmission competence, or zoonotic spillover. Cats are classically recognized as important reservoirs for B. henselae and B. clarridgeiae, but the Tunisian corpus is more strongly supported by flea-based data than by large-scale molecular surveys of cats themselves (Zouari et al., 2017). Cats should therefore remain part of the One Health interpretation, particularly because cat-associated fleas may carry zoonotic Bartonella DNA, but more direct studies on owned and stray cats are still needed.

Furthermore, fleas collected from sheep and goats were found carrying zoonotic Bartonella DNA, suggesting that pastoral systems may contribute to exposure through flea-infested animals (Zouari et al., 2017). However, direct molecular surveys of livestock blood or tissues remain limited. Livestock should therefore be discussed mainly as hosts for infected ectoparasites rather than as confirmed mammalian reservoirs. This distinction is important because animal-associated ectoparasite positivity may identify a risk interface without proving that the animal host maintains the bacterium.

A particularly original contribution of Tunisian research is the molecular detection of B. henselae DNA in dromedary camels (Selmi et al., 2020a). This finding expands the currently recognized host range of Bartonella in Tunisia beyond the classical cat-flea-human interface and suggests that camel-associated ecosystems may represent previously underexplored ecological compartments. From a One Health perspective, the relevance of this finding lies in the position of dromedaries at the interface between animal production systems, arthropod communities and human activities in southern Tunisia, including camel husbandry, transport, slaughtering practices and tourism-related interactions. However, the epidemiological significance of camel-associated Bartonella remains unresolved. The detection of B. henselae DNA in dromedaries does not demonstrate that camels act as competent maintenance reservoirs or that they contribute directly to human infection. Moreover, Bartonella DNA was not detected in the examined Hyalomma ticks, leaving the transmission pathway uncertain and suggesting that other arthropod groups, particularly fleas that were not investigated in this study, may contribute to bacterial maintenance and transmission dynamics (Selmi et al., 2020a).

Wildlife evidence for Bartonella in Tunisia is more limited than the evidence from humans, fleas and dogs, but it remains relevant for understanding peri-domestic transmission interfaces. B. henselae DNA was detected in Rattus rattus, suggesting that synanthropic rodents may participate in local Bartonella ecology (Selmi et al., 2021a). This observation is important because R. rattus often lives near human dwellings, livestock facilities and food-storage environments, potentially linking wildlife, domestic animals and humans. Nevertheless, the evidence should not be generalized to other rodents’ species as the reported prevalence was low, and other sampled rodent species, including Meriones shawi and Mus musculus, were tested negative in the available study (Selmi et al., 2021a). In addition, associated rodent ectoparasites tested negative for Bartonella DNA, which means that the main transmission route in rodent-associated cycles remains unclear. Rodents should therefore be framed as a potential but incompletely characterized wildlife compartment rather than as confirmed reservoirs.

In Tunisian studies, Bartonella was detected by different diagnostic tools, including serology, PCR, sequencing and phylogenetic analysis (Znazen et al., 2005; Hammami et al., 2013; Belkhiria et al., 2017; Zouari et al., 2017; Selmi et al., 2020a; Selmi et al., 2021a). Human endocarditis investigations relied on serological and molecular tools, whereas animal and ectoparasite studies used PCR targeting genes such as ftsZ, rpoB, gltA, ITS, rnpB and fur, followed by sequencing step. This molecular evidence is one of the strengths of the Tunisian literature, particularly for distinguishing classified and non classified zoonotic species (Belkhiria et al., 2017; Zouari et al., 2017; Selmi et al., 2020a). However, PCR sensitivity may vary according to sample type, and some weak positive flea pools could not be fully characterized. Besides, serological tests may be affected by cross-reactivity, thereby limiting the accuracy of pathogen identification (Znazen et al., 2005; Belkhiria et al., 2017; Zouari et al., 2017). Standardized multi-target diagnostic workflows would improve comparability between human, animal and ectoparasite studies.

The Tunisian Bartonella evidence base is scientifically valuable but still incomplete. The most important gap is the lack of integration between human clinical data, animal studies, ectoparasite screening and environmental information. Human data are mostly retrospective and concentrated in Sfax, whereas animal and ectoparasite studies are cross-sectional and often based on convenience sampling (Znazen et al., 2005; Hammami et al., 2013; Belkhiria et al., 2017; Zouari et al., 2017). As a result, no national estimate of the burden of bartonellosis is currently available in Tunisia, and both its true prevalence and regional spatial distribution remain insufficiently characterized.

A second major gap concerns transmission pathways. B. quintana dominates Tunisian endocarditis cohorts, but the classical body-louse route was not directly demonstrated in the studied patients, many of whom had rural origin or animal contact (Znazen et al., 2005; Hammami et al., 2013). Similarly, fleas collected from domestic animals carried Bartonella DNA, but experimental proof of local flea competence is lacking (Zouari et al., 2017). Dromedary and rodent findings also suggest additional ecological compartments, but their reservoir status and spillover potential remain unresolved (Selmi et al., 2020a; Selmi et al., 2021a).

The Tunisian Bartonella profile differs clearly from the tick-centered Rickettsia model, as it is shaped mainly by severe human endocarditis, domestic and peri-domestic ectoparasites, stray animals and rural exposure settings. The evidence is strongest for human endocarditis, stray dogs and flea-associated molecular detections, whereas dromedaries and rodents represent emerging but still insufficiently characterized compartments. Future work should prioritize systematic Bartonella testing in blood culture-negative endocarditis, prospective multicenter clinical studies, expanded surveys in owned pets, stray animals and domestic animal fleas, targeted investigations in camel-exposed populations, and broader screening of synanthropic rodents and their ectoparasites. The taxon-level evidence for Bartonella spp. is summarized in Table 1, while the main critical gaps and surveillance priorities for this genus are presented in Table 2.

3.3 Anaplasmataceae

The Tunisian evidence on the Anaplasmataceae family is dominated by animal and tick-associated data, with a marked contrast between the relatively rich evidence for Anaplasma spp., limited evidence for Ehrlichia spp., and absence for other genera. Several Anaplasma taxa with confirmed, suspected or emerging zoonotic relevance have been reported in domestic animals, wildlife or ticks, including Anaplasma phagocytophilum, Anaplasma ovis, A. phagocytophilum-like genotypes, Anaplasma platys, A. platys-like strains and camel-associated strains related to Candidatus Anaplasma camelii (Sarih et al., 2005; M’ghirbi et al., 2012; Belkahia et al., 2014; Belkahia et al., 2015; Ben Said et al., 2015; Ben Said et al., 2017a,b; Belkahia et al., 2019; Selmi et al., 2019a; Selmi et al., 2020c; Ben Said et al., 2020; Selmi et al., 2021a; M’ghirbi et al., 2022; Rjeibi et al., 2022). In contrast, evidence for Ehrlichia circulation in Tunisia is restricted to historical canine serological studies associated with E. canis, sporadic detections of Ehrlichia-like DNA in ticks, and recent molecular findings in wildlife. Human data remain scarce and are limited to a few serological observations, while no molecularly confirmed human cases have been reported (Ghorbel et al., 1991; Ghorbel et al., 1994; Sarih et al., 2005; Balti et al., 2021). Consequently, for both genera, the lack of molecular confirmation of human infection represents a major One Health knowledge gap.

A. phagocytophilum is globally recognized as the agent of the human granulocytic anaplasmosis. In Tunisia, however, the evidence remains animal- and tick-centered. Serological and/or molecular evidence has been reported in horses, cattle, small ruminants, dromedaries and ticks, especially in humid northern and livestock-associated settings (Sarih et al., 2005; M’ghirbi et al., 2012; Ben Said et al., 2014; M’ghirbi et al., 2016, 2022; Rjeibi et al., 2022). In horses, high serological exposure and PCR-positive animals were reported in northern humid areas, while cattle and small ruminants were also shown to be exposed or infected, although their role as maintenance hosts for zoonotic strains remains unclear (M’ghirbi et al., 2012, 2016, 2022). In dromedaries, serological evidence of exposure was reported, but molecular confirmation was inconsistent or absent in some camel studies, raising the possibility of transient exposure or cross-reactivity with related Anaplasma genotypes (Ben Said et al., 2014; Belkahia et al., 2015; Selmi et al., 2019a). Ticks provide an additional ecological signal for A. phagocytophilum and related genotypes. A. phagocytophilum or closely related DNA has been detected in several tick contexts, including I. ricinus, Rhipicephalus spp., Hyalomma spp. and Hyalomma aegyptium collected from tortoises (Sarih et al., 2005; M’ghirbi et al., 2012; Belkahia et al., 2019; Rjeibi et al., 2022). These detections suggest that several tick-host systems may contribute to ecological circulation. However, tick positivity should not be equated with confirmed biological vector competence. The detected DNA may reflect an infected blood meal, transient carriage or true tick infection, and these possibilities require experimental or longitudinal evidence to be distinguished (Belkahia et al., 2019; Selmi et al., 2019a; Rjeibi et al., 2022). Despite the global zoonotic status of A. phagocytophilum, no confirmed human granulocytic anaplasmosis cases, human PCR detections or human serosurveys were documented in the reviewed Tunisian corpus (M’ghirbi et al., 2012).

Tunisian studies have also reported unclassified strains described as A. phagocytophilum-like 1 and A. phagocytophilum-like 2. These genotypes are important because they show that Anaplasma diversity in Tunisia is broader than classical species-level categories. Molecular surveys in sheep, goats and cattle detected strains genetically related to A. phagocytophilum, whereas specific A. phagocytophilum was not confirmed in the tested ruminants (Ben Said et al., 2015; Ben Said et al., 2017b). Additional detections in ticks and wildlife-associated contexts further support their ecological relevance (Belkahia et al., 2019; Selmi et al., 2021a). The major interpretive challenge is that “A. phagocytophilum-like” should not be treated as equivalent to zoonotic A. phagocytophilum. These strains may be related to known lineages, but their host range, pathogenicity and ability to infect humans remain undetermined, both in Tunisia and globally. They should therefore be described as uncertain zoonotic-relevance strains rather than confirmed human health threats (Ben Said et al., 2018).

A. platys, A. platys-like strains and camel-associated Anaplasma strains represent another unresolved issue. Classical A. platys is associated with canine infectious cyclic thrombocytopenia, and Tunisian evidence includes molecular detection in dog-associated ticks or canine-related contexts (Sarih et al., 2005). However, the most original Tunisian contribution concerns A. platys-like strains detected in dromedaries and ruminants (Belkahia et al., 2015; Ben Said et al., 2017a; Selmi et al., 2019a). In dromedaries, novel strains genetically related to A. platys were reported using molecular and phylogenetic approaches, while later work improved differentiation between classical A. platys and camel-associated A. platys-like strains using PCR-RFLP and sequencing (Belkahia et al., 2015; Selmi et al., 2019a). A. platys-like strains were also detected in cattle, sheep and goats, suggesting a broader host range than initially expected (Ben Said et al., 2017a). Findings related to A. platys-like are highly relevant for One Health surveillance because they occur at the interface between domestic animals, dromedaries, ticks and emerging zoonotic uncertainty. However, their human pathogenicity remains unconfirmed, and infected animals were often asymptomatic or not clearly linked to clinical disease (Belkahia et al., 2015; Ben Said et al., 2017a; Selmi et al., 2019b). The term “Candidatus” Anaplasma camelii should also be used cautiously. Tunisian camel strains are genetically related to camel-associated A. platys-like strains discussed in relation to Ca. A. camelii, but the taxonomic status and zoonotic implications of these isolates remain unresolved (Selmi et al., 2019c).

Anaplasma ovis is the most extensively documented Anaplasma species in Tunisian small ruminants. Molecular studies have reported high prevalence and genetic diversity in sheep and goats from northern and central regions, confirming that A. ovis is deeply established in Tunisian small-ruminant systems and represents an important veterinary concern (Belkahia et al., 2014, 2017; Ben Said et al., 2015; ElHamdi et al., 2022; M’ghirbi et al., 2022). Additional molecular characterization based on msp1a, groEL and gltA further confirmed the diversity of Tunisian A. ovis isolates from small ruminants, camels and associated ticks, supporting the value of high-resolution typing when comparing strains across hosts and regions (Ben Said et al., 2020). From a One Health perspective, A. ovis is important because it is increasingly discussed as a species of suspected zoonotic relevance (Chochlakis et al., 2010). However, its zoonotic role in Tunisia remains inferred rather than demonstrated, since no local human infection has been documented in the reviewed corpus (Ben Said et al., 2018; M’ghirbi et al., 2022). Therefore, the most appropriate interpretation is that A. ovis is highly prevalent in animals and may represent a potential exposure signal, but its clinical significance for humans in Tunisia remains unknown. The ecological range of A. ovis appears broader than small-ruminant systems alone. A. ovis DNA has been detected in Tunisian dromedaries, suggesting that camel-rearing systems may participate in arid-zone transmission cycles (Selmi et al., 2020c). Wildlife-associated detections further complicate the picture. A. ovis DNA was reported in wild rodents and in a rat flea pool, suggesting that synanthropic rodents and their ectoparasites may participate in additional ecological interfaces (Selmi et al., 2021a). However, these data should not be overinterpreted as proof of a stable sylvatic cycle or confirmed wildlife reservoir status. Larger longitudinal studies are needed to determine whether rodents maintain A. ovis independently or reflect spillover from domestic animal environments.

For Ehrlichia, the strongest Tunisian evidence concerns E. canis in dogs. Historical seroepidemiological work reported high levels of canine exposure across several Tunisian regions, with seroprevalence ranging from 48.6% to 85.4% depending on the region or dog population studied (Ghorbel et al., 1994). These findings establish canine ehrlichiosis as an important veterinary problem in Tunisia and suggest that the domestic dog-tick interface has long been involved in local Ehrlichia ecology (Ghorbel et al., 1994). Nevertheless, most historical canine data are based on indirect immunofluorescence rather than species-specific molecular characterization, and serological cross-reactivity between closely related Ehrlichia species can complicate species attribution (Ghorbel et al., 1994).

Human evidence is one of the weakest components of the Tunisian Ehrlichia literature. An early preliminary study reported seropositivity to E. canis in 9 of 89 patients with fever of unknown origin, suggesting possible exposure to Ehrlichia spp. (Ghorbel et al., 1991). Additional serological signals among dog owners have also been mentioned, but such findings do not establish confirmed human infection. This distinction is essential because E. canis, E. chaffeensis and other Ehrlichia species may show serological cross-reactivity, and older IFA/IFI-based studies cannot reliably identify the infecting species without molecular confirmation (Ghorbel et al., 1991, 1994). Based on the reviewed corpus, human exposure to Ehrlichia spp. has been suggested by serology, but molecular confirmation has not yet been documented in Tunisia.

Molecular evidence from ticks shows that Tunisian Ehrlichia diversity extends beyond classical canine serology. In a North African tick survey, Ehrlichia-like DNA was detected in I. ricinus and Hyalomma scupense, formerly referred to as Hyalomma detritum, with sequences related to the E. chaffeensis/E. ruminantium cluster (Sarih et al., 2005). These findings indicate the presence of uncharacterized Ehrlichia-related organisms in Tunisian tick fauna, particularly in northwestern humid environments. However, detection of Ehrlichia DNA in ticks does not prove that these tick species are competent biological vectors (Sarih et al., 2005; Balti et al., 2021). The brown dog tick, Rh. sanguineus s.l., remains central to the domestic ecology of E. canis, and suitability modeling suggests that northern and eastern coastal regions of Tunisia provide favorable conditions for this tick, supporting the relevance of the dog-tick interface in urban and peri-urban environments. However, vector suitability should not be interpreted as pathogen prevalence or transmission intensity without molecular screening of ticks and dogs.

In Tunisian context, the most original recent contribution to the Ehrlichia literature comes from wildlife. Molecular screening reported the detection of E. ewingii and Candidatus Ehrlichia shimanensis in North African hedgehogs and their ectoparasites, and E. canis in the common Pipistrelle bat tissues (Balti et al., 2021; Selmi et al., 2021b). These findings expand the recognized diversity of Ehrlichia taxa in Tunisian wildlife and highlight wild vertebrates as potentially important but still poorly characterized ecological compartments. E. ewingii is recognized globally as a zoonotic species, whereas Ca. E. shimanensis remains an emerging and incompletely characterized taxon (Balti et al., 2021). However, the epidemiological significance of these detections remains uncertain, as the contribution of wildlife hosts to Ehrlichia circulation cannot be established without longitudinal studies evaluating infection dynamics, host competence and transmission pathways.

In contrast to Anaplasma, for which ruminants and camels are strongly represented in Tunisia, the Ehrlichia evidence in livestock and dromedaries remains weak, indirect or negative. Some tick surveys involving small-ruminant or camel-associated contexts did not detect Ehrlichia DNA, suggesting that these compartments may be less central for Ehrlichia than for Anaplasma, although limited sampling prevents firm conclusions (Sarih et al., 2005). E. ruminantium should also be treated cautiously. Although it is a major veterinary pathogen in some African regions, the reviewed Tunisian corpus does not document confirmed infection of Tunisian livestock with E. ruminantium (Sarih et al., 2005).

Anaplasmataceae illustrate one of the most important One Health gaps identified in this review. Animal and tick signals are abundant, especially for A. ovis, A. phagocytophilum, A. phagocytophilum-like and A. platys-like strains, but the absence of human data prevents any conclusion on the real public health burden. For Ehrlichia, the strongest evidence concerns E. canis exposure in dogs, tick-associated Ehrlichia-like DNA and recent wildlife-associated detections, while human exposure remains serological and unconfirmed. Future work should prioritize targeted screening of febrile patients with tick exposure, molecular comparison of human, dog, livestock, camel, wildlife and tick strains, and high-resolution characterization of Anaplasma-like and Ehrlichia-like organisms. The taxon-level evidence for Anaplasma and Ehrlichia spp. is summarized in Table 1, while their main critical gaps and surveillance priorities are presented in Table 2. Their contrasting One Health evidence profiles are also illustrated in Figure 2, which highlights the strong animal-vector signal for Anaplasma and the more fragmented canine, tick and wildlife signal for Ehrlichia.

3.4 Borrelia spp.

Compared with Rickettsia and Bartonella, the evidence on Borrelia spp. in Tunisia shows a distinct One Health profile. The available literature supports environmental, vector-associated and animal-associated circulation of both the B. burgdorferi s.l. complex and relapsing fever-group Borrelia, but the human clinical burden remains insufficiently defined (Younsi et al., 2001, 2005; Bouattour et al., 2010; Ben Said et al., 2016; Selmi et al., 2024a). This makes Borrelia a particularly important genus for critical interpretation: Tunisia has clear evidence of enzootic circulation, yet human disease evidence remains mainly serological, fragmented or limited to small clinical reports (Mahfoudhi et al., 2015).

The epidemiology of Borrelia in Tunisia encompasses two distinct ecological and clinical systems. The first involves the B. burgdorferi s.l. complex, the causative agents of Lyme borreliosis, which has been primarily documented in the humid and sub-humid regions of northern Tunisia (Younsi et al., 2001, 2005; Selmi et al., 2024a). Among the Lyme borreliosis genospecies, B. lusitaniae is consistently the most frequently detected in Tunisian Ixodes ticks, whereas B. garinii, B. burgdorferi sensus tricto, and B. valaisiana have been reported only sporadically (Younsi et al., 2001, 2005). Although molecular analyses have revealed a high degree of genetic homogeneity among North African B. lusitaniae isolates, the pathogenic significance of these strains for human health in Tunisia remains uncertain (Younsi et al., 2005).

The second system comprises relapsing fever borreliae, notably B. crocidurae, which has been detected in Ornithodoros erraticus ticks collected from rodent burrows in arid and Saharan regions of the country (Bouattour et al., 2010). Additional detections of relapsing fever-group Borrelia DNA in ticks parasitizing North African hedgehogs suggest the existence of broader wildlife-tick transmission networks, despite the absence of Borrelia DNA in hedgehog tissue samples (Balti et al., 2021). These findings indicate that the Borrelia epidemiology extends beyond Lyme borreliosis and includes enzootic cycles of tick-borne relapsing fever in southern arid ecosystems that warrant dedicated surveillance and research efforts. Hard ticks, particularly I. ricinus, provide the main vector-associated evidence for the B. burgdorferi s.l. complex in Tunisia. Early studies documented B. lusitaniae in I. ricinus collected from northwestern Tunisia, especially in humid and sub-humid areas such as Béja, Bizerte, Siliana, Ain Draham and Jbelel Jouza (Younsi et al., 2001, 2005). These findings identify the northern humid forest belt as the principal ecological focus for Lyme-group Borrelia circulation in Tunisia. More recently, B. lusitaniae was detected and genetically identified in questing I. inopinatus ticks, suggesting that this tick species may also participate in local Lyme-agent cycles (Selmi et al., 2024a). However, the correct interpretation is that I. inopinatus is currently a tick species carrying B. lusitaniae DNA and a potential vector, not a biologically confirmed vector. Experimental transmission studies and longitudinal field monitoring are therefore required before its role in human risk can be firmly established.

The tick component of the Tunisian Borrelia evidence is mainly represented by B. crocidurae in O. erraticus. This species was detected in O. erraticus collected from rodent burrows in arid and Saharan regions, whereas O. normandi tested negative in the available study (Bouattour et al., 2010). This distinction is important because soft tick species should not be treated as epidemiologically equivalent. The O. erraticus-B. crocidurae association defines a southern transmission cycle that is ecologically distinct from the northern Ixodes-B. burgdorferi s.l. cycle. However, recent human cases of tick-borne relapsing fever are poorly documented in the available Tunisian corpus, and the disease may be underdiagnosed in rural or Saharan settings. The appropriate conclusion is therefore not that relapsing fever is common in southern Tunisia, but that B. crocidurae circulates in infected soft ticks and should be considered in the differential diagnosis of recurrent or unexplained febrile illnesses in exposed populations (Bouattour et al., 2010).

A previous molecular study reported B. burgdorferi s.l. DNA in goats, sheep, cattle and camels, with higher detection in goats and sheep than in cattle and camels (Ben Said et al., 2016). These findings suggest exposure or possible involvement in enzootic cycles, but they do not prove that these animals are competent reservoirs able to infect ticks. Livestock should therefore be discussed as exposed hosts or possible maintenance hosts for tick populations rather than confirmed reservoirs of Lyme borreliosis agents. This distinction is important because livestock may support tick abundance and human exposure without necessarily maintaining Borrelia transmission.

Wildlife-associated evidence provides an important ecological context for interpreting the apparent discrepancy between the frequent detection of B. lusitaniae in Tunisian Ixodes ticks and the limited documentation of human Lyme borreliosis. In contrast to classical Lyme borreliosis transmission systems in temperate regions of the Northern Hemisphere, where small mammals and some bird species represent highly competent reservoirs for several human-associated B. burgdorferi s.l. genospecies (Radolf et al., 2021), North African Lyme-group Borrelia cycles appear to involve a greater contribution of reptilian hosts (Doss et al., 2024). In Tunisia, the lizard Psammodromus algirus is a frequent host for immature I. ricinus stages, and experimental xenodiagnostic studies demonstrated that this reptile can maintain B. lusitaniae infection and infect feeding ticks, supporting its role as a competent reservoir within the Tunisian ecological context (Dsouli et al., 2006). Therefore, the predominance of a reptile-associated transmission cycle may represent one ecological factor contributing to the limited overlap between intense enzootic circulation of B. lusitaniae and documented human Lyme borreliosis cases. However, this interpretation should remain cautious because human infection risk is determined by multiple interacting factors, including the pathogenic potential of circulating Borrelia genotypes, vector competence, reservoir community composition, tick-host preferences and patterns of human exposure (Mead, 2022).

In Tunisia, the available wildlife data should therefore be interpreted as evidence of a specific ecological maintenance system rather than as direct evidence of human transmission. The association between P. algirus and B. lusitaniae is particularly relevant because this lizard is abundant in Mediterranean environments and frequently parasitized by immature Ixodes ticks (Dsouli et al., 2006). Nevertheless, other wildlife compartments remain insufficiently characterized. In the North African hedgehog study, ticks collected from A. algirus carried relapsing fever-group Borrelia DNA, whereas hedgehog tissue samples were negative, indicating that hedgehogs should not currently be considered confirmed reservoirs but rather hosts supporting infected ectoparasites (Balti et al., 2021).

Despite this evidence of enzootic circulation, the human epidemiological impact of Borrelia infection in Tunisia remains incompletely defined. Earlier investigations reported suspected Lyme borreliosis cases and serological evidence suggesting exposure among individuals from northern forested areas, including forest workers, but these observations did not provide molecular confirmation of active infection (Bouattour et al., 2004). A subsequent clinical report described an atypical Lyme borreliosis-like presentation responding to doxycycline; however, this remained an isolated observation mainly supported by serological findings (Mahfoudhi et al., 2015). Consequently, the current evidence supports the existence of human exposure and possible clinical cases, but does not allow estimation of the true national burden or confirmation of the predominant Borrelia genospecies responsible for human infection.

Based on the available corpus, there is no clear evidence of successful PCR or culture confirmation of B. burgdorferi s.l. directly from Tunisian human tissue, blood or cerebrospinal fluid (Bouattour et al., 2004; Mahfoudhi et al., 2015). The most likely conclusion is therefore that suspected human Lyme borreliosis exists in Tunisia, but its true frequency, genospecies attribution and clinical spectrum remain poorly characterized. Similarly, although B. crocidurae is documented in soft ticks from arid and Saharan settings, the contemporary human burden of tick-borne relapsing fever is not well quantified (Bouattour et al., 2010).

The Tunisian Borrelia evidence base therefore reveals a major One Health paradox. Environmental and animal-associated circulation is well documented, but human clinical relevance remains poorly defined. The main gap concerns the pathogenicity and clinical impact of B. lusitaniae, which dominates the vector evidence but has not been clearly linked to molecularly confirmed human disease in Tunisia (Younsi et al., 2005; Selmi et al., 2024a). A second gap concerns livestock, because molecular detection in goats, sheep, cattle and camels is important but does not prove reservoir competence (Ben Said et al., 2016). A third gap concerns vector competence, especially for I. inopinatus, which carries B. lusitaniae DNA but requires experimental confirmation as a biological vector (Selmi et al., 2024a). For relapsing fever, the main priority is improved clinical awareness and molecular diagnosis of recurrent fever syndromes in southern rural and Saharan settings, particularly near rodent burrows or soft tick habitats (Bouattour et al., 2010).

Therefore, Borrelia evidence in the Tunisian context should be considered as at least two distinct groups: a northern Ixodes-Lyme-Borrelia group and a southern Ornithodoros-relapsing fever-Borrelia group. The taxon-level evidence for Borrelia spp. is summarized in Table 3, while the main critical gaps and surveillance priorities for this genus are presented in Table 4.

3.5 Coxiella burnetii and coxiella-like organisms

Coxiella burnetii occupies a particular position in this review. Unlike classical vector-borne bacteria such as Rickettsia, Borrelia, Anaplasma or Ehrlichia, C. burnetii is primarily a livestock-associated and environmentally transmitted zoonotic bacterium, responsible for Q fever in humans and coxiellosis in animals (Kaabia and Letaief, 2009; Kaabia and Bahloul, 2021). Its inclusion in a review of vector-associated zoonotic bacteria in Tunisia is justified by repeated arthropod-associated detections and by the historical association of Coxiella with tick-screening studies, but its epidemiology should not be interpreted as that of a classical tick-borne pathogen (Selmi et al., 2019b; Balti et al., 2021). The Tunisian evidence indicates that the major One Health relevance of C. burnetii lies at the interface between livestock, reproductive products, milk, contaminated environments, occupational exposure and human Q fever, whereas tick-associated detections require cautious interpretation (Rekiki et al., 2005; Barkallah et al., 2018; Selmi et al., 2018).

Livestock represents the most significant evidence compartment for C. burnetii in Tunisia. In small ruminants, serological studies have identified C. burnetii as one of the infectious agents associated with abortion problems in sheep and goats (Rekiki et al., 2005; Guesmi et al., 2023). Early serological data reported flock-level exposure in small ruminants, while more recent work in organized sheep flocks with a history of abortion confirmed that C. burnetii remains relevant in reproductive disorders (Rekiki et al., 2005; Guesmi et al., 2023). However, because some studies targeted flocks with abortion histories, these findings should not be generalized as national baseline prevalence without broader population-based surveys.

Molecular evidence further underscores the veterinary and public health significance of C. burnetii. DNA of this bacterium was detected in vaginal swabs and milk from sheep incentral-eastern Tunisia. This confirms that infected animals may shed the bacterium through vaginal secretions and birth products (Barkallah et al., 2018). Nevertheless, molecular detection in milk should be interpreted as evidence of potential food-chain relevance rather than direct proof of milk-borne human transmission in Tunisia, particularly because shedding may be intermittent and exposure routes are often multiple (Barkallah et al., 2018; Selmi et al., 2018).

A serological study reported high C. burnetii exposure in Tunisian dromedaries and identified age, seasonality and abortion history as impactful factors related to seropositivity (Selmi et al., 2018). These findings suggest that camels should be incorporated into Q fever surveillance in southern Tunisia, especially because camel husbandry, slaughtering, raw milk consumption and close contact during parturition may create occupational exposure contexts (Selmi et al., 2018). However, direct camel-to-human transmission remains insufficiently documented, and seropositivity indicates exposure rather than active shedding. Therefore, camel-associated C. burnetii evidence should be interpreted as an important arid-zone One Health signal requiring further integration with human and environmental data (Selmi et al., 2018; Selmi et al., 2019b).

Human Q fever is documented in Tunisia, but its true burden remains poorly defined. Hospital-based data from Sfax described confirmed cases, with clinical presentations dominated by pneumonia and hepatitis, and a proportion of patients reporting direct livestock contact (Ben Arab et al., 2008). These clinical features make Q fever difficult to recognize when laboratory suspicion is low. The key limitation is that available human data are mainly hospital-based or review-based and do not provide a reliable national incidence estimate (Ben Arab et al., 2008). Moreover, Q fever is considered underestimated in Tunisia, partly because it is not routinely suspected, clinical signs are non-specific and the disease is not systematically integrated into mandatory notification systems (Naoufel and Abdelkarim, 2021).

Ticks and other ectoparasites are relevant to the Coxiella discussion, but their role must be interpreted cautiously. C. burnetii DNA was detected in Hy. dromedarii and Hy. impeltatum ticks collected from dromedary camels, whereas camel blood was tested negative by PCR (Selmi et al., 2019b). This suggests tick-associated molecular detection in camel systems, but it does not prove biological tick-borne transmission. Similarly, C. burnetii DNA was detected in North African hedgehogs and their ectoparasites, including ticks and fleas, but the reservoir competence of hedgehogs and the transmission role of their arthropods remain unclear (Balti et al., 2021). In the Tunisian context, these findings should therefore be described as arthropod-associated or wildlife-associated molecular detections, not as proof of tick-borne Q fever transmission.

A separate issue concerns Coxiella-like organisms and tick endosymbionts. These organisms should not be confused with pathogenic C. burnetii, because their zoonotic significance is uncertain and they may represent tick-associated symbiotic bacteria rather than human pathogens (Selmi et al., 2019b; Benyedem et al., 2022). In Tunisian cattle-associated Hyalomma ticks, a microbiome study did not detect Coxiella endosymbionts and instead reported predominance of other symbionts, including Francisella-like organisms (Benyedem et al., 2022). This negative finding is important because it prevents overgeneralization about the presence of Coxiella-like organisms in all Tunisian ticks and shows that tick microbiomes may vary according to tick species, host and region.

Wildlife data provide an additional but still exploratory perspective on Coxiella ecology in Tunisia. The detection of C. burnetii DNA in the North African hedgehog and its associated ectoparasites suggests that peri-domestic wildlife interfaces may contribute to the ecological complexity of Coxiella circulation beyond traditional livestock settings (Balti et al., 2021). However, the epidemiological importance of these wildlife-associated findings remains uncertain, and further longitudinal studies are required to clarify the contribution of wild hosts to pathogen maintenance and transmission dynamics. In the current Tunisian context, the main public health relevance of C. burnetii remains associated with domestic ruminants, camels, reproductive materials, milk and environmental contamination pathways, which are supported by stronger evidence from livestock-associated studies (Rekiki et al., 2005; Barkallah et al., 2018; Selmi et al., 2018; Guesmi et al., 2023).

Tunisian studies used complementary diagnostic approaches, including complement fixation test, indirect ELISA, indirect immunofluorescence, PCR, real-time PCR, microfluidic PCR, sequencing and next-generation sequencing (Rekiki et al., 2005; Ben Arab et al., 2008; Barkallah et al., 2018; Selmi et al., 2018; Selmi et al., 2019c; Balti et al., 2021; Benyedem et al., 2022; Guesmi et al., 2023). In humans, serology is commonly used. However, the distinction between phase I and phase II antibody responses is important for differentiating acute from chronic Q fever, particularly in suspected endocarditis (Kaabia and Bahloul, 2021). In animals, serology indicates exposure, whereas PCR on milk, vaginal swabs, placentas or abortion material provides stronger evidence of shedding or active infection (Barkallah et al., 2018; Rekiki et al., 2005; Guesmi et al., 2023). The main methodological challenge is therefore the interpretation of the sample type: seropositive animals are not necessarily active shedders, negative blood PCR does not exclude infection, and tick PCR positivity does not prove biological transmission.

Despite substantial evidence, several gaps limit the translation of C. burnetii research into surveillance. First, human Q fever is documented but remains underestimated because of non-specific clinical signs, limited routine testing and absence of systematic notification (Ben Arab et al., 2008). Second, many livestock studies are based on flocks with abortion problems or selected farm systems, limiting the ability to estimate national baseline prevalence (Rekiki et al., 2005; Vanderburg et al., 2014; Guesmi et al., 2023). Third, environmental transmission is strongly plausible but insufficiently measured, because direct dust or aerosol sampling is lacking in the reviewed corpus (Selmi et al., 2018). Future studies should therefore combine human clinical surveillance, animal reproductive-sample testing, milk monitoring, environmental sampling and occupational exposure assessment, especially in small-ruminant and camel-rearing systems.

For surveillance purposes, C. burnetii should be considered as a major One Health pathogen in Tunisia, but not as a classical vector-borne bacterium. Its public health importance is primarily linked to livestock, birth products, milk, aerosols and contaminated environments, whereas tick-associated detections and Coxiella-like organisms should be interpreted cautiously and should not be presented as evidence of confirmed tick-borne Q fever transmission (Selmi et al., 2019b; Balti et al., 2021; Benyedem et al., 2022). The taxon-level evidence for C. burnetii, Coxiella-like organisms and related tick-associated signals is summarized in Table 3, while the main critical gaps and surveillance priorities are presented in Table 4.

3.6 Other vector-associated bacteria

Beyond the major bacterial groups treated separately in this review, recent Tunisian studies have revealed a diverse set of additional bacteria associated with ticks, livestock and vector-related ecological interfaces. These taxa include Francisella-like endosymbionts, M. mitochondrii, Rickettsiella spp., opportunistic genera such as Staphylococcus, Acinetobacter, Pseudomonas and Corynebacterium, as well as bacteria specifically investigated but not detected, including Neoehrlichia mikurensis and Spiroplasma ixodetis (Benyedem et al., 2022; Baede et al., 2024; Kratou et al., 2025b). This group also includes the hemotropic mycoplasma, M. ovis, molecularly detected in Tunisian sheep and usually considered as an arthropod-borne pathogen rather than a classical zoonotic agent (Rjeibi et al., 2015). These bacteria are discussed here not because they represent confirmed human zoonoses in Tunisia, but because they illustrate the complexity of vector, livestock and microbiome-associated bacterial signals and their potential relevance for future One Health surveillance.

Most available evidence for tick microbiome-associated taxa is based on molecular detection of bacterial DNA, mainly through 16S rRNA next-generation sequencing, qPCR or conventional PCR, rather than culture, species-level confirmation, animal disease, human infection or proven vector competence (Benyedem et al., 2022; Baede et al., 2024; Kratou et al., 2025b). These taxa should therefore be discussed as vector-associated or arthropod-associated bacteria of uncertain zoonotic significance rather than established human pathogens in Tunisia.

The most prominent bacterial signal in Tunisian Hyalomma tick microbiome studies is the genus Francisella. In cattle-associated Hyalomma ticks, particularly Hy. scupense, Hy. excavatum and Hy. marginatum, Francisella-related sequences were among the dominant bacterial components, with high relative abundance in some tick groups (Benyedem et al., 2022). In Hy. dromedarii collected from dromedary camels in Kébili, Francisella also dominated the bacterial microbiome of Tunisian ticks and differed from the microbiome profile observed in Saudi Arabian ticks (Kratou et al., 2025b). These findings indicate that Francisella-like bacteria are core components of several Tunisian Hyalomma microbiomes.

However, this result must not be interpreted as evidence of tularemia circulation. The genus Francisella includes F. tularensis, the confirmed zoonotic agent of tularemia, but the sequences detected in Tunisian tick microbiome studies most likely correspond to Francisella-like endosymbionts rather than confirmed pathogenic F. tularensis (Benyedem et al., 2022; Kratou et al., 2025b). The reviewed corpus does not provide direct evidence that F. tularensis was isolated, molecularly confirmed or linked to human disease in Tunisia. The appropriate conclusion is therefore that ticks harbor abundant Francisella-like endosymbionts, while the presence of pathogenic F. tularensis remains unproven. Future work should use species-specific PCR, multilocus sequencing or whole-genome sequencing to distinguish Francisella-like endosymbionts from pathogenic F. tularensis and to assess whether these endosymbionts influence tick fitness or pathogen interactions.

Midichloria mitochondrii is another important tick-associated bacterium detected in Tunisia. It was reported in both Hyalomma and Ixodes ticks, including I. ricinus and I. inopinatus from northwestern Tunisia (Benyedem et al., 2022; Baede et al., 2024). In adult Ixodes ticks collected from humid forest environments, molecular detection rates of M. mitochondrii were approximately 35–37.1%, indicating that this endosymbiont is common in Tunisian Ixodes populations (Baede et al., 2024). The biological significance of M. mitochondrii remains uncertain. As an endosymbiont, it may influence tick physiology, reproduction or microbial community structure, but the reviewed corpus does not demonstrate that it causes disease in humans or animals in Tunisia.

Among the other bacterial taxa, Rickettsiella spp. represent an emerging tick-associated signal. Rickettsiella DNA was detected in questing I. ricinus and I. inopinatus ticks from northwestern Tunisia, with reported molecular detection rates around 6.5–7.5% (Baede et al., 2024). This finding suggests that Rickettsiella spp. are present in Tunisian Ixodes populations and may form part of the bacterial community associated with humid forest tick habitats. However, the available evidence remains limited to molecular detection in ticks. The reviewed corpus does not provide evidence of human infection, animal disease, reservoir hosts, clinical cases or transmission routes for Rickettsiella spp. in Tunisia. A more accurate interpretation is therefore that Rickettsiella spp. are tick-associated bacteria of uncertain medical significance that warrant further characterization rather than confirmed zoonotic agents.

Microbiome studies of Tunisian ticks also detected opportunistic or environmental bacterial genera such as Staphylococcus, Acinetobacter, Pseudomonas, Corynebacterium, Bacillus, Erwinia, Serratia and Janthinobacterium (Benyedem et al., 2022; Kratou et al., 2025b). Some of these genera include species of medical or veterinary importance, especially Staphylococcus and Acinetobacter, which may include opportunistic or antimicrobial-resistant organisms (Kratou et al., 2025b). Nevertheless, these findings should be interpreted as microbiome signals rather than evidence of tick-borne disease. Detection by 16S rRNA sequencing usually does not provide enough resolution to identify pathogenic species or virulence traits, and these bacteria may represent environmental contamination, host skin flora, blood-meal remnants or transient microbial passengers (Benyedem et al., 2022; Kratou et al., 2025b). It would therefore be inappropriate to state that Tunisian ticks transmit Staphylococcus, Acinetobacter or Pseudomonas to humans or animals based only on microbiome detection.

Negative findings remain informative in this group because they help define current surveillance boundaries. N. mikurensis, an emerging tick-borne human pathogen in Europe, was specifically investigated in northwestern Tunisian Ixodes ticks but was not detected (Baede et al., 2024). S. ixodetis was also investigated in the same tick context and was not detected (Baede et al., 2024). These results do not prove absence from Tunisia as a whole, because they are limited to specific tick species, regions, sampling periods and sample sizes, but they provide an important baseline for future surveillance.

The main strength of this corpus is the application of modern molecular tools, especially 16S rRNA next-generation sequencing, qPCR, conventional PCR, sequencing and microbial network analysis (Benyedem et al., 2022; Baede et al., 2024; Kratou et al., 2025b; Rjeibi et al., 2015). The major weakness is that much of the evidence remains descriptive. Genus-level microbiome profiling is often insufficient for distinguishing harmless symbionts from pathogenic species, especially in genera such as Francisella, Staphylococcus or Acinetobacter (Benyedem et al., 2022; Kratou et al., 2025b).

The evidence for other vector-associated bacteria, including tick endosymbionts, microbiome-associated taxa, opportunistic genera, M. ovis and informative negative findings, is summarized in Table 3, while related critical gaps and surveillance priorities are presented in Table 4. Their predominantly vector/microbiome- and livestock-centered profile is illustrated in Figure 3, emphasizing that these taxa should be interpreted mainly as ecological, endosymbiotic or veterinary signals rather than confirmed zoonotic threats.

4 Discussion: toward an integrated one health surveillance and action framework for Tunisia

This review highlights key challenges for implementing an integrated One Health surveillance framework for vector-borne zoonotic bacteria in Tunisia. The main challenge is no longer pathogen detection, but integrating clinical, veterinary, entomological, wildlife, environmental, and microbiome data to assess epidemiological significance and guide public health action.

4.1 Tunisia as a mosaic of One Health systems

The Tunisian evidence should not be interpreted as a uniform national epidemiological pattern. Instead, vector-associated zoonotic bacteria appear to circulate through a mosaic of ecological and socio-economic interfaces that differ in climate, landscape, dominant vectors, animal hosts and human exposure pathways. This interface-based interpretation is summarized in Figure 4, which should be viewed as a conceptual gradient rather than as an administrative map.

Several major One Health systems can be distinguished. Domestic and peri-domestic settings are mainly shaped by dog-tick-human and cat/dog-flea-human interfaces, which are particularly relevant for Rickettsia, Ehrlichia and Bartonella circulation (Ghorbel et al., 1994; Znazen et al., 2005; Hammami et al., 2013; Znazen et al., 2013; Khrouf et al., 2016; Zouari et al., 2017; Gharbi et al., 2024). Rural livestock systems involve sheep, goats, cattle, horses, ticks, farmers and veterinarians, and are especially important for Anaplasma, Borrelia, Rickettsia and C. burnetii (M’ghirbi et al., 2012; Ben Said et al., 2016; Belkahia et al., 2017; Barkallah et al., 2018; Guesmi et al., 2023; Kratou et al., 2023). Arid and Saharan systems are characterized by dromedaries, Hyalomma ticks, camel-associated occupational exposure and raw milk practices, with relevance for Coxiella, Anaplasma, Rickettsia and tick microbiome studies (Belkahia et al., 2015; Selmi et al., 2018, 2019a,b, 2020b,c; Kratou et al., 2025b). In northern humid and forested areas, Ixodes-associated systems involve wildlife, reptiles and humans exposed during forest-related activities, and are particularly relevant for Lyme-group Borrelia, some Rickettsia spp., Anaplasma-related taxa and tick endosymbionts (Younsi et al., 2001; Sfar et al., 2008; M’ghirbi et al., 2012; Baede et al., 2024; Selmi et al., 2024a; Kratou et al., 2025a). Finally, wildlife and peri-domestic interfaces add further ecological complexity through rodents, hedgehogs, reptiles, wild carnivores and their ectoparasites (Balti et al., 2021; Selmi et al., 2021a,b, 2025; Rjeibi et al., 2022).

Surveillance should therefore be adapted to ecological interfaces rather than administrative regions alone: Ixodes surveillance in northern forests; dog, cat, flea and Rhipicephalus monitoring in peri-urban areas; integrated animal-tick surveillance in livestock regions; targeted protocols in camel-rearing areas; and systematic monitoring of wildlife interfaces. This provides a practical basis for the proposed One Health surveillance workflow.

4.2 Proposed surveillance workflow

An integrated One Health surveillance workflow for Tunisia should follow a stepwise logic that moves from signal detection to targeted sampling, laboratory confirmation, evidence ranking and public health action. This approach is consistent with evidence-mapping and One Health risk-interpretation principles, which emphasize diagnostic certainty, compartmental convergence and cautious interpretation of heterogeneous evidence (Page et al., 2021; Tricco et al., 2018; FAO, WHO, and WOAH, 2019, 2020). The purpose of this workflow is not to multiply isolated screenings, but to connect signals generated in different compartments and interpret them according to their epidemiological relevance. Potential signals may arise from compatible human clinical syndromes, animal or wildlife sentinels, arthropod screening, environmental contamination, food-chain exposure or microbiome datasets. These signals should guide targeted investigations within the relevant ecological interface rather than broad, non-contextual sampling.

Sampling should be tailored to the suspected pathogen and epidemiological interface. Human sampling should prioritize patients with compatible syndromes and relevant exposure histories, while animal and wildlife surveillance should target epidemiologically relevant sentinel hosts. Arthropod surveillance should focus on ticks, fleas, and other ectoparasites across domestic, livestock, camel-associated, forest, and wildlife interfaces (Ghorbel et al., 1994; Znazen et al., 2005; M’ghirbi et al., 2012; Ben Said et al., 2016; Zouari et al., 2017; Balti et al., 2021; Gharbi et al., 2024; Selmi et al., 2024a). For C. burnetii, surveillance should extend beyond arthropods and include reproductive materials, milk and environmental samples, because the main risk is livestock-associated and largely aerosol- or environment-based rather than classically vector-borne (Rekiki et al., 2005; Barkallah et al., 2018; Selmi et al., 2018; Guesmi et al., 2023).

Laboratory confirmation should combine serology, PCR/qPCR, sequencing and, when necessary, multilocus typing or whole-genome sequencing. High-resolution molecular tools are particularly important for poorly resolved or emerging taxa, such as Anaplasma-like, Coxiella-like and Francisella-like organisms, as well as for distinguishing closely related bacterial taxa detected in animals, vectors or microbiome datasets (Belkahia et al., 2015; Ben Said et al., 2017a,b; Benyedem et al., 2022; Baede et al., 2024; Kratou et al., 2025a,b). Diagnostic findings should be interpreted within an evidence hierarchy distinguishing exposure, active infection, molecular detection, vector and reservoir competence, and confirmed human disease. This prevents overinterpretation of DNA detection in vectors, animals, or microbiome datasets as evidence of zoonotic transmission. The final step should translate classified evidence into proportionate One Health action. Confirmed or strongly suspected zoonotic signals should trigger clinical awareness, targeted diagnostic testing, veterinary control measures, ectoparasite management, occupational prevention and risk communication. By contrast, microbiome-associated endosymbionts, poorly characterized environmental signals and opportunistic taxa should primarily guide research prioritization and surveillance refinement rather than immediate public health alerts. This proposed workflow is summarized in Figure 5 and is supported by the evidence hierarchy and priority gaps presented in Tables 1–4.

4.3 Strategic recommendations for surveillance, research and public health action

The Tunisian evidence supports a progressive One Health strategy focused on priority interfaces, compatible clinical syndromes and evidence-based risk classification. Human surveillance should prioritize patients showing unexplained fever, rash, eschar, recurrent fever, pneumonia, hepatitis, neurological signs after arthropod exposure, or blood culture-negative endocarditis. Particular attention should be given to exposed groups, including farmers, shepherds, veterinarians, abattoir workers, camel breeders, forest workers, military personnel and people living in close contact with dogs, cats, livestock, camels, rodents or ectoparasites. This targeted approach would improve clinical recognition while avoiding non-specific or poorly contextualized screening. Veterinary and entomological surveillance should be directly connected to human clinical investigation. Dogs, cats, livestock, dromedaries, wildlife and ectoparasites should be used as sentinel compartments according to the ecological interface involved. However, animal or arthropod positivity should not be interpreted in isolation. Each finding should be assessed through an evidence hierarchy that distinguishes exposure, active infection, DNA carriage, vector competence, reservoir competence and confirmed human disease.

Future research should therefore prioritize studies that connect compartments rather than studies that only add isolated detections. Longitudinal investigations, paired human-animal-vector sampling, strain comparison across compartments and systematic reporting of negative findings would substantially improve the Tunisian evidence base. Under-sampled regions, especially central, southern, Saharan and peri-urban settings, should receive greater attention. These areas are particularly important because they include livestock systems, camel-associated environments, peri-domestic exposure and wildlife interfaces that remain incompletely characterized.

A shared national One Health data platform would help transform scattered findings into actionable surveillance. Such a platform should link human cases, animal infections, arthropod detections, wildlife findings, environmental samples, diagnostic methods, geographical information, seasonality and sequence data. Its main value would be to classify evidence according to public health relevance, separating confirmed zoonotic pathogens, suspected zoonotic agents, animal pathogens with uncertain human significance, vector-associated DNA signals, endosymbionts, opportunistic microbiome taxa and informative negative findings. Operationally, this strategy should support proportionate public health action. Confirmed or strongly suspected zoonotic signals should lead to clinical awareness, targeted diagnostics, veterinary control, ectoparasite management, occupational prevention and risk communication. In contrast, weakly characterized microbiome or environmental signals should mainly guide research priorities and surveillance refinement. These practical priorities are aligned with the genus-specific and cross-cutting gaps summarized in Tables 2, 4 and with the surveillance workflow proposed in Figure 5.

4.4 Priority research questions

A national One Health strategy for vector-associated zoonotic bacteria in Tunisia should be guided by priority questions that directly connect research, surveillance and public health action. The objective is not only to document additional bacterial detections, but to determine which signals correspond to confirmed zoonotic disease, which remain animal- or vector-associated findings, and which mainly reflect ecological, environmental or microbiome background. Several priority questions emerge from the Tunisian evidence base. First, the true human burden of non-classical rickettsioses, Q fever, bartonellosis, Lyme borreliosis, tick-borne relapsing fever, anaplasmosis and ehrlichiosis remains incompletely defined. Second, the zoonotic relevance of animal-centered taxa such as A. ovis, A. platys-like strains, A. phagocytophilum-like genotypes and camel-associated Anaplasma strains requires clarification. Third, the role of specific vectors, including I. inopinatus, Rh. sanguineuss.l., Hyalomma spp., O. erraticus, fleas and wildlife-associated ticks, should be evaluated beyond DNA detection. Fourth, the contribution of dromedaries, canids, ruminants, rodents and wildlife to maintenance or spillover cycles remains insufficiently resolved. Finally, the functional role of tick endosymbionts, including Francisella-like organisms and M. mitochondrii, in vector fitness or pathogen interactions represents an important emerging research direction.

Addressing these questions will require study designs that connect compartments rather than examine them separately. Priority should be given to longitudinal surveillance, paired human-animal-arthropod sampling, molecular comparison of strains across hosts and vectors, environmental sampling in high-risk interfaces and systematic reporting of negative findings. Such approaches would help distinguish incidental detection from epidemiologically meaningful circulation and would improve the interpretation of animal, vector, wildlife and microbiome signals. The added value of this research agenda is that it links unresolved scientific questions to decision-making. A bacterium detected only in a tick microbiome should be classified as an ecological signal requiring characterization, not as an immediate public health threat. A bacterium detected in animals and vectors but not in humans would justify targeted human surveillance in exposed groups. Conversely, a bacterium detected in compatible human disease, animals and arthropods would become a high-priority One Health surveillance target.

5 Conclusion

Tunisia has generated substantial evidence on vector-associated zoonotic bacteria, but important gaps remain across bacterial groups, ecological compartments, and levels of epidemiological confirmation. While Rickettsia spp., Bartonella spp., and C. burnetii are the best-documented pathogens, evidence for Anaplasma, Ehrlichia, and Borrelia spp. is derived mainly from animal, vector, and environmental studies, with limited data on human infection. Future progress will depend less on additional detections than on integrating existing evidence within a One Health surveillance framework. Molecular detection should be distinguished from active infection, vector competence, reservoir competence, and confirmed human disease. Strengthening the integration of clinical, veterinary, entomological, wildlife, and environmental surveillance, supported by high-resolution molecular confirmation, will improve risk assessment and public health preparedness. Such an approach could position Tunisia as a regional model for evidence-based surveillance of vector-associated zoonotic bacteria in North Africa and the Mediterranean.

Generative AI statement

The author(s) declared that Generative AI was used in the creation of this manuscript. During manuscript preparation, the authors used an AI-assisted tool (Google NotebookLM, developed by Google) for language editing and scientific writing refinement to enhance the manuscript’s cohesion and coherence. Following the use of this tool, the authors reviewed and edited the content as needed and take full responsibility for the technical accuracy and the final version of the publication.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmicb.2026.1951493/full#supplementary-material

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Footnotes

Edited by: Juan Luis Muñoz-Bellido, University of Salamanca, Spain

Reviewed by: Giusto Trevisan, University of Trieste, Italy

Moaz M. Amer, Animal Health Research Institute (AHRI), Egypt