Expanding the wildlife reservoir landscape: A high-throughput molecular surveillance study of zoonotic pathogens in Korean water deer

Abstract

The movement of wildlife across geographic boundaries increasingly drives the emergence and dissemination of transboundary zoonotic diseases. Here, we applied the TaqMan Array Card (TAC) platform for the first time in wildlife disease surveillance, enabling high-throughput screening of multiple pathogens in Korean water deer (Hydropotes inermis argyropus, KWD)—a species widely distributed and frequently interacting with humans and livestock. Between April and November 2023, we collected 192 spleen tissue samples from 12 regions. Pathogen screening was performed using the TAC system, with positive detections subsequently validated via phylogenetic analysis. Seven zoonotic pathogens were identified: Ehrlichia canis (0.5 %), E. muris (1.6 %), Neoehrlichia mikurensis (4.2 %), Candidatus Rickettsia longicornii (26.0 %), Rickettsia raoultii (2.1 %), Bartonella schoenbuchensis (6.8 %), and Giardia duodenalis (1.0 %). Notably, E. muris was detected in KWD for the first time globally, and G. duodenalis was identified in spleen tissue—marking the first such report from any cervid species. Spatial and temporal analyses revealed distinct region-specific and seasonal trends in pathogen prevalence, providing key epidemiological insights. These results underscore the ecological significance of KWD as potential reservoirs and amplifiers of emerging zoonoses. Furthermore, the study demonstrates the utility of the TAC system as a rapid, scalable diagnostic platform for wildlife pathogen surveillance, supporting proactive disease monitoring under the One Health framework.

Highlights

• •First use of TAC system for pathogen surveillance in wildlife (Korean water deer)
• •Ehrlichia muris detected in Korean water deer for the first time globally.
• •Giardia duodenalis found in spleen tissue of cervids for the first time.
• •Seven zoonotic pathogens identified via high-throughput molecular screening.
• •Findings support One Health-based integration of wildlife into surveillance systems.

1. Introduction

Pathogen monitoring in wastewater has become a critical approach for early outbreak detection and plays a significant role in global infectious disease surveillance [1]. In South Korea, the Korea Disease Control and Prevention Agency (KDCA) implemented wastewater surveillance (WS) using the TaqMan Array Card (TAC) platform, first introduced during the COVID-19 pandemic [2], enabling earlier identification of emerging pathogens than hospital-based surveillance [3]. Despite its demonstrated success in human health, the WS concept has not yet been adapted in the veterinary field, including livestock and wildlife. In this study, we applied this concept to wildlife for the first time, integrating the TAC platform to screen for multiple pathogens in Korean water deer (Hydropotes inermis argyropus, KWD).

The KWD is widely distributed across South Korea and is recognized as an important host for tick-borne pathogens (TBPs). Our previous nationwide survey, conducted with the National Institute of Wildlife Disease Control and Prevention (NIWDC), demonstrated that KWD harbor multiple pathogens capable of infecting humans and animals [4]. In addition, KDCA surveillance has reported circulation of severe fever with thrombocytopenia syndrome virus (SFTSV) and other tick-borne diseases [5], while the Animal and Plant Quarantine Agency has documented highly pathogenic avian influenza in wild birds [6]. These findings underscore the zoonotic potential of wildlife reservoirs. Moreover, KWD often inhabit forest edges, agricultural landscapes, and peri-urban areas [7], where they overlap with humans and domestic animals, further increasing the relevance of monitoring.

In South Korea, wildlife pathogen surveillance has mainly relied on PCR or serological testing, which provide baseline data but are limited in scope and efficiency [8]. The TAC platform, a 384-well microfluidic multiplex qPCR system, enables simultaneous detection of multiple pathogens from a single specimen [3], with reduced handling time, improved cost-efficiency, and high throughput [9,10]. In this study, we applied TAC to wildlife for the first time by screening for 45 selected pathogens in KWD, demonstrating its potential to enhance wildlife pathogen surveillance frameworks by improving sensitivity, efficiency, and breadth of coverage, ultimately contributing to early warning systems and public health preparedness.

2. Materials and methods

2.1. Ethical approval

Between April and November 2023, spleen samples from KWD were collected by licensed hunters under the supervision of the NIWDC. Of the 1035 samples obtained in a previous nationwide survey [4], 192 were selected for this study. All procedures were approved by the Institutional Animal Care and Use Committee of Kyungpook National University (Approval No. KNU 2024–0407).

2.2. Sample collection

Between April and November 2023, spleen samples were collected as part of the NIWDC wildlife surveillance program targeting SFTS, and the same samples were also used in our previous study on Anaplasma, Borrelia, and Theileria [4]. For this study, 192 samples were selected from 12 regions, ensuring both regional and seasonal representativeness within available resources. The detailed sample distribution and geographic coordinates are shown in Fig. 1. Selection prioritized individuals previously positive for Anaplasma, Borrelia, and Theileria to increase the likelihood of detecting additional pathogens.

Fig. 1.

Sampling sites of Korean water deer across South Korea. (a) Twelve collection sites were classified as northern, central, or southern regions. The northern region includes Wonju (WJ, n = 27), Chuncheon (CC, n = 12), Gapyeong (GP, n = 12), Namyangju (NY, n = 13), Yangpyeong (YP, n = 13), Yeoncheon (YC, n = 13), and Pocheon (PC, n = 20). The central region comprises Andong (AD, n = 17), Yeongyang (YY, n = 13), and Gongju (GJ, n = 21), while Jinju (JJ, n = 14) and Hapcheon (HC, n = 17) constitute the southern region. (b) Spatial distribution map displaying GPS coordinates for all 192 sampled Korean water deer collected from April to November 2023. Each red point indicates an individual sampling site.

2.3. Gene extraction and processing

Approximately 10 mg of spleen tissue was homogenized (Precellys CK28-R Lysing kit, Bertin Technologies, Bretonneus, France) and DNA/RNA were simultaneously extracted using the Clear-S™ Quick DNA Extraction Kit (Invirustech, Gwangju, Korea). To verify successful extraction and amplification, TaqMan Universal Extraction Control Organism (Bacillus atrophaeus, 5 × 10³ copies/μL, Thermo Fisher Scientific, Waltham, USA) was added to each homogenate. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) served as a housekeeping gene to assess sample processing adequacy, and a manufacturer-provided 18S rRNA quality control assay was included to confirm proper array spotting and plate performance.

2.4. Detection of diseases using TAC system

To optimize and customize the TAC system, we selected 45 infectious diseases for analysis based on their known prevalence in KWD, zoonotic potential, and relevance to public and veterinary health. Of these, some assays were manufacturer-validated commercial assays provided by Thermo Fisher Scientific (TFS), whereas others were custom-designed in collaboration with the manufacturer to target pathogens of particular concern in South Korea. All assays were manufactured and quality-controlled by TFS; however, primer and probe sequences remain proprietary and are not publicly disclosed. The full list of 45 pathogens included in the panel is provided in Supplementary Table S1.

The TAC platform was used to detect these pathogens with TaqPath™ 1-Step RT-qPCR Master Mix (TFS). Preliminary validation was performed using customized synthetic positive controls (TFS), consisting of a linearized multi-target plasmid DNA pool representing 20 selected pathogens (listed in Supplementary Table S1). These plasmid DNA constructs included both DNA and RNA targets, with RNA viruses represented as DNA constructs according to the manufacturer's specifications. The controls were serially diluted from 1 × 10⁵ to 1 × 10¹ copies/μL. Negative controls (distilled water) were included in each run.

Each TAC plate (48 wells) contained 45 pathogen assays, GAPDH for host species identification, a B. atrophaeus extraction control, and one manufacturer-provided quality control (QC) assay targeting 18S rRNA, which served as a manufacturing control for the array card. Cards passed QC if the 18S assay showed a Cq value within 8–15 and a standard deviation ≤0.5, as specified by the manufacturer. Samples were considered positive if the Amp score exceeded 1.0, the Cq confidence exceeded 0.7, and the Cq value was <35. All TAC-positive results were subsequently re-tested using conventional PCR assays targeting the same pathogens.

2.5. Detection of diseases using conventional PCR

Conventional PCR was used to confirm TAC-positive results and to generated sequences for phylogenetic analysis. For validation, specific target gene regions and primer sets corresponding to TAC-detected pathogens were required. Because the TAC platform is commercially developed and does not normally disclose target gene information, the relevant details were provided by the manufacturer under a research agreement, enabling PCR validation and direct sequencing. Ehrlichia and Rickettsia were detected with the AccuPower Rickettsiales 3-Plex PCR (Bioneer, Daejeon, South Korea), Bartonella with primers supplied by TFS, and Giardia using previously published protocols [11]. The sequenced products corresponded to partial regions of the 16S rRNA gene (Ehrlichia, Rickettsia), the 23S rRNA gene (Bartonella), and the 18S rRNA gene (Giardia), which were subsequently used for phylogenetic tree construction. Both positive and negative controls were included in each PCR to ensure assay reliability.

2.6. Genetic sequencing and phylogenetic study

PCR-positive products were sequenced at Macrogen (Seoul, South Korea) using their respective primers. Obtained sequences were compared with NCBI GenBank database, aligned with CLUSTAL Omega, and edited in BioEdit. Phylogenetic trees were constructed in MEGA software [12] using the Kimura two-parameter model [13] and maximum likelihood method.

2.7. Statistical examination

Statistical analyses were performed using GraphPad Prism version 5.04 (GraphPad Software Inc., La Jolla, CA, USA). Pearson's chi-square was applied to contingency tables with >2 variables. For multiple comparisons, p-values were adjusted using the Bonferroni correction (regional: adjusted α = 0.0042 [0.05/12 regions]; seasonal: adjusted α = 0.00625 [0.05/8 months]). To avoid spurious results caused by insufficient sample sizes, pathogens with <10 positive detections were excluded from statistical testing. A p-value ≤0.05 was considered statistically significant. Ninety-five percent confidence intervals (CIs) were calculated. To evaluate assay performance and reproducibility in preliminary tests, standard curves were generated using 20 synthetic positive controls serially diluted from 1 × 10⁵ to 1 × 10¹ copies/μL. Linear regression analysis was performed to calculate correlation coefficients (R²) for each pathogen target.

3. Results

3.1. Validation of experiment

To assess experimental performance, 20 customized positive controls were tested. All targets were reliably detected at concentrations down to 10² copies/μL, whereas detection at 10¹ copies/μL was inconsistent. The standard curves for each target pathogen showed high correlation coefficients (R² > 0.90), confirming reliable assay performance within the dynamic range tested. Bacillus atrophaeus and GAPDH were consistently detected as internal controls. In addition, the 18S rRNA quality control assay was consistently detected, with Amp scores and Cq confidence values falling within the recommended ranges. No amplification was observed in negative controls.

3.2. Prevalence analysis using TAC system

Of the 45 pathogens tested, seven were detected by the TAC assay (Table 1). Anaplasma, Borrelia, and Theileria were detected, consistent with our previous study [4]. Additional pathogens detected included Ehrlichia (68.2 %), Rickettsia (31.8 %), Bartonella (9.9 %), and Giardia (3.1 %).

Table 1.

Detection of tick-borne and other pathogens in Korean water deer by TAC assay. Pathogens previously reported in our earlier study are indicated separately.

Group	Pathogen	Positive samples (%)
Previously reported in our previous study	Anaplasma spp.	177 (92.2)
	Borrelia spp.	68 (35.4)
	Theileria spp.	183 (95.3)
Newly detected in this study	Ehrlichia spp.	131 (68.2)
	Rickettsia spp.	61 (31.8)
	Bartonella spp.	19 (9.9)
	Giardia spp.	6 (3.1)

3.3. Prevalence analysis using conventional PCR

Conventional PCR confirmed the TAC results for seven pathogens (Tables 2 and 3). Prevalence estimates were as follows: E. canis, 0.5 % (95 % CI 0–1.5 %); E. muris, 1.6 % (95 % CI 0–3.3 %); Neoehrlichia mikurensis, 4.2 % (95 % CI 1.3–7.0 %); Candidatus R. longicornii, 26.0 % (95 % CI 19.8–32.2 %); R. raoultii, 2.1 % (95 % CI 0.1–4.1 %); B. schoenbuchensis, 6.8 % (95 % CI 3.2–10.3 %); and G. duodenalis, 1.0 % (95 % CI 0–2.5 %).

Table 2.

Regional distribution of pathogens detected in Korean water deer.

Area	Region		Number tested	Number infected (%)							Total
				Ehrlichia canis	Ehrlichia muris	Neoehrlichia mikurensis	Candidatus Rickettsia longicornii	Rickettsia raoultii	Bartonella schoenbuchensis	Giardia duodenalis
Northern	Gangwon	Wonju	27	0	1 (3.7)	0	14 (51.9)	0	0	0	15 (55.6)
		Chuncheon	12	0	0	2 (16.7)	3 (25.0)	0	0	1 (8.3)	6 (50.0)
	Gyeonggi	Gapyeong	12	0	0	1 (8.3)	3 (25.0)	0	0	0	4 (33.3)
		Namyangju	13	0	0	0	9 (69.2)⁎	0	1 (7.7)	0	10 (76.9)⁎
		Yangpyeong	13	0	0	0	8 (61.5)	0	0	0	8 (61.5)
		Yeoncheon	13	0	0	1 (7.7)	6 (46.2)	0	1 (7.7)	0	8 (61.5)
		Pocheon	20	0	1 (5.0)	1 (5.0)	1 (5.0)	0	0	0	3 (15.0)
Central	Gyeongbuk	Andong	17	1 (5.9)	0	0	1 (5.9)	0	0	0	2 (11.8)
		Yeongyang	13	0	0	1 (7.7)	1 (7.7)	3 (23.1)	0	0	5 (38.5)
	Chungnam	Gongju	21	0	0	0	4 (19.0)	0	0	1 (4.8)	5 (23.8)
Southern	Gyeongnam	Jinju	14	0	0	1 (7.1)	0	0	7 (50.0)⁎	0	8 (57.1)
		Hapcheon	17	0	1 (5.9)	1 (5.9)	0	1 (5.9)	4 (23.5)	0	7 (41.2)
Total			192	1 (0.5)	3 (1.6)	8 (4.2)	50 (26.0)	4 (2.1)	13 (6.8)	2 (1.0)	81 (42.2)

^⁎Indicates statistically significant difference in prevalence among regions after Bonferroni correction for multiple comparisons (C. R. longicornii, B. schoenbuchensis, and total positives; adjusted α < 0.0042). Pathogens with fewer than 10 positive detections (e.g., R. raoultii) were reported descriptively but excluded from statistical testing.

Spatial distribution maps illustrated the geographic locations of positive individuals for each pathogen (Fig. 2). C. R. longicornii was widely distributed except in a few regions. R. raoultii was found in central and southern regions. B. schoenbuchensis was concentrated in the southern region.

Fig. 2.

Geographic distribution of Korean water deer positive for tick-borne pathogens detected by conventional PCR. Each dot is colour-coded by pathogen: Candidatus Rickettsia longicornii (red), Rickettsia raoultii (blue), Bartonella schoenbuchensis (green), Ehrlichia canis (orange), Ehrlichia muris (yellow), Neoehrlichia mikurensis (purple), and Giardia duodenalis (brown). Maps plot GPS coordinates of infected animals, highlighting regional clustering and distribution patterns. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

For regional comparisons (Table 2), pathogens with fewer than 10 positives were excluded from statistical testing. C. R. longicornii showed the highest prevalence in Namyangju (69.2 %; 95 % CI 44.1–94.3 %), and remained significant after Bonferroni correction (p < 0.001; adjusted α = 0.0042). B. schoenbuchensis prevalence was highest in Jinju (50.0 %; 95 % CI 23.8–76.2 %) and also remained significant (p < 0.001; adjusted α = 0.0042). Total positives were highest in Namyangju (76.9 %; 95 % CI 54.0–99.8 %) and also remained significant (p = 0.0017, adjusted α = 0.0042). No other pathogens differed significantly by region.

For seasonal comparisons (Table 3; Fig. 3), pathogens with fewer than 10 positives were excluded from statistical testing. Total positives peaked in April (93.3 %; 95 % CI 80.7–100 %) and remained significant after Bonferroni correction (p = 0.0004; adjusted α = 0.00625). C. R. longicornii was most frequent in April (66.7 %; 95 % CI 42.8–90.5 %) and showed a significant seasonal trend (p = 0.0005; adjusted α = 0.00625). No other pathogens showed meaningful seasonal variation. Coinfections were detected in 21 individuals, including 20 with two pathogens and one with three.

Table 3.

Seasonal distribution of pathogens detected in Korean water deer.

Group		Number tested	Number infected (%)							Total
			Ehrlichia canis	Ehrlichia muris	Neoehrlichia mikurensis	Candidatus Rickettsia longicornii	Rickettsia raoultii	Bartonella schoenbuchensis	Giardia duodenalis
Season	April	15	0	3 (20.0)	0	10 (66.7)⁎	0	1 (6.7)	0	14 (93.3)⁎
	May	7	0	0	0	4 (57.1)	0	0	0	4 (57.1)
	June	10	0	0	0	1 (10.0)	2 (20.0)	2 (20.0)	0	5 (50.0)
	July	10	0	0	0	0	1 (10.0)	0	0	1 (10.0)
	August	13	0	0	3 (23.1)	0	1 (7.7)	1 (7.7)	0	5 (38.5)
	September	40	0	0	3 (7.5)	10 (25.0)	0	5 (12.5)	2 (5.0)	20 (50.0)
	October	56	1 (1.8)	0	1 (1.8)	15 (26.8)	0	4 (7.1)	0	21 (37.5)
	November	41	0	0	1 (2.4)	10 (24.4)	0	0	0	11 (26.8)
Total		192	1 (0.5)	3 (1.6)	8 (4.2)	50 (26.0)	4 (2.1)	13 (6.8)	2 (1.0)	81 (42.2)

^⁎Indicates statistically significant difference in prevalence among months (corresponding to the month of sample collection) after Bonferroni correction for multiple comparisons (C. R. longicornii and total positives; adjusted α < 0.00625). Pathogens with fewer than 10 positive detections (e.g., E. muris and R. raoultii) were reported descriptively but excluded from statistical testing.

Fig. 3.

Monthly sampling and detection rates of seven pathogens in Korean water deer from April to November 2023.

3.4. Molecular and phylogenetic analyses

Phylogenetic analysis of 16S rRNA gene sequences from Ehrlichia grouped sequences into E. canis, E. muris, and N. mikurensis (Supplementary Fig. S1). One E. canis sequence exhibited 98.8 %–99.4 % identity with GenBank references. Three E. muris sequences showed 99.4 %–100 % identity among themselves and 99.1 %–100 % similarity to GenBank references. Six representative N. mikurensis sequences demonstrated 98.8 %–100 % identity to each other and 97.6 %–98.8 % similarity to GenBank references. Rickettsia sequences clustered into C. R. longicornii and R. raoultii based on 16S rRNA (Supplementary Fig. S2). Four representative C. R. longicornii sequences shared 99.6 %–100 % identity with one another and GenBank sequences. One R. raoultii sequence showed 100 % identity with previous reports. Bartonella sequences, based on 23S rRNA, were classified as B. schoenbuchensis (Supplementary Fig. S3). Four representative sequences showed 99.1 %–100 % identity among themselves and 97.2 %–99.4 % similarity to GenBank references. G. duodenalis was identified by the 18S rRNA gene (Supplementary Fig. S4). Two sequences were identical to each other and to GenBank references.

Supplementary Fig. S1.

Phylogenetic analysis of Ehrlichia spp. was performed using 16S rRNA gene sequences. Sequences obtained in this study are highlighted in green, with corresponding GenBank accession numbers in parentheses. Rickettsia raoultii was used as an outgroup. Bootstrap support values (1000 replicates) are shown at the nodes. The scale bar indicates evolutionary distance. The phylogeny was reconstructed using the maximum likelihood method.

Supplementary Fig. S2.

Phylogenetic analysis of Rickettsia spp. was performed using 16S rRNA gene sequences. Sequences obtained in this study are highlighted in green, with corresponding GenBank accession numbers in parentheses. Anaplasma phagocytophilum was used as an outgroup. Bootstrap support values (1000 replicates) are shown at the nodes. The scale bar indicates evolutionary distance. The phylogeny was reconstructed using the maximum likelihood method.

Supplementary Fig. S3.

Phylogenetic analysis of Bartonella spp. was performed using 23S rRNA gene sequences. Sequences obtained in this study are highlighted in green, with corresponding GenBank accession numbers in parentheses. Brucella canis was used as an outgroup. Bootstrap support values (1000 replicates) are shown at the nodes. The scale bar indicates evolutionary distance. The phylogeny was reconstructed using the maximum likelihood method.

Supplementary Fig. S4.

Phylogenetic analysis of Giardia spp. was performed using 18S rRNA gene sequences. Sequences obtained in this study are highlighted in green, with corresponding GenBank accession numbers in parentheses. An 18S rRNA gene sequence from Theileria sergenti was used as an outgroup. Bootstrap support values (1000 replicates) are shown at the nodes. The scale bar indicates evolutionary distance. The phylogeny was reconstructed using the maximum likelihood method.

4. Discussion

TAC technology has been applied to diagnose human infectious diseases such as acute febrile illness [14] and mumps [15]. In South Korea, wildlife pathogen surveillance has largely relied on PCR or serological testing of captured animals, usually targeting a limited number of pathogens in restricted regions. These approaches provide baseline data but are constrained by sensitivity, scope, and resources. In this study, we applied TAC for the first time to wildlife, representing the first global attempt to use TAC-based pathogen surveillance in KWD. By enabling simultaneous detection of multiple pathogens in a single run [16], TAC addresses key limitations of current frameworks, offering improved sensitivity, efficiency, and scalability for surveillance. Our findings demonstrate that TAC is a promising tool for high-throughput monitoring of free-ranging wildlife such as KWD, which are widely distributed and known reservoirs of zoonotic pathogens [17].

Four pathogens—Ehrlichia spp., Rickettsia spp., Bartonella spp., and Giardia spp.—were consistently confirmed by TAC and conventional PCR. Among Ehrlichia spp., E. canis is a major tick-borne zoonotic agent [18] and the causative agent of canine monocytic ehrlichiosis [19]. Human cases have been reported in abroad [19,20], while in South Korea, it has been detected in KWD [21], and their ticks [22]. In this study, only one spleen sample tested positive, indicating limited circulation, yet providing further molecular evidence of E. canis in wild cervids in South Korea. E. muris, typically associated with rodents [23], has been detected in Sika deer in Japan [24]. In South Korea, recent surveillance identified the pathogen in rodents and mites [25], indicating ongoing environmental circulation. Here, E. muris was detected in KWD across multiple regions, representing the first molecular evidence of infection in cervids within the country and highlighting an expanded host range. Neoehrlichia mikurensis is an emerging human pathogen [26], primarily associated with rodents [27]. In South Korea, it has been reported at a high prevalence in rodents (82.0 %) [28]. In Europe, it has also been detected in dogs [29] and hedgehogs [30]. In this study, N. mikurensis was identified in KWD across multiple regions, providing rare evidence of infection in cervids and expanding its known host spectrum. Taken together with the detection of other Ehrlichia species, these findings suggest that KWD may contribute to the natural ecology of TBPs, underscoring the importance of continued wildlife surveillance.

Rickettsia are important arthropod-borne pathogens with significant zoonotic potential [31]. In South Korea, C. R. longicornii is the most prevalent spotted fever group Rickettsia, with 31.2 % positivity in ticks from KWD [22]. Consistent with this, it showed the highest prevalence (26.0 %) among pathogens detected in this study, and was broadly distributed across both regions and seasons. Similar sequences have been reported from ticks in China and Japan [32], although the pathogenicity remains unclear. Given its high prevalence and broad distribution, it should be considered a priority target for future wildlife surveillance. Rickettsia raoultii, a known cause of human tick-borne lymphadenitis in Europe [33], has been reported across Asia, including ticks from Mongolia [34]. In South Korea, it has been detected in dogs [35], rodents, mites [25], and ticks parasitizing KWD [22]. In this study, we provide the first molecular evidence of R. raoultii infection in KWD, suggesting systemic infection and potential reservoir status. These findings enhance our understanding of R. raoultii ecology and raise the possibility of wildlife-to-human transmission.

Bartonella shoenbuchensis, a zoonotic Bartonella species [36], has been reported in roe deer [37] and deer keds [38]. In South Korea, B. schoenbuchensis-related species were previously detected in KWD spleens [39]. In this study, it was identified in KWD across four regions, supporting its continued circulation in cervid populations. Given its presence in ectoparasites such as deer keds and ticks, these findings highlight the need to further investigate the role of arthropod vectors in Bartonella transmission within wildlife and their potential impact on zoonotic risk.

Giardia duodenalis is a globally distributed protozoan parasite that infects the intestinal tract of humans and animals [40]. In wildlife, assemblage A has been detection in sika deer in Japan [41], while in South Korea, assemblages A and B were reported in wild boars, with genotypes identical to human strains [42], suggesting zoonotic potential. In this study, G. duodenalis assemblage A was identified in the spleen of two KWD, representing the first organ-level detection in cervids in South Korea. Both cases occurred in autumn, consistent with seasonal patterns observed in wild boars. Although Giardia is typically detected in fecal samples, our spleen-based detection suggests possible extra-intestinal dissemination. This interpretation is supported by previous histological and experimental studies demonstrating mucosal and even systemic tissue invasion under certain conditions [43,44]. These findings underscore the need for further investigation into the pathogenicity, transmission dynamics, and zoonotic implications of G. duodenalis in wild ruminants.

Spleen tissue was selected because it is a major lymphoid organ where pathogens accumulate during systemic infection, increasing detection sensitivity. Indeed, previous work showed higher pathogen prevalence in spleen than in kidney, particularly for Bartonella spp. [45]. Nonetheless, future surveillance should explore non-lethal approaches such as fecal sampling, as well as specimens from road-killed or rescued animals, to enhance sustainability. The detection of coinfections in this study also highlights the need to consider complex transmission dynamics in wildlife surveillance frameworks.

Several limitations must be acknowledged. Although over 1000 KWD samples were available, only 192 were analyzed due to resource constraints, and selection was not random but designed to ensure geographic and seasonal coverage. This may limit generalizability. Moreover, seasonal patterns should be interpreted cautiously, as more deer were captured between September and November due to hunting availability, and tick infestation was generally higher in summer–autumn. These factors may act as confounders, partly influencing apparent peaks. In addition, while TAC has been validated, sensitivity may vary by pathogen, and primer/probe sequences are proprietary. However, reproducibility is ensured as identical panels can be ordered. Preliminary testing showed reliable detection to 10² copies/μL (R² > 0.90), with inconsistent results at 10¹, indicating very low loads may be missed. Still, sensitivity was comparable to conventional qPCR, and specificity was supported by consistent confirmatory PCR. Finally, while TAC is costlier upfront than singleplex PCR, its high-throughput multiplex capacity substantially reduces per-target cost and turnaround time. Future studies should expand sample numbers and include further optimization and independent validation to strengthen its applicability in wildlife and zoonotic surveillance.

5. Conclusions

This study is the first in South Korea to apply the TAC system for wildlife disease surveillance, using KWD as a sentinel species. The platform enabled high-throughput detection of multiple zoonotic pathogens from spleen samples, including the first molecular evidence of E. muris in KWD and the first global report of organ-level G. duodenalis infection in cervids, suggesting possible extra-intestinal dissemination. The widespread detection of multiple TBPs highlights the role of KWD as a reservoir in the wildlife–tick–pathogen cycle. These findings demonstrate the diagnostic value of TAC for wildlife surveillance and emphasize its potential to strengthen early warning systems and One Health strategies in the face of emerging zoonotic threats.

The following are the supplementary data related to this article.

Supplementary Table S1

Pathogens targeted by the customized TaqMan Array Card system and synthetic positive controls.

CRediT authorship contribution statement

Beoul Kim: Writing – original draft, Conceptualization. Su-Jin Chae: Writing – original draft, Conceptualization. You-Jeong Lee: Methodology, Formal analysis. Haksub Shin: Visualization, Validation. Sunmin Kwak: Visualization, Resources. Hyesung Jeong: Software, Investigation. Suwoong Lee: Resources, Project administration. Yong-Myung Kang: Supervision, Formal analysis. Dongmi Kwak: Writing – review & editing, Supervision. Min-Goo Seo: Writing – review & editing, Funding acquisition.

Ethics statement

Between April and November 2023, spleen samples from KWD were collected by licensed hunters under the supervision of the NIWDC. Of the 1035 samples obtained in a previous nationwide survey, 192 were selected for this study. All procedures were approved by the Institutional Animal Care and Use Committee of Kyungpook National University (Approval No. KNU 2024-0407).

Funding

This study was supported by the National Institute of Wildlife Disease Control and Prevention under a research support project.

Declaration of competing interest

The authors declare that they have no conflicts of interest.

Data availability

Data supporting the conclusions of this article are included within the article. The newly generated sequences were submitted to the GenBank database under the accession numbers PV628751–PV628765, PV637156–PV637160, PV637183 and PV637184. The datasets used and/or analyzed during the present study are available from the corresponding author upon reasonable request.