Detection of Besnoitia besnoiti DNA by real-time PCR in individuals or in pools of Stomoxys calcitrans (Diptera: Muscidae), reveals parasite circulation on cattle farms

Abstract

Bovine besnoitiosis is an emerging disease in Europe, transmitted mechanically by the stable fly Stomoxys calcitrans. Infected cattle harbouring large quantities of parasitic cysts in their dermis are considered super-spreaders. The objective of this study was to demonstrate that Besnoitia besnoiti DNA can be detected in stable flies captured in the immediate vicinity of these hosts. In this case, we wanted to assess how far away from the super-spreaders we could capture infected Stomoxys flies. On two cattle farms affected by besnoitiosis in the south of France, super-spreaders were identified with skin biopsies. In spring, these cattle were kept indoors, and Vavoua traps and BioFlyTrap were deployed in their immediate vicinity and then at increasing distances up to 130 m. DNA extracted from the abdomens of captured stable flies was subjected to real-time PCR to detect B. besnoiti DNA. The apparent densities of S. calcitrans were very high in the immediate vicinity of cattle, but fell sharply as the trap was placed further away, thus demonstrating the sedentary nature of this species. Besnoitia besnoiti DNA was detected in 3 out of 42 batches of Stomoxys flies on one farm (only in the vicinity of the cattle) and in 10 out of 134 batches in the other farm (five in the vicinity of the cattle, the others up to 70 m). No infected S. calcitrans were found 130 m away from the parasite carriers. The detection of pathogen DNA in S. calcitrans, captured on farms, could serve as a tool for epidemiological surveillance of their circulation.

Graphical abstract

Highlights

• •Besnoitia besnoiti DNA is detectable in stable flies captured near super-spreader cattle.
• •Infected Stomoxys calcitrans were detected up to 70 m from super-spreader cattle.
• •Sedentary nature of stable flies was observed in this study.
• •Epidemiological surveillance of pathogens in cattle herds could be achieved with stable flies.

1. Introduction

Over the past twenty years, Western Europe has experienced the emergence of several vector-borne diseases in cattle, including bluetongue disease (Kundlacz et al., 2019), epizootic haemorrhagic disease (Gondard et al., 2024), lumpy skin disease (Yeşilbağ et al., 2026) and besnoitiosis (EFSA, 2010). The latter two diseases, lumpy skin disease and besnoitiosis, are known to be transmitted mechanically by horseflies and Stomoxys calcitrans (Baldacchino et al., 2013, 2014), two types of blood-feeding flies that also cause nuisance and direct economic losses in livestock (Taylor et al., 2012; Desquesnes et al., 2022).

Bovine besnoitiosis is a disease caused by Besnoitia besnoiti, an apicomplexan protozoan of the family Sarcocystidae, which has spread rapidly in the southern half of France since the 2000s (Jacquiet et al., 2010; Grisez et al., 2020). Long confined to the foothills of the Pyrenees, bovine besnoitiosis has spread to the Auvergne-Rhône-Alpes region (Lahondes et al., 2025). More sporadic cases are found in central and eastern France. It is also found in enzootic form in the Iberian Peninsula (Cortes et al., 2014; Esteban-Gil et al., 2017) and Italy (Neve et al., 2022) and more sporadically in central and northern European countries (Basso et al., 2013; Ryan et al., 2016; Delooz et al., 2021).

When clinically apparent, infection of cattle occurs in three phases: a febrile phase marked by epiphora, nasal discharge and anorexia, followed by a phase of oedema, and thirdly a chronic phase in which the parasite concentrates in bradyzoite cysts, located mostly in the connective tissue of the upper respiratory tract and the dermis. The third phase, also known as the chronic phase of the disease, is characterised by significant thickening of the dermis (scleroderma), which persists for at least several months or years, sometimes throughout the animal’s life. During the febrile phase, a very brief period of parasitaemia can account for parasite transmission by vectors. But in reality, the dermal location of the bradyzoite cysts explains why cattle besnoitiosis is, among other possible ways of transmission, a vector-borne parasitic disease, as horseflies and stable flies pierce the dermal cysts with their mouthparts when feeding on the blood of chronically infected cattle; and as a result, become contaminated (Bigalke, 1968). Horse flies and stable flies are persistent feeders, inflicting painful bites, leading in most cases to an interruption of blood-feeding by host defensive behaviour (Schofield and Torr, 2002). When blood-feeding is interrupted and resumed on a healthy animal, vector transmission is possible and has been experimentally demonstrated on several occasions (Bigalke, 1968; Sharif et al., 2019) using severely affected animals in the scleroderma phase of the disease as parasite donors.

This shows that vector-borne transmission is one of the possible routes of transmission for B. besnoiti. Blood-feeding flies act only as “flying syringes”, without modification or multiplication of the parasite in the vectors. Therefore, for mechanical transmission to occur, it is essential that at least one but more likely several hosts carrying a large number of parasites in their skin are located in close proximity to uninfected animals, combined with significant vector activity.

Among besnoitiosis-infected cattle, there is considerable heterogeneity in the abundance of bradyzoite cysts in the dermis. This abundance of dermal cysts can be assessed using real-time PCR analysis on skin biopsies (Schares et al., 2016; Grisez et al., 2020). While chronically infected individuals showing severe scleroderma have a very high number of cysts in their dermis (Ct-values commonly ranging from 15 to 20 in skin biopsy samples), a minority of asymptomatic animals also have parasite cysts in their dermis, as indicated by Ct-values ranging from 20 to 35 (Grisez et al., 2020). These clinically infected animals can therefore be described as super-spreaders (VanderWaal and Ezenwa, 2016), but the asymptomatic ones as well, even if they show no clinical signs of infection. In our experience, these asymptomatic animals carry cysts in their dermis for very long periods of time, months or even years, such as symptomatic cattle in the clinical chronic phase (Grisez et al., 2020). Besides, the majority of asymptomatic seropositive individuals do not test positive in skin tests and are considered to be low-level carriers or even epidemiological dead ends for the parasite.

Regarding the role of stable flies in the parasite’s transmission, B. besnoiti DNA has already been found in one specimen of S. calcitrans out of 48, caught directly on an acutely affected Simmental heifer (Gollnick et al., 2015) and in pooled (n = 10) abdominal contents of stable flies artificially fed on a chronically affected cow showing severe scleroderma (Sharif et al., 2019). However, no attempt has ever been made to detect the parasite’s DNA in Stomoxys flies feeding on asymptomatic animals. These missing data are needed to fully understand the epidemiology of besnoitiosis. A second issue concerns the active dispersal of B. besnoiti-infected stable flies, which could lead to the contamination of a parasite-free farm neighbouring the contaminated one, due to flies flying from one herd to another. In the course of blood-feeding, the active dispersal of these flies is known to be relatively limited (Taylor et al., 2010; Lempereur et al., 2018); in their observation under natural conditions, Gollnick et al. (2015) did not record seroconversion in cattle kept at 20 m from infected ones. Therefore, it would be interesting to look for parasite DNA in stable flies captured at increasing distances from an infection source consisting of a group of asymptomatic but super-spreader cattle, strictly confined to a barn.

In this context, our study had two objectives: (i) to evaluate the feasibility of B. besnoiti DNA detection in stable flies caught near asymptomatic but super-spreader cattle and, if so, (ii) to assess the maximum distance at which S. calcitrans flies infected with B. besnoiti can still be found from the strictly confined source of infection. The surveys took place on two commercial Besnoitia-infected farms in south-western France featuring super-spreaders, under field conditions of parasite transmission, in spring that is known to be the first peak of adult S. calcitrans activity (Jacquiet et al., 2014) in this region, the second one occurring in autumn.

2. Materials and methods

2.1. Experimental design

2.1.1. Study farms

Two commercial farms (A and B), located in the southwest of France (région Occitanie), were enrolled in the study. The first was studied during 2023 (Aveyron Département) and the second during 2024 (Tarn Département). The main characteristics of these farms and their history of besnoitiosis are presented in Table 1. Both farms experienced their first clinical cases of besnoitiosis following the introduction of an infected animal. Farm A is an isolated farm, whereas Farm B is located in close proximity to another farm which is free of besnoitiosis, as no clinical cases or seropositive animals have been detected there. Part of the results obtained on Farm A (2023) was used to adjust the protocol applied on Farm B (2024).

Table 1.

Characteristics of the studied farms.

	Farm A	Farm B
Production	Beef cattle	Beef cattle
Cattle breed	Limousine	Limousine
Number of adult cows	84	85
Reproduction	Artificial insemination (100%)	Natural mating (90%); artificial insemination (10%)
Probable origin of besnoitiosis	Purchase of an infected animal in 2018	Purchase of an infected animal in 2022
Date of the first clinical case on the farm	August 2019	February 2023
No. of clinical cases per year since the introduction	1–4 (variable according to year)	2–3
No. of animals testing positive in serology at the beginning of the study (date of serological analysis)	32 (February 2023)	21 (February 2024)
No. of animals testing positive in real-time PCR on skin biopsies (Mean Ct ± SD)	4 (23.3 ± 6.8)	6 (27 ± 4.5)

Abbreviations: Ct, cycle threshold; SD, standard deviation.

2.1.2. Animals

Serological testing for besnoitiosis on all animals over six months of age on the farms, as well as the detection of super-spreaders by real-time PCR on skin biopsies on seropositive individuals, were carried out during February, the month preceding the season of stable flies’ activity and their capture in the traps (February 2023 for Farm A and February 2024 for Farm B). Once diagnostics of super-spreaders were established, to avoid the spreading of the disease during the time required for their selective slaughterhouse reform (livestock sanitation strategy in accordance with the cattle breeders and veterinary practitioners), the super-spreader animals of both farms were kept in stalls, without access to pasture, for the duration of the study (23 May to 7 July 2023 in Farm A and 19 March to 10 July 2024 in Farm B). The area comprising the stalls where the PCR-positive cattle were concentrated and the immediately adjacent pasture are hereinafter referred to as ‘infection focus’.

2.1.3. Monitoring of adult stable flies’ activity and dispersion

Two types of hematophagous fly traps were used in the study, a Vavoua trap (Laveissiere and Grebaut, 1990) obtained from Vestergaard® (Lausanne, Switzerland) and two prototypes (P10 and P23) of BioFlyTraps (Fig. 1A and 2A, B), new traps under development (Desquesnes et al., 2021). Prototype P10 is a cage trap including a large internal blue panel, while P23 is an open trap including 2 blue and 2 black panels.

Fig. 1.

Trap setting in Farm A (BioFlyTraps P10 and Vavoua traps). A P10 BioFlyTraps set up in Farm A. B Field view of L2 and L3 made of Vavoua and P10. C Map view showing the infection focus and the four trap lines: L0 (1 Vavoua), L1, L2 and L3 (4 P10 and 1 Vavoua trap in each).

Fig. 2.

Trap settings in Farm B (BioFlyTraps P23 and Vavoua traps). A Field view of the 5 trap-lines (L0-L4). B Vavoua trap equipped with a collecting flask (ethanol container). C Map view showing the “infection focus” and the 5 trap lines: L0 (3 Vavoua traps and 1 P23 BioFlyTrap), L1, L2, L3 (4 P23 and 1 Vavoua trap in each) and L4 (one Vavoua trap), as well as the neighbouring farm and pasture.

In the immediate vicinity of the cattle super-spreaders housing, i.e. in the infection foci, which were named Line 0 (L0), traps were set up as follows: in Farm A, only one Vavoua trap; in Farm B, three Vavoua traps and one BioFlyTrap P23. The other traps were set up in lines, in the adjacent and unoccupied pastures, by batches of 5 traps: 1 reference Vavoua trap and 4 experimental BioFlyTraps (BioFlyTraps P10 in Farm A and P23 in Farm B). Line 1 was set up at 10 m apart from the infection focus. Lines 2 and 3 were set up at 30 m from L1 and L2, respectively (i.e. 40 and 70 m from the infection focus, respectively; Fig. 1B and C). In Farm B, an additional line (L4) was made of a single Vavoua trap set up at 60 m from L3, at the extremity of the pasture, in close vicinity to the neighbouring Besnoitia-free farm (i.e. 130 m from the infection focus; Fig. 2A–C).

In Farm A, the topography of the areas where the traps were placed was flat for Lines 1 and 2 but declining towards Line 3 (Fig. 1B); in Farm B, the topography of the areas where the traps were placed was roughly flat (Fig. 2A), and, in both cases, there were no obstacles (buildings or woods) separating the different trap lines (Fig. 1B and C and 2A, C).

In each line, the Vavoua trap and one of the BioFlyTraps were equipped with collecting flasks (see Fig. 2B) (no collection of flies was conducted in the other traps, which were only used to kill the flies). The insect capture devices consisted of one horizontal water bottle connected to a set of 2 plastic flasks sealed head-to-head, the upper one being sealed to the water bottle and the lower one, removable, containing a preservative solution. Two types of preservative solutions were tested in parallel in Farm A only: soap in water (2% vol/vol of a washing-up liquid, trade name “Paic citron”®, Colgate Palmolive company) or 50% ethanol (see Fig. 1A and 2B). Traps were checked at various time intervals ranging from 1 h to 1 week (see Table 2), according to our collection schedule, and to evaluate and determine the optimal protocol to ensure the best quality of fly DNA samples for real-time PCR. To take account of the results from Farm A in 2023, the sampling and conservation procedures were standardised in Farm B during the spring of 2024: flies were regularly collected once a week from the traps, equipped with 50% ethanol, from 27 March to 8 July, and then identified and counted.

Table 2.

Results of the real-time-PCR for Besnoitia besnoiti and cattle DNA in samples collected at L0 from Farm A according to various sampling conditions.

Type of conservation liquid	Duration of trapping period	No. of samples tested	No. of samples positive for bovine DNA	No. of Besnoitia-positive samples in real-time PCR	Type of fly batches
Soap and water	1 week	2	2	0
	3 days	6	6	0
	1 day	10	9	2	1 pool of 5 flies + 1 individual fly
	Total	18	17 (94%)	2 (11%)
50% ethanol	1 week	24	14	0
	4 days	3	3	0
	1 h	2	2	1	1 pool of 10 flies
	Total	29	19 (66%)	1 (3%)

Back in the laboratory, stable flies were identified using morphological keys (Zumpt, 1973) and sorted according to the number of flies captured in the traps. Mean apparent densities (ADT), corresponding to the mean number of stable flies captured per day and per trap, were evaluated for each line. Wherever possible, pools of 10 Stomoxys flies were prepared and studied; however, at the start of the study, the small number of Stomoxys flies caught in the traps meant that pools of 2 or 5 individuals had to be formed, and on very few occasions, that work had to be carried out on individual specimens. As soon as catch numbers became more substantial (from May to July), only pools of 10 Stomoxys flies were formed. These pools of 10 Stomoxys flies represented 26 out of 42 samples in Farm A, and 112 out of 126 samples in Farm B. Finally, the stable flies were stored in 50% ethanol and refrigerated at 4 °C until DNA extraction.

2.2. Laboratory procedures

2.2.1. Serology in cattle

Blood samples were taken by the veterinary practitioner during February 2023 (Farm A) and February 2024 (Farm B and neighbouring non-infected farm). Sera were separated by centrifugation and tested for B. besnoiti antibodies using the same commercial ELISA kit (ID Screen® Besnoitia Indirect 2.0, IDVET) for both farms. Serological analyses were performed by the Departmental Veterinary Laboratory of Aveyron (2023 study) and Tarn (2024 study).

2.2.2. Skin biopsies and real-time PCR for Besnoitia besnoiti DNA detection in seropositive cattle

Skin samples for real-time PCR were taken at the tail base using biopsy punches of 8 mm in diameter (Kruuse, Langeskov, Denmark) by the veterinary practitioner as part of the herd sanitation programme. The skin fragments were placed individually in a dry tube (Coring MCT-150-G), identified with the bovine individual ear tag and stored at 4 °C; analyses were carried out within two days after collection. DNA was extracted from tissue biopsies using a commercial kit (QIAmp® DNA Mini Kit, Qiagen, Courtaboeuf, France). Following the manufacturer’s recommendations, 50 mg aliquots of tissue biopsies were processed after an over-night incubation with proteinase K. A duplex real-time PCR was performed to detect simultaneously cattle DNA (bovine GAPDH gene) and B. besnoiti DNA (internal transcribed spacer 1 (ITS1) of ribosomal DNA). Bovine GAPDH and Besnoitia spp. ITS1 amplifications were performed with the commercial PCR kit AdiaVetTM Besnoitia (AES Chemunex, Bruz, France). The quantitative PCR was performed with the Stratagene MX3005P thermal cycler (Agilent Technologies, La Jolla, CA, USA). Positive and negative controls were provided by the manufacturer. Results were analyzed using the MxPro QPCR v.4.10 software (Agilent Technologies). When the Ct-value was inferior or equal to 36, the cattle or parasite DNA was considered detected. Animals with Ct-values inferior or equal to 36 for parasite DNA were deemed to be a super-spreaders; when Ct-values ranged between 36 and 40, parasite DNA was at the limit of detection; and a Ct-value superior to 40 was considered a negative real-time PCR result. A standard curve of B. besnoiti DNA detection was established as previously described in Sharif et al. (2017). Briefly, DNA extraction of a suspension of 10⁵ B. besnoiti culture tachyzoites was performed as described above and serially diluted 5 times 10-fold in PBS, till equivalent to DNA of a single parasite cell. The standard curve was generated by plotting the Ct-values against the log of the number of parasites.

2.2.3. Detection of cattle and Besnoitia besnoiti DNA in Stomoxys calcitrans

In Farm A (spring 2023), the insects captured in the traps were collected at various time intervals (1 h to 1 week) to explore the quality of DNA obtained from the flies after short or longer intervals. Long intervals (1 week) would be favoured due to the fact that farms were quite far from the laboratory; indeed, lowering the frequency of fly collection was one of the objectives of increasing intervals between collections. Once collected, insects were taken to the INRAE-ENVT IHAP laboratory to be identified based on morphological characteristics (Zumpt, 1973) and sorted. Based on weekly intervals between collections in Farm B, a mean apparent density (mean number of stable flies captured per day and per trap) was evaluated for each line (L0 to L4). This was done only in Farm B, as frequent torn traps in Farm A on several occasions prevented accurate estimates of apparent densities.

The stable flies captured were either treated individually or in pools of 2, 5 or 10 flies. Stable flies that were too damaged in the traps were not included in the study. Dissection of the insects was carried out using forceps, cleaned by dipping successively in (i) distilled water, (ii) bleach and (iii) distilled water again between each sample or batch of samples. The abdomens of the stable flies in each batch were separated from the rest of the body and then subjected to DNA extraction individually or in pools of 2, 5 or 10 abdomens. Extraction and detection of B. besnoiti and cattle DNA were performed as described above.

2.3. Statistical analyses

Statistical analyses of entomological data focused on Farm B, as Farm A underwent several losses of data during the sampling seasons. The effectiveness (attractiveness) of the trap types, the variation in S. calcitrans density per month, and per lines of collection was assessed using a Generalised Linear Mixed Model (GLMM) with a negative binomial error distribution, using the total number of insects captured as the response variable. Trap type, month and line were included as fixed effects without interaction terms, while an offset of log (number of trapping days) was added to take into account the trapping effort. Trapping date was included as a random effect to account for sampling taken on the same day from multiple lines/traps (temporal pseudo-replication).

The significance of the fixed effects (trap type, month and line) was evaluated using a Wald chi-square test (Type II). Pairwise comparisons between months and lines were performed using estimated marginal means (emmeans package) with Tukey adjustment for multiple comparisons. As trap type comprised only two levels, a single pairwise comparison was performed without correction (Sidak adjustment, equivalent to an unadjusted test in this two-level case).

The proportion of positive abdomen pools for cattle and B. besnoiti DNA in stable flies was compared between farms using a two-sample test for equality of proportions with continuity correction (prop.test, R base stats). R software (v.4.5.3) was used for all statistical analyses (R Core Team, 2026).

3. Results

3.1. Detection of Besnoitia besnoiti-specific antibodies and parasite DNA in cattle skin biopsies

Serological analysis showed a total of 32 out of 84 and 21 out of 85 seropositive cattle, respectively, in farms A and B. As a consequence, all seropositive cattle underwent a skin biopsy at the base of the tail using the method described by Grisez et al. (2020). Four out of 32 (Farm A) and six out of 21 (Farm B) seropositive individuals tested positive in real-time PCR (mean Ct-values of 23.3 and 27, respectively) on the skin fragments. These animals were therefore considered super-spreaders. The 28 and 15 remaining seropositive animals showing negative results in real-time PCR were placed in dedicated pastures, 300 m from the super-spreader cattle during the whole experiment and did not take part in the trial any longer. All animals belonging to the neighbouring farm of Farm B tested negative in besnoitiosis serology at the beginning and at the end of the experiment, which suggests that there was no transmission to the animals on this farm from the super-spreaders of Farm B.

3.2. Captures and collection of stable flies in the traps

In Farm A, due to harsh meteorological conditions (strong wind and heavy rains), establishing reliable fly density with apparent densities per trap was not possible (traps fallen down or collection flask filled with rainwater). Therefore, it was decided to first focus on the flies collected in the L0 Vavoua trap, close to super-spreader cattle, for Besnoitia DNA detection using various conditions (soap/ethanol) and time periods. Flies from the other lines were submitted to real-time PCR only when the fly collection was possible. During the study, the time period and the conservation liquid were progressively adjusted. Rotten material was often obtained when using water and soap; this material was discarded, and only a small number of water and soap-collected flies could be submitted to real-time PCR. For this reason, water and soap in initial attempts were later replaced by 50% ethanol. In both cases, the initial time period of 1 week was reduced to 4, 3 and 1 day and even 1 h, to increase the probability of obtaining positive results (Table 2).

In farm B, collections were done once a week from 27 March to 8 July. Apparent densities per trap (ADT) were estimated for S. calcitrans only; tabanids, which were very occasionally found, and sucking flies are not reported here. The mean S. calcitrans ADT are presented in Fig. 3; ADT showed the highest values in line L0 (18.2 for P23 and 28.1 for Vavoua trap) and L4 (a single Vavoua trap, 22.8), likely in relation to the vicinity of the neighbouring Besnoitia-free farm. Pairwise comparisons of estimated marginal means showed that L0 and L4 did not differ significantly from each other (Wald z-test, z-ratio = 1.89, P = 0.324), but both were significantly higher than L1-L3 (Wald z-test, z-ratio = 9.34, 9.94, 9.06, respectively, and z-ratio = 5.49, 6.08, 5.32, respectively, all P < 0.0001), which exhibited mean values ranging from 0.3 to 2.7. ADT in Vavoua and P23 traps ranged from 5 to 51.1 across the sampling period (raw monthly means: April: 5.6 vs 10.1; May: 20.7 vs 24.3; June: 51.1 vs 20.4; July: 20.4 vs 5.0, for Vavoua and P23, respectively). A trap type × month interaction was tested but not retained (Likelihood ratio test, χ² = 6.55, df = 3, P = 0.088; ΔAIC = 0.6 in favour of the additive model), indicating no statistical evidence that the relative performance of the two trap types differed across months, despite some visual variation in the raw monthly means, likely influenced by the limited sample size in July (n = 2 for P23). Averaged across the whole sampling period, Vavoua trap caught significantly more flies than P23 (Wald z-test, z-ratio = 2.16, P < 0.031), with a mean ADT of 17.1 and 4.9, respectively (Fig. 4). Considering L0 (infection focus traps), the ADT evolved over time; initially low in April (mean ADT 7.9), it increased in May (20.7) to reach a peak in June (37.6) and decreased in July (12.7), with only a significant difference between April and June (Wald z-test, z-ratio = 4.404, P < 0.001).

Fig. 3.

Mean apparent density per trap per day (ADT) of Stomoxys calcitrans using Vavoua trap and BioFlyTraps P23 in 5 trapping lines of Farm B, between 27th March and 8th July 2024.

Fig. 4.

Mean monthly apparent density per trap (ADT) of Stomoxys calcitrans on Farm B from April to July 2024 in Lines 0 to 4, with Vavoua traps (red colour for L0, and decreasing red colour density from L1 to L3, and purple colour for L4), and BioFlytraps (blue colour for L0 and decreasing blue colour density from L1 to L3). Error bars represent 95% confidence intervals.

Overall, high catches were recorded in the vicinity of the farm under investigation, L0 (mean ADT 20.2) and in the vicinity of the neighbouring Besnoitia-free farm, L4 (mean ADT 23.3), while pasture area always exhibited low ADT (mean ADT 1.4).

3.3. Detection of cattle- and Besnoitia besnoiti DNA in stable flies

In Farm A, a total of 42 batches of S. calcitrans from different time-trapping intervals, corresponding to 291 individuals, were assembled, including 26 batches from L0 (62% of batches; Table 3). Of all batches, 32 tested positive for bovine DNA, but only 3, located in L0, tested positive for B. besnoiti DNA (Table 3). The mean and SD of Ct-values for parasite DNA in the pooled abdomens of flies were 33 ± 3.2. In non-rotten samples collected in water and soap, the quality of bovine DNA was almost always detected regardless of the length of the trapping period (1–7 days), while in 50% ethanol (never rotten), bovine DNA was poorly detected with a 1-week trapping period (i.e. 58% positive samples); however, it was always positive for shorter periods (4 days and 1 h lead to 100% positive samples) (Table 2). Concerning Besnoitia DNA detection, only samples collected within short trapping periods were positive (1 day with water and soap, and 1 h with ethanol). Regarding the type of sample, all types exhibited a Besnoitia-positive result in real-time PCR: 1 fly, as well as pools of 5 and 10 flies.

Table 3.

Results of the real-time PCR for Besnoitia besnoiti and cattle DNA in samples from Farm A and B along various trapping lines.

Farm	Stable fly batches	Trap line					Total (Mean Ct ± SD)
		L0	L1 (10 m)	L2 (40 m)	L3 (70 m)	L4 (130 m)
A	Total no. of examined fly batches	26	6	4	6	–	42
	No. of fly batches testing positive for cattle DNA	22	5	1	4	–	32 (31.2 ± 3.9)
	No. of fly batches testing positive for Besnoitia DNA	3	0	0	0	–	3 (33.0 ± 3.2)
B	Total no. of examined fly batches	83	15	12	12	12	134
	No. of fly batches testing positive for cattle DNA	23	3	1	2	1	30 (33.3 ± 2.9)
	No. of fly batches testing positive for Besnoitia DNA	5	2	2	1	0	10 (38.3 ± 0.7)

Abbreviations: Ct, cycle threshold; SD, standard deviation.

In Farm B, a larger number of stable fly batches (n = 134), corresponding to 1187 individuals, were obtained, including 83 in L0 (62% of batches; Table 3). All samples were collected in 50% ethanol, and the insects were collected after trapping periods of 6–7 days. Of all batches, 33 tested positive for bovine DNA (25%) and 10 for B. besnoiti DNA (7.46%). These 10 batches were spread across L0 (5 batches, i.e. 3.73%), L1 and L2 (2 batches each, i.e. 1.49%) and L3 (1 positive batch, i.e. 0.74%). All batches of flies captured in L4 (n = 12) were negative for B. besnoiti DNA (Table 3). Eight positive batches came from pools of 10 flies, while 1 came from a pool of 5 flies, and the last one came from a pool of 2 flies.

Regarding the two farms, the proportion of positive abdomen pools for cattle DNA was significantly higher on Farm A (76.2%, 32/42) than on Farm B (22.4%, 30/134) (Chi-square test for equality of proportions; χ² = 38.24, df = 1, P < 0.0001; difference = 53.8 percentage points, 95% CI: 37.6–70.1). For B. besnoiti DNA, the proportion of positive abdomen pools was not significantly different between Farm A (7.1%, 3/42) and Farm B (7.4%, 10/134) (Chi-square test for equality of proportions; χ² < 0.001, df = 1, P = 1; difference = −0.3 percentage points, 95% CI: −9.6–9.0). In the two farms, circulation of Besnoitia DNA in S. calcitrans was demonstrated in the vicinity of asymptomatic and super-spreader cattle (L0) and occasionally as far as 70 m from the infection focus. No positive results were obtained at 130 m from the super-spreader focus (all negative results in insects from L4 in Farm B, located at 130 m from the infection focus in a total of 106 flies mainly tested in pools of 10).

4. Discussion

The present study demonstrates that B. besnoiti DNA is detectable in S. calcitrans stable flies captured near asymptomatic and super-spreader cattle, with infected flies recovered at distances of up to 70 m from the infection source.

The first stage of this study was to identify super-spreader animals on the two farms using skin biopsies. The real-time PCR analysis detected individuals considered to be super-spreaders on each farm: 4 out of 32 seropositive cattle on Farm A (12.5%) and 6 out of 21 cattle on Farm B (28%). These proportions are comparable to those found in the previous study by Grisez et al. (2020) in southwestern France. Moreover, the mean cycle threshold (Ct) values of Besnoitia real-time PCR in this study were within the range usually found in asymptomatic cattle carrying bradyzoite cysts in their dermis (Grisez et al., 2020), namely 23 on Farm A (representing a DNA equivalent of 10⁵ parasites in 50 mg of dermis used for DNA extraction) and 27 on Farm B (DNA equivalent of 10⁴ parasites) (Sharif et al., 2017). Consequently, the stable flies of these two farms had the opportunity to feed on infected cattle carrying cysts in their skin.

Although the GAPDH system is not species-specific (it can detect DNA from both cattle, sheep, goat, deer etc.; as indicated in the technical notes of PCR kit AdiaVetTM Besnoitia; https://www.biox.com/fr/adi391-50-adiavet-btv-type-1-real-time-50r-p-398/), considering the topography and limited natural environment of these farms, it is highly probable that mammal DNA detected comes mainly from the cattle.

Therefore, it can be assumed that DNA from cattle and from B. besnoiti could be found in the abdomen of captured flies that took a blood meal on these animals. The choice of the abdomen for the B. besnoiti DNA detection is explained by the greater abundance of parasites in this anatomical compartment of the fly compared to the mouthparts, as the entire blood meal accumulates in the crop and is then slowly directed towards the fly’s midgut (Sharif et al., 2017). The digestive enzymes in the midgut probably degrade the parasites and their DNA, which may both reduce the proportion of positive pools and increase the Ct-values obtained in real-time PCR. Indeed, the living parasite is no longer found in the abdominal contents more than 3 h after feeding on contaminated blood (Bigalke, 1968) and B. besnoiti DNA is no longer detected beyond a period of 96 h between a blood meal on an animal with severe scleroderma and real-time PCR analysis of the abdominal contents in living flies (Sharif et al., 2017).

It was only possible to assess the dynamics of the ADTs on Farm B correctly, for the reasons mentioned above. The highest numbers of S. calcitrans were caught in the immediate vicinity of cattle kept in stalls (L0), regardless of the capture period or trap type. As soon as we moved away from the animals (Lines 1, 2 and 3), the number of stable flies caught in the traps decreased significantly. However, capture levels were higher again in Line 4, i.e. near a paddock where cattle from the neighbouring farm grazed. This observation highlights the sedentary nature of stable flies, already mentioned by Bailey et al. (1973) and Rochon et al. (2021), which do not tend to stray far from areas where they find both egg-laying sites, resting places and cattle for their blood meals. However, in the specific context of this study, the active dispersal of S. calcitrans was therefore artificially reduced because the super-spreaders do not have access to pasture. Thus, these results cannot be extrapolated to other livestock farming contexts; for example, in South Ontario, Canada, in spring, recolonisation of farms from refuge farms was observed (Beresford and Sutcliffe, 2009). Overall, on Farm B, the numbers of S. calcitrans captured per trap and per day were similar to those observed in the study of the seasonal dynamics of S. calcitrans populations on the campus of the National Veterinary School of Toulouse (Jacquiet et al., 2014). Greater numbers of flies were trapped in the Vavoua traps and BioFlyTraps prototypes in Line 0; in Lines L1, L2 and L3, low numbers of flies were captured, indicating a very limited spreading of the flies from the sedentary cattle group.

A total of 42 batches of flies were analyzed on Farm A and 134 on Farm B. Most of these batches (62%) came from line L0 on both farms. It is on this Line 0 that we would expect to find a higher proportion of stable fly pools in which cattle and B. besnoiti DNA would be detected.

Concerning the detection of cattle DNA in captured stable flies, the results differed significantly between the two farms: on Farm A, 76% of the abdomen pools were positive with average Ct-values of 31.2 ± 4, compared with only 22% on Farm B with average Ct-values of 33.3 ± 3. This difference is difficult to explain but could be the result of a greater DNA degradation in flies collected on Farm B, in link with a longer ethanol conservation for a duration of one week. Indeed, the lower bovine DNA detection was observed in ethanol versus water and soap in the comparative study carried out in Farm A, and better bovine DNA recovery was obtained with shorter trapping periods. This should have led to the decision to preserve the samples in a water-soap solution, but due to the huge number of rotten samples, the decision was ultimately made to preserve the flies in 50% ethanol in the study of Farm B.

On Farm A, only pools of stable flies captured in Line 0 tested positive for B. besnoiti DNA in real-time PCR, with average Ct-values of 33 ± 3.2, corresponding to a DNA equivalent of 100 parasites. This result was obtained from a pool of 10 abdomens, which means that the possibility of having a highly infected fly among the 10 cannot be ruled out.

Farm B had the highest proportion of real-time PCR B. besnoiti-positive S. calcitrans pools (5/10) in Line 0, but parasite DNA was still found in fly pools collected in Lines 1, 2 and 3. However, the Ct-values observed on Farm B were very high, corresponding to less than 10 parasite DNA equivalents. When collecting the flies once a week, it is not known how long (from some minutes to several days) the flies stayed in the dry part of the trapping device (water plastic bottle); indeed, some flies were still walking inside, and some were already dead and drying in this part of the device. Besides, it is also not known for how long time (from some minutes to several days) the flies found in the ethanol flask stayed in it. Therefore, the quality of fly conservation may be quite variable from a fly to another. Alternative collection systems should also be evaluated, such as molecular xenomonitoring based on the analysis of excreta from live collected specimens (L’Ambert et al., 2023). This approach has been successfully used in mosquito and virus surveillance (Bigeard et al., 2024), and a proof of concept was also demonstrated in tsetse flies and trypanosomes (Pilotte et al., 2020). It could be adapted to Stomoxys flies and tabanids for other pathogens. In terms of distance, Besnoitia DNA-positive flies were found at decreasing rates from the vicinity of the infection focus (50%), to 10 m (20%), 40 m (20%) and 70 m (10%), while none was detected at 130 m from the infection focus. This is in alignment with the security distance generally recommended to avoid transmission of pathogens by mechanical vectors (Barros and Foil, 2007).

This study has some limitations linked with field conditions, where some parameters are out of control. Amongst them are: (i) meteorological conditions and their impact on DNA conservation, which evolved during the study; (ii) farms being far from the laboratory; the irregular frequency of farm visits had consequences on the state of fly conservation (periods of time during which flies stay in ethanol or water and soap); (iii) unknown status of the flies trapped in link with random parasite loading from the super-spreaders, etc.

For these reasons, short periods of time (1 h) will be applied to fly trapping in future field studies using 50% ethanol; however, other types of conservation solutions could be evaluated in future studies in order to extend the time interval of farm visits from 1 h to several days, which would be more convenient for veterinary services in a surveillance context. Batches of 5 flies should be favoured to avoid a strong dilution effect, and finally, small groups of animals placed under a space-limited area located at least 200 m apart from other animals should be explored to ensure catching the exact picture of the group under study.

5. Conclusions

The first important finding of this study is the possibility of B. besnoiti DNA detection in stable flies captured on a B. besnoiti-contaminated farm, at least in the immediate vicinity of asymptomatic and super-spreader cattle. This highlights the importance of these individuals in parasite transmission within a herd. The second important finding is that stable flies show a high degree of sedentarity around hosts, sources of blood meals. Indeed, given the feeding habits of S. calcitrans, characterised by frequent interrupted blood meals, this result suggests transmission over short distances around super-spreaders. The implications for control are significant, as rapid and selective elimination of these super-spreaders makes sense, as does the physical separation between infected and non-infected cattle, for example in different pastures at a minimum distance of 200 to 300 m. Furthermore, this study provides a proof of concept for the possible epidemiological surveillance of the presence of pathogens in cattle herds through xenodiagnosis of infection, applicable to besnoitiosis but possibly also to other pathogens. Trapping and using S. calcitrans and other biting flies to implement epidemiological surveillance would be a way to investigate at low cost a number of circulating pathogens, especially but maybe not exclusively those mechanically transmitted.

Ethical approval

This study was conducted on two commercial cattle farms with owners’ consent. Blood samples for ELISA testing and skin biopsies to identify super-spreaders were taken as part of a herd culling programme led by the practising veterinarian, rather than as part of an experimental study.

CRediT authorship contribution statement

Marc Desquesnes: Conceptualization, Methodology, Writing - review & editing. Geoffrey Gimonneau: Conceptualization, Methodology, Writing - review & editing. Philippe Jacquiet: Conceptualization, Methodology, Writing - review & editing. Christophe Lacz: Conceptualization, Methodology, Writing - review & editing. Mariana Ribeiro Sato: Investigation, Data curation, Writing - original draft, Writing - review & editing. Théotime Guiffray: Investigation, Data curation, Writing - original draft, Writing - review & editing. Leïla Graba: Investigation, Data curation, Writing - original draft, Writing - review & editing. Christelle Grisez: Investigation, Data curation, Writing - original draft, Writing - review & editing. Fany Trotier: Investigation, Data curation, Writing - original draft, Writing - review & editing. Julie Petermann: Visualization, Investigation, Writing - review & editing.

Funding

This study has been funded by Groupement de Défense Sanitaire, France.

Declaration of competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this study.

Data availability

The data supporting the conclusions of this article are included within the article.