Seroprevalence and risk factors for Trypanosoma evansi, the causative agent of surra, in the dromedary camel (Camelus dromedarius) population in Southeastern Algeria

Surra, caused by Trypanosoma evansi, is a re-emerging animal trypanosomosis, which is of special concern for camel-rearing regions of Africa and Asia. Surra decreases milk yield, lessens animal body condition score and reduces market value of exported animals resulting in substantial economic losses. A cross-sectional seroprevalence study of dromedary camels was conducted in Algeria, and major risk factors associated with infection were identified by collecting data on animal characteristics and herd management practices. The seroprevalence of T. evansi infection was determined in sera of 865 camels from 82 herds located in eastern Algeria using an antibody test (card agglutination test for Trypanosomiasis – CATT/T. evansi). Individual and herd seroprevalence were 49.5% and 73.2%, respectively, indicating substantial exposure of camels to T. evansi in the four districts studied. Five significant risk factors for T. evansi hemoparasite infection were identified: geographical area, herd size, husbandry system, accessibility to natural water sources and type of watering. There was no association between breed, sex or age with T. evansi infection. Results of this study provide baseline information that will be useful for launching control programmes in the region and potentially elsewhere.


Introduction
Trypanosomiasis is a significant animal and human health problem worldwide. Human African trypanosomiasis (sleeping sickness) is caused by infestation with the flagellate protozoan Trypanosoma brucei, whereas animal trypanosomiasis is caused by different species and subspecies of Trypanosoma (T. congolense, T. vivax, T. brucei brucei and T. brucei evansi). Amongst all of the trypanosomiases, 'surra' (caused by Trypanosoma brucei evansi) has the widest host range and the broadest geographical distribution. It is transmitted mechanically by the bite of infected tabanid flies; Family: Tabanidae (Desquesnes et al. 2009); Genus: Tabanus, Muscidae stomoxys, Haematopota, Lyperosia spp. and Chrysops spp. (Luckins 1998), or infected vampire bats (Stoco et al. 2017). Transmission occasionally occurs iatrogenically through herd identification and vaccination procedures (Gutierrez et al. 2005), and oral transmission was confirmed recently in a mouse model (Mandal et al. 2017).
http://www.ojvr.org Open Access limited geographical area with a limited number of samples (Benfodil et al. 2020;Bennoune et al. 2013;Boushaki et al. 2019); those studies concluded only that the hemoparasite can be detected in blood samples. Limited data are available currently concerning the epidemiology and distribution of the T. evansi parasite in Algeria. To gauge levels of endemic stability, it is important to assess the seroprevalence of T. evansi and to characterise local variations in risk factors.
To help fill the scientific knowledge gap regarding this important disease, the present study was conducted with the objectives of determining the seroprevalence of T. evansi and investigating potential risk factors for infection with T. evansi in the one-humped camel (Camelus dromedarius) population in eastern Algeria.

Study area
This study was carried out in four districts (Biskra, El-Oued, Ouargla and Ghardaia) in eastern Algeria. These provinces are located at 00204-00735 E and 28°32-34°56 N ( Figure 1). This region is considered as one of the most significant camel-rearing areas in Algeria, and it is a region where camel milk is increasingly commercialised and consumed. The climate of this region is arid and is characterised by long, hot summers and short, mild winters.

Data collection
Data were collected using a paper-based questionnaire given to herd owners by trained investigators. The study questionnaire focused on 23 potential explanatory variables grouped by: (1)

Sample size determination
The minimal sample size required for this study was determined using the random sampling method (Thrusfield 2018), considering an expected prevalence of 35% (based on unpublished survey data collected in 2008 in seven districts in Algeria) with 5% precision at the 95% confidence level. The formula used for calculation of sample size was: where N = number of sample size, P = expected prevalence and d 2 = absolute precision.
Although the minimum sample size calculated for this study was 350 animals, we increased the number of samples collected to improve the degree of accuracy and to account for some potential sample losses.

Sample collection and processing
A total of 865 blood samples were collected in a randomised manner. A total of 10 mL of blood was collected by jugular venipuncture. Blood samples remained at room temperature for 3-5 h to allow clotting, followed by centrifugation for 15 min at 3000 revolution per minute (rpm). The serum fraction was extracted with a micropipette, and serum samples were frozen at −20 °C until analysis.

Ethical consideration
Animal sampling was implemented with the approval of the national veterinary authorities and was exclusively done by veterinarians; the animals were blood-sampled without suffering and were subsequently released. Farmers in each zone gave verbal consent to participate in our study and gave permission for the blood sample collection from camels on their property. We followed the European guidelines (European directives EU 86/609-STE123 and 2010/63/EU) for animal handling. Results of the univariate analysis (Chi-square) revealed that individual animal seroprevalence for T. evansi was affected by geographical area, herd size, husbandry system, accessibility to natural water sources and type of watering (Table 1). Contrarily, no differences (p > 0.05) were demonstrated in trypanosomosis seroprevalence because of genetic sex, breed or animal age.

Individual level
There was a very strong association between T. evansi seroprevalence and geographical area (χ 2 = 4.80, p < 0.0001). The highest infection rate (62.5%, 125/200) was found in Ghardaia (northern part of the Sahara region), whereas the district of Ouargla had the lowest infection rate (27.5%, 55/200). The analysis showed that camels raised in large herds had a higher (p < 0.007) incidence of T. evansi infection compared with those raised in small herds.
Regarding the multivariate logistic regression analyses (Table 2), three variables remained in the final model: (1) herd size, (2) husbandry system and (3) water source. The odds of T. evansi infection were 2.2 times higher amongst large camel farms (p = 0.007, odds ratio [OR] = 2.159; 95% CI = 1262-3693) compared with small herds. Herds with more than 50 camels were 2.1 times more likely to have a seropositive animal than herds with fewer than 10 head. The presence of natural sources of water in nearby grazing areas was associated with an increased risk of T. evansi infection (OR = 9.415; 95% CI = 1.256-70, 583). Animals in a semi-intensive management system were more likely to be infected (OR = 2.95; 95% CI: 1.878-4.645) than extensively managed animals.

Herd level
The univariate herd level analysis revealed five factors associated with T. evansi seropositivity (Table 3): geographical location (p = 0.004), husbandry system (p = 0.038), history of abortion (p = 0.005), herd size (p = 0.010), and nearby natural water sources (p < 0.05; Table 3). Trypanosoma evansi seroprevalence was higher (p < 0.001) in Ghardaia, Biskra and El-Oued than in Ouargla. In contrast, no association was found between T. evansi seropositivity of the herd and contact with other camel herds, husbandry system (nomadic) or introduction of purchased camels (Table 3).
The multivariate logistic regression model showed that the main risk factor associated with T. evansi infection in camel herds was geographical location (  (Table 4).

Discussion
To our knowledge, few studies have been performed in Algeria aimed at assessing risk factors associated with T. evansi infection of dromedary camels. Our results, which document high seroprevalence rates at the individual and herd levels, could ultimately guide the design and implementation of enhanced surveillance programmes, control measures and prevention strategies for T. evansi not only in eastern Algeria but throughout major camel producing regions of the world.
In Algeria, T. evansi infection in camels was confirmed for the first time more than a century ago when parasitological examinations discovered that 10% of 282 camels were infected (Sergent & Sergent 1905). Outbreaks of trypanosomosis in dromedary herds associated with mortalities and abortions have been documented ). The T. evansi seropositivity rate in our study (49.5%) was higher than previous serological studies conducted in Algeria by Benfodil et al. (2020) and Boushaki et al. (2019), who reported seroprevalence up to 32.4%, and 9.9%, respectively. The higher seroprevalence in the present study could be because of the differences in the topography of the study areas, age of tested animals, proportion of males versus females, sample size, environmental conditions during the sampling period, or differences in diagnostic methods.   (Tehseen et al. 2015). One study in Somalia reported a higher T. evansi seroprevalence rate (56.4%) than that found in our study (Baumann & Zessin 1992   Our study documents a high prevalence of this diseasecausing parasite and corroborates field observations of substantial waves of mortality and abortion reported by camel farmers and veterinarians in the region. The high herdlevel seroprevalence supports field reports of marked financial impacts of this disease and indicates the need to develop field-practical control and prevention programmes. Although few studies have addressed herd-level seroprevalence, our seropositivity result was much higher than the 42.3% reported earlier in Algeria (Boushaki 2007). Given the span of time between studies, our study highlights the continued circulation of the parasite within herds in the region as well as its endemicity. In uncontrolled situations, herd-level seroprevalence could reach as high as 94.9% (Delafosse & Doutoum 2004). In fact, two important risk factors identified at the individual level in the former study (contact between herds and introduction of new camels into the herd) support that infection can easily spread from one herd to the other.
No effect of camel age on T. evansi seropositivity was observed in our study. Our finding agrees with some surveys Pathak & Khanna 1995;Shah et al. 2004), but differs from those of other research groups who reported increased seroprevalence with increasing age (Atarhouch et al. 2003;Bogale et al. 2012;Dia et al. 1997;Eshetu, Desta & Amare 2013;Gutierrez et al. 2000;Mirshekar, Yakhchali & Shariati-Sharifi 2017;Olani et al. 2016;Tehseen et al. 2015). In contrast to the above findings, one study in Ethiopia revealed higher infection rates in younger camels than in older adult camels (Lemecha, Lidetu & Hussein 2008). These conflicting results suggest that future investigations need to be designed to specifically test the hypothesis of age effects on T. evansi seropositivity.
The risk of contracting trypanosomosis in semi-intensive and intensive systems was approximately three and two times higher, respectively, than in extensive management systems in our study. This result agrees with previous reports that husbandry systems and practices have a great impact on circulation of T. evansi (Dia et al. 1997). Serological surveys conducted in Sudan detected a higher prevalence of T. evansi in nomadic camels than in camels managed in an agropastoral system (Elamin, El Bashir & Saeed 1998). Camel husbandry systems are frequently characterised by a remarkably high level of animal mobility and exchange between farms and areas, and even between countries (e.g., Algeria, Tunisia, Libya, Mali, and Niger). These highly common movements (controlled or uncontrolled) of camels could explain the wide distribution of T. evansi infection.
An association between camel seropositivity for T. evansi and herd history of abortion was observed in our study. This finding is in good agreement with a recent epidemiological study where abortion was associated with T. evansi seropositivity ). This result was not unexpected, as abortions induced by T. evansi have been reported in other species. Unfortunately, data concerning the incidence of abortions and other productivity measures are frequently lacking in camel-dependent livestock production settings.
Herd T. evansi seropositivity rates varied greatly between geographical areas. The highest risk of camel trypanosomosis was observed in Ghardaia (a region with relatively cold climate). Ghardaia and Biskra, situated in the northern Algerian Sahara, have an abundance of naturally occurring water sources that are associated with higher rates of T. evansi seropositivity. Although our study area belongs to the arid climate stage, it remains rich in natural water resources, with several rivers as well as temporary lakes and wadis appearing during the rainy season. We interpret our results to mean that these natural water sources likely enhance the spread of trypanosomosis through the development and proliferation of tabanids, which are indiscriminate in the transmission of T. evansi. Because these vectors seem subservient to standing water (ponds and lakes), regions with wetlands and saline lakes can be identified as higher risk areas as reported recently in a study in seven others provinces of southern Algeria ).
Worldwide, several authors have reported considerable differences in the seroprevalence of T. evansi in different geographical areas within the same country (Aregawi et al. 2015;Bogale et al., 2012;Delafosse & Doutoum 2004;Dia et al. 1997;Mirshekar et al. 2017;Olani et al. 2016;Salah, Robertson & Mohamed 2019;Tehseen et al. 2015). In most cases, these within country differences can be explained by the observed variation in ecological factors (e.g., pastoral camel densities, hottest, most humid and wettest districts in the region) that affect vector densities and therefore infection prevalence (Hagos et al. 2009). Interestingly, geographical variation was not considered in some recent animal-level studies of Trypanosoma seroprevalence (Ghaemi, Zavarib & Jannati Pirouz 2019;Hassan-Kadle et al. 2019;Salah et al. 2019). The geographical distribution of the parasite and vectors, and thus the occurrence of trypanosomosis, is largely dependent on the nature of the agro-ecological environment. Trypanosomosis, therefore, is regarded as endemic in certain regions (Holt et al. 2016). The climate is thought to permit and promote high densities of vectors such as the Tabanides fly (James et al. 1985).
Seropositivity for T. evansi infection was not associated with breed in our study. This result is in contrast with the studies that identified breed as a risk factor for T. evansi infection (e.g., in Chad; Delafosse & Doutoum 2004). It seems likely, however, that the confounding effects of breed and hair colour make interpretation of any potential breed effect difficult. Animals with dark hair ('Zerga') were three times more likely to be infected than camels with other colours of hair coat . Similarly, white-coloured camels in Saudi Arabia were more likely to be infected with T. evansi than animals with a dark coat (El-Wathig et al., 2016). Further complicating the interpretation of potential breed effects is the notion that various colour coat phenotypes may belong to the same genotypes (Almathen, Mwaracharo & Hanotte 2012;Cherifi et al. 2017). It should be noted that some other studies showed that camels were equally susceptible to Trypanosoma infection regardless of breed (Pathak & Khanna 1995).
No difference in T. evansi seropositivity was observed between male and female camels in our study. There are conflicting reports in the literature, however, regarding the effect of sex. Some studies reported that Trypanosoma infection rate was higher for females than males (Dia et al. 1997;Shah et al. 2004;Sobhy et al. 2017). Contrarily, one study in Mali (Ndoutamia et al. 1999) found higher rates of infection amongst males -a finding supported by a Kenyan study where Trypanosoma infection rates were 2.6 times higher in males than females (Njiru et al. 2004). These conflicting results may be explained by interactions with other factors, disproportionate sampling of males versus females or differential application of husbandry practices to male versus female camels. In many countries, male camels are housed in a pen or tied when females go to the rangelands for grazing. This differential management approach may explain apparent sex differences rather than there being an actual biological predisposition to infection linked to camel sex.
Serologic testing is an important method for detecting parasitic infections. The direct card agglutination test (CATT test) is a validated test that is considered a rapid tool of choice for detection of T. evansi infection in camels (Dia et al. 1997;Ngaira et al. 2003), buffalos (Davison et al. 1999) and cattle (Reid & Copeman 2003). The CATT test has been the test of choice for field investigations, despite the low sensitivity reported in some studies: 65.5% (Ngaira et al. 2003) and 68.6% (Njiru et al. 2004). The T. evansi direct card agglutination test (CATT) uses latex agglutination of beads coated with recombinant antigens RoTat 1.2 VSG (Manual 2012). Moreover, other reports (Benfodil et al. 2020;Diall et al. 1994;Gutierrez et al. 2000;Pathak et al. 1997) have shown a moderate to strong agreement between CATT and other diagnostic methods such as trypanosome lysis (TL), WCL ELISA and RoTat 1.2 type A ELISA. Recent validation studies on serologic tests for T. evansi infection in camels and other ruminants showed the degree of concordance was highest between TL and T. evansi CATT (k = 0.784), thus confirming the interest of using T. evansi CATT for serodiagnosis of surra in the field (Benfodil et al. 2020).
Because of its mechanical means of transmission, T. evansi may be disseminated through a wide range of vectors. The multitude of potential vectors allowed Trypanosoma to become the most widespread pathogens in tropical and subtropical areas (Hoare 1972). Although T. evansi is not pathogenic to humans, human infestation by this parasite was first reported in India (Joshi et al. 2005). A follow-up study (Shegokar et al. 2006) reported a 22.7% T. evansi seroprevalence in humans in India. This finding is probably the consequence of frequent exposure of humans to T. evansi-infected livestock.
Camel farmers can experience considerable economic losses because of T. evansi infections because they result in outbreaks of abortions (and subsequent loss of replacement breeding stock or meat animals), decreased milk production, loss of body condition, decline of production efficiency and increased mortality. Additional financial losses occur because of premature culling of sick animals and biannual treatments administered to T. evansi seropositive animals. These devastating financial impacts point out the need to establish effective surveillance, control and prevention programmes.
Conditions in which contact with a vector is enhanced, such as sedentary camels sharing a common water source in an oasis or contact between animals and target potential reservoirs, are common in camel rearing (Elamin et al. 1998).
In the particular case of camel husbandry in Africa, nomadism is characterised by migration during the dry season to areas with bioclimatic areas or conditions favouring survival of the vectors (e.g., woodlands, valleys, permanent water). Nomadism, therefore, is an important risk factor (Baumann & Zessin, 1992;Delafosse & Doutoum, 2004;Dia et al. 1997;Ngaira et al. 2003).
In the present study, T. evansi seroprevalence in individual camels in large camel herds reached 52%, which corresponded to a risk of infection that was 2.1 times higher than camels from small herds. This higher risk is probably because of the fact that, in large farms, the infection pressure is stronger with more frequent contact between infected and non-infected animals. Our results agree with a study in Chad, which revealed that the largest herds were associated with a higher risk of trypanosomosis (Delafosse & Doutoum 2004).
The source of water (e.g., rivers versus artificial wells) was a major risk factor for exposure to T. evansi in our study. Herds whose primary water source was a river had greater rates of infection than herds that drank from wells. Also, the presence of a dense canopy and permanent ponds create ideal breeding conditions for various species of tabanids, thus promoting spread of the disease (Doutoum et al. 2002).
Where feasible, camel farmers should reduce access to naturally occurring water sources and instead provide well water.

Conclusion
Dromedary camels in eastern Algeria showed high rates of seropositivity for T. evansi at both the individual and herd levels. At the individual level, the three main risk factors for trypanosomosis were herd size, animal management system and accessibility to naturally occurring sources of water (rivers, lakes). Herd seroprevalence varied to a much lesser extent because of geographical location (district). These findings stress the importance of an integrated disease surveillance system for T. evansi. Preventive measures are needed to improve the health status of camel herds and to reduce the economic losses associated with outbreaks of camel trypanosomosis (which include abortions, cachexia, decreased productivity and mortality).
Preventive measures may include surveillance of animal movements, particularly in rural areas. Studies employing more sensitive molecular diagnostic techniques, as well as epidemiological surveys to estimate seroprevalence at a national level, are needed to develop an efficient control strategy to avoid an endemic state of disease. In the context of global climate change, trypanosomiasis has been recognised as a re-emerging disease and a serious and economically important health issue with particular interest for Africa and Asia.