Seroprevalence of Oestrus ovis (Diptera, Oestridae) Infestation and Associated Risk Factors in Ovine Livestock from Southwestern Spain

MOLECULAR BIOLOGY/GENOMICS Seroprevalence of Oestrus ovis (Diptera, Oestridae) Infestation and Associated Risk Factors in Ovine Livestock from Southwestern Spain MARÍA ALCAIDE, DAVID REINA, JAVIER SÁNCHEZ-LÓPEZ, EVA FRONTERA, AND IGNACIO NAVARRETE. Parasitology Section, Department of Medicine and Animal Health, Faculty of Veterinary Sciences, University of Extremadura, 10071 Cáceres, Spain J. Med. Entomol. 42(3): 327Ð331 (2005) ABSTRACT This survey was carried out to determine the seroprevalence of nasal infestation by sheep bot ßy, Oestrus ovis L., and to identify the risk factors associated with the disease in ßocks in southwestern Spain. In total, 5,878 sera samples of adult sheep were collected at random in 551 farms from four provinces in the southwestern Spain: Badajoz, Cáceres, Córdoba, and Sevilla. Sera were tested by enzyme-linked immunosorbent assay for O. ovis antibodies, by using a crude L2 larval as antigen. The seropositive mean prevalence was 69.30%, and mean percentage of optical densities was 61.83%. There were signiþcant differences between the provinces studied; Córdoba and Sevilla were the provinces with more infested animals and higher seroprevalences. The correlation between seroprevalence and percentage of antibodies by farms was signiþcant. There were only 18 farms free of seropositive animals, and 115 of the total 551 farms had all sampled animals seropositive, an indication of the high importance of this parasitosis in the investigated areas. Altitude, latitude, ßock size, and ovine population density were the potential risk factors associated with the detection of O. ovis antibodies. Those animals breeding in regions located at low altitudes ( 500 m), meridian latitudes ( 39.5 N), and on farms with medium-to-large ßock size ( 250 sheep) and high ovine population density ( 100 sheep per km 2 ) were more likely to be seropositive. These Þndings conþrm that these studied factors should be considered as potential risk factors to the presence of O. ovis in ovines from southwestern Spain. KEY WORDS Oestrus ovis, ovine, seroprevalence, risk factors, Spain SHEEP BOT FLY, Oestrus ovis L., is a cosmopolitan parasite that affects ovine and caprine livestock. The success of this myasis agent to thrive in Mediterranean, tropical, or desert environments is mainly due to high reproductive rate and the ability to adjust its developmental rhythms and energy economy to particular climactic conditions (Cepeda-Palacios 2001). The presence of larvae is detrimental to the health of infested animals (Cobbett and Mitchell 1941, Horak 1981). They invade the sinu-nasal passages to cause sinu-nasal myiasis, especially in heavy infestations. This situation predisposes the host to other maladies of the respiratory system, secondary bacterial invasion such as septic ovine pneumonia (Bissette 1932, Bekele and Mukasa-Mugerwa 1994), or complication caused by sinu-nasal tumors or lung abscesses (Dorchies et al. 1993) and cerebral insolvency (Urquhart et al. 1996). Recent surveys have demonstrated the high prevalence of ovine oestrosis in numerous areas all over the world, e.g., 81.3% in India (Pathak 1992), 30.6% in Mexico (Murguõ a et al. 2000), 86.3% in Burkina-Faso (Ouattara and Dorchies 1996), and 40% in Slovenia (Brglez and Polajner 1997). In Europe, there are also many studies about the presence of O. ovis, especially in Mediterranean countries such as Italy, where the prevalence described was 91% (Scala et al. 2001), and in France, with 43.4% of sheep parasitized (Dorchies et al. 2000). Due to the estimated importance of ovine oestrosis and its severe consequences in livestock in Spain, the aim of the current study was to determine the seroprevalence of O. ovis infestation and the inßuence of potential risk factors associated with this myiasis in sheep ßocks from southwestern Spain. Materials and Methods Data Collection and Study Area. In this investigation, a total of 5,878 serum samples from adult sheep ( 6 mo of age) were examined. These samples were obtained from the Animal Health Regional Laboratories in Badajoz, Cáceres, Córdoba, and Sevilla (Fig. 1). They were routinely collected for the brucellosis surveillance program in these regions between November 2000 and September 2002. Data about location, number of animals, and approximate age were recorded from each checked farm. 0022-2585/05/0327Ð0331$04.00/0 2005 Entomological Society of America

328 JOURNAL OF MEDICAL ENTOMOLOGY Vol. 42, no. 3 Fig. 1. Map of Spain showing the investigated southwestern provinces. (a) Cáceres. (b) Badajoz. (c) Sevilla. (d) Córdoba. According to the ofþcial census in December 1999, the total number of sheep in Badajoz was 2,700,044; in Cáceres, 1,675,364; in Córdoba, 651,918; and in Sevilla, 302,045. The number of animals investigated represented 0.09% in Badajoz (n 2,318), 0.12% in Cáceres (n 2,000), 0.16% in Córdoba (n 1,035), and 0.17% in Sevilla (n 525) from the total ovine population of each province from southwestern Spain. The number of animals per square kilometer in the studied regions was between 0.75 and 242 (median 114.88, mean 105.64). In total, 551 farms were surveyed, 245 from Badajoz, 170 from Cáceres, 83 from Córdoba, and 53 from Sevilla. The number of ewes per ßock ranged from 2 to 2,275 (median 400, mean 451.56). The sera samples collected from each ßock were not proportional to ßock size. The study was carried out in southwesternern Spain in an area located between the 40.2 and 36.7 N parallel and the 7.2 and 4.1 W meridian (Fig. 1). The majority of the ßocks studied were maintained primarily in grazing lands of traditional Mediterranean forest, whereas the remainder made use of cultivated lands, essentially dry farming and mountain pastures no more than 1,100 m above sea level. The mean altitude in the surveyed area was 431 m, and the median was 472 m. The weather in southwestern Spain is characterized by irregular rainfall in spring and fall. Summer is long, hot, and extremely dry, with an average temperature and rainfall of 30 C and 19 mm/mo. Winter is calm and dry, with a mean temperature 12.5 C and rainfall of 80 mm. Enzyme-Linked Immunosorbent Assay (ELISA). Sera samples were obtained and stored at 20 C until use. The ELISA used was previously described by Yilma 1992. Second instars were obtained from heads of naturally infected sheep at slaughterhouses in Cáceres. After the larvae were homogenized, the extract was centrifuged at 450 g for 20 min at 4 C, the supernatant was removed, and its protein concentration was determined by a modiþed method of Bradford (Coomassie Plus protein assay, ref-1856210, Pierce Chemical, Rockford, IL). Plates were incubated with the antigen for 3 h at 37 C. After washing, 5% phosphate-buffered saline (PBS) skimmed milk solution was added for blocking the wells. After triple washings, 100 l of duplicated sera diluted at 1:200 in PBS with 0.1% Tween 20 was added to each well and incubated at 37 C for 45 min. After washing, 100 lof anti-sheep IgG (whole molecule) peroxidase conjugate (A-3415, Sigma, St. Louis, MO) was added per well, diluted at 1:4000. Finally, 100 l of substrate solution with 0.05 M citrate buffer, ph 5.0, hydrogen peroxide, and 3,3,5,5 -tetrametilbencidine (T-3405, Sigma) was incubated for 20 min and stopped with 3 N sulfuric acid. Positive controls were evaluated by repeated titration of several serum samples belonging to naturally infested animals from the studied regions. For negative control, we obtained serum samples from young treated animals or sheep kept indoors that had had no contact with adult bot ßies. The cut-off value was determined as a value of 2 standard deviations (P 0.05) above the mean optical densities for negative control serum samples, and in study was 36%. Data Analysis. The percentage of antibodies (mean percentage of optical density [OD]) was calculated for each sera sample by comparison with positive and negative reference sera as follows: % antibodies OD (serum sample) OD (negative sample) OD (positive control sera) OD (negative control sera) 100 The prevalence data were calculated as follows: % seroprevalence Total no. of serum samples with OD 0.36 Total no. of sheep examined 100 Data analyses were performed using the commercial statistical program SPSS 11.0 for Windows (SPSS Inc., Chicago, IL). Quantitative variables (seroprevalence and percentage of antibodies) were categorized by latitude, altitude, ßock size, and ovine population density. Statistical comparisons were evaluated by analysis of variance (ANOVA)-test, to investigate the signiþcant differences among groups. Pearson correlation was estimated between seroprevalence or percentage of antibodies and the potential risk factors (latitude, altitude, ßock size, and ovine population density). Results Seroprevalence and Percentage of Antibodies. The results of seroepidemiological survey showed that 4,070 animals were seropositive and the estimated seroprevalence was 69.30 27.88%. When analyzed by province, the seroprevalence was 67.19% in

May 2005 ALCAIDE ET AL.: OVINE OESTROSIS IN SOUTHWESTERN SPAIN 329 Table 1. Values of the Pearson correlation (r) and the level of significance (P) between risk factors (latitude, altitude, flock size, and ovine pop density) and the seroprevalence and the percentage of antibodies of O. ovis infection in sheep from southwestern Spain Risk factor Seroprevalence % antibodies r P r P Latitude 0.108 0.011 0.126 0.003 Altitude 0.116 0.007 0.121 0.004 Flock size 0.304 0.000 0.227 0.000 Ovine pop density 0.069 0.106 0.001 0.988 Fig. 2. Mean seroprevalence and percentage of antibodies of sheep in southwestern Spain by surveyed provinces: Badajoz, Cáceres, Córdoba, and Sevilla. Cáceres, 64.89% in Badajoz, 75.40% in Córdoba, and 78.84% in Sevilla (Fig. 2). These differences were signiþcant between Cáceres and Badajoz relative to Córdoba and Sevilla. The percentage of antibodies ranged between 53.31% in Badajoz and 76.6% in Sevilla (Fig. 2). By farms, 96.73% (n 533) of sampled ßocks had parasitized sheep because only 18 farms were found to be free of infested animals. A total of 115 had all their animals seropositive (Fig. 3). The Normal distribution calculated for the ovine population investigated was clearly sloped to the right. Therefore, the majority of farms surveyed had 60% of their animals seropositive. The detection of anti-oestrus antibodies was significantly associated with the seroprevalence (r 0.886, P 0.001). The high number of animals seropositive showed the highest levels of antibodies. Potential Risk Factors. There was a signiþcantly negative correlation between seroprevalence and latitude (Table 1). It was more frequent to Þnd seropositive animals in locations under 39.5 N parallel (Table 2). The correlation between seroprevalence or Fig. 3. Within-ßock seroprevalence of O. ovis. (0, no seropositive sheep; 1, prevalence 25%; 2, prevalence 25Ð 50%; 3, 51Ð75%; 4, 76Ð99%; and 5, all animals seropositive) and the Normal distribution estimated in ovine livestock from the southwestern Spain. percentage of antibodies and altitude was also significantly negative (r 0.108). The lower number of seropositive animals was found at higher altitudes. However, the differences among the described ranges (Table 2) were signiþcant between the altitudes under 500 m relative to the higher altitudes. Positive and signiþcant correlation was found between ßock size and seroprevalence or percentage of antibodies (Table 1). Large ßocks (66.12%) and medium and very large ßocks (71.41 and 82.95%, respectively) were more likely to have seropositive sheep and higher percentages of antibodies than small ßocks (55.59%). No signiþcant correlation was found between ovine population density and seroprevalence. However, a signiþcant association was observed between the detection of anti-oestrus antibodies and ovine density. Hence, in regions with 100 sheep per km 2 there was a greater probability of Þnding seropositive animals than in less populated areas. Discussion Despite the importance of parasitism by O. ovis in small ruminants and the high number of ewes breed- Table 2. Seroprevalence and percentage of antibodies of O. ovis infection in sheep from southwestern Spain for risk factors latitude, altitude, flock size, and ovine population density Risk factor No. of sheep examined No. of ßocks examined % seroprevalence % antibodies Latitude 40.2Ð39.5 N 579 66 58.87a 50.76a 39.5Ð38.8 N 1,551 292 68.03b 60.35a,b 38.8Ð38.1 N 3,036 121 71.15b 56.64a,b,c 38.1Ð37.4 N 283 28 72.58b 65.02b,c 37.4Ð36.7 N 429 44 76.49b 68.57c Altitude (m) 0Ð200 434 46 70.92 61.72a 200Ð500 1,646 272 70.28 60.43a 500Ð750 3,691 223 66.31 55.14b 750 107 10 59.14 45.83b Flock size 5Ð250 1,868 182 55.59a 49.67a 250Ð500 1,499 153 77.41b 65.59b 500Ð750 1,447 126 66.12c 52.54a 750 1,064 90 82.95b 69.52b Ovine pop density 0Ð100 sheep/km 2 2,060 220 64.71a 57.29 100 sheep/km 2 3,818 331 71.07b 58.69 Values with different lowercase letters for a risk factor in the same column are signiþcantly different (P 0.05).

330 JOURNAL OF MEDICAL ENTOMOLOGY Vol. 42, no. 3 ing in Spain, the knowledge of prevalence of this myasis has only recently been recorded. The Þrst published works were in 1992, when Ruõ z Martõ nez et al. (1992) carried out a slaughterhouse survey in sheep from southern Spain, detecting a prevalence of 56.5%. Later, Lucientes (2000) reported that 71.18% of sheep surveyed in northeastern Spain were infested. The Þrst serological survey in Spain was conducted by Reina et al. (2001). Their preliminary results showed a prevalence of 80% (n 1,104) in ovines from southwestern Spain, a result similar to the prevalence observed in the current study. This large-scale (n 5,878) seroepidemiological study was carried out using ELISA as described by Yilma (1992). The results demonstrated that O. ovis was a very frequent parasite among the ßocks evaluated (96.73%). The infestation was widespread inside the ovine farms, with a high prevalence of 69.30% and a mean OD of 61.83%. These values indicate that seropositive farms had the most individual animals with high reactivity, possibly due to the high risk of infestation and reinfestation for susceptible animals gathered closely. Many options of diagnosing O. ovis antibodies have been described, such as dot-elisa (Duranton et al. 1995), direct-elisa (Goddard et al. 1999), or tests using thin layers (Murguõ a et al. 1998). These serodiagnostic procedures have been validated by their effectiveness in the diagnosis of oestrosis relies, in conjunction with subjective observations of clinical signs or the demonstration of larvae postmortem. The overall seroprevalence estimated for the Spanish ovine population was not as high as what had been determined in other countries. For example, in Senegal, Deconinck et al. (1995) reported a prevalence of 88%, and Burkina Faso Ouattara and Dorchies (1996) detected 86.3% seropositive sheep. The inßuence of climatic conditions (e.g., temperature, rainfall, and relative humidity) on the presence of O. ovis in ovines is well known (Jagannath et al. 1989, Dorchies et al. 1996, Abo-Shehada et al. 2000); thus, in this study we considered other risk factors. Our study reported that sheep from meridian locations were more likely to have higher prevalences than sheep from northern locations (Tables 1 and 2). The relationship of the prevalence according to latitude was previously observed by Ruiz-Martõ nez et al. (1992)O. ovis as described previously Bergeaud et al. (1994). According to our results (Tables 1 and 2), the number of seropositive sheep was signiþcantly associated with ßock size. Many studies have reported no association between prevalence and ßock size (Gupta et al. 1985, Marchenko and Marchenko 1989, Ribeiro et al. (1990)). However, Bauer et al. (2002) in Germany, conþrmed this association, where large ßocks ( 50) were more likely to be seropositives than small ßocks ( 10). They suggested that it was easier to Þnd at least one infested animal in an endemic area with a larger number of ewes in a pen. Other work carried out by Murguõ a et al. (2000) indicated that medium size ßocks (11Ð25) had a higher probability of having the disease compared with either small ( 11) or large ( 25) ßocks due to the better husbandry systems of small and large ßocks. In our study, small ßocks ( 250) had signiþcantly fewer parasitized animals. The seroprevalence in large ßocks (500Ð750) was lower than either medium (250Ð500) or very large ßocks ( 750). This Þnding also could be explained because the farmers of large ßocks typically provide better health care. But, another reason could be the very high number of ewes per pen raised in the investigated regions, 250 sheep in medium-sized ßocks, for detecting a progressive relationship between medium, large, or very large ßock size and seroprevalence or percentage of antibodies. In fact, in caprine herds, Cepeda-Palacios and Scholl (2000) observed that the number of strikes by O. ovis gravid females was related to the number of animals gathered together grazing. SpeciÞcally, when there were 80 goats per pen, the number of strikes increased relative to the small ßocks ( 45). Finally, there was no signiþcant correlation between seroprevalence levels and ovine population density. However, the results obtained showed differences between both factors, because higher seroprevalence was observed in regions with 100 sheep per km 2. A possible explanation of this is the high ovine density in the study area because those regions surveyed are the most important ovine production regions in Spain. In conclusion, O. ovis is very commonly found and widely disseminated inside and among ovine ßocks in Southwest Spain. The results indicated that meridian latitudes ( 39.5 N), low altitudes ( 500 m), ßock size ( 250 sheep per ßock), and ovine population density ( 100 sheep per km 2 ) were the most important risk factors to this parasite in these investigated regions. Acknowledgments We recognize the exceptional help offered by Philip Scholl (USDA, Lincoln, NE) in the correction of the manuscript and in the contribution of useful ideas for the Þnal version of this article. We thank the personnel from the Animal Health Regional Laboratories from Badajoz, Cáceres, Córdoba, and Sevilla and colleagues from Veterinary Faculty of Córdoba for help in the sampling process. References Cited Abo-Shehada, N. M., B. Arab, R. Mekbel, D. Williams, and P. R. Torgeson. 2000. Age and seasonal variations in the prevalence of Oestrus ovis larvae among sheep in the northern Jordan. Prev. Vet. Med. 47: 205Ð212. Bauer, C., G. Steng, F. Prevot, and P. Dorchies. 2002. Seroprevalence of Oestrus ovis infection in sheep in southwestern Germany. Vet. Parasitol. 110: 137Ð143. Bekele, T., and E. Mukasa-Mugerwa. 1994. Oestrus ovis infection in Ethiopian highland sheep. Vet. Res. Commun. 18: 439Ð442. Bergeaud, J. P., C. Duranton, and P. Dorchies. 1994. Study of Oestrus ovis in south-west of France: a slaughter-house survey in Aveyron. Rev. Med. Vet-Toulouse 145: 863Ð866.

May 2005 ALCAIDE ET AL.: OVINE OESTROSIS IN SOUTHWESTERN SPAIN 331 Bissette, N. 1932. The possible association of the nostril ßy (Oestrus ovis) with pneumonia in sheep. Vet. J. 88: 220Ð 223. Brglez, J., and V. Polajner. 1997. Oestrosis in sheep. Veterinarske Novice 23: 393Ð394. Cepeda-Palacios, R. 2001. Recent advances in Oestrus ovis developmental biology and ecology, pp. 1Ð6. In M. Good, M. J. Hall, B. Losson, D. OÕBrien, K. Pithan, and J. Sol [eds.], Mange and myiasis in livestock. COST Action 833, European Commission, Brussels, Belgium. Cepeda-Palacios, R., and P. J. Scholl. 2000. Factors affecting the larvipositional activity of Oestrus ovis gravid females (Diptera: Oestridae). Vet. Parasitol. 91: 93Ð105. Cobbett, N. G., and W. C. Mitchell. 1941. Further observations on the life cycle and incidence of the sheep bot, O. ovis, in New Mexico and Texas. Am. J. Vet. Res. 2: 358Ð366. Deconinck, P., L. J. Pangui, L. Carriere, and P. Dorchies. 1995. Detection of sheep nasal-bot (Oestrus ovis) in Senegal with speciþc ELISA-test. Revue Méd. Vét. 146: 265Ð268. Dorchies, P., J. P. Bergueaud, G. Tabouret, F. Prevot, and P. Jacquiet. 2000. Prevalence and larval burden of Oestrus ovis (Linné, 1761) in sheep and goat in northern Mediterranean region of France. Vet. Parasitol. 88: 269Ð 273. Dorchies, P., C. Duranton, J. P. Bergueaud, and J. P. Alzieu. 1996. Kinetics of Oestrus ovis (Linné, 1758) larval development in naive lambs after natural infection in Þeld. Bulletin Soc. Française Parasitol. 14: 20Ð27. Dorchies, P., J. M. Yilma, and J. Savey. 1993. Prevalence of lung abscesses and interstitial pneumonia in ovine oestrosis. Vet. Rec. 133: 325. Duranton, C., J. P. Bergueaud, and P. Dorchies. 1995. Dotenzyme-linked-immunosorbent assay for the rapid diagnosis of ovine oestrosis. Revue Méd. Vét. 146: 283Ð286. Goddard, P., P. Bates, and K. A. Webster. 1999. Evaluation of a direct ELISA for the serodiagnosis of Oestrus ovis infections in sheep. Vet. Rec. 144: 497Ð501. Gupta, J. R., N. S. Ruprah, and M. B. Chhabra. 1985. Bionomics of Oestrus ovis (sheep nasal ßy). Indian Vet. Med. J. 9: 71Ð75. Horak, I. G. 1981. The similarity between arrested development in parasitic nematodes and diapause in insects. J. S. Afr. Vet. Assoc. 52: 299Ð303. Jagannath, M. S., N. Cozab, S. A. Rahman, and T. G. Honnappa. 1989. Factors inßuencing Oestrus ovis infestation in sheep and goats. Curr. Res. 18: 118Ð119. Lucientes, J. 2000. La oestrosis de los pequeños rumiantes. pp. 31Ð33. In Proceedings, XXV Jornadas cientõ Þcas y IV Internacionales de la Sociedad Española de Ovinotécnia y Caprinotécnia, 28Ð30 September 2000, Teruel, Spain. Marchenko, V. P., and V. A. Marchenko. 1989. Survival of larvae of Oestrus ovis L. depends on the state of the immune system of sheep. Parazitologiya 23: 129Ð133. Murguía, M. L., A. Alzina, S. Villegas, and J. C. Rodríguez. 1998. Evaluation of the immunoassay test in thin layer, using microplates of polyestirene for the detection of circulating antibodies of Oestrus ovis naturally infected ovines. Rev. Biomed. 9: 236Ð241. Murguía, M. L., J. C. Rodríguez, F. J. Torres, and J. C. Segura. 2000. Detection of Oestrus ovis and associated risk factors in sheep from the central region of Yucatan, Mexico. Vet. Parasitol. 88: 73Ð78. Ouattara, L., and P. Dorchies. 1996. Serological prevalence of Oestrus ovis infestation in Burkina Faso: evaluation using the ELISA technique. Rev. Elev. Med. Vet. Pays. Trop. 49: 219Ð221. Pathak, K.M.L. 1992. Incidence of Oestrus ovis in sheep and goats in Rajasthan state of India. Indian J. Anim. Sci. 62: 50. Reina, D., F. Bonilla, J. Martínez-Moreno, P. Dorchies, S. Hernández, and I. Navarrete. 2001. First approach to determining the prevalence of Oestrus ovis infestation of sheep in Spain, pp. 145Ð150. In M. Good, M. J. Hall, B. Losson, D. OÕBrien, K. Pithan, and J. Sol [eds.], Mange and myiasis in livestock. COST Action 833, European Commision, Brussels, Belgium. Ribeiro, V.L.S., C.M.B. Oliveira, and F.P.J. Alves-Branco. 1990. Prevalence and monthly variations in larvae of Oestrus ovis in sheep from Bage municipality, Rio Grande do Soul, Brazil. Arq. Bras. Med. Vet. Zool. 41: 211Ð221. Ruiz-Martínez, I., J. M. Pérez-Jiménez, and M. Louassini. 1992. Comparative analysis of prevalence of Oestrus ovis L. (Diptera, Oestridae) in sheep horns of southern Spain and northern Morocco. Boletim Soc. Port. Entomol. 3: 421Ð428. Scala, A., G. Solinas, C. V. Citterio, L. H. Kramer, and C. Genchi. 2001. Sheep oestrosis (Oestrus ovis, Linné 1761, Diptera, Oestridae) in Sardania, Italy. Vet. Parasitol. 102: 133Ð141. Urquhart, G. M., J. Amour, J. L. Duncan, A. M. Dunn, and F. W. Jennings. 1996. Veterinary parasitology, 2nd ed. Longman ScientiÞc, London, United Kingdom. Yilma, J. M. 1992. Contribution à l étude de l épidémilogie, du diagnostic immunologique et de la physiopathologie de l oestrose ovine. Thèse Institut National Polytechnique de Toulouse, France. Received 25 May 2004; accepted 17 January 2005.