Diagnostic rates in ruminant abortions are low worldwide, reaching approximately 50% of the cases ( Anderson, 2007 ; Moeller, 2012 ; Matthews, 2016 ). Most diagnostic laboratories use a standardized diagnostic protocol consisting of a panel of tests ( Anderson, 2007 ) that generally include pathogens that are considered important in abortion outbreaks, such as Bovine Viral Diarrhoea (BVD) viruses type 1 (BVDV-1) and type 2 (BVDV-2), Bovine Herpesvirus-1 (BoHV-1), Neospora caninum and Leptospira spp. in cattle, and Chlamydia abortus and Toxoplasma gondii in small ruminants. Other pathogens such as Campylobacter spp., Salmonella spp., or Listeria spp. are considered occasional ( Borel et al. , 2014 ), so their presence in panels of diagnosis is variable and it depends on the laboratory.

Additionally, control programmes carried out by Health Defence Groups (ADSG, Agrupación de Defensa Sanitaria Ganadera ) in cattle usually include some of the most important reproductive pathogens (BVDV, BoHV-1 and N. caninum ). However, abortions are frequently an important problem on these farms and the relative situation to the presence of other pathogens potentially involved in reproductive failure is often unknown.

In this context, a narrow range of pathogens tested or controlled may contribute to the low efficiency in the diagnosis. To avoid this, in addition to the aforementioned important pathogens for each species, we have included in the differential diagnosis pathogens variably considered by laboratories such as Campylobacter spp., Salmonella spp., Coxiella burnetii or Listeria spp. and pathogens rarely included in panels of diagnosis in Spain, such as U. diversum in cattle and N. caninum and Leptospira spp. in small ruminants. This panel was applied to a group of bovine, ovine and caprine farms within ADSG and suffering of abortion outbreaks, in spite of the existence of control programmes. Our aim was to identify the presence of pathogens that are not frequently investigated or present in control programmes in abortion outbreaks in farms with ongoing control programmes for some reproductive pathogens.

Thirty-one of 42 available farms were defined as cases of abortion outbreaks and included in this analysis. There were 16 bovine, 7 ovine and 8 caprine farms and a total of 52 analysed foetuses: 25 bovine, 11 ovine and 16 caprine. Compatible causes for abortion were identified in 21/31 farms: 9/16 bovine, 6/7 ovine and 6/8 caprine (Figure 1). The results are broken down by pathogen and sample type in Table 1. Overall, these results prove the presence of atypical pathogens in ruminant abortions, highlighting the diagnosis of U. diversum in 4/16 bovine farms, BVDV-2 in 2/7 ovine farms and Campylobacter spp. in 3/16 bovine farms, 2/7 ovine farms and 4/8 caprine farms. In cattle, both U. diversum and Campylobacter spp. have been considered a sporadic abortifacient ( Anderson, 2007 ), but they may also constitute one of the most frequently diagnosed causes in a region ( Campero et al. , 2003 ; Syrjälä et al. , 2007 ). Speciation of Campylobacter spp. was not performed, but since all bovine herds performed artificial insemination and there is no evidence of venereal transmission of Campylobacter spp. in small ruminants ( Timoney et al. , 1988 ), faecal origin is the most likely source of infection. In small ruminants, some of the most important abortifacients in ovine/caprine were not extensively identified ( T. gondii, C. abortus ). In contrast, Campylobacter spp., Leptospira spp., Listeria spp. and N. caninum were found herein.

Twenty farms with a diagnosis provided information regarding the duration of problems. An apparent difference between the three species was observed. The six ovine diagnosed cases were all short-term outbreaks with less than 1 month of duration, which is consistent with lambing management in this species. In bovine, a large duration of problems was the most frequent situation (4/8 > 6 months; 3/8 between 1-6 months; 1/8 < 1 month), indicating persistent herd complications. In caprine the duration of problems was < 1 month in 2/6, 1/6 between 1-6 months and 3/6 more than 6 months, which may indicate episodes of persistence between lambing seasons.

The incidence of abortion is not consistently documented in this region. After consulting practitioners, mean and median values of annual percentage of abortions in cattle could be estimated around 7.59% and 4.9% In small ruminants, around 10% of the farms enrolled in the ADSG would submit samples to the laboratory for abortion diagnosis yearly. The identification of an aetiologic agent is not achieved in about 55-70% of the cases, according to practitioners in the zone. The prevalence of pathogens in this region is not extensively reported. A study underlined C. burnetii in 9/44 cattle abortions, but the diagnosis was only achieved in the 34.1% of cases ( González-Warleta et al. , 2016 ). This situation was not found in our study; however the differences may due to the low sample sizes or the different methodologies used in both studies. The exposure to some reproductive pathogens has been reported. Animal seroprevalences of 25% for BVD ( Eiras, 2010 ), 38.4-35.7% for IBR ( Eiras, 2010 ) and 15.7-24.1% for N. caninum ( González-Warleta et al. , 2008 ; Eiras, 2010 ; Panadero et al. , 2010 ) are documented in cattle. In small ruminants, animal seroprevalences for T. gondii were 38.1-57% in sheep ( Panadero et al. , 2010 ; Díaz et al. , 2014 ) and 48% in goats ( Díaz et al. , 2016 ); and for N. caninum , 5.5-10.1% in sheep ( Panadero et al. , 2010 ; Diaz et al. , 2014 ) and 6% in goats ( Diaz et al. , 2016 ). In sheep, a serological study showed high seropositivity to T. gondii (38.1%) and low for N. caninum, C. burnetii, C. abortus and Pestivirus (< 8%) with only 1/44 farm with problems of abortion ( Diaz et al. , 2014 ). Prevalence studies for other pathogens are sparse and focused on limited areas ( Díaz Cao, 2016 ).

Our study offers some information regarding the problematic of abortions in this region. However, it should be noted that due to the sampling method and the low number of cases analysed, this study cannot aim to be representative of the prevalence of abortive agents. Actually, it could be expected that the control programmes for some pathogens could increase the relative importance of other infections, as it has been shown to occur in bovine mastitis ( Fernandez et al. , 2013 ). Nevertheless, our results prove that other pathogens not routinely considered in the differential diagnosis of abortions may be present in these farms and consequently, they should be considered in the differential diagnosis of abortion. The addition of U. diversum , as well as N. caninum and Leptospira spp. in small ruminants, and Campylobacter spp. in the three species increased our aetiologic diagnostic rates from 4/16 (25%) to 9/16 (56.25%) in bovine, from 3/7 (42.86%) to 6/7 (85.71%) in ovine and from 2/8 (25%) to 7/8 (87.5%) in goats. Low sample sizes do not make these percentages very reliable, but they indicate that some gain in the diagnostic rates may be obtained by adding some pathogens to the diagnosis.

The low number of cases submitted to our laboratory could have been due to the costs of analysing a wide panel of pathogens. In addition, low rates of diagnosis of abortions can make it economically unattractive. However, narrow-range diagnostic panels can fail to make a diagnosis with subsequent negative economic consequences for farms. We recommend the inclusion of these commonly excluded pathogens as a second step, after discarding those others of recognized importance. In summary, our results emphasize the importance of continuing research to reveal the presence of overlooked pathogens in the field.