Farms and sampling

One hundred and sixty cows from five certified organic farms representative of the Cantabrian Region were monthly monitored from February 2006 to January 2008 and a total of 8,496 foremilk samples were collected. These farms were selected from the whole population (n=17) on the basis of being representative of the sector (size, Holstein-Friesian breed, feeding and management practices, etc.), being in the organic system for more than 2 years and agreeing to participate in the study. Overall all the farms were small (<50 cows in lactation) traditional farms, with a high degree of pasture (66-82% dry matter intake (DMI)) and a low milk production (average milk yield: 5950 L) and used homeopathic treatments to some degree. Details of the farms and sampling collection are described in Villar & López-Alonso (2015).

Laboratory analysis

Microbiological milk cultures. Single-quarter foremilk samples (0.01 mL) were plated on blood agar containing 0.1% aesculin and incubated at 37ºC for 48 h. Bacteria were identified by colony morphology, gram staining, haemolysis patterns, catalase test, the aesculin reaction, the Camp test, API system (Biomérieux, España)and other standard microbiological methods. A quarter was considered culture-positive when more than three microbial colonies were detected, except for Corynebacterium and coagulase negative staphylococci (CNS), where ten microbial colonies were needed. Samples yielding three or more different bacterial species were considered contaminated.

SCC. Somatic cell count was determined in foremilk samples using a Fossomatic FC fluorescence-optical counting system at the Dairy Interprofessional Laboratory (Guarnizo, Cantabria).

Diagnosis and intramammary infection criteria

A quarter was diagnosed with “clinical mastitis” when any macroscopic changes in the milk or palpatory abnormalities of the udder were observed by the farmer/veterinarian during the last month (receiving or not medical treatment) or the research team during the monthly sampling. A quarter was diagnosed with “subclinical mastitis” if no detectable changes in the udder and milk were observed but the quarter milk SCC exceeded 300,000 cells/mL for two consecutive monthly controls and the same mastitis pathogen was isolated in both examinations. The only exception was Staphylococcus aureus: since SCC can be low in cows infected by this pathogen (Green & Bradley, 2004) a threshold of 300,000 cells/mL was not required for the diagnosis. A quarter was defined as suffering “chronic mastitis” when the duration exceeded 3 monthly controls, both remaining in a subclinical phase all the process, or alternating between clinical and subclinical phases (NMC, 1998). The pathogen involved in the mastitis process was considered “uncertain” when (i) the isolated microorganism varied along the mastitis process or (ii) for short-duration clinical mastitis when the microbiological culture was negative during the monthly control in spite of having been observed both udder and milk abnormalities.

Each cow was monitored over one or two consecutive lactations; animals with less than six monthly controls per lactation were excluded of the analysis. No monitoring was performed during dry periods. To calculate the duration of the infection, only complete mastitis processes (beginning and ending at the monitoring period) were considered. The quarter samples taken at the last control before the dry period and at the first control after calving allowed the progress of infection to be followed: when the same pathogen was isolated in both samples the mastitis was considered as pre-existent from the previous lactation.

Corynebacterium bovis, other Corynebacterium spp. and CNS were considered minor pathogens. Within the major pathogens, Streptococcus agalactiae, S. aureus, and Mycoplasma spp were considered contagious pathogens, whereas coliform bacteria (including Escherichia coli, Klebsiella spp. and Enterobacter spp.) and species of streptococci other than Strep. agalactiae were considered environmental pathogens (NMC, 2015).

Definitions

Incidence rate at cow level: Newly infected cows per 100 cows at risk per control. Cows at risk included all healthy animals at the beginning of the particular period.

Incidence rate at quarter level: Newly infected quarters per 100 quarters at risk per control.

Prevalence rate at cow level: Cows with at least one infected quarter per 100 cows per unit of time.

Prevalence rate at quarter level: Infected quarters per 100 quarters at risk per unit of time.

Duration of infection: Period (in months) between 1^st positive control (mastitis diagnosis) until cure.

Statistical analysis

Statistical analysis was performed using SPSS v 20.0. SCCs were transformed to base-10 logarithmic scale prior to statistical analyses. Proportions of pathogen prevalences depending on mastitis diagnosis and time of beginning of infection were compared by using z-test for proportions; differences among seasons on prevalence and new infection rates by using chi-squared test; and differences on SCC depending on diagnosis of mastitis (no pathogen isolation, pathogen isolation and no-mastitis and pathogen isolation and mastitis) and number of lactation (1, 2, 3 and >3 lactations) were evaluated by two way ANOVA and post-hoc HSD Tukey tests. A regression analysis was carried out to evaluate the effect of number of parturition (1, 2, 3 and >3), stage of lactation (number of monthly control), pathogen isolation (0: –; 1: +), type of pathogen (1 to 11 as ordered in Table 4) and quarter position (1: front right, 2: front left 3: rear right, 4: rear left) on the SCC at a single-quarter level.

Table 1. Pathogen prevalence in culture-positive foremilk samples according to the diagnosis criteria “no mastitis” or “mastitis” (details in the text).

Table 2. Main pathogen prevalence in foremilk samples diagnosed as mastitis by farm.

Table 3. Mastitis variables (in %).

Table 4. Pathogen prevalence according to time of beginning of infection and to type of mastitis.

Table 5. Duration of infection according to pathogen.

Table 6. Foremilk SCC (expressed as mean log SCC ± SD and geometric mean in parenthesis) according to diagnosis of mastitis and number of lactation.

Table 7. SCC (expressed as LogSCC ± SD and geometric mean) from infected quarter according to pathogen isolation.

Table 8. Summary of the regression model for the SCC in foremilk samples according to number of parturition, time of lactation, pathogen isolation, type of pathogen and quarter position (R²=0.382).

Microbiological status of mammary glands

Only a slight comparison can be made between our data on mastitis in organic farming in Northern Spain and those reported in the literature based on differences in the criteria used to diagnose subclinical mastitis. If the criteria of diagnosis were the presence of pathogens in the foremilk sample (as made by Bradley et al., 2007) 79.3% of quarters would be infected in our study, whereas if the diagnosis criteria were a SCC over a threshold (usually 300,000 cells/mL; Pitkälä et al., 2004), mastitis diagnosed quarters would be 24.3%; moreover, if the criteria were the presence of contained pathogens plus SCC >300,000 cells/mL in a foremilk sample (Hogan et al., 1989a) the figure would be 23.2%. Our definition of subclinical mastitis is very restrictive to try to minimize false positive diagnosis based on the high number of positive microbiological cultures, so that the second control should be understood as “confirmatory”; nevertheless we are aware that the best way for confirmation would be to take and analyse foremilk samples by duplicate, and to establish as diagnostic criteria isolation of the same pathogen from both duplicate foremilk samples (Hogan et al., 1989a). The difference between considering one (23.2%) or two controls (21.2%) is small and would include both incorrectly diagnosed udder quarters and short duration mastitis. In spite of the difficulties to compare data, particularly for gram-negative microorganisms, most studies, both in conventional and organic farms, have shown lower mastitis rates than in our organic farms (Hovi & Roderick, 1998; Vaarst, 2001; Lopez-Villalobos et al., 2003; Sato et al., 2005; Piepers et al., 2007; Ruegg, 2009).

The proportion (nearly 80%) of pathogen-positive cultures in foremilk samples was very high compared with data reported in conventional farming in our region (García, 1990) and most studies throughout the world (Hogan et al., 1989a,b; Wilson et al., 1997; Middleton et al., 2004; Tenhagen et al., 2006; Piepers et al., 2007). In most of these cultures C. bovis (69%) was the pathogen involved. According to Watts et al. (2000), C. bovis is a teat canal-region pathogen, frequently isolated in herds with low hygiene during milking and especially no post-milking teat disinfection, so the high prevalence of C. bovis could be a consequence of a high teat canal colonization associated to the management and sanitary practices of organic farming, mainly the lack of regular post-milking teat disinfection. In a nationwide survey in Finland, Pitkälä et al. (2004) found that the number of culture-positive quarter foremilk samples increased from 21% (1995) to 33.5% (2001), and related this to a high frequency of infection by coryneform bacteria.

Mastitis control programs carried out in conventional farming led to a change in the farm pathogen profile (Piepers et al., 2007). They have demonstrated to be very effective methods to control contagious mastitis pathogens —allowing to eradicate Strep. agalactiae from a herd and effectively reduce S. aureus infected quarters (Hogan et al., 1989a)— but being less effective to control mastitis caused by environmental pathogens that become the prevalent pathogens. The pathogen prevalence observed in our study, with a similar weight of both environmental and contagious microorganisms, is probably related to the follow-up of a strategy of low employment of antibiotics and other conventional mastitis control tools (that effectively control the contagious ones) by organic farmers. In fact, when conventional measures (namely allopathic treatments) are used for the mastitis control (as indicated for farm #2) the incidence of contagious pathogens is very low.

Intramammary infection dynamics

Again, only a broad comparison of the infection dynamics can be made between our data and those reported in the literature, based on differences on the diagnosis criteria and the way in which data are presented (for review see Ruegg, 2009). In general, prevalence of clinical mastitis is similar to those reported in conventional farms in Northern Spain (Pérez-Cabal et al., 2008) although no data of subclinical mastitis is available to compare. Limited information is available in the literature comparing the mastitis prevalence in organic and conventional systems worldwide, however, in contrast to some inconsistencies among studies when comparing SCC, virtually all studies have reported fewer cases of clinical mastitis for organic when compared to conventional farms (for review see Ruegg, 2009). A completely different scenario has been observed for the subclinical mastitis, which is assumed to be a frequent problem in organic farming (Krutzinna et al., 1997; Weller & Davies, 1998; Busato et al., 2000; Fehlings & Deneke, 2000; Hovi & Roderick, 2000; Zwald et al., 2004; Roesch et al., 2006, 2007; Doherr et al., 2007). Environmental pathogens are generally associated with higher percentages of clinical mastitis than contagious pathogens (Bradley et al., 2007). Thereby, in organic farms without the traditional control measures for preventing mastitis, it would not be strange that differences in clinical mastitis do not exist when comparing with conventional systems. In contrast, differences in the prevalence are expected when subclinical mastitis are compared, since this type of mastitis are often associated with contagious pathogens (more sensitive to control measures; Hertzberg et al., 2003). Moreover, other factors related to husbandry, management, genetics, nutrition and associated metabolism and endocrine changes (Elbers et al., 1998; Bielfeldt et al., 2004) could also have a significant contribution. In fact, in studies conducted in identical experimental conditions, no differences were observed between the prevalence of subclinical mastitis in organic and conventional managed herds (López Villalobos, 2003; Fall et al., 2008).

The mean mastitis prevalence in our study (47.4%) was higher than in the conventional dairy sector in North Spain (24-29%, using a less strict diagnosis criteria of pathogen isolation plus a SCC>200,000 cells/mL in a simple control (Luis M. Jiménez-Galán, pers. comm.). This result indicates again a worse udder sanitary condition of the organic herd, though this could be related in part with the profile of the pathogens observed in the organic farms. It is well known that contagious mastitis have a longer duration that the environmental mastitis (Escobal et al., 2004); even with a similar incidence rate, herds affected by contagious mastitis can have a quarter prevalence of 40-60%, whereas in herds predominantly affected by environmental mastitis the quarter prevalence could be ~15-20%. Moreover, the low incidence rate in our study (Hogan et al., 1989a) together with the high prevalence rate (specifically for subclinical infections), indicates that the main problem of the organic herd is the chronification of the infections. The duration of mastitis in our organic herd was longer that in other studies (Zadoks et al., 2001; McDougall et al., 2004), especially if we take into account that the prevalent pathogens were environmental (Escobal et al., 2004). However, our data may have been overestimatedsince they were calculated over the period of study and not over the period of lactation, that means infections reaching drying off and persisting after calving (accounting for a 30% of the total) were counted as a single process. Moreover, the monthly control could not detect short-term infections, and some long-term infections may not have been accurately identified, e.g., processes in which cows may have been cured but then re-infected by the same pathogen (by identical or different strains) between controls.In this context, it should be considered that without conducting molecular methods, it cannot be distinguished between cure-reinfection with the same pathogen and chronification. A recent study in Northern Spain (Lavín, 2013) applying Molecular Typification Methods (Pulsed Field Gel Electrophoresis-PFGE and GTG5-PCR) in the study of epidemiology of infections by Strep. uberis showed thatthe same strains were isolated both in the udder and the environment and more than 70% of the monitored udder quarters showed different strains during the infection process, as well as after and before drying, supporting the cure-reinfection hypothesis against persistent infections. Everything seen underlines the need to use typing methods to distinguish between persistency and cure-reinfection during lactation (monthly controls) and throughout the dry period.

No seasonal trend was observed, either for the prevalence or the incidence rate in dairy farming in North Spain. It is commonly assumed that the risk of mastitis is higher in summer (Hogan et al., 1989a; Vaarst, 2001; Escobal et al., 2004): hot and humid weather increases mastitis germ loads, as well as decreasing animal immunity due to heat stress, lower dry matter intake, and other stressors (Hammami et al., 2013) being especially true for environmental mastitis pathogens such as E. coli and Strep. uberis that grow where it is warm and moist. The weather conditions in our region without extreme summers could explain our results.

Single-quarter somatic cell counts

Although SCC have been largely studied in organic farming at a cow or herd levels (bulk tank SCC) only a few information is available at single-quarter level. Our results indicate that, as expected, single-quarter SCC dramatically increases in case of mastitis (Reneau, 1986; Schepers et al., 1997), but also moderately in animals chronically exposed to sources of infection. Older cows tend to have mastitis that are longer and cause more extensive tissue damage; animals with previous histories of mastitis elicit greater cellular response than uninfected cattle (Reneau, 1986). Major pathogens are generally associated to higher SCC at quarter and cow levels compared with minor pathogens (Reneau, 1986; Schepers et al., 1997). Although classically C. bovis is considered a minor pathogen, causing subclinical mastitis and low SCC (Schepers et al., 1997), in our study SCC associated to C. bovis were very high, similar to other major pathogens. By contrast, SCC in foremilk samples infected by S. aureus were very low: whereas some studies found very high SCC in cows infected with S. aureus, similar to other major pathogens (Schepers et al., 1997), infected animals do not necessarily have elevated SCC; in this sense Jones et al. (1984) found that only 60% of the infections by S. aureus were found in cows producing milk with SCC greater than 200,000 cells/mL.

Our results also indicate that non-bacteriological factors, mainly number and stage of lactation are important when establishing thresholds of udder health. SCC significantly increases with number of lactation and is inversely related to milk yield (dilution factor) (Reneau, 1986). Both factors are very important in organic farming since herds are older and produce less milk (understood as a low dilution effect) that the conventional counterpart, and should be considered when establishing relative thresholds for mastitis control.

In conclusion, our results indicate that the higher SCC observed in organic farms in North Spain compared to the conventional ones (Villar & López-Alonso, 2015) are mainly associated to a high prevalence of chronic subclinical mastitis, even though the high number of parturitions and the low production in part, could also explain this. The high number of pathogen positive cultures, mainly by Corynebacterium bovis, in foremilk samples that did not fit our mastitis diagnoses criteria, highlights the frequent pathogen colonization of the teat-canal region and could be a consequence of the absence of teat dipping after milking in these farms. The following of the mastitis processes has revealed that a lot of infections start at the dry off and periparturient period, and are maintained during most of lactation (even to the next lactation), probably being a consequence of the low use of antibiotics in dry and lactating periods in these farms. All of the above mentioned is indicative of an inefficient mastitis control program in the studied farms and highlights the importance of a strict mastitis control program including a rational (selective) use of antibiotics in dry and lactating cows, the culling of cows with chronic disease, the acquisition of more suitable breeds, the improvement of hygiene and preventive measures in animal housing, management and feeding.