Abstract
An observational study was conducted in 63 Canadian dairy farms to evaluate the association of bulk tank milk selenium (Se) concentration (BTSe) with average daily milk yield, milk fat and milk protein, bulk tank somatic cell count (BTSCC), and the probability of being a Staphylococcus aureus-positive herd. Bulk tank milk samples collected between March 2007 and February 2008 were evaluated for BTSCC, S. aureus culture status, and BTSe. Mean BTSe was 0.51 ± 0.15 μmol/L; no herds were classified as deficient or marginal based on BTSe. Bulk tank milk Se was unconditionally associated with milk production; however, adjusting by region indicated a confounding effect of this variable. There was no relationship between BTSe and BTSCC. Higher values of BTSe were associated with lower risk of being a S. aureus-positive herd, possibly as a result of a more robust udder immune response, or as a result of providing Se at a higher rate as one component of an udder health program that reduces risk of being S. aureus positive.
Résumé
Sélénium dans le lait de réservoir et son association avec les paramètres de production du lait dans les troupeaux laitiers canadiens. Une étude par observation a été réalisée dans 63 fermes laitières canadiennes afin d’évaluer l’association entre les concentrations de sélénium (Se) dans le lait de réservoir (SeLR) et la production quotidienne de lait, des matières grasses du lait et des protéines du lait, la numération des cellules somatiques de réservoir (NCSR) et la probabilité de la présence d’un troupeau positif pour Staphylococcus aureus. Des échantillons prélevés dans le lait de réservoir entre mars 2007 et février 2008 ont été évalués pour la NCSR, le statut de culture de S. aureus et de SeLR. Le SeLR moyen était de 0,51 ± 0,15 μmol/l; aucun troupeau n’a été classé comme ayant une carence ou des doses minimes de SeLR. Le Se dans le lait de réservoir était inconditionnellement associé à la production de lait; cependant, l’ajustement par région a indiqué un effet parasite de cette variable. Il n’y avait aucune relation entre le SeLR et la NCSR. Des valeurs supérieures de SeLR ont été associées à un risque inférieur d’être un troupeau positif pour S. aureus, peut-être en raison d’une réaction immunitaire robuste du pis ou en raison de l’ajout d’un taux supérieur de Se comme composante d’un programme de santé des pis qui réduit le risque d’un résultat positif pour S. aureus.
(Traduit par Isabelle Vallières)
Introduction
Soils of the eastern and western Canadian coasts contain relatively low selenium (Se) concentrations. Consequently, a low concentration of Se [< 0.05 mg/kg dry matter (DM)] has been reported in pastures and crops grown on these soils (1,2). A low Se concentration in the forage, which constitutes at least 50% of the total dry matter intake of the diet, has been associated with marginal serum Se concentrations in ruminants (1). Taken together, these results indicate that Se deficiency may be a widespread problem in Canada, particularly in the coastal provinces where livestock are fed entirely on locally grown feeds.
Milk yield is reported to be a sensitive indicator of Se status in dairy cattle, thus a low milk yield may be an important economic consequence of subclinical Se deficiency in cattle (3). Milk yield increased by 4.6% in cows in New Zealand that were supplemented with Se (4). In Prince Edward Island (PEI), milk yield was, on average, 7.6% higher in Se-adequate dairy herds than in herds with marginal Se status (5). Further, 2 North American clinical trials have indicated an increase between 2% and 5% in milk yield in dairy cows supplemented with organic Se at the onset of lactation (6).
A dietary Se concentration < 0.1 mg Se/kg DM has been associated with an increased risk to Se-responsive disorders, such as mastitis in cattle (7), and that risk is higher when dietary Se decreases to 0.05 mg/kg DM (8). Two North American herd-level studies identified a negative association between Se status and bulk tank somatic cell count (BTSCC) in dairy herds with good control of clinical mastitis (CM) caused by major pathogens (9,10). Staphylococcus aureus is frequently isolated in herds with high BTSCC and there is a herd-level prevalence of S. aureus of 83% in Canadian dairy herds, with a higher geometric mean BTSCC in S. aureus-positive herds than in negative herds (11). Although 59% of PEI dairy herds were marginal for Se status in 1998–1999, there was no association of bulk tank milk Se concentration (BTSe) with any measure of udder health at herd level (5).
Selenium management practices have changed over the past decade, and more Se sources are available, but little is known about the association between herd-level Se status measured by BTSe, and udder health in Canadian dairy herds. Because no studies of the association of BTSe with milk yield and udder health have been conducted in Canada, and because of previous evidence indicating that Se intake by dairy cows may be lower than the current recommended level (8), we hypothesized that Se status may be associated with milk production parameters and measures of udder health. The objectives of this study were: 1) to determine the Se status in selected Canadian dairy herds by measuring BTSe; and 2) to evaluate the association between BTSe concentration and milk production parameters, BTSCC, and the probability of being a S. aureus-positive herd.
Materials and methods
A total of 63 dairy farms from Alberta (n = 15), Ontario (n = 3), and Québec (n = 28), which were considered as one region, and Atlantic Canada (n = 17) participating in the Canadian Bovine Mastitis Research Network (CBMRN) cohort study were selected as a stratified convenience sample to represent, as closely as possible, the Canadian commercial dairy farm population. Selection criteria, production data, and housing management have been described elsewhere (12). The selection process was based on the relative contribution of each region to Canadian milk production, and on strata of low, intermediate and high 12-month rolling average BTSCC in 2006, with cut-offs < 150 000, between 150 000 and 300 000, and > 300 000 cells/mL, respectively. The farms were grouped according to the dairy production system as follows: intensively managed herds with cows housed in tie-stall or free-stall barns over the year, and intensively managed herds with cows on pasture during the grazing season (between spring and summer). All farms had > 80% Holstein-Friesian cows, the cows were milked twice daily, and the farms subscribed to Dairy Herd Improvement (DHI) recording (12).
Sampling and data collection
Sampling procedures were discussed between the CBMRN personnel and dairy producers before the beginning of the study (12). Technicians from the CBMRN project aseptically collected bulk tank milk samples from the selected farms on a monthly basis. Immediately after collection, samples were frozen and sent to the Maritime Quality Milk laboratory of the University of Prince Edward Island (Charlottetown, PEI, Canada) for analysis. Samples were cultured for S. aureus, and then preserved with Bronopol and refrigerated or re-frozen in preparation for SCC analysis. An aliquot (5 mL) of the samples collected in April, July, October of 2007 and January 2008, which typically represented spring, summer, fall, and winter periods, respectively, was used to evaluate BTSe concentration. However, a full set of the seasonal samples from each herd was not always available for BTSe analysis (Table 1).
Table 1.
Breakdown by region and season of unavailable samples for the evaluation of bulk tank milk selenium concentration in Canadian dairy herds
| Region | Season | Available | Unavailable |
|---|---|---|---|
| Western Canada | Winter | 7 | 8 |
| Spring | 6 | 9 | |
| Summer | 7 | 8 | |
| Fall | 0 | 15 | |
| Ontario | Winter | 1 | 2 |
| Spring | 0 | 3 | |
| Summer | 3 | 0 | |
| Fall | 0 | 3 | |
| Québec | Winter | 28 | 0 |
| Spring | 22 | 6 | |
| Summer | 0 | 28 | |
| Fall | 0 | 28 | |
| Atlantic provinces | Winter | 11 | 6 |
| Spring | 17 | 0 | |
| Summer | 15 | 2 | |
| Fall | 6 | 11 |
Specific farm data, such as herd size (number of lactating cows), average days in milk (DIM), and daily milk yield corresponding to the study period (from March 2007 to February 2008) were obtained from the regional DHI organizations.
Laboratory analysis
For each bulk tank sample, an aliquot of 0.05 mL was plated on blood esculin agar to detect S. aureus (13). Plates were examined after 24 and 48 h of incubation, then colonies were counted when ≤ 10/plate (equivalent to ≤ 200 cfu/mL). Staphylococcus aureus was identified by Gram stain, a positive catalase test, and double-zone hemolysis on blood agar. Preserved samples were analyzed for SCC using a Foss 4000 cell counter (Foss Electric, Hillerød, Denmark).
Milk Se concentration in fresh samples was evaluated by graphite furnace atomic absorption spectroscopy (14). Readings were carried out using a spectrometer (PerkinElmer AAnalyst 800; PerkinElmer, Waltham, Massachusetts, USA), reporting milk Se concentration as micromoles per liter (μmol/L). Herd Se status was interpreted using previously suggested cut-off points (5). Briefly, the reference range for BTSe was established by fitting a regression model of the mean herd serum Se concentration on the BTSe concentration in 15 dairy herds from PEI with widely differing BTSe concentrations. The cutoff points for BTSe were selected by calculating the point at which the reference values for serum Se from the peer-reviewed literature intersected the regression line of the model. Thus a BTSe concentration < 0.12 μmol/L was considered to represent deficiency, 0.20 μmol/L divided the marginal range equally into high- and low-marginal ranges, and 0.28 μmol/L or higher was taken to represent an adequate status.
Statistical analysis
A total of 192 samples were collected to evaluate BTSe; however, due to the uneven yearly distribution of samples within herds, a single annual BTSe value was estimated for each herd from the available data for the period between March 2007 and February 2008. Unconditional associations between independent and dependent variables were established. All independent variables unconditionally associated (P ≤ 0.10) with the dependent variables were included in a multivariable model which was manually reduced by backwards selection of the significant variables (P ≤ 0.05). The estimation was based on a multivariable linear mixed model for BTSe, including all independent variables: region, season [winter (December 21st to March 20th), spring, summer, and fall], housing system (tie-stall, free-stall, and bedding pack barns), and whether or not cows had access to pasture in the grazing season. Clustering within herds was accounted for by including herd random effects (15).
The dependent variables used in the statistical models were: yearly average of daily milk yield and BTSCC (log scale), as well as isolation of S. aureus in bulk tank milk samples during the study period (March 2007 to February 2008). Milk production parameters and the probability of being a S. aureus-positive herd were analyzed by fixed-effects linear and logistic regression models, with region and the single estimated BTSe value as predictors. A S. aureus-positive herd was defined by the isolation of S. aureus (≥ 1 cfu/0.05 mL) in 2 consecutive bulk tank milk samples collected no more than 2 mo apart during the study period (16).
All model assumptions were evaluated by examining the standardized residuals. A natural logarithmic transformation of SCC values (1000 cells/mL) was used to approximate the normal distribution. Analyses were carried out in SAS version 9.2 (SAS Institute, Cary, North Carolina, USA) using the MIXED, GLM, and LOGISTIC procedures.
Results
Most farms used a tie-stall barn system (57%), while free-stall and bedding-pack barn systems were used in 35% and 8% of the farms, respectively. Cows were housed throughout the year in 71% of the farms. Most of the farms (56%) where cows had access to pasture during the grazing season were located in Atlantic Canada. Some herd-level measures of productivity are presented in Table 2.
Table 2.
Measures of herd productivity based on DHI information for 63 Canadian dairy farms between March of 2007 and February of 2008
| Variable | Mean | s | Median | Range |
|---|---|---|---|---|
| Herd size | ||||
| Lactating cows | 73 | 42 | 59 | 30–251 |
| Dry cows | 13 | 8 | 11 | 3–39 |
| Culled cows | 3 | 2 | 3 | 1–15 |
| Average days in milk | 200 | 16 | 196 | 176–268 |
| Daily milk yield (kg/cow) | 31.4 | 3.1 | 31.2 | 23.8–40.4 |
| Milk fat (%) | 3.7 | 0.2 | 3.7 | 2.4–4.2 |
| Milk protein (%) | 3.2 | 0.1 | 3.2 | 3.0–3.4 |
| BTSCCa (1000 cells/mL) | 219 | 77 | 211 | 81–416 |
| Calving to first service (days) | 75 | 19 | 72 | 26–166 |
| Calving to conception (days) | 129 | 29 | 130 | 35–221 |
Bulk tank somatic cell count.
s — standard deviation.
No BTSe values compatible with Se deficiency were found; however, 8 samples (4%) were considered marginal (BTSe < 0.28 μmol/L). These marginal values were found in 1 farm from Québec, and in 7 farms from Atlantic Canada; all of them used a tie-stall barn with cows having access to pasture during the grazing season. Five of the 8 BTSe marginal values were found in the summer, 2 in the spring, and 1 in the winter.
Bulk tank milk Se concentration tended to differ among the 3 Canadian regions (Table 3). Bulk tank milk Se significantly differed between seasons, housing systems, and grazing. Mean BTSe was significantly lower in the summer, in farms using tie-stall barns, and in farms where lactating cows had access to pasture (Table 3).
Table 3.
Final multivariable regression model of the association between bulk tank milk selenium concentration and region, season, housing system, and whether or not cows had access to pasture in 63 dairy farms from 3 Canadian regions
| Effect | Coefficient | 95% CI | P-value |
|---|---|---|---|
| Intercept | −0.53 | 0.38, 0.68 | <0.01 |
| Region | 0.10 | ||
| Atlantic Canada | −0 | ||
| Alberta | −0.01 | −0.09, 0.06 | |
| Ontario/Québec | −0.07 | −0.14, −0.00 | |
| Season | < 0.01 | ||
| Fall | 0 | ||
| Winter | 0.11 | −0.01, 0.23 | |
| Spring | 0.16 | 0.03, 0.28 | |
| Summer | 0.03 | −0.10, 0.16 | |
| Housing system | <0.01 | ||
| Bedding pack | 0 | ||
| Tie-stall | −0.15 | −0.26, −0.05 | |
| Free-stall | −0.20 | −0.30, −0.09 | |
| Grazing | 0.04 | ||
| No | 0 | ||
| Yes | −0.07 | −0.14, 0.00 |
CI — confidence interval.
The estimated yearly BTSe values were lower than the observed values (Table 4). Most of the bulk tank samples from dairy farms in Alberta and Ontario/Québec were collected in the winter and spring (Table 1), corresponding to the time of the year with higher BTSe values; therefore, the estimated yearly average values were lower than the observed averages. This difference was less marked in Atlantic Canada where herds had BTSe data collected in the summer and fall.
Table 4.
Average bulk tank milk selenium concentration (BTSe in μmol/L) and yearly mean BTSe estimated for 63 dairy farms from 3 Canadian regions
| BTSe observed | BTSe estimated | ||||
|---|---|---|---|---|---|
| Region | Number of farms | Mean | s | Mean | s |
| Alberta | 15 | 0.53 | 0.17 | 0.45 | 0.09 |
| Ontario/Québec | 31 | 0.49 | 0.12 | 0.38 | 0.04 |
| Atlantic Canada | 17 | 0.52 | 0.18 | 0.47 | 0.08 |
s — standard deviation.
There was an unconditional (univariable) positive association between mean daily milk yield and BTSe (Table 5). Daily milk yield increased 12.3 kg for each unit (μmol/L) increase in BTSe (P = 0.01). However, there was no association between BTSe and other milk production parameters, such as milk fat and milk protein (Table 5). The association between BTSe and milk production parameters was adjusted by geographical region, leading to substantial changes in the coefficients (Table 5), and thus suggesting confounding effects of region on the relationship between BTSe and production.
Table 5.
Unconditional associations and multivariable model of estimated bulk tank milk selenium concentration (BTSe in μmol/L) and region with yearly average of milk production parameters in 63 dairy farms from 3 Canadian regions
| Daily milk yield (kg/d) | Milk fat (%) | Milk protein (%) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Coefficient | 95% CI | P-value | Coefficient | 95% CI | P-value | Coefficient | 95% CI | P-value | |
| Unconditional associationsa | |||||||||
| BTSe | 12.3 | 2.9, 21.8 | 0.01 | −0.15 | −0.9, 0.6 | 0.68 | −0.16 | −0.45, 0.12 | 0.26 |
| Region | 0.05 | < 0.01 | 0.02 | ||||||
| Alberta | 0 | 0 | 0 | ||||||
| Ontario/Québec | −2.9 | −4.8, −1.1 | 0.19 | 0.05, 0.32 | 0.05 | −0.00, 0.11 | |||
| Atlantic Canada | −1.2 | −3.2, 0.9 | 0.01 | −0.14, 0.16 | −0.02 | −0.08, 0.04 | |||
| Multivariable model | |||||||||
| Intercept | 30.0 | 25.0, 35.1 | < 0.01 | 3.35 | 2.98, 3.72 | < 0.01 | 3.16 | 3.01, 3.31 | < 0.01 |
| BTSe | 7.0 | −3.8, 17.8 | 0.20 | 0.62 | −0.18, 1.14 | 0.13 | 0.08 | −0.25, 0.40 | 0.64 |
| Region | 0.05 | < 0.01 | 0.03 | ||||||
| Alberta | 0 | 0 | 0 | ||||||
| Ontario/Québec | 1.1 | −4.4, −0.5 | 0.23 | 0.08, 0.37 | 0.06 | −0.00, 0.12 | |||
| Atlantic Canada | −1.3 | −3.3, 0.7 | 0.00 | −0.15, 0.15 | −0.02 | −0.08, 0.04 | |||
CI — confidence interval.
Univariable results were shown only to illustrate the confounding effects of region.
Unconditionally, bulk milk Se concentration had no effect on BTSCC (P = 0.29), while regional differences were found (Table 6). Farms in Ontario/Québec had a higher mean BTSCC than did farms in Alberta (P = 0.05) and the Atlantic region (P = 0.03). No differences were observed between Alberta and Atlantic regions (P = 0.91). The adjustment of the association between BTSe and BTSCC by the effect of geographical region showed a substantial change in the coefficient, suggesting confounding effects of region on that relationship (Table 6).
Table 6.
Unconditional association and multivariable model of estimated bulk tank milk selenium concentration (BTSe in μmol/L) and region with log bulk tank milk somatic cell count (BTSCC) in 63 dairy farms from 3 Canadian regions
| BTSCC (1000 cells/mL) | |||
|---|---|---|---|
| Coefficient | 95% CI | P-value | |
| Unconditional associationsa | |||
| BTSe | −0.62 | −1.78, 0.54 | 0.29 |
| Region | 0.05 | ||
| Alberta | 0 | ||
| Ontario/Québec | 0.22 | 0.00, 0.44 | |
| Atlantic Canada | −0.01 | −0.26, 0.23 | |
| Multivariable model | |||
| Intercept | 5.09 | 4.46, 5.72 | < 0.01 |
| BTSe | 0.19 | −1.14, 1.53 | 0.77 |
| Region | 0.08 | ||
| Alberta | 0 | ||
| Ontario/Québec | 0.23 | −0.01, 0.47 | |
| Atlantic Canada | −0.02 | −0.27, 0.23 | |
CI — confidence interval.
Univariable results were shown only to illustrate the confounding effects of region.
Staphylococcus aureus was found at least once in 47 (75%) herds during the study period. Bulk tank milk Se concentration was unconditionally associated with the odds of being a S. aureus-positive herd (P = 0.002). When BTSe increases, the odds of being a S. aureus-positive herd are reduced (Table 7). After controlling for region, the protective effect of BTSe was slightly reduced, but remained significant. Higher values of BTSe were associated with lower risk of being a S. aureus-positive herd (Table 7).
Table 7.
Unconditional associations and final multivariable model of estimated bulk tank milk selenium concentration (BTSe in μmol/L) and region with the probability of being a Staphylococcus aureus-positive herd
| Odds ratio | 95% CI | P-value | |
|---|---|---|---|
| Unconditional associations | |||
| BTSea | 0.035 | 0.004, 0.288 | < 0.01 |
| Region | < 0.01 | ||
| Alberta | 1 | ||
| Ontario/Québec | 21.8 | 3.7, 127.2 | |
| Atlantic Canada | 2.8 | 0.7, 11.5 | |
| Multivariable model | |||
| BTSe | 0.054 | 0.004, 0.704 | 0.03 |
| Region | 0.02 | ||
| Alberta | 1 | ||
| Ontario/Québec | 11.2 | 1.7, 72.0 | |
| Atlantic Canada | 5.0 | 0.9, 28.4 | |
CI — confidence interval.
Odds ratio — calculated for an increase of 0.20 μmol/L in BTSe.
Discussion
Bulk tank milk Se concentration responds rapidly to changes in Se intake, and has been satisfactorily used as a screening test for evaluating lactating cows Se adequacy (5). In the United States mean milk Se concentrations in cattle ranged from 0.09 to 1.11 μmol/L (17). In Canada, mean cow-level milk Se concentration was 0.35 μmol/L (18), ranging between 0.23 and 0.34 μmol/L in herd-level studies (5). In the present study, 8 of 192 samples had BTSe < 0.28 μmol/L; 7 of these samples were from farms in Atlantic Canada. Selenium-marginal values may be related to low Se concentrations found in forages and crops grown in the Atlantic provinces, where approximately 90% of forage samples contain < 0.05 ppm of Se (2), which is insufficient to meet the Se requirement in cattle. No BTSe values compatible with Se deficiency were found.
These results must be interpreted with caution because Se supplementation practices have changed markedly since earlier studies established the relationships between concentrations of Se in blood and milk. Earlier data were derived from herds where the principal or only source of supplementary Se was inorganic (for example, sodium selenite). Organic sources of Se have been widely used in North American dairy herds in recent years. At the recommended rate of Se supplementation (0.3 mg/kg DM), replacing selenite with organic Se will almost double the BTSe (19), while increasing plasma or serum Se concentrations only 10% to 20%. Milk Se concentrations observed in our study are generally higher than those previously reported in North America (5,17,20), supporting the argument that more organic Se is now being fed. The observation of marginal Se status in Atlantic Canada herds may reflect supplementation practices rather than the Se content of the base feeds. Thus, in interpreting these results, we should acknowledge that the relationships between Se intake, milk Se concentration, and Se status (based on established reference ranges for Se in blood) would differ between herds depending on the source of supplementary Se (organic or inorganic) being fed (19). It was beyond the scope of this study to collect information on Se supplementation practices in the study herds.
Seasonal variations in milk Se concentrations have been found, with greater risk of finding Se deficiency in the fall and winter compared with other seasons (5). Nevertheless, our results concur with previous studies in the United States (20), which found marginal Se values in the summer. A seasonal variation in Se concentration in pastures, and differences in the amount of concentrate fed to dairy cows may partially explain the seasonal changes found in BTSe.
No association between BTSe and yearly average of daily milk yield was found in our study after adjusting for the confounding effect of region. Milk yield can be a sensitive indicator of Se status, and reduced milk yield may be the most important economic consequence of marginal Se status (3). However, our results concur with former observational studies (21,22), and Se supplementation trials (23,24) which found no association between Se status and milk yield parameters. Contrary to our results, in New Zealand milk yield and milk fat were higher in Se-adequate herds (blood Se > 0.15 μmol/L) compared with Se-deficient herds; additionally, milk yield in response to Se supplementation was higher when Se levels were low before supplementation (4,25). Mean BTSe in the herds of our study was higher compared with milk Se concentration in other studies where a positive response to Se supplementation was found, which may explain the lack of association between BTSe and milk yield.
There was no association between BTSe and BTSCC in our study, which concurred with results from Norway (21), New Zealand (26), and Canada (5). Nevertheless, other studies have reported a negative association between Se status and BTSCC in herds with geometric mean BTSCC > 750 000 cells/mL (9) or > 250 000 cells/mL (10). The dairy population in our study had a Se status much higher than that in studies where a negative association between Se and udder health was first reported. A high BTSCC might induce the farmer to intentionally supplement Se at a higher rate resulting in a higher BTSe. This latter situation has resulted in observations of weak relationships between impaired udder health and Se status (22). Moreover, geometric mean BTSCC in our dairy population was also much lower (196 000 cells/mL) than that in dairy herds 25 years ago, when a negative association between Se status and udder health was first reported.
Higher BTSe levels were negatively associated with the odds of being a S. aureus-positive herd in our study. For example, increasing BTSe by 0.2 μmol/L reduced the odds by a factor of 0.95. Earlier studies found that an adequate Se status was associated with a decreased prevalence of mastitis pathogens, and lower BTSCC (9,10). Further, a protective effect of supplementing minerals, especially Se, on udder health and leukocyte function has been found (27). Even though biomarkers of Se status (selenoproteins with specific functions, for example) were not evaluated in our study, a reduction in the odds of being a S. aureus-positive herd can be explained by the biological function of Se on the immune system.
The effect of Se against mammary pathogens is mediated through several mechanisms: more rapid and massive influx of polymorphonuclear cells (PMN) to the udder (28), bacteria (such as S. aureus) being killed more efficiently by PMN (29), high anti-bacterial activity in whey sufficient to inhibit S. aureus growth (30), and high expression of selenoproteins with antioxidant properties in the mammary gland (31). These mechanisms, acting together in Se-adequate cows, enhance the immune response of the udder against mammary pathogens, reducing their herd prevalence.
Current mean BTSe was higher than that previously reported for Atlantic Canadian dairy herds (5). The effectiveness of Se supplementation practices may have increased over time (22). In addition, herds with a higher BTSCC (and higher odds of being S. aureus-positive) are often supplemented with Se, or provided with Se at a higher rate, as a management intervention to control BTSCC.
The results of this study largely support the contention that, whereas supplementing cows consuming Se-deficient diets may elicit a favorable production response or a reduction in disease prevalence, super-supplementation of Se-adequate diets may not result in additional benefits to animal performance (6). Although no herds were considered to be Se-deficient based on accepted reference ranges for Se status, a reduction in the odds of being a S. aureus-positive herd was observed in this study in herds with the highest BTSe concentrations. This latter finding suggests that feeding cows to increase the milk Se concentration (for example using organic Se) in excess of that currently considered adequate for support of growth and milk production, may optimize udder defense against S. aureus infection. Further study is required to confirm this hypothesis.
Acknowledgments
The Canadian Bovine Mastitis Research Network (CBMRN) and the Atlantic Veterinary College Research Fund funded this research. Authors wish also to thank to Dr. Kristen Reyher, Dr. Raphaël Vanderstichel, Natasha Robinson, Orysia Dawydiak, and Doris Poole from the University of Prince Edward Island; and Dr. Simon Dufour from the University of Montreal for their valuable help. We would like to also thank all technicians who collected the samples. CVJ
Footnotes
Use of this article is limited to a single copy for personal study. Anyone interested in obtaining reprints should contact the CVMA office (hbroughton@cvma-acmv.org) for additional copies or permission to use this material elsewhere.
References
- 1.Fenimore RL, Adams DS, Puls R. Selenium levels of beef cattle in southeastern British Columbia relative to supplementation and type of pasture. Can Vet J. 1983;24:41–45. [PMC free article] [PubMed] [Google Scholar]
- 2.Winter KA, Gupta UC. Selenium content of forages grown in Nova Scotia, New Brunswick, and Newfoundland. Can J Anim Sci. 1979;59:107–111. [Google Scholar]
- 3.Wichtel JJ. A review of selenium deficiency in grazing ruminants. Part 2: Towards a more rational approach to diagnosis and prevention. N Z Vet J. 1998;46:54–58. doi: 10.1080/00480169.1998.36056. [DOI] [PubMed] [Google Scholar]
- 4.Fraser AJ, Ryan TJ, Sproule R, Clark RG, Anderson D, Pederson EO. The effect of selenium supplementation on milk production in dairy cattle. Proc N Z Soc Anim Prod. 1987;47:61–64. [Google Scholar]
- 5.Wichtel JJ, Keefe GP, Van Leeuwen JA, Spangler E, McNiven MA, Ogilvie TH. The selenium status of dairy herds in Prince Edward Island. Can Vet J. 2004;45:124–132. [PMC free article] [PubMed] [Google Scholar]
- 6.Silvestre FT, Rutigliano HM, Thatcher WW, Santos JE-P, Staples CR. Effect of selenium source on production, reproduction, and immunity of lactating dairy cows. Fla Rumin Nutr Symp. 2007:55–67. [Google Scholar]
- 7.Smith K, Hogan J, Weiss W. Dietary vitamin E and selenium affect mastitis and milk quality. J Anim Sci. 1997;75:1659–1665. doi: 10.2527/1997.7561659x. [DOI] [PubMed] [Google Scholar]
- 8.Subcommittee on Dairy Cattle Nutrition, Committee on Animal Nutrition, National Research Council. Nutrient Requirements of Dairy Cattle. 7th ed. Washington DC: National Academy Press; 2001. [Google Scholar]
- 9.Erskine RJ, Eberhart RJ, Hutchinson LJ, Scholz RW. Blood selenium concentrations and glutathione peroxidase activities in dairy herds with high and low somatic cell counts. J Am Vet Med Assoc. 1987;190:1417–1421. [PubMed] [Google Scholar]
- 10.Weiss WP, Hogan JS, Smith KL, Hoblet KH. Relationships among selenium, vitamin E, and mammary gland health in commercial dairy herds. J Dairy Sci. 1990;73:381–390. doi: 10.3168/jds.S0022-0302(90)78684-5. [DOI] [PubMed] [Google Scholar]
- 11.Olde Riekerink RG, Barkema HW, Scholl DT, Poole DE, Kelton DF. Management practices associated with the bulk-milk prevalence of Staphylococcus aureus in Canadian dairy farms. Prev Vet Med. 2010;97:20–28. doi: 10.1016/j.prevetmed.2010.07.002. [DOI] [PubMed] [Google Scholar]
- 12.Reyher KK, Dufour S, Barkema HW, et al. The National Cohort of Dairy Farms — A data collection platform for mastitis research in Canada. J Dairy Sci. 2011;94:1616–1626. doi: 10.3168/jds.2010-3180. [DOI] [PubMed] [Google Scholar]
- 13.Hogan JS, Gonzalez RN, Harmon RJ, et al. Laboratory Handbook on Bovine Mastitis. Verona, Wisconsin: Natl. Mastitis Counc. Inc; 1999. revised edition. [Google Scholar]
- 14.Oster O, Prellwitz W. A methodological comparison of hydride and carbon furnace atomic absorption spectroscopy for the determination of selenium in serum. Clin Chim Acta. 1982;124:277–291. doi: 10.1016/0009-8981(82)90422-3. [DOI] [PubMed] [Google Scholar]
- 15.Dohoo IR, Martin W, Stryhn H. Veterinary Epidemiologic Research. 2nd ed. Charlottetown, Prince Edward Island: VER Inc; 2009. p. 865. [Google Scholar]
- 16.Jayarao BM, Pillai SR, Sawant AA, Wolfgang DR, Hegde NV. Guidelines for monitoring bulk tank milk somatic cell and bacterial counts. J Dairy Sci. 2004;87:3561–3573. doi: 10.3168/jds.S0022-0302(04)73493-1. [DOI] [PubMed] [Google Scholar]
- 17.Sanz M, Diaz C. Selenium concentration in milks. Food Chem. 1995;52:1–18. [Google Scholar]
- 18.Fisher LJ, Hoogendoorn C, Montemurro J. The effect of added dietary selenium on the selenium content of milk, urine and feces. Can J Anim Sci. 1980;60:79–86. [Google Scholar]
- 19.Ceballos A, Sanchez J, Stryhn H, Montgomery JB, Barkema HW, Wichtel JJ. Meta-analysis of the effect of oral selenium supplementation on milk selenium concentration in cattle. J Dairy Sci. 2009;92:324–342. doi: 10.3168/jds.2008-1545. [DOI] [PubMed] [Google Scholar]
- 20.Miller GY, Bartlett PC, Erskine RJ, Smith KL. Factors affecting serum selenium and vitamin E concentrations in dairy cows. J Am Vet Med Assoc. 1995;206:1369–1373. [PubMed] [Google Scholar]
- 21.Kommisrud E, Osterås O, Vatn T. Blood selenium associated with health and fertility in Norwegian dairy herds. Acta Vet Scand. 2005;46:229–240. doi: 10.1186/1751-0147-46-229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Enjalbert F, Lebreton P, Salat O. Effects of copper, zinc and selenium status on performance and health in commercial dairy and beef herds: Retrospective study. J Anim Physiol Anim Nutr. 2006;90:459–466. doi: 10.1111/j.1439-0396.2006.00627.x. [DOI] [PubMed] [Google Scholar]
- 23.Coe PH, Maas J, Reynolds J, Gardner I. Randomized field trial to determine the effects of oral selenium supplementation on milk production and reproductive performance of Holstein heifers. J Am Vet Med Assoc. 1993;202:875–881. [PubMed] [Google Scholar]
- 24.Grace ND, Ankenbauer-Perkins KL, Alexander AM, Marchant RM. Relationship between blood selenium concentration or glutathione peroxidase activity, and milk selenium concentrations in New Zealand dairy cows. N Z Vet J. 2001;49:24–28. doi: 10.1080/00480169.2001.36198. [DOI] [PubMed] [Google Scholar]
- 25.Ellison RS. A review of copper and selenium reference ranges in cattle and sheep. Vet Cont Educ Massey Univ. 1992;145:3–27. [Google Scholar]
- 26.Grace N, Knowles S, Lee J. Relationships between blood Se concentrations and milk somatic cell counts in dairy cows. N Z Vet J. 1997;45:171–172. doi: 10.1080/00480169.1997.36020. [DOI] [PubMed] [Google Scholar]
- 27.Piepers S, Opsomer G, Meyer E, et al. Heifer and quarter characteristics associated with periparturient blood and milk neutrophil apoptosis in healthy heifers and in heifers with subclinical mastitis. J Dairy Sci. 2009;92:4330–4339. doi: 10.3168/jds.2009-2029. [DOI] [PubMed] [Google Scholar]
- 28.Erskine RJ, Eberhart RJ, Grasso PJ, Scholz RW. Induction of Escherichia coli mastitis in cows fed selenium-deficient or selenium-supplemented diets. Am J Vet Res. 1989;50:2093–2100. [PubMed] [Google Scholar]
- 29.Erskine RJ, Eberhart RJ, Scholz RW. Experimentally induced Staphylococcus aureus mastitis in selenium-deficient and selenium-supplemented dairy cows. Am J Vet Res. 1990;51:1107–1111. [PubMed] [Google Scholar]
- 30.Ali-Vehmas T, Vikerpuur M, Fang W, Sandholm M. Giving selenium supplements to dairy cows strengthens the inflammatory response to intramammary infection and induces a growth-suppressing effect on mastitis pathogens in whey. Zentralbl Veterinarmed A. 1997;44:559–571. doi: 10.1111/j.1439-0442.1997.tb01142.x. [DOI] [PubMed] [Google Scholar]
- 31.Sordillo LM, O’Boyle N, Gandy JC, Corl CM, Hamilton E. Shifts in thioredoxin reductase activity and oxidant status in mononuclear cells obtained from transition dairy cattle. J Dairy Sci. 2007;90:1186–1192. doi: 10.3168/jds.S0022-0302(07)71605-3. [DOI] [PubMed] [Google Scholar]