Material & methods

Colostrum samples were obtained from Lacaune ewes from one commercial dairy farm (Granja Cerromonte S.L., Spain) with 4100 ewes and 450 l of annual average production per ewe. Animal management on this farm has been described by Hernandez et al. (Reference Hernandez, Elvira, Gonzalez-Martin, Gonzalez-Bulnes and Astiz2011). Briefly, sheep are housed indoors without access to pasture. No suckling period is allowed: lambs are separated from mothers at delivery and fed with an automatic lamb feeder after colostrum administration with oesophageal feeder. From the day after lambing, ewes are milked twice a day until production drops below 0·5 l/d or until 30 d before the next lambing, when they are dried off.

First milking colostrum was obtained from 536 ewes within the first 6 h post-delivery by machine milking and stored at −20 °C until analysis. Thawed samples were measured in triplicate [intra-assay coefficient of variation (CV), 0·98%] using a digital Brix refractometer (Pocket Pal-1, Atago, Spain), and the mean value for each sample was used in statistical analyses. Average colostral temperature during analysis was 24·2 °C (range, 22·7–26·4 °C).

Data on age and reproductive life were extracted from the electronic farm management system, and colostrum samples were stratified according to potential risk factors of poor quality: parturition year, delivery season (considering seasonal lambing, deliveries between December and August and counter-seasonal lambing, deliveries between September and November, in the North Hemisphere), parity, PDP-length and age at first lambing (AFL).

We analysed 142 colostra from lambings in 2011, 191 colostra from 2012 and 203 from 2013. The quality of samples from seasonal lambings (December-August, n = 344) was compared with that of samples from counter-seasonal lambings (September-November, n = 192). Ewes were primi- or multiparous with this distribution: first lambing, n = 117; second lambing, 116; third lambing, 87; fourth lambing, 84; and fifth lambing and beyond, 132. Quality was compared among colostrum samples from ewes showing different PDP lengths: no PDP (primiparous ewes), 117; short (0–30 d), 23; adequate (31–60 d), 214; long (61–90 d), 134; and extremely long (>90 d), 48. Finally, quality was compared among colostrum samples from primiparous ewes with different AFLs: <13 months, 88; 13–15 months, 20; and >15 months, 9.

Results and discussion

Our results provide some of the first detailed insights into factors affecting ovine colostrum using Brix refractometry on-farm. Brix refractometer measurements of colostrum quality across all samples are shown in Fig. 1. Most samples (336 of 536, 62·7%) showed a solids percentage >18%. The remaining 200 samples (37·3%) showed a solids percentage <18%. With just a few drops of colostrum, measurements can be obtained rapidly and with excellent repeatability; in the present study, mean CV for repeated measures was 0·41%. Our results support the robustness of Brix refractometry, since colostrum quality showed an expected distribution across all samples, similar to that observed in sheep and cattle (Alves et al. Reference Alves, Alves, Ascari, Junqueira, Coutinho, Lima, Pérez, De Paula, Furusho-Garcia and Abreu2015; Bartier et al. Reference Bartier, Windeyer and Doepel2015). However Brix refractometry in ovine colostrum is not a fully qualified method since solid percentage has not been yet correlated to actual IgG concentrations in ovine colostrum as it has already been done in bovine colostrum (Bartier et al. Reference Bartier, Windeyer and Doepel2015).

Fig. 1. Distribution of colostrum samples analysed with a Brix refractometer.

Parity significantly affected colostrum quality. Quality was highest in colostrum from primiparous ewes (22·6 ± 5·6%, P < 0·0001) and was lowest in colostrum from ewes with two lambings (18·8 ± 3·8%), rising gradually with three lambings (19·1 ± 4·8%), four (19·5 ± 5·9%) and finally five or more (19·7 ± 5·1%). This variation with number of lambings did not achieve statistical significance. Our results suggest that colostrum quality is highest at the first lambing, which contrasts with other studies reporting higher IgG concentration in sheep colostrum from multiparous than primiparous ewes (Chniter et al. Reference Chniter, Salhi, Harrabi, Khorchani, Lainé, Nowak and Hammadi2016) or no variances (Alves et al. Reference Alves, Alves, Ascari, Junqueira, Coutinho, Lima, Pérez, De Paula, Furusho-Garcia and Abreu2015). This discrepancy may reflect differences in breed, nutrition, method of analysis, and method of colostrum collection. In agreement with our results, Gilbert et al. (Reference Gilbert, Gaskins, Hillers and Parker1988) measured higher IgG concentration in colostrum from primiparous ewes. Our observation that colostrum quality was lowest after the second lambing may reflect characteristics of the Lacaune animals in our study (Hernandez et al. Reference Hernandez, Elvira, Gonzalez-Martin and Astiz2012). The first lactation in these animals lasts the longest and is the most productive.

Among multiparous sheep, mean dry period length was 65·1 ± 41·5 d. PDP length significantly affected colostrum quality. Quality was significantly higher in ewes at first lambing (22·5 ± 5·6%) and in ewes with extremely long PDP (20·9 ± 6·3%) than in other ewes (P < 0·0001). Quality was significantly lower in ewes with the shortest PDP (16·8 ± 4·2%) than in other ewes (P < 0·0001). Quality was intermediate in ewes with adequate PDP (18·7 ± 4·1%) or long PDP (19·9 ± 5·4%). PDP length showed weak positive correlation with Brix values (r = 0·235, P < 0·0001). Our results are consistent with the idea that colostrum quality increases with PDP length. This is consistent with reports that omitting the dry period leads to lower IgG concentration in colostrum from goats (Caja et al. Reference Caja, Salama and Such2006). One possible explanation is that selective transfer of IgG from the dam blood to colostrum begins in the udder a few weeks before parturition, and this transfer decreases if milk secretion is maintained (Guy et al. Reference Guy, McFadden, Cockrell and Besser1994).

We found similar colostrum quality between ewes with adequate PDP length (30–60 d) and those with long PDP (61–90 d). This is consistent with similar quality of cattle colostrum with PDPs of 30–40 d or the traditional 60 d (Cermakova et al. Reference Cermakova, Kudrna, Simeckova, Vyborna, Dolezal and Illek2014; Shoshani et al. Reference Shoshani, Rozen and Doekes2014; Mayasari et al. Reference Mayasari, de Vries, Nieuwland, Remmelink, Parmentier, Kemp and van Knegsel2015). Similar results have been observed in goats (Caja et al. Reference Caja, Salama and Such2006). Our data suggests even higher IgG concentration can be achieved in colostrum if PDP is prolonged beyond 90 d, but this is not advisable for production. Taken together, our results support the conclusion of Hernandez et al. (Reference Hernandez, Elvira, Gonzalez-Martin and Astiz2012) that a 30 d prepartum is the most appropriate drying period for milk production in Lacaune ewes reared under intensive conditions.

Mean AFL was 12·7 ± 1·7 months. Colostrum quality was similar among primiparous ewes with an AFL < 13 months (22·8 ± 5·8%), 13–15 months (21·9 ± 4·7%) or >15 months (22·2 ± 6·2%; P = 0·77). No significant correlation was observed between AFL and Brix-values. These results suggest that AFL does not influence colostrum quality and run counter to the idea that older ewes are likely to have experienced more immune challenges that increase IgG content in colostrum, as demonstrated for cows (Bartier et al. Reference Bartier, Windeyer and Doepel2015). In part, our negative result may reflect the narrow range of AFL (11·5–19 months) among our ewes. It may also reflect the intensive management conditions, which may reduce animal exposure to new immune challenges. Finally, this result suggests that our previous proposal of 13–15 months as optimal AFL for reproductive and productive performance of the Lacaune breed under intensive management conditions (Hernandez et al. Reference Hernandez, Elvira, Gonzalez-Martin, Gonzalez-Bulnes and Astiz2011) is reasonable.

Similarly, season did not affect colostrum quality: colostrum quality was 20·11 ± 5·0% in the seasonal lambing and 19·7 ± 5·7% in th; counter-seasonal lambing (P = 0·47). This is despite the fact that season has been shown to influence over- and under-nutrition as well as body condition (Gilbert et al. Reference Gilbert, Gaskins, Hillers and Parker1988; al-Sabbagh et al. Reference al-Sabbagh, Swanson and Thompson1995). Variations in energy intake during late pregnancy influence colostral IgG content (Castro et al. Reference Castro, Capote, Bruckmaier and Argüello2011). Our negative result probably reflects the intensive management that likely provides balanced nutrition throughout the year, independent of seasonal variations in food availability.

However, year of study affected colostrum quality: quality was significantly higher in 2011 (21·0 ± 5·2%) than in 2012 (19·8 ± 5·7%) or 2013 (19·5 ± 4·7%; P < 0·0001). Our observation of higher quality in 2011 may be due to the higher percentage of animals with very long PDP that year. In any event, we did not measure body condition score in the present study, so future work should examine whether this parameter may be associated with colostrum quality.

High-quality colostrum means sufficiently high IgG concentration, since newborns would be unable to absorb sufficient volumes of colostrum with low IgG concentrations. Unfortunately, our data do not allow us to determine appropriate Brix refractometer cut-off(s) to classify colostrum as adequate, and we are unaware of any such standards for sheep. On the basis of available evidence, we suggest threshold values of <18% (discard) and 21–22% (retain). With a cut-off of 22%, 164 of 536 of our samples (30·6%) would be classified as high-quality; with a cut-off of 20%, 253 of 536 (47·2%) would be so classified; with a cut-off of 18%, 336 of 536 (62·7%) would be so classified. Studies are needed to determine whether colostrum meeting our proposed cut-offs would provide the minimum of 30 g of IgG during the first 24 h of life needed for successful passive immune transfer in lambs (Alves et al. Reference Alves, Alves, Ascari, Junqueira, Coutinho, Lima, Pérez, De Paula, Furusho-Garcia and Abreu2015). The proportions of samples in our study meeting these cut-offs for high quality colostrum are smaller than the corresponding proportions reported for cattle (Bartier et al. Reference Bartier, Windeyer and Doepel2015). This may reflect loss of colostrum quality during the time from lambing to colostrum sampling: IgG concentration in sheep colostrum can drop by over 30% between 1 and 6 h post-delivery (Chniter et al. Reference Chniter, Salhi, Harrabi, Khorchani, Lainé, Nowak and Hammadi2016). Consistent with this possibility, mortality and morbidity of lambs bred artificially since birth were low throughout the study period.