Prediction of likelihood of conception in dairy cows using milk mid-infrared spectra collected before the first insemination and machine learning algorithms

Accurate and ex-ante prediction of cows' likelihood of conception (LC) based on milk composition information could improve reproduction management on dairy farms. Milk composition is already routinely measured by mid-infrared (MIR) spectra, which are known to change with advancing stages of pregnancy. For lactating cows, MIR spectra may also be used for predicting the LC. Our objectives were to classify the LC at first insemination using milk MIR spectra data collected from calving to first insemination and to identify the spectral regions that contribute the most to the prediction of LC at first insemination. After quality control, 4,866 MIR spectra, milk production, and reproduction records from 3,451 Holstein cows were used. The classification accuracy and area under the curve (AUC) of 6 models comprising different predictors and 3 machine learning methods were estimated and compared. The results showed that partial least square discriminant analysis (PLS-DA) and random forest had higher prediction accuracies than logistic regression. The classification accuracy of good and poor LC cows and AUC in herd-by-herd validation of the best model were 76.35 ± 10.60% and 0.77 ± 0.11, respectively. All wavenumbers with values of variable importance in the projection higher than 1.00 in PLS-DA belonged to 3 spectral regions, namely from 1,003 to 1,189, 1,794 to 2,260, and 2,300 to 2,660 cm-1. In conclusion, the model can predict LC in dairy cows from a high productive TMR system before insemination with a relatively good accuracy, allowing farmers to intervene in advance or adjust the insemination schedule for cows with a poor predicted LC.

Evaluation of breeding objectives, breeding practices and reproductive performance of indigenous dairy cows in selected districts of Kaffa Zone, South West Ethiopia

BACKGROUND

Breeding objectives are designed to achieve targeted dairy cow production goals, which can be affected by production type, farmer preferences, environmental factors and genetic factors individually or in combination. Breeding practices, such as both controlled and uncontrolled, and artificial insemination (AI) are the tools used to obtain the desired breeding objectives. The lower reproductive performance of indigenous dairy cows affects the total milk production and calf crops that are produced during their lifetime. Designing appropriate breeding objectives and breeding practices can improve the reproductive performance of dairy cows and their overall production performance.

MATERIALS AND METHODS

The current study was conducted with the objective of evaluating the breeding, practices and performance of indigenous dairy cattle in the south western part of Ethiopia. The districts of Gesha and Chena were purposefully chosen. The study design for the 384 household surveys was a cross-sectional survey with a simple random sample approach. Data analysis was carried out by MS-Excel (2010) and the general linear model procedure of SAS of 2008.

RESULTS

The current study revealed that methods of breeding were predominantly natural-controlled mating, followed by natural-uncontrolled mating and AI in descending order. Breeding objectives were input function, output function, sociocultural and economic functions and assets and security functions in decreasing order of rank. Reproduction performance indexes of indigenous dairy cows age at first service (3.72 ± 0.05 years), age at first calving (AFC) (4.71 ± 0.07 years), calving interval (CI) (1.58 ± 0.03 years), days open (DO) (4.26 ± 0.11 months), services per conception in natural mating (1.4 ± 0.08) and AI (2.73 ± 0.14), age of bull at maturity (4.17 ± 0.74 years), interoestrus interval (23.18 ± 0.61 days), calves crop (7.53 ± 0.22) and the life span of indigenous dairy cow (11.94 ± 0.26 years) were significant (p < 0.01) between two districts, whereas the values of age of bull at maturity and number of services per conception in natural mating were significant (p < 0.05) between districts.

CONCLUSIONS

Using AI and major reproduction performances, such as AFC, CI and DO of indigenous dairy cows in the study area, were very low. Therefore, concerned bodies should intervene to improve reproduction performance through the utilization of AI techniques, with the integration of forage development activities and improvements in livestock health care.

Predicting Pregnancy Outcome in Dairy Cows: The Role of IGF-1 and Progesterone

The purpose of this study was to determine the link between insulin-like growth factor 1 (IGF-1), progesterone (PROG), non-esterified fatty acids (NEFAs), β-hydroxybutyrate (BHB), and glucose (GLU) and pregnancy probability after the first artificial insemination (AI) and during the first 100 days in milk (DIM), during the critical transition period. We determined levels of serum IGF-1, PROG, NEFA, BHB, and GLU in Holstein dairy cows via ELISA, using blood samples collected 7 days before parturition (DAP) until 21 days postparturition (DPP). The group was split into cows diagnosed pregnant at 100 DIM (PREG) and those that did not conceive at 100 and 150 DIM (NPREG). Serum IGF-1 and PROG median levels at 7 DAP were significantly higher in PREG vs. NPREG (p = 0.029), the only statistically significant differences across the subgroups. At 7 DAP, IGF-1 levels within the initial group showed a strong negative correlation with PROG (r = -0.693; p = 0.006), while for the PREG subgroup, the IGF-1 levels exhibited a very strong positive correlation with GLU (r = 0.860; p = 0.011) and NEFA (r = 0.872; p = 0.013). IGF-1 and PROG levels detected at 7 DAP may be useful to predict pregnancy at 100 DIM. The positive correlation of NEFA and GLU levels during the transition period demonstrates that the initial group is not in NEB; thus, the NEFA level was not a decisive factor for reproduction success.

Changes in metabolic and hormonal profiles during transition period in dairy cattle - the role of spexin

BACKGROUND

This study aimed to evaluate spexin as a novel blood marker and to describe the relationship of this peptide with selected biochemical metabolites measured during the transition period in dairy cows. Additionally, mRNA expression of the spexin gene as well as spexin receptors - galanin receptor type 2 and galanin receptor type 3, was investigated in several bovine tissues. Blood samples were collected at weekly intervals starting at 21 days before the estimated parturition day until 21 days in milk to determine concentrations of spexin, nonesterified fatty acids, β-hydroxybutyrate acid, total and active ghrelin, progesterone, glucose, insulin, IGF-I, triglycerides, cholesterol, leptin, corticosterone and 17-β-estradiol as well as the activity of aspartate transaminase, alkaline phosphatase and gamma-glutamyl transferase.

RESULTS

Spexin concentration decreased from 21 d before parturition to calving day and next it rose during the first 14 d of lactation. The lowest concentration of spexin was recorded on the calving day and it differed from the mean level of this peptide before parturition as well as postpartum. Moreover, differences were observed between mean spexin concentrations before and after calving. Spexin levels were moderately negatively correlated with NEFA (r = - 0.39) and total ghrelin contents (r = - 0.41), weakly correlated with BHBA (r = - 0.35) while they showed a moderate positive relationship with progesterone concentrations (r = 0.42). Moreover, we detected that mRNA expression of GALR2, GALR3 and SPX is present in various bovine tissues (kidney, bowel, rumen, spinal cord, lung, skeletal muscle, liver, heart, fat and spleen).

CONCLUSION

A negative correlation between spexin concentration and NEFA, BHBA and total ghrelin contents as well as a positive relationship with levels of progesterone, metabolites and hormones, which are key players in the dairy cow transition period, may confirm an important function of this peptide in metabolism regulation. Thus measurement of spexin concentration could provide useful supplementary information for dairy cow herd health monitoring.

Genetic parameters for reproductive losses estimated from in-line milk progesterone profiles in Swedish dairy cattle

This study assessed the extent of reproductive losses and associated genetic parameters in dairy cattle, using in-line milk progesterone records for 14 Swedish herds collected by DeLaval's Herd Navigator. A total of 330,071 progesterone samples were linked to 10,219 inseminations (AI) from 5,238 lactations in 1,457 Swedish Red and 1,847 Swedish Holstein cows. Pregnancy loss traits were defined as early embryonic loss (1-24 d after AI), late embryonic loss (25-41 d after AI), fetal loss (42 d after AI until calving), and total pregnancy loss (from d 1 after AI until calving). The following classical fertility traits were also analyzed: interval from calving to first service, interval from calving to last service, interval between first and last service, calving interval, and number of inseminations per service period. Least squares means with standard error (LSM ± SE), heritabilities, and genetic correlations were estimated in a mixed linear model. Fixed effects included breed, parity (1, 2, ≥3), estrus cycle number when the AI took place, and a linear regression on 305-d milk yield. Herd by year and season of AI, cow, and permanent environmental effect were considered random effects. Extensive (approximately 45%) early embryonic loss was found, but with no difference between the breeds. Swedish Red was superior to Swedish Holstein in the remaining pregnancy loss traits with, respectively: late embryonic loss of 6.1 ± 1.2% compared with 13.3 ± 1.1%, fetal loss of 7.0 ± 1.2% compared with 12.3 ± 1.2%, and total pregnancy loss of 54.4 ± 1.4% compared with 60.6 ± 1.4%. Swedish Red also had shorter calving to first service and calving to last service than Swedish Holstein. Estimated heritability was 0.03, 0.06, and 0.02 for early embryonic, late embryonic, and total pregnancy loss, respectively. Milk yield was moderately genetically correlated with both early and late embryonic loss (0.52 and 0.39, respectively). The pregnancy loss traits were also correlated with several classical fertility traits (-0.46 to 0.92). In conclusion, Swedish Red cows had lower reproductive loss during late embryonic stage, fetal stage, and in total, and better fertility than Swedish Holstein cows. The heritability estimates for pregnancy loss traits were of the same order of magnitude as previously reported for classical fertility traits. These findings could be valuable in work to determine genetic variation in reproductive loss and its potential usefulness as an alternative fertility trait to be considered in genetic or genomic evaluations.

Predicting the likelihood of conception to first insemination of dairy cows using milk mid-infrared spectroscopy

The objective of this study was to examine the ability of milk mid-infrared (MIR) spectroscopy and other on-farm data, such as milk yield, milk composition, stage of lactation, calving age, days in milk at insemination, and somatic cell count, to identify cows that were most or least likely to conceive to first insemination. A total of 16,628 spectral and milk production records of 7,040 cows from 29 commercial dairy herds across 3 Australian states were used. Three models, comprising different explanatory variables, were tested. Model 1 included features that are readily available on farms participating in milk recording, such as milk yield, milk composition, somatic cell count, days from calving to insemination, and calving season. Days in milk and age at calving were incorporated into model 1 to form model 2. In model 3, MIR was added to model 2, but to avoid double counting, milk composition traits of model 2 were removed. The models were first trained on extreme data [i.e., including cows that (1) conceived to first insemination and (2) cows with no conception event recorded and with only 1 insemination]. Then, the models were validated in a fresh data set with all cows regardless of conception outcomes present to test for their ability to identify cows that conceived or did not conceive to first insemination. To do this, we ranked the predicted probability of all cows in the validation set and then selected the top and bottom records in varying proportions from 5 to 40% (i.e., where the model predicted the highest versus lowest likelihood of conception to first insemination, respectively) and compared with the actual values. The model's performance was evaluated through herd-year by herd-year external validation and measured as the proportion of selected records being correct. The results show that when more cows are selected (i.e., descending confidence), the accuracy of the models was reduced, and selecting the 10% of cows with the highest confidence of predictions produces optimal accuracy. Irrespective of the proportions, none of the models could predict cows that conceived to first insemination, with an accuracy around 0.48. When attempting to predict the bottom 10% of cows, which had the least likelihood of conception to first insemination, model 1 had prediction accuracy around 0.64. Compared with model 1, the addition of days in milk and calving age (model 2) resulted in a negligible improvement in prediction accuracy (0.01 to 0.03). Model 3 had the highest prediction accuracy (0.76), which implies that in the models tested, MIR is of primary importance in the prediction of fertility of dairy cows. In conclusion, this study indicates that MIR and other milk recording data could be used to identify cows with potential difficulty in getting pregnant to first insemination with promising accuracy.

Extent and pattern of pregnancy losses and progesterone levels during gestation in Swedish Red and Swedish Holstein dairy cows

Background

Pregnancy loss is a major source of infertility in dairy cows. Despite a fertilization rate after insemination (AI) of approximately 90%, calving rates are 30%–50%, indicating the occurrence of extensive embryonic and foetal losses. The aim of this study was to establish the extent and pattern of embryonic and foetal loss in Swedish Red (SR) and Swedish Holstein (SH) dairy cows, as well as, the relationship to oestrus intensity (OI) and progesterone (P4) concentration. In total, 2130 AIs and 16,176 milk P4 samples from 359 SR and 212 SH dairy cows were included in the study. Pregnancy losses were estimated using data from P4 values combined with AI information and calving data.

Results

Total pregnancy loss from AI to the day of calving was 65%. Early embryonic loss, late embryonic loss and foetal loss were estimated to be 29, 14 and 13%, respectively. There is strong evidence in the literature that P4 concentrations at different time points are associated with pregnancy loss. In the present study, cows with pregnancy losses had significantly higher P4 levels at the day of AI and significantly lower P4 concentration at days 10, 21 and 30 after AI compared to pregnant cows. Swedish Red cows had significantly lower total pregnancy losses compared to SH cows (62% and 68% respectively, P = 0.017). Early embryonic loss was 6.7% points lower for cows inseminated at a stronger OI (OI = 3) compared to at a weaker OI (OI = 2, P = 0.006). Cows inseminated at ovulation number ≥ 5 had significantly lower early pregnancy losses compared to cows inseminated at first or second ovulation (11.5 and 8% points, respectively, P < 0.05). With an increase of one SD of milk (448 kg ECM) during the first 60 days in milk, early embryonic loss increased by 4.7% points (P = 0.006).

Conclusions

It is important to increase the number of cows calving per insemination by reducing embryo/foetal loss. This outcome can be achieved by management and breeding for optimal P4 levels at critical time points, and by considering oestrus expression in the breeding programmes to facilitate the correct timing of insemination.