Benefits of ultrasound scanning during gestation in the small ruminant

Evaluation of a Numerical, Real-Time Ultrasound Imaging Model for the Prediction of Litter Size in Pregnant Sows, with Machine Learning

Simple Summary

The use of a numeric model to quantify real-time ultrasonographic (RTU) imaging is a prominent methodfor predicting the expected litter size by training an artificial neural network (ANN) to minimize the error of the prediction measured by metrics, such as root square mean error and mean absolute error. Time of the RTU application is a critical factor for such a prediction. Rated scale values (RSV) obtained from the RTU images relate to the accurate diagnosis of pregnancy and of litter size, suggesting the potential of a generalized use of the model in various farm conditions. Ultimately, the employment of the model in machine learning for an automated prediction of litter size can be used as a routine on-farm procedure for the efficient management of gestating sows.

Abstract

The present study aimed to evaluate the accuracy of a numerical model, quantifying real-time ultrasonographic (RTU) images of pregnant sows, to predict litter size. The time of the test with the least error was also considered. A number of 4165 pregnancies in Farm 1 and 438 in Farm 2 were diagnosed twice, with the quality of the RTU images translated into rated-scale values (RSV1 and RSV2). When a deep neural network (DNN) was trained, the evaluation of the method showed that the prediction of litter size can be performed with little error. Root square mean error (RMSE) for training, validation with data from Farm 1, and testing on the data from Farm 2 were 0.91, 0.97, and 1.05, respectively. Corresponding mean absolute errors (MAE) were 2.27, 2.41, and 2.58. Time appeared to be a critical factor for the accuracy of the model. The smallest MAE was achieved when the RTU was performed at days 20–22. It is concluded that a numerical, RTU imaging model is a prominent predictor of litter size, when a DNN is used. Therefore, early routinely evaluated RTU images of pregnant sows can predict litter size, with machine learning, in an automated manner and provide a useful tool for the efficient management of pregnant sows.

Interactions between Ewes and Rams during Mating Can Be Used to Predict Lambing Dates Accurately, but Not Sire

Ewes often lamb over extended periods so the level of nutrition during pregnancy and lambing may be suboptimal for ewes that conceived later during mating. Predicting lambing dates would allow cohorts of ewes with similar gestational ages to be managed more precisely to achieve targets for ewe nutrition, feed on offer, mob sizes and access to shelter to improve lamb survival. The interactions between ewes and rams during mating have been used to predict the time of oestrus and lambing dates successfully, but this has not been tested at a commercial scale. In this study, proximity sensors were used to measure interactions between inexperienced Merino ewes (n = 317) and experienced rams (n = 9) during a 27-day mating period under commercial production conditions. When the gestation length was assumed to be 150 days, 91% of lambing dates were predicted within ±6 days of the actual birth date of lambs and 84% of lambing dates were predicted within ±3 days. The use of proximity sensors during mating was an effective means of predicting lambing dates, and there was no significant difference in accuracy for single bearing verses multiple bearing ewes. However, DNA parentage data showed the ram corresponding with the maximum daily interactions ratio was the sire for only 16% of all progeny, suggesting they could not be used to indicate the sire of the progeny.

Predictive blood biomarkers of sheep pregnancy and litter size

Early detection of sheep pregnancy and the prediction of how many lambs a pregnant ewe delivers affects sheep farmers in a number of ways, most notably with regard to feed management, lambing rate, and sheep/lamb health. The standard practice for direct detection of sheep pregnancy and litter size (PLS) is ultrasonography. However, this approach has a number of limitations. Indirect measurement of PLS using blood biomarkers could offer a simpler, faster and earlier route to PLS detection. Therefore, we undertook a large-scale metabolomics study to identify and validate predictive serum biomarkers of sheep PLS. We conducted a longitudinal experiment that analyzed 131 serum samples over five timepoints (from seven days pre-conception to 70 days post-conception) from six commercial flocks in Alberta and Ontario, Canada. Using LC–MS/MS and NMR, we identified and quantified 107 metabolites in each sample. We also identified three panels of serum metabolite biomarkers that can predict ewe PLS as early as 50 days after breeding. These biomarkers were then validated in separate flocks consisting of 243 animals yielding areas-under-the-receiver-operating-characteristic-curve (AU-ROC) of 0.81–0.93. The identified biomarkers could lead to the development of a simple, low-cost blood test to measure PLS at an early stage of pregnancy, which could help optimize reproductive management on sheep farms.

Melatonin treatment during late gestation of undernourished ewes: lamb body temperature and mother–young behaviours after birth

ContextIn extensive grazing sheep systems, pregnant ewes undergo periods of undernutrition because gestation coincides with winter when natural pasture is of lowest quantity and poorest quality. The lamb’s weight and thermoregulatory capacity, and the ewe–lamb bond at birth, may be compromised. Maternal melatonin treatment during gestation may reverse these effects.AimThe aim was to determine the effects of melatonin treatment of single-lambing, undernourished ewes during the last third of gestation on lamb birthweights and body temperatures, and on ewe–lamb interactive behaviour after birth.MethodsAt Day 100 of gestation, 39 single-bearing ewes received a subcutaneous melatonin implant, and 54 ewes served as controls with no implant. Throughout gestation, the ewes remained under extensive conditions grazing on natural pasture. Measurements were made of lamb birthweight, body temperatures (surface temperature by infrared thermography and rectal temperature), and ewe–lamb behaviours during a handling test at 6–17h after lambing.Key resultsThere was no effect of melatonin treatment on lamb birthweight or rectal temperature, or on ewe–lamb interaction behaviours. Hip minimum surface temperature was greater in lambs from melatonin-treated ewes than lambs from control ewes (21.2°C±0.9°C vs 18.8°C±0.8°C; P=0.05), and there was a similar trend for hip mean surface temperature (24.6°C±0.9°C vs 22.3°C±0.7°C; P=0.06). Rump surface temperatures were greater in male than female lambs: maximum (27.9°C±1.2°C vs 22.9°C±1.2°C; P=0.01), minimum (22.2°C±1.5°C vs 16.7°C±1.5°C; P=0.02) and mean (25.4°C±1.3°C vs 20.5°C±1.3°C; P=0.02).ConclusionMelatonin treatment during the last third of ewe pregnancy slightly enhanced the surface temperature of lambs at birth but did not influence ewe–lamb interaction behaviour after birth (i.e. after establishment of the ewe–lamb bond).ImplicationsFurther study in more depth is warranted into the possible effects of maternal supplementation with commercial melatonin implants on lamb development, thermoregulatory capacity, behaviour and survival rates in extensive grazing systems, including the effect on ewe–lamb behaviours immediately after birth for both singletons and twins.

A Survey of Farm Management Practices Relating to the Risk Factors, Prevalence, and Causes of Lamb Mortality in Ireland

Lamb mortality is a key factor influencing ewe productivity and profitability. The current study investigated risk factors associated with and management practices implemented on sheep farms to reduce lamb mortality. A survey consisting of 13 multiple-part questions (57 separate questions) was administered to all sheep farmers participating in the Teagasc National Farm Survey, representative of the Irish national population of sheep farms. A total of 60% of respondents identify mating or lambing date, and this practice tended to be associated with reduced lamb mortality (1.2%, p = 0.08). Individual lambing pens were used by 88% of farmers, but 26% did not clean or disinfect them. A total of 79% and 9.5% of farmers applied iodine to all lambs' navels and administered antibiotics to all lambs to treat and/or prevent diseases, respectively. Most farmers vaccinated their ewes (86%) and lambs (79%) against clostridial diseases and/or pasteurellosis; 13% vaccinated against abortion agents. Lamb mortality tended to be lower (Kruskal-Wallis (KW) = 2.749; p = 0.09) on farms that used stomach tubing, heat box, iodine, hospital, and individual pens compared with farms that do not implement all those practices. Predators, lamb birth weight, and diseases were perceived by respondents to be the three main causes of live-born lamb mortality. The gross margin is significantly higher on lowland farms by €37 per ewe compared with hill farms (Kruskal-Wallis (KW) = 4.056; p < 0.001). The combination of full-time farming and the use of hospital and individual pens improved gross margin (€18/ewe, p = 0.028). It is concluded that on-farm management practices affect both lamb mortality and flock gross margin.

Ultrasonographic and anatomical examination of normal thyroid and internal parathyroid glands in goats

Ultrasonographic examination of the normal thyroid and parathyroid glands has been described for humans and many animal species. However, similar reports for goats are still missing. The aim of the study was to present ultrasound features of the normal thyroid and internal parathyroid glands in goats with the determination of their dimensions and volume, followed by a comparison of the results to the gross examination. Seventy-two goats were used in the study. The echostructure and echogenicity of the thyroid and parathyroid glands were assessed. The length, width and height of the thyroid and the length and width of the parathyroid glands were measured. The thyroid volume was calculated using the ellipsoid formula, basing on the ultrasonographic dimensions. Size and volume of the dissected thyroid glands were established grossly, followed by a histological examination. In order to accurately describe the anatomy of the thyroid, new anatomical terminology characterizing this gland was proposed. The mean dimensions of the thyroid lobes were 30.2 x 10.5 x 6.3 mm. There were no statistically significant differences between the right and left lobe. Parathyroid glands measured an average of 3.6 x 2.4 mm. The percentage Root Mean Square Error between the results of ultrasonographic and gross examination was 16.73%, 20.65% and 17.01% for thyroid length, width and height, respectively, and 46.30% for volume. In order to obtain more precise calculation of the thyroid volume, a modified correction factor for the ellipsoid formula was introduced. For the first time, the normal ultrasonographic characteristics and dimensions of the caprine thyroid and internal parathyroid glands were presented. The results may serve as a radiological reference and be the basis for further research.

Consideration of breed specific fetal trunk diameter for the estimation of gestation age using trans-abdominal ultrasonography in sheep

The study aim was to estimate gestational age (GA), expected parturition date (EPD) and growth rate by determining fetal trunk diameter (TD). Effects of fetal-dam pelvis alignment and in utero fetal position at time of ultrasonography (UG) on fetal numbers and sex determination were also studied. Trans-abdominal UG (3-6.5 MHz) was conducted on 37 ewes with known breeding dates from Days 25-120 of pregnancy. Errors in GA and EPD were studied using an equation in the same ewes at their successive breeding when date of breeding was unknown. There were four equations, Y = 1.28861X+32.656 (R² = 0.92), for Indigenous; Y = 1.2603X+38.075 (R² = 0.85), for Indigenous × Garole; and Y = 0.8932X+45.916 (R² = 0.99), for Garole fetuses; and the equation, Y = 1.3565X + 32.604 (R² = 0.94), independent of breed were computed to estimate GA and the relationship between GA and TD of different breeds. The error in estimated GA and EPD using these four equations was determined and there was comparison with the data collected using US and the previously described equations. Results indicate there was the greatest (P <  0.01) error for GA and EPD values using the US TD equation for all breeds. There was the least error in estimated EPD using the breed specific equations. Error in the sex determination was 4.8 % and fetal number determination was 16.7 % with singleton and 7.7 % twin fetuses. The results indicate there is a breed specific fetal TD that is useful for predicting GA in sheep.

Effects of month of kidding, parity number, and litter size on milk yield of commercial dairy goats in Australia

The aim of this observational study was to identify the influence of key nongenetic factors such as month of kidding, parity, and litter size on milk yield and composition of Australian dairy goats throughout lactation. The study was conducted over 4 consecutive kidding seasons from June 2016 to March 2017. Data from 940 lactations of Saanen goats from a commercial herd were used to observe the effects of month of kidding, parity number, and litter size on total milk yield (L/goat) in early lactation (kidding to 90 d in milk; DIM), mid lactation (91-180 DIM), and late lactation (181-270 DIM), cumulative milk yield (from kidding to 270 DIM; CMY), average lactation length, proportion (%) of does reaching their target lactation length (270 DIM), somatic cell count (SCC), and percentages of milk fat and protein in early lactation. The mean herd responses throughout the entire study were as follows: CMY = 519 L/goat; lactation length = 233 d, with 70% of does reaching 270 DIM; milk fat = 4.2%; milk protein = 2.9%; and SCC = 6.2 × 10⁵ cells/mL. Average milk production peaked in February and was lowest in June (2.4 vs. 1.8 L/goat per day, respectively). Milk yield was affected by month of kidding, parity number, and litter size in all phases of lactation. November kidders had the greatest CMY, and March kidders had the lowest CMY. March kidders had the shortest lactation length and the lowest proportion of does reaching 270 DIM. June kidders had the longest lactation length, whereas September kidders had the highest proportion of does reaching 270 DIM. Maximum milk yield was attained in third parity. Goats in fourth or greater parity had the shortest lactation length, the lowest proportion of does reaching 270 DIM, and the highest SCC. Goats delivering single kids had lower CMY, lower SCC, and higher percentages of fat and protein than does delivering multiple kids. Our findings indicate that milk yield was primarily influenced by month of kidding, and the effects of month of kidding on milk yield were accentuated during mid lactation. However, the effects of month of kidding on milk yield varied significantly among parities.

Early pregnancy diagnosis by semiquantitative evaluation of luteal vascularity using power Doppler ultrasonography in sheep

The aim of the study was to measure the area of coloured pixels captured in power Doppler examination of corpus luteum (CL) by a computer-assisted image analysis software and to evaluate the efficiency of the semiquantitative assessment of CL in early pregnancy diagnosis in sheep. Assessment of luteal vascularity was carried out on 30 Kivircik ewes with power Doppler ultrasonography daily from Day 12 to Day 19 post breeding. The mean area of coloured pixels of CL with pregnancy was found significantly higher than the area of coloured pixels of CL undergoing luteolysis (0.38 ± 0.012 mm2 and 0.31 ± 0.017 mm2, respectively, P < 0.001). Assessment of day-based examination showed a significant difference between the area of coloured pixels of CL with continuing pregnancy and luteolysis on Day 17 (P < 0.05), which was followed by Days 18 and 19 (P < 0.001). It is concluded that pregnancy diagnosis can be done accurately by semiquantitative assessment of luteal vascularization as early as 17 days post breeding in ewes. According to the authors’ knowledge, this is the first report on the efficacy of semiquantitative assessment of ovine CL in early pregnancy diagnosis.

Ultrasonographic foetometry and maternal serum progesterone concentrations during pregnancy in Turkish Saanen goats

The objectives of the present study were (1) to generate reliable equations to estimate the gestational age of Saanen goats by measuring embryonic and foetal structures via B-Mode ultrasonography (USG), (2) determine the maternal serum progesterone (P₄) concentrations during pregnancy, and (3) compare the relationship between P₄ concentrations and gestational age, litter size, and birth weight of kids. Seven pregnant goats were used in this study. Ultrasonographic examinations were performed transrectally from Days 15 to 27 of pregnancy at 3-day intervals and transabdominally from Days 30 to 128 at 7-day intervals. Blood samples were collected daily between Day 15 and 24 of gestation, and weekly from Day 27 to 128 to determine the serum P₄ concentrations. Outer placentome diameter, biparietal diameter, occipito-nasal length, chest diameter, foetal heart rate, transversal heart diameter, orbit diameter, umbilical cord diameter, and kidney diameter were highly correlated with gestational age (P <  0.0001). Goats with twins had greater progesterone concentrations than those with a single foetus (P <  0.001). Furthermore, maternal serum progesterone concentrations detected on Days 51, 107, 114, and 128 of gestation were greater in goats carrying twins than those with a single foetus (P <  0.05). A negative correlation was evident between serum progesterone concentrations and birth weights of the kids (r = -0.78; P =  0.007). In conclusion, gestational age can be estimated using ultrasonographic foetometry, and litter size at specific times during pregnancy can be determined based on maternal serum progesterone concentration in Turkish Saanen goats.

Use of ultrasonographic fetometry for the estimation of days to kidding in dairy does

The objective of this study was to predict the number of days to kidding in dairy goats by ultrasonographic measurement of three consecutive ribs and intercostal spaces or trunk diameter of fetuses. Two groups of primiparous and multiparous goats were examined by transabdominal ultrasonography. Actual kidding date was used to compute the true number of days to kidding relative to the ultrasound examination day. Linear regression models were built to predict days to kidding based on fetal measurements, parity, and number of kids. The model using the width of three consecutive ribs and intercostal spaces (n = 29 goats) showed a curvilinear relationship between days to kidding and rib distance (R² = 0.60; P < 0.01). The trunk diameter model (n = 135 goats) showed a curvilinear relationship between days to kidding and trunk diameter with an effect of parity (R² = 0.81; P < 0.01).

Evaluation of the Nova Vet Meter for sheep-side monitoring of β-hydroxybutyric acid (BHBA) and description of ewe BHBA during late gestation in three flocks from the Northeastern U.S

Prevention of metabolic diseases in small ruminants may improve production efficiency and profitability, yet ewes carrying multiples or who are in poor body condition are at increased susceptibility to develop ketosis. This study evaluated the hand-held Nova Vet Meter to accurately detect β-hydroxybutyric acid (BHBA) concentrations in ewes and determined the percentage of ewes at moderate (0.8 to 1.5 mmol/L BHBA) and greatest (≥1.6 mmol/L BHBA) risk to develop ketosis during late gestation. To validate the Nova Vet Meter, BHBA concentrations of 104 paired blood samples were measured using the Nova Vet Meter and gold-standard laboratory analysis. Receiver operating characteristics were calculated. The accuracy and sensitivity of detecting BHBA concentrations at 0.8 to 1.5 mmol/L were 94.2% and 97.3%, respectively. The accuracy and sensitivity of detecting BHBA concentrations ≥ 1.6 mmol/L were 98.0% and 50.0%, respectively. Ewe body weight (BW), body condition score (BCS), and BHBA of 117 ewes from three flocks were determined weekly during the four weeks before parturition. During the last three weeks of gestation >20% of ewes were identified with moderate risk to develop ketosis. During the last four weeks of gestation, ewes carrying triplets had reduced BCS (P = 0.0002) and increased BHBA concentrations (P < 0.0001) compared with singleton and twin pregnancies. Ewe BHBA did not correlate with lamb birth weight (R² = 0.003; P = 0.41). In conclusion, the Nova Vet Meter is suitable for sheep-side BHBA monitoring between 0.8 and 1.5 mmol/L, but further testing is necessary to evaluate BHBA readings ≥1.6 mmol/L.