A machine vision system to predict individual cow feed intake of different feeds in a cowshed

Data on individual feed intake of dairy cows, an important variable for farm management, are currently unavailable in commercial dairies. A real-time machine vision system including models that are able to adapt to multiple types of feed was developed to predict individual feed intake of dairy cows. Using a Red-Green-Blue-Depth (RGBD) camera, images of feed piles of two different feed types (lactating cows' feed and heifers' feed) were acquired in a research dairy farm, for a range of feed weights under varied configurations and illuminations. Several models were developed to predict individual feed intake: two Transfer Learning (TL) models based on Convolutional Neural Networks (CNNs), one CNN model trained on both feed types, and one Multilayer Perceptron and Convolutional Neural Network model trained on both feed types, along with categorical data. We also implemented a statistical method to compare these four models using a Linear Mixed Model and a Generalised Linear Mixed Model, showing that all models are significantly different. The TL models performed best and were trained on both feeds with TL methods. These models achieved Mean Absolute Errors (MAEs) of 0.12 and 0.13 kg per meal with RMSE of 0.18 and 0.17 kg per meal for the two different feeds, when tested on varied data collected manually in a cowshed. Testing the model with actual cows’ meals data automatically collected by the system in the cowshed resulted in a MAE of 0.14 kg per meal and RMSE of 0.19 kg per meal. These results suggest the potential of measuring individual feed intake of dairy cows in a cowshed using RGBD cameras and Deep Learning models that can be applied and tuned to different types of feed.     

A context-free encoding scheme of protein sequences for predicting antigenicity of diverse influenza A viruses

Background

The evolution of influenza A viruses leads to the antigenic changes. Serological diagnosis of the antigenicity is usually labor-intensive, time-consuming and not suitable for early-stage detection. Computational prediction of the antigenic relationship between emerging and old strains of influenza viruses using viral sequences can facilitate large-scale antigenic characterization, especially for those viruses requiring high biosafety facilities, such as H5 and H7 influenza A viruses. However, most computational models require carefully designed subtype-specific features, thereby being restricted to only one subtype.

Methods

In this paper, we propose a Context-FreeEncoding Scheme (CFreeEnS) for pairs of protein sequences, which encodes a protein sequence dataset into a numeric matrix and then feeds the matrix into a downstream machine learning model. CFreeEnS is not only free from subtype-specific selected features but also able to improve the accuracy of predicting the antigenicity of influenza. Since CFreeEnS is subtype-free, it is applicable to predicting the antigenicity of diverse influenza subtypes, hopefully saving the biologists from conducting serological assays for highly pathogenic strains.

Results

The accuracy of prediction on each subtype tested (A/H1N1, A/H3N2, A/H5N1, A/H9N2) is over 85%, and can be as high as 91.5%. This outperforms existing methods that use carefully designed subtype-specific features. Furthermore, we tested the CFreeEnS on the combined dataset of the four subtypes. The accuracy reaches 84.6%, much higher than the best performance 75.1% reported by other subtype-free models, i.e. regional band-based model and residue-based model, for predicting the antigenicity of influenza. Also, we investigate the performance of CFreeEnS when the model is trained and tested on different subtypes (i.e. transfer learning). The prediction accuracy using CFreeEnS is 84.3% when the model is trained on the A/H1N1 dataset and tested on the A/H5N1, better than the 75.2% using a regional band-based model.

Conclusions

The CFreeEnS not only improves the prediction of antigenicity on datasets with only one subtype but also outperforms existing methods when tested on a combined dataset with four subtypes of influenza viruses.

     

不同月龄绵羊对食物营养利用效率研究

利用祁连山地和河西走廊绿洲作物生态系统的时空差和生产方式所提供的饲草养畜,建立资源互补效应,提高生物效率．结果表明,在相同的营养条件下,随着羊龄的增大,食入营养量、体增重、单位增重消耗的营养量均增多,但营养物质消化率、可消化营养物质的利用率、饲料报酬均有所下降．F测验食入营养物质的差异程度表明,34月龄羊的干物质和蛋白质食入量显著大于22月龄羊(P<0．01),22月龄羊显著大于10月龄羊(P<0．01),食入能为34月龄羊与22月龄羊的差异显著(P<0．05),22月龄羊与10月龄羊差异极显著(P<0．01)．体增重是34月龄与22月龄羊和22月龄羊与10月龄羊差异不显著(P>0．05),34月龄显著大于10月龄羊(P<0．05)．饲料转化率三组间的差异均不显著,表明10月龄到22月龄羊的生产效率高,能生产高质量的风味羊肉,是适宰羊的年龄段．     

Effect of herbal supplementation on growth,immunity, rumen histology,serum antioxidants and meat quality of sheep

There is a wide range of feed additives deliberately designed to be used in sheep diets that can improve production performance. Whereas herbal supplementation is gaining popularity not only for improving sheep productivity and mutton quality but also for safe application without any harmful residual effects. The present study was designed to investigate the effect of plantain (Plantago lanceolata L.) and/or garlic leaf (Allium sativum) dietary supplementation on growth performance, immunity, rumen histology, serum antioxidants and meat quality of sheep. The experiment consisted of a completely randomized design with 32 one-year-old sheep (initial mean live weight 9 ± 0.2 kg) allocated to four groups (8 sheep per group). Rice straw and concentrates-based total mixed ration pellets (2390 kcal/kg DM, CP = 15.1%) were offered as a control diet (CL diet). Herbal treatment diets included (i) CL diet + 10 g DM of plantain herb (PL diet), (ii) CL diet + 10 g DM of garlic leaf (GL diet) and (iii) CL diet + 5 g DM of PL and 5 g DM of GL (PG diet). Compared with the CL diet group, the live weight gain and feed conversion ratio were 18% to 26% and 13% to 20% higher in herbal-supplemented groups, respectively. Moreover, the herbal-supplemented groups, especially the PL diet group had higher serum immunoglobulin concentration, antioxidant capacity and rumen papillae size compared to the control. Besides, the lowest caul fat and pelvic fat levels were observed in the PL diet group followed by PG, GL and CL diet groups. In addition, lower mutton ether extract and saturated fatty acid along with higher polyunsaturated fatty acid levels were found in all herbal-supplemented groups. In conclusion, dietary supplementation with PL and/or GL might be used as an alternative in sheep to promote growth, health status and lean mutton production.     

基于性诱和深度学习的草地贪夜蛾成虫自动识别计数方法

【目的】探究深度学习在草地贪夜蛾Spodoptera frugiperda成虫自动识别计数上的可行性,并评估模型的识别计数准确率,为害虫机器智能监测提供图像识别与计数方法。【方法】设计一种基于性诱的害虫图像监测装置,定时自动采集诱捕到的草地贪夜蛾成虫图像,结合采集船形诱捕器粘虫板上草地贪夜蛾成虫图像,构建数据集;应用YOLOv5深度学习目标检测模型进行特征学习,通过草地贪夜蛾原始图像、清除边缘残缺目标、增加相似检测目标(斜纹夜蛾成虫)、无检测目标负样本等不同处理的数据集进行模型训练,得到Yolov5s-A1, Yolov5s-A2, Yolov5s-AB, Yolov5s-ABC 4个模型,对比在不同遮挡程度梯度下的测试样本不同模型检测结果,用准确率(P)、召回率(R)、F1值、平均准确率(average precision, AP)和计数准确率(counting accuracy, CA)评估各模型的差异。【结果】通过原始图像集训练的模型Yolov5s-A1的识别准确率为87.37%,召回率为90.24%,F1值为88.78;清除边缘残缺目标图像集训练得到的模型Yolov5s-A2的识别准确率为93.15%,召回率为84.77%,F1值为88.76;增加斜纹夜蛾成虫样本图像训练的模型Yolov5s-AB的识别准确率为96.23%,召回率为91.85%,F1值为93.99;增加斜纹夜蛾成虫和无检测对象负样本训练的模型Yolov5s-ABC的识别准确率为94.76%,召回率为88.23%,F1值为91.38。4个模型的AP值从高到低排列如下：Yolov5s-AB>Yolov5s-ABC> Yolov5s-A2>Yolov5s-A1,其中Yolov5s-AB与Yolov5s-ABC结果相近;CA值从高到低排列如下：Yolov5s-AB>Yolov5s-ABC>Yolov5s-A2>Yolov5s-A1。【结论】结果表明本文提出的方法应用于控制条件下害虫图像监测设备及诱捕器粘虫板上草地贪夜蛾成虫的识别计数是可行的,深度学习技术对于草地贪夜蛾成虫的识别和计数是有效的。基于深度学习的草地贪夜蛾成虫自动识别与计数方法对虫体姿态变化、杂物干扰等有较好的鲁棒性,可从各种虫体姿态及破损虫体中自动统计出草地贪夜蛾成虫的数量,在害虫种群监测中具有广阔的应用前景。     

One Month Weight Loss Predicts the Efficacy of Liraglutide in Obese Patients: Data from A Single Center

Objective: Liraglutide is a glucagon-like peptide 1 receptor agonist which acts through peripheral and central receptor pathways affecting food intake. Preliminary identification of responder patients represents a crucial point to reduce an inappropriate exposure to the drug and the health expenditure. The primary endpoint of our study was to identify predictors of liraglutide efficacy in the short term follow-up. The secondary endpoint was to evaluate the treatment efficacy stratified by the underlying psychiatric disorder.Methods: We evaluated a cohort of 100 patients (77 females, 23 males, mean body mass index 38.6 ± 3.2 kg/m²) who were evaluated at baseline, and after 1, 3, and 6 months of treatment. Liraglutide efficacy was defined by a weight loss ≥5% of initial weight. Sociodemographic/metabolic parameters, food intake, smoking habit, and physical activity were correlated with liraglutide efficacy.Results: There was a significant weight loss after 1 month of therapy, as well as after 3 and 6 months when compared to the baseline (P<.0001; 27%, 45%, and 57% of patients showed a weight loss ≥5%, respectively). No difference was found in weight loss between the 3 groups of patients (with binge eating, with/without psychiatric disorders). The weight loss at 1 month was the only predictor of a positive response to the treatment.Conclusion: Our results confirm the efficacy of liraglutide even at a lower dose than conventional. The early response to the drug seems to be a good predictor of long-term efficacy and it might be useful in clinical practice to identify patients in whom liraglutide may induce a significant weight loss.Abbreviations: BMI = body mass index; EMA = European Medicine Agency; FDA = Food and Drug Administration; GLP-1 RA = glucagon-like peptide 1 receptor agonist     

Social facilitation of predatory, sheep-chasing behaviour in Norwegian Elkhounds, grey

Sheep grazing on unfenced mountain pasture may be attacked by loose dogs, which may chase and kill sheep and separate lambs from their mothers. We have earlier shown that testing dogs individually towards sheep with an electronic dog collar may effectively reduce the chasing propensity 1 year later. The aim of this study was to investigate whether and how a non-chasing and a chasing companion dog influence sheep chasing in test dogs (20 Norwegian Elkhounds, grey), and whether this varies with sex or age. The test dogs' predatory behaviour towards sheep was examined by observation of the dogs in a fenced enclosure with sheep in 5min tests together with, first, a non-chasing (Hamilton stoever, a hound breed) and 2-3 days later a chasing (Border collie) companion dog. Physical contact with sheep was prevented. Initially, the test dogs exhibited a higher chasing motivation towards sheep in tests with a chasing, as compared with a non-chasing, companion (P<0.001). During the entire test, 60% of the dogs attacked sheep when accompanied by the non-chaser. All dogs attacked sheep when the chasing companion was present (P=0.008), although in only 8 of 20 cases the companion dog chased simultaneously with the test dog. In the chasing-companion tests, the attack latency was shorter (P<0.001) than in tests with a non-chasing companion. The attack severity was higher when attacking alone than when attacking together with the chasing companion (P=0.033). In these tests, the higher the attack severity, the shorter was the attack latency (P=0.006). A sequence analysis on chasing-companion tests showed that test dogs generally started with observing or showing interest in sheep, followed by attacks, which increased in severity. No sex differences were observed. In non-chasing companion tests, dogs aged up to 2.5 years exhibited a weaker initial hunting motivation than older dogs (P=0.025), but during the entire test were more inclined to attack (P=0.019). Taken together, our findings indicate that a companion dog showing intentions of predatory behaviour stimulates predatory chase in another dog, while a non-chasing companion has a limited influence on this. In tests certifying dogs for their refrainment of chasing sheep, well trained Border collies approaching sheep on command might be used to reveal the full predatory potential of the dog being tested.     

Estimation of (co)variance components,genetic parameters,and genetic trends of growth traits in community-based breeding programs of Bonga sheep

Community-based sheep breeding programs (CBBPs) have been adopted strategically to improve Bonga sheep, the most popular sheep breed in Ethiopia. The present study was undertaken to estimate genetic parameters and genetic trends for growth traits and inbreeding levels in each Bonga sheep CBBP. Data pertaining to growth traits, spanning a period of seven years (2012–2017), were collected from 14 Bonga sheep CBBPs. Data were analyzed using the General Linear Model procedure of SAS to study the performance of the breed over the years. The genetic parameters were estimated by univariate and multivariate animal model using restricted maximum likelihood method of WOMBAT software. The genetic trends were estimated by the regression of the average breeding values of the animals on the year of birth. The overall least square means ± SE of BW (kg) were 3.10 ± 0.010, 16.1 ± 0.07, 24.7 ± 0.20, 30.4 ± 0.40 and 34.0 ± 0.84 for birth weight (BWT), weaning weight (WWT), six-month weight (SMWT), nine-month weight (NMWT) and yearling weight (YWT), respectively. Direct heritability estimates from selected models were 0.56 ± 0.030, 0.36 ± 0.030, 0.22 ± 0.040, 0.17 ± 0.070 and 0.13 ± 0.150 for BWT, WWT, SMWT, NMWT and YWT, respectively. Six-month weight was the selection trait and presented positive trends for 10 CBBPs, and negative trends for four CBBPs. Moderate to high heritability estimates and positive genetic trends indicated scope for further improvement of BW. Additionally, the positive and high correlation between BW traits indicated that selection for just one trait would also improve the other traits through correlated responses.     

Genetic and non-genetic parameter estimates for growth traits and Kleiber ratios in Dorper × indigenous sheep

Genetic improvement programme will only be successful when accompanied by a good understanding of the influence of different environmental factors, knowledge of the genetic parameters and the genetic relationships between the traits of interest. This study aimed to evaluate the influence of non-genetic factors on growth traits and Kleiber ratios and to estimate genetic parameters for early growth traits in Dorper × indigenous crossbred sheep. The effects of fixed factors were analysed by the general linear model procedure of statistical analysis system, while the genetic parameters were estimated using a WOMBAT computer program fitted animal model. The overall least-square mean for birth weight (BRW), weaning weight (3MW), six-month weight, nine-month weight, and yearling weight were 3.03 ± 0.02, 14.5 ± 0.18, 20.4 ± 0.26, 24.8 ± 0.31, and 28.3 ± 0.40 kg, respectively. The overall least-square mean for Kleiber ratio from birth to weaning (KR1), weaning to six months, six to nine months and nine months to yearling age were 16.8 ± 0.10, 6.41 ± 0.17, 4.55 ± 0.21 and 3.38 ± 0.20 g/kg of metabolic weight, respectively. The inclusion of maternal genetic effect had a significant influence on BRW, and it explains 20% of the phenotypic variation. The total heritability estimates for BRW, 3MW, birth to weaning average daily weight gain and KR1 were 0.10, 0.14, 0.16 and 0.12, respectively. The phenotypic correlation varied from ?0.11 ± 0.05 to 0.98 ± 0.02, whereas the direct genetic correlation ranged from ?0.32 ± 0.40 to 0.98 ± 0.17. The mean inbreeding coefficient was 0.105% with an annual rate of 0.02%. The heritability estimates for growth traits and Kleiber ratio suggest that slow genetic progress would be expected from the selection. However, the integration of selection with crossbreeding programme can enhance genetic gain. Therefore, selection should be conducted based on breeding values estimated from multiple information sources to increase the selection response.     

Dual energy X-ray absorptiometry precisely and accurately predicts lamb carcass composition at abattoir chain speed across a range of phenotypic and genotypic variables

Dual energy X-ray absorptiometry (DEXA) is an imaging modality that has been used to predict the computed tomography (CT)-determined carcass composition of multiple species, including sheep and pigs, with minimal inaccuracies, using medical grade DEXA scanners. An online DEXA scanner in an Australian abattoir has shown that a high level of precision can be achieved when predicting lamb carcass composition in real time. This study investigated the accuracy of that same online DEXA when predicting fat and lean percentages as determined by CT over a wide range of phenotypic and genotypic variables across 454 lambs over 6 kill groups and contrasted these results against the current Australian industry standard of grade-rule (GR) measurements to grade carcasses. Lamb carcasses were DEXA scanned and then CT scanned to determine CT Fat % and CT Lean %. All phenotypic traits and genotypic information, including Australian Sheep Breeding Values, were recorded for each carcass. Residuals of the DEXA predicted CT Fat % and Lean %, and the actual CT Fat % and Lean % were calculated and tested against all phenotypic and genotypic variables. Excellent overall precision was recorded when predicting CT Fat % (R² = 0.91, RMSE = 1.19%). Small biases present for sire breed, sire type, dam breed, hot carcass weight and c-site eye muscle area could be explained by a regression paradox; however, biases among kill group (−0.73% to 1.01% for CT Fat %, −1.48% to 0.76% for CT Lean %) and the Merino sire type (0.36% for CT Fat %, −0.73% for CT Lean %) could not be explained by this effect. Over the large range of phenotypic and genotypic variation, there was excellent precision when predicting CT Fat % and CT Lean % by an online DEXA, with only minor biases, showing superiority to the existing Australian standard of GR measurements.     

The Benefit of Short-Term Weight Loss with Anti-Obesity Medications in Real-World Clinical Practice

Objective: The effectiveness of anti-obesity medications (AOMs) outside of clinical trials is unclear. The objective of this study was to compare the short-term effectiveness of AOMs in real-world practice.Methods: This retrospective study included adults aged ≥18 years, with body mass index ≥30 kg/m² or ≥27 kg/m² with at least one obesity-related comorbidity who were prescribed phentermine hydrochloride, phenterminetopiramate, bupropion-naltrexone, or lorcaserin for 12 consecutive weeks between 2006 and 2016 at a large tertiary healthcare system. Propensity score–matched cohorts were created for each pair of AOMs. The primary outcomes were percent and absolute weight loss from baseline after 12 weeks. A prediction model was constructed to estimate weight loss with different AOMs based on demographic and clinical data.Results: Of the 3,411 patients included in this study, patients lost an average of 3.45% of body weight from baseline. All AOMs were associated with a significant weight loss from baseline (P<.0001). Patients lost the highest percentage of body weight on phentermine hydrochloride (3.75 ± 5.66%), followed by phentermine-topiramate (3.63 ± 5.7%), bupropion-naltrexone (2.66 ± 5.03%), and lorcaserin (1.84 ± 6.69%). In propensity-matched cohorts, patients taking phentermine hydrochloride lost more weight than those taking lorcaserin or bupropion-naltrexone, and patients taking phentermine topiramate lost more weight than patients taking lorcaserin.Conclusion: In real-world practice, AOMs are associated with clinically meaningful weight loss of 2 to 4% after 12 weeks. In this study, phentermine hydrochloride and phentermine topiramate produced the most weight loss. AOMs should be seriously considered as part of the armamentarium to treat patients with obesity.Abbreviations: AOM = anti-obesity medication; BMI = body mass index; EMR = electronic medical record; FDA = Food and Drug Administration; T2D = type 2 diabetes     

Arrhythmic Heartbeat Classification Using 2D Convolutional Neural Networks

BackgroundElectrocardiogram (ECG) is a method of recording the electrical activity of the heart and it provides a diagnostic means for heart-related diseases. Arrhythmia is any irregularity of the heartbeat that causes an abnormality in the heart rhythm. Early detection of arrhythmia has great importance to prevent many diseases. Manual analysis of ECG recordings is not practical for quickly identifying arrhythmias that may cause sudden deaths. Hence, many studies have been presented to develop computer-aided-diagnosis (CAD) systems to automatically identify arrhythmias.MethodsThis paper proposes a novel deep learning approach to identify arrhythmias in ECG signals. The proposed approach identifies arrhythmia classes using Convolutional Neural Network (CNN) trained by two-dimensional (2D) ECG beat images. Firstly, ECG signals, which consist of 5 different arrhythmias, are segmented into heartbeats which are transformed into 2D grayscale images. Afterward, the images are used as input for training a new CNN architecture to classify heartbeats.ResultsThe experimental results show that the classification performance of the proposed approach reaches an overall accuracy of 99.7%, sensitivity of 99.7%, and specificity of 99.22% in the classification of five different ECG arrhythmias. Further, the proposed CNN architecture is compared to other popular CNN architectures such as LeNet and ResNet-50 to evaluate the performance of the study.ConclusionsTest results demonstrate that the deep network trained by ECG images provides outstanding classification performance of arrhythmic ECG signals and outperforms similar network architectures. Moreover, the proposed method has lower computational costs compared to existing methods and is more suitable for mobile device-based diagnosis systems as it does not involve any complex preprocessing process. Hence, the proposed approach provides a simple and robust automatic cardiac arrhythmia detection scheme for the classification of ECG arrhythmias.     

Purification and characterization of carbonic anhydrase from sheep kidney and effects of sulfonamides on enzyme activity

Carbonic anhydrase (CA, EC: 4.2.1.1) was purified from sheep kidney by affinity chromatography on a Sepharose 4B-tyrosine-sulfanilamide column. By means of two consecutive procedures, the enzyme (sCA) was purified 227.61-fold with a yield of 60.75%, and a specific activity of 838.89 U/mg proteins. The optimum temperature, ionic strength and pH were determined to be 35 °C, 20 mM and 8.5, respectively. The molecular weight determined by SDS–PAGE was found to be 29 kDa. The kinetic parameters, K_M and V_max values were determined for the 4-nitrophenyl acetate (p-NpA) hydrolysis reaction. Some sulfonamides were tested as inhibitors against the purified CAs enzyme. The K_i constants for benzenesulfonamide (1), sulfanilamide (2), mafenide (3), 4-(2-aminoethyl) benzenesulfonamide (4), 4-methyl-benzenesulfonamide (5), 2-bromo-benzenesulfonamide (6), naphthalene-2-sulfonamide (7), 4-amino-6-chlorobenzene-1,3-disulfonamide (8) and saccharin (9) were in the range 1.348–69.31 μM.     

Previous Delivery of Macrosomia is Associated with Maternal Adiposity in Later Life in Chinese Parous Women with Normal Weight Before and/or After Pregnancy

Objective: Obesity has become a major worldwide health challenge. Macrosomic infants are more likely to experience type 2 diabetes mellitus, obesity and hypertension in adulthood. However, whether macrosomia increases the risk of maternal adiposity later in life is still unknown.Methods: One thousand nine hundred eighty-six unrelated parous women of Chinese Han ancestry aged from 40 to 76 years were enrolled. Self-reported information about reproductive status, including age at menarche, number of children, previous delivery of macrosomic infants, and body weight before and after pregnancy were obtained from personal interview by trained interviewers using a standard questionnaire. Macrosomia was defined as birth weight greater than 4,000 g. Adiposity indexes were measured or calculated.Results: Prior delivery of macrosomia was associated with an increased risk of having obesity in parous women with normal weight before pregnancy (odds ratio &lsqb;OR] = 1.840; 95% confidence interval &lsqb;CI] 1.028, 3.294; P = .040), as well as a higher risk of overweight/obesity in parous women with normal weight after pregnancy (OR = 1.777; 95% CI 1.131, 2.794; P = .013). In addition, previous delivery of macrosomia was related with 1.919 (95% CI 1.207, 3.050; P = .006) times higher risk of overweight/obesity in parous women with normal weight before and after pregnancy.Conclusion: The present study suggests that prior delivery of macrosomia may be an independent risk factor for adiposity later in life in parous women with normal weight before and/or after pregnancy.Abbreviations: BMI = body mass index; CI = confidence interval; OR = odds ratio; WC = waist circumference; WHR = waist-to-hip ratio; WHtR = waist-to-height ratio     

A Retrospective Study Examining Phentermine on Preconception Weight Loss and Pregnancy Outcomes

Objective: Obesity is a well-known risk factor for infertility. However, the use of weight loss medications prior to conception is underutilized. The objectives of our study are to describe weight loss, pregnancy rates, and live birth rates after short-term phentermine use in women with obesity and infertility.Methods: This was a retrospective analysis of 55 women (18 to 45 years old) who were overweight or obese, diagnosed with infertility, and prescribed phentermine for weight loss in an ambulatory endocrinology clinic at a single, tertiary level academic medical center. Main outcome measures were mean percent weight change at 3 months after starting phentermine, and pregnancy, and live birth rates from start of phentermine to June 30, 2017.Results: Median duration of phentermine use was 70 days (Q1, Q3 &lsqb;33, 129]). Mean ± SD percent weight change at 3 months after starting phentermine was -5.3 ± 4.1% (P<.001). The pregnancy rate was 60% and the live birth rate was 49%. There was no significant difference in pregnancy rates (52% versus 68%; P = .23) or live birth rates (44% versus 54%; P = .50) in women who lost ≥5% versus <5% of their baseline weight. The number of metabolic comorbidities was negatively associated with the pregnancy rate. Phentermine was generally well-tolerated with no serious adverse events.Conclusion: Phentermine can produce clinically significant weight loss in women with obesity during the preconception period. Higher pregnancy or live birth rates were not observed with a greater degree of weight loss with phentermine.     

Community-based sheep breeding programs generated substantial genetic gains and socioeconomic benefits

Community-based breeding programs (CBBPs) for small ruminants have been suggested as alternatives to centralised, government-controlled breeding schemes which have been implemented in many developing countries. An innovative methodological framework on how to design, implement and sustain CBBPs was tested in three sites in Ethiopia: Bonga, Horro and Menz. In these CBBPs, the main selection trait identified through participatory approaches was 6-month weight in all three sites. In Horro and Bonga, where resources such as feed and water permitted larger litter sizes, twinning rate was included. Ten-year (2009 to 2018) performance data from the breeding programs were analysed using Average Information Restricted Maximum Likelihood method (AI-REML). Additionally, the socioeconomic impact of CBBPs was assessed. Results indicated that 6-month weight increased over the years in all breeds. In Bonga, the average increase was 0.21 ± 0.018 kg/year, followed by 0.18 ± 0.007 and 0.11 ± 0.003 kg/year in Horro and Menz, respectively. This was quite substantial in an on-farm situation. The birth weight of lambs did not improve over the years in Bonga and Horro sheep but significant increases occurred in Menz. Considering that there was no direct selection on birth weight in the community flock, the increased weights observed in Menz could be due to correlated responses, but this was not the case in Bonga and Horro. The genetic trend for prolificacy over the years in both Bonga and Horro flocks was positive and significant (P < 0.01). This increase in litter size, combined with the increased 6-month body weight, increased income by 20% and farm-level meat consumption from slaughter of one sheep per year to three. The results show that CBBPs are technically feasible, result in measurable genetic gains in performance traits and impact the livelihoods of farmers.     

Estimation of calcium requirements for optimal productive and reproductive performance,eggshell and tibial quality in egg-type duck breeders

Optimizing the dietary calcium (Ca) level is essential to maximize the eggshell quality, egg production and bone formation in poultry. This study aimed to establish the Ca requirements of egg-type duck breeders from 23 to 57 weeks of age on egg production, eggshell, incubation, tibial, plasma and ovary-related indices, as well as the expression of matrix protein-related genes. Totally, 450 Longyan duck breeders aged 21 weeks of age were allotted randomly into five treatments, each with six replicates of 15 individually caged birds. The data collection started from 23 weeks of age and continued over the following 35 weeks. The five groups corresponded to five dietary treatments containing either 2.8%, 3.2%, 3.6%, 4.0% or 4.4% Ca. The tested dietary Ca levels increased (linear, P <0.01) egg production and egg mass, and linearly improved (P <0.01) the feed conversion ratio (FCR). Increasing the dietary Ca levels from 2.8% to 4.4% increased (P <0.01) the eggshell thickness and eggshell content. The tested Ca levels showed a quadratic effect on eggshell thickness and ovarian weight (P <0.01); the highest values were obtained with the Ca levels 4.0% and 3.6%, respectively. Dietary Ca levels affected the small yellow follicles (SYF) number and SYF weight/ovarian weight, and the linear response (P <0.01) was significant vis-à-vis SYF number. In addition, dietary Ca levels increased (P <0.05) the tibial dry weight, breaking strength, mineral density and ash content. Plasma and tibial phosphorus concentration exhibited a quadratic (P <0.01) response to dietary Ca levels. Plasma calcitonin concentration linearly (P <0.01) increased as dietary Ca levels increased. The relative expression of carbonic anhydrase 2 in the uterus rose (P <0.01) with the increment of dietary Ca levels, and the highest value was obtained with 3.2% Ca. In conclusion, Longyan duck breeders fed a diet with 4.0% Ca had superior eggshell and tibial quality, while those fed a diet with 3.6% Ca had the heaviest ovarian weights. The regression model indicated that the dietary Ca levels 3.86%, 3.48% and 4.00% are optimal levels to obtain maximum eggshell thickness, ovarian weight and tibial mineral density, respectively.     

Three critical factors affecting automated image species recognition performance for camera traps

Ecological camera traps are increasingly used by wildlife biologists to unobtrusively monitor an ecosystems animal population. However, manual inspection of the images produced is expensive, laborious, and time‐consuming. The success of deep learning systems using camera trap images has been previously explored in preliminary stages. These studies, however, are lacking in their practicality. They are primarily focused on extremely large datasets, often millions of images, and there is little to no focus on performance when tasked with species identification in new locations not seen during training. Our goal was to test the capabilities of deep learning systems trained on camera trap images using modestly sized training data, compare performance when considering unseen background locations, and quantify the gradient of lower bound performance to provide a guideline of data requirements in correspondence to performance expectations. We use a dataset provided by Parks Canada containing 47,279 images collected from 36 unique geographic locations across multiple environments. Images represent 55 animal species and human activity with high‐class imbalance. We trained, tested, and compared the capabilities of six deep learning computer vision networks using transfer learning and image augmentation: DenseNet201, Inception‐ResNet‐V3, InceptionV3, NASNetMobile, MobileNetV2, and Xception. We compare overall performance on “trained” locations where DenseNet201 performed best with 95.6% top‐1 accuracy showing promise for deep learning methods for smaller scale research efforts. Using trained locations, classifications with <500 images had low and highly variable recall of 0.750 ± 0.329, while classifications with over 1,000 images had a high and stable recall of 0.971 ± 0.0137. Models tasked with classifying species from untrained locations were less accurate, with DenseNet201 performing best with 68.7% top‐1 accuracy. Finally, we provide an open repository where ecologists can insert their image data to train and test custom species detection models for their desired ecological domain.     

Growth modelling of different ram breeds using computer tomography

In this study the growth of rams of 3 different sheep breeds with extremely different growth patterns (Hungarian Merino, HM; American Suffolk, AS; and Dutch Texel, DT) were modelled. Two models (modified Gompertz and logistic) were fitted to experimental data generated from successive computer tomography (CT) measurements. In each breed the data fitted the Gompertz model with better precision (lower RSS values). However, the symmetric logistic function fitted the late maturing fat and long loin of the AS lambs more accurately. These models have consequently been validated to give in vivo predictions of market specific products and the optimal slaughter weight of each breed. Based on the results of the Gompertz model, medium sized sheep breeds with low growth capacity (e.g. HM) should be slaughtered when their live weight is between 21 and 33 kg at an age of 3–5 months. The large-sized, slow maturing breeds (AS) should be slaughtered between 27 and 45 kg, at a similar age to the HM lambs. The early maturing breed (DT) should not be reared older than 2–3.5 months of age, with a live weight varying between 24 and 36 kg.