Caleb O. Lemley

Estrus detection remains a challenge in the dairy industry due to the labor-intensive management of traditional methods and the high cost of wearable biosensors. Computer vision offers an affordable alternative for behavior recognition and individual animal identification. However, concerns of poor lighting conditions degrading model performance are prevalent in recent literature. This study evaluated the relative importance of training dataset diversity for computer-vision-based individual animal identification and estrus detection utilizing YOLOv5su, YOLOv8s, YOLO11s, and YOLO26s. Twenty non-pregnant Holstein-Friesian cows were monitored over two 7-day trials using a 5-camera surveillance system with near-infrared capabilities. From 29 h of estrus behavior-containing footage, datasets were created for low-light, night-vision and combined conditions. Thirty-six models were fine-tuned, and precision, recall mAP50, mAP50-95, F1-score, and inference time were evaluated. Lighting condition had a significant effect on model performance, with models trained on night-vision images underperformed compared to those trained on low-light images for precision and mean average precision (p = 0.02, p < 0.01). Additionally, combined models achieved superior performance on night-vision validation datasets when compared to low-light models and better on low-light validation datasets than night-vision models for all accuracy metrics (p < 0.05 for all). In contrast, differences between YOLO architectures were limited to inference times. Collectively, these results indicate that the deliberate diversification of the training dataset, particularly with respect to lighting, has a greater impact on model robustness and practicality than selection among recent YOLO iterations, making combined-lighting training an efficient strategy for continuous estrus monitoring in commercial dairy facilities.

Summary: Colostrum is the first and essential meal in a calf's life; however, research investigating the microbial and metabolic profiles of colostrum in relation to calf gut microbiota remains limited. In this study, 15 bacterial species identified in colostrum were persistently detected in calf feces throughout the first month of life, indicating their potential to colonize the gut and influence the microbiota development of calves. Based on bacterial composition, colostrum samples were classified into 3 clusters, though no significant differences were observed in the metabolomic profiles between the clusters. Nevertheless, multiple associations were detected between colostrum bacteria and metabolites, including 4 of the 15 shared species that were positively correlated with microbial-derived metabolites. These microbial-metabolite interactions warrant further investigation to clarify their potential roles in colostrum quality and calf health. Created in BioRender.

Gestational nutrient restriction can reduce fetal growth efficiency, while melatonin supplementation can mitigate some of these negative outcomes. This study investigated placental fatty acid (FA) transporter gene expression and maternal-fetal FA concentrations in cattle supplemented with dietary melatonin during late gestational nutrient restriction. Brangus heifers (n = 29) were fed a diet of 100% of National Research Council (NRC) requirements (ADQ) or 60% of NRC requirements (RES) from day 160 to day 240 of gestation. These groups were then randomly assigned to receive 20 mg of melatonin daily or no supplement, resulting in four treatment groups (ADQ-CON, n = 7; ADQ-MEL, n = 7; RES-CON, n = 7; RES-MEL, n = 8). On d 240, heifers underwent cesarean sections, in which amniotic fluid and blood samples from both the dam and fetus were collected to determine FA profiles. Placentomes were collected to determine transporter transcript abundance. Maternal concentrations of monounsaturated FA (P = 0.001) and total omega-3 (P = 0.041) were increased in RES versus ADQ fed dams. Melatonin supplementation did not alter maternal FA profiles (P ≥ 0.081). Fetal concentrations of total omega-6 (n-6) were increased (P < 0.05) in RES-CON versus ADQ-CON, whereas ADQ-MEL and RES-MEL did not differ. Fetal concentrations of total branch chain fatty acids (BCFA) were increased (P < 0.05) in ADQ-MEL versus RES-MEL. Interestingly, maternal concentrations of C16:0 were increased (P = 0.015) in RES vs ADQ fed dams, while the opposite was observed for amniotic fluid concentrations of C16:0 which were decreased (P = 0.003) in RES vs ADQ fed dams. Amniotic fluid concentrations of total BCFA were decreased (P = 0.019) in RES versus ADQ fed dams. Both caruncular and cotyledonary transcript abundance of CASR were decreased (P < 0.05) in RES-CON versus ADQ-CON, while melatonin supplementation rescued this depression in the cotyledon. Both caruncular and cotyledonary transcript abundance of SLC27A1 were increased (P < 0.05) in RES versus ADQ fed dams. These results indicate that maternal melatonin supplementation differentially altered placental FA transporter transcript abundance and fetal FA profiles, demonstrating context-dependent effects under adequate versus restricted maternal nutrition. Maternal nutrition during gestation, especially during key periods of fetal growth and development, is vital for producing a calf capable of meeting its genetic potential. Reduced growth efficiency in offspring is detrimental to producer profitability because maternal nutrient repartitioning prioritizes development of central nervous system over skeletal, muscular, and immune systems. Previous studies have investigated the use of melatonin supplementation to mitigate the effects of nutrient restriction during gestation. Melatonin increased fetal birth weights as well as negating cellular stress prior to birth. The present study investigated the effect of maternal nutrition and dietary melatonin supplementation on fatty acid concentrations and placental fatty acid transporters. Nutrient restriction increased maternal fatty acid concentrations, while melatonin did not influence maternal levels. Melatonin altered gene expression of nutrient transporters in the placenta leading to several nutrient restriction by melatonin interactions for fetal concentrations of fatty acids. In general, fetal fatty acids were altered in nutrient restricted animals supplemented with dietary melatonin. Furthermore, amnionic fluid experienced a depression in branch chain fatty acids when dams were nutrient restricted which was not rescued by melatonin supplementation.

Fetal development is a critical period that establishes reproductive efficiency and herd performance depending on in-utero epigenetic modifications. Dietary restrictions may affect fetal testis development and the offspring fertility. Several studies have connected genetic instability to circadian cycle disruptions, including epigenetic modifications to melatonin, a key regulator. On day 160 of gestation, 17 male-bearing Brangus heifers were assigned to one of four groups in a 2 × 2 factorial treatment arrangement: adequately fed (ADQ; 100% NRC recommendation, n = 3), nutrient restricted (RES; 60% NRC recommendation, n = 5), or ADQ or RES supplemented with 20 mg/d melatonin (ADQ-MEL, n = 5; RES-MEL, n = 4). On day 240 of gestation, heifers underwent Cesarean sections to collect fetuses and testicular tissues. The fetal testicular tissue was processed and analyzed using the Methyl-MiniSeq Service: Genome-wide bisulfite sequencing (Methyl MiniSeq-GWBS). Sequence reads from Methyl Mini-Seq libraries were identified using standard Illumina platform calling software for methylome profile. RNA-Seq libraries were then sequenced on the Illumina platform for transcriptome profile. The common genes between differentially methylated regions (DMRs) and differentially expressed genes (DEGs) across different treatment groups were identified by an overlap analysis using bedtools v2.31.1. There were 413 DMRs in RES-CON and ADQ-CON testicular tissues, without differential gene expression. Compared with the ADQ-CON group, the ADQ-MEL group showed 411 DMRs and a higher KYAT1 gene expression (P-adj <0.05) without methylation changes. Comparing RES-MEL with RES-CON showed that 9 genes (DAAM1, COL28A1, RPL10, TRPM3, SLIT, ARHGEF40, SYT1, TMEM35B, CSPG4B) were expressed more in the former (P-adj <0.05). The only hypomethylated gene was DAAM1 located on chromosome 10. However, 13 genes (PTPRU, snRNP-E, TMEM59L, MUC5B, ANAPC15, FAM221A, SHCBPiL, PAQR5, PPP4R3C, DTNB, LncRNA, SHANK2, RIC3) showed increased expression in RES-CON vs. RES-MEL without differential methylation alterations, yet there were 370 DMRs. Five genes showed increased expression in RES-MEL compared with ADQ-MEL (P-adj <0.05), including histone H2B on chromosome 23. Two genes (PTPRU, TDRD10) showed increased expression in ADQ-MEL compared with RES-MEL (P-adj <0.05) without affecting methylation and 344 DMRs. In conclusion, dietary melatonin supplementation to nutrient restricted dams may influence fetal development as epigenomic and transcriptomic regulators are altered.

This study investigated the influence of maternal nutrient restriction and dietary melatonin supplementation on DNA methylation and gene expression in bovine placental cotyledons, with a focus on sex-specific changes. On day 160 of gestation, 29 Brangus heifers (bred to a single sire by AI) were subjected to a 2 × 2 factorial design: adequately fed (ADQ-CON, n = 7), nutrient-restricted (RES-CON, n = 7), and adequately fed or nutrient-restricted supplemented with 20 mg/d of melatonin (ADQ-MEL, n = 7; RES-MEL, n = 8). Cotyledons were collected at day 240 from 12 female and 17 male conceptuses for Methyl MiniSeq-GWBS and RNA-Seq. In RES-CON vs. ADQ-CON, 93 hypomethylated and 143 hypermethylated DMRs were identified, primarily in exonic, intronic, and promoter regions. Melatonin altered the methylation patterns of male and female cotyledons, respectively, with 203 and 460 DMRs associated with axon guidance, RHOC GTPase cycle, and BDNF signaling pathways. RES-MEL showed higher expression of the UBOX5 gene compared with RES-CON. Moreover, 15 DEGs (5 upregulated and 10 downregulated) were observed in the male vs. female comparison. In melatonin-treated males, PIGX, ATP11C, snoRNA U2-19, ZNF82 genes were upregulated. Thus, melatonin may modulate conceptus growth and development in a sex-specific manner.

The vaginal microbiota (VM) is the neonate’s first microbial exposure, playing a role in early health outcomes. However, the impact of VM disruption at parturition on neonatal microbial inoculation and passive transfer success remains unclear. Betadine lavages (BL) are commonly used to control reproductive tract infections but can induce microbial dysbiosis. Disrupting the VM before calving may reduce microbial competition during colonization, increasing the risk of pathogen establishment. This study aimed to determine whether VM dysbiosis affects neonatal bacterial communities and passive transfer in beef cow-calf pairs. Cows (n = 12) were randomly assigned to a control group (CON) or BL treatment group (BLG) two weeks before calving. BLG cows received 1-2 treatments based on calving date. Samples (dam colostrum, calf sera, dam-vaginal swab, dam-udder swab, dam-haircoat swab, and calf-nasal swab) were collected within 24 hours of birth and stored at -80°C. The V3-V4 region of the 16S rRNA gene was sequenced using Illumina MiSeq. Alpha diversity was assessed via two-way ANOVA, β-diversity via PERMANOVA, and taxonomic differences via LEfSe. IgG concentrations in calf serum and colostrum were analyzed using ELISA. Data were evaluated using the R software package (v. 2023.03 + 386). BL-induced VM dysbiosis did not alter α-diversity in the dam-vaginal (P = 0.42), dam-udder (P = 0.53), or dam-haircoat (P = 0.21) microbiota but did affect the calf-nasal microbiota (P = 0.03). Beta diversity was unchanged in the dam-vaginal (P = 0.66) and dam-udder (P = 0.56) microbiota but trended toward greater variation in BLG calf-nasal (P = 0.08) and dam-haircoat (P = 0.09) samples. BLG calves exhibited increased the relative abundance of the phylum Actinobacteria and decreased the relative abundance of the phylum Proteobacteria in the nasal microbiota compared with CON calves. Passive transfer status did not differ between groups (P = 0.88), but BLG dams had higher colostrum IgG concentrations than CON dams (P = 0.04), suggesting an immune response to pre-calving microbial disruption. Together, these results are indicative of microbial dysbiosis prior to parturition leading to an increased immune response in BLG dams which altered colostrum IgG. Thus, dam colostrum composition could drive immune responses against neonatal inoculating bacteria resulting in differing nasal microbial communities between BLG and CON calves.

Young beef bulls are often fed high-energy diets to maximize postweaning growth; however, overnutrition has been associated with decreased semen quality. Seminal plasma has been shown to prime the uterus for early pregnancy events in mammalian species; thus, alterations in this response caused by overnutrition could be detrimental or beneficial. The present study evaluated the impact of seminal plasma from bulls exposed to overnutrition on the transcriptome of endometrial epithelial cells in vitro. Yearling half-sibling beef bulls (n = 20) were stratified by age, body weight, and body condition score before being randomly assigned to one of two diets: 1) a diet formulated to elicit a target average daily gain (ADG) of 1.81 kg/d (High Gain, n = 10) or 2) a diet formulated to elicit a target ADG of 1.22 kg/d (Moderate Gain, n = 10). Seminal plasma was collected from each treatment group at three timepoints and then pooled and passed through a 22 μm sterile filter. Based on preliminary experiments, cells were randomly assigned to 1 of the 3 treatment groups: no treatment (Control), 1% seminal plasma from moderate gain bulls (Moderate Gain), or 1% seminal plasma from high gain bulls (High Gain) and cultured at 38.5 °C, 95% O2, 5% CO2 for 12 h. Experiments were conducted in duplicate per treatment, 3 times for a total of 6 experimental units per treatment group. Following the treatment period, total RNA was extracted from cells and underwent sequencing on the Illumina platform. Raw reads were mapped to the ARS-UCD1.3 Bos taurus RefSeq transcriptome, then converted to gene-level counts. The quasi-likelihood negative binomial generalized log-linear model (glmOL) from edgeR was used to test for significantly differentially expressed genes (DEG) at Benjamin-Hochberg adjusted (B.H. adj) P ≤ 0.05. There were 7,905 DEG (Padj ≤ 0.05) in bEND cells treated with seminal plasma from High Gain bulls compared with the Control. There were 8,933 (Padj ≤ 0.05) DEG in bEND cells treated with seminal plasma from Moderate Gain bulls compared with Control. There were 28 DEG (Padj ≤ 0.05) in bEND cells treated with seminal plasma from High Gain compared with Moderate Gain bulls. Metabolic pathways were depleted in each of these comparisons. In conclusion, seminal plasma alters the bEND cell transcriptome; however, the plane of nutrition had minimal effects.

Gossypol is a phenolic compound found in cottonseed. Phenolic compounds are plant secondary metabolites, known to have both, beneficial and negative effects on animals. The present experiment evaluated the beneficial and negative effects of cottonseed on growth, daily weight gain, and parasite infection measured by fecal egg counts, liver function, and ruminal microbiome of growing female meat goats. Forty Boer × Spanish doelings were assigned to 4 groups (n = 10 per group) in a 2 × 2 factorial design consisting of cottonseed supplementation (CS)-NP CS (0.5% body weight-BW in kg) + no parasite infection (NP) CON-NP commercial pellets or control (CON) + no parasite infection (NP); CON-P commercial pellets or control (CON) + Haemonchus contortus artificial infection (P); and CS-P CS supplementation (CS) + H. contortus artificial infection (P). Artificial infection consisted of 4,000 H. contortus infective larvae (L3). The FEC were not affected by diet. The liver function showed a diet by parasite interaction (P = 0.03) for the mean velocity of the hepatic portal vein, which was decreased in CON-P compared with CS-NP indicating a reduction in metabolism. A diet by parasite interaction (P = 0.014) was also observed for the diameter of the hepatic portal vein, which was increased in the CS-P goats compared to all other groups. Liver weights were affected by diet, being decreased in the CS animals vs the CON ones (P = 0.00), as well as the liver weight relative to the animal’s BW (P = 0.01). Diet effect (P = 0.01) was observed for the liver blood flow relative to liver weight. For the rumen microbiome, the number of bacterial species was unaffected by CS but decreased by the parasite challenge (P = 0.03). In conclusion, the supplementation of female meat goats with CS did not affect BW, daily weight gain, parasite infection, and FAMACHA©, but reduced the liver weights, influenced fiber-digesting bacteria in the rumen, and restored lipolytic bacteria in the rumen of goats with parasite challenge.

Holstein cattle carrying a prolactin receptor gene mutation (SLICK) exhibit short and sleek hair coats (short-haired Holstein [SLK]) enhancing thermotolerance and productivity compared with wild type-haired Holstein (WT) under tropical conditions. The objectives were to unravel the physiological and molecular mechanisms that confer an advantage to this slick genotype in Puerto Rico and determine potential correlations between metabolites and physiological variables. At 160 ± 3 DIM we compared vaginal temperatures (VT) and voluntary solar radiation exposure (VSRE) during 48 h between 9 SLK and 9 WT Holsteins, whereas a subsample of 7 SLK and 7 WT were used to assess udder skin temperature, mammary gland hemodynamics and transcriptomics, and blood plasma untargeted metabolomics at a single time point. The SLK cattle showed lower VT throughout the day and greater VSRE at 1000 h and 1100 h compared with their WT counterparts. Total mammary blood flow (MBF) was greater in SLK Holsteins compared with WT. The metabolite 9-nitrooctadecenoic acid was identified as a potential biomarker for MBF; moreover, SLK cattle had greater amounts of this metabolite in their plasma. Prostaglandin D2 synthase (PTGS) was upregulated in the slick mammary gland, while plasma prostaglandin D2 was positively correlated with milk yield and increased in SLK Holsteins compared with WT. Interestingly, the arachidonic acid metabolism pathway was enriched in the mammary gland transcriptome and perturbed in the blood metabolome in the SLK Holsteins. In conclusion, SLK Holsteins exhibited lower body temperatures, greater VSRE, enhanced blood supply to the mammary gland, and alterations in genes and metabolites involved in arachidonic acid metabolism at the mammary gland and blood plasma. The usage of the SLK Holstein cattle genetics in dairy operations could be a feasible alternative to mitigate the adverse consequences of heat stress.