About

Isaac Salfer is an Assistant Professor in the Department of Animal Science at the University of Minnesota Twin Cities. He holds a Ph.D. in Animal Science from The Pennsylvania State University, earned in 2019, and both a Master's and Bachelor's degree in Animal Science from the University of Minnesota-Twin Cities, obtained in 2015 and 2013 respectively. His research focuses on understanding how circadian and seasonal rhythms impact metabolism and behavior in dairy cows to influence feed efficiency. He also studies the effects of novel feeds and feed additives on rumen nitrogen and energy balance, as well as neonatal nutrition's impact on calf growth, feed efficiency, and health. Salfer is actively involved in teaching courses such as Applied Dairy Nutrition, Animal Bioenergetics, and Ruminant Nutrition, often serving as a guest lecturer. His contributions include numerous peer-reviewed publications and conference presentations that advance knowledge in dairy nutritional physiology and animal welfare.

Research topics

• Biology
• Animal science
• Medicine
• Internal medicine
• Physics
• Endocrinology
• Food science
• Chemistry
• Mechanics
• Environmental science
• Biochemistry
• Mathematics

Selected publications

• Evaluation of 30 novel plant secondary compounds for gas production, methane concentration, nutrient digestibility and volatile fatty acid concentrations using in vitro batch culture fermentation

  The Journal of Agricultural Science · 2026-01-01

  articleOpen accessSenior authorCorresponding

  Abstract The objective of the current study was to evaluate 30 previously uncharacterised pure plant secondary metabolites (PSM) for effects on in vitro gas production (GP) and methane concentration. Purified compounds ( n = 4) were incubated in buffered rumen fluid for 48 h at a rate of 25 mg per g substrate. Gas production was measured using ANKOM RF pressure analysers and gas composition was measured using gas chromatography. Dry matter and fibre digestibility and volatile fatty acid (VFA) concentrations were determined. Data were analysed using a linear mixed model with fixed effect of treatment, random effect of experimental run and blank as a covariate and Dunnett’s test to compare each treatment to a control. Sabenine, apigenin, galangin, isoliquiritigenin, quercetin, rutin, vitexin, abscisic acid and uridine reduced methane concentration, with a tendency for reduction by kaempferol. Sabenine, thymol, apigenin, quercetin, rutin, vitexin, abscisic acid and uridine reduced methane production, with the greatest reduction occurring for rutin (81.8 %), vitexin (81.7 %) and thymol (80.5 %). Eight compounds increased methane production compared to control, with stigmasterol having the greatest increase (173.0 %). Out of the compounds that reduced methane emissions, a reduction in fibre digestibility was observed for all except quercetin and thymol. Minimal effects of PSM on VFA profile were observed, with myrtenal increasing proportion of acetate, alpha-humulene, alpha-longipinene and beta-caryophyllene increasing proportion of propionate and thymol, apigenin, hyperoside and verbenone increasing proportion of butyrate. Results suggest that quercetin and thymol warrant further exploration as potential feed additives to reduce methane emissions.

  PublisherDOI

• Comparison of glucose concentration measured in samples collected with different anticoagulants and analyzed using 2 glucose quantification methods

  JDS Communications · 2025-04-03

  articleOpen access

  -EDTA and Li-Hep anticoagulant could be used for glucose quantification to avoid cumbersome collection of several tubes, but results must be reviewed with caution as concentrations would be higher compared with those quantified with NaF. This observation is also conserved when quantifying glucose using the PGO and HK methods.

  PublisherDOI

• Comparison of glucose concentrations in samples frozen for varying durations and subjected to freeze-thaw cycles using 2 assay systems

  JDS Communications · 2025-02-20 · 1 citations

  articleOpen access

  <h2>Abstract</h2> Although laboratory best practices recommend that total storage time and number of freeze-thaw cycles should be minimized before analysis of metabolites within plasma, the specific impacts of storage time and freeze-thaw cycles on apparent glucose concentrations in bovine plasma have not been well quantified. Additionally, within the field of dairy science, multiple assay systems for glucose concentration analysis are used, yet these assays have not been directly compared with each other. Thus, the first objective was to investigate the stability of glucose in samples that have been frozen for an increased duration or frozen and thawed up to 4 times. The second objective was to investigate the agreement between 2 assay systems for glucose measurement: the peroxidase and glucose oxidase (PGO) assay system and the hexokinase reaction (HK) assay system. Blood samples were collected into sodium fluoride (NaF) tubes from 166 dairy cows, across different lactation stages, on 3 farms over a 2-d interval. After centrifugation, plasma was separated into 4 aliquots. One aliquot was used to determine the impacts of up to 4 freeze-thaw cycles, and the first measurement from this aliquot along with the other 3 aliquots were used to investigate potential glucose degradation in aliquots frozen for different lengths of time, including 2, 3, 4, and 16 wk. All samples were analyzed using both the PGO and HK assay systems. Bland–Altman plots, Pearson correlations, and paired <i>t</i>-tests were used to compare the effects of storage time and freeze-thaw cycles. Additionally, a linear mixed model with the fixed effect of either storage time or freeze-thaw cycle and the random effect of cow within farm was used to consider the effect of each of these factors on measured glucose concentrations. For comparisons between assay systems, a linear mixed model was used with the fixed effect of assay system, time, frozen or freeze-thaw cycles depending on the model, their interaction, and cow within farm as a random effect. Storage time and freeze-thaw cycles influenced glucose concentration measured by the PGO and HK. Concentrations of glucose both increased and decreased with increasing storage time and freeze-thaw cycles and compared with the initial quantification. Apparent glucose concentration was more variable across all aliquots when the HK assay (± 0.14 mmol/L) was used compared with the PGO assay (± 0.06 mmol/L). The HK assay continuously yielded lower concentrations of glucose compared with the PGO assay. Overall, glucose concentrations can be variable over different storage times and freeze-thaw cycles; however, they cannot be compared across the PGO and HK assay systems.

  PublisherOA PDFDOI

• The effects of saturated fatty acid supplements on plasma and milk concentration of fatty acid esters of hydroxy fatty acids in dairy cows

  JDS Communications · 2025-09-11

  articleOpen access

  Fatty acid esters of hydroxy fatty acids (FAHFA) are a novel class of bioactive lipids with demonstrated antidiabetic and anti-inflammatory properties in rodent models and humans but have not been investigated in cows. The major FAHFA are synthesized from palmitic (PA), stearic (SA), and oleic acid. The PA can be esterified to hydroxy fatty acids, such as hydroxy PA or hydroxy SA, or SA can be esterified to hydroxy SA, forming PAHPA, PAHSA, or SAHSA, respectively. The objective of the current study was to determine the effect of increasing intake of PA, SA, or both PA and SA on nonesterified FAHFA in the plasma and milk of dairy cows. We hypothesized that increasing PA and SA in the diet would increase PA and SA containing FAHFA in plasma and milk. Samples were analyzed from a previous experiment that used 12 multiparous Holstein cows in a 4 × 4 Latin square design. Treatments were a no-fat supplement control (CON) and fat supplements that were high in PA (91% C16:0), high in SA (92.6% C18:0), or contained a blend of PA and SA (PA/SA; 45.3% C16:0 and 49.1% C18:0) at 1.95% of diet DM. The concentrations of nonesterified FAHFA in plasma and milk fat were quantified using liquid chromatography tandem MS, and data were analyzed using a mixed model that included treatment as a fixed effect and cow and period as random effects. The relationship between plasma FAHFA and milk production variables were analyzed using regression analysis. Five nonesterified FAHFA (9-PAHPA, 5-PAHSA, 9-PAHSA, 10-PAHSA, and 9-SAHSA) were quantified in plasma and all were affected by treatment. Plasma concentration of 9-PAHPA was increased 2.9-fold by PA compared with CON, whereas 9-SAHSA was increased 2.7-fold by SA compared with CON. The concentrations of 5-PAHSA, 9-PAHSA, and 10-PAHSA were highest with PA/SA. In milk, 8 nonesterified FAHFA were quantified, and only 12-PAHSA was increased by SA and 12-PAHPA tended to be increased by PA. Plasma 9-PAHPA was positively associated with milk fat yield and mixed FA and negatively associated with milk preformed FA, de novo FA, and odd- and branched-chain fatty acids, whereas plasma 9-SAHSA was positively associated with milk preformed FA. Overall, FA supplements affected nonesterified FAHFA concentration in plasma, demonstrating a direct effect of dietary FA on this emerging regulator of metabolism. There were limited effects of FA supplements on nonesterified FAHFA in milk fat. Functional roles for these lipids require further exploration.

  PublisherOA PDFDOI

• Impacts of feeding milk replacer containing 20% or 24% protein and fat on growth and feed efficiency of beef × dairy crossbred calves

  JDS Communications · 2025-05-11 · 1 citations

  articleOpen accessSenior author

  It is becoming more prominent for dairy producers to breed their lower genetic merit cows and heifers to beef sires to increase the value of calves when they are sold to beef markets. However, data are currently lacking on optimal feeding strategies for rearing these crossbred calves until weaning. Therefore, the objective of the current study is to compare the effects of feeding 2 milk replacers (20% CP and 20% fat [MR20] vs. 24% CP and 24% fat [MR24]) on calf growth, feed efficiency, and scouring frequency. Forty individually housed 2- to 5-d-old Angus × Holstein crossbred calves with a BW ranging from 34.0 to 47.6 kg were randomly assigned to 1 of 2 milk replacer (MR) treatments. Calves were fed 2.84 L of their respective MR twice per day (5.68 L total) from d 1 to 42 of the experiment and 2.84 L once per day from d 43 to 49. Feeding crossbred calves MR24 increased ADG and gain per feed by 23.6% and 9.80%, respectively, compared with calves fed MR20 from d 1 to 56. Feeding crossbred calves MR24 tended to increase calf starter intake and total feed intake compared with calves receiving MR20. Results suggest feeding a MR of higher nutritive value is a simple and effective strategy to increase weight gain and feed efficiency of beef × dairy calves.

  PublisherOA PDFDOI

• Effects of timing of abomasal infusion of fatty acids on the daily rhythms of milk synthesis and plasma hormones and metabolites in dairy cows

  Journal of Dairy Science · 2025-05-16

  articleOpen access1st authorCorresponding

  Dairy cows display daily rhythms of milk synthesis that appear to be driven by a circadian clock located in the mammary gland. These rhythms are altered by the time of feed availability. Fatty acids have been shown to entrain circadian rhythms in liver and adipose tissue in experimental models, but their role in the mammary gland has not been well investigated. Our objective was to determine the effects of the timing of fatty acid absorption on the daily rhythms of milk synthesis. Nine lactating Holstein cows were arranged in a 3 × 3 Latin square design. Treatments were abomasal infusions of 350 g/d of a free fatty acid stock enriched in cis-9 18:1 either continuously throughout the day for 22 h (CON) or for 8 h from 0900 to 1700 h (DAY) or from 2100 to 0500 h (NGT). Treatment periods were 12 d with a 5-d washout. Cows were milked every 6 h during the final 7 d of each period to determine the daily patterns of milk synthesis. A 24-h rhythm was fit to time course data using cosine analysis, and the amplitude and acrophase (time at peak) were determined. Daily milk and milk protein yields were decreased by DAY and NGT compared with CON, whereas milk fat yield was not changed. Milk yield fit a 24-h rhythm in CON and DAY but not in NGT. Furthermore, DAY delayed the peak of the daily rhythm of milk yield by 2 h compared with CON. Fat and protein concentrations exhibited daily rhythms in CON and NGT but not DAY. Fat yield only fit a 24-h rhythm in DAY. Both de novo and mixed-source fatty acid yields were reduced by DAY and NGT, suggesting that the faster infusion rates may have resulted in concentrations of fatty acids that exceeded a threshold sufficient to inhibit de novo fatty acid synthesis. Plasma glucose concentration failed to display a daily rhythm in any treatment, whereas nonesterified fatty acids showed a rhythm in CON and NGT, but this rhythm was abolished by DAY. Insulin fit a rhythm in NGT and tended to fit a rhythm with a lower amplitude in CON, but no rhythm was present in DAY. Blood urea nitrogen exhibited a daily rhythm under all treatments, and both the mean and amplitude were increased by DAY. Daily rhythms of milk synthesis were also modified by DAY, with a slight delay in the daily peak of milk yield and elimination of the rhythms of milk fat and protein concentrations. Infusion at night had little effect. Daytime infusion also modified the daily rhythms of plasma metabolites by reducing the amplitude of nonesterified fatty acid concentration and increased the amplitude of blood urea nitrogen.

  PublisherOA PDFDOI

• Assessment of lactoferrin supplementation in milk replacer on calf growth, feed efficiency and scouring incidence preweaning and postweaning

  Journal of Dairy Research · 2025-02-01

  articleOpen accessSenior authorCorresponding

  We investigated the hypothesis that supplementing milk replacer (MR) with exogenous lactoferrin (LF) would improve average daily gain (ADG) and feed efficiency and decrease scouring incidence in dairy calves. Lactoferrin is an antimicrobial and anti-inflammatory glycoprotein naturally found in bovine colostrum and milk that is low in MR. Previous studies suggest that supplementing LF to MR enhances ADG and feed efficiency while reducing disease occurrence in pre-weaning dairy calves. In our experiment, 103 Holstein heifer calves were randomly assigned to 1 of 4 treatments in a randomized complete block design from birth to d56 of age. Each calf received 340.1 g/d of 24% protein, 20% fat basal MR fed twice daily from d1 to 42 and once daily from d43 to 49, supplemented with 0 (L0), 1 (L1), 2 (L2) or 4 (L4) g/hd/d of LF treatment (45% purity). Calves were weaned at 49d of age. Body weight was measured at d1, 14, 28, 42, 49 and 56 of age. Faecal scores were measured weekly. Milk replacer and calf starter intake was measured daily and calculated biweekly. Data were analyzed using a linear mixed model with fixed effects of LF inclusion, and random effects of source herd and nursery room. In the first two weeks of life, ADG and gain-to-feed ratio (G:F) were numerically (non-statistically) increased in L4 tended compared with L1 and L2, but this effect was not maintained throughout the rest of the pre-weaning period or entire experiment. Average faecal score during the entire 56d experiment was greater in L2 compared with L0, L1 and L4, although faecal scores of all treatment groups were generally low. Under the conditions of the present study, LF supplementation at the inclusion levels provided showed minimal effects on feed intake, growth rate or calf health.

  PublisherOA PDFDOI

• Daily rhythms of glucose, insulin, and nonesterified fatty acid responses to an intravenous glucose tolerance test in dairy cows

  Journal of Dairy Science · 2025-03-25 · 2 citations

  articleOpen accessSenior author

  In nonruminant species, glucose tolerance and insulin sensitivity are known to be regulated by circadian rhythms, which are repeating ∼24-h cycles that govern many aspects of behavior, physiology, and metabolism. However, it is unknown if these rhythms exist in dairy cows. Our objective was to determine the fit of a daily rhythm of glucose, insulin, and nonesterified fatty acid (NEFA) clearance rates independent of daily patterns of nutrient intake. To accomplish our objective, 12 multiparous lactating Holstein cows were enrolled in a within-subject design conducted over 2 experimental periods (n = 6/period). Within each period, cows were subjected to intravenous glucose tolerance tests (IVGTT) at 4 timepoints, representing different times of the day (0300, 0900, 1500, and 2100 h). The 0900 and 2100 h IVGTT were performed 36 h apart, followed by a 7-d washout, and then the 1500 and 0300 h IVGTT were performed 36 h apart. Cows were fed 12 times/d at 2-h intervals beginning 24 h before the first IVGTT in each set until the second IVGTT in each set to stabilize feed intake across the day, with 1 time/d feeding occurring during the washout period. For each IVGTT, 250 g of glucose was infused as a 50% (wt/vol) d-glucose solution via a jugular catheter and blood was collected at -15, -5, immediately before, 0, 5, 10, 15, 20, 30, 45, 60, 90, and 120 min relative to infusion. A linear mixed model with the fixed effects of cosine and sine and random effect of cow within period was used for the outcomes of clearance rate, half-life, baseline concentration, time to baseline concentration, and area under the curve (AUC) for glucose, insulin, and NEFA. A zero-amplitude test was used to determine the fit of a 24-h cosine function and cosinor rhythmometry was used to determine the amplitude and acrophase of the 24-h rhythm. Insulin concentrations at baseline followed a diurnal rhythm. Glucose and insulin clearance rate, half-life, and AUC also followed a diurnal rhythm. Glucose and insulin clearance rates peaked at 1247 h and 0944 h, respectively. No circadian rhythm was detected for plasma NEFA concentrations. Results suggest that insulin-stimulated glucose uptake is controlled differently throughout the day by circadian rhythms.

  PublisherOA PDFDOI

• Associations between hyperketonemia and the rumen metagenome and metabolome in Holstein cows during the first 2 weeks postpartum

  Animal Microbiome · 2025-06-04 · 2 citations

  articleOpen access

  Hyperketonemia (HYK) is characterized by elevated levels of blood ketone bodies in dairy cows. Despite previous research on the physiological events related to HYK, associations between the rumen metagenome, metabolome, and HYK have not been well described. Therefore, the objective of this study was to compare rumen metagenome and metabolome profiles of cows with naturally occurring HYK to those without HYK during the first two weeks postpartum. Blood beta-hydroxybutyrate (BHB) concentrations at d 5, 10 and 14 postpartum were used to classify 16 rumen cannulated Holstein cows as either hyperketonemic (+ HYK; ≥1.2 mmol/L BHB at any of the collection days) or non-hyperketonemic (-HYK). Five + HYK cows were identified and were paired with 5 -HYK cows based on parity and calving date. Microbial DNA was extracted from rumen fluid and sequenced using shotgun metagenomics with the Illumina platform. Kraken2 was used to map reads to microbial taxonomic groups and Humann3.8 was used to predict potential functions. Metabolome profiling of rumen fluid was conducted using high-resolution liquid chromatography-mass spectroscopy. MetaboAnalyst6.0 was used to identify potential changes in metabolic pathways. Metagenomes and metabolomes comparisons were conducted using mixed models that included the fixed effects of group, day, their interaction, and the random effect of cow. There was minimal difference detected in alpha diversity for the metagenome, but differences in the metabolome were detected by HYK status. The concentration of asparagine and p-cresol was greater in + HYK cows compared to -HYK, but citrulline was greater in -HYK cows throughout all days considered. On d5 the concentration of ornithine was greater in + HYK compared to-HYK cows, and on d10 acetate was greater in the -HYK cows. There were no differences detected in the pathway analysis from the metabolites quantified by HYK status. Overall, modest differences in rumen metabolome were observed between + HYK and -HYK cows in early lactation. Future studies should explore associations between the rumen environment and HYK as this could be informative for treatment and management practices.

  PublisherOA PDFDOI

• Comparisons of Feed Bunk Nutrient Consistency, Milk Production and Cow Behavior Between Herds Using Automated Milking Systems With or Without Automated Feeding Robots

  Animals · 2025-04-11 · 1 citations

  articleOpen accessSenior authorCorresponding

  Automated feeding robots (AFR) are increasingly being used on North American dairy farms to reduce dependency on human labor for feeding. These systems mix, deliver, and push up feed to cows at any frequency or interval desired, allowing for more frequent feed delivery than conventional feeding systems (CFS). This observational study investigated differences in ration consistency, milk components, milk fatty acid profile, and cow behavior between herds using AFR and those using CFS. Sixteen commercial dairies with automated milking systems (AMS) in the upper Midwest United States were paired based on herd size and location into eight blocks each consisting of one CFS and one AFR herd. Feed bunk samples were collected at four equally spaced time points for 3 consecutive d and analyzed for coefficient of variation (CV) of nutrient composition and particle size distribution. Bulk tank milk samples were collected 1 ×/d for 3 d and analyzed for fat, protein, milk urea nitrogen (MUN), lactose, and milk fatty acid (FA) profile. Daily AMS visit intervals, milk yield and composition, and rumination time data were collected from AMS software. A linear mixed model tested fixed effects of feeding system, block, and the random effect of day nested within block. The CV of feed bunk DM, ADF, NDF, and lignin was lower in AFR. Bulk tank milk fat, protein, and MUN were not different between AFR or CFS. AFR had a greater proportion of de novo synthesized FA, but no difference in preformed or mixed FA. Herds with AFR had a shorter AMS visit interval with more AMS refusals per day than CFS. Results imply that AFR may be associated with lower daily variation in fiber concentration at the feed bunk, increased mammary de novo fatty acid synthesis, and increased frequency of cow visits to the AMS compared to conventional PMR feeding.

  PublisherOA PDFDOI

Frequent coauthors

• K.J. Harvatine

  Cornell University

  16 shared

• M.D. Stern

  5 shared

• Paul A. Bartell

  Pennsylvania State University

  4 shared

• K. Nedelkov

  Trakia University

  4 shared

• C. Matamoros

  Pennsylvania State University

  4 shared

• C.D. Dechow

  Pennsylvania State University

  3 shared

• Ghazanfar Ali Chishti

  University of Veterinary and Animal Sciences

  3 shared

• Yun Ying

  University of Pennsylvania

  2 shared

Labs

• Department of Animal SciencePI

Education

• Ph.D., Animal Science

  Pennsylvania State University

  2019

• M.S., Animal Science

  University of Minnesota System

  2015

• B.S., Animal Science

  University of Minnesota System

  2013

Awards & honors

• Investigating Circadian Rhythms’ Impact on Animal Welfare in…