Douglas M. Liebe
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Douglas M. Liebe has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Robin R. White
School of Animal Sciences, Virginia Tech, Blacksburg, VA, United States
4 shared publications - Albert De Vries
Department of Animal Sciences, University of Florida, Gainesville, FL 32611, USA
2 shared publications - C. S. Petersson-Wolfe
Department of Dairy Science, Virginia Tech, Blacksburg 24061
2 shared publications - Claire B. Gleason
Department of Animal and Poultry Sciences, Virginia Tech, Blacksburg, VA 24061, USA
2 shared publications - Kristy M. Daniels
School of Animal Sciences, Virginia Tech, Blacksburg, VA 24061
2 shared publications - Mary Beth Hall
US Dairy Forage Research Center, USDA-ARS, Madison, WI 53706
2 shared publications
Latest Publications
- Open Access
- 1 March 2022
- Published by American Dairy Science Association in Journal of Dairy Science
- Vol. 105 (3) , 2369-2379
Practical challenges and potential approaches to predicting low-incidence diseases on farm using individual cow data: A clinical mastitis example
Clinical mastitis (CM) incidence is considerable in terms of cows affected per year, but cases are much less common in terms of detections per cow per milking. From a modeling perspective, where predictions are made every time any cow is milked, low CM incidence per cow day makes training, evaluating, and applying CM prediction models a challenge. The objective of this study was to build models for predicting CM incidence using time-series sensor data and choose models that maximize net return based on a cost matrix. Data collected from 2 university dairy farms, the University of Florida and Virginia Polytechnic Institute and State University, were used to gather representative data, including 110,156 milkings and 333 CM cases. Variables used in the models were milk yield, protein, lactose, fat, electrical conductivity, days in milk, lactation number, and activity as the number of steps, lying time, lying bouts, and lying bout duration. Models that predicted either likelihood of CM caused by gram-negative (GN) or gram-positive (GP) bacteria on each day were derived using extreme gradient boosting with weighting favoring true-positive cases, logistic responses, and log-loss errors. Model accuracies were determined using data randomly held out from the training set on each run. All variables considered were in terms of change (slope) over previous days, including the day CM was visually detected. The GN models had a median sensitivity (Se) of 52.6% and specificity (Sp) of 99.8%, whereas the GP models had a median Se of 37.5% and Sp of 99.9% when tested on the held-out data. In our models optimized to reduce cost from predictions, the Se was much less than Sp, suggesting that CM models might benefit from greater model weighting placed on Sp. Results also highlight the importance of positive predictive value (true positive cases per predicted positive case) along with Sp and Se, as models built on sparse data tend to predict too many false-positive cases. The calculated partial net return of our GN and GP models were -$0.15 and -$0.10 per cow per lactation, respectively, whereas International Organization for Standardization (ISO) standard models with Se of 80% and Sp of 99% would return -$1.32 per cow per lactation. Models chosen that minimized the cost to the farmer differed markedly from models that met ISO guidelines, showing asymmetry in targets between Sp and Se when the disease incidence rate is low. Because of the unique challenges that low-incidence diseases like CM present, we recommend that future CM predictive models consider the economic and practical implications in addition to the traditional model evaluation metrics.
- 1 March 2022
- Published by American Dairy Science Association in Journal of Dairy Science
- Vol. 105 (3) , 2369-2379
Open AccessPractical challenges and potential approaches to predicting low-incidence diseases on farm using individual cow data: A clinical mastitis example
Clinical mastitis (CM) incidence is considerable in terms of cows affected per year, but cases are much less common in terms of detections per cow per milking. From a modeling perspective, where predictions are made every time any cow is milked, low CM incidence per cow day makes training, evaluating, and applying CM prediction models a challenge. The objective of this study was to build models for predicting CM incidence using time-series sensor data and choose models that maximize net return based on a cost matrix. Data collected from 2 university dairy farms, the University of Florida and Virginia Polytechnic Institute and State University, were used to gather representative data, including 110,156 milkings and 333 CM cases. Variables used in the models were milk yield, protein, lactose, fat, electrical conductivity, days in milk, lactation number, and activity as the number of steps, lying time, lying bouts, and lying bout duration. Models that predicted either likelihood of CM caused by gram-negative (GN) or gram-positive (GP) bacteria on each day were derived using extreme gradient boosting with weighting favoring true-positive cases, logistic responses, and log-loss errors. Model accuracies were determined using data randomly held out from the training set on each run. All variables considered were in terms of change (slope) over previous days, including the day CM was visually detected. The GN models had a median sensitivity (Se) of 52.6% and specificity (Sp) of 99.8%, whereas the GP models had a median Se of 37.5% and Sp of 99.9% when tested on the held-out data. In our models optimized to reduce cost from predictions, the Se was much less than Sp, suggesting that CM models might benefit from greater model weighting placed on Sp. Results also highlight the importance of positive predictive value (true positive cases per predicted positive case) along with Sp and Se, as models built on sparse data tend to predict too many false-positive cases. The calculated partial net return of our GN and GP models were -$0.15 and -$0.10 per cow per lactation, respectively, whereas International Organization for Standardization (ISO) standard models with Se of 80% and Sp of 99% would return -$1.32 per cow per lactation. Models chosen that minimized the cost to the farmer differed markedly from models that met ISO guidelines, showing asymmetry in targets between Sp and Se when the disease incidence rate is low. Because of the unique challenges that low-incidence diseases like CM present, we recommend that future CM predictive models consider the economic and practical implications in addition to the traditional model evaluation metrics.
- Open Access
Short-Term Adaptation of Dairy Cattle Production Parameters to Individualized Changes in Dietary Top Dress
Immediate and short-term changes in diet composition can support individualized, real-time interventions in precision dairy production systems, and might increase feed efficiency (FE) of dairy cattle in the short-term. The objective of this study was to determine immediate and short-term effects of changes in diet composition on production parameters of dairy cattle fed varying amounts of top dressed commodities. A 4 × 4 replicated Latin square design was used to evaluate responses of twenty-four Holstein cows fed either no top dress (Control) or increasing amounts of: corn grain (CG), soybean meal (SBM), or chopped mixed grass hay (GH) top dressed on a total mixed ration (TMR) over four, 9-day periods. Throughout each period, top dressed commodities were incrementally increased, providing 0% to 20% of calculated net energy of lactation (NEL) intake. Measured production responses were analyzed for each 9-d period using a mixed-effects model considering two different time ranges. Samples collected from d 3 and 4 and from d 7 and 8 of each period were averaged and used to reflect “immediate” vs. “short-term” responses, respectively. In the immediate response time frame, control fed cows had lower milk yield, milk fat yield, and milk true protein yield than CG and SBM supplemented animals but similar responses to GH supplemented animals. Milk fat and protein percentages were not affected by top dress type in the immediate term. In the short-term response time-frame, GH supplemented animals had lower DMI and milk fat yield than all other groups. Control and GH supplemented cows had lower milk yield than CG and SBM fed cows. In the immediate response time frame, FE of SBM supplemented cows was superior to other groups. In the short-term time frame, FE of GH and SBM groups was improved over the control group. Results suggest that lactating dairy cows show rapid performance responses to small (L) changes in dietary composition, which may be leveraged within automated precision feeding systems to optimize efficiency of production. Before this potential can be realized, further research is needed to examine integration of such strategies into automatic feeding systems and downstream impacts on individual animal FE and farm profitability.
Open AccessShort-Term Adaptation of Dairy Cattle Production Parameters to Individualized Changes in Dietary Top Dress
Immediate and short-term changes in diet composition can support individualized, real-time interventions in precision dairy production systems, and might increase feed efficiency (FE) of dairy cattle in the short-term. The objective of this study was to determine immediate and short-term effects of changes in diet composition on production parameters of dairy cattle fed varying amounts of top dressed commodities. A 4 × 4 replicated Latin square design was used to evaluate responses of twenty-four Holstein cows fed either no top dress (Control) or increasing amounts of: corn grain (CG), soybean meal (SBM), or chopped mixed grass hay (GH) top dressed on a total mixed ration (TMR) over four, 9-day periods. Throughout each period, top dressed commodities were incrementally increased, providing 0% to 20% of calculated net energy of lactation (NEL) intake. Measured production responses were analyzed for each 9-d period using a mixed-effects model considering two different time ranges. Samples collected from d 3 and 4 and from d 7 and 8 of each period were averaged and used to reflect “immediate” vs. “short-term” responses, respectively. In the immediate response time frame, control fed cows had lower milk yield, milk fat yield, and milk true protein yield than CG and SBM supplemented animals but similar responses to GH supplemented animals. Milk fat and protein percentages were not affected by top dress type in the immediate term. In the short-term response time-frame, GH supplemented animals had lower DMI and milk fat yield than all other groups. Control and GH supplemented cows had lower milk yield than CG and SBM fed cows. In the immediate response time frame, FE of SBM supplemented cows was superior to other groups. In the short-term time frame, FE of GH and SBM groups was improved over the control group. Results suggest that lactating dairy cows show rapid performance responses to small (L) changes in dietary composition, which may be leveraged within automated precision feeding systems to optimize efficiency of production. Before this potential can be realized, further research is needed to examine integration of such strategies into automatic feeding systems and downstream impacts on individual animal FE and farm profitability.
- 1 August 2021
- Published by American Dairy Science Association in Journal of Dairy Science
- Vol. 104 (8) , 8857-8869
Rumen volatile fatty acid molar proportions, rumen epithelial gene expression, and blood metabolite concentration responses to ruminally degradable starch and fiber supplies
The objective of this work was to characterize rumen volatile fatty acid (VFA) concentrations, rumen epithelial gene expression, and blood metabolite responses to diets with different starch and fiber sources. Six ruminally cannulated yearling Holstein heifers (body weight = 330 ± 11.3 kg) were arranged in a partially replicated Latin square experiment with 4 treatments consisting of different starch [barley (BAR) or corn (CRN)] and fiber [timothy hay (TH) or beet pulp (BP)] sources. Treatments were arranged as a 2 × 2 factorial. Beet pulp and TH were used to create relative changes in apparent ruminal fiber disappearance, whereas CRN and BAR were used to create relative changes in apparent ruminal starch disappearance. Each period consisted of 3 d of diet adaptation and 15 d of dietary treatment. In situ disappearance of fiber and starch were estimated from bags incubated in the rumen from d 10 to 14. From d 15 to 17, rumen fluid was collected every hour from 0500 to 2300 h. Rumen fluid samples were pooled by animal/period and analyzed for pH and VFA concentrations. On d 18, 60 to 80 papillae were biopsied from the epithelium and preserved for gene expression analysis. On d 18, one blood sample per heifer was collected from the coccygeal vessel. In situ ruminal starch disappearance rate (7.30 to 8.72%/h for BAR vs. 7.61 to 10.5%/h for CRN) and the extent of fiber disappearance (22.2 to 33.4% of DM for TH vs. 34.4 to 38.7% of DM for BP) were affected by starch and fiber source, respectively. Analysis of VFA molar proportions showed a shift from propionate to acetate, and valerate to isovalerate on TH diets compared with BP. Corn diets favored propionate over butyrate in comparison to BAR diets. Corn diets also had higher molar proportions of valerate. Expression of 1 gene (SLC9A3) were increased in BP diets and 2 genes (BDH1 and SLC16A4) tended to be increased in TH diets. Plasma acetate demonstrated a tendency for a starch by fiber interaction with BAR-BP diets having the highest plasma acetate, but other metabolites measured were not significant. These results suggest that TH has the greatest effect on shifts in VFA molar proportions and epithelial transporters, but does not demonstrate shifts in blood metabolite concentrations.
- 1 August 2021
- Published by American Dairy Science Association in Journal of Dairy Science
- Vol. 104 (8) , 8857-8869
Rumen volatile fatty acid molar proportions, rumen epithelial gene expression, and blood metabolite concentration responses to ruminally degradable starch and fiber supplies
The objective of this work was to characterize rumen volatile fatty acid (VFA) concentrations, rumen epithelial gene expression, and blood metabolite responses to diets with different starch and fiber sources. Six ruminally cannulated yearling Holstein heifers (body weight = 330 ± 11.3 kg) were arranged in a partially replicated Latin square experiment with 4 treatments consisting of different starch [barley (BAR) or corn (CRN)] and fiber [timothy hay (TH) or beet pulp (BP)] sources. Treatments were arranged as a 2 × 2 factorial. Beet pulp and TH were used to create relative changes in apparent ruminal fiber disappearance, whereas CRN and BAR were used to create relative changes in apparent ruminal starch disappearance. Each period consisted of 3 d of diet adaptation and 15 d of dietary treatment. In situ disappearance of fiber and starch were estimated from bags incubated in the rumen from d 10 to 14. From d 15 to 17, rumen fluid was collected every hour from 0500 to 2300 h. Rumen fluid samples were pooled by animal/period and analyzed for pH and VFA concentrations. On d 18, 60 to 80 papillae were biopsied from the epithelium and preserved for gene expression analysis. On d 18, one blood sample per heifer was collected from the coccygeal vessel. In situ ruminal starch disappearance rate (7.30 to 8.72%/h for BAR vs. 7.61 to 10.5%/h for CRN) and the extent of fiber disappearance (22.2 to 33.4% of DM for TH vs. 34.4 to 38.7% of DM for BP) were affected by starch and fiber source, respectively. Analysis of VFA molar proportions showed a shift from propionate to acetate, and valerate to isovalerate on TH diets compared with BP. Corn diets favored propionate over butyrate in comparison to BAR diets. Corn diets also had higher molar proportions of valerate. Expression of 1 gene (SLC9A3) were increased in BP diets and 2 genes (BDH1 and SLC16A4) tended to be increased in TH diets. Plasma acetate demonstrated a tendency for a starch by fiber interaction with BAR-BP diets having the highest plasma acetate, but other metabolites measured were not significant. These results suggest that TH has the greatest effect on shifts in VFA molar proportions and epithelial transporters, but does not demonstrate shifts in blood metabolite concentrations.
- Open Access
- 1 November 2020
- Published by American Dairy Science Association in Journal of Dairy Science
- Vol. 103 (11) , 10867-10881
Contributions of dairy products to environmental impacts and nutritional supplies from United States agriculture
Questions regarding the balance between the contribution to human nutrition and the environmental impact of livestock food products rarely evaluate specific species or how to accomplish the recommended depopulation. The objective of this study was to assess current contributions of the US dairy industry to the supply of nutrients and environmental impact, characterize potential impacts of alternative land use for land previously used for crops for dairy cattle, and evaluate the impacts of these approaches on US dairy herd depopulation. We modeled 3 scenarios to reflect different sets of assumptions for how and why to remove dairy cattle from the US food production system coupled with 4 land-use strategies for the potential newly available land previously cropped for dairy feed. Scenarios also differed in assumptions of how to repurpose land previously used to grow grain for dairy cows. The current system provides sufficient fluid milk to meet the annual energy, protein, and calcium requirements of 71.2, 169, and 254 million people, respectively. Vitamins supplied by dairy products also make up a high proportion of total domestic supplies from foods, with dairy providing 39% of the vitamin A, 54% of the vitamin D, 47% of the riboflavin, 57% of the vitamin B12, and 29% of the choline available for human consumption in the United States. Retiring (maintaining animals without milk harvesting) dairy cattle under their current management resulted in no change in absolute greenhouse gas emissions (GHGE) relative to the current production system. Both depopulation and retirement to pasture resulted in modest reductions (6.8-12.0%) in GHGE relative to the current agricultural system. Most dairy cow removal scenarios reduced availability of essential micronutrients such as α-linolenic acid, Ca, and vitamins A, D, B12, and choline. Those removal scenarios that did not reduce micronutrient availability also did not improve GHGE relative to the current production system. These results suggest that removal of dairy cattle to reduce GHGE without reducing the supply of the most limiting nutrients to the population would be difficult.
- 1 November 2020
- Published by American Dairy Science Association in Journal of Dairy Science
- Vol. 103 (11) , 10867-10881
Open AccessContributions of dairy products to environmental impacts and nutritional supplies from United States agriculture
Questions regarding the balance between the contribution to human nutrition and the environmental impact of livestock food products rarely evaluate specific species or how to accomplish the recommended depopulation. The objective of this study was to assess current contributions of the US dairy industry to the supply of nutrients and environmental impact, characterize potential impacts of alternative land use for land previously used for crops for dairy cattle, and evaluate the impacts of these approaches on US dairy herd depopulation. We modeled 3 scenarios to reflect different sets of assumptions for how and why to remove dairy cattle from the US food production system coupled with 4 land-use strategies for the potential newly available land previously cropped for dairy feed. Scenarios also differed in assumptions of how to repurpose land previously used to grow grain for dairy cows. The current system provides sufficient fluid milk to meet the annual energy, protein, and calcium requirements of 71.2, 169, and 254 million people, respectively. Vitamins supplied by dairy products also make up a high proportion of total domestic supplies from foods, with dairy providing 39% of the vitamin A, 54% of the vitamin D, 47% of the riboflavin, 57% of the vitamin B12, and 29% of the choline available for human consumption in the United States. Retiring (maintaining animals without milk harvesting) dairy cattle under their current management resulted in no change in absolute greenhouse gas emissions (GHGE) relative to the current production system. Both depopulation and retirement to pasture resulted in modest reductions (6.8-12.0%) in GHGE relative to the current agricultural system. Most dairy cow removal scenarios reduced availability of essential micronutrients such as α-linolenic acid, Ca, and vitamins A, D, B12, and choline. Those removal scenarios that did not reduce micronutrient availability also did not improve GHGE relative to the current production system. These results suggest that removal of dairy cattle to reduce GHGE without reducing the supply of the most limiting nutrients to the population would be difficult.
- 1 March 2020
- Published by American Dairy Science Association in Journal of Dairy Science
- Vol. 103 (3) , 2602-2614
Identifying gram-negative and gram-positive clinical mastitis using daily milk component and behavioral sensor data
Opportunities exist for automated animal health monitoring and early detection of diseases such as mastitis with greater on-farm adoption of precision technologies. Our objective was to evaluate time series changes in individual milk component or behavioral variables for all clinical mastitis (CM) cases (ACM), for CM caused by gram-negative (GN) or gram-positive (GP) pathogens, or CM cases in which no pathogen was isolated (NPI). We developed algorithms using a combination of milk and activity parameters for predicting each of these infection types. Milk and activity data were collated for the 14 d preceding a CM event (n = 170) and for controls (n = 166) matched for breed, parity, and days in milk. Explanatory variables in the univariate and multiple regression models were the slope change in milk (milk yield, conductivity, somatic cell count, lactose percentage, protein percentage, and fat percentage) and activity parameters (steps, lying time, lying bout duration, and number of lying bouts) over 7 d. Slopes were estimated using linear regression between d -7 and -5, d -7 and -4, d -7 and -3, d -7 and -2, and d -7 and -1 relative to CM detection for all parameters. Univariate analyses determined significant slope ranges for explanatory variables against the 4 responses: ACM, GN, GP, and NPI. Next, all slope ranges were offered into the multivariate models for the same 4 responses using 3 baselines: d -10, -7, and -3 relative to CM detection. In the univariate analysis, no explanatory variables were significant indicators of ACM, whereas at least 1 parameter was significant for each of GN, GP, and NPI models. Superior sensitivity (Se) and specificity (Sp) estimates were observed for the best GP (Se = 82%, Sp = 87%) and NPI (Se = 80%, Sp = 94%) multiple regression models compared with the best ACM (Se = 73%, Sp = 75%) and GN (Se = 71%, Sp = 74%) models. Sensitivity for the GN model was greater at the baseline closest to the day of CM detection (d -3), whereas the opposite was observed for the GP and NPI model as Se was maximized at the d -10 baseline. Based on this screening of relationships, milk and activity sensor data could be used in CM detection systems.
- 1 March 2020
- Published by American Dairy Science Association in Journal of Dairy Science
- Vol. 103 (3) , 2602-2614
Identifying gram-negative and gram-positive clinical mastitis using daily milk component and behavioral sensor data
Opportunities exist for automated animal health monitoring and early detection of diseases such as mastitis with greater on-farm adoption of precision technologies. Our objective was to evaluate time series changes in individual milk component or behavioral variables for all clinical mastitis (CM) cases (ACM), for CM caused by gram-negative (GN) or gram-positive (GP) pathogens, or CM cases in which no pathogen was isolated (NPI). We developed algorithms using a combination of milk and activity parameters for predicting each of these infection types. Milk and activity data were collated for the 14 d preceding a CM event (n = 170) and for controls (n = 166) matched for breed, parity, and days in milk. Explanatory variables in the univariate and multiple regression models were the slope change in milk (milk yield, conductivity, somatic cell count, lactose percentage, protein percentage, and fat percentage) and activity parameters (steps, lying time, lying bout duration, and number of lying bouts) over 7 d. Slopes were estimated using linear regression between d -7 and -5, d -7 and -4, d -7 and -3, d -7 and -2, and d -7 and -1 relative to CM detection for all parameters. Univariate analyses determined significant slope ranges for explanatory variables against the 4 responses: ACM, GN, GP, and NPI. Next, all slope ranges were offered into the multivariate models for the same 4 responses using 3 baselines: d -10, -7, and -3 relative to CM detection. In the univariate analysis, no explanatory variables were significant indicators of ACM, whereas at least 1 parameter was significant for each of GN, GP, and NPI models. Superior sensitivity (Se) and specificity (Sp) estimates were observed for the best GP (Se = 82%, Sp = 87%) and NPI (Se = 80%, Sp = 94%) multiple regression models compared with the best ACM (Se = 73%, Sp = 75%) and GN (Se = 71%, Sp = 74%) models. Sensitivity for the GN model was greater at the baseline closest to the day of CM detection (d -3), whereas the opposite was observed for the GP and NPI model as Se was maximized at the d -10 baseline. Based on this screening of relationships, milk and activity sensor data could be used in CM detection systems.
- 9 September 2019
- Published by Wageningen Academic Publishers
- p. 131-132
Dairy contributions to environmental impacts and nutritional supplies in United States agriculture
- 9 September 2019
- Published by Wageningen Academic Publishers
- p. 131-132
Dairy contributions to environmental impacts and nutritional supplies in United States agriculture