The DairyPrint Model: A Decision-Support Model to Help Dairy Farmers and Other Stakeholders Towards Improved Sustainability
Tadeu E da Silva,Victor E Cabrera
DOI: https://doi.org/10.3168/jds.2024-24946
2024-08-02
Abstract:Dairy farmers face increasing pressure to reduce greenhouse gas (GHG) emissions [i.e., carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O)], but measuring on-farm GHG emissions directly is costly or impractical. Therefore, the dairy industry has relied upon mathematical models to estimate those emissions. However, current models tend to be not user-friendly, difficult to access or sometimes very research-focused, limiting their practical use. To address this, we introduce the DairyPrint model, a user-friendly tool designed to estimate GHG emissions from dairy farming. The model integrates herd dynamics, manure management, crop, and feed costs considerations, simplifying the estimation process while providing comprehensive insights. The herd module simulates monthly herd dynamics based on inputs as total cows, calving interval, and culling rate, outputting average annual demographics and estimating various animal related variables (i.e., dry matter intake, milk yield, manure excretion, and enteric CH4 emissions). These outputs feed into other modules, such as the manure module, which calculates emissions based on manure, weather data, and facility type. The manure module processes manure according to farm practices, and the crop module accounts for GHG emissions from manure, fertilizers, and limestone application, also estimating nutrient balances. The DairyPrint model was developed using the Shiny framework and the Golem package for robust production-grade shiny applications in the R programming language. We evaluated the model across 32 simulation scenarios by combining various factors and considering a standard free-stall system with 1000 dairy cows averaging 40 kg/day of milk production. These factors included 2 levels of NDF-ADF in the diet (28-22.8% and 24-19.5%), the presence or absence of 3-NOP dietary addition (yes or no) at an average dose of 70 mg/kg DM per cow daily, the type of bedding used (sawdust or sand), the frequency of manure pond emptying [once (only Fall) or twice a year (Fall and Spring)], and the utilization or non-utilization of a biodigester plus solid-liquid separator (Biod + SL). In our results across the 32 scenarios simulated, the average GHG emission was 0.811 kgCO2eq/kg of milk corrected for fat and protein contents (4% and 3.3%, respectively), ranging from 0.644 to 1.082. Notably, the scenario yielding the lowest GHG emission (i.e., 0.644 kgCO2eq/kg) involved a combination of factors, including a lower level of NDF-ADF in the diet in addition to incorporation of 3-NOP, utilization of sand as bedding, application of Biod + SL, and strategic manure pond emptying in both Fall and Spring. Conversely, the scenario that resulted in the highest GHG emission (i.e., 1.082 kgCO2eq/kg) involved a combination of higher level of NDF-ADF in the diet and excluded the incorporation of 3-NOP, utilization of sawdust as bedding, no application of Biod + SL, and manure pond emptying only in Fall. All these scenarios can be easily simulated in the DairyPrint model and results obtained immediately for user evaluation. Therefore, the DairyPrint model can help farmers move toward improved sustainability, providing a user-friendly and intuitive graphical user interface allowing the user to ask what-if questions.