GreenFeed records sensor data on a one-second basis. Automated processors have been developed to aggregate the data to various levels so that users can quickly and effectively analyze the results. The data can be useful on a number of different levels including:
- High Resolution Data
- Feed Period Data
- Diurnal Emission Patterns over the Day
- Evaluation of Long Term Trends
- Individual Animal Emissions
- Herd Average Trends
High Resolution Data
During each animal visit, high resolution data is automatically collected and stored. GreenFeed software uses the second-by-second measurements to automatically calculate mass fluxes for the sampling period for each animal visit. Second-by-second emissions data is providing new insights into metabolic processes not previously possible with any alternative emissions sampling method. Figure 1 shows a 25 minute snapshot of CH4 and CO2 concentrations measured using GreenFeed from four different animals. With GreenFeed, it is possible to measure both the lung and rumen CO2 and CH4 from each eructation and the average time between eructation events.
Figure 1. Rumen and Lung CO2 and CH4 from Four
Animals over a 25 Minute Period
Typically operators program GreenFeed to reward each animal with feed when they voluntarily visit, three to six times per day, each day. Animals typically visit at all times of the day. During a week of sampling, one animal could be measured on as many as 42 separate occasions. In the same week, a herd of twenty animals would be measured as many as 840 times! Data for each individual feeding period can be aggregated and analyzed in a number of ways to determine herd and individual animal behavior and trends.
Below are examples of herd averaged CH4 emission patterns measured by GreenFeed from (Figure 2) lactating cows fed TMR in a confinement dairy, and (Figure 3) animals on low quality pasture. From these plots, it is possible to conclude that the emissions from hour to hour are predictable, and can be characterized well with GreenFeed spot sampling multiple times throughout the day.
Figure 2. Herd Averaged CH4 Emissions Pattern, Measured by GreenFeed
from Lactating Cows in Confinement Fed TMR
Figure 3. Herd Averaged CH4 Emissions Pattern, Measured
by GreenFeed from Angus Cows on Pasture
Long Term CH4 and CO2 Emissions Trends
The GreenFeed system requires only about thirty minutes per week to maintain and calibrate. Therefore it is practical to track animal emissions over time periods of weeks, months, and even years. Figure 4 below shows GreenFeed measurements of the daily average emission rates from about 50 lactating American Holstein cows at the Kellogg Biological Station operated by Michigan State University. The research was directed by Dr. Santiago Utsumi. The GreenFeed data clearly shows the day to day effects on CH4 and CO2 fluxes of three different feeding regimens implemented over a period of five weeks. The plot also shows the excellent repeatability of the measurements day to day.
Figure 4. Herd Emissions Trends during Transition from TMR to Pasture
Individual Animal CH4 and CO2 Emissions
Using the periodic measurements, it is possible to determine the total daily emission rates for each animal, and compare each animal’s emission rates to the rest of the herd’s. Table 1 shows an example of CH4 emission rates in grams per day for a period of nine days. The cows are listed by column, and the emission rates by day are listed in the rows. *
* The emission rates shown are moving averages over 3 days.
GreenFeed Accuracy and Precision
The questions below detail how periodic GreenFeed measurements can be used to accurately represent animal emissions over the entire day.
How does the “Bait” Dispensed by GreenFeed Impact Animal Emission Rates?
GreenFeed incorporates a precision system that dispenses a small amount feed or supplements that is “bait” to attract the animal to GreenFeed at specified intervals. Normally users specify that GreenFeed slowly delivers from about 150 to 300 grams of bait to each animal three to six times per day. The overall sampling goal is to feed as little bait as possible to the animals with GreenFeed to not significantly impacting the animal’s normal diet or feeding pattern.
Through testing, we have found that animals can be attracted to GreenFeed with bait that has similar composition to the natural diet, which minimizes interference from the bait on normal CH4 production.
What Sampling Strategy is used to Produce Valid Results?
The general GreenFeed sampling strategy is to attract each animal to the baiting station several times per day at random intervals. After several days, emissions at varied times of the day can be precisely defined. The sampling interval for each animal can be controlled by controlling the time interval in which the animal received the bait.
Statistically, gathering large numbers of samples over long time periods significantly reduces uncertainty. Within a week of GreenFeed sampling, each animal’s diurnal pattern and average emission rates can be determined with a high level of certainty.
Can Animal Emissions Patterns over the Day Actually be Measured?
CH4 and CO2 emission rates from individual animals can change during the day. Rumen gas production is related to how much the animal has eaten and the interval between meals. Generally, gas production maximizes one to three hours after the animal eats, and decreases predictably until the next meal.
A ruminant’s digestive system operates as a relatively large, continuous vat processor, gas production rates change in a predictable way, over the course of hours, as a result of changes in intake or feed composition. In production environments, where animals have continuous access to feed, emission rates typically do not change more than about 40% during a 24 hour cycle.
What about the Statistical Validity of GreenFeed Spot Sampling?
With minimal effort, GreenFeed can be used with large emissions data sets from individual animals and groups of animals over long time periods. Consider that often users are able to gather about 90-180 GreenFeed spot measurement per month from one animal, at all times of the day. For a herd of 30 animals, users can gather emissions from between 2,700 and 5,400 feeding periods. Statistically, the uncertainty of the cumulative data set becomes very small with large numbers of data points and is often better than any other method.