Differentiated identification of behavioral elements during herbage intake – a tool for intake estimation of grazing dairy cows
Buch in deiner Nähe kaufen
Beschreibung
The estimation of the intake of herbage dry matter in grazing dairy cows is challenging, as it is influenced by many different animal, environmental, and feed characteristics. The resultant uncertainty about the nutrient supply on pasture impedes the adaptation of supplementation in the barn. Even though different methods for the estimation of herbage dry matter intake (HDMI) exist, they are expensive, inaccurate, and impracticable under farm conditions. Moreover, they derive only the mean estimation of a herd’s HDMI. A more effective method of direct herbage dry matter intake estimation (HDMIE) could be based on different behavioral characteristics, as the herbage dry matter intake is the product of grazing time, biting rate, and bite mass.
In enabling the adequate estimation of the HDMI based on behavioral characteristics, the first step was to validate the effectiveness of the RumiWatch system (RWS) as measuring tool for behavioral characteristics. The validation was performed on pasture and in the barn in 425 and 121 visual observations, respectively, each with a duration of 10 min. It was found that the RWS and its evaluation software did not enable the satisfactory evaluation of the ingestive and rumination behaviors on pasture. However, the adaptation of the evaluation software showed a significant improvement in the mean absolute percentage error compared to the visual observations. The evaluation using the adapted evaluation software showed the following mean absolute percentage error measured in grazing dairy cows: 11% for eating chews, 7% for eating time, 6% for rumination chews, and 4% for rumination time.
In the second step, the influence of different herbage masses per hectare at the same herbage allowance on different ingestive characteristics was considered. In addition, it was determined whether the further adaptation of the RWS evaluation software would enable the differentiation of prehension bites and mastication chews while eating and whether there was a relation between the number of prehension bites and eating chews, respectively, and the intake of herbage dry matter. The findings showed that the different herbage masses in the same herbage allowance had no effect on the different ingestive characteristics. However, the adaptation of the RWS evaluation software showed the possibility of differentiating between prehension bites and mastication chews while eating. Compared to the visual observation, the mean absolute percentage error was 12.4%. In addition, just a minimal part of the scattering of the HDMI could be explained by the number of prehension bites or eating chews respectively.
As only a minimal part of the scattering of the HDMI could be explained by the number of prehension bites or eating chews, in the third step, further characteristics were used to develop two types of models for the estimation of the intake of herbage dry matter in grazing dairy cows under a global approach (GA) and an approach without information about the amount of supplements fed (WSB). The dataset that was used to develop and validate the two types of models for the HDMIE consisted of 52 and 50 variables, respectively, as well as 130 observations that were collected in four preceding experiments. The target variable used for the development and validation of the two model types was the HDMI, which was estimated by the n-alkane method. Models were developed with an accuracy of 88.9% for the GA and of 86.8% for the WSB.
Briefly, within the given framework and based on animals and dataset of this study, an acceptable estimation of the individual HDMI was possible by using the developed models. However, the further development and validation of the models could be achieved by using a worldwide dataset.
Autor*in
Markus Rombach