Skip to main content
Report, 2007

Sensorer för bestämning av ensilagekvalitet i samband med utfodring

Stenberg, Bo; Hetta, Mårten; Sundberg, Martin


Variations in silage quality affect the daily composition of the diet and thereby animal feed intake and milk production ability. To improve cost efficiency and precision in the Swedish dairy industry, there is a need for tools that can provide continuous information on variations in different silage quality parameters. In order to calculate dietary requirements correctly, there is also a need for information on cow intake quantities of certain feeds. NIR (Near Infrared Reflectance) and NIT (Near Infrared Transmittance) are two variants of spectrometry that could provide the potential for non-complex nutritional analysis of fresh silage at farm level. This report describes a joint project carried out by SLU and JTI aimed at investigating and developing the potential for analysing silage quality in conjunction with feeding. The potential to predict intake by daily analysis was also investigated in the project in a feed intake study on dairy cows. During winter 2003-2004, a total of 69 silage samples were collected from 47 dairy farms in northern Sweden. These samples represented a wide variation in both phenological development and botanical composition and were taken from several different forage conservation systems. In conjunction with the feed intake study in dairy cows, a further 59 samples were taken during spring 2005 to monitor the daily variation in silage quality in an individual dairy unit. All forage samples were taken as mature silage in conjunction with feeding. Bulk samples of approx. 5 litres of fresh material were homogenised manually and divided into four equal portions. One of these portions was dried (60C) and milled for chemical and in vitro analyses. A second portion was extracted in distilled water for analysis of fermentation products, pH and ammonium nitrogen. The two remaining portions were analysed fresh using NIR or NIT. The dried and milled sub-samples were also analysed by NIR, as was dried but non-milled silage from the farm samples taken in 2003/04. The feed intake study investigated the effects of variations in silage quality on silage intake in dairy cows. The study lasted for two months and used silage from both pit silos and round bales. The results showed that both NIT and NIR can be used to determine dry matter content with good accuracy. In analyses of other quality parameters using the instruments investigated in the present study, NIR produced better results than NIT. For silage with a limited physiological variation, it was possible using NIR to determine crude protein and NDF in the first instance, while analyses of metabolisable energy, soluble protein, lactate and acetate also functioned reasonably well. Analyses of sugars and ammonium nitrogen functioned poorly, while ADF was intermediate. The feed intake study showed that using NIR analyses alone functioned poorly as a predictor of silage intake. However, good results were obtained when predictions were based on both silage analyses and animal data. Quality parameters determined using NIR then produced similar or even better results than wet chemical analysis methods


NIR; nära infraröd spektroskopi; grovfoder; on-line analys; vallfoderkvalitet

Published in

JTI-rapport. Lantbruk & industri
2007, number: 355
Publisher: JTI

    SLU Authors

      • Hetta, Mårten

        • Department of Agricultural Research for Northern Sweden, Swedish University of Agricultural Sciences

      UKÄ Subject classification

      Veterinary Science
      Animal and Dairy Science
      Food Science
      Agricultural Science

      Permanent link to this page (URI)