Agron. Sustain. Dev.
Volume 30, Number 4, October-December 2010
|Page(s)||789 - 795|
|Published online||29 June 2010|
Variables influencing nitrogen surplus of dairy farms in Flanders
Department of Plant Production, Faculty of Bioscience Engineering, Ghent
Coupure Links 653, 9000
* Corresponding author:
Accepted: 11 March 2010
Nitrogen losses are major concerns for agriculture and policy-makers. Intensification of livestock production has contributed to an increase in nutrient surpluses. Here, we performed an exploratory analysis of the variables influencing the nitrogen surplus in Flemish dairy farms. We used the large dataset of the Farm Accountancy Data Network, holding technical and economic data of Flemish farms. A statistical model is proposed by performing multiple linear regression with several variable selection procedures. This approach focuses on a deep statistical analysis and interpretation of the model. The final model contains the following variables: N in fertilizers (kg/ha), N in concentrates (kg/ha), N in by-products (kg/ha) and N in straw (kg/ha), which refer to purchased inputs, livestock units of dairy cows per ha and percentage of arable crops. The input variables show a positive sign, indicating that the higher the nitrogen inputs, the higher the nitrogen surplus. Contrary to current knowledge, a lower nitrogen surplus was observed for farms with a higher number of livestock units of dairy cows per ha, holding the rest of the N inputs constant. A higher stocking density is compatible with a higher agricultural sustainability. The unexpected negative correlation of livestock units of dairy cows per ha with the dependent variable surplus per ha means that the higher the stocking density – under a certain limit – the lower the surplus of nitrogen will be, provided that feed inputs to the farm and cows are kept at a constant level.
Key words: statistical model / farm gate nitrogen surplus / dairy farm / livestock units of dairy cows ha−1 / multiple linear regression
© INRA, EDP Sciences, 2010