Competition for light, water and nitrogen in an association of hazel (Corylus avellana L.) and cocksfoot (Dactylis glomerata L.)

Competition for light, water and nitrogen between hazel trees and cocksfoot grown under the trees was studied in two hazel plantations managed in different ways. The first plantation with a light transmission of at least 70% at grass level was compared with two control hazel and cocksfoot monocrops. The soil was a calcareous heavy clay, 1.4 m deep. Water stress during summer was severe in the intercropped stand as well as the grass-only plots, while it was much less in the sole stand of trees (without grass). Hazel tree water potential in the intercropped stand improved in the fourth year as the trees extended their roots similarly to the grass-free ones, despite their much lower canopy growth rate. The fourth year, local N fertilisation began to be effective as evidenced by the N content per leaf area unit and growth of the intercropped trees: grass root competition was high and root barriers were needed to exclude it. The second plantation was managed on a deep soil with a water table at approximately 2.5 m depth. The trees did not suffer from the presence of grass, and allowed a 15 to 75% light transmission rate. A control monocropped grass plot was established for comparison. On the basis of the multiple limitation hypothesis, a grass dry matter production model was built which accounts for radiation transmitted to the understorey, water supply, temperature and air moisture deficit in interaction with the radiation transmission rate (r(2) = 0.716; 590 observations). The limits of such models are discussed with regard to insufficient knowledge on root dynamics.