Etude de la productivité des parcours de la zone d'Ain-Leuh (Moyen Atlas, plateau central). I. Effets de la fréquence d'exploitation et du taux de couvert arboré sur la productivité herbacée

Rough pastures in Middle Atlas spread over four bioclimatic levels: 1, Quercus ilex and Juniperus oxycedrus forest; 2, pure Quercus ilex forest; 3, Cedrus atlantica and Quercus ilex forest; 4, unsylvatic low grass, from 900 to 2 200 m above sea level. Dry matter production of grass was measured from and underneath the tree canopy in 1983 and 1984 in order to compare the effect of a variety of intervals between cuts and to measure shadow influence. Grass growth mainly occured from late January to early July; it began later at Cedrus-Quercus and low grass sites and stopped earlier at pure Quercus and Quercus-Juniperus sites, depending on longer frost persistence at highest levels, on earlier arrival of drought at lowest levels. Mean dry matter production was 2.5 kg h 1a d-! on light 30 cm deep soils at Quercus-Juniperus level and dolomitic soils at Cedrus-Quercus level. It was 5.5 kg ha-! d i- on colluvial soils at pure Quercus level and basaltic soil at lowgrass level Water balance and legume species abundance explained a great part of D.M. variance. Four cuts per year were more productive than three or five. Missing the earliest cuts resulted in much lower D.M. yields: a half with two cuts; three of five with an only cut per year. Shadow poorly reduced the yield up to a forest canopy cover of about 60%: from 35 to 65% cover, the canopy resulted in 30% fall in seasonal yield as water supply was not limiting, but had no effect as relative water deficit occurred. A pasture and forestry mixed system, aiming a canopy of 60 to 70% cover, would result in better conditions for cattle breeding and timber as a whole.