The impact of close spacing on yield of arabica coffee under contrasting agro-ecologies of Ethiopia

In an attempt to investigate the efficiency of close spacing and determine the optimum density for the approved coffee berry disease (CBD) resistant selections, field experiments were conducted at Tepi, Wenago, and Gera Research Centers for eleven consecutive years (1988/89-1998/99). A randomised complete block design was used to assign seven population densities that ranged between 4,006 and 10,000 trees ha-1. Coffee yield results of seven years revealed significant yield responses to close spacing in different crop seasons at each site, specifically when the trees bore heavy crops. Subsequently, coffee yield increased with increasing population densities, though the magnitude varied across crop years and locations. The combined analysis of variance at Wenago and Tepi showed non-significant yield difference, but differences were highly significant at Gera. The interactions of crop season and spacing were also significant at Tepi and Gera, but this was not the case at Wenago. At Tepi, yield initially increased with increasing tree populations up to the fifth crop but, thereafter, tended to decrease above the population densities of 7,062 trees ha-1, at hot and humid agroecology indicating early mutual shading effects. In contrast, in medium (Wenago) and high altitude (Gera) areas, yield increased from 13.22 to 21.84 and 17.05 to 25.84 Q ha-1, respectively, with increasing population density from 4,006 to 9,066 tress ha-1. Such impacts of close spacing on coffee yield performances were largely associated with the prevailing climatic factors that determined the rate of vegetative growth and subsequent adverse mutual shading effects. RESUME En vue de chercher l'efficacite d'un espacement etroit et determiner la densite optimale des cafes selectionnes resistants a la maladie des cerises du cafe, des essais en champs ont ete conduits aux Centres de Recherche de Tepi, Wenago et de Gera pendant onze annees consecutives (1988/89-1998/99). Un dispositif des blocs completement randomises a ete utulise avec sept densites de populations qui variaient entre 4.006 et 10.000 arbres ha-1. Les resultats de rendements de cafe de sept annees ont revele des reponses de rendements significativement differentes a l'espacement etroit au cours des differentes saisons culturales et pour les differentes localites et particulierement quand les arbres ont produits beacoup de plantes. En consequence, le rendemenet du cafe a augmentee avec l'augementation des densites de populations, bien que l'ampleur variait a travers les annees culturales et les localites. L'analyse combinee de la variance a Wenago et a Tepi a montre des differences non significatives pour le rendement, mais il y avaient des differences significatives a Gera. Des interactions entre la saison culturale et l'espacement etaient significatives a Tepi et a Gera, mais n'etait pas la cas a Wenago. A Tepi, le rendement a augmente initiallement avec l'augmentation de la population d'arbres jusqu'a la 5eme culture, et par apres il y eut une tendance de reduction au dela de la densite de population de 7.062 arbres ha-1, aux agroecologies humides et seches indiquant des effets d'ombrages mutuels precoces. Au contraire, dans les regions de moyenne (Wenago) et de haute altitude (Gera), le rendement a augmente de 13,33 a 21.84 et de 17,84 a 25,84 Q ha-1, respectivement, avec une augmentation de la densite de population de 4,006 a 9,066 arbres ha-1. De tels impacts d'espacements etroits sur les performances de rendement du cafe ont ete largement associes aux facteurs climatiques ambiants qui determinaient le taux de croissance vegetative et consequement des effets nefastes d'ombrage mutual. (African Crop Science Journal 2001 9(2): 401-410)