The relationships between plant architecture and flower production of
rose were studied for a cropping period of 2.5 years. Four types of pl
ants, varying in number of basal shoots and their architecture were cr
eated, which were combined with three plant densities (7.7-11.6 and 17
.4 plants per m(2)). Plant-architecture was mainly managed by manipula
tion of the basal-shoot formation in combination with height of prunin
g and deshooting practices. The investment in stem mass significantly
reduced the number and weight of flowers for the first eight months. T
his was due to a delay in time of basal shoot growth, restriction of n
umber of outgrowing shoots and disbudding of shoots. However, this ini
tial financial loss of about DFl 20 per m(2) amply paid itself in the
next two cropping years. Regression analyses showed that number and di
ameter of second-order laterals as formed after eight months, could ex
plain more than 70% of the variation in number and weight of flowers h
arvested in more than two cropping years. mower production was much le
ss related to the number and diameter of basal shoots than to number a
nd diameter of second-order laterals. The importance of the diameter a
nd cross-sectional areas of shoots of different heights with respect t
o assimilate storage and transport capacity is discussed in relation t
o flower production. Increasing plant densities lead to a higher bioma
ss and more flower production but with a reduced individual plant weig
ht. Furthermore, weight of harvested flowers was reduced. The harvest
index was neither affected by the applied treatments nor by plant dens
ity.