In a multi-strata agroforestry system in central Amazonia, we studied the n
itrogen (N) use of two indigenous fruit tree species, Theobroma grandifloru
m Willd. (ex Spreng.) K. Schum. (cupuacu) and Bactris gasipaes Kunth. (peac
hpalm) for heart of palm production, and a legume cover crop, Pueraria phas
eoloides Roxb. (Benth.) (pueraria). N-15 was applied at a rate of 1 kg ha(-
1) twice at the beginning and at the peak of the rainy season, in a split p
lot design under either cupuacu, peachpalm or pueraria together with fertil
izer N usually applied (95.4 and 42.4 g N tree(-1) for cupuacu and peachpal
m, respectively). Plant and soil N-15 content and total N-15 uptake were me
asured for 1 year. The highest N uptake by the trees occurred from areas un
derneath their canopy being more than 70% of their total N uptake. During t
he dry season, pueraria also took up most of its N (more than 70%) from the
area underneath its own canopy. During the rainy season, however, pueraria
utilized N from the area under cupuacu (27-40%) and peachpalm (34-47% of t
he total N uptake by pueraria). Cupuacu took up between 12 and 26% of its N
from the area covered by pueraria, peachpalm slightly less with 10 to 18%
(significant only at the end of the rainy season; P <0.05). Competition for
N uptake between the trees was negligible. The above-ground recovery was h
ighest in cupuacu (15% of the applied N-15), followed by pueraria (11%) and
peachpalm (3%). Pueraria proved to be very important for the N cycling in
the mixed tree cropping system recovering most (31%) of the applied N-15 in
plant and soil in comparison to cupuacu (20%) and peachpalm (21%). However
, the natural N-15 abundance of the tree leaves did not show a significant
transfer of biologically fixed N-2 from pueraria to the trees (P >0.05) and
the cover crop did not improve tree N nutrition. The investigated fruit tr
ees did not benefit from biologically fixed N-2 of the legume cover crop du
e to their low lateral root activity and the high available soil N contents
largely being an effect of the amount and placement of mineral fertilizer.