Forage tree legumes. II. Investigation of nitrogen transfer to an associated grass using a split-root technique

In an earlier field experiment in this series, we found no significant transfer of N between Leucaena or Gliricidia and Panicum maximum. The glasshouse study reported here employed a split-root technique, whereby trees of leucaena and gliricidia were grown in boxes with 15N fed to one half of the root system and the transfer of N to the other half of the box was measured by sampling tree and planted grass. Detection of l5N in the grass tops and roots from the unlabelled half of the box was used to indicate N transfer from the tree roots to the grass. Transfer of labelled N to the grass amounted to 4.1% in the first 6 week period when 15N was being injected into the tree root zone. A harvest of the tree and grass was made at 6 weeks and both allowed to regrow for a further 6 weeks with no further addition of l5N. Over the entire 12 week experimental period 7.6% of the labelled N from the tree was transferred to the grass. The low proportion of N transferred from tree legume to the grass in this experiment, where herbage was cut and removed, is similar to the findings in the earlier field experiment and indicates that, in such a system, little direct beneficial effect of N fixation would be expected in an understorey grass or food crop.