Plant and Soil 255: 495–502, 2003. © 2003 Kluwer Academic Publishers. Printed in the Netherlands. 495 Influence of mycorrhizal associations on foliar δ 15 N values of legume and non-legume shrubs and trees in the fynbos of South Africa: Implications for estimating N 2 fixation using the 15 N natural abundance method Amy C. Spriggs 1 , William D. Stock 1 & Felix D. Dakora 2,3 1 Botany Department, University of Cape Town, Private Bag, Rondebosch, 7701, Cape Town, South Africa. 2 Research Development, Cape Technikon, P.O. Box 652, Cape Town, 8000, South Africa. 3 Corresponding author ∗ Received 13 August 2002. Accepted in revised form 21 March. 2003 Key words: ericoid and arbuscular mycorrhiza, fynbos, isotope fractionation, 15 N natural abundance, N 2 fixation, N 2 -fixing shrubs and trees, reference plants Abstract In this study, we examined the use of the 15 N natural abundance method to quantify the percentage N derived from fixation of atmospheric N 2 in honeybush (Cyclopia spp.) shrubs and trees in the fynbos, South Africa. Non-fixing shrubs and trees of similar phenology to the Cyclopia species were chosen as reference plants. These reference plants were selected to cover a range of mycorrhizal associations (ericoid mycorrhizal, arbuscular mycorrhizal and non-mycorrhizal). Isotopic analysis revealed a wide range of foliar δ 15 N values for the reference plants, including many very negative values. The marked differences in δ 15 N values were defined by the mycorrhizal status of the reference plant species, with the ericoid and arbuscular mycorrhizal plants showing lower foliar δ 15 N values relative to their non-mycorrhizal counterparts. In contrast, the δ 15 N values of the N 2 -fixing Cyclopia species were uniformly clustered around zero, from −0.11‰ to −1.43‰. These findings are consistent with the observation that mycorrhizal fungi discriminate against the heavier 15 N isotope during transfer of N from the fungus to the host plant, leaving the latter depleted in 15 N (i.e. with a more negative δ 15 N value). However, a major assumption of the 15 N natural abundance method for estimating N 2 fixation is that both legume and reference plant should have the same level of fractionation associated with N uptake. But, because mycorrhizal associations may strongly affect the level of fractionation during N uptake and transfer, the test legume should belong to the same mycorrhizal group as the chosen reference plant species. As shown in this study, if the mycorrhizal status of the legume and the reference plant differs, or cannot be assessed, then the 15 N natural abundance technique cannot be used to quantitatively estimate N 2 fixation. Introduction There are many problems associated with estimating the proportion of N 2 fixed by leguminous shrubs and trees in the field (Danso, 1985; Domenach et al., 1989; Sanginga et al., 1988). It is not only difficult to harvest all the nodules from a deep-rooted woody legume, it is also destructive and time-consuming to remove the whole plant for analysis. Because of these technical problems, the most widely accepted methods used for ∗ FAX No: 27-21-460 3887. E-mail: dakora@ctech.ac.za measuring N 2 fixation (such as the acetylene reduc- tion assay) are not suitable for use on shrubs and trees (Danso, 1985; Ham, 1977; Sanginga et al., 1988). The 15 N natural abundance method, which is based on differences in the natural abundance of 15 N between atmospheric N and other available N sources, has be- come increasingly popular as a method for estimating levels of N 2 fixation (Kohl et al., 1980). The method is widely used in studies on grain and pasture legumes and has been suggested in many studies to be the most reliable method for quantifying N 2 fixation (Bergersen and Turner, 1983; Peoples et al., 1997; Unkovich et