Natural abundance of 15N varies greatly and unpredictably within and between environments. The unpredictable nature of 15N limits the use of N isotope natural abundance (d15N) in tracing the flow and fate of N in environments. Recent investigations have, however, revealed consistent and repeatable patterns of 15N in some ecosystem components. These patterns suggest that d15N may yet provide a tool to investigate and illuminate ecosystem N cycling processes. Identifying and quantifying the sources of isotopic variation must precede any significant advance in the application of this technique, and to this end an assessment of isotopic variation associated with major ecosystem components has been carried out in this thesis. d15N patterns have been established, hypotheses proposed and tested, and conclusions about the application of the technique are presented. 15N patterns in surface and groundwater were measured in a variety of different land-use catchments in an attempt to identify distinct isotopic 'fingerprints'. High levels of 15N variation were measured in both stream and groundwaters, resulting in strongly overlapping land-use 'fingerprints'. Environmental 15N variation in streams and groundwaters was found to be too great to differentiate between land-uses based on d15N alone. In contrast, the artificially 15N enriched signature of effluent N was used to trace its flow and fate, following irrigation, in a forested catchment. The effluent d15N signature allowed it to be traced into the major ecosystem components, permitting a first order N budget to be determined for effluent N storage and loss. N sources with significantly different 15N signatures to that of 'background ecosystem N' can therefore be used to trace the flow and fate of N in ecosystems. During the course of this work a number of higher and lower order plants were observed to have highly depleted (lt; -8 ) d15N signatures. Epiphytes and lithophytes, strongly reliant on atmospheric N sources, were consistently depleted in 15N, with signatures as low as -24 , measured in a range of environments. A similar level of depletion was measured in a wide range of plants growing in early primary succession sites (as low as -22.3 ), which could not be accounted for by any abiotic or biotic factor or significantly depleted N source. The absence of any measurable driver of depletion suggested a universal fractionating mechanism which acts in a wide range of environments and vegetation types. Diffusive uptake of atmospheric NH3(g) and the proportional uptake of a supplied N source were two proposed mechanisms that could theoretically account for the level and universal nature of depletion. Diffusive uptake of atmospheric NH3(g) was tested as a primary fractionating mechanism in plants. Strongly N deficient plants were capable of utilising NH3(g) as a nutritional source, but the level of 15N depletion measured in these plants closely approximated that of the inherent NH3(g) d15N signature. No significant additional fractionation is associated with NH3(g) diffusive uptake. Diffusive uptake of atmospheric NH3(g) by plants cannot alone account for the level of depletion measured in early primary succession plant communities. Proportional uptake of a N source as a primary fractionating mechanism was tested by growing plants in various concentrations and rates of applied N. Fractionation attributed to the proportional uptake of a supplied N source, as a consequence of P limitation or rapid flow over roots, resulted in a significant level of 15N depletion in plants. The level of depletion attributed to this mechanism was, however, not sufficient to account for the level measured in early primary succession plant communities. Individual 15N fractionating mechanisms cannot alone explain the level of depletion observed in early primary succession plants, however a combination of fractionating mechanisms can. Fractionation attributed to the proportional uptake of an already depleted N source, i.e., wet deposited N, largely accounts for the level of depletion measured in early succession plant communities. This two-step fractionation model can act on both higher and lower plants, independent of ecosystem biotic and abiotic factors. Additional, and less dramatic fractionations attributed to atmospheric NH3(g) uptake, mycorrhizal associations, internal remobilisation, and taxon-specific N acquisition strategies, will contribute to the level of d15N depletion. This thesis presents the first extensive survey of highly depleted d15N signatures in terrestrial vegetation. Furthermore, thorough testing of theoretically plausible mechanisms has resulted in a full account of the highly depleted d15N signatures measured in a wide range of vegetation types and environments.