RATIONALE: Recent studies indicate that plants can partition electron transport through alternative oxidase (AOX) and cytochrome c oxidase (COX) in response to environmental cues, thus modulating respiratory efficiency. The ¹⁸O discrimination method necessary for measuring electron partitioning in vivo, however, has been restricted to laboratory settings due to equipment constraints. Since plants grown in more natural and variable environments may not respond as predicted by laboratory experiments, I developed a new field-compatible analytical method and then applied it to three ecophysiological studies. METHODS: To address these needs, I developed a field-compatible method in which plant tissue was incubated in 12 mL septum-capped vials for 0.5 – 3 h, after which the incubation air was transferred to 3.7 mL storage vials for subsequent measurement by mass spectrometry. I also developed mathematical tools to correct for unavoidable contamination, and to detect and address curvature in the data – whether intrinsic to the respiration or due to contamination, – and to extend the usable dynamic range of the mass spectrometer. These methods were used to investigate respiratory responses (1) in canopy trees growing across a soil nutrient gradient at the Franz Josef chronosequence, New Zealand; (2) in a nutrient manipulation experiment of Griselinea littoralis; and (3) in a long-term nutrient-, temperature-, and light manipulation at Toolik, Alaska, USA. Leaf dry matter content, specific leaf area, nitrogen (N), phosphorus (P), sugars, starch, and AOX/COX protein concentrations were also measured as explanatory variables. (Leaf Cu and Fe were measured at the Franz Josef chronosequence.) RESULTS: Discrimination values computed using my methods replicated previously reported results over a range of 10 – 31‰, with precision generally better than ±0.5‰, thus demonstrating its validity as tool for measuring respiratory electron partitioning. Foliar respiration declined with site age across the soil chronosequence, increasing with leaf N levels, r² = 0.8, but electron partitioning declined with increasing N/P, r² = 0.23. AP activity was positively correlated with leaf P, Cu, and starch, r² = 0.71. In younger soils, however, declines in respiration were attributed to declines in cytochrome pathway (CP) activity, whereas across the older sites respiration declined due to a reduction in AOX pathway (AP) activity. The Griselinia nutrient-manipulation experiment partially confirmed these results: AOX protein levels were highest in N-deficient plants rather than in plants deficient in both N and P. AP activity was very low in all leaves, however, possibly due to low illumination. In support of this claim, leaves that had developed in the sun had higher AOX/COX protein ratios than those that had developed in the greenhouse. In Griselinia roots, CP activity declined by more than half in response to nutrient deficiency, whereas AP activity was unchanged. At the Arctic site, only one species changed electron partitioning in response to nutrient addition. Betula nana, the most successful adapter to improved mineral nutrition, doubled leaf CP activity without changing AP activity. Species grown in full sun at that site also had higher AOX/COX protein ratios than those that grew in enclosures. CONCLUSIONS: This is the first study of how engagement of terminal respiratory oxidases in plants responds to multiple nutrient deficiencies, both in nature and in a controlled environment. I have uncovered some intriguing relationships, including the possible importance of N/P to electron partitioning, as well as a role for Cu. The results also suggest that electron partitioning is sensitive to plant energy balance, as suggested by the low AP activities and low AOX/COX protein ratios in shaded plants. Perhaps most significantly, the AP and CP appeared to act independently of each other, rather than through a concerted “partitioning” process. In addition to their own scientific merit, these results illustrate the value of using the new field-compatible method to conduct ecophysiological investigations of plant respiratory electron partitioning on a much large scale, and under more realistic conditions, than has been previously possible.