Background/Question/Methods The concept of competition is central to plant ecology: it regulates success of individuals and populations, composition of plant communities (and hence biodiversity), and the evolution of plant strategies and traits. A central ecological debate concerns the changing role of plant competition in relation to environmental gradients. However, arguments concerning the regulation of plant competition by environmental drivers are severely hampered by dependence on indirect (proxy) measurements of competition – competition is assessed by measuring what we assume is its outcome, rather than by measuring the process itself. Here we used an existing method – isotope pool-dilution – in a novel context to measure directly changes in competition for a limiting resource - nitrogen (N) – under contrasting levels of two types of environmental severity, soil resource-driven (low vs. high N availability) and climatic (lowland vs. upland sites). We applied our approach to a classic pot-based competition experiment to measure competition between two species common in UK temperate grassland systems: Dactylis glomerata and Plantago lanceolata. Competition was measured directly as absolute uptake of either NO₃^- or NH₄⁺, and indirectly during the same time period as RGR (relative growth rate).

Results/Conclusions We found that competition for N is determined by its form (NO₃^- or NH₄⁺) and by both nutrient and climatic variation in environmental severity. A critical issue regulating measured competition is probably the influence of these factors on the developmental trajectories of the competing plants. Furthermore we found little overlap between patterns detected by our direct measurement of competition, and those from an analysis of a simultaneously-measured indirect response variable RGR. Our results neither fully support nor refute existing models of plant competition that assume either a reduction or no change in the degree of competition in relation to environmental severity. We observe changes in competition dependent on both soil-resource and climatic environmental conditions, but these results indicate the substantial impact of the point within a plant’s annual growth cycle at which the measurements of such responses are made. Resolution of debate concerning existing models of the changing role of plant competition along environmental gradients may necessitate the realisation that current models are unable to encompass the processes that are being revealed by the application of this new technique, or that such models relate simply to patterns in the products of competition, rather than the process of competition itself.