1. We report results from a long-term experiment in which additional nitrogen has been deposited on a peat bog in central Scotland for over 14 years, in three different forms: as ammonia (NH 3 ) gas, as ammonium (NH + 4 ) solution, or as nitrate (NO − 3 ) solution. The automated experiment was designed to apply nitrogen in such a way that mimics real-world nitrogen deposition. Background nitrogen deposition at the site was 0.8 g N m −2 year −1 ).
2.Observations of cover for 46 species were made. We analysed the change in six common species in relation to nitrogen dose and form. The responses differed among species and nitrogen forms, but five out of the six species declined, and NH 3 produced the biggest change in cover per unit of nitrogen addition. The exception was the graminoid sedge Eriophorum vaginatum, which increased dramatically in the NH 3 treatment. Multivariate analyses identified responses to nitrogen dose across treatments which were consistent with the univariate results.3. We surmised that the larger experimental response to nitrogen observed in the NH 3 treatment (cf. the NH + 4 and NO − 3 treatments) was because of the higher nitrogen concentrations at the vegetation surface produced by dry deposition. NH + 4 and NO − 3 were sprayed in solution, but much of this will enter the peat porewater, and be further diluted. Because NH 3 deposits directly to the leaf, it stays contained within the small volume of water on and in the leaf, producing a high internal concentration of nitrogen ions. 4. Synthesis. Consistent trends with nitrogen were discernible across species. All species showed a decline with NH 3 treatment, except for Eriophorum vaginatum which increased. In the absence of phosphorous and potassium (PK), all species declined with NH + 4 and NO − 3 , except for Calluna vulgaris and Hypnum jutlandicum. The effect of PK was not consistent across species. Per unit of nitrogen deposited, NH 3 generally had a larger impact on vegetation composition than NH + 4 or NO − 3 .However, the actual deposition rate of NH 3 on UK peat bogs is lower than the other forms. In the case of the most common species of the peat-forming genus Sphagnum, we estimate that NH + 4 deposition has the largest impact, followed by NO − 3 and NH 3 . 2008, 2011, 2014). Here, we build on this work in several ways.Firstly, we present a longer term analysis, with six additional years of experimental treatment. Secondly, we apply a linear mixedmodel approach (Pinheiro & Bates, 2006), allowing us to treat nitrogen deposition as a continuous variate, accounting appropriately for the correlation in residuals which arises from making repeated measurements on the same locations (quadrats nested within plots, nested within blocks). Thirdly, a phosphorus and potassium addition treatment was included in the experiment, but was excluded from most previous analyses. We include this interaction in our analysis. Fourthly, we apply multivariate analyses of species cover K E Y W O R D S air pollution, ammonia, multivariate analysis, nitrate, ...