Several temporal and spatial scales have been considered during the Symposium, and the presentations by Wallenda & Kottke (1998) and by Norby (1998) represent two extreme points in a wide range. They both deal with short and long-term impacts on the local as well as the global scale. I present here a few additional thoughts, and a subject that both papers have touched only slightly will be expanded: the differential impact of different forms of nitrogen deposition.

Ammonia was necessary for the origin of life. However, NH3 would not have been a significant component of the neutral or mildly reducing (N2, CO2, H2O) atmosphere which characterized the Hadean earth (>3·8 Gyr ago), especially in view of the u.v. lability of NH3, the greater u.v. output of the young sun than pertains today, and the absence of an atmospheric u.v. screen. Heterogeneous phase reactions, following atmospheric chemistry, have been proposed as a prebiotic NH3 source. Lightning, or bolides (meteorites and comets), generated NOx., and Fe2+ in the sea could have reduced NO2− resulting from the NOx. to NH4+; this disequilibrium NH4+ level could have supported the production of the nitrogenous organic building blocks of life. NOx. and NHy thus could both have had important roles in the origin and evolution of life. Burgeoning biota could soon have depleted abiotically generated NH4+, and biological N2 fixation could have evolved in its present Fe-demanding, O2-sensitive forms in the Archaean O2-free, Fe2+-rich environment. Organic matter could have driven biological denitrification reactions based on NO2− and NO3− generated abiologically using some redox components which evolved in earlier chemolithotrophs and photolithotrophs, regenerating atmospheric NOx. and producing N2O. Any atmospheric NH3 leaking from oceanic biology would have been subject to u.v. breakdown and rain-out. Although O2-evolving photosynthesis probably began in the Archean some 3·5 Gyr ago, any O2 accumulation was local until 2·0 Gyr ago (Proterozoic) due to consumption by oxidation of Fe2+ and S2−. However, localized O2 accumulation before 2·0 Gyr ago could account for the observed early evolution of cytochrome oxidase and the possibility of O2-consuming chemolithotrophic and chemoorganotrophic nitrification, with further possibilities of NOx. production. Oxygen accumulation globally from 2·0 Gyr onwards coincided approximately with the evolution of eukaryotes, which contributed phagotrophy to the reactions of the N cycle as well as the nutrification-like aerobic production of NO. by nitric oxide synthetase, while lightning and bolides could now generate NO. from N2 and O2.

Evidence for terrestrial ecosystems is found from 1·0 Gyr onwards; NOx. and NH3 generated by terrestrial biota stands a greater chance of escaping to the atmosphere than do these compounds generated in the sea where recycling within the water body is likely. As CO2 levels fell and O2 levels rose, NH3 cycling in the photorespiratory carbon oxidation cycle might have been evident as early as 1 Gyr ago, although this does not seem to be a major contributor to atmospheric NH3 today. Embryophyte evolution on land 450 Myr ago, together with symbionts and biophages, increased primary productivity and N cycling on land, with greater quantitative possibilities for NOx. and NH3 escape to the atmosphere. The evolution of lignin (and related phenylpropanoids) at least 400 Myr ago, with associated NH3 recycling in vascular land plant, does not seem (on present evidence) to increase NH3 loss to the atmosphere significantly. Biomass burning occurred at least 350 Myr ago with lightning as the likely ignition source; such burning yielded NO., NH3, N2O and N2 from organic N. As in earlier times, the existence of terrestrial embryophyte vegetation has been punctuated by major bolide impacts, with the Cretaceous-Tertiary boundary impact generating perhaps 104 as much NO. as a year's thunderstorms do today, although the toxicity of NOx.per se to vegetation might not have been the major effect of the NOx. on biota. Terrestrial plants suffered fewer extinctions at this time than did many other major taxa. Despite such very significant generators of atmospheric combined N as major bolide impacts, the mean levels of NHy, NOx. and N2O (and O3) in the atmosphere during the 450 Myr existence of embryophyte vegetation was lower than the current, globally averaged, anthropogenically influenced values. Current globally averaged levels of atmospheric combined N are thus not without precedent in the history of life, globally for NOx., and locally for NHy, so vegetation has had evolutionary experience of high atmospheric combined N. However, this should not make us complacent about the impact of current wide-spread and continuing anthropogenic inputs of combined N, in view of the rate and extent of the inputs, and their combination with other aspects of local anthropogenic influence (e.g. SO2 from burning of high-S coal), stratospheric O3 depletion, and global environmental change (increasing CO2, temperature and sea-level).

The final paper of the Symposium (delivered by Professor Raven) reviewed the rôle of nitrogen in the development of terrestrial life, and gave a timely reminder of the context of the current debate on the effects of anthropogenic N; current man-made perturbations of the N cycle are approaching the scale of catastrophic perturbations caused by major meteorite impacts.

Faced with such a stark evaluation of the current position one is forced to consider how the scientific community is reacting to the challenge of understanding and predicting the effects of man's activities.

This paper discusses ways in which plant metabolism, development and survival influence, and are influenced by, gaseous pollutants. When amino acids are used as carbon and energy sources by plants and other organisms, reduced inorganic nitrogen is released, with wide-ranging implications for the N cycle. Plants can detect the presence and amounts of deposited N and ozone in the environment, not only by direct metabolic incorporation but probably also via distinct signalling systems. Low-nutrient communities and ecosystems, such as temperate conifer forest, are often protected against N overnutrition by adaptations that insulate plant metabolism and growth from the fertilizing effects of deposited pollutant N. Changes in ecosystem balance and biodiversity are not related in any simple way to differential injury responses of constituent species. The question of what happens when a particular component of a changing plant community fails is considered in terms of the relationship between phases of development and the turnover of tissues and organisms.