Wet deposition of nitrogen (N) occurs in oxidized (nitrate) and reduced (ammonium) forms. Whether one form drives vegetation change more than the other is widely debated, as field evidence has been lacking. We are manipulating N form in wet deposition to an ombrotrophic bog, Whim (Scottish Borders), and here report nine years of results. Ammonium and nitrate were provided in rainwater spray as NH4Cl or NaNO3 at 8, 24 or 56kgNha(-1)yr(-1), plus a rainwater only control, via an automated system coupled to site meteorology. Detrimental N effects were observed in sensitive nonvascular plant species, with higher cumulative N loads leading to more damage at lower annual doses. Cover responses to N addition, both in relation to form and dose, were species specific and mostly dependent on N dose. Some species were generally indifferent to N form and dose, while others were dose sensitive. Calluna vulgaris showed a preference for higher N doses as ammonium N and Hypnum jutlandicum for nitrate N. However, after 9years, the magnitude of change from wet deposited N on overall species cover is small, indicating only a slow decline in key species. Nitrogen treatment effects on soil N availability were likewise small and rarely correlated with species cover. Ammonium caused most N accumulation and damage to sensitive species at lower N loads, but toxic effects also occurred with nitrate. However, because different species respond differently to N form, setting of ecosystem level critical loads by N form is challenging. We recommend implementing the lowest value of the critical load range where communities include sensitive nonvascular plants and where ammonium dominates wet deposition chemistry. In the context of parallel assessment at the same site, N treatments for wet deposition showed overall much smaller effects than corresponding inputs of dry deposition as ammonia.

Although the effects of atmospheric nitrogen deposition on species composition are relatively well known, the roles of the different forms of nitrogen, in particular gaseous ammonia ( NH 3 ), have not been tested in the field. Since 2002, we have manipulated the form of N deposition to an ombrotrophic bog, W him, on deep peat in southern S cotland, with low ambient N (wet + dry = 8 kg N ha −1 yr −1 ) and S (4 kg S ha −1 yr −1 ) deposition. A gradient of ammonia ( NH 3 , dry N ), from 70 kg N ha −1 yr −1 down to background, 3–4 kg N ha −1 yr −1 was generated by free air release. Wet ammonium ( NH 4 + , wet N) was provided to replicate plots in a fine rainwater spray ( NH 4 Cl at +8, +24, +56 kg N ha −1 yr −1 ). Automated treatments are coupled to meteorological conditions, in a globally unique, field experiment. Ammonia concentrations were converted to NH 3 ‐ N deposition (kg N ha −1 ) using a site/vegetation specific parameterization. Within 3 years, exposure to relatively modest deposition of NH 3 , 20–56 kg NH 3 ‐N ha −1 yr −1 led to dramatic reductions in species cover, with almost total loss of C alluna vulgaris , S phagnum capillifolium and C ladonia portentosa . These effects appear to result from direct foliar uptake and interaction with abiotic and biotic stresses, rather than via effects on the soil. Additional wet N by contrast, significantly increased C alluna cover after 5 years at the 56 kg N dose, but reduced cover of S phagnum and C ladonia . Cover reductions caused by wet N were significantly different from and much smaller than those caused by equivalent dry N doses. The effects of gaseous NH 3 described here, highlight the potential for ammonia to destroy acid heathland and peat bog ecosystems. Separating the effects of gaseous ammonia and wet ammonium deposition, for a peat bog, has significant implications for regulatory bodies and conservation agencies.

A line Source of ammonia, simulating NH3 emissions from an intensive livestock unit, was established in 2002 on an ombrotrophic bog, Whim bog in the Scottish Borders. The site is at 55 degrees 46'N, 3 degrees 16'W and (based on 4.5 years) has a mean monthly temperature of 8.8 degrees C and mean annual rainfall 971 min. Supporting a Calluna Eriophorum vegetation, NVC M19. Release of NH3 is controlled by meteorological conditions i.e. wind direction and speed. NH3 concentrations were measured at >10 distances along the transect of release using passive ALPHA samplers, fixed at two or more heights above the vegetation. Effects were recorded for the ericoids, mosses, Sphagnum and Cladonia lichens with respect to cover changes, visible damage, nutrients and interactions with abiotic stress. These data, collected over >4 years, have been used to evaluate the Critical Level (CLE) for NH3 and compared with values derived from short-term (weeks) exposures at high concentrations in controlled conditions. Results Suggest the annual Critical Level of 8 mu g m(-3) is too high to protect sensitive species experiencing long-term exposures to NH3. In particular, they show exposure periods longer than 1 year lead to detrimental effects at lower mean NH3 concentrations through relationships between CLE and exposure period that are still linear after 4.5 years exposure. These results highlight the need to set a CLE representing the mean concentration of NH3 in air for long term protection of sensitive habitats (e.g. over 20-30 years), that will be substantially smaller than the current annual CLE.

This paper describes a field manipulation experiment where the effects of a simulated decline in acidified S inputs on the fate of N on Sitka spruce growing on an organo‐mineral soil were investigated, alongside those of the original treatments: ammonium nitrate, with and without sulphuric acid and sodium sulphate. Five years of treatment, at canopy height, with up to 100 kg N and S ha−1y−1 were extended for a further three years, for half the plots, while the remaining plots were deprived of N, sulphuric acid or S. Stem area increment was unresponsive, whereas foliar N and Mg concentrations and fine roots were sensitive to the removal of N and acidity. This recovery experiment confirmed that the presence of acidified S modifies the fate of N and suggests that the reduction in acidified S deposition will increase the bioavailability of N.

It has repeatedly been shown that increased nitrogen (N) deposition results in dramatic shifts in vegetation composition. The sources of N-deposition vary from agriculture (mainly NH3 and NH4) to industry and traffic (mainly NOx). Effects of these different N forms on the vegetation and biogeochemistry of an ombrotrophic peat bog, Whim Moss (~15km southwest of Edinburgh), have been investigated since April 2002, by employing an automate N manipulation system. This field experiment, uniquely, offers the possibility to investigate the effects of the different N forms at the same site and at application rates and deposition scenarios simulating natural variation in rainfall. Within the manipulation system there are two N manipulations: Dry, where gaseous NH3 is released over a 60 m transect at concentrations that simulate a 100,000 bird poultry unit (0.4-200 μg m-3), and Wet, as soluble nitrate or ammonium, covering the full range of UK wet N-deposition (8 – 64 kg N ha-1yr-1). The effects of dry N deposition on the vegetation and biogeochemistry at different distances from the NH3 source have been analysed. In the wet N deposition experiment, 5 treatments were followed, ranging from 8-64 kg N ha-1yr-1 and differing in N form as either oxidised or reduced N. Samples of soil water were obtained using mini-rhizon samplers and were tested for pH, NH4, NO3, P and base cations. Young (1 year old) shoots of Calluna vulgaris L. (Hull), Erica tetralix L. and Sphagnum capillifolium Ehrh. (Hedw.) were harvested and tested for chlorophyll concentrations, aminoacids and P and base cation concentrations in acid digests. The preliminary results show differences between treatment N forms and N doses and along the NH3 gradient. In this presentation we will focus on the plant responses to the changes in the biochemistry. CAPER Conference

Providing an accurate estimate of the dry component of N deposition to low N background, semi-natural habitats, such as bogs and upland moors dominated by Calluna vulgaris is difficult, but essential to relate nitrogen deposition to effects in these communities. To quantify the effects of NH3 inputs to moorland vegetation growing on a bog at a field scale, a field release NH3 fumigation system was established at Whim Moss (Scottish Borders) in 2002. Gaseous NH3 from a line source was released along of a 60 m transect, when meteorological conditions (wind speed >2.5 m s−1 and wind direction in the sector 180–215°) were met, thereby providing a profile of decreasing NH3 concentration with distance from the source. In a complementary study, using a NH3 flux chamber system, the relationships between NH3 concentrations and cuticular resistances were quantified for a range of NH3 concentrations and micrometeorological conditions for moorland vegetation. Cuticular resistances increased with NH3 concentration from 11 s m−1 at 3.0 µg m−3 to 30 s m−1 at 30 µg m−3. The NH3 concentration data and the concentration-dependent canopy resistance are used to calculate NH3 deposition taking into account leaf surface wetness. The implications of using an NH3 concentration-dependent cuticular resistance and the importance for refining critical loads are discussed.

Providing an accurate estimate of the dry component of N deposition to low N background, semi-natural habitats, such as bogs and upland moors dominated by Calluna vulgaris is difficult, but essential to relate nitrogen deposition to effects in these communities. To quantify the effects of NH3 inputs to moorland vegetation growing on a bog at a field scale, a field release NH3 fumigation system was established at Whim Moss (Scottish Borders) in 2002. Gaseous NH3 from a line source was released along of a 60 m transect, when meteorological conditions (wind speed >2.5 m s-1 and wind direction in the sector 180–215°) were met, thereby providing a profile of decreasing NH3 concentration with distance from the source. In a complementary study, using a NH3 flux chamber system, the relationships between NH3 concentrations and cuticular resistances were quantified for a range of NH3 concentrations and micrometeorological conditions for moorland vegetation. Cuticular resistances increased with NH3 concentration from 11 s m-1 at 3.0 µg m-3 to 30 s m-1 at 30 µg m-3. The NH3 concentration data and the concentration-dependent canopy resistance are used to calculate NH3 deposition taking into account leaf surface wetness. The implications of using an NH3 concentration-dependent cuticular resistance and the importance for refining critical loads are discussed.

A mixed provenance Sitka spruce plantation, planted in 1986 on a drained deep peat, has been exposed to 6 different simulated mist treatments in 4 replicated blocks since 1996. Treatments provided N and/or S at a concentration of 1.6 mol m −3 , supplying ca. 50 kg S and/or N ha −1 yr −1 as N (NH 4 NO 3 ), S (Na 2 SO 4 ), NS Acid (NH 4 NO 3 + H 2 SO 4 at pH 2.5), 2NS Acid (double dose by application at twice frequency), a control treatment supplied with additional rainwater only and a 'no treatment' set of plots. Throughfall, preserved with thymol in the field, was collected using gutters with a surface area of 1 m 2 in all the replicate plots, and was analysed for all major ions. Prior to treatment in 1999, S deposition in throughfall exceeded that in rain because of dry deposition of SO 2 and SO 4 2− to the canopy; NH 4 + and NO 3 − ions were both retained in the canopy. During treatment, only 20–40% of the applied N in the high-N treatments was retained in the canopy. Acidity in the applied mist was partly neutralised by the canopy, but not primarily through exchange of base cations, leading to the conclusion that weak organic acids, in solution or in situ in the canopy, contributed to the buffering of the H + ion deposition in the acid treatments.

Effects of enhanced N, S and NS Acid additions, up to approximately 100 kg N and S ha−1 y−1, are described for a 15-year-old Sitka spruce forest growing on an acid peat in Scotland. Groups of 10 trees, replicated over 4 blocks, have been treated at canopy height on approximately 50 or 100 occasions with 2 mm precipitation equivalent, between April and November, since 1996. Relative stem volume increment (RSVI) has been consistently higher in the NS Acid treated trees compared with control or N treated trees since the second year of treatment, although no dose response was found. Litterfall was also considerably increased in the NS Acid treatments and showed a clear dose effect but was not correlated with RSVI. Base cation concentrations in soil waters, collected using zero tension lysimeters reflected the presence or absence of the spray treatments and showed a dose related increase in response to NS Acid inputs. Treatment with 2NS Acid increased NH4-N and NO3-N by an order of magnitude. Results after 4 years of treatment showed a positive growth and litterfall response to NS Acid inputs but no effect of N alone. Enhanced stemwood growth may be linked to the higher base cation and phosphate concentrations measured in the soil water in the presence of NS Acid and S treatments.

1 In a field acid mist simulation experiment, Sitka spruce (Picea sitchensis) was sprayed with different pollutant treatments: N, NH4NO3; S, Na2SO4; NS Acid, NH4NO3 + H2SO4 and control, no spray. Treatment effects on the abundance of the green spruce aphid Elatobium abietinum and honeydew production were assessed. In addition, needles were sampled for phyllosphere micro‐organisms. In a manipulative experiment, shoots were established and maintained as with or without E. abietinum infestation in order to determine the effects of infestation on needle loss and throughfall nutrient fluxes.

Monodisperse particles (particle diameter approximate to 0.5 mu m) were produced by a particle generator and tagged with a fluorescent dye. The particles were injected into a small wind tunnel, into which single needles or small branches of Picea abies, Pinus sylvestris, and Abies alba had been introduced. The needles were examined after treatment, using fluorescence microscopy, and the spatial patterns of particle deposition determined.The particles deposited preferentially in the stomatal regions of the needles. In these areas the incidence of micro-roughness due to epicuticular waxes is highest, reducing the laminar boundary layer of the needle. Atmospheric particles of less than 1 pm diameter are mostly hygroscopic and the potential effect of substantial deposition of these particles to stomatal surroundings and their influence on plant water relations is discussed.

Intra- and inter-tree variations in 13C/12C ratios were studied within a single clone plantation of 20-year-old Sitka spruce, some of which were treated with mist simulating acidic cloud water. For groups of trees of similar height and the same treatment, sampled at the same whorl height, δ13C values for current year needles showed variations (1 SD) of between 0.2 and 0.7‰. The variations reflect the seasonally averaged influences, on intercellular CO2 concentrations, of slight variations in the microhabitat within a group. For a typical intra-group variation of 0.4‰ one may be able to distinguish between groups whose mean intercellular CO2 concentrations differ by only 8 ppm. Acid misting resulted in a lowering of δ13C values by c. 0.7‰ (significant at the P≤0.05 level). This reflects higher intercellular CO2 concentrations for acid misted trees, which can be interpreted in terms of their having assimilation rates c. 10% lower than those of control trees, and might explain the observed reduction in stem growth for acid-misted trees. Without careful attention to sampling strategy, however, these small inter-tree δ13C variations can be easily masked by the much larger intra-tree variations with height. Large gradients of increasing needle δ13C with height, of c. 0.5‰ m-1, were observed in two untreated trees of different total height. The gradient was similar for both trees so, though δ13C values of both trees were identical close to their leaders (−27‰), the taller tree displayed much lower values close to the ground (−31‰). The gradients are believed to reflect lower light levels close to the ground, rather than the accumulation of respired CO2 in the atmosphere. The different height response of stems versus needles, reflected by an increase in δ13Cstems−δ13Cneedles with height (for cellulose), is discussed in terms of stem photosynthetic recapture of internally respired CO2.