2. Literature Review:

This chapter will provide an overview of; the scientific basis behind why adding nitrogen will cause an increase in carbon stocks, the impacts of nitrogen deposition on tree carbon stocks documented in the current literature, the impacts of nitrogen deposition on soil carbon stocks documented in the current literature and information more specifically targeted at the UK and the South Downs.

2.1. The global nitrogen cycle

The nitrogen cycle is strongly coupled with the carbon cycle, this dependence on one another and key feedbacks is illustrated in figure 2. 1. If the availability of one of these elements changes it will affect the biochemical cycle of the other element and eventually the functioning of the entire ecosystem (Gruber & Galloway, 2008). Reactive nitrogen (Nr) is a requirement for the growth of plants and soil microorganisms, the processes of both play key roles in the global carbon cycle. The rise in anthropogenic CO₂ has been observed to increase plant photosynthesis and therefore plant growth and carbon storage (Oren et al., 2001). This growth however can be limited by the levels of Nr available in soils (Vitousek and Howarth, 1991), hence in Nr poor ecosystems the potential carbon sink could be limited.

In some regions it is conceivable that anthropogenic production of Nr could provide the nutrients necessary to reduce limitations on plant growth (Ciais et al., 1995). Prior to the industrial revolution reactive nitrogen, any form of nitrogen species other than N₂, was only made available to the terrestrial ecosystem by natural processes, such as biological nitrogen fixation (BNF) and lightening, since industrialisation a number of sources of Nr have become much more significant (Figure 2.2). BNF allows the terrestrial ecosystem to acquire essential nitrogen compounds a through a series of reactions that convert N₂ into ammonia (Ciais et al., 2013). Initially there was an equilibrium between the input of Nr to the ecosystem and its loss through the process of denitrification, however since industrialisation this equilibrium no longer exists. Humans produce a quantity of Nr that is much greater than that produced naturally in ecosystems. This Nr is produced by humans in a number of ways: 1) it is produced industrially by the Haber-Bosch process, producing NH₃ as a fertiliser for crops, this nitrogen is then spread through run off and emission into the atmosphere; 2) the growth of crops such as legumes that are associated with mycorrhizal fungi and higher levels of BNF (Hayman, 1986); and 3) the combustion of fossil fuels which converts N₂ and fossil fuel nitrogen into nitrous oxides (NO_x) which are emitted into the atmosphere and then deposited on terrestrial ecosystems and the ocean (Ciais et al., 2013).

This undeniable evidence of the anthropogenic perturbation of the nitrogen cycle and the close relationship between the nitrogen cycle and the carbon cycle highlights the need to understand how additions of nitrogen will interact with the carbon cycle, and in what way this will impact carbon sinks and therefore feedback to the climate. The increased supply of Nr can be expected to increase terrestrial CO₂ uptake by increasing NPP (net primary productivity) (chapter 2.2.1) or reducing the rate of organic matter breakdown (chapter 2.2.2).

It must be noted however that additions of Nr will not exclusively increase CO₂ uptake by the terrestrial biosphere:

negative direction (in situations where it accelerates organic matter breakdown)->

) O3 formed in the troposphere as a result of NOx and volatile organic compound emissions reduces plant productivity, and therefore reduces CO2 uptake from the atmosphere. On the global scale the net influence of the direct and indirect contributions of Nr on the radiative balance was estimated to be –0.24 W m–2 (with an uncertainty range of +0.2 to –0.5 W m–2)(Erisman et al., 2011).

The balance between the directions will determine the potential.

2.2.1 Forest Carbon Uptake- Trees

Various studies have assessed the possible impact of increased nitrogen on tree carbon stock and a variety of methods have been employed. One approach involves assessing the relationship between the spatial trends of carbon uptake, found by study of forest growth or net ecosystem production (NEP), and nitrogen deposition. In their 2007 study Magnani et al. employed a variation of this technique. Carbon stocks and their fluxes were measured in 5 representative chronosequences in Europe. In addition data from the literature, from a further 13 chronosequences and two uneven aged stands were used. Estimates of wet deposition of nitrogen were calculated from various data sets. The relationship between wet nitrogen deposition and NEP was then analysed graphically and statistically. Studies of the influence of nitrogen deposition at stand level have also been carried out, one example of this is a study by Solberg et al. (2009). The impact of nitrogen deposition was evaluated by using deposition values from the growth period (1993-2000). These values were then correlated with the values for relative volume which was calculated as actual increment in % of expected increment.

In their study Magnani et al. (2007) found a strong relationship between C sequestration and wet N-deposition (Figure 2.3d) with an R² value of 0.97. Though they found this relationship was largely obscured by age effects when individual stands were considered. Solberg et at. (2009) found that nitrogen deposition had a fertilising effect of slightly higher than 1% increase in volume increment per kg N ha^-1 yr^-1. The results of these papers can be converted into nitrogen uptake efficiency (NUE) in order to make them comparable, where NUE is the response of carbon sequestration to nitrogen deposition in kg C/kg N (de Vries et al., 2009). Thus it was calculated that Magnani et al. (2007) found an NUE_eco value of approximately 475 kg C/ kg N (de Vries et al., 2009). Sutton et al. (2008) found this value to be unlikely and tested the data against more appropriate N-depostion values for the period suggested and produced results calculated by de Vries et al. (2009) to be between 91 and 177 kg/C/N. The NUE value for trees in the study by Solberg et al. (2009) were calculated to be equivalent to 19-38 kg C/ kg N.

Another important area of research in this field is the evaluation of C-N stoichiometry of ecosystem compartments because it strongly influences the potential for carbon fixation to respond to nitrogen deposition (de Vries et al., 2009). Nitrogen entering the ecosystem can be traced by applying isotopically labelled nitrogen (¹⁵N) to the forest floor and tracking its movement. Melin et al. (1983) applied this techniques in nitrogen-limited Scots pine stand in Sweden, to study the distribution and recovery of the labelled fertiliser. The application rate was 100kg of ammonium nitrate-N/ha. Nadelhoffer et al. (1999) carried out similar ¹⁵N tracer experiments in nine temperate forests for three years. Further to this de Vries et al. (2006) used the same methodology as Nadelhoffer et al. (1999) at over 6000 level I plots (From a large-scale forest condition monitoring scheme based on a 16 x 16 km gridnet across Europe). Site specific soil C/N ratios were utilised and the assumption of an increase of upto 0.1 of N retention fractions in stem wood that are influenced by N deposition.

Melin et al. (1983) reported a value of between 12 and 28% of nitrogen applied recovered in trees. De Vries et al. (2009) used the C/N ratio of 500 for stem wood estimated by Nadelhoffer et al.(1999) to calculate a NUE_tree of 30-70 kg C/kg N. A NUE_tree of 25 kg C/kg N was calculated in the same way for the results presented by Nadelhoffer et al. (1999) which showed 5% of nitrogen applied to be recovered in woody biomass.

The results of direct fertilisation experiment methodologies are important to consider for this research paper, low doses of N fertiliser are applied to selected sites for a long-term (8-30 years) study. Hyvonen et al. (2008) ran experiments in Sweden and Finland in Picea abies and Pinus sylvestris stands. 15 sites were selected for long-term experimentation that ranged from 14 to 30 years. Low (30-50 kg N ha^-1 yr^-1) or high (50-200 kg N ha^-1) doses of N-fertiliser either alone or in combination with other nutrients were applied to plots no smaller than 30 x 30 m. Diameter at breast height (DBH), number of trees ha^-1 and tree height (when available) at each plot on each occasion of measurement were utilised to calculate biomass. The amount of carbon at each site was estimated to be 0.5 of the biomass. Changes in the C-pool was calculated for each site and the effect of N-fertilisation was calculated as the difference between the fertilised plots and the control plots. Pregitzer et al. (2007) applied 30 kg N ha^-1 yr^-1 from 1994 to 2004 to four different hardwood forests in Michigan. Tree growth was measured yearly and compared with control sites. Hogberg et al. (2006) ran a 30 year experiment in unpolluted boreal forest. N fertiliser (ammonium nitrate) was applied to replicated (N=3) 0.09 ha plots. 3 doses were prescribed, 34, 68 and 108 kg N ha^-1 yr^-1. The highest level of application was cancelled after 20 years to allow recovery to be assessed in the subsequent decade.

Hyvonen et al. (2008) presented their results as kg C/kg N and so no further calculations were necessary. They found that for the low dose application of N NUE_tree was 25 kg C/kg N whereas for high dose it was 11 kg C/kg N. NPK addition produced values of 38 and 11 kg C/kg N for low and high doses respectively. Pregitzer at al. (2007) found an increase of 5000 kg C ha^-1 in woody biomass, with a total application of N of 300 kg N ha^-1 for the whole study period this equates to 17 kg C/kg N (de Vries et al, 2009). Hogberg et al. (2006) found an initial increase in tree growth at all doses of nitrogen treatment. However in the long-term the impact of fertilisation was found to be highly rate dependant. The high dose showed no gain, medium dose gave an increase of 50 m³ ha^-1 and low dose a growth increase of 100 m³ ha^-1 as compared to the control. In order to calculate NUE_tree de Vries (2009) assumed a wood density of 500 kg m^-3 and C content at 50% to calculate a net C gain of 25, 000 kg C the total N input was also calculated and net C gain was divided by this value to give 25 kg C/kg N for the lowest levels of N application.

Model simulations have been used extensively in the literature. Levy et al. (2004) used three models (CENTURY, BCG and Hybrid) that employ a Monte Carlo approach, utilising conceptual algorithms that depend on repeated random sampling to obtain numerical results. The models were applied to a coniferous forest in Sweden for a 100 year period. Simulated nitrogen deposition levels were from data by Schulze (2000) the current ambient nitrogen deposition of 12 kg N ha^-1 year^-1 or 10 × current ambient nitrogen deposition, 120 kg N ha^-1 year^-1. Sutton et al. (2008) used the same inputs as Levy et al. (2004) but calculated a smaller value of total N-deposition (6-26 kg N ha^-1 yr^-1). Milne and Van Oijen (2005) used a complex forest growth model (EFM) directly parameterised to 22 specific sites selected across Europe, because growing conditions such as soil nitrogen levels were available for the sites. Model simulations were run for an 80 year period, control runs maintained N-deposition values at their 1920 levels whereas environmental change scenarios ran observed values of change averaging 10.5 ± 5.2 kg N ha^-1 yr^-1 for the whole period.