Literally many thousands of papers have been published on every aspect of the nitrogen cycle. Critically reviewing them all would (even if possible) have produced an excessively long thesis introductory chapter, and then have resulted in potentially wasteful repetition later, in the more selective and focused introductions in subsequent chapters. The author therefore decided to make this first chapter relatively concise, and use his literature survey selectively to introduce where the ideas that form the basis of subsequent chapters emerged from. More detailed literature surveys of the literature underpinning the research in individual chapters are included in those chapters, and excluded from this one, with the hope that this would make the thesis more readable.
1.2 Forms and sources of plant-available nitrogen
The forms in which N occurs are very important as their responses to environmental factors vary markedly. Inorganic nitrogen occurs in four main chemical forms in the soil, alongside the massively predominant organic N. The main inorganic N forms, apart from N2, are as follows:
All the above forms of inorganic N are available to plants to a greater or lesser extent. In addition, there will be traces of gases such as N2O in the soil atmosphere and of course N2. On the other hand organic N includes compounds like amino acids, proteins and more complex N compounds of humus. The organic N forms are not available to plants directly.
The N cycle is very complex and includes many transformations. It is actively biotic in nature and therefore organisms influence it directly. There are also abiotic transformations and processes in the cycle which include, for example, ionic adsorption of NH4+ to clay particles, transformations attributable to fires, and oxidation of N2 to NO3- by lightning. In most ecosystems (e.g. streams, lakes, coniferous forests, prairies, salt marshes) the processes in the basic nitrogen cycle are similar but specific organisms play different vital roles in transformations and the relative importance of individual processes varies.
In soils the availability of nutrient elements depends upon a range of biotic and abiotic factors. These include soil moisture, acidity, salinity, soil particle size, nutrient input and activity of roots and microbes Binkley and Vitousek, 1989; Mengel and Kirkby, 2001). Fitter and Hay (2002) emphasised that the availability of N is largely under biological control while that of other nutrients is determined predominantly by inorganic equilibria. The key procedures governing the formation and mobility of nitrogen species are mineralization, volatilization, nitrification, immobilization and denitrification (Black, 1968; Hellebrand, 1998). Figures 1.1. and 1.2 are typical schematic representations of the N cycle, and indicate its highly dynamic nature.
Fig: 1.1 Key processes in the natural nitrogen cycle.
Fig. 1.2 A more pictorial representation of the terrestrial nitrogen cycle, drawn from a localised and predominantly microbiological perspective.
In 1890 Winogradsky discovered that nitrification was a two-step process and that Nitrosomonas and Nitrobacter were the organisms involved in this oxidative reaction (Stevenson and Cole, 1999). Nitrification (part of the overall mineralization process) is the microbially mediated oxidation of NH4+ to NO3- carried out by autotrophic bacteria (Simek, 2000). The two steps are performed by different groups of aerobic bacteria, the ammonia-oxidizing bacteria (AOB) and nitrite-oxidizing bacteria (Prosser, 1989). The equations may be represented as follows (Brady and Weils, 1999).
Step 1) NH4++1.5 O2â†’ NO2-+ 2H+ +H2O+275 kJ energy
Step 2) NO2-+ 1/2 O2â†’ NO3-+ 76 kJ energy
Several researchers have reported results showing that acid coniferous forest soils nitrify, as discussed later in section 1.7. The main ammonia-oxidizing bacteria (AOB) have been reported to be Nitrosomonas and the main nitrite-oxidizing bacteria are Nitrobacter in soils (Bock et al., 1991). Nitrification is a soil acidifying process (van Miegroet and Cole, 1985). AOB play important roles in the nitrogen cycle because of the high bioavailability and relatively high mobility of nitrate. Distinct from autotrophic nitrification, ammonia can also be aerobically oxidized by a number of heterotrophic fungi and bacteria (Prosser, 1989; Killham, 1990). Several studies have noted that heterotrophic nitrifiers are responsible for nitrification in some acidic coniferous soils (Lang and Jagnow, 1986; Stroo et al., 1986; Killham, 1987 & 1990; Duggin 1991; Papen and von Berg, 1998; Brierley et al., 2001; Jordan et al., 2005). The results from a large number of studies applying specific inhibitors of autotrophic nitrification have revealed that heterotrophic nitrification does not play an important role in acidic coniferous forest soils (Stams et al., 1990; De Boer et al., 1992; Martikainen et al., 1993; Pennington and Ellis, 1993; Stark and Hart, 1997; Paavolainen and Smolander, 1998; Rudebeck and Persson, 1998; Pedersen et al., 1999; Laverman et al., 2000; Ross et al., 2004; Burns and Murdoch, 2005).
Relatively recently, the concept of AOB being primarily responsible for nitrification has been questioned. High numbers of active ammonia-oxidizing Archaea (AOA) were discovered in calcareous grassland soil (Treusch et al., 2005) and, more recently, in agriculture and grassland soils with pH 5.5 to 7.3 (Leininger et al., 2006).
The importance of nitrogen to plants
Nitrogen is one of the most abundant elements in the Earth's biosphere and a major constituent of living cells. It is a key element in the metabolism of plants and essential for plant growth as a constituent of proteins, nucleic acid, chlorophyll and growth hormones. It is required by plants in larger amounts than other nutrients and its deficiency in soil results in lower crop yields (Buresh et al., 1993; George et al., 1993). Soil nutrient availabilities, more often than any other environmental factor, limit the growth of agricultural plants and forest trees, but are also essential to soil microorganisms. As a result, soil nutrients have wide-ranging and often surprising effects on ecosystem function, for example, by changing plant community composition when supplies are either deficient or excessive (Vitousek et al., 1997). Nutrient deficiencies may limit the potential growth stimulation response of an ecosystem to elevated CO2 (Zak et al., 2000). Human-induced global change is likely to affect soil nutrient availability, but much remains to be learned about the magnitude and even the direction of these effects. To the author, climate change, and particularly potential disrupting effects of increased incidence and extent of droughts, therefore seemed a worthwhile research topic. Hence the effects of drying and re-wetting are considered in some detail in Chapters 3 and 4.
The amounts of N required by plants differ greatly from one species to another, and, for any given species, with genotype characteristics and the environment (Viets and Hageman, 1971). This suggests that, for a PhD study, it is sensible to concentrate on a single ecosystem type such as the grassland discussed in Chapter 2.
Importance of N transformations in soil
Many scientists, for example Stevenson and Cole (1999), Compton and Boone (2002), Templer et al. (2003) and Grenon et al. (2004), have reported that soil N transformations, and hence N bioavailability to plants, are microbially mediated processes influenced by composition and diversity of the soil microbial community, substrate quality and quantity, and environmental conditions. These concepts are universally accepted and must therefore underpin the work in this thesis. Several researchers (Tilman, 1987; Aerts and Berendse, 1988; Wedin and Tilman, 1996; Vitousek et al., 1997) have also concluded that changes in N availability can lead in changing dynamics of plant populations and their primary consumers and ultimately all species that depend on plants. This, and my supervisor's interest in discussing the Gaia hypothesis (Cresser et al., 2008), lead to the idea behind Chapter 6, that the N cycle may be the heart of James Lovelock's Gaia, and plants have evolved to have a low C:N ratio at senescence and litter fall so that litter decomposes slowly to conserve nutrient N until needed later by the same plant. Increased N deposition has serious repercussions on all processes of the nitrogen cycle and consequently on many other elemental biogeochemical cycles (Aber et al., 1989; Vitousek et al., 1997; Aber et al., 1998; Tietema et al., 1998; Lovett et al., 2000). In this context it could clearly damage ecological niche by providing bio-available N too early, giving competitor plants a competitive edge.
According to several researchers, excess nitrogen can lead to eutrophication or levels of ammonia (NH3), nitrite (NO2-), and nitrate (NO3-) potentially toxic to humans, livestock, and wildlife in aquatic systems (Cairns et al. 1990, Carpenter et al. 1998, Marco et al., 1999). Lovelock (2006) would regard this as part of the "revenge of Gaia".
1.5 Nitrification in drainage waters and its links to ammonium
Nitrifying bacteria are important in both soil and drainage water because they ammonium-N (NH4+-N) to nitrate-N (NO3--N), which is generally accepted to be a more mobile chemical species in soil (Sprent, 1987). Tank et al. (2000) and Webster et al. (2003) investigated what influences NH4-N dynamics along streams. They found uptake of ammonium by nitrifying bacteria negligible compared with its removal by other processes such as heterotrophic metabolism in-stream. Others though have described nitrification as a quantitatively important in streams (Mulholland et al., 2000; Findlay and Sinsabaugh, 2003). Despite these unresolved contrary opinions, few studies have been conducted of potential variations in nitrification rate and what controls it (Bernhardt et al., 2002; Strauss et al., 2002). Generally though, temperature, dissolved oxygen (DO), NH4+-N availability and pH have been perceived as the potential regulators of nitrification rates (Wild et al., 1971; Kuenen and Robertson, 1987). Nitrification has been suggested as a bioindicator in streams impacted by mining activities (Niyogi et al., 2003). Strauss et al. (2002) examined 13 variables that might affect nitrification rates in sediments in 36 streams in northern Wisconsin and Michigan. NH4+-N availability and pH predicted nitrification rates most effectively. Kemp and Dodds (2001) concluded that nitrification rate and DO availability were significantly and positively correlated for Kansas prairie streams. These studies, and the findings of Cresser et al. (2004) that NH4+-N is more mobile in N-impacted upland soils in the UK than the vast majority of soil scientists believe, stimulated the author in Chapter 3 to investigate ammonium mobility in, and hence potentially from, acid soils under grassland near York.
1.6 Disruptions of the natural N cycle
The key processes that interact to regulate N species concentrations in soils include: nitrification, immobilization, nitrogen fixation, atmospheric deposition, mineralization, denitrification and leaching (Stoddard, 1994). Nitrogen is the most abundant element in the atmosphere as molecular dinitrogen (N2). Only after N2 is converted into reactive forms such as ammonium (NH4+) and nitrate (NO3-) however is it available to support the growth of plants and microbes (Galloway et al., 2003). Ammonium has been found to be the preferred form of N for assimilation by microbes in many cultivated soils (Azam et al., 1993). Nitrogen deficiency frequently limits forest productivity (Binkley and Hart, 1989; Paul and Clark, 1989; Reich et al., 1997), in spite of its great abundance in the atmosphere. Over the last century, however, human activities have more than doubled the global rate of mineral nitrogen production (Vitousek et al., 1997) through industrial production of N fertilizers, through atmospheric emissions of N species associated with fossil fuel combustion, and through cultivation of crops that host microorganisms capable of fixing N biologically (Smil, 2001).
1.7 Forest soils
There have been numerous reports of nitrification in acidic forest soils (Killham, 1990; Tietema et al., 1992; Pennington and Ellis, 1993; Stark and Hart, 1997; Degrange et al., 1998; Brierley et al., 2001; De Boer and Kowalchuk, 2001; Laverman et al., 2002; Bäckman and Klemedtsson, 2003; Bottomley et al., 2004; Stoddard et al., 2004; Laverman et al., 2005; Hart, 2006). For a long time nitrification was thought to be largely insignificant to nitrogen cycling in coniferous forest soils (Mintie et al., 2003) because several soil factors were regarded as suboptimal for nitrifying microorganisms. High soil acidity, high C/N ratio, low nitrogen availability and/or the presence of chemical compounds from coniferous litter could all impede net nitrate production (De Boer and Kowalchuk, 2001; Kowalchuk and Stephen, 2001). The cycling of forest litter therefore seemed an important research topic to the author, and features at least to some extent in Chapter 6.
1.8 Minimally managed/natural ecosystems.
Nitrogen is very often the most limiting nutrient in terrestrial ecosystems (Stark, 2000) and often limits their biological production (Schlesinger, 1997). However, surplus nitrogen can have harmful effects. For example, surplus N can facilitate increased losses of nutrient cations and increase soil and water acidity in forest ecosystems (Vitousek et al., 1997), while in aquatic ecosystems it may cause eutrophication (Carpenter et al., 1998; Marco et al., 1999). Almost all the N that enters a terrestrial ecosystem by natural processes is derived from biological nitrogen fixation and atmospheric deposition (Stevenson and Cole, 1999).
1.9 Importance of measuring nitrification rates
To better understand soil fertility and ecosystem function it is necessary to be able to accurately assess nitrification rates in soils. It is important to know how nitrogen is transformed from one nitrogenous compound to another and what factors regulate the transformation dynamics.
1.9.1 Methods for measuring nitrification rates
There are several possible approaches to quantifying nitrification rate. Laboratory incubations under controlled moisture and temperature conditions are often employed. Net inorganic nitrogen species accumulation is then monitored after a selected time period from days to several months (e.g. Carnol et al., 2002; Laverman et al., 2005; Kanerva et al., 2006). In situ incubations of enclosed soils at field sites are regarded as more realistic by some researchers. Again net inorganic nitrogen species accumulation is measured at the end of a few weeks or a few months (e.g. De Boer et al., 1993; Tietema et al., 1993; Vestgarden et al., 2003; Jussy et al., 2004; Fenn et al., 2005; Hart, 2006). However, the extent to which "realism" is enhanced if the soil has been removed from its associated vegetation must be regarded as questionable.
Laboratory- or field-based incubations using isotopic labeling, in which changes in a 15N-labeled ammonium-N pool are measured over 1-3 days of incubation may be preferable (e.g. Ross et al., 2004; Scowcroft et al., 2004; Perakis et al., 2005). Some authors just measure net nitrification rates over a short period (typically 1 day) in the laboratory (e.g. Bäckman and Klemedtsson, 2003; Ross et al., 2006).
1.9.2 Overview on methods
The major criterion for choosing the suitable method for each study is based on the objectives of the study, particularly on the selected element of the nitrogen cycle under focus. According to Binkley and Hart (1989), aerobic incubation under controlled environmental conditions is the most commonly employed method for assessment of nitrification. They further concluded that none of the methods gives an exact, accurate assessment of the nitrification rates in a forest soil (see Binkley and Hart, 1989 for a comprehensive review on the methodology).
1.9.3 Factors affecting nitrification
Under appropriate environmental conditions (e.g., temperature and pH), and in the presence of the two vital substrates for nitrification (oxygen and NH4+), the process of nitrification takes place extensively. Numerous rate-regulating variables have been suggested to be potentially influencing activities of nitrifying bacteria, including: NH4+ availability, the competition for NH4+ from other sinks (Riha et al., 1986; Jones et al., 1995; Verhagen et al., 1995; Strauss and Dodds, 1997), soil pH (Sarathchandra, 1978), soil temperature (Paul and Clark, 1989), oxygen concentration in the soil atmosphere (Wild et al., 1971; Stenstrom and Poduska, 1980; Kuenen and Robertson, 1987), and the availability (quantity and quality) of organic carbon (Verhagen and Laanbroek, 1991; Strauss and Dodds, 1997; Butturini et al., 2000).
Bianchi et al. (1999) could explain > 74% of the variability in nitrification in a single area by variation in NH4+ concentration. However, it is improbable that such a single factor would control net nitrification over a large scale because other environmental factors would then be much more variable.
One regulatory factor that potentially could strongly affect net nitrification is the availability of organic carbon. Carbon availability is both highly dynamic and spatially variable in streams. Its concentration varies with the abundance of wetland zones in a drainage basin (Kortelainen, 1993; Gergel et al., 1999), with catchment slope characteristics (Rasmussen, 1989), with water retention times (Sedell and Dahm, 1990), with drainage area contributing to discharge at different times during a storm event (Engstrom, 1987; Kortelainen, 1993), with discharge (and whether it is rising or falling (Sedell and Dahm, 1990), with primary production (Kaplan and Bott, 1989), and with litter deposition and the subsequent leaching of organic molecules (McDowell and Fisher, 1976; Meyer et al., 1998). In-stream nitrification is of interest to the author because of his interest in controls on seasonality trends for nitrate concentrations in the River Derwent in North Yorkshire, as discussed in Chapter 8.
The conversion of inorganic N (NH4-N and NO3-N) to organic forms is biotic immobilization. The reverse process of mineralization, although the two processes must generally occur simultaneously in soils.
The process of NO3- conversion to gaseous forms of N such as N2O and N2 by facultative and obligate anaerobes in soil is termed denitrification (Brady and Weil, 1999). Denitrification is important not only because it results in a loss of available N for plants, but also because N2O is a greenhouse gas 230-fold more potent than CO2 at trapping infrared radiation and it survives in the atmosphere 3-5 times longer than CO2 (Powlson, 1993; IPCC, 1995). The N2O and CO2 together account for 18 - 50% of current global warming and are also involved in reactions leading to stratospheric ozone depletion (Warneck et al., 1988). Besides denitrification, leaching of NO3--N is regarded as another way that NO3--N is lost from terrestrial ecosystems. It occurs because NO3- ions are not adsorbed significantly by negatively charged surfaces that occur in soils (Brady and Weil, 1999). Nitrate leaching is a concern in the present context not just because it decreases available N supply for plants and may acidify freshwaters and cause eutrophication of estuaries and coastal waters (Murdoch and Stoddard, 1992; Henriksen and Hessen, 1997). It also may contribute to denitrification in sub-soils and sediments. The heterotrophic process of denitrification and can itself be limited by the availability of organic carbon in some aquatic environments however (Seitzinger, 1988).
Nitrogen mineralization or ammonification (the release of NH4+ from decomposing organic matter) is thought by many to be controlled by the C:N ratio of the environment, although it should not be assumed that soil organic matter (and hence its C:N ratio) is homogeneous: under high C:N ratio conditions nitrogen is more likely to be imobilized in microbial biomass, whereas under low C:N ratio conditions a net flux of NH4+ into the environment is more likely (Schlesinger, 1997).
1.12 Identification of Gaps in Knowledge
To know the impacts of nitrogen on all biological systems it is necessary to understand how nitrogen is transformed from one nitrogenous compound to another and what factors regulate the dynamics of these transformations. The key processes may be the same or different, their relative importance changing according to different stages and/or different conditions. A firm understanding of the bio-geochemical nitrogen cycle is needed to address all the environmental challenges associated with assessing the importance of anthropogenically induced imbalances in ecosystem N cycling, such as those induced by global climate change (Houghton, 1997) or by acid rain (Driscoll et al., 2001).
A specific way to quantify changes brought in the soil N cycle by enhanced N availability is to measure rates of N mineralization and nitrification, two important microbial processes that govern the availability of N to plants and micro-organisms. These processes, usually measured as net N mineralization and net nitrification, can provide an accurate benchmark as to where the system is in terms of saturation, a condition where N availability exceeds biotic demand. Although several researchers have been working on the different aspects of the N cycle (Vitousek et al., 1997, Smil, 2001, Galloway et al., 2003), a fully comprehensive study allowing the quantification of nitrogen processes, is still one of the greatest challenges in N research. Other scientists (Jarvis et al., 1996; Powlson, 1997) have reported that mechanistic approaches of N processes may improve our understanding of the relationship between soil organic matter and N mineralization.
Mechanistic approaches take the internal N cycle as a starting point where gross N mineralization (N supply) and concurrent N immobilization (N removal from mineral N pool) are two fundamental processes that largely determine net N mineralization. Soil organic matter consists of various heterogeneous pools with different rates of decomposition. Unless the relative distribution of organic matter between these pools is the same in all soils, the total amount of soil organic matter will be a poor predictor for N mineralization.
The soil organic matter pool may be split into an 'active' pool and a 'passive' pool (Jansson, 1958). Indeed, mechanistic models (e.g. Jenkinson and Rayner, 1977; Smith et al. 1997; Jansson & Karlberg, 2001; Kätterer & Andrén, 2001) often divide the organic matter pool into a whole series of organic C pools (e.g. a pool of organic C from crop residues or manure, a microbial biomass C pool and a stabilized soil organic matter C pool). Each pool has a different turnover rate and assumes a characteristic C:N ratio (e.g. Parton et al., 1987; Hansen et al., 1991; Rijtema and Kroes, 1991). Each organic C pool is treated as a homogenous substrate following first order kinetics to simplify the model production. The turnover rateof each individual pool may be modified by abiotic factors such as temperature, soil moisture or soil texture, generally by using empirically based relationships. The C:N ratio of the organic matter in the individual pools determines whether net N mineralization or net N immobilization occurs.
Validating these models precisely is not possible, as the different pools of soil organic matter, in reality, can never be measured directly and they are therefore conceptual rather than real. A model by Bosatta and Ågren (1985) and Ågren and Bosatta (1996) considers the decomposition of soil organic matter as progressing through a continuum, so organic matter is assumed to progress down a quality scale. The mathematics of this approach is complex. However, the concept is used to some extent in Chapter 6. There it is assumed that forest litter, because of its high C:N ratio, initially immobilizes mineral N, but as the decomposing litter component C:N ratio progresses to lower values, eventually mineral N starts to become progressively more available.
1.13 The research in this thesis
As discussed in this chapter, because of the complexity of the N cycle and the timescale and resources available to complete a PhD in the Environment Department at the University of York, the author decided to concentrate his efforts upon unfertilized soils close to York and predominantly on N cycling under acid grassland at Hob Moor, as discussed in Chapter 2. Chapters 3 and 4 were stimulated by the thoughts that climate change is likely to lead to increased occurrence of periods of drought in the UK, and hence more drying and re-wetting cycles that will disrupt the N cycle.
Chapter 5 was triggered by the desire to explain my supervisor's observation that ammonium-N seemed to being mobilized at unexpectedly high concentrations into a stream that runs beside Hob Moor. It seemed that measuring ammonium absorption/desorption characteristics in soils over a range of depth was the most appropriate way to answer this question. This was made possible by allowing the author to use groups of second year environmental science students for two days to process the large number of samples in a short period of time, an essential requirement for this study. The author designed and tested the experimental protocol, and closely supervised the students in the laboratory. He performed all the ammonium determinations himself, however.
Chapter 6 was prompted by a developing interest in viewing the N cycle from the perspective of the Gaia hypothesis. The nature of this chapter and its results are unusual for a PhD thesis since several aspects remain speculative as it was a very preliminary evaluation of a set of novel ideas. The author nevertheless thinks that this is an exciting contribution to understanding how atmospheric N pollutant deposition may be causing biodiversity change.
Chapter 7 was stimulated by observations of higher than anticipated variation in ammonium and nitrate concentrations some experiments, and the need to know more about the sample preparation and storage condition constraints to the widely used operationally defined procedures for assessing mineral N species in field moist soils. This experiment too was only possible because my supervisor arranged for me to get access to a group of second year environmental science students for a 2-day practical session so that very tight processing time constraints could be met. They were supervised closely throughout by the author.
Chapter 8 looks at spatial and temporal variations in nitrate-N concentrations in the River Derwent in North Yorkshire from the perspective of its having been declared a nitrate vulnerable zone, and to see if any of the preceding research helped to explain temporal trends found in 20-year runs of Environmental Agency data.
Finally, Chapter 9 briefly summarizes the conclusions from all the research and discusses their significance, but also suggests some possible future research avenues.