1
1 Introduction
1.1 Context of this work
Nitrogen used for plant growth can come from external N sources such as fertiliser, mineralisation of soil organic matter or atmospheric deposition. Furthermore, a major contribution may originate from N stored within the plants themselves through the internal cycling of N (Millard 1996; Clark 1977). In extensively managed systems without fertiliser inputs plants are subjected to low soil nitrogen availability, therefore storage of N, remobilization and its subsequent reuse could be an important prerequisite for the plants in these environments. In contrast to trees (Millard 1996) and forage species (Volenec et al. 1996), very little is known about the importance of N storage and remobilization in grassland species representative of natural plant communities. The aim of this work was to investigate these plant traits in species coexisting on the nutrient-poor, slightly acidic soils of the hill and upland regions of Britain. Rumex acetosa was chosen as an example of the dicot flora of nutrient-poor grasslands, whereas the grasses were represented by the dominating species Agrostis capillaris and Festuca rubra. Different genotypes of these species, isolated from two Scottish semi-natural grasslands, provided the possibility to assess not only interspecific differences in storage and remobilization, but also intraspecific variation.
The quantification of both storage and remobilization is not straightforward, requiring sequential harvesting and - as a technical tool - 15N labelling. Furthermore, these 15N tracer techniques are difficult to apply in the field due to problems in uniformly labelling the many different soil N pools that exist. Parts of this work were initiated due to the hope that naturally occurring d15N signatures could be exploited for the measurement of N remobilization, providing an easy means with which to work under field conditions and, additionally, screen numerous genotypes.
The following sections of the introduction will (1) give a short description of the relevant ecosystem and the experimental species, (2) provide an overview of N cycling within this ecosystem, (3) summarize the literature on N storage and remobilization in herbaceous perennials, (4) provide information on d15N signatures in plants and (5) briefly tackle the issue of genotypic diversity in natural populations.
1.2 Temperate Grasslands
1.2.1 Types of grasslands in Britain
While many grasslands in the drier regions of the earth constitute the natural climax vegetation (e.g. prairies of North America, steppes of central Asia), grasslands in the temperate regions are mainly man-made and occupy areas which were once covered by forests (Whitehead 2000). They are maintained either by grazing of animals, or by cutting for hay. The species composition as well as the productivity of a grassland will depend on soil characteristics, hydrology, climate and management strategies (fertilizer inputs, stocking density and species of grazing animals, time and frequency of cutting) (e.g. Barthram et al. 1992; Bobbink 1991; Rodwell 1992; Silvertown 1980). The National Vegetation Classification (NVC; Rodwell 1992), which describes and classifies the full range of British vegetation, divides grasslands into four main groups characterised by soils: (1) mesotrophic grasslands (on neutral soils and grasslands with fertilizers and liming), (2) calcicolous grasslands (nutrient-poor on calcereous soils), (3) calcifugous grasslands (nutrient-poor on acid soils) and (4) wet, acid grasslands and fen meadows. Their productivities, reflected by the amount of herbage produced, range between 15000 and 1000 kg dry matter ha-1 year-1 (Whitehead 2000). Systems of intensive productions rely heavily on fertilizer inputs. They often consist of a single species and are reseeded frequently. Grasslands with a moderate productivity are often grass-clover swards receiving limited inputs of fertilizer, and derive much of their N requirements from atmospheric N2 fixation through clover-associated Rhizobium strains. Extensively managed grasslands with no fertilizer inputs have low productivities, and are mainly found in the hill and mountain areas of Britain. On acidic soils, grasslands such as Agrostis-Festuca grasslands and those dominated by Nardus stricta or Molinia caerulea are belonging to this group. The average production (averaged over 3 years, measured between April and October) in Agrostis-Festuca grasslands at six different locations in Wales and England was only 2400 kg dry matter ha-1 (Milne et al. 2001).
1.2.2 Agrostis-Festuca grasslands
The species examined in this work are derived from grasslands that can be predominantly characterised as U4b in the National Vegetation Classification. These are Festuca ovina – Agrostis capillaris – Galium saxatile grasslands. The description given here is based upon the extensive treatise by Rodwell (1992). They are generally dominated by the two mentioned grass species, but in some cases Festuca ovina can be replaced by F. rubra. These grasses form short swards, which sometimes are a little rough and tussocky, but often close-cropped into a tight, fine-textured sward. The dicotyledon flora of the Festuca-Agrostis-Gallium grassland is generally not very numerous or diverse. Among the associated species – apart from G. saxatile - are Potentilla erecta, Viola riviniana, Rumex acetosa or Ranunculus acris. Festuca-Agrostis-Galium grasslands form the dominant type of pasture on better-drained, more base-poor soils through the cool and wet sub-montane zone of Britain, and are of major agricultural importance. They consist both of unenclosed pastures over steeper slopes at moderate altitudes through the upland fringes, providing much of the better-quality rough grazing, and of enclosed, slightly improved, pastures. They form a plagioclimax community, which takes much of its distinctness from the influence of grazing, in the absence of which progression to woody vegetation or bracken would be most likely. They occupy areas originally covered by a variety of more calcifuge woodland types, and which through clearance, burning and grazing were turned into grasslands. Sporadic clearances probably stretch as far back as the Neolithic period. Substantial areas of some kind of pasture (similarities to those found today not known) were established as early as the medieval period in the more accessible parts of Wales, the Lake District and the Southern Uplands, and were extended over the next few centuries to the remoter regions like the Highlands. The human treatment of the early grasslands was different than today by being used mainly for the summer period with a much greater importance of cattle grazing. The introduction of improved sheep breeds from mid-1700s meant a shift in the management of the grasslands, which from then on consisted of almost year-round sheep grazing.
Two Agrostis-Festuca grasslands provided the species used within this PhD work. They are located on the Kirkton farm of the Scottish Agricultural College (National Grid Reference NN360310; 56°25'N 4°38'W) in West Perthshire, and on MLURI-owned grounds in the Cleish Hills in Fife (NGR NT080936; 56°7'N 3°28'W). The sampling plots, 50 x 50 m in size, were established on either a slope with an altitude range of 20 m (Kirkton), or on a shallow ridge with an altitude range of 5 m (Cleish). The Kirkton site contained about 80 species, and also contained vegetation other than that classified as U4 (M23a, M25, U5c, U20). For the Cleish plot, only 25 different species were recorded, and the vegetation was consistently U4. The Kirkton site is situated on a farmland, which was managed extensively for sheep production for the last 200 years, and due to its slope and rocky outcrops, was probably never ploughed. At least for the last 30 years, since their management by either SAC or MLURI, both grasslands have been maintained by sheep grazing, without receiving fertilizer inputs.
1.2.3 Experimental species
From these Agrostis-Festuca-Galium grasslands, three species were chosen for experimentation. They are the two dominant grass species, A. capillaris and F. rubra, and R. acetosa as the dicot component. The two grass species were chosen as they are well distinguished by their different leaf habit (broad-leafed versus narrow-leafed), growth form (spreading versus “sessile”) and phenology (late versus early development). All three species are polycarpic, outbreeding perennials, and within the plant growth strategy system devised by Grime et al. (1988), group as C-S-R strategists. The species are further characterised in Table 1, and their habit is shown in Fig. 1.
Table 1: Characterization of the three experimental species. Data are derived from Grime et al. (1988).
SpeciesFamily / Phenology / Habitats / Soil pH / Specials
Agrostis capillaris L. (Common Bentgrass); Gramineae / winter green; leaf expansion in late spring; flowering June-August; seed set August-October / permanent pastures, heaths and waste places; spoil heaps of coal and lead mines; paths, road verges, outcrops / pH 4.0-6.0 / regeneration by means of rhizomes or stolons, allowing rapid lateral spread; persistent seed bank; can become dominant or co-dominant species, especially together with F. ovina or F. rubra; high potential for developing genetically specilized populations
Festuca rubra L. (Red Fescue); Gramineae / winter green; early spring and late autumn growth; flowering May-June, seed set July-August / base-rich grasslands; road verger, wasteland, meadows and pastures; rocky habitats, paths, spoils / absent from most acidic soils; most frequent at pH > 5.0 / regeneration by means of rhizomes and seeds (no persistent seed bank); taxonomically complex, with many subgroups or ecotypes; hybridization with L. perenne
Rumex acetosa L. (Common Sorrel); Polygon-aceae / winter green; flowering May-June; seed set June-September / meadows and pastures, lead-mine spoil, wasteland; outcrops / wide tolerance; most frequently pH 5.0-7.0 / dioecious, with natural populations frequently dominated by female plants; leaves contain oxalates (can be toxic to livestock); regeneration mainly by seeds (no persistent seed bank), but can produce daughter ramets
1.3 N cycling in grasslands (from a plant’s perspective)
In natural and semi-natural grasslands, in the absence of fertilizer applications, only minor inputs of nutrients occur from the atmosphere and the weathering of soil minerals. Therefore plants derive their nutrients almost entirely from recycling processes (Whitehead 2000). N will become available for plant uptake through the decomposition of organic matter (derived form plant litter, dead microorganisms and animals, animal excreta) and conversion of organic N into the inorganic N forms ammonium and (unless the soil pH is very acidic) nitrate. While plants are potentially able to take up organic N in the form of amino acids (Chapin et al. 1993; Streeter et al. 2000), and thereby short-cut the mineralization steps, the quantitative importance of amino acid uptake under natural conditions in temperate climates is not known (Hodge et al. 2000). N decomposition and mineralization depend on the activity of (and the release of extracellular enzymes by) soil microorganisms, which is greatest under warm and moist conditions. Earthworms and other soil fauna enhance composition by mixing organic material into the soil. They reduce the size of the detritus particles and make them available to microbes (Swift et al. 1979). Also soil fauna feeding on the bacterial and fungal community in the soil can increase mineralization rates (Bardgett & Chan 1999). The nature of the vegetation can have an effect on N cycling (Hobbie 1992; Van Breemen & Finzi 1998), probably through factors like litter quality, quantity and timing of inputs. Wedin & Timan (1990) demonstrated that different grass species, planted on initially identical soils, could lead to 10-fold differences in net mineralization rates and influence seasonal patterns of mineralization.
A proportion of the N produced by decomposition will be immobilised in the microbial biomass as a result of their N requirements, and will become available with the death of microorganisms. The rate of N cycling in the ecosystem is increased by the presence of grazing animals because the breakdown of organic matter is faster in the guts of animals than in the soil. Apart from the large herbivores, invertebrates such as slugs or the root-feeding Tipula-larvae (leatherjackets) can be of importance (Lutman 1978; Coulson & Whittaker 1978). Animals utilise only a small proportion of the nutrients that they consume (less than 15% of the N intake (Floate 1987)), with the remainder being returned in the form of dung and urine. The N in urine of vertebrate grazers is mainly as urea and amino acids, and therefore potentially available for plant uptake immediately or soon after deposition, while nutrients in dung are mainly insoluble and become available slowly (Whitehead 2000). Through the faeces and urine of grazers and dead bodies of the soil fauna nutrients are distributed in patches, which can result in parts of a grazed sward being deficient in one or more nutrients, while other parts might have an excess (West et al. 1989).
In general, N supply through mineralization and urine deposition is variable, unpredictable and not necessarily coinciding with peaks of plant demand. Therefore the ability of a plant to support growth through internal N sources at times of nutrient limitation, and to accumulate N beyond immediate requirements at times of nutrient excess should be important both for the plant and the ecosystem.
1.4 Nitrogen storage in herbaceous perennials
1.4.1 Definitions
Millard (1988) makes a distinction between accumulation and storage of N. Accumulation occurs when the plant uptake exceeds the demand, in this case increases in N taken up do not lead to an increase in the growth rate of the plant. Nitrogen is stored if it can be mobilized from one tissue and subsequently be reused for the growth or maintenance of another. N storage is independent of the growth rate of the plant. The ability of a plant to store N is not dependent upon its N status.
Chapin et al. (1990) define storage as "resources that build up in the plant and can be mobilized in the future to support biosynthesis for growth or other plant functions". They distinguish between three different classes of storage. Firstly, accumulation as an increase in compounds that do not directly promote growth, due to supply exceeding demand. Secondly, reserve formation as the "metabolically regulated compartmentation or synthesis of storage compounds from resources that would otherwise directly promote growth". Thirdly, recycling as "the reutilization of compounds whose immediate physiological function contributes to growth or defense but which can subsequently be broken down to support future growth".
In temperate regions, the cold temperatures in winter either result in reduced growth or dormancy, often associated with loss of aboveground plant parts. Overwintering plant parts are therefore places where both accumulation and reserve formation of N compounds occur. The next chapter gives an overview of N storage in perennials in a seasonal context. The chapter after that will illustrate with a few examples where N recycling is an important process.