PLANTS IN THE WILD seldom grow in isolation; they interact with other plants and with their physical environment in all manner of ways. Plant ecology as a systematic study is only a little over a century old. In the first half of the twentieth century, plant-ecological thinking in the English-speaking countries was heavily influenced by the ideas of the great American ecologist Frederic E. Clements (1874–1945). Clements (1916, 1936) envisaged vegetation – plant communities – in terms of ‘climatic climaxes’, stable communities in equilibrium with the climate of particular regions (e.g. temperate deciduous forests, boreal coniferous forests, tundra), and ‘succession’ towards these climaxes. Succession might begin from many starting points (e.g. an abandoned field, bare rock, a lake), and he called the course of succession in a particular instance a ‘sere’ – a ‘xerosere’ from bare dry ground, a ‘hydrosere’ from open water. Clements likened the climax to ‘a complex organism inseparably connected with its climate and often continental in extent’, and developed an elaborate intellectual framework to describe it – and a terminology to match. Another American ecologist, Henry A. Gleason (1882–1975), who had begun his career in the Clementsian tradition, increasingly found that tradition both frustrating and inadequate, and argued for ‘the individualistic concept of the plant association’, seeing the plant community as simply the sum of the tolerances of the individual species (Gleason 1917, 1939). Both points of view have merits. Gleason’s ideas made little impact at the time, but were immensely influential in the later development of ecology.
It was perhaps natural that North American ecologists, in a continent relatively newly colonised by Europeans, should have seen vegetation in terms of climatic climaxes and succession, but American ideas were influential in Britain too. Tansley (1939) largely uses Clements’s terminology and successional concepts. A notable early contribution from Europe was the elucidation by the German agricultural botanist and peatland ecologist Carl Albert Weber (1856–1931) of the succession leading to the raised bogs of the North German Plain (1902). However, in the long-settled landscape of continental Europe the emphasis tended to fall more on description and classification of plant communities. This work became particularly associated with the name of the Swiss ecologist Josias Braun-Blanquet (1884–1980) – although Braun-Blanquet’s own Plant Sociology (1927, 1932 and many subsequent editions) is a wide-ranging textbook of all aspects of plant ecology. Over the decades, Continental ‘phytosociologists’ have built up a body of description and classification from which we can draw many useful insights into our own vegetation.
A less controversial division exists between autecology, the study of the responses of individual plant species to their environment, and synecology, which deals with the composition and functioning of vegetation – of entire plant communities – whatever view one takes of their nature.
PLANT LIFE-FORMS AND LIFE STRATEGIES
Plants vary greatly in shape and functional architecture. The Danish botanist Christen C. Raunkiaer (1860–1938) a century ago devised a scheme of ‘life-forms’ based on the position of the buds by which the plant passes the unfavourable season – winter in northern Europe, summer in arid climates (Raunkiaer 1907, 1934). In our climate, most plants start growth in spring from winter buds at about the soil surface – they are hemicryptophytes in Raunkiaer’s terminology (H in Fig. 43) – either with a rosette at the base or with evenly leafy stems. Some plants have their perennating buds on underground rhizomes (Grh), or in corms or bulbs (Gb). These geophytes are commoner in Mediterranean and semi-arid climates where the soil surface becomes not only dry but intolerably hot in the summer sun. Nevertheless, plants as diverse as bluebell, ramsons (Allium ursinum), bracken and many of our orchids are geophytes. Subshrubs, chamaephytes, have their perennating buds at the tips of the shoots, up to about 25 cm above the soil surface; they may be entirely woody (Chw) like the heathers, or a proportion of the shoots may die back in the unfavourable season (Chh). They are prominent in the Mediterranean region where the ground surface becomes too hot in summer, and in the Arctic where the soil is frozen in winter and growth can begin when the surface is thawed and shoots are warmed by the first spring sunshine. Therophytes evade the unfavourable season by perennating as seeds; we have many annuals in our climate, but they become more prominent in the Mediterranean. Shrubs and trees with their buds more than 25 cm above the ground are phanerophytes (P); shrubs less than 2 m high may be distinguished as nanophanerophytes. The humid tropics, with no unfavourable season, have a preponderance of phanerophytes. Plants of all life-forms have diversified in the course of evolution. Subtropical and temperate shrubs and trees evolved protected perennating buds to survive summer drought or winter cold, and many went beyond that to become deciduous, losing their leaves in winter – or in the dry season. Most trees at our latitude are deciduous, but in humid temperate climates only a little warmer than ours many trees are evergreen with laurel-like leaves, as in Madeira and the Canary Islands, and New Zealand.
FIG 43. Raunkiaer life-forms. Light outlines show the parts of the plant that die in the unfavourable season, solid black the parts that remain alive. The diagrams represent (left to right) a phanerophyte (P), herbaceous (Chh) and woody chamaephytes (Chw), a hemicryptophyte (H), and rhizomatous (Grh) and bulbous geophytes (Gb). An important life-form omitted from the diagram is the therophytes, which pass the unfavourable season as seeds; further life-forms are helophytes and hydrophytes, with perennating buds under mud and water, respectively. Simplified from Raunkiaer (1934).
Plants interact also with animals, and a very important innovation was the evolution of the grasses, which probably took place hand-in-hand with the evolution of the large grazing mammals in the early Cenozoic era. The essential feature of grass architecture in relation to grazing is that the growing-points of the shoot, and the growing-points of the individual leaves, are at the base, not at the apex as in most other plants. Various other groups of monocotyledons have the same growth-habit as the grasses, notably the sedges (Cyperaceae).
Raunkiaer’s life-form concept is illuminating, but tells only a part of the story. If habitats are grouped in relation to disturbance and ‘stress’ (anything that limits productivity, e.g. drought, salt, lack of nutrients or light), there are four possibilities. Stable but unstressed habitats are inevitably competitive, bearing vigorous closed vegetation. Stable but stressed habitats are tenable to specialised species that can tolerate the unfavourable conditions. Disturbed but otherwise favourable habitats are open to species that can establish, grow and set seed quickly. No plant can cope with the fourth possibility, the habitat that is both disturbed and unfavourable. Professor Philip Grime (1979) suggested that plants could be seen in terms of three ‘strategies’, competitors, stress-tolerators and ruderals.
This threefold C-S-R model has the merit that it can be set out as a triangular diagram like the conventional diagrams for the mechanical composition of soils (Chapter 1). The sides of the C-S-R triangle can usefully be thought of as the inverse of the opposite vertices – the left-hand side as (at least potentially) ‘productive’, limited by neither water, nutrients or light, the right-hand side as ‘stable’, and yield along the base as limited, by either habitat constraints or disturbance. Figure 44 shows the most-frequent occurrence of some species from the Sheffield region in vegetation types classified by the C-S-R strategies of their component species. Stinging nettle (Urtica dioica), meadowsweet (Filipendula ulmaria), rosebay willowherb (Chamerion angustifolium) and bracken (Pteridium aquilinum) grow in communities predominantly of C-strategists. Bell heather (Erica cinerea), mat-grass (Nardus stricta), salad burnet (Sanguisorba minor), biting stonecrop (Sedum acre) and bluebell (Hyacinthoides non-scripta) characterise vegetation of habitats imposing various kind of ‘stress’. Scarlet pimpernel (Anagallis arvensis), shepherd’s-purse (Capsella bursa-pastoris) and knotgrass (Polygonum aviculare) are species of weed and ruderal communities. Bulrush (Typha latifolia), cuckooflower (Cardamine pratensis), creeping buttercup (Ranunculus repens) and colt’s-foot (Tussilago farfara) compromise between C and R strategies, and dog’s mercury (Mercurialis perennis) and heather (Calluna vulgaris) are dominants of communities that are stable but stressed in different ways. Thale-cress (Arabidopsis thaliana) and rue-leaved saxifrage (Saxifraga tridactylites) are small annuals of seasonally-desiccated wall-tops and similar places. Common grassland species occupy the centre of the diagram – here, red fescue (Festuca rubra), Yorkshire-fog (Holcus lanatus), sweet vernal-grass (Anthoxanthum odoratum), daisy (Bellis perennis), cat’s-ear (Hypochaeris radicata) and autumn hawkbit (Leontodon autumnalis).
FIG 44. C-S-R diagram for some common species from the Sheffield region. For further explanation see text. Based on Grime, Hodgson & Hunt (2007).
The C-S-R concept is elegant, and a productive framework for thought, but, like Raunkiaer’s, it tells only a part of the story and should not be pressed beyond its limitations. In particular, stress is many-faceted, and ‘one man’s meat is another’s poison.’
A related concept comes from studies of growth. Both individual organisms and populations tend to follow S-shaped (sigmoid) growth curves. Growth begins slowly, then gathers pace and passes though a rapid phase until it slows again with the approach of the carrying capacity of the habitat (or adulthood). Often the growth curve roughly fits a logistic relationship, of which one formulation is the equation dN/dt = rN(1 – N/K), where dN/dt is the rate of change of N with time, N is the number of individuals at time t, r is the intrinsic rate of increase, and K is the carrying capacity of the habitat (Fig. 45). Clearly, starting with an empty habitat, the species with the highest intrinsic rate of increase will win in the short term, but as the carrying capacity is approached competition becomes progressively more important in determining the fraction of the habitat a particular species occupies in the long term. Species adapted to open habitats can be thought of as r-selected: they are primarily adapted to acquire resources (Grime’s ‘ruderals’, many of them annuals, biennials or short-lived perennials, belong here). Species adapted to stable, closed habitats are K-selected: their primary adaptation is to retain resources (generally perennials, these embrace Grime’s ‘competitors’ and ‘stress-tolerators’). Evidently there is a trade-off between these two adaptive patterns; compromise is possible, but no species can have the best of both worlds.
FIG 45. Two logistic growth curves. Species 1 will win in the short term because of its greater intrinsic rate of increase (r), but will be outcompeted in the long term by the slower-growing species 2. Weedy species are adapted to disturbed habitats in which growth-rate and pre-emption of resources are paramount: they are r-selected. For dominant species of stable communities, selection pressure is for share of the carrying capacity (K) of the habitat, and retention of nutrient capital: they are K-selected.
HABITAT TOLERANCES AND PREFERENCES OF SPECIES, ECOLOGICAL NICHES AND COMPETITION
The basic needs of plants for water and nutrients from the soil, and light and a tolerable microclimate from their above-ground environment, have already been sketched out in Chapter 1. No two plant species respond to the factors of the habitat in exactly the same way, which is one reason why so many species can grow together in the same community. Each has its own ecological niche, which may be likened in a general way to its ‘job’ in the community. It is often useful to think of a plant’s ecological niche in terms of its preferred place along gradients of habitat factors such as soil pH, moisture, light and temperature. Professor Heinz Ellenberg (1988) gave scores on a 10-point scale for seven habitat factors: light, temperature, continentality, moisture, soil-reaction (pH), nitrogen and salinity, for all the species in the central European flora, and Dr Mark Hill and his colleagues have adapted these (and summarised a great deal of other information) for the British and Irish flora (Hill et al. 2004). Plants compete with one another for resources both above and below ground, and generally show a broader tolerance of habitat factors in laboratory experiments (or in gardens), in which they are free of competition, than they do in the wild; the potential niche is generally broader than the realised niche. It is in fact uncommon for plants to be limited in a simple way by a single environmental factor – if only because most factors vary so much, and interact. Nevertheless, it is often possible to infer likely governing factors from the distribution limits of species. Some examples were given in Chapter 1.
Two salient points may be made about plant competition. One is that in single-species stands seedling densities are often immensely greater than can ever come to maturity. Growth is invariably accompanied by self-thinning, and yield of the stand is in general proportional to (1/density)–3/2, flattening off as maximum size is approached. This relationship also holds in the thinning tables used by foresters to maximise yield from plantations. The other point is that if individuals of two species are competing at a range of densities, three possibilities can be envisaged. If the two species are competing for the same mix of resources, the one with the more vigorous growth (and uptake) will over generations (or seasons) progressively oust the other. If the two species are not competing for exactly the same resources, they will progressively converge on a stable equilibrium between them. This must be a common situation in species-rich communities. The third possibility is that one or both species is actively inimical to the other (e.g. by producing toxic root exudates, or slow-to-decay leaf litter), and in such a case the weaker competitor will be eliminated at an accelerating rate. This sort of process is probably often implicated in plant succession – and (arguably) underlies the success of Sphagnum.
PLANT COMMUNITIES
The structure of plant communities
Plant communities, except the very simplest, are not just haphazard aggregations of individuals; they have at least some degree of structure. This is most obviously shown by layering. A population of weeds is a one-layered community, and so are many grasslands. Some grasslands and many heaths are two-layered, with a field-layer of the dominant flowering plants, and a ground-layer of mosses, liverworts and lichens. Most forests have three or more layers; an oakwood often has a tree-layer, a shrub-layer of (mainly) hazel, a herbaceous field-layer (of e.g. primrose, wood anemone), and a ground-layer of woodland mosses (Fig. 46). The lower layers must be tolerant of the shade cast by the layers above, but benefit from the shelter they provide. Each layer generally has a dominant species; sometimes two or more species share dominance, but almost always the bulk of the herbage is made up of relatively few species. Plants may also be layered below ground. T. W. Woodhead (1906) described how in the oakwoods around Huddersfield creeping soft-grass (Holcus mollis) roots in the surface layers of the soil, the rhizomes of bracken are deeper, and the bulbs of bluebell are deepest of all. In chalk and limestone heaths, and in the Burren grasslands, calcifuges (e.g. bell heather) are rooted in the leached surface layers, while the calcicoles root in the lime-rich soil underneath (Chapter 11). It is tempting to suppose that in richly diverse communities the species must be at least to some degree complementary to one another, but in species-rich road verges near Bibury in Gloucestershire monitored for almost half a century by the late Professor Arthur Willis without showing significant change, Thompson and his colleagues (2010) could find little evidence for this. The resolution of this enigma awaits new insights.
FIG 46. A traditional coppice-with-standards wood near Sixpenny Handley, Dorset, April 1974. This is a three-layered community, with a rather open tree-layer of large oaks, a shrublayer of hazel, which would have been regularly coppiced, and a field-layer of wood anemone (Anemone nemorosa).
Plants are never perfectly evenly distributed on the ground (except in plantations or orchards), neither are they distributed truly at random. Some degree of clumping is almost always present. This may arise from irregularities of seed distribution, from underlying patterns in the soil or bedrock, or from the growth-form of the dominant or other prominent plants (Fig. 47). Bracken can form striking circles in the course of invading grassland; the disturbed patch (rabbit burrow or whatever) where the spores germinated and growth began can often be made out in the centre. White clover (Trifolium repens), another species that spreads by rhizome growth, often forms circles in newly sown lawns in a similar way (Fig. 48). As the community ages, the circles break up into an all-over pattern, but patchiness persists at a scale related to the rhizome systems of the plants. Patchiness is also created as individual dominant trees and shrubs age, die and are replaced. As there are many causes of pattern and patchiness, some degree of patchiness is virtually universal, and appears at every scale.
FIG 47. Daisies (Bellis perennis) in lawns. (Left) A relatively newly sown lawn, probably 3–5 years old, with the daisies still conspicuously clumped around the points of initial colonisation. (Right) An old-established lawn; the initial colonisation pattern has disappeared, but an irregular finer-scale pattern persists, related to the size of the daisy plants. Exeter University campus, May 1976.
FIG 48. (Left) Bracken ring on a hillside near Aberayron, Cardiganshire, September 1956; the rhizomes have grown radially from the point of establishment, and in this case have died away behind the vigorous invading front. (Right) Rings of white clover in a recently seeded lawn; clover is still present in the interior of the rings, but is less vigorous. Exeter University campus, June 1973.
Some numerical relationships: how many species can we expect in a given area?
One of the questions asked by the early plant ecologists was ‘what is the minimal area of sample we must take for it to be representative of the community?’ It became customary to plot species–area curves, which were found to be initially steep at small areas, but flattened off as larger and larger areas were examined. The degree of flattening that satisfied the investigator was taken as the minimal area. If species-number is plotted against area on a logarithmic (as against a linear) scale, it was found that the flattening was illusory, and that the number of species generally continued to increase far beyond any practicable area for routine quadrat samples. Species recording from squares of the Ordnance Survey national grid (‘square-bashing’) renewed interest in species–area relations, but at a larger scale. How many species should be expected from a complete survey of a 10 km (or 2 km) square? In preparation for a Flora of Hertfordshire, John Dony (1963) surveyed some available quadrat data, along with the number of species recorded from various larger defined areas, up to the total flora of Britain and Ireland. Plotted on a graph of log(species number) against log(area), all the (somewhat variable) sets of data lie remarkably close to a straight line (Fig. 49). This gave him an estimate of the number of species to be expected in a 2 × 2 km square (‘tetrad’). Species–area relationships have a voluminous literature, of which the consensus is that, overall, the double logarithmic relationship generally gives the best approximation to reality, even though data from particular places can depart quite widely from it (Connor & McCoy 1979). This implies that, generally, ‘minimal area’ as an intrinsic property of the community is an illusion, and that we are at liberty to choose any scale we like to describe vegetation – and that our human scale, a metre or two, is as valid as any other. We may adjust the scale on which we look at vegetation upwards or downwards when dealing with organisms very different in scale from ourselves, such as forest trees, or bryophytes and lichens.
FIG 49. Plant species numbers from a wide range of defined areas, plotted on a logarithmic scale on both axes. Included are quadrat data, starting at 1 cm2, from Hopkins (1955, crosses and broken line), the median number of species recorded by Dony in 160 samples (66.7 m2) from various habitats, the number of species recorded for the BSBI maps scheme (1955–62) in Hertfordshire (1631 km2) and the British Isles (310,600 km2). Separate lines link the higher points on the graph. Slightly simplified from Dony (1963)
The distribution of commonness and rarity
It is characteristic of many communities of plants and animals that they are made up of relatively few common species; and large numbers of species that are less common or rare. The maps in the New Atlas of the British and Irish Flora illustrate this in a general way. It is instructive to consider the species that are most-nearly ubiquitous in Britain and Ireland (Table 3). Predictably, the list contains many species abundant in ordinary farm grassland: the common pasture grasses, buttercups (Robert Browning’s ‘the little children’s dower’), sorrel, tormentil, several common clovers and vetches, self-heal, plantains, common daisy, yarrow and a clutch of common yellow composites. (St Patrick might have had several clovers at his feet as a symbol of the Trinity, and white clover (seamaìr bhán) is as likely as any other; the Irish diminutive seamróg (shamrock) could have referred to any small clover – and the legend does not appear in written form until many centuries after his death.) The table also contains a group of plants of wetter places, such as lesser spearwort, meadowsweet, silverweed, cuckooflower, wild angelica, marsh bedstraw, marsh thistle, colt’s-foot, several common rushes, glaucous sedge and yellow iris, a versatile group with their headquarters in shady places, such as bracken, male and broad-buckler ferns, common dog-violet, primrose, brambles, broad-leaved willowherb, herb-robert, ivy, cleavers and honeysuckle, and some common weeds such as stinging nettle, chickweed, docks, dandelions and annual meadow-grass. The ‘near-misses’ that just fail to make 85% of the possible 10 km squares tell much the same story. Notably few woody plants appear in the list – common sallow, hawthorn and common gorse. The only near-ubiquitous tree is the ‘non-native’ sycamore, which has been with us for centuries. We shall now never know whether the odd winged seed has blown across the Channel and established with us, and sycamore could surely now be granted UK and Irish citizenship for its long residence and impeccable EU credentials! None of our long-established forest trees makes 85% of the grid squares or more. Those that come nearest are listed in the last part of Table 3.
TABLE 3. The most-nearly ubiquitous plants in Britain and Ireland. Species are listed in the order of the New Atlas of the British and Irish Flora. The numerals show the number of 10 km squares from which the species was recorded in 1987–99 in Britain/Ireland.
However, the distribution of commonness and rarity depends greatly on how you view the data. The species lists from the 3859 10 km squares of the Atlas do not accurately reflect the abundance of different plants, because a widely distributed but a minor ingredient of a common vegetation type (e.g. common mouse-ear) may appear as abundant as a widespread dominant grass (e.g. red fescue) – indeed more abundant than a forest tree or other plant that makes a far greater impact on our landscape (e.g. oak, heather). A minor but evenly scattered species, beyond a certain abundance, will inevitably be found in almost every grid square.
The engineer, glass-technologist, and spare-time ecologist and conservationist Frank W. Preston (1896–1989) suggested that, in general, species-abundance follows a bell-shaped ‘normal’ distribution on a logarithmic scale – i.e. a scale in which equal distance along the x-axis represents multiplication by a constant factor (e.g. doubling) (Preston 1948, 1962). He pursued the mathematical consequences of his theory for the structure and properties of animal and plant communities, including species–area relationships. Mathematically inclined readers will find his papers fascinating. He tested his ideas mainly on data from birds. Most species-lists from quadrat samples of vegetation (or from mapping projects) are unsuited to testing whether Preston’s hypothesis holds for plants for the reason outlined above, but weight or percentage cover are appropriate. In Figure 50 a small set of data from one of the communities described in Chapter 9 is plotted in this way.
FIG 50. Number of species plotted against the logarithm of the number of ‘hits’ of 100 random pins in native localities of Jacob’s-ladder (Polemonium caeruleum) in Britain (omitting Polemonium itself); most of the sites are false oat-grass (Arrhenatherum) grasslands (MG2). (a) All data from 13 native sites in Britain; (b) The Winnats, Castleton, Derbyshire; (c) Arnber Scar, Littondale, Yorkshire. The curves are normal distributions with the same mean and standard deviation as the points. The frequencies are calculated in intervals centred on 1, 2, 4, 8, 16 … ‘hits’ – each figure twice the one before (‘octaves’). The figures for the 13 sites are summed for graph (a). This is a small dataset, so random ‘noise’ has a large effect, but the graphs support Preston’s hypothesis rather than otherwise. Data from Pigott (1958).
A log-normal (or similar) distribution implies that most species are moderately abundant, with few species very abundant, and few species very rare. Some characteristics of both common and rare species can be recognised. Dominant species in cool-temperate climates tend to be wind-pollinated (forest trees, grasses), and often have seeds with no particular adaptations to wide dispersal. Common and gregarious species are often insect-pollinated, but not specialised to particular pollinators, and often conform to one of a few common flower forms (e.g. yellow disc, white disc with yellow centre, massed small white flowers). Less-common plants often have flowers adapted to particular pollinators (bumblebees, butterflies, night-flying moths), and rely on the flower-constancy of their pollinators and the precise placing of pollen on their bodies for effective pollination. Many of the precise pollination systems in orchids could only have evolved in relatively rare species. But many plants that are ‘rare’ with us are common elsewhere – Diapensia lapponica as its name implies in Lapland (and elsewhere in northern circumpolar regions), hoary rock-rose (Helianthemum oelandicum) on limestone in the central and south European mountains (and of course on Öland). Plants that are rare everywhere are rare indeed!
SOME DYNAMIC ASPECTS OF PLANT COMMUNITIES
Climatic climaxes were (and are) close to the American consciousness (though Clements’ generation may have underestimated the impact of the native Americans). In Europe it was different. Virtually the only truly ‘natural’ vegetation was on high mountains, and in remote boreal forests and bogs. Human influence was all-pervasive in habitable country, and had been for centuries or millennia. We too tend to underestimate the influence of our ancestors in moulding the landscape we live in, and which we are prone to regard as ‘natural’.
The ‘potential natural vegetation’ of Europe has long been a preoccupation of Continental ecologists. John Cross (2006) has published a map of the potential natural vegetation of Ireland. For Britain the data are more scattered, but from the pollen in peats and lake sediments (Chapter 2) we can reconstruct some picture of what Britain might have been like in the absence of man. Deep fertile soils in the lowlands of England would probably be dominated by (pedunculate) oak forest with wych elm and locally a substantial proportion of small-leaved lime; alder would be abundant in wetter places and beech in drier sites. On the acid soils of the rainier north and west acid (sessile) oakwood would probably be dominant, giving way to birch and pine in the Scottish Highlands. Oakwood of one sort or another would cover most of the Irish lowlands – sessile oak with an understorey of bluebells on the more acid soils, pedunculate oak with wych elm, ash and hazel on the more calcareous soils. As in Britain, the hard rocks of the rainy uplands and west would be dominated by acid woodland of sessile oak. In both islands, the forest cover would be broken by expanses of ombrogenous bog (Chapter 15), and treeless vegetation on the highest summits (Chapter 17).
Plant succession and the acquisition and cycling of nutrients by vegetation
The concepts of ‘climax’ and ‘potential natural vegetation’ imply that a piece of ground in any other state will tend to develop towards the climax – in other words to undergo vegetational succession. A distinction is often drawn between primary successions, from bare ground (as at Glacier Bay, Chapter 1) or open water (Chapter 13), and secondary successions, from some kind of pre-existing vegetation. The same kinds of processes are involved whatever the starting point. Individual plants germinate, establish, grow, compete and die. In so doing, they take up nutrients, photosynthesise and build up organic matter, which is incorporated in their leaves and roots. Some part of the vegetation is eaten by grazing animals – farm livestock, insects – but most is usually returned to the soil by decay organisms. Nature is good at recycling. Succession can be thought of as the balance between upgrade and downgrade processes, and the climax as the point at which these are in balance. Generally, the total mass of living material in the plant community rises in the course of succession, as do the quantities of organic matter and nutrients stored in the vegetation and the soil. An important caveat is that successions often pass through stages in which all the dominant plants are young and vigorous together; these stages may be more productive than the climax. Succession takes time. The dominant plants become established relatively quickly, but many years may pass before the less-common K-selected species (which may be very characteristic of the mature community) colonise and establish. Other things being equal, long-established plant communities are generally the richest in species.
Plant nutrients are added to the soil by weathering of soil minerals, and brought in by rain and as airborne dust. Animals may bring nutrients in (seabird colonies) or lead to their net loss (grazers). Nutrients are constantly lost from the soil in the drainage water, so in rainy climates soils tend to leach and become acid. Over time, the gains and losses must come into balance; change will only cease when a sustainable equilibrium is reached. It could be argued the ‘climax’ is a will-o’-the-wisp, because these processes need centuries or millennia to approach equilibrium, and on that timescale climate itself changes.
All of this presupposes that the plant community (and the ecosystem as whole) can ‘do its own thing’ without outside interference. But forest fires and volcanic eruptions have always happened, and grasslands are where herds of grazing mammals evolved. This was long before our ancestors began to use fire as a management tool, and before tools and domesticated animals gave them increasing control over the landscape. Many of our plant communities are maintained by some combination of grazing, cutting or burning – and secondary successions from these are very common (Chapter 8). Primary successions to forest do not generally go through a grassland stage; woody plants are often among the early colonists.
Even a ‘climax’ community is far from static. Quite apart from the constant turnover of organic matter and cycling of nutrients, plants, even forest trees, do not live for ever. In the course of centuries there must be a constant turnover of trees in the canopy. We can visualise two processes by which this can happen. In a mixed-age forest, individual trees age and die. Their canopy thins, they may lose branches, and they become liable to windthrow, leaving a gap in which regeneration can take place. Seedlings of many climax forest trees (e.g. oak, beech) commonly exist for many years in a suppressed state in the understorey. An opening in the leafy canopy lets in light and gives them the opportunity to grow to maturity. Many conifers are prone to forest fire and windthrow, leaving wide areas for regeneration at one time. These generally form extensive even-aged stands. Some conifers indeed are fire-adapted to the extent that the cones remain closed until the stimulus of fire causes the cone-scales to gape apart and release the seeds. Australian Eucalyptus and Banksia species behave similarly. Trees with this pattern of regeneration generally have light, easily-dispersed seeds, as do early-successional trees such as birch, ash and the maples.
It is a general principle that a plant cannot occupy the same piece of ground indefinitely. Either it has a limited life span, and must sooner or later give up its place to others, or it must move on, usually year by year. The ‘herbaceous perennial’ growth form is potentially immortal, because the plant in effect renews itself every year from lateral buds, so it is never in precisely the same place two years in succession. Some plants produce new shoots only millimetres from the old, so forming tight clumps; some produce extensive (above-ground) runners or (underground) rhizomes, appearing as looser patches or scattered shoots.
SPATIAL VARIATION AND BOUNDARIES IN VEGETATION
We all recognise some broad plant communities, such as oakwood, heath, chalk down, saltmarsh. But can boundaries be drawn between plant communities, or is variation in vegetation continuous? This was a matter of some controversy in the 1950s, with the American ecologists John Curtis and Robert Whittaker arguing strongly for a continuum view of vegetation (heir to Gleason’s ‘individualistic concept of the plant association’). Duncan Poore came to a similar conclusion from his work in the Scottish Highlands. He coined the non-committal term nodum for a plant community of any rank, and used it for commonly recurring combinations in the continuous field of variation. Most plant ecologists now would accept that variation in vegetation is essentially continuous.
However, recognisable boundaries can arise in vegetation for two reasons. Some are imposed by the physical environment, such as spring-lines, abrupt changes in soil depth, or geological boundaries affecting the physical and chemical properties of the soil. Often the effect of these physical boundaries has been sharpened by land use – most of all where they are followed by a fence or hedge (Fig. 51). Boundaries also arise from the meeting of communities of different growth or life-form. Gregarious tussocky dominants, such as purple moor-grass (Molinia caerulea) and black bog-rush (Schoenus nigricans) often grow in rather sharply delimited patches, and their growth-form dictates the range of microhabitats available to the associated species. Cycling of organic matter and nutrients is radically different in a wood and a grassland. The transition between a grassland with scattered hawthorns and a hawthorn scrub which is in all essentials a wood can take place in remarkably few years, and within a few metres on the ground. Woodland herbs need the shade and shelter of the dominant trees; tree-seedlings spend their early life in the ground flora, with which they may compete, upon which they may depend for shelter and support – and the ground flora is attuned to the annual leaf-fall and the yearly flush of nutrients that it brings. The contrast between heath and chalk grassland depends much more on the tolerances of the individual species to pH and mineral nutrients in the soil (i.e. on their autecology), but these are worked out in the context of competition and other interactions with neighbouring plants.
FIG 51. Looking east along the scarp of the Mid-Craven Fault, near Malham, Yorkshire, August 1963. Limestone grassland dominated by sheep’s fescue (Festuca ovina) and Sesleria covers the steep Carboniferous limestone slopes to the left, with enclosed meadow/pasture with red fescue (Festuca rubra) and yellow oat-grass (Trisetum flavescens) on the deeper soils of the gently sloping fields below. The slopes to the right of the road are on base-poor Bowland Shales; the enclosed fields near the road are damp acid pasture, giving way to mat-grass (Nardus) and heath rush (Juncus squarrosus) moorland above. The influence on the vegetation of the geological and topographical boundaries has been sharpened by land-use.
Even if no discontinuities existed, when we talk about anything as complex and diverse as vegetation we are compelled to use words – and words (if they are to be useful) need definitions, and definitions imply limits. A good analogy is colour. Colour varies continuously, but it can be precisely specified by three measurements: hue, saturation and brightness in one system, the density of yellow, magenta and cyan printing inks in another, but always just three. (The human eye has receptors sensitive to the primary colours red, green and blue.) Yet when we talk about colour we use words, and do not worry about the fact that there are no sharp boundaries in the spectrum. We can do the same with vegetation; much of the ensuing chapters can be thought of as describing and defining ‘colours’ in the landscape. When we actually encounter recognisable boundaries in vegetation, that is a bonus!
