17. Development of Plant Communities in Time

17.2.4.1 Early Land-Use Influences

The area surrounding the Bay of Bengal, the northern Savanna landscapes of India, the Yangtze region of China and the highlands of Mexico are some of the oldest centres of agriculture (von Wissmann 1957). There is now proof that these agricultural centres had access to a number of different agricultural techniques since at least 14,000 years (Bar-Yosef 1998). Agriculturalists are thought to have come from the Levant about a thousand years later, where they arrived via Iran in Mesopotamia and essentially replaced the hunter-gatherers already located there (Hillman et al. 2001). The first technical advance was irrigation.

From this region of the Fertile Crescent agriculture and animal husbandry (cultivation) reached Central Europe via the Mediterranean region. The important steps in the shift away from a culture based on pure hunting-gathering to one that developed through the use of many techniques to improve crop growth had already taken place outside Europe. Neolithic agriculture in southern-central Europe has been found to have started about 7000 years ago, while the northern region began roughly 5500 years ago, generally limited to areas with the best climatic and edaphic conditions (dry, warm, but often nutrient-poor sites) as present in sparse mixed oak forests.

17.2.4.2 Land-Use Influences from Neolithic to Middle Ages

For a long time the effect of prehistoric man on the cover of vegetation was severely underestimated. Today, it is possible to prove the effect of these human activities for the periods in which, during postglacial times, the edges of the ice sheet retreated and woodland species immigrated. Agricultural activities (clearing and burning) contributed already during late-Neolithic times to the excessive release of greenhouse gases (GHGs) in the atmosphere and leading to human-induced climate warming. Furthermore, the intensified rice cultivation in Asia has consistently contributed to GHG emissions, where significant levels of methane are a primary by-product (Kerr 2013). Deforestation, soil erosion, nutrient depletion of soils, and land-use changes have led to often irreversible impacts worldwide, especially since the industrialisation of many larger countries over the past 300 years.

In the Neolithic Age, “unregulated forest-field management” was the most important type of land use in Central Europe. After clearing, agriculture was practised for a few years. Then the area was left fallow for a long period. Regulated rotation with sowing, grazing and use of wood in certain restricted periods in an area was only developed in the Bronze and Iron Ages (about 3700 BP), when sickles and iron ploughs replaced the wooden hook plough. The major crops grown during this period were primarily wheat, emmer and spelt. At the regional level, buckwheat was important, as were barley and rye later. The primary livestock of cattle and pigs were used to graze in forests, where the canopy trees provided wood for housing and firewood.

The first coppiced forests were developed in the Bronze Age. Wood was used for charcoal kilns and the smelting of iron ore. Hedges grew along species-rich edges of forests and nutrient-poor meadows, and dwarf shrub heathland expanded, leaving the regeneration of grazed forest areas endangered. Many deciduous trees were harvested for their branches (lopping) to provide livestock food for the winter. Thus, natural vegetation was not only considerably changed on arable fields and replaced by crop plants and their accompanying vegetation, but also in structure and species composition. In addition, large areas around settlements were influenced by grazing, trampling and tree felling and the use of other non-timber forest products, such as leaf litter.

For the last 2000 years, the youngest history of climate and vegetation development has been the focus of IGBP’s “Past Global Change” (PAGES 2k) project (Kaufman et al. 2013). A continued cooling event has been confirmed with modest fluctuations (Little Ice Age between 1550 and 1850, with three warmer intervals) until the beginning of the industrial period (circa 1700 AD). During this period, on average, global temperature was still 0.3 K lower than that found over the last millennium. These findings should be taken with a grain of salt, as considerable variability and a combination of anthropogenic and natural drivers skew the effects at the local scale.

In Central Europe, until the beginning of modern times, human activity focused on clearing areas for agricultural purposes and settlements. Forests were degraded by grazing animals and by the increasing demand for wood, which was the sole energy source for cooking and heat prior to the use of coal. The ratio of forest to open land at the end of the first clearing period was approximately 70:30, despite Neolithic and Roman settlements. In the thirteenth century this ratio flipped to 30:70. Natural forests were restricted to areas unfavourable for agriculture. Not only was the vegetation cover changed drastically, but considerable soil erosion also occurred, outlined by the alluvial deposits in most European valleys (Dotterweich 2008). Towards the end of the Middle Ages, considerable breakdowns within the social system would lead to slight increases in forested areas. At the dawn of the sixteenth century, some of the first forest regulations were enacted as a means of protecting and limiting the use of forests. As early as the twelfth century, in some regions, the grazing of goats in forests was already forbidden.

The remaining forests were structurally and floristically changed. To maintain the energy requirements and to consciously save areas of forests for grazing, coppiced forests were managed. Oak and hornbeam were particularly suited to this kind of forest management because they regenerate easily through coppicing. The fruits of oak, beech and hazel were not only important feed for livestock but were also consumed by humans. These species were thus cultivated in open forests so shade-demanding woody species decreased. Oaks were grown not only for acorns (animal feed) but also for their bark, which was used for tanning. Large, emergent trees were retained as standards in the coppice system for timber, which was required for construction (Box 17.1A).

Forest management strategies like that of the “thinned coppice” allows several light-demanding herbaceous and woody plants to become established. However, losses of species can occur owing to heavy grazing from livestock, in addition to the use of leaf litter and humus-rich topsoil for barns. The excessive collection of humus and leaf litter contributed significantly to the loss of nutrients in the system and led to large differences in fertility between forests and arable land. Excessive use of forests on sandy soils caused the formation of heathland in Northern Europe. Chronicles describe the scarcity of wood, which led to the use of peat from bogs.

Today grazing and browsing in forests occur worldwide and are regarded as the anthropogenic factor that has had the most significant influence on vegetation (Ellenberg 2009) (Fig. 17.13). The increasing level of grazing had two consequences. First, depending on the original forest type and the number and type of grazing animals, many different plant communities, dominated by herbs and grasses, developed. Moist meadows, nutrient-deficient meadows, hay meadows and pastures with grasses able to regenerate rapidly expanded and excluded woody species. Occasional burning stimulated some herbaceous species, while others disappeared. More animals led to an increase of weeds in pastures, and plants sensitive to trampling disappeared. Poisonous species and those possessing protection (i.e. thorns) flourished and became dominant in the system. Animals contributed to the expansion of shrubs, dwarf shrubs and herbaceous species. To control the species composition within their fields, farmers would regulate the frequency of mowing, periods of grazing and the number of animals allowed to graze. A particular feature of this vegetation type is the large number of nitrophilic species in areas where many animals were kept (“Lägerflur”, Sect. 20.​3).

Previously unknown plant communities also developed with agriculture. During the Middle Ages, the permanent cultivation of rye without crop rotation, together with the growth of other cereals and buckwheat, led to so-called three-field crop rotation. The succession was winter cereals, summer cereals and usually a compulsory grazing fallow period that would regenerate the balance of nutrients through the input of manure as fertiliser. The first communities of weeds on arable land still contained many perennial species because of the fallow period and the relatively rare disruption from ploughing the soil. The highest species diversity in Central Europe occurred at the end of the Middle Ages as a result of human activities, which also led to a higher diversity of plant communities and habitat. In valley landscapes of Central Europe, a number of different important processes occurred that led to a number of different communities (Fig. 17.14). The valleys were important not only for the expansion of economic innovations but also for the dispersal of plants because of their small-scale differentiation of site conditions. The increase in the richness of the flora was only possible because of the disturbance of natural sites and resulted in the creation of new sites. The plant communities within the valleys were much different from those occurring within a forest or open fields because the depletion and accumulation of nutrients, irrigation and drainage were controlled by many different microclimatic processes.

The agricultural landscape of the Middle Ages was driven primarily by two developments. The first one was the introduction of crops from the Americas (e.g. potatoes, maize, quinoa). The second one concerns an extremely varied mosaic of plant communities, resulting in high compositional and structural diversity, which became very much endangered as a result of subsequent mechanisation in agriculture. The temporal development of species that have become naturalised (archaeophytes and neophytes), as well as those that have become extinct, have been summarised for Central Europe (Fig. 17.15). Despite the fact that species extinctions occurred mainly after the nineteenth century, the total number of plant species is still increasing owing to the continued emergence of more neophytes.

17.2.4.3 Agriculture and Forestry in Industrial Times

Since the onset of the Industrial Revolution and continued urban sprawl, significant changes have occurred in plant communities, where even the smallest of areas have now been affected (Sukopp and Wittig 1998). In the past, localised ploughing, watering and draining had an effect that was only apparent at a specific scale; however, as the influence of supplying and managing resources and their waste products increases, so does the effect on living systems. Pollutants and the effects of fertilisation have undoubtably influenced vegetation. For example, atmospheric deposition of nitrogen currently exceeds natural background levels by three orders of magnitude in some regions, leading to a steadily increasing eutrophication of terrestrial ecosystems (Galloway et al. 2004).

The beginning of the industrial period is linked with new anthropogenic influences on plant cover. At the onset of industrialisation, most of the remaining forests of Central Europe had already been degraded and the soils were much eroded or depleted of nutrients, which left the upper layers acidified. Although the conditions of the forest were maintained, the change in dominant energy source (coal) led to a significant makeover in terms of people thought about forest management. As the use of coal began to increase, many traditional forestry practices were no longer used; however, the demand for timber, paper and energy grew with the onset of the steam engine and railway.

Attempts of reforestation occurred as early as the fourteenth century. However, it was only in the year 1713, when von Carlowitz put forth the “principle of forest sustainability”, that large-scale afforestation projects were undertaken (Pretzsch 2005). It was at this time that the transition from deciduous forests to coniferous tree plantations commenced. Only some mixed forests were retained until the nineteenth century; however, they were considered to have become degraded. The so-called new forest management started with the intensive plantation of conifers on degraded lands and within abandoned fields, where this practice was used to meet the increasing requirements for construction-grade lumber. In Fig. 17.16, characteristic forms of forest management are shown. On the other hand, there are also individual steps of degradation from a natural forest to wasteland (Fig. 17.13). Experiments with trees, such as the Douglas fir (Pseudotsuga menziesii), were successful, and spruce (Picea), pine (Abies) and larch (Larix) were planted beyond their natural ranges. Coppiced forests slowly transitioned to high forests (mature). Various forms of forest management practices were used, for example, clear-cutting, shelterwood-compartment systems and shelterwood-selection systems (Box 17.1). The aim was to produce commercial timber as quickly as possible while maintaining a sustainable production yield (harvesting < regrowth).

Managing the forest using these strategies was successful for many reasons: forests were no longer used for livestock grazing, pruning or sod-cutting, and the use of leaf litter was significantly reduced, which in turn returned substantial amounts of nutrients back to the system. The most important change in these forests was the change from the once dominant deciduous trees to conifers, which triggered a change in vegetation dynamics. All coniferous species, particularly pine and spruce, are relatively undemanding in terms of regeneration and can be sown without major drawbacks. Furthermore, conifers tend to grow quite rapidly, resulting in much shorter rotational periods (80 years) than those of oak, which can be twice as long. The extent of this change from predominantly deciduous forests to coniferous forests is considerable, as depicted on maps of German forests over the last few centuries (Fig. 17.9).

Reforestation with only conifers can cause significant problems for forest communities. Soil erosion and acidification are higher in spruce plantations than in mixed deciduous forests. Bird diversity is also significantly reduced, where the number of species and their abundance is half as much in conifer stands as in deciduous forests (Goudie 1994). These findings were reinforced in the Chilean deciduous forest, where differences were more abrupt compared with exotic pine plantations (Finckh 1995). It is well known that coniferous plantations typically have much less food resources for birds and also lack the multi-tiered structures found in deciduous-dominated stands.

The separation of forest use and grazing was also important for the development of vegetation in open spaces. To protect forests and arable land from animals, fences were put around grazing areas. In early summer with a surplus of food, animals tend to favour the most palatable species. Areas of eutrophication due to the excrement deposits of livestock and grazing areas gave rise to a characteristic pattern. Particularly noticeable was the damage caused by trampling and erosion by too many animals. This can be counteracted to some extent through the use of mineral fertilisers, and changing the variety of crops can improve yields. Soils of low fertility were fertilised and sown with grass (Vergrünlandung). Meadows were mowed, often several times per year, to provide enough fodder for animals that were kept in stables during the winter months. Cereal fields were also used for grazing after harvests, which affected segetal flora. Nutrient-poor meadows were invaded by scrub, and sites for thermophilic and xerophilic species became rarer. These examples show how new agricultural techniques and trade patterns affected the composition of plant communities.

In agriculture, the improved three-field crop rotation was increasingly used. The fallow was abandoned and replaced by root crops, in particular potato. A significant agricultural advancement occurred in the middle of the nineteenth century, when N and P mineral fertilisers became available. As a result, the diversity of site conditions decreased, and flora diversity was reduced with fewer fallow fields occurring owing to the increased use of fertilisers and herbicides, where many perennial species of segetal flora disappeared. Marshes, bogs and river forests were drained and converted to meadows or pastures. Edges of fields, hedges, terraces and marginal biotopes, all with important interconnecting functions for non-crop plant and animal species, were drastically reduced. Schreiber (1995) demonstrated how various central European grassland communities, which had developed up to the middle of the nineteenth century, were converted into intensive grassland by new agricultural practices (Fig. 17.17). Currently, intensive land use has led to drastic losses of diversity in the landscape and plant communities in all regions of Europe. The colourful mosaic of a rural landscape developed for extensive use since the Middle Ages has only been sustained in some remote areas.

Already 20 years ago, Vitousek et al. (1997) estimated that between one-third and a half of all terrestrial ecosystems had been transformed by human activities, more than half of the total surface of freshwater resources is used by humans, and one-quarter of the bird species on Earth have already gone extinct as a result of anthropogenic influences. More recently, Ellis et al. (2010) characterised the artificial changes in the terrestrial biosphere since the Industrial Revolution on the basis of gridded global data for human population density and land use. According to their evaluation of the past three centuries, the terrestrial biosphere passed from mainly “wild” to mainly “anthropogenic”. With fewer small subsistence farms it is no longer necessary to combine arable crops with animal husbandry. The former “balance” of forest, grazing land and arable land once required in the rural landscape for reasons of greater self-sufficiency has now been lost. The impacts of human transformation of ecosystems over such a short time span during the last few decades has no parallel in the past with respect to intensity and consequences. Nowadays, even more drastic and irreversible shifts of global ecosystems are being discussed, potentially threatening Earth’s ability to sustain humans and other species in the near future (Barnosky et al. 2012).

17.2.4.4 Recent Developments in Forests, Arable Fields and Settlements

Intensifying production in agriculture continues apace. Machines in agriculture are becoming increasingly heavier and cause increasing concerns regarding soil compaction. As a result, soil organisms suffer, particularly the macrofauna with Lumbricides. However, yields are still increasing, because modern plant-breeding programmes continuously provide new varieties (strains) with higher yield potential. Varieties must be continuously improved as they become resistant towards xenobiotica, which also require continuous changes. An unwanted “co-evolution” takes place, and the WBGU risk assessment (2000) points to the dangers posed by genetically modified crops that are resistant to herbicides and viruses if such transgenic crop plants become wild or cross with indigenous species.

Intensive agricultural practices also result in fewer species in the associated flora, which are often resistant to pesticides and herbicides. Therefore, measures for the protection of segetal flora are required. Various attempts have been made in this regarding, including mixed cropping, intercropping and managing strips at the edges of fields. However, the present-day agricultural landscape is characterised not only by mechanisation, specialisation and high production but also by homogenisation and contamination.

The use of land for settlement and industrial sites has grown enormously. Habitats and communities, which are without parallel in the natural landscape, are investigated in urban ecology. Changes are especially apparent for flora and fauna within an urban setting (Fig. 17.18). These areas are basically warm islands where warmth- and light-demanding species find refuge, but they need to be relatively insensitive to mechanical disturbance and pollutants.

Urban flora is fairly diverse and develops many rather constant plant communities owing to new non-natural sites being colonised mainly by ruderal species. Lichen flora is particularly well represented in towns. Especially epiphytic lichens are often taken as the standard for assessment of SO₂ pollution and evaluated regarding their toxic tolerance. New characteristic plant communities are also found in villages. Wittkamp et al. (1995) compared the vegetation of northern Bavarian and southern Thuringian villages on both sides of the former border between the western and eastern parts of Germany and found that variation in land use was closely linked to variation in vegetation. The comparison also showed that social structure and economic activity were important for this variability. Quite often, differences in vegetation caused by site conditions are of secondary importance compared with differences due to land use. The marked differences in the actual vegetation cover of landscapes north and south of the Strait of Gibraltar are strongly determined by different agro-political measures, culture-specific characteristics and economic activities. With almost identical natural initial conditions, a completely different inventory of communities developed with the intensive, market-orientated land use of Spain compared with the subsistence-orientated land use of Morocco (Deil 1995).

Recent human influences on vegetation in Central Europe is also well illustrated by the introduction of a new type of forest decline. Damage to trees and complete forest areas by smoke and other pollutants from industry has been reported since the beginning of the industrial age. It is also known that rejuvenation of forests in Central Europe has become more difficult, for several reasons, including damage by pests, attacks by fungi or other pathogens, climatic abnormalities and above all from acidification of the upper soil layer. It is now certain that the soils of many semi-natural woodlands and managed forests were depleted of nutrients because of litter removal, erosion and monoculture practices over the centuries.

The “new forest decline” affecting complete forest ecosystems has only been observed for the past 50 years. Significant damage has been attributed to the influx of acid from the atmosphere via long-distance transport in the form of acid rain and complex interactions with other drivers of environmental change (Sect. 16.​4). However, analysis of the recent history of vegetation gives much evidence that the key to understanding present-day vegetation is a thorough recognition of past management methods and their effects. This is relevant not only for highly modified ecosystems such as agricultural fields, managed grasslands and heathlands but also for more natural vegetation types such as forests or bogs (Ellenberg 2009).

According to the results of the PAGES 2k project, the twentieth century was the warmest in the last 1400 years. The increase in temperature between the nineteenth and twentieth centuries was the highest ever, and the long-term cooling trend was reversed as a result. Cooling in pre-industrial times was certainly mainly due to natural forces. Because these factors did not change, the actual warming, which has exceeded the known dimension until now, can only be explained by new anthropogenic influences, leading to actual global change (Chap. 21), and there will also be more changes in vegetation cover in the future.

17.2.6.1 Mediterranean Region

Plants in the Mediterranean region are particularly stressed because of the dry periods in summer. The influence of fire must also be considered, particularly when human intervention prevents it. Tree species found within the upper mountainous zones of the Mediterranean listed in the scheme of altitudinal zones in Fig. 17.19 typically have high drought resistance and are stimulated by fire (e.g. Quercus suber). Land-use changes began affecting large areas in the western Mediterranean region (bays of Alicante and Catalonia) in the early Neolithic Age, circa 7000 BP (Badal et al. 1994; Riera-Mora and Esteban-Amat 1994). The evergreen oak forest (Quercus rotundifolia) near the coast was mixed with deciduous oaks (Quercus faginea, Q. pubescens) and eventually thinned, and heliophilic shrubs became dominant.

Typical Mediterranean shrub communities (macchie, garrigues), developed. Inland areas remained untouched for a long time, disregarding the first transhumance, for example, shepherds moving in the hot, dry summers from the plain into the mountains near the coast, and establishing a “vertical migratory pattern” of grazing management. The farms on the coastal plain, with favourable soils and availability of water, for example, the Spanish huertas, were the first areas with intensive agricultural crops. Evidence of crop plants introduced from the Mediterranean into North Africa demonstrates the development of an increasingly richer flora in agriculturally managed landscapes. In the Palaeolithic period, barley (Hordeum vulgare) was evident, whereas in the Neolithic period, wheat (Triticum spp.) emerged as well as fruit trees (Phoenix dactylifera, Ziziphus spp., Lawsonia inermis) and Fabaceae (Lens culinaris, Pisum sativum, Vicia faba).

The Phoenicians cultivated olives, vine, pomegranates and figs, while the Romans added garlic, onions, apples and pears. During the Arab period, species from Central and Eastern Asia arrived, for example, citrus, mulberry and carob trees, hemp and sugar cane. Species from the New World followed after the discovery of the Americas. Only at that time did some plant species often regarded as characteristic of the region reach the Mediterranean basin: agave, opuntia and maize, tobacco and tomato. In the nineteenth century many ornamental plants were added for economic reasons, and Acacia and Eucalyptus species were introduced. Today, about 200 exotic species are naturalised in North Africa, corresponding to 4% of the total regional flora (Le Floch et al. 1990). Some species were introduced unintentionally and, along with crop plants, became established. The number of those species originating from the New World is considerable, among them so-called aggressive invaders that became established particularly on ruderal sites (species of the genera Amaranthus, Cuscuta and Conyza). Currently, rapid expansion of Heliotropium curassavicum, Solanum eleagnifolium and Xanthium spinosum are occurring, as well as of some shrub species (Ricinus communis and Nicotiana glauca) into North Africa.

In the segetal flora, on the other hand, indigenous species found new niches. The approx. 50 species found within fields and fallow land include Ammi majus, Ammi visnaga, Anagallis arvensis and Anagallis foemina, Ridolfia segetum, Sinapis arvensis and Carduncellus, Convolvulus, Cyperus and Diplotaxis species. The number of sometimes thorny weeds on pastures is just as high (Astragalus armatus, Calycotome villosa, Scolymus hispanicus and S. grandiflorum) and also partly poisonous or at least non-edible (div. Asphodelus spec., Peganum harmala, Solanum nigrum, Stipagrostis pungens and Stipagrostis capensis). The nutrient indicators (nitrophils) include Aizoon and Mesembryanthemum species, Chenopodium murale, Hyoscyamus albus and Hyoscyamus niger and Withania somnifera.

As for the Mediterranean regions, the same applies as for Central Europe: Naturalisation, expansion and disappearance of individual species are closely connected to the agricultural practices performed in those regions. Deil (1997) showed for the areas bordering the Strait of Gibraltar that vegetated landscapes can be read as a history book. Further, they can show how historic and actual forms of use demonstrate religious regulation as well as subsistence-orientated management of the indigenous population, or the market-orientated management of colonial times. Even today, it is possible to identify from the spectrum of different species an Islamic cemetery from the time before the reconquista. Similarly, it is possible to show the different uses of herbicides and the mechanical control of weeds in the fields of southern Spain or northern Morocco.

The recent history of forests in the Mediterranean regions is completely determined by human activities. During the Greek colonisation of Dalmatia around 3000 years BP, forests were intensively cleared. The Italian peninsula was largely deforested during Roman times, and the Iberian Peninsula mostly from the sixth century BC onwards. Severe consequences for vegetation and site quality have often been documented (Müller-Hohenstein 1973). Forest degradation starts with structural changes (age structure, density of stand and closure of canopy); livestock grazing in a forest leads to the invasion of ruderal species and the eventual decline of the forest, which causes a thinning of the stand. Forests on more favourable sites were cleared first, while arable fields extended on land with not very productive rain-fed agriculture. Clearing was also often supported by fires to take over grazing areas. Mediterranean dwarf shrub communities (macchie with various regionally different terms such as garrigue, tomillares, phrygana, matorral) and pastures differing in their floristic compositions are quite often determined by human activities.

Clearing was not just an intervention with irreversible consequences for vegetation; soil erosion down to the bed rock and sedimentation in valleys generally followed. The negative effects of soil degradation also has serious economic consequences, and a growing lack of wood has been an issue in the past 50 years. Reforestation efforts have increased in many southern European regions and North African countries, but so far such efforts have borne no discernible results. Exotic, fast-growing types of trees are chosen for reforestation, in particular some pine species (Pinus radiata), but a variety of Eucalyptus species are favoured for their fast growth rates (Fig. 17.19).

Where the reforestation has been successful, “forest” areas increased, but a uniformity also arose. The expected economic yields have not always materialised. The negative side effects on fauna and soil organic matter, attacks by pests and fires, for example, known from monocultures, increased. With the high diversity of Mediterranean woody species, the preferential use of exotic species would not have been necessary. The choice of fast-growing exotic species stemmed mainly from economic considerations, while ecological considerations were ignored. One could have limited those species to in areas where urgent protection from further erosion was required. In southern European countries, the pressure to harvest wood from forests is presently declining; the recovery of sclerophyllic and deciduous species can be observed. In North African countries, however, the pressure on the last near-natural remnants of forests is growing.

Also, in the European/North African-Mediterranean region, it is obvious that subsistence-oriented land management in small family farming systems coincides with a greater diversity of vegetation. Modern market-oriented management is always linked to a loss of diversity at all levels. Typical forms of Mediterranean land use are shown in Fig. 17.10.

17.2.6.2 Desertification on the Border of the Sahara

The term desertification has been used to describe environmental problems in arid areas, at least since the many years of drought in the Sahel in the late 1970s. Today it is obvious that this term describes not only climatic stress in long-lasting drought periods but also the complex interactions that include in particular human impacts on vegetation and landscapes. In the context of ecosystems in arid areas, and of the often-cited man-made deserts, it should be underlined that extended dry periods are a characteristic feature of arid areas. It is not only the small amount of available water after a precipitation event coupled with high temperatures but also the highly variable temporal, episodic and spatial distribution of this precipitation that cannot be anticipated.

In the Sahara region, it is not uncommon for prolonged periods of dryness to last for several years, as depicted historically through the interpretation of lake sediments (Nicholson 1978). The recent dry period in the 1970s in the Sahel had catastrophic consequences as it preceded a relatively moist period. During the moist period the human population grew considerably, but it found itself maladapted when the dry period arrived. Improved medical supplies and technical innovations (deep wells to tap fossil water) have improved grazing conditions and increased the number of animals being farmed in this area (Müller-Hohenstein 1993). However, it is not correct to regard ecosystems of arid areas as particularly labile systems. These ecosystems are adapted to extreme climatic variability. Autochthonous plants and animals are able to adapt in many ways, and thus are able to survive under such conditions. The human population in these areas has also adapted to maintain supplies and is prepared for variations in crop yield. A “desertification scheme” (Fig. 17.20) indicates the most important causes and consequences of land degradation in arid regions.

It is important to consider the discrepancy between population growth and—despite all the technological progress—the limited availability of renewable resources, in addition, as well as the substitution of traditional forms of management (nomadic life style) by modern forms of grazing and rain-fed agriculture. Up until the end of the twentieth century, the human population around the Sahara has grown five-fold (Goudie 1994). Today, owing to these increases in population, many examples of poor agricultural practices can be found in all areas near deserts, for example, agriculture in north-eastern Syria or eastern Jordan, which receives less than 200 mm of precipitation annually, or growing animal fodder through the use of fossil water from depths more than 1 km under the Algerian oases, all of which can further contribute to the degradation of vegetation in the area.

Human influence on the vegetation in arid areas is particularly linked to the disturbance of sites and original plant cover. As a result, many of the woody plants previously present in arid regions disappear as the requirements for energy are not met. The naturally sparse contracted vegetation, growing in areas with an above-average water supply, is damaged by overgrazing. While vegetation in drier areas with a high fodder quality typically becomes locally extinct, toxic or thorny plants can strongly expand. Local changes resulting from the increased loss of vegetation are seen in the remobilisation of dunes and increased number of dust storms. Also, water relations of areas are affected by modern irrigation installations, which result in the salinisation of soils and a rise in halophytes (Fig. 17.21).

The Sahara Desert area has probably grown by 15% in the last 100 years; indeed, it is believed that, just between the years 1958 and 1975, the Sahara extended 100 km to the north and to the south. However, such rates of desert growth are also questioned. Hellden (1991) found on the basis of satellite photos a close relationship between precipitation and vegetation, but no evidence for expansion of desert areas. It is still not exactly known whether the observed changes are permanent or whether regeneration is possible. The latter is understood as the sum of processes in an ecosystem by which lost elements may be regained and thus re-establish themselves to the point where they achieve a status equivalent to their original status. There are many positive and negative interactions between climatic events, the development of vegetation and abiotic site factors and human influences, all of which remain poorly understood. However, in all dry areas of the Old World, there have been examples of a rather rapid recovery of vegetation following a precipitation event. Obviously, the ability of species to regenerate in arid regions depends on the lower and upper variability of precipitation a plant can tolerate; however, this still remains largely unknown. Nevertheless, the existing alterations in the albedo, increasing dust load in the atmosphere and changes in soil moisture in arid regions also contribute to global climate changes and influence the regeneration of desert vegetation.

17.2.6.3 Destruction of Tropical Forests

The destruction of tropical forests has been a topic of public discussion even more than the problem of desertification. Not so much climatic but edaphic preconditions, particularly the availability of nutrients, are the limiting factors in this biome. It has long been known that soils weather intensively and nutrients are leached under conditions of high temperatures and continuous moisture, two prominent features of tropical forest biomes. Although litter production is high, fast decomposition impedes the build-up of nutrient reserves, for example, in the form of humus. Even fertilisation is not sustainable in tropical soils with two-layered clay minerals, which have only weak ion-exchange capacities. The system can remain functional only if the short-cut nutrient cycle is maintained, with the help of mycorrhizae or fine roots present in the soil. The exportation of organic substances by harvesting has far-reaching consequences for the stability of soils and plant growth. Exceptions exist only where primary cations are released because of the mineralogical conditions, for example, in some regions of Java and Sulawesi, relatively young, nutrient-rich volcanic soils favour the establishment of permanent rice crops.

Clearing and changes in land use have taken place in tropical rain forests for thousands of years. However, the consequences of these early influences are considered to be much smaller than those occurring today. Intervention over large areas began with colonial times. Of the approximate 1.5 million described animal and plant species of the perhumid tropics—globally 10–11 million (Mora et al. 2011)—many thousands probably went extinct before they could be scientifically classified. In contrast to the forest management strategies used in temperate forests, these strategies have had negative effects in the tropics, as perhaps only one single tree trunk is used per hectare and many neighbouring trees are destroyed. Even the mangrove stands in tropical coastlands were intensively used for wood extraction, despite their protective functions against normal floods and tsunamis. Sites of these unique plant communities are used in East Asia, as well as Ecuador, for the breeding of shrimp.

Not only has the management of forests led to losses of species. The settlement that followed forest harvesting limited the development of secondary forests. Governments of several rain forest states saw the forests as a “valve” for the growing population pressure. Thus, many states subsidised their populations to help clear large patches of forest for agricultural purposes, more specifically ranching livestock. Grass seed was even sown from planes, but subsequent grazing was mostly given up after a few years as substitute communities with sclerophyllic grasses (e.g. Imperata cylindrica) developed, which were not suitable species for animal fodder. The practice of shifting cultivation with slash and burn was commonly used in all moist, tropical areas and was initially thought to be less damaging. Today this form of land clearing is regarded as less damaging, particularly if the periods of fallow last long enough, and the effect of fire is less harmful to the soil mycorrhizae, which is well known to help stabilise tropical forests and promote species biodiversity (Kottke et al. 2013). However, regeneration requires rather long fallow periods, which were often shortened because of the growing pressure from the human populations. Bruenig (1991) compared the consequences of “mild” intervention with “modern” intervention and assessed the consequences using two transects (Fig. 17.22). There are typical phases of the development of land use from the natural forest to unproductive fallow areas (Fig. 17.11).

Growing evidence supports significant losses of tropical moist forests, where in the period 2000–2012 it was found using Landsat satellite imagery a total of 2.3 million km² or 32% of the total biome was lost (Hansen et al. 2013). The tropical zone not only experienced the greatest total losses, it also suffered an estimated annual forest loss of 2101 km². Land-use changes are considered to have the highest impact on the decline of species richness and are still considered a major driver of change in ecosystem functions (Murphy and Romanuk 2014; Hooper et al. 2012).

If one starts with the assumption of a relatively low β-diversity (differences of diversity between different sites, Box 20.​1), these habitat losses affect whole plant communities with highly endemic species, and particularly the specialists among plants and animals. Considerable losses in genetic material for breeding, species for medicinal purposes, food and various commodities have occurred (WBGU 2000). Cleared forest results in increases in soil erosion, which in turn leads to changes in water relations. These deleterious effects are expected to intensify with increases in atmospheric CO₂ concentrations associated with global climate change, which is expected to increase mean temperatures and the variability in precipitation (IPCC 2013) (Chap. 21).

In the wake of the destruction of tropical rain forests, various social movements have become involved in demanding increasing efforts to establish larger protected areas. It has been difficult to convince the many millions of native people in these areas to change their way of life. It is often overlooked that concepts for sustainable management already exist in some areas. Earlier hunters-gatherers and the first settled farmers developed well-adapted forms of land use in most tropical regions. Low population densities allowed long fallow periods, which had a minimal effect on the land. The population was less mobile, stimulating a closed cover of several “storeys” from different crop plants. The original land use simulated (intuitively?) the structure and biodiversity of a natural rainforest with useful plants for daily life. Today, few examples exist of sustainable land use in permanent and seasonally moist rain forests of the tropics (Fig. 17.12).

Sustainable forest harvesting is extremely difficult to maintain owing to the imbalance of nutrient content in the soil post-harvest. New integrated approaches that take advantage of local crops that are adapted to current conditions (e.g. supply with nutrients), while limiting the effect of the site (e.g. creation of particular microclimatic conditions), must be developed to educate the local indigenous populations to improve management in the area. However, in doing this, the expectations of higher yields often linked to land use in the tropics must be revised. Sustainability must be thought of more than just as a rate of production or bottom line and more in terms of the maintenance of cultural and natural heritages in a particular environment or region. Some elements of sustainable use are generally valid. These are, for example, agroforestry, mixed cropping, mulching and biological pest control.

Many questions remain open about the scientific basis for nature protection in the tropics. It is assumed that a suitable number of plant individuals, able to reproduce, must be available to maintain a species. Assuming this number to be only 500, this could require, in the rain forest ecosystem, probably several hundred hectares for some species. This concerns the protection not only of species but of plant communities as a whole. Little is known about the protection of complete communities, about the minimum size of such areas and the bordering buffer zones required to ensure the persistence of a species. The borders of national parks and biosphere reserves are usually roughly estimated, even if not determined according to political and economic criteria. To define these borders is an important task for applied ecology, which is concerned with the protection of tropical ecosystems.

17.3.2.1 Beginnings of Succession Research

Systematic research on vegetation dynamics is connected with the names of the ecologists Clements and Gleason. Clements (1916) is regarded as the founder of the discipline of successional research, starting from the idea that only undemanding species grow first on bare substrates. Once established, species change and influence the conditions and community of their site, which then facilitates invasion by others. Clear temporal periods are recognisable until the development of a uniform, final community, the so-called climax community, representing a relative equilibrium. Small fluctuations in community assemblages are possible, but a reversal of the clearly directed processes leading to this climax community is not.

Later Clements assumed three succession sequences. One of these sequences has its origin in a freshwater basin. During the processes of receding water and emerging land in temperate latitudes, deciduous forests develop as a climax via a boggy and pre-forest stage (hydroseries). Another sequence starts from a saline substrate (haloseries) and a third on bare rock (xeroseries), but all end in the same forest climax. Changes in the site always originate from plants themselves. Also, in larger areas with the same climatic conditions, the same final community will always emerge (monoclimax). This concept was well received, but there were critics who could not empirically confirm this theory. The reason for the discrepancy was that Clements concentrated on the dynamics of undisturbed vegetation (primary succession) and interferences of any kind, and developments arising from them (secondary succession), were disregarded.

Clements’ theory, which regarded plant communities as “organisms”, was rejected by Gleason (1926), who stressed that there was also an individualistic type of vegetation dynamics without clear steps and without formation of the different vegetation units. He started from the assumption that changes in species at a site depended on the species composition and the influx of diaspores to that site.

Definitions of succession or vegetation dynamics are today broader and more general. These dynamics comprise changes in structure and species in vegetation units in space and time. Short-term, daily and seasonal changes are included. All forms may be described by the term patch dynamics (Pickett and White 1985). Another general definition underlines the notion that successions are medium- to long-term dynamic processes in a directed, temporal sequence of plant communities in an ecosystem. It is often stressed that these sequences are irreversible—within limits—and predictable and may be triggered by endogenous or exogenous processes. During succession, self-organisation in the ecosystem (on the precondition that there is no disturbance) increases and a balance between primary production and mineralisation develops.

Odum (1980) lists a total of 24 different ecosystem characteristics that are able to change during a succession sequence. He regarded development as a well-ordered process resulting from changes caused by vegetation and ultimately leading to a stable ecosystem, where maximum functional interconnections are achieved. The hypothetical final state is understood as a complex community of organisms with complex links and interactions, in which once equilibrium is achieved, it can change again, for example, after a new disturbance.

However, successions are often also understood as stochastic processes, basically because the seed input in a biocoenosis at a site is random; the same applies to the germination and establishment rates or the availability of necessary safe sites and the occurrence of disturbance—their type, intensity and degree—which are generally not predictable.

17.3.2.2 Succession Types

Succession may be considered over large or small areas, with fast or slow changes. It may be directed or cyclical and may be affected by endogenous or exogenous processes, which may differ by type, intensity, duration or extent. It is important to consider the geographical location within different climatic zones and whether development occurs on newly formed soils or on those previously colonised. Finally, participating plant species will differ in their ability to disperse and compete and in their life history strategies and resource demands/strategies. Dynamic processes are triggered if net production and mineralisation in an ecosystem are not balanced. In such cases it is possible, for example, to accumulate organic matter (formation of bogs, autotrophic succession) or for organic matter to be consumed (use of humus by agricultural systems, heterotrophic succession).

Considering all these aspects, many succession types on different spatial and temporal scales have been described. Spatial and temporal relations and how these aspects are related and differ are shown in Table 17.4. In Fig. 17.29, the successional types are schematically arranged. These are the most important examples of progressive and regressive, directed as well as cyclical vegetation dynamics.

Table 17.4

Relationships between spatial and temporal dimensions in changes in vegetation

	Individual	Patch	Population	Community	Landscape	Region
Fluctuation	×	××	××
Gap dynamics	×	××	××
Patch dynamics	×	×	×
Cyclical succession		×	××	×
Regeneration succession		××	××	××
Secondary succession		×	××	××	××
Primary succession			××	××	××
Secular succession				×	×	×

The terms in the table headers refer to spatial dimensions. A patch is a small homogeneous area (ecotope), a gap a small open space in the forest (after van der Maarel 1988)

Primary Successions

Primary successions are important for understanding processes of vegetation dynamics, but in reality they are rarely observed because they typically occur over long time periods where an uninhabited abiotic substrate needs to be present. Knowledge of individual stages of the sequence of primary successions derives mainly from newly formed sites, for example, following a volcanic eruption (Fig. 17.30) or in newly developed coastal regions or at the receding edge of a melting glacier (Fig. 17.31). Also anthropogenic sites, such as rubble heaps, quarries, road cuts and so forth, may be regarded as starting points of primary succession. Despite the differences in sites, similar trends and basic patterns may be discerned in such situations.

Very few plants are able to cope with the hostile conditions of a site during the initial phases. On rocks, cryptogamic communities develop first, perhaps lithophilic lichens, then fruticose lichens, mosses and ferns. In unfavourable exposed positions, for example on steep rock walls or on mobile debris, development rarely goes beyond this phase (permanent pioneer communities). Usually, the pioneer communities, which themselves depend on the input of propagules from the surrounding areas, change the site over the course of time, which allows for the formation of a soil substrate. Many Leguminosae are herbaceous pioneers that facilitate the accumulation of nitrogen in this new substrate. Upon the establishment of later pioneer plants, these will help change the local microclimate. If flowering plants have not become established during the first phase of succession, they will now follow. These are predominantly heliophilic, epigaeically germinating annuals or, where climatic conditions do not allow such establishment, undemanding (i.e. able to grow on a nutrient-deficient substrate) herbaceous hemicryptophytes and chamaephytes.

As succession progresses, the pioneer species are excluded more and more with each step by more demanding species. When this change takes place solely because of the changes in the vegetation itself, it is called autogenic succession. If other causes are more important— for example, natural climate change or a lower groundwater table caused by human extraction of water—this is called allogenic succession. If successions proceed under steady external conditions, there are further stages, after the pioneer stage where competing species occur, that are superior to the pioneer species in vegetative and reproductive growth. These species grow taller, despite lower rates of growth, because of their longer life span (occurrence and increase of shrubs and trees), and they are more shade tolerant but often have less effective mechanisms for distribution, despite their more effective occupation of space in the long-term. Trees are superior to shrubs, forming polycormones, which enable them to dominate for a certain time. In the final phase of such development, trees dominate. This type of primary succession is understood to be progressive because the diversity increases at the level of the species as well as with regard to structure.

Common dynamic trends, despite the apparent individuality of responses, are summarised in Fig. 17.32. At the beginning of primary succession, chaotic interactions dominate, which cannot be predicted, because input and the establishment of propagules are, to a high degree, stochastic; widely distributed annuals dominate. Between the development of vegetation and changes in the site, feedback mechanisms trigger (Table 17.5). This means that plants themselves contribute, to a considerable degree, to creating environmental conditions that favour conspecific species. The dominance of exogenous factors is only slowly substituted by endogenous factors and is accompanied by a change in species and life forms. In the labile pioneer stages, intra- and interspecific competition is not very important, but in the later stages competition plays a decisive role in the development of communities that become increasingly resistant to external disturbances.

Table 17.5

Environmental changes during succession (after Klötzli 1993)

1. Species structure
Species composition Number of autotrophic species (autotrophic diversity) Number of heterotrophic species (heterotrophic diversity) Species diversity Niche specialists Size of organism Life cycle Selection pressure Type of production	First rapid, then stepwise change Increasing, often early in secondary succession Increasing, until late in secondary succession Initially increasing, then stable or again decreasing Initially broad then narrow Initially small then large Initially brief and simple, then long and complex Initially r-selection then K-selective Initially quantitative then qualitative
2. Nutrient cycles
Cycles Nutrient exchange Role of organic debris (detritus)	Initially open then closed Initially rapid then slow Initially unimportant then very important
3. Organic structure
Total biomass Stratification Inorganic nutrients Recalcitrant organic material (e.g. humus) Biochemical diversity (e.g. pigments) Chlorophyll amount	Initially small then large Initially one layer then multiple Initially in soil then in biomass Initially little then much Initially low then high (increasing accumulation of toxins!)Initially small then large (in secondary succession little difference!)
4. Energy flow
Relation to food chain Gross primary production Net primary productionTotal respiration Primary production per unit respiration Primary production per unit biomass Biomass per unit of energy	Initially simple (food chain) then complex (food web) During early phase of primary succession increasing (only limited or no increase in secondary succession) Continuously decreasing Continuously increasing Initially primary production greater than equal to respiration (steady-state condition) Initially large then small Initially small then large; therefore, minimal use of energy per unit of biomass
5. Homeostasis (ecological steady-state, result of irreversible processes)
Internal symbiosis Nutrient storage Stability (resistance to disturbance) Entropy (increasing entropy) Information	Initially not developed then developed Initially small then large Initially small then large Initially large (small) then small (constant) Initially small then large

Disturbances and Secondary Successions

Primary successions occur faster under warm, humid climatic conditions than under cold or dry conditions. However, it is most important that development of these successions are not affected by disturbances, which can only be rarely excluded. Thus, primary successions are particularly important for a better understanding of the processes of secondary successions, which dominate the actual vegetation dynamics in our environment.

Disturbances are the main causes of spatial heterogeneity and also affect environmental conditions and plant competition. They may disturb but at the same time also maintain biodiversity. For instance, grazing by large herbivores is an important disturbance because it can control vegetative dynamics, structure and diversity. All successional concepts, from patch dynamics to mosaic cycles, are based on disturbance regimes. Disturbances can have a profound influence on the species composition, changes and vegetative matter fluxes at a site, which can regulate habitat fragmentation. These factors concern plant communities as well as entire ecosystems, and the loss, gain and turnover of organic matter and disturbances may be the result of natural events or human interventions (Fig. 17.33).

There are several definitions and a multitude of different types of disturbances (Box 17.4).

Box 17.4: Disturbances: Definitions and Properties

A disturbance is a relatively discrete event in time that concerns an ecosystem, community or population structure (of vegetation) and changes the resources, substrate availability or physical environment (Pickett and White 1985). Because this is an absolute definition, all changes must be measured and accounted for. In contrast, by some relative definitions, disturbances are regarded as events leading to deviations from the normal dynamics in an ecosystem or plant community, based on the assumption that a “normal” state is known. Furthermore, severe disturbances can range from having very far-reaching consequences, up to a complete floristically and structurally different plant community, to very small ones, for example, affecting the ability of a single plant to resprout.

Richter (1997) distinguishes between endogenous and exogenous disturbances. The former occur regularly; they are a feature of a system, so organisms in the community adapt. This is the case for seasonal changes and for changes such as frequent, regular flooding of meadows. Most disturbances that then lead to secondary successions are exogenous and are typically natural in origin (e.g. catastrophic storms or landslides and avalanches); however, they can also be caused by human activity.

To overcome the high complexity of disturbance regimes, Buhk et al. (2007) established a classification scheme that takes into account the disturbance type (e.g. felling, flooding), disturbance space, form and distribution (spatial dimensions) and disturbance frequency, seasonality and duration (time dimensions).

The intermediate disturbance hypothesis (IDH) puts forth an often discussed theoretical background to evaluate the effects of a given disturbance (Sect. 13.​5 and Sect. 20.​3). This hypothesis describes the relations between the type and force of the disturbance and the species diversity in the area concerned (Huston 1979). It states that at intermediate levels of disturbance, species diversity is at its highest. No species will be lost by being overly disturbed, and the competitive exclusion of a species is not very likely. Field tests of the IDH showed mixed support for and against this hypothesis (Mackey and Currie 2001).

In our managed landscapes, anthropogenic disturbance dominates because it begins with processes that immediately affect individual plants or that alter the environment and affect the whole community. It is therefore useful to differentiate between natural and anthropogenic disturbances, such as unusual weather conditions, flooding and landslides versus logging, mowing and trampling. However, some disturbances may be a combination of both, for example, fire (Fig. 17.33). There are spatial and temporal patterns in disturbances that help to understand the patterns of succession. Richter (1997) attempted to characterise disturbances in different climate zones (Fig. 17.34).

Secondary succession sequences may be classified according to the responses to disturbance (regarding the material balance of an ecosystem):

• Disturbances may interrupt the dynamics and the vegetation but leave the ecosystem at a certain stage (permanent communities).

• Changes occur without clear direction within vegetation, for example, because of long-term fluctuations in the water level.

Influences mentioned can lead to faster or slower, as well as irreversible, alterations. A response that reverses the process or sets it back to a previous step or direction of change is known as regressive succession. These include successions that occur as a consequence of over-exploitation or selective use of forests and pastures.

Most often secondary succession leads to reestablishment of the state that preeceded the disturbance. Here, too, the changes in direction lead to a more complex form of organisation in the plant community. However, secondary successions have different initial stages. Their dynamic changes start from an already settled substrate and may build on an existing seed bank. This means that secondary successions progress faster than primary successions; right from the start there is competition for space and resources, and only species that can compete are able to invade.

In Central Europe, numerous studies investigate secondary succession on fallow land and meadows. Schmidt (1993) describes succession over 25 years on (previously sterilised) fallow fields and found that after only 2 years of occupation by annuals, the persistence of herbaceous plants allowed them to become established by the eighth year, after which dwarf shrubs and shrubs occurred up to the twentieth year. Finally, successions continued with pioneer forests, or a pre-forest stage. These phases were initially limited by nutrients. Even if individual phases develop relatively regularly, it is difficult to analyse this development and to define the decisive factors determining the succession because of the many factors influencing and feeding back into the process (Fig. 17.35).

Secondary succession in fallow pastures is delayed because of the dense grass layer. Schreiber (1995) shows from observations over 20 years that the establishment of woody plants can occur in several ways. Occasionally, well-known basic processes in the succession of different functional types of plants from annuals to woody species occur. However, pastures, initially relatively homogeneous, often show the formation of dominance patterns with few species that cover the area well after just a few years. With good nutrient supply, herbs dominate because biomass is not removed (auto-eutrophication), but with poor nutrient supply grasses dominate. The temporal sequence of individual stages differs considerably. In some areas, 10–15 m high pre-forest vegetation occurs, while on other potential forest sites neither trees nor shrubs grow. The difference in the establishment of woody plants does not fit any succession model so far described; a prediction about changes in life forms and species is not yet possible, even if occasionally such phases in vegetation dynamics may be clearly seen.

A special case of secondary succession is polycormophyte succession. Vegetative lateral shoots of plants form colonies that are able to expand in closely covered herbaceous plant communities even faster than via seeds. This type of succession usually starts with a pioneer woody species possessing a superior defence against herbivores, for example, Prunus spinosa. Such plants are able to expand over large areas for decades. Finally, taller, unprotected shrubs and trees become established in centres of such areas, so that forest islands and, ultimately, closed forest areas may result. This development can be interpreted as an autogenous succession, regulated by the vegetation itself. The study of this form of succession allows strategies of competition in woody species to be recognised (Fig. 17.36).

The concept of succession as a directed, more or less deterministic, although predictable, process must therefore be corrected according to empirical results obtained in recent decades. These changes over time are very complex, with many variables influencing development, for example, site, biological reproduction, space and time. Particularly important are the type, intensity and duration of disturbance. All these factors make it difficult to regulate and predict successions.

Cyclical Vegetation Dynamics

Secondary successions not only develop on fallow or abandoned land but also in plant communities affected by all kinds of disturbances. Here processes directed towards a re-establishment of the original status may occur and fit into a cyclical regeneration scheme. Examples originate from managed forests where humans are the driving force, ensuring that the cycle is maintained. Pignatti and Pignatti (1984) analysed such regeneration cycles for Mediterranean forests (Fig. 17.37) and showed two variants—one for regeneration after clearing and one, more regressive, succession after several fires. After clearing, various weeds become established that are later on outcompeted by oaks. Reoccurring fires will lead to a permanent stage where regeneration takes place—if at all—only after a long period.

The final phase of secondary succession can be understood as a kind of self-preservation cycle. In a cyclical succession, plants of different ages enter and replace each other in the same vegetation community. Structural changes with the same or different species at one site are important. Further concepts about cyclical vegetation dynamics originated from Watt (1947), who introduced the “gap” into the discussion, that is, sites where these processes take place. Remmert (1991) coined the term mosaic-cycle concept to describe such cyclical dynamics. An important observation was that the climax stage (mature stage) does not extend over enormous areas in natural forests, but that various phases of development are spread out like a mosaic. All stages co-occur adjacent to each other in small patches. For Central European forest areas three phases are distinguished:

• A rejuvenating and juvenile phase, with young plants of the same or different tree species, with the previous phases occurring again. Often light-demanding trees invade first, followed by shade-tolerant trees.

• An optimal phase corresponding to a forest of uniform age with few species, a closed stand with little undergrowth.

• Ageing and decaying phases when the tree layer is disrupted over large areas and species requiring light are able to invade.

These phases may be regarded as a continuous process of self-thinning.

Figure 17.38 shows schematically the cyclical vegetation dynamics for a tropical rain forest. The size of the pieces of the mosaic differs, depending on the diversity of species. In nemoral deciduous forests an average size of 1–2 ha is assumed; in boreal forests the pieces of the mosaic cover several square kilometres because of the influence of fires. In contrast, in tropical rain forests they are hardly ever larger than 100 m² (treefall gaps). The duration of cycles also differs. For the well-analysed forests of the boreal zone several centuries are assumed; in the North American sequoia forests cycles last several thousand years. Processes of vegetation dynamics similar to mosaic cycles also exist in North Atlantic heath and bog complexes.

Disturbance is again seen as the driving force for cyclic vegetation dynamics. The life span of “key organisms” is important, and competition following their death is centred on obtaining light and nutrients. The same is true for mechanical influences (wind breakage, fire). Constant conditions are not to be expected in a pristine forest, which is also a mosaic of asynchronous phases of cyclical vegetation dynamics. During the individual phases, the structural characteristics of forests and their species diversity change.

Today, additional types of cyclical vegetation dynamics are distinguished based on questions of scale and preferential occurrence in different biomes. The formation of mosaics over large areas in Hawaii is known as a consequence of the “death of cohorts” (Müller-Dombois 1995). Demographically unfavourable situations caused and maintained by fire and storm damage are the driving force. In this context, the mass die-off of bamboo-like grasses in all tropical regions should also be included. However, it is still not clear whether only endogenous (age) or also exogenous (climate) influences are responsible; it is likely that a combination of both is responsible.

In cohort dynamics, as in the mosaic cycle, the species composition may change. However, in so-called carousel dynamics, the pattern formation in grasslands occurring over small areas is not included according to van der Maarel and Sykes (1993). The basic idea of this change over time is that species are able to recycle in a very short time period on the smallest of scales. A regeneration gap is a sufficient free space for seeds to germinate and become established. These spaces become available through the death of individual plants and are conquered in a sort of “guerrilla strategy” or even according to the lottery principle of species in the community in which all have occupied the same regeneration niche, or a first-come, first-served approach. Because these species are short lived, the carousel model describes vegetation dynamics in the smallest space and of the shortest duration. Populations of a species are always available, but they are very mobile, and the species composition remains constant. Figure 17.39 compares different spatial and temporal scales of cohort, gap and carousel dynamics schematically.

17.3.2.3 Aspects of Applied Succession Research

The dynamic nature of plant communities has often been identified via bio-indicators. It is assumed that the response of a plant community to disturbances is specific, and perhaps quantifiable. The vitality of species of the plant community and shifts in the species spectrum are measured. Knowledge about the formation of plant communities and their regulation is gained by empirical findings of many succession studies. For instance, pioneer plants with uniform seeds are selected for the greening of open spaces or to stabilise open slopes after road building (Fig. 17.40).

Today, we know about asynchronous vegetation cycles and therefore understand the limits to the practical application of current knowledge of vegetation dynamics. Management recommendations for the maintenance of certain ecosystems also need to be revisited, as do the minimum spatial dimensions of protected areas. Current knowledge of vegetation dynamics helps in the selection of “replacement areas”, where disturbed or lost communities may develop again.

Biomonitoring was developed in order to understand syndynamic processes, to identify damaging influences so that they could be quickly counteracted or minimised. Examples are the indication of air pollution and the eutrophication of water. Vegetation changes resulting from interventions at certain sites can also be recorded by continuous biomonitoring. Multiple scales on the hemeroby (level of naturalness) of spaces for vegetation have been devised. Red List species, loss of species, proportion of annuals and neophytes are important values for structuring vegetation units according to the degree of human intervention. Ultimately, knowledge about vegetation dynamics is applied in attempts to improve areas to be used for agriculture or forestry and to control such attempts. Thus, for example, the application of Ca-containing fertilisers in order to combat the acidification of soils in forests and the effects of herbicide application in vineyards, leading to the loss of endangered species, were assessed in this way.

17.3.3.1 Initial Floristic Composition

The model of initial floristic composition (IFC), based on Egler (1954), is still used to describe the sequence of successions because it represents sequences that are physiognomically recognisable. According to this model, the complete set of species is present from the beginning, and a sequence of starts results in only a few of these species becoming dominant. Figure 17.42 shows the course of growth for six species from germination via establishment of seedlings, through to the vegetative stage, and finally to the reproductive growth phase. Species 1–4 are relatively short-lived therophytes to short-lived woody species, and species 5 and 6 are long-lived, shade-tolerant species. When the latter species are fully developed, the short-lived ones are no longer able to persist. IFC is still accepted as a descriptive model of primary as well as of secondary successions. There are no generally valid succession sequences because of external disturbances and stress situations caused by them. Egler introduced this concept of gradual change as the centre point and prefers the term “development of vegetation” to that of succession. The limitation of such a model is the exclusion of the propagule influx.

17.3.3.2 r and K Strategies

The concept of r and K selection is based on models of population growth in animals, which was developed by MacArthur and Wilson (1967). This concept is used by botanists and zoologists to stress particularly contrasting characteristics of species. “r” is the intrinsic rate of population growth or rate of reproduction, while “K” represents the carrying capacity or maximum population size. Essentially, the r and K concepts stress the allocation patterns of limiting resources that are inversely related.

r-strategists form a large number of very small seeds representing a high proportion of their phytomass; they also tend to have an efficient mechanism for dispersal and grow fast without high biomass accumulation. This leads to a much quicker time to reach reproductive maturity, and such plants also tolerate long periods of seed dormancy, but the seeds can be activated very quickly. Species that are r-strategists are short-lived opportunists that are able to become established faster in their environment than shade-tolerant species, but they are inferior competitors to K-strategists.

K-strategists are long-lived and slow-growing and accumulate large amounts of biomass over time. For this reason, and because they are shade tolerant, they are able to outcompete r-strategists. They invest less into reproduction and will typically produce a small number of seeds that are larger than those of r-strategists. However, they need to form metabolites for defence to protect long-lived leaves and other organs from herbivory. In terms of successional stages, it is easy to determine that early stages are dominated by r-strategists, while later stages are dominated by K-strategists. On fallow land, for example, annual and biennial species are replaced over time by shrubs, which are then taken over by trees.

The aforementioned characteristics are accepted as important trade-offs, but a continuum is seen between the extreme positions of r- and K-strategists. This continuum also indicates not only that disturbed habitats are colonised by r- strategists in the short-term (Fig. 17.43) but also that those habitats permanently dominated by K- strategists do not necessarily allow stress-free growth. With the requirement for stress tolerance, a further strategy type is indicated.

17.3.3.3 Strategy Types of Grime (C-S-R Model)

Grime (1974) and Grime et al. (1988) worked out one of the most well-known models of vegetation dynamics: the triangle model of primary ecological strategies. Stress tolerance, adaptation to unfavourable conditions and reaction to disturbance missing in the r/K strategy model were regarded as particularly important. Three strategy types were introduced (Sect. 19.​3):

Competitor strategist (C): long-lived, competitive species on favourable sites without limitation of resources and almost stress free. They use resources particularly well, often possess storage organs and show considerable plasticity in root and shoot formation, continuously produce leaves that live for only a short time, and have low seed production. These are perennial herbs, shrubs and trees, which form in mid to late successional stages.

Stress-Tolerant Strategist (S): species adapted to unfavourable sites with limited or poor availability of resources (e.g. inadequate light, drought, nutrient deficiency, frequent frosts). They are long-lived, but with low productivity and reproductive rates and are often restricted to sites with little or no competition. Extremely stress-tolerant species are lichens on cold and dry sites.

Ruderal Strategist (R): short-lived, usually herbaceous species with fast growth rates and high seed production, usually self-pollinated with rapid seed ripening and dispersal, disturbance-tolerant opportunists and pioneers, but weak competitors. Therefore, they grow on sites with frequent, natural disturbances.

However, it is difficult to classify many plant species into only one of these three strategies. Intermediates are frequently observed, where plants with characteristics of several strategic types in different combinations are possible. On favourable sites where competitor strategists should dominate, C-R transition forms are found if the favourable conditions are temporarily disturbed, for example, by mowing meadows. Stress-tolerant competitor strategists (C-S) are long-lived species; for example, many tree species of nemoral forests have adapted to non-optimum sites. Stress-tolerant ruderal strategists (S-R) grow on unfavourable sites that are often disturbed. Included is also an intermediate type (C-S-R) combining several strategies and may also occur in temporal sequence within the same species.

Grime et al. (1988) attributed life forms to some of his strategy types and included them in areas of his so-called strategic triangle (Fig. 17.44). These strategy types are linked to succession sequences, assuming different productivities. On productive sites, development proceeded from ruderals via competitively strong types to those plants with competitive-stress strategies. This applies in Central Europe to beech forests on limestone, where on unproductive sites the final stage is determined by stress-tolerant species. The more productive the site, the higher the proportion of competitively strong species in the mid-successional period. The intensity of competition increases with increasing productivity at the site. However, Tilman (1990) argues against this conclusion.

17.3.3.4 Resource Ratio Model of Tilman

All plant species are limited in their distribution by resources, which are often difficult to acquire but are required for survival. A species is deemed successful in competition for such a limiting resource if it is able to reduce the level of nutrients to the lowest level (called R*), resulting in competitive exclusion (Sect. 19.​3). Tilman (1982) developed this model to describe the growth of algae, while for terrestrial systems the assumption is made that the most important limiting resource is within the soil (typically nitrogen) or is light. Any limitation in either of these resources will lead to increased stress within the plant. Often, these resources are inversely related, that is, a low nitrogen supply is typically associated with favourable light conditions, and low light levels (in the understorey) are commonly found at sites with favourable conditions for plant nutrition. Gradients between those two extremes lead to succession stages where plants replace each other according to their demands and their specific competitiveness. Each species grows fastest and reproduces best with the resources to which it is best adapted and where it can outcompete other species. Sufficient root biomass is very important for obtaining sufficient nutrients in the soil, but also sufficient above-ground biomass is essential for light capture by producing leaves. In primary succession, species able to cope with a poor nitrogen supply typically exclude species that require a great deal of nitrogen. As time passes, however, soil nutrition gradually improves thanks to mineralisation and nitrogen input into the system. Now, species able to cope with less light are more competitive. This hypothesis is the basis of the resource ratio model or the nutrient-light relation model.

Simpler than the ideas of this model are those of the nutrient-colonisation and light-colonisation models, or the “competition-colonisation trade-off”. The former again applies to nutrient-limited habitats. Colonists producing many seeds, which they are able to disperse quickly, are initially successful. However, they are weak in the competition for nutrients, and thus species that allocate their resources in such a way as to establish their rooting system to gain more nutrients are favoured. The corresponding situation applies to nutrient-rich habitats for light as the limiting factor. These interactions for five grass species demonstrating their ability for competition and colonisation are shown in Fig. 17.45.

From these models, the consequences for biodiversity within areas may be derived. The more heterogeneous a space to be colonised is—particularly regarding its environmental conditions (e.g. nutrients, light, water) and natural and anthropogenic disturbances—the more species are able to coexist. It follows that with increasing heterogeneity not only does biodiversity increase but, in the habitat, resources should ideally be available at intermediate levels (Sect. 20.​3).

17.3.3.5 Facilitation-Tolerance-Inhibition Models of Connell and Slatyer

Competition for resources may change by type and amount over time, especially when disturbance regimes are considered. Connell and Slatyer (1977) pointed out that the driving force behind vegetation dynamics and stress is the issue of scale, where certain dynamics only occur at certain scales. As such, three models, or succession pathways, are considered (Box 17.5):

Box 17.5: Three “Succession Pathways” of Connell and Slatyer (1977) (after McCook 1994)

Facilitation model: Facilitation is understood as the “promotion” or “enabling” of other species, starting from the assumption that the first colonists change site conditions over the course of time autogenically in such a way (e.g. change in substrate, formation of humus) that colonisation by more demanding species is stimulated. This is usually linked to the displacement of the first colonists as they become unable to cope with the competitive advantage of the incoming species.

Tolerance model: Late successional species are neither stimulated nor inhibited. However, they are only able to establish themselves if previously established individuals die or are removed, leaving an area to invade. The species now present compete with one another and behave similarly to the resource ratio model of Tilman, where species with lower R* gain dominance.

Inhibition model: Earlier colonists suppress or inhibit subsequent invaders, which will only be able to become established in the event of a disturbance or by the creation of open space for short time spans due to an individual’s death and not because of their competitive advantage.

All three models have been tested in field experiments. The facilitation model may be observed in primary successions, where pioneer species prepare the substrate, as well as following disturbance, because the succession is thrown back into an earlier phase. Both of the other model versions were tested in later, species-rich successional stages. The final steady-state stage is regarded as relatively stable, but cyclical dynamics are not excluded.

Also, the question of stability is taken into consideration. Consequently, succession is regarded as a balancing reaction of a “stable” system that has been heavily disturbed. The spatial and temporal extent of the disturbance as well as their type and intensity are decisive for the following new initial stage of succession.