Vegetation Changes in Space and Time—A Special Issue on Plant Succession and Vegetation Dynamics

1. Introduction

Plant cover on Earth is far from static. Rather it is subject to continuous alterations on different spatial and temporal scales. Changes can be very obvious or barely noticeable [1], they can be gradual or abrupt [2,3], they can be locally restricted or extend over large areas [4], they can be running over different periods of time, i.e., from days or weeks to millions of years [5], they might include variation in plant physiological processes or shifts of entire biomes [6], and they can be linked to a variety of drivers [7]. These processes are commonly subsumed under the term vegetation dynamics and cover a wide spatio-temporal range [8,9,10,11,12,13], from local or seasonal changes (symphenology [14], i.e., at a recurring rhythm linked to climate) to secular developments on a spatially large and temporally very long-term scale (synchronology [15]) controlled by vegetation history, e.g., the post-glacial vegetation development. Between the two extremes, cyclic processes (regeneration) and processes of directional change (succession) are arranged. they are indicative of changes over time in species populations, communities, and whole ecosystems, for example, on new substrates or after disturbances. While succession is one of the most fundamental and longest-studied concepts in plant community ecology [5], a comprehensive successional theory is still lacking [16,17]. The processes and patterns of succession and vegetation dynamics are common subjects in the field of biogeography, aimed at recording and explaining the distribution of species, population communities, and ecosystems, as well as their changes in space and time [18]. In times of substantial global changes, this topic is more relevant than ever, to better predict how ecosystems will respond to both land-use and climate change, and to improve the design of promising ecosystem restoration strategies [13,16,17]. This paper aims to review the existing concepts and models on types, trajectories, and processes recognized within vegetation dynamics [19,20], and to put the papers published in the MDPI Diversity Special Issue “Plant Succession and Vegetation Dynamics” into a contextual and conceptual framework. The Special Issue was intended to offer a platform to showcase recent findings in the study of plant succession and vegetation dynamics. with case studies from various plant communities, starting from different origins, covering different spatial and temporal scales, and employing different research methods, this SI provides a nice compilation of recent studies on a classical yet highly relevant object.

2. Views and Hypotheses on Vegetation Dynamics over Time

The early records of (bio)geographical observation date back to Greek scholars like Theoprastus and Aristotle in 300 BC (for a detailed overview on the science history of biogeography, see [18,21]). Perceptions of nature at that time differed widely, from a virtually static nature (equilibrium) to a dynamic one (non-equilibrium). During the Renaissance, European scientists started studying biogeographical foundations, which basically followed two fundamentally different biogeographical paths, a historical–evolutionary and an ecologically oriented one (Figure 1, [18]).

Two important representatives of the formerwere Charles Darwin (1809–1882) and Alfred Russel Wallace (1823–1913), whose insights through their exceptional journeys challenged the creationist view of the unchangeability of nature [22,23]. The recognition of cyclical climate fluctuations [24,25] and the drift of the lithospheric plates [26] subsequently underpinned their visionary views. While the historical–evolutionary biogeography made long-lasting evolutionary processes responsible for the species composition and distribution, main drivers within an ecologically oriented biogeography at that time were the prevailing environmental conditions. Introducing an ecological approach for explaining the distribution of species and communities, Alexander von Humboldt (1796–1859) is regarded as the founder of modern ecological biogeography ([27], Figure 1).

Based on these early biogeographical accounts, the concept of succession (from Latin, “succession” meaning progress) of plants was developed during the 20th century, despite the basic idea being introduced earlier by the French naturalist Adolphe Dureau de la Malle [28], who described the vegetation development after a forest clearcut. The term succession, however, was not yet used. Later, the Danish botanist Eugen Warming described the vegetation development on sand dunes on the coast of Denmark in his work Plantesamfund [29]. His interpretation considerably influenced two American contemporaries, Henry Cowles and Frederik Clements, who are generally regarded as the forefathers of the scientific concept of succession, although without any doubt Warming deserves that honor. In a broader understanding, all vegetation changes in a certain direction via various stages, and the underlying processes are referred to as successions.

Figure 1. The development of successional models with time. Fundamentals come from early accounts of biogeographical pioneers ([22,23,25,26,27,28,29], in blue); during the last century and a half, many different models evolved (each of these continues to develop) that might be arranged into four groups—succession of plants ([30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45] in green), succession of communities and ecosystems [5,46], (in yellow), succession in landscapes ([47,48], in orange), and succession with people ([49,50], in red), successively increasing in complexity. Date of milestone publications for the different groups of models are indicated by the pin including publication year and authors’ name. The arrows below the timeline indicate the domain, discipline, and time of development (adapted and modified from [16]).

Figure 1. The development of successional models with time. Fundamentals come from early accounts of biogeographical pioneers ([22,23,25,26,27,28,29], in blue); during the last century and a half, many different models evolved (each of these continues to develop) that might be arranged into four groups—succession of plants ([30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45] in green), succession of communities and ecosystems [5,46], (in yellow), succession in landscapes ([47,48], in orange), and succession with people ([49,50], in red), successively increasing in complexity. Date of milestone publications for the different groups of models are indicated by the pin including publication year and authors’ name. The arrows below the timeline indicate the domain, discipline, and time of development (adapted and modified from [16]).

Cowles interpreted the vegetation of differently aged sand dunes as stages of a directional vegetation sequence [30], so he can be credited to be the founder of the chronosequence approach or space for time substitution in the sense of Pickett’s [51], where a temporal sequence is derived from a spatial juxtaposition of different locations [52]. Clements instead became famous for his mono-climax concept [31,53], which ever since has been disputed. According to his conception, succession always develops towards a climate-controlled final plant association called climax, and colonization of open ground is taking place through the immigration of species from the surrounding area (primary succession) or due to a diaspore bank already present on the site (secondary succession). Schematically, he divided the controlling processes of succession into the following steps [9,31]:

• Nudation (disclosure of the substrate by disturbance);
• Migration (immigration of species);
• Ecesis (establishment);
• Competition;
• Reaction (location modification);
• Stabilization (establishment of a stable climax society).

Clements regarded the two factors of competition and site modification as central mechanisms for the evolving stages of succession. Clements’ holistic view was controversially discussed. In particular, the European School of Vegetation Science at that time doubted the existence of a single final plant association, because variable site conditions or different disturbances should result in different endpoints even under equal climatic conditions [20]. Consequently, the British plant ecologist Sir Arthur George Tansley attempted to address this dilemma by introducing the term polyclimax [32].

In the Clements’ “Relay Floristics” model, particular sets of plant species appear and disappear on a site during succession (Figure 2). Each set of species appears at a specific stage, making the site conditions unsuitable for itself but more suitable for the next set of species. While generally considering this to be possible, Egler [33] proposed the “Initial Floristic Composition” model. In this model, diaspores of all future species are already present at the beginning of a successional sequence but certain groups of species and/or lifeforms become dominant at certain stages during succession (Figure 2).

In the 1970s, Connell and Slatyer [34] presented their widely recognized and still influential hypothesis, trying to integrate all concepts of succession developed so far in three trajectories (Figure 3), with early settlers promoting, tolerating, or inhibiting the establishment of later successional species. The three successional pathways presented [34] were not considered exclusive [35,36], rather sub-elements of the three models are complementary and might explain the different phases of the succession or for different species.

Henry Gleason, a contemporary of Frederik Clements, vehemently challenged Clements’ perception of a climax. He proposed an individualistic view of the succession of plants [37] (Figure 1), assuming that the biotic and abiotic heterogeneity in space and time (i.e., non-equilibrium) makes biocenoses largely unpredictable. Plant communities would represent random conglomerates of plant species controlled by environmental conditions, arranged along ecological gradients. In addition, interspecific competition for resources as well as disturbances and the species pool of the surrounding play an important role in Gleason’s succession model.

The two early concepts of Clements and Gleason—though far apart in their basic assumptions—had a very stimulating effect on the development of succession theories. Initially treated as alternatives, they are considered complementary today [20]. In essence, the current doctrines on succession and vegetation dynamics are founded on opinions and concepts that have been developed in the aftermath of Clements and Gleason.

Although initially not much supported, Gleason’s ideas became more popular in the middle of the 20th century with an increasing focus on population ecology. Data-driven mathematical–statistical models replaced the “philosophical” views on succession and underpinned Gleason’s individualistic view of the arrangement of species along ecological gradients. Grime’s CSR model [38] is well known, which distinguishes three groups of plant strategies dominating in particular successional stages (Figure 4):

• Ruderals (R, mainly annual and perennial herbs) immediately colonize after disturbances, grow fast, reproduce rapidly, and have a short lifespan.
• Competitors (C, including various lifeforms from perennial herbs to trees) obtain limiting resources through fast horizontal or vertical expansion and growth, and are adapted to productive, little disturbed habitats.
• Stress tolerators (S, lichens, mosses, perennial plants) persist through slow growth and conservative resource use, and are adapted to unproductive, stressful habitats with little disturbance only.

This group of successional theories belong to the models of Noble and Slatyer [39], Tilman [40], and Huston and Smith [41], making “vital” life history attributes, plant traits, and below- (e.g., nitrogen, water) or aboveground resources (e.g., light) responsible for the (dis)appearance of species during succession, thus explaining the large variety of successional pathways observed in nature.

Still other theories on the succession of plants focus on structural changes during succession (Figure 1). In many vegetation types, a mosaic of patches in different successional stages is present, for example, after a disturbance or when the dominant plants become senescent and die [42]. Such cyclic successions are a characteristic feature of forests where small-scale disturbances like treefall gaps initiate patch development. Vegetation dynamics following large-scale disturbances such as clearcutting for agriculture and subsequent abandonment were described by Budowski [43] and Oliver [44] who recognized different successional stages dominated by particular functional groups and life history traits. Yarranton and Morrison [45] proposed the idea that shrubs and trees might serve as nuclei, as safe sites for regeneration, enabling other species to establish in a favorable environment (less solar radiation, cooler temperature, and more humidity, entrapped organic material, improved soil conditions) below their crowns. Due to parallels to the nucleation in physics where the formation of a new organized structure is initiated by interactions of pre-existing disordered components as in a self-assembling process, Yarranton and Morrison [45] called this process nucleation.

In the 1960s, the idea evolved that succession can be observed not only from a species or lifeform point of view, but also with regard to ecosystems (Figure 1). Maybe the first heading in this direction was Odum [46]. Changes in ecosystem characteristics (energetics, nutrient cycling, community structure, life history, selection pressure, and homeostasis) during succession were made responsible for ecosystem development over time. Temporal scales are also key to the concept of Walker and Wardle [5], where plant succession is explained by processes operating at different timescales. Short-term processes such as soil nutrient fluxes and plant physiology are believed to drive plant succession while long-term geological (parent material), evolutionary (species pool), and soil-forming processes (soil properties) operating on a timescale of thousands to millions of years limit succession, having a modifying effect.

More recently, different spatial scales are considered to influence succession ([47,48], Figure 1). A larger regional scale determines climate, topography, disturbance regime, and species pool, while a smaller landscape scale refers to spatially heterogeneous areas consisting of a mosaic of different land cover types (such as forests, agricultural areas, meadows, etc.), determining mesoclimate, seed availability, dispersal agents, etc. Spatial scales explain the large variation in successional pathways observed across and within landscapes.

The latest add-on in the study of succession and vegetation dynamics is the integration of humans as relevant factor, understanding the land cover on Earth as a socio-ecological system (Figure 1). Land-use intensity (duration, extent, type, use of chemicals, etc.) affects the local species pool, which is the seed source for colonizers and, in turn, affects the direction of successional pathways [49,50]. A high land-use intensity commonly results in a high number of plant species adapted to disturbance such as weeds, lianas, or invasive species. As natural regeneration would lead to an undesirable outcome, planting or at least a human-assisted regeneration should be carried out. Low land-use, in contrast, preserve more intact landscapes, allowing natural regeneration from more diverse and pristine seed sources.

3. Current Perspectives on Vegetation Dynamics

Vegetation dynamics cover a wide spatio-temporal range, from small-scale or annually recurring symphenological processes to large-scale or secular synchronological developments. Between the two extremes, short-to-long-term vegetation changes are found as a result of gradual site modifications and/or sudden disturbances that initiate succession or regeneration, but the transitions are usually gradual, and their different types can occur together, next to each other or one after the other (Figure 5).

3.1. Directional Succession

Succession as a directional sequence of different vegetation types or plant communities at a site [17,56] can be subdivided into progressive and regressive successions with regard to the direction of development (Figure 5). In both cases, different stages can be distinguished by species composition, vegetation structure, and/or stand physiognomy. Progressive successions are usually characterized by an increase in the number of species, the number of individuals, ground cover, increased structural complexity (e.g., lifeform diversity), and a replacement of earlier colonizers by succeeding species [57]. In regressive succession, changes occur backwards from a more advanced to an earlier stage of development [58]. Reverse developments are often caused by anthropogenic factors (e.g., overgrazing, lowering of the groundwater table, etc.), but also might be a result of natural events (drought, fire, etc.) or the gradual deterioration of the site (e.g., soil acidification caused by the vegetation itself).

Vegetation development on new terrain without any diaspores and soil formation (e.g., fresh lava, freshly piled sandbanks surrounded by the sea, new islands, recently deglaciated glacier forelands, etc.) are called primary succession ([59], Figure 5), characterized by the following main processes [60]:

• Getting there: diaspores must reach the new ground, which requires suitable plants to be nearby for dispersal.
• Establish: seeds must germinate and establish themselves; success might also depend on the ability of species to dormancy, the stress tolerance of the species, and the presence of “safe sites”.
• Growth: largely controlled by the site resources (nutrients, water, light, etc.).

Early colonizers are commonly short-lived, often annual taxa, which are fast-growing (“R-strategists” according to Grime ([38,61,62], Figure 4) and often wind- or self-pollinated species.

Directional vegetation developments on sites already covered by plants and/or possessing a seed bank are called secondary successions ([62], Figure 5) and proceed considerably faster than primary successions. It is a continuously ongoing process in undisturbed plant communities, affecting the dominant species composition, environmental characteristics, and plant community dynamics. This is shown by Duan et al. [63] in this Special Issue for Picea jezoensis forests in Northeast China. The quantification of changes in Picea jezoensis dominance status provides a scientific basis for evaluating the status and for the protection and restoration of Picea jezoensis forests. More obvious and somewhat transitional to regeneration dynamics are secondary successions after natural or anthropogenic disturbances and in (semi-)natural systems.

Primary succession is commonly slower than secondary succession, but proceeds more quickly, if locations adjoin existing vegetation from where diaspores can access the new ground via plants’ natural diaspore dispersal pathways. Differences between primary succession and secondary succession also exist in terms of species diversity. While in the former, the species diversity commonly increases with ongoing development, in the latter, the early stages are often significantly more species-rich than later shrub and forest formations [54]. While secondary succession is usually swift, there are exceptions as De Giuli et al. [64] report in this Special Issue. Even 80 years after the subalpine forest fire in the Northern European Alps, the fire affected slopes are still not wooded and show higher species numbers and clearly distinct floristic communities compared to unburned sites. This pattern is likely driven by disturbance-related environmental changes, such as increased light availability and highly eroded soils, preventing a fast return to the subalpine forest under this so-called arrested succession.

3.2. Cyclic Regeneration

Directional (acyclical) succession is contrasted with cyclical changes due to regularly occurring disturbances and site-internal modifications of environmental factors. Vegetation changes of endogenous or exogenous origin are summarized under the term regeneration (Figure 5). Endogenous causes are those that arise from the ecosystem itself (e.g., “mosaic cycles”, [65], including continuous self-reproduction in plant communities with a more or less constant species stock, or small-to-medium-sized “disturbance-like” events such as gaps by species’ die-off (Remmert (1991 [65], Figure 6). Sometimes large areas are affected by the so-called “cohort die-offs” if individuals in even-aged populations reach the end of their lifespan at the same time, as described for Metrosideros forests in Hawaii [66] or Nothofagus populations in Patagonia [67].

Equally important for the dynamics in plant communities are exogenous disturbances that affect plant communities from outside (see also Figure 5). As defined by White and Pickett [68], disturbances are discrete events in time that disrupt ecosystem, community, or population structure and change resources, substrate availability, or the physical environment. Due to their patchy nature, dynamic processes caused by the disturbances are called “patch dynamics”, independent of their dimension [9,68]. Garate-Quispe et al. [69] in this Special Issue describe the factors influencing the natural regeneration in abandoned small-scale gold mining areas in the Peruvian Amazon. The focus is on the influence of the distance to the forest edge and the time since abandonment for spontaneous natural succession. Basal area and species diversity here are related to time after abandonment and inversely related to the distance to forest edges. Here, variation in species composition is mainly explained by the distance to the forest edge rather than the abandonment time. In a subsequent contribution to this Special Issue [70] by the same team of researchers, aboveground biomass, forest structure, and species diversity during secondary succession were analyzed in the same study area, where two different types of goldmining (heavy machinery and suction pumping) were formerly used. The study shows that heavy machinery more negatively affects natural regeneration than suction pumping. Tropical secondary forests give important insights on how to support restoration and enhance conservation as Rodríguez-León et al. [71] show in this Special Issue for Colombian Amazon. The authors evaluated aboveground biomass, species diversity, forest structure, and soil properties in secondary forests along a chronosequence to identify key indicators for effective restoration management.

In order to structure the confusing variety of disturbances, Böhmer and Richter [55] propose a rough classification by three size classes (Figure 7), which are responsible for different regeneration models:

• Macro-disturbances such as wildfires or strong storms trigger mass extinctions comparable to endogenously triggered cohort extinctions.
• Meso-disturbances with individual trees or groups of trees dying/falling due to exogenous factors (gap dynamics sensu stricto).
• Micro-disturbances allowing an interim invasion of certain species due to the small-scale gaps caused by the death of individual, mostly non-tree species. While these species disappear in one location, they easily colonize newly created gaps (“carousel dynamics” [72]).

Moreno-Gonzalez et al. [73], in this Special Issue, describe the early vegetation recovery after a macro-disturbance by the 2008–2009 explosive eruption of the Chaitén Volcano in Chile. There was a gradient of decreasing disturbance with distance from the crater. High levels of devastation without surviving species, scarcely standing-dead trees and logs, and no tree regeneration were found close to the crater. Farther away from the crater, trees were resprouting, understory plants were regrowing, and seedlings were present, creating completely different starting points for regeneration/succession. Large-scale disturbances often also result from anthropogenic activities such as forest clearcutting. This is highly problematic if climate-resistant tree species such as Quercus robur are involved. Nasibullina et al. [74] report, in this Special Issue, that the natural regeneration of the oaks in the clearcut oak–lime forests of European Russia is successful, so planting oak seedlings or sowing acorns, i.e., active restoration, in combination with the natural restoration of lime trees are recommended to regain climate-resistant mixed forests here. Vegetation dynamics (and associated vegetation patterns such as species diversity) are sometimes controlled by co-varying disturbance and environmental gradients as shown by Abella et al. [75] in this Special Issue. In their case study, fire history (i.e., number of fires) and parent material (i.e., soil coarseness and chemistry) in combination were responsible for diversity patterns and plant community features including fuel, tree structure, and understory vegetation within the pinyon–juniper woodland of the Southwestern USA.

Last but not least, phytoglobalization [76] has become an important topic in the current ecological science. The highly dynamic spread of invasive alien plant species such as Impatiens glandulifera has become a serious threat to native, less-competitive taxa worldwide. Allelopathy has been suggested to be one of the possible mechanisms supporting the invasion of exotic plants. The study of Kupcinskiene et al. [77], in this Special Issue, about the allelopathic potential and phenolic content of I. glandulifera confirms the supreme competitive ability of this species over I. parviflora and native co-occurring species.

4. A Broad Range of Questions Needs a Broad Range of Methods

The processes of cyclic regeneration and directional succession play a major role for species composition and vegetation structure in plant communities. The controlling factors, however, are controversially discussed since the early accounts by pioneers of biological and biogeographical research in the 19th and 20th century. A better understanding of successional pathways, the processes involved, and vegetation dynamics, in general, is crucial in times of rapid environmental change, increasing the frequency and intensity of natural and man-made disturbances, as it has important implications for ecosystem restoration and overall vegetation management. Due to the broad array of dynamic processes and patterns of vegetation change in space and time involved, several different investigation methods are available [59]. The appropriate choice is eventually depending on the temporal and spatial scales of the study (see Figure 8), each with its respective pros and cons. Generally, a distinction is made between direct and indirect recording methods. Indirect methods include comparisons between the current state and the conditions documented in historical documents (records, photographs, paintings, etc.). Diaspore banks, datable macrofossil remains, and/or pollen analyses have been proven useful for a reconstruction over long periods of secular development. This method is used by O’Connell and Wolters in this Special Issue for the reconstruction of vegetation history and land use during the last 4000 years in Kilmore townland, Dingle peninsula, south-western Ireland in this Special Issue [78]. They found the main factor determining the direction and magnitude of change of woody vegetation and biodiversity to be farming activity.

The use of chronosequences (space for time substitution according to [51]), assuming a temporal sequence derived from spatially different sites, is another well-established indirect method. It is useful when the site age can be determined with sufficient accuracy, for example, in the case of lava flows, exogenous disturbance events (fire, storm, flooding, insect calamities, etc.), or in the case of retreating glaciers. The comparisons of current vegetation recordings with historical surveys of the same area (without precisely fixed long-term sample sites) allow for an accurate assessment of the degree and direction of the vegetation change. Such comparisons are transitional from the indirect to the direct methods. Direct methods include the comparisons of vegetation maps of the same areas but from different dates or repeated recordings of precisely localized and dated sample sites set up in the past (“Permanent Plots”), which allow a very precise evaluation of the course of succession.

The papers presented in this Special Issue give evidence of the broad range of spatial and temporal scales on which vegetation dynamics might be studied, regarding type, plant communities, geographical location, and investigation methods employed. We wish all readers an insightful reading and many new findings.