like Artemisia and Carduus species. On severely trampled sites Polygonum aviculare reaches high abundance. The opposite pathway of degradation with canopy closure and nutrient accumulation typically result in shrub encroachment. Besides the shrubs, broad-leaved species of loess steppes such as Thalictrum minus, Salvia pratensis and Centaurea sadleriana or ruderal species such as Calamagrostis epigeios, Cardaria draba, Salvia verticillata, Cynoglossum officinale or even Urtica dioica can proliferate. In case of shrub encroach- ment, it is questioned whether it is a degradation process of the loess steppe, or the opposite, regen- eration towards a forest. In fact, nutrient accumu- lation always bears a risk, since it enables invasive alien species (for example Ailathus altissima, As- clepias syriaca or Solidago spp.) to colonize and forage for the surplus resources.
It is important to emphasize that the history of a given stand is always complex, the actual out- look is always the result of several different, some- times opposing factors. In the case of grazing, the biomass decreases, the leftovers of the herd, how-
ever, give some extra nutrient supply. As grazing intensity typically varies over time, with periods of overgrazing and undergrazing, the conse- quences are layered upon one another. Overgraz- ing, though, can lead not only to erosion, but also to the proliferation of nitrophilous weeds .
Degradation not necessarily means total de- struction and disappearance of steppes. Since natural grasslands are evolutionary adapted to natural disturbances, they possess the ability of regeneration. Regeneration potential depends on site conditions (exposition, steepness, soil type) and on the species pool of the surrounding landscape. According to our present knowledge, landscape context is the major determinant of the speed of regeneration processes. According to studies by Zsolt Molnár secondary succession is retarded in the great Hungarian Plain close to the village Pitvaros, where the loess flora has declined in the last few centuries. Due to the lack of prop- agulum sources of loess steppe specialists in the neighbourhood of oldfields, regeneration reach- es only the phase dominated by generalists and does not go further. The opposite was found by Eszter Ruprecht in the Transsylvanian Mezőség (Romania) where almost all characteristic species of loess steppes were present in oldfields 40 years after the abandonment. However, significant correlations were revealed between the amount of (semi-)natural grasslands around the studied oldfields within a 500 meter radius circle and the success of regeneration. Thus, the amount of available propagules is a limiting factor even in that more favorable landscape. Changes in land use, ploughing of the verges and collectiviza- tion have lead to severe decrease in the regen- eration potential of steppes. In some cases, when the (semi)natural stands serving as propagule sources for regeneration processes have already disappeared from the landscape, succession has stopped at a stage dominated by ruderals only.
Regeneration processes change according to the scale of disturbance. As an example we present the regeneration process of a loess steppe near the village Albertirsa (Fig xxx). Grazing causes dis- turbances similar in size to the individuals or the ramets of steppe plants. As a consequence , the grassland opens up, it becomes drier, the domi- nant Festuca rupicola becomes suppressed and
the formerly subordinate Bothriochloa ischae- mum becomes dominant. When grazing ceases, the canopy becomes closed again and the grass- land regenerates into a Festuca rupicola domi- nated form. If the extent of disturbance is larger, reaches few square meters, the vegetatively well- spreading clonal forbs, such as Hieracium, Trifo- lium, Astragalus or Fragaria species spread first into the gap and Bothriochloa ischaemum reaches high abundance. If the extent of the disturbance is further increased, the annual grass species Bro- mus arvensis and Setaria species or occasionally Poa pratensis s.l. colonize the gap first, followed by Festuca rupicola and Bothriochloa ischaemum. Regeneration of the oldfields follows the same sequence, although there is another phase domi- nated by Elymus repens inserted into the series. (Figure xxx.
4ab. ábra)
The early stage of secondary succession on abandoned fields are profoundly determined by the last crop and the related management. On abandoned cereal fields, Consolida regalis, Pa- paver rhoeas, Tripleurospermum inodorum and Elymus repens are abundant in the first years. After the abandonment of row crops (corn for
example) Amarathus retroflexus, Chenopodium species, Cirsium arvense and Elymus repens estab-
lish first. Later (in ca. 5-10 years) the successional pathways converge to a stage typically dominated by Elymus repens accompanied by ruderal spe- cies like Cichorium inthybus, Plantago lanceolata, and Convulvulus arvensis. The abundance of dry grasslands species, such as Koeleria cristata, Medicago lupulina, Hieracium spp., Salvia nemo- rosa, Leontodon hispidus, Dorycnium herbaceum, Lotus corniculatus increase gradually over time. Poa pratensis is often a co-dominant with Elymus repens, while later successional stages are char- acterized by a patchwork of Bothriochloa ischae- mum and Festuca rupicola.
Succession processes of abandoned vineyards differ significantly from those of oldfields (Fig- ure xxx 4.b). The slope of the stand is usually steeper, more eroded and has lower nutrient con- tent. Consequently, the early successional stage of segetal weeds is less pronounced. After 2-8 years Erigeron annus and Calamagrostis epigeios become dominant. Verges among the parcels act as refugia for steppe species, thus re-colonization is quicker. Species such as Dorycnium herbaceum, Bothriochloa ischaemum, Chrysopogon gryllus and Stipa spp. are typical in the intermediate suc- cessional stages followed later by Festuca rupicola dominance. Festuca Festuca Festuca Festuca Bromus Bromus Poa Elymus növekvő zavarás Bothriochloa Bothriochloa Bothriochloa egyéves gyomok (1–3 év )
Évelő füvek és társulásaik ( –20 év) tarackos füvek és kétszikűek foltosan ( –10 év)
erdősödés visszaalakul az eredeti társulás
száraz természetes gyepek másodlagos változatai
a)
b)
Fig. 13 a) An example regeneration dynamics of a loess valley of the Great Plain near Albertirsa. To depict a clear view, only the most
important grass species are indicated. In case of the most serious disturbance (oldfield succession) the annual Bromus arvensis colonizes at the earliest stage in the succession. It is followed by Elymus repens and Poa pratensis, which are replaced by Both- riochloa ischaemum in the next step. The last colonizer is Festuca rupicola. Should the disturbance be less serious, regeneration is quicker, and less grass species has enough time to develop a distinct successional stage on their own.
b) Regeneration of abandoned vineyards in the Medium Mountains was described by Zoltán Baráth, as early as in the 1960’s. It is worth taking notice of the similarity between the recently describedsuccessional sequence of oldfields on the Plain (annual grass, stoloniferous grass, perennial grass, natural grass species) and that of abandoned vineyards described 40 years earlier.
For the success of regeneration, not only prop- agule sources are important, but also the state of the receiving vegetation. It is a well-known fact that the rate of immigration decreases exponen- tially over time. Supposing that landscape scale species availability is constant in the timescale of succession, species should be able to colonize much easier at the beginning of the succession. There are so called “succession windows” in the
course of the succession, when the vegetation is particularily susceptible to the establishment of new species. Several grasses form very dense patches after 2-3 years of their colonization with high competitive pressure to the subordinates, but later this matrix of dominant grasses opens up again, mainly due to architectural changes of the growth form of these clonal grass spe- cies. The detailed mechanism is not known so far, although , the negative feedback caused by accumulating litter probably plays an important role in the process. Colonization and survival of subordinate species depends on the characteris- tics of the matrix of the dominant grass species. The matrix is affected by factors like weather, fire, grazing and trampling. Multiple interactions and indirect effects are common among these factors. Accumulating litter for example can suppress
subordinate species in the short run, while in the long run it suppresses the dominant grasses and thus indirectly favors the subordinates. Fire has a positive effect on subordinates in the beginning by eliminating litter and mobilizing nutrients, while its long-term effect is opposite by favoring dominant grasses. The same can be said about grazing. Grazing opens the grass matrix and thus favors the colonization of other species, mainly forbs. However, if the grazing animal prefers special forbs, it will lead again to the spread of the dominant grass. The presence and timing of a “succession window” is an important question from the point of shrub invasion as well. If prop- agules are present at the beginning of the succes- sion, shrubs can easily colonize and closed shrub- lands can develop within 10-15 years. However, if grasses form a dense matrix in the beginning, shrub establishment will fail.
The time span of regeneration varies signifi- cantly. Regeneration processes close to equilib- rium usually last not longer than 5-15 years. On
the contrary, successional dynamics take decades or more than 100 years in a degraded landscape. If the propagules of late successional species are missing, succession processes might be retarded even for centuries.
How many different regeneration and degra- dation pathways can be distinguished? The an- swer depends on the type and intensity of the disturbance. Fine-scale disturbances induce a stochastic manifold of micro-processes without visible changes at stand level. The larger the grain and the more intense disturbances are (i.e., the further we are from the equilibrium), the smaller the number of potential states and transitions is. Far from the equilibrium, in ruderal phases wide- spread species dominate (e.g. Elymus repens and Poa pratensis s.l.) resulting in similar successional stages even at different points of the globe (for example in Central Europe and North America). Vegetation patterns are most diverse if distur-
bances occur with intermediate frequency and in the intermediatephase of regeneration, when stands are differentiating into dominance types. Although, patches can be classified according to the dominant species or a few abundant spe- cies , we have to be careful, since the direction and speed of species turnover is often defined by the neglected subordinate species and not by the studied dominants. The compositional variability of these dominance types is usually high. Differ- ent Bothriochloa ischaemum dominated stands can have considerably different species combina- tions with different behaviour in different regions of the country or in different positions within the same loess valley. It can be important to know the state of a vegetation stand in detail because the degradation process can cross a critical thresh- old where ruderalization starts. By crossing this threshold the equilibrium dynamics change to non-equilibrium ones, where community regula- tion breaks down and ecosystem services become much less effective. Aggressive spread of invasive alien species is another challenge. Since their be- haviour is unknown or unpredictable, they can cause unexpected problems. From the point of view of the nature conservation, the study of the local characteristics of these invasive alien spe- cies is one of the most urgent and most important tasks of our days.