Fire and succession in the ultramafic maquis of New Caledonia

Aim This study investigates the role of fire and post fire succession in determining the structure and composition of vegetation on ultramafic iron crust soils. Location The study was conducted in the Plaines des Lacs region of southern New Caledonia. Methods A survey was made of eighty-eight sites, recording floristic composition, trunk size-class distributions, regeneration after fire, growth ring counts of Dacrydium araucarioides (Podocarpaceae) and historical information on past fires. Floristic data was ordinated using multidimensional scaling and an index of succession based on structural and historical information. A transition matrix model was developed to predict the effect of fire frequency on vegetation composition. Results The vegetation is undergoing postfire succession from maquis to forest, after about 75 years, and eventually to rainforest. Gymnostoma deplancheanum has a key role as an early colonist that produces shade, the bulk of the litter, and forms nitrogen fixing nodules with Frankia sp. However, the open canopy of Gymnostoma and slow litter decay creates flammable conditions. Though many species resprout from rootstocks, only thirty-nine persist through fires while 114 others colonize at later successional stages, as the litter layer and shade increase. Some early successional species are later excluded but these can persist locally in swamps and on rocky hill tops. Forest and rainforest are less flammable and the matrix model suggests that ignition frequency has a critical role in determining the abundance of maquis or forest. Main conclusions The vegetation mosaic represents a post fire succession from open maquis to forest. Palynological and charcoal records from late Pleistocene sediments suggest that fire has been a major factor determining the development of maquis vegetation since before the arrival of humans. Recently, frequent fires have converted much of the vegetation to maquis, posing a threat to some forest species and largely eliminating rainforest from iron crust soils.     

Phyllostomid bat assemblages in different successional stages of tropical rain forest in Chiapas,Mexico

Due to their role in seed dispersal, changes in the community of phyllostomid bats have direct consequences on ecological succession. The objective of this work was to document changes in the structure of bat assemblages among secondary successional stages of tropical rain forest in Chiapas, Mexico. Bats were mist-netted at ground level during 18 months in 10 sites belonging to 3 successional stages: four sites represented early succession (2–8 years of abandonment), four intermediate succession (10–20 years of abandonment), and two late succession (mature old-growth forest).We captured 1,179 phyllostomids comprising 29 species. Phyllostomid species richness was 17 (58% of all species) in the early stage, 18 (62%) in the intermediate stage and 24 (83%) in the late stage. The late successional mature forest possessed nine species that were exclusively found there, whereas early and intermediate successional stages contained only one exclusive species. Sturnira lilium, Artibeus lituratus, Carollia perpicillata, Artibeus jamaicensis and Glossophaga soricina represented 88% of all captured phyllostomid bats. Frugivores made up more than 90% of the species captured in early and intermediate successional stages and 84% in late successional forest. The Bray–Curtis index of dissimilarity showed a replacement of species through successional stages with the largest dissimilarity between early and late stages, followed by intermediate and late, and the lowest dissimilarity between early and intermediate stages. The number of gleaning insectivore species increased during succession. The carnivorous guild was exclusively found in the late stage (three species). We conclude that the late successional mature forest was the main reservoir for the gleaning insectivore and carnivore guilds; however, early and intermediate successional stages possessed a great diversity of species including many frugivores.     

Plant Recruitment in Early Development Stages on Rehabilitated Quarries in Hong Kong

To understand vegetation development during the ecological succession of rehabilitated quarries, floristic composition and structure were evaluated at different restoration phases on three quarries in Hong Kong that were planted with exotic woody species over the course of 2–14 years. A total of 113 species, of which 82 were woody species, were recorded. Exotic species dominated the overstory, and species number, richness, and diversity increased with age. Some light‐demanding early successional species were becoming dominant in the overstory vegetation at the older phases of revegetation. These species could be potential candidates for early enrichment planting with Acacia spp. Common secondary forest species occurred naturally in the understory vegetation, and were more abundant and dominant after 10 years of ecological development. The most successful restorations were on scree slopes using leguminous Acacia spp. as nurse species.     

Changes in the vegetation of Quercus pubescens woodland after cessation of coppicing and grazing

Abstract. The cessation of coppicing and grazing in Quercus woodlands, with its subsequent changes in the structure, composition and functioning of vegetation communities, is becoming more frequent throughout the Mediterranean Basin. In southern France, we have studied successional changes in Quercus pubescens woodlands by visiting previously studied sites 18 yr later. Changes in vertical structure, species richness, floristic composition, life form and dispersal type were analysed and compared between woodlands that had previously been grazed or ungrazed. Both successions showed a decline in vegetation cover in the 0–25 cm height class and in the height class immediately under the canopy layer, due to oak litter accumulation and tree ageing. In post‐grazing succession, the abandonment of grazing and associated burning has allowed the vegetation cover to increase in the 0.25‐2 m height class. In both successions, grassland species decreased in frequency and forest species increased, a trend which was stronger in undisturbed succession. Species richness decreased with time in the undisturbed succession, but remained stable in the post‐grazing succession mainly because of the slow decline of plants linked to grazing. In undisturbed succession, therophytes and hemicryptophytes decreased. In contrast, therophytes and hemicryptophytes remained stable in post‐grazing succession. In both successions, endozoochorous species (notably Ruscus aculeatus) increased. Plants dispersed by non‐animal vectors decreased in undisturbed succession, but stability was observed in most of the dispersal types in post‐grazing succession. These results showed that a time‐lag existed between undisturbed succession and post‐grazing succession, the latter remaining at a younger stage of successional development due to more recent impact of grazing. However, both successions have converged suggesting that most of the traces of grazing on vegetation will disappear within a few years. The vegetation of these coppices, regardless of the previous grazing regime, will become increasingly similar to the vegetation of undisturbed woodlands. However, their floristic composition will probably never be identical to that of undisturbed woodlands, mainly because of the rarity of these undisturbed woodlands and of the short‐distance dispersal of many forest plant species.     

Is microclimate-driven turnover of dung beetle assemblage structure in regenerating coastal vegetation a precursor to re-establishment of a forest fauna?

We questioned the capability of post-mining rehabilitation and successional changes in coastal vegetation to achieve restoration of dune forest, dung beetle assemblages in the Maputaland Centre of Endemism, South Africa. A repeat 2010 study of structural turnover between dung beetle assemblages across a 33 year successional sere of rehabilitating vegetation and old-growth forest (>73 years) produced comparable results to an earlier study across the 23 year chronosequence of 2000. Despite overlap, three structural patterns along the 33 year chronosequence were associated with specific stages of vegetation succession and their characteristic microclimates as in 2000. Although species biased to unshaded habitat dominated the earliest succession, there was rapid re-establishment of dominance by shade-associated forest species. In concert with progression from unshaded, post-mining vegetation to strongly shaded, early successional, Acacia shrub-woodland, there was an initial increase in similarity of the dung beetle fauna (species-poor, low abundance) to that in strongly-shaded forest (also species-poor, low abundance). However, in concert with decreasing shade cover in late successional woodland, the dung beetle fauna became species-rich with high abundance so that the early successional trajectory of increasing similarity to forest fauna either levelled off to a plateau (species in 2000; abundance in 2010) or declined (species in 2010, abundance in 2000). It remains to be seen if gaps forming in the oldest Acacia woodland permit forest tree saplings of the exposed understorey to recreate a forest canopy that would be tracked by dung beetles to re-establish a typically species-poor, deep shade, forest assemblage with low abundance.     

Interactions with successional stage and nutrient status determines the life‐form‐specific effects of increased soil temperature on boreal forest floor vegetation

The boreal forest is one of the largest terrestrial biomes and plays a key role for the global carbon balance and climate. The forest floor vegetation has a strong influence on the carbon and nitrogen cycles of the forests and is sensitive to changes in temperature conditions and nutrient availability. Additionally, the effects of climate warming on forest floor vegetation have been suggested to be moderated by the tree layer. Data on the effects of soil warming on forest floor vegetation from the boreal forest are, however, very scarce. We studied the effects on the forest floor vegetation in a long‐term (18 years) soil warming and fertilization experiment in a Norway spruce stand in northern Sweden. During the first 9 years, warming favored early successional species such as grasses and forbs at the expense of dwarf shrubs and bryophytes in unfertilized stands, while the effects were smaller after fertilization. Hence, warming led to significant changes in species composition and an increase in species richness in the open canopy nutrient limited forest. After another 9 years of warming and increasing tree canopy closure, most of the initial effects had ceased, indicating an interaction between forest succession and warming. The only remaining effect of warming was on the abundance of bryophytes, which contrary to the initial phase was strongly favored by warming. We propose that the suggested moderating effects of the tree layer are specific to plant life‐form and conclude that the successional phase of the forest may have a considerable impact on the effects of climate change on forest floor vegetation and its feedback effects on the carbon and nitrogen cycles, and thus on the climate.     

Secondary Succession and Indigenous Management in Semideciduous Forest Fallows of the Amazon Basin1

To the discussion on secondary succession in tropical forests, we bring data on three under‐addressed issues: understory as well as overstory changes, continuous as opposed to phase changes, and integration of forest succession with indigenous fallow management and plant uses. Changes in vegetation structure and species composition were analyzed in secondary forests following swidden agriculture in a semideciduous forest of Bolivian lowlands. Twenty‐eight fallows, stratified by four successional stages (early = 1–5 yr, intermediate = 6–10 yr, advanced = 12–20 yr, and older = 22–36 yr), and ten stands of mature forests were sampled. The overstory (plants ≥5 cm diameter at breast height [DBH]) was sampled using a 20 × 50 m plot and the understory (plants <5 cm DBH) in three nested 2 × 5 m subplots. Semistructured interviews provided information on fallow management. Canopy height, basal area, and liana density of the overstory increased with secondary forest age. The early stage had the lowest species density and diversity in the overstory, but the highest diversity in the understory. Species composition and abundance differentiated mature forests and early successional stage from other successional stages; however, species showed individualistic responses across the temporal gradient. A total of 123 of 280 species were useful with edible, medicinal, and construction plants being the most abundant for both over‐ and understories. Most of Los Gwarayo preferred mature forests for making new swidden, while fallows were valuable for crops, useful species, and regenerating timber species.     

Succession following the catastrophic eruption of Ksudach volcano (Kamchatka, 1907)

Ksudach Volcano, southern Kamchatka Peninsula, erupted in 1907 and impacted over 2000 km² of forests with air-fall pumice deposits. We identified three impact zones. In Zone I, deposits deeper than 100 cm destroyed all vegetation. Two early successional stages occur, a lichen-dominated desert and isolated patches of a pioneer herb stage. Zone II is defined by pumice deposits 30 to 100 cm deep. Deposits of 70 to 100 cm destroyed all vegetation, but left scattered snags. Here primary succession dominates recovery, but its rate varies. Isolated trees survived in deposits of 30 to 70 cm and primary and secondary successional stages form a complex mosaic termed an intermediate succession. In Zone II, the primary stages found in Zone I are joined by a dwarf shrub-herb stage and a secondary birch forest stage. Zone III occurs where thinner deposits permitted some vegetation to survive in all locations. Secondary succession dominates in deposits of 10 to 30 cm. Trees suffered damage, but survived deposits of 20 to 30 cm, while other vegetation layers were eliminated. Deposits of 10 to 20 cm eliminated mosses and lichens and but only reduced the number of dwarf shrubs and herbs. Deposits of less than 10 cm damaged herb, moss and lichen layers but did not eliminate any species. All sampled vegetation remains in a pre-climax state, having yet to recover fully from earlier eruptions. Reconstructed vegetation maps for before 1907 and for ca. 1925 are compared to the map of vegetation in 1994. Based on degree of soil formation, vegetation recovery and colonization rates at different pumice depths, and the current vegetation, we estimate that full recovery of the soil-vegetation system will take more than 2000 years.     

Pattern analysis in successional communities – An approach for studying shifts in ecological interactions

Abstract. Many ecological studies have addressed issues of vegetation spatial patterns in attempts to understand the processes generating them. We investigated changes in ecological processes during succession via the analysis of shrubs’ spatial patterns in a system of linear sand dunes, an arid ecosystem located in the Negev Desert in Israel during three consecutive years. We hypothesized that spatial patterns change from clustered to regular as succession progresses due to changes in the relative importance of facilitation and competition in this environment. In this ecosystem communities of early successional stages are frequently disturbed by high rates of sand movement, whereas in later successional stages sand stability is high. We mapped in the field individual shrubs on high‐resolution aerial photographs, and converted the digital images to a GIS data set. Using Ripley's K‐function we analysed spatial patterns at three levels: the single‐species level, among species and at the individual level, in three communities characterizing different successional stages. In the early successional communities we found clustered spatial patterns, in comparison with stable habitats where spatial patterns tended to be regular. We argue that these shifts in spatial patterns are indicative of the assumption that in this sand‐dune system ecological interactions change from facilitation to competition as succession progresses. Further, we argue that these interactions operate in different spatial scales at the different successional stages, and that the study of these processes should be conducted at the spatial scales specific to each community.     

Patterns and mechanisms of plant succession after fire on Artemisia-grass sites in southeastern Idaho

Cover data for plant species on eight environmentally similar sites that were each burned in a different year (from 2 to 36 years ago) were used to construct a composite sequence of vegetational change after fire on Artemisia-grassland sites in southeastern Idaho. Some species were early successional such as Lithospermum ruderale, and some late successional: Artemisia tridentata, A. tripartita, and Gutierreza sarothrae. But many species: Purshia tridentata, Symphoricarpos oreophilus, Amelanchier alnifolia, Chrysothamnus viscidiflorus, Achillea millefolium, Agropyron dasystachyum, and A. spicatum were present in both early and later stages. Shannon and Simpson indices of diversity and species richness indicated little change in alpha diversity through time. This was attributed mainly to the limited change in species composition from early to later stages. The general pattern of succession is compatible with the tolerance model of Connell & Slatyer (1977) in most respects. Species traits relating to persistence through a disturbance or re-establishment on the site, and tolerance of competition shape the course of succession on a site. Perennial grasses and forbs which sprout from the base after fire are the first species to dominate the sites. Sprouting shrubs, which require some years to regrow to their pre-fire form, are prominent by the sixth year. Shrubs which rely on dispersal become co-dominants in later stages, at which time some herbaceous species are reduced oreliminated. The pattern of succession can differ due to presence or absence of species with particular traits.     

Early seral trends in plant community diversity on a recontoured surface mine

The initial floristic composition and structure of vegetation on a recontoured surface mine in southwestern Wyoming was experimentally manipulated to test ideas concerning successional convergence. Experimental treatments involving the planting pattern and density of shrubs and the storage of topsoil were evaluated on the basis of their ability to divert the successional trajectories of the redeveloping vegetation and to increase species diversity. The change in compositional similarity along a topographic gradient was also monitored.The influence of initial composition and soil storage generally declined with time; successional convergence among treatments over the four years of the study was largely monotonic. The planting of shrubs and the storage of topsoil did, however, influence the rate of plant succession. There were no persistent differences in alpha diversity among the experimental treatments. Compositional dissimilarity along the topographic gradient, as expressed by beta diversity, increased with time.     

Secondary forest succession differs through naturalised gorse and native kānuka near Wellington and Nelson

The dominant native woody species forming early-successional vegetation on formerly forested sites in lowland New Zealand were kānuka (Kunzea ericoides) and mānuka (Leptospermum scoparium) (Myrtaceae). These have been replaced extensively by gorse (Ulex europaeus, Fabaceae), a naturalised species in New Zealand. Because gorse typically gives way to native broadleaved (angiosperm) forest in about 30 years, it is often considered desirable for facilitating native forest restoration. We tested three hypotheses, derived from the New Zealand literature, on gorse and kānuka: (1) kānuka stands have a different species composition and greater species richness than gorse stands at comparable successional stages; (2) differences between gorse and kānuka stands do not lessen over time; and (3) several native plant taxa are absent from or less common in gorse than in kānuka stands. We sampled 48 scrub or low-forest sites in two regions, Wellington and Nelson. Sites were classified into one of four predefined categories – young gorse, young kānuka, old gorse, old kānuka – based on canopy height of the succession and the dominant early-successional woody species. Few characteristics of the sites and surrounding landscapes differed significantly among site categories, and none consistently across regions. The vegetation composition of gorse and kānuka and their immediate successors differed in both regions, mainly in native woody species. Species richness was often lower in gorse and there were fewer smallleaved shrubs and orchids in gorse. Persistent differences at the older sites suggest the successional trajectories will not converge in the immediate future; gorse leads to different forest from that developed through kānuka. Gorse-dominated succession is therefore not a direct substitute for native successions. We suggest areas of early native succession should be preserved, and initiated in landscapes where successions are dominated by gorse or other naturalised shrubs.     

Primary succession on lava domes on Terceira (Azores)

Questions: Are the vegetation attributes significantly different among lava domes and among geomorphologic units as a result of age and soil features? Are the successional rates equal in all the geomorphologic units of the domes? Are the colonizer species of lava domes totally replaced by other species in the late successional stages? Location: Terceira Island, Azores (Portugal). Methods: Three comparable domes of 240, 370 and 2080 yr old were selected. Data on floristic composition, vegetation bioarea (area occupied by plant species in transects), structure, demography and soil nutrients were collected. Quantitative and qualitative changes along the succession gradient were also analyzed. Results: Vegetation attributes were consistent with the successional stage of each dome in the primary sere; however, the geomorphologic units did not follow the same pattern. The influence of the rates of plant colonization and soil formation are responsible for the decrease of the successional rates from footslopes, to summits, to slopes. The vegetation successionally changes from Juniperus scrub, to Juniperus wood and forest, and there is little species replacement since the similarity in species composition is high between the 3 domes. Conclusions: This is a special type of direct succession that takes place mainly through an increase in biomass and structural complexity. We observed small wooded areas in the fissures of very young domes that are samples of later successional forests — a phenomenon that we call ‘zoom effect’.     

Impact of a hurricane on the herpetofaunal assemblages of a successional chronosequence in a tropical dry forest

Land‐use change is the main cause of deforestation and degradation of tropical forest in Mexico. Frequently, these lands are abandoned leading to a mosaic of natural vegetation in secondary succession. Further degradation of the natural vegetation in these lands could be exacerbated by stochastic catastrophic events such as hurricanes. Information on the impact of human disturbance parallel to natural disturbance has not yet been evaluated for faunal assemblages in tropical dry forests. To evaluate the response of herpetofaunal assemblages to the interaction of human and natural disturbances, we used information of pre‐ and post‐hurricane herpetofaunal assemblages inhabiting different successional stages (pasture, early forest, young forest, intermediate forest, and old growth forest) of dry forest. Herpetofaunal assemblages were surveyed in all successional stages two years before and two years after the hurricane Jova that hit the Pacific Coast of Mexico on October 2011. We registered 4093 individuals of 61 species. Overall, there were only slight effects of successional stage, hurricane Jova or the interaction between them on abundance, observed species richness and diversity of the herpetofauna. However, we found marked changes in estimated richness and composition of frogs, lizards, and snakes among successional stages in response to hurricane Jova. Modifications in vegetation structure as result of hurricane pass promoted particular changes in each successional stage and taxonomic group (anurans, lizards, and snakes). Secondary forests at different stages of succession may attenuate the negative effects of an intense, short‐duration, and low‐frequency natural disturbance such as hurricane Jova on successional herpetofaunal trajectories and species turnover.     

Succession changes in diversity and assemblages composition of planthoppers and leafhoppers in natural ancient peat bogs in Belarus

Succession has a strong influence on species diversity and composition of terrestrial ecosystems. Peat bogs are among them. They have a large area in Belarus compared to other Central European countries. While in several studies have analyzed the effects of succession on vegetation in peat bog ecosystems, the response of peatland insects to succession has not been investigated yet. To address this issue were sampled Auchenorrhyncha abundance and environmental parameters on the ancient and one of the largest natural peat bog along a successional gradient from the margin to the bog dome. The results provide evidence that succession of peat bogs has influence on planthoppers and leafhoppers abundance, diversity and species composition. Along the successional gradient from younger towards older successional stages an increase abundance of specialized peat bog species, chamebionts, oligophagous and monophagous was observed. On the contrary, the younger stages of natural peat bog succession offer favorable conditions to eurytopic, polyphagous and chortobiont planthoppers and leafhoppers. The highest abundance and species richness of Auchenorrhyncha were in the lagg zone followed by early stages of natural peat bog succession. The highest diversity was in the middle stages of succession. A determinant of Auchenorrhyncha diversity was the cover of ericaceous dwarf shrubs. Linear models shrub cover and number of plants species had a positive effect on planthoppers and leafhoppers diversity and a negative effect on their abundance. Amount of ericaceous dwarf shrubs within the peat bog could be as a measure of heterogeneity.     

亚热带常绿阔叶林不同演替阶段土壤种子库的初步研究

 本文把四川缙云山亚热带常绿阔叶林的演替系列分成6个阶段：(1)弃耕地;(2)灌草丛;(3)马尾松林;(4)以马尾松为优势的常绿针、阔叶混交林;(5)以常绿阔叶树为优势的针、阔叶混交林;(6)常绿阔叶林。分别从每个阶段分3层采集0.5m2,l0cm厚度的5个小样地的土样,带回实验室,通过萌发法研究土壤种子库。得到如下结论：A、种子数量和物种多样性在演替早期阶段(从弃耕地到马尾松林)一般随演替发展而减少,种子数目随土层加深而降低;后期阶段(马尾松林到常绿阔叶林)种子数则大致随演替发展而增加;同一阶段各土层种子数差异不大。B、各演替阶段土壤种子库的种类组成均以草本植物为主。C、在湿度和光照比较一致的情况下,萌发种子数与温度成正相关。大多数种类的种子80%以上在实验的头六个星期内萌发。D、演替早期阶段的地上植物种类组成与土壤种子库中种类组成基本一致,演替后期阶段相关甚少。本文还讨论了土壤种子库研究的取样体积问题,以及土壤种子库在植被动态中的作用。     

Colonization and early succession on anthropogenic soils

Abstract. This paper reports on vegetation development on permanent experimental plots during five years of succession. Nine (1 m²) plots were filled with three typical substrates from man-made habitats of urban and industrial areas in the region of Berlin. The three substrates (a commercial ‘topsoil’, a ruderal ‘landfill’ soil and a sandy soil), differ in organic matter and nutrient contents. Relevés of species composition and percent cover of each species present were made monthly during the growing season from the start of vegetation development. This paper describes the different successional pathways on topsoil and ruderal soil and the colonization process on sandy soil. On topsoil, ruderal annuals are dominant in the first year and are replaced by short-lived perennials from the second year. Those species were replaced by long-lived perennial herbs (Ballota nigra or Urtica dioica) from the third year of succession onwards. On the ruderal land-fill soil the early successional stages are less sharp and the perennial Solidago canadensis is able to dominate within one year after the succession was initiated. On sandy soil there is still an ongoing colonization process, where pioneer tree species like Betula pendula and Populus nigra play a main role. The importance of ‘initial floristic composition’, the role of substrate for community structure and the peculiarities of successional sequences on anthropogenic soils in the context of primary and secondary successions are discussed.     

Temporal trends in species composition and plant traits in natural grasslands of Uruguay

Abstract. We report the successional trends of the major life‐forms (graminoids and forbs) in natural grasslands of Uruguay over a 9‐yr period after the removal of domestic herbivores. For the whole community, species richness and diversity decreased over the successional period. In graminoids we observed clear temporal trajectories in floristic composition; the rate of floristic change decreased with time and was associated with a shift in plant traits. The exclusion of large herbivores promoted erect and tall grasses with narrow leaves and greater seed length, vegetative growth constrained to the cool season and increased frequency of annual species. Forbs did not show a clear temporal trend in species composition, but there was, nevertheless, a plot‐specific species turnover of this functional group that was reflected in their attributes. Species spreading by means of rhizomes, with vegetative growth restricted to the warm season. Species with larger seeds increased under grazing exclusion, as did annual and nitrogen‐fixing forbs. The floristic changes induced by cattle exclusion occurred early in the succession. This early high rate of change has practical implications for management and conservation programs of the natural grasslands of Uruguay. Additionally, the shift in plant traits may be helpful in devising simple indicators of grazing impact.     

Can Establishment Characteristics Explain the Poor Colonization Success of Late Successional Grassland Species on Ex‐Arable Land?

Abstract Many grassland restoration projects on former arable land face problems because early successional grassland species establish vigorously and persistently but late successional grassland species fail to establish. Differences in establishment characteristics of early and late successional species might provide an explanation for the failure of many late successional species to colonize grasslands on ex‐arable land. I examined whether early and late successional species had different establishment rates in the initial years of a grassland succession, whether a specific establishment stage (seedling emergence, mortality or growth) could be identified as the key process controlling establishment, and what management would enhance the establishment of late successional grassland species. Seeds of three early and three late successional species were sown separately in ex‐arable plots with bare soil, 1‐year‐old vegetation, and 2‐year‐old vegetation. Emergence, mortality, and seedling growth were monitored for 1 year. Early successional species established successfully in the bare soil plots but failed to establish in plots with 1‐ and 2‐year‐old vegetation. Late successional species showed either lower establishment rates in the younger succession stages or decreased establishment with succession that nevertheless resulted in significant establishment in the oldest plots. Seedling emergence proved to be the key factor determining the establishment pattern of early and late successional species. In absolute numbers, emergence of late successional species was, however, similar or higher than that of early successional species, even in the earliest succession stage. The poor establishment of late successional species on former arable land could therefore not be explained solely by differences in establishment characteristics between early and late successional grassland species. Competitive processes between early and late successional species later in the life cycle probably play an important role. The results do point out that establishment of late successional species can be promoted by creating vegetative cover from the start of the restoration effort.