Investigating the direction of changes in functional diversity involving successional pathways and historical disturbances may be a promising tool for predictions of the effectiveness of the semi-natural meadows conservation, with great emphasis on formulation of more cost-effective restoration strategies. The goal of this research was to assess the differences in long-term shifts in the functional diversity of plant species in semi-natural wet meadows unmanaged for the last 40 years, under the influence of different successional pathways and historical disturbances. Using ordination techniques, linear mixed-effect models, a set of plant functional traits and parameters of functional diversity, we assessed the importance of habitat filtering, competition and niche partitioning in shaping community assembly changes over time. The most dramatic shifts in functional diversity were found in the Carex acutiformis successional pathway after topsoil removal, where colonization by successional inhibitors was the main driver causing decreases in functional dispersion and divergence. This was expressed as a decrease in the importance of habitat filtering and replacement of specialized species by competitors with heavier seeds and higher specific leaf area. Regarding the C. cespitosa and Salix cinerea pathways, the magnitudes of shifts in functional diversity were milder and differed less between the historical topsoil removal and mowing treatments, thereby maintaining a large role for niche partitioning in shaping the vegetation structure. The results of our study highlight the importance of tussock sedges and shrubs as effective buffers against the biotic homogenization of meadows driven by the simplification of functional composition, even from a long-term perspective.

The study was conducted on unmown meadows in the Reski Range, with an area of approximately 15 ha. The Reski Range is located in the northwestern part of the Białowieża Glade in Białowieża National Park (N 52⁰42’32’’; E 23⁰50’06’’) in the Narewka River valley (NE Poland). Meadows of the Reski Range occupy the habitats of old ash-alder riparian forests. More than 200 years ago, in the part of the Narewka River valley, forests were cut down, and extensively mowed meadows were formed (Falińska 2003). Due to relatively low utility values and early spring flooding, the meadows in the Reski Range were gradually abandoned by farmers. The process of scything cessation began in the 1960s and ended in 1978. After stopping the use of these meadows, spontaneous secondary succession began. This created an opportunity to establish the first long-term permanent plot experiments, aiming to assess the population dynamics of single plant species over time, as well as the mechanisms of the replacement of populations of individual species by other plants during secondary succession (Falińska 1991). This research was initiated by K. Falińska in 1972 (Falińska 1991). After approximately 20 years of these detailed investigations, she identified three main pathways, which were the most commonly occurring pathways in the study area, of secondary succession, based on life-history traits (e.g., clonal architecture, seed bank characteristics and reproductive strategies) of the dominant and co-occurring species (Falińska 2003). The pathway of succession, with the participation of Salix cinerea (hereafter ‘S. cinerea pathway’), involves the exchange of species composition according to the following scheme: meadow – tall herb communities – thicket – forest. In turn, the Carex cespitosa (hereafter ‘C. cespitosa pathway’) and Carex acutiformis (hereafter ‘C. acutiformis pathway’) pathways involve the simultaneous coexistence of species composition characteristics of rush, tall herb and thicket communities (Falińska 2003).

An experiment was conducted in 1995-2015, approximately 20 years after the initial long-term research of plant demography during secondary succession (Falińska 1991, 2003). To study each of the three successional pathways identified by K. Falińska as the most common in the study area, one transect consisting of 30 permanent plots (1x1 m) was established and divided into three sub-transects, each one being 10 m in length and 1 m in width. Therefore, the total number of permanent plots for the three successional pathways and three treatments was 90. At the beginning of the experiment, different types of disturbances were applied once in each sub-transect for each successional pathway. The first sub-transect was subjected to strong disturbance (n=10): the plants, including roots and topsoil surface, were removed. In the second sub-transect (n=10), weak disturbance was applied: the aerial parts of the plants were cut down with a knife/scythe. The third sub-transect (n=10) was left intact as a control area, without the application of any disturbance. Thus, as each successional pathway was identified prior to our long-term experiment and application of different disturbance types, we did not take into account interactions between the successional pathway and disturbance, assuming that observations within each transect were independent. Permanent plots were established in 1994, and the first field survey took place in 1995. Then, within five-year intervals (to 2015), analyses of species composition were carried out on each plot (1 x 1 m) within each sub-transect and each successional pathway, and species abundance (the total number of ramets per species recorded) was noted. Plant species names were adopted after Mirek et al. (2002).

Using the LEDA database, we compiled information on seven functional traits related to plant size and leaf morphology (i.e., canopy height, shoot growth form, leaf distribution along stem, and specific leaf area; SLA) and dispersal abilities (dispersal mode, seed mass and seed bank longevity).