Dynamics of montane grasslands and species coexistence



Fine scale dynamics
Within-species functional differentiation
Genet diversity
Belowground processes and structures
Population dynamics
Mechanisms of dominance of expansive species
Management design
Research sites



The Krkonose Mts. (a mountain range in the northern part of the Czech Republic, the highest peak 1602 m) reach above the timberline and hence harbour natural treeless vegetation (alpine meadows above timberline and in avalanche tracks). In addition to that, there are numerous larger and smaller grassland areas below timberline that are due to systematic felling in the area. (These meadows are 300-400 years old.)

A view of several grassland sites surrounded by forests (the altitude of the peak, Snezka, is 1602 m; it is the highest peak of the Krkonose range)

A meadow (at the Jana site) in June with Geranium sylvaticum, Ranunculus acris (and many other plants including grasses) in flower


The flora of these secondary grasslands is very rich.

Gymnadenia conopsea


Primula elatior

The flora contains also numerous species that descended from natural treeless habitats, particularly from glacial cirques (e.g. Hypochaeris uniflora, Viola lutea ssp. sudetica etc.). The species composition of the grasslands has been maintained by human activity, particularly primarily by mowing (once or twice a year, depending on productivity of the site) and irregular manuring. However, these meadows have been in steady decline after the WW II (when most of the German population had been expelled and new settlers came into the area) that was speeded up particularly following intensification of agriculture in 70ies and 80ies. Since mountain grasslands are largely not suitable for high-input agriculture, a major share of grasslands have been abandoned by now; the abandonments results in developing coarser grain in vegetation structure (i.e. reducing richness at the fine scale), in dominance of some competitive species (particularly Polygonum bistorta, sometimes also Holcus mollis, Calamagrostis villosa or Hypericum maculatum). Getting more involved in the study of these grasslands opened a number of issues of plant population biology (namely of clonal plants), ecological genetics, plant-level interactions, spatiotemporal dynamics of clonal plant communities etc.

A small research group in the Institute has been interested in dynamics and species coexistence in these grasslands since Frantisek Krahulec had inventoried community types there (1979-1985; see Krahulec 1990, Krahulec et al. 1997). Since then, our research has concentrated on the following focal areas:

* Fine-scale dynamics. We began recording fine-scale dynamics using permanent plots of different sizes (either 50 x 50 cm or 30 cm x 30 cm) with grid of cells of 3.3 cm x 3.3 cm. Number of modules (shoots for grasses, leaves for larger herbs, rosettes for small rosette herbs) of each species rooting in these cells have been counted every year. The flowering shoots are counted separately. Plots are recorded yearly in ca. mid July; after the recording the plots are clipped and the clippings are sorted into species. These data revealed a lot of fine-scale dynamics at the cell level. These data enabled to quantify between-species differences in growth form (Herben et al. 1993a, 1994, 1995), relate fine-scale and gross-scale dynamics (Herben et al. 1993b), identify climate-specific responses (Herben et al. 1995), identify species-specific replacements and competitive interactions (Herben et al. 1997b, Law et al. 1997).

Overview of fine-scale data currently available/being recorded for the fine-scale dynamics
Severka, four 50 x 50 permanent plots of 15x15 cells 
Jana, four 50 x 50 permanent plots of 15x15 cells
Severka, fertilization experiment, 21 30 x 30 cm plots of 8x8 cells 
Severka, removal experiment, 15 30 x 30 cm plots of 8x8 cells 
Jana, removal experiment, 18 30 x 30 cm plots of 8x8 cells

NB. The long-term fine-scale permanent plot recording at the Severka site has produced an uninterrupted data series of 18 years of 50 cm x 50 cm plots with shoot counts recorded at the level 3.3 cm x 3.3 cm. In particular, the data contain information on spatio-temporal dynamics of all species of coexisting clonal plants in one community at a fine scale. After we do some preliminary analyses, we would like to make the data set public and to invite everybody to analyse this data set. Please let me know if you are interested. More information of the recording techniques and the data set is available here.

An example of the data structure is here (each row is one recording; each column is one species; rainbow spectrum in the small squares indicates ramet numbers; small squares correspond to 3.3 x 3.3 squares of the field recording). Note the change in local density of species over time, while the overall density changes only little.


* Species interactions and their mechanisms. The fine-scale dynamics can be seen in two different contexts: (i) as a result of clonal growth form of component species, and (ii) as the result of competitive interactions between species. While there is a correlation between the clonal growth form and the rate of the fine-scale dynamics across species, the role of interspecific competitive interactions has to be demonstrated by manipulative experiments. A removal experiment with fine-scale recording revealed that fine-scale spatial dynamics of some species changed when their neighbours are removed (Deschampsia flexuosa, Festuca rubra), some other species were not affected at the fine scale (Nardus stricta, Anthoxanthum alpinum; Herben et al. 1997a). Removal experiments with onsets in three subsequent years revealed that the year-to-year variation at the site has a high potential to affect species coexistence and richness in the community (Herben et al., submitted). An implant experiment with Festuca rubra demonstrated some species-specific neighbourhood effects on tillering; vertical growth is affected by biomass of the neighbours, but shows no response to neighbour identity (Herben et al., Oikos in press).

* Modelling of the spatiotemporal structure. A simulation model has been built to enable to link the following kinds of data

The modelling framework is based on a combination of an architectural model with an individual-based model of ramet competition and is calibrated by the field data of these four species. First results are just being published (Herben & Suzuki in press 2002). Some details on the model are also in our talk at the IAVS conference at Freising in 2001. Some   information on simulation modelling can be found here.

* Functional differentiation of dominant species at the genet level. Hana Skalova demonstrated in her MSc thesis that Festuca rubra clones coming from different vegetation types differ in the intensity of their response to red/far red ratio of the incident light (Skalova & Krahulec 1992). Further experiments have shown that a similar type of variation exists even among clones within one vegetation type: the clones differed in response in tillering rate, structure and size of the rhizome system. There was a strong GxE interaction in most of these parameters (Skalova et al. 1997). Implications of this fact for the species coexistence within the grassland is now being explored. Current experiments are testing the role of this functional differentiation by implanting vegetatively propagated material of different genets in different microenvironment. In contrast to strong GxE interactions found in culture, there does not seem to be a lot of interaction of genet identity with the natural microenvironment (Herben et al. 2001). Currently a similar study is conducted on two closely related species of Anthoxanthum (A. alpinum and A. odoratum) which show pronounced altitudinal differentiation (Monika Flegrova-Durasova, MSc thesis).

* Genetic structure of dominant species. DNA RAPD was used to determine number of genotypes within the meadow and their mean size (i.e. distance between the most distant individuals belonging to the same clone); it turned out that the number of clones was quite high and the spatial extent of clones was rather small. A simulation study showed that a constant input of new genotypes by recombination (i.e. seed reproduction) is necessary to maintain this population structure (about 4-10 seeds establishing per 1m2 and year; Suzuki et al. 1999). A project  to link genetic structure of dominant species (Festuca rubra, Anthoxanthum alpinum and Deschampsia flexuosa; it is impossible for Nardus stricta since this species is likely to be agamic) with the fine-scale spatiotemporal structure known from the permanent plots  is under way (collaboration with Jun-ichirou Suzuki and Toshihiko Hara). As a part of this research, we examine the potential of different sampling strategies to recover the 'true' numbers and sizes of genets in clonal plants.

* Belowground processes and structure. Sylvie Pechackova , Radka Wildova and Stanislav Brezina have been studying  rhizome architecture under different levels of competition, tracer element transport horizontally and vertically, root spatial extension, fine-scale heterogeneity in root growth and associated heterogeneity in nutrient availability. Most of the work has concentrated on three main topics:

* Population dynamics and shoot growth dynamics of major species. Special attention has been paid to growth dynamics of Festuca rubra. Its shoot growth and tillering in culture has been studied by Ales Hajek (MSc. thesis). Fate of individual shoots in the field is almost age-independent; since the variation in size among shoots is low, shoot size plays also a minor role (Hara & Herben 1997). The only major factor affecting shoot fate is their (morphological) origin, i.e. whether the shoot is formed intravaginally or extravaginally (Herben et al. 1994). Hana Skalova-Kotrbova (another MSc thesis) demonstrated pronounced response in the parameters of shoot growth and tillering to changed red/far red ratio of the incident light (see also below). A similar study conducted on other three species (Deschampsia caespitosa, Silene inflata, Polygonum bistorta) showed that red/far red response is species-specific (Daniel Abazid, MSc. thesis). Two MSc. theses (by Katerina Jancarikova-Macurova and Jana Samcova) identified rates of dicot seedling establishment in the field and fate of seedlings in different types of microsites; the establishment seems to be highest in sites already occupied by dicots.

* Seedling establishment in grasses (Anthoxanthum spp., and Festuca rubra) is been recorded to (i) check the predictions obtained from the long term estimates by the genetical study (here the study involves also estimation of the year-to-year variation in the seedling establishment, and some characteristics of the micro-sites, such as density and structure of the neighbouring stand and light conditions), and (ii) to investigate interspecific differentiation seedling recruitment rate in Anthoxanthum spp.

* Population level mechanisms through which some plant species attain dominance after abandonment. Abandoned meadows of the Central European mountain are usually characterized by dominance of several herbs or grass species. The dominance mechanism and reasons leading for the dominance of particular species are mostly unknown. In the Krkonose Mts, the following species dominate these abandoned grasslands:


Herbs: Polygonum bistorta (=Bistorta major), Hypericum maculatum, Vaccinium myrtillus, rarely Galium harcynicum (= G. saxatile)
Grasses: Poa chaixii, Deschampsia cespitosa, Calamagrostis villosa, Holcus mollis, Deschampsia flexuosa.


To understand what conditions lead to a dominance of particular species is important for the meadow management and for the restoration of species rich meadows. It seems that local soil conditions, management regime before and at time of meadow abandonment, and biological properties of individual species play important role for future development. As a model system, the meadows dominated by Polygonum bistorta were studied in detail, especially with respect to its influence on nutrient (especially nitrogen) cycling. Polygonum has an efficient nitrogen metabolism and during the time after abandonment about 40% of total nitrogen is stored within its biomass, mostly within its rhizomes (Pechackova & Krahulec 1995). Cutting of meadows leads to poor conditions, which are poor to other species than to Polygonum. Efficient management should include both suppression of Polygonum and fertilization of other species. Cutting combined with manuring or grazing seems to be sufficiently effective to restore herb rich meadows during three to five years. A manipulative experiment (see also here) has indeed shown that mowing and nitrogen addition were the most effective treatment to suppress Polygonum bistorta and restore species-rich grassland.

Polygonum bistorta

Similar experiment are going at present to study the Holcus mollis and Hypericum maculatum (diploma theses by K. Scharffova and A. Moravcova ).

* Grazing as an alternative to the traditional management. Since the traditional management by mowing and manuring is largely unattainable due to its high labour costs, sheep grazing is tested as an alternative. The experiments have been carried out in two localities: (i) Predni Rennerovky - sheep grazing in combination with mowing once in several years (MSc thesis of Radka Patkova), (ii) Zadni Rennerovky (50o41'53"- 50o42'05"N, 15o40'00"-15o40'10"E) - sheep grazing with different duration. The second experiment was done in co-operation with M. Bilek (Institute of Animal Husbandry) and with the Krkonose National Park Administration. Grazing leads to the suppression of the species that dominate the degradation phases (Polygonum bistorta) and to an increase in grass species; with the increase being positively correlated with an increase in several herbs. However, different vegetation changes following grazing cessation (i.e. persistence of nitriphilous herbs and grasses) in two localities indicate limited extrapolation of the results. No negative influence on rare and protected plants was observed. The sheep grazing, particularly when done on a rotating basis, seems to maintain most of the species of the species-rich grasslands. The main and only conservation risk is the expansion of a competitive nitrophilous (and unpalatable when old) species Deschampsia cespitosa. Grazing has thus to be combined with occasional mowing that is needed to suppress spreading of this unpalatable species.

* Mulching as an alternative to the traditional management. Mulching became a way of management of mountain meadows in the Czech Republic during a last decade. Unfortunately, there is no experience with it on longer time scale. It is evident that on the shorter time scale (of years) it can substitute hay-making and that it is better system than abandonment. There is a danger of long term effects connected (i) with absence of export of biomass (and nutrients) from enriched part of meadows and (ii) with increase of undecomposed litter in meadow part with low decomposition rate. For that reason we started in cooperation with administration of the Krkonose National Park with a long-term experiment testing the influence of mulching and manuring (to stimulate decomposition) on the species composition of mountain meadows. Three blocks were established in different communities differing in their position on the productivity gradient: mulched-manured, mulched-unmanured, mown-manured and mown-unmanured.
The experiment was established in in 1997 in Snezne Domky settlement on the ridge of Rychory, in the eastern part of the Krkonose Mts, at the altitude of 960-1000 m, in the following communities: Sileno-Nardetum pleurozietosum (oligotrophic, lowest production), Geranio-Trisetetum melandrietosum (mesotrophic, medium production), and degradated meadow covered by Poa chaixii (eutrophic, high production). During the first three years of the experiment, we studied decomposition of mulched material to compare native material within particular block and cellulose (filter paper) to compare between plots. There were no significant differences in the decomposition during first three years of experiment (the years were wet and good for decomposition processes). On the other hand, significant changes were found in the species composition with decrease of herbs on mulched plots. Diploma thesis of Martin Lexa was defended at Charles University in 2000, at present the research is continued by Andrea Moravcova (student of Dpt. of Ecology and Environment Protection, Palacky University at Olomouc).

* Control of invasive species. Currently experiments to control Rumex longifolius and Rumex alpinus are done by one MSc. student ( A. Bucharova)

The intensive research is carried out at two major study sites:

  1. the Severka study site is located in a mountain grassland in the Krkonose Mts. (ca. 3 km NW of Pec pod Snezkou, latitude 50o 41' 42'' N, longitude 15o 42' 25'' E, altitude approx. 1100 m). The site has a harsh climate; mean temperature in the warmest month (July) was 13.6o C (1988), 13.70 C (1989) at the nearby climatic station (Pec pod Snizkou, ca. 900 m a.s.l.). There is a long winter with thick snow cover, which lasts generally from November till the end of April. There are only five principal species at the site: Anthoxanthum alpinum, Deschampsia flexuosa, Festuca rubra, Nardus stricta and Polygonum bistorta. Though the system is species poor at the large scale, the species are coexisting at the very fine scale; the species density is 2-4 species/10 cm2 and 6-10 species/2500 cm2. In the Braun-Blanquet classification of the Krkonose grasslands (Krahulec 1990) the studied grassland is classified as Sileno-Nardetum (Nardo-Agrostion, Nardetalia), subassociation pleurozietosum. Most of the research carried out here terminated in 2001. Also  recording of the permanent plots S1 to S4 finished that year.
  2. (2) the Jana study site (Janovy boudy, 3.75 km ESE of Pec pod Snezkou, latitude 500 41' 28'' N, longitude 15o 47' 35'' E, altitude 880 m, slope 50). This is a species-rich locality, with ca. 4- 7 species per 10 cm2 and 25-30 species per 2500 cm0). The locality has a milder climate with a longer growing season (ca. 7-8 months) than the Severka locality. Also soils are richer in magnesium and calcium and have a lower C/N ratio in the species-rich locality. Except for Euphrasia rostkoviana at the species-rich locality there are only perennial species at both localities.

The Jana site in early summer


Please mail me if you have any comments or if you wish to learn more.

Back to Institute of Botany