«Diversity patterns in transitional grassland areas in floodplain landscapes with different heterogeneity Diversitätsmuster in Graslandkomplexen von ...»
Tuexenia 33: 347–369. Göttingen 2013.
available online at www.tuexenia.de
Diversity patterns in transitional grassland areas
in floodplain landscapes with different heterogeneity
Diversitätsmuster in Graslandkomplexen von Auen
mit unterschiedlicher Landschaftsheterogenität
Solvita Rūsiņa*, Ilze Pušpure, Lauma Gustiņa
Faculty of Geography and Earth Sciences, University of Latvia,
10 Alberta Str., LV-1010 Riga, Latvia
*Corresponding author, e-mail:email@example.com
The complex topography of floodplains provides conditions for high diversity and density of transitional areas between different grassland plant communities. Nevertheless, transitions have been almost completely neglected in previous studies of diversity patterns in semi-natural floodplain grasslands.
We analyzed α-, β- and γ-components of plant species diversity in transitional areas between neighbouring wet and dry grassland communities in two landscapes of the Gauja River floodplain (Latvia) differing in landscape heterogeneity created by land use history and current management type (grazing versus mowing). In total 9 transition areas and their adjacent vegetation were sampled in 1 m wide and 8 to 28 m long belt transects gridded into 0.5 m × 1 m plots. Cluster analysis was used to analyze variation in species composition of transitional areas and adjacent vegetation. Indicator species analysis was used to determine species specific to transitional areas (ecotonal species).
Transitional areas of the homogeneous site had a more distinct species composition (clear division in clusters by cluster analysis) and significantly lower β-diversity than those of the heterogeneous site.
α-Diversity was significantly higher in transitions than in wet grasslands and lower than in dry grasslands in both sites. Comparing the two sites, α-diversity was significantly higher in wet grasslands of the heterogeneous site, but no differences were found between transitions and dry grasslands in the two sites. Higher β-diversity of transitional areas in the heterogeneous site could be attributed to a higher density of different habitats per unit area in combination with grazing. No species were restricted to transitions, and no differences were apparent in the number of generalist species (indifferent species sensu Ellenberg indicator values for edaphic factors) between transitions and adjacent grasslands.
Nevertheless, the total number of generalist species was considerably higher in the investigated floodplain grasslands than usually reported for respective vegetation types in Latvia.
It was concluded that transitions of neighboring floodplain grassland plant communities were more important in shaping β-diversity comparedto α- and γ-diversity. Our results suggest that destroying transitional areas between dry and wet grasslands by leveling the ground or by abandonment of the management practices will decrease habitat heterogeneity and lead to less pronounced ecotonal processes in the whole riverine landscape. Thus further studies about the role of fine-scale transitional areas for biodiversity of floodplains are important for ecological restoration of floodplain grasslands.
Keywords: boundary, diversity, ecotone, generalists, grazing, mowing Erweiterte deutsche Zusammenfassung am Ende des Textes Manuscript received 17 September 2012, accepted 17 April 2013 Co-ordinating Editor: Triin Reitalu
1. Introduction In the boreal and nemoral zone of Europe, a high proportion of semi-natural grasslands are located in floodplains. For example, EU Habitats Directive habitat type 6450 (Northern Boreal Alluvial Meadows) covers 43% of all EU-protected semi-natural grassland habitats in Latvia, 22% in Lithuania and 27% in Estonia (ANONYM 2012). Floodplains are disturbancedominated ecosystems characterised by a high level of habitat and landscape heterogeneity (BAKER 1989, TOCKNER & STANFORD 2002). Microtopography formed by erosional and depositional processes of rivers provide conditions for high diversity and density of transitional areas between different plant communities. Transitional areas (ecotones, ecoclines, boundaries (KENT et al. 1997, STRAYER et al. 2003, KARK & VAN RENSBURG 2006)) are important components of landscape heterogeneity (AHLQVIST & SHORTRIDGE 2010) because much of the floristic diversity at the habitat and landscape scale is attributable to transitional areas (TOCKNER & STANFORD 2002). Transitional areas are influenced both by natural factors and by management activities. Management is a driving factor in the development and patterning of ecotones/ecoclines in human-dominated landscapes (KARK & VAN RENSBURG 2006, DUTOIT et al. 2007).
Theoretical frameworks of ecotones include a hierarchy of spatial scales. Three main spatial levels of transitional areas can be distinguished. Macroecotones are transitions between biomes. Mesoecotones are ecotones on the scale of a landscape (forest-grassland, grassland-fen etc.) or a catena (a sequence of soils and plant communities down a slope, created by the balance of processes such as precipitation, infiltration and runoff). Microecotones are boundaries between individual plants and populations (CADENASSO et al. 2003, PETERS et al. 2006, ERDÖS et al. 2011). Generalisations about transitions may apply only to a subset of boundaries with common attributes (STRAYER et al. 2003).
With regard to structure, mesoecotones differ in the degree of structural complexity. Two main categories can be distinguished. Transitions between two structurally different ecosystems (such as the transition from a system dominated by one life form to a system dominated by another life form, e.g. forest-grassland, grassland-fen, grassland-mire etc.) are boundaries between two systems (KOLASA & ZALEWSKI 1995) and can be considered ecotones of asymmetric interface (VAN DER MAAREL 1990, MARGALEF 1994). The second category includes transitions that are part of a single system (KOLASA & ZALEWSKI 1995). Such transitions develop between two structurally similar ecosystems (both of which are dominated by the same life form, e.g. grassland-grassland, forest-forest etc.) and can be considered environmental ecoclines (VAN DER MAAREL 1990) or symmetric interfaces (MARGALEF 1994). Our research was focused on the symmetric grassland-grassland interface in floodplains, i.e. on mesoscale transitional areas.
Mesoecotones between plant communities within the same vegetation formation at fine spatial scale have gained far less attention in biodiversity studies (e.g. ZALATNAI & KÖRMÖCZI 2004, ZALATNAI et al. 2007) than other types of transitional areas (see KENT et al. 1997, KARK & VAN RENSBURG 2006, HUFKENS et al. 2009 for a review). To our knowledge no biodiversity data are available about catena-scale transitional areas in grassland ecosystem complexes. However, it should be emphasised that transitional areas, regardless of their spatial resolution, have important ecological functions including maintenance of biodiversity (PETTS 1990, PINAY et al. 1990, KOLASA & ZALEWSKI 1995, NAIMAN & DÉCAMPS 1997, MARSHALL & MOONEN 2002, BERGER et al. 2003, STRAYER et al. 2003).
The aim of this research was to assess the diversity of transitional areas between dry and wet semi-natural grasslands in a riverine landscape with reference to landscape heterogeneity.
To reveal the diversity patterns of dry-wet grassland transitional areas, we analyzed species diversity by partitioning it into α-, β- and γ-components. β-Diversity is both turnover (the change in species composition along a spatial or an environmental gradient) and variation (dissimilarity among communities, variance in community structure) (ANDERSON et al.
2011). To address the latter and to reveal the importance of transitional areas in creating grassland community diversity, we also examined community structure by means of cluster analysis. We expected higher spatial heterogeneity of vegetation in transitional areas (higher β-diversity and more clusters in transitional areas than in adjacent vegetation) because of the spatial mosaic of plant assemblages created both from ecotonal species and species of adjacent vegetation (STOWE et al. 2003).
Ecotonal and generalist species are important components of transitional area species diversity. True ecotonal species are species that are restricted to the transitional area and are absent from the two communities flanking the ecotone (WALKER et al. 2003). Transitional areas can possess more generalist species than the neighboring communities because environmental conditions in transitional areas are more stochastic and species with higher dispersal ability and ecological plasticity can establish more easily in transitional areas (ZELENÝ et al. 2010). We hypothesized that transitional areas of dry and wet grasslands will be characterized by the existence of ecotonal species and by higher numbers of generalist species than dry and wet grasslands on either side of the transitional areas.
Finally we examined whether the diversity patterns of dry-wet grassland transitional areas differ depending on landscape heterogeneity. Landscape heterogeneity, alongside with vegetation and land-use history and management practices, is one of the most important landscape-scale factors creating high biodiversity in semi-natural grasslands (COUSINS & ERIKSSON 2001, LINDBORG & ERIKSSON 2004, OKLAND et al. 2006, COUSINS et al. 2007, WELLSTEIN et al. 2007). Landscape heterogeneity is scale-dependent (GOSZ 1993, WIENS
2000) and determined both by natural environmental heterogeneity and by management history (BENTON et al. 2003). The importance of landscape characteristics for species diversity has been shown in several studies: β-diversity is positively associated with spatial heterogeneity (VAN DEN BOS & BAKKER 1990, DE BELLO et al. 2007, KOMAC et al. 2011), species richness is higher in geomorphologically more diverse landscapes (WELLSTEIN et al. 2007), and species richness is promoted by the mass effect or vicinism, which is higher in more heterogeneous landscapes (ZELENÝ et al. 2010). We therefore expected higher plant species diversity in a heterogeneous landscape than in a homogeneous landscape.
2. Material and methods
2.1 Study area The research was carried out in the middle reaches of the Gauja River Valley, Northern Latvia, in the Protected Landscape Area „Northern Gauja” (Fig. 1). Due to the light sedimentary rocks in the area, the river has formed a characteristic valley landscape with many meanders, oxbow lakes and floodplains. The river flow rate is 800 m3 s-1 during flooding (5–20 days in spring) and 6 m3 s-1 in winter;
fluctuations of the water table are 3.2–4.8 m (ĀBOLTIŅŠ 1971). The climate is relatively cool and moist.
The average temperature is +5.5 °C (the coldest month is January with -6.5 °C, the warmest July with +16.5 °C). The average annual precipitation is 600 mm.
Fig. 1. Location of the study area.
Abb. 1. Lage des Untersuchungsgebiets.
2.2 Landscape characteristics of the study sites Two sites differing in landscape heterogeneity and management regime were chosen; they are referred to as the heterogeneous site (grazed) and the homogeneous site (mown) throughout this paper (Fig. 2).
The heterogeneous site was located in the Middle Gauja forested landscape region with a forest cover of 60% (Pinus sylvestris-dominated on poor fluvioglacial sandy podzols). Agricultural lands are concentrated mainly in the valley of the Gauja River on sod-gley and alluvial soils.
The homogeneous site was located about 21 km upstream from the heterogeneous site in the Gaujiena-Lejasciems half-open landscape region with extensive tracts of arable fields on sod-gleyic and sod-pseudogley soils and a forest cover of 58% (mostly Picea abies-dominated on sod-podzolic soils).
The width of the valley was 3.5 km in the heterogeneous site. Floodplains and terraces were characterized by a much segmented floodplain relief (height differences were 2 m) with many oxbow-lakes of different sizes and depths, oxbow-lake fens, abandoned river beds (channels), levees and small depressions created by floods.
The width of the valley was 1–1.5 km in the homogeneous site. Floodplains were wide and flat with uniform relief. Height differences were 2 m. Oxbow-lakes were mainly wide and not as numerous as in the heterogeneous site.
At a local scale, both sites were dominated by semi-dry grasslands (further referred to as dry grasslands) of the North European alliance Filipendulo vulgaris-Helictotrichion pratensis Dengler & Löbel 2003 (Festuco-Brometea Br.-Bl. et Tx. ex Klika et Hadač 1944) (DENGLER et al. 2003), mesic Arrhenatherion Koch 1926 and Calthion Tx. 1937 grasslands on slopes of depressions and wet Magnocaricion Koch 1926 communities in depressions and old oxbow-lakes. The homogeneous site contained mostly semi-natural grasslands and old fallow lands and only some trees (Fig. 2, 3). The heterogeneous site was more diverse with a mosaic of secondary scrub, fallow lands and dry and wet grasslands (Fig. 2, 4).
Landscape heterogeneity of both sites was compared using the Shannon-Wiener diversity index, −Σpi*ln pi, where pi is the proportion of the area covered by patch type i (FORMAN 1995). The different habitats included in the calculation were rivers; forests; secondary scrub; dry, moist and wet grasslands Fig. 2. Location of transects and land-use of study sites between 1950 and 1990. Note the very articulated microrelief creating high habitat density in the heterogeneous site (best visible in the area of transect no. 4).
Abb. 2. Lage der Transekte sowie Landnutzung der Untersuchungsflächen von 1950 bis1990. Beachte das stark ausgeprägte Mikrorelief in dem heterogenen Gebiet (untere Abb.) mit hoher Dichte an verschiedenen Habitaten (besonders gut sichtbar im Bereich von Transekt Nr. 4).
Fig. 3. Characteristic landscape of the heterogeneous site with dry grassland vegetation on levees and wet grasslands in depressions. The left photo is taken in the area where transect no.4. is located. The right photo shows the location of transect no. 2, which was located on the right side of the depression.
Wet grassland with Carex vulpina was found in the depression. Differences of transitional grassland and dry grassland are smoothed out by grazing.