ACTA BOT. CROAT. 76 (2), 2017 111 Acta Bot. Croat. 76 (2), 111–119, 2017 CODEN: ABCRA 25 DOI: 10.1515/botcro-2017-0001 ISSN 0365-0588 eISSN 1847-8476 Relation between plant species diversity and landscape variables in Central-European dry grassland fragments and their successional derivates Igor Paušič, Danijel Ivajnšič, Mitja Kaligarič, Nataša Pipenbaher Biology Department, Faculty of Natural Sciences and Mathematics, University of Maribor, Koroška c. 160, SI-2000 Maribor, Slovenia Abstract – A systematic field survey of an area of 843 ha in the traditional Central-European agricultural landscape of Goričko Nature Park in Slovenia revealed 80 fragments of dry semi-natural grasslands. Vascular plant species diversity was studied in relation to landscape variables and to threat (Slovenian red-listed spe- cies). Our results show that fragment size does not affect plant species diversity. In addition, fragment shape index is not related to Alpha diversity. Higher Alpha diversity was observed for abandoned grassland frag- ments. The lowest Alpha diversity was perceived on more mesic fragments, where habitat specialists are much scarcer. It was confirmed that the highest diversity of specialists are in the driest fragments, both still mowed and abandoned. With an increase in the number of distinctly different bordering habitat types, the to- tal number of species per fragment generally does not increase, except in the case of those fragments that are already in different succession stages. Abandoned and typical dry grasslands are associated with a higher number of bordering habitats. Typical dry grassland fragments and abandoned ones, which probably derived mostly from drier (less productive) grasslands, are found on lower altitude and have a lower shape index. Habitat specialists Sedum sexangulare, Polygala vulgaris and Spiranthes spiralis have higher frequency in fragments with a lower shape index. This means that these oligotrophic specialists occur in smaller frag- ments. But Orchis morio has higher frequencies of occurrence on polygons with a higher shape index, which confirms the observation that this species occurs in larger and more irregular fragments, as well as close to houses and fields and along the roads. Key words: Goričko Nature Park, habitat specialists, NE Slovenia, semi-natural dry grasslands, traditional agricultural landscape * Corresponding author, e-mail: igor.pausic@um.si Introduction Over the past century, grasslands and other semi-natural plant communities in Southern and temperate Europe have suffered a dramatic decline in the area they cover, owing to land use changes (Luoto et al. 2003). However, the primary driver of the current decrease in species densities (and their populations) and species loss is habitat fragmentation (Fat- torini and Borges 2012). Habitat fragmentation not only causes habitat loss or reduction, but also promotes habitat isolation, diminishes habitat quality, and increases edge ef- fects (Lomolino et al. 2010). The alteration of landscapes throughout the world has resulted in fragmentation of natu- ral and semi-natural vegetation, imposing a great threat on numerous plant and animal species (Kiviniemi and Eriks- son 2002). Many researchers have concluded that habitat fragmentation poses an important additional threat to biodi- versity (Hanski 2005, Janišova et al. 2014, Zulka et al. 2014), but others claim that fragmentation per se is of sec- ondary importance (Fahrig 2003). One reason why the ef- fect of connectivity (or fragmentation itself) might not be significant is the slow response of populations to environ- mental change (Hanski 2005). One consequence of the in- tensification of agriculture in Europe over the last half-cen- tury has been the loss and fragmentation of traditionally managed habitats, such as semi-natural grasslands (Kivini- emi and Eriksson 2002, Cousins 2005). In fragmented land- scapes, plant species persistence depends on functional PAUŠIČ I., IVAJNŠIČ D., KALIGARIČ M., PIPENBAHER N. 112 ACTA BOT. CROAT. 76 (2), 2017 connectivity in terms of pollen flow to maintain genetic di- versity within populations and seed dispersal to re-colonize habitat patches following local extinction (Rico et al. 2011). Habitat fragmentation is predicted to lead to area-related re- ductions in population size and decreasing colonization rates as a result of isolation (Kiviniemi 2008). According to island biogeography theory, the species richness of patches is determined by their size and spatial isolation, while in conservation practice, it is patch quality that determines protection and guides management (Zulka et al. 2014). Dry grassland species-area relationships can be explained by two ecological hypotheses: the first predicts higher species richness due to the higher habitat heteroge- neity and the second claims that colonisation (extinction) dynamics causes increasing species numbers with increas- ing habitat area independent of habitat heterogeneity (Rosenzweig 1995, Krauss et al. 2004). The area effect per se cannot be easily disentangled from the habitat diversity effect and so it is difficult to quantify the interrelationship (Steinmann et al. 2011). Previous studies have already re- ported a positive correlation between habitat heterogeneity and species diversity (Rosenzweig 1995, Ricklefs and Lovette 1999). There are only a few studies dealing with habitat specialists in general while there are several studies of the frequency of particular grassland species in relation to fragment area, shape index and to habitat heterogeneity (Paušič and Kaligarič 2015). The landscape structure in Goričko, i.e., the spatial dis- tribution of ecotopes, is regarded as a mosaic of “frozen processes”. In other words, landscape structure is assumed to mirror the processes that have been going on in a land- scape (Forman 1995). Small-scale patchiness and its diver- sity reveal the “typical” traditional agricultural landscape, which was fragmented for socio-economic reasons at the end of the 19th and at the beginning of the 20th century. Goričko Nature Park was established in 2003 in order to protect this well-preserved, traditional Central-European agricultural landscape. The mosaic structure of landscape and the variety of secondary habitats in Goričko were caused by the simultaneous effect of natural and anthropo- genic-geopolitical, economic and social factors (Kaligarič et al. 2008). In the traditional agricultural landscape of Goričko, there were basically four major land use (manage- ment) categories: arable fields, grasslands (including or- chard grasslands), vineyards and gardens. During the past century, traditional land use changed drastically in Central Europe, and gradually also within the study area inside Goričko Nature Park (Škornik 2003). The area used for grasslands decreased mainly because of land abandonment and forest progression (Cousins et al. 2014). The aim of this study was to investigate the relation be- tween plant species diversity, fragment area and fragment shape index. The following methodological steps were un- dertaken: (1) analyse the differences in floristic composi- tion of dry grassland fragments using TWINSPAN soft- ware, (2) compare the plant composition/frequency in the fragments with selected landscape variables, (3) study the differences between TWINSPAN groups and all three Shannon-Weaver diversity indices (Alpha diversity, Alpha diversity of habitat specialists and Alpha diversity of habi- tat generalists), (4) test the relations between the landscape variables and frequency of single habitat specialists, and (5) find differences between threatened (red-listed) plant spe- cies and TWINSPAN groups. Material and methods Study area and field methods The study area lies in NE Slovenia within the borders of Goričko Nature Park (Fig.1), at approximately 46˚50’– 46˚52’ N, 16˚15’–16˚52’ E and covers 843 ha. The area has a Central-European climate with a relatively dry winter. Av- erage annual rainfall is between 750 and 820 mm (ARSO 2014). The driest months are January and February, while the most rainfall occurs in July and August (ARSO 2014). Mean annual temperature is between 9 and 10 °C (ARSO 2014). Dry and semi-dry grasslands in Goričko are not as rich in species as similar grasslands elsewhere in Slovenia, owing to the low pH value of the soil (Škornik 2003). We analyzed 800 plots on 80 dry grasslands fragments (including some earlier or middle successional stages after abandonment) systematically mapped in the study area. They mostly belong to the association Hypochoerido-Fes- tucetum rupicolae or drier stands of the association Ranun- culo bulbosi-Arrhenatheretum elatioris. All dry grassland fragments (80) were sampled and analysed on randomly dispersed square meter plots (10 per each fragment). Spe- cies composition was recorded on 800 plots, where vascu- lar plants were sampled using presence/absence data. For each plant species we calculate frequencies; 10 plot sam- ples on 80 fragments. The vegetation composition data were collected in only one year (2013) by the first author. Plant species occurring only in one sample were removed (we exclude 11 plant species), in order to exclude casual occurrences from the analyses. Taxonomic nomenclature followed Martinčič et al. (2007). Plant species specialists and generalists were defined according to expert knowl- edge, by help of the syntaxonomic classification of the taxa. Plant species characteristic for the alliances Bromion erecti (and order Brometalia erecti) and Violion caninae represent species of drier, nutrient-poor neutral to acid soils of Cen- tral European provenance and climate. The rest of the spe- cies were considered generalists, although some species from e.g. ruderal or forest habitats are also considered to be in this group. Fig. 1. Map of Slovenia with bordering countries (A) and study area (B) with 80 dry grassland fragments. SPAtIAL PAttERN of DRy GRASSLAND fRAGMENtS ACTA BOT. CROAT. 76 (2), 2017 113 Species diversity Shannon-Weaver diversity index (H) was calculated for all species (called Alpha diversity), and further indices were calculated only for species specialists (called Alpha diver- sity specialists) and for species generalists (called Alpha diversity generalists) (Shannon and Weaver 1963). We also identified threatened plant species using the Red Lists of the threatened flora of Slovenia (Anonymous 2002). Of the 169 plant species included in the study, 7 were included in the Red List and from this we calculated threatened species frequency (Frk_red). Threatened species frequency was noted as the number of threatened plant spe- cies recorded in 10 plots on each dry grassland fragment. Landscape variables The spatial geometry (area and perimeter) of the grass- land fragments was obtained using ArcGIS 9.3 spatial ana- lyst tools (ESRI 2010) by digitalizing the plots – drawing polygons in vector format. The plots were first drawn on print- ed Orto-photo imagery in 2009 (pixel size=0.5 m) (GURS 2010). Later, the relation between fragment area and its pe- rimeter – a straightforward “shape index”, perimeter-area ratio for each grassland fragments was calculated (PARA index, referred to as “shape index” later on in the text). All different bordering habitat types (polygons) (NofHT) for each grassland fragment were determined using an ex- isting 2012 habitat type map owned by the national Institute for Nature Conservation (ZRSVN) which is based on the PHYSIS typology (Devillers and Devillers-Terschuren 1996), adopted and modified for Slovenian conditions (Jogan et al. 2004). The buffer analysis tool within ArcGIS 9.3 (ESRI 2010) was applied in order to quantify habitat types that share the border with the target grassland fragments. Addi- tionally, the above-mentioned habitat type map was used to calculate the distance from the settlement/houses variable (DIS2H). In this regard, all polygons determined as build- ings (houses) were selected and extracted from the initial habitat type map. The distance surfaces between all houses were calculated in the next step by using the Euclidean dis- tance algorithm in the ArcGIS spatial analyst tools (ESRI 2010). The mean distance from the settlement/houses for each grassland fragment in the study area was determined in the end with the Zonal statistics tool in ArcGIS and was additionally log transformed for the purpose of statistical analysis. Accordingly, a digital elevation model (DEM) in a 5 m horizontal resolution (GURS 2010) served as the input data with a calculation of the mean height above sea level variable (HASL) for each grassland fragment and the ap- plication of the zonal statistics tool in ArcGIS 9.3 (ESRI 2010). All variables were z standardized for the purpose of landscape variables analysis. Data analysis To explain the relations between Alpha diversity of all 169 plant species and 46 habitat specialists the relations be- tween fragment area and fragment shape index were tested separately with Pearson’s correlation coefficient. The ana- lysis was carried out in R (R Development Core Team 2009). To classify grassland fragments according to their simi- larity we used two-way indicator species analysis (TWINS- PAN) (Hill 1979). This analysis was run with frequencies data of 80 dry grassland fragments and 169 plant species. Further, redundancy analysis (RDA) ordination was conducted with 80 grassland fragments, to relate variation in the whole of the species data set (169) to landscape vari- ables and threatened species richness. TWINSPAN groups were added as supplementary variable without any effect on the analysis. Ordination method (RDA) and visualiza- tion of their results were carried out using the Canoco and CanoDraw programs (ter Braak and Šmilauer 2002). Additionally, a comparison was performed between the TWINSPAN groups and all three Shannon-Weaver diversi- ty indexes (Alpha diversity, Alpha diversity specialist and Alpha diversity generalist) using one-way ANOVA. One- way ANOVA was also performed for TWINSPAN groups and landscape variables. The analysis was carried out in R (R Development Core Team 2009). We used Poisson or QuasiPoisson regression (depend- ing on the data dispersal) to test the effect of landscape vari- ables (HASL, DIS2H, NofHT and shape index) on specialist species. We implemented Poisson or QuasiPoisson regres- sion within a generalized linear modelling (GLM) frame- work, using a Poisson link or QuasiPoisson link function. We tested only specialist species with frequencies higher than ten. The analysis was carried out in R (R Development Core Team 2009). In order to present differences between threatened plant species frequency from the Red List of Slovenia we calcu- lated sum of 10 plots on each dry grassland fragment. In addition we performed the Kruskal- Wallis test to find pos- sible differences between threatened species frequency (Frk_red) and determined TWINSPAN groups. Results From 80 dry grassland fragments, we included in the dataset 169 vascular plant species (On-line Suppl. Tab. 1) with an average of 43 plant species per fragment. The total fragment area covers 25.56 ha. The area of the smallest fragment is 0.028 ha and the largest is 1.12 ha. From 169 vascular plant species, 46 were considered habitat special- ists with an average of 13 plant specialists per fragment, mostly characteristic species of Mesobromion and Violion alliances. The average number of plant generalists per dry grassland fragment is 31 plant species. As the first step, the relations between the Alpha diver- sity of all 169 plant species and 46 habitat specialists be- tween fragment area and fragment shape index were tested separately. However, no significant relations between all mentioned parameters were found (results not shown). At the second step, differences in floristic composition were analysed with TWINSPAN classification of matrix with 169 plant species and 80 dry grassland fragments (Fig. 2). Four groups were distinguished on the basic of this anal- ysis. The first TWINSPAN group represents the most typi- cal extensively used oligotrophic and mostly dry grasslands fragments (referred to as DRY), regularly mowed, with typ- PAUŠIČ I., IVAJNŠIČ D., KALIGARIČ M., PIPENBAHER N. 114 ACTA BOT. CROAT. 76 (2), 2017 ical and abundant species such as Briza media, Peuceda- num oreoselinum and Festuca rupicola. The second TWIN- SPAN group represents also extensive and still dry, but relatively more mesic, grassland fragments (referred as MES) with typical and abundant species as Sanguisorba of- ficinalis, Ranunculus bulbosus and Centaurea jacea. The third TWINSPAN group with the most typical and abun- dant species Quercus petraea, Prunus domestica (remnant of orchards), Tanacetum vulgare, Lychnis flos-cuculi, Ve- ronica camaedrys and Hypericum peforatum represents abandoned grassland fragments (referred as ABA), mostly already overgrown with young woody perennials. The fourth TWINSPAN group is constituted by grassland frag- ments, which are ruderalised (disturbed) fragments (re- ferred as RUD), also containing weeds or ruderals as typi- cal and abundant species: Plantago major, Cirsium arvense, Setaria viridis or Erigeron annuus. In the third step, we compared the plant composition/ frequency in the fragments with landscape variables. In the RDA ordination of two matrixes: first one with 169 plant species and 80 dry grassland fragments and second one with 4 landscape variables (NofHT, DIS2H, HASL and shape index) plus threatened species richness was per- formed in order to differentiate the analysed dry grasslands fragment on the basic of selected landscape variables and their endangerment (threat) (Fig. 3). The first two axes of RDA analysis explained 33.1% of variation. The first axes were related positively to the shape of the fragments (shape index), altitude (HASL) and threatened species richness (Frk_red). However, the second axes were related positive- ly to the number of different bordering habitat types (NofHT), threatened species richness (Frk_red) and nega- tively to shape index, distance from the settlements/houses (DIS2H) and altitude (HASL). Furthermore, the DRY and MES groups had more species from the Red List than ABA and RUD groups and the DRY and the ABA group have highest frequencies of bordering habitat types than the MES and RUD groups, in contrast to landscape variables shape index and altitude. Additionally, as the fourth step, we were looking for the differences between TWINSPAN groups and all three Shannon-Weaver diversity indexes (Alpha diversity, Alpha diversity specialist and Alpha diversity generalist). We found significant differences for Alpha diversity (all 169 species considered) in the first three TWINSPAN groups – DRY, MES and ABA (p<0.001); only the RUD group it is not significantly different from the DRY and the MES group (Fig 4a). Very similar results were found when Alpha diversity was calculated for generalists only. We found sig- nificant differences between MES and ABA group (p< 0.001) and between the DRY and the ABA group (p<0.001). However, the Alpha diversity, calculated for generalists does not differentiate DRY from MES and ABA groups. We found also no significant differences between the ABA and RUD group, when calculating the Alpha diversity for gen- eralists (Fig. 4b). What is more, Alpha diversity for special- ists does not differentiate between the DRY and the ABA group or between the MES and RUD group, but shows sig- nificant differences between MES and RUD groups togeth- er and the ABA group (Fig. 4c). We found also significant differences in landscape variable “distance from the settle- ments/houses” (DIS2H), but we have not found differences Fig. 2. Simplified TWINSPAN classification tree (dendrogram) of matrix with 169 plant species and 80 dry grassland fragments. For each division number of groups, number of dry grassland frag- ment (in brackets) in group and indicator plant species are shown. Grassland fragments were divided into four groups: DRY – exten- sively used oligotrophic and mostly dry grasslands fragments, MES – more mesic grassland fragments, ABA – abandoned grass- land fragments mostly overgrown with young woody perennials, RUD – ruderalised grassland fragments. Abbreviation of plant species are explained in On-line Suppl. Tab. 1. Fig. 3. Redundancy analysis (RDA) ordination diagram of two matrixes: first with 169 plant species and 80 dry grassland frag- ments and second matrix with 4 landscape variables, frequency of the species from the Red List and 80 dry grassland fragments. Ei- genvalues: axis 1=0.158, axis 2=0.097, 25.5 % of variance in spe- cies explained by both axes. Shown species have the highest weight. Grassland fragments were divided into four groups with TWINSPAN analysis: DRY group, MES group, ABA group, RUD group; explained in Fig. 2 in detail. HASL – height above sea lev- el; SHAPE – shape index; NofHT – number of bordering habitat types, DIS2H – distance from the settlements/houses, Frk_red – threatened species richness. SPAtIAL PAttERN of DRy GRASSLAND fRAGMENtS ACTA BOT. CROAT. 76 (2), 2017 115 in TWINSPAN groups between any other landscape vari- ables (Fig 4d). Significant differences in log landscape vari- able “distance from the settlements/houses” were found be- tween the DRY and the RUD group (p<0.05). When testing the relations between the landscape vari- ables and frequency of single habitat specialists, we found only some statistically significant correlations (results not shown). Higher number of neighbouring habitats is corre- lated with higher frequency of Trifolium montanum, Spi- ranthes spiralis and Centaurium erythrea. More distant from the settlements/houses are polygons with Potentila ar- gentea, Carex caryophyllea and Thymus pulegioides. Rela- tion with shape index gave opposite results: Sedum sexan- gulare, Polygala vulgaris and Spiranthes spirales have higher frequency on plots with lower shape index; while Orchis morio have higher frequencies of occurrence on polygons with higher shape index (On line-Suppl. Fig. 1). Finally, we present differences between Red listed (threat- ened) plant species frequency with highest frequency of species Anacamptis morio and Fragaria viridis (Fig. 5). We Fig. 5. Threatened species frequency of seven plant species from Red List of Slovenia (left) and frequency of threatened species in each TWINSPAN group (right). The abbreviations of species names: Centaurea cyanus, Daphne cneorum, Fragaria viridis, Ophioglossum vulgatum, Orchis morio, Orchis ustulata, Spiranthes spiralis. Abbreviations of TWINSPAN groups are explained in Fig. 2. Fig. 4. Differences between TWINSPAN groups and all three Shannon-Weaver diversity indexes: Alpha diversity (A), Alpha diversity generalists (B) and Alpha diversity specialists (C) and for TWINSPAN groups and distance from the settlements/houses (DIS2H) (D). TWINSPAN groups: 1 – DRY group, 2 – MES group, 3 – ABA group, 4 – RUD group; explained in Fig. 2 in detail. Letter in box-plot are from ANOVA, Duncan post- hoc test. PAUŠIČ I., IVAJNŠIČ D., KALIGARIČ M., PIPENBAHER N. 116 ACTA BOT. CROAT. 76 (2), 2017 were looking for differences between the TWINSPAN groups and found the highest values for DRY and MES TWINSPAN group, but there were no significant differenc- es among the four TWINSPAN groups. Discussion A large proportion of European biodiversity today de- pends on habitat provided by low-intensity farming prac- tices, yet this resource is declining as European agriculture intensifies (Sutcliffe et al. 2014). One consequence of the rationalization of agriculture in Europe over the last half century has been the loss and fragmentation of traditionally managed habitats such as semi-natural grasslands (Cousins 2001). In many agricultural landscapes, as also in the Goričko area, only fragments have remained. The crucial question in the situation of the declining number and area of grasslands in the landscape is whether fragment size af- fects plant species diversity. This problem has received dif- ferent answers in studies. However, our results show that the fragment size (large vs. small) does not affect species diversity of plants. This has been confirmed clearly already by Helm et al. (2006) – no effect of current grassland size on plant species diversity in Estonian alvar grasslands was confirmed. Our results indicate that there is no relation be- tween species diversity (here measured as Alpha diversity of all species and habitat specialists only) and grassland area, which could be explained by the assumptions pro- posed by Helm et al. (2006). Long leaved perennial plants may also persist for relatively long time after the habitat has been degraded or converted to a later successional stage. In other words – since grassland fragments were assembled a long time ago and their structure is homogenous, these present remnants are still enough large to harbour more or less the entire species diversity of the previous (larger) area. The authors (e.g. Cousins and Eriksson 2002, Eriksson et al. 2002, Lindborg 2007, Kuussaari et al. 2009, Reitalu et al. 2010) summarized that it is not current but past grass- land areas (prior to the drastic decline of grasslands) and the connectivity among them that affect the current species diversity in fragmented grasslands across Europe. Our study revealed that both fragment size and fragment shape index are not correlated to Alpha diversity: the shapes of the remnant grasslands are different, in many cases elon- gated or irregular, but this does not influence Alpha diver- sity. In semi-natural grasslands, the land use treatment (man- agement) is one of the major ecological factors, known to affect plant reproductive success and vegetation patterns (Lopez-Marino et al. 2000, Eriksson et al. 2002, Lindborg et al. 2005, Wissman 2006, Gustavsson et al. 2007). Sites with long land use continuity had higher plant species di- versities than sites with shorter continuity (Gustavsson et al. 2007). Thus, we questioned if the fragments of dry grasslands are floristically homogenous. The TWINSPAN classification revealed four TWINSPAN groups, which show the influence of moisture/nutrient level (dryer – DRY; more mesic – MES), abandonment (ABA) and disturbance (ruderal – RUD). In the next step we correlated the species composition in the fragments with some landscape vari- ables. The influence of landscape context in habitat frag- mentation studies is often ignored (Wiens 1997, Hanski 1999). High habitat diversity in the surroundings might en- hance the species diversity of a local site. Landscape con- text might affect gene flow via pollen and seed dispersal, the pressure of herbivory and the functional connectivity between habitat patches (Ricketts 2001). The complex ef- fect of landscape context on local plant species diversity revealed the importance of the bordering habitats in many studies (Kollman and Schneider 1999, Metzger 2000, Söderström et al. 2001). High landscape diversity in the surrounding matrix provides more distinct habitat types for generalists or species with other habitat preferences, sup- porting the prediction that landscape diversity enhances the number of generalists, especially at the edges, but hardly of the specialists (Jonsen and Fahrig 1997). In our study area, the fragments (thought to be bigger and more connected polygons in the past) are currently surrounded by numerous habitat types: arable fields, forest edges, mesic meadow or- chards, intensive meadows, vineyards, gardens, roads and even settlements. The RDA ordination diagram shows (Fig. 3) that the ABA and DRY groups are associated with a higher number of different bordering habitats types (NofHT). It is also worth mentioning that the most typical, dry fragments (DRY) and abandoned ones (ABA), which probably also derived mostly from drier (less productive) places, are found at lower altitudes and have lower shape indices. It has been established by Tscharntke et al. (2002) that not all species depend equally on habitat area, isolation and landscape context. Habitat specialists proved to be more af- fected by habitat loss than generalists (Warren et al. 2001). The comparison of Alpha diversity considering all species, habitat specialists only and generalists only between the RUD groups revealed interesting results. Higher Alpha di- versity was calculated for ABA, because abandonment re- sults in colonization of species from the later colonization stages, while grassland species persist. The lowest Alpha diversity was perceived on more mesic fragments (MES), where habitat specialists are much scarcer. It was empha- sized many times that early successional stages after aban- donment represent the most species-rich assemblages (Denslow 1980, Balmer and Erhardt 2000). A similar pat- tern was revealed, when only generalists were considered. Interesting results were obtained when only habitat special- ists were considered. It was confirmed again that the high- est diversity of specialists are in the driest fragments, both still mowed (DRY) and abandoned (ABA), which means that among habitat specialists are mostly long-lived peren- nials which persist also during the earlier successional stag- es. Long-lived plants decline slowly and may survive for decades after environmental change (Eriksson 1996, Dahl- strom et al. 2006, Helm et al. 2006, Lindborg 2007, Hy- lander and Ehrlén 2013). It could be concluded that, with a reduction in grassland fragment area, the majority of spe- cies still conform to previous distribution patterns, when grassland areas were larger. There are no habitat maps from past decades to support this with exact figures. However, SPAtIAL PAttERN of DRy GRASSLAND fRAGMENtS ACTA BOT. CROAT. 76 (2), 2017 117 another study in the NW part of the Goričko Park revealed that at the beginning of the 20th century the area was domi- nated by open agricultural land, with a forest cover around 39%; a landscape composition today shows forests cover- ing around 57% of the landscape (Cousins et al. 2014). Fur- thermore, 65% of the forests in NW Goričko are between only 15 and 30 years old. Young forests are primarily devel- oped on former arable fields or grasslands (Cousins et al. 2014). We claim that this is a strong proof that the grass- lands in Goričko area were much larger in the past, although exact spatial data on their distribution are lacking. High habitat diversity in the surroundings might en- hance the species diversity of a local site. Furthermore, the surrounding matrix provides more distinct habitat types for generalists or species with other habitat preferences, sup- porting the prediction that landscape diversity enhances the number of generalists, especially at the edges, but hardly of the specialists (Jonsen and Fahrig 1997). Interesting results were obtained when the TWINSPAN groups were con- firmed on the basis of the distance of polygons from settle- ments/houses. It seems peculiar at first sight, but the shorter distance to the settlements/houses of the most dry and regu- larly mowed meadows in Goričko can be explained by the fact that still mowed “best preserved” dry (oligotrophic!) fragments are close to the houses, due to the fact that hous- es are built on less productive soils in order to save better land for more productive (fertile) meadows and fields. When testing the relation between frequencies of single habitat specialists we found interesting results: Sedum sex- angulare, Polygala vulgaris and Spiranthes spirales have higher frequency on plots with a lower shape index. This means very small remnants of formerly larger polygons. But Orchis morio has higher frequencies of occurrence on polygons with higher shape index, which confirm the ob- servation that this species occurs on fragments of different complex shapes, also close to houses and fields and along the roads. Semi-dry grasslands, the remnants of traditional agri- cultural landscape, are today threatened mainly due to quick overgrowing, but also due to intensification or con- version to arable land. Disturbance in the form of ruderali- sation (occasional fertilization or mulching) and the early stages of secondary succession emerged as much stronger drivers of community change than decreases in patch size per se. The selection of red-listed species on national level is not really supporting the conservation efforts for grass- lands in Goričko area; only a few such species are present in the area. It can only be concluded that in still mowed (DRY and MES) fragments the number of red-listed species tends to be higher, although not significant and despite the relatively low number of such species. The dilemma wheth- er biodiversity conservation should focus on a single or several large habitat fragments, or whether the protection of many small fragments (covering the same habitat area in a landscape) is of equal or even greater importance is still open (Tscharntke et al. 2012). The dilemma is commonly referred as SLOSS (single large or several small) and has been widely discussed by many authors (e.g. Ovaskainen 2002, Tjørve 2010, Tscharntke et al. 2012). In the case of Goričko Park it is of a crucial importance to preserve the regularly managed medium- and small- sized fragments, which still represent important centres for plant species di- versity within the agricultural matrix of Goričko Nature Park. They are important as refugia and for future restora- tion activities within the park and in the wider region. References Anonymous, 2002: Principles of the inclusion of endangered plant and animal species in the Red List. The Slovenian official ga- zette 82/2002. pp. 8893–8910 (in Slovenian). ARSO, Ministry of the environment and spatial planning of the Republic of Slovenia, Slovenian environment agency, 30-year climate data, Veliki Dolenci. Retrieved on July 2014 from http://www.arso.gov.si/vreme/napovedi%20in%20podatki/ veliki_dolenci.htm. Balmer, O., Erhardt, A., 2000: Consequences of succession on ex- tensively grazed grasslands for Central European butterfly communities: Rethinking conservation practices. Conserva- tion Biology 14, 746–757. Cousins, S. A. O., 2001: Analysis of land cover-transition based on 17th and 18th century cadastral maps and aerial photo- graphs. Landscape Ecology 16, 41–54. Cousins, S. A. O., 2005: Plant species richness in midfield islets and road verges- The effect of landscape fragmentation. Bio- logical Conservation 127, 500–509. Cousins, S. A. O., Eriksson, O., 2002: The influence of management history and habitat on plant species richness in a rural hemibo- real landscape, Sweden. Landscape Ecology 17, 517–529. Cousins, S. A. O., Kaligarič, M., Bakan, B., Lindborg, R., 2014: Political systems affect mobile and sessile species diversity – a legacy from the post-WWII period. Plos One 9/8, e103367. Dahlstrom, A., Cousins, S. A. O., Eriksson, O., 2006: The history (1620–2003) of land use, people and livestock, and the rela- tionship to present plant species diversity in a rural landscape in Sweden. Environment and History 12, 191–212. Denslow, J. S., 1980: Patterns of plant species diversity during succession under different disturbance regimes. Oecologia 46, 18–21. Devillers, P., Devillers-Terschuren, J., 1996: A classification of pa- laearctic habitats. Council of Europe, Strasbourg. 1996. Eriksson, O., 1996: Regional dynamics of plants: a review of evidence for remnant, source–sink and metapopulations. Oikos 77, 248–258. Eriksson, O., Cousins, S. A. O., Bruun, H. H., 2002: Land use history and fragmentation of traditionally managed grasslands in Scan- dinavia. Journal of Vegetation Science 3, 743–748. ESRI, 2010: ArcGIS Desktop, Release 9.3, Redlands, CA: Envi- ronmental Systems Research Institute. Fahrig, L., 2003: Effects of habitat fragmentation on biodiversity. Annual Review of Ecology, Evolution, and Systematics 34, 487–515. Fattorini, S., Borges, P. A. V., 2012: Species-area relationship un- derestimate extinction rates. Acta Oecologica 40, 27–30. Forman, R. T. T., 1995: Some general principles of landscape and regional ecology. Landscape Ecology 10, 133–142. Goodall, D. W., 1954: Objective methods for the classification of vegetation. III. An essay in the use of factor analysis. Austra- lian Journal of Botany 2/3, 304–324. PAUŠIČ I., IVAJNŠIČ D., KALIGARIČ M., PIPENBAHER N. 118 ACTA BOT. CROAT. 76 (2), 2017 GURS 2010: Ministry of Infrastructure, Surveying and Mapping Authority of the Republic of Slovenia, Orthophoto and aerial pictures 2009–2010 (in Slovenian). Gustavsson, E., Lennartsson, T., Emanuelsson, M., 2007: Land use more than 200 years ago explains current grassland plant diversity in a Swedish agricultural landscape. Biological Con- servation 138, 47–59. Hanski, I., 1999: Habitat connectivity, habitat continuity, and metapopulations in dynamic landscapes. Oikos 87, 209–219. Hanski, I., 2005: The shrinking world: ecological consequences of habitat loss. International ecology institute, Oldendorf / Luhe, Germany. Helm, A., Hanski, I., Pärtel, M., 2006: Slow response of plant spe- cies richness to habitat loss and fragmentation. Ecology Let- ters 9, 72–77. Hill, M.O., 1979: TWINSPAN-a Fortran program for arranging multivariate data in an ordered two-way table by calssification of the individuals and attributes. Ecology and Systematics, Cornell University, New York. Hylander, K., Ehrlén, J., 2013: The mechanisms causing extinc- tion debts. Trends in Ecology and Evolution 28, 341–346. Janišova, M., Michalcova, D., Bacaro, G., Ghisla, A., 2014: Land- scape effects on diversity of semi-natural grasslands. Agricul- ture Ecosystems and Environment 182, 47–58. Jogan, N., Kaligarič, M., Leskovar, I., Seliškar, A., Dobravec, J., 2004: Slovenian habitat types typology 2004. Ministrstvo za okolje, prostor in energijo, Agencija Republike Slovenije za okolje (in Slovenian). Jonsen, I. D., Fahrig, L., 1997: Response of generalist and special- ist insect herbivores to landscape spatial structure. Landscape Ecology 12, 185–197. Kaligarič, M., Sedonja, J., Šajna, N., 2008: Traditional agricultur- al landscape in Goričko Landscape Park (Slovenia): distribu- tion and variety of riparian stream corridors and patches. Landscape and Urban Planning 85, 71–78. Kiviniemi, K., Eriksson, O., 2002: Size-related deterioration of semi-natural grassland fragments in Sweden. Diversity and Distributions 8, 21–29. Kiviniemi, K., 2008: Effects of fragment size and isolation on the occurrence of four short-lived plants in semi-natural grass- lands. Acta Oecologica 33, 56–65. Kollmann, J., Schneider, B., 1999: Landscape structure and diver- sity of fleshy-fruited species at forest edges. Plant Ecology 144, 37–48. Krauss, J., Klein, A.M., Dewenter, I.S., Tscharntke, T., 2004: Ef- fects of habitat area, isolation, and landscape diversity on plant species richness of calcareous grasslands. Biodiversity and Conservation 13, 1427–1439. Kuussaari, M., Bommarco, R., Heikkinen, R. K., Helm, A., Krauss, J., Lindborg, R., Öckinger, E., Pärtel, M., Pino, J., Rodà, F., Stefanescu, C., Teder, T., Zobel, M., Steffan-Dewenter, I., 2009: Extinction debt: a challenge for biodiversity conserva- tion. Trends in Ecology and Evolution 24, 564–571. Lindborg, R., Cousins, S.A.O., Eriksson, O., 2005: Plant species response to land use change- Campanula rotundifolia, Primu- la veris and Rhinanthus minor. Ecography 28, 29–36. Lindborg, R., 2007: Evaluating the distribution of plant life-histo- ry traits in relation to current and historical landscape configu- rations. Journal of Ecology 95, 555–564. Lomolino, M. V., Riddle, B. R., Whittaker, R. J., Brown, J. H., 2010: Biogeography, fourth ed. Sinauer Associates, Sunder- land. Lopez-Marino, A., Luis-Calabuig, E., Fillat, F., Bermudez, F.F., 2000: Floristic composition of established vegetation and the soil seed bank in pasture communities under different tradi- tional management regimes. Agriculture, Ecosystems and En- vironment 78, 273–282. Luoto, M., Pykäla, J., Kuusaari, M., 2003: Decline of landscape- scale habitat and species diversity after the end of cattle graz- ing. Journal for Nature Conservation 11, 171–178. Martinčič, A., Wraber, T., Jogan, N., Podobnik, A., Turk, B., Vreš, B., Ravnik, V., Frajman, B., Strgulc Krajšek, S., Trčak, B., Bačič, T., Fischer, M.A., Eler, K, Surina, B., 2007: Flora of Slovenia in brief. Identification key for the ferns and flower- ing plants. Tehniška založba Slovenije, Ljubljana (in Slove- nian). Metzger, J. P., 2000: Tree functional group richness and landscape structure in a Brazilian tropical fragmented landscape. Eco- logical Applications 10, 1147–1161. Ovaskainen, O., 2002: The effective size of a metapopulation liv- ing in a heterogeneous patch network. The American Natural- ist 160/5, 612–628. Paušič, I., Kaligarič, M., 2015: Dry grassland land use treatment regime explains the occurrence of the green winged orchid, Anacamptis morio (L.) R.M. Bateman, Pridgeon & M.W. Chase in the Goričko nature park, NE Slovenia. Folia Biologi- ca et Geologica 56/3, 137–148. R Development Core Team, 2008: R: A language and environment for statistical computing. R Foundation for Statistical Com- puting, Vienna, Austria. Reitalu, T., Johansson, L. J., Sykes, M. T., Hall, K., Prentice, H. C., 2010: History matters: village distances, grazing and grassland species diversity. Journal of Applied Ecology 47, 1216–1224. Ricketts, T. H., 2001: The matrix matters: effective isolation in fragmented landscapes. The American Naturalist 158, 87–99. Ricklefs, R. E., Lovette, I. J., 1999: The role of island area per se and habitat diversity in the species–area relationship of four lesser Antillean faunal groups. Journal of Animal Ecology 68, 1142–1160. Rico, Y., Boehmer, H. J., Wagner, H. H., 2011: Determinants of actual functional connectivity for calcareous grassland com- munities linked by rotational sheep grazing. Landscape Ecol- ogy 27, 199–209. Rosenzweig, M. L., 1995: Species diversity in space and time. Cambridge University Press, Cambridge, UK. Shannon, C. E., Weaver, W., 1963: The mathematical theory of communication. University of Illinois Press, Urbana, Söderström, B., Svensson, B., Vessby, K., Glimskӓr, A., 2001: Plants, insects and birds in semi-natural pastures in relation to local habitat and landscape factors. Biodiversity and Conser- vation 10, 1839–1863. Steinmann, K., Eggenberg, S., Wohlgemuth, T., Linder, H.P., Zim- mermann, N. E., 2011: Niches and noise—Disentangling hab- itat diversity and area effect on species diversity. Ecological Complexity 8, 313–319. Sutcliffe, L. M. E., Batáry, P., Kormann, U., Báldi, A., Dicks, L. V., Herzon, I., Kleijn, D., Tryjanowski, P., Apostolova, I., Ar- lettaz, R., Aunins, A., Aviron, S., Baležentienè, L., Fischer, C., Halada, L., Hartel, T., Helm, A., Hristov, I., Jelaska, S.D., Kaligarič, M., Kamp, J., Klimek, S., Koorberg, P., Kosti- uková, J., Kovács-Hostyánszki, A., Kuemmerle, T., Leusch- ner, C., Lindborg, R., Loos, J., Maccherini, S., Marja, R., Máthé, O., Paulini, I., Proença, V., Rey-Benayas, J., Sans, F. X., Seifert, C., Stalenga, J., Timaeus, J., Török, P., van Swaay, SPAtIAL PAttERN of DRy GRASSLAND fRAGMENtS ACTA BOT. CROAT. 76 (2), 2017 119 C., Viik, E., Tscharntke, T., 2014: Harnessing the biodiversity value of Central and Eastern European farmland. Diversity and Distributions 21, 722–730. Škornik, S., 2003: Dry grasslands of the Brometalia erecti Koch (1926) order in Goričko region (NE Slovenia). Hacquetia 2, 71–90. ter Braak, C.J.F., Šmilauer, P., 2002: CANOCO reference manual and CanoDraw for Windows user's guide: software for canon- ical community ordination (version 4.5). Microcomputer Power (Ithaca NY, USA). Tjørve, E., 2010: How to resolve the SLOSS debate: lessons from species-diversity models. Journal of Theoretical Biology 264, 604–612. Tscharntke, T., Steffan-Dewenter, I., Kruess, A., Thies, C., 2002: Contribution of small habitat fragments to conservation of in- sect communities of grassland–cropland landscapes. Ecologi- cal Applications 12, 354–363. Tscharntke, T., Tylianakis, J. M., Rand, T. A., Didham, R.K., Fah- rig, L., Batáry, P., Bengtsson, J., Clough, Y., Crist, T. O., Dor- mann, C. F., Ewers, R. M., Fründ, J., Holt, R. D., Holzschuh, A., Klein, A. M., Kleijn, D., Kremen, C., Landis, D. A., Laur- ance, W., Lindenmayer, D., Scherber, C., Sodhi, N., Steffan- Dewenter, I., Thies, C., van der Putten, W. H., Westphal, C., 2012: Landscape moderation of biodiversity patterns and pro- cesses – eight hypotheses. Biological Reviews 87, 661–685. Warren, M. S., Hill, J. K., Thomas, J. A., Asher, J., Fox, R., Hunt- ley, B., Roy, D. B., Telfer, M. G., Jeffcoate, S., Harding, P., Jeffcoate, G., Willis, S. G., Greatorex-Davies, J. N., Moss, D., Thomas, C. D., 2001: Rapid responses of British butterflies to opposing forces of climate and habitat change. Nature 414, 65–69. Wiens, J. A., 1997: Metapopulation dynamics and landscape ecol- ogy. In: Hanski, I., Gilpin, M., (ed.), Metapopulation biology: ecology, genetics and evolution. Academic Press, San Diego, CA, 43– 62. Wissman, J., 2006: Grazing regimes and plant reproduction in semi-natural grasslands. Doctoral thesis, Department of con- servation biology, Swedish University of agricultural scienc- es, Uppsala. Acta Universitatis Agriculturae Sueciae 40, 13– 20. ZRSVN – Institute of the Republic of Slovenia for Nature Conser- vation. Retrived on May 2014 from http://www.zrsvn.si/sl/in- formacija.asp?id_meta_type=62&id_ informacija=705. Zulka, K. P., Abensperg-Traun, M., Milasowszky, N., Bieringer, G., Gereben-Krenn, B. A., Holzinger, W., Hӓlzler, G., Ra- bitsch, W., Reischütz, A., Querner, P., Sauberer, N., Schmitz- berger, I., Willner, W., Wrbka, T., Zechmeister, H., 2014: Spe- cies richness in dry grassland patches of eastern Austria: A multi-taxon study on the role of local, landscape and habitat quality variables. Agriculture, Ecosystems and Environment 182, 25–36.