Urban Forestry & Urban Greening 24 (2017) 101–108 Contents lists available at ScienceDirect Urban Forestry & Urban Greening journal homepage: www.elsevier.com/locate/ufug Floristic diversity, composition and invasibility of riparian habitats with Amorpha fruticosa: A case study from Belgrade (Southeast Europe) MARK Nataša Radovanović, Nevena Kuzmanović, Snežana Vukojičić, Dmitar Lakušić, ⁎ Slobodan Jovanović University of Belgrade, Faculty of Biology, Institute of Botany and Botanical Garden, Takovska 43, 11000 Belgrade, Serbia A R T I C L E I N F O A B S T R A C T Keywords: City forests Flooded lowland meadows Invasive neophytes Invasibility-coverage index relations Species richness Willow and poplar forests Amorpha fruticosa L. represents one of the most dangerous invasive neophytes spreading quickly in many countries and cities of southeastern Europe where it aggressively penetrates into newly invaded sites and establishes permanently. It prefers moist and periodically flooded terrains, being therefore a serious threat for fragile wet habitats. Considering this, the main aim of this research was to determine the floristic diversity, composition and level of invasibility of urban and suburban riparian forests and open habitats with domination of A. fruticosa at the mouth of the Sava and Danube Rivers in Belgrade, and to assess the impact of all invasive neophytes in the analyzed habitats. Two hundred fifty seven (257) relevés, made according to Braun-Blanquet (1964) methodology, were subjected to different statistical analyses. The obtained results showed that urban wet habitats with domination of A. fruticosa were differentiated into 7 coenological groups, with the total of 222 registered taxa, out of which 29 (13.06%) were invasive neophytes. These coenoses are developed within willow and poplar habitats, wet lowland meadows and reedbed habitats. We found a direct negative correlation between the change in the number of species and the proportion of invasive species i.e. their coverage indexes in the analysed stands. The identified coenological group with domination of Rubus caesius and A. fruticosa represents the most dangerous hotspot of invasive species, which might seriously threaten native species and their urban riparian habitats, as well as similar habitats downstream. 1. Introduction It is well-known that river floodplains are among the most threatened habitats (Pyšek and Prach, 1994; Hood and Naiman, 2000; Schnitzler et al., 2007). Namely, within the group of the plant species that grow almost exclusively in the corridors of large rivers (river corridor plants), we can find a high proportion of threatened species (Burkart, 2001). However, rivers also transport vegetative parts and seeds of some hydrophilic invasive plants, which can develop very quickly in the fertile riparian zones (Gallé et al., 1995; Säumel and Kowarik, 2010; Pedashenko et al., 2012). In this sense, the Danube with its characteristics is absolutely one of the most important routes for spreading these species in Europe (Pedashenko et al., 2012). Amorpha fruticosa L. (false indigo or indigo bush) is a deciduous shrub which originates from central and eastern part of North America and was introduced into Europe in 1724 as an ornamental species. It was brought to the Balkan Peninsula at the beginning of the twentieth century when it started to colonise alluvial forests and other habitats in large lowland river valleys (Gagić-Serdar et al., 2013), seriously ⁎ Corresponding author. E-mail address: sjov@bio.bg.ac.rs (S. Jovanović). http://dx.doi.org/10.1016/j.ufug.2017.04.006 Received 25 October 2016; Received in revised form 7 April 2017; Accepted 8 April 2017 Available online 10 April 2017 1618-8667/ © 2017 Elsevier GmbH. All rights reserved. threatening the ecological balance of native ecosystems (Krpan and Benko, 2009). Although Weber and Gut (2004) assessed that A. fruticosa represents a potentially invasive plant species in central Europe, nowadays it is one of the most dangerous invasive neophytes spreading rapidly in many countries and cities of south-eastern Europe as well (Anastasiu et al., 2007; Grbić et al., 2007; Pedashenko et al., 2012; Anačkov et al., 2013). The false indigo, growing mainly in wet habitats, is becoming very dangerous especially in fragile wet habitats of protected areas (Török et al., 2003; Botta-Dukát and Mihály, 2006; Dumitraşcu et al., 2012; Batanjski et al., 2016), e.g. the Danube Delta, one of the most important Ramsar sites of Europe (Protopopova et al., 2006; Anastasiu et al., 2007). As a semi-aquatic species, A. fruticosa prefers moist and periodically flooded habitats regardless of the level of their degradation (Doroftei, 2009; Anačkov et al., 2013). As it can reproduce both generatively and vegetatively, it is growing faster than most forest-cultural species (Tucović and Isajev, 2000; Gagić-Serdar et al., 2013). Amorpha fruticosa aggressively penetrates into newly invaded sites, where it establishes permanently (Radulović et al., 2008). Urban Forestry & Urban Greening 24 (2017) 101–108 N. Radovanović et al. Riparian forests in urban areas, as well as different open wet habitats, play a significant environmental role in many cities which lie on banks of large rivers. In addition to many human-made influences which threaten these fragile ecosystems, presence and spread of invasive species, as a secondary consequence, may limit their basic functions, ecosystem services and role in biodiversity conservation. The investigations and knowledge about spread and effects of invasive plants, especially woody species, as well as management plans for invaders are crucial in sustainable urban forestry (Alvey, 2006; Dyderski et al., 2015). The main aims of this study were: 1) to determine the floristic diversity and composition of stands with domination of Amorpha fruticosa in urban riparian forests and open habitats at the mouth of the Sava and Danube Rivers in Belgrade, 2) to assess the impact of all invasive neophytes on floristic diversity in the analysed habitats, and finally 3) to determine the level of invasibility of the stands with domination of A. fruticosa. 2.2. Vegetation sampling The phytosociological investigations at the confluence of the Sava and Danube Rivers were carried out in the period 2011–2013, during the summer (from the middle of June until the end of August). Altogether two hundred fifty-seven (257) relevés were made according to Braun-Blanquet (1964) methodology (Fig. 1). The sites were selected systematically, with the aim to cover the invaded areas, as well as different habitats in which A. fruticosa dominates. The size of the sampling plots was c. 25 square metres, whereas the minimal distance between two studied sites of the same habitat types was c. 100 m. Only in the cases of sites with habitat patches narrower than 5 m, the sample plot corresponded to a rectangle, whose metrics depended on the patch shape. In order to determine the invasibility of stands with Amorpha fruticosa, we extracted a subset of 173 relevés in which this species had the highest cover-abundance values (8 and 9 in Van der Maarel scale). This subset was subjected to further analyses. All field data were georeferenced using a GPS device eTrex VistaC (Garmin). The collected plant material was deposited in the Herbarium of the University of Belgrade − BEOU (Thiers 2016). 2. Materials and methods 2.1. Study area 2.3. Data analysis The city of Belgrade and its surroundings lies on the border between the Pannonian plain and the Balkans-Šumadija region (44°48′52” – 44°50′67” N and 20°31′68” – 20°37′22” E), at the confluence of the Danube and Sava Rivers (Fig. 1). The city territory covers an area of 322.268 ha divided into 17 municipalities where 1.689.000 inhabitants or 22.5% of the total Serbian population live (Jovanović et al., 2014). Its geographic location gives it a mild-continental climate with an average annual temperature of 12.7 °C and annual precipitation of 750 mm (Đurđić et al., 2011). Regarding the phytogeographical affiliation, Belgrade lies on the border of two phytogeographical regions: 1. Pontic-South-Siberian floristic-vegetation region and 2. Middle European region-Balkan subregion-Moesian province (Jovanović, 1994). The investigated area occupies four urban municipalities: Čukarica (Makiš), Zemun (Veliko Ratno ostrvo and Zemunski kej), Novi Beograd (the Sava embankment) and Palilula (Ada Huja, Višnjica, Veliko Selo, Krnjača, Borča, Crvenka, fishpond Mika Alas and puddle Reva). The nomenclature source for the names of vascular plants was the Flora Europaea (Tutin et al., 1968–1980; Tutin et al., 1993). Life forms of plants were determined according to Raunkier (1934). Terminology used in this paper is in accordance with Pyšek et al. (2004). Under the term “invasive plants” here we considered “alien plants that produce reproductive offspring, often in very large numbers, at considerable distances from the parent plants, and thus have the potential to spread over a large area”, whereas under the term “invasive neophytes” we considered “invasive alien plants introduced after ca. 1500”. Furthermore, the invasive status of registered plants is defined in accordance with the relevant sources for the Serbian region published by Lazarević et al. (2012) and Anačkov et al. (2011–2013). Prior to numerical analyses, we transformed Braun–Blanquet combined alpha-numeric scale into numerical scale as proposed by Van Der Maarel (1979). Groups of vegetation types were ascertained using cluster analyses in the programme package PAST (Hammer et al., Fig. 1. Map of Amorpha fruticosa stands (black relevés points) in the study area of Belgrade (Serbia). 102 Urban Forestry & Urban Greening 24 (2017) 101–108 N. Radovanović et al. group C) (Table 2). A post-hoc test showed that most of the differences in diversity indices between the obtained groups were not statistically significant (Table 1). 2001). The arrangement of relevés was done according to the results of the cluster analysis, and as a non-parametric test of significance of differences between the identified groups of relevés we used ANOSIM (analysis of similarity) in the programme package PAST (Hammer et al., 2001). The significance was computed by permutation of group membership, with 10.000 replicates. The results were considered as significant if the probability of the null hypothesis was less than 0.05. We used SIMPER (Similarity Percentage), an algorithm implemented in the programme package PAST (Hammer et al., 2001), for assessing which taxa are primarily responsible for observed differences, as well as to measure the level of dissimilarities between the groups obtained in cluster analysis. Furthermore, for each of these groups, we determined the number of invasive neophytes per group, the percentage of invasive neophytes, coverage index according to Surina (D%, Surina, 2004), as well as the following parameters of diversity − total species number, heterogeneity index (logS/logA) and Shannon-Wiener diversity index (H). We employed Tukey’s HSD post-hoc test for unequal group sizes in order to test if the obtained differences in diversity index (ShannonWiener diversity index H) among the groups are statistically significant. Finally, in order to test the effects of coverage index of invasive neophytes on the diversity of analysed stands (expressed through the total species number), we used generalized linear model (GLM). A Poisson probability distribution with a log link function was used (Crawley, 1993). The statistical significance was tested by Monte Carlo permutation test with 999 runs. GLM was performed by means of CANOCO 5 (Ter Braak and Šmilauer, 2012). 3.1. Floristic diversity and composition of stands with domination of Amorpha fruticosa Cluster A (Vitis vulpina-Populus nigra group) consisted of 19 relevés in which 65 species were recorded. The most frequent and the most abundant plants were Amorpha fruticosa (Fr% = 100, Mean abund. 8.68), Vitis vulpina (Fr% = 89, Mean abund. 6.68), Rubus caesius (Fr% = 79, Mean abund. 5.26) and Populus nigra (Fr% = 42, Mean abund. 3.58). Other coenologically important plants were Acer negundo (Fr% = 37, Mean abund. 1.84), Echinocystis lobata (Fr% = 32, Mean abund. 1.79), Fraxinus pennsylvanica (Fr% = 32, Mean abund. 1.58), Echinochloa crus-galli (Fr% = 26, Mean abund. 0.58), Bidens tripartita (Fr% = 21, Mean abund. 0.74), Elymus repens (Fr% = 21, Mean abund. 0.95) and Aristolochia clematitis (Fr% = 21, Mean abund. 0.58). From the total of 65 recorded species, 14 were invasive (21.54%) prevailing in coverage (D% = 55.66) (Tables 2 and 3). In number of vascular plant taxa, hemicryptophytes predominate (37.5%), followed by therophytes (28.13%) and phanerophytes (18.75%) (Table 4). Cluster B (Rubus caesius-Amorpha fruticosa group) consisted of 55 relevés in which 135 species were recorded. The most frequent and the most abundant plants were Amorpha fruticosa (Fr% = 100, Mean abund. 8.62), Rubus caesius (Fr% = 96, Mean abund. 6.93) and Aster lanceolatus (Fr% = 65, Mean abund. 3.07). Other coenologically important plants were Calystegia sepium (Fr% = 38, Mean abund. 1.35) and Galium aparine (Fr% = 31, Mean abund. 1.49). From the total number of recorded species, 21 were invasive neophytes (15.56%, D% = 42.47, Tables 2 and 3), while native and other non-indigenous species prevail in number and coverage (114 taxa-84.44%, D% = 57.53). In the number of vascular plant taxa, hemicryptophytes predominate (42.54%), followed by therophytes (27.61%) and phanerophytes (17.16%) (Table 4). Cluster C (Aster lanceolatus-Amorpha fruticosa group) consisted of 35 relevés in which 113 species were recorded. The most frequent and the most abundant plants were Amorpha fruticosa (Fr% = 100, Mean abund. 8.77), Aster lanceolatus (Fr% = 97, Mean abund. 6.83) and Elymus repens (Fr% = 54, Mean abund. 2.71). Other coenologically important plants were Acer negundo (Fr% = 29, Mean abund. 1.40), Equisetum arvense (Fr% = 23, Mean abund. 1.00), Rubus caesius (Fr% = 23, Mean abund. 0.80), Galium aparine (Fr% = 20, Mean abund. 0.83), Echinocystis lobata (Fr% = 20, Mean abund. 0.57) and Polygonum aviculare (Fr% = 20, Mean abund. 0.43). From the total number of recorded species, 20 were invasive neophytes (17.70%) that predominated in coverage (D% = 52.75) (Tables 2 and 3). Like in previous groups, hemicryptophytes predominate (45.95%), followed by therophytes (24.32%) and phanerophytes (19.82%) (Table 4). Cluster D (Salix alba-Amorpha fruticosa group) consisted of 17 3. Results Within the 173 analysed relevés, 222 taxa of vascular plants were recorded. Hemicryptophytes with 94 taxa completely prevailed representing 42.34% of all registered plants (D% = 25.28), followed by therophytes (24.77%, D% = 14.21), and nanophanerophytes (8.56%, D % = 32.10). Cluster analysis revealed seven groups of relevés (Fig. 2, groups AG). SIMPER analysis showed that overall average dissimilarity between these seven groups was 69.83%, with moderate to relatively high values of dissimilarities between pairs of groups (62.53–77.86% − Table 1). Taxa that contributed most to the obtained differences were Rubus caesius (7.46%), Aster lanceolatus (6.42%), Elymus repens (4.83%), Vitis vulpina (3.17%), Calystegia sepium (2.64%), Salix alba (2.63%), Galium aparine (2.51%), Populus alba (2.41%), Acer negundo (2.27%), Fraxinus pennsylvanica (2.06%), Phragmites australis (1.89%) and Populus nigra (1.73%). The analysis of similarity (ANOSIM) showed statistically significant differences between almost all the obtained groups (Table 1), yielding mostly moderate to high values of R (0.24–0.95).The heterogeneity index (logS/logA) varied between 1.29 and 1.50 (the highest value was for the group D, and the lowest for the group F), while Shannon-Wiener diversity index (H) varied between 1.55 and 2.02 (the highest value was for the group D, the lowest for the Fig. 2. The result of cluster analysis (Ward’s method, Euclidean distances), with identified seven groups of relevés (groups A-G). 103 Urban Forestry & Urban Greening 24 (2017) 101–108 N. Radovanović et al. Table 1 Dissimilarities between the groups (A–G) obtained in cluster analysis. Dissimilarity percentages from SIMPER are in the upper right hand corners, while in the lower left hand corners are Bonferroni-corrected p values from ANOSIM. *Statistically significant differences in diversity index (H) between the groups (p < 0.05). A B C D E F G SIMPER%/ANOSIM Bonferroni-corrected p values A B C D E F G Vitis vulpina-Populus nigra Rubus caesius-Amorpha fruticosa Aster lanceolatus-Amorpha fruticosa Salix alba-Amorpha fruticosa Solidago gigantea-Phragmites australis Populus alba-Amorpha fruticosa Elymus repens-Amorpha fruticosa x 0.002 0.002 0.002 0.002 0.002 0.002 69.83 x 0.002 0.002 0.002 0.002 0.002 67.12 65.75 x 0.002 0.002 0.294 0.002* 71.75 70.31 70.58 x 0.002 0.0021 0.002* 77.86 75.08 74.38 77.81 x 0.0021 0.032 70.61 68.47 62.53 69.68 73.66 x 1 74.15 73.05 70.16 74.63 72.39 69.86 – Table 2 Basic parameters of floristic diversity for the groups obtained in cluster analysis (No. – number; D% – coverage index according to Surina, 2004). Group Group name No. of relevés No. of species heterogeneity index (logS/ logA) Shannon-Wiener diversity index (H) No. of invasive species % of invasive species D% of invasive species A B Vitis vulpina-Populus nigra Rubus caesius-Amorpha fruticosa Aster lanceolatusAmorpha fruticosa Salix alba-Amorpha fruticosa Solidago giganteaPhragmites australis Populus alba-Amorpha fruticosa Elymus repens-Amorpha fruticosa 19 55 65 135 1.48 1.47 1.78 1.84 14 21 21.54 15.56 55.66 42.47 35 113 1.43 1.55 20 17.70 52.76 17 57 1.50 2.02 13 22.81 40.56 12 56 1.42 1.80 7 12.50 31.42 8 60 1.29 1.96 12 20.00 34.37 27 110 1.39 2.01 17 15.45 32.21 C D E F G therophytes (23.64%) and geophytes (16.36%) (Table 4). Cluster F (Populus alba-Amorpha fruticosa group) consisted of 8 relevés in which 60 species were recorded. The most frequent and the most abundant plants were Amorpha fruticosa (Fr% = 100, Mean abund. 8.63), Populus alba (Fr% = 100, Mean abund. 7.00) and Elymus repens (Fr% = 88, Mean abund. 5.00). Other coenologically important plants were Aster lanceolatus (Fr% = 75, Mean abund. 4.00), Rubus caesius (Fr% = 63, Mean abund. 1.38), Lactuca serriola (Fr% = 63, Mean abund. 1.13), Cirsium arvense (Fr% = 38, Mean abund. 0.75), Chenopodium album (Fr% = 38, Mean abund. 0.75), Acer negundo (Fr% = 38, Mean abund. 0.75), Trifolium pratense (Fr% = 25, Mean abund. 1.25), Equisetum arvense (Fr% = 25, Mean abund. 1.00), Aristolochia clematitis (Fr% = 25, Mean abund. 0.88), Calystegia sepium (Fr% = 25, Mean abund. 0.50), Setaria viridis (Fr% = 25, Mean abund. 0.50), Melilotus alba (Fr% = 25, Mean abund. 0.50), Dipsacus fullonum (Fr% = 25, Mean abund. 0.50), Xanthium strumarium subsp. italicum (Fr % = 25, Mean abund. 0.50), Sonchus oleraceus (Fr% = 25, Mean abund. 0.50), Plantago major (Fr% = 25, Mean abund. 0.50), Artemisia vulgaris (Fr% = 25, Mean abund. 0.50), Erigeron annuus (Fr % = 25, Mean abund. 0.50), Crepis setosa (Fr% = 25, Mean abund. 0.50) and Matricaria perforata (Fr% = 25, Mean abund. 0.38). From the total number of recorded species, 12 were invasive neophytes (20.00%) having the coverage index D% = 34.37 (Tables 2 and 3). Like in some of the previous groups, hemicryptophytes predominated (41.07%), followed by therophytes (35.71%) and phanerophytes (12.50%) (Table 4). The last group – cluster G (Elymus repens-Amorpha fruticosa group) consisted of 27 relevés in which 110 species were recorded. The most frequent and the most abundant plants were Amorpha fruticosa (Fr% = 100, Mean abund. 8.74), Elymus repens (Fr% = 70, Mean abund. 3.30) and Aster lanceolatus (Fr% = 52, Mean abund. 1.44). Other coenologically important plants were Convolvulus arvensis (Fr% = 41, Mean abund. 1.33), Sorghum halepense (Fr% = 33, Mean abund. 1.07), Carduus acanthoides (Fr% = 30, Mean abund. 0.89), Bromus sterilis (Fr % = 26, Mean abund. 1.04), Galium aparine (Fr% = 26, Mean abund. 1.11), Cornus sanguinea (Fr% = 22, Mean abund. 1.30), Aristolochia relevés in which 57 species were recorded. The most frequent and the most abundant plants were Amorpha fruticosa (Fr% = 100, Mean abund. 8.65), Salix alba (Fr% = 71, Mean abund. 5.53) and Aster lanceolatus (Fr% = 53, Mean abund. 1.59). Other coenologically important plants were Xanthium strumarium subsp. italicum (Fr% = 47, Mean abund. 1.82), Bidens tripartita (Fr% = 47, Mean abund. 1.65), Chenopodium album (Fr% = 47, Mean abund. 1.29), Echinochloa crusgalli (Fr% = 35, Mean abund. 1.41), Calystegia sepium (Fr% = 29, Mean abund. 1.12), Polygonum lapathifolium (Fr% = 29, Mean abund. 1.12), Rubus caesius (Fr% = 29, Mean abund. 0.941), Amaranthus retroflexus (Fr% = 29, Mean abund.0.65), Amaranthus lividus (Fr% = 24, Mean abund. 1.29) and Solanum nigrum (Fr% = 24, Mean abund. 0.82). From the total number of recorded species, 22.81% were invasive neophytes (13), having coverage index D% = 40.56 (Tables 2 and 3). Hemicryptophytes and therophytes shared the equal portions (30.91%), followed by phanerophytes (23.64%) (Table 4). Cluster E (Solidago gigantea-Phragmites australis group) consisted of 12 relevés in which 56 species were recorded. The most frequent and the most abundant plants were Amorpha fruticosa (Fr% = 100, Mean abund. 8.58), Solidago gigantea (Fr% = 92, Mean abund. 5.25) and Elymus repens (Fr% = 75, Mean abund. 3.08). Other coenologically important plants were Cynodon dactylon (Fr% = 67, Mean abund. 2.33), Phragmites australis (Fr% = 58, Mean abund. 3.92), Calystegia sepium (Fr% = 58, Mean abund. 2.08), Althaea officinalis (Fr% = 58, Mean abund. 1.17), Lolium perenne (Fr% = 50, Mean abund. 1.75), Glycyrrhiza echinata (Fr% = 50, Mean abund. 1.75), Carduus acanthoides (Fr% = 50, Mean abund. 1.67), Lythrum salicaria (Fr% = 50, Mean abund. 1.00), Equisetum arvense (Fr% = 33, Mean abund. 1.33), Calamagrostis epigejos (Fr% = 33, Mean abund. 0.92), Linaria vulgaris (Fr% = 33, Mean abund. 0.67), Setaria viridis (Fr% = 25, Mean abund. 0.75), Aster lanceolatus (Fr% = 25, Mean abund. 0.50), Torilis arvensis (Fr% = 25, Mean abund. 0.42) and Lactuca serriola (Fr% = 25, Mean abund. 0.33). From the total number of recorded species, only seven were invasive neophytes (12.73%), having the coverage index D % = 31.42 (Tables 2 and 3). To the life form of hemicryptophytes belonged almost half of the recorded species (49.10%), followed by 104 Urban Forestry & Urban Greening 24 (2017) 101–108 N. Radovanović et al. Table 3 Invasive species, their frequencies (Fr%) and coverage index according to Surina, 2004 (D%) within the groups obtained in cluster analysis. Taxon Acer negundo L. Acer saccharinum L. Ailanthus altissima (Mill.) Swingle Amaranthus retroflexus L. Ambrosia artemisiifolia L. Amorpha fruticosa L. Asclepias syriaca L. Aster lanceolatus Willd. Bidens frondosa L. Broussonetia papyrifera (L.) Vent. Cirsium arvense (L.) Scop. Conyza canadensis (L.) Cronq. Datura stramonium Thunb. Echinochloa crus-galli (L.) Beauv. Echinocystis lobata (Michx) Torrey & A. Gray Eleusine indica (L.) Gaertner Erigeron annuus (L.) Pers. Fraxinus americana L. Fraxinus pennsylvanica Marshall Oenothera biennis L. Panicum miliaceum L. Portulaca oleracea L. Reynoutria japonica Houtt. Robinia pseudoacacia L. Solidago canadensis L. Solidago gigantea Aiton Sorghum halepense (L.) Pers. Vitis vulpina L. Xanthium strumarium subsp. italicum (Moretti) D. Löve Total D% per group Vitis vulpinaPopulus nigra Group A Rubus caesiusAmorpha fruticosa Group B Aster lanceolatusAmorpha fruticosa Group C Salix albaAmorpha fruticosa Group D Solidago giganteaPhragmites australis Group E Populus albaAmorpha fruticosa Group F Elymus repensAmorpha fruticosa Group G Fr% D% Fr% D% Fr% D% Fr% D% Fr% Fr% D% Fr% D% 37 4.04 13 13 1.28 1.92 29 3.51 18 6 0.59 0.29 38 1.55 3 0.07 4 0.08 3 100 11 97 0.14 21.98 1.43 17.11 0.16 0.72 18.96 0.72 3.13 3 0.14 11 100 0.46 19.05 4 9 100 0.14 0.82 21.62 74 21 6.47 0.58 65 13 7.71 0.73 5 11 0.58 0.46 7 4 0.36 0.18 17 9 1 0.43 26 1.27 16 1.05 9 0.86 32 3.93 13 0.78 20 1.43 0.5 3.46 0.09 0.32 0.09 2.55 6 0 32 2 7 2 15 11 1.22 3 0.36 5 0.23 0.09 0.23 3 0.14 0.32 0.73 0.73 0.73 3 3 17 6 9 0.57 0.14 1.07 0.29 0.36 2 2 5 89 5 4 15 7 16 0.23 14.67 0.23 55.66 42.47 P H G Hyd NP T A B Vitis vulpina-Populus nigra Rubus caesius-Amorpha fruticosa Aster lanceolatus-Amorpha fruticosa Salix alba-Amorpha fruticosa Solidago gigantea-Phragmites australis Populus alba-Amorpha fruticosa Elymus repens-Amorpha fruticosa 18.75 17.16 37.50 42.54 10.94 10.45 3.13 2.24 1.56 2.24 28.13 27.61 19.82 45.95 5.41 1.80 2.70 24.32 23.64 7.27 30.91 49.09 9.09 16.36 3.64 3.64 1.82 0.00 30.91 23.64 12.50 41.07 7.14 1.79 1.79 35.71 16.51 42.20 9.17 2.75 2.75 26.61 G 3.95 0.59 100 16.86 100 17.83 25 0.98 75 8.27 4 11 100 11 52 17 8 1.64 0.16 38 13 1.55 0.52 22 11 1.45 1.12 13 13 0.52 0.52 7 0.56 25 1.03 11 0.48 11 0.72 4 0.24 4 0.4 7 33 4 4 0.56 2.33 0.4 0.16 0.29 3.51 8 18 18 47 0.82 0.73 0.88 92 17 52.75 Group name F 53 47 18 Group D E 1.61 0.88 21.52 6 35 10.31 0.65 1.17 4.54 40.55 31.42 13 0.52 13 0.52 13 0.52 25 1.03 34.38 32.19 3.2. Invasibility of stands with domination of Amorpha fruticosa Table 4 Life form spectra for the groups obtained in cluster analysis (%). Life forms: P – Phanerophytes, H – Hemicryptophytes, G – Geophytes, Hyd – Hydrophytes, NP – Nanophanerophytes, T – Therophytes. C 29 18 100 D% From the total of 222 taxa registered within the analysed relevés, 29 taxa were invasive (13.06%). The number of invasive neophytes per relevés varied between one and eight, with the average value of three species per relevé. Regarding coenological groups obtained in the cluster analysis, the number of invasive taxa varied between seven in group E (Solidago gigantea-Phragmites australis) and 21 in group B (Rubus caesius-Amorpha fruticosa – Table 2). Although all invasive taxa made up only 13.06% of the total flora, they had high total coverage index of D % = 42.72. Coverage index D% per relevés varied between 18 and 100, with average value D% = 45 per relevé. Within the coenological groups, coverage index D% varied between 31.42 in group E (Solidago gigantea-Phragmites australis) and 55.66 in group A (Vitis vulpina-Populus nigra – Table 2). In addition to Amorpha fruticosa which dominated in all the investigated stands, other invasive neophytes with high frequency in the analysed relevés were Aster lanceolatus (63.0%), Acer negundo (19.3%), Vitis vulpina (17.7%), Solidago gigantea (17.0%), Cirsium arvense (15.1%), Echinochloa crus-galli (15.1%), Xanthium strumarium subsp. italicum (15.1%), Fraxinus pennsylvanica (12.4%), Sorghum halepense (12.4%), Bidens frondosa (11.6%) and Echinocystis lobata (10.4%). Regarding the abundance expressed through the coverage index D %, absolutely dominant species was Amorpha fruticosa. In addition to A. clematitis (Fr% = 22, Mean abund. 0.93), Cichorium intybus (Fr% = 22, Mean abund. 0.37), Lactuca serriola (Fr% = 22, Mean abund. 0.63) and Cirsium arvense (Fr% = 22, Mean abund. 0.67). From the total number of recorded species, 17 were invasive neophytes (15.45%) having the coverage index D% = 32.21 (Tables 2 and 3). Again hemicryptophytes predominated (42.20%), followed by therophytes (26.61%) and phanerophytes (16.51%) (Table 4). 105 Urban Forestry & Urban Greening 24 (2017) 101–108 N. Radovanović et al. b) coenoses developed in the wet lowland meadows (C, G) and c) coenoses developed within the tall reedbed habitats (E). Additionally, we distinguished three groups within the coenoses in willow and poplar habitats, which follow hydric regime gradient of hygrophilous forests developed in the valleys of large lowland rivers. Thus, group D, Salix alba-Amorpha fruticosa, corresponded to communities developed within the white willow habitats (Salicion albae Soó (1930) 1940); group A–Vitis vulpina-Populus nigra corresponded to communities developed in the black poplar habitats, while group F (Populus alba-Amorpha fruticosa) to communities developed on the white poplar habitats. The latter two can be phytosociologically classified as belonging to the alliance Populion albae Br.-Bl. 1931. Group B Rubus caesius-Amorpha fruticosa, within which stands the largest number of elements of willow and poplar forests occurred can be interpreted as the final stage of degradation of floodplain willow and poplar forests. Phytosociological interpretation of coenoses developed in the wet lowland meadows is rather difficult. A significant contribution of the species Elymus repens, Rumex crispus and Agrostis stolonifera, especially in group G Elymus repens-Amorpha fruticosa, pointed to the possibility that these stands are developing in habitats that corresponded to the alliance Agropyro-Rumicion crispi Nordh. 1940 of the order Agrostietalia stoloniferae Oberd. 1967. Nevertheless, the species Carex hirta, Mentha aquatica, Poa pratensis, Poa trivialis, Potentilla reptans, Ranunculus repens, Stachys palustris and others, which occurred with significant contribution in both groups of stands (C and G) indicated that part of these stands probably originated from the lowland meadow communities that could be classified in the alliance Trifolion resupinati K. Micevski 1957 of the order Trifolio-Hordeetalia H-ić 1963. Similar observation was published by Radulović et al. (2008), who reported that A. fruticosa builds large populations in meadows of the alliances Agropyro-Rumicion crispi and Trifolion resupinati, regardless of their differences in terms of the impact of flood and ground water. Furthermore, these authors pointed out that A. fruticosa was absent from the communities of the alliance Magnocaricion Br.-Bl, suggesting that further investigation is needed in order to find whether Carex spp. have mechanisms that inhibit the growth of false indigo. Finally, group E Solidago gigantea-Phragmites australis are undoubtedly part of the communities originally belonging to the alliance Phragmition communis W. Koch 1926 of the order Phragmitetalia communis W. Koch 1926. The analyses on our dataset showed that the total number of species per stands significantly decreases with the increase of the total cover of invasive neophytes (Fig. 3). The same results were obtained in similar studies regarding some types of grazing grasslands from western Romania, where a negative correlation was observed between the indigo bush coverage index and the number of the species, as well Shannon-Wiener diversity index (Sărăţeanu, 2010). It is important to emphasise that the highest species diversity was recorded in the group B Rubus caesius-Amorpha fruticosa, which represented the final stage of degradation of floodplain willow and poplar forests. Namely, in this group 135 species were recorded, being approximately twice more than the number of species recorded in white willow (D Salix alba-Amorpha fruticosa = 57 sp.), black poplar (A Vitis vulpina-Populus nigra = 65 sp.) and white poplar forests (F Populus albaAmorpha fruticosa = 60 sp.). This can be explained by the change of ecological conditions, as a result of cutting down the dominant tall tree species – Salix alba, Populus alba and P. nigra – and opening the forest canopy favouring the species of open habitats to colonise these disturbed forest habitats. Although such a secondary increase of biodiversity may seem as a positive effect, the fact is that in this increase a significant contribution have the alien species, especially invasive neophytes (Deutschewitz et al., 2003; Olden and Rooney, 2006; Knapp et al., 2010; Jarošík et al., 2011; Thomas, 2013). If we bear in mind that in the stands of this group, out of the total 29 recorded invasive neophytes 21 are present, it becomes clear that these Fig. 3. Scatterplot of GLM showing the effects of coverage index (D%) of invasive neophytes on the floristic diversity of analysed stands expressed through the total species number. fruticosa, coverage index D% higher than 10, had only Aster lanceolatus (D% = 17.11) in group C – Aster lanceolatus-Amorpha fruticosa, Solidago gigantea (D% = 10.31) in group E – Solidago gigantea-Phragmites australis and Vitis vulpina (D% = 14.67) in group A – Vitis vulpinaPopulus nigra (Table 3). Much lower, but still significant coverage index varying between 2 and 10 in different coenological groups, had Acer negundo, Echinochloa crus-galli, Echinocystis lobata, Fraxinus pennsylvanica, Sorghum halepense and Xanthium strumarium subsp. italicum (Table 3). Finally, although a post-hoc test showed that most of the differences in diversity index (H) between the obtained groups were not statistically significant (Table 1), generalized linear model showed strong influence of invasive species on the species diversity between the analysed stands. Namely, GLM performed on coverage index of invasive neophytes as predictor, showed statistically significant negative correlation between this parameter and species diversity expressed through the total number of species within the analysed stands (p < 0.00001; Fig. 3). 4. Discussion Some previous studies reported that riparian forests were more invaded by alien species in urban than rural areas (e.g. Moffatt et al., 2004; Yong Sung et al., 2011). According to Jovanović et al. (2014), along regulated and mainly ruderalized banks of the Sava and Danube Rivers in Belgrade, numerous invasive neophytes were inhabited in the last 60 years; among them the most aggressive is Amorpha fruticosa. Due to the significant occurrence of false indigo and other invasive neophytes, floristic composition of the investigated urban riparian forests and open stands in Belgrade was considerably changed compared to potential characteristics of habitats within the study area (Jovanović et al., 1984; Jovanović et al., 1985). A similar situation was identified in some grassland and forest habitat types along the lower Danube flow in Romania and Bulgaria, and its delta (Doroftei, 2009; Sărăţeanu, 2010; Pedashenko et al., 2012). However, on the basis of the composition of primarily native species, coenological groups obtained in the cluster analysis can be interpreted as follows: a) coenoses developed within the willow and poplar habitats (groups A, B, D, F); 106 Urban Forestry & Urban Greening 24 (2017) 101–108 N. Radovanović et al. Anačkov, G., Bjelić-Čabrilo, O., Karaman, I., Karaman, M., Radenković, S., Radulović, S., Vukov, D., Boža, P. (Eds.), 2011–2013. Lista invazivnih vrsta na području AP Vojvodine (Invasive Alien Species in AP Vojvodina). Prirodno-matematički fakultet, University of Novi Sad, Novi Sad, URL: http://iasv.dbe.pmf.uns.ac.rs/. Anačkov, G., Rat, M., Radak, B., Igić, R., Vukov, D., Rućando, M., Krstivojević, M., Radulović, S., Cvijanović, D., Milić, D., Panjković, B., Szabados, K., Perić, R., Kiš, A., Stojšić, V., Boža, P., 2013. Alien invasive neophytes of the southeastern part of the Pannonian plain. Cent. Eur. J. Biol. 8, 1032–1043. Anastasiu, P., Negrean, G., Basnou, C., Sîrbu, C., Oprea, A., 2007. A preliminary study on the neophytes of wetlands in Romania. In: Rabitsch, W., Essl, F., Klingenstein, F. (Eds.), Biological Invasions – from Ecology to Conservation 7. NeoBiota, pp. 181–192. Batanjski, V., Batrićević, A., Purger, D., Alegro, A., Jovanović, S., Joldžić, V., 2016. Critical legal and environmental review on Ramsar convention in protection from invasive plant species – an example of the Southern Pannonia region. Int. Environ. Agreem.: Politics Law Econ. 16, 833–848. Botta-Dukát, Z., Mihály, B. (Eds.), 2006. Biológiai Inváziók Magyarországon. Özönnövények II. (Biological Invasion in Hungary. Invasive Plants II). KvVM TvH Tanulmánykötetei 10. Természet BÚVÁR Alapítvány Kiadó, Budapest, In Hungarian, pp. 396. Braun-Blanquet, J., 1964. Pflanzensoziologie: Grundzüge Der Vegetationskunde 3 Springer, Verlag, Wien, New York. Brennan, R.S., Withgott, J., 2011. Environment: the Science Behind the Stories. Pearson, Benjamin – Cummings, San Francisco, pp. 280–312. Burkart, M., 2001. River corridor plants (Stromtalpflanzen) in Central European lowland: a review of a poorly understood plant distribution pattern. Global Ecol. Biogeogr. 10, 449–468. Celesti-Grapow, L., Pyšek, P., Jarošík, V., Blasi, C., 2006. Determinants of native and neophytes species richness in the urban flora of Rome. Divers. Distrib. 12, 490–501. http://dx.doi.org/10.1111/j.1366-9516.2006.00282.x. Crawley, M.J., 1993. GLIM for Ecologists. Blackwell Scientific Publications, Oxford, UK, pp. 392. Deutschewitz, K., Lausch, A., Kühn, I., Klotz, S., 2003. Native and neophytes plant species richness in relation to spatial heterogeneity on a regional scale in Germany. Global Ecol. Biogeogr. 12, 299–311. Doroftei, M., 2009. Chorology of Amorpha fruticosa in the Danube Delta. Romanian J. Biol.-Plant Biol. 54 (1), 61–67. Dumitraşcu, M., Grigorescu, I., Kucsicsa Gh Dragotă, C.S., Năstase, M., 2012. Non-native and native invasive terrestrial plant species in Comana Natural Park: case-studies: Amorpha fruticosa and Crategus monogyna. Romanian J. Geogr. 55 (2), 81–89. Đurđić, S., Stojković, S., Šabić, D., 2011. Nature conservation in urban conditions: a case study from Belgrade, Serbia. Maejo Int. J. Sci. Technol. 5 (1), 129–145. Dyderski, К.М., Gdula, К.А., Jagodzinski, М.А., 2015. The rich get richer concept in riparian woody species – A case study of the Warta River Valley (Poznan, Poland). Urban For. Urban Green. 14, 107–114. Gagić-Serdar, R., Poduška, Z., Đorđević, I., Češljar, G., Bilibajkić, S., Rakonjac, L.J., Nevenić, R., 2013. Suppression of Indigo bush with pod pests. Arch. Biol. Sci. 65 (2), 801–806. Gallé, L., Margóczy, K., Kovács, E., Gyrffy Gy Körmöczy, L., Németh, L., 1995. River valleys: are they ecological corridors? Tiscia 29, 53–58. Genovesi, P., Shine, C., 2003. European strategy on invasive alien species: convention on the conservation of European wildlife and habitats final version. In: Standing Committee 23rd Meeting. Strasbourg, 1–5 December, 2003. . Grbić, M., Đukić, M., Skočajić, D., Đunisijević-Bojović, D., 2007. Role of invasive plant species in landscapes of Serbia. In: 18th International Annual ECLAS Conference Landscape Assessment – From Theory to Practice: Applications in Planning and Design Proceedings, Faculty of Forestry. Belgrade,. pp. 219–228. Hammer, O., Harper, D.A.T., Ryan, P.D., 2001. Paleontological statistics software package for education and data analysis. Palaeontol. Electron. 4, 1–9. Hood, W.G., Naiman, R.J., 2000. Vulnerability of riparian zones to invasion by exotic vascular plants. Plant Ecol. 148, 105–114. Jarošík, V., Pyšek, P., Kadlec, T., 2011. Alien plants in urban nature reserves: from red-list species to future invaders? NeoBiota 10, 27–46. http://dx.doi.org/10.3897/neobiota. 10.1262. Jovanović, B., Vukićević, E., Radulović, S., 1984. Prvobitna, postojeća i potencijalna prirodna vegetacija Ade Ciganlije sa okolinom i vegetacijske karte (The original, existing and potential natural vegetation of Ada Ciganlija with the surroundings and vegetation maps). Bull. Fac. For. 63, 4–47 (in Serbian with English summary). Jovanović, B., Vukićević, E., Radulović, S., 1985. Vegetacija i vegetacijske karte Ade Huje kod Beograda (The vegetation and the vegetation maps of Ada Huja, near Belgrade). Bull. Fac. For. 64, 289–317 (in Serbian with English summary). Jovanović, S., Stojanović, V., Lazarević, P., Jelić, I., Vukojičić, S., Jakovljević, K., 2014. Flora of Belgrade surroundings (Serbia) 150 years after Pančić’s monograph – a comparative overview. Bot. Serbica 38 (2), 201–207. Jovanović, S., 1994. Ekološka studija ruderalne flore i vegetacije Beograda (Ecological Study of Ruderal Flora and Vegetation in the City of Belgrade). Biološki fakultet Univerziteta u Beogradu, Belgrade, pp. 222 (in Serbian with English summary). Knapp, S., Kühn, I., Stolle, J., Klotz, S., 2010. Changes in the functional composition of a Central European urban flora over three centuries. Perspect. Plant Ecol. Evolut. Syst. 12, 235–244. Kowarik, I., 2008. On the role of alien species in urban flora and vegetation. In: Marzluff, J., et, al. (Eds.), Urban Ecology. Springer, US, pp. 21–338. Krpan, A.P.B., Benko, M., 2009. Preface, biological – ecological and energetic characteristics of Indigo bush (Amorpha fruticosa L.). In: Proc. of Scientific Symposium with International Participation. Zagreb, Croatia, March 12th 2009. pp. 4. Lazarević, P., Stojanović, V., Jelić, I., Perić, R., Krsteski, B., Ajtić, R., Sekulić, N., stands are the most dangerous hotspot of invasive species, which might threaten native species and riparian habitats at the mouth of the Sava and Danube rivers in the future. In lower, Bulgarian part of Danube, the cover value of A. fruticosa as well the number of invasive species are also higher in man-made habitats compared to natural habitats (Pedashenko et al., 2012). In addition, according to data for the Danube Delta, A. fruticosa has been identified with much more frequency and average cover percentage in forest and shrub communities such as Salicetum albae-fragilis, Calamagrostio-Salicetum cinereae, Salicetum triandrae, Salicetum triandrae subass. amorphosum fruticosae, than in different Phragmitetum s.l. communities as well in some grassland or ruderal communities (Doroftei, 2009). The obtained results confirm the general opinion that invasive neophytes are the serious risk factor for biodiversity loss (Brennan and Withgott, 2011), as well as that invasive alien species are one of the biggest challenges in the preservation of biodiversity in Europe (Genovesi and Shine, 2003). The cities have long been recognised as hotspots of alien plants (Pyšek, 1998; Celesti-Grapow et al., 2006; Kowarik, 2008), whose invasive probability and invasive potential must not be underestimated (Dyderski et al., 2015). Therefore, many authors recommend that alien species should be avoided in urban forestry and greening, especially near the natural riparian forests which are greatly vulnerable to biological invasion (e.g. Richardson et al., 2007; Dyderski et al., 2015). 5. Conclusions In the studied area Amorpha fruticosa is threatening the most a) willow and poplar habitats; b) wet lowland meadows and c) reedbed habitats. Communities with domination of A. fruticosa are differentiated into 7 coenological groups in the investigated area. Four groups are developed within the willow and poplar habitats, two in the wet lowland meadows, while one group is developed within the tall reedbed habitats. In all analysed stands 222 taxa were recorded, out of which 29 taxa (13.06%) were invasive neophytes. In addition to Amorpha fruticosa which dominates in all the investigated stands, other invasive neophytes with high frequency in the analysed relevés are Aster lanceolatus, Acer negundo, Vitis vulpina, Solidago gigantea, Cirsium arvense, Echinochloa crus-galli, Xanthium strumarium subsp. italicum, Fraxinus pennsylvanica, Sorghum halepense, Bidens frondosa and Echinociystis lobata. We established a direct negative correlation between the change in the number of species and the proportion of invasive species i.e. their coverage indexes in the analysed stands. Coenological group B Rubus caesius-Amorpha fruticosa represents the final stage of degradation of urban floodplain willow and poplar forests in Belgrade, in which secondary increase of biodiversity was recorded, as well as the presence of 21 of the total of 29 recorded invasive species. Therefore, habitats of this coenological group represent the most dangerous hotspots of invasive species, which might in the future threaten native species and riparian habitats at the mouth of the Sava and Danube rivers, as well as similar habitats downstream. Acknowledgements This study was supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia, project no. 173030. We are grateful to the anonymous reviewer for helpful comments on the manuscript. References Alvey, A.A., 2006. Promoting and preserving biodiversity in the urban forest. Urban For. Urban Green. 5, 195–201. 107 Urban Forestry & Urban Greening 24 (2017) 101–108 N. Radovanović et al. on some grassland vegetation types from western Romania. Res. J. Agric. Sci. 42 (1), 536–540. Schnitzler, A., Haleb, W.B., Alsum, M.E., 2007. Examining native and exotic species diversity in European riparian forests. Biol. Conserv. 138, 146–156. Surina, B., 2004. The association Gentiano terglouensis-Caricetum firmae T. Wraber 1970 in the Krn Mts. (Julian Alps). AnnalesSeries historia naturalis. Ser. Historia Nat. 14, 99–112. Ter Braak, C.J., Šmilauer, P., 2012. Canoco Reference Manual and User’s Guide: Software for Ordination (version 5.0). Microcomputer Power, Ithaca NY, USA. Török, K., Botta-Dukát, Z., Dancza, I., Németh, I., Kiss, J., Mihály, B., Magyar, D., 2003. Invasion gateways and corridors in the Carpathian basin: biological invasions in Hungary. Biol. Invasions 5, 349–356. Thiers, B., 2016. Index Herbariorum: A global directory of public herbaria and associated staff. New York Botanical Garden’s Virtual Herbarium. http://sweetgum.nybg.org/ (continuously updated). Thomas, C., 2013. The Anthropocene could raise biological diversity. Nature 502, 7. Tucović, A., Isajev, V., 2000. Colonization of forest habitats with Amorpha fruticosa L. in Serbia and its biological properties. Proc. of 6th Weed Congress of Serbian Herbological Society 217–227. Tutin, T.G., Heywood, V.H., Burges, N.A., Moore, D.M., Valentine, D.H., Walters, S.M., Webb, D.A. (Eds.), 1968–1980. Flora Europaea 2–5. University Press Cambridge. Tutin, T.G., Burges, N.A., Chater, O.A., Edmondson, J.R., Heywood, V.H., Moore, D.M., Valentine, D.H., Walters, S.M., Webb, D.A. (Eds.), 1993. Flora Europaea. University Press, Cambridge Ed. 2, 1. Van Der Maarel, E., 1979. Transformation of cover-abundance values in phytosociology and its effects on community similarity. Vegetatio 39, 97–114. Weber, E., Gut, D., 2004. Assessing the risk of potentially invasive plant species in central Europe. J. Nat. Conserv. 12 (3), 171–179. Yong Sung, C., Li, M., Rogers, G.O., Volder, A., Wang, Z., 2011. Investigating alien plant invasion in urban riparian forests in a hot and semi-arid region. Landsc. Urban Plan. 100, 278–286. Branković, S., Sekulić, G., Bjedov, V., 2012. A preliminary list of invasive species in Serbia, with general measures of control and reduction as a basis of future legal acts. Prot. of Nat. 62, 5–32 (In Serbian with English summary). Moffatt, S.F., McLachlan, S.M., Kenkel, N.C., 2004. Impacts of land use on riparian forest along an urban–rural gradient in southern Manitoba. Plant Ecol. 174, 119–135. Olden, J.D., Rooney, T.R., 2006. On defining and quantifying biotic homogenization. Global Ecol. Biogeogr. 15, 113–120. Pedashenko, H., Apostolova, I., Vassilev, K., 2012. Amorpha fruticosa invasibility of different habitats in lower Danube. Phytol. Balc. 18 (3), 285–291. Protopopova, V.V., Shevera, M.V., Mosyakin, S.L., 2006. Deliberate and unintentional introduction of invasive weeds: a case study of the alien flora of Ukraine. Euphytica 148, 17–33. Pyšek, P., Prach, K., et al., 1994. How important are rivers for supporting plant invasions? In: de Waal, L.C. (Ed.), Ecology and Management of Invasive Riverside Plants. John Wiley & Sons, New York, pp. 19–26. Pyšek, P., Richardson, D.M., Rejmánek, M., Webster, G.L., Williamson, M., Kirschner, J., 2004. Alien plants in checklists and floras: towards better communication between taxonomists and ecologists. Taxon 51 (1), 131–143. Pyšek, P., 1998. Neophytes and native species in Central European urban floras: a quantitative comparison. J. Biogeogr. 25, 155–163. Radulović, S., Skočajić, D., Bjedov, I., Đunisijević-Bojović, D., 2008. Amorpha fruticosa L. on wet sites in Belgrade. Bull. Fac. For. 97, 221–233 (in Serbian with English summary). Raunkier, C., 1934. The Life Forms of Plants and Statistical Plant Geography. pp. 632 Oxford at the Clarendon Press. Richardson, M., Holmes, P.M., Esler, K.J., Galatowitsch, S.M., Stromberg, J.C., Kirkman, S.P., Pyšek, P., Hoobs, R.J., 2007. Riparian vegetation: degradation, alien plant invasions, and restoration prospects. Divers. Distrib. 13, 126–139. Säumel, I., Kowarik, I., 2010. Urban rivers as dispersal corridors for primarily winddispersed invasive tree species. Landsc. Urban Plan. 94, 244–249. Sărăţeanu, V., 2010. Assessing the influence of Amorpha fruticosa L. invasive shrub species 108