Vascular flora of eight water reservoir areas in southern Italy

Artificial lakes play an important role in maintaining the valuable biodiversity linked to water bodies and related habitats. The vascular plant diversity of eight reservoirs and surrounding areas in southern Italy was inventoried and further analysed in terms of biodiversity. A total of 730 specific and subspecific taxa were recorded, with 179 taxa in the poorest area and 303 in the richest one. The results indicate a good richness of the habitats surrounding the water basins, with some species of nature conservation interest and only a few alien species.


INTRODUCTION
Wetlands are increasingly contracting under the influence of anthropogenic pressures (Egertson et al. 2004).Indications of biodiversity loss and habitat loss have been recorded worldwide (Moyle and Leidy 1992;Croce et al. 2012;Kozlowski and Bondallaz 2012;Azzella et al. 2013).Even small ponds and lakes (Semlitsch and Bodie 1998;Williams et al. 2004) can be of strategic importance for freshwater biodiversity at a regional scale (Gibbs 2000).
In large parts of the Mediterranean basin, reservoirs are the most important water bodies.There are 70 artificial lakes of varying size and shape in southern Italy, accounting for approximately 4 billion cubic metres of water (Romano and Costantini 2010).Construction of these water reservoirs led to the inundation of farm land and rural areas in the valley bottoms.Shores were often reforested and protected by enclosures and other protective measures.The succession of secondary vegetation in the abandoned areas led to the establishment of natural and seminatural habitats, while the disturbance of the existing natural habitats decreased.Thus, the areas surrounding the reservoirs evolved into isolated patches of nature in an agriculturally dominated landscape.
Although many authors have reported the negative impact of dams on rivers and their ecosystems (e.g., McAllister et al. 2001;Nilsson et al. 2005), dams are very important for wildlife, such as birds (Mancuso 2010).Artificial lakes fulfill an important role as water reservoirs for agricultural irrigation; however, their other functions, such as recreation, fishing, and biodiversity conservation, should not be overlooked.The Italian National Institute for Economic Agriculture (INEA) launched the project "Azione 7" (Romano and Costantini 2010) to assess the suitability of reservoirs in southern Italy for nature conservation purposes.Nine basins were chosen as a case study that would implement aspects of the researches concerning landscape features, aquatic biology, vascular plants, and birds.The present work is based on a botanic survey carried out in 2007 for the above-mentioned project (Croce 2010).The aims of this study were to explore the floristic diversity of the reservoir areas, and to define their importance for plant conservation.

Study areas
Eight reservoirs located in six regions in southern Italy were selected for the present study (Figure 1 and Table 1).These reservoirs were the focus of an INEA "Azione 7" project (Romano and Costantini 2010).The Marsico Nuovo area was excluded because the maximum water level was always under the operating level, resulting in the widespread emergence of artificial shores.According to the European Environment Agency (EEA 2011), all of the sites are located in the Mediterranean Biogeographic Region, with the exception of Penne Lake, which is located in the Continental Biogeographic Region.Altitudes range from 43 -434 m above sea level (a.s.l.).
A 500-m buffer zone was established around the lake perimeters.Within these zones, the land use and floristic composition were described.The following land use categories were applied to the various landscapes of the basins:
The land cover composition of the eight areas is shown in Figure 2. Cartographic analyses were realised with the software Qgis 1.8.0 (Quantum GIS Development Team 2013).

Floristic procedures
The database built for the botanic survey of the INEA project was used to analyse the floristic diversity.Each basin was explored during the spring, summer, and autumn of 2007 to compile an inventory of the vascular flora that was as complete as possible.For each habitat, random plots were established for floristic sampling, and a random exploration of the shores and the adjacent habitats was carried out.Species were identified in the field, or they were collected (with the exclusion of protected species) and identified by using Flora d'Italia (Pignatti 1982) and Flora Europaea (Tutin et al. 1964(Tutin et al. -1980(Tutin et al. , 1993)).All specimens were stored in the herbarium of the Department of Environmental, Biological, and Pharmaceutical Sciences and Technologies of the Second University of Naples (Caserta).
Nomenclature of the inventoried flora mainly follows Conti et al. (2005Conti et al. ( , 2007)), but the Index Plantarum Florae Italicae (IPFI) (Acta Plantarum 2013) was also consulted.For the Orchidaceae family, the nomenclature of GIROS (2009) was used.For the European white oak the name Quercus pubescens s.l.(including Q. virgiliana Ten., Q.   dalechampii Ten., Q. amplifolia Guss, etc.) has been used.The systematic scheme used in this study follows Peruzzi (2010), but families, genera, and intraspecific taxa are listed in alphabetical order.The plant life form, according to Raunkiaer (1934) and modified by Pignatti (1982), was assigned to each taxon as follows: P Phanerophytes.Woody perennials (trees, shrubs), including Epiphytes (EP).The endemic taxa have been checked according to Peruzzi et al. 2014.To analyse the importance of each species objectively, criteria were selected that consider the species value at a national or international level, without consideration of regional protection laws.Endemic and subendemic taxa, species listed in Annexes II, IV, and V of EU Directive 92/43 /CEE, Appendices I, II, and III of the CITES Convention, and Annex I of the Bern Convention were considered patrimonial species.

Statistical analysis
The Sørensen-Dice coefficient (Sørensen 1948) was computed to measure the similarity between the eight floras.The matrix obtained was correlated with the matrix of the geographic distances.A Mantel Test with 10,000 permutations was carried out.Mantel tests were performed with the use of zt (Bonnet and Van de Peer 2002).Principal component analysis (PCA) was performed on both the life form and chorotype compositions.The eigenvalues of the two main components were used as factors to test for a correlation (Spearman rs) with altitude, latitude, longitude, distance from sea, and land cover types.PCA and the correlation test were performed in PAST (Hammer et al. 2001).

RESULTS
A total of 730 specific and subspecific taxa were recorded for the studied habitats surrounding the lakes (Table 3).The flora of the basins represented 9.5% of the Italian flora (Conti et al. 2005).The richness varied from 179 taxa for the Locone basin to 303 taxa for Monte Cotugno.The species/area ratio was highly correlated with the area of each basin in a log-log space (Figure 3).
Aquatic species were absent or scarce.Only three hydrophytes (Alisma plantago-aquatica L., Lemna minor L., and Ranunculus trichophyllus Chaix subsp.trichophyllus) and one helophyte (Schoenoplectus tabernaemontani (C.C.Gmel.)Palla)) were recorded in three basins, although several plants associated with aquatic habitats were present (e.g., Lycopus europaeus L., Phragmites australis (Cav.)Trin.ex Steud.subsp.australis, Typha spp., Veronica anagallis-aquatica L.).The variation in water level, with excursions of up to several meters during the year (and complete drying up of some lakes in the summer), was not compatible with the life of aquatic plants, especially hydrophytes.Therefore, the major interests lie in the species collected from the shores or from the wetlands at the confluence of small streams and rivers.
A total of 89 families were represented.Three families (i.e., Fabaceae, Asteraceae, and Poaceae) accounted for 33.3% to 38.7% of the total flora, with few differences between different areas (Table 4).The number of families showed a statistically significant correlation with the species richness (Spearman r s = 0.85; P < 0.05).

Basin
Protected area                  Considering the life form composition (Figure 4), herb aceous species (Therophytes, Hemicryptophytes, and Geophytes) dominated woody species (Chamaephytes and Phanerophytes).Therophytes were the most represented life form, ranging from 35.3% to 41.9%, except in the Penne reservoir (28%), which is in agreement with its position in the Continental Biogeographic Region.Phanerophytes, on the other hand, accounted for less than 20% in the studied areas, with the exception of the Penne and Alento Lakes.

Penne
The first two components of the PCA performed on the life form spectra explained 87% of the variability (Figure 6A).The life form composition typically varies along the Italian Peninsula (Pignatti 1994).However, the first component was not statistically correlated with latitude, even though Penne Lake was situated on the extreme end of another quadrant in the diagram.On the other hand, the second axis was correlated with the (1) Note: previously reported in Iocchi et al. (2011).
(2) Identified using Danin (2004) Croce | Vascular flora of eight water reservoir areas in southern Italy   Croce | Vascular flora of eight water reservoir areas in southern Italy distance to the sea (Spearman rs = 0.8; P < 0.05), so it is assumed to be related to the life form compositions.
Although the Mantel test did not show a correlation between distance and the Sørensen-Dice coefficient for similarity (r = -0.216399,P = 0.18), the chorological spectra (Figure 5) were representative of the geographic positions of the lakes.Mediterranean species (Euri-Mediterranean + Steno-Mediterranean) tended to increase southward, representing only 28% of the variability for Lake Penne and exceeding 60% for Monte Cotugno.Furthermore, Eurasiatic chorotypes were dominant around Lake Penne (32%) and Lake Conza (27%), whereas they were less present around the other basins.
The first component of the PCA performed on the chorotype composition explained more than 86% of the total variability and was significantly correlated with longitude (Spearman rs = -0.78,P < 0.05; see Figure 6B).This result indicates the presence of the well-known eastwest-directed vegetation gradient, in which the eastern part of the Italian Peninsula received an admixture of species from Eastern Europe (e.g., Balcanic, Illyric, SE-European and Pontic species; see Trotter 1912, Pezzetta 2010, Peruzzi et al. 2014;Wagensommer et al. 2014).On the other hand, the PCA diagram showed a high correlation of this axis with an abundance of Eurasiatic floristic species.
Thirty-nine taxa were selected as patrimonial species (Table 5).Twenty-three taxa were orchid species, related to open habitats, such as meadows, grasslands, and shrublands.The Locone and Conza Lakes showed a high number of patrimonial taxa, despite these areas having relatively low floristic biodiversity.The Penne and Alento Lakes harboured the lowest number of patrimonial species.On the other hand, the number of invasive species and their impact on the floristic biodiversity were very limited.There were only 22 invasive species in a total of 125 (sub)cosmopolite species (Celesti-Grapow et al 2010; see Table 6).The most common ones were Arundo donax L. and Robinia pseudacacia L. (six areas), and Ailanthus altissima (Mill.)Swingle (five areas).The highest number of invasive species occurred in the Lake Penne and Ponte Liscione areas (9 taxa), whereas only two taxa were found around Lake Alento and only three taxa around Lake Conza.

DISCUSSION
The INEA "Azione 7" project represents an important case study of the natural value of large reservoir areas.Similar surveys have been conducted in South America (Tundisi and Matsumura-Tundisi 2003;Tundisi et al. 2008;Alves-da-Silva et al. 2014) and China (Tian et al. 2007).These studies underlined the positive effects of the presence of inundated areas on the aquatic biodiversity and the surrounding habitats, although the construction of dams had a high impact on the landscape.The project provided opportunities to collect biological information in agricultural areas that remained poorly studied because they were of little interest to botanists.
Seven of the eight basins have long been designated as either Sites of Community Importance (SCIs) or Special Protection Areas (SPAs), according to the European Council Directive 92/43/CEE (known as the "Habitats Directive"), or as Ramsar sites or WWF Oases (Table 2).These designations were done mainly taking account of the occurrence of rare bird species and the European otter (Lutra lutra L.).Knowledge of the floristic diversity is scarce for most of the areas (Scoppola and Blasi 2005).The floristic survey of the selected reservoir areas resulted in an inventory of 730 specific and subspecific taxa belonging to 89 families.The high biodiversity of the areas surrounding the reservoirs, varying from 179 to 303 taxa, counterbalanced the scarcity of aquatic plant species in the reservoirs.The two main factors related to the floristic composition of the areas were the distance from the sea and the east-west gradient across the Italian peninsula.The importance of this gradient for the floristic diversity of areas in southern Italy was recognised previously.The high plant biodiversity was probably underestimated because the present study was carried out over only 1 year.The biodiversity could be increased by increasing the natural state of some of the habitats, for example, by using reforestation, creating or increasing the size of ponds, preventing the complete drying up of some of the lakes, and enabling the establishment and growth of aquatic vegetation.
The floristic inventory identified several species that are of interest in terms of nature conservation.Therefore, the basins can be considered as biodiversity hotspots in agriculturally dominated landscapes.The contribution of such (semi-)natural habitat remnants is critical for the regional biodiversity (Duelli and Obrist 2003).In addition, the invasive plant species were limited in number, and their facilitation was not observed (Johnson et al. 2008).The lakes and their surrounding habitats acted not only as water reservoirs, but also as reserves of plant species.

Figure 2 .
Figure 2. Land cover of the 500-m buffer surrounding the studied reservoirs.
IT9220144 "Lago di S. Giuliano e Timmari" Natural Oriented Reserve "Lago S. Giuliano e Timmari" Monte Cotugno SPA IT9210275 "Massiccio del Monte Pollino e Monte Alpi" Angitola Wetland of International Importance (RAMSAR) WWF Oasis Table 2. Protected areas established in the studied areas (SCI = Site of Community importance; SPA = Special Protection Area).

Figure 3 .
Figure 3. Linear regression between the areas (log) and the number of taxa/area ratio (log).

Figure 5 .
Figure 5. Chorological spectra of the studied areas.

Figure 6 .
Figure 6.PCA diagrams of the eight reservoir areas, based on analyses of the life form spectra (A) and the chorological spectra (B).

Table 1 .
Information about the eight reservoirs analysed.
Figure 1.Location of the eight analysed basins.Biogeographic regions after EEA (2011).See Table1for the names and features of each basin.
Croce | Vascular flora of eight water reservoir areas in southern Italy

Table 3 .
Vascular flora of the eight studied areas.The inventory shows the scientific name, the life form (Ch = Chamephyte, G =Geophyte, He = Helophyte, Angitola) and the voucher number (accession number).All specimens have been collected by the author and deposited in the Herbarium of the Department of Environmental, Biological, and Pharmaceutical Sciences and Technologies of the Second University of Naples (Caserta).
ContinuedCroce | Vascular flora of eight water reservoir areas in southern Italy
ContinuedCroce | Vascular flora of eight water reservoir areas in southern Italy
ContinuedCroce | Vascular flora of eight water reservoir areas in southern Italy
ContinuedCroce | Vascular flora of eight water reservoir areas in southern Italy
ContinuedCroce | Vascular flora of eight water reservoir areas in southern Italy
ContinuedCroce | Vascular flora of eight water reservoir areas in southern Italy

Table 3 .
Continued.Vascular flora of eight water reservoir areas in southern Italy ContinuedCroce |

Table 3 .
Continued.Vascular flora of eight water reservoir areas in southern Italy ContinuedCroce |

Table 3 .
Continued.Vascular flora of eight water reservoir areas in southern Italy ContinuedCroce |

Table 4 .
Number and relative abundance of the three richest families of the analyzed flora.

Table 5 .
Croce | Vascular flora of eight water reservoir areas in southern Italy Patrimonial species of the eight basins.B = Bern Convention, Annex I, C = Cites Convention, Annex II; D-II = Directive 92/43/CEE, Annex II; D-V = Directive 92/43/CEE, Annex V; and E = Endemic.