Environ Monit Assess (2015) 187: 301 DOI 10.1007/s10661-015-4521-7
The effect of river damming on vegetation: is it always unfavourable? A case study from the River Tiber (Italy) Simona Ceschin & Ilaria Tombolini & Silverio Abati & Vincenzo Zuccarello
Received: 25 November 2014 / Accepted: 9 April 2015 / Published online: 29 April 2015 # Springer International Publishing Switzerland 2015
Abstract River damming leads to strong hydromorphological alterations of the watercourse, consequently affecting river vegetation pattern. A multitemporal and spatial analysis of the dam effect on composition, structure and dynamic of the upstream vegetation was performed on Tiber River at Nazzano-dam (Rome). The main research questions were as follows: How does plant landscape vary over time and along the river? Where does the dam effect on vegetation end? How does naturalistic importance of the vegetation affected by damming change over time? Data collection was performed mapping the vegetation in aerial photos related to the period before (1944), during (1954) and after dam construction (1984, 2000). The plant landscape has significantly changed over time and along the river, particularly as a result of the dam construction (1953). The major vegetation changes have involved riparian forests and macrophytes. Dam effect on vegetation is evident up to 3 km, and gradually decreases along an attenuation zone for about another 3 km. Despite the fact that the damming has caused strong local hydromorphological modification of the river ecosystem transforming it into a sub-lacustrine habitat, it has also led to the formation S. Ceschin (*) : I. Tombolini : S. Abati Department of Sciences, Roma Tre University, V.le G. Marconi 446, 00146 Rome, Italy e-mail: [email protected] V. Zuccarello Department of Sciences and Biological and Environmental Technology, Salento University, St. Prov. Lecce-Monteroni, Polo Ecotekne, 73100 Lecce, Italy
of wetlands of considerable naturalistic importance. Indeed, in these man-made wetlands, optimal hydrological conditions have been created by favouring both the expansion of pre-existing riparian communities and the rooting of new aquatic communities, albeit typical of lacustrine ecosystems. Some of these plant communities have become an important food resource, refuge or nesting habitats for aquatic fauna, while others fall into category of Natura 2000 habitats. Therefore, river damming seems to have indirectly had a “favourable” effect for habitat conservation and local biodiversity. Keywords Hydroelectric dam . Habitats directive . River vegetation . Multitemporal-spatial analysis . Land use change
Introduction River damming is one of the most serious anthropogenic disturbances impacting on river ecosystem, since it generally leads to strong alterations in the watercourse’s hydromorphological and biological features. Some studies have focused on the dam effects on the upstream river stretch, among which the most important are the increase in water level with subsequent flooding of the surrounding riparian lands (Crivelli et al. 1995; Nilsson and Berggren 2000; New and Xie 2008) and the reduction in the river water velocity with formation of sublacustrine habitats (Nilsson and Jansson 1995; Jansson et al. 2000a, b; Evans et al. 2007; Franchi et al. 2014).
301 Page 2 of 12
These hydromorphological changes of the river ecosystem strongly influence the structure and distribution of the upstream aquatic and riparian macrophyte communities (Nilsson and Berggren 2000; New and Xie 2008). Despite these studies, research into multitemporal and spatial effects of the river damming on composition, structure and dynamic of the plant communities is scarce, especially as regards upstream vegetation (Nilsson et al. 1997; Tombolini et al. 2014). In our own recent study about the vegetation of the lower course of the Tiber River at Nazzano-dam (Rome) (Tombolini et al. 2014), we came to the conclusion that the aquatic and riparian communities are those most affected by the dam operation, i.e. by the resulting alteration of some hydrological parameters, in particular the river water level. In this paper, we have investigated the dam effects on the upstream river landscape; in particular, we aimed to answer to the following main research questions:
Environ Monit Assess (2015) 187: 301
period in July and August and a cold stress period that lasts from October to May (Blasi 1994). From a hydrological point of view, at the analysed river stretch, the estimated average flow is 189 m3/s (Gallo 1983). The Nazzano-dam has a limited capacity for modifying the water regime, considering the similar flow water between the related upstream and downstream river stretches. At Nazzano Lake, the monthly variability of the water level is kept low by the maintenance of the reservoir water level at the spillway crest elevation throughout the year. Moreover, the reservoir’s storage capacity represents a small fraction of the annual average runoff (about 5.9 billion m3) and allows only a regulation on a daily basis. For these features, the Nazzano Lake can be considered as a run-of-the-river impoundment (Tombolini et al. 2014).
Methods 1. How do vegetation and river landscape vary upstream from the dam over time and along the river? 2. Where does the damming effect on vegetation end? 3. How does the naturalistic importance of the vegetation affected by damming change over time?
Study area The study area is located along the lower course of the Tiber River at Nazzano (40 km North of Rome, Latium) and it is immediately upstream of a hydroelectric dam, the Nazzano-dam, operational since 1956 (Fig. 1). The selected river stretch is 8.5 km long and covers a surface of about 300 ha. After the dam came into operation, a sector of this river stretch became transformed into a sub-lacustrine ecosystem, the Nazzano Lake, whose storage capacity is about 18 million m3 (AA.VV. 1985). This river stretch is included in the Tiber-Farfa Natural Regional Reserve, which was founded in 1979 (1979/21 R.L.) and includes a Site of Community Importance and a Special Protection Area (SCI/SPA IT6030012) (92/43/CEE, 79/409/CEE Directives) (AA.VV. 2006). As regards the climatic features, the area falls within the Transitional Temperate region, characterized by lower humid ombrotype and upper Mesomediterranean thermotype. It is a semi-arid mesoclimate with an arid
Data collection and mapping The multitemporal and spatial analysis of the vegetation along the selected river stretch was performed by using aerial photos of the summer season from 1944, 1954, 1984 to 2000, related to the period before (1944), during (1954) and after the Nazzano dam construction (1984, 2000). Firstly, we digitalized the vegetation in the four aerial photos and analysed the data with ArcGis 9.0 (ESRI inc. Redland, CA). Different vegetation types were mapped at an approximate scale of 1:10.000. Subsequently, the mapped vegetation types were verified in the field by using the phytosociological method of Braun-Blanquet (1932) and consulting the vegetation surveys available in the literature concerning the area (D’Antoni et al. 2002; Spada 2006; Spada and Casella 2006; Fanelli et al. 2007; Ceschin and Salerno 2008; Ceschin et al. 2010). The vegetation mapping in the historical aerial photos was carried backward, starting from the 2000 vegetation map, considered as representative of the current vegetation. The characterization of the vegetation types through phytosociological surveys allowed assigning a land use category to each identified vegetation type. We used the CORINE Land Cover Legend (APAT 2005; Sambucini et al. 2010), according to ISPRA (2010) for the fourth level of detail.
Environ Monit Assess (2015) 187: 301
Page 3 of 12 301
Fig. 1 Lower course of the Tiber River crossing the Latium region in Central Italy (a). The investigated river stretch subdivided into 153 transects. The dam location is indicated (arrow) (b)
On each of the obtained land use maps, a continuous buffer area starting from the Nazzanodam (lower limit) to the border of the Tiber-Farfa Natural Regional Reserve (higher limit), placed at a distance of about 8.5 km, was defined. The buffer area, including the river bed and both the riparian zones along the selected river stretch, was divided into a series of transversal transects 50 m wide on each riverbank and with a constant surface area (153 transects) (Fig. 1b). The extension of each land use category falling in each transect was calculated in the four land use maps and compared over the 1944–2000 time period. The set of these data has been the basis on which we carried out the subsequent statistical analyses. The land use categories for each time period were related to the habitats defined by Habitats Directive 92/ 43/EEC (Natura 2000 Habitat) in order to assign them a different naturalistic importance.
Statistical analysis Statistical analyses were performed in order to evaluate how the dam operation has affected the river landscape and in particular the vegetation pattern. The analyses were as follows: (i) Analysis of the land use transformations over time and along space. The matrix of the transects relative to different periods (1944, 1954, 1984, 2000), described by the percentages of the land use categories, was performed by fuzzy c-means clustering (Bezdek 1981). Identification of optimal fuzzy partition of the transects based on normalized Dunn coefficient (Dunn 1973) was carried out. In this way, each fuzzy cluster identifies a characteristic land use type. (ii) Identification of homogenous river sectors in a land use transformation perspective. The matrix
301 Page 4 of 12
of the degrees of belonging to the fuzzy clusters in the different years was analysed by hierarchical classification (complete linkage), based on chord distance (Orloci 1978). From the resulting dendrogram, we identified homogenous river sectors. The matrix of the river sector centroids related to the degrees of belonging to land use types (identified in the previous step) during the four periods was used to identify the land use types prevailing in each time period. The land use type that showed the greatest degree of belonging at time t was defined as dominant in the river sector at time t. For the statistical analyses, the software GINKGO ver. 1.7.0 (De Caceres et al. 2007) was utilized.
Results How do river vegetation and landscape vary upstream from the dam over time and along the river? During the four time periods in question, the land use mapping in the buffer area has allowed the identification of 13 cartographic types related to the same number of land use categories (Table 1). These categories refer to three main typologies: (1) artificial areas and cultivated vegetation, (2) spontaneous forests or grasslands and (3) riparian forests and riverbank and aquatic macrophyte vegetation. The variations in extension of the 13 land use categories in pre- and post-dam period are shown in Table 2. The period 1954–1984 appears to be that in which the major land use changes occurred. The most evident changes have affected, on the one hand, the inland waters, and riverbank and aquatic macrophyte vegetation which have greatly increased in surface, and, on the other hand, the agricultural areas and natural grasslands, whose coverage has instead been decreasing. The optimal partition obtained by fuzzy clustering identifies 5 fuzzy groups corresponding to the different land use modalities. The centroids of the land use categories of each group are shown in Table 3. On the basis of these data, it is possible to describe the significance of each group. Groups 1, 2, and 4 are characterized by a predominantly agricultural land use: this is highest in group 2 and intermediate in groups 1 and 4. In the latter group, the agricultural use is combined with wide pasture areas and natural grassland replacement. The
Environ Monit Assess (2015) 187: 301
artificial areas are greater in group 3, in which, to compensate the presence of infrastructures, there is the most abundant presence of riparian forests. In this group, the agricultural exploitation is very low. Group 5 is characterized by aquatic and riverbank macrophyte vegetation, as well as by spontaneous deciduous woods. Also in this case, the areas for agricultural use are of very limited extent. The inland waters category, showing high values in all the five fuzzy groups, was not considered as a characteristic element of any fuzzy cluster, although its area has increased significantly as a result of the dam operation (Table 2). The hierarchical classification of transects, described on the basis of the degrees of belonging to the fuzzy clusters during the considered period (1944–2000), distinguishes five main groups (not to be confused with the five fuzzy clusters identified previously). Each group identifies a river sector consisting of several transects more or less spatially contiguous. The five identified river sectors are shown in Fig. 2. Sector 1 (S1) represents mainly the river stretch immediately upstream of the confluence Tiber-Farfa; sector 2 (S2) corresponds to two small river stretches in the meandering zone with double curve of the River Tiber, relatively distant from the dam; sector 3 (S3) identifies different sections of the river, although all these are upstream of the Tiber-Farfa confluence; sector 4 (S4) is exclusively a contiguous river stretch not far from the dam and includes both the confluence Tiber-Farfa zone and some transects just upstream and downstream of it; and sector 5 (S5) mainly identifies the first river stretch upstream of the dam, as well as few transects in the meandering zone. The dynamic of the prevailing land use type in the various river sectors, on the basis of the matrix of the fuzzy clusters centroids in the different time periods, is shown in Table 4. In 1944 and 1954 (pre-dam period), agricultural land use is evident throughout the investigated area, i.e. along all the river sectors identified (from S1 to S5). The only difference between 1944 and 1954 is linked to the different extension of the natural grasslands in respect to agricultural areas in the sectors S1 and S2; indeed, if natural grasslands are prevalent in 1944 in S2, they are dominant in 1954 in S1. In 1984, considering the river sectors more distant from the dam, it is observed that S2 maintains a mainly agricultural land use, while S1 and S3 remain of agricultural use although not exclusively, evidence of a reduction of agricultural activities. In general, however,
Environ Monit Assess (2015) 187: 301
Page 5 of 12 301
Table 1 Land use categories recognized in the study area with the corresponding Corine Land-Cover categories, Natura 2000 Habitat types and syntaxonomical classification (phytosociological class) Land use category
Description
Corine LandCover codex
Natura Habitat 2000 (codex)
Phytosociological class
Artificial surface
Urban area, artificial fabric, mine, dump, sand and gravel drainage Cereal and tree crops
1
–
–
2.1+2.2
–
–
–
Quercetea ilicis
–
Querco-Fagetea
Agricultural areas
Broad-leaved forest Forest with Quercus ilex 3.1.1.1 with evergreen oaks Broad-leaved forest Mixed deciduous forest with Quercus 3.1.1.2 with deciduous oaks cerris Broad-leaved forest Riparian forest with Populus alba, 3.1.1.6 with hygrophilous Salix alba or Salix purpurea species
Natural grassland
3.2.3
–
Quercetea ilicis/ Rosmarinetea/ Helianthemetea guttatae
3.2.4
–
Artemisietea vulgaris
3.3
–
–
4.1.3
–
PhragmitoMagnocaricetea/ MolinioArrhenatheretea
4.1.3
Eutrophic natural lakes with Magnopotamion or Hydrocharition vegetation (3150), Mountain rivers and plains with Ranunculion fluitantis and CallitrichoBatrachion vegetation. –
Potametea pectinati
Meadows subject to flooding, 3.2.1 grasslands replacement and pasture areas
Bushes and heathland
Bushes of substitution with Crataegus monogyna, Cornus sanguinea, Cornus mas, Pyrus amygdaliformis, Spartium junceum), thickened by tangles of bramble (Rubus ulmifolius, R. caesius) Sclerophyllous Mosaic of evergreen bushes and vegetation vegetation of substitution of oak forests or of early evolutive stages of secondary succession towards the holm oak forest Transitional wood Vegetation of substitution in shrub anthropized areas with Arundo donax or communities with woody (Ulmus minor, Rubus ulmifolius, Robinia pseudoacacia, Ailanthus altissima) or herbaceous species, in landfill environments or near to abandoned fields Open spaces with little Areas sparsely vegetated or subject to or no vegetation sedimentation. Riverbank macrophyte Riverbank herbaceous vegetation vegetation with Phragmites australis, Iris pseudoacorus, Typha latifolia, Carex pseudocyperus or with Paspalum paspaloides and Agrostis semiverticillatum Aquatic macrophyte Aquatic vegetation with vegetation Potamogeton nodosus, Najas marina, Ceratophyllym demersum, Callitriche stagnalis
Inland waters
3.2.2
Forest dominated by Salix alba and Querco-Fagetea/ Populus alba (92A0) / Standing Salicetea flowing Mediterranean rivers purpurea with Paspalo-Agrostidion and with riparian tree row dominated by Salix spp. and Populus alba (3280). Secondary perennial grasslands Festuco-Brometea with Bromus erectus or Dactilys glomerata growing on increased edaphic water retention (6210). – Rhamno-Prunetea
Waters of the Tiber and Farfa rivers
5.1
–
301 Page 6 of 12 Table 2 Variations in extension (ha) of the 13 land use categories in pre- and post-dam period
Environ Monit Assess (2015) 187: 301
Land use category
1944–1954
1984–2000
Aquatic macrophyte vegetation
1.81
7.29
5.73
Riverbank macrophyte vegetation
4.69
11.62
−7.48
Riparian forest
1.80
5.06
−7.06
−10.29
37.89
6.24
0.82
−10.89
1.76
Transitional wood shrub
−0.02
−1.22
0.44
Sclerophyllous vegetation
−0.30
−0.51
0.64
Broad-leaved forest with evergreen oaks
−0.53
−0.07
0.61
Broad-leaved forest with deciduous oaks
Inland waters Natural grassland
−1.39
2.03
−2.39
Bushes and heathland
1.33
1.18
3.94
Agricultural areas
0.35
−56.27
1.21
−3.26
−0.97
−0.28
5.00
4.86
−3.35
Open spaces with little or no vegetation Artificial surface
agricultural land use continues to be predominant in the river sectors upstream of the Tiber-Farfa confluence. Instead, sectors S4 and S5, closer to the dam, have undergone a major change in land use; indeed in these sectors, it is possible to observe a transformation from a mainly agricultural use during the pre-dam period to a strong development of macrophyte vegetation (and deciduous forests) (S4) as well as of riparian forests (S5) in the post-dam period. Therefore, macrophyte and riparian vegetation have been the vegetation types most affected by the dam operation. In S4, the extension of the deciduous forest area is related to the abandonment of agriculture activities.
In 2000, we observe the same situation recorded in 1984 with no further changes in the land use type. Therefore, there are two factors driving land use changes in the river stretches closer to the dam: the dam effect with the consequent local expansion of inland waters, and the reduction of the agricultural activities. In the river stretches more distant from the dam, a more limited land use change has been observed, being a mainly agricultural use conserved over time. Multitemporal analysis of coverage variations of the aquatic and riverbank macrophytes as well as riparian
Table 3 Centroid values of the 13 land use categories for each of 5 fuzzy groups
In italic, the most substantial centroid values
1954–1984
gr1
gr2
gr3
gr4
gr5
Aquatic macrophyte vegetation
0.9
0.9
2.6
0.7
13.9
Riverbank macrophyte vegetation
8.9
4.6
4.7
5.1
11.0
Riparian forest
12.6
10.3
15.3
12.6
7.6 31.9
Inland waters
33.5
24.7
42.6
26.9
Natural grassland
1.4
1.5
3.0
21.8
4.1
Transitional wood shrub
2.1
1.3
3.4
1.7
2.7
Sclerophyllous vegetation
0.3
0.2
0.2
0.1
0.1
Broad-leaved forest with evergreen oaks
0.2
0.3
0.1
0.1
0.1
Broad-leaved forest with deciduous oaks
1.7
2.5
2.9
1.6
21.0
Bushes and heathland
1.7
0.5
5.4
1.7
1.0
Agricultural areas
30.4
50.1
5.6
23.0
3.4
Open spaces with little or no vegetation
0.6
0.2
0.3
0.2
0.5
Artificial surface
5.7
2.9
13.9
4.7
2.7
Environ Monit Assess (2015) 187: 301
Fig. 2 The five river sectors (S) identified along the Tiber River
Page 7 of 12 301
301 Page 8 of 12
Environ Monit Assess (2015) 187: 301
Table 4 Land use changes over time (1944–2000) along the identified river sectors (S1–S5)
river sector S1 S2 S3 S4 S5
1944
1954
1984
2000
g2 g4 g2 g2 g2
g4 g2 g2 g2 g2
g1 g2 g1 g5 g3
g1 g2 g1 g5 g3
forests (Table 2) confirms these findings and underlines the importance of the dam effect on river vegetation, showing local increases in coverage of these vegetation types closer to the river. Where does the damming effect on vegetation end? The results indicate that damming has a strong influence on vegetation in the river sectors closer to the infrastructure itself (i.e. S4 and S5), affecting mainly the aquatic and riverbank macrophyte vegetation and riparian forests. The temporal and spatial dynamics of these vegetation types, more sensitive to the dam effect, were analysed in detail for identifying the spatial threshold of this effect. Therefore, the multitemporal analyses of the variations in extension of these vegetation types along the river are shown in Fig. 3. Specifically, in the post-dam period, the aquatic macrophytes are mostly expanded in the river stretch ranging from the dam to the Tiber-Farfa confluence zone (approximately transects 10–30) (Fig. 3a). The riverbank macrophyte vegetation dominated by reed beds shows an important expansion over time too, especially in this river stretch, with the difference that its expansion area is more extensive than aquatic macrophte in the upstream zone, including approximately the transects 16–64 (Fig. 3b). The riparian forests, such as willow and poplar woods, display a similar multitemporal trend along the entire buffer area, even if they show a general increase in the post-dam period, with the exception of a few transects (Fig. 3c). In the transects closest to the dam, the riverbank vegetation and the riparian forests are decreased or have even disappeared; probably, this is a result of cleaning activities along the banks for the
dam construction, including the removal of riverbank vegetation. The difference in surface covered totally by the typical vegetation of the river ecosystem (i.e. aquatic macrophytes, riverbank macrophytes, riparian forests) between pre- and post-dam operation periods is shown in Fig. 4. This area is calculated as cumulative, i.e. the sum of the differences in extension of these vegetation typologies in the single transects included between the dam and the transect furthest from the dam (i.e. transect 153). It is possible to observe that the area covered by these vegetation types increases in most of the river stretch. Only in transects closest to the dam (up to transect 17), and in those farthest from it, does this area decrease, albeit moderately. The river stretch in which the increase is greater is approximately from transect 18 to transect 30, followed by a significant increase up to transect 60 and a final stretch of attenuated increase up to transect 120. Therefore, excluding those transects closest to the dam, the dam effect on vegetation is evident up to transect 60 (located at a distance of about 3 km from the dam), this continuing although moderately, up to transect 120. The river stretch included between transect 60 and 120 can be considered as attenuation zone of the dam effect, which, instead, is clearly manifested in the previous river stretch.
How does naturalistic importance of the vegetation affected by damming change over time? The classification of the 13 land use categories on the basis of Habitats Directive has made possible to distinguish in the study area (1) vegetation types related to habitats of community interest, (2) natural vegetation types not included in the previous category and (3) anthropogenic vegetation or artificial surfaces, devoid of naturalistic interest (planted forests, crops, etc.). In the first category, there are included locally hygrophilous riparian forests dominated by Salix alba and Populus alba (Natura 2000 Habitat: 92A0), river woods with Salix purpurea (Natura 2000 Habitat: 3280), and aquatic vegetation with pleustophytes or submerged or surfacing rizophytes belonging to Magnopotamion or Hydrochariton of lentic water (Natura 2000 Habitat: 3150) or Ranunculion fluitantis and Callitricho-Batrachion of running waters
Environ Monit Assess (2015) 187: 301
Page 9 of 12 301
Fig. 3 Multitemporal and spatial analysis of the variations in extension of the vegetation types along the river stretch investigated upstream of the dam: aquatic macrophyte vegetation (a), riverbank macrophyte vegetation (b) and riparian forests (c)
(Natura 2000 Habitat: 3260). The area covered by this vegetation type of conservation interest has increased
significantly over time after the dam construction, as the map shows (Fig. 5).
Fig. 4 Cumulative area difference covered by the typical river vegetation (aquatic macrophytes, riverbank macrophytes, riparian forests) between pre- and post-dam period
301 Page 10 of 12
Environ Monit Assess (2015) 187: 301
Fig. 5 Vegetation mapping on the basis of the conservation importance (high, medium, low) of the vegetation types recorded in the buffer area
Discussion Generally, the plant landscape of the investigated river stretch has significantly changed over time. The main transformations in land use and river vegetation recorded therein are due to the multiple action of concurrent factors. Some factors have greatly influenced the area’s vegetation dynamics. In particular: –
–
– Dam construction in 1953: in the first river sectors upstream of the dam, the riverbed widened considerably, leading to a significant expansion of inland waters resulting in the flooding of many agricultural areas adjacent to the riverbanks. The aquatic macrophytes began to colonize and spread in the newly
formed wetlands, while the riverbank macrophytes and riparian forests began to settle and develop along the new banks. Adoption of an economic policy that has led to the progressive abandonment of agricultural and pastoral activities along the riverbanks, also due to the flooding of many such areas as a result of the dam construction. Adoption of a protection policy of the area as Natural Reserve in the 1979. The reserve management, aimed at protecting the natural heritage and thus the plant and animal communities present therein, has probably led to a further expansion of the macrophytes, both aquatic and riverbank communities, and a settling of the new riparian forests
Environ Monit Assess (2015) 187: 301
that began to expand. More recently (2000), the positive effect linked to the protection policy adopted in the area enhanced the “favouring” dam effect on the river vegetation, causing indirectly its expansion. In 2000, therefore, the combined effects of these factors were more evident than in 1984. It should be underlined that, locally-speaking, the dam-factor is undoubtedly the main force driving the major changes in river vegetation. Although the dam construction has created a strong local hydromorphological alteration of the river ecosystem, thus transforming it into a sub-lacustrine habitat, it has also created optimal hydrological conditions for macrophyte development. On the one hand, such conditions have favoured the expansion of some river plant communities already existing there; on the other hand, they have enabled the development of new aquatic communities typical of lentic environments, especially favouring the rooting of aquatic macrophytes as well as helophytes and geophytes (e.g. Phragmites australis, Iris pseudoacorus) able to build reasonably sized islets within the main channel. It should be noted that some of these vegetation types linked to the river habitat are also included among communities of high naturalistic interest sensu Habitats Directive (e.g. riparian forest with S. alba, S. purpurea and P. alba, and aquatic macrophyte vegetation with Potamogeton nodosus, Najas marina, Ceratophyllum demersum and Callitriche stagnalis) (see Tabel 1). Therefore, although causing severe alterations in the river ecosystem’s hydromorphological and biological features, the human impact following the hydroelectric dam construction has favoured the development of riparian and aquatic communities. A greater number of vegetation types than the natural aquatic environment pre-existing the dam construction. The dam operation has also allowed the formation of wetlands (i.e. man-made wetland) of considerable naturalistic importance not only for the aquatic and riparian vegetation types occurring therein but also as a food resource or refuge and/or nesting habitat for several species of fish and water birds. This evidence was confirmed by the recognition of the area as Regional Natural Reserve, including SCI and SPA, and wetlands of interest according to Birds Directive on wetlands of international importance.
Page 11 of 12 301
Conclusions The investigated river stretch, being subject on the one hand to damming and on the other hand to naturalistic protection as a Natural Reserve, has shown itself to be an ideal ecosystem as a study model. Firstly, this study provided many insights into how a dam and its major hydromorphological alteration in a river ecosystem can impact upon on vegetation pattern over time and in space. Secondly, analysis of this river sector has enabled us to highlight how the adoption of nature conservation strategies can positively influence vegetation dynamics, ensuring maintenance and protection of the community interest vegetation types. It should be stressed that many vegetation types with high naturalistic importance (and so falling in the categories of Habitats Directive) have developed and expanded after the dam construction. Therefore, the damming of the river seems to have had a “favourable” effect from the viewpoint of conservation and increase in biodiversity, at least in terms of vegetation. However, it is should also note that many of the locally developed plant communities are described by the literature as typical of lacustrine ecosystems (e.g. Iberite et al. 1995; Venanzoni et al. 2003; Ceschin and Salerno 2008), therefore different in structure and composition from those expected for the river ecosystem analysed. Acknowledgments The authors are grateful to the staff of the Tiber-Farfa Natural Regional Reserve for their help during fieldwork and the director Dr. Umberto Pessolano of the “Museo del Fiume” at Nazzano, for providing useful historical documentation on site. They also thank Dr. Ilaria Mazzini for the linguistic review of this manuscript.
References AA.VV. (1985). Il Tevere Natura, storia e territorio da Nazzano a Castel Giubileo. Roma: Savelli Editore. AA.VV. (2006). SIC-ZPS IT 6030012 “Riserva Naturale Tevere Farfa”. Arpino, FR: Sintesi del piano di gestione, Arpinate Stampa S.R.L. APAT (2005). La realizzazione in Italia del progetto europeo Corine Land Cover 2000. Roma: Rapporti APAT 36. Bezdek, J. C. (1981). Pattern recognition with fuzzy objective function algorithms. Norwell: Kluwer Academic Publishers. Blasi, C. (1994). Fitoclimatologia del Lazio. Fitosociologia, 27, 151–175. Braun-Blanquet, J. (1932). Plant sociology: The study of plant communities. New York: McGraw-Hill.
301 Page 12 of 12 Ceschin, S., & Salerno, G. (2008). La vegetazione del basso corso del Fiume Tevere e dei suoi affluenti (Lazio, Italia). Fitosociologia, 45(1), 39–74. Ceschin, S., Zuccarello, V., & Caneva, G. (2010). Role of macrophyte communities as bioindicators of water quality: application on the Tiber River basin (Italy). Plant Biosystems, 144(3), 528–536. Crivelli, A. J., Grillas, P., & Lacaze, B. (1995). Responses of vegetation to a rise in water level at Kerkini Reservoir (1982–1991), a Ramsar site in Northern Greece. Environmental Management, 19(3), 417–430. D’Antoni, S., Pacini, A., Cocchieri, G., Pittiglio, C., & Reggiani, G. (2002). L’impatto della nutria (Myocastor coypus) nella Riserva Naturale Tevere-Farfa (RM). Firenze: Atti Conv. Naz. La gestione delle specie alloctone in Italia: il caso della nutria e del gambero rosso della Louisiana. De Caceres, M., Oliva, F., Font, X., & Vives, S. (2007). Ginkgo, a program for non-standard multivariate fuzzy analysis. Fuzzy Sets and Systems, 2(1), 41–56. Dunn, J. C. (1973). A fuzzy relative of the ISODATA process and its use in detecting compact well-separated clusters. Cybernetics and Systems, 3(3), 32–57. Evans, J. E., Huxley, J. M., & Vincent, R. K. (2007). Upstream channel changes following dam construction and removal using a GIS/remote sensing approach. Journal of the American Water Resources Association, 43(2), 683–697. Fanelli, G., Bianco, P. M., Cazzagon, P., D’Angeli, D., De Sanctis, M., Bertarelli, M., et al. (2007). Banche dati vegetazionali della Provincia di Roma. Roma: Memoria illustrativa della carta della vegetazione della Provincia di Roma, Assessorato alle Politiche del Territorio. Franchi, E., Carosi, A., Ghetti, L., Giannetto, D., Pedicillo, G., Pompei, L., & Lorenzoni, M. (2014). Changes in the fish community of the upper Tiber River after construction of a hydro-dam. Journal of Limnology, 73(2), 203–210. Gallo, M. (1983). La riserva naturale Nazzano-Tevere-Farfa. Roma: Il Comune democratico 2, Edizione delle Autonomie. Iberite, M., Palozzi, A. M., & Resini, A. M. (1995). La vegetazione del Lago di Bolsena (Viterbo, Italia centrale). Fitosociologia, 29, 151–164. ISPRA. (2010). Analisi conclusive relative alla cartografia Corine Land Cover 2000. Roma: Rapporti ISPRA.
Environ Monit Assess (2015) 187: 301 Jansson, R., Nilsson, C., Dynesius, M., & Andersson, E. (2000a). Effects of river regulation on river-margin vegetation: a comparison of eight boreal rivers. Ecological Applications, 10(1), 203–224. Jansson, R., Nilsson, C., & Renöfält, B. (2000b). Fragmentation of riparian floras in rivers with multiple dams. Ecology, 81(4), 899–903. New, T., & Xie, Z. (2008). Impacts of large dams on riparian vegetation: applying global experience to the case of China’s Three Gorges Dam. Biodiversity and Conservation, 17(13), 3149–3163. Nilsson, C., & Berggren, K. (2000). Alterations of riparian ecosystems caused by river regulation. Bioscience, 50(9), 783–792. Nilsson, C., & Jansson, R. (1995). Floristic differences between riparian corridors of regulated and free-flowing b o r e a l r i v e r s . R e g u l a t e d R i v e r s : R e s e a rc h & Management, 11(1), 55–66. Nilsson, C., Jansson, R., & Zinko, U. (1997). Long-term responses of river-margin vegetation to water level regulation. Science, 276, 798–800. Orloci, L. (1978). Multivariate analysis in vegetation research. Hague: W. Junk BV. Sambucini, V., Marinosci, I., Bonora, N., & Chirici, G. (2010). La realizzazione in Italia del Progetto europeo Corine Land Cover 2006. Roma: Rapporti ISPRA. Spada, F. (2006). Carta della vegetazione della Riserva Naturale Regionale “Tevere-Farfa” scala 1:10.000. Roma: Regione Lazio, R.N.R. Tevere Farfa. Spada, F., & Casella, L. (2006). Memoria illustrativa alla Carta della Vegetazione SIC/ZPS IT 6030012 “Riserva Naturale Tevere Farfa”. Roma: Regione Lazio Parchi e Riserve Naturali, R.N.R. Tombolini, I., Caneva, G., Cancellieri, L., Abati, S., & Ceschin, S. (2014). Damming effects on riparian and aquatic vegetation: the Nazzano case study (Tiber River, central Italy). Knowledge and Management of Aquatic Ecosystems 412. doi:10.1051/kmae/2013085. Venanzoni, R., Apruzzese, A., Gigante, D., Suanno, G., & Vale, F. (2003). Contributo alla conoscenza della vegetazione acquatica e igrofitica del Laghi di Monticchio. Informatore Botanico Italiano, 35, 69–80.
The effect of river damming on vegetation: is it always unfavourable? A case study from the River Tiber (Italy) Simona Ceschin & Ilaria Tombolini & Silverio Abati & Vincenzo Zuccarello
Received: 25 November 2014 / Accepted: 9 April 2015 / Published online: 29 April 2015 # Springer International Publishing Switzerland 2015
Abstract River damming leads to strong hydromorphological alterations of the watercourse, consequently affecting river vegetation pattern. A multitemporal and spatial analysis of the dam effect on composition, structure and dynamic of the upstream vegetation was performed on Tiber River at Nazzano-dam (Rome). The main research questions were as follows: How does plant landscape vary over time and along the river? Where does the dam effect on vegetation end? How does naturalistic importance of the vegetation affected by damming change over time? Data collection was performed mapping the vegetation in aerial photos related to the period before (1944), during (1954) and after dam construction (1984, 2000). The plant landscape has significantly changed over time and along the river, particularly as a result of the dam construction (1953). The major vegetation changes have involved riparian forests and macrophytes. Dam effect on vegetation is evident up to 3 km, and gradually decreases along an attenuation zone for about another 3 km. Despite the fact that the damming has caused strong local hydromorphological modification of the river ecosystem transforming it into a sub-lacustrine habitat, it has also led to the formation S. Ceschin (*) : I. Tombolini : S. Abati Department of Sciences, Roma Tre University, V.le G. Marconi 446, 00146 Rome, Italy e-mail: [email protected] V. Zuccarello Department of Sciences and Biological and Environmental Technology, Salento University, St. Prov. Lecce-Monteroni, Polo Ecotekne, 73100 Lecce, Italy
of wetlands of considerable naturalistic importance. Indeed, in these man-made wetlands, optimal hydrological conditions have been created by favouring both the expansion of pre-existing riparian communities and the rooting of new aquatic communities, albeit typical of lacustrine ecosystems. Some of these plant communities have become an important food resource, refuge or nesting habitats for aquatic fauna, while others fall into category of Natura 2000 habitats. Therefore, river damming seems to have indirectly had a “favourable” effect for habitat conservation and local biodiversity. Keywords Hydroelectric dam . Habitats directive . River vegetation . Multitemporal-spatial analysis . Land use change
Introduction River damming is one of the most serious anthropogenic disturbances impacting on river ecosystem, since it generally leads to strong alterations in the watercourse’s hydromorphological and biological features. Some studies have focused on the dam effects on the upstream river stretch, among which the most important are the increase in water level with subsequent flooding of the surrounding riparian lands (Crivelli et al. 1995; Nilsson and Berggren 2000; New and Xie 2008) and the reduction in the river water velocity with formation of sublacustrine habitats (Nilsson and Jansson 1995; Jansson et al. 2000a, b; Evans et al. 2007; Franchi et al. 2014).
301 Page 2 of 12
These hydromorphological changes of the river ecosystem strongly influence the structure and distribution of the upstream aquatic and riparian macrophyte communities (Nilsson and Berggren 2000; New and Xie 2008). Despite these studies, research into multitemporal and spatial effects of the river damming on composition, structure and dynamic of the plant communities is scarce, especially as regards upstream vegetation (Nilsson et al. 1997; Tombolini et al. 2014). In our own recent study about the vegetation of the lower course of the Tiber River at Nazzano-dam (Rome) (Tombolini et al. 2014), we came to the conclusion that the aquatic and riparian communities are those most affected by the dam operation, i.e. by the resulting alteration of some hydrological parameters, in particular the river water level. In this paper, we have investigated the dam effects on the upstream river landscape; in particular, we aimed to answer to the following main research questions:
Environ Monit Assess (2015) 187: 301
period in July and August and a cold stress period that lasts from October to May (Blasi 1994). From a hydrological point of view, at the analysed river stretch, the estimated average flow is 189 m3/s (Gallo 1983). The Nazzano-dam has a limited capacity for modifying the water regime, considering the similar flow water between the related upstream and downstream river stretches. At Nazzano Lake, the monthly variability of the water level is kept low by the maintenance of the reservoir water level at the spillway crest elevation throughout the year. Moreover, the reservoir’s storage capacity represents a small fraction of the annual average runoff (about 5.9 billion m3) and allows only a regulation on a daily basis. For these features, the Nazzano Lake can be considered as a run-of-the-river impoundment (Tombolini et al. 2014).
Methods 1. How do vegetation and river landscape vary upstream from the dam over time and along the river? 2. Where does the damming effect on vegetation end? 3. How does the naturalistic importance of the vegetation affected by damming change over time?
Study area The study area is located along the lower course of the Tiber River at Nazzano (40 km North of Rome, Latium) and it is immediately upstream of a hydroelectric dam, the Nazzano-dam, operational since 1956 (Fig. 1). The selected river stretch is 8.5 km long and covers a surface of about 300 ha. After the dam came into operation, a sector of this river stretch became transformed into a sub-lacustrine ecosystem, the Nazzano Lake, whose storage capacity is about 18 million m3 (AA.VV. 1985). This river stretch is included in the Tiber-Farfa Natural Regional Reserve, which was founded in 1979 (1979/21 R.L.) and includes a Site of Community Importance and a Special Protection Area (SCI/SPA IT6030012) (92/43/CEE, 79/409/CEE Directives) (AA.VV. 2006). As regards the climatic features, the area falls within the Transitional Temperate region, characterized by lower humid ombrotype and upper Mesomediterranean thermotype. It is a semi-arid mesoclimate with an arid
Data collection and mapping The multitemporal and spatial analysis of the vegetation along the selected river stretch was performed by using aerial photos of the summer season from 1944, 1954, 1984 to 2000, related to the period before (1944), during (1954) and after the Nazzano dam construction (1984, 2000). Firstly, we digitalized the vegetation in the four aerial photos and analysed the data with ArcGis 9.0 (ESRI inc. Redland, CA). Different vegetation types were mapped at an approximate scale of 1:10.000. Subsequently, the mapped vegetation types were verified in the field by using the phytosociological method of Braun-Blanquet (1932) and consulting the vegetation surveys available in the literature concerning the area (D’Antoni et al. 2002; Spada 2006; Spada and Casella 2006; Fanelli et al. 2007; Ceschin and Salerno 2008; Ceschin et al. 2010). The vegetation mapping in the historical aerial photos was carried backward, starting from the 2000 vegetation map, considered as representative of the current vegetation. The characterization of the vegetation types through phytosociological surveys allowed assigning a land use category to each identified vegetation type. We used the CORINE Land Cover Legend (APAT 2005; Sambucini et al. 2010), according to ISPRA (2010) for the fourth level of detail.
Environ Monit Assess (2015) 187: 301
Page 3 of 12 301
Fig. 1 Lower course of the Tiber River crossing the Latium region in Central Italy (a). The investigated river stretch subdivided into 153 transects. The dam location is indicated (arrow) (b)
On each of the obtained land use maps, a continuous buffer area starting from the Nazzanodam (lower limit) to the border of the Tiber-Farfa Natural Regional Reserve (higher limit), placed at a distance of about 8.5 km, was defined. The buffer area, including the river bed and both the riparian zones along the selected river stretch, was divided into a series of transversal transects 50 m wide on each riverbank and with a constant surface area (153 transects) (Fig. 1b). The extension of each land use category falling in each transect was calculated in the four land use maps and compared over the 1944–2000 time period. The set of these data has been the basis on which we carried out the subsequent statistical analyses. The land use categories for each time period were related to the habitats defined by Habitats Directive 92/ 43/EEC (Natura 2000 Habitat) in order to assign them a different naturalistic importance.
Statistical analysis Statistical analyses were performed in order to evaluate how the dam operation has affected the river landscape and in particular the vegetation pattern. The analyses were as follows: (i) Analysis of the land use transformations over time and along space. The matrix of the transects relative to different periods (1944, 1954, 1984, 2000), described by the percentages of the land use categories, was performed by fuzzy c-means clustering (Bezdek 1981). Identification of optimal fuzzy partition of the transects based on normalized Dunn coefficient (Dunn 1973) was carried out. In this way, each fuzzy cluster identifies a characteristic land use type. (ii) Identification of homogenous river sectors in a land use transformation perspective. The matrix
301 Page 4 of 12
of the degrees of belonging to the fuzzy clusters in the different years was analysed by hierarchical classification (complete linkage), based on chord distance (Orloci 1978). From the resulting dendrogram, we identified homogenous river sectors. The matrix of the river sector centroids related to the degrees of belonging to land use types (identified in the previous step) during the four periods was used to identify the land use types prevailing in each time period. The land use type that showed the greatest degree of belonging at time t was defined as dominant in the river sector at time t. For the statistical analyses, the software GINKGO ver. 1.7.0 (De Caceres et al. 2007) was utilized.
Results How do river vegetation and landscape vary upstream from the dam over time and along the river? During the four time periods in question, the land use mapping in the buffer area has allowed the identification of 13 cartographic types related to the same number of land use categories (Table 1). These categories refer to three main typologies: (1) artificial areas and cultivated vegetation, (2) spontaneous forests or grasslands and (3) riparian forests and riverbank and aquatic macrophyte vegetation. The variations in extension of the 13 land use categories in pre- and post-dam period are shown in Table 2. The period 1954–1984 appears to be that in which the major land use changes occurred. The most evident changes have affected, on the one hand, the inland waters, and riverbank and aquatic macrophyte vegetation which have greatly increased in surface, and, on the other hand, the agricultural areas and natural grasslands, whose coverage has instead been decreasing. The optimal partition obtained by fuzzy clustering identifies 5 fuzzy groups corresponding to the different land use modalities. The centroids of the land use categories of each group are shown in Table 3. On the basis of these data, it is possible to describe the significance of each group. Groups 1, 2, and 4 are characterized by a predominantly agricultural land use: this is highest in group 2 and intermediate in groups 1 and 4. In the latter group, the agricultural use is combined with wide pasture areas and natural grassland replacement. The
Environ Monit Assess (2015) 187: 301
artificial areas are greater in group 3, in which, to compensate the presence of infrastructures, there is the most abundant presence of riparian forests. In this group, the agricultural exploitation is very low. Group 5 is characterized by aquatic and riverbank macrophyte vegetation, as well as by spontaneous deciduous woods. Also in this case, the areas for agricultural use are of very limited extent. The inland waters category, showing high values in all the five fuzzy groups, was not considered as a characteristic element of any fuzzy cluster, although its area has increased significantly as a result of the dam operation (Table 2). The hierarchical classification of transects, described on the basis of the degrees of belonging to the fuzzy clusters during the considered period (1944–2000), distinguishes five main groups (not to be confused with the five fuzzy clusters identified previously). Each group identifies a river sector consisting of several transects more or less spatially contiguous. The five identified river sectors are shown in Fig. 2. Sector 1 (S1) represents mainly the river stretch immediately upstream of the confluence Tiber-Farfa; sector 2 (S2) corresponds to two small river stretches in the meandering zone with double curve of the River Tiber, relatively distant from the dam; sector 3 (S3) identifies different sections of the river, although all these are upstream of the Tiber-Farfa confluence; sector 4 (S4) is exclusively a contiguous river stretch not far from the dam and includes both the confluence Tiber-Farfa zone and some transects just upstream and downstream of it; and sector 5 (S5) mainly identifies the first river stretch upstream of the dam, as well as few transects in the meandering zone. The dynamic of the prevailing land use type in the various river sectors, on the basis of the matrix of the fuzzy clusters centroids in the different time periods, is shown in Table 4. In 1944 and 1954 (pre-dam period), agricultural land use is evident throughout the investigated area, i.e. along all the river sectors identified (from S1 to S5). The only difference between 1944 and 1954 is linked to the different extension of the natural grasslands in respect to agricultural areas in the sectors S1 and S2; indeed, if natural grasslands are prevalent in 1944 in S2, they are dominant in 1954 in S1. In 1984, considering the river sectors more distant from the dam, it is observed that S2 maintains a mainly agricultural land use, while S1 and S3 remain of agricultural use although not exclusively, evidence of a reduction of agricultural activities. In general, however,
Environ Monit Assess (2015) 187: 301
Page 5 of 12 301
Table 1 Land use categories recognized in the study area with the corresponding Corine Land-Cover categories, Natura 2000 Habitat types and syntaxonomical classification (phytosociological class) Land use category
Description
Corine LandCover codex
Natura Habitat 2000 (codex)
Phytosociological class
Artificial surface
Urban area, artificial fabric, mine, dump, sand and gravel drainage Cereal and tree crops
1
–
–
2.1+2.2
–
–
–
Quercetea ilicis
–
Querco-Fagetea
Agricultural areas
Broad-leaved forest Forest with Quercus ilex 3.1.1.1 with evergreen oaks Broad-leaved forest Mixed deciduous forest with Quercus 3.1.1.2 with deciduous oaks cerris Broad-leaved forest Riparian forest with Populus alba, 3.1.1.6 with hygrophilous Salix alba or Salix purpurea species
Natural grassland
3.2.3
–
Quercetea ilicis/ Rosmarinetea/ Helianthemetea guttatae
3.2.4
–
Artemisietea vulgaris
3.3
–
–
4.1.3
–
PhragmitoMagnocaricetea/ MolinioArrhenatheretea
4.1.3
Eutrophic natural lakes with Magnopotamion or Hydrocharition vegetation (3150), Mountain rivers and plains with Ranunculion fluitantis and CallitrichoBatrachion vegetation. –
Potametea pectinati
Meadows subject to flooding, 3.2.1 grasslands replacement and pasture areas
Bushes and heathland
Bushes of substitution with Crataegus monogyna, Cornus sanguinea, Cornus mas, Pyrus amygdaliformis, Spartium junceum), thickened by tangles of bramble (Rubus ulmifolius, R. caesius) Sclerophyllous Mosaic of evergreen bushes and vegetation vegetation of substitution of oak forests or of early evolutive stages of secondary succession towards the holm oak forest Transitional wood Vegetation of substitution in shrub anthropized areas with Arundo donax or communities with woody (Ulmus minor, Rubus ulmifolius, Robinia pseudoacacia, Ailanthus altissima) or herbaceous species, in landfill environments or near to abandoned fields Open spaces with little Areas sparsely vegetated or subject to or no vegetation sedimentation. Riverbank macrophyte Riverbank herbaceous vegetation vegetation with Phragmites australis, Iris pseudoacorus, Typha latifolia, Carex pseudocyperus or with Paspalum paspaloides and Agrostis semiverticillatum Aquatic macrophyte Aquatic vegetation with vegetation Potamogeton nodosus, Najas marina, Ceratophyllym demersum, Callitriche stagnalis
Inland waters
3.2.2
Forest dominated by Salix alba and Querco-Fagetea/ Populus alba (92A0) / Standing Salicetea flowing Mediterranean rivers purpurea with Paspalo-Agrostidion and with riparian tree row dominated by Salix spp. and Populus alba (3280). Secondary perennial grasslands Festuco-Brometea with Bromus erectus or Dactilys glomerata growing on increased edaphic water retention (6210). – Rhamno-Prunetea
Waters of the Tiber and Farfa rivers
5.1
–
301 Page 6 of 12 Table 2 Variations in extension (ha) of the 13 land use categories in pre- and post-dam period
Environ Monit Assess (2015) 187: 301
Land use category
1944–1954
1984–2000
Aquatic macrophyte vegetation
1.81
7.29
5.73
Riverbank macrophyte vegetation
4.69
11.62
−7.48
Riparian forest
1.80
5.06
−7.06
−10.29
37.89
6.24
0.82
−10.89
1.76
Transitional wood shrub
−0.02
−1.22
0.44
Sclerophyllous vegetation
−0.30
−0.51
0.64
Broad-leaved forest with evergreen oaks
−0.53
−0.07
0.61
Broad-leaved forest with deciduous oaks
Inland waters Natural grassland
−1.39
2.03
−2.39
Bushes and heathland
1.33
1.18
3.94
Agricultural areas
0.35
−56.27
1.21
−3.26
−0.97
−0.28
5.00
4.86
−3.35
Open spaces with little or no vegetation Artificial surface
agricultural land use continues to be predominant in the river sectors upstream of the Tiber-Farfa confluence. Instead, sectors S4 and S5, closer to the dam, have undergone a major change in land use; indeed in these sectors, it is possible to observe a transformation from a mainly agricultural use during the pre-dam period to a strong development of macrophyte vegetation (and deciduous forests) (S4) as well as of riparian forests (S5) in the post-dam period. Therefore, macrophyte and riparian vegetation have been the vegetation types most affected by the dam operation. In S4, the extension of the deciduous forest area is related to the abandonment of agriculture activities.
In 2000, we observe the same situation recorded in 1984 with no further changes in the land use type. Therefore, there are two factors driving land use changes in the river stretches closer to the dam: the dam effect with the consequent local expansion of inland waters, and the reduction of the agricultural activities. In the river stretches more distant from the dam, a more limited land use change has been observed, being a mainly agricultural use conserved over time. Multitemporal analysis of coverage variations of the aquatic and riverbank macrophytes as well as riparian
Table 3 Centroid values of the 13 land use categories for each of 5 fuzzy groups
In italic, the most substantial centroid values
1954–1984
gr1
gr2
gr3
gr4
gr5
Aquatic macrophyte vegetation
0.9
0.9
2.6
0.7
13.9
Riverbank macrophyte vegetation
8.9
4.6
4.7
5.1
11.0
Riparian forest
12.6
10.3
15.3
12.6
7.6 31.9
Inland waters
33.5
24.7
42.6
26.9
Natural grassland
1.4
1.5
3.0
21.8
4.1
Transitional wood shrub
2.1
1.3
3.4
1.7
2.7
Sclerophyllous vegetation
0.3
0.2
0.2
0.1
0.1
Broad-leaved forest with evergreen oaks
0.2
0.3
0.1
0.1
0.1
Broad-leaved forest with deciduous oaks
1.7
2.5
2.9
1.6
21.0
Bushes and heathland
1.7
0.5
5.4
1.7
1.0
Agricultural areas
30.4
50.1
5.6
23.0
3.4
Open spaces with little or no vegetation
0.6
0.2
0.3
0.2
0.5
Artificial surface
5.7
2.9
13.9
4.7
2.7
Environ Monit Assess (2015) 187: 301
Fig. 2 The five river sectors (S) identified along the Tiber River
Page 7 of 12 301
301 Page 8 of 12
Environ Monit Assess (2015) 187: 301
Table 4 Land use changes over time (1944–2000) along the identified river sectors (S1–S5)
river sector S1 S2 S3 S4 S5
1944
1954
1984
2000
g2 g4 g2 g2 g2
g4 g2 g2 g2 g2
g1 g2 g1 g5 g3
g1 g2 g1 g5 g3
forests (Table 2) confirms these findings and underlines the importance of the dam effect on river vegetation, showing local increases in coverage of these vegetation types closer to the river. Where does the damming effect on vegetation end? The results indicate that damming has a strong influence on vegetation in the river sectors closer to the infrastructure itself (i.e. S4 and S5), affecting mainly the aquatic and riverbank macrophyte vegetation and riparian forests. The temporal and spatial dynamics of these vegetation types, more sensitive to the dam effect, were analysed in detail for identifying the spatial threshold of this effect. Therefore, the multitemporal analyses of the variations in extension of these vegetation types along the river are shown in Fig. 3. Specifically, in the post-dam period, the aquatic macrophytes are mostly expanded in the river stretch ranging from the dam to the Tiber-Farfa confluence zone (approximately transects 10–30) (Fig. 3a). The riverbank macrophyte vegetation dominated by reed beds shows an important expansion over time too, especially in this river stretch, with the difference that its expansion area is more extensive than aquatic macrophte in the upstream zone, including approximately the transects 16–64 (Fig. 3b). The riparian forests, such as willow and poplar woods, display a similar multitemporal trend along the entire buffer area, even if they show a general increase in the post-dam period, with the exception of a few transects (Fig. 3c). In the transects closest to the dam, the riverbank vegetation and the riparian forests are decreased or have even disappeared; probably, this is a result of cleaning activities along the banks for the
dam construction, including the removal of riverbank vegetation. The difference in surface covered totally by the typical vegetation of the river ecosystem (i.e. aquatic macrophytes, riverbank macrophytes, riparian forests) between pre- and post-dam operation periods is shown in Fig. 4. This area is calculated as cumulative, i.e. the sum of the differences in extension of these vegetation typologies in the single transects included between the dam and the transect furthest from the dam (i.e. transect 153). It is possible to observe that the area covered by these vegetation types increases in most of the river stretch. Only in transects closest to the dam (up to transect 17), and in those farthest from it, does this area decrease, albeit moderately. The river stretch in which the increase is greater is approximately from transect 18 to transect 30, followed by a significant increase up to transect 60 and a final stretch of attenuated increase up to transect 120. Therefore, excluding those transects closest to the dam, the dam effect on vegetation is evident up to transect 60 (located at a distance of about 3 km from the dam), this continuing although moderately, up to transect 120. The river stretch included between transect 60 and 120 can be considered as attenuation zone of the dam effect, which, instead, is clearly manifested in the previous river stretch.
How does naturalistic importance of the vegetation affected by damming change over time? The classification of the 13 land use categories on the basis of Habitats Directive has made possible to distinguish in the study area (1) vegetation types related to habitats of community interest, (2) natural vegetation types not included in the previous category and (3) anthropogenic vegetation or artificial surfaces, devoid of naturalistic interest (planted forests, crops, etc.). In the first category, there are included locally hygrophilous riparian forests dominated by Salix alba and Populus alba (Natura 2000 Habitat: 92A0), river woods with Salix purpurea (Natura 2000 Habitat: 3280), and aquatic vegetation with pleustophytes or submerged or surfacing rizophytes belonging to Magnopotamion or Hydrochariton of lentic water (Natura 2000 Habitat: 3150) or Ranunculion fluitantis and Callitricho-Batrachion of running waters
Environ Monit Assess (2015) 187: 301
Page 9 of 12 301
Fig. 3 Multitemporal and spatial analysis of the variations in extension of the vegetation types along the river stretch investigated upstream of the dam: aquatic macrophyte vegetation (a), riverbank macrophyte vegetation (b) and riparian forests (c)
(Natura 2000 Habitat: 3260). The area covered by this vegetation type of conservation interest has increased
significantly over time after the dam construction, as the map shows (Fig. 5).
Fig. 4 Cumulative area difference covered by the typical river vegetation (aquatic macrophytes, riverbank macrophytes, riparian forests) between pre- and post-dam period
301 Page 10 of 12
Environ Monit Assess (2015) 187: 301
Fig. 5 Vegetation mapping on the basis of the conservation importance (high, medium, low) of the vegetation types recorded in the buffer area
Discussion Generally, the plant landscape of the investigated river stretch has significantly changed over time. The main transformations in land use and river vegetation recorded therein are due to the multiple action of concurrent factors. Some factors have greatly influenced the area’s vegetation dynamics. In particular: –
–
– Dam construction in 1953: in the first river sectors upstream of the dam, the riverbed widened considerably, leading to a significant expansion of inland waters resulting in the flooding of many agricultural areas adjacent to the riverbanks. The aquatic macrophytes began to colonize and spread in the newly
formed wetlands, while the riverbank macrophytes and riparian forests began to settle and develop along the new banks. Adoption of an economic policy that has led to the progressive abandonment of agricultural and pastoral activities along the riverbanks, also due to the flooding of many such areas as a result of the dam construction. Adoption of a protection policy of the area as Natural Reserve in the 1979. The reserve management, aimed at protecting the natural heritage and thus the plant and animal communities present therein, has probably led to a further expansion of the macrophytes, both aquatic and riverbank communities, and a settling of the new riparian forests
Environ Monit Assess (2015) 187: 301
that began to expand. More recently (2000), the positive effect linked to the protection policy adopted in the area enhanced the “favouring” dam effect on the river vegetation, causing indirectly its expansion. In 2000, therefore, the combined effects of these factors were more evident than in 1984. It should be underlined that, locally-speaking, the dam-factor is undoubtedly the main force driving the major changes in river vegetation. Although the dam construction has created a strong local hydromorphological alteration of the river ecosystem, thus transforming it into a sub-lacustrine habitat, it has also created optimal hydrological conditions for macrophyte development. On the one hand, such conditions have favoured the expansion of some river plant communities already existing there; on the other hand, they have enabled the development of new aquatic communities typical of lentic environments, especially favouring the rooting of aquatic macrophytes as well as helophytes and geophytes (e.g. Phragmites australis, Iris pseudoacorus) able to build reasonably sized islets within the main channel. It should be noted that some of these vegetation types linked to the river habitat are also included among communities of high naturalistic interest sensu Habitats Directive (e.g. riparian forest with S. alba, S. purpurea and P. alba, and aquatic macrophyte vegetation with Potamogeton nodosus, Najas marina, Ceratophyllum demersum and Callitriche stagnalis) (see Tabel 1). Therefore, although causing severe alterations in the river ecosystem’s hydromorphological and biological features, the human impact following the hydroelectric dam construction has favoured the development of riparian and aquatic communities. A greater number of vegetation types than the natural aquatic environment pre-existing the dam construction. The dam operation has also allowed the formation of wetlands (i.e. man-made wetland) of considerable naturalistic importance not only for the aquatic and riparian vegetation types occurring therein but also as a food resource or refuge and/or nesting habitat for several species of fish and water birds. This evidence was confirmed by the recognition of the area as Regional Natural Reserve, including SCI and SPA, and wetlands of interest according to Birds Directive on wetlands of international importance.
Page 11 of 12 301
Conclusions The investigated river stretch, being subject on the one hand to damming and on the other hand to naturalistic protection as a Natural Reserve, has shown itself to be an ideal ecosystem as a study model. Firstly, this study provided many insights into how a dam and its major hydromorphological alteration in a river ecosystem can impact upon on vegetation pattern over time and in space. Secondly, analysis of this river sector has enabled us to highlight how the adoption of nature conservation strategies can positively influence vegetation dynamics, ensuring maintenance and protection of the community interest vegetation types. It should be stressed that many vegetation types with high naturalistic importance (and so falling in the categories of Habitats Directive) have developed and expanded after the dam construction. Therefore, the damming of the river seems to have had a “favourable” effect from the viewpoint of conservation and increase in biodiversity, at least in terms of vegetation. However, it is should also note that many of the locally developed plant communities are described by the literature as typical of lacustrine ecosystems (e.g. Iberite et al. 1995; Venanzoni et al. 2003; Ceschin and Salerno 2008), therefore different in structure and composition from those expected for the river ecosystem analysed. Acknowledgments The authors are grateful to the staff of the Tiber-Farfa Natural Regional Reserve for their help during fieldwork and the director Dr. Umberto Pessolano of the “Museo del Fiume” at Nazzano, for providing useful historical documentation on site. They also thank Dr. Ilaria Mazzini for the linguistic review of this manuscript.
References AA.VV. (1985). Il Tevere Natura, storia e territorio da Nazzano a Castel Giubileo. Roma: Savelli Editore. AA.VV. (2006). SIC-ZPS IT 6030012 “Riserva Naturale Tevere Farfa”. Arpino, FR: Sintesi del piano di gestione, Arpinate Stampa S.R.L. APAT (2005). La realizzazione in Italia del progetto europeo Corine Land Cover 2000. Roma: Rapporti APAT 36. Bezdek, J. C. (1981). Pattern recognition with fuzzy objective function algorithms. Norwell: Kluwer Academic Publishers. Blasi, C. (1994). Fitoclimatologia del Lazio. Fitosociologia, 27, 151–175. Braun-Blanquet, J. (1932). Plant sociology: The study of plant communities. New York: McGraw-Hill.
301 Page 12 of 12 Ceschin, S., & Salerno, G. (2008). La vegetazione del basso corso del Fiume Tevere e dei suoi affluenti (Lazio, Italia). Fitosociologia, 45(1), 39–74. Ceschin, S., Zuccarello, V., & Caneva, G. (2010). Role of macrophyte communities as bioindicators of water quality: application on the Tiber River basin (Italy). Plant Biosystems, 144(3), 528–536. Crivelli, A. J., Grillas, P., & Lacaze, B. (1995). Responses of vegetation to a rise in water level at Kerkini Reservoir (1982–1991), a Ramsar site in Northern Greece. Environmental Management, 19(3), 417–430. D’Antoni, S., Pacini, A., Cocchieri, G., Pittiglio, C., & Reggiani, G. (2002). L’impatto della nutria (Myocastor coypus) nella Riserva Naturale Tevere-Farfa (RM). Firenze: Atti Conv. Naz. La gestione delle specie alloctone in Italia: il caso della nutria e del gambero rosso della Louisiana. De Caceres, M., Oliva, F., Font, X., & Vives, S. (2007). Ginkgo, a program for non-standard multivariate fuzzy analysis. Fuzzy Sets and Systems, 2(1), 41–56. Dunn, J. C. (1973). A fuzzy relative of the ISODATA process and its use in detecting compact well-separated clusters. Cybernetics and Systems, 3(3), 32–57. Evans, J. E., Huxley, J. M., & Vincent, R. K. (2007). Upstream channel changes following dam construction and removal using a GIS/remote sensing approach. Journal of the American Water Resources Association, 43(2), 683–697. Fanelli, G., Bianco, P. M., Cazzagon, P., D’Angeli, D., De Sanctis, M., Bertarelli, M., et al. (2007). Banche dati vegetazionali della Provincia di Roma. Roma: Memoria illustrativa della carta della vegetazione della Provincia di Roma, Assessorato alle Politiche del Territorio. Franchi, E., Carosi, A., Ghetti, L., Giannetto, D., Pedicillo, G., Pompei, L., & Lorenzoni, M. (2014). Changes in the fish community of the upper Tiber River after construction of a hydro-dam. Journal of Limnology, 73(2), 203–210. Gallo, M. (1983). La riserva naturale Nazzano-Tevere-Farfa. Roma: Il Comune democratico 2, Edizione delle Autonomie. Iberite, M., Palozzi, A. M., & Resini, A. M. (1995). La vegetazione del Lago di Bolsena (Viterbo, Italia centrale). Fitosociologia, 29, 151–164. ISPRA. (2010). Analisi conclusive relative alla cartografia Corine Land Cover 2000. Roma: Rapporti ISPRA.
Environ Monit Assess (2015) 187: 301 Jansson, R., Nilsson, C., Dynesius, M., & Andersson, E. (2000a). Effects of river regulation on river-margin vegetation: a comparison of eight boreal rivers. Ecological Applications, 10(1), 203–224. Jansson, R., Nilsson, C., & Renöfält, B. (2000b). Fragmentation of riparian floras in rivers with multiple dams. Ecology, 81(4), 899–903. New, T., & Xie, Z. (2008). Impacts of large dams on riparian vegetation: applying global experience to the case of China’s Three Gorges Dam. Biodiversity and Conservation, 17(13), 3149–3163. Nilsson, C., & Berggren, K. (2000). Alterations of riparian ecosystems caused by river regulation. Bioscience, 50(9), 783–792. Nilsson, C., & Jansson, R. (1995). Floristic differences between riparian corridors of regulated and free-flowing b o r e a l r i v e r s . R e g u l a t e d R i v e r s : R e s e a rc h & Management, 11(1), 55–66. Nilsson, C., Jansson, R., & Zinko, U. (1997). Long-term responses of river-margin vegetation to water level regulation. Science, 276, 798–800. Orloci, L. (1978). Multivariate analysis in vegetation research. Hague: W. Junk BV. Sambucini, V., Marinosci, I., Bonora, N., & Chirici, G. (2010). La realizzazione in Italia del Progetto europeo Corine Land Cover 2006. Roma: Rapporti ISPRA. Spada, F. (2006). Carta della vegetazione della Riserva Naturale Regionale “Tevere-Farfa” scala 1:10.000. Roma: Regione Lazio, R.N.R. Tevere Farfa. Spada, F., & Casella, L. (2006). Memoria illustrativa alla Carta della Vegetazione SIC/ZPS IT 6030012 “Riserva Naturale Tevere Farfa”. Roma: Regione Lazio Parchi e Riserve Naturali, R.N.R. Tombolini, I., Caneva, G., Cancellieri, L., Abati, S., & Ceschin, S. (2014). Damming effects on riparian and aquatic vegetation: the Nazzano case study (Tiber River, central Italy). Knowledge and Management of Aquatic Ecosystems 412. doi:10.1051/kmae/2013085. Venanzoni, R., Apruzzese, A., Gigante, D., Suanno, G., & Vale, F. (2003). Contributo alla conoscenza della vegetazione acquatica e igrofitica del Laghi di Monticchio. Informatore Botanico Italiano, 35, 69–80.
