Aquatic macrophytes in the large , sub-tropical Itaipu Reservoir , Brazil

In the last three decades, rapid assessment surveys have become an important approach for measuring aquatic ecosystem biodiversity. These methods can be used to detect anthropogenic impacts and recognize local or global species extinctions. We present a floristic survey of the aquatic macrophytes along the Brazilian margin of the Itaipu Reservoir conducted in 2008 and compare this with a floristic survey conducted ten years earlier. We used ordination analysis to determine whether assemblage composition differed among reservoir arms. Macrophyte species were sampled in each of the 235 sampling stations using a boat, which was positioned inside three places of each macrophyte stand to record species and search for small plants. We also collected submerged plants using a rake with the boat moving at constant velocity for ten minutes. We assigned individual macrophyte species to life form and identified representative species for each life form. A total of 87 macrophyte taxa were identified. The “emergent” life forms contained the highest number of species, followed by “rooted submerged” life forms. The extensive survey of macrophytes undertaken in September 2008 recorded more species than a survey conducted between 1995 and 1998. This could be due to changes in water physico-chemistry, disturbances due to water drawdown and the long period between surveys, which may have allowed natural colonization by other species. Additionally, differences in the classification systems and taxonomic resolution used in the surveys may account for differences in the number of species recorded. Assemblage composition varied among the arms and was affected by underwater radiation (as measured using a Secchi disk) and fetch. Five non-native species were found. Two of these non-native species (Urochloa subquadripara and Hydrilla verticillata) are of special concern because they have a high frequency of occurrence and occupy large marginal areas of the reservoir. Future surveys should be conducted to determine the habitat most frequently colonized by these species. This would allow management strategies to be developed to protect native aquatic biota and prevent interference with the recreational and commercial uses of the Itaipu Reservoir. Rev. Biol. Trop. 58 (4): 1437-1452. Epub 2010 December 01.

Rapid assessment style surveys have become important tools to assess the status of biodiversity, especially in the last three decades because of the recognition that anthropogenic impacts are leading several species to local or global extinction (Maltchik & Callisto 2004).By conducting floristic surveys, one can quantify species richness, identify possible threats to biodiversity, assess rare species and detect the presence of non-native, nuisance species in an ecosystem.
Macrophytes are important components of freshwater ecosystems because they enhance the physical structure of habitats and biological complexity, which increases biodiversity within littoral zones (Esteves 1998, Wetzel 2001, Agostinho et al. 2007, Pelicice et al. 2008).For example, fish diversity is positively related to macrophyte diversity in the Upper Paraná River Basin at both the large (between different ecosystems) and fine (within specific ecosystems) spatial scales (Agostinho et al. 2003, 2007, Pelicice et al. 2008).In addition, both live and dead material (detritus) from aquatic macrophytes may serve as a food resource for aquatic and terrestrial organisms (Lopes et al. 2007).
Surveys of macrophytes have been conducted in several types of ecosystems in Brazil, including rivers (Pedralli et al. 1993), lakes (Pott et al. 1992, Kita & Souza 2003) and reservoirs (Thomaz et al. 1999, Martins et al. 2008).In addition to quantifying biodiversity, surveys can provide ecological information regarding, for example, the presence of introduced species, increases or decreases in the frequency of native species with time and identify locations that have been colonised by species that could cause excessive, damaging growth and surveys are particularly useful when they are conducted in more than one time period within the same ecosystem.For example, surveys of the Itaipu Reservoir were conducted prior to its construction and 15 years after its construction (Thomaz et al. 1999).Because the Itaipu reservoir has been undergoing temporal changes (Thomaz et al. 2006), surveys could provide ecological information on the temporal trends of macrophytes within this reservoir.
Different abiotic factors affect aquatic macrophytes.Submerged species, for example, are in general limited by underwater radiation, while floating forms are more limited by nutrients (Bini et al. 1999, Camargo et al. 2003).Despite the ecological importance of aquatic macrophytes, the excessive growth of some species may be considered a nuisance to some reservoir uses, such as recreation and electrical generation (Pieterse & Murphy 1990).The most frequent impacts are caused by the floating species Eichhornia crassipes (Mart.)Solms and Salvinia spp.and the submerged species Egeria densa Planch.and Egeria najas Planch.
The construction of reservoirs causes major changes to rivers; for example, current velocity is reduced, shorelines may become more developed and sediment stability and water transparency may be affected (Agostinho et al. 2007).Moreover, depending on the limnological characteristics of the river, the reservoir may undergo rapid eutrophication, which is typically worst in the first reservoir of a series (Barbosa et al. 1999).These changes can lead to the invasion and rapid development of different species and life forms of aquatic macrophytes (Pieterse & Murphy 1990).In addition to nutrient contents, other physical characteristics, such as underwater radiation and fetch (a surrogate of wave disturbance), may also affect the composition of macrophyte assemblages (Thomaz et al. 2003, Bini et al. 1999).
The Itaipu Reservoir is located downstream of many conservation areas on the Paraná River Basin, which has high biodiversity.It is the largest hydroelectric reservoir in the world in terms of electric energy generation, producing 20% of the total energy consumed in Brazil and 94% of the energy in Paraguay.Macrophytes may become detached and obstruct turbines (although it never happened in Itaipu) and consequently, the monitoring of aquatic plants is important to energy production.
In this study, we present a floristic survey of the aquatic macrophytes along the Eastern margin (Brazilian border) of the Itaipu Reservoir; we compared the species composition of aquatic macrophytes in this survey with a floristic register collected between 1995 and 1998.We assigned individual macrophyte species to life form groups to determine the dominant life forms within the reservoir.Finally, using ordination techniques, we compared the composition among the reservoir arms and assessed the influence of abiotic factors (underwater light and fetch) on macrophyte composition.Using Cook's classification, aquatic macrophytes were defined as plants with photosynthetic parts that are permanently or temporarily submerged or floating in water and visible with the human eye (Cook 1990).

MATERIALS AND METHODS
Study Area: This study was conducted in eight arms of the Brazilian margin of the Itaipu Reservoir (Eastern margin; 24°15' and 25°33' S; 54°00' and 54°37' W (Fig. 1).The reservoir was formed in October 1982 in the Paraná River.It has an area of 1 350km 2 at the normal operation water level (ca.220m.a.s.l.) and a theoretical residence time of approximately 40 days.The water levels are relatively stable and typically fluctuate less than one meter per year; however, a conspicuous water drawdown of five meters occurred between November 1999 and March 2000 (Thomaz et al. 2006).

Sampling:
The main results in this study were collected in a survey of aquatic macrophytes conducted from August 28 to September the 11, 2008.Surveys in this reservoir began in 1995 and in 1999, 235 permanent sampling stations were positioned in eight arms on the Brazilian side of the reservoir.Since 1999, surveys were conducted every six months at these stations and euhydrophytes (especially free floating and submerged species) were prioritized in these surveys.In this paper, we used only data obtained in August-September 2008 because during this time, the taxonomic resolution was improved with macrophyte taxonomists participation.
Thirty sampling points were selected in each arm except at the Arroio Guaçu River and Pinto River, where we selected 26 and 29 sampling points, respectively, because their areas are smaller.The points were distributed to maximize the diversity of abiotic conditions (pH, conductivity, turbidity, dissolved oxygen and nutrient availability) within each arm.The geographic position of each sampling point (latitude and longitude) was obtained by GPS (Global Positioning System).Macrophyte species were sampled in each of the 235 sampling stations using a boat, which was positioned inside three places of each macrophyte stand to record species and search for small plants (Lemna sp., Spirodela sp. and Utricularia sp.).The species presence/absence was recorded by three persons from a boat moving at a constant, slow velocity along the shoreline for ca.80-100m.In order to record data for submerged plants, we lifted these plants out of the water using a rake attached to a 4m long pipe.Although this method may not be precise in terms of covering the surveyed area in each stand, we used this same procedure in all stands.In addition, previous observations have shown that this procedure usually accurately estimates the number of species per stand (data not shown).
Species that could not be identified in the field were collected for later identification in the laboratory and kept in the Herbarium of the University of Maringá (HUEM).The submerged species were fixed in 70% alcohol.Taxonomic identification followed the specialized literature (Cook 1990, Kissman 1997, Kissman & Groth 1999, 2000, Lorenzi 2000, Pott & Pott 2000).Scientific names followed the APG II classification system (2003) and the spelling of names was revised according to the database of the Missouri Botanical Garden (2009).
Species frequencies were obtained by dividing the number of occurrence of a species by 235 (total number of sampling stations).We also classified species according to their life forms because each life form colonises and uses water and sediment resources quite differently.In fact, different life forms occupy distinct positions in the water column (free floating, submerged and emergent), have different access to light (underwater and/or abovewater) and nutrients (sediment and/or water column) and consequently, these groups have different roles (or functions) in littoral regions.We based our seven life forms classification on Pedralli (1990): emergent, rooted submerged, free floating, amphibious, floating leaved, free submerged and epiphyte.
Analysis: To investigate patterns in assemblage composition between reservoir arms, we calculated axis scores using a detrended correspondence analysis (DCA) and used an analysis of variance (ANOvA) to assess if there were significant differences among arms.In addition, correlations between the original data matrix and the DCA scores matrix were calculated and the species correlated to each axis were examined.We correlated the scores of the first (most important) axis of the DCA with two abiotic variables, underwater light (Secchi disk) and effective fetch, to assess if these variables affected macrophyte assemblage composition.Measurements of fetch followed Thomaz et al. (2003).We only used these two variables because previous investigations showed that they are important factors in structuring macrophyte assemblages in Itaipu (Bini et al. 1999, Thomaz et al. 2003) and because measurements of these two variables were obtained in each of the 235 sampling stations.

RESULTS
A total of 87 taxa of aquatic macrophytes belonging to 34 families and 57 genera were identified (Appendix 1).Of these, 60 taxa were identified to species level, but the lack of reproductive structures prevented some plants from being identified to species.The families with the highest number of species were Poaceae (14 taxa), Cyperaceae and Polygonaceae (seven taxa each) and Characeae, Araceae, Hydrocharitaceae, Onagraceae (four taxa each).These families contributed 50.5% of the species recorded and together with Pontederiaceae, Salviniaceae and Potamogetonaceae, had the highest frequencies of occurrence in all arms (Fig. 2a).
The genus with the highest number of species was Polygonum (seven species), followed by Eleocharis (five), Ludwigia and Panicum (four species each).Two genera had three species each, ten genera had two species each and 41 genera were represented by a single species.The life form containing the highest number of species was emergent (52.9% of total), followed by rooted submerged (14.9%), amphibious (12.6%), free floating (9.2%), floating leaved (5.7%), free submerged (3.5%) and epiphyte (1.2%).Similarly, the frequency of occurrence of the life forms followed the same sequence (Fig. 2b).The most frequent genera  It is worth noting that we found five non-native species: U. subquadripara, H. verticillata, Nymphaea caerulea Savigny, Coix lacryma-jobi L. and Hedychium coronarium J. König.The first species occurred in 165 points and in elevated frequencies in all arms and the second species was recorded in 71 points in five arms.The other three non-native species occurred in low frequencies in the arms (28, three and three sampling stations, respectively).N. caerulea occurred in six arms, C. lacrymajobi was recorded only in one arm and H. coronarium in two arms.In addition, although the arm with the highest frequency of non-native species was the Arroio Guaçu (all sampling points), the arms with the highest richness of non-native species were Ocoí and Passo Cuê River (four species).
The ordination analysis showed that species composition differed in the different reservoir arms (Fig. 3).The DCA axis 1 shows the reservoir arms ordered according to their longitudinal position along the reservoir main body (riverine-lacustrine gradient): arms farther from the dam (Arroio Guaçu and São Francisco verdadeiro) were positioned on the right side of axis 1, while arms closer to the dam (Pinto and Passo Cuê) were concentrated on the left side.In fact, there were significant differences among scores of axis 1 (ANOvA, F=60.2, p<0.001).The main species found to be positively correlated to axis 1 were Lemna valdiviana Phil.(r=0.85),Salvinia biloba Raddi (r=0.64) and Salvinia minima Baker (r=0.63).
Secchi disk depth varied between 0.5 and 6m and fetch varied between 0.45 and 30km.The scores of axis 1 were significantly and negatively correlated to Secchi disk depth and fetch (r=0.43,p<0.001 and r=-0.36,p<0.001, respectively), indicating that these two abiotic variables are important determinants of macrophyte assemblage composition.

DISCUSSION
Despite conducting the survey in a single, short (15 day) period, we found 87 macrophyte taxa in the eight arms located on the Brazilian side of the Itaipu Reservoir.In the first survey of macrophytes conducted ten years earlier in this reservoir, Thomaz et al. (1999) found 62 taxa belonging to 25 families and 42 genera.Among this total, 47 were identified at the species level and 33 species were common between both surveys.Similar to what we found in 2008, the Poaceae family was the best represented in terms of taxa in 1995-1998.There were many Polygonum species in both surveys; however, the other most specious genera in 2008 differed from those in 1995-1998.Differences in the number of species and the identity of genera between these two surveys indicate that macrophyte assemblages are still changing and are dynamic in this reservoir.In fact, conspicuous changes due to disturbances caused by water drawdown and alterations in water physico-chemistry (mainly underwater radiation) occurred in the Itaipu Reservoir (Thomaz et al. 2006(Thomaz et al. , 2009)).The long period between surveys may have also allowed the natural colonization of other species.However, in addition to these environmental causes, differences between the two surveys may be attributable to distinct sampling protocols because the early surveys in the Itaipu Reservoir focused mainly on the euhydrophytes (especially rooted submerged and free floating species), while in the second survey, we followed Cook's classification.In addition, many taxa were identified only to genus in the first survey; therefore, comparisons should be made with caution.
Compared with other reservoirs of the Upper Paraná River Basin, Itaipu has a rich macrophyte assemblage.For example, Martins et al. (2008) studied 18 reservoirs and found a total of 39 species in all of them and Thomaz et al. (2005) recorded 37 species in the Rosana Reservoir (Paranapanema River).Differences may be due to several non-independent factors, such as differences in area, physico-chemistry and even taxonomic resolution.
The families Poaceae and Cyperaceae, which are among the best-represented families in Itaipu, are also the most important families in other freshwater ecosystems of the Upper Paraná River, like floodplain lakes (Pott et al. 1992, Bini et al. 1999, Kita & Souza 2003) and reservoirs (Pedralli et al. 1993, Martins et al. 2008).Despite having similar families, the composition of the macrophyte assemblage in the Itaipu Reservoir differs from natural habitats in other parts of the Upper Paraná River.For example, in one lake in the Upper Paraná River floodplain, the genera Cyperus, Paspalum and Solanum had the highest number of species (Kita & Souza 2003).
It is worth noting that free floating species, like E. crassipes, Salvinia spp.and Pistia stratiotes L., are very common in other Brazilian reservoirs (Carvalho et al. 2003, 2005, Gastal Jr. et al. 2003, Martins et al. 2003) but are not among the most frequent species in Itaipu.Some arms of the Itaipu Reservoir have mesotrophic characteristics (Thomaz et al. 2003) and thus the physico-chemistry in the Itaipu Reservoir is not favourable to this specific life form, which depends on water nutrients.Although considered to be one of the greatest nuisance species of macrophytes, free floating species have rarely grown extensively and caused trouble in the Itaipu Reservoir (Thomaz et al. 2003).In contrast to free floating species, H. amplexicaulis, an emergent species, is the most frequent species in Itaipu, but it is not common in other reservoirs (Cavenaghi et al. 2003, Carvalho et al. 2003).
Although there is little threat from the fast growth of native species, two non-native species of macrophytes in Itaipu are of great concern: the emergent species, U. subquadripara, which has floating stems and the rooted submerged species, H. verticillata.U. subquadripara was also very frequent in the reservoir during a previous survey, but it was mistakenly named by Thomaz et al. (1999) as Urochloa plantaginea (Link) R.D. Webster.U. subquadripara is native to Africa and is one of the most common species in Itaipu and in several other Brazilian reservoirs (e.g., Carvalho et al. 2003, 2005, Martins et al. 2003, 2008).In Itaipu, it covers large areas and creates extensive stands, which compromise access to the water table and reduces the diversity of native species of macrophytes (Michelan et al. 2010).U. subquadripara, recorded since the first survey in 1995, is continuously increasing in frequency of occurrence and is colonizing new areas in Itaipu (Thomaz et al. 2009).
In contrast with U. subquadripara, H. verticillata was recorded only in the second extensive survey (2008).H. verticillata was registered for the first time in Itaipu in January 2007 (Thomaz et al. 2009) and has quickly colonized new sites since then.The invasion by H. verticillata, which is native to Asia and North Africa, has been facilitated in Itaipu since 2001 by increase in underwater radiation, which enables the rapid growth of this macrophyte (Thomaz et al. 2009).According to Thomaz et al. (2009), H. verticillata arrived from the main river (Paraná River), where this species was recorded earlier than it was recorded in this reservoir.Together with U. subquadripara, H. verticillata causes concern because it can affect native aquatic assemblages (Hofstra et al. 1999, Mony et al. 2007, Theel et al. 2008).
N. caerulea (native to South Africa), C. lacryma-jobi (native to Asia) and H. coronarium (native to the Himalayas and Madagascar) were rare in Itaipu and despite being non-native, cannot be considered invasive in this reservoir.This also appears to be the case in other reservoirs in the Upper Paraná Basin, where they have not been recorded (Cavenaghi et al. 2003, Carvalho et al. 2005, Martins et al. 2008).However, H. coronarium has been found in natural ecosystems in the State of Minas Gerais (Santos et al. 2005).
Differences in the species composition between arms showed a pattern of organization along the reservoir main axis.Assemblages were preferentially composed by free floating species in the arms closer to the riverine zone and by submerged species in the arms positioned from the middle portion of the reservoir down to the dam.These differences may reflect the higher inputs of nutrients in the first arms through the Paraná River and higher light penetration in the others.In fact, there is a conspicuous gradient of sedimentation along the Itaipu Reservoir, with higher inputs of solids and nutrients carried by the Paraná River (Pagioro & Thomaz 2002).Both solids and nutrients decrease as the distance to the dam decreases, resulting in higher water transparency in the lacustrine zone of this reservoir (Pagioro & Thomaz 2002).In accordance with this finding, the underwater light (measured with the Secchi disk) was negatively correlated to scores of the DCA axis 1 and thus, positively correlated with the predominance of two of the most important submerged species: E. densa and H. verticillata.It shows that underwater light is an important determinant of macrophyte assemblage composition in the Itaipu Reservoir.
Fetch was also an important variable in determining assemblage composition.Fetch is a surrogate of waves, which represent a disturbance to macrophytes; a correlation between fetch and the attributes of macrophyte populations and assemblages have been shown elsewhere (Chambers 1987, Rea et al. 1998).Earlier investigations in the Itaipu Reservoir, for example, showed that macrophyte richness was negatively affected by fetch (Thomaz et al. 2003).
In summary, the extensive survey of aquatic macrophytes conducted in September 2008 recorded more species than the survey conducted in 1995-1998.The increased number of species may be primarily due to changes in the water physico-chemistry and water drawdown disturbances, although differences in classification systems (sensu Cook 1990) and taxonomic resolutions between the surveys may also account for such differences in species data between the years.Additionally, five of the 87 species recorded were non-native.Two of these non-natives are of special concern because they occur frequently and occupy large marginal areas in the reservoir.Finally, assemblage composition varied among arms and was at least in part determined by underwater radiation and fetch.Future surveys should examine the effects of other abiotic factors (e.g., nutrients) on assemblages and examine assemblage alterations in sites that have been colonized by exotic species.Together with our results, those of such future studies will help provide a scientific basis for the management of aquatic macrophytes and could be used to avoid damage to the reservoir biota and also to the potential water source that is the Itaipu Reservoir.

Fig. 1 .
Fig. 1.Sampling points in the eight arms of the Itaipu Reservoir.

Fig. 2 .
Fig. 2. (A) Frequency of families occurrence as recorded in 235 sampling stations in the Itaipu Reservoir; other families were included in the column "others".(B) Ranking of life forms occurring in the Itaipu Reservoir.Em=emergent, RS=rooted submerged, Am=amphibious, FF=free floating, FL=floating leaved, FS=free submerged and Ep=epiphyte.