Morphometry and retention time as forcing functions to establishment and maintenance of aquatic macrophytes in a tropical reservoir

Macrophytes may constitute an important resource for several chemical, physical and biological processes within aquatic ecosystems. This study considers that in tropical reservoirs with low retention time and with low values of shoreline development (DL), the expansion and persistence of aquatic macrophytes are mainly reported to local conditions (e.g., hydrodynamic and wind exposure) rather than trophic status and depth of the euphotic zone. In this context, this study aimed at describing and comparing the incidence of aquatic macrophytes in a throughflowing, non-dendritic tropical reservoir. During February 2006 to November 2007, eight limnological surveys were performed quarterly within the Ourinhos Reservoir, and in the mouth areas of its tributaries. At the six sampling stations 30 variables were measured. The number of sites with plants varied between 21 and 38 and at the end of the 1st year the total richness was found. The sampling survey outcome the recognition of 18 species of aquatic macrophytes; Cyperaceae (2 genera and 1 species), Pontederiaceae (3 species) and Onarograceae (3 genera) were the families with higher diversity. Seven species (Typha domingensis Pers., Myriophyllum aquaticum (Vell.) Verdec, Salvinia auriculata Aubl., Eichhornia azurea (Sw.) Kunth, Eleocharis sp1, Eichhornia crassipes (Mart.) Solms, Oxycaryum cubense (Poepp. & Kunth) Lye) always were present and were more frequent in the sites. The occurrence of emergent species predominated (45.9%), followed by submersed rooted (24.5%), free floating (19.5%), floating rooted (9.7%) and free submersed (0.3%). Although limnological variables and the distribution of macrophytes have discriminated the same sampling points, the stepwise multiple linear regressions did not pointed out strong correspondences (or coherence) among the most constant and distributed macrophyte species and the selected limnological variables, as well the trophic statuses. Seeing the low relationship among limnological variables and macrophytes distribution, in the case of Ourinhos Reservoir, the results pointed out that the water turbulence, low DL and wind exposure are the main driving forces that determine its aquatic plant distribution, life forms and species composition.

The high ability of macrophytes to explore distinct compartments (water column and sediment) within a aquatic systems is reflected by different life forms such as emergent, submersed and floating plants (Riemer, 1984), and for encompass diverse types of plants, macrophytes are included in different plant divisions: algae, bryophytes and vascular plants (Chambers et al., 2008).Thus, the macrophytes are key resources to habitat structuring, and in supporting the trophic chains and several chemical, physical and biological processes within aquatic ecosystems (Thomaz and Cunha, 2010).
Aquatic macrophytes are affected by spatial and temporal scales, which may drive their successful establishment and development (Ferreira et al., 2015); the changes in species composition between habitat patches are related to a number of factors, including environmental heterogeneity, connectivity, disturbance and productivity (Partanen et al., 2009;Lopes et al., 2014).A number of explanations have been raised to account for the high proportion of widely distributed taxa among the aquatic plants.They include: i) the uniformity of aquatic environment; ii) widespread clonality; iii) high phenotypic plasticity (Santamaria, 2002).Furthermore, according to the bottom-up control concept (Glibert, 1998), the growth and establishment of aquatic plants are associated with physical and chemical factors such as: temperature, underwater radiation, nutrients availability, sediment composition, littoral slope, water speed, waves, wind exposure (Carr et al., 1997;Hudon et al., 2000).
Aquatic plants integrate temporal, spatial, chemical, physical and biological characteristics of an ecosystem, and their spreading and richness are influenced by changes in environmental conditions (Lacoul and Freedman, 2006).Taking into account the many factors that contribute to the establishment of aquatic plants, this study considered the following hypothesis: in a tropical reservoir with low retention time (i.e., RT < two weeks;Straškraba, 1999) and with low values of shoreline development (D L ), the expansion and maintenance of aquatic macrophytes are mainly due to local conditions (e.g., hydrodynamic and wind exposure) rather than customary factors such as, trophic status and depth of the euphotic zone.In this circumstance, the location and the establishment of aquatic macrophyte assemblages basically arise from the physical protection of the environment (shelter), and complementarily the availability of the light and nutrients.In this context, this study aimed at describing and comparing the incidence of aquatic macrophytes in a throughflowing, non-dendritic (i.e., with relatively low value of shoreline development) tropical reservoir.
The reservoir was formed in 2005 by the damming of Paranapanema River, in order to support a hydroelectric power plant (HPP Ourinhos).It is a small (area: 6.82 km 2 ; maximum length: 17.35 km), narrow (maximum width: The fetch values of reservoir range from 0.7 to 3.4 km (mean: 2.2 km).The Ourinhos Reservoir with volume of ca.20.82×10 6 m 3 (maximum operational elevation: 398 m a.s.l.; average flow : 289.2 m 3 s -1 ) presents a low retention time (average: 0.83 days); the low retention time of this reservoir leads to ca. 438 water renewals per year.It is the fifth reservoir (in upstream-downstream) of the complex of dams that succeed in Paranapanema River.This watershed stretch (16,751 km 2 , ANA, 2006) encompasses several small cities and large rural occupations, whose predominant activity is intensive agriculture (Nogueira, 2005), pasture and reforestation.In each SS, the following variables were measured at each 10 cm (from the surface to the bottom) with a multi-parameter sonde (YSI, model 6600): electrical conductivity (EC), dissolved oxygen (DO), temperature (air: Ta; water: Tw), oxidation-reduction potential (ORP), total dissolved solids (TDS), turbidity, pH, chlorophyll-a. Wter samples were also collected for determination of color, alkalinity, coliforms, BOD, COD and Org-N concentrations were determined according to APHA et al. (1998).Inorganic nitrogen compounds (Inorg-N = NH 4 -N + NO 2 -N + NO 3 -N) and phosphorus were determined according to colorimetric methods (Koroleff, 1976;Mackereth et al., 1978).DOC, DIC (TC = DOC + DIC) were determinate by combustion/non-dispersive infrared gas analysis method (Shimadzu analyzer, model 5000A).Maximum depth (z max ) and Secchi disk transparency (z sd ) depth were also measured.The depths of euphotic zone (z eu ) were calculated from the Secchi disk disappearance depths (z eu = z sd ×1.88), according to French et al. (1982).

Limnological surveys
The sum of tributaries flow was calculated from the budget between the output flows of hydroelectric power plant and the affluent flows from the Paranapanema River; the change in water level was also recorded.According to the monthly average output flows (January 2006 to December 2007) and morphometric characteristics of reservoir (volume, length and mean depth), the chronological variations in retention time (RT) and in densimetric Froude number (F d ) were calculated (Orlob et al., 1969).
The mean occurrence percentages were calculated based on the presence of macrophytes in each location (the ratio between the number of sites with a given species and the total number of sites containing macrophytes), considering the eight surveys.Constancy was calculated according to the occurrence of the species (anywhere) in the eight inventories (Bodenheimer, 1938; Equation 1).The mean frequencies refer to the occurrence of the species (or family and life form) compared to the total of plants that have been identified in the each survey.Thus, occurrence refers to spatial distribution and frequency with composition of species.
where: P = number of surveys containing the species; N = total number of surveys.
In order to verify the differences among the SS, discriminant function analysis was employed by using chemical, physical and biological variables.After the SS differentiation by discriminant function analysis, the mean values were used for the similar SS.The macrophytes distributions were also used in discriminant function analysis to distinguish the SS (Legendre and Legendre, 1998).
Stepwise multiple linear regressions were employed to attest the dependence of macrophytes and the limnological variables (confidence interval: 95%), using XLSTAT Version 2008.3.01 software.The ANOVA F-test was also used to compare all variables among the sampling stations.

Results
During the study period, the reservoir displayed low retention times (range: 0.61 to 1.31 days; Figure 2) without a clear trend of variation during rainy period (ca.October-February) and dry season (March to September).On average, the densimetric Froude number was 22.6 ± 5.4 and ranged between 15.7 and 34.6 (Figure 2).The reservoir level was relatively stable at the operational level, ranging between -29 and +32 cm in the region near to the dam (CBA, 2007).The budget between the output and Paranapanema River input flows (CBA, 2007) showed that, on average, the tributaries support with ca.0.13 ± 1.67% of water content.Thus, the Paranapanema River constitutes the main source of water of Ourinhos Reservoir.
Basing on the variables measured, the discriminant function analysis did not point out differences among the results obtained in SS along the Ourinhos Reservoir (Figure 3); therefore, the results of these locations (SS2, SS4, SS6), shown in Table 1, were pooled.High values of turbidity, color, alkalinity, DIC, TC, TDS, EC, Ta, Tw, chlorophyll-a were observed in SS1, while SS3 (high values) and SS5 (less intensity than SS3) were segregated by: NO 3 -N, Inorg-N; coliforms (total and fecal), DP, TP and low N:P ratios.The SS of reservoir were discriminated by the high values of following variables: z max , z eu , z sd and N:P ratio.
The Ourinhos Reservoir is an oligo-mesoeutrophic (sensu Vollenweider, 1968); clear water (i.e., low turbidity, low color, z eu ca. 5 m) and neutral-alkaline environment (pH: 7.2-7.7;Table 1).The DO concentrations are always near to the saturation values (range: 95.2-123%), maintaining high ORP values (range: 97-155 mV).The water of reservoir presents low alkalinity (23.3 ± 2.6 mg L -1 ) and low concentration of nitrogen compounds (with nitrate being the predominant form of inorganic nitrogen (88.8-99.6%)and N:P ca.16.The water of this reservoir also presents low concentrations of carbon (organic and inorganic), low values of the EC, BOD, COD, total and fecal coliforms and chlorophyll-a.On the other hand, the tributaries tended to present highest values to EC, turbidity, TDS, TP, Inorg-N, Org-N, color, TC, alkalinity, coliforms (total and fecal), and lower values to DO, ORP, chlorophyll-a and z eu .The N:P rate in all tributaries were lower than that registered to reservoir.Comparing the variables of the reservoir and its tributaries, the analysis of variance (F-test) indicated significant differences (P: 0.459 to 0.001) among all the sites.The reservoir and its tributaries presented the same order of magnitude to BOD (0.0-3.0 mg L -1 ), COD (3.0-32.0mg L -1 ), and pH (7.0-7.9).
The number of sites with macrophytes ranged from 21 to 38, and of the species between 10 and 15 (Figure 4).Eighteen species of aquatic macrophytes were identified comprising 12 families (Table 2), and in the final of 1st year the total number of species was found (Figure 4).Cyperaceae, Pontederiaceae and Onarograceae were the families with highest number of species.5), the occurrence of emergent species predominated in Ourinhos Reservoir (45.9%), followed by submersed rooted (24.5%), free floating (19.5%), floating rooted (9.7%) and free submersed (0.3%).
Regardless the set of variables used (i.e., limnological variables or macrophytes), the discriminant function analysis did not indicate differences among the SS along the Ourinhos Reservoir (Figure 3).Taking into account the macrophytes distribution, the discriminant function analysis   indicated that the SS1 was characterized by the incidence of N. elegans, in spite of low frequency; H. ranunculoides, H. multiflora, Ludwigia sp3, E. macrophyllus discriminated SS3 and SS5, with higher incidences in SS3 than SS5.Higher incidence of macrophytes species discriminated the reservoir sites.Overall, the differences were basically found among the tributaries SS1, SS3 and SS5 and the reservoir sites (SS2; SS4 and SS6).
The morphometric (shallow, narrow, wind-swept, not dendritic) and hydraulic features (low RT and high F d values), and the type of HPP operation (that encompass low variation in water level) categorized this aquatic system as run-of-the-river (Arcifa and Esguícero, 2012) and rapidly throughflowing reservoir (Straškraba, 1999), in which the vertical circulation of water is continually favored; thus, thermal stratification is not expected (i.e., ). Due to the morphometric features (length and mean depth) of Ourinhos Reservoir, the thermal stratification can only occur when its RT would be higher than 66 days.The contribution of tributaries do not exerts influence on chemical and physical characteristics of reservoir water, once its flows are very low (less than 1%) compared with from the Paranapanema River.In addition, the low residence times did not allow the occurrence of a well-marked longitudinal gradient of limnological variables, as expected to artificial reservoirs (Kimmel et al., 1990).Thus, the chemical and physical constraints imposed by Ourinhos Reservoir waters to aquatic macrophytes community generate a predominance of emergent species, from Typhaceae, Cyperaceae and Alismataceae families.The species related with these families are predominantly located in the littoral region of the reservoir.
Although limnological variables and the distribution of macrophytes have discriminated similarly the sampling points, the stepwise multiple linear regressions did not pointed out strong correspondences (or coherence) among the most constant and distributed macrophyte species and the selected limnological variables, as well the trophic status (Table 3).The incidence degree of submersed species owed the constancy and distribution of M. aquaticum.Because it is rooted, this species may not be directly dependent on the chemical characteristics of the water (i.e., nutrient availability), but may be affected by the availability of light (Hussner et al., 2009;Wersal and Madsen, 2011).On the other hand, the free submersed species (Utricularia sp), that depends on a more sheltered place and dissolved nutrients availability, did not present a good settlement performance (Lacoul and Freedman, 2006).Low constancy and incidence of Utricularia sp are linked with the low DIC and alkalinity concentrations (Carr et al., 1997;Vestergaard and Sand-Jensen, 2000).With regard the distribution and frequency, floating species (rooted or free) are of secondary status in this reservoir.As verified to almost emergent species, the main floating species (E.azurea) also has linked with palustrine area, which can present local characteristics, not necessarily close to that prevail in water (Hudon et al., 2000;Santamaria, 2002;Hawes et al., 2003).Of the free-floating species, two showed high constancy (S. auriculata and E. azurea) but its incidences did not were reported to any selected variable, as well verified to P. stratiotes and Ludwigia sp1 that showed lower constancy (87%).
In conclusion, in the shallow reservoir with low retention time and low DL, the hydrodynamic regime (i.e., turbulent environment) tend to prevail in order to determine the sites morphometry and, consequently, the predominant species and their distribution.In addition, the wind exposure (i.e., fetch length) determines the type of thermal circulation, as well, the favorable sites for the 1.53 km; mean width: 0.39 km; shore line: 44.05 km; shoreline development (D L ): 4.76), and shallow (base quota: 386 m a.s.l.; maximum depth: 12 m; mean depth: 3.1 m; relative depth: 0.41%; volume development (D V ): 0.76).

Figure 2 .
Figure 2. Temporal changes in the average of retention time (bars: standard deviation) and in the densimetric Froude number from Ourinhos Reservoir.

Figure 3 .
Figure 3.The sampling stations differentiation using the relative macrophytes distribution (a) and chemical, physical and biological variables (b) through discriminant function analysis.

Figure 4 .
Figure 4. Change in the number of sites with macrophytes in the Ourinhos Reservoir (a), number of the species registered per survey (b), and accumulate number of species (c).

Figure 5 .
Figure 5. Average of occurrence (bars: standard deviation) and constancy of aquatic macrophytes in the Ourinhos Reservoir from February 2006 to November 2007.

Figure 6 .
Figure 6.Average of frequencies (bars: standard deviation) of aquatic macrophytes according their life forms (a), species (b) and families (c in Ourinhos Reservoir.

Table 1 .
Means and standard deviations (SD) of limnological variables of Ourinhos Reservoir (from February 2006 to November 2007; n = 8).

Table 3 .
t-Student and determination coefficient (r 2 ) from the stepwise multiple linear regressions for most frequent and constant macrophytes species and limnological variables selected by this analysis.Considering the weak relationship among chemical and physical variables of water and macrophytes distribution, in the case of Ourinhos Reservoir the results pointed out that the water turbulence, low DL and wind exposure are main drive forces that determine its aquatic plant distribution, life forms and species composition.In this context, four species among the most frequent and constant (T.domingensis, M. aquaticum, E. azurea, Eleocharis sp1) did not depends directly of characteristics of reservoir water.Among S. auriculata, E. crassipes, and O. cubense, just S. auriculata presented correlation with a variable directly linked with its physiology and water characteristics (i.e., alkalinity).