Aquatic Botany 124 (2015) 85–91

Contents lists available at ScienceDirect

Aquatic Botany

journa l homepage: www.e lsev ier .com/ locate /aquabot

Growth of a native versus an invasive submerged aquatic macrophytediffers in relation to mud and organic matter concentrations insediment

Márcio José Silveira ∗, Sidinei Magela ThomazUniversidade Estadual de Maringá, Núcleo de Pesquisas em Limnologia, Ictiologia e Aquicultura – Nupelia, Av. Colombo 5790, Maringá, PR, 87020-900,Brazil

a r t i c l e i n f o

Article history:Received 3 October 2014Received in revised form 14 March 2015Accepted 17 March 2015Available online 19 March 2015

Keywords:Non-native speciesExotic speciesInvasionAnoxic sedimentMacrophyte growthExperiment

a b s t r a c t

Eutrophication has reduced the colonization of submerged plants in some freshwater ecosystems, andorganic matter (OM) together with grain size in sediment may mediate such decreases. We tested theisolated and combined effects of sediment mud and OM concentrations on the early growth of two speciesof submerged macrophytes, Egeria najas (native) and Hydrilla verticillata (invasive). We hypothesizedthat the sediment OM concentration has more important effects than the mud concentration on plantgrowth and that E. najas is more successful than H. verticillata in highly organic sediment. We testedthese hypotheses using mesocosms with several combinations of mud and OM concentrations in thesediment. We used plant length, the number of lateral branches, root dry mass and total plant dry massas response variables. Both OM and mud were found to be important determinants of the growth ofboth species, but the former had a stronger influence than the latter. However, the responses of allplant attributes differed between the species. For example, the growth of E. najas increased linearly withincreasing OM concentration, but H. verticillata responded to this variable with a quadratic tendency.The decrease in the growth of the invasive species at high OM concentrations may be associated withits lower tolerance to phytotoxic compounds released by the decomposition that occurs in more organicsediment. In conclusion, our experimental data support the importance of the sediment OM in the growthof the native E. najas and invasive H. verticillata. However, future studies should investigate the adaptationmechanisms that allow both plants to colonise such distinct sediment types.

© 2015 Elsevier B.V. All rights reserved.

1. Introduction

Eutrophication has reduced the presence of submerged macro-phytes in some freshwater ecosystems over the last 100 years(Sand-Jensen et al., 2000; Seddon et al., 2000). Sediment organicmatter (OM), which usually increases with eutrophication, is oneof the factors used to explain this tendency (Irfanullah and Moss,2004; Liu et al., 2004). Nutrients typically increase with the OMconcentration in the sediment (Barko and Smart, 1986; Newboltet al., 2008), which could account for higher plant growth in moreorganic sediment (Pulido et al., 2011). However, after a threshold,OM may hinder plant growth through the release of phytotoxins(Barko and Smart, 1986; Wu et al., 2009). Because of its importancefor macrophytes growth, sediment OM, along with other environ-mental factors, can influence entire macrophyte communities by

∗ Corresponding author. Tel.: +55 443011 5952; fax: +55 443011 4625.E-mail address: s.marciojs@gmail.com (S.M. Thomaz).

determining the patterns of species distribution (Chappuis et al.,2014).

Several studies have indicated that an increase in sediment OMmay inhibit the growth and survival of some submerged plantspecies (Barko and Smart, 1986; Terrados et al., 1999). Thus, sed-iment OM could be a filter preventing invasion by non-nativemacrophytes, such as the extremely invasive submerged macro-phyte Hydrilla verticillata (L.f.) Royle, which is highly productive inits introduced ranges (Bianchini Jr. et al., 2010) but is sensitive tophytotoxins released by organic sediment (Barko and Smart, 1986;Wu et al., 2009). This species has successfully invaded ecosystemsand caused ecological and economic impacts worldwide (Cook andLüönd, 1982; Sousa, 2011); thus, studies of its ecology are of greatinterest. It has recently appeared in and colonised the Upper ParanáRiver (Brazil), but it grows poorly in sites with highly organic sed-iment (Sousa et al., 2009, 2010). The decline of H. verticillata inits native range (Asia) is also attributed to eutrophication and theconsequent increase in sediment OM (Yan et al., 1997; Qiu and Wu,1998). However, the OM concentration in sediment does not seem

http://dx.doi.org/10.1016/j.aquabot.2015.03.0040304-3770/© 2015 Elsevier B.V. All rights reserved.

86 M.J. Silveira, S.M. Thomaz / Aquatic Botany 124 (2015) 85–91

to compromise the success of several Neotropical macrophytes,such as Egeria najas (Planch), Cabomba furcata Schult. & Schult. f. andMyriophyllum aquaticum (Vell.) Verdcourt, which are found in habi-tats with high OM concentrations, such as floodplain lakes (Sousaet al., 2009; Sousa et al., 2009).

In addition to OM, other sediment features, such as grain size,are also determinants of submerged macrophyte success. Increasedplant growth in fine sediment appears to be a response to nutrientavailability (Boeger, 1992). Most aquatic macrophytes grow poorlyin coarse sediment compared with fine sediment because the latterincreases nutrient diffusion and exchange between the roots andthe sediment (Barko and Smart, 1986).

Sub-tropical lakes in the Upper Paraná River floodplain accu-mulate sediment with a high mud concentration (grain size<0.062 mm) together with high quantities of OM (Rosin et al., 2010;Ragonha et al., 2013). Some of these lakes are colonized by thenative submerged species E. najas, but they are apparently lessprone to colonization by the non-native H. verticillata (Sousa et al.,2009; Sousa et al., 2009). Although OM and mud may explain thedistribution of these and other species of macrophytes, it is difficultto distinguish the effects of both variables on plant success becausethey are positively correlated in natural ecosystems. Thus, exper-imental manipulations of these two important sediment featuresare necessary to determine their individual and interactive roles inplant growth.

In this study, we assessed both isolated and interaction effectsbetween OM and mud on the success of E. najas and H. verticillatain mesocosms. We chose these two species because the former isvery common in several natural and man-made ecosystems in theNeotropical region, while the latter has a great invasive potential.We asked the following questions: (i) which sediment feature (OMor mud) is the most important determinant of plant establishmentand growth? (ii) How do these species respond to increases in OMand mud in sediment? We hypothesized that (i) sediment OM isthe variable that best explains plant performance compared withmud and that (ii) E. najas growth will be less affected than H. verti-cillata growth in OM and mud-rich sediment. We predicted that drymass, plant length and the number of lateral branches, which aresurrogates of plant success, are not limited by high OM and mudconcentrations for E. najas, but that these factors are negativelyaffected in H. verticillata.

2. Materials and methods

Our experiments were conducted on the campus of the Uni-versity of Maringá (South Brazil). Apical portions of healthymacrophytes (15 cm long), sediment and riparian forest litter werecollected from the Upper Paraná River floodplain (22◦45′S, 53◦15′Wand 22◦45′S, 53◦30′W). Fragments of both species of macrophyteswere collected from a site at the main river channel where the twospecies co-existed. Litter was dried at 60 ◦C to a constant weightand ground in a mill to obtain a homogeneous powder, which wasused to manipulate the sediment OM concentrations. All naturalsediment was incinerated at 560 ◦C for four hours to eliminateOM, and it was then sieved according to the method described byWentworth (1922) and Suguio (1973) to remove all mud (silt + clay;grain size <0.062 mm). The two fractions of OM-free sediment (onewith particle sizes of <0.062 mm and the other with particle sizesof >0.062 mm) were saved for later use in the experiment. Thisprocedure produced mud-free and OM-free sediment.

To assess the independent effect of OM, litter powder was addedto the OM-free sediment >0.062 mm to produce a gradient of 0, 2,4, 7, 12 and 22% g DW OM. Then, these different sediments weremixed with OM-free mud sediment (<0.062 mm particle size, col-lected by sieving as described above) to obtain a gradient of 0, 5,

Fig. 1. Experimental procedures and addition matrix, showing the percentages oforganic matter (OM) and mud (M) added to the sediment of the sampling units.*The approximate levels of OM and mud found in the sediment where both speciesof macrophytes co-existed in situ. Note that comparisons between six levels of bothmud and OM produced the 36 treatments.

15, 25, 40 and 55% g DW mud. This procedure allowed for all possi-ble mixtures of mud and OM, producing a total of 36 combinations,which are (the procedure and all combinations) shown in Fig. 1.These percentages were chosen to simulate the range found in thefield in the Paraná River floodplain (Sousa et al., 2009, 2010 Rosinet al., 2010; Ragonha et al., 2013). The treatment with 2% g DWmud and 5% g DW OM can be considered to be a control because itis similar to the condition found at the site where both macrophytespecies co-existed (Sousa et al., 2010).

The apical portions of E. najas and H. verticillata without adven-titious roots or branches were planted in pots (8 × 8 × 10 cm) withthe sediment treatments shown in Fig. 1. Each pot contained oneapical portion. We used three replicates for the treatment pots(a total of 108 pots for each species). Although the number ofreplicates can be considered to be low, our procedure produceda gradient of sediment OM and mud that permitted the use of

M.J. Silveira, S.M. Thomaz / Aquatic Botany 124 (2015) 85–91 87

regression analysis to evaluate our data (see below). This analy-sis circumvented the limited number of replicates. The pots wererandomized inside of three tanks (161 × 111 × 73 cm), and eachtank received a replicate to ensure replicate independence. Thetanks were filled with tap water (approximately 500 L), which wasreplaced every 10 days to avoid phytoplankton accumulation. Alltanks remained outdoors; thus, the incidence of light was similarto that experienced by plants grown in natural ecosystems.

Water temperature, dissolved oxygen and saturation, pH andconductivity were measured at the sub-surface every 10 days. Inaddition, sediment was collected from each pot to measure totalnitrogen (N) (Mackereth et al., 1978) and total phosphorus (P) con-centrations (Golterman et al., 1978). Mud content was determinedas previously stated and OM content was measured after incinera-tion at 560 ◦C for four hours.

The experiment ended when the plants reached the water sur-face (60 days). Macrophytes were removed from the tanks tomeasure the following attributes: plant length (cm), the numberof lateral branches, root dry mass and total dry mass (obtained bydrying in an oven at 60 ◦C until a constant weight was reached).

2.1. Data analyses

Correlations between OM, mud, and N and P concentrationswere examined to assess the effects of mud and OM on nutrientconcentrations. These correlations also allowed us to assess theindependence of the OM and mud concentrations in the treatments.

To test the effects of OM and mud on the plant attributes, weused multiple regression analyses. We assessed the assumptions ofmulticollinearity and homogeneity through the visual inspection ofresiduals. This exploratory analysis revealed distinct tendencies forboth species. A linear tendency was observed for E. najas, whereasa quadratic curve in response to OM was clear for H. verticillata(Figs. 3 and 4). However, the linear and quadratic models weretested for both species, after which we selected the best model(based on residual analyses) that explained their development.Because OM and mud usually co-vary in the field (Burone et al.,2003; Magni et al., 2008), we included a coefficient in the modelsto test the interaction between OM × mud. The linear models weredescribed as follows:

z = constant + b(OM) + c(mud) + d(OM × mud) (1)

where,z = macrophyte attribute;b = partial coefficient for the rela-tionship with OM;c = partial coefficient for the relationship withmud; and d = partial coefficient for the interaction of OM × mud.Themodel with the quadratic parameter was determined by the follow-ing equation:

z = constant + b(OM) + c(OM∧2) + d(mud) + f (OM × mud) (2)

where,z = macrophyte attribute;b = partial coefficient for the rela-tionship with OM;c = partial coefficient for the relationshipwith OM2 (quadratic term);d = partial coefficient for the relation-ship with mud; and f = partial coefficient for the interaction ofOM × mud.

All analyses were conducted using Statistica TM 7.0.a softwarepackage.

No statistical comparison of growth between the two specieswas performed, i.e., no interaction terms were investigated,because the species were fitted by two different models (quadraticversus linear). We believe that these results are sufficient to drawconclusions about interactions, i.e., differing species responses inthe presence of varying OM and mud concentrations.

Table 1Means and standard deviations (SDs) of the abiotic water variables measured in thetanks during the experiment. The means were obtained from samples collected overtime.

Abiotic water variables Means and standard deviations (SDs)

Tank 1 Tank 2 Tank 3

Temperature (◦C) 28.3 ± 0.57 28.1 ± 0.53 28.3 ± 0.75Oxygen (%) 134.5 ± 15.25 121.4 ± 8.91 99.1 ± 40.06Oxygen (mg l−1) 6.9 ± 0.76 6.5 ± 0.57 6.2 ± 0.56pH 8.3 ± 0.37 8.0 ± 0.20 8.0 ± 0.28Conductivity (�S cm−1) 150.7 ± 12.34 151.5 ± 12.53 150.7 ± 11.9

3. Results

The water features were very similar among the tanks, indi-cating that all of the treatments were conducted under similarabiotic conditions (Table 1). High oxygen levels, alkaline pH levelsand warm temperatures predominated, which are similar to theconditions found in the natural habitat of the Upper Paraná Riverfloodplain, where plants grow during the spring and summer.

The sediment N concentrations varied between 1.59 and2.98 mg/g DW, and P concentrations ranged from 0.03 to 2.37 mg/gDW. The treatments with lower OM and mud concentrations alsohad low levels of N and P (Fig. 2). Indeed, nutrient concentrationswere highly correlated with that of OM (N: r = 0.94; p < 0.001; P:r = 0.78; p < 0.001). Correlations between nutrient and mud concen-trations were less strong, although they were also significant (N:r = 0.18; p = 0.05; P: r = 0.48; p < 0.001). Mud and OM concentrationswere not correlated (r = 0.01; p = 0.901), indicating that these twopredictive variables varied independently in our treatments. Thisindependence is an important pre-assumption to test our hypothe-ses.

The responses of E. najas and H. verticillata to changes in the OMconcentration were stronger than those for mud (Figs. 3 and 4).However, the attributes of E. najas exhibited different tendenciesthan those of H. verticillata in response to the varying OM and mudconcentrations (Figs. 3 and 4).

For E. najas, all attributes (except for root DW) increased lin-early and significantly with increasing OM (Table 2 and Fig. 3). Bycontrast, mud affected only the length of E. najas, and this effectwas less noticeable than the effects caused by OM (Table 2 andFig. 3). Indeed, the OM × mud interaction term was significant, indi-cating that all of the response variables increased much faster withincreasing OM when there was an also increase in the mud concen-tration (Table 2 and Fig. 3).

Similar to the observations in E. najas, OM significantly influ-enced all attributes of H. verticillata (Table 2 and Fig. 4). However,all attributes of the latter species increased with changes in OMuntil a threshold was reached, after which a decrease was observed(Fig. 4). Indeed, the quadratic term was significant for all attributesmeasured in H. verticillata (Table 2). The mud concentration had asignificant effect only on plant length and total dry mass, and itseffects on these attributes were also weaker than those caused byOM (Table 2 and Fig. 4).

4. Discussion

Our results showed that both OM and mud are determinantsof macrophyte success but that the former has a greater impor-tance than the latter. The results also showed that both macrophytespecies responded differently to increase in OM. Thus, our hypothe-ses that (i) OM is more influential than mud on macrophyte growthand that (ii) the native macrophyte E. najas is more resistant to OMaccumulation in sediment were supported. Our results are rele-vant because our experimental design allowed us to distinguish the

88 M.J. Silveira, S.M. Thomaz / Aquatic Botany 124 (2015) 85–91

Fig. 2. The means and standard deviations of the total nitrogen (N) and phosphorus (P) concentrations of the sediment analyzed for all treatments (mg/g DW). *The levelsof (N) and (P) found in the sediment where both species of macrophytes co-existed in situ.

effects of OM and mud on macrophyte growth, which is not pos-sible in the field, where both components usually co-vary (Buroneet al., 2003; Magni et al., 2008), making it difficult to assess theirindependent effects.

In contrast with E. najas, the invasive H. verticillata showeddecreases in all attributes after an OM threshold was reached, asdemonstrated by the significance of the quadratic term. It appearsthat approximately 7% OM is the optimum level for H. verticillatadevelopment. These results are in accordance with others obtainedin situ, showing that this species is still able to grow in sites with<10% OM in the sediment but does not colonize sites with >13% OM(Sousa et al., 2009; Sousa et al., 2009).

Although there was a decrease in H. verticillata growth in highlyorganic sediment in our experiment, this species was still able togrow in the high OM treatments. Thus, sediment OM is not the onlyfactor that explains the low success of H. verticillata in the lakesof the Upper Paraná, and other factors are certainly involved (seediscussion below).

The increases in plant length, the number of lateral branches andplant biomass in sediments with <10% OM support the view thatthese attributes may be involved in the dominance of H. verticillataat sites with low to medium concentrations of organic matter inthe sediment in situ. First, the growth of this macrophyte allowsit to reach the water surface and to reduce underwater light. This

Table 2The results of equations (linear model for Egeria najas and quadratic model for Hydrilla verticillata) with the best fit for the attributes of plant length, the number of lateralbranches, root dry mass and plant dry mass for Egeria najas and Hydrilla verticillata. The model parameters of the equation used for the species were as follows: b = partialcoefficient for the relationship with OM; c = partial coefficient for the relationship with OM2 (quadratic term); d = partial coefficient for the relationship with mud; andf = partial coefficient for the interaction of OM × mud. Only the significant coefficients are presented.

Equations

Attributes Species Constant b(OM) c(OM2) d(mud) f(OM × mud) R2 F p

Plant length E. najas 17.52 0.46 – 0.07 0.0077 0.55 44.94 <0.001H. verticillata 22.52 5.38 −0.22 0.37 – 0.39 24.39 <0.001

Number of lateral branches E. najas 0.90 0.12 – – 0.0031 0.59 78.29 <0.001H. verticillata 3.73 1.25 −0.05 – – 0.50 55.60 <0.001

Root dry mass E. najas 0.008 0.00010 – 0.000067 0.75 167.17 <0.001H. verticillata 0.041 0.028 −0.0011 – – 0.35 30.89 <0.001

Plant dry mass E. najas 0.16 0.018 – – 0.00047 0.80 221.57 <0.001H. verticillata 0.351 0.28 −0.011 0.010 – 0.41 26.10 <0.001

M.J. Silveira, S.M. Thomaz / Aquatic Botany 124 (2015) 85–91 89

Fig. 3. The effects of OM and mud concentrations on the measured attributes of Egeria najas, including plant length (a), the number of lateral branches (b), root dry mass (c)and plant dry mass (d).

is a typical characteristic of canopy-forming species that concen-trate the photosynthetic tissues close to the water surface. Second,the development of a greater number of lateral branches, whosefragmentation contributes to plant dispersion (Madsen and Smith,1999; Martin and Valentine, 2014), further enhance the dominanceof this species in the Paraná River. On the other hand, a decreasein H. verticillata growth in highly organic sediment is most likelyrelated to its sensitivity to phytotoxins released during the decom-position of OM, such as methane and sulfur dioxide (Barko andSmart, 1986; Wu et al., 2009; Marín-Muniz et al., 2014). Thus,after a threshold point, the negative effects of OM outperformthe positive effect of nutrients, which increases continuously withincreasing OM. There is also evidence that biomechanical proper-ties of submerged macrophytes (e.g., tensile force, bending forceand structural stiffness) are negatively affected by nutrient-richsediment (Zhu et al., 2013).

Although data from previous studies in the field in sites of theUpper Paraná River (Sousa et al., 2009; Sousa, 2011) and the resultsof the present study suggest that H. verticillata is affected nega-tively by high sediment OM concentrations, other investigationshave demonstrated that this species is able to grow in a wide rangeof OM concentrations (0–60%; Barko and Smart, 1986; Sousa, 2011).The unsuccessful colonization of sites with high levels of OM inthe sediment in the Upper Paraná floodplain may have occurredbecause the plants that invaded this area originated from popula-

tions that were highly sensitive to sediment OM, as observed in itsnative range (Wu et al., 2009). Alternatively, other environmentalconditions, such as the elevated temperatures found in the UpperParaná (up to 30 ◦C during summer; Sousa et al., 2010) can con-tribute to the acceleration of the release of phytotoxins, even atlow OM concentrations, causing toxic effects on H. verticillata.

The different responses of H. verticillata and E. najas to the vary-ing OM concentrations may be related to physiological adaptationsand root morphology (Sorrell, 1999; Vretare and Weisner, 2000).The acclimatisation of macrophytes to anoxic sediment can beattributed to their ability to translocate oxygen to roots (Pedersenet al., 1998), and this ability may be more developed in E. najasconsidering its evolutionary history in warm Neotropical regions,where unsaturated oxygen concentrations are very common (Sousaet al., 2010).

Despite the great importance of OM determined in our experi-ment, the mud concentration also had a positive, although lesser,role on macrophyte development, affecting only the lengths of bothspecies and the biomass of H. verticillata. The influence of grain sizeon plant growth is well known (Barko and Smart, 1986; Xiao et al.,2007) and occurs because this factor affects several physical andchemical sediment properties (Volkenborn et al., 2007). For exam-ple, fine sediment grains increase nutrient diffusion and exchange,which enhance plant growth (Barko and Smart, 1986). In addition,grain size affects the rates of nutrient infiltration and nutrient loss

90 M.J. Silveira, S.M. Thomaz / Aquatic Botany 124 (2015) 85–91

in sediment (Sculthorpe, 1967). As a result, coarse sediment sup-ports lower macrophyte growth rates (Wheeler and Giller, 1982;Barko and Smart, 1986).

Our results also demonstrated that in addition to having iso-lated effects of OM and mud, these factors interacted to determineE. najas growth. Studies conducted in situ in the Upper Paraná flood-plain have shown that E. najas is able to greatly accumulate biomassand attain dominance in floodplain lakes, where sediments withhigh organic matter and mud concentrations predominate (Sousaet al., 2009, 2010; Rosin et al., 2010; Ragonha et al., 2013). This find-ing was most likely caused by increased nutrient concentrationswith increasing OM concentrations. In addition, nutrient reten-tion increases and nutrient loss decreases with increasing mudconcentrations due to its fine particle size. This fact may explainthe observation of the highest growth of E. najas at the great-est concentrations of OM and mud and the highest growth of H.verticillata at intermediate concentrations of OM and high concen-trations of mud. The interactions between sediment mud and OMalso indicated that data on macrophyte growth with sediment char-acteristics obtained in the field should be interpreted with caution

because it is difficult to make conclusions on interactions betweenfactors in the absence of controlled experiments.

Extrapolation of experimental results to the field are not alwayseasy because mesocosms represent a simplification of reality.Although our results add confidence that the OM concentrationin sediment may play a role in the patterns of distribution of E.najas and H. verticillata recorded in the Paraná River floodplain,interactions with other factors that were not manipulated in ourexperiment may also occur, influencing the field results in theUpper Paraná floodplain (Sousa et al., 2009; Sousa, 2011). At leasttwo potential, and not mutually exclusive, factors may interact withsediment, including underwater light and herbivory. Indeed, thesites colonized by H. verticillata in the Upper Paraná are locationsat which Secchi depth values are greater, whereas the opposite istrue for E. najas (Sousa et al., 2009; Sousa et al., 2009). Regardingherbivory, it has been shown that native macrophytes are pre-ferred by native herbivores over introduced macrophytes (Xionget al., 2008). Non-choice experiments have indicated that there isat least one species of snail that prefers E. najas to H. verticillata inthe Paraná River floodplain (T. Meurer, unpublished). Thus, a com-

Fig. 4. The effects of OM and mud concentrations on the measured attributes of Hydrilla verticillata, including plant length (a), the number of lateral branches (b), root drymass (c) and plant dry mass (d).

M.J. Silveira, S.M. Thomaz / Aquatic Botany 124 (2015) 85–91 91

bination of sediment type, underwater light and herbivory (enemyrelease hypothesis) might explain the differential distribution ofnative and non-native species in the Paraná River floodplain.

In conclusion, our experimental data support the importance ofsediment OM in the growth of the native E. najas and invasive H. ver-ticillata, at least in their early stages (two months) of development.In addition, our findings show that these two species respond differ-ently to OM increases in sediment and that granulometry interactswith OM to determine plant growth. Finally, our results indicatethat sediment type alone is not sufficient to explain the absenceof H. verticillata in floodplain lakes, where the sediment is rich inOM. Future studies integrating sediment type with other potentialfactors that affect H. verticillata success, such as biotic resistanceand underwater light, are necessary to better explain the distri-bution of this species in the lakes of the Upper Paraná floodplain.In addition, the assessment of the morphologies and physiologiesof these two species, combined with experiments that manipulatesediment OM and oxygen concentrations, will help to explain themechanisms underlying our results.

Acknowledgements

We thank L.M. Bini (Universidade Estadual de Goiás), E.R. Cunha(Universidade Estadual de Maringá) and R.P. Mormul for the fruit-ful discussions and for helping with the selection of data analyses.We also thank Dr. Elisabeth Gross (editor) for her constructive com-ments. S.M. Thomaz is especially thankful to the National Councilfor Scientific and Technological Development (CNPq) for providingcontinuous funding through a Research Productivity Grant. M.J. Sil-veira thanks CAPES, an organization of the Brazilian Government forthe training of human resources, for providing a PhD scholarship.

References

Barko, J.W., Smart, R.M., 1986. Sediment-related mechanisms of growth limitationin submersed macrophytes. Ecology 65, 1328–1340.

Bianchini Jr., I., Cunha Santino, M.B., Milan, J.A.M., Rodrigues, C.J., Dias, J.H.P., 2010.Growth of Hydrilla verticillata (L.f.) Royle under controlled conditions.Hydrobiologia 644, 301–312.

Boeger, R., 1992. The influence of substratum and water velocity on growth ofRanunculus aquatilis L. (Ranunculaceae). Aquat. Bot. 42, 351–359.

Burone, L., Muniz, P., Pires-Vanini, A.M.S., Rodrigues, M., 2003. Spatial distributionof organic matter in the surface sediments of Ubatuba Bay(Southeastern–Brazil). Annu. Braz. Acad. Sci. 75, 77–90.

Chappuis, E., Garcia, E., Ballesteros, E., 2014. Environmental factors explaining thedistribution and diversity of vascular aquatic macrophytes in a highlyheterogeneous Mediterranean region. Aquat. Bot. 113, 72–82.

Cook, C.D.K., Lüönd, R., 1982. A revision of the genus Hydrilla (Hydrocharitaceae).,Aquat. Bot. 13, 485–504.

Golterman, H.L., Clymo, R.S., Ohmstad, M.A.M., 1978. Methods for Physical andChemical Analysis of Freshwaters. Blackwell Scientific Publication, Oxford.

Irfanullah, H.M., Moss, B., 2004. Factors influencing the return of submerged plantsto a clear-water shallow temperate lake. Aquat. Bot. 80, 177–191.

Liu, X., Hu, D., Chen, W., 2004. Studies on the submersed vegetation of Beijing. J.Cap. Norm. Univ. (Nat. Sci.) 25, 46–50.

Mackereth, F.Y.H., Heron, J.G., Talling, J.J., 1978. Water analysis some revisedmethods for limnologist, Ambleside. In: Freshwater Bological Association.Scientific Publication.

Madsen, J.D., Smith, D.H., 1999. Vegetative spread of dioecious Hydrilla colonies inexperimental ponds. J. Aquat. Plant Manage. 37, 25–29.

Magni, P., de Falco, G., Como, S., Casu, D., Floris, A., Petrov, A.N., Castelli, A., Perilli,A., 2008. Distribution and ecological relevance of fine sediments in organicenriched lagoons: the case study of the Cabras lagoon (Sardinia, Italy). Mar.Pollut. Bull. 56, 549–564.

Marín-Muniz, J.S., Hernández, M.E., Moreno-Casasola, P., 2014. Comparing soilcarbon sequestration in coastal freshwater wetlands with various geomorphicfeatures and plant communities in Veracruz. Mex. Plant Soil 378, 189–203.

Martin, C.W., Valentine, J.F., 2014. Sexual and asexual reproductive strategies ofinvasive Eurasian milfoil (Myriophyllum spicatum) in estuarine environments.Hydrobiologia 727, 177–184.

Newbolt, C.H., Hepp, G.R., Wood, C.W., 2008. Characteristics of sedimentsassociated with submersed plant communities in the lower Mobile River Delta,Alabama. J. Aquat. Plant Manage. 46, 107–113.

Pedersen, O., Borum, J., Duarte, C.M., 1998. Oxygen dynamics in the rhizosphere ofCymodocea rotundata. Mar. Ecol. Prog. Ser. 169, 283–288.

Pulido, C., Lucassen, E., Pedersen, O., Roelofs, J.G.M., 2011. Influence of quantity andlability of sediment organic matter on the biomass of two isoetids, Littorellauniflora and Echinodorus repens. Freshwater Biol. 56, 939–951.

Qiu, D.R., Wu, Z.B., 1998. Ecological studies on aquatic macrophytes in lake Donghuof Wuhan – feasibility for rehabilitation of submerged macrophytes in the lake.Resour. Environ. Yangtze Basin 7, 42–47.

Ragonha, F.H., Chiaramonte, J.B., Junior Fontes, H.M., Cunha, E.R., Benedito, E.,Takeda, A.M., 2013. Spatial distribution of aquatic Oligochaeta in Ilha GrandeNational Park. Braz. Acta Sci. Biol. Sci. 35, 63–70.

Rosin, G.C., Mangarotti, D.P.O., Takeda, A.M., 2010. Chironomidae (Diptera)community structure in two subsystems with different states of conservationin a floodplain of southern Brazil. Acta Limnol. Bras. 22, 276–286.

Sand-Jensen, K., Riis, T., Vestergaard, O., Larsen, S.E., 2000. Macrophyte decline inDanish lakes and streams over the past 100 years. J. Ecol. 88, 1030–1040.

Seddon, S., Connolly, R.S., Edyvane, K.S., 2000. Large-scale seagrass die back innorthern Spencer Gulf. South Aust. Aquat. Bot. 66, 297–310.

Sorrell, B.K., 1999. Effect of external oxygen demand on radial oxygen loss byJuncus roots in titanium citrate solutions. Plant Cell Environ. 22, 1583–1587.

Sousa, W.T.Z., Thomaz, S.M., Murphy, K.J., Silveira, M.J., Mormul, R.P., 2009.Environmental predictors of exotic Hydrilla verticillata L.f. Royle and a nativeEgeria najas Planch. occurrence in a sub-tropical river floodplain: the upperRiver Paraná, Brazil. Hydrobiologia 632, 65–78.

Sousa, W.T.Z., Thomaz, S.M., Murphy, K.J., 2010. Response of native Egeria najasPlanch: and invasive Hydrilla verticillata (L.f.) Role to altered hydroecologicalregime in a subtropical river. Aquat. Bot. 92, 40–48.

Sousa, W.T.Z., 2011. Hydrilla verticillata (Hydrocharitaceae): a recent invaderthreatening Brazil’s freshwater environments: a review of the extent of theproblem Hydrilla verticillata (Hydrocharitaceae): a recent invader threateningBrazil’s freshwater environments: a review of the extent of the problem.Hydrobiologia 669, 1–20.

Sculthorpe, C.D., 1967. The Biology of Aquatic Vascular Plants. Edward ArnoldPublishers, London.

Suguio, K., 1973. Introducão a sedimentologia. Edgard Blucher, São Paulo, pp. 317.Terrados, J., Duarte, C.M., Kamp-Nielsen, L., Agawin, N.S.R., 1999. Are seagrass

growth and survival constrained by the reducing conditions of the sediment?Aquat. Bot. 65, 175–197.

Vretare, V., Weisner, S.E.B., 2000. Influence of pressurized ventilation onperformance of an emergent macrophyte (Phragmites australis). J. Ecol. 88,978–987.

Volkenborn, N., Hedtkamp, S.I.C., van Beusekom, J.E.E., Reise, K., 2007. Effects ofbioturbation and bioirrigation by lugworms (Arenicola marina) on physical andchemical sediment properties and implications for intertidal habitatsuccession. Estuar. Coast. Shelf. Sci. 74, 331–343.

Wentworth, C.K., 1922. A scale of grade and class trems for clastic sediments. J.Geol. 30, 377–392.

Wheeler, B.D., Giller, K.E., 1982. Status of aquatic macrophytes in an undrainedarea of fen in the Norfolk Broads. Engl. Aquat. Bot. 12, 277–296.

Wu, J., Cheng, S., Liang, W., Wu, Z., 2009. Effects of organic-rich sediment andbelow-ground sulfide exposure on submerged macrophyte: Hydrillaverticillata. Bull. Environ. Contam. Toxicol. 83, 497–501.

Xiao, K., Yu, D., Wu, Z., 2007. Differential effects of water depth and sediment typeon clonal growth of the submersed macrophyte Vallisneria natans.Hydrobiologia 589, 265–272.

Xiong, W., Yu, D., Wang, Q., Liu, C., Wang, L., 2008. A snail prefers native over exoticfreshwater plants: implications for the enemy release hypotheses. FreshwaterBiol. 5, 2256–2263.

Yan, G., Ma, J.M., Qiu, D.R., Wu, Z.B., 1997. Succession and species replacement ofaquatic plant community in East Lake. Acta Phytoecol. Sin. 21, 319–327.

Zhu, G., Cao, T., Zhang, M., Ni, L., Zhang, X., 2013. Fertile sediment and ammoniumenrichment decrease the growth and biomechanical strength of submersedmacrophyte Myriophyllum spicatum in an experiment. Hydrobiologia 727,109–120.